diff options
Diffstat (limited to 'arch/x86/kernel')
227 files changed, 28682 insertions, 13926 deletions
diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile index 5369059c07a..047f9ff2e36 100644 --- a/arch/x86/kernel/Makefile +++ b/arch/x86/kernel/Makefile @@ -2,14 +2,13 @@ # Makefile for the linux kernel. # -extra-y := head_$(BITS).o head$(BITS).o head.o init_task.o vmlinux.lds +extra-y := head_$(BITS).o head$(BITS).o head.o vmlinux.lds CPPFLAGS_vmlinux.lds += -U$(UTS_MACHINE) ifdef CONFIG_FUNCTION_TRACER # Do not profile debug and lowlevel utilities CFLAGS_REMOVE_tsc.o = -pg -CFLAGS_REMOVE_rtc.o = -pg CFLAGS_REMOVE_paravirt-spinlocks.o = -pg CFLAGS_REMOVE_pvclock.o = -pg CFLAGS_REMOVE_kvmclock.o = -pg @@ -17,25 +16,31 @@ CFLAGS_REMOVE_ftrace.o = -pg CFLAGS_REMOVE_early_printk.o = -pg endif +CFLAGS_irq.o := -I$(src)/../include/asm/trace + obj-y := process_$(BITS).o signal.o entry_$(BITS).o obj-y += traps.o irq.o irq_$(BITS).o dumpstack_$(BITS).o obj-y += time.o ioport.o ldt.o dumpstack.o nmi.o obj-y += setup.o x86_init.o i8259.o irqinit.o jump_label.o obj-$(CONFIG_IRQ_WORK) += irq_work.o obj-y += probe_roms.o -obj-$(CONFIG_X86_32) += sys_i386_32.o i386_ksyms_32.o +obj-$(CONFIG_X86_32) += i386_ksyms_32.o obj-$(CONFIG_X86_64) += sys_x86_64.o x8664_ksyms_64.o -obj-y += syscall_$(BITS).o +obj-$(CONFIG_X86_64) += mcount_64.o +obj-y += syscall_$(BITS).o vsyscall_gtod.o obj-$(CONFIG_X86_64) += vsyscall_64.o obj-$(CONFIG_X86_64) += vsyscall_emu_64.o +obj-$(CONFIG_X86_ESPFIX64) += espfix_64.o +obj-$(CONFIG_SYSFS) += ksysfs.o obj-y += bootflag.o e820.o obj-y += pci-dma.o quirks.o topology.o kdebugfs.o obj-y += alternative.o i8253.o pci-nommu.o hw_breakpoint.o -obj-y += tsc.o io_delay.o rtc.o +obj-y += tsc.o tsc_msr.o io_delay.o rtc.o obj-y += pci-iommu_table.o obj-y += resource.o -obj-y += trampoline.o trampoline_$(BITS).o +obj-$(CONFIG_PREEMPT) += preempt.o + obj-y += process.o obj-y += i387.o xsave.o obj-y += ptrace.o @@ -48,8 +53,6 @@ obj-$(CONFIG_STACKTRACE) += stacktrace.o obj-y += cpu/ obj-y += acpi/ obj-y += reboot.o -obj-$(CONFIG_X86_32) += reboot_32.o -obj-$(CONFIG_MCA) += mca_32.o obj-$(CONFIG_X86_MSR) += msr.o obj-$(CONFIG_X86_CPUID) += cpuid.o obj-$(CONFIG_PCI) += early-quirks.o @@ -65,12 +68,13 @@ obj-$(CONFIG_X86_REBOOTFIXUPS) += reboot_fixups_32.o obj-$(CONFIG_DYNAMIC_FTRACE) += ftrace.o obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += ftrace.o obj-$(CONFIG_FTRACE_SYSCALLS) += ftrace.o +obj-$(CONFIG_X86_TSC) += trace_clock.o obj-$(CONFIG_KEXEC) += machine_kexec_$(BITS).o obj-$(CONFIG_KEXEC) += relocate_kernel_$(BITS).o crash.o obj-$(CONFIG_CRASH_DUMP) += crash_dump_$(BITS).o -obj-$(CONFIG_KPROBES) += kprobes.o +obj-y += kprobes/ obj-$(CONFIG_MODULES) += module.o -obj-$(CONFIG_DOUBLEFAULT) += doublefault_32.o +obj-$(CONFIG_DOUBLEFAULT) += doublefault.o obj-$(CONFIG_KGDB) += kgdb.o obj-$(CONFIG_VM86) += vm86_32.o obj-$(CONFIG_EARLY_PRINTK) += early_printk.o @@ -83,23 +87,25 @@ obj-$(CONFIG_DEBUG_RODATA_TEST) += test_rodata.o obj-$(CONFIG_DEBUG_NX_TEST) += test_nx.o obj-$(CONFIG_DEBUG_NMI_SELFTEST) += nmi_selftest.o -obj-$(CONFIG_KVM_GUEST) += kvm.o -obj-$(CONFIG_KVM_CLOCK) += kvmclock.o +obj-$(CONFIG_KVM_GUEST) += kvm.o kvmclock.o obj-$(CONFIG_PARAVIRT) += paravirt.o paravirt_patch_$(BITS).o obj-$(CONFIG_PARAVIRT_SPINLOCKS)+= paravirt-spinlocks.o obj-$(CONFIG_PARAVIRT_CLOCK) += pvclock.o obj-$(CONFIG_PCSPKR_PLATFORM) += pcspeaker.o -microcode-y := microcode_core.o -microcode-$(CONFIG_MICROCODE_INTEL) += microcode_intel.o -microcode-$(CONFIG_MICROCODE_AMD) += microcode_amd.o -obj-$(CONFIG_MICROCODE) += microcode.o - obj-$(CONFIG_X86_CHECK_BIOS_CORRUPTION) += check.o obj-$(CONFIG_SWIOTLB) += pci-swiotlb.o obj-$(CONFIG_OF) += devicetree.o +obj-$(CONFIG_UPROBES) += uprobes.o +obj-y += sysfb.o +obj-$(CONFIG_X86_SYSFB) += sysfb_simplefb.o +obj-$(CONFIG_EFI) += sysfb_efi.o + +obj-$(CONFIG_PERF_EVENTS) += perf_regs.o +obj-$(CONFIG_TRACING) += tracepoint.o +obj-$(CONFIG_IOSF_MBI) += iosf_mbi.o ### # 64 bit specific files diff --git a/arch/x86/kernel/acpi/Makefile b/arch/x86/kernel/acpi/Makefile index 6f35260bb3e..163b2258147 100644 --- a/arch/x86/kernel/acpi/Makefile +++ b/arch/x86/kernel/acpi/Makefile @@ -1,14 +1,7 @@ -subdir- := realmode - obj-$(CONFIG_ACPI) += boot.o -obj-$(CONFIG_ACPI_SLEEP) += sleep.o wakeup_rm.o wakeup_$(BITS).o +obj-$(CONFIG_ACPI_SLEEP) += sleep.o wakeup_$(BITS).o ifneq ($(CONFIG_ACPI_PROCESSOR),) obj-y += cstate.o endif -$(obj)/wakeup_rm.o: $(obj)/realmode/wakeup.bin - -$(obj)/realmode/wakeup.bin: FORCE - $(Q)$(MAKE) $(build)=$(obj)/realmode - diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c index ce664f33ea8..86281ffb96d 100644 --- a/arch/x86/kernel/acpi/boot.c +++ b/arch/x86/kernel/acpi/boot.c @@ -44,20 +44,15 @@ #include <asm/mpspec.h> #include <asm/smp.h> +#include "sleep.h" /* To include x86_acpi_suspend_lowlevel */ static int __initdata acpi_force = 0; -u32 acpi_rsdt_forced; int acpi_disabled; EXPORT_SYMBOL(acpi_disabled); #ifdef CONFIG_X86_64 # include <asm/proto.h> -# include <asm/numa_64.h> #endif /* X86 */ -#define BAD_MADT_ENTRY(entry, end) ( \ - (!entry) || (unsigned long)entry + sizeof(*entry) > end || \ - ((struct acpi_subtable_header *)entry)->length < sizeof(*entry)) - #define PREFIX "ACPI: " int acpi_noirq; /* skip ACPI IRQ initialization */ @@ -67,6 +62,7 @@ EXPORT_SYMBOL(acpi_pci_disabled); int acpi_lapic; int acpi_ioapic; int acpi_strict; +int acpi_disable_cmcff; u8 acpi_sci_flags __initdata; int acpi_sci_override_gsi __initdata; @@ -141,16 +137,8 @@ static u32 irq_to_gsi(int irq) } /* - * Temporarily use the virtual area starting from FIX_IO_APIC_BASE_END, - * to map the target physical address. The problem is that set_fixmap() - * provides a single page, and it is possible that the page is not - * sufficient. - * By using this area, we can map up to MAX_IO_APICS pages temporarily, - * i.e. until the next __va_range() call. - * - * Important Safety Note: The fixed I/O APIC page numbers are *subtracted* - * from the fixed base. That's why we start at FIX_IO_APIC_BASE_END and - * count idx down while incrementing the phys address. + * This is just a simple wrapper around early_ioremap(), + * with sanity checks for phys == 0 and size == 0. */ char *__init __acpi_map_table(unsigned long phys, unsigned long size) { @@ -160,6 +148,7 @@ char *__init __acpi_map_table(unsigned long phys, unsigned long size) return early_ioremap(phys, size); } + void __init __acpi_unmap_table(char *map, unsigned long size) { if (!map || !size) @@ -195,24 +184,31 @@ static int __init acpi_parse_madt(struct acpi_table_header *table) return 0; } -static void __cpuinit acpi_register_lapic(int id, u8 enabled) +/** + * acpi_register_lapic - register a local apic and generates a logic cpu number + * @id: local apic id to register + * @enabled: this cpu is enabled or not + * + * Returns the logic cpu number which maps to the local apic + */ +static int acpi_register_lapic(int id, u8 enabled) { unsigned int ver = 0; - if (id >= (MAX_LOCAL_APIC-1)) { + if (id >= MAX_LOCAL_APIC) { printk(KERN_INFO PREFIX "skipped apicid that is too big\n"); - return; + return -EINVAL; } if (!enabled) { ++disabled_cpus; - return; + return -EINVAL; } if (boot_cpu_physical_apicid != -1U) ver = apic_version[boot_cpu_physical_apicid]; - generic_processor_info(id, ver); + return generic_processor_info(id, ver); } static int __init @@ -239,7 +235,7 @@ acpi_parse_x2apic(struct acpi_subtable_header *header, const unsigned long end) * to not preallocating memory for all NR_CPUS * when we use CPU hotplug. */ - if (!cpu_has_x2apic && (apic_id >= 0xff) && enabled) + if (!apic->apic_id_valid(apic_id) && enabled) printk(KERN_WARNING PREFIX "x2apic entry ignored\n"); else acpi_register_lapic(apic_id, enabled); @@ -422,12 +418,14 @@ acpi_parse_int_src_ovr(struct acpi_subtable_header * header, return 0; } - if (intsrc->source_irq == 0 && intsrc->global_irq == 2) { + if (intsrc->source_irq == 0) { if (acpi_skip_timer_override) { - printk(PREFIX "BIOS IRQ0 pin2 override ignored.\n"); + printk(PREFIX "BIOS IRQ0 override ignored.\n"); return 0; } - if (acpi_fix_pin2_polarity && (intsrc->inti_flags & ACPI_MADT_POLARITY_MASK)) { + + if ((intsrc->global_irq == 2) && acpi_fix_pin2_polarity + && (intsrc->inti_flags & ACPI_MADT_POLARITY_MASK)) { intsrc->inti_flags &= ~ACPI_MADT_POLARITY_MASK; printk(PREFIX "BIOS IRQ0 pin2 override: forcing polarity to high active.\n"); } @@ -558,6 +556,12 @@ static int acpi_register_gsi_ioapic(struct device *dev, u32 gsi, int (*__acpi_register_gsi)(struct device *dev, u32 gsi, int trigger, int polarity) = acpi_register_gsi_pic; +#ifdef CONFIG_ACPI_SLEEP +int (*acpi_suspend_lowlevel)(void) = x86_acpi_suspend_lowlevel; +#else +int (*acpi_suspend_lowlevel)(void); +#endif + /* * success: return IRQ number (>=0) * failure: return < 0 @@ -572,6 +576,12 @@ int acpi_register_gsi(struct device *dev, u32 gsi, int trigger, int polarity) return irq; } +EXPORT_SYMBOL_GPL(acpi_register_gsi); + +void acpi_unregister_gsi(u32 gsi) +{ +} +EXPORT_SYMBOL_GPL(acpi_unregister_gsi); void __init acpi_set_irq_model_pic(void) { @@ -599,95 +609,43 @@ static void acpi_map_cpu2node(acpi_handle handle, int cpu, int physid) int nid; nid = acpi_get_node(handle); - if (nid == -1 || !node_online(nid)) - return; - set_apicid_to_node(physid, nid); - numa_set_node(cpu, nid); + if (nid != -1) { + set_apicid_to_node(physid, nid); + numa_set_node(cpu, nid); + } #endif } -static int __cpuinit _acpi_map_lsapic(acpi_handle handle, int *pcpu) +static int _acpi_map_lsapic(acpi_handle handle, int physid, int *pcpu) { - struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL }; - union acpi_object *obj; - struct acpi_madt_local_apic *lapic; - cpumask_var_t tmp_map, new_map; - u8 physid; int cpu; - int retval = -ENOMEM; - - if (ACPI_FAILURE(acpi_evaluate_object(handle, "_MAT", NULL, &buffer))) - return -EINVAL; - - if (!buffer.length || !buffer.pointer) - return -EINVAL; - - obj = buffer.pointer; - if (obj->type != ACPI_TYPE_BUFFER || - obj->buffer.length < sizeof(*lapic)) { - kfree(buffer.pointer); - return -EINVAL; - } - - lapic = (struct acpi_madt_local_apic *)obj->buffer.pointer; - - if (lapic->header.type != ACPI_MADT_TYPE_LOCAL_APIC || - !(lapic->lapic_flags & ACPI_MADT_ENABLED)) { - kfree(buffer.pointer); - return -EINVAL; - } - - physid = lapic->id; - kfree(buffer.pointer); - buffer.length = ACPI_ALLOCATE_BUFFER; - buffer.pointer = NULL; - - if (!alloc_cpumask_var(&tmp_map, GFP_KERNEL)) - goto out; - - if (!alloc_cpumask_var(&new_map, GFP_KERNEL)) - goto free_tmp_map; - - cpumask_copy(tmp_map, cpu_present_mask); - acpi_register_lapic(physid, lapic->lapic_flags & ACPI_MADT_ENABLED); - - /* - * If mp_register_lapic successfully generates a new logical cpu - * number, then the following will get us exactly what was mapped - */ - cpumask_andnot(new_map, cpu_present_mask, tmp_map); - if (cpumask_empty(new_map)) { - printk ("Unable to map lapic to logical cpu number\n"); - retval = -EINVAL; - goto free_new_map; + cpu = acpi_register_lapic(physid, ACPI_MADT_ENABLED); + if (cpu < 0) { + pr_info(PREFIX "Unable to map lapic to logical cpu number\n"); + return cpu; } acpi_processor_set_pdc(handle); - - cpu = cpumask_first(new_map); acpi_map_cpu2node(handle, cpu, physid); *pcpu = cpu; - retval = 0; - -free_new_map: - free_cpumask_var(new_map); -free_tmp_map: - free_cpumask_var(tmp_map); -out: - return retval; + return 0; } /* wrapper to silence section mismatch warning */ -int __ref acpi_map_lsapic(acpi_handle handle, int *pcpu) +int __ref acpi_map_lsapic(acpi_handle handle, int physid, int *pcpu) { - return _acpi_map_lsapic(handle, pcpu); + return _acpi_map_lsapic(handle, physid, pcpu); } EXPORT_SYMBOL(acpi_map_lsapic); int acpi_unmap_lsapic(int cpu) { +#ifdef CONFIG_ACPI_NUMA + set_apicid_to_node(per_cpu(x86_cpu_to_apicid, cpu), NUMA_NO_NODE); +#endif + per_cpu(x86_cpu_to_apicid, cpu) = -1; set_cpu_present(cpu, false); num_processors--; @@ -732,7 +690,7 @@ static int __init acpi_parse_sbf(struct acpi_table_header *table) #ifdef CONFIG_HPET_TIMER #include <asm/hpet.h> -static struct __initdata resource *hpet_res; +static struct resource *hpet_res __initdata; static int __init acpi_parse_hpet(struct acpi_table_header *table) { @@ -945,10 +903,6 @@ static int __init acpi_parse_madt_lapic_entries(void) #ifdef CONFIG_X86_IO_APIC #define MP_ISA_BUS 0 -#ifdef CONFIG_X86_ES7000 -extern int es7000_plat; -#endif - void __init mp_override_legacy_irq(u8 bus_irq, u8 polarity, u8 trigger, u32 gsi) { int ioapic; @@ -989,7 +943,7 @@ void __init mp_config_acpi_legacy_irqs(void) int i; struct mpc_intsrc mp_irq; -#if defined (CONFIG_MCA) || defined (CONFIG_EISA) +#ifdef CONFIG_EISA /* * Fabricate the legacy ISA bus (bus #31). */ @@ -998,14 +952,6 @@ void __init mp_config_acpi_legacy_irqs(void) set_bit(MP_ISA_BUS, mp_bus_not_pci); pr_debug("Bus #%d is ISA\n", MP_ISA_BUS); -#ifdef CONFIG_X86_ES7000 - /* - * Older generations of ES7000 have no legacy identity mappings - */ - if (es7000_plat == 1) - return; -#endif - /* * Use the default configuration for the IRQs 0-15. Unless * overridden by (MADT) interrupt source override entries. @@ -1071,9 +1017,7 @@ static int mp_config_acpi_gsi(struct device *dev, u32 gsi, int trigger, if (!acpi_ioapic) return 0; - if (!dev) - return 0; - if (dev->bus != &pci_bus_type) + if (!dev || !dev_is_pci(dev)) return 0; pdev = to_pci_dev(dev); @@ -1101,6 +1045,7 @@ int mp_register_gsi(struct device *dev, u32 gsi, int trigger, int polarity) int ioapic; int ioapic_pin; struct io_apic_irq_attr irq_attr; + int ret; if (acpi_irq_model != ACPI_IRQ_MODEL_IOAPIC) return gsi; @@ -1130,7 +1075,9 @@ int mp_register_gsi(struct device *dev, u32 gsi, int trigger, int polarity) set_io_apic_irq_attr(&irq_attr, ioapic, ioapic_pin, trigger == ACPI_EDGE_SENSITIVE ? 0 : 1, polarity == ACPI_ACTIVE_HIGH ? 0 : 1); - io_apic_set_pci_routing(dev, gsi_to_irq(gsi), &irq_attr); + ret = io_apic_set_pci_routing(dev, gsi_to_irq(gsi), &irq_attr); + if (ret < 0) + gsi = INT_MIN; return gsi; } @@ -1333,17 +1280,12 @@ static int __init dmi_disable_acpi(const struct dmi_system_id *d) } /* - * Force ignoring BIOS IRQ0 pin2 override + * Force ignoring BIOS IRQ0 override */ static int __init dmi_ignore_irq0_timer_override(const struct dmi_system_id *d) { - /* - * The ati_ixp4x0_rev() early PCI quirk should have set - * the acpi_skip_timer_override flag already: - */ if (!acpi_skip_timer_override) { - WARN(1, KERN_ERR "ati_ixp4x0 quirk not complete.\n"); - pr_notice("%s detected: Ignoring BIOS IRQ0 pin2 override\n", + pr_notice("%s detected: Ignoring BIOS IRQ0 override\n", d->ident); acpi_skip_timer_override = 1; } @@ -1437,7 +1379,7 @@ static struct dmi_system_id __initdata acpi_dmi_table_late[] = { * is enabled. This input is incorrectly designated the * ISA IRQ 0 via an interrupt source override even though * it is wired to the output of the master 8259A and INTIN0 - * is not connected at all. Force ignoring BIOS IRQ0 pin2 + * is not connected at all. Force ignoring BIOS IRQ0 * override in that cases. */ { @@ -1472,6 +1414,14 @@ static struct dmi_system_id __initdata acpi_dmi_table_late[] = { DMI_MATCH(DMI_PRODUCT_NAME, "HP Compaq 6715b"), }, }, + { + .callback = dmi_ignore_irq0_timer_override, + .ident = "FUJITSU SIEMENS", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "FUJITSU SIEMENS"), + DMI_MATCH(DMI_PRODUCT_NAME, "AMILO PRO V2030"), + }, + }, {} }; @@ -1595,7 +1545,7 @@ static int __init parse_acpi(char *arg) } /* acpi=rsdt use RSDT instead of XSDT */ else if (strcmp(arg, "rsdt") == 0) { - acpi_rsdt_forced = 1; + acpi_gbl_do_not_use_xsdt = TRUE; } /* "acpi=noirq" disables ACPI interrupt routing */ else if (strcmp(arg, "noirq") == 0) { @@ -1604,6 +1554,10 @@ static int __init parse_acpi(char *arg) /* "acpi=copy_dsdt" copys DSDT */ else if (strcmp(arg, "copy_dsdt") == 0) { acpi_gbl_copy_dsdt_locally = 1; + } + /* "acpi=nocmcff" disables FF mode for corrected errors */ + else if (strcmp(arg, "nocmcff") == 0) { + acpi_disable_cmcff = 1; } else { /* Core will printk when we return error. */ return -EINVAL; @@ -1694,3 +1648,9 @@ int __acpi_release_global_lock(unsigned int *lock) } while (unlikely (val != old)); return old & 0x1; } + +void __init arch_reserve_mem_area(acpi_physical_address addr, size_t size) +{ + e820_add_region(addr, size, E820_ACPI); + update_e820(); +} diff --git a/arch/x86/kernel/acpi/cstate.c b/arch/x86/kernel/acpi/cstate.c index f50e7fb2a20..4b28159e042 100644 --- a/arch/x86/kernel/acpi/cstate.c +++ b/arch/x86/kernel/acpi/cstate.c @@ -14,6 +14,7 @@ #include <acpi/processor.h> #include <asm/acpi.h> #include <asm/mwait.h> +#include <asm/special_insns.h> /* * Initialize bm_flags based on the CPU cache properties @@ -86,7 +87,9 @@ static long acpi_processor_ffh_cstate_probe_cpu(void *_cx) num_cstate_subtype = edx_part & MWAIT_SUBSTATE_MASK; retval = 0; - if (num_cstate_subtype < (cx->address & MWAIT_SUBSTATE_MASK)) { + /* If the HW does not support any sub-states in this C-state */ + if (num_cstate_subtype == 0) { + pr_warn(FW_BUG "ACPI MWAIT C-state 0x%x not supported by HW (0x%x)\n", cx->address, edx_part); retval = -1; goto out; } @@ -149,29 +152,6 @@ int acpi_processor_ffh_cstate_probe(unsigned int cpu, } EXPORT_SYMBOL_GPL(acpi_processor_ffh_cstate_probe); -/* - * This uses new MONITOR/MWAIT instructions on P4 processors with PNI, - * which can obviate IPI to trigger checking of need_resched. - * We execute MONITOR against need_resched and enter optimized wait state - * through MWAIT. Whenever someone changes need_resched, we would be woken - * up from MWAIT (without an IPI). - * - * New with Core Duo processors, MWAIT can take some hints based on CPU - * capability. - */ -void mwait_idle_with_hints(unsigned long ax, unsigned long cx) -{ - if (!need_resched()) { - if (this_cpu_has(X86_FEATURE_CLFLUSH_MONITOR)) - clflush((void *)¤t_thread_info()->flags); - - __monitor((void *)¤t_thread_info()->flags, 0, 0); - smp_mb(); - if (!need_resched()) - __mwait(ax, cx); - } -} - void acpi_processor_ffh_cstate_enter(struct acpi_processor_cx *cx) { unsigned int cpu = smp_processor_id(); diff --git a/arch/x86/kernel/acpi/realmode/.gitignore b/arch/x86/kernel/acpi/realmode/.gitignore deleted file mode 100644 index 58f1f48a58f..00000000000 --- a/arch/x86/kernel/acpi/realmode/.gitignore +++ /dev/null @@ -1,3 +0,0 @@ -wakeup.bin -wakeup.elf -wakeup.lds diff --git a/arch/x86/kernel/acpi/realmode/Makefile b/arch/x86/kernel/acpi/realmode/Makefile deleted file mode 100644 index 6a564ac67ef..00000000000 --- a/arch/x86/kernel/acpi/realmode/Makefile +++ /dev/null @@ -1,59 +0,0 @@ -# -# arch/x86/kernel/acpi/realmode/Makefile -# -# This file is subject to the terms and conditions of the GNU General Public -# License. See the file "COPYING" in the main directory of this archive -# for more details. -# - -always := wakeup.bin -targets := wakeup.elf wakeup.lds - -wakeup-y += wakeup.o wakemain.o video-mode.o copy.o bioscall.o regs.o - -# The link order of the video-*.o modules can matter. In particular, -# video-vga.o *must* be listed first, followed by video-vesa.o. -# Hardware-specific drivers should follow in the order they should be -# probed, and video-bios.o should typically be last. -wakeup-y += video-vga.o -wakeup-y += video-vesa.o -wakeup-y += video-bios.o - -targets += $(wakeup-y) - -bootsrc := $(src)/../../../boot - -# --------------------------------------------------------------------------- - -# How to compile the 16-bit code. Note we always compile for -march=i386, -# that way we can complain to the user if the CPU is insufficient. -# Compile with _SETUP since this is similar to the boot-time setup code. -KBUILD_CFLAGS := $(LINUXINCLUDE) -g -Os -D_SETUP -D_WAKEUP -D__KERNEL__ \ - -I$(srctree)/$(bootsrc) \ - $(cflags-y) \ - -Wall -Wstrict-prototypes \ - -march=i386 -mregparm=3 \ - -include $(srctree)/$(bootsrc)/code16gcc.h \ - -fno-strict-aliasing -fomit-frame-pointer \ - $(call cc-option, -ffreestanding) \ - $(call cc-option, -fno-toplevel-reorder,\ - $(call cc-option, -fno-unit-at-a-time)) \ - $(call cc-option, -fno-stack-protector) \ - $(call cc-option, -mpreferred-stack-boundary=2) -KBUILD_CFLAGS += $(call cc-option, -m32) -KBUILD_AFLAGS := $(KBUILD_CFLAGS) -D__ASSEMBLY__ -GCOV_PROFILE := n - -WAKEUP_OBJS = $(addprefix $(obj)/,$(wakeup-y)) - -LDFLAGS_wakeup.elf := -T - -CPPFLAGS_wakeup.lds += -P -C - -$(obj)/wakeup.elf: $(obj)/wakeup.lds $(WAKEUP_OBJS) FORCE - $(call if_changed,ld) - -OBJCOPYFLAGS_wakeup.bin := -O binary - -$(obj)/wakeup.bin: $(obj)/wakeup.elf FORCE - $(call if_changed,objcopy) diff --git a/arch/x86/kernel/acpi/realmode/bioscall.S b/arch/x86/kernel/acpi/realmode/bioscall.S deleted file mode 100644 index f51eb0bb56c..00000000000 --- a/arch/x86/kernel/acpi/realmode/bioscall.S +++ /dev/null @@ -1 +0,0 @@ -#include "../../../boot/bioscall.S" diff --git a/arch/x86/kernel/acpi/realmode/copy.S b/arch/x86/kernel/acpi/realmode/copy.S deleted file mode 100644 index dc59ebee69d..00000000000 --- a/arch/x86/kernel/acpi/realmode/copy.S +++ /dev/null @@ -1 +0,0 @@ -#include "../../../boot/copy.S" diff --git a/arch/x86/kernel/acpi/realmode/regs.c b/arch/x86/kernel/acpi/realmode/regs.c deleted file mode 100644 index 6206033ba20..00000000000 --- a/arch/x86/kernel/acpi/realmode/regs.c +++ /dev/null @@ -1 +0,0 @@ -#include "../../../boot/regs.c" diff --git a/arch/x86/kernel/acpi/realmode/video-bios.c b/arch/x86/kernel/acpi/realmode/video-bios.c deleted file mode 100644 index 7deabc144a2..00000000000 --- a/arch/x86/kernel/acpi/realmode/video-bios.c +++ /dev/null @@ -1 +0,0 @@ -#include "../../../boot/video-bios.c" diff --git a/arch/x86/kernel/acpi/realmode/video-mode.c b/arch/x86/kernel/acpi/realmode/video-mode.c deleted file mode 100644 index 328ad209f11..00000000000 --- a/arch/x86/kernel/acpi/realmode/video-mode.c +++ /dev/null @@ -1 +0,0 @@ -#include "../../../boot/video-mode.c" diff --git a/arch/x86/kernel/acpi/realmode/video-vesa.c b/arch/x86/kernel/acpi/realmode/video-vesa.c deleted file mode 100644 index 9dbb9672226..00000000000 --- a/arch/x86/kernel/acpi/realmode/video-vesa.c +++ /dev/null @@ -1 +0,0 @@ -#include "../../../boot/video-vesa.c" diff --git a/arch/x86/kernel/acpi/realmode/video-vga.c b/arch/x86/kernel/acpi/realmode/video-vga.c deleted file mode 100644 index bcc81255f37..00000000000 --- a/arch/x86/kernel/acpi/realmode/video-vga.c +++ /dev/null @@ -1 +0,0 @@ -#include "../../../boot/video-vga.c" diff --git a/arch/x86/kernel/acpi/realmode/wakemain.c b/arch/x86/kernel/acpi/realmode/wakemain.c deleted file mode 100644 index 883962d9eef..00000000000 --- a/arch/x86/kernel/acpi/realmode/wakemain.c +++ /dev/null @@ -1,81 +0,0 @@ -#include "wakeup.h" -#include "boot.h" - -static void udelay(int loops) -{ - while (loops--) - io_delay(); /* Approximately 1 us */ -} - -static void beep(unsigned int hz) -{ - u8 enable; - - if (!hz) { - enable = 0x00; /* Turn off speaker */ - } else { - u16 div = 1193181/hz; - - outb(0xb6, 0x43); /* Ctr 2, squarewave, load, binary */ - io_delay(); - outb(div, 0x42); /* LSB of counter */ - io_delay(); - outb(div >> 8, 0x42); /* MSB of counter */ - io_delay(); - - enable = 0x03; /* Turn on speaker */ - } - inb(0x61); /* Dummy read of System Control Port B */ - io_delay(); - outb(enable, 0x61); /* Enable timer 2 output to speaker */ - io_delay(); -} - -#define DOT_HZ 880 -#define DASH_HZ 587 -#define US_PER_DOT 125000 - -/* Okay, this is totally silly, but it's kind of fun. */ -static void send_morse(const char *pattern) -{ - char s; - - while ((s = *pattern++)) { - switch (s) { - case '.': - beep(DOT_HZ); - udelay(US_PER_DOT); - beep(0); - udelay(US_PER_DOT); - break; - case '-': - beep(DASH_HZ); - udelay(US_PER_DOT * 3); - beep(0); - udelay(US_PER_DOT); - break; - default: /* Assume it's a space */ - udelay(US_PER_DOT * 3); - break; - } - } -} - -void main(void) -{ - /* Kill machine if structures are wrong */ - if (wakeup_header.real_magic != 0x12345678) - while (1); - - if (wakeup_header.realmode_flags & 4) - send_morse("...-"); - - if (wakeup_header.realmode_flags & 1) - asm volatile("lcallw $0xc000,$3"); - - if (wakeup_header.realmode_flags & 2) { - /* Need to call BIOS */ - probe_cards(0); - set_mode(wakeup_header.video_mode); - } -} diff --git a/arch/x86/kernel/acpi/realmode/wakeup.S b/arch/x86/kernel/acpi/realmode/wakeup.S deleted file mode 100644 index b4fd836e405..00000000000 --- a/arch/x86/kernel/acpi/realmode/wakeup.S +++ /dev/null @@ -1,170 +0,0 @@ -/* - * ACPI wakeup real mode startup stub - */ -#include <asm/segment.h> -#include <asm/msr-index.h> -#include <asm/page_types.h> -#include <asm/pgtable_types.h> -#include <asm/processor-flags.h> -#include "wakeup.h" - - .code16 - .section ".jump", "ax" - .globl _start -_start: - cli - jmp wakeup_code - -/* This should match the structure in wakeup.h */ - .section ".header", "a" - .globl wakeup_header -wakeup_header: -video_mode: .short 0 /* Video mode number */ -pmode_return: .byte 0x66, 0xea /* ljmpl */ - .long 0 /* offset goes here */ - .short __KERNEL_CS -pmode_cr0: .long 0 /* Saved %cr0 */ -pmode_cr3: .long 0 /* Saved %cr3 */ -pmode_cr4: .long 0 /* Saved %cr4 */ -pmode_efer: .quad 0 /* Saved EFER */ -pmode_gdt: .quad 0 -pmode_misc_en: .quad 0 /* Saved MISC_ENABLE MSR */ -pmode_behavior: .long 0 /* Wakeup behavior flags */ -realmode_flags: .long 0 -real_magic: .long 0 -trampoline_segment: .word 0 -_pad1: .byte 0 -wakeup_jmp: .byte 0xea /* ljmpw */ -wakeup_jmp_off: .word 3f -wakeup_jmp_seg: .word 0 -wakeup_gdt: .quad 0, 0, 0 -signature: .long WAKEUP_HEADER_SIGNATURE - - .text - .code16 -wakeup_code: - cld - - /* Apparently some dimwit BIOS programmers don't know how to - program a PM to RM transition, and we might end up here with - junk in the data segment descriptor registers. The only way - to repair that is to go into PM and fix it ourselves... */ - movw $16, %cx - lgdtl %cs:wakeup_gdt - movl %cr0, %eax - orb $X86_CR0_PE, %al - movl %eax, %cr0 - jmp 1f -1: ljmpw $8, $2f -2: - movw %cx, %ds - movw %cx, %es - movw %cx, %ss - movw %cx, %fs - movw %cx, %gs - - andb $~X86_CR0_PE, %al - movl %eax, %cr0 - jmp wakeup_jmp -3: - /* Set up segments */ - movw %cs, %ax - movw %ax, %ds - movw %ax, %es - movw %ax, %ss - lidtl wakeup_idt - - movl $wakeup_stack_end, %esp - - /* Clear the EFLAGS */ - pushl $0 - popfl - - /* Check header signature... */ - movl signature, %eax - cmpl $WAKEUP_HEADER_SIGNATURE, %eax - jne bogus_real_magic - - /* Check we really have everything... */ - movl end_signature, %eax - cmpl $WAKEUP_END_SIGNATURE, %eax - jne bogus_real_magic - - /* Call the C code */ - calll main - - /* Restore MISC_ENABLE before entering protected mode, in case - BIOS decided to clear XD_DISABLE during S3. */ - movl pmode_behavior, %eax - btl $WAKEUP_BEHAVIOR_RESTORE_MISC_ENABLE, %eax - jnc 1f - - movl pmode_misc_en, %eax - movl pmode_misc_en + 4, %edx - movl $MSR_IA32_MISC_ENABLE, %ecx - wrmsr -1: - - /* Do any other stuff... */ - -#ifndef CONFIG_64BIT - /* This could also be done in C code... */ - movl pmode_cr3, %eax - movl %eax, %cr3 - - movl pmode_cr4, %ecx - jecxz 1f - movl %ecx, %cr4 -1: - movl pmode_efer, %eax - movl pmode_efer + 4, %edx - movl %eax, %ecx - orl %edx, %ecx - jz 1f - movl $MSR_EFER, %ecx - wrmsr -1: - - lgdtl pmode_gdt - - /* This really couldn't... */ - movl pmode_cr0, %eax - movl %eax, %cr0 - jmp pmode_return -#else - pushw $0 - pushw trampoline_segment - pushw $0 - lret -#endif - -bogus_real_magic: -1: - hlt - jmp 1b - - .data - .balign 8 - - /* This is the standard real-mode IDT */ -wakeup_idt: - .word 0xffff /* limit */ - .long 0 /* address */ - .word 0 - - .globl HEAP, heap_end -HEAP: - .long wakeup_heap -heap_end: - .long wakeup_stack - - .bss -wakeup_heap: - .space 2048 -wakeup_stack: - .space 2048 -wakeup_stack_end: - - .section ".signature","a" -end_signature: - .long WAKEUP_END_SIGNATURE diff --git a/arch/x86/kernel/acpi/realmode/wakeup.h b/arch/x86/kernel/acpi/realmode/wakeup.h deleted file mode 100644 index 97a29e1430e..00000000000 --- a/arch/x86/kernel/acpi/realmode/wakeup.h +++ /dev/null @@ -1,48 +0,0 @@ -/* - * Definitions for the wakeup data structure at the head of the - * wakeup code. - */ - -#ifndef ARCH_X86_KERNEL_ACPI_RM_WAKEUP_H -#define ARCH_X86_KERNEL_ACPI_RM_WAKEUP_H - -#ifndef __ASSEMBLY__ -#include <linux/types.h> - -/* This must match data at wakeup.S */ -struct wakeup_header { - u16 video_mode; /* Video mode number */ - u16 _jmp1; /* ljmpl opcode, 32-bit only */ - u32 pmode_entry; /* Protected mode resume point, 32-bit only */ - u16 _jmp2; /* CS value, 32-bit only */ - u32 pmode_cr0; /* Protected mode cr0 */ - u32 pmode_cr3; /* Protected mode cr3 */ - u32 pmode_cr4; /* Protected mode cr4 */ - u32 pmode_efer_low; /* Protected mode EFER */ - u32 pmode_efer_high; - u64 pmode_gdt; - u32 pmode_misc_en_low; /* Protected mode MISC_ENABLE */ - u32 pmode_misc_en_high; - u32 pmode_behavior; /* Wakeup routine behavior flags */ - u32 realmode_flags; - u32 real_magic; - u16 trampoline_segment; /* segment with trampoline code, 64-bit only */ - u8 _pad1; - u8 wakeup_jmp; - u16 wakeup_jmp_off; - u16 wakeup_jmp_seg; - u64 wakeup_gdt[3]; - u32 signature; /* To check we have correct structure */ -} __attribute__((__packed__)); - -extern struct wakeup_header wakeup_header; -#endif - -#define WAKEUP_HEADER_OFFSET 8 -#define WAKEUP_HEADER_SIGNATURE 0x51ee1111 -#define WAKEUP_END_SIGNATURE 0x65a22c82 - -/* Wakeup behavior bits */ -#define WAKEUP_BEHAVIOR_RESTORE_MISC_ENABLE 0 - -#endif /* ARCH_X86_KERNEL_ACPI_RM_WAKEUP_H */ diff --git a/arch/x86/kernel/acpi/realmode/wakeup.lds.S b/arch/x86/kernel/acpi/realmode/wakeup.lds.S deleted file mode 100644 index d4f8010a5b1..00000000000 --- a/arch/x86/kernel/acpi/realmode/wakeup.lds.S +++ /dev/null @@ -1,62 +0,0 @@ -/* - * wakeup.ld - * - * Linker script for the real-mode wakeup code - */ -#undef i386 -#include "wakeup.h" - -OUTPUT_FORMAT("elf32-i386", "elf32-i386", "elf32-i386") -OUTPUT_ARCH(i386) -ENTRY(_start) - -SECTIONS -{ - . = 0; - .jump : { - *(.jump) - } = 0x90909090 - - . = WAKEUP_HEADER_OFFSET; - .header : { - *(.header) - } - - . = ALIGN(16); - .text : { - *(.text*) - } = 0x90909090 - - . = ALIGN(16); - .rodata : { - *(.rodata*) - } - - .videocards : { - video_cards = .; - *(.videocards) - video_cards_end = .; - } - - . = ALIGN(16); - .data : { - *(.data*) - } - - . = ALIGN(16); - .bss : { - __bss_start = .; - *(.bss) - __bss_end = .; - } - - .signature : { - *(.signature) - } - - _end = .; - - /DISCARD/ : { - *(.note*) - } -} diff --git a/arch/x86/kernel/acpi/sleep.c b/arch/x86/kernel/acpi/sleep.c index 103b6ab368d..31368207837 100644 --- a/arch/x86/kernel/acpi/sleep.c +++ b/arch/x86/kernel/acpi/sleep.c @@ -14,8 +14,9 @@ #include <asm/desc.h> #include <asm/pgtable.h> #include <asm/cacheflush.h> +#include <asm/realmode.h> -#include "realmode/wakeup.h" +#include "../../realmode/rm/wakeup.h" #include "sleep.h" unsigned long acpi_realmode_flags; @@ -25,20 +26,27 @@ static char temp_stack[4096]; #endif /** - * acpi_suspend_lowlevel - save kernel state + * x86_acpi_enter_sleep_state - enter sleep state + * @state: Sleep state to enter. + * + * Wrapper around acpi_enter_sleep_state() to be called by assmebly. + */ +acpi_status asmlinkage __visible x86_acpi_enter_sleep_state(u8 state) +{ + return acpi_enter_sleep_state(state); +} + +/** + * x86_acpi_suspend_lowlevel - save kernel state * * Create an identity mapped page table and copy the wakeup routine to * low memory. */ -int acpi_suspend_lowlevel(void) +int x86_acpi_suspend_lowlevel(void) { - struct wakeup_header *header; - /* address in low memory of the wakeup routine. */ - char *acpi_realmode; + struct wakeup_header *header = + (struct wakeup_header *) __va(real_mode_header->wakeup_header); - acpi_realmode = TRAMPOLINE_SYM(acpi_wakeup_code); - - header = (struct wakeup_header *)(acpi_realmode + WAKEUP_HEADER_OFFSET); if (header->signature != WAKEUP_HEADER_SIGNATURE) { printk(KERN_ERR "wakeup header does not match\n"); return -EINVAL; @@ -46,41 +54,39 @@ int acpi_suspend_lowlevel(void) header->video_mode = saved_video_mode; - header->wakeup_jmp_seg = acpi_wakeup_address >> 4; - - /* - * Set up the wakeup GDT. We set these up as Big Real Mode, - * that is, with limits set to 4 GB. At least the Lenovo - * Thinkpad X61 is known to need this for the video BIOS - * initialization quirk to work; this is likely to also - * be the case for other laptops or integrated video devices. - */ - - /* GDT[0]: GDT self-pointer */ - header->wakeup_gdt[0] = - (u64)(sizeof(header->wakeup_gdt) - 1) + - ((u64)__pa(&header->wakeup_gdt) << 16); - /* GDT[1]: big real mode-like code segment */ - header->wakeup_gdt[1] = - GDT_ENTRY(0x809b, acpi_wakeup_address, 0xfffff); - /* GDT[2]: big real mode-like data segment */ - header->wakeup_gdt[2] = - GDT_ENTRY(0x8093, acpi_wakeup_address, 0xfffff); + header->pmode_behavior = 0; #ifndef CONFIG_64BIT - store_gdt((struct desc_ptr *)&header->pmode_gdt); + native_store_gdt((struct desc_ptr *)&header->pmode_gdt); - if (rdmsr_safe(MSR_EFER, &header->pmode_efer_low, - &header->pmode_efer_high)) - header->pmode_efer_low = header->pmode_efer_high = 0; + /* + * We have to check that we can write back the value, and not + * just read it. At least on 90 nm Pentium M (Family 6, Model + * 13), reading an invalid MSR is not guaranteed to trap, see + * Erratum X4 in "Intel Pentium M Processor on 90 nm Process + * with 2-MB L2 Cache and Intel® Processor A100 and A110 on 90 + * nm process with 512-KB L2 Cache Specification Update". + */ + if (!rdmsr_safe(MSR_EFER, + &header->pmode_efer_low, + &header->pmode_efer_high) && + !wrmsr_safe(MSR_EFER, + header->pmode_efer_low, + header->pmode_efer_high)) + header->pmode_behavior |= (1 << WAKEUP_BEHAVIOR_RESTORE_EFER); #endif /* !CONFIG_64BIT */ header->pmode_cr0 = read_cr0(); - header->pmode_cr4 = read_cr4_safe(); - header->pmode_behavior = 0; + if (__this_cpu_read(cpu_info.cpuid_level) >= 0) { + header->pmode_cr4 = read_cr4(); + header->pmode_behavior |= (1 << WAKEUP_BEHAVIOR_RESTORE_CR4); + } if (!rdmsr_safe(MSR_IA32_MISC_ENABLE, &header->pmode_misc_en_low, - &header->pmode_misc_en_high)) + &header->pmode_misc_en_high) && + !wrmsr_safe(MSR_IA32_MISC_ENABLE, + header->pmode_misc_en_low, + header->pmode_misc_en_high)) header->pmode_behavior |= (1 << WAKEUP_BEHAVIOR_RESTORE_MISC_ENABLE); header->realmode_flags = acpi_realmode_flags; @@ -88,10 +94,9 @@ int acpi_suspend_lowlevel(void) #ifndef CONFIG_64BIT header->pmode_entry = (u32)&wakeup_pmode_return; - header->pmode_cr3 = (u32)__pa(&initial_page_table); + header->pmode_cr3 = (u32)__pa_symbol(initial_page_table); saved_magic = 0x12345678; #else /* CONFIG_64BIT */ - header->trampoline_segment = trampoline_address() >> 4; #ifdef CONFIG_SMP stack_start = (unsigned long)temp_stack + sizeof(temp_stack); early_gdt_descr.address = @@ -121,6 +126,8 @@ static int __init acpi_sleep_setup(char *str) #endif if (strncmp(str, "nonvs", 5) == 0) acpi_nvs_nosave(); + if (strncmp(str, "nonvs_s3", 8) == 0) + acpi_nvs_nosave_s3(); if (strncmp(str, "old_ordering", 12) == 0) acpi_old_suspend_ordering(); str = strchr(str, ','); diff --git a/arch/x86/kernel/acpi/sleep.h b/arch/x86/kernel/acpi/sleep.h index 416d4be13fe..65c7b606b60 100644 --- a/arch/x86/kernel/acpi/sleep.h +++ b/arch/x86/kernel/acpi/sleep.h @@ -2,14 +2,20 @@ * Variables and functions used by the code in sleep.c */ -#include <asm/trampoline.h> +#include <asm/realmode.h> extern unsigned long saved_video_mode; extern long saved_magic; extern int wakeup_pmode_return; +extern u8 wake_sleep_flags; + extern unsigned long acpi_copy_wakeup_routine(unsigned long); extern void wakeup_long64(void); extern void do_suspend_lowlevel(void); + +extern int x86_acpi_suspend_lowlevel(void); + +acpi_status asmlinkage x86_acpi_enter_sleep_state(u8 state); diff --git a/arch/x86/kernel/acpi/wakeup_32.S b/arch/x86/kernel/acpi/wakeup_32.S index 13ab720573e..665c6b7d2ea 100644 --- a/arch/x86/kernel/acpi/wakeup_32.S +++ b/arch/x86/kernel/acpi/wakeup_32.S @@ -1,4 +1,4 @@ - .section .text..page_aligned + .text #include <linux/linkage.h> #include <asm/segment.h> #include <asm/page_types.h> @@ -18,7 +18,6 @@ wakeup_pmode_return: movw %ax, %gs # reload the gdt, as we need the full 32 bit address - lgdt saved_gdt lidt saved_idt lldt saved_ldt ljmp $(__KERNEL_CS), $1f @@ -44,7 +43,6 @@ bogus_magic: save_registers: - sgdt saved_gdt sidt saved_idt sldt saved_ldt str saved_tss @@ -75,7 +73,7 @@ ENTRY(do_suspend_lowlevel) call save_processor_state call save_registers pushl $3 - call acpi_enter_sleep_state + call x86_acpi_enter_sleep_state addl $4, %esp # In case of S3 failure, we'll emerge here. Jump @@ -93,7 +91,6 @@ ENTRY(saved_magic) .long 0 ENTRY(saved_eip) .long 0 # saved registers -saved_gdt: .long 0,0 saved_idt: .long 0,0 saved_ldt: .long 0 saved_tss: .long 0 diff --git a/arch/x86/kernel/acpi/wakeup_64.S b/arch/x86/kernel/acpi/wakeup_64.S index 8ea5164cbd0..ae693b51ed8 100644 --- a/arch/x86/kernel/acpi/wakeup_64.S +++ b/arch/x86/kernel/acpi/wakeup_64.S @@ -73,7 +73,7 @@ ENTRY(do_suspend_lowlevel) addq $8, %rsp movl $3, %edi xorl %eax, %eax - call acpi_enter_sleep_state + call x86_acpi_enter_sleep_state /* in case something went wrong, restore the machine status and go on */ jmp resume_point diff --git a/arch/x86/kernel/acpi/wakeup_rm.S b/arch/x86/kernel/acpi/wakeup_rm.S deleted file mode 100644 index 63b8ab524f2..00000000000 --- a/arch/x86/kernel/acpi/wakeup_rm.S +++ /dev/null @@ -1,12 +0,0 @@ -/* - * Wrapper script for the realmode binary as a transport object - * before copying to low memory. - */ -#include <asm/page_types.h> - - .section ".x86_trampoline","a" - .balign PAGE_SIZE - .globl acpi_wakeup_code -acpi_wakeup_code: - .incbin "arch/x86/kernel/acpi/realmode/wakeup.bin" - .size acpi_wakeup_code, .-acpi_wakeup_code diff --git a/arch/x86/kernel/alternative.c b/arch/x86/kernel/alternative.c index 1f84794f075..703130f469e 100644 --- a/arch/x86/kernel/alternative.c +++ b/arch/x86/kernel/alternative.c @@ -1,14 +1,16 @@ +#define pr_fmt(fmt) "SMP alternatives: " fmt + #include <linux/module.h> #include <linux/sched.h> #include <linux/mutex.h> #include <linux/list.h> #include <linux/stringify.h> -#include <linux/kprobes.h> #include <linux/mm.h> #include <linux/vmalloc.h> #include <linux/memory.h> #include <linux/stop_machine.h> #include <linux/slab.h> +#include <linux/kdebug.h> #include <asm/alternative.h> #include <asm/sections.h> #include <asm/pgtable.h> @@ -21,19 +23,6 @@ #define MAX_PATCH_LEN (255-1) -#ifdef CONFIG_HOTPLUG_CPU -static int smp_alt_once; - -static int __init bootonly(char *str) -{ - smp_alt_once = 1; - return 1; -} -__setup("smp-alt-boot", bootonly); -#else -#define smp_alt_once 1 -#endif - static int __initdata_or_module debug_alternative; static int __init debug_alt(char *str) @@ -63,8 +52,11 @@ static int __init setup_noreplace_paravirt(char *str) __setup("noreplace-paravirt", setup_noreplace_paravirt); #endif -#define DPRINTK(fmt, args...) if (debug_alternative) \ - printk(KERN_DEBUG fmt, args) +#define DPRINTK(fmt, ...) \ +do { \ + if (debug_alternative) \ + printk(KERN_DEBUG fmt, ##__VA_ARGS__); \ +} while (0) /* * Each GENERIC_NOPX is of X bytes, and defined as an array of bytes @@ -160,7 +152,7 @@ static const unsigned char * const k7_nops[ASM_NOP_MAX+2] = #endif #ifdef P6_NOP1 -static const unsigned char __initconst_or_module p6nops[] = +static const unsigned char p6nops[] = { P6_NOP1, P6_NOP2, @@ -219,7 +211,7 @@ void __init arch_init_ideal_nops(void) ideal_nops = intel_nops; #endif } - + break; default: #ifdef CONFIG_X86_64 ideal_nops = k8_nops; @@ -279,7 +271,7 @@ void __init_or_module apply_alternatives(struct alt_instr *start, replacement = (u8 *)&a->repl_offset + a->repl_offset; BUG_ON(a->replacementlen > a->instrlen); BUG_ON(a->instrlen > sizeof(insnbuf)); - BUG_ON(a->cpuid >= NCAPINTS*32); + BUG_ON(a->cpuid >= (NCAPINTS + NBUGINTS) * 32); if (!boot_cpu_has(a->cpuid)) continue; @@ -312,7 +304,7 @@ static void alternatives_smp_lock(const s32 *start, const s32 *end, /* turn DS segment override prefix into lock prefix */ if (*ptr == 0x3e) text_poke(ptr, ((unsigned char []){0xf0}), 1); - }; + } mutex_unlock(&text_mutex); } @@ -321,9 +313,6 @@ static void alternatives_smp_unlock(const s32 *start, const s32 *end, { const s32 *poff; - if (noreplace_smp) - return; - mutex_lock(&text_mutex); for (poff = start; poff < end; poff++) { u8 *ptr = (u8 *)poff + *poff; @@ -333,7 +322,7 @@ static void alternatives_smp_unlock(const s32 *start, const s32 *end, /* turn lock prefix into DS segment override prefix */ if (*ptr == 0xf0) text_poke(ptr, ((unsigned char []){0x3E}), 1); - }; + } mutex_unlock(&text_mutex); } @@ -354,7 +343,7 @@ struct smp_alt_module { }; static LIST_HEAD(smp_alt_modules); static DEFINE_MUTEX(smp_alt); -static int smp_mode = 1; /* protected by smp_alt */ +static bool uniproc_patched = false; /* protected by smp_alt */ void __init_or_module alternatives_smp_module_add(struct module *mod, char *name, @@ -363,19 +352,18 @@ void __init_or_module alternatives_smp_module_add(struct module *mod, { struct smp_alt_module *smp; - if (noreplace_smp) - return; + mutex_lock(&smp_alt); + if (!uniproc_patched) + goto unlock; - if (smp_alt_once) { - if (boot_cpu_has(X86_FEATURE_UP)) - alternatives_smp_unlock(locks, locks_end, - text, text_end); - return; - } + if (num_possible_cpus() == 1) + /* Don't bother remembering, we'll never have to undo it. */ + goto smp_unlock; smp = kzalloc(sizeof(*smp), GFP_KERNEL); if (NULL == smp) - return; /* we'll run the (safe but slow) SMP code then ... */ + /* we'll run the (safe but slow) SMP code then ... */ + goto unlock; smp->mod = mod; smp->name = name; @@ -387,11 +375,10 @@ void __init_or_module alternatives_smp_module_add(struct module *mod, __func__, smp->locks, smp->locks_end, smp->text, smp->text_end, smp->name); - mutex_lock(&smp_alt); list_add_tail(&smp->next, &smp_alt_modules); - if (boot_cpu_has(X86_FEATURE_UP)) - alternatives_smp_unlock(smp->locks, smp->locks_end, - smp->text, smp->text_end); +smp_unlock: + alternatives_smp_unlock(locks, locks_end, text, text_end); +unlock: mutex_unlock(&smp_alt); } @@ -399,66 +386,36 @@ void __init_or_module alternatives_smp_module_del(struct module *mod) { struct smp_alt_module *item; - if (smp_alt_once || noreplace_smp) - return; - mutex_lock(&smp_alt); list_for_each_entry(item, &smp_alt_modules, next) { if (mod != item->mod) continue; list_del(&item->next); - mutex_unlock(&smp_alt); - DPRINTK("%s: %s\n", __func__, item->name); kfree(item); - return; + break; } mutex_unlock(&smp_alt); } -bool skip_smp_alternatives; -void alternatives_smp_switch(int smp) +void alternatives_enable_smp(void) { struct smp_alt_module *mod; -#ifdef CONFIG_LOCKDEP - /* - * Older binutils section handling bug prevented - * alternatives-replacement from working reliably. - * - * If this still occurs then you should see a hang - * or crash shortly after this line: - */ - printk("lockdep: fixing up alternatives.\n"); -#endif - - if (noreplace_smp || smp_alt_once || skip_smp_alternatives) - return; - BUG_ON(!smp && (num_online_cpus() > 1)); + /* Why bother if there are no other CPUs? */ + BUG_ON(num_possible_cpus() == 1); mutex_lock(&smp_alt); - /* - * Avoid unnecessary switches because it forces JIT based VMs to - * throw away all cached translations, which can be quite costly. - */ - if (smp == smp_mode) { - /* nothing */ - } else if (smp) { - printk(KERN_INFO "SMP alternatives: switching to SMP code\n"); + if (uniproc_patched) { + pr_info("switching to SMP code\n"); + BUG_ON(num_online_cpus() != 1); clear_cpu_cap(&boot_cpu_data, X86_FEATURE_UP); clear_cpu_cap(&cpu_data(0), X86_FEATURE_UP); list_for_each_entry(mod, &smp_alt_modules, next) alternatives_smp_lock(mod->locks, mod->locks_end, mod->text, mod->text_end); - } else { - printk(KERN_INFO "SMP alternatives: switching to UP code\n"); - set_cpu_cap(&boot_cpu_data, X86_FEATURE_UP); - set_cpu_cap(&cpu_data(0), X86_FEATURE_UP); - list_for_each_entry(mod, &smp_alt_modules, next) - alternatives_smp_unlock(mod->locks, mod->locks_end, - mod->text, mod->text_end); + uniproc_patched = false; } - smp_mode = smp; mutex_unlock(&smp_alt); } @@ -535,40 +492,22 @@ void __init alternative_instructions(void) apply_alternatives(__alt_instructions, __alt_instructions_end); - /* switch to patch-once-at-boottime-only mode and free the - * tables in case we know the number of CPUs will never ever - * change */ -#ifdef CONFIG_HOTPLUG_CPU - if (num_possible_cpus() < 2) - smp_alt_once = 1; -#endif - #ifdef CONFIG_SMP - if (smp_alt_once) { - if (1 == num_possible_cpus()) { - printk(KERN_INFO "SMP alternatives: switching to UP code\n"); - set_cpu_cap(&boot_cpu_data, X86_FEATURE_UP); - set_cpu_cap(&cpu_data(0), X86_FEATURE_UP); - - alternatives_smp_unlock(__smp_locks, __smp_locks_end, - _text, _etext); - } - } else { + /* Patch to UP if other cpus not imminent. */ + if (!noreplace_smp && (num_present_cpus() == 1 || setup_max_cpus <= 1)) { + uniproc_patched = true; alternatives_smp_module_add(NULL, "core kernel", __smp_locks, __smp_locks_end, _text, _etext); - - /* Only switch to UP mode if we don't immediately boot others */ - if (num_present_cpus() == 1 || setup_max_cpus <= 1) - alternatives_smp_switch(0); } -#endif - apply_paravirt(__parainstructions, __parainstructions_end); - if (smp_alt_once) + if (!uniproc_patched || num_possible_cpus() == 1) free_init_pages("SMP alternatives", (unsigned long)__smp_locks, (unsigned long)__smp_locks_end); +#endif + + apply_paravirt(__parainstructions, __parainstructions_end); restart_nmi(); } @@ -611,7 +550,7 @@ void *__init_or_module text_poke_early(void *addr, const void *opcode, * * Note: Must be called under text_mutex. */ -void *__kprobes text_poke(void *addr, const void *opcode, size_t len) +void *text_poke(void *addr, const void *opcode, size_t len) { unsigned long flags; char *vaddr; @@ -646,97 +585,93 @@ void *__kprobes text_poke(void *addr, const void *opcode, size_t len) return addr; } -/* - * Cross-modifying kernel text with stop_machine(). - * This code originally comes from immediate value. - */ -static atomic_t stop_machine_first; -static int wrote_text; +static void do_sync_core(void *info) +{ + sync_core(); +} -struct text_poke_params { - struct text_poke_param *params; - int nparams; -}; +static bool bp_patching_in_progress; +static void *bp_int3_handler, *bp_int3_addr; -static int __kprobes stop_machine_text_poke(void *data) +int poke_int3_handler(struct pt_regs *regs) { - struct text_poke_params *tpp = data; - struct text_poke_param *p; - int i; + /* bp_patching_in_progress */ + smp_rmb(); - if (atomic_dec_and_test(&stop_machine_first)) { - for (i = 0; i < tpp->nparams; i++) { - p = &tpp->params[i]; - text_poke(p->addr, p->opcode, p->len); - } - smp_wmb(); /* Make sure other cpus see that this has run */ - wrote_text = 1; - } else { - while (!wrote_text) - cpu_relax(); - smp_mb(); /* Load wrote_text before following execution */ - } + if (likely(!bp_patching_in_progress)) + return 0; - for (i = 0; i < tpp->nparams; i++) { - p = &tpp->params[i]; - flush_icache_range((unsigned long)p->addr, - (unsigned long)p->addr + p->len); - } - /* - * Intel Archiecture Software Developer's Manual section 7.1.3 specifies - * that a core serializing instruction such as "cpuid" should be - * executed on _each_ core before the new instruction is made visible. - */ - sync_core(); - return 0; -} + if (user_mode_vm(regs) || regs->ip != (unsigned long)bp_int3_addr) + return 0; + + /* set up the specified breakpoint handler */ + regs->ip = (unsigned long) bp_int3_handler; + + return 1; -/** - * text_poke_smp - Update instructions on a live kernel on SMP - * @addr: address to modify - * @opcode: source of the copy - * @len: length to copy - * - * Modify multi-byte instruction by using stop_machine() on SMP. This allows - * user to poke/set multi-byte text on SMP. Only non-NMI/MCE code modifying - * should be allowed, since stop_machine() does _not_ protect code against - * NMI and MCE. - * - * Note: Must be called under get_online_cpus() and text_mutex. - */ -void *__kprobes text_poke_smp(void *addr, const void *opcode, size_t len) -{ - struct text_poke_params tpp; - struct text_poke_param p; - - p.addr = addr; - p.opcode = opcode; - p.len = len; - tpp.params = &p; - tpp.nparams = 1; - atomic_set(&stop_machine_first, 1); - wrote_text = 0; - /* Use __stop_machine() because the caller already got online_cpus. */ - __stop_machine(stop_machine_text_poke, (void *)&tpp, cpu_online_mask); - return addr; } /** - * text_poke_smp_batch - Update instructions on a live kernel on SMP - * @params: an array of text_poke parameters - * @n: the number of elements in params. + * text_poke_bp() -- update instructions on live kernel on SMP + * @addr: address to patch + * @opcode: opcode of new instruction + * @len: length to copy + * @handler: address to jump to when the temporary breakpoint is hit * - * Modify multi-byte instruction by using stop_machine() on SMP. Since the - * stop_machine() is heavy task, it is better to aggregate text_poke requests - * and do it once if possible. + * Modify multi-byte instruction by using int3 breakpoint on SMP. + * We completely avoid stop_machine() here, and achieve the + * synchronization using int3 breakpoint. * - * Note: Must be called under get_online_cpus() and text_mutex. + * The way it is done: + * - add a int3 trap to the address that will be patched + * - sync cores + * - update all but the first byte of the patched range + * - sync cores + * - replace the first byte (int3) by the first byte of + * replacing opcode + * - sync cores + * + * Note: must be called under text_mutex. */ -void __kprobes text_poke_smp_batch(struct text_poke_param *params, int n) +void *text_poke_bp(void *addr, const void *opcode, size_t len, void *handler) { - struct text_poke_params tpp = {.params = params, .nparams = n}; + unsigned char int3 = 0xcc; + + bp_int3_handler = handler; + bp_int3_addr = (u8 *)addr + sizeof(int3); + bp_patching_in_progress = true; + /* + * Corresponding read barrier in int3 notifier for + * making sure the in_progress flags is correctly ordered wrt. + * patching + */ + smp_wmb(); + + text_poke(addr, &int3, sizeof(int3)); - atomic_set(&stop_machine_first, 1); - wrote_text = 0; - __stop_machine(stop_machine_text_poke, (void *)&tpp, cpu_online_mask); + on_each_cpu(do_sync_core, NULL, 1); + + if (len - sizeof(int3) > 0) { + /* patch all but the first byte */ + text_poke((char *)addr + sizeof(int3), + (const char *) opcode + sizeof(int3), + len - sizeof(int3)); + /* + * According to Intel, this core syncing is very likely + * not necessary and we'd be safe even without it. But + * better safe than sorry (plus there's not only Intel). + */ + on_each_cpu(do_sync_core, NULL, 1); + } + + /* patch the first byte */ + text_poke(addr, opcode, sizeof(int3)); + + on_each_cpu(do_sync_core, NULL, 1); + + bp_patching_in_progress = false; + smp_wmb(); + + return addr; } + diff --git a/arch/x86/kernel/amd_gart_64.c b/arch/x86/kernel/amd_gart_64.c index b1e7c7f7a0a..8e3842fc8be 100644 --- a/arch/x86/kernel/amd_gart_64.c +++ b/arch/x86/kernel/amd_gart_64.c @@ -477,7 +477,7 @@ error: /* allocate and map a coherent mapping */ static void * gart_alloc_coherent(struct device *dev, size_t size, dma_addr_t *dma_addr, - gfp_t flag) + gfp_t flag, struct dma_attrs *attrs) { dma_addr_t paddr; unsigned long align_mask; @@ -500,7 +500,8 @@ gart_alloc_coherent(struct device *dev, size_t size, dma_addr_t *dma_addr, } __free_pages(page, get_order(size)); } else - return dma_generic_alloc_coherent(dev, size, dma_addr, flag); + return dma_generic_alloc_coherent(dev, size, dma_addr, flag, + attrs); return NULL; } @@ -508,10 +509,10 @@ gart_alloc_coherent(struct device *dev, size_t size, dma_addr_t *dma_addr, /* free a coherent mapping */ static void gart_free_coherent(struct device *dev, size_t size, void *vaddr, - dma_addr_t dma_addr) + dma_addr_t dma_addr, struct dma_attrs *attrs) { gart_unmap_page(dev, dma_addr, size, DMA_BIDIRECTIONAL, NULL); - free_pages((unsigned long)vaddr, get_order(size)); + dma_generic_free_coherent(dev, size, vaddr, dma_addr, attrs); } static int gart_mapping_error(struct device *dev, dma_addr_t dma_addr) @@ -700,8 +701,8 @@ static struct dma_map_ops gart_dma_ops = { .unmap_sg = gart_unmap_sg, .map_page = gart_map_page, .unmap_page = gart_unmap_page, - .alloc_coherent = gart_alloc_coherent, - .free_coherent = gart_free_coherent, + .alloc = gart_alloc_coherent, + .free = gart_free_coherent, .mapping_error = gart_mapping_error, }; @@ -767,10 +768,9 @@ int __init gart_iommu_init(void) aper_base = info.aper_base; end_pfn = (aper_base>>PAGE_SHIFT) + (aper_size>>PAGE_SHIFT); - if (end_pfn > max_low_pfn_mapped) { - start_pfn = (aper_base>>PAGE_SHIFT); + start_pfn = PFN_DOWN(aper_base); + if (!pfn_range_is_mapped(start_pfn, end_pfn)) init_memory_mapping(start_pfn<<PAGE_SHIFT, end_pfn<<PAGE_SHIFT); - } pr_info("PCI-DMA: using GART IOMMU.\n"); iommu_size = check_iommu_size(info.aper_base, aper_size); diff --git a/arch/x86/kernel/amd_nb.c b/arch/x86/kernel/amd_nb.c index be16854591c..f04dbb3069b 100644 --- a/arch/x86/kernel/amd_nb.c +++ b/arch/x86/kernel/amd_nb.c @@ -2,6 +2,9 @@ * Shared support code for AMD K8 northbridges and derivates. * Copyright 2006 Andi Kleen, SUSE Labs. Subject to GPLv2. */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/types.h> #include <linux/slab.h> #include <linux/init.h> @@ -16,12 +19,19 @@ const struct pci_device_id amd_nb_misc_ids[] = { { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_K8_NB_MISC) }, { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_10H_NB_MISC) }, { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_15H_NB_F3) }, + { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_15H_M10H_F3) }, + { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_15H_M30H_NB_F3) }, + { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_16H_NB_F3) }, + { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_16H_M30H_NB_F3) }, {} }; EXPORT_SYMBOL(amd_nb_misc_ids); -static struct pci_device_id amd_nb_link_ids[] = { +static const struct pci_device_id amd_nb_link_ids[] = { { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_15H_NB_F4) }, + { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_15H_M30H_NB_F4) }, + { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_16H_NB_F4) }, + { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_16H_M30H_NB_F4) }, {} }; @@ -75,14 +85,20 @@ int amd_cache_northbridges(void) next_northbridge(misc, amd_nb_misc_ids); node_to_amd_nb(i)->link = link = next_northbridge(link, amd_nb_link_ids); - } + } - /* some CPU families (e.g. family 0x11) do not support GART */ + /* GART present only on Fam15h upto model 0fh */ if (boot_cpu_data.x86 == 0xf || boot_cpu_data.x86 == 0x10 || - boot_cpu_data.x86 == 0x15) + (boot_cpu_data.x86 == 0x15 && boot_cpu_data.x86_model < 0x10)) amd_northbridges.flags |= AMD_NB_GART; /* + * Check for L3 cache presence. + */ + if (!cpuid_edx(0x80000006)) + return 0; + + /* * Some CPU families support L3 Cache Index Disable. There are some * limitations because of E382 and E388 on family 0x10. */ @@ -165,7 +181,7 @@ int amd_get_subcaches(int cpu) return (mask >> (4 * cuid)) & 0xf; } -int amd_set_subcaches(int cpu, int mask) +int amd_set_subcaches(int cpu, unsigned long mask) { static unsigned int reset, ban; struct amd_northbridge *nb = node_to_amd_nb(amd_get_nb_id(cpu)); @@ -258,7 +274,7 @@ void amd_flush_garts(void) } spin_unlock_irqrestore(&gart_lock, flags); if (!flushed) - printk("nothing to flush?\n"); + pr_notice("nothing to flush?\n"); } EXPORT_SYMBOL_GPL(amd_flush_garts); @@ -269,11 +285,10 @@ static __init int init_amd_nbs(void) err = amd_cache_northbridges(); if (err < 0) - printk(KERN_NOTICE "AMD NB: Cannot enumerate AMD northbridges.\n"); + pr_notice("Cannot enumerate AMD northbridges\n"); if (amd_cache_gart() < 0) - printk(KERN_NOTICE "AMD NB: Cannot initialize GART flush words, " - "GART support disabled.\n"); + pr_notice("Cannot initialize GART flush words, GART support disabled\n"); return err; } diff --git a/arch/x86/kernel/apb_timer.c b/arch/x86/kernel/apb_timer.c index afdc3f756de..af5b08ab3b7 100644 --- a/arch/x86/kernel/apb_timer.c +++ b/arch/x86/kernel/apb_timer.c @@ -40,7 +40,7 @@ #include <asm/fixmap.h> #include <asm/apb_timer.h> -#include <asm/mrst.h> +#include <asm/intel-mid.h> #include <asm/time.h> #define APBT_CLOCKEVENT_RATING 110 @@ -157,13 +157,13 @@ static int __init apbt_clockevent_register(void) adev->num = smp_processor_id(); adev->timer = dw_apb_clockevent_init(smp_processor_id(), "apbt0", - mrst_timer_options == MRST_TIMER_LAPIC_APBT ? + intel_mid_timer_options == INTEL_MID_TIMER_LAPIC_APBT ? APBT_CLOCKEVENT_RATING - 100 : APBT_CLOCKEVENT_RATING, adev_virt_addr(adev), 0, apbt_freq); /* Firmware does EOI handling for us. */ adev->timer->eoi = NULL; - if (mrst_timer_options == MRST_TIMER_LAPIC_APBT) { + if (intel_mid_timer_options == INTEL_MID_TIMER_LAPIC_APBT) { global_clock_event = &adev->timer->ced; printk(KERN_DEBUG "%s clockevent registered as global\n", global_clock_event->name); @@ -240,7 +240,7 @@ static int apbt_cpuhp_notify(struct notifier_block *n, dw_apb_clockevent_pause(adev->timer); if (system_state == SYSTEM_RUNNING) { pr_debug("skipping APBT CPU %lu offline\n", cpu); - } else if (adev) { + } else { pr_debug("APBT clockevent for cpu %lu offline\n", cpu); dw_apb_clockevent_stop(adev->timer); } @@ -253,7 +253,7 @@ static int apbt_cpuhp_notify(struct notifier_block *n, static __init int apbt_late_init(void) { - if (mrst_timer_options == MRST_TIMER_LAPIC_APBT || + if (intel_mid_timer_options == INTEL_MID_TIMER_LAPIC_APBT || !apb_timer_block_enabled) return 0; /* This notifier should be called after workqueue is ready */ @@ -311,7 +311,6 @@ void __init apbt_time_init(void) #ifdef CONFIG_SMP int i; struct sfi_timer_table_entry *p_mtmr; - unsigned int percpu_timer; struct apbt_dev *adev; #endif @@ -341,18 +340,15 @@ void __init apbt_time_init(void) } #ifdef CONFIG_SMP /* kernel cmdline disable apb timer, so we will use lapic timers */ - if (mrst_timer_options == MRST_TIMER_LAPIC_APBT) { + if (intel_mid_timer_options == INTEL_MID_TIMER_LAPIC_APBT) { printk(KERN_INFO "apbt: disabled per cpu timer\n"); return; } pr_debug("%s: %d CPUs online\n", __func__, num_online_cpus()); - if (num_possible_cpus() <= sfi_mtimer_num) { - percpu_timer = 1; + if (num_possible_cpus() <= sfi_mtimer_num) apbt_num_timers_used = num_possible_cpus(); - } else { - percpu_timer = 0; + else apbt_num_timers_used = 1; - } pr_debug("%s: %d APB timers used\n", __func__, apbt_num_timers_used); /* here we set up per CPU timer data structure */ diff --git a/arch/x86/kernel/aperture_64.c b/arch/x86/kernel/aperture_64.c index 6e76c191a83..76164e173a2 100644 --- a/arch/x86/kernel/aperture_64.c +++ b/arch/x86/kernel/aperture_64.c @@ -10,6 +10,8 @@ * * Copyright 2002 Andi Kleen, SuSE Labs. */ +#define pr_fmt(fmt) "AGP: " fmt + #include <linux/kernel.h> #include <linux/types.h> #include <linux/init.h> @@ -18,9 +20,7 @@ #include <linux/pci_ids.h> #include <linux/pci.h> #include <linux/bitops.h> -#include <linux/ioport.h> #include <linux/suspend.h> -#include <linux/kmemleak.h> #include <asm/e820.h> #include <asm/io.h> #include <asm/iommu.h> @@ -55,18 +55,6 @@ int fallback_aper_force __initdata; int fix_aperture __initdata = 1; -static struct resource gart_resource = { - .name = "GART", - .flags = IORESOURCE_MEM, -}; - -static void __init insert_aperture_resource(u32 aper_base, u32 aper_size) -{ - gart_resource.start = aper_base; - gart_resource.end = aper_base + aper_size - 1; - insert_resource(&iomem_resource, &gart_resource); -} - /* This code runs before the PCI subsystem is initialized, so just access the northbridge directly. */ @@ -88,21 +76,14 @@ static u32 __init allocate_aperture(void) */ addr = memblock_find_in_range(GART_MIN_ADDR, GART_MAX_ADDR, aper_size, aper_size); - if (!addr || addr + aper_size > GART_MAX_ADDR) { - printk(KERN_ERR - "Cannot allocate aperture memory hole (%lx,%uK)\n", - addr, aper_size>>10); + if (!addr) { + pr_err("Cannot allocate aperture memory hole [mem %#010lx-%#010lx] (%uKB)\n", + addr, addr + aper_size - 1, aper_size >> 10); return 0; } memblock_reserve(addr, aper_size); - /* - * Kmemleak should not scan this block as it may not be mapped via the - * kernel direct mapping. - */ - kmemleak_ignore(phys_to_virt(addr)); - printk(KERN_INFO "Mapping aperture over %d KB of RAM @ %lx\n", - aper_size >> 10, addr); - insert_aperture_resource((u32)addr, aper_size); + pr_info("Mapping aperture over RAM [mem %#010lx-%#010lx] (%uKB)\n", + addr, addr + aper_size - 1, aper_size >> 10); register_nosave_region(addr >> PAGE_SHIFT, (addr+aper_size) >> PAGE_SHIFT); @@ -146,10 +127,11 @@ static u32 __init read_agp(int bus, int slot, int func, int cap, u32 *order) u64 aper; u32 old_order; - printk(KERN_INFO "AGP bridge at %02x:%02x:%02x\n", bus, slot, func); + pr_info("pci 0000:%02x:%02x:%02x: AGP bridge\n", bus, slot, func); apsizereg = read_pci_config_16(bus, slot, func, cap + 0x14); if (apsizereg == 0xffffffff) { - printk(KERN_ERR "APSIZE in AGP bridge unreadable\n"); + pr_err("pci 0000:%02x:%02x.%d: APSIZE unreadable\n", + bus, slot, func); return 0; } @@ -173,16 +155,18 @@ static u32 __init read_agp(int bus, int slot, int func, int cap, u32 *order) * On some sick chips, APSIZE is 0. It means it wants 4G * so let double check that order, and lets trust AMD NB settings: */ - printk(KERN_INFO "Aperture from AGP @ %Lx old size %u MB\n", - aper, 32 << old_order); + pr_info("pci 0000:%02x:%02x.%d: AGP aperture [bus addr %#010Lx-%#010Lx] (old size %uMB)\n", + bus, slot, func, aper, aper + (32ULL << (old_order + 20)) - 1, + 32 << old_order); if (aper + (32ULL<<(20 + *order)) > 0x100000000ULL) { - printk(KERN_INFO "Aperture size %u MB (APSIZE %x) is not right, using settings from NB\n", - 32 << *order, apsizereg); + pr_info("pci 0000:%02x:%02x.%d: AGP aperture size %uMB (APSIZE %#x) is not right, using settings from NB\n", + bus, slot, func, 32 << *order, apsizereg); *order = old_order; } - printk(KERN_INFO "Aperture from AGP @ %Lx size %u MB (APSIZE %x)\n", - aper, 32 << *order, apsizereg); + pr_info("pci 0000:%02x:%02x.%d: AGP aperture [bus addr %#010Lx-%#010Lx] (%uMB, APSIZE %#x)\n", + bus, slot, func, aper, aper + (32ULL << (*order + 20)) - 1, + 32 << *order, apsizereg); if (!aperture_valid(aper, (32*1024*1024) << *order, 32<<20)) return 0; @@ -238,7 +222,7 @@ static u32 __init search_agp_bridge(u32 *order, int *valid_agp) } } } - printk(KERN_INFO "No AGP bridge found\n"); + pr_info("No AGP bridge found\n"); return 0; } @@ -330,7 +314,8 @@ void __init early_gart_iommu_check(void) if (e820_any_mapped(aper_base, aper_base + aper_size, E820_RAM)) { /* reserve it, so we can reuse it in second kernel */ - printk(KERN_INFO "update e820 for GART\n"); + pr_info("e820: reserve [mem %#010Lx-%#010Lx] for GART\n", + aper_base, aper_base + aper_size - 1); e820_add_region(aper_base, aper_size, E820_RESERVED); update_e820(); } @@ -374,7 +359,7 @@ int __init gart_iommu_hole_init(void) !early_pci_allowed()) return -ENODEV; - printk(KERN_INFO "Checking aperture...\n"); + pr_info("Checking aperture...\n"); if (!fallback_aper_force) agp_aper_base = search_agp_bridge(&agp_aper_order, &valid_agp); @@ -415,8 +400,9 @@ int __init gart_iommu_hole_init(void) aper_base = read_pci_config(bus, slot, 3, AMD64_GARTAPERTUREBASE) & 0x7fff; aper_base <<= 25; - printk(KERN_INFO "Node %d: aperture @ %Lx size %u MB\n", - node, aper_base, aper_size >> 20); + pr_info("Node %d: aperture [bus addr %#010Lx-%#010Lx] (%uMB)\n", + node, aper_base, aper_base + aper_size - 1, + aper_size >> 20); node++; if (!aperture_valid(aper_base, aper_size, 64<<20)) { @@ -427,9 +413,9 @@ int __init gart_iommu_hole_init(void) if (!no_iommu && max_pfn > MAX_DMA32_PFN && !printed_gart_size_msg) { - printk(KERN_ERR "you are using iommu with agp, but GART size is less than 64M\n"); - printk(KERN_ERR "please increase GART size in your BIOS setup\n"); - printk(KERN_ERR "if BIOS doesn't have that option, contact your HW vendor!\n"); + pr_err("you are using iommu with agp, but GART size is less than 64MB\n"); + pr_err("please increase GART size in your BIOS setup\n"); + pr_err("if BIOS doesn't have that option, contact your HW vendor!\n"); printed_gart_size_msg = 1; } } else { @@ -450,12 +436,8 @@ int __init gart_iommu_hole_init(void) out: if (!fix && !fallback_aper_force) { - if (last_aper_base) { - unsigned long n = (32 * 1024 * 1024) << last_aper_order; - - insert_aperture_resource((u32)last_aper_base, n); + if (last_aper_base) return 1; - } return 0; } @@ -470,13 +452,10 @@ out: force_iommu || valid_agp || fallback_aper_force) { - printk(KERN_INFO - "Your BIOS doesn't leave a aperture memory hole\n"); - printk(KERN_INFO - "Please enable the IOMMU option in the BIOS setup\n"); - printk(KERN_INFO - "This costs you %d MB of RAM\n", - 32 << fallback_aper_order); + pr_info("Your BIOS doesn't leave a aperture memory hole\n"); + pr_info("Please enable the IOMMU option in the BIOS setup\n"); + pr_info("This costs you %dMB of RAM\n", + 32 << fallback_aper_order); aper_order = fallback_aper_order; aper_alloc = allocate_aperture(); diff --git a/arch/x86/kernel/apic/Makefile b/arch/x86/kernel/apic/Makefile index 0ae0323b1f9..dcb5b15401c 100644 --- a/arch/x86/kernel/apic/Makefile +++ b/arch/x86/kernel/apic/Makefile @@ -18,10 +18,7 @@ obj-y += apic_flat_64.o endif # APIC probe will depend on the listing order here -obj-$(CONFIG_X86_NUMAQ) += numaq_32.o -obj-$(CONFIG_X86_SUMMIT) += summit_32.o obj-$(CONFIG_X86_BIGSMP) += bigsmp_32.o -obj-$(CONFIG_X86_ES7000) += es7000_32.o # For 32bit, probe_32 need to be listed last obj-$(CONFIG_X86_LOCAL_APIC) += probe_$(BITS).o diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c index 2eec05b6d1b..ad28db7e6bd 100644 --- a/arch/x86/kernel/apic/apic.c +++ b/arch/x86/kernel/apic/apic.c @@ -35,6 +35,8 @@ #include <linux/smp.h> #include <linux/mm.h> +#include <asm/trace/irq_vectors.h> +#include <asm/irq_remapping.h> #include <asm/perf_event.h> #include <asm/x86_init.h> #include <asm/pgalloc.h> @@ -56,10 +58,11 @@ unsigned int num_processors; -unsigned disabled_cpus __cpuinitdata; +unsigned disabled_cpus; /* Processor that is doing the boot up */ unsigned int boot_cpu_physical_apicid = -1U; +EXPORT_SYMBOL_GPL(boot_cpu_physical_apicid); /* * The highest APIC ID seen during enumeration. @@ -72,10 +75,17 @@ unsigned int max_physical_apicid; physid_mask_t phys_cpu_present_map; /* + * Processor to be disabled specified by kernel parameter + * disable_cpu_apicid=<int>, mostly used for the kdump 2nd kernel to + * avoid undefined behaviour caused by sending INIT from AP to BSP. + */ +static unsigned int disabled_cpu_apicid __read_mostly = BAD_APICID; + +/* * Map cpu index to physical APIC ID */ -DEFINE_EARLY_PER_CPU(u16, x86_cpu_to_apicid, BAD_APICID); -DEFINE_EARLY_PER_CPU(u16, x86_bios_cpu_apicid, BAD_APICID); +DEFINE_EARLY_PER_CPU_READ_MOSTLY(u16, x86_cpu_to_apicid, BAD_APICID); +DEFINE_EARLY_PER_CPU_READ_MOSTLY(u16, x86_bios_cpu_apicid, BAD_APICID); EXPORT_EARLY_PER_CPU_SYMBOL(x86_cpu_to_apicid); EXPORT_EARLY_PER_CPU_SYMBOL(x86_bios_cpu_apicid); @@ -87,23 +97,8 @@ EXPORT_EARLY_PER_CPU_SYMBOL(x86_bios_cpu_apicid); * used for the mapping. This is where the behaviors of x86_64 and 32 * actually diverge. Let's keep it ugly for now. */ -DEFINE_EARLY_PER_CPU(int, x86_cpu_to_logical_apicid, BAD_APICID); +DEFINE_EARLY_PER_CPU_READ_MOSTLY(int, x86_cpu_to_logical_apicid, BAD_APICID); -/* - * Knob to control our willingness to enable the local APIC. - * - * +1=force-enable - */ -static int force_enable_local_apic __initdata; -/* - * APIC command line parameters - */ -static int __init parse_lapic(char *arg) -{ - force_enable_local_apic = 1; - return 0; -} -early_param("lapic", parse_lapic); /* Local APIC was disabled by the BIOS and enabled by the kernel */ static int enabled_via_apicbase; @@ -132,6 +127,29 @@ static inline void imcr_apic_to_pic(void) } #endif +/* + * Knob to control our willingness to enable the local APIC. + * + * +1=force-enable + */ +static int force_enable_local_apic __initdata; + +/* Control whether x2APIC mode is enabled or not */ +static bool nox2apic __initdata; + +/* + * APIC command line parameters + */ +static int __init parse_lapic(char *arg) +{ + if (config_enabled(CONFIG_X86_32) && !arg) + force_enable_local_apic = 1; + else if (arg && !strncmp(arg, "notscdeadline", 13)) + setup_clear_cpu_cap(X86_FEATURE_TSC_DEADLINE_TIMER); + return 0; +} +early_param("lapic", parse_lapic); + #ifdef CONFIG_X86_64 static int apic_calibrate_pmtmr __initdata; static __init int setup_apicpmtimer(char *s) @@ -148,8 +166,7 @@ int x2apic_mode; /* x2apic enabled before OS handover */ int x2apic_preenabled; static int x2apic_disabled; -static int nox2apic; -static __init int setup_nox2apic(char *str) +static int __init setup_nox2apic(char *str) { if (x2apic_enabled()) { int apicid = native_apic_msr_read(APIC_ID); @@ -164,7 +181,7 @@ static __init int setup_nox2apic(char *str) } else setup_clear_cpu_cap(X86_FEATURE_X2APIC); - nox2apic = 1; + nox2apic = true; return 0; } @@ -269,8 +286,12 @@ u32 native_safe_apic_wait_icr_idle(void) void native_apic_icr_write(u32 low, u32 id) { + unsigned long flags; + + local_irq_save(flags); apic_write(APIC_ICR2, SET_APIC_DEST_FIELD(id)); apic_write(APIC_ICR, low); + local_irq_restore(flags); } u64 native_apic_icr_read(void) @@ -314,6 +335,7 @@ int lapic_get_maxlvt(void) /* Clock divisor */ #define APIC_DIVISOR 16 +#define TSC_DIVISOR 32 /* * This function sets up the local APIC timer, with a timeout of @@ -332,6 +354,9 @@ static void __setup_APIC_LVTT(unsigned int clocks, int oneshot, int irqen) lvtt_value = LOCAL_TIMER_VECTOR; if (!oneshot) lvtt_value |= APIC_LVT_TIMER_PERIODIC; + else if (boot_cpu_has(X86_FEATURE_TSC_DEADLINE_TIMER)) + lvtt_value |= APIC_LVT_TIMER_TSCDEADLINE; + if (!lapic_is_integrated()) lvtt_value |= SET_APIC_TIMER_BASE(APIC_TIMER_BASE_DIV); @@ -340,6 +365,11 @@ static void __setup_APIC_LVTT(unsigned int clocks, int oneshot, int irqen) apic_write(APIC_LVTT, lvtt_value); + if (lvtt_value & APIC_LVT_TIMER_TSCDEADLINE) { + printk_once(KERN_DEBUG "TSC deadline timer enabled\n"); + return; + } + /* * Divide PICLK by 16 */ @@ -383,20 +413,25 @@ static inline int eilvt_entry_is_changeable(unsigned int old, unsigned int new) static unsigned int reserve_eilvt_offset(int offset, unsigned int new) { - unsigned int rsvd; /* 0: uninitialized */ + unsigned int rsvd, vector; if (offset >= APIC_EILVT_NR_MAX) return ~0; - rsvd = atomic_read(&eilvt_offsets[offset]) & ~APIC_EILVT_MASKED; + rsvd = atomic_read(&eilvt_offsets[offset]); do { - if (rsvd && - !eilvt_entry_is_changeable(rsvd, new)) + vector = rsvd & ~APIC_EILVT_MASKED; /* 0: unassigned */ + if (vector && !eilvt_entry_is_changeable(vector, new)) /* may not change if vectors are different */ return rsvd; rsvd = atomic_cmpxchg(&eilvt_offsets[offset], rsvd, new); } while (rsvd != new); + rsvd &= ~APIC_EILVT_MASKED; + if (rsvd && rsvd != vector) + pr_info("LVT offset %d assigned for vector 0x%02x\n", + offset, rsvd); + return new; } @@ -447,6 +482,16 @@ static int lapic_next_event(unsigned long delta, return 0; } +static int lapic_next_deadline(unsigned long delta, + struct clock_event_device *evt) +{ + u64 tsc; + + rdtscll(tsc); + wrmsrl(MSR_IA32_TSC_DEADLINE, tsc + (((u64) delta) * TSC_DIVISOR)); + return 0; +} + /* * Setup the lapic timer in periodic or oneshot mode */ @@ -514,7 +559,7 @@ static DEFINE_PER_CPU(struct clock_event_device, lapic_events); * Setup the local APIC timer for this CPU. Copy the initialized values * of the boot CPU and register the clock event in the framework. */ -static void __cpuinit setup_APIC_timer(void) +static void setup_APIC_timer(void) { struct clock_event_device *levt = &__get_cpu_var(lapic_events); @@ -527,7 +572,15 @@ static void __cpuinit setup_APIC_timer(void) memcpy(levt, &lapic_clockevent, sizeof(*levt)); levt->cpumask = cpumask_of(smp_processor_id()); - clockevents_register_device(levt); + if (this_cpu_has(X86_FEATURE_TSC_DEADLINE_TIMER)) { + levt->features &= ~(CLOCK_EVT_FEAT_PERIODIC | + CLOCK_EVT_FEAT_DUMMY); + levt->set_next_event = lapic_next_deadline; + clockevents_config_and_register(levt, + (tsc_khz / TSC_DIVISOR) * 1000, + 0xF, ~0UL); + } else + clockevents_register_device(levt); } /* @@ -655,7 +708,9 @@ static int __init calibrate_APIC_clock(void) * in the clockevent structure and return. */ - if (lapic_timer_frequency) { + if (boot_cpu_has(X86_FEATURE_TSC_DEADLINE_TIMER)) { + return 0; + } else if (lapic_timer_frequency) { apic_printk(APIC_VERBOSE, "lapic timer already calibrated %d\n", lapic_timer_frequency); lapic_clockevent.mult = div_sc(lapic_timer_frequency/APIC_DIVISOR, @@ -668,6 +723,9 @@ static int __init calibrate_APIC_clock(void) return 0; } + apic_printk(APIC_VERBOSE, "Using local APIC timer interrupts.\n" + "calibrating APIC timer ...\n"); + local_irq_disable(); /* Replace the global interrupt handler */ @@ -805,9 +863,6 @@ void __init setup_boot_APIC_clock(void) return; } - apic_printk(APIC_VERBOSE, "Using local APIC timer interrupts.\n" - "calibrating APIC timer ...\n"); - if (calibrate_APIC_clock()) { /* No broadcast on UP ! */ if (num_possible_cpus() > 1) @@ -826,7 +881,7 @@ void __init setup_boot_APIC_clock(void) setup_APIC_timer(); } -void __cpuinit setup_secondary_APIC_clock(void) +void setup_secondary_APIC_clock(void) { setup_APIC_timer(); } @@ -873,24 +928,42 @@ static void local_apic_timer_interrupt(void) * [ if a single-CPU system runs an SMP kernel then we call the local * interrupt as well. Thus we cannot inline the local irq ... ] */ -void __irq_entry smp_apic_timer_interrupt(struct pt_regs *regs) +__visible void __irq_entry smp_apic_timer_interrupt(struct pt_regs *regs) { struct pt_regs *old_regs = set_irq_regs(regs); /* * NOTE! We'd better ACK the irq immediately, * because timer handling can be slow. + * + * update_process_times() expects us to have done irq_enter(). + * Besides, if we don't timer interrupts ignore the global + * interrupt lock, which is the WrongThing (tm) to do. */ - ack_APIC_irq(); + entering_ack_irq(); + local_apic_timer_interrupt(); + exiting_irq(); + + set_irq_regs(old_regs); +} + +__visible void __irq_entry smp_trace_apic_timer_interrupt(struct pt_regs *regs) +{ + struct pt_regs *old_regs = set_irq_regs(regs); + /* + * NOTE! We'd better ACK the irq immediately, + * because timer handling can be slow. + * * update_process_times() expects us to have done irq_enter(). * Besides, if we don't timer interrupts ignore the global * interrupt lock, which is the WrongThing (tm) to do. */ - irq_enter(); - exit_idle(); + entering_ack_irq(); + trace_local_timer_entry(LOCAL_TIMER_VECTOR); local_apic_timer_interrupt(); - irq_exit(); + trace_local_timer_exit(LOCAL_TIMER_VECTOR); + exiting_irq(); set_irq_regs(old_regs); } @@ -1171,7 +1244,7 @@ void __init init_bsp_APIC(void) apic_write(APIC_LVT1, value); } -static void __cpuinit lapic_setup_esr(void) +static void lapic_setup_esr(void) { unsigned int oldvalue, value, maxlvt; @@ -1218,7 +1291,7 @@ static void __cpuinit lapic_setup_esr(void) * Used to setup local APIC while initializing BSP or bringin up APs. * Always called with preemption disabled. */ -void __cpuinit setup_local_APIC(void) +void setup_local_APIC(void) { int cpu = smp_processor_id(); unsigned int value, queued; @@ -1320,11 +1393,13 @@ void __cpuinit setup_local_APIC(void) acked); break; } - if (cpu_has_tsc) { - rdtscll(ntsc); - max_loops = (cpu_khz << 10) - (ntsc - tsc); - } else - max_loops--; + if (queued) { + if (cpu_has_tsc) { + rdtscll(ntsc); + max_loops = (cpu_khz << 10) - (ntsc - tsc); + } else + max_loops--; + } } while (queued && max_loops > 0); WARN_ON(max_loops <= 0); @@ -1411,7 +1486,7 @@ void __cpuinit setup_local_APIC(void) #endif } -void __cpuinit end_local_APIC_setup(void) +void end_local_APIC_setup(void) { lapic_setup_esr(); @@ -1436,8 +1511,7 @@ void __init bsp_end_local_APIC_setup(void) * Now that local APIC setup is completed for BP, configure the fault * handling for interrupt remapping. */ - if (intr_remapping_enabled) - enable_drhd_fault_handling(); + irq_remap_enable_fault_handling(); } @@ -1512,7 +1586,7 @@ void enable_x2apic(void) int __init enable_IR(void) { #ifdef CONFIG_IRQ_REMAP - if (!intr_remapping_supported()) { + if (!irq_remapping_supported()) { pr_debug("intr-remapping not supported\n"); return -1; } @@ -1523,7 +1597,7 @@ int __init enable_IR(void) return -1; } - return enable_intr_remapping(); + return irq_remapping_enable(); #endif return -1; } @@ -1532,10 +1606,13 @@ void __init enable_IR_x2apic(void) { unsigned long flags; int ret, x2apic_enabled = 0; - int dmar_table_init_ret; + int hardware_init_ret; + + /* Make sure irq_remap_ops are initialized */ + setup_irq_remapping_ops(); - dmar_table_init_ret = dmar_table_init(); - if (dmar_table_init_ret && !x2apic_supported()) + hardware_init_ret = irq_remapping_prepare(); + if (hardware_init_ret && !x2apic_supported()) return; ret = save_ioapic_entries(); @@ -1551,7 +1628,7 @@ void __init enable_IR_x2apic(void) if (x2apic_preenabled && nox2apic) disable_x2apic(); - if (dmar_table_init_ret) + if (hardware_init_ret) ret = -1; else ret = enable_IR(); @@ -1632,9 +1709,11 @@ static int __init apic_verify(void) mp_lapic_addr = APIC_DEFAULT_PHYS_BASE; /* The BIOS may have set up the APIC at some other address */ - rdmsr(MSR_IA32_APICBASE, l, h); - if (l & MSR_IA32_APICBASE_ENABLE) - mp_lapic_addr = l & MSR_IA32_APICBASE_BASE; + if (boot_cpu_data.x86 >= 6) { + rdmsr(MSR_IA32_APICBASE, l, h); + if (l & MSR_IA32_APICBASE_ENABLE) + mp_lapic_addr = l & MSR_IA32_APICBASE_BASE; + } pr_info("Found and enabled local APIC!\n"); return 0; @@ -1652,13 +1731,15 @@ int __init apic_force_enable(unsigned long addr) * MSR. This can only be done in software for Intel P6 or later * and AMD K7 (Model > 1) or later. */ - rdmsr(MSR_IA32_APICBASE, l, h); - if (!(l & MSR_IA32_APICBASE_ENABLE)) { - pr_info("Local APIC disabled by BIOS -- reenabling.\n"); - l &= ~MSR_IA32_APICBASE_BASE; - l |= MSR_IA32_APICBASE_ENABLE | addr; - wrmsr(MSR_IA32_APICBASE, l, h); - enabled_via_apicbase = 1; + if (boot_cpu_data.x86 >= 6) { + rdmsr(MSR_IA32_APICBASE, l, h); + if (!(l & MSR_IA32_APICBASE_ENABLE)) { + pr_info("Local APIC disabled by BIOS -- reenabling.\n"); + l &= ~MSR_IA32_APICBASE_BASE; + l |= MSR_IA32_APICBASE_ENABLE | addr; + wrmsr(MSR_IA32_APICBASE, l, h); + enabled_via_apicbase = 1; + } } return apic_verify(); } @@ -1860,12 +1941,10 @@ int __init APIC_init_uniprocessor(void) /* * This interrupt should _never_ happen with our APIC/SMP architecture */ -void smp_spurious_interrupt(struct pt_regs *regs) +static inline void __smp_spurious_interrupt(void) { u32 v; - irq_enter(); - exit_idle(); /* * Check if this really is a spurious interrupt and ACK it * if it is a vectored one. Just in case... @@ -1880,15 +1959,30 @@ void smp_spurious_interrupt(struct pt_regs *regs) /* see sw-dev-man vol 3, chapter 7.4.13.5 */ pr_info("spurious APIC interrupt on CPU#%d, " "should never happen.\n", smp_processor_id()); - irq_exit(); +} + +__visible void smp_spurious_interrupt(struct pt_regs *regs) +{ + entering_irq(); + __smp_spurious_interrupt(); + exiting_irq(); +} + +__visible void smp_trace_spurious_interrupt(struct pt_regs *regs) +{ + entering_irq(); + trace_spurious_apic_entry(SPURIOUS_APIC_VECTOR); + __smp_spurious_interrupt(); + trace_spurious_apic_exit(SPURIOUS_APIC_VECTOR); + exiting_irq(); } /* * This interrupt should never happen with our APIC/SMP architecture */ -void smp_error_interrupt(struct pt_regs *regs) +static inline void __smp_error_interrupt(struct pt_regs *regs) { - u32 v0, v1; + u32 v; u32 i = 0; static const char * const error_interrupt_reason[] = { "Send CS error", /* APIC Error Bit 0 */ @@ -1901,29 +1995,42 @@ void smp_error_interrupt(struct pt_regs *regs) "Illegal register address", /* APIC Error Bit 7 */ }; - irq_enter(); - exit_idle(); /* First tickle the hardware, only then report what went on. -- REW */ - v0 = apic_read(APIC_ESR); - apic_write(APIC_ESR, 0); - v1 = apic_read(APIC_ESR); + if (lapic_get_maxlvt() > 3) /* Due to the Pentium erratum 3AP. */ + apic_write(APIC_ESR, 0); + v = apic_read(APIC_ESR); ack_APIC_irq(); atomic_inc(&irq_err_count); - apic_printk(APIC_DEBUG, KERN_DEBUG "APIC error on CPU%d: %02x(%02x)", - smp_processor_id(), v0 , v1); + apic_printk(APIC_DEBUG, KERN_DEBUG "APIC error on CPU%d: %02x", + smp_processor_id(), v); - v1 = v1 & 0xff; - while (v1) { - if (v1 & 0x1) + v &= 0xff; + while (v) { + if (v & 0x1) apic_printk(APIC_DEBUG, KERN_CONT " : %s", error_interrupt_reason[i]); i++; - v1 >>= 1; - }; + v >>= 1; + } apic_printk(APIC_DEBUG, KERN_CONT "\n"); - irq_exit(); +} + +__visible void smp_error_interrupt(struct pt_regs *regs) +{ + entering_irq(); + __smp_error_interrupt(regs); + exiting_irq(); +} + +__visible void smp_trace_error_interrupt(struct pt_regs *regs) +{ + entering_irq(); + trace_error_apic_entry(ERROR_APIC_VECTOR); + __smp_error_interrupt(regs); + trace_error_apic_exit(ERROR_APIC_VECTOR); + exiting_irq(); } /** @@ -2015,13 +2122,45 @@ void disconnect_bsp_APIC(int virt_wire_setup) apic_write(APIC_LVT1, value); } -void __cpuinit generic_processor_info(int apicid, int version) +int generic_processor_info(int apicid, int version) { int cpu, max = nr_cpu_ids; bool boot_cpu_detected = physid_isset(boot_cpu_physical_apicid, phys_cpu_present_map); /* + * boot_cpu_physical_apicid is designed to have the apicid + * returned by read_apic_id(), i.e, the apicid of the + * currently booting-up processor. However, on some platforms, + * it is temporarily modified by the apicid reported as BSP + * through MP table. Concretely: + * + * - arch/x86/kernel/mpparse.c: MP_processor_info() + * - arch/x86/mm/amdtopology.c: amd_numa_init() + * + * This function is executed with the modified + * boot_cpu_physical_apicid. So, disabled_cpu_apicid kernel + * parameter doesn't work to disable APs on kdump 2nd kernel. + * + * Since fixing handling of boot_cpu_physical_apicid requires + * another discussion and tests on each platform, we leave it + * for now and here we use read_apic_id() directly in this + * function, generic_processor_info(). + */ + if (disabled_cpu_apicid != BAD_APICID && + disabled_cpu_apicid != read_apic_id() && + disabled_cpu_apicid == apicid) { + int thiscpu = num_processors + disabled_cpus; + + pr_warning("APIC: Disabling requested cpu." + " Processor %d/0x%x ignored.\n", + thiscpu, apicid); + + disabled_cpus++; + return -ENODEV; + } + + /* * If boot cpu has not been detected yet, then only allow upto * nr_cpu_ids - 1 processors and keep one slot free for boot cpu */ @@ -2035,7 +2174,7 @@ void __cpuinit generic_processor_info(int apicid, int version) " Processor %d/0x%x ignored.\n", max, thiscpu, apicid); disabled_cpus++; - return; + return -ENODEV; } if (num_processors >= nr_cpu_ids) { @@ -2046,7 +2185,7 @@ void __cpuinit generic_processor_info(int apicid, int version) " Processor %d/0x%x ignored.\n", max, thiscpu, apicid); disabled_cpus++; - return; + return -EINVAL; } num_processors++; @@ -2091,6 +2230,8 @@ void __cpuinit generic_processor_info(int apicid, int version) #endif set_cpu_possible(cpu, true); set_cpu_present(cpu, true); + + return cpu; } int hard_smp_processor_id(void) @@ -2108,6 +2249,42 @@ void default_init_apic_ldr(void) apic_write(APIC_LDR, val); } +int default_cpu_mask_to_apicid_and(const struct cpumask *cpumask, + const struct cpumask *andmask, + unsigned int *apicid) +{ + unsigned int cpu; + + for_each_cpu_and(cpu, cpumask, andmask) { + if (cpumask_test_cpu(cpu, cpu_online_mask)) + break; + } + + if (likely(cpu < nr_cpu_ids)) { + *apicid = per_cpu(x86_cpu_to_apicid, cpu); + return 0; + } + + return -EINVAL; +} + +/* + * Override the generic EOI implementation with an optimized version. + * Only called during early boot when only one CPU is active and with + * interrupts disabled, so we know this does not race with actual APIC driver + * use. + */ +void __init apic_set_eoi_write(void (*eoi_write)(u32 reg, u32 v)) +{ + struct apic **drv; + + for (drv = __apicdrivers; drv < __apicdrivers_end; drv++) { + /* Should happen once for each apic */ + WARN_ON((*drv)->eoi_write == eoi_write); + (*drv)->eoi_write = eoi_write; + } +} + /* * Power management */ @@ -2167,8 +2344,7 @@ static int lapic_suspend(void) local_irq_save(flags); disable_local_APIC(); - if (intr_remapping_enabled) - disable_intr_remapping(); + irq_remapping_disable(); local_irq_restore(flags); return 0; @@ -2184,16 +2360,15 @@ static void lapic_resume(void) return; local_irq_save(flags); - if (intr_remapping_enabled) { - /* - * IO-APIC and PIC have their own resume routines. - * We just mask them here to make sure the interrupt - * subsystem is completely quiet while we enable x2apic - * and interrupt-remapping. - */ - mask_ioapic_entries(); - legacy_pic->mask_all(); - } + + /* + * IO-APIC and PIC have their own resume routines. + * We just mask them here to make sure the interrupt + * subsystem is completely quiet while we enable x2apic + * and interrupt-remapping. + */ + mask_ioapic_entries(); + legacy_pic->mask_all(); if (x2apic_mode) enable_x2apic(); @@ -2204,10 +2379,12 @@ static void lapic_resume(void) * FIXME! This will be wrong if we ever support suspend on * SMP! We'll need to do this as part of the CPU restore! */ - rdmsr(MSR_IA32_APICBASE, l, h); - l &= ~MSR_IA32_APICBASE_BASE; - l |= MSR_IA32_APICBASE_ENABLE | mp_lapic_addr; - wrmsr(MSR_IA32_APICBASE, l, h); + if (boot_cpu_data.x86 >= 6) { + rdmsr(MSR_IA32_APICBASE, l, h); + l &= ~MSR_IA32_APICBASE_BASE; + l |= MSR_IA32_APICBASE_ENABLE | mp_lapic_addr; + wrmsr(MSR_IA32_APICBASE, l, h); + } } maxlvt = lapic_get_maxlvt(); @@ -2219,7 +2396,7 @@ static void lapic_resume(void) apic_write(APIC_SPIV, apic_pm_state.apic_spiv); apic_write(APIC_LVT0, apic_pm_state.apic_lvt0); apic_write(APIC_LVT1, apic_pm_state.apic_lvt1); -#if defined(CONFIG_X86_MCE_P4THERMAL) || defined(CONFIG_X86_MCE_INTEL) +#if defined(CONFIG_X86_MCE_INTEL) if (maxlvt >= 5) apic_write(APIC_LVTTHMR, apic_pm_state.apic_thmr); #endif @@ -2234,8 +2411,7 @@ static void lapic_resume(void) apic_write(APIC_ESR, 0); apic_read(APIC_ESR); - if (intr_remapping_enabled) - reenable_intr_remapping(x2apic_mode); + irq_remapping_reenable(x2apic_mode); local_irq_restore(flags); } @@ -2250,7 +2426,7 @@ static struct syscore_ops lapic_syscore_ops = { .suspend = lapic_suspend, }; -static void __cpuinit apic_pm_activate(void) +static void apic_pm_activate(void) { apic_pm_state.active = 1; } @@ -2275,7 +2451,7 @@ static void apic_pm_activate(void) { } #ifdef CONFIG_X86_64 -static int __cpuinit apic_cluster_num(void) +static int apic_cluster_num(void) { int i, clusters, zeros; unsigned id; @@ -2320,10 +2496,10 @@ static int __cpuinit apic_cluster_num(void) return clusters; } -static int __cpuinitdata multi_checked; -static int __cpuinitdata multi; +static int multi_checked; +static int multi; -static int __cpuinit set_multi(const struct dmi_system_id *d) +static int set_multi(const struct dmi_system_id *d) { if (multi) return 0; @@ -2332,7 +2508,7 @@ static int __cpuinit set_multi(const struct dmi_system_id *d) return 0; } -static const __cpuinitconst struct dmi_system_id multi_dmi_table[] = { +static const struct dmi_system_id multi_dmi_table[] = { { .callback = set_multi, .ident = "IBM System Summit2", @@ -2344,7 +2520,7 @@ static const __cpuinitconst struct dmi_system_id multi_dmi_table[] = { {} }; -static void __cpuinit dmi_check_multi(void) +static void dmi_check_multi(void) { if (multi_checked) return; @@ -2361,7 +2537,7 @@ static void __cpuinit dmi_check_multi(void) * multi-chassis. * Use DMI to check them */ -__cpuinit int apic_is_clustered_box(void) +int apic_is_clustered_box(void) { dmi_check_multi(); if (multi) @@ -2462,3 +2638,12 @@ static int __init lapic_insert_resource(void) * that is using request_resource */ late_initcall(lapic_insert_resource); + +static int __init apic_set_disabled_cpu_apicid(char *arg) +{ + if (!arg || !get_option(&arg, &disabled_cpu_apicid)) + return -EINVAL; + + return 0; +} +early_param("disable_cpu_apicid", apic_set_disabled_cpu_apicid); diff --git a/arch/x86/kernel/apic/apic_flat_64.c b/arch/x86/kernel/apic/apic_flat_64.c index 8c3cdded6f2..7c1b2947951 100644 --- a/arch/x86/kernel/apic/apic_flat_64.c +++ b/arch/x86/kernel/apic/apic_flat_64.c @@ -14,16 +14,13 @@ #include <linux/string.h> #include <linux/kernel.h> #include <linux/ctype.h> -#include <linux/init.h> #include <linux/hardirq.h> #include <linux/module.h> #include <asm/smp.h> #include <asm/apic.h> #include <asm/ipi.h> -#ifdef CONFIG_ACPI -#include <acpi/acpi_bus.h> -#endif +#include <linux/acpi.h> static struct apic apic_physflat; static struct apic apic_flat; @@ -36,25 +33,6 @@ static int flat_acpi_madt_oem_check(char *oem_id, char *oem_table_id) return 1; } -static const struct cpumask *flat_target_cpus(void) -{ - return cpu_online_mask; -} - -static void flat_vector_allocation_domain(int cpu, struct cpumask *retmask) -{ - /* Careful. Some cpus do not strictly honor the set of cpus - * specified in the interrupt destination when using lowest - * priority interrupt delivery mode. - * - * In particular there was a hyperthreading cpu observed to - * deliver interrupts to the wrong hyperthread when only one - * hyperthread was specified in the interrupt desitination. - */ - cpumask_clear(retmask); - cpumask_bits(retmask)[0] = APIC_ALL_CPUS; -} - /* * Set up the logical destination ID. * @@ -92,7 +70,7 @@ static void flat_send_IPI_mask(const struct cpumask *cpumask, int vector) } static void - flat_send_IPI_mask_allbutself(const struct cpumask *cpumask, int vector) +flat_send_IPI_mask_allbutself(const struct cpumask *cpumask, int vector) { unsigned long mask = cpumask_bits(cpumask)[0]; int cpu = smp_processor_id(); @@ -180,12 +158,13 @@ static struct apic apic_flat = { .name = "flat", .probe = flat_probe, .acpi_madt_oem_check = flat_acpi_madt_oem_check, + .apic_id_valid = default_apic_id_valid, .apic_id_registered = flat_apic_id_registered, .irq_delivery_mode = dest_LowestPrio, .irq_dest_mode = 1, /* logical */ - .target_cpus = flat_target_cpus, + .target_cpus = online_target_cpus, .disable_esr = 0, .dest_logical = APIC_DEST_LOGICAL, .check_apicid_used = NULL, @@ -209,8 +188,7 @@ static struct apic apic_flat = { .set_apic_id = set_apic_id, .apic_id_mask = 0xFFu << 24, - .cpu_mask_to_apicid = default_cpu_mask_to_apicid, - .cpu_mask_to_apicid_and = default_cpu_mask_to_apicid_and, + .cpu_mask_to_apicid_and = flat_cpu_mask_to_apicid_and, .send_IPI_mask = flat_send_IPI_mask, .send_IPI_mask_allbutself = flat_send_IPI_mask_allbutself, @@ -220,12 +198,13 @@ static struct apic apic_flat = { .trampoline_phys_low = DEFAULT_TRAMPOLINE_PHYS_LOW, .trampoline_phys_high = DEFAULT_TRAMPOLINE_PHYS_HIGH, - .wait_for_init_deassert = NULL, + .wait_for_init_deassert = false, .smp_callin_clear_local_apic = NULL, .inquire_remote_apic = default_inquire_remote_apic, .read = native_apic_mem_read, .write = native_apic_mem_write, + .eoi_write = native_apic_mem_write, .icr_read = native_apic_icr_read, .icr_write = native_apic_icr_write, .wait_icr_idle = native_apic_wait_icr_idle, @@ -260,17 +239,6 @@ static int physflat_acpi_madt_oem_check(char *oem_id, char *oem_table_id) return 0; } -static const struct cpumask *physflat_target_cpus(void) -{ - return cpu_online_mask; -} - -static void physflat_vector_allocation_domain(int cpu, struct cpumask *retmask) -{ - cpumask_clear(retmask); - cpumask_set_cpu(cpu, retmask); -} - static void physflat_send_IPI_mask(const struct cpumask *cpumask, int vector) { default_send_IPI_mask_sequence_phys(cpumask, vector); @@ -292,38 +260,6 @@ static void physflat_send_IPI_all(int vector) physflat_send_IPI_mask(cpu_online_mask, vector); } -static unsigned int physflat_cpu_mask_to_apicid(const struct cpumask *cpumask) -{ - int cpu; - - /* - * We're using fixed IRQ delivery, can only return one phys APIC ID. - * May as well be the first. - */ - cpu = cpumask_first(cpumask); - if ((unsigned)cpu < nr_cpu_ids) - return per_cpu(x86_cpu_to_apicid, cpu); - else - return BAD_APICID; -} - -static unsigned int -physflat_cpu_mask_to_apicid_and(const struct cpumask *cpumask, - const struct cpumask *andmask) -{ - int cpu; - - /* - * We're using fixed IRQ delivery, can only return one phys APIC ID. - * May as well be the first. - */ - for_each_cpu_and(cpu, cpumask, andmask) { - if (cpumask_test_cpu(cpu, cpu_online_mask)) - break; - } - return per_cpu(x86_cpu_to_apicid, cpu); -} - static int physflat_probe(void) { if (apic == &apic_physflat || num_possible_cpus() > 8) @@ -337,18 +273,19 @@ static struct apic apic_physflat = { .name = "physical flat", .probe = physflat_probe, .acpi_madt_oem_check = physflat_acpi_madt_oem_check, + .apic_id_valid = default_apic_id_valid, .apic_id_registered = flat_apic_id_registered, .irq_delivery_mode = dest_Fixed, .irq_dest_mode = 0, /* physical */ - .target_cpus = physflat_target_cpus, + .target_cpus = online_target_cpus, .disable_esr = 0, .dest_logical = 0, .check_apicid_used = NULL, .check_apicid_present = NULL, - .vector_allocation_domain = physflat_vector_allocation_domain, + .vector_allocation_domain = default_vector_allocation_domain, /* not needed, but shouldn't hurt: */ .init_apic_ldr = flat_init_apic_ldr, @@ -367,8 +304,7 @@ static struct apic apic_physflat = { .set_apic_id = set_apic_id, .apic_id_mask = 0xFFu << 24, - .cpu_mask_to_apicid = physflat_cpu_mask_to_apicid, - .cpu_mask_to_apicid_and = physflat_cpu_mask_to_apicid_and, + .cpu_mask_to_apicid_and = default_cpu_mask_to_apicid_and, .send_IPI_mask = physflat_send_IPI_mask, .send_IPI_mask_allbutself = physflat_send_IPI_mask_allbutself, @@ -378,12 +314,13 @@ static struct apic apic_physflat = { .trampoline_phys_low = DEFAULT_TRAMPOLINE_PHYS_LOW, .trampoline_phys_high = DEFAULT_TRAMPOLINE_PHYS_HIGH, - .wait_for_init_deassert = NULL, + .wait_for_init_deassert = false, .smp_callin_clear_local_apic = NULL, .inquire_remote_apic = default_inquire_remote_apic, .read = native_apic_mem_read, .write = native_apic_mem_write, + .eoi_write = native_apic_mem_write, .icr_read = native_apic_icr_read, .icr_write = native_apic_icr_write, .wait_icr_idle = native_apic_wait_icr_idle, diff --git a/arch/x86/kernel/apic/apic_noop.c b/arch/x86/kernel/apic/apic_noop.c index 775b82bc655..8c7c98249c2 100644 --- a/arch/x86/kernel/apic/apic_noop.c +++ b/arch/x86/kernel/apic/apic_noop.c @@ -15,7 +15,6 @@ #include <linux/string.h> #include <linux/kernel.h> #include <linux/ctype.h> -#include <linux/init.h> #include <linux/errno.h> #include <asm/fixmap.h> #include <asm/mpspec.h> @@ -100,12 +99,12 @@ static unsigned long noop_check_apicid_present(int bit) return physid_isset(bit, phys_cpu_present_map); } -static void noop_vector_allocation_domain(int cpu, struct cpumask *retmask) +static void noop_vector_allocation_domain(int cpu, struct cpumask *retmask, + const struct cpumask *mask) { if (cpu != 0) pr_warning("APIC: Vector allocated for non-BSP cpu\n"); - cpumask_clear(retmask); - cpumask_set_cpu(cpu, retmask); + cpumask_copy(retmask, cpumask_of(cpu)); } static u32 noop_apic_read(u32 reg) @@ -124,6 +123,7 @@ struct apic apic_noop = { .probe = noop_probe, .acpi_madt_oem_check = NULL, + .apic_id_valid = default_apic_id_valid, .apic_id_registered = noop_apic_id_registered, .irq_delivery_mode = dest_LowestPrio, @@ -158,8 +158,7 @@ struct apic apic_noop = { .set_apic_id = NULL, .apic_id_mask = 0x0F << 24, - .cpu_mask_to_apicid = default_cpu_mask_to_apicid, - .cpu_mask_to_apicid_and = default_cpu_mask_to_apicid_and, + .cpu_mask_to_apicid_and = flat_cpu_mask_to_apicid_and, .send_IPI_mask = noop_send_IPI_mask, .send_IPI_mask_allbutself = noop_send_IPI_mask_allbutself, @@ -173,13 +172,13 @@ struct apic apic_noop = { .trampoline_phys_low = DEFAULT_TRAMPOLINE_PHYS_LOW, .trampoline_phys_high = DEFAULT_TRAMPOLINE_PHYS_HIGH, - .wait_for_init_deassert = NULL, - + .wait_for_init_deassert = false, .smp_callin_clear_local_apic = NULL, .inquire_remote_apic = NULL, .read = noop_apic_read, .write = noop_apic_write, + .eoi_write = noop_apic_write, .icr_read = noop_apic_icr_read, .icr_write = noop_apic_icr_write, .wait_icr_idle = noop_apic_wait_icr_idle, diff --git a/arch/x86/kernel/apic/apic_numachip.c b/arch/x86/kernel/apic/apic_numachip.c index 09d3d8c1cd9..a5b45df8bc8 100644 --- a/arch/x86/kernel/apic/apic_numachip.c +++ b/arch/x86/kernel/apic/apic_numachip.c @@ -22,15 +22,17 @@ #include <linux/hardirq.h> #include <linux/delay.h> +#include <asm/numachip/numachip.h> #include <asm/numachip/numachip_csr.h> #include <asm/smp.h> #include <asm/apic.h> #include <asm/ipi.h> #include <asm/apic_flat_64.h> +#include <asm/pgtable.h> static int numachip_system __read_mostly; -static struct apic apic_numachip __read_mostly; +static const struct apic apic_numachip __read_mostly; static unsigned int get_apic_id(unsigned long x) { @@ -56,6 +58,12 @@ static unsigned int read_xapic_id(void) return get_apic_id(apic_read(APIC_ID)); } +static int numachip_apic_id_valid(int apicid) +{ + /* Trust what bootloader passes in MADT */ + return 1; +} + static int numachip_apic_id_registered(void) { return physid_isset(read_xapic_id(), phys_cpu_present_map); @@ -66,18 +74,7 @@ static int numachip_phys_pkg_id(int initial_apic_id, int index_msb) return initial_apic_id >> index_msb; } -static const struct cpumask *numachip_target_cpus(void) -{ - return cpu_online_mask; -} - -static void numachip_vector_allocation_domain(int cpu, struct cpumask *retmask) -{ - cpumask_clear(retmask); - cpumask_set_cpu(cpu, retmask); -} - -static int __cpuinit numachip_wakeup_secondary(int phys_apicid, unsigned long start_rip) +static int numachip_wakeup_secondary(int phys_apicid, unsigned long start_rip) { union numachip_csr_g3_ext_irq_gen int_gen; @@ -151,38 +148,6 @@ static void numachip_send_IPI_self(int vector) __default_send_IPI_shortcut(APIC_DEST_SELF, vector, APIC_DEST_PHYSICAL); } -static unsigned int numachip_cpu_mask_to_apicid(const struct cpumask *cpumask) -{ - int cpu; - - /* - * We're using fixed IRQ delivery, can only return one phys APIC ID. - * May as well be the first. - */ - cpu = cpumask_first(cpumask); - if (likely((unsigned)cpu < nr_cpu_ids)) - return per_cpu(x86_cpu_to_apicid, cpu); - - return BAD_APICID; -} - -static unsigned int -numachip_cpu_mask_to_apicid_and(const struct cpumask *cpumask, - const struct cpumask *andmask) -{ - int cpu; - - /* - * We're using fixed IRQ delivery, can only return one phys APIC ID. - * May as well be the first. - */ - for_each_cpu_and(cpu, cpumask, andmask) { - if (cpumask_test_cpu(cpu, cpu_online_mask)) - break; - } - return per_cpu(x86_cpu_to_apicid, cpu); -} - static int __init numachip_probe(void) { return apic == &apic_numachip; @@ -201,8 +166,11 @@ static void __init map_csrs(void) static void fixup_cpu_id(struct cpuinfo_x86 *c, int node) { - c->phys_proc_id = node; - per_cpu(cpu_llc_id, smp_processor_id()) = node; + + if (c->phys_proc_id != node) { + c->phys_proc_id = node; + per_cpu(cpu_llc_id, smp_processor_id()) = node; + } } static int __init numachip_system_init(void) @@ -213,6 +181,7 @@ static int __init numachip_system_init(void) return 0; x86_cpuinit.fixup_cpu_id = fixup_cpu_id; + x86_init.pci.arch_init = pci_numachip_init; map_csrs(); @@ -233,23 +202,24 @@ static int numachip_acpi_madt_oem_check(char *oem_id, char *oem_table_id) return 0; } -static struct apic apic_numachip __refconst = { +static const struct apic apic_numachip __refconst = { .name = "NumaConnect system", .probe = numachip_probe, .acpi_madt_oem_check = numachip_acpi_madt_oem_check, + .apic_id_valid = numachip_apic_id_valid, .apic_id_registered = numachip_apic_id_registered, .irq_delivery_mode = dest_Fixed, .irq_dest_mode = 0, /* physical */ - .target_cpus = numachip_target_cpus, + .target_cpus = online_target_cpus, .disable_esr = 0, .dest_logical = 0, .check_apicid_used = NULL, .check_apicid_present = NULL, - .vector_allocation_domain = numachip_vector_allocation_domain, + .vector_allocation_domain = default_vector_allocation_domain, .init_apic_ldr = flat_init_apic_ldr, .ioapic_phys_id_map = NULL, @@ -267,8 +237,7 @@ static struct apic apic_numachip __refconst = { .set_apic_id = set_apic_id, .apic_id_mask = 0xffU << 24, - .cpu_mask_to_apicid = numachip_cpu_mask_to_apicid, - .cpu_mask_to_apicid_and = numachip_cpu_mask_to_apicid_and, + .cpu_mask_to_apicid_and = default_cpu_mask_to_apicid_and, .send_IPI_mask = numachip_send_IPI_mask, .send_IPI_mask_allbutself = numachip_send_IPI_mask_allbutself, @@ -279,12 +248,13 @@ static struct apic apic_numachip __refconst = { .wakeup_secondary_cpu = numachip_wakeup_secondary, .trampoline_phys_low = DEFAULT_TRAMPOLINE_PHYS_LOW, .trampoline_phys_high = DEFAULT_TRAMPOLINE_PHYS_HIGH, - .wait_for_init_deassert = NULL, + .wait_for_init_deassert = false, .smp_callin_clear_local_apic = NULL, .inquire_remote_apic = NULL, /* REMRD not supported */ .read = native_apic_mem_read, .write = native_apic_mem_write, + .eoi_write = native_apic_mem_write, .icr_read = native_apic_icr_read, .icr_write = native_apic_icr_write, .wait_icr_idle = native_apic_wait_icr_idle, diff --git a/arch/x86/kernel/apic/bigsmp_32.c b/arch/x86/kernel/apic/bigsmp_32.c index 521bead0113..e4840aa7a25 100644 --- a/arch/x86/kernel/apic/bigsmp_32.c +++ b/arch/x86/kernel/apic/bigsmp_32.c @@ -26,15 +26,6 @@ static int bigsmp_apic_id_registered(void) return 1; } -static const struct cpumask *bigsmp_target_cpus(void) -{ -#ifdef CONFIG_SMP - return cpu_online_mask; -#else - return cpumask_of(0); -#endif -} - static unsigned long bigsmp_check_apicid_used(physid_mask_t *map, int apicid) { return 0; @@ -105,32 +96,6 @@ static int bigsmp_check_phys_apicid_present(int phys_apicid) return 1; } -/* As we are using single CPU as destination, pick only one CPU here */ -static unsigned int bigsmp_cpu_mask_to_apicid(const struct cpumask *cpumask) -{ - int cpu = cpumask_first(cpumask); - - if (cpu < nr_cpu_ids) - return cpu_physical_id(cpu); - return BAD_APICID; -} - -static unsigned int bigsmp_cpu_mask_to_apicid_and(const struct cpumask *cpumask, - const struct cpumask *andmask) -{ - int cpu; - - /* - * We're using fixed IRQ delivery, can only return one phys APIC ID. - * May as well be the first. - */ - for_each_cpu_and(cpu, cpumask, andmask) { - if (cpumask_test_cpu(cpu, cpu_online_mask)) - return cpu_physical_id(cpu); - } - return BAD_APICID; -} - static int bigsmp_phys_pkg_id(int cpuid_apic, int index_msb) { return cpuid_apic >> index_msb; @@ -177,12 +142,6 @@ static const struct dmi_system_id bigsmp_dmi_table[] = { { } /* NULL entry stops DMI scanning */ }; -static void bigsmp_vector_allocation_domain(int cpu, struct cpumask *retmask) -{ - cpumask_clear(retmask); - cpumask_set_cpu(cpu, retmask); -} - static int probe_bigsmp(void) { if (def_to_bigsmp) @@ -198,19 +157,20 @@ static struct apic apic_bigsmp = { .name = "bigsmp", .probe = probe_bigsmp, .acpi_madt_oem_check = NULL, + .apic_id_valid = default_apic_id_valid, .apic_id_registered = bigsmp_apic_id_registered, .irq_delivery_mode = dest_Fixed, /* phys delivery to target CPU: */ .irq_dest_mode = 0, - .target_cpus = bigsmp_target_cpus, + .target_cpus = default_target_cpus, .disable_esr = 1, .dest_logical = 0, .check_apicid_used = bigsmp_check_apicid_used, .check_apicid_present = bigsmp_check_apicid_present, - .vector_allocation_domain = bigsmp_vector_allocation_domain, + .vector_allocation_domain = default_vector_allocation_domain, .init_apic_ldr = bigsmp_init_apic_ldr, .ioapic_phys_id_map = bigsmp_ioapic_phys_id_map, @@ -228,8 +188,7 @@ static struct apic apic_bigsmp = { .set_apic_id = NULL, .apic_id_mask = 0xFF << 24, - .cpu_mask_to_apicid = bigsmp_cpu_mask_to_apicid, - .cpu_mask_to_apicid_and = bigsmp_cpu_mask_to_apicid_and, + .cpu_mask_to_apicid_and = default_cpu_mask_to_apicid_and, .send_IPI_mask = bigsmp_send_IPI_mask, .send_IPI_mask_allbutself = NULL, @@ -240,13 +199,13 @@ static struct apic apic_bigsmp = { .trampoline_phys_low = DEFAULT_TRAMPOLINE_PHYS_LOW, .trampoline_phys_high = DEFAULT_TRAMPOLINE_PHYS_HIGH, - .wait_for_init_deassert = default_wait_for_init_deassert, - + .wait_for_init_deassert = true, .smp_callin_clear_local_apic = NULL, .inquire_remote_apic = default_inquire_remote_apic, .read = native_apic_mem_read, .write = native_apic_mem_write, + .eoi_write = native_apic_mem_write, .icr_read = native_apic_icr_read, .icr_write = native_apic_icr_write, .wait_icr_idle = native_apic_wait_icr_idle, diff --git a/arch/x86/kernel/apic/es7000_32.c b/arch/x86/kernel/apic/es7000_32.c deleted file mode 100644 index 5d513bc47b6..00000000000 --- a/arch/x86/kernel/apic/es7000_32.c +++ /dev/null @@ -1,755 +0,0 @@ -/* - * Written by: Garry Forsgren, Unisys Corporation - * Natalie Protasevich, Unisys Corporation - * - * This file contains the code to configure and interface - * with Unisys ES7000 series hardware system manager. - * - * Copyright (c) 2003 Unisys Corporation. - * Copyright (C) 2009, Red Hat, Inc., Ingo Molnar - * - * All Rights Reserved. - * - * This program is free software; you can redistribute it and/or modify it - * under the terms of version 2 of the GNU General Public License as - * published by the Free Software Foundation. - * - * This program is distributed in the hope that it would be useful, but - * WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. - * - * You should have received a copy of the GNU General Public License along - * with this program; if not, write the Free Software Foundation, Inc., 59 - * Temple Place - Suite 330, Boston MA 02111-1307, USA. - * - * Contact information: Unisys Corporation, Township Line & Union Meeting - * Roads-A, Unisys Way, Blue Bell, Pennsylvania, 19424, or: - * - * http://www.unisys.com - */ - -#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt - -#include <linux/notifier.h> -#include <linux/spinlock.h> -#include <linux/cpumask.h> -#include <linux/threads.h> -#include <linux/kernel.h> -#include <linux/module.h> -#include <linux/reboot.h> -#include <linux/string.h> -#include <linux/types.h> -#include <linux/errno.h> -#include <linux/acpi.h> -#include <linux/init.h> -#include <linux/gfp.h> -#include <linux/nmi.h> -#include <linux/smp.h> -#include <linux/io.h> - -#include <asm/apicdef.h> -#include <linux/atomic.h> -#include <asm/fixmap.h> -#include <asm/mpspec.h> -#include <asm/setup.h> -#include <asm/apic.h> -#include <asm/ipi.h> - -/* - * ES7000 chipsets - */ - -#define NON_UNISYS 0 -#define ES7000_CLASSIC 1 -#define ES7000_ZORRO 2 - -#define MIP_REG 1 -#define MIP_PSAI_REG 4 - -#define MIP_BUSY 1 -#define MIP_SPIN 0xf0000 -#define MIP_VALID 0x0100000000000000ULL -#define MIP_SW_APIC 0x1020b - -#define MIP_PORT(val) ((val >> 32) & 0xffff) - -#define MIP_RD_LO(val) (val & 0xffffffff) - -struct mip_reg { - unsigned long long off_0x00; - unsigned long long off_0x08; - unsigned long long off_0x10; - unsigned long long off_0x18; - unsigned long long off_0x20; - unsigned long long off_0x28; - unsigned long long off_0x30; - unsigned long long off_0x38; -}; - -struct mip_reg_info { - unsigned long long mip_info; - unsigned long long delivery_info; - unsigned long long host_reg; - unsigned long long mip_reg; -}; - -struct psai { - unsigned long long entry_type; - unsigned long long addr; - unsigned long long bep_addr; -}; - -#ifdef CONFIG_ACPI - -struct es7000_oem_table { - struct acpi_table_header Header; - u32 OEMTableAddr; - u32 OEMTableSize; -}; - -static unsigned long oem_addrX; -static unsigned long oem_size; - -#endif - -/* - * ES7000 Globals - */ - -static volatile unsigned long *psai; -static struct mip_reg *mip_reg; -static struct mip_reg *host_reg; -static int mip_port; -static unsigned long mip_addr; -static unsigned long host_addr; - -int es7000_plat; - -/* - * GSI override for ES7000 platforms. - */ - - -static int __cpuinit wakeup_secondary_cpu_via_mip(int cpu, unsigned long eip) -{ - unsigned long vect = 0, psaival = 0; - - if (psai == NULL) - return -1; - - vect = ((unsigned long)__pa(eip)/0x1000) << 16; - psaival = (0x1000000 | vect | cpu); - - while (*psai & 0x1000000) - ; - - *psai = psaival; - - return 0; -} - -static int es7000_apic_is_cluster(void) -{ - /* MPENTIUMIII */ - if (boot_cpu_data.x86 == 6 && - (boot_cpu_data.x86_model >= 7 && boot_cpu_data.x86_model <= 11)) - return 1; - - return 0; -} - -static void setup_unisys(void) -{ - /* - * Determine the generation of the ES7000 currently running. - * - * es7000_plat = 1 if the machine is a 5xx ES7000 box - * es7000_plat = 2 if the machine is a x86_64 ES7000 box - * - */ - if (!(boot_cpu_data.x86 <= 15 && boot_cpu_data.x86_model <= 2)) - es7000_plat = ES7000_ZORRO; - else - es7000_plat = ES7000_CLASSIC; -} - -/* - * Parse the OEM Table: - */ -static int parse_unisys_oem(char *oemptr) -{ - int i; - int success = 0; - unsigned char type, size; - unsigned long val; - char *tp = NULL; - struct psai *psaip = NULL; - struct mip_reg_info *mi; - struct mip_reg *host, *mip; - - tp = oemptr; - - tp += 8; - - for (i = 0; i <= 6; i++) { - type = *tp++; - size = *tp++; - tp -= 2; - switch (type) { - case MIP_REG: - mi = (struct mip_reg_info *)tp; - val = MIP_RD_LO(mi->host_reg); - host_addr = val; - host = (struct mip_reg *)val; - host_reg = __va(host); - val = MIP_RD_LO(mi->mip_reg); - mip_port = MIP_PORT(mi->mip_info); - mip_addr = val; - mip = (struct mip_reg *)val; - mip_reg = __va(mip); - pr_debug("host_reg = 0x%lx\n", - (unsigned long)host_reg); - pr_debug("mip_reg = 0x%lx\n", - (unsigned long)mip_reg); - success++; - break; - case MIP_PSAI_REG: - psaip = (struct psai *)tp; - if (tp != NULL) { - if (psaip->addr) - psai = __va(psaip->addr); - else - psai = NULL; - success++; - } - break; - default: - break; - } - tp += size; - } - - if (success < 2) - es7000_plat = NON_UNISYS; - else - setup_unisys(); - - return es7000_plat; -} - -#ifdef CONFIG_ACPI -static int __init find_unisys_acpi_oem_table(unsigned long *oem_addr) -{ - struct acpi_table_header *header = NULL; - struct es7000_oem_table *table; - acpi_size tbl_size; - acpi_status ret; - int i = 0; - - for (;;) { - ret = acpi_get_table_with_size("OEM1", i++, &header, &tbl_size); - if (!ACPI_SUCCESS(ret)) - return -1; - - if (!memcmp((char *) &header->oem_id, "UNISYS", 6)) - break; - - early_acpi_os_unmap_memory(header, tbl_size); - } - - table = (void *)header; - - oem_addrX = table->OEMTableAddr; - oem_size = table->OEMTableSize; - - early_acpi_os_unmap_memory(header, tbl_size); - - *oem_addr = (unsigned long)__acpi_map_table(oem_addrX, oem_size); - - return 0; -} - -static void __init unmap_unisys_acpi_oem_table(unsigned long oem_addr) -{ - if (!oem_addr) - return; - - __acpi_unmap_table((char *)oem_addr, oem_size); -} - -static int es7000_check_dsdt(void) -{ - struct acpi_table_header header; - - if (ACPI_SUCCESS(acpi_get_table_header(ACPI_SIG_DSDT, 0, &header)) && - !strncmp(header.oem_id, "UNISYS", 6)) - return 1; - return 0; -} - -static int es7000_acpi_ret; - -/* Hook from generic ACPI tables.c */ -static int __init es7000_acpi_madt_oem_check(char *oem_id, char *oem_table_id) -{ - unsigned long oem_addr = 0; - int check_dsdt; - int ret = 0; - - /* check dsdt at first to avoid clear fix_map for oem_addr */ - check_dsdt = es7000_check_dsdt(); - - if (!find_unisys_acpi_oem_table(&oem_addr)) { - if (check_dsdt) { - ret = parse_unisys_oem((char *)oem_addr); - } else { - setup_unisys(); - ret = 1; - } - /* - * we need to unmap it - */ - unmap_unisys_acpi_oem_table(oem_addr); - } - - es7000_acpi_ret = ret; - - return ret && !es7000_apic_is_cluster(); -} - -static int es7000_acpi_madt_oem_check_cluster(char *oem_id, char *oem_table_id) -{ - int ret = es7000_acpi_ret; - - return ret && es7000_apic_is_cluster(); -} - -#else /* !CONFIG_ACPI: */ -static int es7000_acpi_madt_oem_check(char *oem_id, char *oem_table_id) -{ - return 0; -} - -static int es7000_acpi_madt_oem_check_cluster(char *oem_id, char *oem_table_id) -{ - return 0; -} -#endif /* !CONFIG_ACPI */ - -static void es7000_spin(int n) -{ - int i = 0; - - while (i++ < n) - rep_nop(); -} - -static int es7000_mip_write(struct mip_reg *mip_reg) -{ - int status = 0; - int spin; - - spin = MIP_SPIN; - while ((host_reg->off_0x38 & MIP_VALID) != 0) { - if (--spin <= 0) { - WARN(1, "Timeout waiting for Host Valid Flag\n"); - return -1; - } - es7000_spin(MIP_SPIN); - } - - memcpy(host_reg, mip_reg, sizeof(struct mip_reg)); - outb(1, mip_port); - - spin = MIP_SPIN; - - while ((mip_reg->off_0x38 & MIP_VALID) == 0) { - if (--spin <= 0) { - WARN(1, "Timeout waiting for MIP Valid Flag\n"); - return -1; - } - es7000_spin(MIP_SPIN); - } - - status = (mip_reg->off_0x00 & 0xffff0000000000ULL) >> 48; - mip_reg->off_0x38 &= ~MIP_VALID; - - return status; -} - -static void es7000_enable_apic_mode(void) -{ - struct mip_reg es7000_mip_reg; - int mip_status; - - if (!es7000_plat) - return; - - pr_info("Enabling APIC mode.\n"); - memset(&es7000_mip_reg, 0, sizeof(struct mip_reg)); - es7000_mip_reg.off_0x00 = MIP_SW_APIC; - es7000_mip_reg.off_0x38 = MIP_VALID; - - while ((mip_status = es7000_mip_write(&es7000_mip_reg)) != 0) - WARN(1, "Command failed, status = %x\n", mip_status); -} - -static void es7000_vector_allocation_domain(int cpu, struct cpumask *retmask) -{ - /* Careful. Some cpus do not strictly honor the set of cpus - * specified in the interrupt destination when using lowest - * priority interrupt delivery mode. - * - * In particular there was a hyperthreading cpu observed to - * deliver interrupts to the wrong hyperthread when only one - * hyperthread was specified in the interrupt desitination. - */ - cpumask_clear(retmask); - cpumask_bits(retmask)[0] = APIC_ALL_CPUS; -} - - -static void es7000_wait_for_init_deassert(atomic_t *deassert) -{ - while (!atomic_read(deassert)) - cpu_relax(); -} - -static unsigned int es7000_get_apic_id(unsigned long x) -{ - return (x >> 24) & 0xFF; -} - -static void es7000_send_IPI_mask(const struct cpumask *mask, int vector) -{ - default_send_IPI_mask_sequence_phys(mask, vector); -} - -static void es7000_send_IPI_allbutself(int vector) -{ - default_send_IPI_mask_allbutself_phys(cpu_online_mask, vector); -} - -static void es7000_send_IPI_all(int vector) -{ - es7000_send_IPI_mask(cpu_online_mask, vector); -} - -static int es7000_apic_id_registered(void) -{ - return 1; -} - -static const struct cpumask *target_cpus_cluster(void) -{ - return cpu_all_mask; -} - -static const struct cpumask *es7000_target_cpus(void) -{ - return cpumask_of(smp_processor_id()); -} - -static unsigned long es7000_check_apicid_used(physid_mask_t *map, int apicid) -{ - return 0; -} - -static unsigned long es7000_check_apicid_present(int bit) -{ - return physid_isset(bit, phys_cpu_present_map); -} - -static int es7000_early_logical_apicid(int cpu) -{ - /* on es7000, logical apicid is the same as physical */ - return early_per_cpu(x86_bios_cpu_apicid, cpu); -} - -static unsigned long calculate_ldr(int cpu) -{ - unsigned long id = per_cpu(x86_bios_cpu_apicid, cpu); - - return SET_APIC_LOGICAL_ID(id); -} - -/* - * Set up the logical destination ID. - * - * Intel recommends to set DFR, LdR and TPR before enabling - * an APIC. See e.g. "AP-388 82489DX User's Manual" (Intel - * document number 292116). So here it goes... - */ -static void es7000_init_apic_ldr_cluster(void) -{ - unsigned long val; - int cpu = smp_processor_id(); - - apic_write(APIC_DFR, APIC_DFR_CLUSTER); - val = calculate_ldr(cpu); - apic_write(APIC_LDR, val); -} - -static void es7000_init_apic_ldr(void) -{ - unsigned long val; - int cpu = smp_processor_id(); - - apic_write(APIC_DFR, APIC_DFR_FLAT); - val = calculate_ldr(cpu); - apic_write(APIC_LDR, val); -} - -static void es7000_setup_apic_routing(void) -{ - int apic = per_cpu(x86_bios_cpu_apicid, smp_processor_id()); - - pr_info("Enabling APIC mode: %s. Using %d I/O APICs, target cpus %lx\n", - (apic_version[apic] == 0x14) ? - "Physical Cluster" : "Logical Cluster", - nr_ioapics, cpumask_bits(es7000_target_cpus())[0]); -} - -static int es7000_cpu_present_to_apicid(int mps_cpu) -{ - if (!mps_cpu) - return boot_cpu_physical_apicid; - else if (mps_cpu < nr_cpu_ids) - return per_cpu(x86_bios_cpu_apicid, mps_cpu); - else - return BAD_APICID; -} - -static int cpu_id; - -static void es7000_apicid_to_cpu_present(int phys_apicid, physid_mask_t *retmap) -{ - physid_set_mask_of_physid(cpu_id, retmap); - ++cpu_id; -} - -static void es7000_ioapic_phys_id_map(physid_mask_t *phys_map, physid_mask_t *retmap) -{ - /* For clustered we don't have a good way to do this yet - hack */ - physids_promote(0xFFL, retmap); -} - -static int es7000_check_phys_apicid_present(int cpu_physical_apicid) -{ - boot_cpu_physical_apicid = read_apic_id(); - return 1; -} - -static unsigned int es7000_cpu_mask_to_apicid(const struct cpumask *cpumask) -{ - unsigned int round = 0; - int cpu, uninitialized_var(apicid); - - /* - * The cpus in the mask must all be on the apic cluster. - */ - for_each_cpu(cpu, cpumask) { - int new_apicid = early_per_cpu(x86_cpu_to_logical_apicid, cpu); - - if (round && APIC_CLUSTER(apicid) != APIC_CLUSTER(new_apicid)) { - WARN(1, "Not a valid mask!"); - - return BAD_APICID; - } - apicid = new_apicid; - round++; - } - return apicid; -} - -static unsigned int -es7000_cpu_mask_to_apicid_and(const struct cpumask *inmask, - const struct cpumask *andmask) -{ - int apicid = early_per_cpu(x86_cpu_to_logical_apicid, 0); - cpumask_var_t cpumask; - - if (!alloc_cpumask_var(&cpumask, GFP_ATOMIC)) - return apicid; - - cpumask_and(cpumask, inmask, andmask); - cpumask_and(cpumask, cpumask, cpu_online_mask); - apicid = es7000_cpu_mask_to_apicid(cpumask); - - free_cpumask_var(cpumask); - - return apicid; -} - -static int es7000_phys_pkg_id(int cpuid_apic, int index_msb) -{ - return cpuid_apic >> index_msb; -} - -static int probe_es7000(void) -{ - /* probed later in mptable/ACPI hooks */ - return 0; -} - -static int es7000_mps_ret; -static int es7000_mps_oem_check(struct mpc_table *mpc, char *oem, - char *productid) -{ - int ret = 0; - - if (mpc->oemptr) { - struct mpc_oemtable *oem_table = - (struct mpc_oemtable *)mpc->oemptr; - - if (!strncmp(oem, "UNISYS", 6)) - ret = parse_unisys_oem((char *)oem_table); - } - - es7000_mps_ret = ret; - - return ret && !es7000_apic_is_cluster(); -} - -static int es7000_mps_oem_check_cluster(struct mpc_table *mpc, char *oem, - char *productid) -{ - int ret = es7000_mps_ret; - - return ret && es7000_apic_is_cluster(); -} - -/* We've been warned by a false positive warning.Use __refdata to keep calm. */ -static struct apic __refdata apic_es7000_cluster = { - - .name = "es7000", - .probe = probe_es7000, - .acpi_madt_oem_check = es7000_acpi_madt_oem_check_cluster, - .apic_id_registered = es7000_apic_id_registered, - - .irq_delivery_mode = dest_LowestPrio, - /* logical delivery broadcast to all procs: */ - .irq_dest_mode = 1, - - .target_cpus = target_cpus_cluster, - .disable_esr = 1, - .dest_logical = 0, - .check_apicid_used = es7000_check_apicid_used, - .check_apicid_present = es7000_check_apicid_present, - - .vector_allocation_domain = es7000_vector_allocation_domain, - .init_apic_ldr = es7000_init_apic_ldr_cluster, - - .ioapic_phys_id_map = es7000_ioapic_phys_id_map, - .setup_apic_routing = es7000_setup_apic_routing, - .multi_timer_check = NULL, - .cpu_present_to_apicid = es7000_cpu_present_to_apicid, - .apicid_to_cpu_present = es7000_apicid_to_cpu_present, - .setup_portio_remap = NULL, - .check_phys_apicid_present = es7000_check_phys_apicid_present, - .enable_apic_mode = es7000_enable_apic_mode, - .phys_pkg_id = es7000_phys_pkg_id, - .mps_oem_check = es7000_mps_oem_check_cluster, - - .get_apic_id = es7000_get_apic_id, - .set_apic_id = NULL, - .apic_id_mask = 0xFF << 24, - - .cpu_mask_to_apicid = es7000_cpu_mask_to_apicid, - .cpu_mask_to_apicid_and = es7000_cpu_mask_to_apicid_and, - - .send_IPI_mask = es7000_send_IPI_mask, - .send_IPI_mask_allbutself = NULL, - .send_IPI_allbutself = es7000_send_IPI_allbutself, - .send_IPI_all = es7000_send_IPI_all, - .send_IPI_self = default_send_IPI_self, - - .wakeup_secondary_cpu = wakeup_secondary_cpu_via_mip, - - .trampoline_phys_low = 0x467, - .trampoline_phys_high = 0x469, - - .wait_for_init_deassert = NULL, - - /* Nothing to do for most platforms, since cleared by the INIT cycle: */ - .smp_callin_clear_local_apic = NULL, - .inquire_remote_apic = default_inquire_remote_apic, - - .read = native_apic_mem_read, - .write = native_apic_mem_write, - .icr_read = native_apic_icr_read, - .icr_write = native_apic_icr_write, - .wait_icr_idle = native_apic_wait_icr_idle, - .safe_wait_icr_idle = native_safe_apic_wait_icr_idle, - - .x86_32_early_logical_apicid = es7000_early_logical_apicid, -}; - -static struct apic __refdata apic_es7000 = { - - .name = "es7000", - .probe = probe_es7000, - .acpi_madt_oem_check = es7000_acpi_madt_oem_check, - .apic_id_registered = es7000_apic_id_registered, - - .irq_delivery_mode = dest_Fixed, - /* phys delivery to target CPUs: */ - .irq_dest_mode = 0, - - .target_cpus = es7000_target_cpus, - .disable_esr = 1, - .dest_logical = 0, - .check_apicid_used = es7000_check_apicid_used, - .check_apicid_present = es7000_check_apicid_present, - - .vector_allocation_domain = es7000_vector_allocation_domain, - .init_apic_ldr = es7000_init_apic_ldr, - - .ioapic_phys_id_map = es7000_ioapic_phys_id_map, - .setup_apic_routing = es7000_setup_apic_routing, - .multi_timer_check = NULL, - .cpu_present_to_apicid = es7000_cpu_present_to_apicid, - .apicid_to_cpu_present = es7000_apicid_to_cpu_present, - .setup_portio_remap = NULL, - .check_phys_apicid_present = es7000_check_phys_apicid_present, - .enable_apic_mode = es7000_enable_apic_mode, - .phys_pkg_id = es7000_phys_pkg_id, - .mps_oem_check = es7000_mps_oem_check, - - .get_apic_id = es7000_get_apic_id, - .set_apic_id = NULL, - .apic_id_mask = 0xFF << 24, - - .cpu_mask_to_apicid = es7000_cpu_mask_to_apicid, - .cpu_mask_to_apicid_and = es7000_cpu_mask_to_apicid_and, - - .send_IPI_mask = es7000_send_IPI_mask, - .send_IPI_mask_allbutself = NULL, - .send_IPI_allbutself = es7000_send_IPI_allbutself, - .send_IPI_all = es7000_send_IPI_all, - .send_IPI_self = default_send_IPI_self, - - .trampoline_phys_low = 0x467, - .trampoline_phys_high = 0x469, - - .wait_for_init_deassert = es7000_wait_for_init_deassert, - - /* Nothing to do for most platforms, since cleared by the INIT cycle: */ - .smp_callin_clear_local_apic = NULL, - .inquire_remote_apic = default_inquire_remote_apic, - - .read = native_apic_mem_read, - .write = native_apic_mem_write, - .icr_read = native_apic_icr_read, - .icr_write = native_apic_icr_write, - .wait_icr_idle = native_apic_wait_icr_idle, - .safe_wait_icr_idle = native_safe_apic_wait_icr_idle, - - .x86_32_early_logical_apicid = es7000_early_logical_apicid, -}; - -/* - * Need to check for es7000 followed by es7000_cluster, so this order - * in apic_drivers is important. - */ -apic_drivers(apic_es7000, apic_es7000_cluster); diff --git a/arch/x86/kernel/apic/hw_nmi.c b/arch/x86/kernel/apic/hw_nmi.c index 31cb9ae992b..6a1e71bde32 100644 --- a/arch/x86/kernel/apic/hw_nmi.c +++ b/arch/x86/kernel/apic/hw_nmi.c @@ -9,6 +9,7 @@ * */ #include <asm/apic.h> +#include <asm/nmi.h> #include <linux/cpumask.h> #include <linux/kdebug.h> @@ -32,34 +33,44 @@ static DECLARE_BITMAP(backtrace_mask, NR_CPUS) __read_mostly; /* "in progress" flag of arch_trigger_all_cpu_backtrace */ static unsigned long backtrace_flag; -void arch_trigger_all_cpu_backtrace(void) +void arch_trigger_all_cpu_backtrace(bool include_self) { int i; + int cpu = get_cpu(); - if (test_and_set_bit(0, &backtrace_flag)) + if (test_and_set_bit(0, &backtrace_flag)) { /* * If there is already a trigger_all_cpu_backtrace() in progress * (backtrace_flag == 1), don't output double cpu dump infos. */ + put_cpu(); return; + } cpumask_copy(to_cpumask(backtrace_mask), cpu_online_mask); + if (!include_self) + cpumask_clear_cpu(cpu, to_cpumask(backtrace_mask)); - printk(KERN_INFO "sending NMI to all CPUs:\n"); - apic->send_IPI_all(NMI_VECTOR); + if (!cpumask_empty(to_cpumask(backtrace_mask))) { + pr_info("sending NMI to %s CPUs:\n", + (include_self ? "all" : "other")); + apic->send_IPI_mask(to_cpumask(backtrace_mask), NMI_VECTOR); + } /* Wait for up to 10 seconds for all CPUs to do the backtrace */ for (i = 0; i < 10 * 1000; i++) { if (cpumask_empty(to_cpumask(backtrace_mask))) break; mdelay(1); + touch_softlockup_watchdog(); } clear_bit(0, &backtrace_flag); - smp_mb__after_clear_bit(); + smp_mb__after_atomic(); + put_cpu(); } -static int __kprobes +static int arch_trigger_all_cpu_backtrace_handler(unsigned int cmd, struct pt_regs *regs) { int cpu; @@ -79,6 +90,7 @@ arch_trigger_all_cpu_backtrace_handler(unsigned int cmd, struct pt_regs *regs) return NMI_DONE; } +NOKPROBE_SYMBOL(arch_trigger_all_cpu_backtrace_handler); static int __init register_trigger_all_cpu_backtrace(void) { diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c index fb072754bc1..81e08eff05e 100644 --- a/arch/x86/kernel/apic/io_apic.c +++ b/arch/x86/kernel/apic/io_apic.c @@ -37,9 +37,6 @@ #include <linux/kthread.h> #include <linux/jiffies.h> /* time_after() */ #include <linux/slab.h> -#ifdef CONFIG_ACPI -#include <acpi/acpi_bus.h> -#endif #include <linux/bootmem.h> #include <linux/dmar.h> #include <linux/hpet.h> @@ -64,6 +61,7 @@ #include <asm/apic.h> #define __apicdebuginit(type) static type __init + #define for_each_irq_pin(entry, head) \ for (entry = head; entry; entry = entry->next) @@ -123,7 +121,7 @@ int mp_irq_entries; /* GSI interrupts */ static int nr_irqs_gsi = NR_IRQS_LEGACY; -#if defined (CONFIG_MCA) || defined (CONFIG_EISA) +#ifdef CONFIG_EISA int mp_bus_id_to_type[MAX_MP_BUSSES]; #endif @@ -208,20 +206,17 @@ int __init arch_early_irq_init(void) count = ARRAY_SIZE(irq_cfgx); node = cpu_to_node(0); - /* Make sure the legacy interrupts are marked in the bitmap */ - irq_reserve_irqs(0, legacy_pic->nr_legacy_irqs); - for (i = 0; i < count; i++) { irq_set_chip_data(i, &cfg[i]); zalloc_cpumask_var_node(&cfg[i].domain, GFP_KERNEL, node); zalloc_cpumask_var_node(&cfg[i].old_domain, GFP_KERNEL, node); /* * For legacy IRQ's, start with assigning irq0 to irq15 to - * IRQ0_VECTOR to IRQ15_VECTOR on cpu 0. + * IRQ0_VECTOR to IRQ15_VECTOR for all cpu's. */ if (i < legacy_pic->nr_legacy_irqs) { cfg[i].vector = IRQ0_VECTOR + i; - cpumask_set_cpu(0, cfg[i].domain); + cpumask_setall(cfg[i].domain); } } @@ -283,17 +278,6 @@ static struct irq_cfg *alloc_irq_and_cfg_at(unsigned int at, int node) return cfg; } -static int alloc_irq_from(unsigned int from, int node) -{ - return irq_alloc_desc_from(from, node); -} - -static void free_irq_at(unsigned int at, struct irq_cfg *cfg) -{ - free_irq_cfg(at, cfg); - irq_free_desc(at); -} - struct io_apic { unsigned int index; unsigned int unused[3]; @@ -308,22 +292,23 @@ static __attribute_const__ struct io_apic __iomem *io_apic_base(int idx) + (mpc_ioapic_addr(idx) & ~PAGE_MASK); } -static inline void io_apic_eoi(unsigned int apic, unsigned int vector) +void io_apic_eoi(unsigned int apic, unsigned int vector) { struct io_apic __iomem *io_apic = io_apic_base(apic); writel(vector, &io_apic->eoi); } -static inline unsigned int io_apic_read(unsigned int apic, unsigned int reg) +unsigned int native_io_apic_read(unsigned int apic, unsigned int reg) { struct io_apic __iomem *io_apic = io_apic_base(apic); writel(reg, &io_apic->index); return readl(&io_apic->data); } -static inline void io_apic_write(unsigned int apic, unsigned int reg, unsigned int value) +void native_io_apic_write(unsigned int apic, unsigned int reg, unsigned int value) { struct io_apic __iomem *io_apic = io_apic_base(apic); + writel(reg, &io_apic->index); writel(value, &io_apic->data); } @@ -334,7 +319,7 @@ static inline void io_apic_write(unsigned int apic, unsigned int reg, unsigned i * * Older SiS APIC requires we rewrite the index register */ -static inline void io_apic_modify(unsigned int apic, unsigned int reg, unsigned int value) +void native_io_apic_modify(unsigned int apic, unsigned int reg, unsigned int value) { struct io_apic __iomem *io_apic = io_apic_base(apic); @@ -343,29 +328,6 @@ static inline void io_apic_modify(unsigned int apic, unsigned int reg, unsigned writel(value, &io_apic->data); } -static bool io_apic_level_ack_pending(struct irq_cfg *cfg) -{ - struct irq_pin_list *entry; - unsigned long flags; - - raw_spin_lock_irqsave(&ioapic_lock, flags); - for_each_irq_pin(entry, cfg->irq_2_pin) { - unsigned int reg; - int pin; - - pin = entry->pin; - reg = io_apic_read(entry->apic, 0x10 + pin*2); - /* Is the remote IRR bit set? */ - if (reg & IO_APIC_REDIR_REMOTE_IRR) { - raw_spin_unlock_irqrestore(&ioapic_lock, flags); - return true; - } - } - raw_spin_unlock_irqrestore(&ioapic_lock, flags); - - return false; -} - union entry_union { struct { u32 w1, w2; }; struct IO_APIC_route_entry entry; @@ -377,6 +339,7 @@ static struct IO_APIC_route_entry __ioapic_read_entry(int apic, int pin) eu.w1 = io_apic_read(apic, 0x10 + 2 * pin); eu.w2 = io_apic_read(apic, 0x11 + 2 * pin); + return eu.entry; } @@ -384,9 +347,11 @@ static struct IO_APIC_route_entry ioapic_read_entry(int apic, int pin) { union entry_union eu; unsigned long flags; + raw_spin_lock_irqsave(&ioapic_lock, flags); eu.entry = __ioapic_read_entry(apic, pin); raw_spin_unlock_irqrestore(&ioapic_lock, flags); + return eu.entry; } @@ -396,8 +361,7 @@ static struct IO_APIC_route_entry ioapic_read_entry(int apic, int pin) * the interrupt, and we need to make sure the entry is fully populated * before that happens. */ -static void -__ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e) +static void __ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e) { union entry_union eu = {{0, 0}}; @@ -409,6 +373,7 @@ __ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e) static void ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e) { unsigned long flags; + raw_spin_lock_irqsave(&ioapic_lock, flags); __ioapic_write_entry(apic, pin, e); raw_spin_unlock_irqrestore(&ioapic_lock, flags); @@ -435,8 +400,7 @@ static void ioapic_mask_entry(int apic, int pin) * shared ISA-space IRQs, so we have to support them. We are super * fast in the common case, and fast for shared ISA-space IRQs. */ -static int -__add_pin_to_irq_node(struct irq_cfg *cfg, int node, int apic, int pin) +static int __add_pin_to_irq_node(struct irq_cfg *cfg, int node, int apic, int pin) { struct irq_pin_list **last, *entry; @@ -450,8 +414,8 @@ __add_pin_to_irq_node(struct irq_cfg *cfg, int node, int apic, int pin) entry = alloc_irq_pin_list(node); if (!entry) { - printk(KERN_ERR "can not alloc irq_pin_list (%d,%d,%d)\n", - node, apic, pin); + pr_err("can not alloc irq_pin_list (%d,%d,%d)\n", + node, apic, pin); return -ENOMEM; } entry->apic = apic; @@ -521,6 +485,7 @@ static void io_apic_sync(struct irq_pin_list *entry) * a dummy read from the IO-APIC */ struct io_apic __iomem *io_apic; + io_apic = io_apic_base(entry->apic); readl(&io_apic->data); } @@ -574,19 +539,10 @@ static void unmask_ioapic_irq(struct irq_data *data) * Otherwise, we simulate the EOI message manually by changing the trigger * mode to edge and then back to level, with RTE being masked during this. */ -static void __eoi_ioapic_pin(int apic, int pin, int vector, struct irq_cfg *cfg) +void native_eoi_ioapic_pin(int apic, int pin, int vector) { if (mpc_ioapic_ver(apic) >= 0x20) { - /* - * Intr-remapping uses pin number as the virtual vector - * in the RTE. Actual vector is programmed in - * intr-remapping table entry. Hence for the io-apic - * EOI we use the pin number. - */ - if (cfg && irq_remapped(cfg)) - io_apic_eoi(apic, pin); - else - io_apic_eoi(apic, vector); + io_apic_eoi(apic, vector); } else { struct IO_APIC_route_entry entry, entry1; @@ -607,14 +563,15 @@ static void __eoi_ioapic_pin(int apic, int pin, int vector, struct irq_cfg *cfg) } } -static void eoi_ioapic_irq(unsigned int irq, struct irq_cfg *cfg) +void eoi_ioapic_irq(unsigned int irq, struct irq_cfg *cfg) { struct irq_pin_list *entry; unsigned long flags; raw_spin_lock_irqsave(&ioapic_lock, flags); for_each_irq_pin(entry, cfg->irq_2_pin) - __eoi_ioapic_pin(entry->apic, entry->pin, cfg->vector, cfg); + x86_io_apic_ops.eoi_ioapic_pin(entry->apic, entry->pin, + cfg->vector); raw_spin_unlock_irqrestore(&ioapic_lock, flags); } @@ -651,7 +608,7 @@ static void clear_IO_APIC_pin(unsigned int apic, unsigned int pin) } raw_spin_lock_irqsave(&ioapic_lock, flags); - __eoi_ioapic_pin(apic, pin, entry.vector, NULL); + x86_io_apic_ops.eoi_ioapic_pin(apic, pin, entry.vector); raw_spin_unlock_irqrestore(&ioapic_lock, flags); } @@ -662,7 +619,7 @@ static void clear_IO_APIC_pin(unsigned int apic, unsigned int pin) ioapic_mask_entry(apic, pin); entry = ioapic_read_entry(apic, pin); if (entry.irr) - printk(KERN_ERR "Unable to reset IRR for apic: %d, pin :%d\n", + pr_err("Unable to reset IRR for apic: %d, pin :%d\n", mpc_ioapic_id(apic), pin); } @@ -836,7 +793,7 @@ static int __init find_isa_irq_apic(int irq, int type) return -1; } -#if defined(CONFIG_EISA) || defined(CONFIG_MCA) +#ifdef CONFIG_EISA /* * EISA Edge/Level control register, ELCR */ @@ -873,12 +830,6 @@ static int EISA_ELCR(unsigned int irq) #define default_PCI_trigger(idx) (1) #define default_PCI_polarity(idx) (1) -/* MCA interrupts are always polarity zero level triggered, - * when listed as conforming in the MP table. */ - -#define default_MCA_trigger(idx) (1) -#define default_MCA_polarity(idx) default_ISA_polarity(idx) - static int irq_polarity(int idx) { int bus = mp_irqs[idx].srcbus; @@ -902,7 +853,7 @@ static int irq_polarity(int idx) } case 2: /* reserved */ { - printk(KERN_WARNING "broken BIOS!!\n"); + pr_warn("broken BIOS!!\n"); polarity = 1; break; } @@ -913,7 +864,7 @@ static int irq_polarity(int idx) } default: /* invalid */ { - printk(KERN_WARNING "broken BIOS!!\n"); + pr_warn("broken BIOS!!\n"); polarity = 1; break; } @@ -936,7 +887,7 @@ static int irq_trigger(int idx) trigger = default_ISA_trigger(idx); else trigger = default_PCI_trigger(idx); -#if defined(CONFIG_EISA) || defined(CONFIG_MCA) +#ifdef CONFIG_EISA switch (mp_bus_id_to_type[bus]) { case MP_BUS_ISA: /* ISA pin */ { @@ -953,14 +904,9 @@ static int irq_trigger(int idx) /* set before the switch */ break; } - case MP_BUS_MCA: /* MCA pin */ - { - trigger = default_MCA_trigger(idx); - break; - } default: { - printk(KERN_WARNING "broken BIOS!!\n"); + pr_warn("broken BIOS!!\n"); trigger = 1; break; } @@ -974,7 +920,7 @@ static int irq_trigger(int idx) } case 2: /* reserved */ { - printk(KERN_WARNING "broken BIOS!!\n"); + pr_warn("broken BIOS!!\n"); trigger = 1; break; } @@ -985,7 +931,7 @@ static int irq_trigger(int idx) } default: /* invalid */ { - printk(KERN_WARNING "broken BIOS!!\n"); + pr_warn("broken BIOS!!\n"); trigger = 0; break; } @@ -1003,7 +949,7 @@ static int pin_2_irq(int idx, int apic, int pin) * Debugging check, we are in big trouble if this message pops up! */ if (mp_irqs[idx].dstirq != pin) - printk(KERN_ERR "broken BIOS or MPTABLE parser, ayiee!!\n"); + pr_err("broken BIOS or MPTABLE parser, ayiee!!\n"); if (test_bit(bus, mp_bus_not_pci)) { irq = mp_irqs[idx].srcbusirq; @@ -1124,8 +1070,7 @@ __assign_irq_vector(int irq, struct irq_cfg *cfg, const struct cpumask *mask) * 0x80, because int 0x80 is hm, kind of importantish. ;) */ static int current_vector = FIRST_EXTERNAL_VECTOR + VECTOR_OFFSET_START; - static int current_offset = VECTOR_OFFSET_START % 8; - unsigned int old_vector; + static int current_offset = VECTOR_OFFSET_START % 16; int cpu, err; cpumask_var_t tmp_mask; @@ -1135,48 +1080,61 @@ __assign_irq_vector(int irq, struct irq_cfg *cfg, const struct cpumask *mask) if (!alloc_cpumask_var(&tmp_mask, GFP_ATOMIC)) return -ENOMEM; - old_vector = cfg->vector; - if (old_vector) { - cpumask_and(tmp_mask, mask, cpu_online_mask); - cpumask_and(tmp_mask, cfg->domain, tmp_mask); - if (!cpumask_empty(tmp_mask)) { - free_cpumask_var(tmp_mask); - return 0; - } - } - /* Only try and allocate irqs on cpus that are present */ err = -ENOSPC; - for_each_cpu_and(cpu, mask, cpu_online_mask) { - int new_cpu; - int vector, offset; + cpumask_clear(cfg->old_domain); + cpu = cpumask_first_and(mask, cpu_online_mask); + while (cpu < nr_cpu_ids) { + int new_cpu, vector, offset; - apic->vector_allocation_domain(cpu, tmp_mask); + apic->vector_allocation_domain(cpu, tmp_mask, mask); + + if (cpumask_subset(tmp_mask, cfg->domain)) { + err = 0; + if (cpumask_equal(tmp_mask, cfg->domain)) + break; + /* + * New cpumask using the vector is a proper subset of + * the current in use mask. So cleanup the vector + * allocation for the members that are not used anymore. + */ + cpumask_andnot(cfg->old_domain, cfg->domain, tmp_mask); + cfg->move_in_progress = + cpumask_intersects(cfg->old_domain, cpu_online_mask); + cpumask_and(cfg->domain, cfg->domain, tmp_mask); + break; + } vector = current_vector; offset = current_offset; next: - vector += 8; + vector += 16; if (vector >= first_system_vector) { - /* If out of vectors on large boxen, must share them. */ - offset = (offset + 1) % 8; + offset = (offset + 1) % 16; vector = FIRST_EXTERNAL_VECTOR + offset; } - if (unlikely(current_vector == vector)) + + if (unlikely(current_vector == vector)) { + cpumask_or(cfg->old_domain, cfg->old_domain, tmp_mask); + cpumask_andnot(tmp_mask, mask, cfg->old_domain); + cpu = cpumask_first_and(tmp_mask, cpu_online_mask); continue; + } if (test_bit(vector, used_vectors)) goto next; - for_each_cpu_and(new_cpu, tmp_mask, cpu_online_mask) - if (per_cpu(vector_irq, new_cpu)[vector] != -1) + for_each_cpu_and(new_cpu, tmp_mask, cpu_online_mask) { + if (per_cpu(vector_irq, new_cpu)[vector] > VECTOR_UNDEFINED) goto next; + } /* Found one! */ current_vector = vector; current_offset = offset; - if (old_vector) { - cfg->move_in_progress = 1; + if (cfg->vector) { cpumask_copy(cfg->old_domain, cfg->domain); + cfg->move_in_progress = + cpumask_intersects(cfg->old_domain, cpu_online_mask); } for_each_cpu_and(new_cpu, tmp_mask, cpu_online_mask) per_cpu(vector_irq, new_cpu)[vector] = irq; @@ -1208,7 +1166,7 @@ static void __clear_irq_vector(int irq, struct irq_cfg *cfg) vector = cfg->vector; for_each_cpu_and(cpu, cfg->domain, cpu_online_mask) - per_cpu(vector_irq, cpu)[vector] = -1; + per_cpu(vector_irq, cpu)[vector] = VECTOR_UNDEFINED; cfg->vector = 0; cpumask_clear(cfg->domain); @@ -1216,11 +1174,10 @@ static void __clear_irq_vector(int irq, struct irq_cfg *cfg) if (likely(!cfg->move_in_progress)) return; for_each_cpu_and(cpu, cfg->old_domain, cpu_online_mask) { - for (vector = FIRST_EXTERNAL_VECTOR; vector < NR_VECTORS; - vector++) { + for (vector = FIRST_EXTERNAL_VECTOR; vector < NR_VECTORS; vector++) { if (per_cpu(vector_irq, cpu)[vector] != irq) continue; - per_cpu(vector_irq, cpu)[vector] = -1; + per_cpu(vector_irq, cpu)[vector] = VECTOR_UNDEFINED; break; } } @@ -1244,12 +1201,6 @@ void __setup_vector_irq(int cpu) cfg = irq_get_chip_data(irq); if (!cfg) continue; - /* - * If it is a legacy IRQ handled by the legacy PIC, this cpu - * will be part of the irq_cfg's domain. - */ - if (irq < legacy_pic->nr_legacy_irqs && !IO_APIC_IRQ(irq)) - cpumask_set_cpu(cpu, cfg->domain); if (!cpumask_test_cpu(cpu, cfg->domain)) continue; @@ -1259,12 +1210,12 @@ void __setup_vector_irq(int cpu) /* Mark the free vectors */ for (vector = 0; vector < NR_VECTORS; ++vector) { irq = per_cpu(vector_irq, cpu)[vector]; - if (irq < 0) + if (irq <= VECTOR_UNDEFINED) continue; cfg = irq_cfg(irq); if (!cpumask_test_cpu(cpu, cfg->domain)) - per_cpu(vector_irq, cpu)[vector] = -1; + per_cpu(vector_irq, cpu)[vector] = VECTOR_UNDEFINED; } raw_spin_unlock(&vector_lock); } @@ -1311,89 +1262,18 @@ static void ioapic_register_intr(unsigned int irq, struct irq_cfg *cfg, fasteoi = false; } - if (irq_remapped(cfg)) { - irq_set_status_flags(irq, IRQ_MOVE_PCNTXT); - irq_remap_modify_chip_defaults(chip); + if (setup_remapped_irq(irq, cfg, chip)) fasteoi = trigger != 0; - } hdl = fasteoi ? handle_fasteoi_irq : handle_edge_irq; irq_set_chip_and_handler_name(irq, chip, hdl, fasteoi ? "fasteoi" : "edge"); } - -static int setup_ir_ioapic_entry(int irq, - struct IR_IO_APIC_route_entry *entry, +int native_setup_ioapic_entry(int irq, struct IO_APIC_route_entry *entry, unsigned int destination, int vector, struct io_apic_irq_attr *attr) { - int index; - struct irte irte; - int ioapic_id = mpc_ioapic_id(attr->ioapic); - struct intel_iommu *iommu = map_ioapic_to_ir(ioapic_id); - - if (!iommu) { - pr_warn("No mapping iommu for ioapic %d\n", ioapic_id); - return -ENODEV; - } - - index = alloc_irte(iommu, irq, 1); - if (index < 0) { - pr_warn("Failed to allocate IRTE for ioapic %d\n", ioapic_id); - return -ENOMEM; - } - - prepare_irte(&irte, vector, destination); - - /* Set source-id of interrupt request */ - set_ioapic_sid(&irte, ioapic_id); - - modify_irte(irq, &irte); - - apic_printk(APIC_VERBOSE, KERN_DEBUG "IOAPIC[%d]: " - "Set IRTE entry (P:%d FPD:%d Dst_Mode:%d " - "Redir_hint:%d Trig_Mode:%d Dlvry_Mode:%X " - "Avail:%X Vector:%02X Dest:%08X " - "SID:%04X SQ:%X SVT:%X)\n", - attr->ioapic, irte.present, irte.fpd, irte.dst_mode, - irte.redir_hint, irte.trigger_mode, irte.dlvry_mode, - irte.avail, irte.vector, irte.dest_id, - irte.sid, irte.sq, irte.svt); - - memset(entry, 0, sizeof(*entry)); - - entry->index2 = (index >> 15) & 0x1; - entry->zero = 0; - entry->format = 1; - entry->index = (index & 0x7fff); - /* - * IO-APIC RTE will be configured with virtual vector. - * irq handler will do the explicit EOI to the io-apic. - */ - entry->vector = attr->ioapic_pin; - entry->mask = 0; /* enable IRQ */ - entry->trigger = attr->trigger; - entry->polarity = attr->polarity; - - /* Mask level triggered irqs. - * Use IRQ_DELAYED_DISABLE for edge triggered irqs. - */ - if (attr->trigger) - entry->mask = 1; - - return 0; -} - -static int setup_ioapic_entry(int irq, struct IO_APIC_route_entry *entry, - unsigned int destination, int vector, - struct io_apic_irq_attr *attr) -{ - if (intr_remapping_enabled) - return setup_ir_ioapic_entry(irq, - (struct IR_IO_APIC_route_entry *)entry, - destination, vector, attr); - memset(entry, 0, sizeof(*entry)); entry->delivery_mode = apic->irq_delivery_mode; @@ -1422,18 +1302,18 @@ static void setup_ioapic_irq(unsigned int irq, struct irq_cfg *cfg, if (!IO_APIC_IRQ(irq)) return; - /* - * For legacy irqs, cfg->domain starts with cpu 0 for legacy - * controllers like 8259. Now that IO-APIC can handle this irq, update - * the cfg->domain. - */ - if (irq < legacy_pic->nr_legacy_irqs && cpumask_test_cpu(0, cfg->domain)) - apic->vector_allocation_domain(0, cfg->domain); if (assign_irq_vector(irq, cfg, apic->target_cpus())) return; - dest = apic->cpu_mask_to_apicid_and(cfg->domain, apic->target_cpus()); + if (apic->cpu_mask_to_apicid_and(cfg->domain, apic->target_cpus(), + &dest)) { + pr_warn("Failed to obtain apicid for ioapic %d, pin %d\n", + mpc_ioapic_id(attr->ioapic), attr->ioapic_pin); + __clear_irq_vector(irq, cfg); + + return; + } apic_printk(APIC_VERBOSE,KERN_DEBUG "IOAPIC[%d]: Set routing entry (%d-%d -> 0x%x -> " @@ -1441,7 +1321,7 @@ static void setup_ioapic_irq(unsigned int irq, struct irq_cfg *cfg, attr->ioapic, mpc_ioapic_id(attr->ioapic), attr->ioapic_pin, cfg->vector, irq, attr->trigger, attr->polarity, dest); - if (setup_ioapic_entry(irq, &entry, dest, cfg->vector, attr)) { + if (x86_io_apic_ops.setup_entry(irq, &entry, dest, cfg->vector, attr)) { pr_warn("Failed to setup ioapic entry for ioapic %d, pin %d\n", mpc_ioapic_id(attr->ioapic), attr->ioapic_pin); __clear_irq_vector(irq, cfg); @@ -1545,12 +1425,10 @@ void setup_IO_APIC_irq_extra(u32 gsi) * Set up the timer pin, possibly with the 8259A-master behind. */ static void __init setup_timer_IRQ0_pin(unsigned int ioapic_idx, - unsigned int pin, int vector) + unsigned int pin, int vector) { struct IO_APIC_route_entry entry; - - if (intr_remapping_enabled) - return; + unsigned int dest; memset(&entry, 0, sizeof(entry)); @@ -1558,9 +1436,13 @@ static void __init setup_timer_IRQ0_pin(unsigned int ioapic_idx, * We use logical delivery to get the timer IRQ * to the first CPU. */ + if (unlikely(apic->cpu_mask_to_apicid_and(apic->target_cpus(), + apic->target_cpus(), &dest))) + dest = BAD_APICID; + entry.dest_mode = apic->irq_dest_mode; entry.mask = 0; /* don't mask IRQ for edge */ - entry.dest = apic->cpu_mask_to_apicid(apic->target_cpus()); + entry.dest = dest; entry.delivery_mode = apic->irq_delivery_mode; entry.polarity = 0; entry.trigger = 0; @@ -1579,9 +1461,68 @@ static void __init setup_timer_IRQ0_pin(unsigned int ioapic_idx, ioapic_write_entry(ioapic_idx, pin, entry); } -__apicdebuginit(void) print_IO_APIC(int ioapic_idx) +void native_io_apic_print_entries(unsigned int apic, unsigned int nr_entries) +{ + int i; + + pr_debug(" NR Dst Mask Trig IRR Pol Stat Dmod Deli Vect:\n"); + + for (i = 0; i <= nr_entries; i++) { + struct IO_APIC_route_entry entry; + + entry = ioapic_read_entry(apic, i); + + pr_debug(" %02x %02X ", i, entry.dest); + pr_cont("%1d %1d %1d %1d %1d " + "%1d %1d %02X\n", + entry.mask, + entry.trigger, + entry.irr, + entry.polarity, + entry.delivery_status, + entry.dest_mode, + entry.delivery_mode, + entry.vector); + } +} + +void intel_ir_io_apic_print_entries(unsigned int apic, + unsigned int nr_entries) { int i; + + pr_debug(" NR Indx Fmt Mask Trig IRR Pol Stat Indx2 Zero Vect:\n"); + + for (i = 0; i <= nr_entries; i++) { + struct IR_IO_APIC_route_entry *ir_entry; + struct IO_APIC_route_entry entry; + + entry = ioapic_read_entry(apic, i); + + ir_entry = (struct IR_IO_APIC_route_entry *)&entry; + + pr_debug(" %02x %04X ", i, ir_entry->index); + pr_cont("%1d %1d %1d %1d %1d " + "%1d %1d %X %02X\n", + ir_entry->format, + ir_entry->mask, + ir_entry->trigger, + ir_entry->irr, + ir_entry->polarity, + ir_entry->delivery_status, + ir_entry->index2, + ir_entry->zero, + ir_entry->vector); + } +} + +void ioapic_zap_locks(void) +{ + raw_spin_lock_init(&ioapic_lock); +} + +__apicdebuginit(void) print_IO_APIC(int ioapic_idx) +{ union IO_APIC_reg_00 reg_00; union IO_APIC_reg_01 reg_01; union IO_APIC_reg_02 reg_02; @@ -1597,7 +1538,6 @@ __apicdebuginit(void) print_IO_APIC(int ioapic_idx) reg_03.raw = io_apic_read(ioapic_idx, 3); raw_spin_unlock_irqrestore(&ioapic_lock, flags); - printk("\n"); printk(KERN_DEBUG "IO APIC #%d......\n", mpc_ioapic_id(ioapic_idx)); printk(KERN_DEBUG ".... register #00: %08X\n", reg_00.raw); printk(KERN_DEBUG "....... : physical APIC id: %02X\n", reg_00.bits.ID); @@ -1635,58 +1575,7 @@ __apicdebuginit(void) print_IO_APIC(int ioapic_idx) printk(KERN_DEBUG ".... IRQ redirection table:\n"); - if (intr_remapping_enabled) { - printk(KERN_DEBUG " NR Indx Fmt Mask Trig IRR" - " Pol Stat Indx2 Zero Vect:\n"); - } else { - printk(KERN_DEBUG " NR Dst Mask Trig IRR Pol" - " Stat Dmod Deli Vect:\n"); - } - - for (i = 0; i <= reg_01.bits.entries; i++) { - if (intr_remapping_enabled) { - struct IO_APIC_route_entry entry; - struct IR_IO_APIC_route_entry *ir_entry; - - entry = ioapic_read_entry(ioapic_idx, i); - ir_entry = (struct IR_IO_APIC_route_entry *) &entry; - printk(KERN_DEBUG " %02x %04X ", - i, - ir_entry->index - ); - printk("%1d %1d %1d %1d %1d " - "%1d %1d %X %02X\n", - ir_entry->format, - ir_entry->mask, - ir_entry->trigger, - ir_entry->irr, - ir_entry->polarity, - ir_entry->delivery_status, - ir_entry->index2, - ir_entry->zero, - ir_entry->vector - ); - } else { - struct IO_APIC_route_entry entry; - - entry = ioapic_read_entry(ioapic_idx, i); - printk(KERN_DEBUG " %02x %02X ", - i, - entry.dest - ); - printk("%1d %1d %1d %1d %1d " - "%1d %1d %02X\n", - entry.mask, - entry.trigger, - entry.irr, - entry.polarity, - entry.delivery_status, - entry.dest_mode, - entry.delivery_mode, - entry.vector - ); - } - } + x86_io_apic_ops.print_entries(ioapic_idx, reg_01.bits.entries); } __apicdebuginit(void) print_IO_APICs(void) @@ -1727,8 +1616,8 @@ __apicdebuginit(void) print_IO_APICs(void) continue; printk(KERN_DEBUG "IRQ%d ", irq); for_each_irq_pin(entry, cfg->irq_2_pin) - printk("-> %d:%d", entry->apic, entry->pin); - printk("\n"); + pr_cont("-> %d:%d", entry->apic, entry->pin); + pr_cont("\n"); } printk(KERN_INFO ".................................... done.\n"); @@ -1741,9 +1630,9 @@ __apicdebuginit(void) print_APIC_field(int base) printk(KERN_DEBUG); for (i = 0; i < 8; i++) - printk(KERN_CONT "%08x", apic_read(base + i*0x10)); + pr_cont("%08x", apic_read(base + i*0x10)); - printk(KERN_CONT "\n"); + pr_cont("\n"); } __apicdebuginit(void) print_local_APIC(void *dummy) @@ -1845,7 +1734,7 @@ __apicdebuginit(void) print_local_APIC(void *dummy) printk(KERN_DEBUG "... APIC EILVT%d: %08x\n", i, v); } } - printk("\n"); + pr_cont("\n"); } __apicdebuginit(void) print_local_APICs(int maxcpu) @@ -1988,30 +1877,14 @@ void __init enable_IO_APIC(void) clear_IO_APIC(); } -/* - * Not an __init, needed by the reboot code - */ -void disable_IO_APIC(void) +void native_disable_io_apic(void) { /* - * Clear the IO-APIC before rebooting: - */ - clear_IO_APIC(); - - if (!legacy_pic->nr_legacy_irqs) - return; - - /* * If the i8259 is routed through an IOAPIC * Put that IOAPIC in virtual wire mode * so legacy interrupts can be delivered. - * - * With interrupt-remapping, for now we will use virtual wire A mode, - * as virtual wire B is little complex (need to configure both - * IOAPIC RTE as well as interrupt-remapping table entry). - * As this gets called during crash dump, keep this simple for now. */ - if (ioapic_i8259.pin != -1 && !intr_remapping_enabled) { + if (ioapic_i8259.pin != -1) { struct IO_APIC_route_entry entry; memset(&entry, 0, sizeof(entry)); @@ -2031,12 +1904,25 @@ void disable_IO_APIC(void) ioapic_write_entry(ioapic_i8259.apic, ioapic_i8259.pin, entry); } + if (cpu_has_apic || apic_from_smp_config()) + disconnect_bsp_APIC(ioapic_i8259.pin != -1); + +} + +/* + * Not an __init, needed by the reboot code + */ +void disable_IO_APIC(void) +{ /* - * Use virtual wire A mode when interrupt remapping is enabled. + * Clear the IO-APIC before rebooting: */ - if (cpu_has_apic || apic_from_smp_config()) - disconnect_bsp_APIC(!intr_remapping_enabled && - ioapic_i8259.pin != -1); + clear_IO_APIC(); + + if (!legacy_pic->nr_legacy_irqs) + return; + + x86_io_apic_ops.disable(); } #ifdef CONFIG_X86_32 @@ -2141,7 +2027,7 @@ void __init setup_ioapic_ids_from_mpc_nocheck(void) reg_00.raw = io_apic_read(ioapic_idx, 0); raw_spin_unlock_irqrestore(&ioapic_lock, flags); if (reg_00.bits.ID != mpc_ioapic_id(ioapic_idx)) - printk("could not set ID!\n"); + pr_cont("could not set ID!\n"); else apic_printk(APIC_VERBOSE, " ok.\n"); } @@ -2252,9 +2138,11 @@ static int ioapic_retrigger_irq(struct irq_data *data) { struct irq_cfg *cfg = data->chip_data; unsigned long flags; + int cpu; raw_spin_lock_irqsave(&vector_lock, flags); - apic->send_IPI_mask(cpumask_of(cpumask_first(cfg->domain)), cfg->vector); + cpu = cpumask_first_and(cfg->domain, cpu_online_mask); + apic->send_IPI_mask(cpumask_of(cpu), cfg->vector); raw_spin_unlock_irqrestore(&vector_lock, flags); return 1; @@ -2286,137 +2174,7 @@ void send_cleanup_vector(struct irq_cfg *cfg) cfg->move_in_progress = 0; } -static void __target_IO_APIC_irq(unsigned int irq, unsigned int dest, struct irq_cfg *cfg) -{ - int apic, pin; - struct irq_pin_list *entry; - u8 vector = cfg->vector; - - for_each_irq_pin(entry, cfg->irq_2_pin) { - unsigned int reg; - - apic = entry->apic; - pin = entry->pin; - /* - * With interrupt-remapping, destination information comes - * from interrupt-remapping table entry. - */ - if (!irq_remapped(cfg)) - io_apic_write(apic, 0x11 + pin*2, dest); - reg = io_apic_read(apic, 0x10 + pin*2); - reg &= ~IO_APIC_REDIR_VECTOR_MASK; - reg |= vector; - io_apic_modify(apic, 0x10 + pin*2, reg); - } -} - -/* - * Either sets data->affinity to a valid value, and returns - * ->cpu_mask_to_apicid of that in dest_id, or returns -1 and - * leaves data->affinity untouched. - */ -int __ioapic_set_affinity(struct irq_data *data, const struct cpumask *mask, - unsigned int *dest_id) -{ - struct irq_cfg *cfg = data->chip_data; - - if (!cpumask_intersects(mask, cpu_online_mask)) - return -1; - - if (assign_irq_vector(data->irq, data->chip_data, mask)) - return -1; - - cpumask_copy(data->affinity, mask); - - *dest_id = apic->cpu_mask_to_apicid_and(mask, cfg->domain); - return 0; -} - -static int -ioapic_set_affinity(struct irq_data *data, const struct cpumask *mask, - bool force) -{ - unsigned int dest, irq = data->irq; - unsigned long flags; - int ret; - - raw_spin_lock_irqsave(&ioapic_lock, flags); - ret = __ioapic_set_affinity(data, mask, &dest); - if (!ret) { - /* Only the high 8 bits are valid. */ - dest = SET_APIC_LOGICAL_ID(dest); - __target_IO_APIC_irq(irq, dest, data->chip_data); - } - raw_spin_unlock_irqrestore(&ioapic_lock, flags); - return ret; -} - -#ifdef CONFIG_IRQ_REMAP - -/* - * Migrate the IO-APIC irq in the presence of intr-remapping. - * - * For both level and edge triggered, irq migration is a simple atomic - * update(of vector and cpu destination) of IRTE and flush the hardware cache. - * - * For level triggered, we eliminate the io-apic RTE modification (with the - * updated vector information), by using a virtual vector (io-apic pin number). - * Real vector that is used for interrupting cpu will be coming from - * the interrupt-remapping table entry. - * - * As the migration is a simple atomic update of IRTE, the same mechanism - * is used to migrate MSI irq's in the presence of interrupt-remapping. - */ -static int -ir_ioapic_set_affinity(struct irq_data *data, const struct cpumask *mask, - bool force) -{ - struct irq_cfg *cfg = data->chip_data; - unsigned int dest, irq = data->irq; - struct irte irte; - - if (!cpumask_intersects(mask, cpu_online_mask)) - return -EINVAL; - - if (get_irte(irq, &irte)) - return -EBUSY; - - if (assign_irq_vector(irq, cfg, mask)) - return -EBUSY; - - dest = apic->cpu_mask_to_apicid_and(cfg->domain, mask); - - irte.vector = cfg->vector; - irte.dest_id = IRTE_DEST(dest); - - /* - * Atomically updates the IRTE with the new destination, vector - * and flushes the interrupt entry cache. - */ - modify_irte(irq, &irte); - - /* - * After this point, all the interrupts will start arriving - * at the new destination. So, time to cleanup the previous - * vector allocation. - */ - if (cfg->move_in_progress) - send_cleanup_vector(cfg); - - cpumask_copy(data->affinity, mask); - return 0; -} - -#else -static inline int -ir_ioapic_set_affinity(struct irq_data *data, const struct cpumask *mask, - bool force) -{ - return 0; -} -#endif - -asmlinkage void smp_irq_move_cleanup_interrupt(void) +asmlinkage __visible void smp_irq_move_cleanup_interrupt(void) { unsigned vector, me; @@ -2426,13 +2184,13 @@ asmlinkage void smp_irq_move_cleanup_interrupt(void) me = smp_processor_id(); for (vector = FIRST_EXTERNAL_VECTOR; vector < NR_VECTORS; vector++) { - unsigned int irq; + int irq; unsigned int irr; struct irq_desc *desc; struct irq_cfg *cfg; irq = __this_cpu_read(vector_irq[vector]); - if (irq == -1) + if (irq <= VECTOR_UNDEFINED) continue; desc = irq_to_desc(irq); @@ -2440,6 +2198,9 @@ asmlinkage void smp_irq_move_cleanup_interrupt(void) continue; cfg = irq_cfg(irq); + if (!cfg) + continue; + raw_spin_lock(&desc->lock); /* @@ -2503,6 +2264,84 @@ void irq_force_complete_move(int irq) static inline void irq_complete_move(struct irq_cfg *cfg) { } #endif +static void __target_IO_APIC_irq(unsigned int irq, unsigned int dest, struct irq_cfg *cfg) +{ + int apic, pin; + struct irq_pin_list *entry; + u8 vector = cfg->vector; + + for_each_irq_pin(entry, cfg->irq_2_pin) { + unsigned int reg; + + apic = entry->apic; + pin = entry->pin; + + io_apic_write(apic, 0x11 + pin*2, dest); + reg = io_apic_read(apic, 0x10 + pin*2); + reg &= ~IO_APIC_REDIR_VECTOR_MASK; + reg |= vector; + io_apic_modify(apic, 0x10 + pin*2, reg); + } +} + +/* + * Either sets data->affinity to a valid value, and returns + * ->cpu_mask_to_apicid of that in dest_id, or returns -1 and + * leaves data->affinity untouched. + */ +int __ioapic_set_affinity(struct irq_data *data, const struct cpumask *mask, + unsigned int *dest_id) +{ + struct irq_cfg *cfg = data->chip_data; + unsigned int irq = data->irq; + int err; + + if (!config_enabled(CONFIG_SMP)) + return -EPERM; + + if (!cpumask_intersects(mask, cpu_online_mask)) + return -EINVAL; + + err = assign_irq_vector(irq, cfg, mask); + if (err) + return err; + + err = apic->cpu_mask_to_apicid_and(mask, cfg->domain, dest_id); + if (err) { + if (assign_irq_vector(irq, cfg, data->affinity)) + pr_err("Failed to recover vector for irq %d\n", irq); + return err; + } + + cpumask_copy(data->affinity, mask); + + return 0; +} + + +int native_ioapic_set_affinity(struct irq_data *data, + const struct cpumask *mask, + bool force) +{ + unsigned int dest, irq = data->irq; + unsigned long flags; + int ret; + + if (!config_enabled(CONFIG_SMP)) + return -EPERM; + + raw_spin_lock_irqsave(&ioapic_lock, flags); + ret = __ioapic_set_affinity(data, mask, &dest); + if (!ret) { + /* Only the high 8 bits are valid. */ + dest = SET_APIC_LOGICAL_ID(dest); + __target_IO_APIC_irq(irq, dest, data->chip_data); + ret = IRQ_SET_MASK_OK_NOCOPY; + } + raw_spin_unlock_irqrestore(&ioapic_lock, flags); + return ret; +} + static void ack_apic_edge(struct irq_data *data) { irq_complete_move(data->chip_data); @@ -2512,21 +2351,96 @@ static void ack_apic_edge(struct irq_data *data) atomic_t irq_mis_count; -static void ack_apic_level(struct irq_data *data) +#ifdef CONFIG_GENERIC_PENDING_IRQ +static bool io_apic_level_ack_pending(struct irq_cfg *cfg) { - struct irq_cfg *cfg = data->chip_data; - int i, do_unmask_irq = 0, irq = data->irq; - unsigned long v; + struct irq_pin_list *entry; + unsigned long flags; - irq_complete_move(cfg); -#ifdef CONFIG_GENERIC_PENDING_IRQ + raw_spin_lock_irqsave(&ioapic_lock, flags); + for_each_irq_pin(entry, cfg->irq_2_pin) { + unsigned int reg; + int pin; + + pin = entry->pin; + reg = io_apic_read(entry->apic, 0x10 + pin*2); + /* Is the remote IRR bit set? */ + if (reg & IO_APIC_REDIR_REMOTE_IRR) { + raw_spin_unlock_irqrestore(&ioapic_lock, flags); + return true; + } + } + raw_spin_unlock_irqrestore(&ioapic_lock, flags); + + return false; +} + +static inline bool ioapic_irqd_mask(struct irq_data *data, struct irq_cfg *cfg) +{ /* If we are moving the irq we need to mask it */ if (unlikely(irqd_is_setaffinity_pending(data))) { - do_unmask_irq = 1; mask_ioapic(cfg); + return true; } + return false; +} + +static inline void ioapic_irqd_unmask(struct irq_data *data, + struct irq_cfg *cfg, bool masked) +{ + if (unlikely(masked)) { + /* Only migrate the irq if the ack has been received. + * + * On rare occasions the broadcast level triggered ack gets + * delayed going to ioapics, and if we reprogram the + * vector while Remote IRR is still set the irq will never + * fire again. + * + * To prevent this scenario we read the Remote IRR bit + * of the ioapic. This has two effects. + * - On any sane system the read of the ioapic will + * flush writes (and acks) going to the ioapic from + * this cpu. + * - We get to see if the ACK has actually been delivered. + * + * Based on failed experiments of reprogramming the + * ioapic entry from outside of irq context starting + * with masking the ioapic entry and then polling until + * Remote IRR was clear before reprogramming the + * ioapic I don't trust the Remote IRR bit to be + * completey accurate. + * + * However there appears to be no other way to plug + * this race, so if the Remote IRR bit is not + * accurate and is causing problems then it is a hardware bug + * and you can go talk to the chipset vendor about it. + */ + if (!io_apic_level_ack_pending(cfg)) + irq_move_masked_irq(data); + unmask_ioapic(cfg); + } +} +#else +static inline bool ioapic_irqd_mask(struct irq_data *data, struct irq_cfg *cfg) +{ + return false; +} +static inline void ioapic_irqd_unmask(struct irq_data *data, + struct irq_cfg *cfg, bool masked) +{ +} #endif +static void ack_apic_level(struct irq_data *data) +{ + struct irq_cfg *cfg = data->chip_data; + int i, irq = data->irq; + unsigned long v; + bool masked; + + irq_complete_move(cfg); + masked = ioapic_irqd_mask(data, cfg); + /* * It appears there is an erratum which affects at least version 0x11 * of I/O APIC (that's the 82093AA and cores integrated into various @@ -2581,69 +2495,9 @@ static void ack_apic_level(struct irq_data *data) eoi_ioapic_irq(irq, cfg); } - /* Now we can move and renable the irq */ - if (unlikely(do_unmask_irq)) { - /* Only migrate the irq if the ack has been received. - * - * On rare occasions the broadcast level triggered ack gets - * delayed going to ioapics, and if we reprogram the - * vector while Remote IRR is still set the irq will never - * fire again. - * - * To prevent this scenario we read the Remote IRR bit - * of the ioapic. This has two effects. - * - On any sane system the read of the ioapic will - * flush writes (and acks) going to the ioapic from - * this cpu. - * - We get to see if the ACK has actually been delivered. - * - * Based on failed experiments of reprogramming the - * ioapic entry from outside of irq context starting - * with masking the ioapic entry and then polling until - * Remote IRR was clear before reprogramming the - * ioapic I don't trust the Remote IRR bit to be - * completey accurate. - * - * However there appears to be no other way to plug - * this race, so if the Remote IRR bit is not - * accurate and is causing problems then it is a hardware bug - * and you can go talk to the chipset vendor about it. - */ - if (!io_apic_level_ack_pending(cfg)) - irq_move_masked_irq(data); - unmask_ioapic(cfg); - } -} - -#ifdef CONFIG_IRQ_REMAP -static void ir_ack_apic_edge(struct irq_data *data) -{ - ack_APIC_irq(); -} - -static void ir_ack_apic_level(struct irq_data *data) -{ - ack_APIC_irq(); - eoi_ioapic_irq(data->irq, data->chip_data); + ioapic_irqd_unmask(data, cfg, masked); } -static void ir_print_prefix(struct irq_data *data, struct seq_file *p) -{ - seq_printf(p, " IR-%s", data->chip->name); -} - -static void irq_remap_modify_chip_defaults(struct irq_chip *chip) -{ - chip->irq_print_chip = ir_print_prefix; - chip->irq_ack = ir_ack_apic_edge; - chip->irq_eoi = ir_ack_apic_level; - -#ifdef CONFIG_SMP - chip->irq_set_affinity = ir_ioapic_set_affinity; -#endif -} -#endif /* CONFIG_IRQ_REMAP */ - static struct irq_chip ioapic_chip __read_mostly = { .name = "IO-APIC", .irq_startup = startup_ioapic_irq, @@ -2651,9 +2505,7 @@ static struct irq_chip ioapic_chip __read_mostly = { .irq_unmask = unmask_ioapic_irq, .irq_ack = ack_apic_edge, .irq_eoi = ack_apic_level, -#ifdef CONFIG_SMP - .irq_set_affinity = ioapic_set_affinity, -#endif + .irq_set_affinity = native_ioapic_set_affinity, .irq_retrigger = ioapic_retrigger_irq, }; @@ -2852,8 +2704,7 @@ static inline void __init check_timer(void) * 8259A. */ if (pin1 == -1) { - if (intr_remapping_enabled) - panic("BIOS bug: timer not connected to IO-APIC"); + panic_if_irq_remap("BIOS bug: timer not connected to IO-APIC"); pin1 = pin2; apic1 = apic2; no_pin1 = 1; @@ -2885,8 +2736,7 @@ static inline void __init check_timer(void) clear_IO_APIC_pin(0, pin1); goto out; } - if (intr_remapping_enabled) - panic("timer doesn't work through Interrupt-remapped IO-APIC"); + panic_if_irq_remap("timer doesn't work through Interrupt-remapped IO-APIC"); local_irq_disable(); clear_IO_APIC_pin(apic1, pin1); if (!no_pin1) @@ -3051,74 +2901,74 @@ static int __init ioapic_init_ops(void) device_initcall(ioapic_init_ops); /* - * Dynamic irq allocate and deallocation + * Dynamic irq allocate and deallocation. Should be replaced by irq domains! */ -unsigned int create_irq_nr(unsigned int from, int node) +int arch_setup_hwirq(unsigned int irq, int node) { struct irq_cfg *cfg; unsigned long flags; - unsigned int ret = 0; - int irq; - - if (from < nr_irqs_gsi) - from = nr_irqs_gsi; + int ret; - irq = alloc_irq_from(from, node); - if (irq < 0) - return 0; cfg = alloc_irq_cfg(irq, node); - if (!cfg) { - free_irq_at(irq, NULL); - return 0; - } + if (!cfg) + return -ENOMEM; raw_spin_lock_irqsave(&vector_lock, flags); - if (!__assign_irq_vector(irq, cfg, apic->target_cpus())) - ret = irq; + ret = __assign_irq_vector(irq, cfg, apic->target_cpus()); raw_spin_unlock_irqrestore(&vector_lock, flags); - if (ret) { + if (!ret) irq_set_chip_data(irq, cfg); - irq_clear_status_flags(irq, IRQ_NOREQUEST); - } else { - free_irq_at(irq, cfg); - } + else + free_irq_cfg(irq, cfg); return ret; } -int create_irq(void) -{ - int node = cpu_to_node(0); - unsigned int irq_want; - int irq; - - irq_want = nr_irqs_gsi; - irq = create_irq_nr(irq_want, node); - - if (irq == 0) - irq = -1; - - return irq; -} - -void destroy_irq(unsigned int irq) +void arch_teardown_hwirq(unsigned int irq) { struct irq_cfg *cfg = irq_get_chip_data(irq); unsigned long flags; - irq_set_status_flags(irq, IRQ_NOREQUEST|IRQ_NOPROBE); - - if (irq_remapped(cfg)) - free_irte(irq); + free_remapped_irq(irq); raw_spin_lock_irqsave(&vector_lock, flags); __clear_irq_vector(irq, cfg); raw_spin_unlock_irqrestore(&vector_lock, flags); - free_irq_at(irq, cfg); + free_irq_cfg(irq, cfg); } /* * MSI message composition */ +void native_compose_msi_msg(struct pci_dev *pdev, + unsigned int irq, unsigned int dest, + struct msi_msg *msg, u8 hpet_id) +{ + struct irq_cfg *cfg = irq_cfg(irq); + + msg->address_hi = MSI_ADDR_BASE_HI; + + if (x2apic_enabled()) + msg->address_hi |= MSI_ADDR_EXT_DEST_ID(dest); + + msg->address_lo = + MSI_ADDR_BASE_LO | + ((apic->irq_dest_mode == 0) ? + MSI_ADDR_DEST_MODE_PHYSICAL: + MSI_ADDR_DEST_MODE_LOGICAL) | + ((apic->irq_delivery_mode != dest_LowestPrio) ? + MSI_ADDR_REDIRECTION_CPU: + MSI_ADDR_REDIRECTION_LOWPRI) | + MSI_ADDR_DEST_ID(dest); + + msg->data = + MSI_DATA_TRIGGER_EDGE | + MSI_DATA_LEVEL_ASSERT | + ((apic->irq_delivery_mode != dest_LowestPrio) ? + MSI_DATA_DELIVERY_FIXED: + MSI_DATA_DELIVERY_LOWPRI) | + MSI_DATA_VECTOR(cfg->vector); +} + #ifdef CONFIG_PCI_MSI static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_msg *msg, u8 hpet_id) @@ -3135,70 +2985,27 @@ static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, if (err) return err; - dest = apic->cpu_mask_to_apicid_and(cfg->domain, apic->target_cpus()); - - if (irq_remapped(cfg)) { - struct irte irte; - int ir_index; - u16 sub_handle; - - ir_index = map_irq_to_irte_handle(irq, &sub_handle); - BUG_ON(ir_index == -1); - - prepare_irte(&irte, cfg->vector, dest); - - /* Set source-id of interrupt request */ - if (pdev) - set_msi_sid(&irte, pdev); - else - set_hpet_sid(&irte, hpet_id); + err = apic->cpu_mask_to_apicid_and(cfg->domain, + apic->target_cpus(), &dest); + if (err) + return err; - modify_irte(irq, &irte); + x86_msi.compose_msi_msg(pdev, irq, dest, msg, hpet_id); - msg->address_hi = MSI_ADDR_BASE_HI; - msg->data = sub_handle; - msg->address_lo = MSI_ADDR_BASE_LO | MSI_ADDR_IR_EXT_INT | - MSI_ADDR_IR_SHV | - MSI_ADDR_IR_INDEX1(ir_index) | - MSI_ADDR_IR_INDEX2(ir_index); - } else { - if (x2apic_enabled()) - msg->address_hi = MSI_ADDR_BASE_HI | - MSI_ADDR_EXT_DEST_ID(dest); - else - msg->address_hi = MSI_ADDR_BASE_HI; - - msg->address_lo = - MSI_ADDR_BASE_LO | - ((apic->irq_dest_mode == 0) ? - MSI_ADDR_DEST_MODE_PHYSICAL: - MSI_ADDR_DEST_MODE_LOGICAL) | - ((apic->irq_delivery_mode != dest_LowestPrio) ? - MSI_ADDR_REDIRECTION_CPU: - MSI_ADDR_REDIRECTION_LOWPRI) | - MSI_ADDR_DEST_ID(dest); - - msg->data = - MSI_DATA_TRIGGER_EDGE | - MSI_DATA_LEVEL_ASSERT | - ((apic->irq_delivery_mode != dest_LowestPrio) ? - MSI_DATA_DELIVERY_FIXED: - MSI_DATA_DELIVERY_LOWPRI) | - MSI_DATA_VECTOR(cfg->vector); - } - return err; + return 0; } -#ifdef CONFIG_SMP static int msi_set_affinity(struct irq_data *data, const struct cpumask *mask, bool force) { struct irq_cfg *cfg = data->chip_data; struct msi_msg msg; unsigned int dest; + int ret; - if (__ioapic_set_affinity(data, mask, &dest)) - return -1; + ret = __ioapic_set_affinity(data, mask, &dest); + if (ret) + return ret; __get_cached_msi_msg(data->msi_desc, &msg); @@ -3209,9 +3016,8 @@ msi_set_affinity(struct irq_data *data, const struct cpumask *mask, bool force) __write_msi_msg(data->msi_desc, &msg); - return 0; + return IRQ_SET_MASK_OK_NOCOPY; } -#endif /* CONFIG_SMP */ /* * IRQ Chip for MSI PCI/PCI-X/PCI-Express Devices, @@ -3222,56 +3028,32 @@ static struct irq_chip msi_chip = { .irq_unmask = unmask_msi_irq, .irq_mask = mask_msi_irq, .irq_ack = ack_apic_edge, -#ifdef CONFIG_SMP .irq_set_affinity = msi_set_affinity, -#endif .irq_retrigger = ioapic_retrigger_irq, }; -/* - * Map the PCI dev to the corresponding remapping hardware unit - * and allocate 'nvec' consecutive interrupt-remapping table entries - * in it. - */ -static int msi_alloc_irte(struct pci_dev *dev, int irq, int nvec) -{ - struct intel_iommu *iommu; - int index; - - iommu = map_dev_to_ir(dev); - if (!iommu) { - printk(KERN_ERR - "Unable to map PCI %s to iommu\n", pci_name(dev)); - return -ENOENT; - } - - index = alloc_irte(iommu, irq, nvec); - if (index < 0) { - printk(KERN_ERR - "Unable to allocate %d IRTE for PCI %s\n", nvec, - pci_name(dev)); - return -ENOSPC; - } - return index; -} - -static int setup_msi_irq(struct pci_dev *dev, struct msi_desc *msidesc, int irq) +int setup_msi_irq(struct pci_dev *dev, struct msi_desc *msidesc, + unsigned int irq_base, unsigned int irq_offset) { struct irq_chip *chip = &msi_chip; struct msi_msg msg; + unsigned int irq = irq_base + irq_offset; int ret; ret = msi_compose_msg(dev, irq, &msg, -1); if (ret < 0) return ret; - irq_set_msi_desc(irq, msidesc); - write_msi_msg(irq, &msg); + irq_set_msi_desc_off(irq_base, irq_offset, msidesc); - if (irq_remapped(irq_get_chip_data(irq))) { - irq_set_status_flags(irq, IRQ_MOVE_PCNTXT); - irq_remap_modify_chip_defaults(chip); - } + /* + * MSI-X message is written per-IRQ, the offset is always 0. + * MSI message denotes a contiguous group of IRQs, written for 0th IRQ. + */ + if (!irq_offset) + write_msi_msg(irq, &msg); + + setup_remapped_irq(irq, irq_get_chip_data(irq), chip); irq_set_chip_and_handler_name(irq, chip, handle_edge_irq, "edge"); @@ -3282,69 +3064,37 @@ static int setup_msi_irq(struct pci_dev *dev, struct msi_desc *msidesc, int irq) int native_setup_msi_irqs(struct pci_dev *dev, int nvec, int type) { - int node, ret, sub_handle, index = 0; - unsigned int irq, irq_want; struct msi_desc *msidesc; - struct intel_iommu *iommu = NULL; + unsigned int irq; + int node, ret; - /* x86 doesn't support multiple MSI yet */ + /* Multiple MSI vectors only supported with interrupt remapping */ if (type == PCI_CAP_ID_MSI && nvec > 1) return 1; node = dev_to_node(&dev->dev); - irq_want = nr_irqs_gsi; - sub_handle = 0; - list_for_each_entry(msidesc, &dev->msi_list, list) { - irq = create_irq_nr(irq_want, node); - if (irq == 0) - return -1; - irq_want = irq + 1; - if (!intr_remapping_enabled) - goto no_ir; - if (!sub_handle) { - /* - * allocate the consecutive block of IRTE's - * for 'nvec' - */ - index = msi_alloc_irte(dev, irq, nvec); - if (index < 0) { - ret = index; - goto error; - } - } else { - iommu = map_dev_to_ir(dev); - if (!iommu) { - ret = -ENOENT; - goto error; - } - /* - * setup the mapping between the irq and the IRTE - * base index, the sub_handle pointing to the - * appropriate interrupt remap table entry. - */ - set_irte_irq(irq, iommu, index, sub_handle); + list_for_each_entry(msidesc, &dev->msi_list, list) { + irq = irq_alloc_hwirq(node); + if (!irq) + return -ENOSPC; + + ret = setup_msi_irq(dev, msidesc, irq, 0); + if (ret < 0) { + irq_free_hwirq(irq); + return ret; } -no_ir: - ret = setup_msi_irq(dev, msidesc, irq); - if (ret < 0) - goto error; - sub_handle++; + } return 0; - -error: - destroy_irq(irq); - return ret; } void native_teardown_msi_irq(unsigned int irq) { - destroy_irq(irq); + irq_free_hwirq(irq); } #ifdef CONFIG_DMAR_TABLE -#ifdef CONFIG_SMP static int dmar_msi_set_affinity(struct irq_data *data, const struct cpumask *mask, bool force) @@ -3352,9 +3102,11 @@ dmar_msi_set_affinity(struct irq_data *data, const struct cpumask *mask, struct irq_cfg *cfg = data->chip_data; unsigned int dest, irq = data->irq; struct msi_msg msg; + int ret; - if (__ioapic_set_affinity(data, mask, &dest)) - return -1; + ret = __ioapic_set_affinity(data, mask, &dest); + if (ret) + return ret; dmar_msi_read(irq, &msg); @@ -3366,19 +3118,15 @@ dmar_msi_set_affinity(struct irq_data *data, const struct cpumask *mask, dmar_msi_write(irq, &msg); - return 0; + return IRQ_SET_MASK_OK_NOCOPY; } -#endif /* CONFIG_SMP */ - static struct irq_chip dmar_msi_type = { .name = "DMAR_MSI", .irq_unmask = dmar_msi_unmask, .irq_mask = dmar_msi_mask, .irq_ack = ack_apic_edge, -#ifdef CONFIG_SMP .irq_set_affinity = dmar_msi_set_affinity, -#endif .irq_retrigger = ioapic_retrigger_irq, }; @@ -3399,16 +3147,17 @@ int arch_setup_dmar_msi(unsigned int irq) #ifdef CONFIG_HPET_TIMER -#ifdef CONFIG_SMP static int hpet_msi_set_affinity(struct irq_data *data, const struct cpumask *mask, bool force) { struct irq_cfg *cfg = data->chip_data; struct msi_msg msg; unsigned int dest; + int ret; - if (__ioapic_set_affinity(data, mask, &dest)) - return -1; + ret = __ioapic_set_affinity(data, mask, &dest); + if (ret) + return ret; hpet_msi_read(data->handler_data, &msg); @@ -3419,48 +3168,31 @@ static int hpet_msi_set_affinity(struct irq_data *data, hpet_msi_write(data->handler_data, &msg); - return 0; + return IRQ_SET_MASK_OK_NOCOPY; } -#endif /* CONFIG_SMP */ - static struct irq_chip hpet_msi_type = { .name = "HPET_MSI", .irq_unmask = hpet_msi_unmask, .irq_mask = hpet_msi_mask, .irq_ack = ack_apic_edge, -#ifdef CONFIG_SMP .irq_set_affinity = hpet_msi_set_affinity, -#endif .irq_retrigger = ioapic_retrigger_irq, }; -int arch_setup_hpet_msi(unsigned int irq, unsigned int id) +int default_setup_hpet_msi(unsigned int irq, unsigned int id) { struct irq_chip *chip = &hpet_msi_type; struct msi_msg msg; int ret; - if (intr_remapping_enabled) { - struct intel_iommu *iommu = map_hpet_to_ir(id); - int index; - - if (!iommu) - return -1; - - index = alloc_irte(iommu, irq, 1); - if (index < 0) - return -1; - } - ret = msi_compose_msg(NULL, irq, &msg, id); if (ret < 0) return ret; hpet_msi_write(irq_get_handler_data(irq), &msg); irq_set_status_flags(irq, IRQ_MOVE_PCNTXT); - if (irq_remapped(irq_get_chip_data(irq))) - irq_remap_modify_chip_defaults(chip); + setup_remapped_irq(irq, irq_get_chip_data(irq), chip); irq_set_chip_and_handler_name(irq, chip, handle_edge_irq, "edge"); return 0; @@ -3473,8 +3205,6 @@ int arch_setup_hpet_msi(unsigned int irq, unsigned int id) */ #ifdef CONFIG_HT_IRQ -#ifdef CONFIG_SMP - static void target_ht_irq(unsigned int irq, unsigned int dest, u8 vector) { struct ht_irq_msg msg; @@ -3494,30 +3224,30 @@ ht_set_affinity(struct irq_data *data, const struct cpumask *mask, bool force) { struct irq_cfg *cfg = data->chip_data; unsigned int dest; + int ret; - if (__ioapic_set_affinity(data, mask, &dest)) - return -1; + ret = __ioapic_set_affinity(data, mask, &dest); + if (ret) + return ret; target_ht_irq(data->irq, dest, cfg->vector); - return 0; + return IRQ_SET_MASK_OK_NOCOPY; } -#endif - static struct irq_chip ht_irq_chip = { .name = "PCI-HT", .irq_mask = mask_ht_irq, .irq_unmask = unmask_ht_irq, .irq_ack = ack_apic_edge, -#ifdef CONFIG_SMP .irq_set_affinity = ht_set_affinity, -#endif .irq_retrigger = ioapic_retrigger_irq, }; int arch_setup_ht_irq(unsigned int irq, struct pci_dev *dev) { struct irq_cfg *cfg; + struct ht_irq_msg msg; + unsigned dest; int err; if (disable_apic) @@ -3525,36 +3255,37 @@ int arch_setup_ht_irq(unsigned int irq, struct pci_dev *dev) cfg = irq_cfg(irq); err = assign_irq_vector(irq, cfg, apic->target_cpus()); - if (!err) { - struct ht_irq_msg msg; - unsigned dest; + if (err) + return err; - dest = apic->cpu_mask_to_apicid_and(cfg->domain, - apic->target_cpus()); + err = apic->cpu_mask_to_apicid_and(cfg->domain, + apic->target_cpus(), &dest); + if (err) + return err; - msg.address_hi = HT_IRQ_HIGH_DEST_ID(dest); + msg.address_hi = HT_IRQ_HIGH_DEST_ID(dest); + + msg.address_lo = + HT_IRQ_LOW_BASE | + HT_IRQ_LOW_DEST_ID(dest) | + HT_IRQ_LOW_VECTOR(cfg->vector) | + ((apic->irq_dest_mode == 0) ? + HT_IRQ_LOW_DM_PHYSICAL : + HT_IRQ_LOW_DM_LOGICAL) | + HT_IRQ_LOW_RQEOI_EDGE | + ((apic->irq_delivery_mode != dest_LowestPrio) ? + HT_IRQ_LOW_MT_FIXED : + HT_IRQ_LOW_MT_ARBITRATED) | + HT_IRQ_LOW_IRQ_MASKED; - msg.address_lo = - HT_IRQ_LOW_BASE | - HT_IRQ_LOW_DEST_ID(dest) | - HT_IRQ_LOW_VECTOR(cfg->vector) | - ((apic->irq_dest_mode == 0) ? - HT_IRQ_LOW_DM_PHYSICAL : - HT_IRQ_LOW_DM_LOGICAL) | - HT_IRQ_LOW_RQEOI_EDGE | - ((apic->irq_delivery_mode != dest_LowestPrio) ? - HT_IRQ_LOW_MT_FIXED : - HT_IRQ_LOW_MT_ARBITRATED) | - HT_IRQ_LOW_IRQ_MASKED; + write_ht_irq_msg(irq, &msg); - write_ht_irq_msg(irq, &msg); + irq_set_chip_and_handler_name(irq, &ht_irq_chip, + handle_edge_irq, "edge"); - irq_set_chip_and_handler_name(irq, &ht_irq_chip, - handle_edge_irq, "edge"); + dev_printk(KERN_DEBUG, &dev->dev, "irq %d for HT\n", irq); - dev_printk(KERN_DEBUG, &dev->dev, "irq %d for HT\n", irq); - } - return err; + return 0; } #endif /* CONFIG_HT_IRQ */ @@ -3577,12 +3308,15 @@ int io_apic_setup_irq_pin_once(unsigned int irq, int node, { unsigned int ioapic_idx = attr->ioapic, pin = attr->ioapic_pin; int ret; + struct IO_APIC_route_entry orig_entry; /* Avoid redundant programming */ if (test_bit(pin, ioapics[ioapic_idx].pin_programmed)) { - pr_debug("Pin %d-%d already programmed\n", - mpc_ioapic_id(ioapic_idx), pin); - return 0; + pr_debug("Pin %d-%d already programmed\n", mpc_ioapic_id(ioapic_idx), pin); + orig_entry = ioapic_read_entry(attr->ioapic, pin); + if (attr->trigger == orig_entry.trigger && attr->polarity == orig_entry.polarity) + return 0; + return -EBUSY; } ret = io_apic_setup_irq_pin(irq, node, attr); if (!ret) @@ -3617,9 +3351,9 @@ static void __init probe_nr_irqs_gsi(void) printk(KERN_DEBUG "nr_irqs_gsi: %d\n", nr_irqs_gsi); } -int get_nr_irqs_gsi(void) +unsigned int arch_dynirq_lower_bound(unsigned int from) { - return nr_irqs_gsi; + return from < nr_irqs_gsi ? nr_irqs_gsi : from; } int __init arch_probe_nr_irqs(void) @@ -3722,7 +3456,8 @@ static int __init io_apic_get_unique_id(int ioapic, int apic_id) /* Sanity check */ if (reg_00.bits.ID != apic_id) { - printk("IOAPIC[%d]: Unable to change apic_id!\n", ioapic); + pr_err("IOAPIC[%d]: Unable to change apic_id!\n", + ioapic); return -1; } } @@ -3828,10 +3563,7 @@ void __init setup_ioapic_dest(void) else mask = apic->target_cpus(); - if (intr_remapping_enabled) - ir_ioapic_set_affinity(idata, mask, false); - else - ioapic_set_affinity(idata, mask, false); + x86_io_apic_ops.set_affinity(idata, mask, false); } } @@ -3871,7 +3603,7 @@ static struct resource * __init ioapic_setup_resources(int nr_ioapics) return res; } -void __init ioapic_and_gsi_init(void) +void __init native_io_apic_init_mappings(void) { unsigned long ioapic_phys, idx = FIX_IO_APIC_BASE_0; struct resource *ioapic_res; @@ -3967,15 +3699,33 @@ int mp_find_ioapic_pin(int ioapic, u32 gsi) static __init int bad_ioapic(unsigned long address) { if (nr_ioapics >= MAX_IO_APICS) { - printk(KERN_WARNING "WARNING: Max # of I/O APICs (%d) exceeded " - "(found %d), skipping\n", MAX_IO_APICS, nr_ioapics); + pr_warn("WARNING: Max # of I/O APICs (%d) exceeded (found %d), skipping\n", + MAX_IO_APICS, nr_ioapics); return 1; } if (!address) { - printk(KERN_WARNING "WARNING: Bogus (zero) I/O APIC address" - " found in table, skipping!\n"); + pr_warn("WARNING: Bogus (zero) I/O APIC address found in table, skipping!\n"); + return 1; + } + return 0; +} + +static __init int bad_ioapic_register(int idx) +{ + union IO_APIC_reg_00 reg_00; + union IO_APIC_reg_01 reg_01; + union IO_APIC_reg_02 reg_02; + + reg_00.raw = io_apic_read(idx, 0); + reg_01.raw = io_apic_read(idx, 1); + reg_02.raw = io_apic_read(idx, 2); + + if (reg_00.raw == -1 && reg_01.raw == -1 && reg_02.raw == -1) { + pr_warn("I/O APIC 0x%x registers return all ones, skipping!\n", + mpc_ioapic_addr(idx)); return 1; } + return 0; } @@ -3995,6 +3745,12 @@ void __init mp_register_ioapic(int id, u32 address, u32 gsi_base) ioapics[idx].mp_config.apicaddr = address; set_fixmap_nocache(FIX_IO_APIC_BASE_0 + idx, address); + + if (bad_ioapic_register(idx)) { + clear_fixmap(FIX_IO_APIC_BASE_0 + idx); + return; + } + ioapics[idx].mp_config.apicid = io_apic_unique_id(id); ioapics[idx].mp_config.apicver = io_apic_get_version(idx); @@ -4015,10 +3771,10 @@ void __init mp_register_ioapic(int id, u32 address, u32 gsi_base) if (gsi_cfg->gsi_end >= gsi_top) gsi_top = gsi_cfg->gsi_end + 1; - printk(KERN_INFO "IOAPIC[%d]: apic_id %d, version %d, address 0x%x, " - "GSI %d-%d\n", idx, mpc_ioapic_id(idx), - mpc_ioapic_ver(idx), mpc_ioapic_addr(idx), - gsi_cfg->gsi_base, gsi_cfg->gsi_end); + pr_info("IOAPIC[%d]: apic_id %d, version %d, address 0x%x, GSI %d-%d\n", + idx, mpc_ioapic_id(idx), + mpc_ioapic_ver(idx), mpc_ioapic_addr(idx), + gsi_cfg->gsi_base, gsi_cfg->gsi_end); nr_ioapics++; } diff --git a/arch/x86/kernel/apic/ipi.c b/arch/x86/kernel/apic/ipi.c index cce91bf2667..62071569bd5 100644 --- a/arch/x86/kernel/apic/ipi.c +++ b/arch/x86/kernel/apic/ipi.c @@ -1,6 +1,5 @@ #include <linux/cpumask.h> #include <linux/interrupt.h> -#include <linux/init.h> #include <linux/mm.h> #include <linux/delay.h> @@ -106,7 +105,7 @@ void default_send_IPI_mask_logical(const struct cpumask *cpumask, int vector) unsigned long mask = cpumask_bits(cpumask)[0]; unsigned long flags; - if (WARN_ONCE(!mask, "empty IPI mask")) + if (!mask) return; local_irq_save(flags); diff --git a/arch/x86/kernel/apic/numaq_32.c b/arch/x86/kernel/apic/numaq_32.c deleted file mode 100644 index c4a61ca1349..00000000000 --- a/arch/x86/kernel/apic/numaq_32.c +++ /dev/null @@ -1,541 +0,0 @@ -/* - * Written by: Patricia Gaughen, IBM Corporation - * - * Copyright (C) 2002, IBM Corp. - * Copyright (C) 2009, Red Hat, Inc., Ingo Molnar - * - * All rights reserved. - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, but - * WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or - * NON INFRINGEMENT. See the GNU General Public License for more - * details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. - * - * Send feedback to <gone@us.ibm.com> - */ -#include <linux/nodemask.h> -#include <linux/topology.h> -#include <linux/bootmem.h> -#include <linux/memblock.h> -#include <linux/threads.h> -#include <linux/cpumask.h> -#include <linux/kernel.h> -#include <linux/mmzone.h> -#include <linux/module.h> -#include <linux/string.h> -#include <linux/init.h> -#include <linux/numa.h> -#include <linux/smp.h> -#include <linux/io.h> -#include <linux/mm.h> - -#include <asm/processor.h> -#include <asm/fixmap.h> -#include <asm/mpspec.h> -#include <asm/numaq.h> -#include <asm/setup.h> -#include <asm/apic.h> -#include <asm/e820.h> -#include <asm/ipi.h> - -int found_numaq; - -/* - * Have to match translation table entries to main table entries by counter - * hence the mpc_record variable .... can't see a less disgusting way of - * doing this .... - */ -struct mpc_trans { - unsigned char mpc_type; - unsigned char trans_len; - unsigned char trans_type; - unsigned char trans_quad; - unsigned char trans_global; - unsigned char trans_local; - unsigned short trans_reserved; -}; - -static int mpc_record; - -static struct mpc_trans *translation_table[MAX_MPC_ENTRY]; - -int mp_bus_id_to_node[MAX_MP_BUSSES]; -int mp_bus_id_to_local[MAX_MP_BUSSES]; -int quad_local_to_mp_bus_id[NR_CPUS/4][4]; - - -static inline void numaq_register_node(int node, struct sys_cfg_data *scd) -{ - struct eachquadmem *eq = scd->eq + node; - u64 start = (u64)(eq->hi_shrd_mem_start - eq->priv_mem_size) << 20; - u64 end = (u64)(eq->hi_shrd_mem_start + eq->hi_shrd_mem_size) << 20; - int ret; - - node_set(node, numa_nodes_parsed); - ret = numa_add_memblk(node, start, end); - BUG_ON(ret < 0); -} - -/* - * Function: smp_dump_qct() - * - * Description: gets memory layout from the quad config table. This - * function also updates numa_nodes_parsed with the nodes (quads) present. - */ -static void __init smp_dump_qct(void) -{ - struct sys_cfg_data *scd; - int node; - - scd = (void *)__va(SYS_CFG_DATA_PRIV_ADDR); - - for_each_node(node) { - if (scd->quads_present31_0 & (1 << node)) - numaq_register_node(node, scd); - } -} - -void __cpuinit numaq_tsc_disable(void) -{ - if (!found_numaq) - return; - - if (num_online_nodes() > 1) { - printk(KERN_DEBUG "NUMAQ: disabling TSC\n"); - setup_clear_cpu_cap(X86_FEATURE_TSC); - } -} - -static void __init numaq_tsc_init(void) -{ - numaq_tsc_disable(); -} - -static inline int generate_logical_apicid(int quad, int phys_apicid) -{ - return (quad << 4) + (phys_apicid ? phys_apicid << 1 : 1); -} - -/* x86_quirks member */ -static int mpc_apic_id(struct mpc_cpu *m) -{ - int quad = translation_table[mpc_record]->trans_quad; - int logical_apicid = generate_logical_apicid(quad, m->apicid); - - printk(KERN_DEBUG - "Processor #%d %u:%u APIC version %d (quad %d, apic %d)\n", - m->apicid, (m->cpufeature & CPU_FAMILY_MASK) >> 8, - (m->cpufeature & CPU_MODEL_MASK) >> 4, - m->apicver, quad, logical_apicid); - - return logical_apicid; -} - -/* x86_quirks member */ -static void mpc_oem_bus_info(struct mpc_bus *m, char *name) -{ - int quad = translation_table[mpc_record]->trans_quad; - int local = translation_table[mpc_record]->trans_local; - - mp_bus_id_to_node[m->busid] = quad; - mp_bus_id_to_local[m->busid] = local; - - printk(KERN_INFO "Bus #%d is %s (node %d)\n", m->busid, name, quad); -} - -/* x86_quirks member */ -static void mpc_oem_pci_bus(struct mpc_bus *m) -{ - int quad = translation_table[mpc_record]->trans_quad; - int local = translation_table[mpc_record]->trans_local; - - quad_local_to_mp_bus_id[quad][local] = m->busid; -} - -/* - * Called from mpparse code. - * mode = 0: prescan - * mode = 1: one mpc entry scanned - */ -static void numaq_mpc_record(unsigned int mode) -{ - if (!mode) - mpc_record = 0; - else - mpc_record++; -} - -static void __init MP_translation_info(struct mpc_trans *m) -{ - printk(KERN_INFO - "Translation: record %d, type %d, quad %d, global %d, local %d\n", - mpc_record, m->trans_type, m->trans_quad, m->trans_global, - m->trans_local); - - if (mpc_record >= MAX_MPC_ENTRY) - printk(KERN_ERR "MAX_MPC_ENTRY exceeded!\n"); - else - translation_table[mpc_record] = m; /* stash this for later */ - - if (m->trans_quad < MAX_NUMNODES && !node_online(m->trans_quad)) - node_set_online(m->trans_quad); -} - -static int __init mpf_checksum(unsigned char *mp, int len) -{ - int sum = 0; - - while (len--) - sum += *mp++; - - return sum & 0xFF; -} - -/* - * Read/parse the MPC oem tables - */ -static void __init smp_read_mpc_oem(struct mpc_table *mpc) -{ - struct mpc_oemtable *oemtable = (void *)(long)mpc->oemptr; - int count = sizeof(*oemtable); /* the header size */ - unsigned char *oemptr = ((unsigned char *)oemtable) + count; - - mpc_record = 0; - printk(KERN_INFO - "Found an OEM MPC table at %8p - parsing it...\n", oemtable); - - if (memcmp(oemtable->signature, MPC_OEM_SIGNATURE, 4)) { - printk(KERN_WARNING - "SMP mpc oemtable: bad signature [%c%c%c%c]!\n", - oemtable->signature[0], oemtable->signature[1], - oemtable->signature[2], oemtable->signature[3]); - return; - } - - if (mpf_checksum((unsigned char *)oemtable, oemtable->length)) { - printk(KERN_WARNING "SMP oem mptable: checksum error!\n"); - return; - } - - while (count < oemtable->length) { - switch (*oemptr) { - case MP_TRANSLATION: - { - struct mpc_trans *m = (void *)oemptr; - - MP_translation_info(m); - oemptr += sizeof(*m); - count += sizeof(*m); - ++mpc_record; - break; - } - default: - printk(KERN_WARNING - "Unrecognised OEM table entry type! - %d\n", - (int)*oemptr); - return; - } - } -} - -static __init void early_check_numaq(void) -{ - /* - * get boot-time SMP configuration: - */ - if (smp_found_config) - early_get_smp_config(); - - if (found_numaq) { - x86_init.mpparse.mpc_record = numaq_mpc_record; - x86_init.mpparse.setup_ioapic_ids = x86_init_noop; - x86_init.mpparse.mpc_apic_id = mpc_apic_id; - x86_init.mpparse.smp_read_mpc_oem = smp_read_mpc_oem; - x86_init.mpparse.mpc_oem_pci_bus = mpc_oem_pci_bus; - x86_init.mpparse.mpc_oem_bus_info = mpc_oem_bus_info; - x86_init.timers.tsc_pre_init = numaq_tsc_init; - x86_init.pci.init = pci_numaq_init; - } -} - -int __init numaq_numa_init(void) -{ - early_check_numaq(); - if (!found_numaq) - return -ENOENT; - smp_dump_qct(); - - return 0; -} - -#define NUMAQ_APIC_DFR_VALUE (APIC_DFR_CLUSTER) - -static inline unsigned int numaq_get_apic_id(unsigned long x) -{ - return (x >> 24) & 0x0F; -} - -static inline void numaq_send_IPI_mask(const struct cpumask *mask, int vector) -{ - default_send_IPI_mask_sequence_logical(mask, vector); -} - -static inline void numaq_send_IPI_allbutself(int vector) -{ - default_send_IPI_mask_allbutself_logical(cpu_online_mask, vector); -} - -static inline void numaq_send_IPI_all(int vector) -{ - numaq_send_IPI_mask(cpu_online_mask, vector); -} - -#define NUMAQ_TRAMPOLINE_PHYS_LOW (0x8) -#define NUMAQ_TRAMPOLINE_PHYS_HIGH (0xa) - -/* - * Because we use NMIs rather than the INIT-STARTUP sequence to - * bootstrap the CPUs, the APIC may be in a weird state. Kick it: - */ -static inline void numaq_smp_callin_clear_local_apic(void) -{ - clear_local_APIC(); -} - -static inline const struct cpumask *numaq_target_cpus(void) -{ - return cpu_all_mask; -} - -static unsigned long numaq_check_apicid_used(physid_mask_t *map, int apicid) -{ - return physid_isset(apicid, *map); -} - -static inline unsigned long numaq_check_apicid_present(int bit) -{ - return physid_isset(bit, phys_cpu_present_map); -} - -static inline int numaq_apic_id_registered(void) -{ - return 1; -} - -static inline void numaq_init_apic_ldr(void) -{ - /* Already done in NUMA-Q firmware */ -} - -static inline void numaq_setup_apic_routing(void) -{ - printk(KERN_INFO - "Enabling APIC mode: NUMA-Q. Using %d I/O APICs\n", - nr_ioapics); -} - -/* - * Skip adding the timer int on secondary nodes, which causes - * a small but painful rift in the time-space continuum. - */ -static inline int numaq_multi_timer_check(int apic, int irq) -{ - return apic != 0 && irq == 0; -} - -static inline void numaq_ioapic_phys_id_map(physid_mask_t *phys_map, physid_mask_t *retmap) -{ - /* We don't have a good way to do this yet - hack */ - return physids_promote(0xFUL, retmap); -} - -/* - * Supporting over 60 cpus on NUMA-Q requires a locality-dependent - * cpu to APIC ID relation to properly interact with the intelligent - * mode of the cluster controller. - */ -static inline int numaq_cpu_present_to_apicid(int mps_cpu) -{ - if (mps_cpu < 60) - return ((mps_cpu >> 2) << 4) | (1 << (mps_cpu & 0x3)); - else - return BAD_APICID; -} - -static inline int numaq_apicid_to_node(int logical_apicid) -{ - return logical_apicid >> 4; -} - -static int numaq_numa_cpu_node(int cpu) -{ - int logical_apicid = early_per_cpu(x86_cpu_to_logical_apicid, cpu); - - if (logical_apicid != BAD_APICID) - return numaq_apicid_to_node(logical_apicid); - return NUMA_NO_NODE; -} - -static void numaq_apicid_to_cpu_present(int logical_apicid, physid_mask_t *retmap) -{ - int node = numaq_apicid_to_node(logical_apicid); - int cpu = __ffs(logical_apicid & 0xf); - - physid_set_mask_of_physid(cpu + 4*node, retmap); -} - -/* Where the IO area was mapped on multiquad, always 0 otherwise */ -void *xquad_portio; - -static inline int numaq_check_phys_apicid_present(int phys_apicid) -{ - return 1; -} - -/* - * We use physical apicids here, not logical, so just return the default - * physical broadcast to stop people from breaking us - */ -static unsigned int numaq_cpu_mask_to_apicid(const struct cpumask *cpumask) -{ - return 0x0F; -} - -static inline unsigned int -numaq_cpu_mask_to_apicid_and(const struct cpumask *cpumask, - const struct cpumask *andmask) -{ - return 0x0F; -} - -/* No NUMA-Q box has a HT CPU, but it can't hurt to use the default code. */ -static inline int numaq_phys_pkg_id(int cpuid_apic, int index_msb) -{ - return cpuid_apic >> index_msb; -} - -static int -numaq_mps_oem_check(struct mpc_table *mpc, char *oem, char *productid) -{ - if (strncmp(oem, "IBM NUMA", 8)) - printk(KERN_ERR "Warning! Not a NUMA-Q system!\n"); - else - found_numaq = 1; - - return found_numaq; -} - -static int probe_numaq(void) -{ - /* already know from get_memcfg_numaq() */ - return found_numaq; -} - -static void numaq_vector_allocation_domain(int cpu, struct cpumask *retmask) -{ - /* Careful. Some cpus do not strictly honor the set of cpus - * specified in the interrupt destination when using lowest - * priority interrupt delivery mode. - * - * In particular there was a hyperthreading cpu observed to - * deliver interrupts to the wrong hyperthread when only one - * hyperthread was specified in the interrupt desitination. - */ - cpumask_clear(retmask); - cpumask_bits(retmask)[0] = APIC_ALL_CPUS; -} - -static void numaq_setup_portio_remap(void) -{ - int num_quads = num_online_nodes(); - - if (num_quads <= 1) - return; - - printk(KERN_INFO - "Remapping cross-quad port I/O for %d quads\n", num_quads); - - xquad_portio = ioremap(XQUAD_PORTIO_BASE, num_quads*XQUAD_PORTIO_QUAD); - - printk(KERN_INFO - "xquad_portio vaddr 0x%08lx, len %08lx\n", - (u_long) xquad_portio, (u_long) num_quads*XQUAD_PORTIO_QUAD); -} - -/* Use __refdata to keep false positive warning calm. */ -static struct apic __refdata apic_numaq = { - - .name = "NUMAQ", - .probe = probe_numaq, - .acpi_madt_oem_check = NULL, - .apic_id_registered = numaq_apic_id_registered, - - .irq_delivery_mode = dest_LowestPrio, - /* physical delivery on LOCAL quad: */ - .irq_dest_mode = 0, - - .target_cpus = numaq_target_cpus, - .disable_esr = 1, - .dest_logical = APIC_DEST_LOGICAL, - .check_apicid_used = numaq_check_apicid_used, - .check_apicid_present = numaq_check_apicid_present, - - .vector_allocation_domain = numaq_vector_allocation_domain, - .init_apic_ldr = numaq_init_apic_ldr, - - .ioapic_phys_id_map = numaq_ioapic_phys_id_map, - .setup_apic_routing = numaq_setup_apic_routing, - .multi_timer_check = numaq_multi_timer_check, - .cpu_present_to_apicid = numaq_cpu_present_to_apicid, - .apicid_to_cpu_present = numaq_apicid_to_cpu_present, - .setup_portio_remap = numaq_setup_portio_remap, - .check_phys_apicid_present = numaq_check_phys_apicid_present, - .enable_apic_mode = NULL, - .phys_pkg_id = numaq_phys_pkg_id, - .mps_oem_check = numaq_mps_oem_check, - - .get_apic_id = numaq_get_apic_id, - .set_apic_id = NULL, - .apic_id_mask = 0x0F << 24, - - .cpu_mask_to_apicid = numaq_cpu_mask_to_apicid, - .cpu_mask_to_apicid_and = numaq_cpu_mask_to_apicid_and, - - .send_IPI_mask = numaq_send_IPI_mask, - .send_IPI_mask_allbutself = NULL, - .send_IPI_allbutself = numaq_send_IPI_allbutself, - .send_IPI_all = numaq_send_IPI_all, - .send_IPI_self = default_send_IPI_self, - - .wakeup_secondary_cpu = wakeup_secondary_cpu_via_nmi, - .trampoline_phys_low = NUMAQ_TRAMPOLINE_PHYS_LOW, - .trampoline_phys_high = NUMAQ_TRAMPOLINE_PHYS_HIGH, - - /* We don't do anything here because we use NMI's to boot instead */ - .wait_for_init_deassert = NULL, - - .smp_callin_clear_local_apic = numaq_smp_callin_clear_local_apic, - .inquire_remote_apic = NULL, - - .read = native_apic_mem_read, - .write = native_apic_mem_write, - .icr_read = native_apic_icr_read, - .icr_write = native_apic_icr_write, - .wait_icr_idle = native_apic_wait_icr_idle, - .safe_wait_icr_idle = native_safe_apic_wait_icr_idle, - - .x86_32_early_logical_apicid = noop_x86_32_early_logical_apicid, - .x86_32_numa_cpu_node = numaq_numa_cpu_node, -}; - -apic_driver(apic_numaq); diff --git a/arch/x86/kernel/apic/probe_32.c b/arch/x86/kernel/apic/probe_32.c index 0787bb3412f..cceb352c968 100644 --- a/arch/x86/kernel/apic/probe_32.c +++ b/arch/x86/kernel/apic/probe_32.c @@ -66,21 +66,6 @@ static void setup_apic_flat_routing(void) #endif } -static void default_vector_allocation_domain(int cpu, struct cpumask *retmask) -{ - /* - * Careful. Some cpus do not strictly honor the set of cpus - * specified in the interrupt destination when using lowest - * priority interrupt delivery mode. - * - * In particular there was a hyperthreading cpu observed to - * deliver interrupts to the wrong hyperthread when only one - * hyperthread was specified in the interrupt desitination. - */ - cpumask_clear(retmask); - cpumask_bits(retmask)[0] = APIC_ALL_CPUS; -} - /* should be called last. */ static int probe_default(void) { @@ -92,6 +77,7 @@ static struct apic apic_default = { .name = "default", .probe = probe_default, .acpi_madt_oem_check = NULL, + .apic_id_valid = default_apic_id_valid, .apic_id_registered = default_apic_id_registered, .irq_delivery_mode = dest_LowestPrio, @@ -104,7 +90,7 @@ static struct apic apic_default = { .check_apicid_used = default_check_apicid_used, .check_apicid_present = default_check_apicid_present, - .vector_allocation_domain = default_vector_allocation_domain, + .vector_allocation_domain = flat_vector_allocation_domain, .init_apic_ldr = default_init_apic_ldr, .ioapic_phys_id_map = default_ioapic_phys_id_map, @@ -122,8 +108,7 @@ static struct apic apic_default = { .set_apic_id = NULL, .apic_id_mask = 0x0F << 24, - .cpu_mask_to_apicid = default_cpu_mask_to_apicid, - .cpu_mask_to_apicid_and = default_cpu_mask_to_apicid_and, + .cpu_mask_to_apicid_and = flat_cpu_mask_to_apicid_and, .send_IPI_mask = default_send_IPI_mask_logical, .send_IPI_mask_allbutself = default_send_IPI_mask_allbutself_logical, @@ -134,13 +119,13 @@ static struct apic apic_default = { .trampoline_phys_low = DEFAULT_TRAMPOLINE_PHYS_LOW, .trampoline_phys_high = DEFAULT_TRAMPOLINE_PHYS_HIGH, - .wait_for_init_deassert = default_wait_for_init_deassert, - + .wait_for_init_deassert = true, .smp_callin_clear_local_apic = NULL, .inquire_remote_apic = default_inquire_remote_apic, .read = native_apic_mem_read, .write = native_apic_mem_write, + .eoi_write = native_apic_mem_write, .icr_read = native_apic_icr_read, .icr_write = native_apic_icr_write, .wait_icr_idle = native_apic_wait_icr_idle, @@ -206,6 +191,9 @@ void __init default_setup_apic_routing(void) if (apic->setup_apic_routing) apic->setup_apic_routing(); + + if (x86_platform.apic_post_init) + x86_platform.apic_post_init(); } void __init generic_apic_probe(void) diff --git a/arch/x86/kernel/apic/probe_64.c b/arch/x86/kernel/apic/probe_64.c index 3fe98669892..1793dba7a74 100644 --- a/arch/x86/kernel/apic/probe_64.c +++ b/arch/x86/kernel/apic/probe_64.c @@ -23,11 +23,6 @@ #include <asm/ipi.h> #include <asm/setup.h> -static int apicid_phys_pkg_id(int initial_apic_id, int index_msb) -{ - return hard_smp_processor_id() >> index_msb; -} - /* * Check the APIC IDs in bios_cpu_apicid and choose the APIC mode. */ @@ -48,10 +43,8 @@ void __init default_setup_apic_routing(void) } } - if (is_vsmp_box()) { - /* need to update phys_pkg_id */ - apic->phys_pkg_id = apicid_phys_pkg_id; - } + if (x86_platform.apic_post_init) + x86_platform.apic_post_init(); } /* Same for both flat and physical. */ diff --git a/arch/x86/kernel/apic/summit_32.c b/arch/x86/kernel/apic/summit_32.c deleted file mode 100644 index 19114423c58..00000000000 --- a/arch/x86/kernel/apic/summit_32.c +++ /dev/null @@ -1,556 +0,0 @@ -/* - * IBM Summit-Specific Code - * - * Written By: Matthew Dobson, IBM Corporation - * - * Copyright (c) 2003 IBM Corp. - * - * All rights reserved. - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or (at - * your option) any later version. - * - * This program is distributed in the hope that it will be useful, but - * WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or - * NON INFRINGEMENT. See the GNU General Public License for more - * details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. - * - * Send feedback to <colpatch@us.ibm.com> - * - */ - -#include <linux/mm.h> -#include <linux/init.h> -#include <asm/io.h> -#include <asm/bios_ebda.h> - -/* - * APIC driver for the IBM "Summit" chipset. - */ -#include <linux/threads.h> -#include <linux/cpumask.h> -#include <asm/mpspec.h> -#include <asm/apic.h> -#include <asm/smp.h> -#include <asm/fixmap.h> -#include <asm/apicdef.h> -#include <asm/ipi.h> -#include <linux/kernel.h> -#include <linux/string.h> -#include <linux/gfp.h> -#include <linux/smp.h> - -static unsigned summit_get_apic_id(unsigned long x) -{ - return (x >> 24) & 0xFF; -} - -static inline void summit_send_IPI_mask(const struct cpumask *mask, int vector) -{ - default_send_IPI_mask_sequence_logical(mask, vector); -} - -static void summit_send_IPI_allbutself(int vector) -{ - default_send_IPI_mask_allbutself_logical(cpu_online_mask, vector); -} - -static void summit_send_IPI_all(int vector) -{ - summit_send_IPI_mask(cpu_online_mask, vector); -} - -#include <asm/tsc.h> - -extern int use_cyclone; - -#ifdef CONFIG_X86_SUMMIT_NUMA -static void setup_summit(void); -#else -static inline void setup_summit(void) {} -#endif - -static int summit_mps_oem_check(struct mpc_table *mpc, char *oem, - char *productid) -{ - if (!strncmp(oem, "IBM ENSW", 8) && - (!strncmp(productid, "VIGIL SMP", 9) - || !strncmp(productid, "EXA", 3) - || !strncmp(productid, "RUTHLESS SMP", 12))){ - mark_tsc_unstable("Summit based system"); - use_cyclone = 1; /*enable cyclone-timer*/ - setup_summit(); - return 1; - } - return 0; -} - -/* Hook from generic ACPI tables.c */ -static int summit_acpi_madt_oem_check(char *oem_id, char *oem_table_id) -{ - if (!strncmp(oem_id, "IBM", 3) && - (!strncmp(oem_table_id, "SERVIGIL", 8) - || !strncmp(oem_table_id, "EXA", 3))){ - mark_tsc_unstable("Summit based system"); - use_cyclone = 1; /*enable cyclone-timer*/ - setup_summit(); - return 1; - } - return 0; -} - -struct rio_table_hdr { - unsigned char version; /* Version number of this data structure */ - /* Version 3 adds chassis_num & WP_index */ - unsigned char num_scal_dev; /* # of Scalability devices (Twisters for Vigil) */ - unsigned char num_rio_dev; /* # of RIO I/O devices (Cyclones and Winnipegs) */ -} __attribute__((packed)); - -struct scal_detail { - unsigned char node_id; /* Scalability Node ID */ - unsigned long CBAR; /* Address of 1MB register space */ - unsigned char port0node; /* Node ID port connected to: 0xFF=None */ - unsigned char port0port; /* Port num port connected to: 0,1,2, or 0xFF=None */ - unsigned char port1node; /* Node ID port connected to: 0xFF = None */ - unsigned char port1port; /* Port num port connected to: 0,1,2, or 0xFF=None */ - unsigned char port2node; /* Node ID port connected to: 0xFF = None */ - unsigned char port2port; /* Port num port connected to: 0,1,2, or 0xFF=None */ - unsigned char chassis_num; /* 1 based Chassis number (1 = boot node) */ -} __attribute__((packed)); - -struct rio_detail { - unsigned char node_id; /* RIO Node ID */ - unsigned long BBAR; /* Address of 1MB register space */ - unsigned char type; /* Type of device */ - unsigned char owner_id; /* For WPEG: Node ID of Cyclone that owns this WPEG*/ - /* For CYC: Node ID of Twister that owns this CYC */ - unsigned char port0node; /* Node ID port connected to: 0xFF=None */ - unsigned char port0port; /* Port num port connected to: 0,1,2, or 0xFF=None */ - unsigned char port1node; /* Node ID port connected to: 0xFF=None */ - unsigned char port1port; /* Port num port connected to: 0,1,2, or 0xFF=None */ - unsigned char first_slot; /* For WPEG: Lowest slot number below this WPEG */ - /* For CYC: 0 */ - unsigned char status; /* For WPEG: Bit 0 = 1 : the XAPIC is used */ - /* = 0 : the XAPIC is not used, ie:*/ - /* ints fwded to another XAPIC */ - /* Bits1:7 Reserved */ - /* For CYC: Bits0:7 Reserved */ - unsigned char WP_index; /* For WPEG: WPEG instance index - lower ones have */ - /* lower slot numbers/PCI bus numbers */ - /* For CYC: No meaning */ - unsigned char chassis_num; /* 1 based Chassis number */ - /* For LookOut WPEGs this field indicates the */ - /* Expansion Chassis #, enumerated from Boot */ - /* Node WPEG external port, then Boot Node CYC */ - /* external port, then Next Vigil chassis WPEG */ - /* external port, etc. */ - /* Shared Lookouts have only 1 chassis number (the */ - /* first one assigned) */ -} __attribute__((packed)); - - -typedef enum { - CompatTwister = 0, /* Compatibility Twister */ - AltTwister = 1, /* Alternate Twister of internal 8-way */ - CompatCyclone = 2, /* Compatibility Cyclone */ - AltCyclone = 3, /* Alternate Cyclone of internal 8-way */ - CompatWPEG = 4, /* Compatibility WPEG */ - AltWPEG = 5, /* Second Planar WPEG */ - LookOutAWPEG = 6, /* LookOut WPEG */ - LookOutBWPEG = 7, /* LookOut WPEG */ -} node_type; - -static inline int is_WPEG(struct rio_detail *rio){ - return (rio->type == CompatWPEG || rio->type == AltWPEG || - rio->type == LookOutAWPEG || rio->type == LookOutBWPEG); -} - -#define SUMMIT_APIC_DFR_VALUE (APIC_DFR_CLUSTER) - -static const struct cpumask *summit_target_cpus(void) -{ - /* CPU_MASK_ALL (0xff) has undefined behaviour with - * dest_LowestPrio mode logical clustered apic interrupt routing - * Just start on cpu 0. IRQ balancing will spread load - */ - return cpumask_of(0); -} - -static unsigned long summit_check_apicid_used(physid_mask_t *map, int apicid) -{ - return 0; -} - -/* we don't use the phys_cpu_present_map to indicate apicid presence */ -static unsigned long summit_check_apicid_present(int bit) -{ - return 1; -} - -static int summit_early_logical_apicid(int cpu) -{ - int count = 0; - u8 my_id = early_per_cpu(x86_cpu_to_apicid, cpu); - u8 my_cluster = APIC_CLUSTER(my_id); -#ifdef CONFIG_SMP - u8 lid; - int i; - - /* Create logical APIC IDs by counting CPUs already in cluster. */ - for (count = 0, i = nr_cpu_ids; --i >= 0; ) { - lid = early_per_cpu(x86_cpu_to_logical_apicid, i); - if (lid != BAD_APICID && APIC_CLUSTER(lid) == my_cluster) - ++count; - } -#endif - /* We only have a 4 wide bitmap in cluster mode. If a deranged - * BIOS puts 5 CPUs in one APIC cluster, we're hosed. */ - BUG_ON(count >= XAPIC_DEST_CPUS_SHIFT); - return my_cluster | (1UL << count); -} - -static void summit_init_apic_ldr(void) -{ - int cpu = smp_processor_id(); - unsigned long id = early_per_cpu(x86_cpu_to_logical_apicid, cpu); - unsigned long val; - - apic_write(APIC_DFR, SUMMIT_APIC_DFR_VALUE); - val = apic_read(APIC_LDR) & ~APIC_LDR_MASK; - val |= SET_APIC_LOGICAL_ID(id); - apic_write(APIC_LDR, val); -} - -static int summit_apic_id_registered(void) -{ - return 1; -} - -static void summit_setup_apic_routing(void) -{ - printk("Enabling APIC mode: Summit. Using %d I/O APICs\n", - nr_ioapics); -} - -static int summit_cpu_present_to_apicid(int mps_cpu) -{ - if (mps_cpu < nr_cpu_ids) - return (int)per_cpu(x86_bios_cpu_apicid, mps_cpu); - else - return BAD_APICID; -} - -static void summit_ioapic_phys_id_map(physid_mask_t *phys_id_map, physid_mask_t *retmap) -{ - /* For clustered we don't have a good way to do this yet - hack */ - physids_promote(0x0FL, retmap); -} - -static void summit_apicid_to_cpu_present(int apicid, physid_mask_t *retmap) -{ - physid_set_mask_of_physid(0, retmap); -} - -static int summit_check_phys_apicid_present(int physical_apicid) -{ - return 1; -} - -static unsigned int summit_cpu_mask_to_apicid(const struct cpumask *cpumask) -{ - unsigned int round = 0; - int cpu, apicid = 0; - - /* - * The cpus in the mask must all be on the apic cluster. - */ - for_each_cpu(cpu, cpumask) { - int new_apicid = early_per_cpu(x86_cpu_to_logical_apicid, cpu); - - if (round && APIC_CLUSTER(apicid) != APIC_CLUSTER(new_apicid)) { - printk("%s: Not a valid mask!\n", __func__); - return BAD_APICID; - } - apicid |= new_apicid; - round++; - } - return apicid; -} - -static unsigned int summit_cpu_mask_to_apicid_and(const struct cpumask *inmask, - const struct cpumask *andmask) -{ - int apicid = early_per_cpu(x86_cpu_to_logical_apicid, 0); - cpumask_var_t cpumask; - - if (!alloc_cpumask_var(&cpumask, GFP_ATOMIC)) - return apicid; - - cpumask_and(cpumask, inmask, andmask); - cpumask_and(cpumask, cpumask, cpu_online_mask); - apicid = summit_cpu_mask_to_apicid(cpumask); - - free_cpumask_var(cpumask); - - return apicid; -} - -/* - * cpuid returns the value latched in the HW at reset, not the APIC ID - * register's value. For any box whose BIOS changes APIC IDs, like - * clustered APIC systems, we must use hard_smp_processor_id. - * - * See Intel's IA-32 SW Dev's Manual Vol2 under CPUID. - */ -static int summit_phys_pkg_id(int cpuid_apic, int index_msb) -{ - return hard_smp_processor_id() >> index_msb; -} - -static int probe_summit(void) -{ - /* probed later in mptable/ACPI hooks */ - return 0; -} - -static void summit_vector_allocation_domain(int cpu, struct cpumask *retmask) -{ - /* Careful. Some cpus do not strictly honor the set of cpus - * specified in the interrupt destination when using lowest - * priority interrupt delivery mode. - * - * In particular there was a hyperthreading cpu observed to - * deliver interrupts to the wrong hyperthread when only one - * hyperthread was specified in the interrupt desitination. - */ - cpumask_clear(retmask); - cpumask_bits(retmask)[0] = APIC_ALL_CPUS; -} - -#ifdef CONFIG_X86_SUMMIT_NUMA -static struct rio_table_hdr *rio_table_hdr; -static struct scal_detail *scal_devs[MAX_NUMNODES]; -static struct rio_detail *rio_devs[MAX_NUMNODES*4]; - -#ifndef CONFIG_X86_NUMAQ -static int mp_bus_id_to_node[MAX_MP_BUSSES]; -#endif - -static int setup_pci_node_map_for_wpeg(int wpeg_num, int last_bus) -{ - int twister = 0, node = 0; - int i, bus, num_buses; - - for (i = 0; i < rio_table_hdr->num_rio_dev; i++) { - if (rio_devs[i]->node_id == rio_devs[wpeg_num]->owner_id) { - twister = rio_devs[i]->owner_id; - break; - } - } - if (i == rio_table_hdr->num_rio_dev) { - printk(KERN_ERR "%s: Couldn't find owner Cyclone for Winnipeg!\n", __func__); - return last_bus; - } - - for (i = 0; i < rio_table_hdr->num_scal_dev; i++) { - if (scal_devs[i]->node_id == twister) { - node = scal_devs[i]->node_id; - break; - } - } - if (i == rio_table_hdr->num_scal_dev) { - printk(KERN_ERR "%s: Couldn't find owner Twister for Cyclone!\n", __func__); - return last_bus; - } - - switch (rio_devs[wpeg_num]->type) { - case CompatWPEG: - /* - * The Compatibility Winnipeg controls the 2 legacy buses, - * the 66MHz PCI bus [2 slots] and the 2 "extra" buses in case - * a PCI-PCI bridge card is used in either slot: total 5 buses. - */ - num_buses = 5; - break; - case AltWPEG: - /* - * The Alternate Winnipeg controls the 2 133MHz buses [1 slot - * each], their 2 "extra" buses, the 100MHz bus [2 slots] and - * the "extra" buses for each of those slots: total 7 buses. - */ - num_buses = 7; - break; - case LookOutAWPEG: - case LookOutBWPEG: - /* - * A Lookout Winnipeg controls 3 100MHz buses [2 slots each] - * & the "extra" buses for each of those slots: total 9 buses. - */ - num_buses = 9; - break; - default: - printk(KERN_INFO "%s: Unsupported Winnipeg type!\n", __func__); - return last_bus; - } - - for (bus = last_bus; bus < last_bus + num_buses; bus++) - mp_bus_id_to_node[bus] = node; - return bus; -} - -static int build_detail_arrays(void) -{ - unsigned long ptr; - int i, scal_detail_size, rio_detail_size; - - if (rio_table_hdr->num_scal_dev > MAX_NUMNODES) { - printk(KERN_WARNING "%s: MAX_NUMNODES too low! Defined as %d, but system has %d nodes.\n", __func__, MAX_NUMNODES, rio_table_hdr->num_scal_dev); - return 0; - } - - switch (rio_table_hdr->version) { - default: - printk(KERN_WARNING "%s: Invalid Rio Grande Table Version: %d\n", __func__, rio_table_hdr->version); - return 0; - case 2: - scal_detail_size = 11; - rio_detail_size = 13; - break; - case 3: - scal_detail_size = 12; - rio_detail_size = 15; - break; - } - - ptr = (unsigned long)rio_table_hdr + 3; - for (i = 0; i < rio_table_hdr->num_scal_dev; i++, ptr += scal_detail_size) - scal_devs[i] = (struct scal_detail *)ptr; - - for (i = 0; i < rio_table_hdr->num_rio_dev; i++, ptr += rio_detail_size) - rio_devs[i] = (struct rio_detail *)ptr; - - return 1; -} - -void setup_summit(void) -{ - unsigned long ptr; - unsigned short offset; - int i, next_wpeg, next_bus = 0; - - /* The pointer to the EBDA is stored in the word @ phys 0x40E(40:0E) */ - ptr = get_bios_ebda(); - ptr = (unsigned long)phys_to_virt(ptr); - - rio_table_hdr = NULL; - offset = 0x180; - while (offset) { - /* The block id is stored in the 2nd word */ - if (*((unsigned short *)(ptr + offset + 2)) == 0x4752) { - /* set the pointer past the offset & block id */ - rio_table_hdr = (struct rio_table_hdr *)(ptr + offset + 4); - break; - } - /* The next offset is stored in the 1st word. 0 means no more */ - offset = *((unsigned short *)(ptr + offset)); - } - if (!rio_table_hdr) { - printk(KERN_ERR "%s: Unable to locate Rio Grande Table in EBDA - bailing!\n", __func__); - return; - } - - if (!build_detail_arrays()) - return; - - /* The first Winnipeg we're looking for has an index of 0 */ - next_wpeg = 0; - do { - for (i = 0; i < rio_table_hdr->num_rio_dev; i++) { - if (is_WPEG(rio_devs[i]) && rio_devs[i]->WP_index == next_wpeg) { - /* It's the Winnipeg we're looking for! */ - next_bus = setup_pci_node_map_for_wpeg(i, next_bus); - next_wpeg++; - break; - } - } - /* - * If we go through all Rio devices and don't find one with - * the next index, it means we've found all the Winnipegs, - * and thus all the PCI buses. - */ - if (i == rio_table_hdr->num_rio_dev) - next_wpeg = 0; - } while (next_wpeg != 0); -} -#endif - -static struct apic apic_summit = { - - .name = "summit", - .probe = probe_summit, - .acpi_madt_oem_check = summit_acpi_madt_oem_check, - .apic_id_registered = summit_apic_id_registered, - - .irq_delivery_mode = dest_LowestPrio, - /* logical delivery broadcast to all CPUs: */ - .irq_dest_mode = 1, - - .target_cpus = summit_target_cpus, - .disable_esr = 1, - .dest_logical = APIC_DEST_LOGICAL, - .check_apicid_used = summit_check_apicid_used, - .check_apicid_present = summit_check_apicid_present, - - .vector_allocation_domain = summit_vector_allocation_domain, - .init_apic_ldr = summit_init_apic_ldr, - - .ioapic_phys_id_map = summit_ioapic_phys_id_map, - .setup_apic_routing = summit_setup_apic_routing, - .multi_timer_check = NULL, - .cpu_present_to_apicid = summit_cpu_present_to_apicid, - .apicid_to_cpu_present = summit_apicid_to_cpu_present, - .setup_portio_remap = NULL, - .check_phys_apicid_present = summit_check_phys_apicid_present, - .enable_apic_mode = NULL, - .phys_pkg_id = summit_phys_pkg_id, - .mps_oem_check = summit_mps_oem_check, - - .get_apic_id = summit_get_apic_id, - .set_apic_id = NULL, - .apic_id_mask = 0xFF << 24, - - .cpu_mask_to_apicid = summit_cpu_mask_to_apicid, - .cpu_mask_to_apicid_and = summit_cpu_mask_to_apicid_and, - - .send_IPI_mask = summit_send_IPI_mask, - .send_IPI_mask_allbutself = NULL, - .send_IPI_allbutself = summit_send_IPI_allbutself, - .send_IPI_all = summit_send_IPI_all, - .send_IPI_self = default_send_IPI_self, - - .trampoline_phys_low = DEFAULT_TRAMPOLINE_PHYS_LOW, - .trampoline_phys_high = DEFAULT_TRAMPOLINE_PHYS_HIGH, - - .wait_for_init_deassert = default_wait_for_init_deassert, - - .smp_callin_clear_local_apic = NULL, - .inquire_remote_apic = default_inquire_remote_apic, - - .read = native_apic_mem_read, - .write = native_apic_mem_write, - .icr_read = native_apic_icr_read, - .icr_write = native_apic_icr_write, - .wait_icr_idle = native_apic_wait_icr_idle, - .safe_wait_icr_idle = native_safe_apic_wait_icr_idle, - - .x86_32_early_logical_apicid = summit_early_logical_apicid, -}; - -apic_driver(apic_summit); diff --git a/arch/x86/kernel/apic/x2apic_cluster.c b/arch/x86/kernel/apic/x2apic_cluster.c index 50079587582..e66766bf164 100644 --- a/arch/x86/kernel/apic/x2apic_cluster.c +++ b/arch/x86/kernel/apic/x2apic_cluster.c @@ -3,7 +3,6 @@ #include <linux/string.h> #include <linux/kernel.h> #include <linux/ctype.h> -#include <linux/init.h> #include <linux/dmar.h> #include <linux/cpu.h> @@ -81,7 +80,7 @@ static void x2apic_send_IPI_mask(const struct cpumask *mask, int vector) } static void - x2apic_send_IPI_mask_allbutself(const struct cpumask *mask, int vector) +x2apic_send_IPI_mask_allbutself(const struct cpumask *mask, int vector) { __x2apic_send_IPI_mask(mask, vector, APIC_DEST_ALLBUT); } @@ -96,36 +95,37 @@ static void x2apic_send_IPI_all(int vector) __x2apic_send_IPI_mask(cpu_online_mask, vector, APIC_DEST_ALLINC); } -static unsigned int x2apic_cpu_mask_to_apicid(const struct cpumask *cpumask) +static int +x2apic_cpu_mask_to_apicid_and(const struct cpumask *cpumask, + const struct cpumask *andmask, + unsigned int *apicid) { - /* - * We're using fixed IRQ delivery, can only return one logical APIC ID. - * May as well be the first. - */ - int cpu = cpumask_first(cpumask); + u32 dest = 0; + u16 cluster; + int i; - if ((unsigned)cpu < nr_cpu_ids) - return per_cpu(x86_cpu_to_logical_apicid, cpu); - else - return BAD_APICID; -} + for_each_cpu_and(i, cpumask, andmask) { + if (!cpumask_test_cpu(i, cpu_online_mask)) + continue; + dest = per_cpu(x86_cpu_to_logical_apicid, i); + cluster = x2apic_cluster(i); + break; + } -static unsigned int -x2apic_cpu_mask_to_apicid_and(const struct cpumask *cpumask, - const struct cpumask *andmask) -{ - int cpu; + if (!dest) + return -EINVAL; - /* - * We're using fixed IRQ delivery, can only return one logical APIC ID. - * May as well be the first. - */ - for_each_cpu_and(cpu, cpumask, andmask) { - if (cpumask_test_cpu(cpu, cpu_online_mask)) - break; + for_each_cpu_and(i, cpumask, andmask) { + if (!cpumask_test_cpu(i, cpu_online_mask)) + continue; + if (cluster != x2apic_cluster(i)) + continue; + dest |= per_cpu(x86_cpu_to_logical_apicid, i); } - return per_cpu(x86_cpu_to_logical_apicid, cpu); + *apicid = dest; + + return 0; } static void init_x2apic_ldr(void) @@ -147,7 +147,7 @@ static void init_x2apic_ldr(void) /* * At CPU state changes, update the x2apic cluster sibling info. */ -static int __cpuinit +static int update_clusterinfo(struct notifier_block *nfb, unsigned long action, void *hcpu) { unsigned int this_cpu = (unsigned long)hcpu; @@ -208,23 +208,50 @@ static int x2apic_cluster_probe(void) return 0; } +static const struct cpumask *x2apic_cluster_target_cpus(void) +{ + return cpu_all_mask; +} + +/* + * Each x2apic cluster is an allocation domain. + */ +static void cluster_vector_allocation_domain(int cpu, struct cpumask *retmask, + const struct cpumask *mask) +{ + /* + * To minimize vector pressure, default case of boot, device bringup + * etc will use a single cpu for the interrupt destination. + * + * On explicit migration requests coming from irqbalance etc, + * interrupts will be routed to the x2apic cluster (cluster-id + * derived from the first cpu in the mask) members specified + * in the mask. + */ + if (mask == x2apic_cluster_target_cpus()) + cpumask_copy(retmask, cpumask_of(cpu)); + else + cpumask_and(retmask, mask, per_cpu(cpus_in_cluster, cpu)); +} + static struct apic apic_x2apic_cluster = { .name = "cluster x2apic", .probe = x2apic_cluster_probe, .acpi_madt_oem_check = x2apic_acpi_madt_oem_check, + .apic_id_valid = x2apic_apic_id_valid, .apic_id_registered = x2apic_apic_id_registered, .irq_delivery_mode = dest_LowestPrio, .irq_dest_mode = 1, /* logical */ - .target_cpus = x2apic_target_cpus, + .target_cpus = x2apic_cluster_target_cpus, .disable_esr = 0, .dest_logical = APIC_DEST_LOGICAL, .check_apicid_used = NULL, .check_apicid_present = NULL, - .vector_allocation_domain = x2apic_vector_allocation_domain, + .vector_allocation_domain = cluster_vector_allocation_domain, .init_apic_ldr = init_x2apic_ldr, .ioapic_phys_id_map = NULL, @@ -242,7 +269,6 @@ static struct apic apic_x2apic_cluster = { .set_apic_id = x2apic_set_apic_id, .apic_id_mask = 0xFFFFFFFFu, - .cpu_mask_to_apicid = x2apic_cpu_mask_to_apicid, .cpu_mask_to_apicid_and = x2apic_cpu_mask_to_apicid_and, .send_IPI_mask = x2apic_send_IPI_mask, @@ -253,12 +279,13 @@ static struct apic apic_x2apic_cluster = { .trampoline_phys_low = DEFAULT_TRAMPOLINE_PHYS_LOW, .trampoline_phys_high = DEFAULT_TRAMPOLINE_PHYS_HIGH, - .wait_for_init_deassert = NULL, + .wait_for_init_deassert = false, .smp_callin_clear_local_apic = NULL, .inquire_remote_apic = NULL, .read = native_apic_msr_read, .write = native_apic_msr_write, + .eoi_write = native_apic_msr_eoi_write, .icr_read = native_x2apic_icr_read, .icr_write = native_x2apic_icr_write, .wait_icr_idle = native_x2apic_wait_icr_idle, diff --git a/arch/x86/kernel/apic/x2apic_phys.c b/arch/x86/kernel/apic/x2apic_phys.c index f5373dfde21..6d600ebf6c1 100644 --- a/arch/x86/kernel/apic/x2apic_phys.c +++ b/arch/x86/kernel/apic/x2apic_phys.c @@ -3,7 +3,6 @@ #include <linux/string.h> #include <linux/kernel.h> #include <linux/ctype.h> -#include <linux/init.h> #include <linux/dmar.h> #include <asm/smp.h> @@ -20,12 +19,19 @@ static int set_x2apic_phys_mode(char *arg) } early_param("x2apic_phys", set_x2apic_phys_mode); +static bool x2apic_fadt_phys(void) +{ + if ((acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID) && + (acpi_gbl_FADT.flags & ACPI_FADT_APIC_PHYSICAL)) { + printk(KERN_DEBUG "System requires x2apic physical mode\n"); + return true; + } + return false; +} + static int x2apic_acpi_madt_oem_check(char *oem_id, char *oem_table_id) { - if (x2apic_phys) - return x2apic_enabled(); - else - return 0; + return x2apic_enabled() && (x2apic_phys || x2apic_fadt_phys()); } static void @@ -70,45 +76,13 @@ static void x2apic_send_IPI_all(int vector) __x2apic_send_IPI_mask(cpu_online_mask, vector, APIC_DEST_ALLINC); } -static unsigned int x2apic_cpu_mask_to_apicid(const struct cpumask *cpumask) -{ - /* - * We're using fixed IRQ delivery, can only return one phys APIC ID. - * May as well be the first. - */ - int cpu = cpumask_first(cpumask); - - if ((unsigned)cpu < nr_cpu_ids) - return per_cpu(x86_cpu_to_apicid, cpu); - else - return BAD_APICID; -} - -static unsigned int -x2apic_cpu_mask_to_apicid_and(const struct cpumask *cpumask, - const struct cpumask *andmask) -{ - int cpu; - - /* - * We're using fixed IRQ delivery, can only return one phys APIC ID. - * May as well be the first. - */ - for_each_cpu_and(cpu, cpumask, andmask) { - if (cpumask_test_cpu(cpu, cpu_online_mask)) - break; - } - - return per_cpu(x86_cpu_to_apicid, cpu); -} - static void init_x2apic_ldr(void) { } static int x2apic_phys_probe(void) { - if (x2apic_mode && x2apic_phys) + if (x2apic_mode && (x2apic_phys || x2apic_fadt_phys())) return 1; return apic == &apic_x2apic_phys; @@ -119,18 +93,19 @@ static struct apic apic_x2apic_phys = { .name = "physical x2apic", .probe = x2apic_phys_probe, .acpi_madt_oem_check = x2apic_acpi_madt_oem_check, + .apic_id_valid = x2apic_apic_id_valid, .apic_id_registered = x2apic_apic_id_registered, .irq_delivery_mode = dest_Fixed, .irq_dest_mode = 0, /* physical */ - .target_cpus = x2apic_target_cpus, + .target_cpus = online_target_cpus, .disable_esr = 0, .dest_logical = 0, .check_apicid_used = NULL, .check_apicid_present = NULL, - .vector_allocation_domain = x2apic_vector_allocation_domain, + .vector_allocation_domain = default_vector_allocation_domain, .init_apic_ldr = init_x2apic_ldr, .ioapic_phys_id_map = NULL, @@ -148,8 +123,7 @@ static struct apic apic_x2apic_phys = { .set_apic_id = x2apic_set_apic_id, .apic_id_mask = 0xFFFFFFFFu, - .cpu_mask_to_apicid = x2apic_cpu_mask_to_apicid, - .cpu_mask_to_apicid_and = x2apic_cpu_mask_to_apicid_and, + .cpu_mask_to_apicid_and = default_cpu_mask_to_apicid_and, .send_IPI_mask = x2apic_send_IPI_mask, .send_IPI_mask_allbutself = x2apic_send_IPI_mask_allbutself, @@ -159,12 +133,13 @@ static struct apic apic_x2apic_phys = { .trampoline_phys_low = DEFAULT_TRAMPOLINE_PHYS_LOW, .trampoline_phys_high = DEFAULT_TRAMPOLINE_PHYS_HIGH, - .wait_for_init_deassert = NULL, + .wait_for_init_deassert = false, .smp_callin_clear_local_apic = NULL, .inquire_remote_apic = NULL, .read = native_apic_msr_read, .write = native_apic_msr_write, + .eoi_write = native_apic_msr_eoi_write, .icr_read = native_x2apic_icr_read, .icr_write = native_x2apic_icr_write, .wait_icr_idle = native_x2apic_wait_icr_idle, diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c index 79b05b88aa1..293b41df54e 100644 --- a/arch/x86/kernel/apic/x2apic_uv_x.c +++ b/arch/x86/kernel/apic/x2apic_uv_x.c @@ -5,7 +5,7 @@ * * SGI UV APIC functions (note: not an Intel compatible APIC) * - * Copyright (C) 2007-2010 Silicon Graphics, Inc. All rights reserved. + * Copyright (C) 2007-2014 Silicon Graphics, Inc. All rights reserved. */ #include <linux/cpumask.h> #include <linux/hardirq.h> @@ -25,6 +25,7 @@ #include <linux/kdebug.h> #include <linux/delay.h> #include <linux/crash_dump.h> +#include <linux/reboot.h> #include <asm/uv/uv_mmrs.h> #include <asm/uv/uv_hub.h> @@ -36,27 +37,21 @@ #include <asm/ipi.h> #include <asm/smp.h> #include <asm/x86_init.h> -#include <asm/emergency-restart.h> #include <asm/nmi.h> -/* BMC sets a bit this MMR non-zero before sending an NMI */ -#define UVH_NMI_MMR UVH_SCRATCH5 -#define UVH_NMI_MMR_CLEAR (UVH_NMI_MMR + 8) -#define UV_NMI_PENDING_MASK (1UL << 63) -DEFINE_PER_CPU(unsigned long, cpu_last_nmi_count); - DEFINE_PER_CPU(int, x2apic_extra_bits); #define PR_DEVEL(fmt, args...) pr_devel("%s: " fmt, __func__, args) static enum uv_system_type uv_system_type; static u64 gru_start_paddr, gru_end_paddr; +static u64 gru_dist_base, gru_first_node_paddr = -1LL, gru_last_node_paddr; +static u64 gru_dist_lmask, gru_dist_umask; static union uvh_apicid uvh_apicid; int uv_min_hub_revision_id; EXPORT_SYMBOL_GPL(uv_min_hub_revision_id); unsigned int uv_apicid_hibits; EXPORT_SYMBOL_GPL(uv_apicid_hibits); -static DEFINE_SPINLOCK(uv_nmi_lock); static struct apic apic_x2apic_uv_x; @@ -72,7 +67,20 @@ static unsigned long __init uv_early_read_mmr(unsigned long addr) static inline bool is_GRU_range(u64 start, u64 end) { - return start >= gru_start_paddr && end <= gru_end_paddr; + if (gru_dist_base) { + u64 su = start & gru_dist_umask; /* upper (incl pnode) bits */ + u64 sl = start & gru_dist_lmask; /* base offset bits */ + u64 eu = end & gru_dist_umask; + u64 el = end & gru_dist_lmask; + + /* Must reside completely within a single GRU range */ + return (sl == gru_dist_base && el == gru_dist_base && + su >= gru_first_node_paddr && + su <= gru_last_node_paddr && + eu == su); + } else { + return start >= gru_start_paddr && end <= gru_end_paddr; + } } static bool uv_is_untracked_pat_range(u64 start, u64 end) @@ -91,10 +99,16 @@ static int __init early_get_pnodeid(void) m_n_config.v = uv_early_read_mmr(UVH_RH_GAM_CONFIG_MMR); uv_min_hub_revision_id = node_id.s.revision; - if (node_id.s.part_number == UV2_HUB_PART_NUMBER) - uv_min_hub_revision_id += UV2_HUB_REVISION_BASE - 1; - if (node_id.s.part_number == UV2_HUB_PART_NUMBER_X) + switch (node_id.s.part_number) { + case UV2_HUB_PART_NUMBER: + case UV2_HUB_PART_NUMBER_X: uv_min_hub_revision_id += UV2_HUB_REVISION_BASE - 1; + break; + case UV3_HUB_PART_NUMBER: + case UV3_HUB_PART_NUMBER_X: + uv_min_hub_revision_id += UV3_HUB_REVISION_BASE; + break; + } uv_hub_info->hub_revision = uv_min_hub_revision_id; pnode = (node_id.s.node_id >> 1) & ((1 << m_n_config.s.n_skt) - 1); @@ -130,13 +144,16 @@ static void __init uv_set_apicid_hibit(void) static int __init uv_acpi_madt_oem_check(char *oem_id, char *oem_table_id) { - int pnodeid, is_uv1, is_uv2; + int pnodeid, is_uv1, is_uv2, is_uv3; is_uv1 = !strcmp(oem_id, "SGI"); is_uv2 = !strcmp(oem_id, "SGI2"); - if (is_uv1 || is_uv2) { + is_uv3 = !strncmp(oem_id, "SGI3", 4); /* there are varieties of UV3 */ + if (is_uv1 || is_uv2 || is_uv3) { uv_hub_info->hub_revision = - is_uv1 ? UV1_HUB_REVISION_BASE : UV2_HUB_REVISION_BASE; + (is_uv1 ? UV1_HUB_REVISION_BASE : + (is_uv2 ? UV2_HUB_REVISION_BASE : + UV3_HUB_REVISION_BASE)); pnodeid = early_get_pnodeid(); early_get_apic_pnode_shift(); x86_platform.is_untracked_pat_range = uv_is_untracked_pat_range; @@ -185,18 +202,7 @@ EXPORT_SYMBOL_GPL(uv_possible_blades); unsigned long sn_rtc_cycles_per_second; EXPORT_SYMBOL(sn_rtc_cycles_per_second); -static const struct cpumask *uv_target_cpus(void) -{ - return cpu_online_mask; -} - -static void uv_vector_allocation_domain(int cpu, struct cpumask *retmask) -{ - cpumask_clear(retmask); - cpumask_set_cpu(cpu, retmask); -} - -static int __cpuinit uv_wakeup_secondary(int phys_apicid, unsigned long start_rip) +static int uv_wakeup_secondary(int phys_apicid, unsigned long start_rip) { #ifdef CONFIG_SMP unsigned long val; @@ -266,34 +272,26 @@ static void uv_send_IPI_all(int vector) uv_send_IPI_mask(cpu_online_mask, vector); } -static int uv_apic_id_registered(void) +static int uv_apic_id_valid(int apicid) { return 1; } -static void uv_init_apic_ldr(void) +static int uv_apic_id_registered(void) { + return 1; } -static unsigned int uv_cpu_mask_to_apicid(const struct cpumask *cpumask) +static void uv_init_apic_ldr(void) { - /* - * We're using fixed IRQ delivery, can only return one phys APIC ID. - * May as well be the first. - */ - int cpu = cpumask_first(cpumask); - - if ((unsigned)cpu < nr_cpu_ids) - return per_cpu(x86_cpu_to_apicid, cpu) | uv_apicid_hibits; - else - return BAD_APICID; } -static unsigned int +static int uv_cpu_mask_to_apicid_and(const struct cpumask *cpumask, - const struct cpumask *andmask) + const struct cpumask *andmask, + unsigned int *apicid) { - int cpu; + int unsigned cpu; /* * We're using fixed IRQ delivery, can only return one phys APIC ID. @@ -303,7 +301,13 @@ uv_cpu_mask_to_apicid_and(const struct cpumask *cpumask, if (cpumask_test_cpu(cpu, cpu_online_mask)) break; } - return per_cpu(x86_cpu_to_apicid, cpu) | uv_apicid_hibits; + + if (likely(cpu < nr_cpu_ids)) { + *apicid = per_cpu(x86_cpu_to_apicid, cpu) | uv_apicid_hibits; + return 0; + } + + return -EINVAL; } static unsigned int x2apic_get_apic_id(unsigned long x) @@ -351,18 +355,19 @@ static struct apic __refdata apic_x2apic_uv_x = { .name = "UV large system", .probe = uv_probe, .acpi_madt_oem_check = uv_acpi_madt_oem_check, + .apic_id_valid = uv_apic_id_valid, .apic_id_registered = uv_apic_id_registered, .irq_delivery_mode = dest_Fixed, .irq_dest_mode = 0, /* physical */ - .target_cpus = uv_target_cpus, + .target_cpus = online_target_cpus, .disable_esr = 0, .dest_logical = APIC_DEST_LOGICAL, .check_apicid_used = NULL, .check_apicid_present = NULL, - .vector_allocation_domain = uv_vector_allocation_domain, + .vector_allocation_domain = default_vector_allocation_domain, .init_apic_ldr = uv_init_apic_ldr, .ioapic_phys_id_map = NULL, @@ -380,7 +385,6 @@ static struct apic __refdata apic_x2apic_uv_x = { .set_apic_id = set_apic_id, .apic_id_mask = 0xFFFFFFFFu, - .cpu_mask_to_apicid = uv_cpu_mask_to_apicid, .cpu_mask_to_apicid_and = uv_cpu_mask_to_apicid_and, .send_IPI_mask = uv_send_IPI_mask, @@ -392,19 +396,20 @@ static struct apic __refdata apic_x2apic_uv_x = { .wakeup_secondary_cpu = uv_wakeup_secondary, .trampoline_phys_low = DEFAULT_TRAMPOLINE_PHYS_LOW, .trampoline_phys_high = DEFAULT_TRAMPOLINE_PHYS_HIGH, - .wait_for_init_deassert = NULL, + .wait_for_init_deassert = false, .smp_callin_clear_local_apic = NULL, .inquire_remote_apic = NULL, .read = native_apic_msr_read, .write = native_apic_msr_write, + .eoi_write = native_apic_msr_eoi_write, .icr_read = native_x2apic_icr_read, .icr_write = native_x2apic_icr_write, .wait_icr_idle = native_x2apic_wait_icr_idle, .safe_wait_icr_idle = native_safe_x2apic_wait_icr_idle, }; -static __cpuinit void set_x2apic_extra_bits(int pnode) +static void set_x2apic_extra_bits(int pnode) { __this_cpu_write(x2apic_extra_bits, pnode << uvh_apicid.s.pnode_shift); } @@ -435,6 +440,20 @@ static __initdata struct redir_addr redir_addrs[] = { {UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_2_MMR, UVH_RH_GAM_ALIAS210_OVERLAY_CONFIG_2_MMR}, }; +static unsigned char get_n_lshift(int m_val) +{ + union uv3h_gr0_gam_gr_config_u m_gr_config; + + if (is_uv1_hub()) + return m_val; + + if (is_uv2_hub()) + return m_val == 40 ? 40 : 39; + + m_gr_config.v = uv_read_local_mmr(UV3H_GR0_GAM_GR_CONFIG); + return m_gr_config.s3.m_skt; +} + static __init void get_lowmem_redirect(unsigned long *base, unsigned long *size) { union uvh_rh_gam_alias210_overlay_config_2_mmr_u alias; @@ -462,26 +481,67 @@ static __init void map_high(char *id, unsigned long base, int pshift, paddr = base << pshift; bytes = (1UL << bshift) * (max_pnode + 1); - printk(KERN_INFO "UV: Map %s_HI 0x%lx - 0x%lx\n", id, paddr, - paddr + bytes); + if (!paddr) { + pr_info("UV: Map %s_HI base address NULL\n", id); + return; + } + pr_debug("UV: Map %s_HI 0x%lx - 0x%lx\n", id, paddr, paddr + bytes); if (map_type == map_uc) init_extra_mapping_uc(paddr, bytes); else init_extra_mapping_wb(paddr, bytes); +} +static __init void map_gru_distributed(unsigned long c) +{ + union uvh_rh_gam_gru_overlay_config_mmr_u gru; + u64 paddr; + unsigned long bytes; + int nid; + + gru.v = c; + /* only base bits 42:28 relevant in dist mode */ + gru_dist_base = gru.v & 0x000007fff0000000UL; + if (!gru_dist_base) { + pr_info("UV: Map GRU_DIST base address NULL\n"); + return; + } + bytes = 1UL << UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR_BASE_SHFT; + gru_dist_lmask = ((1UL << uv_hub_info->m_val) - 1) & ~(bytes - 1); + gru_dist_umask = ~((1UL << uv_hub_info->m_val) - 1); + gru_dist_base &= gru_dist_lmask; /* Clear bits above M */ + for_each_online_node(nid) { + paddr = ((u64)uv_node_to_pnode(nid) << uv_hub_info->m_val) | + gru_dist_base; + init_extra_mapping_wb(paddr, bytes); + gru_first_node_paddr = min(paddr, gru_first_node_paddr); + gru_last_node_paddr = max(paddr, gru_last_node_paddr); + } + /* Save upper (63:M) bits of address only for is_GRU_range */ + gru_first_node_paddr &= gru_dist_umask; + gru_last_node_paddr &= gru_dist_umask; + pr_debug("UV: Map GRU_DIST base 0x%016llx 0x%016llx - 0x%016llx\n", + gru_dist_base, gru_first_node_paddr, gru_last_node_paddr); } + static __init void map_gru_high(int max_pnode) { union uvh_rh_gam_gru_overlay_config_mmr_u gru; int shift = UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR_BASE_SHFT; gru.v = uv_read_local_mmr(UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR); - if (gru.s.enable) { - map_high("GRU", gru.s.base, shift, shift, max_pnode, map_wb); - gru_start_paddr = ((u64)gru.s.base << shift); - gru_end_paddr = gru_start_paddr + (1UL << shift) * (max_pnode + 1); + if (!gru.s.enable) { + pr_info("UV: GRU disabled\n"); + return; + } + if (is_uv3_hub() && gru.s3.mode) { + map_gru_distributed(gru.v); + return; } + map_high("GRU", gru.s.base, shift, shift, max_pnode, map_wb); + gru_start_paddr = ((u64)gru.s.base << shift); + gru_end_paddr = gru_start_paddr + (1UL << shift) * (max_pnode + 1); } static __init void map_mmr_high(int max_pnode) @@ -492,23 +552,146 @@ static __init void map_mmr_high(int max_pnode) mmr.v = uv_read_local_mmr(UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR); if (mmr.s.enable) map_high("MMR", mmr.s.base, shift, shift, max_pnode, map_uc); + else + pr_info("UV: MMR disabled\n"); +} + +/* + * This commonality works because both 0 & 1 versions of the MMIOH OVERLAY + * and REDIRECT MMR regs are exactly the same on UV3. + */ +struct mmioh_config { + unsigned long overlay; + unsigned long redirect; + char *id; +}; + +static __initdata struct mmioh_config mmiohs[] = { + { + UV3H_RH_GAM_MMIOH_OVERLAY_CONFIG0_MMR, + UV3H_RH_GAM_MMIOH_REDIRECT_CONFIG0_MMR, + "MMIOH0" + }, + { + UV3H_RH_GAM_MMIOH_OVERLAY_CONFIG1_MMR, + UV3H_RH_GAM_MMIOH_REDIRECT_CONFIG1_MMR, + "MMIOH1" + }, +}; + +static __init void map_mmioh_high_uv3(int index, int min_pnode, int max_pnode) +{ + union uv3h_rh_gam_mmioh_overlay_config0_mmr_u overlay; + unsigned long mmr; + unsigned long base; + int i, n, shift, m_io, max_io; + int nasid, lnasid, fi, li; + char *id; + + id = mmiohs[index].id; + overlay.v = uv_read_local_mmr(mmiohs[index].overlay); + pr_info("UV: %s overlay 0x%lx base:0x%x m_io:%d\n", + id, overlay.v, overlay.s3.base, overlay.s3.m_io); + if (!overlay.s3.enable) { + pr_info("UV: %s disabled\n", id); + return; + } + + shift = UV3H_RH_GAM_MMIOH_OVERLAY_CONFIG0_MMR_BASE_SHFT; + base = (unsigned long)overlay.s3.base; + m_io = overlay.s3.m_io; + mmr = mmiohs[index].redirect; + n = UV3H_RH_GAM_MMIOH_REDIRECT_CONFIG0_MMR_DEPTH; + min_pnode *= 2; /* convert to NASID */ + max_pnode *= 2; + max_io = lnasid = fi = li = -1; + + for (i = 0; i < n; i++) { + union uv3h_rh_gam_mmioh_redirect_config0_mmr_u redirect; + + redirect.v = uv_read_local_mmr(mmr + i * 8); + nasid = redirect.s3.nasid; + if (nasid < min_pnode || max_pnode < nasid) + nasid = -1; /* invalid NASID */ + + if (nasid == lnasid) { + li = i; + if (i != n-1) /* last entry check */ + continue; + } + + /* check if we have a cached (or last) redirect to print */ + if (lnasid != -1 || (i == n-1 && nasid != -1)) { + unsigned long addr1, addr2; + int f, l; + + if (lnasid == -1) { + f = l = i; + lnasid = nasid; + } else { + f = fi; + l = li; + } + addr1 = (base << shift) + + f * (unsigned long)(1 << m_io); + addr2 = (base << shift) + + (l + 1) * (unsigned long)(1 << m_io); + pr_info("UV: %s[%03d..%03d] NASID 0x%04x ADDR 0x%016lx - 0x%016lx\n", + id, fi, li, lnasid, addr1, addr2); + if (max_io < l) + max_io = l; + } + fi = li = i; + lnasid = nasid; + } + + pr_info("UV: %s base:0x%lx shift:%d M_IO:%d MAX_IO:%d\n", + id, base, shift, m_io, max_io); + + if (max_io >= 0) + map_high(id, base, shift, m_io, max_io, map_uc); } -static __init void map_mmioh_high(int max_pnode) +static __init void map_mmioh_high(int min_pnode, int max_pnode) { union uvh_rh_gam_mmioh_overlay_config_mmr_u mmioh; - int shift; + unsigned long mmr, base; + int shift, enable, m_io, n_io; - mmioh.v = uv_read_local_mmr(UVH_RH_GAM_MMIOH_OVERLAY_CONFIG_MMR); - if (is_uv1_hub() && mmioh.s1.enable) { - shift = UV1H_RH_GAM_MMIOH_OVERLAY_CONFIG_MMR_BASE_SHFT; - map_high("MMIOH", mmioh.s1.base, shift, mmioh.s1.m_io, - max_pnode, map_uc); + if (is_uv3_hub()) { + /* Map both MMIOH Regions */ + map_mmioh_high_uv3(0, min_pnode, max_pnode); + map_mmioh_high_uv3(1, min_pnode, max_pnode); + return; } - if (is_uv2_hub() && mmioh.s2.enable) { + + if (is_uv1_hub()) { + mmr = UV1H_RH_GAM_MMIOH_OVERLAY_CONFIG_MMR; + shift = UV1H_RH_GAM_MMIOH_OVERLAY_CONFIG_MMR_BASE_SHFT; + mmioh.v = uv_read_local_mmr(mmr); + enable = !!mmioh.s1.enable; + base = mmioh.s1.base; + m_io = mmioh.s1.m_io; + n_io = mmioh.s1.n_io; + } else if (is_uv2_hub()) { + mmr = UV2H_RH_GAM_MMIOH_OVERLAY_CONFIG_MMR; shift = UV2H_RH_GAM_MMIOH_OVERLAY_CONFIG_MMR_BASE_SHFT; - map_high("MMIOH", mmioh.s2.base, shift, mmioh.s2.m_io, - max_pnode, map_uc); + mmioh.v = uv_read_local_mmr(mmr); + enable = !!mmioh.s2.enable; + base = mmioh.s2.base; + m_io = mmioh.s2.m_io; + n_io = mmioh.s2.n_io; + } else + return; + + if (enable) { + max_pnode &= (1 << n_io) - 1; + pr_info( + "UV: base:0x%lx shift:%d N_IO:%d M_IO:%d max_pnode:0x%x\n", + base, shift, m_io, n_io, max_pnode); + map_high("MMIOH", base, shift, m_io, max_pnode, map_uc); + } else { + pr_info("UV: MMIOH disabled\n"); } } @@ -559,7 +742,7 @@ static void uv_heartbeat(unsigned long ignored) mod_timer_pinned(timer, jiffies + SCIR_CPU_HB_INTERVAL); } -static void __cpuinit uv_heartbeat_enable(int cpu) +static void uv_heartbeat_enable(int cpu) { while (!uv_cpu_hub_info(cpu)->scir.enabled) { struct timer_list *timer = &uv_cpu_hub_info(cpu)->scir.timer; @@ -576,7 +759,7 @@ static void __cpuinit uv_heartbeat_enable(int cpu) } #ifdef CONFIG_HOTPLUG_CPU -static void __cpuinit uv_heartbeat_disable(int cpu) +static void uv_heartbeat_disable(int cpu) { if (uv_cpu_hub_info(cpu)->scir.enabled) { uv_cpu_hub_info(cpu)->scir.enabled = 0; @@ -588,8 +771,8 @@ static void __cpuinit uv_heartbeat_disable(int cpu) /* * cpu hotplug notifier */ -static __cpuinit int uv_scir_cpu_notify(struct notifier_block *self, - unsigned long action, void *hcpu) +static int uv_scir_cpu_notify(struct notifier_block *self, unsigned long action, + void *hcpu) { long cpu = (long)hcpu; @@ -659,7 +842,7 @@ int uv_set_vga_state(struct pci_dev *pdev, bool decode, * Called on each cpu to initialize the per_cpu UV data area. * FIXME: hotplug not supported yet */ -void __cpuinit uv_cpu_init(void) +void uv_cpu_init(void) { /* CPU 0 initilization will be done via uv_system_init. */ if (!uv_blade_info) @@ -671,107 +854,47 @@ void __cpuinit uv_cpu_init(void) set_x2apic_extra_bits(uv_hub_info->pnode); } -/* - * When NMI is received, print a stack trace. - */ -int uv_handle_nmi(unsigned int reason, struct pt_regs *regs) -{ - unsigned long real_uv_nmi; - int bid; - - /* - * Each blade has an MMR that indicates when an NMI has been sent - * to cpus on the blade. If an NMI is detected, atomically - * clear the MMR and update a per-blade NMI count used to - * cause each cpu on the blade to notice a new NMI. - */ - bid = uv_numa_blade_id(); - real_uv_nmi = (uv_read_local_mmr(UVH_NMI_MMR) & UV_NMI_PENDING_MASK); - - if (unlikely(real_uv_nmi)) { - spin_lock(&uv_blade_info[bid].nmi_lock); - real_uv_nmi = (uv_read_local_mmr(UVH_NMI_MMR) & UV_NMI_PENDING_MASK); - if (real_uv_nmi) { - uv_blade_info[bid].nmi_count++; - uv_write_local_mmr(UVH_NMI_MMR_CLEAR, UV_NMI_PENDING_MASK); - } - spin_unlock(&uv_blade_info[bid].nmi_lock); - } - - if (likely(__get_cpu_var(cpu_last_nmi_count) == uv_blade_info[bid].nmi_count)) - return NMI_DONE; - - __get_cpu_var(cpu_last_nmi_count) = uv_blade_info[bid].nmi_count; - - /* - * Use a lock so only one cpu prints at a time. - * This prevents intermixed output. - */ - spin_lock(&uv_nmi_lock); - pr_info("UV NMI stack dump cpu %u:\n", smp_processor_id()); - dump_stack(); - spin_unlock(&uv_nmi_lock); - - return NMI_HANDLED; -} - -void uv_register_nmi_notifier(void) -{ - if (register_nmi_handler(NMI_UNKNOWN, uv_handle_nmi, 0, "uv")) - printk(KERN_WARNING "UV NMI handler failed to register\n"); -} - -void uv_nmi_init(void) -{ - unsigned int value; - - /* - * Unmask NMI on all cpus - */ - value = apic_read(APIC_LVT1) | APIC_DM_NMI; - value &= ~APIC_LVT_MASKED; - apic_write(APIC_LVT1, value); -} - void __init uv_system_init(void) { union uvh_rh_gam_config_mmr_u m_n_config; - union uvh_rh_gam_mmioh_overlay_config_mmr_u mmioh; union uvh_node_id_u node_id; unsigned long gnode_upper, lowmem_redir_base, lowmem_redir_size; - int bytes, nid, cpu, lcpu, pnode, blade, i, j, m_val, n_val, n_io; - int gnode_extra, max_pnode = 0; + int bytes, nid, cpu, lcpu, pnode, blade, i, j, m_val, n_val; + int gnode_extra, min_pnode = 999999, max_pnode = -1; unsigned long mmr_base, present, paddr; - unsigned short pnode_mask, pnode_io_mask; + unsigned short pnode_mask; + unsigned char n_lshift; + char *hub = (is_uv1_hub() ? "UV1" : + (is_uv2_hub() ? "UV2" : + "UV3")); - printk(KERN_INFO "UV: Found %s hub\n", is_uv1_hub() ? "UV1" : "UV2"); + pr_info("UV: Found %s hub\n", hub); map_low_mmrs(); m_n_config.v = uv_read_local_mmr(UVH_RH_GAM_CONFIG_MMR ); m_val = m_n_config.s.m_skt; n_val = m_n_config.s.n_skt; - mmioh.v = uv_read_local_mmr(UVH_RH_GAM_MMIOH_OVERLAY_CONFIG_MMR); - n_io = is_uv1_hub() ? mmioh.s1.n_io : mmioh.s2.n_io; + pnode_mask = (1 << n_val) - 1; + n_lshift = get_n_lshift(m_val); mmr_base = uv_read_local_mmr(UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR) & ~UV_MMR_ENABLE; - pnode_mask = (1 << n_val) - 1; - pnode_io_mask = (1 << n_io) - 1; node_id.v = uv_read_local_mmr(UVH_NODE_ID); gnode_extra = (node_id.s.node_id & ~((1 << n_val) - 1)) >> 1; gnode_upper = ((unsigned long)gnode_extra << m_val); - printk(KERN_INFO "UV: N %d, M %d, N_IO: %d, gnode_upper 0x%lx, gnode_extra 0x%x, pnode_mask 0x%x, pnode_io_mask 0x%x\n", - n_val, m_val, n_io, gnode_upper, gnode_extra, pnode_mask, pnode_io_mask); + pr_info("UV: N:%d M:%d pnode_mask:0x%x gnode_upper/extra:0x%lx/0x%x n_lshift 0x%x\n", + n_val, m_val, pnode_mask, gnode_upper, gnode_extra, + n_lshift); - printk(KERN_DEBUG "UV: global MMR base 0x%lx\n", mmr_base); + pr_info("UV: global MMR base 0x%lx\n", mmr_base); for(i = 0; i < UVH_NODE_PRESENT_TABLE_DEPTH; i++) uv_possible_blades += hweight64(uv_read_local_mmr( UVH_NODE_PRESENT_TABLE + i * 8)); /* uv_num_possible_blades() is really the hub count */ - printk(KERN_INFO "UV: Found %d blades, %d hubs\n", + pr_info("UV: Found %d blades, %d hubs\n", is_uv1_hub() ? uv_num_possible_blades() : (uv_num_possible_blades() + 1) / 2, uv_num_possible_blades()); @@ -806,6 +929,7 @@ void __init uv_system_init(void) uv_blade_info[blade].nr_possible_cpus = 0; uv_blade_info[blade].nr_online_cpus = 0; spin_lock_init(&uv_blade_info[blade].nmi_lock); + min_pnode = min(pnode, min_pnode); max_pnode = max(pnode, max_pnode); blade++; } @@ -828,8 +952,7 @@ void __init uv_system_init(void) uv_cpu_hub_info(cpu)->hub_revision = uv_hub_info->hub_revision; uv_cpu_hub_info(cpu)->m_shift = 64 - m_val; - uv_cpu_hub_info(cpu)->n_lshift = is_uv2_1_hub() ? - (m_val == 40 ? 40 : 39) : m_val; + uv_cpu_hub_info(cpu)->n_lshift = n_lshift; pnode = uv_apicid_to_pnode(apicid); blade = boot_pnode_to_blade(pnode); @@ -868,11 +991,11 @@ void __init uv_system_init(void) map_gru_high(max_pnode); map_mmr_high(max_pnode); - map_mmioh_high(max_pnode & pnode_io_mask); + map_mmioh_high(min_pnode, max_pnode); + uv_nmi_setup(); uv_cpu_init(); uv_scir_register_cpu_notifier(); - uv_register_nmi_notifier(); proc_mkdir("sgi_uv", NULL); /* register Legacy VGA I/O redirection handler */ diff --git a/arch/x86/kernel/apm_32.c b/arch/x86/kernel/apm_32.c index f76623cbe26..58487445141 100644 --- a/arch/x86/kernel/apm_32.c +++ b/arch/x86/kernel/apm_32.c @@ -201,6 +201,8 @@ * http://www.microsoft.com/whdc/archive/amp_12.mspx] */ +#define pr_fmt(fmt) "apm: " fmt + #include <linux/module.h> #include <linux/poll.h> @@ -230,8 +232,8 @@ #include <linux/acpi.h> #include <linux/syscore_ops.h> #include <linux/i8253.h> +#include <linux/cpuidle.h> -#include <asm/system.h> #include <asm/uaccess.h> #include <asm/desc.h> #include <asm/olpc.h> @@ -359,24 +361,44 @@ struct apm_user { * idle percentage above which bios idle calls are done */ #ifdef CONFIG_APM_CPU_IDLE -#warning deprecated CONFIG_APM_CPU_IDLE will be deleted in 2012 #define DEFAULT_IDLE_THRESHOLD 95 #else #define DEFAULT_IDLE_THRESHOLD 100 #endif #define DEFAULT_IDLE_PERIOD (100 / 3) +static int apm_cpu_idle(struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index); + +static struct cpuidle_driver apm_idle_driver = { + .name = "apm_idle", + .owner = THIS_MODULE, + .states = { + { /* entry 0 is for polling */ }, + { /* entry 1 is for APM idle */ + .name = "APM", + .desc = "APM idle", + .flags = CPUIDLE_FLAG_TIME_VALID, + .exit_latency = 250, /* WAG */ + .target_residency = 500, /* WAG */ + .enter = &apm_cpu_idle + }, + }, + .state_count = 2, +}; + +static struct cpuidle_device apm_cpuidle_device; + /* * Local variables */ -static struct { +__visible struct { unsigned long offset; unsigned short segment; } apm_bios_entry; static int clock_slowed; static int idle_threshold __read_mostly = DEFAULT_IDLE_THRESHOLD; static int idle_period __read_mostly = DEFAULT_IDLE_PERIOD; -static int set_pm_idle; static int suspends_pending; static int standbys_pending; static int ignore_sys_suspend; @@ -486,11 +508,11 @@ static void apm_error(char *str, int err) if (error_table[i].key == err) break; if (i < ERROR_COUNT) - printk(KERN_NOTICE "apm: %s: %s\n", str, error_table[i].msg); + pr_notice("%s: %s\n", str, error_table[i].msg); else if (err < 0) - printk(KERN_NOTICE "apm: %s: linux error code %i\n", str, err); + pr_notice("%s: linux error code %i\n", str, err); else - printk(KERN_NOTICE "apm: %s: unknown error code %#2.2x\n", + pr_notice("%s: unknown error code %#2.2x\n", str, err); } @@ -819,24 +841,12 @@ static int apm_do_idle(void) u32 eax; u8 ret = 0; int idled = 0; - int polling; int err = 0; - polling = !!(current_thread_info()->status & TS_POLLING); - if (polling) { - current_thread_info()->status &= ~TS_POLLING; - /* - * TS_POLLING-cleared state must be visible before we - * test NEED_RESCHED: - */ - smp_mb(); - } if (!need_resched()) { idled = 1; ret = apm_bios_call_simple(APM_FUNC_IDLE, 0, 0, &eax, &err); } - if (polling) - current_thread_info()->status |= TS_POLLING; if (!idled) return 0; @@ -883,8 +893,6 @@ static void apm_do_busy(void) #define IDLE_CALC_LIMIT (HZ * 100) #define IDLE_LEAKY_MAX 16 -static void (*original_pm_idle)(void) __read_mostly; - /** * apm_cpu_idle - cpu idling for APM capable Linux * @@ -893,35 +901,36 @@ static void (*original_pm_idle)(void) __read_mostly; * Furthermore it calls the system default idle routine. */ -static void apm_cpu_idle(void) +static int apm_cpu_idle(struct cpuidle_device *dev, + struct cpuidle_driver *drv, int index) { static int use_apm_idle; /* = 0 */ static unsigned int last_jiffies; /* = 0 */ static unsigned int last_stime; /* = 0 */ + cputime_t stime; int apm_idle_done = 0; unsigned int jiffies_since_last_check = jiffies - last_jiffies; unsigned int bucket; - WARN_ONCE(1, "deprecated apm_cpu_idle will be deleted in 2012"); recalc: + task_cputime(current, NULL, &stime); if (jiffies_since_last_check > IDLE_CALC_LIMIT) { use_apm_idle = 0; - last_jiffies = jiffies; - last_stime = current->stime; } else if (jiffies_since_last_check > idle_period) { unsigned int idle_percentage; - idle_percentage = current->stime - last_stime; + idle_percentage = stime - last_stime; idle_percentage *= 100; idle_percentage /= jiffies_since_last_check; use_apm_idle = (idle_percentage > idle_threshold); if (apm_info.forbid_idle) use_apm_idle = 0; - last_jiffies = jiffies; - last_stime = current->stime; } + last_jiffies = jiffies; + last_stime = stime; + bucket = IDLE_LEAKY_MAX; while (!need_resched()) { @@ -949,10 +958,7 @@ recalc: break; } } - if (original_pm_idle) - original_pm_idle(); - else - default_idle(); + default_idle(); local_irq_disable(); jiffies_since_last_check = jiffies - last_jiffies; if (jiffies_since_last_check > idle_period) @@ -962,7 +968,7 @@ recalc: if (apm_idle_done) apm_do_busy(); - local_irq_enable(); + return index; } /** @@ -1185,7 +1191,7 @@ static void queue_event(apm_event_t event, struct apm_user *sender) static int notified; if (notified++ == 0) - printk(KERN_ERR "apm: an event queue overflowed\n"); + pr_err("an event queue overflowed\n"); if (++as->event_tail >= APM_MAX_EVENTS) as->event_tail = 0; } @@ -1234,8 +1240,7 @@ static int suspend(int vetoable) struct apm_user *as; dpm_suspend_start(PMSG_SUSPEND); - - dpm_suspend_noirq(PMSG_SUSPEND); + dpm_suspend_end(PMSG_SUSPEND); local_irq_disable(); syscore_suspend(); @@ -1259,9 +1264,9 @@ static int suspend(int vetoable) syscore_resume(); local_irq_enable(); - dpm_resume_noirq(PMSG_RESUME); - + dpm_resume_start(PMSG_RESUME); dpm_resume_end(PMSG_RESUME); + queue_event(APM_NORMAL_RESUME, NULL); spin_lock(&user_list_lock); for (as = user_list; as != NULL; as = as->next) { @@ -1277,7 +1282,7 @@ static void standby(void) { int err; - dpm_suspend_noirq(PMSG_SUSPEND); + dpm_suspend_end(PMSG_SUSPEND); local_irq_disable(); syscore_suspend(); @@ -1291,7 +1296,7 @@ static void standby(void) syscore_resume(); local_irq_enable(); - dpm_resume_noirq(PMSG_RESUME); + dpm_resume_start(PMSG_RESUME); } static apm_event_t get_event(void) @@ -1449,7 +1454,7 @@ static void apm_mainloop(void) static int check_apm_user(struct apm_user *as, const char *func) { if (as == NULL || as->magic != APM_BIOS_MAGIC) { - printk(KERN_ERR "apm: %s passed bad filp\n", func); + pr_err("%s passed bad filp\n", func); return 1; } return 0; @@ -1588,7 +1593,7 @@ static int do_release(struct inode *inode, struct file *filp) as1 = as1->next) ; if (as1 == NULL) - printk(KERN_ERR "apm: filp not in user list\n"); + pr_err("filp not in user list\n"); else as1->next = as->next; } @@ -1602,11 +1607,9 @@ static int do_open(struct inode *inode, struct file *filp) struct apm_user *as; as = kmalloc(sizeof(*as), GFP_KERNEL); - if (as == NULL) { - printk(KERN_ERR "apm: cannot allocate struct of size %d bytes\n", - sizeof(*as)); + if (as == NULL) return -ENOMEM; - } + as->magic = APM_BIOS_MAGIC; as->event_tail = as->event_head = 0; as->suspends_pending = as->standbys_pending = 0; @@ -2315,16 +2318,16 @@ static int __init apm_init(void) } if (apm_info.disabled) { - printk(KERN_NOTICE "apm: disabled on user request.\n"); + pr_notice("disabled on user request.\n"); return -ENODEV; } if ((num_online_cpus() > 1) && !power_off && !smp) { - printk(KERN_NOTICE "apm: disabled - APM is not SMP safe.\n"); + pr_notice("disabled - APM is not SMP safe.\n"); apm_info.disabled = 1; return -ENODEV; } if (!acpi_disabled) { - printk(KERN_NOTICE "apm: overridden by ACPI.\n"); + pr_notice("overridden by ACPI.\n"); apm_info.disabled = 1; return -ENODEV; } @@ -2358,8 +2361,7 @@ static int __init apm_init(void) kapmd_task = kthread_create(apm, NULL, "kapmd"); if (IS_ERR(kapmd_task)) { - printk(KERN_ERR "apm: disabled - Unable to start kernel " - "thread.\n"); + pr_err("disabled - Unable to start kernel thread\n"); err = PTR_ERR(kapmd_task); kapmd_task = NULL; remove_proc_entry("apm", NULL); @@ -2384,9 +2386,9 @@ static int __init apm_init(void) if (HZ != 100) idle_period = (idle_period * HZ) / 100; if (idle_threshold < 100) { - original_pm_idle = pm_idle; - pm_idle = apm_cpu_idle; - set_pm_idle = 1; + if (!cpuidle_register_driver(&apm_idle_driver)) + if (cpuidle_register_device(&apm_cpuidle_device)) + cpuidle_unregister_driver(&apm_idle_driver); } return 0; @@ -2396,15 +2398,9 @@ static void __exit apm_exit(void) { int error; - if (set_pm_idle) { - pm_idle = original_pm_idle; - /* - * We are about to unload the current idle thread pm callback - * (pm_idle), Wait for all processors to update cached/local - * copies of pm_idle before proceeding. - */ - cpu_idle_wait(); - } + cpuidle_unregister_device(&apm_cpuidle_device); + cpuidle_unregister_driver(&apm_idle_driver); + if (((apm_info.bios.flags & APM_BIOS_DISENGAGED) == 0) && (apm_info.connection_version > 0x0100)) { error = apm_engage_power_management(APM_DEVICE_ALL, 0); diff --git a/arch/x86/kernel/asm-offsets.c b/arch/x86/kernel/asm-offsets.c index 68de2dc962e..9f6b9341950 100644 --- a/arch/x86/kernel/asm-offsets.c +++ b/arch/x86/kernel/asm-offsets.c @@ -32,7 +32,6 @@ void common(void) { OFFSET(TI_flags, thread_info, flags); OFFSET(TI_status, thread_info, status); OFFSET(TI_addr_limit, thread_info, addr_limit); - OFFSET(TI_preempt_count, thread_info, preempt_count); BLANK(); OFFSET(crypto_tfm_ctx_offset, crypto_tfm, __crt_ctx); @@ -69,4 +68,7 @@ void common(void) { OFFSET(BP_kernel_alignment, boot_params, hdr.kernel_alignment); OFFSET(BP_pref_address, boot_params, hdr.pref_address); OFFSET(BP_code32_start, boot_params, hdr.code32_start); + + BLANK(); + DEFINE(PTREGS_SIZE, sizeof(struct pt_regs)); } diff --git a/arch/x86/kernel/asm-offsets_32.c b/arch/x86/kernel/asm-offsets_32.c index 85d98ab15cd..d67c4be3e8b 100644 --- a/arch/x86/kernel/asm-offsets_32.c +++ b/arch/x86/kernel/asm-offsets_32.c @@ -28,7 +28,6 @@ void foo(void) OFFSET(CPUINFO_x86_vendor, cpuinfo_x86, x86_vendor); OFFSET(CPUINFO_x86_model, cpuinfo_x86, x86_model); OFFSET(CPUINFO_x86_mask, cpuinfo_x86, x86_mask); - OFFSET(CPUINFO_hard_math, cpuinfo_x86, hard_math); OFFSET(CPUINFO_cpuid_level, cpuinfo_x86, cpuid_level); OFFSET(CPUINFO_x86_capability, cpuinfo_x86, x86_capability); OFFSET(CPUINFO_x86_vendor_id, cpuinfo_x86, x86_vendor_id); @@ -60,6 +59,9 @@ void foo(void) OFFSET(IA32_RT_SIGFRAME_sigcontext, rt_sigframe, uc.uc_mcontext); BLANK(); + OFFSET(saved_context_gdt_desc, saved_context, gdt_desc); + BLANK(); + /* Offset from the sysenter stack to tss.sp0 */ DEFINE(TSS_sysenter_sp0, offsetof(struct tss_struct, x86_tss.sp0) - sizeof(struct tss_struct)); diff --git a/arch/x86/kernel/asm-offsets_64.c b/arch/x86/kernel/asm-offsets_64.c index 834e897b1e2..e7c798b354f 100644 --- a/arch/x86/kernel/asm-offsets_64.c +++ b/arch/x86/kernel/asm-offsets_64.c @@ -1,6 +1,12 @@ #include <asm/ia32.h> #define __SYSCALL_64(nr, sym, compat) [nr] = 1, +#define __SYSCALL_COMMON(nr, sym, compat) [nr] = 1, +#ifdef CONFIG_X86_X32_ABI +# define __SYSCALL_X32(nr, sym, compat) [nr] = 1, +#else +# define __SYSCALL_X32(nr, sym, compat) /* nothing */ +#endif static char syscalls_64[] = { #include <asm/syscalls_64.h> }; @@ -67,6 +73,7 @@ int main(void) ENTRY(cr3); ENTRY(cr4); ENTRY(cr8); + ENTRY(gdt_desc); BLANK(); #undef ENTRY diff --git a/arch/x86/kernel/check.c b/arch/x86/kernel/check.c index 5da1269e8dd..83a7995625a 100644 --- a/arch/x86/kernel/check.c +++ b/arch/x86/kernel/check.c @@ -27,21 +27,29 @@ static int num_scan_areas; static __init int set_corruption_check(char *arg) { - char *end; + ssize_t ret; + unsigned long val; - memory_corruption_check = simple_strtol(arg, &end, 10); + ret = kstrtoul(arg, 10, &val); + if (ret) + return ret; - return (*end == 0) ? 0 : -EINVAL; + memory_corruption_check = val; + return 0; } early_param("memory_corruption_check", set_corruption_check); static __init int set_corruption_check_period(char *arg) { - char *end; + ssize_t ret; + unsigned long val; - corruption_check_period = simple_strtoul(arg, &end, 10); + ret = kstrtoul(arg, 10, &val); + if (ret) + return ret; - return (*end == 0) ? 0 : -EINVAL; + corruption_check_period = val; + return 0; } early_param("memory_corruption_check_period", set_corruption_check_period); @@ -83,7 +91,7 @@ void __init setup_bios_corruption_check(void) corruption_check_size = round_up(corruption_check_size, PAGE_SIZE); - for_each_free_mem_range(i, MAX_NUMNODES, &start, &end, NULL) { + for_each_free_mem_range(i, NUMA_NO_NODE, &start, &end, NULL) { start = clamp_t(phys_addr_t, round_up(start, PAGE_SIZE), PAGE_SIZE, corruption_check_size); end = clamp_t(phys_addr_t, round_down(end, PAGE_SIZE), diff --git a/arch/x86/kernel/cpu/Makefile b/arch/x86/kernel/cpu/Makefile index 25f24dccdcf..7fd54f09b01 100644 --- a/arch/x86/kernel/cpu/Makefile +++ b/arch/x86/kernel/cpu/Makefile @@ -14,8 +14,8 @@ CFLAGS_common.o := $(nostackp) obj-y := intel_cacheinfo.o scattered.o topology.o obj-y += proc.o capflags.o powerflags.o common.o -obj-y += vmware.o hypervisor.o sched.o mshyperv.o obj-y += rdrand.o +obj-y += match.o obj-$(CONFIG_X86_32) += bugs.o obj-$(CONFIG_X86_64) += bugs_64.o @@ -30,20 +30,29 @@ obj-$(CONFIG_CPU_SUP_UMC_32) += umc.o obj-$(CONFIG_PERF_EVENTS) += perf_event.o ifdef CONFIG_PERF_EVENTS -obj-$(CONFIG_CPU_SUP_AMD) += perf_event_amd.o -obj-$(CONFIG_CPU_SUP_INTEL) += perf_event_p6.o perf_event_p4.o perf_event_intel_lbr.o perf_event_intel_ds.o perf_event_intel.o +obj-$(CONFIG_CPU_SUP_AMD) += perf_event_amd.o perf_event_amd_uncore.o +ifdef CONFIG_AMD_IOMMU +obj-$(CONFIG_CPU_SUP_AMD) += perf_event_amd_iommu.o endif +obj-$(CONFIG_CPU_SUP_INTEL) += perf_event_p6.o perf_event_knc.o perf_event_p4.o +obj-$(CONFIG_CPU_SUP_INTEL) += perf_event_intel_lbr.o perf_event_intel_ds.o perf_event_intel.o +obj-$(CONFIG_CPU_SUP_INTEL) += perf_event_intel_uncore.o perf_event_intel_rapl.o +endif + obj-$(CONFIG_X86_MCE) += mcheck/ obj-$(CONFIG_MTRR) += mtrr/ +obj-$(CONFIG_MICROCODE) += microcode/ obj-$(CONFIG_X86_LOCAL_APIC) += perfctr-watchdog.o perf_event_amd_ibs.o +obj-$(CONFIG_HYPERVISOR_GUEST) += vmware.o hypervisor.o mshyperv.o + quiet_cmd_mkcapflags = MKCAP $@ - cmd_mkcapflags = $(PERL) $(srctree)/$(src)/mkcapflags.pl $< $@ + cmd_mkcapflags = $(CONFIG_SHELL) $(srctree)/$(src)/mkcapflags.sh $< $@ cpufeature = $(src)/../../include/asm/cpufeature.h targets += capflags.c -$(obj)/capflags.c: $(cpufeature) $(src)/mkcapflags.pl FORCE +$(obj)/capflags.c: $(cpufeature) $(src)/mkcapflags.sh FORCE $(call if_changed,mkcapflags) diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c index f4773f4aae3..ce8b8ff0e0e 100644 --- a/arch/x86/kernel/cpu/amd.c +++ b/arch/x86/kernel/cpu/amd.c @@ -1,23 +1,55 @@ #include <linux/export.h> -#include <linux/init.h> #include <linux/bitops.h> #include <linux/elf.h> #include <linux/mm.h> #include <linux/io.h> +#include <linux/sched.h> #include <asm/processor.h> #include <asm/apic.h> #include <asm/cpu.h> #include <asm/pci-direct.h> #ifdef CONFIG_X86_64 -# include <asm/numa_64.h> # include <asm/mmconfig.h> # include <asm/cacheflush.h> #endif #include "cpu.h" +static inline int rdmsrl_amd_safe(unsigned msr, unsigned long long *p) +{ + u32 gprs[8] = { 0 }; + int err; + + WARN_ONCE((boot_cpu_data.x86 != 0xf), + "%s should only be used on K8!\n", __func__); + + gprs[1] = msr; + gprs[7] = 0x9c5a203a; + + err = rdmsr_safe_regs(gprs); + + *p = gprs[0] | ((u64)gprs[2] << 32); + + return err; +} + +static inline int wrmsrl_amd_safe(unsigned msr, unsigned long long val) +{ + u32 gprs[8] = { 0 }; + + WARN_ONCE((boot_cpu_data.x86 != 0xf), + "%s should only be used on K8!\n", __func__); + + gprs[0] = (u32)val; + gprs[1] = msr; + gprs[2] = val >> 32; + gprs[7] = 0x9c5a203a; + + return wrmsr_safe_regs(gprs); +} + #ifdef CONFIG_X86_32 /* * B step AMD K6 before B 9730xxxx have hardware bugs that can cause @@ -25,17 +57,18 @@ * contact AMD for precise details and a CPU swap. * * See http://www.multimania.com/poulot/k6bug.html - * http://www.amd.com/K6/k6docs/revgd.html + * and section 2.6.2 of "AMD-K6 Processor Revision Guide - Model 6" + * (Publication # 21266 Issue Date: August 1998) * * The following test is erm.. interesting. AMD neglected to up * the chip setting when fixing the bug but they also tweaked some * performance at the same time.. */ -extern void vide(void); -__asm__(".align 4\nvide: ret"); +extern __visible void vide(void); +__asm__(".globl vide\n\t.align 4\nvide: ret"); -static void __cpuinit init_amd_k5(struct cpuinfo_x86 *c) +static void init_amd_k5(struct cpuinfo_x86 *c) { /* * General Systems BIOSen alias the cpu frequency registers @@ -53,10 +86,10 @@ static void __cpuinit init_amd_k5(struct cpuinfo_x86 *c) } -static void __cpuinit init_amd_k6(struct cpuinfo_x86 *c) +static void init_amd_k6(struct cpuinfo_x86 *c) { u32 l, h; - int mbytes = num_physpages >> (20-PAGE_SHIFT); + int mbytes = get_num_physpages() >> (20-PAGE_SHIFT); if (c->x86_model < 6) { /* Based on AMD doc 20734R - June 2000 */ @@ -93,7 +126,6 @@ static void __cpuinit init_amd_k6(struct cpuinfo_x86 *c) "system stability may be impaired when more than 32 MB are used.\n"); else printk(KERN_CONT "probably OK (after B9730xxxx).\n"); - printk(KERN_INFO "Please see http://membres.lycos.fr/poulot/k6bug.html\n"); } /* K6 with old style WHCR */ @@ -146,7 +178,7 @@ static void __cpuinit init_amd_k6(struct cpuinfo_x86 *c) } } -static void __cpuinit amd_k7_smp_check(struct cpuinfo_x86 *c) +static void amd_k7_smp_check(struct cpuinfo_x86 *c) { /* calling is from identify_secondary_cpu() ? */ if (!c->cpu_index) @@ -159,11 +191,11 @@ static void __cpuinit amd_k7_smp_check(struct cpuinfo_x86 *c) /* Athlon 660/661 is valid. */ if ((c->x86_model == 6) && ((c->x86_mask == 0) || (c->x86_mask == 1))) - goto valid_k7; + return; /* Duron 670 is valid */ if ((c->x86_model == 7) && (c->x86_mask == 0)) - goto valid_k7; + return; /* * Athlon 662, Duron 671, and Athlon >model 7 have capability @@ -176,7 +208,7 @@ static void __cpuinit amd_k7_smp_check(struct cpuinfo_x86 *c) ((c->x86_model == 7) && (c->x86_mask >= 1)) || (c->x86_model > 7)) if (cpu_has_mp) - goto valid_k7; + return; /* If we get here, not a certified SMP capable AMD system. */ @@ -186,14 +218,10 @@ static void __cpuinit amd_k7_smp_check(struct cpuinfo_x86 *c) */ WARN_ONCE(1, "WARNING: This combination of AMD" " processors is not suitable for SMP.\n"); - if (!test_taint(TAINT_UNSAFE_SMP)) - add_taint(TAINT_UNSAFE_SMP); - -valid_k7: - ; + add_taint(TAINT_CPU_OUT_OF_SPEC, LOCKDEP_NOW_UNRELIABLE); } -static void __cpuinit init_amd_k7(struct cpuinfo_x86 *c) +static void init_amd_k7(struct cpuinfo_x86 *c) { u32 l, h; @@ -205,9 +233,7 @@ static void __cpuinit init_amd_k7(struct cpuinfo_x86 *c) if (c->x86_model >= 6 && c->x86_model <= 10) { if (!cpu_has(c, X86_FEATURE_XMM)) { printk(KERN_INFO "Enabling disabled K7/SSE Support.\n"); - rdmsr(MSR_K7_HWCR, l, h); - l &= ~0x00008000; - wrmsr(MSR_K7_HWCR, l, h); + msr_clear_bit(MSR_K7_HWCR, 15); set_cpu_cap(c, X86_FEATURE_XMM); } } @@ -238,7 +264,7 @@ static void __cpuinit init_amd_k7(struct cpuinfo_x86 *c) * To workaround broken NUMA config. Read the comment in * srat_detect_node(). */ -static int __cpuinit nearby_node(int apicid) +static int nearby_node(int apicid) { int i, node; @@ -263,14 +289,14 @@ static int __cpuinit nearby_node(int apicid) * (2) AMD processors supporting compute units */ #ifdef CONFIG_X86_HT -static void __cpuinit amd_get_topology(struct cpuinfo_x86 *c) +static void amd_get_topology(struct cpuinfo_x86 *c) { u32 nodes, cores_per_cu = 1; u8 node_id; int cpu = smp_processor_id(); /* get information required for multi-node processors */ - if (cpu_has(c, X86_FEATURE_TOPOEXT)) { + if (cpu_has_topoext) { u32 eax, ebx, ecx, edx; cpuid(0x8000001e, &eax, &ebx, &ecx, &edx); @@ -310,10 +336,10 @@ static void __cpuinit amd_get_topology(struct cpuinfo_x86 *c) #endif /* - * On a AMD dual core setup the lower bits of the APIC id distingush the cores. + * On a AMD dual core setup the lower bits of the APIC id distinguish the cores. * Assumes number of cores is a power of two. */ -static void __cpuinit amd_detect_cmp(struct cpuinfo_x86 *c) +static void amd_detect_cmp(struct cpuinfo_x86 *c) { #ifdef CONFIG_X86_HT unsigned bits; @@ -330,9 +356,9 @@ static void __cpuinit amd_detect_cmp(struct cpuinfo_x86 *c) #endif } -int amd_get_nb_id(int cpu) +u16 amd_get_nb_id(int cpu) { - int id = 0; + u16 id = 0; #ifdef CONFIG_SMP id = per_cpu(cpu_llc_id, cpu); #endif @@ -340,7 +366,7 @@ int amd_get_nb_id(int cpu) } EXPORT_SYMBOL_GPL(amd_get_nb_id); -static void __cpuinit srat_detect_node(struct cpuinfo_x86 *c) +static void srat_detect_node(struct cpuinfo_x86 *c) { #ifdef CONFIG_NUMA int cpu = smp_processor_id(); @@ -352,10 +378,11 @@ static void __cpuinit srat_detect_node(struct cpuinfo_x86 *c) node = per_cpu(cpu_llc_id, cpu); /* - * If core numbers are inconsistent, it's likely a multi-fabric platform, - * so invoke platform-specific handler + * On multi-fabric platform (e.g. Numascale NumaChip) a + * platform-specific handler needs to be called to fixup some + * IDs of the CPU. */ - if (c->phys_proc_id != node) + if (x86_cpuinit.fixup_cpu_id) x86_cpuinit.fixup_cpu_id(c, node); if (!node_online(node)) { @@ -391,7 +418,7 @@ static void __cpuinit srat_detect_node(struct cpuinfo_x86 *c) #endif } -static void __cpuinit early_init_amd_mc(struct cpuinfo_x86 *c) +static void early_init_amd_mc(struct cpuinfo_x86 *c) { #ifdef CONFIG_X86_HT unsigned bits, ecx; @@ -417,7 +444,7 @@ static void __cpuinit early_init_amd_mc(struct cpuinfo_x86 *c) #endif } -static void __cpuinit bsp_init_amd(struct cpuinfo_x86 *c) +static void bsp_init_amd(struct cpuinfo_x86 *c) { if (cpu_has(c, X86_FEATURE_CONSTANT_TSC)) { @@ -445,7 +472,7 @@ static void __cpuinit bsp_init_amd(struct cpuinfo_x86 *c) } } -static void __cpuinit early_init_amd(struct cpuinfo_x86 *c) +static void early_init_amd(struct cpuinfo_x86 *c) { early_init_amd_mc(c); @@ -456,6 +483,8 @@ static void __cpuinit early_init_amd(struct cpuinfo_x86 *c) if (c->x86_power & (1 << 8)) { set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC); + if (!check_tsc_unstable()) + set_sched_clock_stable(); } #ifdef CONFIG_X86_64 @@ -476,15 +505,22 @@ static void __cpuinit early_init_amd(struct cpuinfo_x86 *c) set_cpu_cap(c, X86_FEATURE_EXTD_APICID); } #endif + + /* F16h erratum 793, CVE-2013-6885 */ + if (c->x86 == 0x16 && c->x86_model <= 0xf) + msr_set_bit(MSR_AMD64_LS_CFG, 15); } -static void __cpuinit init_amd(struct cpuinfo_x86 *c) +static const int amd_erratum_383[]; +static const int amd_erratum_400[]; +static bool cpu_has_amd_erratum(struct cpuinfo_x86 *cpu, const int *erratum); + +static void init_amd(struct cpuinfo_x86 *c) { u32 dummy; - -#ifdef CONFIG_SMP unsigned long long value; +#ifdef CONFIG_SMP /* * Disable TLB flush filter by setting HWCR.FFDIS on K8 * bit 6 of msr C001_0015 @@ -492,11 +528,8 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) * Errata 63 for SH-B3 steppings * Errata 122 for all steppings (F+ have it disabled by default) */ - if (c->x86 == 0xf) { - rdmsrl(MSR_K7_HWCR, value); - value |= 1 << 6; - wrmsrl(MSR_K7_HWCR, value); - } + if (c->x86 == 0xf) + msr_set_bit(MSR_K7_HWCR, 6); #endif early_init_amd(c); @@ -522,12 +555,10 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) * (AMD Erratum #110, docId: 25759). */ if (c->x86_model < 0x14 && cpu_has(c, X86_FEATURE_LAHF_LM)) { - u64 val; - clear_cpu_cap(c, X86_FEATURE_LAHF_LM); - if (!rdmsrl_amd_safe(0xc001100d, &val)) { - val &= ~(1ULL << 32); - wrmsrl_amd_safe(0xc001100d, val); + if (!rdmsrl_amd_safe(0xc001100d, &value)) { + value &= ~(1ULL << 32); + wrmsrl_amd_safe(0xc001100d, value); } } @@ -576,6 +607,33 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) } } + /* re-enable TopologyExtensions if switched off by BIOS */ + if ((c->x86 == 0x15) && + (c->x86_model >= 0x10) && (c->x86_model <= 0x1f) && + !cpu_has(c, X86_FEATURE_TOPOEXT)) { + + if (msr_set_bit(0xc0011005, 54) > 0) { + rdmsrl(0xc0011005, value); + if (value & BIT_64(54)) { + set_cpu_cap(c, X86_FEATURE_TOPOEXT); + pr_info(FW_INFO "CPU: Re-enabling disabled Topology Extensions Support.\n"); + } + } + } + + /* + * The way access filter has a performance penalty on some workloads. + * Disable it on the affected CPUs. + */ + if ((c->x86 == 0x15) && + (c->x86_model >= 0x02) && (c->x86_model < 0x20)) { + + if (!rdmsrl_safe(0xc0011021, &value) && !(value & 0x1E)) { + value |= 0x1E; + wrmsrl_safe(0xc0011021, value); + } + } + cpu_detect_cache_sizes(c); /* Multi core CPU? */ @@ -588,12 +646,7 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) detect_ht(c); #endif - if (c->extended_cpuid_level >= 0x80000006) { - if (cpuid_edx(0x80000006) & 0xf000) - num_cache_leaves = 4; - else - num_cache_leaves = 3; - } + init_amd_cacheinfo(c); if (c->x86 >= 0xf) set_cpu_cap(c, X86_FEATURE_K8); @@ -621,12 +674,10 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) * benefit in doing so. */ if (!rdmsrl_safe(MSR_K8_TSEG_ADDR, &tseg)) { + unsigned long pfn = tseg >> PAGE_SHIFT; + printk(KERN_DEBUG "tseg: %010llx\n", tseg); - if ((tseg>>PMD_SHIFT) < - (max_low_pfn_mapped>>(PMD_SHIFT-PAGE_SHIFT)) || - ((tseg>>PMD_SHIFT) < - (max_pfn_mapped>>(PMD_SHIFT-PAGE_SHIFT)) && - (tseg>>PMD_SHIFT) >= (1ULL<<(32 - PMD_SHIFT)))) + if (pfn_range_is_mapped(pfn, pfn + 1)) set_memory_4k((unsigned long)__va(tseg), 1); } } @@ -639,34 +690,42 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) if (c->x86 > 0x11) set_cpu_cap(c, X86_FEATURE_ARAT); - /* - * Disable GART TLB Walk Errors on Fam10h. We do this here - * because this is always needed when GART is enabled, even in a - * kernel which has no MCE support built in. - */ if (c->x86 == 0x10) { /* - * BIOS should disable GartTlbWlk Errors themself. If - * it doesn't do it here as suggested by the BKDG. + * Disable GART TLB Walk Errors on Fam10h. We do this here + * because this is always needed when GART is enabled, even in a + * kernel which has no MCE support built in. + * BIOS should disable GartTlbWlk Errors already. If + * it doesn't, do it here as suggested by the BKDG. * * Fixes: https://bugzilla.kernel.org/show_bug.cgi?id=33012 */ - u64 mask; - int err; + msr_set_bit(MSR_AMD64_MCx_MASK(4), 10); - err = rdmsrl_safe(MSR_AMD64_MCx_MASK(4), &mask); - if (err == 0) { - mask |= (1 << 10); - checking_wrmsrl(MSR_AMD64_MCx_MASK(4), mask); - } + /* + * On family 10h BIOS may not have properly enabled WC+ support, + * causing it to be converted to CD memtype. This may result in + * performance degradation for certain nested-paging guests. + * Prevent this conversion by clearing bit 24 in + * MSR_AMD64_BU_CFG2. + * + * NOTE: we want to use the _safe accessors so as not to #GP kvm + * guests on older kvm hosts. + */ + msr_clear_bit(MSR_AMD64_BU_CFG2, 24); + + if (cpu_has_amd_erratum(c, amd_erratum_383)) + set_cpu_bug(c, X86_BUG_AMD_TLB_MMATCH); } + if (cpu_has_amd_erratum(c, amd_erratum_400)) + set_cpu_bug(c, X86_BUG_AMD_APIC_C1E); + rdmsr_safe(MSR_AMD64_PATCH_LEVEL, &c->microcode, &dummy); } #ifdef CONFIG_X86_32 -static unsigned int __cpuinit amd_size_cache(struct cpuinfo_x86 *c, - unsigned int size) +static unsigned int amd_size_cache(struct cpuinfo_x86 *c, unsigned int size) { /* AMD errata T13 (order #21922) */ if ((c->x86 == 6)) { @@ -682,12 +741,68 @@ static unsigned int __cpuinit amd_size_cache(struct cpuinfo_x86 *c, } #endif -static const struct cpu_dev __cpuinitconst amd_cpu_dev = { +static void cpu_set_tlb_flushall_shift(struct cpuinfo_x86 *c) +{ + tlb_flushall_shift = 6; +} + +static void cpu_detect_tlb_amd(struct cpuinfo_x86 *c) +{ + u32 ebx, eax, ecx, edx; + u16 mask = 0xfff; + + if (c->x86 < 0xf) + return; + + if (c->extended_cpuid_level < 0x80000006) + return; + + cpuid(0x80000006, &eax, &ebx, &ecx, &edx); + + tlb_lld_4k[ENTRIES] = (ebx >> 16) & mask; + tlb_lli_4k[ENTRIES] = ebx & mask; + + /* + * K8 doesn't have 2M/4M entries in the L2 TLB so read out the L1 TLB + * characteristics from the CPUID function 0x80000005 instead. + */ + if (c->x86 == 0xf) { + cpuid(0x80000005, &eax, &ebx, &ecx, &edx); + mask = 0xff; + } + + /* Handle DTLB 2M and 4M sizes, fall back to L1 if L2 is disabled */ + if (!((eax >> 16) & mask)) + tlb_lld_2m[ENTRIES] = (cpuid_eax(0x80000005) >> 16) & 0xff; + else + tlb_lld_2m[ENTRIES] = (eax >> 16) & mask; + + /* a 4M entry uses two 2M entries */ + tlb_lld_4m[ENTRIES] = tlb_lld_2m[ENTRIES] >> 1; + + /* Handle ITLB 2M and 4M sizes, fall back to L1 if L2 is disabled */ + if (!(eax & mask)) { + /* Erratum 658 */ + if (c->x86 == 0x15 && c->x86_model <= 0x1f) { + tlb_lli_2m[ENTRIES] = 1024; + } else { + cpuid(0x80000005, &eax, &ebx, &ecx, &edx); + tlb_lli_2m[ENTRIES] = eax & 0xff; + } + } else + tlb_lli_2m[ENTRIES] = eax & mask; + + tlb_lli_4m[ENTRIES] = tlb_lli_2m[ENTRIES] >> 1; + + cpu_set_tlb_flushall_shift(c); +} + +static const struct cpu_dev amd_cpu_dev = { .c_vendor = "AMD", .c_ident = { "AuthenticAMD" }, #ifdef CONFIG_X86_32 - .c_models = { - { .vendor = X86_VENDOR_AMD, .family = 4, .model_names = + .legacy_models = { + { .family = 4, .model_names = { [3] = "486 DX/2", [7] = "486 DX/2-WB", @@ -698,9 +813,10 @@ static const struct cpu_dev __cpuinitconst amd_cpu_dev = { } }, }, - .c_size_cache = amd_size_cache, + .legacy_cache_size = amd_size_cache, #endif .c_early_init = early_init_amd, + .c_detect_tlb = cpu_detect_tlb_amd, .c_bsp_init = bsp_init_amd, .c_init = init_amd, .c_x86_vendor = X86_VENDOR_AMD, @@ -715,8 +831,7 @@ cpu_dev_register(amd_cpu_dev); * AMD_OSVW_ERRATUM() macros. The latter is intended for newer errata that * have an OSVW id assigned, which it takes as first argument. Both take a * variable number of family-specific model-stepping ranges created by - * AMD_MODEL_RANGE(). Each erratum also has to be declared as extern const - * int[] in arch/x86/include/asm/processor.h. + * AMD_MODEL_RANGE(). * * Example: * @@ -726,32 +841,28 @@ cpu_dev_register(amd_cpu_dev); * AMD_MODEL_RANGE(0x10, 0x9, 0x0, 0x9, 0x0)); */ -const int amd_erratum_400[] = +#define AMD_LEGACY_ERRATUM(...) { -1, __VA_ARGS__, 0 } +#define AMD_OSVW_ERRATUM(osvw_id, ...) { osvw_id, __VA_ARGS__, 0 } +#define AMD_MODEL_RANGE(f, m_start, s_start, m_end, s_end) \ + ((f << 24) | (m_start << 16) | (s_start << 12) | (m_end << 4) | (s_end)) +#define AMD_MODEL_RANGE_FAMILY(range) (((range) >> 24) & 0xff) +#define AMD_MODEL_RANGE_START(range) (((range) >> 12) & 0xfff) +#define AMD_MODEL_RANGE_END(range) ((range) & 0xfff) + +static const int amd_erratum_400[] = AMD_OSVW_ERRATUM(1, AMD_MODEL_RANGE(0xf, 0x41, 0x2, 0xff, 0xf), AMD_MODEL_RANGE(0x10, 0x2, 0x1, 0xff, 0xf)); -EXPORT_SYMBOL_GPL(amd_erratum_400); -const int amd_erratum_383[] = +static const int amd_erratum_383[] = AMD_OSVW_ERRATUM(3, AMD_MODEL_RANGE(0x10, 0, 0, 0xff, 0xf)); -EXPORT_SYMBOL_GPL(amd_erratum_383); -bool cpu_has_amd_erratum(const int *erratum) + +static bool cpu_has_amd_erratum(struct cpuinfo_x86 *cpu, const int *erratum) { - struct cpuinfo_x86 *cpu = __this_cpu_ptr(&cpu_info); int osvw_id = *erratum++; u32 range; u32 ms; - /* - * If called early enough that current_cpu_data hasn't been initialized - * yet, fall back to boot_cpu_data. - */ - if (cpu->x86 == 0) - cpu = &boot_cpu_data; - - if (cpu->x86_vendor != X86_VENDOR_AMD) - return false; - if (osvw_id >= 0 && osvw_id < 65536 && cpu_has(cpu, X86_FEATURE_OSVW)) { u64 osvw_len; @@ -776,5 +887,3 @@ bool cpu_has_amd_erratum(const int *erratum) return false; } - -EXPORT_SYMBOL_GPL(cpu_has_amd_erratum); diff --git a/arch/x86/kernel/cpu/bugs.c b/arch/x86/kernel/cpu/bugs.c index 46674fbb62b..03445346ee0 100644 --- a/arch/x86/kernel/cpu/bugs.c +++ b/arch/x86/kernel/cpu/bugs.c @@ -17,24 +17,6 @@ #include <asm/paravirt.h> #include <asm/alternative.h> -static int __init no_halt(char *s) -{ - WARN_ONCE(1, "\"no-hlt\" is deprecated, please use \"idle=poll\"\n"); - boot_cpu_data.hlt_works_ok = 0; - return 1; -} - -__setup("no-hlt", no_halt); - -static int __init no_387(char *s) -{ - boot_cpu_data.hard_math = 0; - write_cr0(X86_CR0_TS | X86_CR0_EM | X86_CR0_MP | read_cr0()); - return 1; -} - -__setup("no387", no_387); - static double __initdata x = 4195835.0; static double __initdata y = 3145727.0; @@ -53,22 +35,13 @@ static void __init check_fpu(void) { s32 fdiv_bug; - if (!boot_cpu_data.hard_math) { -#ifndef CONFIG_MATH_EMULATION - printk(KERN_EMERG "No coprocessor found and no math emulation present.\n"); - printk(KERN_EMERG "Giving up.\n"); - for (;;) ; -#endif - return; - } - kernel_fpu_begin(); /* * trap_init() enabled FXSR and company _before_ testing for FP * problems here. * - * Test for the divl bug.. + * Test for the divl bug: http://en.wikipedia.org/wiki/Fdiv_bug */ __asm__("fninit\n\t" "fldl %1\n\t" @@ -84,91 +57,38 @@ static void __init check_fpu(void) kernel_fpu_end(); - boot_cpu_data.fdiv_bug = fdiv_bug; - if (boot_cpu_data.fdiv_bug) - printk(KERN_WARNING "Hmm, FPU with FDIV bug.\n"); -} - -static void __init check_hlt(void) -{ - if (boot_cpu_data.x86 >= 5 || paravirt_enabled()) - return; - - printk(KERN_INFO "Checking 'hlt' instruction... "); - if (!boot_cpu_data.hlt_works_ok) { - printk("disabled\n"); - return; + if (fdiv_bug) { + set_cpu_bug(&boot_cpu_data, X86_BUG_FDIV); + pr_warn("Hmm, FPU with FDIV bug\n"); } - halt(); - halt(); - halt(); - halt(); - printk(KERN_CONT "OK.\n"); -} - -/* - * Most 386 processors have a bug where a POPAD can lock the - * machine even from user space. - */ - -static void __init check_popad(void) -{ -#ifndef CONFIG_X86_POPAD_OK - int res, inp = (int) &res; - - printk(KERN_INFO "Checking for popad bug... "); - __asm__ __volatile__( - "movl $12345678,%%eax; movl $0,%%edi; pusha; popa; movl (%%edx,%%edi),%%ecx " - : "=&a" (res) - : "d" (inp) - : "ecx", "edi"); - /* - * If this fails, it means that any user program may lock the - * CPU hard. Too bad. - */ - if (res != 12345678) - printk(KERN_CONT "Buggy.\n"); - else - printk(KERN_CONT "OK.\n"); -#endif } -/* - * Check whether we are able to run this kernel safely on SMP. - * - * - In order to run on a i386, we need to be compiled for i386 - * (for due to lack of "invlpg" and working WP on a i386) - * - In order to run on anything without a TSC, we need to be - * compiled for a i486. - */ - -static void __init check_config(void) -{ -/* - * We'd better not be a i386 if we're configured to use some - * i486+ only features! (WP works in supervisor mode and the - * new "invlpg" and "bswap" instructions) - */ -#if defined(CONFIG_X86_WP_WORKS_OK) || defined(CONFIG_X86_INVLPG) || \ - defined(CONFIG_X86_BSWAP) - if (boot_cpu_data.x86 == 3) - panic("Kernel requires i486+ for 'invlpg' and other features"); -#endif -} - - void __init check_bugs(void) { identify_boot_cpu(); #ifndef CONFIG_SMP - printk(KERN_INFO "CPU: "); + pr_info("CPU: "); print_cpu_info(&boot_cpu_data); #endif - check_config(); - check_fpu(); - check_hlt(); - check_popad(); + + /* + * Check whether we are able to run this kernel safely on SMP. + * + * - i386 is no longer supported. + * - In order to run on anything without a TSC, we need to be + * compiled for a i486. + */ + if (boot_cpu_data.x86 < 4) + panic("Kernel requires i486+ for 'invlpg' and other features"); + init_utsname()->machine[1] = '0' + (boot_cpu_data.x86 > 6 ? 6 : boot_cpu_data.x86); alternative_instructions(); + + /* + * kernel_fpu_begin/end() in check_fpu() relies on the patched + * alternative instructions. + */ + if (cpu_has_fpu) + check_fpu(); } diff --git a/arch/x86/kernel/cpu/centaur.c b/arch/x86/kernel/cpu/centaur.c index 159103c0b1f..d8fba5c15fb 100644 --- a/arch/x86/kernel/cpu/centaur.c +++ b/arch/x86/kernel/cpu/centaur.c @@ -1,6 +1,5 @@ #include <linux/bitops.h> #include <linux/kernel.h> -#include <linux/init.h> #include <asm/processor.h> #include <asm/e820.h> @@ -9,236 +8,6 @@ #include "cpu.h" -#ifdef CONFIG_X86_OOSTORE - -static u32 __cpuinit power2(u32 x) -{ - u32 s = 1; - - while (s <= x) - s <<= 1; - - return s >>= 1; -} - - -/* - * Set up an actual MCR - */ -static void __cpuinit centaur_mcr_insert(int reg, u32 base, u32 size, int key) -{ - u32 lo, hi; - - hi = base & ~0xFFF; - lo = ~(size-1); /* Size is a power of 2 so this makes a mask */ - lo &= ~0xFFF; /* Remove the ctrl value bits */ - lo |= key; /* Attribute we wish to set */ - wrmsr(reg+MSR_IDT_MCR0, lo, hi); - mtrr_centaur_report_mcr(reg, lo, hi); /* Tell the mtrr driver */ -} - -/* - * Figure what we can cover with MCR's - * - * Shortcut: We know you can't put 4Gig of RAM on a winchip - */ -static u32 __cpuinit ramtop(void) -{ - u32 clip = 0xFFFFFFFFUL; - u32 top = 0; - int i; - - for (i = 0; i < e820.nr_map; i++) { - unsigned long start, end; - - if (e820.map[i].addr > 0xFFFFFFFFUL) - continue; - /* - * Don't MCR over reserved space. Ignore the ISA hole - * we frob around that catastrophe already - */ - if (e820.map[i].type == E820_RESERVED) { - if (e820.map[i].addr >= 0x100000UL && - e820.map[i].addr < clip) - clip = e820.map[i].addr; - continue; - } - start = e820.map[i].addr; - end = e820.map[i].addr + e820.map[i].size; - if (start >= end) - continue; - if (end > top) - top = end; - } - /* - * Everything below 'top' should be RAM except for the ISA hole. - * Because of the limited MCR's we want to map NV/ACPI into our - * MCR range for gunk in RAM - * - * Clip might cause us to MCR insufficient RAM but that is an - * acceptable failure mode and should only bite obscure boxes with - * a VESA hole at 15Mb - * - * The second case Clip sometimes kicks in is when the EBDA is marked - * as reserved. Again we fail safe with reasonable results - */ - if (top > clip) - top = clip; - - return top; -} - -/* - * Compute a set of MCR's to give maximum coverage - */ -static int __cpuinit centaur_mcr_compute(int nr, int key) -{ - u32 mem = ramtop(); - u32 root = power2(mem); - u32 base = root; - u32 top = root; - u32 floor = 0; - int ct = 0; - - while (ct < nr) { - u32 fspace = 0; - u32 high; - u32 low; - - /* - * Find the largest block we will fill going upwards - */ - high = power2(mem-top); - - /* - * Find the largest block we will fill going downwards - */ - low = base/2; - - /* - * Don't fill below 1Mb going downwards as there - * is an ISA hole in the way. - */ - if (base <= 1024*1024) - low = 0; - - /* - * See how much space we could cover by filling below - * the ISA hole - */ - - if (floor == 0) - fspace = 512*1024; - else if (floor == 512*1024) - fspace = 128*1024; - - /* And forget ROM space */ - - /* - * Now install the largest coverage we get - */ - if (fspace > high && fspace > low) { - centaur_mcr_insert(ct, floor, fspace, key); - floor += fspace; - } else if (high > low) { - centaur_mcr_insert(ct, top, high, key); - top += high; - } else if (low > 0) { - base -= low; - centaur_mcr_insert(ct, base, low, key); - } else - break; - ct++; - } - /* - * We loaded ct values. We now need to set the mask. The caller - * must do this bit. - */ - return ct; -} - -static void __cpuinit centaur_create_optimal_mcr(void) -{ - int used; - int i; - - /* - * Allocate up to 6 mcrs to mark as much of ram as possible - * as write combining and weak write ordered. - * - * To experiment with: Linux never uses stack operations for - * mmio spaces so we could globally enable stack operation wc - * - * Load the registers with type 31 - full write combining, all - * writes weakly ordered. - */ - used = centaur_mcr_compute(6, 31); - - /* - * Wipe unused MCRs - */ - for (i = used; i < 8; i++) - wrmsr(MSR_IDT_MCR0+i, 0, 0); -} - -static void __cpuinit winchip2_create_optimal_mcr(void) -{ - u32 lo, hi; - int used; - int i; - - /* - * Allocate up to 6 mcrs to mark as much of ram as possible - * as write combining, weak store ordered. - * - * Load the registers with type 25 - * 8 - weak write ordering - * 16 - weak read ordering - * 1 - write combining - */ - used = centaur_mcr_compute(6, 25); - - /* - * Mark the registers we are using. - */ - rdmsr(MSR_IDT_MCR_CTRL, lo, hi); - for (i = 0; i < used; i++) - lo |= 1<<(9+i); - wrmsr(MSR_IDT_MCR_CTRL, lo, hi); - - /* - * Wipe unused MCRs - */ - - for (i = used; i < 8; i++) - wrmsr(MSR_IDT_MCR0+i, 0, 0); -} - -/* - * Handle the MCR key on the Winchip 2. - */ -static void __cpuinit winchip2_unprotect_mcr(void) -{ - u32 lo, hi; - u32 key; - - rdmsr(MSR_IDT_MCR_CTRL, lo, hi); - lo &= ~0x1C0; /* blank bits 8-6 */ - key = (lo>>17) & 7; - lo |= key<<6; /* replace with unlock key */ - wrmsr(MSR_IDT_MCR_CTRL, lo, hi); -} - -static void __cpuinit winchip2_protect_mcr(void) -{ - u32 lo, hi; - - rdmsr(MSR_IDT_MCR_CTRL, lo, hi); - lo &= ~0x1C0; /* blank bits 8-6 */ - wrmsr(MSR_IDT_MCR_CTRL, lo, hi); -} -#endif /* CONFIG_X86_OOSTORE */ - #define ACE_PRESENT (1 << 6) #define ACE_ENABLED (1 << 7) #define ACE_FCR (1 << 28) /* MSR_VIA_FCR */ @@ -247,7 +16,7 @@ static void __cpuinit winchip2_protect_mcr(void) #define RNG_ENABLED (1 << 3) #define RNG_ENABLE (1 << 6) /* MSR_VIA_RNG */ -static void __cpuinit init_c3(struct cpuinfo_x86 *c) +static void init_c3(struct cpuinfo_x86 *c) { u32 lo, hi; @@ -318,7 +87,7 @@ enum { EAMD3D = 1<<20, }; -static void __cpuinit early_init_centaur(struct cpuinfo_x86 *c) +static void early_init_centaur(struct cpuinfo_x86 *c) { switch (c->x86) { #ifdef CONFIG_X86_32 @@ -337,7 +106,7 @@ static void __cpuinit early_init_centaur(struct cpuinfo_x86 *c) #endif } -static void __cpuinit init_centaur(struct cpuinfo_x86 *c) +static void init_centaur(struct cpuinfo_x86 *c) { #ifdef CONFIG_X86_32 char *name; @@ -363,20 +132,6 @@ static void __cpuinit init_centaur(struct cpuinfo_x86 *c) fcr_clr = DPDC; printk(KERN_NOTICE "Disabling bugged TSC.\n"); clear_cpu_cap(c, X86_FEATURE_TSC); -#ifdef CONFIG_X86_OOSTORE - centaur_create_optimal_mcr(); - /* - * Enable: - * write combining on non-stack, non-string - * write combining on string, all types - * weak write ordering - * - * The C6 original lacks weak read order - * - * Note 0x120 is write only on Winchip 1 - */ - wrmsr(MSR_IDT_MCR_CTRL, 0x01F0001F, 0); -#endif break; case 8: switch (c->x86_mask) { @@ -393,40 +148,12 @@ static void __cpuinit init_centaur(struct cpuinfo_x86 *c) fcr_set = ECX8|DSMC|DTLOCK|EMMX|EBRPRED|ERETSTK| E2MMX|EAMD3D; fcr_clr = DPDC; -#ifdef CONFIG_X86_OOSTORE - winchip2_unprotect_mcr(); - winchip2_create_optimal_mcr(); - rdmsr(MSR_IDT_MCR_CTRL, lo, hi); - /* - * Enable: - * write combining on non-stack, non-string - * write combining on string, all types - * weak write ordering - */ - lo |= 31; - wrmsr(MSR_IDT_MCR_CTRL, lo, hi); - winchip2_protect_mcr(); -#endif break; case 9: name = "3"; fcr_set = ECX8|DSMC|DTLOCK|EMMX|EBRPRED|ERETSTK| E2MMX|EAMD3D; fcr_clr = DPDC; -#ifdef CONFIG_X86_OOSTORE - winchip2_unprotect_mcr(); - winchip2_create_optimal_mcr(); - rdmsr(MSR_IDT_MCR_CTRL, lo, hi); - /* - * Enable: - * write combining on non-stack, non-string - * write combining on string, all types - * weak write ordering - */ - lo |= 31; - wrmsr(MSR_IDT_MCR_CTRL, lo, hi); - winchip2_protect_mcr(); -#endif break; default: name = "??"; @@ -468,10 +195,10 @@ static void __cpuinit init_centaur(struct cpuinfo_x86 *c) #endif } -static unsigned int __cpuinit +#ifdef CONFIG_X86_32 +static unsigned int centaur_size_cache(struct cpuinfo_x86 *c, unsigned int size) { -#ifdef CONFIG_X86_32 /* VIA C3 CPUs (670-68F) need further shifting. */ if ((c->x86 == 6) && ((c->x86_model == 7) || (c->x86_model == 8))) size >>= 8; @@ -484,16 +211,18 @@ centaur_size_cache(struct cpuinfo_x86 *c, unsigned int size) if ((c->x86 == 6) && (c->x86_model == 9) && (c->x86_mask == 1) && (size == 65)) size -= 1; -#endif return size; } +#endif -static const struct cpu_dev __cpuinitconst centaur_cpu_dev = { +static const struct cpu_dev centaur_cpu_dev = { .c_vendor = "Centaur", .c_ident = { "CentaurHauls" }, .c_early_init = early_init_centaur, .c_init = init_centaur, - .c_size_cache = centaur_size_cache, +#ifdef CONFIG_X86_32 + .legacy_cache_size = centaur_size_cache, +#endif .c_x86_vendor = X86_VENDOR_CENTAUR, }; diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index d43cad74f16..ef1b93f18ed 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -8,6 +8,7 @@ #include <linux/delay.h> #include <linux/sched.h> #include <linux/init.h> +#include <linux/kprobes.h> #include <linux/kgdb.h> #include <linux/smp.h> #include <linux/io.h> @@ -18,7 +19,9 @@ #include <asm/archrandom.h> #include <asm/hypervisor.h> #include <asm/processor.h> +#include <asm/debugreg.h> #include <asm/sections.h> +#include <asm/vsyscall.h> #include <linux/topology.h> #include <linux/cpumask.h> #include <asm/pgtable.h> @@ -28,6 +31,7 @@ #include <asm/apic.h> #include <asm/desc.h> #include <asm/i387.h> +#include <asm/fpu-internal.h> #include <asm/mtrr.h> #include <linux/numa.h> #include <asm/asm.h> @@ -35,6 +39,8 @@ #include <asm/mce.h> #include <asm/msr.h> #include <asm/pat.h> +#include <asm/microcode.h> +#include <asm/microcode_intel.h> #ifdef CONFIG_X86_LOCAL_APIC #include <asm/uv/uv.h> @@ -59,7 +65,7 @@ void __init setup_cpu_local_masks(void) alloc_bootmem_cpumask_var(&cpu_sibling_setup_mask); } -static void __cpuinit default_init(struct cpuinfo_x86 *c) +static void default_init(struct cpuinfo_x86 *c) { #ifdef CONFIG_X86_64 cpu_detect_cache_sizes(c); @@ -76,13 +82,13 @@ static void __cpuinit default_init(struct cpuinfo_x86 *c) #endif } -static const struct cpu_dev __cpuinitconst default_cpu = { +static const struct cpu_dev default_cpu = { .c_init = default_init, .c_vendor = "Unknown", .c_x86_vendor = X86_VENDOR_UNKNOWN, }; -static const struct cpu_dev *this_cpu __cpuinitdata = &default_cpu; +static const struct cpu_dev *this_cpu = &default_cpu; DEFINE_PER_CPU_PAGE_ALIGNED(struct gdt_page, gdt_page) = { .gdt = { #ifdef CONFIG_X86_64 @@ -142,6 +148,8 @@ static int __init x86_xsave_setup(char *s) { setup_clear_cpu_cap(X86_FEATURE_XSAVE); setup_clear_cpu_cap(X86_FEATURE_XSAVEOPT); + setup_clear_cpu_cap(X86_FEATURE_AVX); + setup_clear_cpu_cap(X86_FEATURE_AVX2); return 1; } __setup("noxsave", x86_xsave_setup); @@ -154,8 +162,8 @@ static int __init x86_xsaveopt_setup(char *s) __setup("noxsaveopt", x86_xsaveopt_setup); #ifdef CONFIG_X86_32 -static int cachesize_override __cpuinitdata = -1; -static int disable_x86_serial_nr __cpuinitdata = 1; +static int cachesize_override = -1; +static int disable_x86_serial_nr = 1; static int __init cachesize_setup(char *str) { @@ -209,12 +217,12 @@ static inline int flag_is_changeable_p(u32 flag) } /* Probe for the CPUID instruction */ -static int __cpuinit have_cpuid_p(void) +int have_cpuid_p(void) { return flag_is_changeable_p(X86_EFLAGS_ID); } -static void __cpuinit squash_the_stupid_serial_number(struct cpuinfo_x86 *c) +static void squash_the_stupid_serial_number(struct cpuinfo_x86 *c) { unsigned long lo, hi; @@ -245,32 +253,45 @@ static inline int flag_is_changeable_p(u32 flag) { return 1; } -/* Probe for the CPUID instruction */ -static inline int have_cpuid_p(void) -{ - return 1; -} static inline void squash_the_stupid_serial_number(struct cpuinfo_x86 *c) { } #endif -static int disable_smep __cpuinitdata; static __init int setup_disable_smep(char *arg) { - disable_smep = 1; + setup_clear_cpu_cap(X86_FEATURE_SMEP); return 1; } __setup("nosmep", setup_disable_smep); -static __cpuinit void setup_smep(struct cpuinfo_x86 *c) +static __always_inline void setup_smep(struct cpuinfo_x86 *c) +{ + if (cpu_has(c, X86_FEATURE_SMEP)) + set_in_cr4(X86_CR4_SMEP); +} + +static __init int setup_disable_smap(char *arg) +{ + setup_clear_cpu_cap(X86_FEATURE_SMAP); + return 1; +} +__setup("nosmap", setup_disable_smap); + +static __always_inline void setup_smap(struct cpuinfo_x86 *c) { - if (cpu_has(c, X86_FEATURE_SMEP)) { - if (unlikely(disable_smep)) { - setup_clear_cpu_cap(X86_FEATURE_SMEP); - clear_in_cr4(X86_CR4_SMEP); - } else - set_in_cr4(X86_CR4_SMEP); + unsigned long eflags; + + /* This should have been cleared long ago */ + raw_local_save_flags(eflags); + BUG_ON(eflags & X86_EFLAGS_AC); + + if (cpu_has(c, X86_FEATURE_SMAP)) { +#ifdef CONFIG_X86_SMAP + set_in_cr4(X86_CR4_SMAP); +#else + clear_in_cr4(X86_CR4_SMAP); +#endif } } @@ -284,7 +305,7 @@ struct cpuid_dependent_feature { u32 level; }; -static const struct cpuid_dependent_feature __cpuinitconst +static const struct cpuid_dependent_feature cpuid_dependent_features[] = { { X86_FEATURE_MWAIT, 0x00000005 }, { X86_FEATURE_DCA, 0x00000009 }, @@ -292,7 +313,7 @@ cpuid_dependent_features[] = { { 0, 0 } }; -static void __cpuinit filter_cpuid_features(struct cpuinfo_x86 *c, bool warn) +static void filter_cpuid_features(struct cpuinfo_x86 *c, bool warn) { const struct cpuid_dependent_feature *df; @@ -330,9 +351,10 @@ static void __cpuinit filter_cpuid_features(struct cpuinfo_x86 *c, bool warn) */ /* Look up CPU names by table lookup. */ -static const char *__cpuinit table_lookup_model(struct cpuinfo_x86 *c) +static const char *table_lookup_model(struct cpuinfo_x86 *c) { - const struct cpu_model_info *info; +#ifdef CONFIG_X86_32 + const struct legacy_cpu_model_info *info; if (c->x86_model >= 16) return NULL; /* Range check */ @@ -340,18 +362,19 @@ static const char *__cpuinit table_lookup_model(struct cpuinfo_x86 *c) if (!this_cpu) return NULL; - info = this_cpu->c_models; + info = this_cpu->legacy_models; - while (info && info->family) { + while (info->family) { if (info->family == c->x86) return info->model_names[c->x86_model]; info++; } +#endif return NULL; /* Not found */ } -__u32 cpu_caps_cleared[NCAPINTS] __cpuinitdata; -__u32 cpu_caps_set[NCAPINTS] __cpuinitdata; +__u32 cpu_caps_cleared[NCAPINTS]; +__u32 cpu_caps_set[NCAPINTS]; void load_percpu_segment(int cpu) { @@ -380,9 +403,9 @@ void switch_to_new_gdt(int cpu) load_percpu_segment(cpu); } -static const struct cpu_dev *__cpuinitdata cpu_devs[X86_VENDOR_NUM] = {}; +static const struct cpu_dev *cpu_devs[X86_VENDOR_NUM] = {}; -static void __cpuinit get_model_name(struct cpuinfo_x86 *c) +static void get_model_name(struct cpuinfo_x86 *c) { unsigned int *v; char *p, *q; @@ -411,7 +434,7 @@ static void __cpuinit get_model_name(struct cpuinfo_x86 *c) } } -void __cpuinit cpu_detect_cache_sizes(struct cpuinfo_x86 *c) +void cpu_detect_cache_sizes(struct cpuinfo_x86 *c) { unsigned int n, dummy, ebx, ecx, edx, l2size; @@ -436,8 +459,8 @@ void __cpuinit cpu_detect_cache_sizes(struct cpuinfo_x86 *c) c->x86_tlbsize += ((ebx >> 16) & 0xfff) + (ebx & 0xfff); #else /* do processor-specific cache resizing */ - if (this_cpu->c_size_cache) - l2size = this_cpu->c_size_cache(c, l2size); + if (this_cpu->legacy_cache_size) + l2size = this_cpu->legacy_cache_size(c, l2size); /* Allow user to override all this if necessary. */ if (cachesize_override != -1) @@ -450,7 +473,37 @@ void __cpuinit cpu_detect_cache_sizes(struct cpuinfo_x86 *c) c->x86_cache_size = l2size; } -void __cpuinit detect_ht(struct cpuinfo_x86 *c) +u16 __read_mostly tlb_lli_4k[NR_INFO]; +u16 __read_mostly tlb_lli_2m[NR_INFO]; +u16 __read_mostly tlb_lli_4m[NR_INFO]; +u16 __read_mostly tlb_lld_4k[NR_INFO]; +u16 __read_mostly tlb_lld_2m[NR_INFO]; +u16 __read_mostly tlb_lld_4m[NR_INFO]; +u16 __read_mostly tlb_lld_1g[NR_INFO]; + +/* + * tlb_flushall_shift shows the balance point in replacing cr3 write + * with multiple 'invlpg'. It will do this replacement when + * flush_tlb_lines <= active_lines/2^tlb_flushall_shift. + * If tlb_flushall_shift is -1, means the replacement will be disabled. + */ +s8 __read_mostly tlb_flushall_shift = -1; + +void cpu_detect_tlb(struct cpuinfo_x86 *c) +{ + if (this_cpu->c_detect_tlb) + this_cpu->c_detect_tlb(c); + + printk(KERN_INFO "Last level iTLB entries: 4KB %d, 2MB %d, 4MB %d\n" + "Last level dTLB entries: 4KB %d, 2MB %d, 4MB %d, 1GB %d\n" + "tlb_flushall_shift: %d\n", + tlb_lli_4k[ENTRIES], tlb_lli_2m[ENTRIES], + tlb_lli_4m[ENTRIES], tlb_lld_4k[ENTRIES], + tlb_lld_2m[ENTRIES], tlb_lld_4m[ENTRIES], + tlb_lld_1g[ENTRIES], tlb_flushall_shift); +} + +void detect_ht(struct cpuinfo_x86 *c) { #ifdef CONFIG_X86_HT u32 eax, ebx, ecx, edx; @@ -501,7 +554,7 @@ out: #endif } -static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c) +static void get_cpu_vendor(struct cpuinfo_x86 *c) { char *v = c->x86_vendor_id; int i; @@ -528,7 +581,7 @@ static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c) this_cpu = &default_cpu; } -void __cpuinit cpu_detect(struct cpuinfo_x86 *c) +void cpu_detect(struct cpuinfo_x86 *c) { /* Get vendor name */ cpuid(0x00000000, (unsigned int *)&c->cpuid_level, @@ -558,7 +611,7 @@ void __cpuinit cpu_detect(struct cpuinfo_x86 *c) } } -void __cpuinit get_cpu_cap(struct cpuinfo_x86 *c) +void get_cpu_cap(struct cpuinfo_x86 *c) { u32 tfms, xlvl; u32 ebx; @@ -609,7 +662,7 @@ void __cpuinit get_cpu_cap(struct cpuinfo_x86 *c) init_scattered_cpuid_features(c); } -static void __cpuinit identify_cpu_without_cpuid(struct cpuinfo_x86 *c) +static void identify_cpu_without_cpuid(struct cpuinfo_x86 *c) { #ifdef CONFIG_X86_32 int i; @@ -668,10 +721,9 @@ static void __init early_identify_cpu(struct cpuinfo_x86 *c) return; cpu_detect(c); - get_cpu_vendor(c); - get_cpu_cap(c); + fpu_detect(c); if (this_cpu->c_early_init) this_cpu->c_early_init(c); @@ -679,10 +731,10 @@ static void __init early_identify_cpu(struct cpuinfo_x86 *c) c->cpu_index = 0; filter_cpuid_features(c, false); - setup_smep(c); - if (this_cpu->c_bsp_init) this_cpu->c_bsp_init(c); + + setup_force_cpu_cap(X86_FEATURE_ALWAYS); } void __init early_cpu_init(void) @@ -727,7 +779,7 @@ void __init early_cpu_init(void) * unless we can find a reliable way to detect all the broken cases. * Enable it explicitly on 64-bit for non-constant inputs of cpu_has(). */ -static void __cpuinit detect_nopl(struct cpuinfo_x86 *c) +static void detect_nopl(struct cpuinfo_x86 *c) { #ifdef CONFIG_X86_32 clear_cpu_cap(c, X86_FEATURE_NOPL); @@ -736,7 +788,7 @@ static void __cpuinit detect_nopl(struct cpuinfo_x86 *c) #endif } -static void __cpuinit generic_identify(struct cpuinfo_x86 *c) +static void generic_identify(struct cpuinfo_x86 *c) { c->extended_cpuid_level = 0; @@ -765,8 +817,6 @@ static void __cpuinit generic_identify(struct cpuinfo_x86 *c) c->phys_proc_id = c->initial_apicid; } - setup_smep(c); - get_model_name(c); /* Default name */ detect_nopl(c); @@ -775,7 +825,7 @@ static void __cpuinit generic_identify(struct cpuinfo_x86 *c) /* * This does the hard work of actually picking apart the CPU stuff... */ -static void __cpuinit identify_cpu(struct cpuinfo_x86 *c) +static void identify_cpu(struct cpuinfo_x86 *c) { int i; @@ -831,6 +881,10 @@ static void __cpuinit identify_cpu(struct cpuinfo_x86 *c) /* Disable the PN if appropriate */ squash_the_stupid_serial_number(c); + /* Set up SMEP/SMAP */ + setup_smep(c); + setup_smap(c); + /* * The vendor-specific functions might have changed features. * Now we do "generic changes." @@ -877,6 +931,10 @@ static void __cpuinit identify_cpu(struct cpuinfo_x86 *c) /* AND the already accumulated flags with these */ for (i = 0; i < NCAPINTS; i++) boot_cpu_data.x86_capability[i] &= c->x86_capability[i]; + + /* OR, i.e. replicate the bug flags */ + for (i = NCAPINTS; i < NCAPINTS + NBUGINTS; i++) + c->x86_capability[i] |= boot_cpu_data.x86_capability[i]; } /* Init Machine Check Exception if available. */ @@ -897,6 +955,38 @@ static void vgetcpu_set_mode(void) else vgetcpu_mode = VGETCPU_LSL; } + +/* May not be __init: called during resume */ +static void syscall32_cpu_init(void) +{ + /* Load these always in case some future AMD CPU supports + SYSENTER from compat mode too. */ + wrmsrl_safe(MSR_IA32_SYSENTER_CS, (u64)__KERNEL_CS); + wrmsrl_safe(MSR_IA32_SYSENTER_ESP, 0ULL); + wrmsrl_safe(MSR_IA32_SYSENTER_EIP, (u64)ia32_sysenter_target); + + wrmsrl(MSR_CSTAR, ia32_cstar_target); +} +#endif + +#ifdef CONFIG_X86_32 +void enable_sep_cpu(void) +{ + int cpu = get_cpu(); + struct tss_struct *tss = &per_cpu(init_tss, cpu); + + if (!boot_cpu_has(X86_FEATURE_SEP)) { + put_cpu(); + return; + } + + tss->x86_tss.ss1 = __KERNEL_CS; + tss->x86_tss.sp1 = sizeof(struct tss_struct) + (unsigned long) tss; + wrmsr(MSR_IA32_SYSENTER_CS, __KERNEL_CS, 0); + wrmsr(MSR_IA32_SYSENTER_ESP, tss->x86_tss.sp1, 0); + wrmsr(MSR_IA32_SYSENTER_EIP, (unsigned long) ia32_sysenter_target, 0); + put_cpu(); +} #endif void __init identify_boot_cpu(void) @@ -909,9 +999,10 @@ void __init identify_boot_cpu(void) #else vgetcpu_set_mode(); #endif + cpu_detect_tlb(&boot_cpu_data); } -void __cpuinit identify_secondary_cpu(struct cpuinfo_x86 *c) +void identify_secondary_cpu(struct cpuinfo_x86 *c) { BUG_ON(c == &boot_cpu_data); identify_cpu(c); @@ -926,14 +1017,14 @@ struct msr_range { unsigned max; }; -static const struct msr_range msr_range_array[] __cpuinitconst = { +static const struct msr_range msr_range_array[] = { { 0x00000000, 0x00000418}, { 0xc0000000, 0xc000040b}, { 0xc0010000, 0xc0010142}, { 0xc0011000, 0xc001103b}, }; -static void __cpuinit print_cpu_msr(void) +static void __print_cpu_msr(void) { unsigned index_min, index_max; unsigned index; @@ -945,14 +1036,14 @@ static void __cpuinit print_cpu_msr(void) index_max = msr_range_array[i].max; for (index = index_min; index < index_max; index++) { - if (rdmsrl_amd_safe(index, &val)) + if (rdmsrl_safe(index, &val)) continue; printk(KERN_INFO " MSR%08x: %016llx\n", index, val); } } } -static int show_msr __cpuinitdata; +static int show_msr; static __init int setup_show_msr(char *arg) { @@ -968,12 +1059,13 @@ __setup("show_msr=", setup_show_msr); static __init int setup_noclflush(char *arg) { - setup_clear_cpu_cap(X86_FEATURE_CLFLSH); + setup_clear_cpu_cap(X86_FEATURE_CLFLUSH); + setup_clear_cpu_cap(X86_FEATURE_CLFLUSHOPT); return 1; } __setup("noclflush", setup_noclflush); -void __cpuinit print_cpu_info(struct cpuinfo_x86 *c) +void print_cpu_info(struct cpuinfo_x86 *c) { const char *vendor = NULL; @@ -988,22 +1080,24 @@ void __cpuinit print_cpu_info(struct cpuinfo_x86 *c) printk(KERN_CONT "%s ", vendor); if (c->x86_model_id[0]) - printk(KERN_CONT "%s", c->x86_model_id); + printk(KERN_CONT "%s", strim(c->x86_model_id)); else printk(KERN_CONT "%d86", c->x86); + printk(KERN_CONT " (fam: %02x, model: %02x", c->x86, c->x86_model); + if (c->x86_mask || c->cpuid_level >= 0) - printk(KERN_CONT " stepping %02x\n", c->x86_mask); + printk(KERN_CONT ", stepping: %02x)\n", c->x86_mask); else - printk(KERN_CONT "\n"); + printk(KERN_CONT ")\n"); -#ifdef CONFIG_SMP + print_cpu_msr(c); +} + +void print_cpu_msr(struct cpuinfo_x86 *c) +{ if (c->cpu_index < show_msr) - print_cpu_msr(); -#else - if (show_msr) - print_cpu_msr(); -#endif + __print_cpu_msr(); } static __init int setup_disablecpuid(char *arg) @@ -1019,13 +1113,17 @@ static __init int setup_disablecpuid(char *arg) } __setup("clearcpuid=", setup_disablecpuid); +DEFINE_PER_CPU(unsigned long, kernel_stack) = + (unsigned long)&init_thread_union - KERNEL_STACK_OFFSET + THREAD_SIZE; +EXPORT_PER_CPU_SYMBOL(kernel_stack); + #ifdef CONFIG_X86_64 struct desc_ptr idt_descr = { NR_VECTORS * 16 - 1, (unsigned long) idt_table }; -struct desc_ptr nmi_idt_descr = { NR_VECTORS * 16 - 1, - (unsigned long) nmi_idt_table }; +struct desc_ptr debug_idt_descr = { NR_VECTORS * 16 - 1, + (unsigned long) debug_idt_table }; DEFINE_PER_CPU_FIRST(union irq_stack_union, - irq_stack_union) __aligned(PAGE_SIZE); + irq_stack_union) __aligned(PAGE_SIZE) __visible; /* * The following four percpu variables are hot. Align current_task to @@ -1035,14 +1133,15 @@ DEFINE_PER_CPU(struct task_struct *, current_task) ____cacheline_aligned = &init_task; EXPORT_PER_CPU_SYMBOL(current_task); -DEFINE_PER_CPU(unsigned long, kernel_stack) = - (unsigned long)&init_thread_union - KERNEL_STACK_OFFSET + THREAD_SIZE; -EXPORT_PER_CPU_SYMBOL(kernel_stack); - DEFINE_PER_CPU(char *, irq_stack_ptr) = init_per_cpu_var(irq_stack_union.irq_stack) + IRQ_STACK_SIZE - 64; -DEFINE_PER_CPU(unsigned int, irq_count) = -1; +DEFINE_PER_CPU(unsigned int, irq_count) __visible = -1; + +DEFINE_PER_CPU(int, __preempt_count) = INIT_PREEMPT_COUNT; +EXPORT_PER_CPU_SYMBOL(__preempt_count); + +DEFINE_PER_CPU(struct task_struct *, fpu_owner_task); /* * Special IST stacks which the CPU switches to when it calls @@ -1076,11 +1175,10 @@ void syscall_init(void) /* Flags to clear on syscall */ wrmsrl(MSR_SYSCALL_MASK, - X86_EFLAGS_TF|X86_EFLAGS_DF|X86_EFLAGS_IF|X86_EFLAGS_IOPL); + X86_EFLAGS_TF|X86_EFLAGS_DF|X86_EFLAGS_IF| + X86_EFLAGS_IOPL|X86_EFLAGS_AC); } -unsigned long kernel_eflags; - /* * Copies of the original ist values from the tss are only accessed during * debugging, no special alignment required. @@ -1096,35 +1194,38 @@ int is_debug_stack(unsigned long addr) (addr <= __get_cpu_var(debug_stack_addr) && addr > (__get_cpu_var(debug_stack_addr) - DEBUG_STKSZ)); } +NOKPROBE_SYMBOL(is_debug_stack); + +DEFINE_PER_CPU(u32, debug_idt_ctr); void debug_stack_set_zero(void) { - load_idt((const struct desc_ptr *)&nmi_idt_descr); + this_cpu_inc(debug_idt_ctr); + load_current_idt(); } +NOKPROBE_SYMBOL(debug_stack_set_zero); void debug_stack_reset(void) { - load_idt((const struct desc_ptr *)&idt_descr); + if (WARN_ON(!this_cpu_read(debug_idt_ctr))) + return; + if (this_cpu_dec_return(debug_idt_ctr) == 0) + load_current_idt(); } +NOKPROBE_SYMBOL(debug_stack_reset); #else /* CONFIG_X86_64 */ DEFINE_PER_CPU(struct task_struct *, current_task) = &init_task; EXPORT_PER_CPU_SYMBOL(current_task); +DEFINE_PER_CPU(int, __preempt_count) = INIT_PREEMPT_COUNT; +EXPORT_PER_CPU_SYMBOL(__preempt_count); +DEFINE_PER_CPU(struct task_struct *, fpu_owner_task); #ifdef CONFIG_CC_STACKPROTECTOR DEFINE_PER_CPU_ALIGNED(struct stack_canary, stack_canary); #endif -/* Make sure %fs and %gs are initialized properly in idle threads */ -struct pt_regs * __cpuinit idle_regs(struct pt_regs *regs) -{ - memset(regs, 0, sizeof(struct pt_regs)); - regs->fs = __KERNEL_PERCPU; - regs->gs = __KERNEL_STACK_CANARY; - - return regs; -} #endif /* CONFIG_X86_64 */ /* @@ -1158,15 +1259,6 @@ static void dbg_restore_debug_regs(void) #endif /* ! CONFIG_KGDB */ /* - * Prints an error where the NUMA and configured core-number mismatch and the - * platform didn't override this to fix it up - */ -void __cpuinit x86_default_fixup_cpu_id(struct cpuinfo_x86 *c, int node) -{ - pr_err("NUMA core number %d differs from configured core number %d\n", node, c->phys_proc_id); -} - -/* * cpu_init() initializes state that is per-CPU. Some data is already * initialized (naturally) in the bootstrap process, such as the GDT * and IDT. We reload them nevertheless, this function acts as a @@ -1175,7 +1267,7 @@ void __cpuinit x86_default_fixup_cpu_id(struct cpuinfo_x86 *c, int node) */ #ifdef CONFIG_X86_64 -void __cpuinit cpu_init(void) +void cpu_init(void) { struct orig_ist *oist; struct task_struct *me; @@ -1184,12 +1276,18 @@ void __cpuinit cpu_init(void) int cpu; int i; + /* + * Load microcode on this cpu if a valid microcode is available. + * This is early microcode loading procedure. + */ + load_ucode_ap(); + cpu = stack_smp_processor_id(); t = &per_cpu(init_tss, cpu); oist = &per_cpu(orig_ist, cpu); #ifdef CONFIG_NUMA - if (cpu != 0 && percpu_read(numa_node) == 0 && + if (this_cpu_read(numa_node) == 0 && early_cpu_to_node(cpu) != NUMA_NO_NODE) set_numa_node(early_cpu_to_node(cpu)); #endif @@ -1211,7 +1309,7 @@ void __cpuinit cpu_init(void) switch_to_new_gdt(cpu); loadsegment(fs, 0); - load_idt((const struct desc_ptr *)&idt_descr); + load_current_idt(); memset(me->thread.tls_array, 0, GDT_ENTRY_TLS_ENTRIES * 8); syscall_init(); @@ -1221,8 +1319,7 @@ void __cpuinit cpu_init(void) barrier(); x86_configure_nx(); - if (cpu != 0) - enable_x2apic(); + enable_x2apic(); /* * set up and load the per-CPU TSS @@ -1262,9 +1359,6 @@ void __cpuinit cpu_init(void) dbg_restore_debug_regs(); fpu_init(); - xsave_init(); - - raw_local_save_flags(kernel_eflags); if (is_uv_system()) uv_cpu_init(); @@ -1272,13 +1366,15 @@ void __cpuinit cpu_init(void) #else -void __cpuinit cpu_init(void) +void cpu_init(void) { int cpu = smp_processor_id(); struct task_struct *curr = current; struct tss_struct *t = &per_cpu(init_tss, cpu); struct thread_struct *thread = &curr->thread; + show_ucode_info_early(); + if (cpumask_test_and_set_cpu(cpu, cpu_initialized_mask)) { printk(KERN_WARNING "CPU#%d already initialized!\n", cpu); for (;;) @@ -1290,7 +1386,7 @@ void __cpuinit cpu_init(void) if (cpu_has_vme || cpu_has_tsc || cpu_has_de) clear_in_cr4(X86_CR4_VME|X86_CR4_PVI|X86_CR4_TSD|X86_CR4_DE); - load_idt(&idt_descr); + load_current_idt(); switch_to_new_gdt(cpu); /* @@ -1317,6 +1413,19 @@ void __cpuinit cpu_init(void) dbg_restore_debug_regs(); fpu_init(); - xsave_init(); } #endif + +#ifdef CONFIG_X86_DEBUG_STATIC_CPU_HAS +void warn_pre_alternatives(void) +{ + WARN(1, "You're using static_cpu_has before alternatives have run!\n"); +} +EXPORT_SYMBOL_GPL(warn_pre_alternatives); +#endif + +inline bool __static_cpu_has_safe(u16 bit) +{ + return boot_cpu_has(bit); +} +EXPORT_SYMBOL_GPL(__static_cpu_has_safe); diff --git a/arch/x86/kernel/cpu/cpu.h b/arch/x86/kernel/cpu/cpu.h index 8bacc7826fb..c37dc37e831 100644 --- a/arch/x86/kernel/cpu/cpu.h +++ b/arch/x86/kernel/cpu/cpu.h @@ -1,12 +1,6 @@ #ifndef ARCH_X86_CPU_H #define ARCH_X86_CPU_H -struct cpu_model_info { - int vendor; - int family; - const char *model_names[16]; -}; - /* attempt to consolidate cpu attributes */ struct cpu_dev { const char *c_vendor; @@ -14,14 +8,31 @@ struct cpu_dev { /* some have two possibilities for cpuid string */ const char *c_ident[2]; - struct cpu_model_info c_models[4]; - void (*c_early_init)(struct cpuinfo_x86 *); void (*c_bsp_init)(struct cpuinfo_x86 *); void (*c_init)(struct cpuinfo_x86 *); void (*c_identify)(struct cpuinfo_x86 *); - unsigned int (*c_size_cache)(struct cpuinfo_x86 *, unsigned int); + void (*c_detect_tlb)(struct cpuinfo_x86 *); int c_x86_vendor; +#ifdef CONFIG_X86_32 + /* Optional vendor specific routine to obtain the cache size. */ + unsigned int (*legacy_cache_size)(struct cpuinfo_x86 *, + unsigned int); + + /* Family/stepping-based lookup table for model names. */ + struct legacy_cpu_model_info { + int family; + const char *model_names[16]; + } legacy_models[5]; +#endif +}; + +struct _tlb_table { + unsigned char descriptor; + char tlb_type; + unsigned int entries; + /* unsigned int ways; */ + char info[128]; }; #define cpu_dev_register(cpu_devX) \ diff --git a/arch/x86/kernel/cpu/cyrix.c b/arch/x86/kernel/cpu/cyrix.c index 4fbd384fb64..aaf152e7963 100644 --- a/arch/x86/kernel/cpu/cyrix.c +++ b/arch/x86/kernel/cpu/cyrix.c @@ -1,4 +1,3 @@ -#include <linux/init.h> #include <linux/bitops.h> #include <linux/delay.h> #include <linux/pci.h> @@ -15,7 +14,7 @@ /* * Read NSC/Cyrix DEVID registers (DIR) to get more detailed info. about the CPU */ -static void __cpuinit __do_cyrix_devid(unsigned char *dir0, unsigned char *dir1) +static void __do_cyrix_devid(unsigned char *dir0, unsigned char *dir1) { unsigned char ccr2, ccr3; @@ -44,7 +43,7 @@ static void __cpuinit __do_cyrix_devid(unsigned char *dir0, unsigned char *dir1) } } -static void __cpuinit do_cyrix_devid(unsigned char *dir0, unsigned char *dir1) +static void do_cyrix_devid(unsigned char *dir0, unsigned char *dir1) { unsigned long flags; @@ -59,25 +58,25 @@ static void __cpuinit do_cyrix_devid(unsigned char *dir0, unsigned char *dir1) * Actually since bugs.h doesn't even reference this perhaps someone should * fix the documentation ??? */ -static unsigned char Cx86_dir0_msb __cpuinitdata = 0; +static unsigned char Cx86_dir0_msb = 0; -static const char __cpuinitconst Cx86_model[][9] = { +static const char Cx86_model[][9] = { "Cx486", "Cx486", "5x86 ", "6x86", "MediaGX ", "6x86MX ", "M II ", "Unknown" }; -static const char __cpuinitconst Cx486_name[][5] = { +static const char Cx486_name[][5] = { "SLC", "DLC", "SLC2", "DLC2", "SRx", "DRx", "SRx2", "DRx2" }; -static const char __cpuinitconst Cx486S_name[][4] = { +static const char Cx486S_name[][4] = { "S", "S2", "Se", "S2e" }; -static const char __cpuinitconst Cx486D_name[][4] = { +static const char Cx486D_name[][4] = { "DX", "DX2", "?", "?", "?", "DX4" }; -static char Cx86_cb[] __cpuinitdata = "?.5x Core/Bus Clock"; -static const char __cpuinitconst cyrix_model_mult1[] = "12??43"; -static const char __cpuinitconst cyrix_model_mult2[] = "12233445"; +static char Cx86_cb[] = "?.5x Core/Bus Clock"; +static const char cyrix_model_mult1[] = "12??43"; +static const char cyrix_model_mult2[] = "12233445"; /* * Reset the slow-loop (SLOP) bit on the 686(L) which is set by some old @@ -87,7 +86,7 @@ static const char __cpuinitconst cyrix_model_mult2[] = "12233445"; * FIXME: our newer udelay uses the tsc. We don't need to frob with SLOP */ -static void __cpuinit check_cx686_slop(struct cpuinfo_x86 *c) +static void check_cx686_slop(struct cpuinfo_x86 *c) { unsigned long flags; @@ -112,7 +111,7 @@ static void __cpuinit check_cx686_slop(struct cpuinfo_x86 *c) } -static void __cpuinit set_cx86_reorder(void) +static void set_cx86_reorder(void) { u8 ccr3; @@ -127,7 +126,7 @@ static void __cpuinit set_cx86_reorder(void) setCx86(CX86_CCR3, ccr3); } -static void __cpuinit set_cx86_memwb(void) +static void set_cx86_memwb(void) { printk(KERN_INFO "Enable Memory-Write-back mode on Cyrix/NSC processor.\n"); @@ -143,7 +142,7 @@ static void __cpuinit set_cx86_memwb(void) * Configure later MediaGX and/or Geode processor. */ -static void __cpuinit geode_configure(void) +static void geode_configure(void) { unsigned long flags; u8 ccr3; @@ -166,7 +165,7 @@ static void __cpuinit geode_configure(void) local_irq_restore(flags); } -static void __cpuinit early_init_cyrix(struct cpuinfo_x86 *c) +static void early_init_cyrix(struct cpuinfo_x86 *c) { unsigned char dir0, dir0_msn, dir1 = 0; @@ -185,7 +184,7 @@ static void __cpuinit early_init_cyrix(struct cpuinfo_x86 *c) } } -static void __cpuinit init_cyrix(struct cpuinfo_x86 *c) +static void init_cyrix(struct cpuinfo_x86 *c) { unsigned char dir0, dir0_msn, dir0_lsn, dir1 = 0; char *buf = c->x86_model_id; @@ -249,7 +248,7 @@ static void __cpuinit init_cyrix(struct cpuinfo_x86 *c) /* Emulate MTRRs using Cyrix's ARRs. */ set_cpu_cap(c, X86_FEATURE_CYRIX_ARR); /* 6x86's contain this bug */ - c->coma_bug = 1; + set_cpu_bug(c, X86_BUG_COMA); break; case 4: /* MediaGX/GXm or Geode GXM/GXLV/GX1 */ @@ -317,7 +316,8 @@ static void __cpuinit init_cyrix(struct cpuinfo_x86 *c) /* Enable MMX extensions (App note 108) */ setCx86_old(CX86_CCR7, getCx86_old(CX86_CCR7)|1); } else { - c->coma_bug = 1; /* 6x86MX, it has the bug. */ + /* A 6x86MX - it has the bug. */ + set_cpu_bug(c, X86_BUG_COMA); } tmp = (!(dir0_lsn & 7) || dir0_lsn & 1) ? 2 : 0; Cx86_cb[tmp] = cyrix_model_mult2[dir0_lsn & 7]; @@ -332,7 +332,7 @@ static void __cpuinit init_cyrix(struct cpuinfo_x86 *c) switch (dir0_lsn) { case 0xd: /* either a 486SLC or DLC w/o DEVID */ dir0_msn = 0; - p = Cx486_name[(c->hard_math) ? 1 : 0]; + p = Cx486_name[(cpu_has_fpu ? 1 : 0)]; break; case 0xe: /* a 486S A step */ @@ -355,7 +355,7 @@ static void __cpuinit init_cyrix(struct cpuinfo_x86 *c) /* * Handle National Semiconductor branded processors */ -static void __cpuinit init_nsc(struct cpuinfo_x86 *c) +static void init_nsc(struct cpuinfo_x86 *c) { /* * There may be GX1 processors in the wild that are branded @@ -404,7 +404,7 @@ static inline int test_cyrix_52div(void) return (unsigned char) (test >> 8) == 0x02; } -static void __cpuinit cyrix_identify(struct cpuinfo_x86 *c) +static void cyrix_identify(struct cpuinfo_x86 *c) { /* Detect Cyrix with disabled CPUID */ if (c->x86 == 4 && test_cyrix_52div()) { @@ -440,7 +440,7 @@ static void __cpuinit cyrix_identify(struct cpuinfo_x86 *c) } } -static const struct cpu_dev __cpuinitconst cyrix_cpu_dev = { +static const struct cpu_dev cyrix_cpu_dev = { .c_vendor = "Cyrix", .c_ident = { "CyrixInstead" }, .c_early_init = early_init_cyrix, @@ -451,7 +451,7 @@ static const struct cpu_dev __cpuinitconst cyrix_cpu_dev = { cpu_dev_register(cyrix_cpu_dev); -static const struct cpu_dev __cpuinitconst nsc_cpu_dev = { +static const struct cpu_dev nsc_cpu_dev = { .c_vendor = "NSC", .c_ident = { "Geode by NSC" }, .c_init = init_nsc, diff --git a/arch/x86/kernel/cpu/hypervisor.c b/arch/x86/kernel/cpu/hypervisor.c index 755f64fb074..36ce402a3fa 100644 --- a/arch/x86/kernel/cpu/hypervisor.c +++ b/arch/x86/kernel/cpu/hypervisor.c @@ -25,11 +25,6 @@ #include <asm/processor.h> #include <asm/hypervisor.h> -/* - * Hypervisor detect order. This is specified explicitly here because - * some hypervisors might implement compatibility modes for other - * hypervisors and therefore need to be detected in specific sequence. - */ static const __initconst struct hypervisor_x86 * const hypervisors[] = { #ifdef CONFIG_XEN_PVHVM @@ -37,6 +32,9 @@ static const __initconst struct hypervisor_x86 * const hypervisors[] = #endif &x86_hyper_vmware, &x86_hyper_ms_hyperv, +#ifdef CONFIG_KVM_GUEST + &x86_hyper_kvm, +#endif }; const struct hypervisor_x86 *x86_hyper; @@ -46,18 +44,22 @@ static inline void __init detect_hypervisor_vendor(void) { const struct hypervisor_x86 *h, * const *p; + uint32_t pri, max_pri = 0; for (p = hypervisors; p < hypervisors + ARRAY_SIZE(hypervisors); p++) { h = *p; - if (h->detect()) { + pri = h->detect(); + if (pri != 0 && pri > max_pri) { + max_pri = pri; x86_hyper = h; - printk(KERN_INFO "Hypervisor detected: %s\n", h->name); - break; } } + + if (max_pri) + printk(KERN_INFO "Hypervisor detected: %s\n", x86_hyper->name); } -void __cpuinit init_hypervisor(struct cpuinfo_x86 *c) +void init_hypervisor(struct cpuinfo_x86 *c) { if (x86_hyper && x86_hyper->set_cpu_features) x86_hyper->set_cpu_features(c); @@ -76,3 +78,10 @@ void __init init_hypervisor_platform(void) if (x86_hyper->init_platform) x86_hyper->init_platform(); } + +bool __init hypervisor_x2apic_available(void) +{ + return x86_hyper && + x86_hyper->x2apic_available && + x86_hyper->x2apic_available(); +} diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c index 3e6ff6cbf42..f9e4fdd3b87 100644 --- a/arch/x86/kernel/cpu/intel.c +++ b/arch/x86/kernel/cpu/intel.c @@ -1,4 +1,3 @@ -#include <linux/init.h> #include <linux/kernel.h> #include <linux/string.h> @@ -17,7 +16,6 @@ #ifdef CONFIG_X86_64 #include <linux/topology.h> -#include <asm/numa_64.h> #endif #include "cpu.h" @@ -27,17 +25,14 @@ #include <asm/apic.h> #endif -static void __cpuinit early_init_intel(struct cpuinfo_x86 *c) +static void early_init_intel(struct cpuinfo_x86 *c) { u64 misc_enable; /* Unmask CPUID levels if masked: */ if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) { - rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable); - - if (misc_enable & MSR_IA32_MISC_ENABLE_LIMIT_CPUID) { - misc_enable &= ~MSR_IA32_MISC_ENABLE_LIMIT_CPUID; - wrmsrl(MSR_IA32_MISC_ENABLE, misc_enable); + if (msr_clear_bit(MSR_IA32_MISC_ENABLE, + MSR_IA32_MISC_ENABLE_LIMIT_CPUID_BIT) > 0) { c->cpuid_level = cpuid_eax(0); get_cpu_cap(c); } @@ -94,7 +89,19 @@ static void __cpuinit early_init_intel(struct cpuinfo_x86 *c) set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC); if (!check_tsc_unstable()) - sched_clock_stable = 1; + set_sched_clock_stable(); + } + + /* Penwell and Cloverview have the TSC which doesn't sleep on S3 */ + if (c->x86 == 6) { + switch (c->x86_model) { + case 0x27: /* Penwell */ + case 0x35: /* Cloverview */ + set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC_S3); + break; + default: + break; + } } /* @@ -119,16 +126,10 @@ static void __cpuinit early_init_intel(struct cpuinfo_x86 *c) * Ingo Molnar reported a Pentium D (model 6) and a Xeon * (model 2) with the same problem. */ - if (c->x86 == 15) { - rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable); - - if (misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING) { - printk(KERN_INFO "kmemcheck: Disabling fast string operations\n"); - - misc_enable &= ~MSR_IA32_MISC_ENABLE_FAST_STRING; - wrmsrl(MSR_IA32_MISC_ENABLE, misc_enable); - } - } + if (c->x86 == 15) + if (msr_clear_bit(MSR_IA32_MISC_ENABLE, + MSR_IA32_MISC_ENABLE_FAST_STRING_BIT) > 0) + pr_info("kmemcheck: Disabling fast string operations\n"); #endif /* @@ -152,7 +153,7 @@ static void __cpuinit early_init_intel(struct cpuinfo_x86 *c) * This is called before we do cpu ident work */ -int __cpuinit ppro_with_ram_bug(void) +int ppro_with_ram_bug(void) { /* Uses data from early_cpu_detect now */ if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL && @@ -165,21 +166,7 @@ int __cpuinit ppro_with_ram_bug(void) return 0; } -#ifdef CONFIG_X86_F00F_BUG -static void __cpuinit trap_init_f00f_bug(void) -{ - __set_fixmap(FIX_F00F_IDT, __pa(&idt_table), PAGE_KERNEL_RO); - - /* - * Update the IDT descriptor and reload the IDT so that - * it uses the read-only mapped virtual address. - */ - idt_descr.address = fix_to_virt(FIX_F00F_IDT); - load_idt(&idt_descr); -} -#endif - -static void __cpuinit intel_smp_check(struct cpuinfo_x86 *c) +static void intel_smp_check(struct cpuinfo_x86 *c) { /* calling is from identify_secondary_cpu() ? */ if (!c->cpu_index) @@ -199,24 +186,28 @@ static void __cpuinit intel_smp_check(struct cpuinfo_x86 *c) } } -static void __cpuinit intel_workarounds(struct cpuinfo_x86 *c) +static int forcepae; +static int __init forcepae_setup(char *__unused) { - unsigned long lo, hi; + forcepae = 1; + return 1; +} +__setup("forcepae", forcepae_setup); +static void intel_workarounds(struct cpuinfo_x86 *c) +{ #ifdef CONFIG_X86_F00F_BUG /* * All current models of Pentium and Pentium with MMX technology CPUs * have the F0 0F bug, which lets nonprivileged users lock up the - * system. - * Note that the workaround only should be initialized once... + * system. Announce that the fault handler will be checking for it. */ - c->f00f_bug = 0; + clear_cpu_bug(c, X86_BUG_F00F); if (!paravirt_enabled() && c->x86 == 5) { static int f00f_workaround_enabled; - c->f00f_bug = 1; + set_cpu_bug(c, X86_BUG_F00F); if (!f00f_workaround_enabled) { - trap_init_f00f_bug(); printk(KERN_NOTICE "Intel Pentium with F0 0F bug - workaround enabled.\n"); f00f_workaround_enabled = 1; } @@ -231,16 +222,26 @@ static void __cpuinit intel_workarounds(struct cpuinfo_x86 *c) clear_cpu_cap(c, X86_FEATURE_SEP); /* + * PAE CPUID issue: many Pentium M report no PAE but may have a + * functionally usable PAE implementation. + * Forcefully enable PAE if kernel parameter "forcepae" is present. + */ + if (forcepae) { + printk(KERN_WARNING "PAE forced!\n"); + set_cpu_cap(c, X86_FEATURE_PAE); + add_taint(TAINT_CPU_OUT_OF_SPEC, LOCKDEP_NOW_UNRELIABLE); + } + + /* * P4 Xeon errata 037 workaround. * Hardware prefetcher may cause stale data to be loaded into the cache. */ if ((c->x86 == 15) && (c->x86_model == 1) && (c->x86_mask == 1)) { - rdmsr(MSR_IA32_MISC_ENABLE, lo, hi); - if ((lo & MSR_IA32_MISC_ENABLE_PREFETCH_DISABLE) == 0) { - printk (KERN_INFO "CPU: C0 stepping P4 Xeon detected.\n"); - printk (KERN_INFO "CPU: Disabling hardware prefetching (Errata 037)\n"); - lo |= MSR_IA32_MISC_ENABLE_PREFETCH_DISABLE; - wrmsr(MSR_IA32_MISC_ENABLE, lo, hi); + if (msr_set_bit(MSR_IA32_MISC_ENABLE, + MSR_IA32_MISC_ENABLE_PREFETCH_DISABLE_BIT) + > 0) { + pr_info("CPU: C0 stepping P4 Xeon detected.\n"); + pr_info("CPU: Disabling hardware prefetching (Errata 037)\n"); } } @@ -273,19 +274,15 @@ static void __cpuinit intel_workarounds(struct cpuinfo_x86 *c) } #endif -#ifdef CONFIG_X86_NUMAQ - numaq_tsc_disable(); -#endif - intel_smp_check(c); } #else -static void __cpuinit intel_workarounds(struct cpuinfo_x86 *c) +static void intel_workarounds(struct cpuinfo_x86 *c) { } #endif -static void __cpuinit srat_detect_node(struct cpuinfo_x86 *c) +static void srat_detect_node(struct cpuinfo_x86 *c) { #ifdef CONFIG_NUMA unsigned node; @@ -305,7 +302,7 @@ static void __cpuinit srat_detect_node(struct cpuinfo_x86 *c) /* * find out the number of processor cores on the die */ -static int __cpuinit intel_num_cpu_cores(struct cpuinfo_x86 *c) +static int intel_num_cpu_cores(struct cpuinfo_x86 *c) { unsigned int eax, ebx, ecx, edx; @@ -320,7 +317,7 @@ static int __cpuinit intel_num_cpu_cores(struct cpuinfo_x86 *c) return 1; } -static void __cpuinit detect_vmx_virtcap(struct cpuinfo_x86 *c) +static void detect_vmx_virtcap(struct cpuinfo_x86 *c) { /* Intel VMX MSR indicated features */ #define X86_VMX_FEATURE_PROC_CTLS_TPR_SHADOW 0x00200000 @@ -358,7 +355,7 @@ static void __cpuinit detect_vmx_virtcap(struct cpuinfo_x86 *c) } } -static void __cpuinit init_intel(struct cpuinfo_x86 *c) +static void init_intel(struct cpuinfo_x86 *c) { unsigned int l2 = 0; @@ -373,6 +370,17 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c) */ detect_extended_topology(c); + if (!cpu_has(c, X86_FEATURE_XTOPOLOGY)) { + /* + * let's use the legacy cpuid vector 0x1 and 0x4 for topology + * detection. + */ + c->x86_max_cores = intel_num_cpu_cores(c); +#ifdef CONFIG_X86_32 + detect_ht(c); +#endif + } + l2 = init_intel_cacheinfo(c); if (c->cpuid_level > 9) { unsigned eax = cpuid_eax(10); @@ -392,7 +400,8 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c) set_cpu_cap(c, X86_FEATURE_PEBS); } - if (c->x86 == 6 && c->x86_model == 29 && cpu_has_clflush) + if (c->x86 == 6 && cpu_has_clflush && + (c->x86_model == 29 || c->x86_model == 46 || c->x86_model == 47)) set_cpu_cap(c, X86_FEATURE_CLFLUSH_MONITOR); #ifdef CONFIG_X86_64 @@ -440,17 +449,6 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c) set_cpu_cap(c, X86_FEATURE_P3); #endif - if (!cpu_has(c, X86_FEATURE_XTOPOLOGY)) { - /* - * let's use the legacy cpuid vector 0x1 and 0x4 for topology - * detection. - */ - c->x86_max_cores = intel_num_cpu_cores(c); -#ifdef CONFIG_X86_32 - detect_ht(c); -#endif - } - /* Work around errata */ srat_detect_node(c); @@ -477,7 +475,7 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c) } #ifdef CONFIG_X86_32 -static unsigned int __cpuinit intel_size_cache(struct cpuinfo_x86 *c, unsigned int size) +static unsigned int intel_size_cache(struct cpuinfo_x86 *c, unsigned int size) { /* * Intel PIII Tualatin. This comes in two flavours. @@ -491,12 +489,209 @@ static unsigned int __cpuinit intel_size_cache(struct cpuinfo_x86 *c, unsigned i } #endif -static const struct cpu_dev __cpuinitconst intel_cpu_dev = { +#define TLB_INST_4K 0x01 +#define TLB_INST_4M 0x02 +#define TLB_INST_2M_4M 0x03 + +#define TLB_INST_ALL 0x05 +#define TLB_INST_1G 0x06 + +#define TLB_DATA_4K 0x11 +#define TLB_DATA_4M 0x12 +#define TLB_DATA_2M_4M 0x13 +#define TLB_DATA_4K_4M 0x14 + +#define TLB_DATA_1G 0x16 + +#define TLB_DATA0_4K 0x21 +#define TLB_DATA0_4M 0x22 +#define TLB_DATA0_2M_4M 0x23 + +#define STLB_4K 0x41 +#define STLB_4K_2M 0x42 + +static const struct _tlb_table intel_tlb_table[] = { + { 0x01, TLB_INST_4K, 32, " TLB_INST 4 KByte pages, 4-way set associative" }, + { 0x02, TLB_INST_4M, 2, " TLB_INST 4 MByte pages, full associative" }, + { 0x03, TLB_DATA_4K, 64, " TLB_DATA 4 KByte pages, 4-way set associative" }, + { 0x04, TLB_DATA_4M, 8, " TLB_DATA 4 MByte pages, 4-way set associative" }, + { 0x05, TLB_DATA_4M, 32, " TLB_DATA 4 MByte pages, 4-way set associative" }, + { 0x0b, TLB_INST_4M, 4, " TLB_INST 4 MByte pages, 4-way set associative" }, + { 0x4f, TLB_INST_4K, 32, " TLB_INST 4 KByte pages */" }, + { 0x50, TLB_INST_ALL, 64, " TLB_INST 4 KByte and 2-MByte or 4-MByte pages" }, + { 0x51, TLB_INST_ALL, 128, " TLB_INST 4 KByte and 2-MByte or 4-MByte pages" }, + { 0x52, TLB_INST_ALL, 256, " TLB_INST 4 KByte and 2-MByte or 4-MByte pages" }, + { 0x55, TLB_INST_2M_4M, 7, " TLB_INST 2-MByte or 4-MByte pages, fully associative" }, + { 0x56, TLB_DATA0_4M, 16, " TLB_DATA0 4 MByte pages, 4-way set associative" }, + { 0x57, TLB_DATA0_4K, 16, " TLB_DATA0 4 KByte pages, 4-way associative" }, + { 0x59, TLB_DATA0_4K, 16, " TLB_DATA0 4 KByte pages, fully associative" }, + { 0x5a, TLB_DATA0_2M_4M, 32, " TLB_DATA0 2-MByte or 4 MByte pages, 4-way set associative" }, + { 0x5b, TLB_DATA_4K_4M, 64, " TLB_DATA 4 KByte and 4 MByte pages" }, + { 0x5c, TLB_DATA_4K_4M, 128, " TLB_DATA 4 KByte and 4 MByte pages" }, + { 0x5d, TLB_DATA_4K_4M, 256, " TLB_DATA 4 KByte and 4 MByte pages" }, + { 0x61, TLB_INST_4K, 48, " TLB_INST 4 KByte pages, full associative" }, + { 0x63, TLB_DATA_1G, 4, " TLB_DATA 1 GByte pages, 4-way set associative" }, + { 0x76, TLB_INST_2M_4M, 8, " TLB_INST 2-MByte or 4-MByte pages, fully associative" }, + { 0xb0, TLB_INST_4K, 128, " TLB_INST 4 KByte pages, 4-way set associative" }, + { 0xb1, TLB_INST_2M_4M, 4, " TLB_INST 2M pages, 4-way, 8 entries or 4M pages, 4-way entries" }, + { 0xb2, TLB_INST_4K, 64, " TLB_INST 4KByte pages, 4-way set associative" }, + { 0xb3, TLB_DATA_4K, 128, " TLB_DATA 4 KByte pages, 4-way set associative" }, + { 0xb4, TLB_DATA_4K, 256, " TLB_DATA 4 KByte pages, 4-way associative" }, + { 0xb5, TLB_INST_4K, 64, " TLB_INST 4 KByte pages, 8-way set ssociative" }, + { 0xb6, TLB_INST_4K, 128, " TLB_INST 4 KByte pages, 8-way set ssociative" }, + { 0xba, TLB_DATA_4K, 64, " TLB_DATA 4 KByte pages, 4-way associative" }, + { 0xc0, TLB_DATA_4K_4M, 8, " TLB_DATA 4 KByte and 4 MByte pages, 4-way associative" }, + { 0xc1, STLB_4K_2M, 1024, " STLB 4 KByte and 2 MByte pages, 8-way associative" }, + { 0xc2, TLB_DATA_2M_4M, 16, " DTLB 2 MByte/4MByte pages, 4-way associative" }, + { 0xca, STLB_4K, 512, " STLB 4 KByte pages, 4-way associative" }, + { 0x00, 0, 0 } +}; + +static void intel_tlb_lookup(const unsigned char desc) +{ + unsigned char k; + if (desc == 0) + return; + + /* look up this descriptor in the table */ + for (k = 0; intel_tlb_table[k].descriptor != desc && \ + intel_tlb_table[k].descriptor != 0; k++) + ; + + if (intel_tlb_table[k].tlb_type == 0) + return; + + switch (intel_tlb_table[k].tlb_type) { + case STLB_4K: + if (tlb_lli_4k[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_4k[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lld_4k[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_4k[ENTRIES] = intel_tlb_table[k].entries; + break; + case STLB_4K_2M: + if (tlb_lli_4k[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_4k[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lld_4k[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_4k[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lli_2m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_2m[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lld_2m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_2m[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lli_4m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_4m[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lld_4m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_4m[ENTRIES] = intel_tlb_table[k].entries; + break; + case TLB_INST_ALL: + if (tlb_lli_4k[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_4k[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lli_2m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_2m[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lli_4m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_4m[ENTRIES] = intel_tlb_table[k].entries; + break; + case TLB_INST_4K: + if (tlb_lli_4k[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_4k[ENTRIES] = intel_tlb_table[k].entries; + break; + case TLB_INST_4M: + if (tlb_lli_4m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_4m[ENTRIES] = intel_tlb_table[k].entries; + break; + case TLB_INST_2M_4M: + if (tlb_lli_2m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_2m[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lli_4m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_4m[ENTRIES] = intel_tlb_table[k].entries; + break; + case TLB_DATA_4K: + case TLB_DATA0_4K: + if (tlb_lld_4k[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_4k[ENTRIES] = intel_tlb_table[k].entries; + break; + case TLB_DATA_4M: + case TLB_DATA0_4M: + if (tlb_lld_4m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_4m[ENTRIES] = intel_tlb_table[k].entries; + break; + case TLB_DATA_2M_4M: + case TLB_DATA0_2M_4M: + if (tlb_lld_2m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_2m[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lld_4m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_4m[ENTRIES] = intel_tlb_table[k].entries; + break; + case TLB_DATA_4K_4M: + if (tlb_lld_4k[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_4k[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lld_4m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_4m[ENTRIES] = intel_tlb_table[k].entries; + break; + case TLB_DATA_1G: + if (tlb_lld_1g[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_1g[ENTRIES] = intel_tlb_table[k].entries; + break; + } +} + +static void intel_tlb_flushall_shift_set(struct cpuinfo_x86 *c) +{ + switch ((c->x86 << 8) + c->x86_model) { + case 0x60f: /* original 65 nm celeron/pentium/core2/xeon, "Merom"/"Conroe" */ + case 0x616: /* single-core 65 nm celeron/core2solo "Merom-L"/"Conroe-L" */ + case 0x617: /* current 45 nm celeron/core2/xeon "Penryn"/"Wolfdale" */ + case 0x61d: /* six-core 45 nm xeon "Dunnington" */ + tlb_flushall_shift = -1; + break; + case 0x63a: /* Ivybridge */ + tlb_flushall_shift = 2; + break; + case 0x61a: /* 45 nm nehalem, "Bloomfield" */ + case 0x61e: /* 45 nm nehalem, "Lynnfield" */ + case 0x625: /* 32 nm nehalem, "Clarkdale" */ + case 0x62c: /* 32 nm nehalem, "Gulftown" */ + case 0x62e: /* 45 nm nehalem-ex, "Beckton" */ + case 0x62f: /* 32 nm Xeon E7 */ + case 0x62a: /* SandyBridge */ + case 0x62d: /* SandyBridge, "Romely-EP" */ + default: + tlb_flushall_shift = 6; + } +} + +static void intel_detect_tlb(struct cpuinfo_x86 *c) +{ + int i, j, n; + unsigned int regs[4]; + unsigned char *desc = (unsigned char *)regs; + + if (c->cpuid_level < 2) + return; + + /* Number of times to iterate */ + n = cpuid_eax(2) & 0xFF; + + for (i = 0 ; i < n ; i++) { + cpuid(2, ®s[0], ®s[1], ®s[2], ®s[3]); + + /* If bit 31 is set, this is an unknown format */ + for (j = 0 ; j < 3 ; j++) + if (regs[j] & (1 << 31)) + regs[j] = 0; + + /* Byte 0 is level count, not a descriptor */ + for (j = 1 ; j < 16 ; j++) + intel_tlb_lookup(desc[j]); + } + intel_tlb_flushall_shift_set(c); +} + +static const struct cpu_dev intel_cpu_dev = { .c_vendor = "Intel", .c_ident = { "GenuineIntel" }, #ifdef CONFIG_X86_32 - .c_models = { - { .vendor = X86_VENDOR_INTEL, .family = 4, .model_names = + .legacy_models = { + { .family = 4, .model_names = { [0] = "486 DX-25/33", [1] = "486 DX-50", @@ -509,7 +704,7 @@ static const struct cpu_dev __cpuinitconst intel_cpu_dev = { [9] = "486 DX/4-WB" } }, - { .vendor = X86_VENDOR_INTEL, .family = 5, .model_names = + { .family = 5, .model_names = { [0] = "Pentium 60/66 A-step", [1] = "Pentium 60/66", @@ -520,7 +715,7 @@ static const struct cpu_dev __cpuinitconst intel_cpu_dev = { [8] = "Mobile Pentium MMX" } }, - { .vendor = X86_VENDOR_INTEL, .family = 6, .model_names = + { .family = 6, .model_names = { [0] = "Pentium Pro A-step", [1] = "Pentium Pro", @@ -534,7 +729,7 @@ static const struct cpu_dev __cpuinitconst intel_cpu_dev = { [11] = "Pentium III (Tualatin)", } }, - { .vendor = X86_VENDOR_INTEL, .family = 15, .model_names = + { .family = 15, .model_names = { [0] = "Pentium 4 (Unknown)", [1] = "Pentium 4 (Willamette)", @@ -544,8 +739,9 @@ static const struct cpu_dev __cpuinitconst intel_cpu_dev = { } }, }, - .c_size_cache = intel_size_cache, + .legacy_cache_size = intel_size_cache, #endif + .c_detect_tlb = intel_detect_tlb, .c_early_init = early_init_intel, .c_init = init_intel, .c_x86_vendor = X86_VENDOR_INTEL, diff --git a/arch/x86/kernel/cpu/intel_cacheinfo.c b/arch/x86/kernel/cpu/intel_cacheinfo.c index 6b45e5e7a90..9c8f7394c61 100644 --- a/arch/x86/kernel/cpu/intel_cacheinfo.c +++ b/arch/x86/kernel/cpu/intel_cacheinfo.c @@ -1,5 +1,5 @@ /* - * Routines to indentify caches on Intel CPU. + * Routines to identify caches on Intel CPU. * * Changes: * Venkatesh Pallipadi : Adding cache identification through cpuid(4) @@ -37,7 +37,7 @@ struct _cache_table { /* All the cache descriptor types we care about (no TLB or trace cache entries) */ -static const struct _cache_table __cpuinitconst cache_table[] = +static const struct _cache_table cache_table[] = { { 0x06, LVL_1_INST, 8 }, /* 4-way set assoc, 32 byte line size */ { 0x08, LVL_1_INST, 16 }, /* 4-way set assoc, 32 byte line size */ @@ -203,7 +203,7 @@ union l3_cache { unsigned val; }; -static const unsigned short __cpuinitconst assocs[] = { +static const unsigned short assocs[] = { [1] = 1, [2] = 2, [4] = 4, @@ -217,10 +217,10 @@ static const unsigned short __cpuinitconst assocs[] = { [0xf] = 0xffff /* fully associative - no way to show this currently */ }; -static const unsigned char __cpuinitconst levels[] = { 1, 1, 2, 3 }; -static const unsigned char __cpuinitconst types[] = { 1, 2, 3, 3 }; +static const unsigned char levels[] = { 1, 1, 2, 3 }; +static const unsigned char types[] = { 1, 2, 3, 3 }; -static void __cpuinit +static void amd_cpuid4(int leaf, union _cpuid4_leaf_eax *eax, union _cpuid4_leaf_ebx *ebx, union _cpuid4_leaf_ecx *ecx) @@ -298,12 +298,11 @@ struct _cache_attr { unsigned int); }; -#ifdef CONFIG_AMD_NB - +#if defined(CONFIG_AMD_NB) && defined(CONFIG_SYSFS) /* * L3 cache descriptors */ -static void __cpuinit amd_calc_l3_indices(struct amd_northbridge *nb) +static void amd_calc_l3_indices(struct amd_northbridge *nb) { struct amd_l3_cache *l3 = &nb->l3_cache; unsigned int sc0, sc1, sc2, sc3; @@ -326,8 +325,7 @@ static void __cpuinit amd_calc_l3_indices(struct amd_northbridge *nb) l3->indices = (max(max3(sc0, sc1, sc2), sc3) << 10) - 1; } -static void __cpuinit amd_init_l3_cache(struct _cpuid4_info_regs *this_leaf, - int index) +static void amd_init_l3_cache(struct _cpuid4_info_regs *this_leaf, int index) { int node; @@ -434,14 +432,14 @@ int amd_set_l3_disable_slot(struct amd_northbridge *nb, int cpu, unsigned slot, /* check if @slot is already used or the index is already disabled */ ret = amd_get_l3_disable_slot(nb, slot); if (ret >= 0) - return -EINVAL; + return -EEXIST; if (index > nb->l3_cache.indices) return -EINVAL; /* check whether the other slot has disabled the same index already */ if (index == amd_get_l3_disable_slot(nb, !slot)) - return -EINVAL; + return -EEXIST; amd_l3_disable_index(nb, cpu, slot, index); @@ -469,8 +467,8 @@ static ssize_t store_cache_disable(struct _cpuid4_info *this_leaf, err = amd_set_l3_disable_slot(this_leaf->base.nb, cpu, slot, val); if (err) { if (err == -EEXIST) - printk(KERN_WARNING "L3 disable slot %d in use!\n", - slot); + pr_warning("L3 slot %d in use/index already disabled!\n", + slot); return err; } return count; @@ -525,13 +523,12 @@ store_subcaches(struct _cpuid4_info *this_leaf, const char *buf, size_t count, static struct _cache_attr subcaches = __ATTR(subcaches, 0644, show_subcaches, store_subcaches); -#else /* CONFIG_AMD_NB */ +#else #define amd_init_l3_cache(x, y) -#endif /* CONFIG_AMD_NB */ +#endif /* CONFIG_AMD_NB && CONFIG_SYSFS */ static int -__cpuinit cpuid4_cache_lookup_regs(int index, - struct _cpuid4_info_regs *this_leaf) +cpuid4_cache_lookup_regs(int index, struct _cpuid4_info_regs *this_leaf) { union _cpuid4_leaf_eax eax; union _cpuid4_leaf_ebx ebx; @@ -539,7 +536,11 @@ __cpuinit cpuid4_cache_lookup_regs(int index, unsigned edx; if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD) { - amd_cpuid4(index, &eax, &ebx, &ecx); + if (cpu_has_topoext) + cpuid_count(0x8000001d, index, &eax.full, + &ebx.full, &ecx.full, &edx); + else + amd_cpuid4(index, &eax, &ebx, &ecx); amd_init_l3_cache(this_leaf, index); } else { cpuid_count(4, index, &eax.full, &ebx.full, &ecx.full, &edx); @@ -558,22 +559,40 @@ __cpuinit cpuid4_cache_lookup_regs(int index, return 0; } -static int __cpuinit find_num_cache_leaves(void) +static int find_num_cache_leaves(struct cpuinfo_x86 *c) { - unsigned int eax, ebx, ecx, edx; + unsigned int eax, ebx, ecx, edx, op; union _cpuid4_leaf_eax cache_eax; int i = -1; + if (c->x86_vendor == X86_VENDOR_AMD) + op = 0x8000001d; + else + op = 4; + do { ++i; - /* Do cpuid(4) loop to find out num_cache_leaves */ - cpuid_count(4, i, &eax, &ebx, &ecx, &edx); + /* Do cpuid(op) loop to find out num_cache_leaves */ + cpuid_count(op, i, &eax, &ebx, &ecx, &edx); cache_eax.full = eax; } while (cache_eax.split.type != CACHE_TYPE_NULL); return i; } -unsigned int __cpuinit init_intel_cacheinfo(struct cpuinfo_x86 *c) +void init_amd_cacheinfo(struct cpuinfo_x86 *c) +{ + + if (cpu_has_topoext) { + num_cache_leaves = find_num_cache_leaves(c); + } else if (c->extended_cpuid_level >= 0x80000006) { + if (cpuid_edx(0x80000006) & 0xf000) + num_cache_leaves = 4; + else + num_cache_leaves = 3; + } +} + +unsigned int init_intel_cacheinfo(struct cpuinfo_x86 *c) { /* Cache sizes */ unsigned int trace = 0, l1i = 0, l1d = 0, l2 = 0, l3 = 0; @@ -589,7 +608,7 @@ unsigned int __cpuinit init_intel_cacheinfo(struct cpuinfo_x86 *c) if (is_initialized == 0) { /* Init num_cache_leaves from boot CPU */ - num_cache_leaves = find_num_cache_leaves(); + num_cache_leaves = find_num_cache_leaves(c); is_initialized++; } @@ -598,36 +617,34 @@ unsigned int __cpuinit init_intel_cacheinfo(struct cpuinfo_x86 *c) * parameters cpuid leaf to find the cache details */ for (i = 0; i < num_cache_leaves; i++) { - struct _cpuid4_info_regs this_leaf; + struct _cpuid4_info_regs this_leaf = {}; int retval; retval = cpuid4_cache_lookup_regs(i, &this_leaf); - if (retval >= 0) { - switch (this_leaf.eax.split.level) { - case 1: - if (this_leaf.eax.split.type == - CACHE_TYPE_DATA) - new_l1d = this_leaf.size/1024; - else if (this_leaf.eax.split.type == - CACHE_TYPE_INST) - new_l1i = this_leaf.size/1024; - break; - case 2: - new_l2 = this_leaf.size/1024; - num_threads_sharing = 1 + this_leaf.eax.split.num_threads_sharing; - index_msb = get_count_order(num_threads_sharing); - l2_id = c->apicid >> index_msb; - break; - case 3: - new_l3 = this_leaf.size/1024; - num_threads_sharing = 1 + this_leaf.eax.split.num_threads_sharing; - index_msb = get_count_order( - num_threads_sharing); - l3_id = c->apicid >> index_msb; - break; - default: - break; - } + if (retval < 0) + continue; + + switch (this_leaf.eax.split.level) { + case 1: + if (this_leaf.eax.split.type == CACHE_TYPE_DATA) + new_l1d = this_leaf.size/1024; + else if (this_leaf.eax.split.type == CACHE_TYPE_INST) + new_l1i = this_leaf.size/1024; + break; + case 2: + new_l2 = this_leaf.size/1024; + num_threads_sharing = 1 + this_leaf.eax.split.num_threads_sharing; + index_msb = get_count_order(num_threads_sharing); + l2_id = c->apicid & ~((1 << index_msb) - 1); + break; + case 3: + new_l3 = this_leaf.size/1024; + num_threads_sharing = 1 + this_leaf.eax.split.num_threads_sharing; + index_msb = get_count_order(num_threads_sharing); + l3_id = c->apicid & ~((1 << index_msb) - 1); + break; + default: + break; } } } @@ -713,6 +730,18 @@ unsigned int __cpuinit init_intel_cacheinfo(struct cpuinfo_x86 *c) #endif } +#ifdef CONFIG_X86_HT + /* + * If cpu_llc_id is not yet set, this means cpuid_level < 4 which in + * turns means that the only possibility is SMT (as indicated in + * cpuid1). Since cpuid2 doesn't specify shared caches, and we know + * that SMT shares all caches, we can unconditionally set cpu_llc_id to + * c->phys_proc_id. + */ + if (per_cpu(cpu_llc_id, cpu) == BAD_APICID) + per_cpu(cpu_llc_id, cpu) = c->phys_proc_id; +#endif + c->x86_cache_size = l3 ? l3 : (l2 ? l2 : (l1i+l1d)); return l2; @@ -725,14 +754,40 @@ static DEFINE_PER_CPU(struct _cpuid4_info *, ici_cpuid4_info); #define CPUID4_INFO_IDX(x, y) (&((per_cpu(ici_cpuid4_info, x))[y])) #ifdef CONFIG_SMP -static void __cpuinit cache_shared_cpu_map_setup(unsigned int cpu, int index) + +static int cache_shared_amd_cpu_map_setup(unsigned int cpu, int index) { - struct _cpuid4_info *this_leaf, *sibling_leaf; - unsigned long num_threads_sharing; - int index_msb, i, sibling; - struct cpuinfo_x86 *c = &cpu_data(cpu); + struct _cpuid4_info *this_leaf; + int i, sibling; + + if (cpu_has_topoext) { + unsigned int apicid, nshared, first, last; + + if (!per_cpu(ici_cpuid4_info, cpu)) + return 0; - if ((index == 3) && (c->x86_vendor == X86_VENDOR_AMD)) { + this_leaf = CPUID4_INFO_IDX(cpu, index); + nshared = this_leaf->base.eax.split.num_threads_sharing + 1; + apicid = cpu_data(cpu).apicid; + first = apicid - (apicid % nshared); + last = first + nshared - 1; + + for_each_online_cpu(i) { + apicid = cpu_data(i).apicid; + if ((apicid < first) || (apicid > last)) + continue; + if (!per_cpu(ici_cpuid4_info, i)) + continue; + this_leaf = CPUID4_INFO_IDX(i, index); + + for_each_online_cpu(sibling) { + apicid = cpu_data(sibling).apicid; + if ((apicid < first) || (apicid > last)) + continue; + set_bit(sibling, this_leaf->shared_cpu_map); + } + } + } else if (index == 3) { for_each_cpu(i, cpu_llc_shared_mask(cpu)) { if (!per_cpu(ici_cpuid4_info, i)) continue; @@ -743,8 +798,24 @@ static void __cpuinit cache_shared_cpu_map_setup(unsigned int cpu, int index) set_bit(sibling, this_leaf->shared_cpu_map); } } - return; + } else + return 0; + + return 1; +} + +static void cache_shared_cpu_map_setup(unsigned int cpu, int index) +{ + struct _cpuid4_info *this_leaf, *sibling_leaf; + unsigned long num_threads_sharing; + int index_msb, i; + struct cpuinfo_x86 *c = &cpu_data(cpu); + + if (c->x86_vendor == X86_VENDOR_AMD) { + if (cache_shared_amd_cpu_map_setup(cpu, index)) + return; } + this_leaf = CPUID4_INFO_IDX(cpu, index); num_threads_sharing = 1 + this_leaf->base.eax.split.num_threads_sharing; @@ -768,7 +839,7 @@ static void __cpuinit cache_shared_cpu_map_setup(unsigned int cpu, int index) } } } -static void __cpuinit cache_remove_shared_cpu_map(unsigned int cpu, int index) +static void cache_remove_shared_cpu_map(unsigned int cpu, int index) { struct _cpuid4_info *this_leaf, *sibling_leaf; int sibling; @@ -781,16 +852,16 @@ static void __cpuinit cache_remove_shared_cpu_map(unsigned int cpu, int index) } } #else -static void __cpuinit cache_shared_cpu_map_setup(unsigned int cpu, int index) +static void cache_shared_cpu_map_setup(unsigned int cpu, int index) { } -static void __cpuinit cache_remove_shared_cpu_map(unsigned int cpu, int index) +static void cache_remove_shared_cpu_map(unsigned int cpu, int index) { } #endif -static void __cpuinit free_cache_attributes(unsigned int cpu) +static void free_cache_attributes(unsigned int cpu) { int i; @@ -801,7 +872,7 @@ static void __cpuinit free_cache_attributes(unsigned int cpu) per_cpu(ici_cpuid4_info, cpu) = NULL; } -static void __cpuinit get_cpu_leaves(void *_retval) +static void get_cpu_leaves(void *_retval) { int j, *retval = _retval, cpu = smp_processor_id(); @@ -821,7 +892,7 @@ static void __cpuinit get_cpu_leaves(void *_retval) } } -static int __cpuinit detect_cache_attributes(unsigned int cpu) +static int detect_cache_attributes(unsigned int cpu) { int retval; @@ -955,7 +1026,7 @@ static struct attribute *default_attrs[] = { }; #ifdef CONFIG_AMD_NB -static struct attribute ** __cpuinit amd_l3_attrs(void) +static struct attribute **amd_l3_attrs(void) { static struct attribute **attrs; int n; @@ -963,7 +1034,7 @@ static struct attribute ** __cpuinit amd_l3_attrs(void) if (attrs) return attrs; - n = sizeof (default_attrs) / sizeof (struct attribute *); + n = ARRAY_SIZE(default_attrs); if (amd_nb_has_feature(AMD_NB_L3_INDEX_DISABLE)) n += 2; @@ -1031,7 +1102,7 @@ static struct kobj_type ktype_percpu_entry = { .sysfs_ops = &sysfs_ops, }; -static void __cpuinit cpuid4_cache_sysfs_exit(unsigned int cpu) +static void cpuid4_cache_sysfs_exit(unsigned int cpu) { kfree(per_cpu(ici_cache_kobject, cpu)); kfree(per_cpu(ici_index_kobject, cpu)); @@ -1040,7 +1111,7 @@ static void __cpuinit cpuid4_cache_sysfs_exit(unsigned int cpu) free_cache_attributes(cpu); } -static int __cpuinit cpuid4_cache_sysfs_init(unsigned int cpu) +static int cpuid4_cache_sysfs_init(unsigned int cpu) { int err; @@ -1072,7 +1143,7 @@ err_out: static DECLARE_BITMAP(cache_dev_map, NR_CPUS); /* Add/Remove cache interface for CPU device */ -static int __cpuinit cache_add_dev(struct device *dev) +static int cache_add_dev(struct device *dev) { unsigned int cpu = dev->id; unsigned long i, j; @@ -1123,7 +1194,7 @@ static int __cpuinit cache_add_dev(struct device *dev) return 0; } -static void __cpuinit cache_remove_dev(struct device *dev) +static void cache_remove_dev(struct device *dev) { unsigned int cpu = dev->id; unsigned long i; @@ -1140,8 +1211,8 @@ static void __cpuinit cache_remove_dev(struct device *dev) cpuid4_cache_sysfs_exit(cpu); } -static int __cpuinit cacheinfo_cpu_callback(struct notifier_block *nfb, - unsigned long action, void *hcpu) +static int cacheinfo_cpu_callback(struct notifier_block *nfb, + unsigned long action, void *hcpu) { unsigned int cpu = (unsigned long)hcpu; struct device *dev; @@ -1160,27 +1231,30 @@ static int __cpuinit cacheinfo_cpu_callback(struct notifier_block *nfb, return NOTIFY_OK; } -static struct notifier_block __cpuinitdata cacheinfo_cpu_notifier = { +static struct notifier_block cacheinfo_cpu_notifier = { .notifier_call = cacheinfo_cpu_callback, }; -static int __cpuinit cache_sysfs_init(void) +static int __init cache_sysfs_init(void) { - int i; + int i, err = 0; if (num_cache_leaves == 0) return 0; + cpu_notifier_register_begin(); for_each_online_cpu(i) { - int err; struct device *dev = get_cpu_device(i); err = cache_add_dev(dev); if (err) - return err; + goto out; } - register_hotcpu_notifier(&cacheinfo_cpu_notifier); - return 0; + __register_hotcpu_notifier(&cacheinfo_cpu_notifier); + +out: + cpu_notifier_register_done(); + return err; } device_initcall(cache_sysfs_init); diff --git a/arch/x86/kernel/cpu/match.c b/arch/x86/kernel/cpu/match.c new file mode 100644 index 00000000000..afa9f0d487e --- /dev/null +++ b/arch/x86/kernel/cpu/match.c @@ -0,0 +1,49 @@ +#include <asm/cpu_device_id.h> +#include <asm/processor.h> +#include <linux/cpu.h> +#include <linux/module.h> +#include <linux/slab.h> + +/** + * x86_match_cpu - match current CPU again an array of x86_cpu_ids + * @match: Pointer to array of x86_cpu_ids. Last entry terminated with + * {}. + * + * Return the entry if the current CPU matches the entries in the + * passed x86_cpu_id match table. Otherwise NULL. The match table + * contains vendor (X86_VENDOR_*), family, model and feature bits or + * respective wildcard entries. + * + * A typical table entry would be to match a specific CPU + * { X86_VENDOR_INTEL, 6, 0x12 } + * or to match a specific CPU feature + * { X86_FEATURE_MATCH(X86_FEATURE_FOOBAR) } + * + * Fields can be wildcarded with %X86_VENDOR_ANY, %X86_FAMILY_ANY, + * %X86_MODEL_ANY, %X86_FEATURE_ANY or 0 (except for vendor) + * + * Arrays used to match for this should also be declared using + * MODULE_DEVICE_TABLE(x86cpu, ...) + * + * This always matches against the boot cpu, assuming models and features are + * consistent over all CPUs. + */ +const struct x86_cpu_id *x86_match_cpu(const struct x86_cpu_id *match) +{ + const struct x86_cpu_id *m; + struct cpuinfo_x86 *c = &boot_cpu_data; + + for (m = match; m->vendor | m->family | m->model | m->feature; m++) { + if (m->vendor != X86_VENDOR_ANY && c->x86_vendor != m->vendor) + continue; + if (m->family != X86_FAMILY_ANY && c->x86 != m->family) + continue; + if (m->model != X86_MODEL_ANY && c->x86_model != m->model) + continue; + if (m->feature != X86_FEATURE_ANY && !cpu_has(c, m->feature)) + continue; + return m; + } + return NULL; +} +EXPORT_SYMBOL(x86_match_cpu); diff --git a/arch/x86/kernel/cpu/mcheck/mce-apei.c b/arch/x86/kernel/cpu/mcheck/mce-apei.c index 507ea58688e..a1aef953315 100644 --- a/arch/x86/kernel/cpu/mcheck/mce-apei.c +++ b/arch/x86/kernel/cpu/mcheck/mce-apei.c @@ -33,22 +33,28 @@ #include <linux/acpi.h> #include <linux/cper.h> #include <acpi/apei.h> +#include <acpi/ghes.h> #include <asm/mce.h> #include "mce-internal.h" -void apei_mce_report_mem_error(int corrected, struct cper_sec_mem_err *mem_err) +void apei_mce_report_mem_error(int severity, struct cper_sec_mem_err *mem_err) { struct mce m; - /* Only corrected MC is reported */ - if (!corrected) + if (!(mem_err->validation_bits & CPER_MEM_VALID_PA)) return; mce_setup(&m); m.bank = 1; - /* Fake a memory read corrected error with unknown channel */ + /* Fake a memory read error with unknown channel */ m.status = MCI_STATUS_VAL | MCI_STATUS_EN | MCI_STATUS_ADDRV | 0x9f; + + if (severity >= GHES_SEV_RECOVERABLE) + m.status |= MCI_STATUS_UC; + if (severity >= GHES_SEV_PANIC) + m.status |= MCI_STATUS_PCC; + m.addr = mem_err->physical_addr; mce_log(&m); mce_notify_irq(); diff --git a/arch/x86/kernel/cpu/mcheck/mce-inject.c b/arch/x86/kernel/cpu/mcheck/mce-inject.c index fc4beb39357..5ac2d1fb28b 100644 --- a/arch/x86/kernel/cpu/mcheck/mce-inject.c +++ b/arch/x86/kernel/cpu/mcheck/mce-inject.c @@ -78,6 +78,7 @@ static void raise_exception(struct mce *m, struct pt_regs *pregs) } static cpumask_var_t mce_inject_cpumask; +static DEFINE_MUTEX(mce_inject_mutex); static int mce_raise_notify(unsigned int cmd, struct pt_regs *regs) { @@ -152,7 +153,7 @@ static void raise_mce(struct mce *m) return; #ifdef CONFIG_X86_LOCAL_APIC - if (m->inject_flags & (MCJ_IRQ_BRAODCAST | MCJ_NMI_BROADCAST)) { + if (m->inject_flags & (MCJ_IRQ_BROADCAST | MCJ_NMI_BROADCAST)) { unsigned long start; int cpu; @@ -166,7 +167,7 @@ static void raise_mce(struct mce *m) cpumask_clear_cpu(cpu, mce_inject_cpumask); } if (!cpumask_empty(mce_inject_cpumask)) { - if (m->inject_flags & MCJ_IRQ_BRAODCAST) { + if (m->inject_flags & MCJ_IRQ_BROADCAST) { /* * don't wait because mce_irq_ipi is necessary * to be sync with following raise_local @@ -194,7 +195,11 @@ static void raise_mce(struct mce *m) put_online_cpus(); } else #endif + { + preempt_disable(); raise_local(); + preempt_enable(); + } } /* Error injection interface */ @@ -225,7 +230,10 @@ static ssize_t mce_write(struct file *filp, const char __user *ubuf, * so do it a jiffie or two later everywhere. */ schedule_timeout(2); + + mutex_lock(&mce_inject_mutex); raise_mce(&m); + mutex_unlock(&mce_inject_mutex); return usize; } diff --git a/arch/x86/kernel/cpu/mcheck/mce-internal.h b/arch/x86/kernel/cpu/mcheck/mce-internal.h index ed44c8a6585..09edd0b65fe 100644 --- a/arch/x86/kernel/cpu/mcheck/mce-internal.h +++ b/arch/x86/kernel/cpu/mcheck/mce-internal.h @@ -24,9 +24,22 @@ struct mce_bank { int mce_severity(struct mce *a, int tolerant, char **msg); struct dentry *mce_get_debugfs_dir(void); -extern int mce_ser; - extern struct mce_bank *mce_banks; +extern mce_banks_t mce_banks_ce_disabled; + +#ifdef CONFIG_X86_MCE_INTEL +unsigned long mce_intel_adjust_timer(unsigned long interval); +void mce_intel_cmci_poll(void); +void mce_intel_hcpu_update(unsigned long cpu); +void cmci_disable_bank(int bank); +#else +# define mce_intel_adjust_timer mce_adjust_timer_default +static inline void mce_intel_cmci_poll(void) { } +static inline void mce_intel_hcpu_update(unsigned long cpu) { } +static inline void cmci_disable_bank(int bank) { } +#endif + +void mce_timer_kick(unsigned long interval); #ifdef CONFIG_ACPI_APEI int apei_write_mce(struct mce *m); diff --git a/arch/x86/kernel/cpu/mcheck/mce-severity.c b/arch/x86/kernel/cpu/mcheck/mce-severity.c index 7395d5f4272..c370e1c4468 100644 --- a/arch/x86/kernel/cpu/mcheck/mce-severity.c +++ b/arch/x86/kernel/cpu/mcheck/mce-severity.c @@ -54,7 +54,7 @@ static struct severity { #define MASK(x, y) .mask = x, .result = y #define MCI_UC_S (MCI_STATUS_UC|MCI_STATUS_S) #define MCI_UC_SAR (MCI_STATUS_UC|MCI_STATUS_S|MCI_STATUS_AR) -#define MCACOD 0xffff +#define MCI_ADDR (MCI_STATUS_ADDRV|MCI_STATUS_MISCV) MCESEV( NO, "Invalid", @@ -102,11 +102,29 @@ static struct severity { SER, BITCLR(MCI_STATUS_S) ), - /* AR add known MCACODs here */ MCESEV( PANIC, "Action required with lost events", SER, BITSET(MCI_STATUS_OVER|MCI_UC_SAR) ), + + /* known AR MCACODs: */ +#ifdef CONFIG_MEMORY_FAILURE + MCESEV( + KEEP, "Action required but unaffected thread is continuable", + SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|MCI_ADDR, MCI_UC_SAR|MCI_ADDR), + MCGMASK(MCG_STATUS_RIPV|MCG_STATUS_EIPV, MCG_STATUS_RIPV) + ), + MCESEV( + AR, "Action required: data load error in a user process", + SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|MCI_ADDR|MCACOD, MCI_UC_SAR|MCI_ADDR|MCACOD_DATA), + USER + ), + MCESEV( + AR, "Action required: instruction fetch error in a user process", + SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|MCI_ADDR|MCACOD, MCI_UC_SAR|MCI_ADDR|MCACOD_INSTR), + USER + ), +#endif MCESEV( PANIC, "Action required: unknown MCACOD", SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR, MCI_UC_SAR) @@ -115,11 +133,11 @@ static struct severity { /* known AO MCACODs: */ MCESEV( AO, "Action optional: memory scrubbing error", - SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|0xfff0, MCI_UC_S|0x00c0) + SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|MCACOD_SCRUBMSK, MCI_UC_S|MCACOD_SCRUB) ), MCESEV( AO, "Action optional: last level cache writeback error", - SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|MCACOD, MCI_UC_S|0x017a) + SER, MASK(MCI_STATUS_OVER|MCI_UC_SAR|MCACOD, MCI_UC_S|MCACOD_L3WB) ), MCESEV( SOME, "Action optional: unknown MCACOD", @@ -145,15 +163,19 @@ static struct severity { }; /* - * If the EIPV bit is set, it means the saved IP is the - * instruction which caused the MCE. + * If mcgstatus indicated that ip/cs on the stack were + * no good, then "m->cs" will be zero and we will have + * to assume the worst case (IN_KERNEL) as we actually + * have no idea what we were executing when the machine + * check hit. + * If we do have a good "m->cs" (or a faked one in the + * case we were executing in VM86 mode) we can use it to + * distinguish an exception taken in user from from one + * taken in the kernel. */ static int error_context(struct mce *m) { - if (m->mcgstatus & MCG_STATUS_EIPV) - return (m->ip && (m->cs & 3) == 3) ? IN_USER : IN_KERNEL; - /* Unknown, assume kernel */ - return IN_KERNEL; + return ((m->cs & 3) == 3) ? IN_USER : IN_KERNEL; } int mce_severity(struct mce *m, int tolerant, char **msg) @@ -166,9 +188,9 @@ int mce_severity(struct mce *m, int tolerant, char **msg) continue; if ((m->mcgstatus & s->mcgmask) != s->mcgres) continue; - if (s->ser == SER_REQUIRED && !mce_ser) + if (s->ser == SER_REQUIRED && !mca_cfg.ser) continue; - if (s->ser == NO_SER && mce_ser) + if (s->ser == NO_SER && mca_cfg.ser) continue; if (s->context && ctx != s->context) continue; diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c index 5a11ae2e9e9..9a79c8dbd8e 100644 --- a/arch/x86/kernel/cpu/mcheck/mce.c +++ b/arch/x86/kernel/cpu/mcheck/mce.c @@ -7,6 +7,9 @@ * Copyright 2008 Intel Corporation * Author: Andi Kleen */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/thread_info.h> #include <linux/capability.h> #include <linux/miscdevice.h> @@ -55,35 +58,24 @@ static DEFINE_MUTEX(mce_chrdev_read_mutex); #define CREATE_TRACE_POINTS #include <trace/events/mce.h> -int mce_disabled __read_mostly; - -#define MISC_MCELOG_MINOR 227 - #define SPINUNIT 100 /* 100ns */ -atomic_t mce_entry; - DEFINE_PER_CPU(unsigned, mce_exception_count); -/* - * Tolerant levels: - * 0: always panic on uncorrected errors, log corrected errors - * 1: panic or SIGBUS on uncorrected errors, log corrected errors - * 2: SIGBUS or log uncorrected errors (if possible), log corrected errors - * 3: never panic or SIGBUS, log all errors (for testing only) - */ -static int tolerant __read_mostly = 1; -static int banks __read_mostly; -static int rip_msr __read_mostly; -static int mce_bootlog __read_mostly = -1; -static int monarch_timeout __read_mostly = -1; -static int mce_panic_timeout __read_mostly; -static int mce_dont_log_ce __read_mostly; -int mce_cmci_disabled __read_mostly; -int mce_ignore_ce __read_mostly; -int mce_ser __read_mostly; - -struct mce_bank *mce_banks __read_mostly; +struct mce_bank *mce_banks __read_mostly; + +struct mca_config mca_cfg __read_mostly = { + .bootlog = -1, + /* + * Tolerant levels: + * 0: always panic on uncorrected errors, log corrected errors + * 1: panic or SIGBUS on uncorrected errors, log corrected errors + * 2: SIGBUS or log uncorrected errors (if possible), log corr. errors + * 3: never panic or SIGBUS, log all errors (for testing only) + */ + .tolerant = 1, + .monarch_timeout = -1 +}; /* User mode helper program triggered by machine check event */ static unsigned long mce_need_notify; @@ -95,13 +87,30 @@ static DECLARE_WAIT_QUEUE_HEAD(mce_chrdev_wait); static DEFINE_PER_CPU(struct mce, mces_seen); static int cpu_missing; -/* MCA banks polled by the period polling timer for corrected events */ +/* CMCI storm detection filter */ +static DEFINE_PER_CPU(unsigned long, mce_polled_error); + +/* + * MCA banks polled by the period polling timer for corrected events. + * With Intel CMCI, this only has MCA banks which do not support CMCI (if any). + */ DEFINE_PER_CPU(mce_banks_t, mce_poll_banks) = { [0 ... BITS_TO_LONGS(MAX_NR_BANKS)-1] = ~0UL }; +/* + * MCA banks controlled through firmware first for corrected errors. + * This is a global list of banks for which we won't enable CMCI and we + * won't poll. Firmware controls these banks and is responsible for + * reporting corrected errors through GHES. Uncorrected/recoverable + * errors are still notified through a machine check. + */ +mce_banks_t mce_banks_ce_disabled; + static DEFINE_PER_CPU(struct work_struct, mce_work); +static void (*quirk_no_way_out)(int bank, struct mce *m, struct pt_regs *regs); + /* * CPU/chipset specific EDAC code can register a notifier call here to print * MCE errors in a human-readable form. @@ -191,7 +200,7 @@ static void drain_mcelog_buffer(void) { unsigned int next, i, prev = 0; - next = rcu_dereference_check_mce(mcelog.next); + next = ACCESS_ONCE(mcelog.next); do { struct mce *m; @@ -210,7 +219,7 @@ static void drain_mcelog_buffer(void) cpu_relax(); if (!m->finished && retries >= 4) { - pr_err("MCE: skipping error being logged currently!\n"); + pr_err("skipping error being logged currently!\n"); break; } } @@ -298,7 +307,7 @@ static void wait_for_panic(void) while (timeout-- > 0) udelay(1); if (panic_timeout == 0) - panic_timeout = mce_panic_timeout; + panic_timeout = mca_cfg.panic_timeout; panic("Panicing machine check CPU died"); } @@ -356,7 +365,7 @@ static void mce_panic(char *msg, struct mce *final, char *exp) pr_emerg(HW_ERR "Machine check: %s\n", exp); if (!fake_panic) { if (panic_timeout == 0) - panic_timeout = mce_panic_timeout; + panic_timeout = mca_cfg.panic_timeout; panic(msg); } else pr_emerg(HW_ERR "Fake kernel panic: %s\n", msg); @@ -368,7 +377,7 @@ static int msr_to_offset(u32 msr) { unsigned bank = __this_cpu_read(injectm.bank); - if (msr == rip_msr) + if (msr == mca_cfg.rip_msr) return offsetof(struct mce, ip); if (msr == MSR_IA32_MCx_STATUS(bank)) return offsetof(struct mce, status); @@ -437,10 +446,18 @@ static inline void mce_gather_info(struct mce *m, struct pt_regs *regs) if (m->mcgstatus & (MCG_STATUS_RIPV|MCG_STATUS_EIPV)) { m->ip = regs->ip; m->cs = regs->cs; + + /* + * When in VM86 mode make the cs look like ring 3 + * always. This is a lie, but it's better than passing + * the additional vm86 bit around everywhere. + */ + if (v8086_mode(regs)) + m->cs |= 3; } /* Use accurate RIP reporting if available. */ - if (rip_msr) - m->ip = mce_rdmsrl(rip_msr); + if (mca_cfg.rip_msr) + m->ip = mce_rdmsrl(mca_cfg.rip_msr); } } @@ -501,18 +518,15 @@ static int mce_ring_add(unsigned long pfn) int mce_available(struct cpuinfo_x86 *c) { - if (mce_disabled) + if (mca_cfg.disabled) return 0; return cpu_has(c, X86_FEATURE_MCE) && cpu_has(c, X86_FEATURE_MCA); } static void mce_schedule_work(void) { - if (!mce_ring_empty()) { - struct work_struct *work = &__get_cpu_var(mce_work); - if (!work_pending(work)) - schedule_work(work); - } + if (!mce_ring_empty()) + schedule_work(&__get_cpu_var(mce_work)); } DEFINE_PER_CPU(struct irq_work, mce_irq_work); @@ -540,6 +554,27 @@ static void mce_report_event(struct pt_regs *regs) irq_work_queue(&__get_cpu_var(mce_irq_work)); } +/* + * Read ADDR and MISC registers. + */ +static void mce_read_aux(struct mce *m, int i) +{ + if (m->status & MCI_STATUS_MISCV) + m->misc = mce_rdmsrl(MSR_IA32_MCx_MISC(i)); + if (m->status & MCI_STATUS_ADDRV) { + m->addr = mce_rdmsrl(MSR_IA32_MCx_ADDR(i)); + + /* + * Mask the reported address by the reported granularity. + */ + if (mca_cfg.ser && (m->status & MCI_STATUS_MISCV)) { + u8 shift = MCI_MISC_ADDR_LSB(m->misc); + m->addr >>= shift; + m->addr <<= shift; + } + } +} + DEFINE_PER_CPU(unsigned, mce_poll_count); /* @@ -562,11 +597,11 @@ void machine_check_poll(enum mcp_flags flags, mce_banks_t *b) struct mce m; int i; - percpu_inc(mce_poll_count); + this_cpu_inc(mce_poll_count); mce_gather_info(&m, NULL); - for (i = 0; i < banks; i++) { + for (i = 0; i < mca_cfg.banks; i++) { if (!mce_banks[i].ctl || !test_bit(i, *b)) continue; @@ -580,6 +615,7 @@ void machine_check_poll(enum mcp_flags flags, mce_banks_t *b) if (!(m.status & MCI_STATUS_VAL)) continue; + this_cpu_write(mce_polled_error, 1); /* * Uncorrected or signalled events are handled by the exception * handler when it is enabled, so don't process those here. @@ -587,13 +623,10 @@ void machine_check_poll(enum mcp_flags flags, mce_banks_t *b) * TBD do the same check for MCI_STATUS_EN here? */ if (!(flags & MCP_UC) && - (m.status & (mce_ser ? MCI_STATUS_S : MCI_STATUS_UC))) + (m.status & (mca_cfg.ser ? MCI_STATUS_S : MCI_STATUS_UC))) continue; - if (m.status & MCI_STATUS_MISCV) - m.misc = mce_rdmsrl(MSR_IA32_MCx_MISC(i)); - if (m.status & MCI_STATUS_ADDRV) - m.addr = mce_rdmsrl(MSR_IA32_MCx_ADDR(i)); + mce_read_aux(&m, i); if (!(flags & MCP_TIMESTAMP)) m.tsc = 0; @@ -601,7 +634,7 @@ void machine_check_poll(enum mcp_flags flags, mce_banks_t *b) * Don't get the IP here because it's unlikely to * have anything to do with the actual error location. */ - if (!(flags & MCP_DONTLOG) && !mce_dont_log_ce) + if (!(flags & MCP_DONTLOG) && !mca_cfg.dont_log_ce) mce_log(&m); /* @@ -623,16 +656,22 @@ EXPORT_SYMBOL_GPL(machine_check_poll); * Do a quick check if any of the events requires a panic. * This decides if we keep the events around or clear them. */ -static int mce_no_way_out(struct mce *m, char **msg) +static int mce_no_way_out(struct mce *m, char **msg, unsigned long *validp, + struct pt_regs *regs) { - int i; + int i, ret = 0; - for (i = 0; i < banks; i++) { + for (i = 0; i < mca_cfg.banks; i++) { m->status = mce_rdmsrl(MSR_IA32_MCx_STATUS(i)); - if (mce_severity(m, tolerant, msg) >= MCE_PANIC_SEVERITY) - return 1; + if (m->status & MCI_STATUS_VAL) { + __set_bit(i, validp); + if (quirk_no_way_out) + quirk_no_way_out(i, m, regs); + } + if (mce_severity(m, mca_cfg.tolerant, msg) >= MCE_PANIC_SEVERITY) + ret = 1; } - return 0; + return ret; } /* @@ -660,11 +699,10 @@ static int mce_timed_out(u64 *t) rmb(); if (atomic_read(&mce_paniced)) wait_for_panic(); - if (!monarch_timeout) + if (!mca_cfg.monarch_timeout) goto out; if ((s64)*t < SPINUNIT) { - /* CHECKME: Make panic default for 1 too? */ - if (tolerant < 1) + if (mca_cfg.tolerant <= 1) mce_panic("Timeout synchronizing machine check over CPUs", NULL, NULL); cpu_missing = 1; @@ -714,7 +752,8 @@ static void mce_reign(void) * Grade the severity of the errors of all the CPUs. */ for_each_possible_cpu(cpu) { - int severity = mce_severity(&per_cpu(mces_seen, cpu), tolerant, + int severity = mce_severity(&per_cpu(mces_seen, cpu), + mca_cfg.tolerant, &nmsg); if (severity > global_worst) { msg = nmsg; @@ -728,7 +767,7 @@ static void mce_reign(void) * This dumps all the mces in the log buffer and stops the * other CPUs. */ - if (m && global_worst >= MCE_PANIC_SEVERITY && tolerant < 3) + if (m && global_worst >= MCE_PANIC_SEVERITY && mca_cfg.tolerant < 3) mce_panic("Fatal Machine check", m, msg); /* @@ -741,7 +780,7 @@ static void mce_reign(void) * No machine check event found. Must be some external * source or one CPU is hung. Panic. */ - if (global_worst <= MCE_KEEP_SEVERITY && tolerant < 3) + if (global_worst <= MCE_KEEP_SEVERITY && mca_cfg.tolerant < 3) mce_panic("Machine check from unknown source", NULL, NULL); /* @@ -765,7 +804,7 @@ static int mce_start(int *no_way_out) { int order; int cpus = num_online_cpus(); - u64 timeout = (u64)monarch_timeout * NSEC_PER_USEC; + u64 timeout = (u64)mca_cfg.monarch_timeout * NSEC_PER_USEC; if (!timeout) return -1; @@ -829,7 +868,7 @@ static int mce_start(int *no_way_out) static int mce_end(int order) { int ret = -1; - u64 timeout = (u64)monarch_timeout * NSEC_PER_USEC; + u64 timeout = (u64)mca_cfg.monarch_timeout * NSEC_PER_USEC; if (!timeout) goto reset; @@ -910,13 +949,58 @@ static void mce_clear_state(unsigned long *toclear) { int i; - for (i = 0; i < banks; i++) { + for (i = 0; i < mca_cfg.banks; i++) { if (test_bit(i, toclear)) mce_wrmsrl(MSR_IA32_MCx_STATUS(i), 0); } } /* + * Need to save faulting physical address associated with a process + * in the machine check handler some place where we can grab it back + * later in mce_notify_process() + */ +#define MCE_INFO_MAX 16 + +struct mce_info { + atomic_t inuse; + struct task_struct *t; + __u64 paddr; + int restartable; +} mce_info[MCE_INFO_MAX]; + +static void mce_save_info(__u64 addr, int c) +{ + struct mce_info *mi; + + for (mi = mce_info; mi < &mce_info[MCE_INFO_MAX]; mi++) { + if (atomic_cmpxchg(&mi->inuse, 0, 1) == 0) { + mi->t = current; + mi->paddr = addr; + mi->restartable = c; + return; + } + } + + mce_panic("Too many concurrent recoverable errors", NULL, NULL); +} + +static struct mce_info *mce_find_info(void) +{ + struct mce_info *mi; + + for (mi = mce_info; mi < &mce_info[MCE_INFO_MAX]; mi++) + if (atomic_read(&mi->inuse) && mi->t == current) + return mi; + return NULL; +} + +static void mce_clear_info(struct mce_info *mi) +{ + atomic_set(&mi->inuse, 0); +} + +/* * The actual machine check handler. This only handles real * exceptions when something got corrupted coming in through int 18. * @@ -930,6 +1014,7 @@ static void mce_clear_state(unsigned long *toclear) */ void do_machine_check(struct pt_regs *regs, long error_code) { + struct mca_config *cfg = &mca_cfg; struct mce m, *final; int i; int worst = 0; @@ -941,7 +1026,7 @@ void do_machine_check(struct pt_regs *regs, long error_code) int order; /* * If no_way_out gets set, there is no safe way to recover from this - * MCE. If tolerant is cranked up, we'll try anyway. + * MCE. If mca_cfg.tolerant is cranked up, we'll try anyway. */ int no_way_out = 0; /* @@ -950,13 +1035,12 @@ void do_machine_check(struct pt_regs *regs, long error_code) */ int kill_it = 0; DECLARE_BITMAP(toclear, MAX_NR_BANKS); + DECLARE_BITMAP(valid_banks, MAX_NR_BANKS); char *msg = "Unknown"; - atomic_inc(&mce_entry); - - percpu_inc(mce_exception_count); + this_cpu_inc(mce_exception_count); - if (!banks) + if (!cfg->banks) goto out; mce_gather_info(&m, regs); @@ -964,12 +1048,15 @@ void do_machine_check(struct pt_regs *regs, long error_code) final = &__get_cpu_var(mces_seen); *final = m; - no_way_out = mce_no_way_out(&m, &msg); + memset(valid_banks, 0, sizeof(valid_banks)); + no_way_out = mce_no_way_out(&m, &msg, valid_banks, regs); barrier(); /* - * When no restart IP must always kill or panic. + * When no restart IP might need to kill or panic. + * Assume the worst for now, but if we find the + * severity is MCE_AR_SEVERITY we have other options. */ if (!(m.mcgstatus & MCG_STATUS_RIPV)) kill_it = 1; @@ -980,8 +1067,10 @@ void do_machine_check(struct pt_regs *regs, long error_code) * because the first one to see it will clear it. */ order = mce_start(&no_way_out); - for (i = 0; i < banks; i++) { + for (i = 0; i < cfg->banks; i++) { __clear_bit(i, toclear); + if (!test_bit(i, valid_banks)) + continue; if (!mce_banks[i].ctl) continue; @@ -997,16 +1086,16 @@ void do_machine_check(struct pt_regs *regs, long error_code) * Non uncorrected or non signaled errors are handled by * machine_check_poll. Leave them alone, unless this panics. */ - if (!(m.status & (mce_ser ? MCI_STATUS_S : MCI_STATUS_UC)) && + if (!(m.status & (cfg->ser ? MCI_STATUS_S : MCI_STATUS_UC)) && !no_way_out) continue; /* * Set taint even when machine check was not enabled. */ - add_taint(TAINT_MACHINE_CHECK); + add_taint(TAINT_MACHINE_CHECK, LOCKDEP_NOW_UNRELIABLE); - severity = mce_severity(&m, tolerant, NULL); + severity = mce_severity(&m, cfg->tolerant, NULL); /* * When machine check was for corrected handler don't touch, @@ -1023,23 +1112,14 @@ void do_machine_check(struct pt_regs *regs, long error_code) continue; } - /* - * Kill on action required. - */ - if (severity == MCE_AR_SEVERITY) - kill_it = 1; - - if (m.status & MCI_STATUS_MISCV) - m.misc = mce_rdmsrl(MSR_IA32_MCx_MISC(i)); - if (m.status & MCI_STATUS_ADDRV) - m.addr = mce_rdmsrl(MSR_IA32_MCx_ADDR(i)); + mce_read_aux(&m, i); /* * Action optional error. Queue address for later processing. * When the ring overflows we just ignore the AO error. * RED-PEN add some logging mechanism when * usable_address or mce_add_ring fails. - * RED-PEN don't ignore overflow for tolerant == 0 + * RED-PEN don't ignore overflow for mca_cfg.tolerant == 0 */ if (severity == MCE_AO_SEVERITY && mce_usable_address(&m)) mce_ring_add(m.addr >> PAGE_SHIFT); @@ -1052,6 +1132,9 @@ void do_machine_check(struct pt_regs *regs, long error_code) } } + /* mce_clear_state will clear *final, save locally for use later */ + m = *final; + if (!no_way_out) mce_clear_state(toclear); @@ -1063,65 +1146,91 @@ void do_machine_check(struct pt_regs *regs, long error_code) no_way_out = worst >= MCE_PANIC_SEVERITY; /* - * If we have decided that we just CAN'T continue, and the user - * has not set tolerant to an insane level, give up and die. - * - * This is mainly used in the case when the system doesn't - * support MCE broadcasting or it has been disabled. + * At insane "tolerant" levels we take no action. Otherwise + * we only die if we have no other choice. For less serious + * issues we try to recover, or limit damage to the current + * process. */ - if (no_way_out && tolerant < 3) - mce_panic("Fatal machine check on current CPU", final, msg); - - /* - * If the error seems to be unrecoverable, something should be - * done. Try to kill as little as possible. If we can kill just - * one task, do that. If the user has set the tolerance very - * high, don't try to do anything at all. - */ - - if (kill_it && tolerant < 3) - force_sig(SIGBUS, current); - - /* notify userspace ASAP */ - set_thread_flag(TIF_MCE_NOTIFY); + if (cfg->tolerant < 3) { + if (no_way_out) + mce_panic("Fatal machine check on current CPU", &m, msg); + if (worst == MCE_AR_SEVERITY) { + /* schedule action before return to userland */ + mce_save_info(m.addr, m.mcgstatus & MCG_STATUS_RIPV); + set_thread_flag(TIF_MCE_NOTIFY); + } else if (kill_it) { + force_sig(SIGBUS, current); + } + } if (worst > 0) mce_report_event(regs); mce_wrmsrl(MSR_IA32_MCG_STATUS, 0); out: - atomic_dec(&mce_entry); sync_core(); } EXPORT_SYMBOL_GPL(do_machine_check); -/* dummy to break dependency. actual code is in mm/memory-failure.c */ -void __attribute__((weak)) memory_failure(unsigned long pfn, int vector) +#ifndef CONFIG_MEMORY_FAILURE +int memory_failure(unsigned long pfn, int vector, int flags) { - printk(KERN_ERR "Action optional memory failure at %lx ignored\n", pfn); + /* mce_severity() should not hand us an ACTION_REQUIRED error */ + BUG_ON(flags & MF_ACTION_REQUIRED); + pr_err("Uncorrected memory error in page 0x%lx ignored\n" + "Rebuild kernel with CONFIG_MEMORY_FAILURE=y for smarter handling\n", + pfn); + + return 0; } +#endif /* - * Called after mce notification in process context. This code - * is allowed to sleep. Call the high level VM handler to process - * any corrupted pages. - * Assume that the work queue code only calls this one at a time - * per CPU. - * Note we don't disable preemption, so this code might run on the wrong - * CPU. In this case the event is picked up by the scheduled work queue. - * This is merely a fast path to expedite processing in some common - * cases. + * Called in process context that interrupted by MCE and marked with + * TIF_MCE_NOTIFY, just before returning to erroneous userland. + * This code is allowed to sleep. + * Attempt possible recovery such as calling the high level VM handler to + * process any corrupted pages, and kill/signal current process if required. + * Action required errors are handled here. */ void mce_notify_process(void) { unsigned long pfn; - mce_notify_irq(); - while (mce_ring_get(&pfn)) - memory_failure(pfn, MCE_VECTOR); + struct mce_info *mi = mce_find_info(); + int flags = MF_ACTION_REQUIRED; + + if (!mi) + mce_panic("Lost physical address for unconsumed uncorrectable error", NULL, NULL); + pfn = mi->paddr >> PAGE_SHIFT; + + clear_thread_flag(TIF_MCE_NOTIFY); + + pr_err("Uncorrected hardware memory error in user-access at %llx", + mi->paddr); + /* + * We must call memory_failure() here even if the current process is + * doomed. We still need to mark the page as poisoned and alert any + * other users of the page. + */ + if (!mi->restartable) + flags |= MF_MUST_KILL; + if (memory_failure(pfn, MCE_VECTOR, flags) < 0) { + pr_err("Memory error not recovered"); + force_sig(SIGBUS, current); + } + mce_clear_info(mi); } +/* + * Action optional processing happens here (picking up + * from the list of faulting pages that do_machine_check() + * placed into the "ring"). + */ static void mce_process_work(struct work_struct *dummy) { - mce_notify_process(); + unsigned long pfn; + + while (mce_ring_get(&pfn)) + memory_failure(pfn, MCE_VECTOR, 0); } #ifdef CONFIG_X86_MCE_INTEL @@ -1154,35 +1263,79 @@ void mce_log_therm_throt_event(__u64 status) * poller finds an MCE, poll 2x faster. When the poller finds no more * errors, poll 2x slower (up to check_interval seconds). */ -static int check_interval = 5 * 60; /* 5 minutes */ +static unsigned long check_interval = 5 * 60; /* 5 minutes */ -static DEFINE_PER_CPU(int, mce_next_interval); /* in jiffies */ +static DEFINE_PER_CPU(unsigned long, mce_next_interval); /* in jiffies */ static DEFINE_PER_CPU(struct timer_list, mce_timer); -static void mce_start_timer(unsigned long data) +static unsigned long mce_adjust_timer_default(unsigned long interval) +{ + return interval; +} + +static unsigned long (*mce_adjust_timer)(unsigned long interval) = + mce_adjust_timer_default; + +static int cmc_error_seen(void) +{ + unsigned long *v = &__get_cpu_var(mce_polled_error); + + return test_and_clear_bit(0, v); +} + +static void mce_timer_fn(unsigned long data) { - struct timer_list *t = &per_cpu(mce_timer, data); - int *n; + struct timer_list *t = &__get_cpu_var(mce_timer); + unsigned long iv; + int notify; WARN_ON(smp_processor_id() != data); if (mce_available(__this_cpu_ptr(&cpu_info))) { machine_check_poll(MCP_TIMESTAMP, &__get_cpu_var(mce_poll_banks)); + mce_intel_cmci_poll(); } /* * Alert userspace if needed. If we logged an MCE, reduce the * polling interval, otherwise increase the polling interval. */ - n = &__get_cpu_var(mce_next_interval); - if (mce_notify_irq()) - *n = max(*n/2, HZ/100); - else - *n = min(*n*2, (int)round_jiffies_relative(check_interval*HZ)); + iv = __this_cpu_read(mce_next_interval); + notify = mce_notify_irq(); + notify |= cmc_error_seen(); + if (notify) { + iv = max(iv / 2, (unsigned long) HZ/100); + } else { + iv = min(iv * 2, round_jiffies_relative(check_interval * HZ)); + iv = mce_adjust_timer(iv); + } + __this_cpu_write(mce_next_interval, iv); + /* Might have become 0 after CMCI storm subsided */ + if (iv) { + t->expires = jiffies + iv; + add_timer_on(t, smp_processor_id()); + } +} - t->expires = jiffies + *n; - add_timer_on(t, smp_processor_id()); +/* + * Ensure that the timer is firing in @interval from now. + */ +void mce_timer_kick(unsigned long interval) +{ + struct timer_list *t = &__get_cpu_var(mce_timer); + unsigned long when = jiffies + interval; + unsigned long iv = __this_cpu_read(mce_next_interval); + + if (timer_pending(t)) { + if (time_before(when, t->expires)) + mod_timer_pinned(t, when); + } else { + t->expires = round_jiffies(when); + add_timer_on(t, smp_processor_id()); + } + if (interval < iv) + __this_cpu_write(mce_next_interval, interval); } /* Must not be called in IRQ context where del_timer_sync() can deadlock */ @@ -1211,18 +1364,11 @@ int mce_notify_irq(void) /* Not more than two messages every minute */ static DEFINE_RATELIMIT_STATE(ratelimit, 60*HZ, 2); - clear_thread_flag(TIF_MCE_NOTIFY); - if (test_and_clear_bit(0, &mce_need_notify)) { /* wake processes polling /dev/mcelog */ wake_up_interruptible(&mce_chrdev_wait); - /* - * There is no risk of missing notifications because - * work_pending is always cleared before the function is - * executed. - */ - if (mce_helper[0] && !work_pending(&mce_trigger_work)) + if (mce_helper[0]) schedule_work(&mce_trigger_work); if (__ratelimit(&ratelimit)) @@ -1234,14 +1380,16 @@ int mce_notify_irq(void) } EXPORT_SYMBOL_GPL(mce_notify_irq); -static int __cpuinit __mcheck_cpu_mce_banks_init(void) +static int __mcheck_cpu_mce_banks_init(void) { int i; + u8 num_banks = mca_cfg.banks; - mce_banks = kzalloc(banks * sizeof(struct mce_bank), GFP_KERNEL); + mce_banks = kzalloc(num_banks * sizeof(struct mce_bank), GFP_KERNEL); if (!mce_banks) return -ENOMEM; - for (i = 0; i < banks; i++) { + + for (i = 0; i < num_banks; i++) { struct mce_bank *b = &mce_banks[i]; b->ctl = -1ULL; @@ -1253,7 +1401,7 @@ static int __cpuinit __mcheck_cpu_mce_banks_init(void) /* * Initialize Machine Checks for a CPU. */ -static int __cpuinit __mcheck_cpu_cap_init(void) +static int __mcheck_cpu_cap_init(void) { unsigned b; u64 cap; @@ -1261,19 +1409,19 @@ static int __cpuinit __mcheck_cpu_cap_init(void) rdmsrl(MSR_IA32_MCG_CAP, cap); b = cap & MCG_BANKCNT_MASK; - if (!banks) - printk(KERN_INFO "mce: CPU supports %d MCE banks\n", b); + if (!mca_cfg.banks) + pr_info("CPU supports %d MCE banks\n", b); if (b > MAX_NR_BANKS) { - printk(KERN_WARNING - "MCE: Using only %u machine check banks out of %u\n", + pr_warn("Using only %u machine check banks out of %u\n", MAX_NR_BANKS, b); b = MAX_NR_BANKS; } /* Don't support asymmetric configurations today */ - WARN_ON(banks != 0 && b != banks); - banks = b; + WARN_ON(mca_cfg.banks != 0 && b != mca_cfg.banks); + mca_cfg.banks = b; + if (!mce_banks) { int err = __mcheck_cpu_mce_banks_init(); @@ -1283,25 +1431,29 @@ static int __cpuinit __mcheck_cpu_cap_init(void) /* Use accurate RIP reporting if available. */ if ((cap & MCG_EXT_P) && MCG_EXT_CNT(cap) >= 9) - rip_msr = MSR_IA32_MCG_EIP; + mca_cfg.rip_msr = MSR_IA32_MCG_EIP; if (cap & MCG_SER_P) - mce_ser = 1; + mca_cfg.ser = true; return 0; } static void __mcheck_cpu_init_generic(void) { + enum mcp_flags m_fl = 0; mce_banks_t all_banks; u64 cap; int i; + if (!mca_cfg.bootlog) + m_fl = MCP_DONTLOG; + /* * Log the machine checks left over from the previous reset. */ bitmap_fill(all_banks, MAX_NR_BANKS); - machine_check_poll(MCP_UC|(!mce_bootlog ? MCP_DONTLOG : 0), &all_banks); + machine_check_poll(MCP_UC | m_fl, &all_banks); set_in_cr4(X86_CR4_MCE); @@ -1309,7 +1461,7 @@ static void __mcheck_cpu_init_generic(void) if (cap & MCG_CTL_P) wrmsr(MSR_IA32_MCG_CTL, 0xffffffff, 0xffffffff); - for (i = 0; i < banks; i++) { + for (i = 0; i < mca_cfg.banks; i++) { struct mce_bank *b = &mce_banks[i]; if (!b->init) @@ -1319,17 +1471,47 @@ static void __mcheck_cpu_init_generic(void) } } +/* + * During IFU recovery Sandy Bridge -EP4S processors set the RIPV and + * EIPV bits in MCG_STATUS to zero on the affected logical processor (SDM + * Vol 3B Table 15-20). But this confuses both the code that determines + * whether the machine check occurred in kernel or user mode, and also + * the severity assessment code. Pretend that EIPV was set, and take the + * ip/cs values from the pt_regs that mce_gather_info() ignored earlier. + */ +static void quirk_sandybridge_ifu(int bank, struct mce *m, struct pt_regs *regs) +{ + if (bank != 0) + return; + if ((m->mcgstatus & (MCG_STATUS_EIPV|MCG_STATUS_RIPV)) != 0) + return; + if ((m->status & (MCI_STATUS_OVER|MCI_STATUS_UC| + MCI_STATUS_EN|MCI_STATUS_MISCV|MCI_STATUS_ADDRV| + MCI_STATUS_PCC|MCI_STATUS_S|MCI_STATUS_AR| + MCACOD)) != + (MCI_STATUS_UC|MCI_STATUS_EN| + MCI_STATUS_MISCV|MCI_STATUS_ADDRV|MCI_STATUS_S| + MCI_STATUS_AR|MCACOD_INSTR)) + return; + + m->mcgstatus |= MCG_STATUS_EIPV; + m->ip = regs->ip; + m->cs = regs->cs; +} + /* Add per CPU specific workarounds here */ -static int __cpuinit __mcheck_cpu_apply_quirks(struct cpuinfo_x86 *c) +static int __mcheck_cpu_apply_quirks(struct cpuinfo_x86 *c) { + struct mca_config *cfg = &mca_cfg; + if (c->x86_vendor == X86_VENDOR_UNKNOWN) { - pr_info("MCE: unknown CPU type - not enabling MCE support.\n"); + pr_info("unknown CPU type - not enabling MCE support\n"); return -EOPNOTSUPP; } /* This should be disabled by the BIOS, but isn't always */ if (c->x86_vendor == X86_VENDOR_AMD) { - if (c->x86 == 15 && banks > 4) { + if (c->x86 == 15 && cfg->banks > 4) { /* * disable GART TBL walk error reporting, which * trips off incorrectly with the IOMMU & 3ware @@ -1337,19 +1519,56 @@ static int __cpuinit __mcheck_cpu_apply_quirks(struct cpuinfo_x86 *c) */ clear_bit(10, (unsigned long *)&mce_banks[4].ctl); } - if (c->x86 <= 17 && mce_bootlog < 0) { + if (c->x86 <= 17 && cfg->bootlog < 0) { /* * Lots of broken BIOS around that don't clear them * by default and leave crap in there. Don't log: */ - mce_bootlog = 0; + cfg->bootlog = 0; } /* * Various K7s with broken bank 0 around. Always disable * by default. */ - if (c->x86 == 6 && banks > 0) + if (c->x86 == 6 && cfg->banks > 0) mce_banks[0].ctl = 0; + + /* + * Turn off MC4_MISC thresholding banks on those models since + * they're not supported there. + */ + if (c->x86 == 0x15 && + (c->x86_model >= 0x10 && c->x86_model <= 0x1f)) { + int i; + u64 val, hwcr; + bool need_toggle; + u32 msrs[] = { + 0x00000413, /* MC4_MISC0 */ + 0xc0000408, /* MC4_MISC1 */ + }; + + rdmsrl(MSR_K7_HWCR, hwcr); + + /* McStatusWrEn has to be set */ + need_toggle = !(hwcr & BIT(18)); + + if (need_toggle) + wrmsrl(MSR_K7_HWCR, hwcr | BIT(18)); + + for (i = 0; i < ARRAY_SIZE(msrs); i++) { + rdmsrl(msrs[i], val); + + /* CntP bit set? */ + if (val & BIT_64(62)) { + val &= ~BIT_64(62); + wrmsrl(msrs[i], val); + } + } + + /* restore old settings */ + if (need_toggle) + wrmsrl(MSR_K7_HWCR, hwcr); + } } if (c->x86_vendor == X86_VENDOR_INTEL) { @@ -1362,7 +1581,7 @@ static int __cpuinit __mcheck_cpu_apply_quirks(struct cpuinfo_x86 *c) * valid event later, merely don't write CTL0. */ - if (c->x86 == 6 && c->x86_model < 0x1A && banks > 0) + if (c->x86 == 6 && c->x86_model < 0x1A && cfg->banks > 0) mce_banks[0].init = 0; /* @@ -1370,25 +1589,28 @@ static int __cpuinit __mcheck_cpu_apply_quirks(struct cpuinfo_x86 *c) * synchronization with a one second timeout. */ if ((c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xe)) && - monarch_timeout < 0) - monarch_timeout = USEC_PER_SEC; + cfg->monarch_timeout < 0) + cfg->monarch_timeout = USEC_PER_SEC; /* * There are also broken BIOSes on some Pentium M and * earlier systems: */ - if (c->x86 == 6 && c->x86_model <= 13 && mce_bootlog < 0) - mce_bootlog = 0; + if (c->x86 == 6 && c->x86_model <= 13 && cfg->bootlog < 0) + cfg->bootlog = 0; + + if (c->x86 == 6 && c->x86_model == 45) + quirk_no_way_out = quirk_sandybridge_ifu; } - if (monarch_timeout < 0) - monarch_timeout = 0; - if (mce_bootlog != 0) - mce_panic_timeout = 30; + if (cfg->monarch_timeout < 0) + cfg->monarch_timeout = 0; + if (cfg->bootlog != 0) + cfg->panic_timeout = 30; return 0; } -static int __cpuinit __mcheck_cpu_ancient_init(struct cpuinfo_x86 *c) +static int __mcheck_cpu_ancient_init(struct cpuinfo_x86 *c) { if (c->x86 != 5) return 0; @@ -1412,6 +1634,7 @@ static void __mcheck_cpu_init_vendor(struct cpuinfo_x86 *c) switch (c->x86_vendor) { case X86_VENDOR_INTEL: mce_intel_feature_init(c); + mce_adjust_timer = mce_intel_adjust_timer; break; case X86_VENDOR_AMD: mce_amd_feature_init(c); @@ -1421,27 +1644,32 @@ static void __mcheck_cpu_init_vendor(struct cpuinfo_x86 *c) } } -static void __mcheck_cpu_init_timer(void) +static void mce_start_timer(unsigned int cpu, struct timer_list *t) { - struct timer_list *t = &__get_cpu_var(mce_timer); - int *n = &__get_cpu_var(mce_next_interval); + unsigned long iv = check_interval * HZ; - setup_timer(t, mce_start_timer, smp_processor_id()); - - if (mce_ignore_ce) + if (mca_cfg.ignore_ce || !iv) return; - *n = check_interval * HZ; - if (!*n) - return; - t->expires = round_jiffies(jiffies + *n); - add_timer_on(t, smp_processor_id()); + per_cpu(mce_next_interval, cpu) = iv; + + t->expires = round_jiffies(jiffies + iv); + add_timer_on(t, cpu); +} + +static void __mcheck_cpu_init_timer(void) +{ + struct timer_list *t = &__get_cpu_var(mce_timer); + unsigned int cpu = smp_processor_id(); + + setup_timer(t, mce_timer_fn, cpu); + mce_start_timer(cpu, t); } /* Handle unconfigured int18 (should never happen) */ static void unexpected_machine_check(struct pt_regs *regs, long error_code) { - printk(KERN_ERR "CPU#%d: Unexpected int18 (Machine Check).\n", + pr_err("CPU#%d: Unexpected int18 (Machine Check)\n", smp_processor_id()); } @@ -1453,9 +1681,9 @@ void (*machine_check_vector)(struct pt_regs *, long error_code) = * Called for each booted CPU to set up machine checks. * Must be called with preempt off: */ -void __cpuinit mcheck_cpu_init(struct cpuinfo_x86 *c) +void mcheck_cpu_init(struct cpuinfo_x86 *c) { - if (mce_disabled) + if (mca_cfg.disabled) return; if (__mcheck_cpu_ancient_init(c)) @@ -1465,7 +1693,7 @@ void __cpuinit mcheck_cpu_init(struct cpuinfo_x86 *c) return; if (__mcheck_cpu_cap_init() < 0 || __mcheck_cpu_apply_quirks(c) < 0) { - mce_disabled = 1; + mca_cfg.disabled = true; return; } @@ -1541,6 +1769,12 @@ static int __mce_read_apei(char __user **ubuf, size_t usize) /* Error or no more MCE record */ if (rc <= 0) { mce_apei_read_done = 1; + /* + * When ERST is disabled, mce_chrdev_read() should return + * "no record" instead of "no device." + */ + if (rc == -ENODEV) + return 0; return rc; } rc = -EFAULT; @@ -1718,6 +1952,25 @@ static struct miscdevice mce_chrdev_device = { &mce_chrdev_ops, }; +static void __mce_disable_bank(void *arg) +{ + int bank = *((int *)arg); + __clear_bit(bank, __get_cpu_var(mce_poll_banks)); + cmci_disable_bank(bank); +} + +void mce_disable_bank(int bank) +{ + if (bank >= mca_cfg.banks) { + pr_warn(FW_BUG + "Ignoring request to disable invalid MCA bank %d.\n", + bank); + return; + } + set_bit(bank, mce_banks_ce_disabled); + on_each_cpu(__mce_disable_bank, &bank, 1); +} + /* * mce=off Disables machine check * mce=no_cmci Disables CMCI @@ -1728,9 +1981,12 @@ static struct miscdevice mce_chrdev_device = { * check, or 0 to not wait * mce=bootlog Log MCEs from before booting. Disabled by default on AMD. * mce=nobootlog Don't log MCEs from before booting. + * mce=bios_cmci_threshold Don't program the CMCI threshold */ static int __init mcheck_enable(char *str) { + struct mca_config *cfg = &mca_cfg; + if (*str == 0) { enable_p5_mce(); return 1; @@ -1738,24 +1994,25 @@ static int __init mcheck_enable(char *str) if (*str == '=') str++; if (!strcmp(str, "off")) - mce_disabled = 1; + cfg->disabled = true; else if (!strcmp(str, "no_cmci")) - mce_cmci_disabled = 1; + cfg->cmci_disabled = true; else if (!strcmp(str, "dont_log_ce")) - mce_dont_log_ce = 1; + cfg->dont_log_ce = true; else if (!strcmp(str, "ignore_ce")) - mce_ignore_ce = 1; + cfg->ignore_ce = true; else if (!strcmp(str, "bootlog") || !strcmp(str, "nobootlog")) - mce_bootlog = (str[0] == 'b'); + cfg->bootlog = (str[0] == 'b'); + else if (!strcmp(str, "bios_cmci_threshold")) + cfg->bios_cmci_threshold = true; else if (isdigit(str[0])) { - get_option(&str, &tolerant); + get_option(&str, &(cfg->tolerant)); if (*str == ',') { ++str; - get_option(&str, &monarch_timeout); + get_option(&str, &(cfg->monarch_timeout)); } } else { - printk(KERN_INFO "mce argument %s ignored. Please use /sys\n", - str); + pr_info("mce argument %s ignored. Please use /sys\n", str); return 0; } return 1; @@ -1781,7 +2038,7 @@ static int mce_disable_error_reporting(void) { int i; - for (i = 0; i < banks; i++) { + for (i = 0; i < mca_cfg.banks; i++) { struct mce_bank *b = &mce_banks[i]; if (b->init) @@ -1859,9 +2116,8 @@ static struct bus_type mce_subsys = { .dev_name = "machinecheck", }; -struct device *mce_device[CONFIG_NR_CPUS]; +DEFINE_PER_CPU(struct device *, mce_device); -__cpuinitdata void (*threshold_cpu_callback)(unsigned long action, unsigned int cpu); static inline struct mce_bank *attr_to_bank(struct device_attribute *attr) @@ -1921,15 +2177,15 @@ static ssize_t set_ignore_ce(struct device *s, if (strict_strtoull(buf, 0, &new) < 0) return -EINVAL; - if (mce_ignore_ce ^ !!new) { + if (mca_cfg.ignore_ce ^ !!new) { if (new) { /* disable ce features */ mce_timer_delete_all(); on_each_cpu(mce_disable_cmci, NULL, 1); - mce_ignore_ce = 1; + mca_cfg.ignore_ce = true; } else { /* enable ce features */ - mce_ignore_ce = 0; + mca_cfg.ignore_ce = false; on_each_cpu(mce_enable_ce, (void *)1, 1); } } @@ -1945,14 +2201,14 @@ static ssize_t set_cmci_disabled(struct device *s, if (strict_strtoull(buf, 0, &new) < 0) return -EINVAL; - if (mce_cmci_disabled ^ !!new) { + if (mca_cfg.cmci_disabled ^ !!new) { if (new) { /* disable cmci */ on_each_cpu(mce_disable_cmci, NULL, 1); - mce_cmci_disabled = 1; + mca_cfg.cmci_disabled = true; } else { /* enable cmci */ - mce_cmci_disabled = 0; + mca_cfg.cmci_disabled = false; on_each_cpu(mce_enable_ce, NULL, 1); } } @@ -1969,9 +2225,9 @@ static ssize_t store_int_with_restart(struct device *s, } static DEVICE_ATTR(trigger, 0644, show_trigger, set_trigger); -static DEVICE_INT_ATTR(tolerant, 0644, tolerant); -static DEVICE_INT_ATTR(monarch_timeout, 0644, monarch_timeout); -static DEVICE_INT_ATTR(dont_log_ce, 0644, mce_dont_log_ce); +static DEVICE_INT_ATTR(tolerant, 0644, mca_cfg.tolerant); +static DEVICE_INT_ATTR(monarch_timeout, 0644, mca_cfg.monarch_timeout); +static DEVICE_BOOL_ATTR(dont_log_ce, 0644, mca_cfg.dont_log_ce); static struct dev_ext_attribute dev_attr_check_interval = { __ATTR(check_interval, 0644, device_show_int, store_int_with_restart), @@ -1979,13 +2235,13 @@ static struct dev_ext_attribute dev_attr_check_interval = { }; static struct dev_ext_attribute dev_attr_ignore_ce = { - __ATTR(ignore_ce, 0644, device_show_int, set_ignore_ce), - &mce_ignore_ce + __ATTR(ignore_ce, 0644, device_show_bool, set_ignore_ce), + &mca_cfg.ignore_ce }; static struct dev_ext_attribute dev_attr_cmci_disabled = { - __ATTR(cmci_disabled, 0644, device_show_int, set_cmci_disabled), - &mce_cmci_disabled + __ATTR(cmci_disabled, 0644, device_show_bool, set_cmci_disabled), + &mca_cfg.cmci_disabled }; static struct device_attribute *mce_device_attrs[] = { @@ -2007,7 +2263,7 @@ static void mce_device_release(struct device *dev) } /* Per cpu device init. All of the cpus still share the same ctrl bank: */ -static __cpuinit int mce_device_create(unsigned int cpu) +static int mce_device_create(unsigned int cpu) { struct device *dev; int err; @@ -2024,21 +2280,23 @@ static __cpuinit int mce_device_create(unsigned int cpu) dev->release = &mce_device_release; err = device_register(dev); - if (err) + if (err) { + put_device(dev); return err; + } for (i = 0; mce_device_attrs[i]; i++) { err = device_create_file(dev, mce_device_attrs[i]); if (err) goto error; } - for (j = 0; j < banks; j++) { + for (j = 0; j < mca_cfg.banks; j++) { err = device_create_file(dev, &mce_banks[j].attr); if (err) goto error2; } cpumask_set_cpu(cpu, mce_device_initialized); - mce_device[cpu] = dev; + per_cpu(mce_device, cpu) = dev; return 0; error2: @@ -2053,9 +2311,9 @@ error: return err; } -static __cpuinit void mce_device_remove(unsigned int cpu) +static void mce_device_remove(unsigned int cpu) { - struct device *dev = mce_device[cpu]; + struct device *dev = per_cpu(mce_device, cpu); int i; if (!cpumask_test_cpu(cpu, mce_device_initialized)) @@ -2064,16 +2322,16 @@ static __cpuinit void mce_device_remove(unsigned int cpu) for (i = 0; mce_device_attrs[i]; i++) device_remove_file(dev, mce_device_attrs[i]); - for (i = 0; i < banks; i++) + for (i = 0; i < mca_cfg.banks; i++) device_remove_file(dev, &mce_banks[i].attr); device_unregister(dev); cpumask_clear_cpu(cpu, mce_device_initialized); - mce_device[cpu] = NULL; + per_cpu(mce_device, cpu) = NULL; } /* Make sure there are no machine checks on offlined CPUs. */ -static void __cpuinit mce_disable_cpu(void *h) +static void mce_disable_cpu(void *h) { unsigned long action = *(unsigned long *)h; int i; @@ -2083,7 +2341,7 @@ static void __cpuinit mce_disable_cpu(void *h) if (!(action & CPU_TASKS_FROZEN)) cmci_clear(); - for (i = 0; i < banks; i++) { + for (i = 0; i < mca_cfg.banks; i++) { struct mce_bank *b = &mce_banks[i]; if (b->init) @@ -2091,7 +2349,7 @@ static void __cpuinit mce_disable_cpu(void *h) } } -static void __cpuinit mce_reenable_cpu(void *h) +static void mce_reenable_cpu(void *h) { unsigned long action = *(unsigned long *)h; int i; @@ -2101,7 +2359,7 @@ static void __cpuinit mce_reenable_cpu(void *h) if (!(action & CPU_TASKS_FROZEN)) cmci_reenable(); - for (i = 0; i < banks; i++) { + for (i = 0; i < mca_cfg.banks; i++) { struct mce_bank *b = &mce_banks[i]; if (b->init) @@ -2110,48 +2368,43 @@ static void __cpuinit mce_reenable_cpu(void *h) } /* Get notified when a cpu comes on/off. Be hotplug friendly. */ -static int __cpuinit +static int mce_cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu) { unsigned int cpu = (unsigned long)hcpu; struct timer_list *t = &per_cpu(mce_timer, cpu); - switch (action) { + switch (action & ~CPU_TASKS_FROZEN) { case CPU_ONLINE: - case CPU_ONLINE_FROZEN: mce_device_create(cpu); if (threshold_cpu_callback) threshold_cpu_callback(action, cpu); break; case CPU_DEAD: - case CPU_DEAD_FROZEN: if (threshold_cpu_callback) threshold_cpu_callback(action, cpu); mce_device_remove(cpu); + mce_intel_hcpu_update(cpu); break; case CPU_DOWN_PREPARE: - case CPU_DOWN_PREPARE_FROZEN: - del_timer_sync(t); smp_call_function_single(cpu, mce_disable_cpu, &action, 1); + del_timer_sync(t); break; case CPU_DOWN_FAILED: - case CPU_DOWN_FAILED_FROZEN: - if (!mce_ignore_ce && check_interval) { - t->expires = round_jiffies(jiffies + - __get_cpu_var(mce_next_interval)); - add_timer_on(t, cpu); - } smp_call_function_single(cpu, mce_reenable_cpu, &action, 1); + mce_start_timer(cpu, t); break; - case CPU_POST_DEAD: + } + + if (action == CPU_POST_DEAD) { /* intentionally ignoring frozen here */ - cmci_rediscover(cpu); - break; + cmci_rediscover(); } + return NOTIFY_OK; } -static struct notifier_block mce_cpu_notifier __cpuinitdata = { +static struct notifier_block mce_cpu_notifier = { .notifier_call = mce_cpu_callback, }; @@ -2159,7 +2412,7 @@ static __init void mce_init_banks(void) { int i; - for (i = 0; i < banks; i++) { + for (i = 0; i < mca_cfg.banks; i++) { struct mce_bank *b = &mce_banks[i]; struct device_attribute *a = &b->attr; @@ -2178,39 +2431,78 @@ static __init int mcheck_init_device(void) int err; int i = 0; - if (!mce_available(&boot_cpu_data)) - return -EIO; + if (!mce_available(&boot_cpu_data)) { + err = -EIO; + goto err_out; + } - zalloc_cpumask_var(&mce_device_initialized, GFP_KERNEL); + if (!zalloc_cpumask_var(&mce_device_initialized, GFP_KERNEL)) { + err = -ENOMEM; + goto err_out; + } mce_init_banks(); err = subsys_system_register(&mce_subsys, NULL); if (err) - return err; + goto err_out_mem; + cpu_notifier_register_begin(); for_each_online_cpu(i) { err = mce_device_create(i); - if (err) - return err; + if (err) { + /* + * Register notifier anyway (and do not unreg it) so + * that we don't leave undeleted timers, see notifier + * callback above. + */ + __register_hotcpu_notifier(&mce_cpu_notifier); + cpu_notifier_register_done(); + goto err_device_create; + } } + __register_hotcpu_notifier(&mce_cpu_notifier); + cpu_notifier_register_done(); + register_syscore_ops(&mce_syscore_ops); - register_hotcpu_notifier(&mce_cpu_notifier); /* register character device /dev/mcelog */ - misc_register(&mce_chrdev_device); + err = misc_register(&mce_chrdev_device); + if (err) + goto err_register; + + return 0; + +err_register: + unregister_syscore_ops(&mce_syscore_ops); + +err_device_create: + /* + * We didn't keep track of which devices were created above, but + * even if we had, the set of online cpus might have changed. + * Play safe and remove for every possible cpu, since + * mce_device_remove() will do the right thing. + */ + for_each_possible_cpu(i) + mce_device_remove(i); + +err_out_mem: + free_cpumask_var(mce_device_initialized); + +err_out: + pr_err("Unable to init device /dev/mcelog (rc: %d)\n", err); return err; } -device_initcall(mcheck_init_device); +device_initcall_sync(mcheck_init_device); /* * Old style boot options parsing. Only for compatibility. */ static int __init mcheck_disable(char *str) { - mce_disabled = 1; + mca_cfg.disabled = true; return 1; } __setup("nomce", mcheck_disable); diff --git a/arch/x86/kernel/cpu/mcheck/mce_amd.c b/arch/x86/kernel/cpu/mcheck/mce_amd.c index 786e76a8632..603df4f7464 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_amd.c +++ b/arch/x86/kernel/cpu/mcheck/mce_amd.c @@ -1,15 +1,17 @@ /* - * (c) 2005, 2006 Advanced Micro Devices, Inc. + * (c) 2005-2012 Advanced Micro Devices, Inc. * Your use of this code is subject to the terms and conditions of the * GNU general public license version 2. See "COPYING" or * http://www.gnu.org/licenses/gpl.html * * Written by Jacob Shin - AMD, Inc. * - * Support : jacob.shin@amd.com + * Maintained by: Borislav Petkov <bp@alien8.de> * * April 2006 * - added support for AMD Family 0x10 processors + * May 2012 + * - major scrubbing * * All MC4_MISCi registers are shared between multi-cores */ @@ -25,12 +27,12 @@ #include <linux/cpu.h> #include <linux/smp.h> +#include <asm/amd_nb.h> #include <asm/apic.h> #include <asm/idle.h> #include <asm/mce.h> #include <asm/msr.h> -#define NR_BANKS 6 #define NR_BLOCKS 9 #define THRESHOLD_MAX 0xFFF #define INT_TYPE_APIC 0x00020000 @@ -45,28 +47,16 @@ #define MASK_BLKPTR_LO 0xFF000000 #define MCG_XBLK_ADDR 0xC0000400 -struct threshold_block { - unsigned int block; - unsigned int bank; - unsigned int cpu; - u32 address; - u16 interrupt_enable; - u16 threshold_limit; - struct kobject kobj; - struct list_head miscj; -}; - -struct threshold_bank { - struct kobject *kobj; - struct threshold_block *blocks; - cpumask_var_t cpus; -}; -static DEFINE_PER_CPU(struct threshold_bank * [NR_BANKS], threshold_banks); - -static unsigned char shared_bank[NR_BANKS] = { - 0, 0, 0, 0, 1 +static const char * const th_names[] = { + "load_store", + "insn_fetch", + "combined_unit", + "", + "northbridge", + "execution_unit", }; +static DEFINE_PER_CPU(struct threshold_bank **, threshold_banks); static DEFINE_PER_CPU(unsigned char, bank_map); /* see which banks are on */ static void amd_threshold_interrupt(void); @@ -83,6 +73,47 @@ struct thresh_restart { u16 old_limit; }; +static inline bool is_shared_bank(int bank) +{ + /* Bank 4 is for northbridge reporting and is thus shared */ + return (bank == 4); +} + +static const char * const bank4_names(struct threshold_block *b) +{ + switch (b->address) { + /* MSR4_MISC0 */ + case 0x00000413: + return "dram"; + + case 0xc0000408: + return "ht_links"; + + case 0xc0000409: + return "l3_cache"; + + default: + WARN(1, "Funny MSR: 0x%08x\n", b->address); + return ""; + } +}; + + +static bool lvt_interrupt_supported(unsigned int bank, u32 msr_high_bits) +{ + /* + * bank 4 supports APIC LVT interrupts implicitly since forever. + */ + if (bank == 4) + return true; + + /* + * IntP: interrupt present; if this bit is set, the thresholding + * bank can generate APIC LVT interrupts + */ + return msr_high_bits & BIT(28); +} + static int lvt_off_valid(struct threshold_block *b, int apic, u32 lo, u32 hi) { int msr = (hi & MASK_LVTOFF_HI) >> 20; @@ -104,8 +135,10 @@ static int lvt_off_valid(struct threshold_block *b, int apic, u32 lo, u32 hi) return 1; }; -/* must be called with correct cpu affinity */ -/* Called via smp_call_function_single() */ +/* + * Called via smp_call_function_single(), must be called with correct + * cpu affinity. + */ static void threshold_restart_bank(void *_tr) { struct thresh_restart *tr = _tr; @@ -128,6 +161,12 @@ static void threshold_restart_bank(void *_tr) (new_count & THRESHOLD_MAX); } + /* clear IntType */ + hi &= ~MASK_INT_TYPE_HI; + + if (!tr->b->interrupt_capable) + goto done; + if (tr->set_lvt_off) { if (lvt_off_valid(tr->b, tr->lvt_off, lo, hi)) { /* set new lvt offset */ @@ -136,9 +175,10 @@ static void threshold_restart_bank(void *_tr) } } - tr->b->interrupt_enable ? - (hi = (hi & ~MASK_INT_TYPE_HI) | INT_TYPE_APIC) : - (hi &= ~MASK_INT_TYPE_HI); + if (tr->b->interrupt_enable) + hi |= INT_TYPE_APIC; + + done: hi |= MASK_COUNT_EN_HI; wrmsr(tr->b->address, lo, hi); @@ -174,7 +214,7 @@ void mce_amd_feature_init(struct cpuinfo_x86 *c) unsigned int bank, block; int offset = -1; - for (bank = 0; bank < NR_BANKS; ++bank) { + for (bank = 0; bank < mca_cfg.banks; ++bank) { for (block = 0; block < NR_BLOCKS; ++block) { if (block == 0) address = MSR_IA32_MC0_MISC + bank * 4; @@ -199,17 +239,18 @@ void mce_amd_feature_init(struct cpuinfo_x86 *c) if (!block) per_cpu(bank_map, cpu) |= (1 << bank); - if (shared_bank[bank] && c->cpu_core_id) - break; - - offset = setup_APIC_mce(offset, - (high & MASK_LVTOFF_HI) >> 20); memset(&b, 0, sizeof(b)); - b.cpu = cpu; - b.bank = bank; - b.block = block; - b.address = address; + b.cpu = cpu; + b.bank = bank; + b.block = block; + b.address = address; + b.interrupt_capable = lvt_interrupt_supported(bank, high); + + if (b.interrupt_capable) { + int new = (high & MASK_LVTOFF_HI) >> 20; + offset = setup_APIC_mce(offset, new); + } mce_threshold_block_init(&b, offset); mce_threshold_vector = amd_threshold_interrupt; @@ -235,7 +276,7 @@ static void amd_threshold_interrupt(void) mce_setup(&m); /* assume first bank caused it */ - for (bank = 0; bank < NR_BANKS; ++bank) { + for (bank = 0; bank < mca_cfg.banks; ++bank) { if (!(per_cpu(bank_map, m.cpu) & (1 << bank))) continue; for (block = 0; block < NR_BLOCKS; ++block) { @@ -298,7 +339,7 @@ struct threshold_attr { #define SHOW_FIELDS(name) \ static ssize_t show_ ## name(struct threshold_block *b, char *buf) \ { \ - return sprintf(buf, "%lx\n", (unsigned long) b->name); \ + return sprintf(buf, "%lu\n", (unsigned long) b->name); \ } SHOW_FIELDS(interrupt_enable) SHOW_FIELDS(threshold_limit) @@ -309,6 +350,9 @@ store_interrupt_enable(struct threshold_block *b, const char *buf, size_t size) struct thresh_restart tr; unsigned long new; + if (!b->interrupt_capable) + return -EINVAL; + if (strict_strtoul(buf, 0, &new) < 0) return -EINVAL; @@ -346,38 +390,21 @@ store_threshold_limit(struct threshold_block *b, const char *buf, size_t size) return size; } -struct threshold_block_cross_cpu { - struct threshold_block *tb; - long retval; -}; - -static void local_error_count_handler(void *_tbcc) -{ - struct threshold_block_cross_cpu *tbcc = _tbcc; - struct threshold_block *b = tbcc->tb; - u32 low, high; - - rdmsr(b->address, low, high); - tbcc->retval = (high & 0xFFF) - (THRESHOLD_MAX - b->threshold_limit); -} - static ssize_t show_error_count(struct threshold_block *b, char *buf) { - struct threshold_block_cross_cpu tbcc = { .tb = b, }; + u32 lo, hi; - smp_call_function_single(b->cpu, local_error_count_handler, &tbcc, 1); - return sprintf(buf, "%lx\n", tbcc.retval); -} + rdmsr_on_cpu(b->cpu, b->address, &lo, &hi); -static ssize_t store_error_count(struct threshold_block *b, - const char *buf, size_t count) -{ - struct thresh_restart tr = { .b = b, .reset = 1, .old_limit = 0 }; - - smp_call_function_single(b->cpu, threshold_restart_bank, &tr, 1); - return 1; + return sprintf(buf, "%u\n", ((hi & THRESHOLD_MAX) - + (THRESHOLD_MAX - b->threshold_limit))); } +static struct threshold_attr error_count = { + .attr = {.name = __stringify(error_count), .mode = 0444 }, + .show = show_error_count, +}; + #define RW_ATTR(val) \ static struct threshold_attr val = { \ .attr = {.name = __stringify(val), .mode = 0644 }, \ @@ -387,13 +414,12 @@ static struct threshold_attr val = { \ RW_ATTR(interrupt_enable); RW_ATTR(threshold_limit); -RW_ATTR(error_count); static struct attribute *default_attrs[] = { - &interrupt_enable.attr, &threshold_limit.attr, &error_count.attr, - NULL + NULL, /* possibly interrupt_enable if supported, see below */ + NULL, }; #define to_block(k) container_of(k, struct threshold_block, kobj) @@ -432,16 +458,14 @@ static struct kobj_type threshold_ktype = { .default_attrs = default_attrs, }; -static __cpuinit int allocate_threshold_blocks(unsigned int cpu, - unsigned int bank, - unsigned int block, - u32 address) +static int allocate_threshold_blocks(unsigned int cpu, unsigned int bank, + unsigned int block, u32 address) { struct threshold_block *b = NULL; u32 low, high; int err; - if ((bank >= NR_BANKS) || (block >= NR_BLOCKS)) + if ((bank >= mca_cfg.banks) || (block >= NR_BLOCKS)) return 0; if (rdmsr_safe_on_cpu(cpu, address, &low, &high)) @@ -467,8 +491,14 @@ static __cpuinit int allocate_threshold_blocks(unsigned int cpu, b->cpu = cpu; b->address = address; b->interrupt_enable = 0; + b->interrupt_capable = lvt_interrupt_supported(bank, high); b->threshold_limit = THRESHOLD_MAX; + if (b->interrupt_capable) + threshold_ktype.default_attrs[2] = &interrupt_enable.attr; + else + threshold_ktype.default_attrs[2] = NULL; + INIT_LIST_HEAD(&b->miscj); if (per_cpu(threshold_banks, cpu)[bank]->blocks) { @@ -480,7 +510,7 @@ static __cpuinit int allocate_threshold_blocks(unsigned int cpu, err = kobject_init_and_add(&b->kobj, &threshold_ktype, per_cpu(threshold_banks, cpu)[bank]->kobj, - "misc%i", block); + (bank == 4 ? bank4_names(b) : th_names[bank])); if (err) goto out_free; recurse: @@ -511,47 +541,56 @@ out_free: return err; } -static __cpuinit long -local_allocate_threshold_blocks(int cpu, unsigned int bank) +static int __threshold_add_blocks(struct threshold_bank *b) { - return allocate_threshold_blocks(cpu, bank, 0, - MSR_IA32_MC0_MISC + bank * 4); -} + struct list_head *head = &b->blocks->miscj; + struct threshold_block *pos = NULL; + struct threshold_block *tmp = NULL; + int err = 0; -/* symlinks sibling shared banks to first core. first core owns dir/files. */ -static __cpuinit int threshold_create_bank(unsigned int cpu, unsigned int bank) -{ - int i, err = 0; - struct threshold_bank *b = NULL; - struct device *dev = mce_device[cpu]; - char name[32]; + err = kobject_add(&b->blocks->kobj, b->kobj, b->blocks->kobj.name); + if (err) + return err; - sprintf(name, "threshold_bank%i", bank); + list_for_each_entry_safe(pos, tmp, head, miscj) { - if (cpu_data(cpu).cpu_core_id && shared_bank[bank]) { /* symlink */ - i = cpumask_first(cpu_llc_shared_mask(cpu)); + err = kobject_add(&pos->kobj, b->kobj, pos->kobj.name); + if (err) { + list_for_each_entry_safe_reverse(pos, tmp, head, miscj) + kobject_del(&pos->kobj); - /* first core not up yet */ - if (cpu_data(i).cpu_core_id) - goto out; + return err; + } + } + return err; +} - /* already linked */ - if (per_cpu(threshold_banks, cpu)[bank]) - goto out; +static int threshold_create_bank(unsigned int cpu, unsigned int bank) +{ + struct device *dev = per_cpu(mce_device, cpu); + struct amd_northbridge *nb = NULL; + struct threshold_bank *b = NULL; + const char *name = th_names[bank]; + int err = 0; - b = per_cpu(threshold_banks, i)[bank]; + if (is_shared_bank(bank)) { + nb = node_to_amd_nb(amd_get_nb_id(cpu)); - if (!b) - goto out; + /* threshold descriptor already initialized on this node? */ + if (nb && nb->bank4) { + /* yes, use it */ + b = nb->bank4; + err = kobject_add(b->kobj, &dev->kobj, name); + if (err) + goto out; - err = sysfs_create_link(&dev->kobj, b->kobj, name); - if (err) - goto out; + per_cpu(threshold_banks, cpu)[bank] = b; + atomic_inc(&b->cpus); - cpumask_copy(b->cpus, cpu_llc_shared_mask(cpu)); - per_cpu(threshold_banks, cpu)[bank] = b; + err = __threshold_add_blocks(b); - goto out; + goto out; + } } b = kzalloc(sizeof(struct threshold_bank), GFP_KERNEL); @@ -559,58 +598,52 @@ static __cpuinit int threshold_create_bank(unsigned int cpu, unsigned int bank) err = -ENOMEM; goto out; } - if (!zalloc_cpumask_var(&b->cpus, GFP_KERNEL)) { - kfree(b); - err = -ENOMEM; - goto out; - } b->kobj = kobject_create_and_add(name, &dev->kobj); - if (!b->kobj) + if (!b->kobj) { + err = -EINVAL; goto out_free; - -#ifndef CONFIG_SMP - cpumask_setall(b->cpus); -#else - cpumask_set_cpu(cpu, b->cpus); -#endif + } per_cpu(threshold_banks, cpu)[bank] = b; - err = local_allocate_threshold_blocks(cpu, bank); - if (err) - goto out_free; - - for_each_cpu(i, b->cpus) { - if (i == cpu) - continue; - - dev = mce_device[i]; - if (dev) - err = sysfs_create_link(&dev->kobj,b->kobj, name); - if (err) - goto out; + if (is_shared_bank(bank)) { + atomic_set(&b->cpus, 1); - per_cpu(threshold_banks, i)[bank] = b; + /* nb is already initialized, see above */ + if (nb) { + WARN_ON(nb->bank4); + nb->bank4 = b; + } } - goto out; + err = allocate_threshold_blocks(cpu, bank, 0, + MSR_IA32_MC0_MISC + bank * 4); + if (!err) + goto out; -out_free: - per_cpu(threshold_banks, cpu)[bank] = NULL; - free_cpumask_var(b->cpus); + out_free: kfree(b); -out: + + out: return err; } /* create dir/files for all valid threshold banks */ -static __cpuinit int threshold_create_device(unsigned int cpu) +static int threshold_create_device(unsigned int cpu) { unsigned int bank; + struct threshold_bank **bp; int err = 0; - for (bank = 0; bank < NR_BANKS; ++bank) { + bp = kzalloc(sizeof(struct threshold_bank *) * mca_cfg.banks, + GFP_KERNEL); + if (!bp) + return -ENOMEM; + + per_cpu(threshold_banks, cpu) = bp; + + for (bank = 0; bank < mca_cfg.banks; ++bank) { if (!(per_cpu(bank_map, cpu) & (1 << bank))) continue; err = threshold_create_bank(cpu, bank); @@ -621,12 +654,6 @@ static __cpuinit int threshold_create_device(unsigned int cpu) return err; } -/* - * let's be hotplug friendly. - * in case of multiple core processors, the first core always takes ownership - * of shared sysfs dir/files, and rest of the cores will be symlinked to it. - */ - static void deallocate_threshold_block(unsigned int cpu, unsigned int bank) { @@ -647,40 +674,42 @@ static void deallocate_threshold_block(unsigned int cpu, per_cpu(threshold_banks, cpu)[bank]->blocks = NULL; } +static void __threshold_remove_blocks(struct threshold_bank *b) +{ + struct threshold_block *pos = NULL; + struct threshold_block *tmp = NULL; + + kobject_del(b->kobj); + + list_for_each_entry_safe(pos, tmp, &b->blocks->miscj, miscj) + kobject_del(&pos->kobj); +} + static void threshold_remove_bank(unsigned int cpu, int bank) { + struct amd_northbridge *nb; struct threshold_bank *b; - struct device *dev; - char name[32]; - int i = 0; b = per_cpu(threshold_banks, cpu)[bank]; if (!b) return; + if (!b->blocks) goto free_out; - sprintf(name, "threshold_bank%i", bank); - -#ifdef CONFIG_SMP - /* sibling symlink */ - if (shared_bank[bank] && b->blocks->cpu != cpu) { - sysfs_remove_link(&mce_device[cpu]->kobj, name); - per_cpu(threshold_banks, cpu)[bank] = NULL; - - return; - } -#endif - - /* remove all sibling symlinks before unregistering */ - for_each_cpu(i, b->cpus) { - if (i == cpu) - continue; - - dev = mce_device[i]; - if (dev) - sysfs_remove_link(&dev->kobj, name); - per_cpu(threshold_banks, i)[bank] = NULL; + if (is_shared_bank(bank)) { + if (!atomic_dec_and_test(&b->cpus)) { + __threshold_remove_blocks(b); + per_cpu(threshold_banks, cpu)[bank] = NULL; + return; + } else { + /* + * the last CPU on this node using the shared bank is + * going away, remove that bank now. + */ + nb = node_to_amd_nb(amd_get_nb_id(cpu)); + nb->bank4 = NULL; + } } deallocate_threshold_block(cpu, bank); @@ -688,7 +717,6 @@ static void threshold_remove_bank(unsigned int cpu, int bank) free_out: kobject_del(b->kobj); kobject_put(b->kobj); - free_cpumask_var(b->cpus); kfree(b); per_cpu(threshold_banks, cpu)[bank] = NULL; } @@ -697,15 +725,16 @@ static void threshold_remove_device(unsigned int cpu) { unsigned int bank; - for (bank = 0; bank < NR_BANKS; ++bank) { + for (bank = 0; bank < mca_cfg.banks; ++bank) { if (!(per_cpu(bank_map, cpu) & (1 << bank))) continue; threshold_remove_bank(cpu, bank); } + kfree(per_cpu(threshold_banks, cpu)); } /* get notified when a cpu comes on/off */ -static void __cpuinit +static void amd_64_threshold_cpu_callback(unsigned long action, unsigned int cpu) { switch (action) { @@ -737,4 +766,24 @@ static __init int threshold_init_device(void) return 0; } -device_initcall(threshold_init_device); +/* + * there are 3 funcs which need to be _initcalled in a logic sequence: + * 1. xen_late_init_mcelog + * 2. mcheck_init_device + * 3. threshold_init_device + * + * xen_late_init_mcelog must register xen_mce_chrdev_device before + * native mce_chrdev_device registration if running under xen platform; + * + * mcheck_init_device should be inited before threshold_init_device to + * initialize mce_device, otherwise a NULL ptr dereference will cause panic. + * + * so we use following _initcalls + * 1. device_initcall(xen_late_init_mcelog); + * 2. device_initcall_sync(mcheck_init_device); + * 3. late_initcall(threshold_init_device); + * + * when running under xen, the initcall order is 1,2,3; + * on baremetal, we skip 1 and we do only 2 and 3. + */ +late_initcall(threshold_init_device); diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel.c b/arch/x86/kernel/cpu/mcheck/mce_intel.c index 38e49bc95ff..9a316b21df8 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_intel.c +++ b/arch/x86/kernel/cpu/mcheck/mce_intel.c @@ -6,15 +6,17 @@ */ #include <linux/gfp.h> -#include <linux/init.h> #include <linux/interrupt.h> #include <linux/percpu.h> #include <linux/sched.h> +#include <linux/cpumask.h> #include <asm/apic.h> #include <asm/processor.h> #include <asm/msr.h> #include <asm/mce.h> +#include "mce-internal.h" + /* * Support for Intel Correct Machine Check Interrupts. This allows * the CPU to raise an interrupt when a corrected machine check happened. @@ -22,21 +24,48 @@ * Also supports reliable discovery of shared banks. */ +/* + * CMCI can be delivered to multiple cpus that share a machine check bank + * so we need to designate a single cpu to process errors logged in each bank + * in the interrupt handler (otherwise we would have many races and potential + * double reporting of the same error). + * Note that this can change when a cpu is offlined or brought online since + * some MCA banks are shared across cpus. When a cpu is offlined, cmci_clear() + * disables CMCI on all banks owned by the cpu and clears this bitfield. At + * this point, cmci_rediscover() kicks in and a different cpu may end up + * taking ownership of some of the shared MCA banks that were previously + * owned by the offlined cpu. + */ static DEFINE_PER_CPU(mce_banks_t, mce_banks_owned); /* * cmci_discover_lock protects against parallel discovery attempts * which could race against each other. */ -static DEFINE_RAW_SPINLOCK(cmci_discover_lock); +static DEFINE_SPINLOCK(cmci_discover_lock); + +#define CMCI_THRESHOLD 1 +#define CMCI_POLL_INTERVAL (30 * HZ) +#define CMCI_STORM_INTERVAL (1 * HZ) +#define CMCI_STORM_THRESHOLD 15 + +static DEFINE_PER_CPU(unsigned long, cmci_time_stamp); +static DEFINE_PER_CPU(unsigned int, cmci_storm_cnt); +static DEFINE_PER_CPU(unsigned int, cmci_storm_state); + +enum { + CMCI_STORM_NONE, + CMCI_STORM_ACTIVE, + CMCI_STORM_SUBSIDED, +}; -#define CMCI_THRESHOLD 1 +static atomic_t cmci_storm_on_cpus; static int cmci_supported(int *banks) { u64 cap; - if (mce_cmci_disabled || mce_ignore_ce) + if (mca_cfg.cmci_disabled || mca_cfg.ignore_ce) return 0; /* @@ -53,6 +82,109 @@ static int cmci_supported(int *banks) return !!(cap & MCG_CMCI_P); } +void mce_intel_cmci_poll(void) +{ + if (__this_cpu_read(cmci_storm_state) == CMCI_STORM_NONE) + return; + machine_check_poll(MCP_TIMESTAMP, &__get_cpu_var(mce_banks_owned)); +} + +void mce_intel_hcpu_update(unsigned long cpu) +{ + if (per_cpu(cmci_storm_state, cpu) == CMCI_STORM_ACTIVE) + atomic_dec(&cmci_storm_on_cpus); + + per_cpu(cmci_storm_state, cpu) = CMCI_STORM_NONE; +} + +unsigned long mce_intel_adjust_timer(unsigned long interval) +{ + int r; + + if (interval < CMCI_POLL_INTERVAL) + return interval; + + switch (__this_cpu_read(cmci_storm_state)) { + case CMCI_STORM_ACTIVE: + /* + * We switch back to interrupt mode once the poll timer has + * silenced itself. That means no events recorded and the + * timer interval is back to our poll interval. + */ + __this_cpu_write(cmci_storm_state, CMCI_STORM_SUBSIDED); + r = atomic_sub_return(1, &cmci_storm_on_cpus); + if (r == 0) + pr_notice("CMCI storm subsided: switching to interrupt mode\n"); + /* FALLTHROUGH */ + + case CMCI_STORM_SUBSIDED: + /* + * We wait for all cpus to go back to SUBSIDED + * state. When that happens we switch back to + * interrupt mode. + */ + if (!atomic_read(&cmci_storm_on_cpus)) { + __this_cpu_write(cmci_storm_state, CMCI_STORM_NONE); + cmci_reenable(); + cmci_recheck(); + } + return CMCI_POLL_INTERVAL; + default: + /* + * We have shiny weather. Let the poll do whatever it + * thinks. + */ + return interval; + } +} + +static void cmci_storm_disable_banks(void) +{ + unsigned long flags, *owned; + int bank; + u64 val; + + spin_lock_irqsave(&cmci_discover_lock, flags); + owned = __get_cpu_var(mce_banks_owned); + for_each_set_bit(bank, owned, MAX_NR_BANKS) { + rdmsrl(MSR_IA32_MCx_CTL2(bank), val); + val &= ~MCI_CTL2_CMCI_EN; + wrmsrl(MSR_IA32_MCx_CTL2(bank), val); + } + spin_unlock_irqrestore(&cmci_discover_lock, flags); +} + +static bool cmci_storm_detect(void) +{ + unsigned int cnt = __this_cpu_read(cmci_storm_cnt); + unsigned long ts = __this_cpu_read(cmci_time_stamp); + unsigned long now = jiffies; + int r; + + if (__this_cpu_read(cmci_storm_state) != CMCI_STORM_NONE) + return true; + + if (time_before_eq(now, ts + CMCI_STORM_INTERVAL)) { + cnt++; + } else { + cnt = 1; + __this_cpu_write(cmci_time_stamp, now); + } + __this_cpu_write(cmci_storm_cnt, cnt); + + if (cnt <= CMCI_STORM_THRESHOLD) + return false; + + cmci_storm_disable_banks(); + __this_cpu_write(cmci_storm_state, CMCI_STORM_ACTIVE); + r = atomic_add_return(1, &cmci_storm_on_cpus); + mce_timer_kick(CMCI_POLL_INTERVAL); + + if (r == 1) + pr_notice("CMCI storm detected: switching to poll mode\n"); + return true; +} + /* * The interrupt handler. This is called on every event. * Just call the poller directly to log any events. @@ -61,64 +193,86 @@ static int cmci_supported(int *banks) */ static void intel_threshold_interrupt(void) { + if (cmci_storm_detect()) + return; machine_check_poll(MCP_TIMESTAMP, &__get_cpu_var(mce_banks_owned)); mce_notify_irq(); } -static void print_update(char *type, int *hdr, int num) -{ - if (*hdr == 0) - printk(KERN_INFO "CPU %d MCA banks", smp_processor_id()); - *hdr = 1; - printk(KERN_CONT " %s:%d", type, num); -} - /* * Enable CMCI (Corrected Machine Check Interrupt) for available MCE banks * on this CPU. Use the algorithm recommended in the SDM to discover shared * banks. */ -static void cmci_discover(int banks, int boot) +static void cmci_discover(int banks) { unsigned long *owned = (void *)&__get_cpu_var(mce_banks_owned); unsigned long flags; - int hdr = 0; int i; + int bios_wrong_thresh = 0; - raw_spin_lock_irqsave(&cmci_discover_lock, flags); + spin_lock_irqsave(&cmci_discover_lock, flags); for (i = 0; i < banks; i++) { u64 val; + int bios_zero_thresh = 0; if (test_bit(i, owned)) continue; + /* Skip banks in firmware first mode */ + if (test_bit(i, mce_banks_ce_disabled)) + continue; + rdmsrl(MSR_IA32_MCx_CTL2(i), val); /* Already owned by someone else? */ if (val & MCI_CTL2_CMCI_EN) { - if (test_and_clear_bit(i, owned) && !boot) - print_update("SHD", &hdr, i); + clear_bit(i, owned); __clear_bit(i, __get_cpu_var(mce_poll_banks)); continue; } - val &= ~MCI_CTL2_CMCI_THRESHOLD_MASK; - val |= MCI_CTL2_CMCI_EN | CMCI_THRESHOLD; + if (!mca_cfg.bios_cmci_threshold) { + val &= ~MCI_CTL2_CMCI_THRESHOLD_MASK; + val |= CMCI_THRESHOLD; + } else if (!(val & MCI_CTL2_CMCI_THRESHOLD_MASK)) { + /* + * If bios_cmci_threshold boot option was specified + * but the threshold is zero, we'll try to initialize + * it to 1. + */ + bios_zero_thresh = 1; + val |= CMCI_THRESHOLD; + } + + val |= MCI_CTL2_CMCI_EN; wrmsrl(MSR_IA32_MCx_CTL2(i), val); rdmsrl(MSR_IA32_MCx_CTL2(i), val); /* Did the enable bit stick? -- the bank supports CMCI */ if (val & MCI_CTL2_CMCI_EN) { - if (!test_and_set_bit(i, owned) && !boot) - print_update("CMCI", &hdr, i); + set_bit(i, owned); __clear_bit(i, __get_cpu_var(mce_poll_banks)); + /* + * We are able to set thresholds for some banks that + * had a threshold of 0. This means the BIOS has not + * set the thresholds properly or does not work with + * this boot option. Note down now and report later. + */ + if (mca_cfg.bios_cmci_threshold && bios_zero_thresh && + (val & MCI_CTL2_CMCI_THRESHOLD_MASK)) + bios_wrong_thresh = 1; } else { WARN_ON(!test_bit(i, __get_cpu_var(mce_poll_banks))); } } - raw_spin_unlock_irqrestore(&cmci_discover_lock, flags); - if (hdr) - printk(KERN_CONT "\n"); + spin_unlock_irqrestore(&cmci_discover_lock, flags); + if (mca_cfg.bios_cmci_threshold && bios_wrong_thresh) { + pr_info_once( + "bios_cmci_threshold: Some banks do not have valid thresholds set\n"); + pr_info_once( + "bios_cmci_threshold: Make sure your BIOS supports this boot option\n"); + } } /* @@ -137,6 +291,19 @@ void cmci_recheck(void) local_irq_restore(flags); } +/* Caller must hold the lock on cmci_discover_lock */ +static void __cmci_disable_bank(int bank) +{ + u64 val; + + if (!test_bit(bank, __get_cpu_var(mce_banks_owned))) + return; + rdmsrl(MSR_IA32_MCx_CTL2(bank), val); + val &= ~MCI_CTL2_CMCI_EN; + wrmsrl(MSR_IA32_MCx_CTL2(bank), val); + __clear_bit(bank, __get_cpu_var(mce_banks_owned)); +} + /* * Disable CMCI on this CPU for all banks it owns when it goes down. * This allows other CPUs to claim the banks on rediscovery. @@ -146,51 +313,33 @@ void cmci_clear(void) unsigned long flags; int i; int banks; - u64 val; if (!cmci_supported(&banks)) return; - raw_spin_lock_irqsave(&cmci_discover_lock, flags); - for (i = 0; i < banks; i++) { - if (!test_bit(i, __get_cpu_var(mce_banks_owned))) - continue; - /* Disable CMCI */ - rdmsrl(MSR_IA32_MCx_CTL2(i), val); - val &= ~(MCI_CTL2_CMCI_EN|MCI_CTL2_CMCI_THRESHOLD_MASK); - wrmsrl(MSR_IA32_MCx_CTL2(i), val); - __clear_bit(i, __get_cpu_var(mce_banks_owned)); - } - raw_spin_unlock_irqrestore(&cmci_discover_lock, flags); + spin_lock_irqsave(&cmci_discover_lock, flags); + for (i = 0; i < banks; i++) + __cmci_disable_bank(i); + spin_unlock_irqrestore(&cmci_discover_lock, flags); } -/* - * After a CPU went down cycle through all the others and rediscover - * Must run in process context. - */ -void cmci_rediscover(int dying) +static void cmci_rediscover_work_func(void *arg) +{ + int banks; + + /* Recheck banks in case CPUs don't all have the same */ + if (cmci_supported(&banks)) + cmci_discover(banks); +} + +/* After a CPU went down cycle through all the others and rediscover */ +void cmci_rediscover(void) { int banks; - int cpu; - cpumask_var_t old; if (!cmci_supported(&banks)) return; - if (!alloc_cpumask_var(&old, GFP_KERNEL)) - return; - cpumask_copy(old, ¤t->cpus_allowed); - - for_each_online_cpu(cpu) { - if (cpu == dying) - continue; - if (set_cpus_allowed_ptr(current, cpumask_of(cpu))) - continue; - /* Recheck banks in case CPUs don't all have the same */ - if (cmci_supported(&banks)) - cmci_discover(banks, 0); - } - set_cpus_allowed_ptr(current, old); - free_cpumask_var(old); + on_each_cpu(cmci_rediscover_work_func, NULL, 1); } /* @@ -200,7 +349,20 @@ void cmci_reenable(void) { int banks; if (cmci_supported(&banks)) - cmci_discover(banks, 0); + cmci_discover(banks); +} + +void cmci_disable_bank(int bank) +{ + int banks; + unsigned long flags; + + if (!cmci_supported(&banks)) + return; + + spin_lock_irqsave(&cmci_discover_lock, flags); + __cmci_disable_bank(bank); + spin_unlock_irqrestore(&cmci_discover_lock, flags); } static void intel_init_cmci(void) @@ -211,7 +373,7 @@ static void intel_init_cmci(void) return; mce_threshold_vector = intel_threshold_interrupt; - cmci_discover(banks, 1); + cmci_discover(banks); /* * For CPU #0 this runs with still disabled APIC, but that's * ok because only the vector is set up. We still do another diff --git a/arch/x86/kernel/cpu/mcheck/p5.c b/arch/x86/kernel/cpu/mcheck/p5.c index 5c0e6533d9b..a3042989398 100644 --- a/arch/x86/kernel/cpu/mcheck/p5.c +++ b/arch/x86/kernel/cpu/mcheck/p5.c @@ -5,11 +5,9 @@ #include <linux/interrupt.h> #include <linux/kernel.h> #include <linux/types.h> -#include <linux/init.h> #include <linux/smp.h> #include <asm/processor.h> -#include <asm/system.h> #include <asm/mce.h> #include <asm/msr.h> @@ -34,7 +32,7 @@ static void pentium_machine_check(struct pt_regs *regs, long error_code) smp_processor_id()); } - add_taint(TAINT_MACHINE_CHECK); + add_taint(TAINT_MACHINE_CHECK, LOCKDEP_NOW_UNRELIABLE); } /* Set up machine check reporting for processors with Intel style MCE: */ diff --git a/arch/x86/kernel/cpu/mcheck/therm_throt.c b/arch/x86/kernel/cpu/mcheck/therm_throt.c index 67bb17a37a0..36a1bb6d1ee 100644 --- a/arch/x86/kernel/cpu/mcheck/therm_throt.c +++ b/arch/x86/kernel/cpu/mcheck/therm_throt.c @@ -25,11 +25,11 @@ #include <linux/cpu.h> #include <asm/processor.h> -#include <asm/system.h> #include <asm/apic.h> #include <asm/idle.h> #include <asm/mce.h> #include <asm/msr.h> +#include <asm/trace/irq_vectors.h> /* How long to wait between reporting thermal events */ #define CHECK_INTERVAL (300 * HZ) @@ -55,12 +55,24 @@ struct thermal_state { struct _thermal_state package_power_limit; struct _thermal_state core_thresh0; struct _thermal_state core_thresh1; + struct _thermal_state pkg_thresh0; + struct _thermal_state pkg_thresh1; }; /* Callback to handle core threshold interrupts */ int (*platform_thermal_notify)(__u64 msr_val); EXPORT_SYMBOL(platform_thermal_notify); +/* Callback to handle core package threshold_interrupts */ +int (*platform_thermal_package_notify)(__u64 msr_val); +EXPORT_SYMBOL_GPL(platform_thermal_package_notify); + +/* Callback support of rate control, return true, if + * callback has rate control */ +bool (*platform_thermal_package_rate_control)(void); +EXPORT_SYMBOL_GPL(platform_thermal_package_rate_control); + + static DEFINE_PER_CPU(struct thermal_state, thermal_state); static atomic_t therm_throt_en = ATOMIC_INIT(0); @@ -182,11 +194,6 @@ static int therm_throt_process(bool new_event, int event, int level) this_cpu, level == CORE_LEVEL ? "Core" : "Package", state->count); - else - printk(KERN_CRIT "CPU%d: %s power limit notification (total events = %lu)\n", - this_cpu, - level == CORE_LEVEL ? "Core" : "Package", - state->count); return 1; } if (old_event) { @@ -194,36 +201,46 @@ static int therm_throt_process(bool new_event, int event, int level) printk(KERN_INFO "CPU%d: %s temperature/speed normal\n", this_cpu, level == CORE_LEVEL ? "Core" : "Package"); - else - printk(KERN_INFO "CPU%d: %s power limit normal\n", - this_cpu, - level == CORE_LEVEL ? "Core" : "Package"); return 1; } return 0; } -static int thresh_event_valid(int event) +static int thresh_event_valid(int level, int event) { struct _thermal_state *state; unsigned int this_cpu = smp_processor_id(); struct thermal_state *pstate = &per_cpu(thermal_state, this_cpu); u64 now = get_jiffies_64(); - state = (event == 0) ? &pstate->core_thresh0 : &pstate->core_thresh1; + if (level == PACKAGE_LEVEL) + state = (event == 0) ? &pstate->pkg_thresh0 : + &pstate->pkg_thresh1; + else + state = (event == 0) ? &pstate->core_thresh0 : + &pstate->core_thresh1; if (time_before64(now, state->next_check)) return 0; state->next_check = now + CHECK_INTERVAL; + + return 1; +} + +static bool int_pln_enable; +static int __init int_pln_enable_setup(char *s) +{ + int_pln_enable = true; + return 1; } +__setup("int_pln_enable", int_pln_enable_setup); #ifdef CONFIG_SYSFS /* Add/Remove thermal_throttle interface for CPU device: */ -static __cpuinit int thermal_throttle_add_dev(struct device *dev, - unsigned int cpu) +static int thermal_throttle_add_dev(struct device *dev, unsigned int cpu) { int err; struct cpuinfo_x86 *c = &cpu_data(cpu); @@ -232,7 +249,7 @@ static __cpuinit int thermal_throttle_add_dev(struct device *dev, if (err) return err; - if (cpu_has(c, X86_FEATURE_PLN)) + if (cpu_has(c, X86_FEATURE_PLN) && int_pln_enable) err = sysfs_add_file_to_group(&dev->kobj, &dev_attr_core_power_limit_count.attr, thermal_attr_group.name); @@ -240,7 +257,7 @@ static __cpuinit int thermal_throttle_add_dev(struct device *dev, err = sysfs_add_file_to_group(&dev->kobj, &dev_attr_package_throttle_count.attr, thermal_attr_group.name); - if (cpu_has(c, X86_FEATURE_PLN)) + if (cpu_has(c, X86_FEATURE_PLN) && int_pln_enable) err = sysfs_add_file_to_group(&dev->kobj, &dev_attr_package_power_limit_count.attr, thermal_attr_group.name); @@ -249,16 +266,13 @@ static __cpuinit int thermal_throttle_add_dev(struct device *dev, return err; } -static __cpuinit void thermal_throttle_remove_dev(struct device *dev) +static void thermal_throttle_remove_dev(struct device *dev) { sysfs_remove_group(&dev->kobj, &thermal_attr_group); } -/* Mutex protecting device creation against CPU hotplug: */ -static DEFINE_MUTEX(therm_cpu_lock); - /* Get notified when a cpu comes on/off. Be hotplug friendly. */ -static __cpuinit int +static int thermal_throttle_cpu_callback(struct notifier_block *nfb, unsigned long action, void *hcpu) @@ -272,24 +286,20 @@ thermal_throttle_cpu_callback(struct notifier_block *nfb, switch (action) { case CPU_UP_PREPARE: case CPU_UP_PREPARE_FROZEN: - mutex_lock(&therm_cpu_lock); err = thermal_throttle_add_dev(dev, cpu); - mutex_unlock(&therm_cpu_lock); WARN_ON(err); break; case CPU_UP_CANCELED: case CPU_UP_CANCELED_FROZEN: case CPU_DEAD: case CPU_DEAD_FROZEN: - mutex_lock(&therm_cpu_lock); thermal_throttle_remove_dev(dev); - mutex_unlock(&therm_cpu_lock); break; } return notifier_from_errno(err); } -static struct notifier_block thermal_throttle_cpu_notifier __cpuinitdata = +static struct notifier_block thermal_throttle_cpu_notifier = { .notifier_call = thermal_throttle_cpu_callback, }; @@ -302,19 +312,16 @@ static __init int thermal_throttle_init_device(void) if (!atomic_read(&therm_throt_en)) return 0; - register_hotcpu_notifier(&thermal_throttle_cpu_notifier); + cpu_notifier_register_begin(); -#ifdef CONFIG_HOTPLUG_CPU - mutex_lock(&therm_cpu_lock); -#endif /* connect live CPUs to sysfs */ for_each_online_cpu(cpu) { err = thermal_throttle_add_dev(get_cpu_device(cpu), cpu); WARN_ON(err); } -#ifdef CONFIG_HOTPLUG_CPU - mutex_unlock(&therm_cpu_lock); -#endif + + __register_hotcpu_notifier(&thermal_throttle_cpu_notifier); + cpu_notifier_register_done(); return 0; } @@ -322,6 +329,39 @@ device_initcall(thermal_throttle_init_device); #endif /* CONFIG_SYSFS */ +static void notify_package_thresholds(__u64 msr_val) +{ + bool notify_thres_0 = false; + bool notify_thres_1 = false; + + if (!platform_thermal_package_notify) + return; + + /* lower threshold check */ + if (msr_val & THERM_LOG_THRESHOLD0) + notify_thres_0 = true; + /* higher threshold check */ + if (msr_val & THERM_LOG_THRESHOLD1) + notify_thres_1 = true; + + if (!notify_thres_0 && !notify_thres_1) + return; + + if (platform_thermal_package_rate_control && + platform_thermal_package_rate_control()) { + /* Rate control is implemented in callback */ + platform_thermal_package_notify(msr_val); + return; + } + + /* lower threshold reached */ + if (notify_thres_0 && thresh_event_valid(PACKAGE_LEVEL, 0)) + platform_thermal_package_notify(msr_val); + /* higher threshold reached */ + if (notify_thres_1 && thresh_event_valid(PACKAGE_LEVEL, 1)) + platform_thermal_package_notify(msr_val); +} + static void notify_thresholds(__u64 msr_val) { /* check whether the interrupt handler is defined; @@ -331,10 +371,12 @@ static void notify_thresholds(__u64 msr_val) return; /* lower threshold reached */ - if ((msr_val & THERM_LOG_THRESHOLD0) && thresh_event_valid(0)) + if ((msr_val & THERM_LOG_THRESHOLD0) && + thresh_event_valid(CORE_LEVEL, 0)) platform_thermal_notify(msr_val); /* higher threshold reached */ - if ((msr_val & THERM_LOG_THRESHOLD1) && thresh_event_valid(1)) + if ((msr_val & THERM_LOG_THRESHOLD1) && + thresh_event_valid(CORE_LEVEL, 1)) platform_thermal_notify(msr_val); } @@ -353,17 +395,19 @@ static void intel_thermal_interrupt(void) CORE_LEVEL) != 0) mce_log_therm_throt_event(msr_val); - if (this_cpu_has(X86_FEATURE_PLN)) + if (this_cpu_has(X86_FEATURE_PLN) && int_pln_enable) therm_throt_process(msr_val & THERM_STATUS_POWER_LIMIT, POWER_LIMIT_EVENT, CORE_LEVEL); if (this_cpu_has(X86_FEATURE_PTS)) { rdmsrl(MSR_IA32_PACKAGE_THERM_STATUS, msr_val); + /* check violations of package thermal thresholds */ + notify_package_thresholds(msr_val); therm_throt_process(msr_val & PACKAGE_THERM_STATUS_PROCHOT, THERMAL_THROTTLING_EVENT, PACKAGE_LEVEL); - if (this_cpu_has(X86_FEATURE_PLN)) + if (this_cpu_has(X86_FEATURE_PLN) && int_pln_enable) therm_throt_process(msr_val & PACKAGE_THERM_STATUS_POWER_LIMIT, POWER_LIMIT_EVENT, @@ -379,15 +423,26 @@ static void unexpected_thermal_interrupt(void) static void (*smp_thermal_vector)(void) = unexpected_thermal_interrupt; -asmlinkage void smp_thermal_interrupt(struct pt_regs *regs) +static inline void __smp_thermal_interrupt(void) { - irq_enter(); - exit_idle(); inc_irq_stat(irq_thermal_count); smp_thermal_vector(); - irq_exit(); - /* Ack only at the end to avoid potential reentry */ - ack_APIC_irq(); +} + +asmlinkage __visible void smp_thermal_interrupt(struct pt_regs *regs) +{ + entering_irq(); + __smp_thermal_interrupt(); + exiting_ack_irq(); +} + +asmlinkage __visible void smp_trace_thermal_interrupt(struct pt_regs *regs) +{ + entering_irq(); + trace_thermal_apic_entry(THERMAL_APIC_VECTOR); + __smp_thermal_interrupt(); + trace_thermal_apic_exit(THERMAL_APIC_VECTOR); + exiting_ack_irq(); } /* Thermal monitoring depends on APIC, ACPI and clock modulation */ @@ -471,9 +526,13 @@ void intel_init_thermal(struct cpuinfo_x86 *c) apic_write(APIC_LVTTHMR, h); rdmsr(MSR_IA32_THERM_INTERRUPT, l, h); - if (cpu_has(c, X86_FEATURE_PLN)) + if (cpu_has(c, X86_FEATURE_PLN) && !int_pln_enable) + wrmsr(MSR_IA32_THERM_INTERRUPT, + (l | (THERM_INT_LOW_ENABLE + | THERM_INT_HIGH_ENABLE)) & ~THERM_INT_PLN_ENABLE, h); + else if (cpu_has(c, X86_FEATURE_PLN) && int_pln_enable) wrmsr(MSR_IA32_THERM_INTERRUPT, - l | (THERM_INT_LOW_ENABLE + l | (THERM_INT_LOW_ENABLE | THERM_INT_HIGH_ENABLE | THERM_INT_PLN_ENABLE), h); else wrmsr(MSR_IA32_THERM_INTERRUPT, @@ -481,9 +540,14 @@ void intel_init_thermal(struct cpuinfo_x86 *c) if (cpu_has(c, X86_FEATURE_PTS)) { rdmsr(MSR_IA32_PACKAGE_THERM_INTERRUPT, l, h); - if (cpu_has(c, X86_FEATURE_PLN)) + if (cpu_has(c, X86_FEATURE_PLN) && !int_pln_enable) wrmsr(MSR_IA32_PACKAGE_THERM_INTERRUPT, - l | (PACKAGE_THERM_INT_LOW_ENABLE + (l | (PACKAGE_THERM_INT_LOW_ENABLE + | PACKAGE_THERM_INT_HIGH_ENABLE)) + & ~PACKAGE_THERM_INT_PLN_ENABLE, h); + else if (cpu_has(c, X86_FEATURE_PLN) && int_pln_enable) + wrmsr(MSR_IA32_PACKAGE_THERM_INTERRUPT, + l | (PACKAGE_THERM_INT_LOW_ENABLE | PACKAGE_THERM_INT_HIGH_ENABLE | PACKAGE_THERM_INT_PLN_ENABLE), h); else diff --git a/arch/x86/kernel/cpu/mcheck/threshold.c b/arch/x86/kernel/cpu/mcheck/threshold.c index aa578cadb94..7245980186e 100644 --- a/arch/x86/kernel/cpu/mcheck/threshold.c +++ b/arch/x86/kernel/cpu/mcheck/threshold.c @@ -8,6 +8,7 @@ #include <asm/apic.h> #include <asm/idle.h> #include <asm/mce.h> +#include <asm/trace/irq_vectors.h> static void default_threshold_interrupt(void) { @@ -17,13 +18,24 @@ static void default_threshold_interrupt(void) void (*mce_threshold_vector)(void) = default_threshold_interrupt; -asmlinkage void smp_threshold_interrupt(void) +static inline void __smp_threshold_interrupt(void) { - irq_enter(); - exit_idle(); inc_irq_stat(irq_threshold_count); mce_threshold_vector(); - irq_exit(); - /* Ack only at the end to avoid potential reentry */ - ack_APIC_irq(); +} + +asmlinkage __visible void smp_threshold_interrupt(void) +{ + entering_irq(); + __smp_threshold_interrupt(); + exiting_ack_irq(); +} + +asmlinkage __visible void smp_trace_threshold_interrupt(void) +{ + entering_irq(); + trace_threshold_apic_entry(THRESHOLD_APIC_VECTOR); + __smp_threshold_interrupt(); + trace_threshold_apic_exit(THRESHOLD_APIC_VECTOR); + exiting_ack_irq(); } diff --git a/arch/x86/kernel/cpu/mcheck/winchip.c b/arch/x86/kernel/cpu/mcheck/winchip.c index 54060f56597..7dc5564d0cd 100644 --- a/arch/x86/kernel/cpu/mcheck/winchip.c +++ b/arch/x86/kernel/cpu/mcheck/winchip.c @@ -5,10 +5,8 @@ #include <linux/interrupt.h> #include <linux/kernel.h> #include <linux/types.h> -#include <linux/init.h> #include <asm/processor.h> -#include <asm/system.h> #include <asm/mce.h> #include <asm/msr.h> @@ -16,7 +14,7 @@ static void winchip_machine_check(struct pt_regs *regs, long error_code) { printk(KERN_EMERG "CPU0: Machine Check Exception.\n"); - add_taint(TAINT_MACHINE_CHECK); + add_taint(TAINT_MACHINE_CHECK, LOCKDEP_NOW_UNRELIABLE); } /* Set up machine check reporting on the Winchip C6 series */ diff --git a/arch/x86/kernel/cpu/microcode/Makefile b/arch/x86/kernel/cpu/microcode/Makefile new file mode 100644 index 00000000000..285c85427c3 --- /dev/null +++ b/arch/x86/kernel/cpu/microcode/Makefile @@ -0,0 +1,7 @@ +microcode-y := core.o +obj-$(CONFIG_MICROCODE) += microcode.o +microcode-$(CONFIG_MICROCODE_INTEL) += intel.o intel_lib.o +microcode-$(CONFIG_MICROCODE_AMD) += amd.o +obj-$(CONFIG_MICROCODE_EARLY) += core_early.o +obj-$(CONFIG_MICROCODE_INTEL_EARLY) += intel_early.o +obj-$(CONFIG_MICROCODE_AMD_EARLY) += amd_early.o diff --git a/arch/x86/kernel/cpu/microcode/amd.c b/arch/x86/kernel/cpu/microcode/amd.c new file mode 100644 index 00000000000..8fffd845e22 --- /dev/null +++ b/arch/x86/kernel/cpu/microcode/amd.c @@ -0,0 +1,492 @@ +/* + * AMD CPU Microcode Update Driver for Linux + * Copyright (C) 2008-2011 Advanced Micro Devices Inc. + * + * Author: Peter Oruba <peter.oruba@amd.com> + * + * Based on work by: + * Tigran Aivazian <tigran@aivazian.fsnet.co.uk> + * + * Maintainers: + * Andreas Herrmann <herrmann.der.user@googlemail.com> + * Borislav Petkov <bp@alien8.de> + * + * This driver allows to upgrade microcode on F10h AMD + * CPUs and later. + * + * Licensed under the terms of the GNU General Public + * License version 2. See file COPYING for details. + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/firmware.h> +#include <linux/pci_ids.h> +#include <linux/uaccess.h> +#include <linux/vmalloc.h> +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/pci.h> + +#include <asm/microcode.h> +#include <asm/processor.h> +#include <asm/msr.h> +#include <asm/microcode_amd.h> + +MODULE_DESCRIPTION("AMD Microcode Update Driver"); +MODULE_AUTHOR("Peter Oruba"); +MODULE_LICENSE("GPL v2"); + +static struct equiv_cpu_entry *equiv_cpu_table; + +struct ucode_patch { + struct list_head plist; + void *data; + u32 patch_id; + u16 equiv_cpu; +}; + +static LIST_HEAD(pcache); + +static u16 __find_equiv_id(unsigned int cpu) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + return find_equiv_id(equiv_cpu_table, uci->cpu_sig.sig); +} + +static u32 find_cpu_family_by_equiv_cpu(u16 equiv_cpu) +{ + int i = 0; + + BUG_ON(!equiv_cpu_table); + + while (equiv_cpu_table[i].equiv_cpu != 0) { + if (equiv_cpu == equiv_cpu_table[i].equiv_cpu) + return equiv_cpu_table[i].installed_cpu; + i++; + } + return 0; +} + +/* + * a small, trivial cache of per-family ucode patches + */ +static struct ucode_patch *cache_find_patch(u16 equiv_cpu) +{ + struct ucode_patch *p; + + list_for_each_entry(p, &pcache, plist) + if (p->equiv_cpu == equiv_cpu) + return p; + return NULL; +} + +static void update_cache(struct ucode_patch *new_patch) +{ + struct ucode_patch *p; + + list_for_each_entry(p, &pcache, plist) { + if (p->equiv_cpu == new_patch->equiv_cpu) { + if (p->patch_id >= new_patch->patch_id) + /* we already have the latest patch */ + return; + + list_replace(&p->plist, &new_patch->plist); + kfree(p->data); + kfree(p); + return; + } + } + /* no patch found, add it */ + list_add_tail(&new_patch->plist, &pcache); +} + +static void free_cache(void) +{ + struct ucode_patch *p, *tmp; + + list_for_each_entry_safe(p, tmp, &pcache, plist) { + __list_del(p->plist.prev, p->plist.next); + kfree(p->data); + kfree(p); + } +} + +static struct ucode_patch *find_patch(unsigned int cpu) +{ + u16 equiv_id; + + equiv_id = __find_equiv_id(cpu); + if (!equiv_id) + return NULL; + + return cache_find_patch(equiv_id); +} + +static int collect_cpu_info_amd(int cpu, struct cpu_signature *csig) +{ + struct cpuinfo_x86 *c = &cpu_data(cpu); + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + struct ucode_patch *p; + + csig->sig = cpuid_eax(0x00000001); + csig->rev = c->microcode; + + /* + * a patch could have been loaded early, set uci->mc so that + * mc_bp_resume() can call apply_microcode() + */ + p = find_patch(cpu); + if (p && (p->patch_id == csig->rev)) + uci->mc = p->data; + + pr_info("CPU%d: patch_level=0x%08x\n", cpu, csig->rev); + + return 0; +} + +static unsigned int verify_patch_size(u8 family, u32 patch_size, + unsigned int size) +{ + u32 max_size; + +#define F1XH_MPB_MAX_SIZE 2048 +#define F14H_MPB_MAX_SIZE 1824 +#define F15H_MPB_MAX_SIZE 4096 +#define F16H_MPB_MAX_SIZE 3458 + + switch (family) { + case 0x14: + max_size = F14H_MPB_MAX_SIZE; + break; + case 0x15: + max_size = F15H_MPB_MAX_SIZE; + break; + case 0x16: + max_size = F16H_MPB_MAX_SIZE; + break; + default: + max_size = F1XH_MPB_MAX_SIZE; + break; + } + + if (patch_size > min_t(u32, size, max_size)) { + pr_err("patch size mismatch\n"); + return 0; + } + + return patch_size; +} + +int __apply_microcode_amd(struct microcode_amd *mc_amd) +{ + u32 rev, dummy; + + native_wrmsrl(MSR_AMD64_PATCH_LOADER, (u64)(long)&mc_amd->hdr.data_code); + + /* verify patch application was successful */ + native_rdmsr(MSR_AMD64_PATCH_LEVEL, rev, dummy); + if (rev != mc_amd->hdr.patch_id) + return -1; + + return 0; +} + +int apply_microcode_amd(int cpu) +{ + struct cpuinfo_x86 *c = &cpu_data(cpu); + struct microcode_amd *mc_amd; + struct ucode_cpu_info *uci; + struct ucode_patch *p; + u32 rev, dummy; + + BUG_ON(raw_smp_processor_id() != cpu); + + uci = ucode_cpu_info + cpu; + + p = find_patch(cpu); + if (!p) + return 0; + + mc_amd = p->data; + uci->mc = p->data; + + rdmsr(MSR_AMD64_PATCH_LEVEL, rev, dummy); + + /* need to apply patch? */ + if (rev >= mc_amd->hdr.patch_id) { + c->microcode = rev; + uci->cpu_sig.rev = rev; + return 0; + } + + if (__apply_microcode_amd(mc_amd)) { + pr_err("CPU%d: update failed for patch_level=0x%08x\n", + cpu, mc_amd->hdr.patch_id); + return -1; + } + pr_info("CPU%d: new patch_level=0x%08x\n", cpu, + mc_amd->hdr.patch_id); + + uci->cpu_sig.rev = mc_amd->hdr.patch_id; + c->microcode = mc_amd->hdr.patch_id; + + return 0; +} + +static int install_equiv_cpu_table(const u8 *buf) +{ + unsigned int *ibuf = (unsigned int *)buf; + unsigned int type = ibuf[1]; + unsigned int size = ibuf[2]; + + if (type != UCODE_EQUIV_CPU_TABLE_TYPE || !size) { + pr_err("empty section/" + "invalid type field in container file section header\n"); + return -EINVAL; + } + + equiv_cpu_table = vmalloc(size); + if (!equiv_cpu_table) { + pr_err("failed to allocate equivalent CPU table\n"); + return -ENOMEM; + } + + memcpy(equiv_cpu_table, buf + CONTAINER_HDR_SZ, size); + + /* add header length */ + return size + CONTAINER_HDR_SZ; +} + +static void free_equiv_cpu_table(void) +{ + vfree(equiv_cpu_table); + equiv_cpu_table = NULL; +} + +static void cleanup(void) +{ + free_equiv_cpu_table(); + free_cache(); +} + +/* + * We return the current size even if some of the checks failed so that + * we can skip over the next patch. If we return a negative value, we + * signal a grave error like a memory allocation has failed and the + * driver cannot continue functioning normally. In such cases, we tear + * down everything we've used up so far and exit. + */ +static int verify_and_add_patch(u8 family, u8 *fw, unsigned int leftover) +{ + struct microcode_header_amd *mc_hdr; + struct ucode_patch *patch; + unsigned int patch_size, crnt_size, ret; + u32 proc_fam; + u16 proc_id; + + patch_size = *(u32 *)(fw + 4); + crnt_size = patch_size + SECTION_HDR_SIZE; + mc_hdr = (struct microcode_header_amd *)(fw + SECTION_HDR_SIZE); + proc_id = mc_hdr->processor_rev_id; + + proc_fam = find_cpu_family_by_equiv_cpu(proc_id); + if (!proc_fam) { + pr_err("No patch family for equiv ID: 0x%04x\n", proc_id); + return crnt_size; + } + + /* check if patch is for the current family */ + proc_fam = ((proc_fam >> 8) & 0xf) + ((proc_fam >> 20) & 0xff); + if (proc_fam != family) + return crnt_size; + + if (mc_hdr->nb_dev_id || mc_hdr->sb_dev_id) { + pr_err("Patch-ID 0x%08x: chipset-specific code unsupported.\n", + mc_hdr->patch_id); + return crnt_size; + } + + ret = verify_patch_size(family, patch_size, leftover); + if (!ret) { + pr_err("Patch-ID 0x%08x: size mismatch.\n", mc_hdr->patch_id); + return crnt_size; + } + + patch = kzalloc(sizeof(*patch), GFP_KERNEL); + if (!patch) { + pr_err("Patch allocation failure.\n"); + return -EINVAL; + } + + patch->data = kzalloc(patch_size, GFP_KERNEL); + if (!patch->data) { + pr_err("Patch data allocation failure.\n"); + kfree(patch); + return -EINVAL; + } + + /* All looks ok, copy patch... */ + memcpy(patch->data, fw + SECTION_HDR_SIZE, patch_size); + INIT_LIST_HEAD(&patch->plist); + patch->patch_id = mc_hdr->patch_id; + patch->equiv_cpu = proc_id; + + pr_debug("%s: Added patch_id: 0x%08x, proc_id: 0x%04x\n", + __func__, patch->patch_id, proc_id); + + /* ... and add to cache. */ + update_cache(patch); + + return crnt_size; +} + +static enum ucode_state __load_microcode_amd(u8 family, const u8 *data, + size_t size) +{ + enum ucode_state ret = UCODE_ERROR; + unsigned int leftover; + u8 *fw = (u8 *)data; + int crnt_size = 0; + int offset; + + offset = install_equiv_cpu_table(data); + if (offset < 0) { + pr_err("failed to create equivalent cpu table\n"); + return ret; + } + fw += offset; + leftover = size - offset; + + if (*(u32 *)fw != UCODE_UCODE_TYPE) { + pr_err("invalid type field in container file section header\n"); + free_equiv_cpu_table(); + return ret; + } + + while (leftover) { + crnt_size = verify_and_add_patch(family, fw, leftover); + if (crnt_size < 0) + return ret; + + fw += crnt_size; + leftover -= crnt_size; + } + + return UCODE_OK; +} + +enum ucode_state load_microcode_amd(u8 family, const u8 *data, size_t size) +{ + enum ucode_state ret; + + /* free old equiv table */ + free_equiv_cpu_table(); + + ret = __load_microcode_amd(family, data, size); + + if (ret != UCODE_OK) + cleanup(); + +#if defined(CONFIG_MICROCODE_AMD_EARLY) && defined(CONFIG_X86_32) + /* save BSP's matching patch for early load */ + if (cpu_data(smp_processor_id()).cpu_index == boot_cpu_data.cpu_index) { + struct ucode_patch *p = find_patch(smp_processor_id()); + if (p) { + memset(amd_ucode_patch, 0, PATCH_MAX_SIZE); + memcpy(amd_ucode_patch, p->data, min_t(u32, ksize(p->data), + PATCH_MAX_SIZE)); + } + } +#endif + return ret; +} + +/* + * AMD microcode firmware naming convention, up to family 15h they are in + * the legacy file: + * + * amd-ucode/microcode_amd.bin + * + * This legacy file is always smaller than 2K in size. + * + * Beginning with family 15h, they are in family-specific firmware files: + * + * amd-ucode/microcode_amd_fam15h.bin + * amd-ucode/microcode_amd_fam16h.bin + * ... + * + * These might be larger than 2K. + */ +static enum ucode_state request_microcode_amd(int cpu, struct device *device, + bool refresh_fw) +{ + char fw_name[36] = "amd-ucode/microcode_amd.bin"; + struct cpuinfo_x86 *c = &cpu_data(cpu); + enum ucode_state ret = UCODE_NFOUND; + const struct firmware *fw; + + /* reload ucode container only on the boot cpu */ + if (!refresh_fw || c->cpu_index != boot_cpu_data.cpu_index) + return UCODE_OK; + + if (c->x86 >= 0x15) + snprintf(fw_name, sizeof(fw_name), "amd-ucode/microcode_amd_fam%.2xh.bin", c->x86); + + if (request_firmware_direct(&fw, (const char *)fw_name, device)) { + pr_debug("failed to load file %s\n", fw_name); + goto out; + } + + ret = UCODE_ERROR; + if (*(u32 *)fw->data != UCODE_MAGIC) { + pr_err("invalid magic value (0x%08x)\n", *(u32 *)fw->data); + goto fw_release; + } + + ret = load_microcode_amd(c->x86, fw->data, fw->size); + + fw_release: + release_firmware(fw); + + out: + return ret; +} + +static enum ucode_state +request_microcode_user(int cpu, const void __user *buf, size_t size) +{ + return UCODE_ERROR; +} + +static void microcode_fini_cpu_amd(int cpu) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + + uci->mc = NULL; +} + +static struct microcode_ops microcode_amd_ops = { + .request_microcode_user = request_microcode_user, + .request_microcode_fw = request_microcode_amd, + .collect_cpu_info = collect_cpu_info_amd, + .apply_microcode = apply_microcode_amd, + .microcode_fini_cpu = microcode_fini_cpu_amd, +}; + +struct microcode_ops * __init init_amd_microcode(void) +{ + struct cpuinfo_x86 *c = &cpu_data(0); + + if (c->x86_vendor != X86_VENDOR_AMD || c->x86 < 0x10) { + pr_warning("AMD CPU family 0x%x not supported\n", c->x86); + return NULL; + } + + return µcode_amd_ops; +} + +void __exit exit_amd_microcode(void) +{ + cleanup(); +} diff --git a/arch/x86/kernel/cpu/microcode/amd_early.c b/arch/x86/kernel/cpu/microcode/amd_early.c new file mode 100644 index 00000000000..617a9e28424 --- /dev/null +++ b/arch/x86/kernel/cpu/microcode/amd_early.c @@ -0,0 +1,395 @@ +/* + * Copyright (C) 2013 Advanced Micro Devices, Inc. + * + * Author: Jacob Shin <jacob.shin@amd.com> + * Fixes: Borislav Petkov <bp@suse.de> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/earlycpio.h> +#include <linux/initrd.h> + +#include <asm/cpu.h> +#include <asm/setup.h> +#include <asm/microcode_amd.h> + +/* + * This points to the current valid container of microcode patches which we will + * save from the initrd before jettisoning its contents. + */ +static u8 *container; +static size_t container_size; + +static u32 ucode_new_rev; +u8 amd_ucode_patch[PATCH_MAX_SIZE]; +static u16 this_equiv_id; + +struct cpio_data ucode_cpio; + +/* + * Microcode patch container file is prepended to the initrd in cpio format. + * See Documentation/x86/early-microcode.txt + */ +static __initdata char ucode_path[] = "kernel/x86/microcode/AuthenticAMD.bin"; + +static struct cpio_data __init find_ucode_in_initrd(void) +{ + long offset = 0; + char *path; + void *start; + size_t size; + +#ifdef CONFIG_X86_32 + struct boot_params *p; + + /* + * On 32-bit, early load occurs before paging is turned on so we need + * to use physical addresses. + */ + p = (struct boot_params *)__pa_nodebug(&boot_params); + path = (char *)__pa_nodebug(ucode_path); + start = (void *)p->hdr.ramdisk_image; + size = p->hdr.ramdisk_size; +#else + path = ucode_path; + start = (void *)(boot_params.hdr.ramdisk_image + PAGE_OFFSET); + size = boot_params.hdr.ramdisk_size; +#endif + + return find_cpio_data(path, start, size, &offset); +} + +static size_t compute_container_size(u8 *data, u32 total_size) +{ + size_t size = 0; + u32 *header = (u32 *)data; + + if (header[0] != UCODE_MAGIC || + header[1] != UCODE_EQUIV_CPU_TABLE_TYPE || /* type */ + header[2] == 0) /* size */ + return size; + + size = header[2] + CONTAINER_HDR_SZ; + total_size -= size; + data += size; + + while (total_size) { + u16 patch_size; + + header = (u32 *)data; + + if (header[0] != UCODE_UCODE_TYPE) + break; + + /* + * Sanity-check patch size. + */ + patch_size = header[1]; + if (patch_size > PATCH_MAX_SIZE) + break; + + size += patch_size + SECTION_HDR_SIZE; + data += patch_size + SECTION_HDR_SIZE; + total_size -= patch_size + SECTION_HDR_SIZE; + } + + return size; +} + +/* + * Early load occurs before we can vmalloc(). So we look for the microcode + * patch container file in initrd, traverse equivalent cpu table, look for a + * matching microcode patch, and update, all in initrd memory in place. + * When vmalloc() is available for use later -- on 64-bit during first AP load, + * and on 32-bit during save_microcode_in_initrd_amd() -- we can call + * load_microcode_amd() to save equivalent cpu table and microcode patches in + * kernel heap memory. + */ +static void apply_ucode_in_initrd(void *ucode, size_t size) +{ + struct equiv_cpu_entry *eq; + size_t *cont_sz; + u32 *header; + u8 *data, **cont; + u16 eq_id = 0; + int offset, left; + u32 rev, eax, ebx, ecx, edx; + u32 *new_rev; + +#ifdef CONFIG_X86_32 + new_rev = (u32 *)__pa_nodebug(&ucode_new_rev); + cont_sz = (size_t *)__pa_nodebug(&container_size); + cont = (u8 **)__pa_nodebug(&container); +#else + new_rev = &ucode_new_rev; + cont_sz = &container_size; + cont = &container; +#endif + + data = ucode; + left = size; + header = (u32 *)data; + + /* find equiv cpu table */ + if (header[0] != UCODE_MAGIC || + header[1] != UCODE_EQUIV_CPU_TABLE_TYPE || /* type */ + header[2] == 0) /* size */ + return; + + eax = 0x00000001; + ecx = 0; + native_cpuid(&eax, &ebx, &ecx, &edx); + + while (left > 0) { + eq = (struct equiv_cpu_entry *)(data + CONTAINER_HDR_SZ); + + *cont = data; + + /* Advance past the container header */ + offset = header[2] + CONTAINER_HDR_SZ; + data += offset; + left -= offset; + + eq_id = find_equiv_id(eq, eax); + if (eq_id) { + this_equiv_id = eq_id; + *cont_sz = compute_container_size(*cont, left + offset); + + /* + * truncate how much we need to iterate over in the + * ucode update loop below + */ + left = *cont_sz - offset; + break; + } + + /* + * support multiple container files appended together. if this + * one does not have a matching equivalent cpu entry, we fast + * forward to the next container file. + */ + while (left > 0) { + header = (u32 *)data; + if (header[0] == UCODE_MAGIC && + header[1] == UCODE_EQUIV_CPU_TABLE_TYPE) + break; + + offset = header[1] + SECTION_HDR_SIZE; + data += offset; + left -= offset; + } + + /* mark where the next microcode container file starts */ + offset = data - (u8 *)ucode; + ucode = data; + } + + if (!eq_id) { + *cont = NULL; + *cont_sz = 0; + return; + } + + /* find ucode and update if needed */ + + native_rdmsr(MSR_AMD64_PATCH_LEVEL, rev, eax); + + while (left > 0) { + struct microcode_amd *mc; + + header = (u32 *)data; + if (header[0] != UCODE_UCODE_TYPE || /* type */ + header[1] == 0) /* size */ + break; + + mc = (struct microcode_amd *)(data + SECTION_HDR_SIZE); + + if (eq_id == mc->hdr.processor_rev_id && rev < mc->hdr.patch_id) { + + if (!__apply_microcode_amd(mc)) { + rev = mc->hdr.patch_id; + *new_rev = rev; + + /* save ucode patch */ + memcpy(amd_ucode_patch, mc, + min_t(u32, header[1], PATCH_MAX_SIZE)); + } + } + + offset = header[1] + SECTION_HDR_SIZE; + data += offset; + left -= offset; + } +} + +void __init load_ucode_amd_bsp(void) +{ + struct cpio_data cp; + void **data; + size_t *size; + +#ifdef CONFIG_X86_32 + data = (void **)__pa_nodebug(&ucode_cpio.data); + size = (size_t *)__pa_nodebug(&ucode_cpio.size); +#else + data = &ucode_cpio.data; + size = &ucode_cpio.size; +#endif + + cp = find_ucode_in_initrd(); + if (!cp.data) + return; + + *data = cp.data; + *size = cp.size; + + apply_ucode_in_initrd(cp.data, cp.size); +} + +#ifdef CONFIG_X86_32 +/* + * On 32-bit, since AP's early load occurs before paging is turned on, we + * cannot traverse cpu_equiv_table and pcache in kernel heap memory. So during + * cold boot, AP will apply_ucode_in_initrd() just like the BSP. During + * save_microcode_in_initrd_amd() BSP's patch is copied to amd_ucode_patch, + * which is used upon resume from suspend. + */ +void load_ucode_amd_ap(void) +{ + struct microcode_amd *mc; + size_t *usize; + void **ucode; + + mc = (struct microcode_amd *)__pa(amd_ucode_patch); + if (mc->hdr.patch_id && mc->hdr.processor_rev_id) { + __apply_microcode_amd(mc); + return; + } + + ucode = (void *)__pa_nodebug(&container); + usize = (size_t *)__pa_nodebug(&container_size); + + if (!*ucode || !*usize) + return; + + apply_ucode_in_initrd(*ucode, *usize); +} + +static void __init collect_cpu_sig_on_bsp(void *arg) +{ + unsigned int cpu = smp_processor_id(); + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + + uci->cpu_sig.sig = cpuid_eax(0x00000001); +} + +static void __init get_bsp_sig(void) +{ + unsigned int bsp = boot_cpu_data.cpu_index; + struct ucode_cpu_info *uci = ucode_cpu_info + bsp; + + if (!uci->cpu_sig.sig) + smp_call_function_single(bsp, collect_cpu_sig_on_bsp, NULL, 1); +} +#else +void load_ucode_amd_ap(void) +{ + unsigned int cpu = smp_processor_id(); + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + struct equiv_cpu_entry *eq; + struct microcode_amd *mc; + u32 rev, eax; + u16 eq_id; + + /* Exit if called on the BSP. */ + if (!cpu) + return; + + if (!container) + return; + + rdmsr(MSR_AMD64_PATCH_LEVEL, rev, eax); + + uci->cpu_sig.rev = rev; + uci->cpu_sig.sig = eax; + + eax = cpuid_eax(0x00000001); + eq = (struct equiv_cpu_entry *)(container + CONTAINER_HDR_SZ); + + eq_id = find_equiv_id(eq, eax); + if (!eq_id) + return; + + if (eq_id == this_equiv_id) { + mc = (struct microcode_amd *)amd_ucode_patch; + + if (mc && rev < mc->hdr.patch_id) { + if (!__apply_microcode_amd(mc)) + ucode_new_rev = mc->hdr.patch_id; + } + + } else { + if (!ucode_cpio.data) + return; + + /* + * AP has a different equivalence ID than BSP, looks like + * mixed-steppings silicon so go through the ucode blob anew. + */ + apply_ucode_in_initrd(ucode_cpio.data, ucode_cpio.size); + } +} +#endif + +int __init save_microcode_in_initrd_amd(void) +{ + unsigned long cont; + enum ucode_state ret; + u32 eax; + + if (!container) + return -EINVAL; + +#ifdef CONFIG_X86_32 + get_bsp_sig(); + cont = (unsigned long)container; +#else + /* + * We need the physical address of the container for both bitness since + * boot_params.hdr.ramdisk_image is a physical address. + */ + cont = __pa(container); +#endif + + /* + * Take into account the fact that the ramdisk might get relocated and + * therefore we need to recompute the container's position in virtual + * memory space. + */ + if (relocated_ramdisk) + container = (u8 *)(__va(relocated_ramdisk) + + (cont - boot_params.hdr.ramdisk_image)); + + if (ucode_new_rev) + pr_info("microcode: updated early to new patch_level=0x%08x\n", + ucode_new_rev); + + eax = cpuid_eax(0x00000001); + eax = ((eax >> 8) & 0xf) + ((eax >> 20) & 0xff); + + ret = load_microcode_amd(eax, container, container_size); + if (ret != UCODE_OK) + return -EINVAL; + + /* + * This will be freed any msec now, stash patches for the current + * family and switch to patch cache for cpu hotplug, etc later. + */ + container = NULL; + container_size = 0; + + return 0; +} diff --git a/arch/x86/kernel/microcode_core.c b/arch/x86/kernel/cpu/microcode/core.c index fda91c30710..dd9d6190b08 100644 --- a/arch/x86/kernel/microcode_core.c +++ b/arch/x86/kernel/cpu/microcode/core.c @@ -86,6 +86,8 @@ #include <asm/microcode.h> #include <asm/processor.h> +#include <asm/cpu_device_id.h> +#include <asm/perf_event.h> MODULE_DESCRIPTION("Microcode Update Driver"); MODULE_AUTHOR("Tigran Aivazian <tigran@aivazian.fsnet.co.uk>"); @@ -95,6 +97,9 @@ MODULE_LICENSE("GPL"); static struct microcode_ops *microcode_ops; +bool dis_ucode_ldr; +module_param(dis_ucode_ldr, bool, 0); + /* * Synchronization. * @@ -223,6 +228,9 @@ static ssize_t microcode_write(struct file *file, const char __user *buf, if (do_microcode_update(buf, len) == 0) ret = (ssize_t)len; + if (ret > 0) + perf_check_microcode(); + mutex_unlock(µcode_mutex); put_online_cpus(); @@ -274,21 +282,18 @@ static struct platform_device *microcode_pdev; static int reload_for_cpu(int cpu) { struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + enum ucode_state ustate; int err = 0; - mutex_lock(µcode_mutex); - if (uci->valid) { - enum ucode_state ustate; - - ustate = microcode_ops->request_microcode_fw(cpu, µcode_pdev->dev); - if (ustate == UCODE_OK) - apply_microcode_on_target(cpu); - else - if (ustate == UCODE_ERROR) - err = -EINVAL; - } - mutex_unlock(µcode_mutex); + if (!uci->valid) + return err; + ustate = microcode_ops->request_microcode_fw(cpu, µcode_pdev->dev, true); + if (ustate == UCODE_OK) + apply_microcode_on_target(cpu); + else + if (ustate == UCODE_ERROR) + err = -EINVAL; return err; } @@ -297,20 +302,31 @@ static ssize_t reload_store(struct device *dev, const char *buf, size_t size) { unsigned long val; - int cpu = dev->id; - int ret = 0; - char *end; + int cpu; + ssize_t ret = 0, tmp_ret; - val = simple_strtoul(buf, &end, 0); - if (end == buf) - return -EINVAL; + ret = kstrtoul(buf, 0, &val); + if (ret) + return ret; + + if (val != 1) + return size; + + get_online_cpus(); + mutex_lock(µcode_mutex); + for_each_online_cpu(cpu) { + tmp_ret = reload_for_cpu(cpu); + if (tmp_ret != 0) + pr_warn("Error reloading microcode on CPU %d\n", cpu); - if (val == 1) { - get_online_cpus(); - if (cpu_online(cpu)) - ret = reload_for_cpu(cpu); - put_online_cpus(); + /* save retval of the first encountered reload error */ + if (!ret) + ret = tmp_ret; } + if (!ret) + perf_check_microcode(); + mutex_unlock(µcode_mutex); + put_online_cpus(); if (!ret) ret = size; @@ -339,7 +355,6 @@ static DEVICE_ATTR(version, 0400, version_show, NULL); static DEVICE_ATTR(processor_flags, 0400, pf_show, NULL); static struct attribute *mc_default_attrs[] = { - &dev_attr_reload.attr, &dev_attr_version.attr, &dev_attr_processor_flags.attr, NULL @@ -352,28 +367,26 @@ static struct attribute_group mc_attr_group = { static void microcode_fini_cpu(int cpu) { - struct ucode_cpu_info *uci = ucode_cpu_info + cpu; - microcode_ops->microcode_fini_cpu(cpu); - uci->valid = 0; } static enum ucode_state microcode_resume_cpu(int cpu) { - struct ucode_cpu_info *uci = ucode_cpu_info + cpu; - - if (!uci->mc) - return UCODE_NFOUND; - pr_debug("CPU%d updated upon resume\n", cpu); - apply_microcode_on_target(cpu); + + if (apply_microcode_on_target(cpu)) + return UCODE_ERROR; return UCODE_OK; } -static enum ucode_state microcode_init_cpu(int cpu) +static enum ucode_state microcode_init_cpu(int cpu, bool refresh_fw) { enum ucode_state ustate; + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + + if (uci && uci->valid) + return UCODE_OK; if (collect_cpu_info(cpu)) return UCODE_ERROR; @@ -382,7 +395,8 @@ static enum ucode_state microcode_init_cpu(int cpu) if (system_state != SYSTEM_RUNNING) return UCODE_NFOUND; - ustate = microcode_ops->request_microcode_fw(cpu, µcode_pdev->dev); + ustate = microcode_ops->request_microcode_fw(cpu, µcode_pdev->dev, + refresh_fw); if (ustate == UCODE_OK) { pr_debug("CPU%d updated upon init\n", cpu); @@ -395,14 +409,11 @@ static enum ucode_state microcode_init_cpu(int cpu) static enum ucode_state microcode_update_cpu(int cpu) { struct ucode_cpu_info *uci = ucode_cpu_info + cpu; - enum ucode_state ustate; if (uci->valid) - ustate = microcode_resume_cpu(cpu); - else - ustate = microcode_init_cpu(cpu); + return microcode_resume_cpu(cpu); - return ustate; + return microcode_init_cpu(cpu, false); } static int mc_device_add(struct device *dev, struct subsys_interface *sif) @@ -418,10 +429,8 @@ static int mc_device_add(struct device *dev, struct subsys_interface *sif) if (err) return err; - if (microcode_init_cpu(cpu) == UCODE_ERROR) { - sysfs_remove_group(&dev->kobj, &mc_attr_group); + if (microcode_init_cpu(cpu, true) == UCODE_ERROR) return -EINVAL; - } return err; } @@ -462,41 +471,48 @@ static struct syscore_ops mc_syscore_ops = { .resume = mc_bp_resume, }; -static __cpuinit int +static int mc_cpu_callback(struct notifier_block *nb, unsigned long action, void *hcpu) { unsigned int cpu = (unsigned long)hcpu; struct device *dev; dev = get_cpu_device(cpu); - switch (action) { + + switch (action & ~CPU_TASKS_FROZEN) { case CPU_ONLINE: - case CPU_ONLINE_FROZEN: microcode_update_cpu(cpu); - case CPU_DOWN_FAILED: - case CPU_DOWN_FAILED_FROZEN: pr_debug("CPU%d added\n", cpu); + /* + * "break" is missing on purpose here because we want to fall + * through in order to create the sysfs group. + */ + + case CPU_DOWN_FAILED: if (sysfs_create_group(&dev->kobj, &mc_attr_group)) pr_err("Failed to create group for CPU%d\n", cpu); break; + case CPU_DOWN_PREPARE: - case CPU_DOWN_PREPARE_FROZEN: /* Suspend is in progress, only remove the interface */ sysfs_remove_group(&dev->kobj, &mc_attr_group); pr_debug("CPU%d removed\n", cpu); break; /* + * case CPU_DEAD: + * * When a CPU goes offline, don't free up or invalidate the copy of * the microcode in kernel memory, so that we can reuse it when the * CPU comes back online without unnecessarily requesting the userspace * for it again. */ - case CPU_UP_CANCELED_FROZEN: - /* The CPU refused to come up during a system resume */ - microcode_fini_cpu(cpu); - break; } + + /* The CPU refused to come up during a system resume */ + if (action == CPU_UP_CANCELED_FROZEN) + microcode_fini_cpu(cpu); + return NOTIFY_OK; } @@ -504,20 +520,47 @@ static struct notifier_block __refdata mc_cpu_notifier = { .notifier_call = mc_cpu_callback, }; +#ifdef MODULE +/* Autoload on Intel and AMD systems */ +static const struct x86_cpu_id __initconst microcode_id[] = { +#ifdef CONFIG_MICROCODE_INTEL + { X86_VENDOR_INTEL, X86_FAMILY_ANY, X86_MODEL_ANY, }, +#endif +#ifdef CONFIG_MICROCODE_AMD + { X86_VENDOR_AMD, X86_FAMILY_ANY, X86_MODEL_ANY, }, +#endif + {} +}; +MODULE_DEVICE_TABLE(x86cpu, microcode_id); +#endif + +static struct attribute *cpu_root_microcode_attrs[] = { + &dev_attr_reload.attr, + NULL +}; + +static struct attribute_group cpu_root_microcode_group = { + .name = "microcode", + .attrs = cpu_root_microcode_attrs, +}; + static int __init microcode_init(void) { struct cpuinfo_x86 *c = &cpu_data(0); int error; + if (dis_ucode_ldr) + return 0; + if (c->x86_vendor == X86_VENDOR_INTEL) microcode_ops = init_intel_microcode(); else if (c->x86_vendor == X86_VENDOR_AMD) microcode_ops = init_amd_microcode(); - - if (!microcode_ops) { + else pr_err("no support for this CPU vendor\n"); + + if (!microcode_ops) return -ENODEV; - } microcode_pdev = platform_device_register_simple("microcode", -1, NULL, 0); @@ -528,16 +571,25 @@ static int __init microcode_init(void) mutex_lock(µcode_mutex); error = subsys_interface_register(&mc_cpu_interface); - + if (!error) + perf_check_microcode(); mutex_unlock(µcode_mutex); put_online_cpus(); if (error) goto out_pdev; + error = sysfs_create_group(&cpu_subsys.dev_root->kobj, + &cpu_root_microcode_group); + + if (error) { + pr_err("Error creating microcode group!\n"); + goto out_driver; + } + error = microcode_dev_init(); if (error) - goto out_driver; + goto out_ucode_group; register_syscore_ops(&mc_syscore_ops); register_hotcpu_notifier(&mc_cpu_notifier); @@ -547,7 +599,11 @@ static int __init microcode_init(void) return 0; -out_driver: + out_ucode_group: + sysfs_remove_group(&cpu_subsys.dev_root->kobj, + &cpu_root_microcode_group); + + out_driver: get_online_cpus(); mutex_lock(µcode_mutex); @@ -556,7 +612,7 @@ out_driver: mutex_unlock(µcode_mutex); put_online_cpus(); -out_pdev: + out_pdev: platform_device_unregister(microcode_pdev); return error; @@ -572,6 +628,9 @@ static void __exit microcode_exit(void) unregister_hotcpu_notifier(&mc_cpu_notifier); unregister_syscore_ops(&mc_syscore_ops); + sysfs_remove_group(&cpu_subsys.dev_root->kobj, + &cpu_root_microcode_group); + get_online_cpus(); mutex_lock(µcode_mutex); diff --git a/arch/x86/kernel/cpu/microcode/core_early.c b/arch/x86/kernel/cpu/microcode/core_early.c new file mode 100644 index 00000000000..5f28a64e71e --- /dev/null +++ b/arch/x86/kernel/cpu/microcode/core_early.c @@ -0,0 +1,178 @@ +/* + * X86 CPU microcode early update for Linux + * + * Copyright (C) 2012 Fenghua Yu <fenghua.yu@intel.com> + * H Peter Anvin" <hpa@zytor.com> + * + * This driver allows to early upgrade microcode on Intel processors + * belonging to IA-32 family - PentiumPro, Pentium II, + * Pentium III, Xeon, Pentium 4, etc. + * + * Reference: Section 9.11 of Volume 3, IA-32 Intel Architecture + * Software Developer's Manual. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ +#include <linux/module.h> +#include <asm/microcode.h> +#include <asm/microcode_intel.h> +#include <asm/microcode_amd.h> +#include <asm/processor.h> +#include <asm/cmdline.h> + +#define QCHAR(a, b, c, d) ((a) + ((b) << 8) + ((c) << 16) + ((d) << 24)) +#define CPUID_INTEL1 QCHAR('G', 'e', 'n', 'u') +#define CPUID_INTEL2 QCHAR('i', 'n', 'e', 'I') +#define CPUID_INTEL3 QCHAR('n', 't', 'e', 'l') +#define CPUID_AMD1 QCHAR('A', 'u', 't', 'h') +#define CPUID_AMD2 QCHAR('e', 'n', 't', 'i') +#define CPUID_AMD3 QCHAR('c', 'A', 'M', 'D') + +#define CPUID_IS(a, b, c, ebx, ecx, edx) \ + (!((ebx ^ (a))|(edx ^ (b))|(ecx ^ (c)))) + +/* + * In early loading microcode phase on BSP, boot_cpu_data is not set up yet. + * x86_vendor() gets vendor id for BSP. + * + * In 32 bit AP case, accessing boot_cpu_data needs linear address. To simplify + * coding, we still use x86_vendor() to get vendor id for AP. + * + * x86_vendor() gets vendor information directly through cpuid. + */ +static int x86_vendor(void) +{ + u32 eax = 0x00000000; + u32 ebx, ecx = 0, edx; + + native_cpuid(&eax, &ebx, &ecx, &edx); + + if (CPUID_IS(CPUID_INTEL1, CPUID_INTEL2, CPUID_INTEL3, ebx, ecx, edx)) + return X86_VENDOR_INTEL; + + if (CPUID_IS(CPUID_AMD1, CPUID_AMD2, CPUID_AMD3, ebx, ecx, edx)) + return X86_VENDOR_AMD; + + return X86_VENDOR_UNKNOWN; +} + +static int x86_family(void) +{ + u32 eax = 0x00000001; + u32 ebx, ecx = 0, edx; + int x86; + + native_cpuid(&eax, &ebx, &ecx, &edx); + + x86 = (eax >> 8) & 0xf; + if (x86 == 15) + x86 += (eax >> 20) & 0xff; + + return x86; +} + +static bool __init check_loader_disabled_bsp(void) +{ +#ifdef CONFIG_X86_32 + const char *cmdline = (const char *)__pa_nodebug(boot_command_line); + const char *opt = "dis_ucode_ldr"; + const char *option = (const char *)__pa_nodebug(opt); + bool *res = (bool *)__pa_nodebug(&dis_ucode_ldr); + +#else /* CONFIG_X86_64 */ + const char *cmdline = boot_command_line; + const char *option = "dis_ucode_ldr"; + bool *res = &dis_ucode_ldr; +#endif + + if (cmdline_find_option_bool(cmdline, option)) + *res = true; + + return *res; +} + +void __init load_ucode_bsp(void) +{ + int vendor, x86; + + if (check_loader_disabled_bsp()) + return; + + if (!have_cpuid_p()) + return; + + vendor = x86_vendor(); + x86 = x86_family(); + + switch (vendor) { + case X86_VENDOR_INTEL: + if (x86 >= 6) + load_ucode_intel_bsp(); + break; + case X86_VENDOR_AMD: + if (x86 >= 0x10) + load_ucode_amd_bsp(); + break; + default: + break; + } +} + +static bool check_loader_disabled_ap(void) +{ +#ifdef CONFIG_X86_32 + return __pa_nodebug(dis_ucode_ldr); +#else + return dis_ucode_ldr; +#endif +} + +void load_ucode_ap(void) +{ + int vendor, x86; + + if (check_loader_disabled_ap()) + return; + + if (!have_cpuid_p()) + return; + + vendor = x86_vendor(); + x86 = x86_family(); + + switch (vendor) { + case X86_VENDOR_INTEL: + if (x86 >= 6) + load_ucode_intel_ap(); + break; + case X86_VENDOR_AMD: + if (x86 >= 0x10) + load_ucode_amd_ap(); + break; + default: + break; + } +} + +int __init save_microcode_in_initrd(void) +{ + struct cpuinfo_x86 *c = &boot_cpu_data; + + switch (c->x86_vendor) { + case X86_VENDOR_INTEL: + if (c->x86 >= 6) + save_microcode_in_initrd_intel(); + break; + case X86_VENDOR_AMD: + if (c->x86 >= 0x10) + save_microcode_in_initrd_amd(); + break; + default: + break; + } + + return 0; +} diff --git a/arch/x86/kernel/microcode_intel.c b/arch/x86/kernel/cpu/microcode/intel.c index 3ca42d0e43a..a276fa75d9b 100644 --- a/arch/x86/kernel/microcode_intel.c +++ b/arch/x86/kernel/cpu/microcode/intel.c @@ -79,7 +79,7 @@ #include <linux/module.h> #include <linux/vmalloc.h> -#include <asm/microcode.h> +#include <asm/microcode_intel.h> #include <asm/processor.h> #include <asm/msr.h> @@ -87,59 +87,6 @@ MODULE_DESCRIPTION("Microcode Update Driver"); MODULE_AUTHOR("Tigran Aivazian <tigran@aivazian.fsnet.co.uk>"); MODULE_LICENSE("GPL"); -struct microcode_header_intel { - unsigned int hdrver; - unsigned int rev; - unsigned int date; - unsigned int sig; - unsigned int cksum; - unsigned int ldrver; - unsigned int pf; - unsigned int datasize; - unsigned int totalsize; - unsigned int reserved[3]; -}; - -struct microcode_intel { - struct microcode_header_intel hdr; - unsigned int bits[0]; -}; - -/* microcode format is extended from prescott processors */ -struct extended_signature { - unsigned int sig; - unsigned int pf; - unsigned int cksum; -}; - -struct extended_sigtable { - unsigned int count; - unsigned int cksum; - unsigned int reserved[3]; - struct extended_signature sigs[0]; -}; - -#define DEFAULT_UCODE_DATASIZE (2000) -#define MC_HEADER_SIZE (sizeof(struct microcode_header_intel)) -#define DEFAULT_UCODE_TOTALSIZE (DEFAULT_UCODE_DATASIZE + MC_HEADER_SIZE) -#define EXT_HEADER_SIZE (sizeof(struct extended_sigtable)) -#define EXT_SIGNATURE_SIZE (sizeof(struct extended_signature)) -#define DWSIZE (sizeof(u32)) - -#define get_totalsize(mc) \ - (((struct microcode_intel *)mc)->hdr.totalsize ? \ - ((struct microcode_intel *)mc)->hdr.totalsize : \ - DEFAULT_UCODE_TOTALSIZE) - -#define get_datasize(mc) \ - (((struct microcode_intel *)mc)->hdr.datasize ? \ - ((struct microcode_intel *)mc)->hdr.datasize : DEFAULT_UCODE_DATASIZE) - -#define sigmatch(s1, s2, p1, p2) \ - (((s1) == (s2)) && (((p1) & (p2)) || (((p1) == 0) && ((p2) == 0)))) - -#define exttable_size(et) ((et)->count * EXT_SIGNATURE_SIZE + EXT_HEADER_SIZE) - static int collect_cpu_info(int cpu_num, struct cpu_signature *csig) { struct cpuinfo_x86 *c = &cpu_data(cpu_num); @@ -147,12 +94,6 @@ static int collect_cpu_info(int cpu_num, struct cpu_signature *csig) memset(csig, 0, sizeof(*csig)); - if (c->x86_vendor != X86_VENDOR_INTEL || c->x86 < 6 || - cpu_has(c, X86_FEATURE_IA64)) { - pr_err("CPU%d not a capable Intel processor\n", cpu_num); - return -1; - } - csig->sig = cpuid_eax(0x00000001); if ((c->x86_model >= 5) || (c->x86 > 6)) { @@ -168,128 +109,25 @@ static int collect_cpu_info(int cpu_num, struct cpu_signature *csig) return 0; } -static inline int update_match_cpu(struct cpu_signature *csig, int sig, int pf) -{ - return (!sigmatch(sig, csig->sig, pf, csig->pf)) ? 0 : 1; -} - -static inline int -update_match_revision(struct microcode_header_intel *mc_header, int rev) -{ - return (mc_header->rev <= rev) ? 0 : 1; -} - -static int microcode_sanity_check(void *mc) -{ - unsigned long total_size, data_size, ext_table_size; - struct microcode_header_intel *mc_header = mc; - struct extended_sigtable *ext_header = NULL; - int sum, orig_sum, ext_sigcount = 0, i; - struct extended_signature *ext_sig; - - total_size = get_totalsize(mc_header); - data_size = get_datasize(mc_header); - - if (data_size + MC_HEADER_SIZE > total_size) { - pr_err("error! Bad data size in microcode data file\n"); - return -EINVAL; - } - - if (mc_header->ldrver != 1 || mc_header->hdrver != 1) { - pr_err("error! Unknown microcode update format\n"); - return -EINVAL; - } - ext_table_size = total_size - (MC_HEADER_SIZE + data_size); - if (ext_table_size) { - if ((ext_table_size < EXT_HEADER_SIZE) - || ((ext_table_size - EXT_HEADER_SIZE) % EXT_SIGNATURE_SIZE)) { - pr_err("error! Small exttable size in microcode data file\n"); - return -EINVAL; - } - ext_header = mc + MC_HEADER_SIZE + data_size; - if (ext_table_size != exttable_size(ext_header)) { - pr_err("error! Bad exttable size in microcode data file\n"); - return -EFAULT; - } - ext_sigcount = ext_header->count; - } - - /* check extended table checksum */ - if (ext_table_size) { - int ext_table_sum = 0; - int *ext_tablep = (int *)ext_header; - - i = ext_table_size / DWSIZE; - while (i--) - ext_table_sum += ext_tablep[i]; - if (ext_table_sum) { - pr_warning("aborting, bad extended signature table checksum\n"); - return -EINVAL; - } - } - - /* calculate the checksum */ - orig_sum = 0; - i = (MC_HEADER_SIZE + data_size) / DWSIZE; - while (i--) - orig_sum += ((int *)mc)[i]; - if (orig_sum) { - pr_err("aborting, bad checksum\n"); - return -EINVAL; - } - if (!ext_table_size) - return 0; - /* check extended signature checksum */ - for (i = 0; i < ext_sigcount; i++) { - ext_sig = (void *)ext_header + EXT_HEADER_SIZE + - EXT_SIGNATURE_SIZE * i; - sum = orig_sum - - (mc_header->sig + mc_header->pf + mc_header->cksum) - + (ext_sig->sig + ext_sig->pf + ext_sig->cksum); - if (sum) { - pr_err("aborting, bad checksum\n"); - return -EINVAL; - } - } - return 0; -} - /* * return 0 - no update found * return 1 - found update */ -static int -get_matching_microcode(struct cpu_signature *cpu_sig, void *mc, int rev) +static int get_matching_mc(struct microcode_intel *mc_intel, int cpu) { - struct microcode_header_intel *mc_header = mc; - struct extended_sigtable *ext_header; - unsigned long total_size = get_totalsize(mc_header); - int ext_sigcount, i; - struct extended_signature *ext_sig; + struct cpu_signature cpu_sig; + unsigned int csig, cpf, crev; - if (!update_match_revision(mc_header, rev)) - return 0; - - if (update_match_cpu(cpu_sig, mc_header->sig, mc_header->pf)) - return 1; - - /* Look for ext. headers: */ - if (total_size <= get_datasize(mc_header) + MC_HEADER_SIZE) - return 0; + collect_cpu_info(cpu, &cpu_sig); - ext_header = mc + get_datasize(mc_header) + MC_HEADER_SIZE; - ext_sigcount = ext_header->count; - ext_sig = (void *)ext_header + EXT_HEADER_SIZE; + csig = cpu_sig.sig; + cpf = cpu_sig.pf; + crev = cpu_sig.rev; - for (i = 0; i < ext_sigcount; i++) { - if (update_match_cpu(cpu_sig, ext_sig->sig, ext_sig->pf)) - return 1; - ext_sig++; - } - return 0; + return get_matching_microcode(csig, cpf, mc_intel, crev); } -static int apply_microcode(int cpu) +int apply_microcode(int cpu) { struct microcode_intel *mc_intel; struct ucode_cpu_info *uci; @@ -306,6 +144,14 @@ static int apply_microcode(int cpu) if (mc_intel == NULL) return 0; + /* + * Microcode on this CPU could be updated earlier. Only apply the + * microcode patch in mc_intel when it is newer than the one on this + * CPU. + */ + if (get_matching_mc(mc_intel, cpu) == 0) + return 0; + /* write microcode via MSR 0x79 */ wrmsr(MSR_IA32_UCODE_WRITE, (unsigned long) mc_intel->bits, @@ -344,6 +190,7 @@ static enum ucode_state generic_load_microcode(int cpu, void *data, size_t size, unsigned int leftover = size; enum ucode_state state = UCODE_OK; unsigned int curr_mc_size = 0; + unsigned int csig, cpf; while (leftover) { struct microcode_header_intel mc_header; @@ -368,11 +215,13 @@ static enum ucode_state generic_load_microcode(int cpu, void *data, size_t size, } if (get_ucode_data(mc, ucode_ptr, mc_size) || - microcode_sanity_check(mc) < 0) { + microcode_sanity_check(mc, 1) < 0) { break; } - if (get_matching_microcode(&uci->cpu_sig, mc, new_rev)) { + csig = uci->cpu_sig.sig; + cpf = uci->cpu_sig.pf; + if (get_matching_microcode(csig, cpf, mc, new_rev)) { vfree(new_mc); new_rev = mc_header.rev; new_mc = mc; @@ -399,6 +248,13 @@ static enum ucode_state generic_load_microcode(int cpu, void *data, size_t size, vfree(uci->mc); uci->mc = (struct microcode_intel *)new_mc; + /* + * If early loading microcode is supported, save this mc into + * permanent memory. So it will be loaded early when a CPU is hot added + * or resumes. + */ + save_mc_for_early(new_mc); + pr_debug("CPU%d found a matching microcode update with version 0x%x (current=0x%x)\n", cpu, new_rev, uci->cpu_sig.rev); out: @@ -411,7 +267,8 @@ static int get_ucode_fw(void *to, const void *from, size_t n) return 0; } -static enum ucode_state request_microcode_fw(int cpu, struct device *device) +static enum ucode_state request_microcode_fw(int cpu, struct device *device, + bool refresh_fw) { char name[30]; struct cpuinfo_x86 *c = &cpu_data(cpu); @@ -421,7 +278,7 @@ static enum ucode_state request_microcode_fw(int cpu, struct device *device) sprintf(name, "intel-ucode/%02x-%02x-%02x", c->x86, c->x86_model, c->x86_mask); - if (request_firmware(&firmware, name, device)) { + if (request_firmware_direct(&firmware, name, device)) { pr_debug("data file %s load failed\n", name); return UCODE_NFOUND; } @@ -463,6 +320,14 @@ static struct microcode_ops microcode_intel_ops = { struct microcode_ops * __init init_intel_microcode(void) { + struct cpuinfo_x86 *c = &cpu_data(0); + + if (c->x86_vendor != X86_VENDOR_INTEL || c->x86 < 6 || + cpu_has(c, X86_FEATURE_IA64)) { + pr_err("Intel CPU family 0x%x not supported\n", c->x86); + return NULL; + } + return µcode_intel_ops; } diff --git a/arch/x86/kernel/cpu/microcode/intel_early.c b/arch/x86/kernel/cpu/microcode/intel_early.c new file mode 100644 index 00000000000..18f739129e7 --- /dev/null +++ b/arch/x86/kernel/cpu/microcode/intel_early.c @@ -0,0 +1,787 @@ +/* + * Intel CPU microcode early update for Linux + * + * Copyright (C) 2012 Fenghua Yu <fenghua.yu@intel.com> + * H Peter Anvin" <hpa@zytor.com> + * + * This allows to early upgrade microcode on Intel processors + * belonging to IA-32 family - PentiumPro, Pentium II, + * Pentium III, Xeon, Pentium 4, etc. + * + * Reference: Section 9.11 of Volume 3, IA-32 Intel Architecture + * Software Developer's Manual. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ +#include <linux/module.h> +#include <linux/mm.h> +#include <linux/slab.h> +#include <linux/earlycpio.h> +#include <linux/initrd.h> +#include <linux/cpu.h> +#include <asm/msr.h> +#include <asm/microcode_intel.h> +#include <asm/processor.h> +#include <asm/tlbflush.h> +#include <asm/setup.h> + +unsigned long mc_saved_in_initrd[MAX_UCODE_COUNT]; +struct mc_saved_data { + unsigned int mc_saved_count; + struct microcode_intel **mc_saved; +} mc_saved_data; + +static enum ucode_state +generic_load_microcode_early(struct microcode_intel **mc_saved_p, + unsigned int mc_saved_count, + struct ucode_cpu_info *uci) +{ + struct microcode_intel *ucode_ptr, *new_mc = NULL; + int new_rev = uci->cpu_sig.rev; + enum ucode_state state = UCODE_OK; + unsigned int mc_size; + struct microcode_header_intel *mc_header; + unsigned int csig = uci->cpu_sig.sig; + unsigned int cpf = uci->cpu_sig.pf; + int i; + + for (i = 0; i < mc_saved_count; i++) { + ucode_ptr = mc_saved_p[i]; + + mc_header = (struct microcode_header_intel *)ucode_ptr; + mc_size = get_totalsize(mc_header); + if (get_matching_microcode(csig, cpf, ucode_ptr, new_rev)) { + new_rev = mc_header->rev; + new_mc = ucode_ptr; + } + } + + if (!new_mc) { + state = UCODE_NFOUND; + goto out; + } + + uci->mc = (struct microcode_intel *)new_mc; +out: + return state; +} + +static void +microcode_pointer(struct microcode_intel **mc_saved, + unsigned long *mc_saved_in_initrd, + unsigned long initrd_start, int mc_saved_count) +{ + int i; + + for (i = 0; i < mc_saved_count; i++) + mc_saved[i] = (struct microcode_intel *) + (mc_saved_in_initrd[i] + initrd_start); +} + +#ifdef CONFIG_X86_32 +static void +microcode_phys(struct microcode_intel **mc_saved_tmp, + struct mc_saved_data *mc_saved_data) +{ + int i; + struct microcode_intel ***mc_saved; + + mc_saved = (struct microcode_intel ***) + __pa_nodebug(&mc_saved_data->mc_saved); + for (i = 0; i < mc_saved_data->mc_saved_count; i++) { + struct microcode_intel *p; + + p = *(struct microcode_intel **) + __pa_nodebug(mc_saved_data->mc_saved + i); + mc_saved_tmp[i] = (struct microcode_intel *)__pa_nodebug(p); + } +} +#endif + +static enum ucode_state +load_microcode(struct mc_saved_data *mc_saved_data, + unsigned long *mc_saved_in_initrd, + unsigned long initrd_start, + struct ucode_cpu_info *uci) +{ + struct microcode_intel *mc_saved_tmp[MAX_UCODE_COUNT]; + unsigned int count = mc_saved_data->mc_saved_count; + + if (!mc_saved_data->mc_saved) { + microcode_pointer(mc_saved_tmp, mc_saved_in_initrd, + initrd_start, count); + + return generic_load_microcode_early(mc_saved_tmp, count, uci); + } else { +#ifdef CONFIG_X86_32 + microcode_phys(mc_saved_tmp, mc_saved_data); + return generic_load_microcode_early(mc_saved_tmp, count, uci); +#else + return generic_load_microcode_early(mc_saved_data->mc_saved, + count, uci); +#endif + } +} + +static u8 get_x86_family(unsigned long sig) +{ + u8 x86; + + x86 = (sig >> 8) & 0xf; + + if (x86 == 0xf) + x86 += (sig >> 20) & 0xff; + + return x86; +} + +static u8 get_x86_model(unsigned long sig) +{ + u8 x86, x86_model; + + x86 = get_x86_family(sig); + x86_model = (sig >> 4) & 0xf; + + if (x86 == 0x6 || x86 == 0xf) + x86_model += ((sig >> 16) & 0xf) << 4; + + return x86_model; +} + +/* + * Given CPU signature and a microcode patch, this function finds if the + * microcode patch has matching family and model with the CPU. + */ +static enum ucode_state +matching_model_microcode(struct microcode_header_intel *mc_header, + unsigned long sig) +{ + u8 x86, x86_model; + u8 x86_ucode, x86_model_ucode; + struct extended_sigtable *ext_header; + unsigned long total_size = get_totalsize(mc_header); + unsigned long data_size = get_datasize(mc_header); + int ext_sigcount, i; + struct extended_signature *ext_sig; + + x86 = get_x86_family(sig); + x86_model = get_x86_model(sig); + + x86_ucode = get_x86_family(mc_header->sig); + x86_model_ucode = get_x86_model(mc_header->sig); + + if (x86 == x86_ucode && x86_model == x86_model_ucode) + return UCODE_OK; + + /* Look for ext. headers: */ + if (total_size <= data_size + MC_HEADER_SIZE) + return UCODE_NFOUND; + + ext_header = (struct extended_sigtable *) + mc_header + data_size + MC_HEADER_SIZE; + ext_sigcount = ext_header->count; + ext_sig = (void *)ext_header + EXT_HEADER_SIZE; + + for (i = 0; i < ext_sigcount; i++) { + x86_ucode = get_x86_family(ext_sig->sig); + x86_model_ucode = get_x86_model(ext_sig->sig); + + if (x86 == x86_ucode && x86_model == x86_model_ucode) + return UCODE_OK; + + ext_sig++; + } + + return UCODE_NFOUND; +} + +static int +save_microcode(struct mc_saved_data *mc_saved_data, + struct microcode_intel **mc_saved_src, + unsigned int mc_saved_count) +{ + int i, j; + struct microcode_intel **mc_saved_p; + int ret; + + if (!mc_saved_count) + return -EINVAL; + + /* + * Copy new microcode data. + */ + mc_saved_p = kmalloc(mc_saved_count*sizeof(struct microcode_intel *), + GFP_KERNEL); + if (!mc_saved_p) + return -ENOMEM; + + for (i = 0; i < mc_saved_count; i++) { + struct microcode_intel *mc = mc_saved_src[i]; + struct microcode_header_intel *mc_header = &mc->hdr; + unsigned long mc_size = get_totalsize(mc_header); + mc_saved_p[i] = kmalloc(mc_size, GFP_KERNEL); + if (!mc_saved_p[i]) { + ret = -ENOMEM; + goto err; + } + if (!mc_saved_src[i]) { + ret = -EINVAL; + goto err; + } + memcpy(mc_saved_p[i], mc, mc_size); + } + + /* + * Point to newly saved microcode. + */ + mc_saved_data->mc_saved = mc_saved_p; + mc_saved_data->mc_saved_count = mc_saved_count; + + return 0; + +err: + for (j = 0; j <= i; j++) + kfree(mc_saved_p[j]); + kfree(mc_saved_p); + + return ret; +} + +/* + * A microcode patch in ucode_ptr is saved into mc_saved + * - if it has matching signature and newer revision compared to an existing + * patch mc_saved. + * - or if it is a newly discovered microcode patch. + * + * The microcode patch should have matching model with CPU. + */ +static void _save_mc(struct microcode_intel **mc_saved, u8 *ucode_ptr, + unsigned int *mc_saved_count_p) +{ + int i; + int found = 0; + unsigned int mc_saved_count = *mc_saved_count_p; + struct microcode_header_intel *mc_header; + + mc_header = (struct microcode_header_intel *)ucode_ptr; + for (i = 0; i < mc_saved_count; i++) { + unsigned int sig, pf; + unsigned int new_rev; + struct microcode_header_intel *mc_saved_header = + (struct microcode_header_intel *)mc_saved[i]; + sig = mc_saved_header->sig; + pf = mc_saved_header->pf; + new_rev = mc_header->rev; + + if (get_matching_sig(sig, pf, ucode_ptr, new_rev)) { + found = 1; + if (update_match_revision(mc_header, new_rev)) { + /* + * Found an older ucode saved before. + * Replace the older one with this newer + * one. + */ + mc_saved[i] = + (struct microcode_intel *)ucode_ptr; + break; + } + } + } + if (i >= mc_saved_count && !found) + /* + * This ucode is first time discovered in ucode file. + * Save it to memory. + */ + mc_saved[mc_saved_count++] = + (struct microcode_intel *)ucode_ptr; + + *mc_saved_count_p = mc_saved_count; +} + +/* + * Get microcode matching with BSP's model. Only CPUs with the same model as + * BSP can stay in the platform. + */ +static enum ucode_state __init +get_matching_model_microcode(int cpu, unsigned long start, + void *data, size_t size, + struct mc_saved_data *mc_saved_data, + unsigned long *mc_saved_in_initrd, + struct ucode_cpu_info *uci) +{ + u8 *ucode_ptr = data; + unsigned int leftover = size; + enum ucode_state state = UCODE_OK; + unsigned int mc_size; + struct microcode_header_intel *mc_header; + struct microcode_intel *mc_saved_tmp[MAX_UCODE_COUNT]; + unsigned int mc_saved_count = mc_saved_data->mc_saved_count; + int i; + + while (leftover) { + mc_header = (struct microcode_header_intel *)ucode_ptr; + + mc_size = get_totalsize(mc_header); + if (!mc_size || mc_size > leftover || + microcode_sanity_check(ucode_ptr, 0) < 0) + break; + + leftover -= mc_size; + + /* + * Since APs with same family and model as the BSP may boot in + * the platform, we need to find and save microcode patches + * with the same family and model as the BSP. + */ + if (matching_model_microcode(mc_header, uci->cpu_sig.sig) != + UCODE_OK) { + ucode_ptr += mc_size; + continue; + } + + _save_mc(mc_saved_tmp, ucode_ptr, &mc_saved_count); + + ucode_ptr += mc_size; + } + + if (leftover) { + state = UCODE_ERROR; + goto out; + } + + if (mc_saved_count == 0) { + state = UCODE_NFOUND; + goto out; + } + + for (i = 0; i < mc_saved_count; i++) + mc_saved_in_initrd[i] = (unsigned long)mc_saved_tmp[i] - start; + + mc_saved_data->mc_saved_count = mc_saved_count; +out: + return state; +} + +static int collect_cpu_info_early(struct ucode_cpu_info *uci) +{ + unsigned int val[2]; + u8 x86, x86_model; + struct cpu_signature csig; + unsigned int eax, ebx, ecx, edx; + + csig.sig = 0; + csig.pf = 0; + csig.rev = 0; + + memset(uci, 0, sizeof(*uci)); + + eax = 0x00000001; + ecx = 0; + native_cpuid(&eax, &ebx, &ecx, &edx); + csig.sig = eax; + + x86 = get_x86_family(csig.sig); + x86_model = get_x86_model(csig.sig); + + if ((x86_model >= 5) || (x86 > 6)) { + /* get processor flags from MSR 0x17 */ + native_rdmsr(MSR_IA32_PLATFORM_ID, val[0], val[1]); + csig.pf = 1 << ((val[1] >> 18) & 7); + } + native_wrmsr(MSR_IA32_UCODE_REV, 0, 0); + + /* As documented in the SDM: Do a CPUID 1 here */ + sync_core(); + + /* get the current revision from MSR 0x8B */ + native_rdmsr(MSR_IA32_UCODE_REV, val[0], val[1]); + + csig.rev = val[1]; + + uci->cpu_sig = csig; + uci->valid = 1; + + return 0; +} + +#ifdef DEBUG +static void __ref show_saved_mc(void) +{ + int i, j; + unsigned int sig, pf, rev, total_size, data_size, date; + struct ucode_cpu_info uci; + + if (mc_saved_data.mc_saved_count == 0) { + pr_debug("no micorcode data saved.\n"); + return; + } + pr_debug("Total microcode saved: %d\n", mc_saved_data.mc_saved_count); + + collect_cpu_info_early(&uci); + + sig = uci.cpu_sig.sig; + pf = uci.cpu_sig.pf; + rev = uci.cpu_sig.rev; + pr_debug("CPU%d: sig=0x%x, pf=0x%x, rev=0x%x\n", + smp_processor_id(), sig, pf, rev); + + for (i = 0; i < mc_saved_data.mc_saved_count; i++) { + struct microcode_header_intel *mc_saved_header; + struct extended_sigtable *ext_header; + int ext_sigcount; + struct extended_signature *ext_sig; + + mc_saved_header = (struct microcode_header_intel *) + mc_saved_data.mc_saved[i]; + sig = mc_saved_header->sig; + pf = mc_saved_header->pf; + rev = mc_saved_header->rev; + total_size = get_totalsize(mc_saved_header); + data_size = get_datasize(mc_saved_header); + date = mc_saved_header->date; + + pr_debug("mc_saved[%d]: sig=0x%x, pf=0x%x, rev=0x%x, toal size=0x%x, date = %04x-%02x-%02x\n", + i, sig, pf, rev, total_size, + date & 0xffff, + date >> 24, + (date >> 16) & 0xff); + + /* Look for ext. headers: */ + if (total_size <= data_size + MC_HEADER_SIZE) + continue; + + ext_header = (struct extended_sigtable *) + mc_saved_header + data_size + MC_HEADER_SIZE; + ext_sigcount = ext_header->count; + ext_sig = (void *)ext_header + EXT_HEADER_SIZE; + + for (j = 0; j < ext_sigcount; j++) { + sig = ext_sig->sig; + pf = ext_sig->pf; + + pr_debug("\tExtended[%d]: sig=0x%x, pf=0x%x\n", + j, sig, pf); + + ext_sig++; + } + + } +} +#else +static inline void show_saved_mc(void) +{ +} +#endif + +#if defined(CONFIG_MICROCODE_INTEL_EARLY) && defined(CONFIG_HOTPLUG_CPU) +static DEFINE_MUTEX(x86_cpu_microcode_mutex); +/* + * Save this mc into mc_saved_data. So it will be loaded early when a CPU is + * hot added or resumes. + * + * Please make sure this mc should be a valid microcode patch before calling + * this function. + */ +int save_mc_for_early(u8 *mc) +{ + struct microcode_intel *mc_saved_tmp[MAX_UCODE_COUNT]; + unsigned int mc_saved_count_init; + unsigned int mc_saved_count; + struct microcode_intel **mc_saved; + int ret = 0; + int i; + + /* + * Hold hotplug lock so mc_saved_data is not accessed by a CPU in + * hotplug. + */ + mutex_lock(&x86_cpu_microcode_mutex); + + mc_saved_count_init = mc_saved_data.mc_saved_count; + mc_saved_count = mc_saved_data.mc_saved_count; + mc_saved = mc_saved_data.mc_saved; + + if (mc_saved && mc_saved_count) + memcpy(mc_saved_tmp, mc_saved, + mc_saved_count * sizeof(struct mirocode_intel *)); + /* + * Save the microcode patch mc in mc_save_tmp structure if it's a newer + * version. + */ + + _save_mc(mc_saved_tmp, mc, &mc_saved_count); + + /* + * Save the mc_save_tmp in global mc_saved_data. + */ + ret = save_microcode(&mc_saved_data, mc_saved_tmp, mc_saved_count); + if (ret) { + pr_err("Cannot save microcode patch.\n"); + goto out; + } + + show_saved_mc(); + + /* + * Free old saved microcod data. + */ + if (mc_saved) { + for (i = 0; i < mc_saved_count_init; i++) + kfree(mc_saved[i]); + kfree(mc_saved); + } + +out: + mutex_unlock(&x86_cpu_microcode_mutex); + + return ret; +} +EXPORT_SYMBOL_GPL(save_mc_for_early); +#endif + +static __initdata char ucode_name[] = "kernel/x86/microcode/GenuineIntel.bin"; +static __init enum ucode_state +scan_microcode(unsigned long start, unsigned long end, + struct mc_saved_data *mc_saved_data, + unsigned long *mc_saved_in_initrd, + struct ucode_cpu_info *uci) +{ + unsigned int size = end - start + 1; + struct cpio_data cd; + long offset = 0; +#ifdef CONFIG_X86_32 + char *p = (char *)__pa_nodebug(ucode_name); +#else + char *p = ucode_name; +#endif + + cd.data = NULL; + cd.size = 0; + + cd = find_cpio_data(p, (void *)start, size, &offset); + if (!cd.data) + return UCODE_ERROR; + + + return get_matching_model_microcode(0, start, cd.data, cd.size, + mc_saved_data, mc_saved_in_initrd, + uci); +} + +/* + * Print ucode update info. + */ +static void +print_ucode_info(struct ucode_cpu_info *uci, unsigned int date) +{ + int cpu = smp_processor_id(); + + pr_info("CPU%d microcode updated early to revision 0x%x, date = %04x-%02x-%02x\n", + cpu, + uci->cpu_sig.rev, + date & 0xffff, + date >> 24, + (date >> 16) & 0xff); +} + +#ifdef CONFIG_X86_32 + +static int delay_ucode_info; +static int current_mc_date; + +/* + * Print early updated ucode info after printk works. This is delayed info dump. + */ +void show_ucode_info_early(void) +{ + struct ucode_cpu_info uci; + + if (delay_ucode_info) { + collect_cpu_info_early(&uci); + print_ucode_info(&uci, current_mc_date); + delay_ucode_info = 0; + } +} + +/* + * At this point, we can not call printk() yet. Keep microcode patch number in + * mc_saved_data.mc_saved and delay printing microcode info in + * show_ucode_info_early() until printk() works. + */ +static void print_ucode(struct ucode_cpu_info *uci) +{ + struct microcode_intel *mc_intel; + int *delay_ucode_info_p; + int *current_mc_date_p; + + mc_intel = uci->mc; + if (mc_intel == NULL) + return; + + delay_ucode_info_p = (int *)__pa_nodebug(&delay_ucode_info); + current_mc_date_p = (int *)__pa_nodebug(¤t_mc_date); + + *delay_ucode_info_p = 1; + *current_mc_date_p = mc_intel->hdr.date; +} +#else + +/* + * Flush global tlb. We only do this in x86_64 where paging has been enabled + * already and PGE should be enabled as well. + */ +static inline void flush_tlb_early(void) +{ + __native_flush_tlb_global_irq_disabled(); +} + +static inline void print_ucode(struct ucode_cpu_info *uci) +{ + struct microcode_intel *mc_intel; + + mc_intel = uci->mc; + if (mc_intel == NULL) + return; + + print_ucode_info(uci, mc_intel->hdr.date); +} +#endif + +static int apply_microcode_early(struct mc_saved_data *mc_saved_data, + struct ucode_cpu_info *uci) +{ + struct microcode_intel *mc_intel; + unsigned int val[2]; + + mc_intel = uci->mc; + if (mc_intel == NULL) + return 0; + + /* write microcode via MSR 0x79 */ + native_wrmsr(MSR_IA32_UCODE_WRITE, + (unsigned long) mc_intel->bits, + (unsigned long) mc_intel->bits >> 16 >> 16); + native_wrmsr(MSR_IA32_UCODE_REV, 0, 0); + + /* As documented in the SDM: Do a CPUID 1 here */ + sync_core(); + + /* get the current revision from MSR 0x8B */ + native_rdmsr(MSR_IA32_UCODE_REV, val[0], val[1]); + if (val[1] != mc_intel->hdr.rev) + return -1; + +#ifdef CONFIG_X86_64 + /* Flush global tlb. This is precaution. */ + flush_tlb_early(); +#endif + uci->cpu_sig.rev = val[1]; + + print_ucode(uci); + + return 0; +} + +/* + * This function converts microcode patch offsets previously stored in + * mc_saved_in_initrd to pointers and stores the pointers in mc_saved_data. + */ +int __init save_microcode_in_initrd_intel(void) +{ + unsigned int count = mc_saved_data.mc_saved_count; + struct microcode_intel *mc_saved[MAX_UCODE_COUNT]; + int ret = 0; + + if (count == 0) + return ret; + + microcode_pointer(mc_saved, mc_saved_in_initrd, initrd_start, count); + ret = save_microcode(&mc_saved_data, mc_saved, count); + if (ret) + pr_err("Cannot save microcode patches from initrd.\n"); + + show_saved_mc(); + + return ret; +} + +static void __init +_load_ucode_intel_bsp(struct mc_saved_data *mc_saved_data, + unsigned long *mc_saved_in_initrd, + unsigned long initrd_start_early, + unsigned long initrd_end_early, + struct ucode_cpu_info *uci) +{ + collect_cpu_info_early(uci); + scan_microcode(initrd_start_early, initrd_end_early, mc_saved_data, + mc_saved_in_initrd, uci); + load_microcode(mc_saved_data, mc_saved_in_initrd, + initrd_start_early, uci); + apply_microcode_early(mc_saved_data, uci); +} + +void __init +load_ucode_intel_bsp(void) +{ + u64 ramdisk_image, ramdisk_size; + unsigned long initrd_start_early, initrd_end_early; + struct ucode_cpu_info uci; +#ifdef CONFIG_X86_32 + struct boot_params *boot_params_p; + + boot_params_p = (struct boot_params *)__pa_nodebug(&boot_params); + ramdisk_image = boot_params_p->hdr.ramdisk_image; + ramdisk_size = boot_params_p->hdr.ramdisk_size; + initrd_start_early = ramdisk_image; + initrd_end_early = initrd_start_early + ramdisk_size; + + _load_ucode_intel_bsp( + (struct mc_saved_data *)__pa_nodebug(&mc_saved_data), + (unsigned long *)__pa_nodebug(&mc_saved_in_initrd), + initrd_start_early, initrd_end_early, &uci); +#else + ramdisk_image = boot_params.hdr.ramdisk_image; + ramdisk_size = boot_params.hdr.ramdisk_size; + initrd_start_early = ramdisk_image + PAGE_OFFSET; + initrd_end_early = initrd_start_early + ramdisk_size; + + _load_ucode_intel_bsp(&mc_saved_data, mc_saved_in_initrd, + initrd_start_early, initrd_end_early, &uci); +#endif +} + +void load_ucode_intel_ap(void) +{ + struct mc_saved_data *mc_saved_data_p; + struct ucode_cpu_info uci; + unsigned long *mc_saved_in_initrd_p; + unsigned long initrd_start_addr; +#ifdef CONFIG_X86_32 + unsigned long *initrd_start_p; + + mc_saved_in_initrd_p = + (unsigned long *)__pa_nodebug(mc_saved_in_initrd); + mc_saved_data_p = (struct mc_saved_data *)__pa_nodebug(&mc_saved_data); + initrd_start_p = (unsigned long *)__pa_nodebug(&initrd_start); + initrd_start_addr = (unsigned long)__pa_nodebug(*initrd_start_p); +#else + mc_saved_data_p = &mc_saved_data; + mc_saved_in_initrd_p = mc_saved_in_initrd; + initrd_start_addr = initrd_start; +#endif + + /* + * If there is no valid ucode previously saved in memory, no need to + * update ucode on this AP. + */ + if (mc_saved_data_p->mc_saved_count == 0) + return; + + collect_cpu_info_early(&uci); + load_microcode(mc_saved_data_p, mc_saved_in_initrd_p, + initrd_start_addr, &uci); + apply_microcode_early(mc_saved_data_p, &uci); +} diff --git a/arch/x86/kernel/cpu/microcode/intel_lib.c b/arch/x86/kernel/cpu/microcode/intel_lib.c new file mode 100644 index 00000000000..ce69320d017 --- /dev/null +++ b/arch/x86/kernel/cpu/microcode/intel_lib.c @@ -0,0 +1,174 @@ +/* + * Intel CPU Microcode Update Driver for Linux + * + * Copyright (C) 2012 Fenghua Yu <fenghua.yu@intel.com> + * H Peter Anvin" <hpa@zytor.com> + * + * This driver allows to upgrade microcode on Intel processors + * belonging to IA-32 family - PentiumPro, Pentium II, + * Pentium III, Xeon, Pentium 4, etc. + * + * Reference: Section 8.11 of Volume 3a, IA-32 Intel? Architecture + * Software Developer's Manual + * Order Number 253668 or free download from: + * + * http://developer.intel.com/Assets/PDF/manual/253668.pdf + * + * For more information, go to http://www.urbanmyth.org/microcode + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + * + */ +#include <linux/firmware.h> +#include <linux/uaccess.h> +#include <linux/kernel.h> +#include <linux/module.h> + +#include <asm/microcode_intel.h> +#include <asm/processor.h> +#include <asm/msr.h> + +static inline int +update_match_cpu(unsigned int csig, unsigned int cpf, + unsigned int sig, unsigned int pf) +{ + return (!sigmatch(sig, csig, pf, cpf)) ? 0 : 1; +} + +int +update_match_revision(struct microcode_header_intel *mc_header, int rev) +{ + return (mc_header->rev <= rev) ? 0 : 1; +} + +int microcode_sanity_check(void *mc, int print_err) +{ + unsigned long total_size, data_size, ext_table_size; + struct microcode_header_intel *mc_header = mc; + struct extended_sigtable *ext_header = NULL; + int sum, orig_sum, ext_sigcount = 0, i; + struct extended_signature *ext_sig; + + total_size = get_totalsize(mc_header); + data_size = get_datasize(mc_header); + + if (data_size + MC_HEADER_SIZE > total_size) { + if (print_err) + pr_err("error! Bad data size in microcode data file\n"); + return -EINVAL; + } + + if (mc_header->ldrver != 1 || mc_header->hdrver != 1) { + if (print_err) + pr_err("error! Unknown microcode update format\n"); + return -EINVAL; + } + ext_table_size = total_size - (MC_HEADER_SIZE + data_size); + if (ext_table_size) { + if ((ext_table_size < EXT_HEADER_SIZE) + || ((ext_table_size - EXT_HEADER_SIZE) % EXT_SIGNATURE_SIZE)) { + if (print_err) + pr_err("error! Small exttable size in microcode data file\n"); + return -EINVAL; + } + ext_header = mc + MC_HEADER_SIZE + data_size; + if (ext_table_size != exttable_size(ext_header)) { + if (print_err) + pr_err("error! Bad exttable size in microcode data file\n"); + return -EFAULT; + } + ext_sigcount = ext_header->count; + } + + /* check extended table checksum */ + if (ext_table_size) { + int ext_table_sum = 0; + int *ext_tablep = (int *)ext_header; + + i = ext_table_size / DWSIZE; + while (i--) + ext_table_sum += ext_tablep[i]; + if (ext_table_sum) { + if (print_err) + pr_warn("aborting, bad extended signature table checksum\n"); + return -EINVAL; + } + } + + /* calculate the checksum */ + orig_sum = 0; + i = (MC_HEADER_SIZE + data_size) / DWSIZE; + while (i--) + orig_sum += ((int *)mc)[i]; + if (orig_sum) { + if (print_err) + pr_err("aborting, bad checksum\n"); + return -EINVAL; + } + if (!ext_table_size) + return 0; + /* check extended signature checksum */ + for (i = 0; i < ext_sigcount; i++) { + ext_sig = (void *)ext_header + EXT_HEADER_SIZE + + EXT_SIGNATURE_SIZE * i; + sum = orig_sum + - (mc_header->sig + mc_header->pf + mc_header->cksum) + + (ext_sig->sig + ext_sig->pf + ext_sig->cksum); + if (sum) { + if (print_err) + pr_err("aborting, bad checksum\n"); + return -EINVAL; + } + } + return 0; +} +EXPORT_SYMBOL_GPL(microcode_sanity_check); + +/* + * return 0 - no update found + * return 1 - found update + */ +int get_matching_sig(unsigned int csig, int cpf, void *mc, int rev) +{ + struct microcode_header_intel *mc_header = mc; + struct extended_sigtable *ext_header; + unsigned long total_size = get_totalsize(mc_header); + int ext_sigcount, i; + struct extended_signature *ext_sig; + + if (update_match_cpu(csig, cpf, mc_header->sig, mc_header->pf)) + return 1; + + /* Look for ext. headers: */ + if (total_size <= get_datasize(mc_header) + MC_HEADER_SIZE) + return 0; + + ext_header = mc + get_datasize(mc_header) + MC_HEADER_SIZE; + ext_sigcount = ext_header->count; + ext_sig = (void *)ext_header + EXT_HEADER_SIZE; + + for (i = 0; i < ext_sigcount; i++) { + if (update_match_cpu(csig, cpf, ext_sig->sig, ext_sig->pf)) + return 1; + ext_sig++; + } + return 0; +} + +/* + * return 0 - no update found + * return 1 - found update + */ +int get_matching_microcode(unsigned int csig, int cpf, void *mc, int rev) +{ + struct microcode_header_intel *mc_header = mc; + + if (!update_match_revision(mc_header, rev)) + return 0; + + return get_matching_sig(csig, cpf, mc, rev); +} +EXPORT_SYMBOL_GPL(get_matching_microcode); diff --git a/arch/x86/kernel/cpu/mkcapflags.pl b/arch/x86/kernel/cpu/mkcapflags.pl deleted file mode 100644 index dfea390e160..00000000000 --- a/arch/x86/kernel/cpu/mkcapflags.pl +++ /dev/null @@ -1,32 +0,0 @@ -#!/usr/bin/perl -# -# Generate the x86_cap_flags[] array from include/asm-x86/cpufeature.h -# - -($in, $out) = @ARGV; - -open(IN, "< $in\0") or die "$0: cannot open: $in: $!\n"; -open(OUT, "> $out\0") or die "$0: cannot create: $out: $!\n"; - -print OUT "#include <asm/cpufeature.h>\n\n"; -print OUT "const char * const x86_cap_flags[NCAPINTS*32] = {\n"; - -while (defined($line = <IN>)) { - if ($line =~ /^\s*\#\s*define\s+(X86_FEATURE_(\S+))\s+(.*)$/) { - $macro = $1; - $feature = $2; - $tail = $3; - if ($tail =~ /\/\*\s*\"([^"]*)\".*\*\//) { - $feature = $1; - } - - if ($feature ne '') { - printf OUT "\t%-32s = \"%s\",\n", - "[$macro]", "\L$feature"; - } - } -} -print OUT "};\n"; - -close(IN); -close(OUT); diff --git a/arch/x86/kernel/cpu/mkcapflags.sh b/arch/x86/kernel/cpu/mkcapflags.sh new file mode 100644 index 00000000000..2bf61650549 --- /dev/null +++ b/arch/x86/kernel/cpu/mkcapflags.sh @@ -0,0 +1,41 @@ +#!/bin/sh +# +# Generate the x86_cap_flags[] array from include/asm/cpufeature.h +# + +IN=$1 +OUT=$2 + +TABS="$(printf '\t\t\t\t\t')" +trap 'rm "$OUT"' EXIT + +( + echo "#ifndef _ASM_X86_CPUFEATURE_H" + echo "#include <asm/cpufeature.h>" + echo "#endif" + echo "" + echo "const char * const x86_cap_flags[NCAPINTS*32] = {" + + # Iterate through any input lines starting with #define X86_FEATURE_ + sed -n -e 's/\t/ /g' -e 's/^ *# *define *X86_FEATURE_//p' $IN | + while read i + do + # Name is everything up to the first whitespace + NAME="$(echo "$i" | sed 's/ .*//')" + + # If the /* comment */ starts with a quote string, grab that. + VALUE="$(echo "$i" | sed -n 's@.*/\* *\("[^"]*"\).*\*/@\1@p')" + [ -z "$VALUE" ] && VALUE="\"$NAME\"" + [ "$VALUE" == '""' ] && continue + + # Name is uppercase, VALUE is all lowercase + VALUE="$(echo "$VALUE" | tr A-Z a-z)" + + TABCOUNT=$(( ( 5*8 - 14 - $(echo "$NAME" | wc -c) ) / 8 )) + printf "\t[%s]%.*s = %s,\n" \ + "X86_FEATURE_$NAME" "$TABCOUNT" "$TABS" "$VALUE" + done + echo "};" +) > $OUT + +trap - EXIT diff --git a/arch/x86/kernel/cpu/mshyperv.c b/arch/x86/kernel/cpu/mshyperv.c index 0a630dd4b62..a450373e8e9 100644 --- a/arch/x86/kernel/cpu/mshyperv.c +++ b/arch/x86/kernel/cpu/mshyperv.c @@ -14,28 +14,80 @@ #include <linux/time.h> #include <linux/clocksource.h> #include <linux/module.h> +#include <linux/hardirq.h> +#include <linux/efi.h> +#include <linux/interrupt.h> +#include <linux/irq.h> #include <asm/processor.h> #include <asm/hypervisor.h> #include <asm/hyperv.h> #include <asm/mshyperv.h> +#include <asm/desc.h> +#include <asm/idle.h> +#include <asm/irq_regs.h> +#include <asm/i8259.h> +#include <asm/apic.h> +#include <asm/timer.h> struct ms_hyperv_info ms_hyperv; EXPORT_SYMBOL_GPL(ms_hyperv); -static bool __init ms_hyperv_platform(void) +#if IS_ENABLED(CONFIG_HYPERV) +static void (*vmbus_handler)(void); + +void hyperv_vector_handler(struct pt_regs *regs) +{ + struct pt_regs *old_regs = set_irq_regs(regs); + + irq_enter(); + exit_idle(); + + inc_irq_stat(irq_hv_callback_count); + if (vmbus_handler) + vmbus_handler(); + + irq_exit(); + set_irq_regs(old_regs); +} + +void hv_setup_vmbus_irq(void (*handler)(void)) +{ + vmbus_handler = handler; + /* + * Setup the IDT for hypervisor callback. Prevent reallocation + * at module reload. + */ + if (!test_bit(HYPERVISOR_CALLBACK_VECTOR, used_vectors)) + alloc_intr_gate(HYPERVISOR_CALLBACK_VECTOR, + hyperv_callback_vector); +} + +void hv_remove_vmbus_irq(void) +{ + /* We have no way to deallocate the interrupt gate */ + vmbus_handler = NULL; +} +EXPORT_SYMBOL_GPL(hv_setup_vmbus_irq); +EXPORT_SYMBOL_GPL(hv_remove_vmbus_irq); +#endif + +static uint32_t __init ms_hyperv_platform(void) { u32 eax; u32 hyp_signature[3]; if (!boot_cpu_has(X86_FEATURE_HYPERVISOR)) - return false; + return 0; cpuid(HYPERV_CPUID_VENDOR_AND_MAX_FUNCTIONS, &eax, &hyp_signature[0], &hyp_signature[1], &hyp_signature[2]); - return eax >= HYPERV_CPUID_MIN && - eax <= HYPERV_CPUID_MAX && - !memcmp("Microsoft Hv", hyp_signature, 12); + if (eax >= HYPERV_CPUID_MIN && + eax <= HYPERV_CPUID_MAX && + !memcmp("Microsoft Hv", hyp_signature, 12)) + return HYPERV_CPUID_VENDOR_AND_MAX_FUNCTIONS; + + return 0; } static cycle_t read_hv_clock(struct clocksource *arg) @@ -68,7 +120,28 @@ static void __init ms_hyperv_init_platform(void) printk(KERN_INFO "HyperV: features 0x%x, hints 0x%x\n", ms_hyperv.features, ms_hyperv.hints); - clocksource_register_hz(&hyperv_cs, NSEC_PER_SEC/100); +#ifdef CONFIG_X86_LOCAL_APIC + if (ms_hyperv.features & HV_X64_MSR_APIC_FREQUENCY_AVAILABLE) { + /* + * Get the APIC frequency. + */ + u64 hv_lapic_frequency; + + rdmsrl(HV_X64_MSR_APIC_FREQUENCY, hv_lapic_frequency); + hv_lapic_frequency = div_u64(hv_lapic_frequency, HZ); + lapic_timer_frequency = hv_lapic_frequency; + printk(KERN_INFO "HyperV: LAPIC Timer Frequency: %#x\n", + lapic_timer_frequency); + } +#endif + + if (ms_hyperv.features & HV_X64_MSR_TIME_REF_COUNT_AVAILABLE) + clocksource_register_hz(&hyperv_cs, NSEC_PER_SEC/100); + +#ifdef CONFIG_X86_IO_APIC + no_timer_check = 1; +#endif + } const __refconst struct hypervisor_x86 x86_hyper_ms_hyperv = { diff --git a/arch/x86/kernel/cpu/mtrr/cleanup.c b/arch/x86/kernel/cpu/mtrr/cleanup.c index ac140c7be39..5f90b85ff22 100644 --- a/arch/x86/kernel/cpu/mtrr/cleanup.c +++ b/arch/x86/kernel/cpu/mtrr/cleanup.c @@ -258,15 +258,15 @@ range_to_mtrr(unsigned int reg, unsigned long range_startk, /* Compute the maximum size with which we can make a range: */ if (range_startk) - max_align = ffs(range_startk) - 1; + max_align = __ffs(range_startk); else - max_align = 32; + max_align = BITS_PER_LONG - 1; - align = fls(range_sizek) - 1; + align = __fls(range_sizek); if (align > max_align) align = max_align; - sizek = 1 << align; + sizek = 1UL << align; if (debug_print) { char start_factor = 'K', size_factor = 'K'; unsigned long start_base, size_base; @@ -714,15 +714,15 @@ int __init mtrr_cleanup(unsigned address_bits) if (mtrr_tom2) x_remove_size = (mtrr_tom2 >> PAGE_SHIFT) - x_remove_base; - nr_range = x86_get_mtrr_mem_range(range, 0, x_remove_base, x_remove_size); /* * [0, 1M) should always be covered by var mtrr with WB * and fixed mtrrs should take effect before var mtrr for it: */ - nr_range = add_range_with_merge(range, RANGE_NUM, nr_range, 0, + nr_range = add_range_with_merge(range, RANGE_NUM, 0, 0, 1ULL<<(20 - PAGE_SHIFT)); - /* Sort the ranges: */ - sort_range(range, nr_range); + /* add from var mtrr at last */ + nr_range = x86_get_mtrr_mem_range(range, nr_range, + x_remove_base, x_remove_size); range_sums = sum_ranges(range, nr_range); printk(KERN_INFO "total RAM covered: %ldM\n", diff --git a/arch/x86/kernel/cpu/mtrr/cyrix.c b/arch/x86/kernel/cpu/mtrr/cyrix.c index 68a3343e579..9e451b0876b 100644 --- a/arch/x86/kernel/cpu/mtrr/cyrix.c +++ b/arch/x86/kernel/cpu/mtrr/cyrix.c @@ -167,7 +167,7 @@ static void post_set(void) setCx86(CX86_CCR3, ccr3); /* Enable caches */ - write_cr0(read_cr0() & 0xbfffffff); + write_cr0(read_cr0() & ~X86_CR0_CD); /* Restore value of CR4 */ if (cpu_has_pge) diff --git a/arch/x86/kernel/cpu/mtrr/generic.c b/arch/x86/kernel/cpu/mtrr/generic.c index 97b26356e9e..0e25a1bc5ab 100644 --- a/arch/x86/kernel/cpu/mtrr/generic.c +++ b/arch/x86/kernel/cpu/mtrr/generic.c @@ -12,7 +12,6 @@ #include <asm/processor-flags.h> #include <asm/cpufeature.h> #include <asm/tlbflush.h> -#include <asm/system.h> #include <asm/mtrr.h> #include <asm/msr.h> #include <asm/pat.h> @@ -362,11 +361,7 @@ static void __init print_mtrr_state(void) } pr_debug("MTRR variable ranges %sabled:\n", mtrr_state.enabled & 2 ? "en" : "dis"); - if (size_or_mask & 0xffffffffUL) - high_width = ffs(size_or_mask & 0xffffffffUL) - 1; - else - high_width = ffs(size_or_mask>>32) + 32 - 1; - high_width = (high_width - (32 - PAGE_SHIFT) + 3) / 4; + high_width = (__ffs64(size_or_mask) - (32 - PAGE_SHIFT) + 3) / 4; for (i = 0; i < num_var_ranges; ++i) { if (mtrr_state.var_ranges[i].mask_lo & (1 << 11)) @@ -515,8 +510,9 @@ generic_get_free_region(unsigned long base, unsigned long size, int replace_reg) static void generic_get_mtrr(unsigned int reg, unsigned long *base, unsigned long *size, mtrr_type *type) { - unsigned int mask_lo, mask_hi, base_lo, base_hi; - unsigned int tmp, hi; + u32 mask_lo, mask_hi, base_lo, base_hi; + unsigned int hi; + u64 tmp, mask; /* * get_mtrr doesn't need to update mtrr_state, also it could be called @@ -537,18 +533,18 @@ static void generic_get_mtrr(unsigned int reg, unsigned long *base, rdmsr(MTRRphysBase_MSR(reg), base_lo, base_hi); /* Work out the shifted address mask: */ - tmp = mask_hi << (32 - PAGE_SHIFT) | mask_lo >> PAGE_SHIFT; - mask_lo = size_or_mask | tmp; + tmp = (u64)mask_hi << (32 - PAGE_SHIFT) | mask_lo >> PAGE_SHIFT; + mask = size_or_mask | tmp; /* Expand tmp with high bits to all 1s: */ - hi = fls(tmp); + hi = fls64(tmp); if (hi > 0) { - tmp |= ~((1<<(hi - 1)) - 1); + tmp |= ~((1ULL<<(hi - 1)) - 1); - if (tmp != mask_lo) { + if (tmp != mask) { printk(KERN_WARNING "mtrr: your BIOS has configured an incorrect mask, fixing it.\n"); - add_taint(TAINT_FIRMWARE_WORKAROUND); - mask_lo = tmp; + add_taint(TAINT_FIRMWARE_WORKAROUND, LOCKDEP_STILL_OK); + mask = tmp; } } @@ -556,8 +552,8 @@ static void generic_get_mtrr(unsigned int reg, unsigned long *base, * This works correctly if size is a power of two, i.e. a * contiguous range: */ - *size = -mask_lo; - *base = base_hi << (32 - PAGE_SHIFT) | base_lo >> PAGE_SHIFT; + *size = -mask; + *base = (u64)base_hi << (32 - PAGE_SHIFT) | base_lo >> PAGE_SHIFT; *type = base_lo & 0xff; out_put_cpu: @@ -687,6 +683,7 @@ static void prepare_set(void) __acquires(set_atomicity_lock) } /* Flush all TLBs via a mov %cr3, %reg; mov %reg, %cr3 */ + count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ALL); __flush_tlb(); /* Save MTRR state */ @@ -700,13 +697,14 @@ static void prepare_set(void) __acquires(set_atomicity_lock) static void post_set(void) __releases(set_atomicity_lock) { /* Flush TLBs (no need to flush caches - they are disabled) */ + count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ALL); __flush_tlb(); /* Intel (P6) standard MTRRs */ mtrr_wrmsr(MSR_MTRRdefType, deftype_lo, deftype_hi); /* Enable caches */ - write_cr0(read_cr0() & 0xbfffffff); + write_cr0(read_cr0() & ~X86_CR0_CD); /* Restore value of CR4 */ if (cpu_has_pge) diff --git a/arch/x86/kernel/cpu/mtrr/if.c b/arch/x86/kernel/cpu/mtrr/if.c index 79289632cb2..a041e094b8b 100644 --- a/arch/x86/kernel/cpu/mtrr/if.c +++ b/arch/x86/kernel/cpu/mtrr/if.c @@ -167,6 +167,7 @@ mtrr_ioctl(struct file *file, unsigned int cmd, unsigned long __arg) { int err = 0; mtrr_type type; + unsigned long base; unsigned long size; struct mtrr_sentry sentry; struct mtrr_gentry gentry; @@ -267,14 +268,14 @@ mtrr_ioctl(struct file *file, unsigned int cmd, unsigned long __arg) #endif if (gentry.regnum >= num_var_ranges) return -EINVAL; - mtrr_if->get(gentry.regnum, &gentry.base, &size, &type); + mtrr_if->get(gentry.regnum, &base, &size, &type); /* Hide entries that go above 4GB */ - if (gentry.base + size - 1 >= (1UL << (8 * sizeof(gentry.size) - PAGE_SHIFT)) + if (base + size - 1 >= (1UL << (8 * sizeof(gentry.size) - PAGE_SHIFT)) || size >= (1UL << (8 * sizeof(gentry.size) - PAGE_SHIFT))) gentry.base = gentry.size = gentry.type = 0; else { - gentry.base <<= PAGE_SHIFT; + gentry.base = base << PAGE_SHIFT; gentry.size = size << PAGE_SHIFT; gentry.type = type; } @@ -321,11 +322,12 @@ mtrr_ioctl(struct file *file, unsigned int cmd, unsigned long __arg) #endif if (gentry.regnum >= num_var_ranges) return -EINVAL; - mtrr_if->get(gentry.regnum, &gentry.base, &size, &type); + mtrr_if->get(gentry.regnum, &base, &size, &type); /* Hide entries that would overflow */ if (size != (__typeof__(gentry.size))size) gentry.base = gentry.size = gentry.type = 0; else { + gentry.base = base; gentry.size = size; gentry.type = type; } diff --git a/arch/x86/kernel/cpu/mtrr/main.c b/arch/x86/kernel/cpu/mtrr/main.c index 6b96110bb0c..f961de9964c 100644 --- a/arch/x86/kernel/cpu/mtrr/main.c +++ b/arch/x86/kernel/cpu/mtrr/main.c @@ -51,9 +51,13 @@ #include <asm/e820.h> #include <asm/mtrr.h> #include <asm/msr.h> +#include <asm/pat.h> #include "mtrr.h" +/* arch_phys_wc_add returns an MTRR register index plus this offset. */ +#define MTRR_TO_PHYS_WC_OFFSET 1000 + u32 num_var_ranges; unsigned int mtrr_usage_table[MTRR_MAX_VAR_RANGES]; @@ -305,7 +309,8 @@ int mtrr_add_page(unsigned long base, unsigned long size, return -EINVAL; } - if (base & size_or_mask || size & size_or_mask) { + if ((base | (base + size - 1)) >> + (boot_cpu_data.x86_phys_bits - PAGE_SHIFT)) { pr_warning("mtrr: base or size exceeds the MTRR width\n"); return -EINVAL; } @@ -524,6 +529,73 @@ int mtrr_del(int reg, unsigned long base, unsigned long size) } EXPORT_SYMBOL(mtrr_del); +/** + * arch_phys_wc_add - add a WC MTRR and handle errors if PAT is unavailable + * @base: Physical base address + * @size: Size of region + * + * If PAT is available, this does nothing. If PAT is unavailable, it + * attempts to add a WC MTRR covering size bytes starting at base and + * logs an error if this fails. + * + * Drivers must store the return value to pass to mtrr_del_wc_if_needed, + * but drivers should not try to interpret that return value. + */ +int arch_phys_wc_add(unsigned long base, unsigned long size) +{ + int ret; + + if (pat_enabled) + return 0; /* Success! (We don't need to do anything.) */ + + ret = mtrr_add(base, size, MTRR_TYPE_WRCOMB, true); + if (ret < 0) { + pr_warn("Failed to add WC MTRR for [%p-%p]; performance may suffer.", + (void *)base, (void *)(base + size - 1)); + return ret; + } + return ret + MTRR_TO_PHYS_WC_OFFSET; +} +EXPORT_SYMBOL(arch_phys_wc_add); + +/* + * arch_phys_wc_del - undoes arch_phys_wc_add + * @handle: Return value from arch_phys_wc_add + * + * This cleans up after mtrr_add_wc_if_needed. + * + * The API guarantees that mtrr_del_wc_if_needed(error code) and + * mtrr_del_wc_if_needed(0) do nothing. + */ +void arch_phys_wc_del(int handle) +{ + if (handle >= 1) { + WARN_ON(handle < MTRR_TO_PHYS_WC_OFFSET); + mtrr_del(handle - MTRR_TO_PHYS_WC_OFFSET, 0, 0); + } +} +EXPORT_SYMBOL(arch_phys_wc_del); + +/* + * phys_wc_to_mtrr_index - translates arch_phys_wc_add's return value + * @handle: Return value from arch_phys_wc_add + * + * This will turn the return value from arch_phys_wc_add into an mtrr + * index suitable for debugging. + * + * Note: There is no legitimate use for this function, except possibly + * in printk line. Alas there is an illegitimate use in some ancient + * drm ioctls. + */ +int phys_wc_to_mtrr_index(int handle) +{ + if (handle < MTRR_TO_PHYS_WC_OFFSET) + return -1; + else + return handle - MTRR_TO_PHYS_WC_OFFSET; +} +EXPORT_SYMBOL_GPL(phys_wc_to_mtrr_index); + /* * HACK ALERT! * These should be called implicitly, but we can't yet until all the initcall @@ -583,6 +655,7 @@ static struct syscore_ops mtrr_syscore_ops = { int __initdata changed_by_mtrr_cleanup; +#define SIZE_OR_MASK_BITS(n) (~((1ULL << ((n) - PAGE_SHIFT)) - 1)) /** * mtrr_bp_init - initialize mtrrs on the boot CPU * @@ -600,13 +673,13 @@ void __init mtrr_bp_init(void) if (cpu_has_mtrr) { mtrr_if = &generic_mtrr_ops; - size_or_mask = 0xff000000; /* 36 bits */ + size_or_mask = SIZE_OR_MASK_BITS(36); size_and_mask = 0x00f00000; phys_addr = 36; /* * This is an AMD specific MSR, but we assume(hope?) that - * Intel will implement it to when they extend the address + * Intel will implement it too when they extend the address * bus of the Xeon. */ if (cpuid_eax(0x80000000) >= 0x80000008) { @@ -619,7 +692,7 @@ void __init mtrr_bp_init(void) boot_cpu_data.x86_mask == 0x4)) phys_addr = 36; - size_or_mask = ~((1ULL << (phys_addr - PAGE_SHIFT)) - 1); + size_or_mask = SIZE_OR_MASK_BITS(phys_addr); size_and_mask = ~size_or_mask & 0xfffff00000ULL; } else if (boot_cpu_data.x86_vendor == X86_VENDOR_CENTAUR && boot_cpu_data.x86 == 6) { @@ -627,7 +700,7 @@ void __init mtrr_bp_init(void) * VIA C* family have Intel style MTRRs, * but don't support PAE */ - size_or_mask = 0xfff00000; /* 32 bits */ + size_or_mask = SIZE_OR_MASK_BITS(32); size_and_mask = 0; phys_addr = 32; } @@ -637,21 +710,21 @@ void __init mtrr_bp_init(void) if (cpu_has_k6_mtrr) { /* Pre-Athlon (K6) AMD CPU MTRRs */ mtrr_if = mtrr_ops[X86_VENDOR_AMD]; - size_or_mask = 0xfff00000; /* 32 bits */ + size_or_mask = SIZE_OR_MASK_BITS(32); size_and_mask = 0; } break; case X86_VENDOR_CENTAUR: if (cpu_has_centaur_mcr) { mtrr_if = mtrr_ops[X86_VENDOR_CENTAUR]; - size_or_mask = 0xfff00000; /* 32 bits */ + size_or_mask = SIZE_OR_MASK_BITS(32); size_and_mask = 0; } break; case X86_VENDOR_CYRIX: if (cpu_has_cyrix_arr) { mtrr_if = mtrr_ops[X86_VENDOR_CYRIX]; - size_or_mask = 0xfff00000; /* 32 bits */ + size_or_mask = SIZE_OR_MASK_BITS(32); size_and_mask = 0; } break; @@ -695,11 +768,16 @@ void mtrr_ap_init(void) } /** - * Save current fixed-range MTRR state of the BSP + * Save current fixed-range MTRR state of the first cpu in cpu_online_mask. */ void mtrr_save_state(void) { - smp_call_function_single(0, mtrr_save_fixed_ranges, NULL, 1); + int first_cpu; + + get_online_cpus(); + first_cpu = cpumask_first(cpu_online_mask); + smp_call_function_single(first_cpu, mtrr_save_fixed_ranges, NULL, 1); + put_online_cpus(); } void set_mtrr_aps_delayed_init(void) diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 5adce1040b1..2879ecdaac4 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -24,27 +24,19 @@ #include <linux/slab.h> #include <linux/cpu.h> #include <linux/bitops.h> +#include <linux/device.h> #include <asm/apic.h> #include <asm/stacktrace.h> #include <asm/nmi.h> -#include <asm/compat.h> #include <asm/smp.h> #include <asm/alternative.h> +#include <asm/timer.h> +#include <asm/desc.h> +#include <asm/ldt.h> #include "perf_event.h" -#if 0 -#undef wrmsrl -#define wrmsrl(msr, val) \ -do { \ - trace_printk("wrmsrl(%lx, %lx)\n", (unsigned long)(msr),\ - (unsigned long)(val)); \ - native_write_msr((msr), (u32)((u64)(val)), \ - (u32)((u64)(val) >> 32)); \ -} while (0) -#endif - struct x86_pmu x86_pmu __read_mostly; DEFINE_PER_CPU(struct cpu_hw_events, cpu_hw_events) = { @@ -73,7 +65,7 @@ u64 x86_perf_event_update(struct perf_event *event) int idx = hwc->idx; s64 delta; - if (idx == X86_PMC_IDX_FIXED_BTS) + if (idx == INTEL_PMC_IDX_FIXED_BTS) return 0; /* @@ -85,7 +77,7 @@ u64 x86_perf_event_update(struct perf_event *event) */ again: prev_raw_count = local64_read(&hwc->prev_count); - rdmsrl(hwc->event_base, new_raw_count); + rdpmcl(hwc->event_base_rdpmc, new_raw_count); if (local64_cmpxchg(&hwc->prev_count, prev_raw_count, new_raw_count) != prev_raw_count) @@ -126,6 +118,9 @@ static int x86_pmu_extra_regs(u64 config, struct perf_event *event) continue; if (event->attr.config1 & ~er->valid_mask) return -EINVAL; + /* Check if the extra msrs can be safely accessed*/ + if (!er->extra_msr_access) + return -ENXIO; reg->idx = er->idx; reg->config = event->attr.config1; @@ -188,8 +183,9 @@ static void release_pmc_hardware(void) {} static bool check_hw_exists(void) { - u64 val, val_new = 0; - int i, reg, ret = 0; + u64 val, val_fail, val_new= ~0; + int i, reg, reg_fail, ret = 0; + int bios_fail = 0; /* * Check to see if the BIOS enabled any of the counters, if so @@ -200,8 +196,11 @@ static bool check_hw_exists(void) ret = rdmsrl_safe(reg, &val); if (ret) goto msr_fail; - if (val & ARCH_PERFMON_EVENTSEL_ENABLE) - goto bios_fail; + if (val & ARCH_PERFMON_EVENTSEL_ENABLE) { + bios_fail = 1; + val_fail = val; + reg_fail = reg; + } } if (x86_pmu.num_counters_fixed) { @@ -210,35 +209,41 @@ static bool check_hw_exists(void) if (ret) goto msr_fail; for (i = 0; i < x86_pmu.num_counters_fixed; i++) { - if (val & (0x03 << i*4)) - goto bios_fail; + if (val & (0x03 << i*4)) { + bios_fail = 1; + val_fail = val; + reg_fail = reg; + } } } /* - * Now write a value and read it back to see if it matches, - * this is needed to detect certain hardware emulators (qemu/kvm) - * that don't trap on the MSR access and always return 0s. + * Read the current value, change it and read it back to see if it + * matches, this is needed to detect certain hardware emulators + * (qemu/kvm) that don't trap on the MSR access and always return 0s. */ - val = 0xabcdUL; - ret = checking_wrmsrl(x86_pmu_event_addr(0), val); - ret |= rdmsrl_safe(x86_pmu_event_addr(0), &val_new); + reg = x86_pmu_event_addr(0); + if (rdmsrl_safe(reg, &val)) + goto msr_fail; + val ^= 0xffffUL; + ret = wrmsrl_safe(reg, val); + ret |= rdmsrl_safe(reg, &val_new); if (ret || val != val_new) goto msr_fail; - return true; - -bios_fail: /* * We still allow the PMU driver to operate: */ - printk(KERN_CONT "Broken BIOS detected, complain to your hardware vendor.\n"); - printk(KERN_ERR FW_BUG "the BIOS has corrupted hw-PMU resources (MSR %x is %Lx)\n", reg, val); + if (bios_fail) { + printk(KERN_CONT "Broken BIOS detected, complain to your hardware vendor.\n"); + printk(KERN_ERR FW_BUG "the BIOS has corrupted hw-PMU resources (MSR %x is %Lx)\n", reg_fail, val_fail); + } return true; msr_fail: printk(KERN_CONT "Broken PMU hardware detected, using software events only.\n"); + printk(KERN_ERR "Failed to access perfctr msr (MSR %x is %Lx)\n", reg, val_new); return false; } @@ -301,15 +306,6 @@ int x86_setup_perfctr(struct perf_event *event) hwc->sample_period = x86_pmu.max_period; hwc->last_period = hwc->sample_period; local64_set(&hwc->period_left, hwc->sample_period); - } else { - /* - * If we have a PMU initialized but no APIC - * interrupts, we cannot sample hardware - * events (user-space has to fall back and - * sample via a hrtimer based software event): - */ - if (!x86_pmu.apic) - return -EOPNOTSUPP; } if (attr->type == PERF_TYPE_RAW) @@ -351,13 +347,43 @@ int x86_setup_perfctr(struct perf_event *event) return 0; } +/* + * check that branch_sample_type is compatible with + * settings needed for precise_ip > 1 which implies + * using the LBR to capture ALL taken branches at the + * priv levels of the measurement + */ +static inline int precise_br_compat(struct perf_event *event) +{ + u64 m = event->attr.branch_sample_type; + u64 b = 0; + + /* must capture all branches */ + if (!(m & PERF_SAMPLE_BRANCH_ANY)) + return 0; + + m &= PERF_SAMPLE_BRANCH_KERNEL | PERF_SAMPLE_BRANCH_USER; + + if (!event->attr.exclude_user) + b |= PERF_SAMPLE_BRANCH_USER; + + if (!event->attr.exclude_kernel) + b |= PERF_SAMPLE_BRANCH_KERNEL; + + /* + * ignore PERF_SAMPLE_BRANCH_HV, not supported on x86 + */ + + return m == b; +} + int x86_pmu_hw_config(struct perf_event *event) { if (event->attr.precise_ip) { int precise = 0; /* Support for constant skid */ - if (x86_pmu.pebs_active) { + if (x86_pmu.pebs_active && !x86_pmu.pebs_broken) { precise++; /* Support for IP fixup */ @@ -367,6 +393,37 @@ int x86_pmu_hw_config(struct perf_event *event) if (event->attr.precise_ip > precise) return -EOPNOTSUPP; + /* + * check that PEBS LBR correction does not conflict with + * whatever the user is asking with attr->branch_sample_type + */ + if (event->attr.precise_ip > 1 && + x86_pmu.intel_cap.pebs_format < 2) { + u64 *br_type = &event->attr.branch_sample_type; + + if (has_branch_stack(event)) { + if (!precise_br_compat(event)) + return -EOPNOTSUPP; + + /* branch_sample_type is compatible */ + + } else { + /* + * user did not specify branch_sample_type + * + * For PEBS fixups, we capture all + * the branches at the priv level of the + * event. + */ + *br_type = PERF_SAMPLE_BRANCH_ANY; + + if (!event->attr.exclude_user) + *br_type |= PERF_SAMPLE_BRANCH_USER; + + if (!event->attr.exclude_kernel) + *br_type |= PERF_SAMPLE_BRANCH_KERNEL; + } + } } /* @@ -423,6 +480,7 @@ static int __x86_pmu_event_init(struct perf_event *event) /* mark unused */ event->hw.extra_reg.idx = EXTRA_REG_NONE; + event->hw.branch_reg.idx = EXTRA_REG_NONE; return x86_pmu.hw_config(event); } @@ -505,7 +563,7 @@ struct sched_state { struct perf_sched { int max_weight; int max_events; - struct event_constraint **constraints; + struct perf_event **events; struct sched_state state; int saved_states; struct sched_state saved[SCHED_STATES_MAX]; @@ -514,7 +572,7 @@ struct perf_sched { /* * Initialize interator that runs through all events and counters. */ -static void perf_sched_init(struct perf_sched *sched, struct event_constraint **c, +static void perf_sched_init(struct perf_sched *sched, struct perf_event **events, int num, int wmin, int wmax) { int idx; @@ -522,10 +580,10 @@ static void perf_sched_init(struct perf_sched *sched, struct event_constraint ** memset(sched, 0, sizeof(*sched)); sched->max_events = num; sched->max_weight = wmax; - sched->constraints = c; + sched->events = events; for (idx = 0; idx < num; idx++) { - if (c[idx]->weight == wmin) + if (events[idx]->hw.constraint->weight == wmin) break; } @@ -572,19 +630,18 @@ static bool __perf_sched_find_counter(struct perf_sched *sched) if (sched->state.event >= sched->max_events) return false; - c = sched->constraints[sched->state.event]; - + c = sched->events[sched->state.event]->hw.constraint; /* Prefer fixed purpose counters */ - if (x86_pmu.num_counters_fixed) { - idx = X86_PMC_IDX_FIXED; - for_each_set_bit_cont(idx, c->idxmsk, X86_PMC_IDX_MAX) { + if (c->idxmsk64 & (~0ULL << INTEL_PMC_IDX_FIXED)) { + idx = INTEL_PMC_IDX_FIXED; + for_each_set_bit_from(idx, c->idxmsk, X86_PMC_IDX_MAX) { if (!__test_and_set_bit(idx, sched->state.used)) goto done; } } /* Grab the first unused counter starting with idx */ idx = sched->state.counter; - for_each_set_bit_cont(idx, c->idxmsk, X86_PMC_IDX_FIXED) { + for_each_set_bit_from(idx, c->idxmsk, INTEL_PMC_IDX_FIXED) { if (!__test_and_set_bit(idx, sched->state.used)) goto done; } @@ -631,7 +688,7 @@ static bool perf_sched_next_event(struct perf_sched *sched) if (sched->state.weight > sched->max_weight) return false; } - c = sched->constraints[sched->state.event]; + c = sched->events[sched->state.event]->hw.constraint; } while (c->weight != sched->state.weight); sched->state.counter = 0; /* start with first counter */ @@ -642,12 +699,12 @@ static bool perf_sched_next_event(struct perf_sched *sched) /* * Assign a counter for each event. */ -static int perf_assign_events(struct event_constraint **constraints, int n, - int wmin, int wmax, int *assign) +int perf_assign_events(struct perf_event **events, int n, + int wmin, int wmax, int *assign) { struct perf_sched sched; - perf_sched_init(&sched, constraints, n, wmin, wmax); + perf_sched_init(&sched, events, n, wmin, wmax); do { if (!perf_sched_find_counter(&sched)) @@ -658,19 +715,23 @@ static int perf_assign_events(struct event_constraint **constraints, int n, return sched.state.unassigned; } +EXPORT_SYMBOL_GPL(perf_assign_events); int x86_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign) { - struct event_constraint *c, *constraints[X86_PMC_IDX_MAX]; + struct event_constraint *c; unsigned long used_mask[BITS_TO_LONGS(X86_PMC_IDX_MAX)]; + struct perf_event *e; int i, wmin, wmax, num = 0; struct hw_perf_event *hwc; bitmap_zero(used_mask, X86_PMC_IDX_MAX); for (i = 0, wmin = X86_PMC_IDX_MAX, wmax = 0; i < n; i++) { + hwc = &cpuc->event_list[i]->hw; c = x86_pmu.get_event_constraints(cpuc, cpuc->event_list[i]); - constraints[i] = c; + hwc->constraint = c; + wmin = min(wmin, c->weight); wmax = max(wmax, c->weight); } @@ -680,7 +741,7 @@ int x86_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign) */ for (i = 0; i < n; i++) { hwc = &cpuc->event_list[i]->hw; - c = constraints[i]; + c = hwc->constraint; /* never assigned */ if (hwc->idx == -1) @@ -701,16 +762,35 @@ int x86_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign) /* slow path */ if (i != n) - num = perf_assign_events(constraints, n, wmin, wmax, assign); + num = perf_assign_events(cpuc->event_list, n, wmin, + wmax, assign); /* + * Mark the event as committed, so we do not put_constraint() + * in case new events are added and fail scheduling. + */ + if (!num && assign) { + for (i = 0; i < n; i++) { + e = cpuc->event_list[i]; + e->hw.flags |= PERF_X86_EVENT_COMMITTED; + } + } + /* * scheduling failed or is just a simulation, * free resources if necessary */ if (!assign || num) { for (i = 0; i < n; i++) { + e = cpuc->event_list[i]; + /* + * do not put_constraint() on comitted events, + * because they are good to go + */ + if ((e->hw.flags & PERF_X86_EVENT_COMMITTED)) + continue; + if (x86_pmu.put_event_constraints) - x86_pmu.put_event_constraints(cpuc, cpuc->event_list[i]); + x86_pmu.put_event_constraints(cpuc, e); } } return num ? -EINVAL : 0; @@ -762,15 +842,17 @@ static inline void x86_assign_hw_event(struct perf_event *event, hwc->last_cpu = smp_processor_id(); hwc->last_tag = ++cpuc->tags[i]; - if (hwc->idx == X86_PMC_IDX_FIXED_BTS) { + if (hwc->idx == INTEL_PMC_IDX_FIXED_BTS) { hwc->config_base = 0; hwc->event_base = 0; - } else if (hwc->idx >= X86_PMC_IDX_FIXED) { + } else if (hwc->idx >= INTEL_PMC_IDX_FIXED) { hwc->config_base = MSR_ARCH_PERFMON_FIXED_CTR_CTRL; - hwc->event_base = MSR_ARCH_PERFMON_FIXED_CTR0 + (hwc->idx - X86_PMC_IDX_FIXED); + hwc->event_base = MSR_ARCH_PERFMON_FIXED_CTR0 + (hwc->idx - INTEL_PMC_IDX_FIXED); + hwc->event_base_rdpmc = (hwc->idx - INTEL_PMC_IDX_FIXED) | 1<<30; } else { hwc->config_base = x86_pmu_config_addr(hwc->idx); hwc->event_base = x86_pmu_event_addr(hwc->idx); + hwc->event_base_rdpmc = x86_pmu_rdpmc_index(hwc->idx); } } @@ -805,7 +887,6 @@ static void x86_pmu_enable(struct pmu *pmu) * hw_perf_group_sched_in() or x86_pmu_enable() * * step1: save events moving to new counters - * step2: reprogram moved events into new counters */ for (i = 0; i < n_running; i++) { event = cpuc->event_list[i]; @@ -831,6 +912,9 @@ static void x86_pmu_enable(struct pmu *pmu) x86_pmu_stop(event, PERF_EF_UPDATE); } + /* + * step2: reprogram moved events into new counters + */ for (i = 0; i < cpuc->n_events; i++) { event = cpuc->event_list[i]; hwc = &event->hw; @@ -868,7 +952,7 @@ int x86_perf_event_set_period(struct perf_event *event) s64 period = hwc->sample_period; int ret = 0, idx = hwc->idx; - if (idx == X86_PMC_IDX_FIXED_BTS) + if (idx == INTEL_PMC_IDX_FIXED_BTS) return 0; /* @@ -956,7 +1040,7 @@ static int x86_pmu_add(struct perf_event *event, int flags) /* * If group events scheduling transaction was started, * skip the schedulability test here, it will be performed - * at commit time (->commit_txn) as a whole + * at commit time (->commit_txn) as a whole. */ if (cpuc->group_flag & PERF_EVENT_TXN) goto done_collect; @@ -971,6 +1055,10 @@ static int x86_pmu_add(struct perf_event *event, int flags) memcpy(cpuc->assign, assign, n*sizeof(int)); done_collect: + /* + * Commit the collect_events() state. See x86_pmu_del() and + * x86_pmu_*_txn(). + */ cpuc->n_events = n; cpuc->n_added += n - n0; cpuc->n_txn += n - n0; @@ -1088,28 +1176,46 @@ static void x86_pmu_del(struct perf_event *event, int flags) int i; /* + * event is descheduled + */ + event->hw.flags &= ~PERF_X86_EVENT_COMMITTED; + + /* * If we're called during a txn, we don't need to do anything. * The events never got scheduled and ->cancel_txn will truncate * the event_list. + * + * XXX assumes any ->del() called during a TXN will only be on + * an event added during that same TXN. */ if (cpuc->group_flag & PERF_EVENT_TXN) return; + /* + * Not a TXN, therefore cleanup properly. + */ x86_pmu_stop(event, PERF_EF_UPDATE); for (i = 0; i < cpuc->n_events; i++) { - if (event == cpuc->event_list[i]) { + if (event == cpuc->event_list[i]) + break; + } - if (x86_pmu.put_event_constraints) - x86_pmu.put_event_constraints(cpuc, event); + if (WARN_ON_ONCE(i == cpuc->n_events)) /* called ->del() without ->add() ? */ + return; - while (++i < cpuc->n_events) - cpuc->event_list[i-1] = cpuc->event_list[i]; + /* If we have a newly added event; make sure to decrease n_added. */ + if (i >= cpuc->n_events - cpuc->n_added) + --cpuc->n_added; + + if (x86_pmu.put_event_constraints) + x86_pmu.put_event_constraints(cpuc, event); + + /* Delete the array entry. */ + while (++i < cpuc->n_events) + cpuc->event_list[i-1] = cpuc->event_list[i]; + --cpuc->n_events; - --cpuc->n_events; - break; - } - } perf_event_update_userpage(event); } @@ -1121,8 +1227,6 @@ int x86_pmu_handle_irq(struct pt_regs *regs) int idx, handled = 0; u64 val; - perf_sample_data_init(&data, 0); - cpuc = &__get_cpu_var(cpu_hw_events); /* @@ -1157,7 +1261,7 @@ int x86_pmu_handle_irq(struct pt_regs *regs) * event overflow */ handled++; - data.period = event->hw.last_period; + perf_sample_data_init(&data, 0, event->hw.last_period); if (!x86_perf_event_set_period(event)) continue; @@ -1183,19 +1287,30 @@ void perf_events_lapic_init(void) apic_write(APIC_LVTPC, APIC_DM_NMI); } -static int __kprobes +static int perf_event_nmi_handler(unsigned int cmd, struct pt_regs *regs) { + u64 start_clock; + u64 finish_clock; + int ret; + if (!atomic_read(&active_events)) return NMI_DONE; - return x86_pmu.handle_irq(regs); + start_clock = sched_clock(); + ret = x86_pmu.handle_irq(regs); + finish_clock = sched_clock(); + + perf_sample_event_took(finish_clock - start_clock); + + return ret; } +NOKPROBE_SYMBOL(perf_event_nmi_handler); struct event_constraint emptyconstraint; struct event_constraint unconstrained; -static int __cpuinit +static int x86_pmu_notifier(struct notifier_block *self, unsigned long action, void *hcpu) { unsigned int cpu = (long)hcpu; @@ -1210,6 +1325,8 @@ x86_pmu_notifier(struct notifier_block *self, unsigned long action, void *hcpu) break; case CPU_STARTING: + if (x86_pmu.attr_rdpmc) + set_in_cr4(X86_CR4_PCE); if (x86_pmu.cpu_starting) x86_pmu.cpu_starting(cpu); break; @@ -1244,12 +1361,163 @@ static void __init pmu_check_apic(void) x86_pmu.apic = 0; pr_info("no APIC, boot with the \"lapic\" boot parameter to force-enable it.\n"); pr_info("no hardware sampling interrupt available.\n"); + + /* + * If we have a PMU initialized but no APIC + * interrupts, we cannot sample hardware + * events (user-space has to fall back and + * sample via a hrtimer based software event): + */ + pmu.capabilities |= PERF_PMU_CAP_NO_INTERRUPT; + +} + +static struct attribute_group x86_pmu_format_group = { + .name = "format", + .attrs = NULL, +}; + +/* + * Remove all undefined events (x86_pmu.event_map(id) == 0) + * out of events_attr attributes. + */ +static void __init filter_events(struct attribute **attrs) +{ + struct device_attribute *d; + struct perf_pmu_events_attr *pmu_attr; + int i, j; + + for (i = 0; attrs[i]; i++) { + d = (struct device_attribute *)attrs[i]; + pmu_attr = container_of(d, struct perf_pmu_events_attr, attr); + /* str trumps id */ + if (pmu_attr->event_str) + continue; + if (x86_pmu.event_map(i)) + continue; + + for (j = i; attrs[j]; j++) + attrs[j] = attrs[j + 1]; + + /* Check the shifted attr. */ + i--; + } +} + +/* Merge two pointer arrays */ +static __init struct attribute **merge_attr(struct attribute **a, struct attribute **b) +{ + struct attribute **new; + int j, i; + + for (j = 0; a[j]; j++) + ; + for (i = 0; b[i]; i++) + j++; + j++; + + new = kmalloc(sizeof(struct attribute *) * j, GFP_KERNEL); + if (!new) + return NULL; + + j = 0; + for (i = 0; a[i]; i++) + new[j++] = a[i]; + for (i = 0; b[i]; i++) + new[j++] = b[i]; + new[j] = NULL; + + return new; +} + +ssize_t events_sysfs_show(struct device *dev, struct device_attribute *attr, + char *page) +{ + struct perf_pmu_events_attr *pmu_attr = \ + container_of(attr, struct perf_pmu_events_attr, attr); + u64 config = x86_pmu.event_map(pmu_attr->id); + + /* string trumps id */ + if (pmu_attr->event_str) + return sprintf(page, "%s", pmu_attr->event_str); + + return x86_pmu.events_sysfs_show(page, config); +} + +EVENT_ATTR(cpu-cycles, CPU_CYCLES ); +EVENT_ATTR(instructions, INSTRUCTIONS ); +EVENT_ATTR(cache-references, CACHE_REFERENCES ); +EVENT_ATTR(cache-misses, CACHE_MISSES ); +EVENT_ATTR(branch-instructions, BRANCH_INSTRUCTIONS ); +EVENT_ATTR(branch-misses, BRANCH_MISSES ); +EVENT_ATTR(bus-cycles, BUS_CYCLES ); +EVENT_ATTR(stalled-cycles-frontend, STALLED_CYCLES_FRONTEND ); +EVENT_ATTR(stalled-cycles-backend, STALLED_CYCLES_BACKEND ); +EVENT_ATTR(ref-cycles, REF_CPU_CYCLES ); + +static struct attribute *empty_attrs; + +static struct attribute *events_attr[] = { + EVENT_PTR(CPU_CYCLES), + EVENT_PTR(INSTRUCTIONS), + EVENT_PTR(CACHE_REFERENCES), + EVENT_PTR(CACHE_MISSES), + EVENT_PTR(BRANCH_INSTRUCTIONS), + EVENT_PTR(BRANCH_MISSES), + EVENT_PTR(BUS_CYCLES), + EVENT_PTR(STALLED_CYCLES_FRONTEND), + EVENT_PTR(STALLED_CYCLES_BACKEND), + EVENT_PTR(REF_CPU_CYCLES), + NULL, +}; + +static struct attribute_group x86_pmu_events_group = { + .name = "events", + .attrs = events_attr, +}; + +ssize_t x86_event_sysfs_show(char *page, u64 config, u64 event) +{ + u64 umask = (config & ARCH_PERFMON_EVENTSEL_UMASK) >> 8; + u64 cmask = (config & ARCH_PERFMON_EVENTSEL_CMASK) >> 24; + bool edge = (config & ARCH_PERFMON_EVENTSEL_EDGE); + bool pc = (config & ARCH_PERFMON_EVENTSEL_PIN_CONTROL); + bool any = (config & ARCH_PERFMON_EVENTSEL_ANY); + bool inv = (config & ARCH_PERFMON_EVENTSEL_INV); + ssize_t ret; + + /* + * We have whole page size to spend and just little data + * to write, so we can safely use sprintf. + */ + ret = sprintf(page, "event=0x%02llx", event); + + if (umask) + ret += sprintf(page + ret, ",umask=0x%02llx", umask); + + if (edge) + ret += sprintf(page + ret, ",edge"); + + if (pc) + ret += sprintf(page + ret, ",pc"); + + if (any) + ret += sprintf(page + ret, ",any"); + + if (inv) + ret += sprintf(page + ret, ",inv"); + + if (cmask) + ret += sprintf(page + ret, ",cmask=0x%02llx", cmask); + + ret += sprintf(page + ret, "\n"); + + return ret; } static int __init init_hw_perf_events(void) { struct x86_pmu_quirk *quirk; - struct event_constraint *c; int err; pr_info("Performance Events: "); @@ -1262,7 +1530,7 @@ static int __init init_hw_perf_events(void) err = amd_pmu_init(); break; default: - return 0; + err = -ENOTSUPP; } if (err != 0) { pr_cont("no PMU driver, software events only.\n"); @@ -1277,46 +1545,37 @@ static int __init init_hw_perf_events(void) pr_cont("%s PMU driver.\n", x86_pmu.name); + x86_pmu.attr_rdpmc = 1; /* enable userspace RDPMC usage by default */ + for (quirk = x86_pmu.quirks; quirk; quirk = quirk->next) quirk->func(); - if (x86_pmu.num_counters > X86_PMC_MAX_GENERIC) { - WARN(1, KERN_ERR "hw perf events %d > max(%d), clipping!", - x86_pmu.num_counters, X86_PMC_MAX_GENERIC); - x86_pmu.num_counters = X86_PMC_MAX_GENERIC; - } - x86_pmu.intel_ctrl = (1 << x86_pmu.num_counters) - 1; - - if (x86_pmu.num_counters_fixed > X86_PMC_MAX_FIXED) { - WARN(1, KERN_ERR "hw perf events fixed %d > max(%d), clipping!", - x86_pmu.num_counters_fixed, X86_PMC_MAX_FIXED); - x86_pmu.num_counters_fixed = X86_PMC_MAX_FIXED; - } - - x86_pmu.intel_ctrl |= - ((1LL << x86_pmu.num_counters_fixed)-1) << X86_PMC_IDX_FIXED; + if (!x86_pmu.intel_ctrl) + x86_pmu.intel_ctrl = (1 << x86_pmu.num_counters) - 1; perf_events_lapic_init(); register_nmi_handler(NMI_LOCAL, perf_event_nmi_handler, 0, "PMI"); unconstrained = (struct event_constraint) __EVENT_CONSTRAINT(0, (1ULL << x86_pmu.num_counters) - 1, - 0, x86_pmu.num_counters, 0); + 0, x86_pmu.num_counters, 0, 0); - if (x86_pmu.event_constraints) { - /* - * event on fixed counter2 (REF_CYCLES) only works on this - * counter, so do not extend mask to generic counters - */ - for_each_event_constraint(c, x86_pmu.event_constraints) { - if (c->cmask != X86_RAW_EVENT_MASK - || c->idxmsk64 == X86_PMC_MSK_FIXED_REF_CYCLES) { - continue; - } + x86_pmu_format_group.attrs = x86_pmu.format_attrs; - c->idxmsk64 |= (1ULL << x86_pmu.num_counters) - 1; - c->weight += x86_pmu.num_counters; - } + if (x86_pmu.event_attrs) + x86_pmu_events_group.attrs = x86_pmu.event_attrs; + + if (!x86_pmu.events_sysfs_show) + x86_pmu_events_group.attrs = &empty_attrs; + else + filter_events(x86_pmu_events_group.attrs); + + if (x86_pmu.cpu_events) { + struct attribute **tmp; + + tmp = merge_attr(x86_pmu_events_group.attrs, x86_pmu.cpu_events); + if (!WARN_ON(!tmp)) + x86_pmu_events_group.attrs = tmp; } pr_info("... version: %d\n", x86_pmu.version); @@ -1360,7 +1619,8 @@ static void x86_pmu_cancel_txn(struct pmu *pmu) { __this_cpu_and(cpu_hw_events.group_flag, ~PERF_EVENT_TXN); /* - * Truncate the collected events. + * Truncate collected array by the number of events added in this + * transaction. See x86_pmu_add() and x86_pmu_*_txn(). */ __this_cpu_sub(cpu_hw_events.n_added, __this_cpu_read(cpu_hw_events.n_txn)); __this_cpu_sub(cpu_hw_events.n_events, __this_cpu_read(cpu_hw_events.n_txn)); @@ -1371,6 +1631,8 @@ static void x86_pmu_cancel_txn(struct pmu *pmu) * Commit group events scheduling transaction * Perform the group schedulability test as a whole * Return 0 if success + * + * Does not cancel the transaction on failure; expects the caller to do this. */ static int x86_pmu_commit_txn(struct pmu *pmu) { @@ -1426,6 +1688,7 @@ static struct cpu_hw_events *allocate_fake_cpuc(void) if (!cpuc->shared_regs) goto error; } + cpuc->is_fake = 1; return cpuc; error: free_fake_cpuc(cpuc); @@ -1542,23 +1805,138 @@ static int x86_pmu_event_init(struct perf_event *event) return err; } +static int x86_pmu_event_idx(struct perf_event *event) +{ + int idx = event->hw.idx; + + if (!x86_pmu.attr_rdpmc) + return 0; + + if (x86_pmu.num_counters_fixed && idx >= INTEL_PMC_IDX_FIXED) { + idx -= INTEL_PMC_IDX_FIXED; + idx |= 1 << 30; + } + + return idx + 1; +} + +static ssize_t get_attr_rdpmc(struct device *cdev, + struct device_attribute *attr, + char *buf) +{ + return snprintf(buf, 40, "%d\n", x86_pmu.attr_rdpmc); +} + +static void change_rdpmc(void *info) +{ + bool enable = !!(unsigned long)info; + + if (enable) + set_in_cr4(X86_CR4_PCE); + else + clear_in_cr4(X86_CR4_PCE); +} + +static ssize_t set_attr_rdpmc(struct device *cdev, + struct device_attribute *attr, + const char *buf, size_t count) +{ + unsigned long val; + ssize_t ret; + + ret = kstrtoul(buf, 0, &val); + if (ret) + return ret; + + if (x86_pmu.attr_rdpmc_broken) + return -ENOTSUPP; + + if (!!val != !!x86_pmu.attr_rdpmc) { + x86_pmu.attr_rdpmc = !!val; + on_each_cpu(change_rdpmc, (void *)val, 1); + } + + return count; +} + +static DEVICE_ATTR(rdpmc, S_IRUSR | S_IWUSR, get_attr_rdpmc, set_attr_rdpmc); + +static struct attribute *x86_pmu_attrs[] = { + &dev_attr_rdpmc.attr, + NULL, +}; + +static struct attribute_group x86_pmu_attr_group = { + .attrs = x86_pmu_attrs, +}; + +static const struct attribute_group *x86_pmu_attr_groups[] = { + &x86_pmu_attr_group, + &x86_pmu_format_group, + &x86_pmu_events_group, + NULL, +}; + +static void x86_pmu_flush_branch_stack(void) +{ + if (x86_pmu.flush_branch_stack) + x86_pmu.flush_branch_stack(); +} + +void perf_check_microcode(void) +{ + if (x86_pmu.check_microcode) + x86_pmu.check_microcode(); +} +EXPORT_SYMBOL_GPL(perf_check_microcode); + static struct pmu pmu = { - .pmu_enable = x86_pmu_enable, - .pmu_disable = x86_pmu_disable, + .pmu_enable = x86_pmu_enable, + .pmu_disable = x86_pmu_disable, + + .attr_groups = x86_pmu_attr_groups, + + .event_init = x86_pmu_event_init, - .event_init = x86_pmu_event_init, + .add = x86_pmu_add, + .del = x86_pmu_del, + .start = x86_pmu_start, + .stop = x86_pmu_stop, + .read = x86_pmu_read, - .add = x86_pmu_add, - .del = x86_pmu_del, - .start = x86_pmu_start, - .stop = x86_pmu_stop, - .read = x86_pmu_read, + .start_txn = x86_pmu_start_txn, + .cancel_txn = x86_pmu_cancel_txn, + .commit_txn = x86_pmu_commit_txn, - .start_txn = x86_pmu_start_txn, - .cancel_txn = x86_pmu_cancel_txn, - .commit_txn = x86_pmu_commit_txn, + .event_idx = x86_pmu_event_idx, + .flush_branch_stack = x86_pmu_flush_branch_stack, }; +void arch_perf_update_userpage(struct perf_event_mmap_page *userpg, u64 now) +{ + struct cyc2ns_data *data; + + userpg->cap_user_time = 0; + userpg->cap_user_time_zero = 0; + userpg->cap_user_rdpmc = x86_pmu.attr_rdpmc; + userpg->pmc_width = x86_pmu.cntval_bits; + + if (!sched_clock_stable()) + return; + + data = cyc2ns_read_begin(); + + userpg->cap_user_time = 1; + userpg->time_mult = data->cyc2ns_mul; + userpg->time_shift = data->cyc2ns_shift; + userpg->time_offset = data->cyc2ns_offset - now; + + userpg->cap_user_time_zero = 1; + userpg->time_zero = data->cyc2ns_offset; + + cyc2ns_read_end(data); +} + /* * callchain support */ @@ -1594,32 +1972,68 @@ perf_callchain_kernel(struct perf_callchain_entry *entry, struct pt_regs *regs) dump_trace(NULL, regs, NULL, 0, &backtrace_ops, entry); } +static inline int +valid_user_frame(const void __user *fp, unsigned long size) +{ + return (__range_not_ok(fp, size, TASK_SIZE) == 0); +} + +static unsigned long get_segment_base(unsigned int segment) +{ + struct desc_struct *desc; + int idx = segment >> 3; + + if ((segment & SEGMENT_TI_MASK) == SEGMENT_LDT) { + if (idx > LDT_ENTRIES) + return 0; + + if (idx > current->active_mm->context.size) + return 0; + + desc = current->active_mm->context.ldt; + } else { + if (idx > GDT_ENTRIES) + return 0; + + desc = __this_cpu_ptr(&gdt_page.gdt[0]); + } + + return get_desc_base(desc + idx); +} + #ifdef CONFIG_COMPAT + +#include <asm/compat.h> + static inline int perf_callchain_user32(struct pt_regs *regs, struct perf_callchain_entry *entry) { /* 32-bit process in 64-bit kernel. */ + unsigned long ss_base, cs_base; struct stack_frame_ia32 frame; const void __user *fp; if (!test_thread_flag(TIF_IA32)) return 0; - fp = compat_ptr(regs->bp); + cs_base = get_segment_base(regs->cs); + ss_base = get_segment_base(regs->ss); + + fp = compat_ptr(ss_base + regs->bp); while (entry->nr < PERF_MAX_STACK_DEPTH) { unsigned long bytes; frame.next_frame = 0; frame.return_address = 0; bytes = copy_from_user_nmi(&frame, fp, sizeof(frame)); - if (bytes != sizeof(frame)) + if (bytes != 0) break; - if (fp < compat_ptr(regs->sp)) + if (!valid_user_frame(fp, sizeof(frame))) break; - perf_callchain_store(entry, frame.return_address); - fp = compat_ptr(frame.next_frame); + perf_callchain_store(entry, cs_base + frame.return_address); + fp = compat_ptr(ss_base + frame.next_frame); } return 1; } @@ -1642,6 +2056,12 @@ perf_callchain_user(struct perf_callchain_entry *entry, struct pt_regs *regs) return; } + /* + * We don't know what to do with VM86 stacks.. ignore them for now. + */ + if (regs->flags & (X86_VM_MASK | PERF_EFLAGS_VM)) + return; + fp = (void __user *)regs->bp; perf_callchain_store(entry, regs->ip); @@ -1658,10 +2078,10 @@ perf_callchain_user(struct perf_callchain_entry *entry, struct pt_regs *regs) frame.return_address = 0; bytes = copy_from_user_nmi(&frame, fp, sizeof(frame)); - if (bytes != sizeof(frame)) + if (bytes != 0) break; - if ((unsigned long)fp < regs->sp) + if (!valid_user_frame(fp, sizeof(frame))) break; perf_callchain_store(entry, frame.return_address); @@ -1669,16 +2089,50 @@ perf_callchain_user(struct perf_callchain_entry *entry, struct pt_regs *regs) } } -unsigned long perf_instruction_pointer(struct pt_regs *regs) +/* + * Deal with code segment offsets for the various execution modes: + * + * VM86 - the good olde 16 bit days, where the linear address is + * 20 bits and we use regs->ip + 0x10 * regs->cs. + * + * IA32 - Where we need to look at GDT/LDT segment descriptor tables + * to figure out what the 32bit base address is. + * + * X32 - has TIF_X32 set, but is running in x86_64 + * + * X86_64 - CS,DS,SS,ES are all zero based. + */ +static unsigned long code_segment_base(struct pt_regs *regs) { - unsigned long ip; + /* + * If we are in VM86 mode, add the segment offset to convert to a + * linear address. + */ + if (regs->flags & X86_VM_MASK) + return 0x10 * regs->cs; + + /* + * For IA32 we look at the GDT/LDT segment base to convert the + * effective IP to a linear address. + */ +#ifdef CONFIG_X86_32 + if (user_mode(regs) && regs->cs != __USER_CS) + return get_segment_base(regs->cs); +#else + if (test_thread_flag(TIF_IA32)) { + if (user_mode(regs) && regs->cs != __USER32_CS) + return get_segment_base(regs->cs); + } +#endif + return 0; +} +unsigned long perf_instruction_pointer(struct pt_regs *regs) +{ if (perf_guest_cbs && perf_guest_cbs->is_in_guest()) - ip = perf_guest_cbs->get_guest_ip(); - else - ip = instruction_pointer(regs); + return perf_guest_cbs->get_guest_ip(); - return ip; + return regs->ip + code_segment_base(regs); } unsigned long perf_misc_flags(struct pt_regs *regs) diff --git a/arch/x86/kernel/cpu/perf_event.h b/arch/x86/kernel/cpu/perf_event.h index 8944062f46e..8ade93111e0 100644 --- a/arch/x86/kernel/cpu/perf_event.h +++ b/arch/x86/kernel/cpu/perf_event.h @@ -14,6 +14,18 @@ #include <linux/perf_event.h> +#if 0 +#undef wrmsrl +#define wrmsrl(msr, val) \ +do { \ + unsigned int _msr = (msr); \ + u64 _val = (val); \ + trace_printk("wrmsrl(%x, %Lx)\n", (unsigned int)(_msr), \ + (unsigned long long)(_val)); \ + native_write_msr((_msr), (u32)(_val), (u32)(_val >> 32)); \ +} while (0) +#endif + /* * | NHM/WSM | SNB | * register ------------------------------- @@ -33,6 +45,8 @@ enum extra_reg_type { EXTRA_REG_RSP_0 = 0, /* offcore_response_0 */ EXTRA_REG_RSP_1 = 1, /* offcore_response_1 */ + EXTRA_REG_LBR = 2, /* lbr_select */ + EXTRA_REG_LDLAT = 3, /* ld_lat_threshold */ EXTRA_REG_MAX /* number of entries needed */ }; @@ -46,7 +60,15 @@ struct event_constraint { u64 cmask; int weight; int overlap; + int flags; }; +/* + * struct hw_perf_event.flags flags + */ +#define PERF_X86_EVENT_PEBS_LDLAT 0x1 /* ld+ldlat data address sampling */ +#define PERF_X86_EVENT_PEBS_ST 0x2 /* st data address sampling */ +#define PERF_X86_EVENT_PEBS_ST_HSW 0x4 /* haswell style st data sampling */ +#define PERF_X86_EVENT_COMMITTED 0x8 /* event passed commit_txn */ struct amd_nb { int nb_id; /* NorthBridge id */ @@ -56,7 +78,7 @@ struct amd_nb { }; /* The maximal number of PEBS events: */ -#define MAX_PEBS_EVENTS 4 +#define MAX_PEBS_EVENTS 8 /* * A debug store configuration. @@ -108,14 +130,17 @@ struct cpu_hw_events { unsigned long running[BITS_TO_LONGS(X86_PMC_IDX_MAX)]; int enabled; - int n_events; - int n_added; - int n_txn; + int n_events; /* the # of events in the below arrays */ + int n_added; /* the # last events in the below arrays; + they've never been enabled yet */ + int n_txn; /* the # last events in the below arrays; + added in the current transaction */ int assign[X86_PMC_IDX_MAX]; /* event to counter assignment */ u64 tags[X86_PMC_IDX_MAX]; struct perf_event *event_list[X86_PMC_IDX_MAX]; /* in enabled order */ unsigned int group_flag; + int is_fake; /* * Intel DebugStore bits @@ -130,6 +155,8 @@ struct cpu_hw_events { void *lbr_context; struct perf_branch_stack lbr_stack; struct perf_branch_entry lbr_entries[MAX_LBR_ENTRIES]; + struct er_account *lbr_sel; + u64 br_sel; /* * Intel host/guest exclude bits @@ -139,6 +166,11 @@ struct cpu_hw_events { struct perf_guest_switch_msr guest_switch_msrs[X86_PMC_IDX_MAX]; /* + * Intel checkpoint mask + */ + u64 intel_cp_status; + + /* * manage shared (per-core, per-cpu) registers * used on Intel NHM/WSM/SNB */ @@ -147,21 +179,24 @@ struct cpu_hw_events { /* * AMD specific bits */ - struct amd_nb *amd_nb; + struct amd_nb *amd_nb; + /* Inverted mask of bits to clear in the perf_ctr ctrl registers */ + u64 perf_ctr_virt_mask; void *kfree_on_online; }; -#define __EVENT_CONSTRAINT(c, n, m, w, o) {\ +#define __EVENT_CONSTRAINT(c, n, m, w, o, f) {\ { .idxmsk64 = (n) }, \ .code = (c), \ .cmask = (m), \ .weight = (w), \ .overlap = (o), \ + .flags = f, \ } #define EVENT_CONSTRAINT(c, n, m) \ - __EVENT_CONSTRAINT(c, n, m, HWEIGHT(n), 0) + __EVENT_CONSTRAINT(c, n, m, HWEIGHT(n), 0, 0) /* * The overlap flag marks event constraints with overlapping counter @@ -185,7 +220,7 @@ struct cpu_hw_events { * and its counter masks must be kept at a minimum. */ #define EVENT_CONSTRAINT_OVERLAP(c, n, m) \ - __EVENT_CONSTRAINT(c, n, m, HWEIGHT(n), 1) + __EVENT_CONSTRAINT(c, n, m, HWEIGHT(n), 1, 0) /* * Constraint on the Event code. @@ -201,11 +236,14 @@ struct cpu_hw_events { * - inv * - edge * - cnt-mask + * - in_tx + * - in_tx_checkpointed * The other filters are supported by fixed counters. * The any-thread option is supported starting with v3. */ +#define FIXED_EVENT_FLAGS (X86_RAW_EVENT_MASK|HSW_IN_TX|HSW_IN_TX_CHECKPOINTED) #define FIXED_EVENT_CONSTRAINT(c, n) \ - EVENT_CONSTRAINT(c, (1ULL << (32+n)), X86_RAW_EVENT_MASK) + EVENT_CONSTRAINT(c, (1ULL << (32+n)), FIXED_EVENT_FLAGS) /* * Constraint on the Event code + UMask @@ -213,11 +251,33 @@ struct cpu_hw_events { #define INTEL_UEVENT_CONSTRAINT(c, n) \ EVENT_CONSTRAINT(c, n, INTEL_ARCH_EVENT_MASK) -#define EVENT_CONSTRAINT_END \ - EVENT_CONSTRAINT(0, 0, 0) +#define INTEL_PLD_CONSTRAINT(c, n) \ + __EVENT_CONSTRAINT(c, n, INTEL_ARCH_EVENT_MASK, \ + HWEIGHT(n), 0, PERF_X86_EVENT_PEBS_LDLAT) +#define INTEL_PST_CONSTRAINT(c, n) \ + __EVENT_CONSTRAINT(c, n, INTEL_ARCH_EVENT_MASK, \ + HWEIGHT(n), 0, PERF_X86_EVENT_PEBS_ST) + +/* DataLA version of store sampling without extra enable bit. */ +#define INTEL_PST_HSW_CONSTRAINT(c, n) \ + __EVENT_CONSTRAINT(c, n, INTEL_ARCH_EVENT_MASK, \ + HWEIGHT(n), 0, PERF_X86_EVENT_PEBS_ST_HSW) + +/* + * We define the end marker as having a weight of -1 + * to enable blacklisting of events using a counter bitmask + * of zero and thus a weight of zero. + * The end marker has a weight that cannot possibly be + * obtained from counting the bits in the bitmask. + */ +#define EVENT_CONSTRAINT_END { .weight = -1 } + +/* + * Check for end marker with weight == -1 + */ #define for_each_event_constraint(e, c) \ - for ((e) = (c); (e)->weight; (e)++) + for ((e) = (c); (e)->weight != -1; (e)++) /* * Extra registers for specific events. @@ -235,19 +295,31 @@ struct extra_reg { u64 config_mask; u64 valid_mask; int idx; /* per_xxx->regs[] reg index */ + bool extra_msr_access; }; #define EVENT_EXTRA_REG(e, ms, m, vm, i) { \ - .event = (e), \ - .msr = (ms), \ - .config_mask = (m), \ - .valid_mask = (vm), \ - .idx = EXTRA_REG_##i \ + .event = (e), \ + .msr = (ms), \ + .config_mask = (m), \ + .valid_mask = (vm), \ + .idx = EXTRA_REG_##i, \ + .extra_msr_access = true, \ } #define INTEL_EVENT_EXTRA_REG(event, msr, vm, idx) \ EVENT_EXTRA_REG(event, msr, ARCH_PERFMON_EVENTSEL_EVENT, vm, idx) +#define INTEL_UEVENT_EXTRA_REG(event, msr, vm, idx) \ + EVENT_EXTRA_REG(event, msr, ARCH_PERFMON_EVENTSEL_EVENT | \ + ARCH_PERFMON_EVENTSEL_UMASK, vm, idx) + +#define INTEL_UEVENT_PEBS_LDLAT_EXTRA_REG(c) \ + INTEL_UEVENT_EXTRA_REG(c, \ + MSR_PEBS_LD_LAT_THRESHOLD, \ + 0xffff, \ + LDLAT) + #define EVENT_EXTRA_END EVENT_EXTRA_REG(0, 0, 0, 0, RSP_0) union perf_capabilities { @@ -257,6 +329,11 @@ union perf_capabilities { u64 pebs_arch_reg:1; u64 pebs_format:4; u64 smm_freeze:1; + /* + * PMU supports separate counter range for writing + * values > 32bit. + */ + u64 full_width_write:1; }; u64 capabilities; }; @@ -266,6 +343,29 @@ struct x86_pmu_quirk { void (*func)(void); }; +union x86_pmu_config { + struct { + u64 event:8, + umask:8, + usr:1, + os:1, + edge:1, + pc:1, + interrupt:1, + __reserved1:1, + en:1, + inv:1, + cmask:8, + event2:4, + __reserved2:4, + go:1, + ho:1; + } bits; + u64 value; +}; + +#define X86_CONFIG(args...) ((union x86_pmu_config){.bits = {args}}).value + /* * struct x86_pmu - generic x86 pmu */ @@ -284,6 +384,8 @@ struct x86_pmu { int (*schedule_events)(struct cpu_hw_events *cpuc, int n, int *assign); unsigned eventsel; unsigned perfctr; + int (*addr_offset)(int index, bool eventsel); + int (*rdpmc_index)(int index); u64 (*event_map)(int); int max_events; int num_counters; @@ -306,12 +408,30 @@ struct x86_pmu { struct event_constraint *event_constraints; struct x86_pmu_quirk *quirks; int perfctr_second_write; + bool late_ack; + + /* + * sysfs attrs + */ + int attr_rdpmc_broken; + int attr_rdpmc; + struct attribute **format_attrs; + struct attribute **event_attrs; + ssize_t (*events_sysfs_show)(char *page, u64 config); + struct attribute **cpu_events; + + /* + * CPU Hotplug hooks + */ int (*cpu_prepare)(int cpu); void (*cpu_starting)(int cpu); void (*cpu_dying)(int cpu); void (*cpu_dead)(int cpu); + void (*check_microcode)(void); + void (*flush_branch_stack)(void); + /* * Intel Arch Perfmon v2+ */ @@ -321,17 +441,25 @@ struct x86_pmu { /* * Intel DebugStore bits */ - int bts, pebs; - int bts_active, pebs_active; + unsigned int bts :1, + bts_active :1, + pebs :1, + pebs_active :1, + pebs_broken :1; int pebs_record_size; void (*drain_pebs)(struct pt_regs *regs); struct event_constraint *pebs_constraints; + void (*pebs_aliases)(struct perf_event *event); + int max_pebs_events; /* * Intel LBR */ unsigned long lbr_tos, lbr_from, lbr_to; /* MSR base regs */ int lbr_nr; /* hardware stack size */ + u64 lbr_sel_mask; /* LBR_SELECT valid bits */ + const int *lbr_sel_map; /* lbr_select mappings */ + bool lbr_double_abort; /* duplicated lbr aborts */ /* * Extra registers for events @@ -357,6 +485,23 @@ do { \ #define ERF_NO_HT_SHARING 1 #define ERF_HAS_RSP_1 2 +#define EVENT_VAR(_id) event_attr_##_id +#define EVENT_PTR(_id) &event_attr_##_id.attr.attr + +#define EVENT_ATTR(_name, _id) \ +static struct perf_pmu_events_attr EVENT_VAR(_id) = { \ + .attr = __ATTR(_name, 0444, events_sysfs_show, NULL), \ + .id = PERF_COUNT_HW_##_id, \ + .event_str = NULL, \ +}; + +#define EVENT_ATTR_STR(_name, v, str) \ +static struct perf_pmu_events_attr event_attr_##v = { \ + .attr = __ATTR(_name, 0444, events_sysfs_show, NULL), \ + .id = 0, \ + .event_str = str, \ +}; + extern struct x86_pmu x86_pmu __read_mostly; DECLARE_PER_CPU(struct cpu_hw_events, cpu_hw_events); @@ -384,28 +529,21 @@ extern u64 __read_mostly hw_cache_extra_regs u64 x86_perf_event_update(struct perf_event *event); -static inline int x86_pmu_addr_offset(int index) +static inline unsigned int x86_pmu_config_addr(int index) { - int offset; - - /* offset = X86_FEATURE_PERFCTR_CORE ? index << 1 : index */ - alternative_io(ASM_NOP2, - "shll $1, %%eax", - X86_FEATURE_PERFCTR_CORE, - "=a" (offset), - "a" (index)); - - return offset; + return x86_pmu.eventsel + (x86_pmu.addr_offset ? + x86_pmu.addr_offset(index, true) : index); } -static inline unsigned int x86_pmu_config_addr(int index) +static inline unsigned int x86_pmu_event_addr(int index) { - return x86_pmu.eventsel + x86_pmu_addr_offset(index); + return x86_pmu.perfctr + (x86_pmu.addr_offset ? + x86_pmu.addr_offset(index, false) : index); } -static inline unsigned int x86_pmu_event_addr(int index) +static inline int x86_pmu_rdpmc_index(int index) { - return x86_pmu.perfctr + x86_pmu_addr_offset(index); + return x86_pmu.rdpmc_index ? x86_pmu.rdpmc_index(index) : index; } int x86_setup_perfctr(struct perf_event *event); @@ -417,13 +555,17 @@ void x86_pmu_disable_all(void); static inline void __x86_pmu_enable_event(struct hw_perf_event *hwc, u64 enable_mask) { + u64 disable_mask = __this_cpu_read(cpu_hw_events.perf_ctr_virt_mask); + if (hwc->extra_reg.reg) wrmsrl(hwc->extra_reg.reg, hwc->extra_reg.config); - wrmsrl(hwc->config_base, hwc->config | enable_mask); + wrmsrl(hwc->config_base, (hwc->config | enable_mask) & ~disable_mask); } void x86_pmu_enable_all(int added); +int perf_assign_events(struct perf_event **events, int n, + int wmin, int wmax, int *assign); int x86_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign); void x86_pmu_stop(struct perf_event *event, int flags); @@ -443,6 +585,38 @@ extern struct event_constraint emptyconstraint; extern struct event_constraint unconstrained; +static inline bool kernel_ip(unsigned long ip) +{ +#ifdef CONFIG_X86_32 + return ip > PAGE_OFFSET; +#else + return (long)ip < 0; +#endif +} + +/* + * Not all PMUs provide the right context information to place the reported IP + * into full context. Specifically segment registers are typically not + * supplied. + * + * Assuming the address is a linear address (it is for IBS), we fake the CS and + * vm86 mode using the known zero-based code segment and 'fix up' the registers + * to reflect this. + * + * Intel PEBS/LBR appear to typically provide the effective address, nothing + * much we can do about that but pray and treat it like a linear address. + */ +static inline void set_linear_ip(struct pt_regs *regs, unsigned long ip) +{ + regs->cs = kernel_ip(ip) ? __KERNEL_CS : __USER_CS; + if (regs->flags & X86_VM_MASK) + regs->flags ^= (PERF_EFLAGS_VM | X86_VM_MASK); + regs->ip = ip; +} + +ssize_t x86_event_sysfs_show(char *page, u64 config, u64 event); +ssize_t intel_event_sysfs_show(char *page, u64 config); + #ifdef CONFIG_CPU_SUP_AMD int amd_pmu_init(void); @@ -487,12 +661,18 @@ extern struct event_constraint intel_core2_pebs_event_constraints[]; extern struct event_constraint intel_atom_pebs_event_constraints[]; +extern struct event_constraint intel_slm_pebs_event_constraints[]; + extern struct event_constraint intel_nehalem_pebs_event_constraints[]; extern struct event_constraint intel_westmere_pebs_event_constraints[]; extern struct event_constraint intel_snb_pebs_event_constraints[]; +extern struct event_constraint intel_ivb_pebs_event_constraints[]; + +extern struct event_constraint intel_hsw_pebs_event_constraints[]; + struct event_constraint *intel_pebs_constraints(struct perf_event *event); void intel_pmu_pebs_enable(struct perf_event *event); @@ -523,10 +703,19 @@ void intel_pmu_lbr_init_nhm(void); void intel_pmu_lbr_init_atom(void); +void intel_pmu_lbr_init_snb(void); + +int intel_pmu_setup_lbr_filter(struct perf_event *event); + int p4_pmu_init(void); int p6_pmu_init(void); +int knc_pmu_init(void); + +ssize_t events_sysfs_show(struct device *dev, struct device_attribute *attr, + char *page); + #else /* CONFIG_CPU_SUP_INTEL */ static inline void reserve_ds_buffers(void) diff --git a/arch/x86/kernel/cpu/perf_event_amd.c b/arch/x86/kernel/cpu/perf_event_amd.c index 0397b23be8e..beeb7cc0704 100644 --- a/arch/x86/kernel/cpu/perf_event_amd.c +++ b/arch/x86/kernel/cpu/perf_event_amd.c @@ -1,4 +1,5 @@ #include <linux/perf_event.h> +#include <linux/export.h> #include <linux/types.h> #include <linux/init.h> #include <linux/slab.h> @@ -131,13 +132,49 @@ static u64 amd_pmu_event_map(int hw_event) return amd_perfmon_event_map[hw_event]; } -static int amd_pmu_hw_config(struct perf_event *event) +/* + * Previously calculated offsets + */ +static unsigned int event_offsets[X86_PMC_IDX_MAX] __read_mostly; +static unsigned int count_offsets[X86_PMC_IDX_MAX] __read_mostly; + +/* + * Legacy CPUs: + * 4 counters starting at 0xc0010000 each offset by 1 + * + * CPUs with core performance counter extensions: + * 6 counters starting at 0xc0010200 each offset by 2 + */ +static inline int amd_pmu_addr_offset(int index, bool eventsel) { - int ret = x86_pmu_hw_config(event); + int offset; - if (ret) - return ret; + if (!index) + return index; + + if (eventsel) + offset = event_offsets[index]; + else + offset = count_offsets[index]; + if (offset) + return offset; + + if (!cpu_has_perfctr_core) + offset = index; + else + offset = index << 1; + + if (eventsel) + event_offsets[index] = offset; + else + count_offsets[index] = offset; + + return offset; +} + +static int amd_core_hw_config(struct perf_event *event) +{ if (event->attr.exclude_host && event->attr.exclude_guest) /* * When HO == GO == 1 the hardware treats that as GO == HO == 0 @@ -147,14 +184,9 @@ static int amd_pmu_hw_config(struct perf_event *event) event->hw.config &= ~(ARCH_PERFMON_EVENTSEL_USR | ARCH_PERFMON_EVENTSEL_OS); else if (event->attr.exclude_host) - event->hw.config |= AMD_PERFMON_EVENTSEL_GUESTONLY; + event->hw.config |= AMD64_EVENTSEL_GUESTONLY; else if (event->attr.exclude_guest) - event->hw.config |= AMD_PERFMON_EVENTSEL_HOSTONLY; - - if (event->attr.type != PERF_TYPE_RAW) - return 0; - - event->hw.config |= event->attr.config & AMD64_RAW_EVENT_MASK; + event->hw.config |= AMD64_EVENTSEL_HOSTONLY; return 0; } @@ -179,20 +211,34 @@ static inline int amd_has_nb(struct cpu_hw_events *cpuc) return nb && nb->nb_id != -1; } -static void amd_put_event_constraints(struct cpu_hw_events *cpuc, - struct perf_event *event) +static int amd_pmu_hw_config(struct perf_event *event) +{ + int ret; + + /* pass precise event sampling to ibs: */ + if (event->attr.precise_ip && get_ibs_caps()) + return -ENOENT; + + if (has_branch_stack(event)) + return -EOPNOTSUPP; + + ret = x86_pmu_hw_config(event); + if (ret) + return ret; + + if (event->attr.type == PERF_TYPE_RAW) + event->hw.config |= event->attr.config & AMD64_RAW_EVENT_MASK; + + return amd_core_hw_config(event); +} + +static void __amd_put_nb_event_constraints(struct cpu_hw_events *cpuc, + struct perf_event *event) { - struct hw_perf_event *hwc = &event->hw; struct amd_nb *nb = cpuc->amd_nb; int i; /* - * only care about NB events - */ - if (!(amd_has_nb(cpuc) && amd_is_nb_event(hwc))) - return; - - /* * need to scan whole list because event may not have * been assigned during scheduling * @@ -201,10 +247,8 @@ static void amd_put_event_constraints(struct cpu_hw_events *cpuc, * when we come here */ for (i = 0; i < x86_pmu.num_counters; i++) { - if (nb->owners[i] == event) { - cmpxchg(nb->owners+i, event, NULL); + if (cmpxchg(nb->owners + i, event, NULL) == event) break; - } } } @@ -240,24 +284,24 @@ static void amd_put_event_constraints(struct cpu_hw_events *cpuc, * * Given that resources are allocated (cmpxchg), they must be * eventually freed for others to use. This is accomplished by - * calling amd_put_event_constraints(). + * calling __amd_put_nb_event_constraints() * * Non NB events are not impacted by this restriction. */ static struct event_constraint * -amd_get_event_constraints(struct cpu_hw_events *cpuc, struct perf_event *event) +__amd_get_nb_event_constraints(struct cpu_hw_events *cpuc, struct perf_event *event, + struct event_constraint *c) { struct hw_perf_event *hwc = &event->hw; struct amd_nb *nb = cpuc->amd_nb; - struct perf_event *old = NULL; - int max = x86_pmu.num_counters; - int i, j, k = -1; + struct perf_event *old; + int idx, new = -1; - /* - * if not NB event or no NB, then no constraints - */ - if (!(amd_has_nb(cpuc) && amd_is_nb_event(hwc))) - return &unconstrained; + if (!c) + c = &unconstrained; + + if (cpuc->is_fake) + return c; /* * detect if already present, if so reuse @@ -269,48 +313,33 @@ amd_get_event_constraints(struct cpu_hw_events *cpuc, struct perf_event *event) * because of successive calls to x86_schedule_events() from * hw_perf_group_sched_in() without hw_perf_enable() */ - for (i = 0; i < max; i++) { - /* - * keep track of first free slot - */ - if (k == -1 && !nb->owners[i]) - k = i; + for_each_set_bit(idx, c->idxmsk, x86_pmu.num_counters) { + if (new == -1 || hwc->idx == idx) + /* assign free slot, prefer hwc->idx */ + old = cmpxchg(nb->owners + idx, NULL, event); + else if (nb->owners[idx] == event) + /* event already present */ + old = event; + else + continue; + + if (old && old != event) + continue; + + /* reassign to this slot */ + if (new != -1) + cmpxchg(nb->owners + new, event, NULL); + new = idx; /* already present, reuse */ - if (nb->owners[i] == event) - goto done; - } - /* - * not present, so grab a new slot - * starting either at: - */ - if (hwc->idx != -1) { - /* previous assignment */ - i = hwc->idx; - } else if (k != -1) { - /* start from free slot found */ - i = k; - } else { - /* - * event not found, no slot found in - * first pass, try again from the - * beginning - */ - i = 0; - } - j = i; - do { - old = cmpxchg(nb->owners+i, NULL, event); - if (!old) + if (old == event) break; - if (++i == max) - i = 0; - } while (i != j); -done: - if (!old) - return &nb->event_constraints[i]; - - return &emptyconstraint; + } + + if (new == -1) + return &emptyconstraint; + + return &nb->event_constraints[new]; } static struct amd_nb *amd_alloc_nb(int cpu) @@ -318,8 +347,7 @@ static struct amd_nb *amd_alloc_nb(int cpu) struct amd_nb *nb; int i; - nb = kmalloc_node(sizeof(struct amd_nb), GFP_KERNEL | __GFP_ZERO, - cpu_to_node(cpu)); + nb = kzalloc_node(sizeof(struct amd_nb), GFP_KERNEL, cpu_to_node(cpu)); if (!nb) return NULL; @@ -357,6 +385,8 @@ static void amd_pmu_cpu_starting(int cpu) struct amd_nb *nb; int i, nb_id; + cpuc->perf_ctr_virt_mask = AMD64_EVENTSEL_HOSTONLY; + if (boot_cpu_data.x86_max_cores < 2) return; @@ -398,31 +428,38 @@ static void amd_pmu_cpu_dead(int cpu) } } -static __initconst const struct x86_pmu amd_pmu = { - .name = "AMD", - .handle_irq = x86_pmu_handle_irq, - .disable_all = x86_pmu_disable_all, - .enable_all = x86_pmu_enable_all, - .enable = x86_pmu_enable_event, - .disable = x86_pmu_disable_event, - .hw_config = amd_pmu_hw_config, - .schedule_events = x86_schedule_events, - .eventsel = MSR_K7_EVNTSEL0, - .perfctr = MSR_K7_PERFCTR0, - .event_map = amd_pmu_event_map, - .max_events = ARRAY_SIZE(amd_perfmon_event_map), - .num_counters = AMD64_NUM_COUNTERS, - .cntval_bits = 48, - .cntval_mask = (1ULL << 48) - 1, - .apic = 1, - /* use highest bit to detect overflow */ - .max_period = (1ULL << 47) - 1, - .get_event_constraints = amd_get_event_constraints, - .put_event_constraints = amd_put_event_constraints, +static struct event_constraint * +amd_get_event_constraints(struct cpu_hw_events *cpuc, struct perf_event *event) +{ + /* + * if not NB event or no NB, then no constraints + */ + if (!(amd_has_nb(cpuc) && amd_is_nb_event(&event->hw))) + return &unconstrained; - .cpu_prepare = amd_pmu_cpu_prepare, - .cpu_starting = amd_pmu_cpu_starting, - .cpu_dead = amd_pmu_cpu_dead, + return __amd_get_nb_event_constraints(cpuc, event, NULL); +} + +static void amd_put_event_constraints(struct cpu_hw_events *cpuc, + struct perf_event *event) +{ + if (amd_has_nb(cpuc) && amd_is_nb_event(&event->hw)) + __amd_put_nb_event_constraints(cpuc, event); +} + +PMU_FORMAT_ATTR(event, "config:0-7,32-35"); +PMU_FORMAT_ATTR(umask, "config:8-15" ); +PMU_FORMAT_ATTR(edge, "config:18" ); +PMU_FORMAT_ATTR(inv, "config:23" ); +PMU_FORMAT_ATTR(cmask, "config:24-31" ); + +static struct attribute *amd_format_attr[] = { + &format_attr_event.attr, + &format_attr_umask.attr, + &format_attr_edge.attr, + &format_attr_inv.attr, + &format_attr_cmask.attr, + NULL, }; /* AMD Family 15h */ @@ -470,6 +507,7 @@ static __initconst const struct x86_pmu amd_pmu = { * 0x023 DE PERF_CTL[2:0] * 0x02D LS PERF_CTL[3] * 0x02E LS PERF_CTL[3,0] + * 0x031 LS PERF_CTL[2:0] (**) * 0x043 CU PERF_CTL[2:0] * 0x045 CU PERF_CTL[2:0] * 0x046 CU PERF_CTL[2:0] @@ -483,10 +521,12 @@ static __initconst const struct x86_pmu amd_pmu = { * 0x0DD LS PERF_CTL[5:0] * 0x0DE LS PERF_CTL[5:0] * 0x0DF LS PERF_CTL[5:0] + * 0x1C0 EX PERF_CTL[5:3] * 0x1D6 EX PERF_CTL[5:0] * 0x1D8 EX PERF_CTL[5:0] * - * (*) depending on the umask all FPU counters may be used + * (*) depending on the umask all FPU counters may be used + * (**) only one unitmask enabled at a time */ static struct event_constraint amd_f15_PMC0 = EVENT_CONSTRAINT(0, 0x01, 0); @@ -536,6 +576,12 @@ amd_get_event_constraints_f15h(struct cpu_hw_events *cpuc, struct perf_event *ev return &amd_f15_PMC3; case 0x02E: return &amd_f15_PMC30; + case 0x031: + if (hweight_long(hwc->config & ARCH_PERFMON_EVENTSEL_UMASK) <= 1) + return &amd_f15_PMC20; + return &emptyconstraint; + case 0x1C0: + return &amd_f15_PMC53; default: return &amd_f15_PMC50; } @@ -555,15 +601,23 @@ amd_get_event_constraints_f15h(struct cpu_hw_events *cpuc, struct perf_event *ev return &amd_f15_PMC20; } case AMD_EVENT_NB: - /* not yet implemented */ + /* moved to perf_event_amd_uncore.c */ return &emptyconstraint; default: return &emptyconstraint; } } -static __initconst const struct x86_pmu amd_pmu_f15h = { - .name = "AMD Family 15h", +static ssize_t amd_event_sysfs_show(char *page, u64 config) +{ + u64 event = (config & ARCH_PERFMON_EVENTSEL_EVENT) | + (config & AMD64_EVENTSEL_EVENT) >> 24; + + return x86_event_sysfs_show(page, config, event); +} + +static __initconst const struct x86_pmu amd_pmu = { + .name = "AMD", .handle_irq = x86_pmu_handle_irq, .disable_all = x86_pmu_disable_all, .enable_all = x86_pmu_enable_all, @@ -571,53 +625,104 @@ static __initconst const struct x86_pmu amd_pmu_f15h = { .disable = x86_pmu_disable_event, .hw_config = amd_pmu_hw_config, .schedule_events = x86_schedule_events, - .eventsel = MSR_F15H_PERF_CTL, - .perfctr = MSR_F15H_PERF_CTR, + .eventsel = MSR_K7_EVNTSEL0, + .perfctr = MSR_K7_PERFCTR0, + .addr_offset = amd_pmu_addr_offset, .event_map = amd_pmu_event_map, .max_events = ARRAY_SIZE(amd_perfmon_event_map), - .num_counters = AMD64_NUM_COUNTERS_F15H, + .num_counters = AMD64_NUM_COUNTERS, .cntval_bits = 48, .cntval_mask = (1ULL << 48) - 1, .apic = 1, /* use highest bit to detect overflow */ .max_period = (1ULL << 47) - 1, - .get_event_constraints = amd_get_event_constraints_f15h, - /* nortbridge counters not yet implemented: */ -#if 0 + .get_event_constraints = amd_get_event_constraints, .put_event_constraints = amd_put_event_constraints, + .format_attrs = amd_format_attr, + .events_sysfs_show = amd_event_sysfs_show, + .cpu_prepare = amd_pmu_cpu_prepare, .cpu_starting = amd_pmu_cpu_starting, .cpu_dead = amd_pmu_cpu_dead, -#endif }; -__init int amd_pmu_init(void) +static int __init amd_core_pmu_init(void) { - /* Performance-monitoring supported from K7 and later: */ - if (boot_cpu_data.x86 < 6) - return -ENODEV; + if (!cpu_has_perfctr_core) + return 0; - /* - * If core performance counter extensions exists, it must be - * family 15h, otherwise fail. See x86_pmu_addr_offset(). - */ switch (boot_cpu_data.x86) { case 0x15: - if (!cpu_has_perfctr_core) - return -ENODEV; - x86_pmu = amd_pmu_f15h; + pr_cont("Fam15h "); + x86_pmu.get_event_constraints = amd_get_event_constraints_f15h; break; + default: - if (cpu_has_perfctr_core) - return -ENODEV; - x86_pmu = amd_pmu; - break; + pr_err("core perfctr but no constraints; unknown hardware!\n"); + return -ENODEV; } + /* + * If core performance counter extensions exists, we must use + * MSR_F15H_PERF_CTL/MSR_F15H_PERF_CTR msrs. See also + * amd_pmu_addr_offset(). + */ + x86_pmu.eventsel = MSR_F15H_PERF_CTL; + x86_pmu.perfctr = MSR_F15H_PERF_CTR; + x86_pmu.num_counters = AMD64_NUM_COUNTERS_CORE; + + pr_cont("core perfctr, "); + return 0; +} + +__init int amd_pmu_init(void) +{ + int ret; + + /* Performance-monitoring supported from K7 and later: */ + if (boot_cpu_data.x86 < 6) + return -ENODEV; + + x86_pmu = amd_pmu; + + ret = amd_core_pmu_init(); + if (ret) + return ret; + /* Events are common for all AMDs */ memcpy(hw_cache_event_ids, amd_hw_cache_event_ids, sizeof(hw_cache_event_ids)); return 0; } + +void amd_pmu_enable_virt(void) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + + cpuc->perf_ctr_virt_mask = 0; + + /* Reload all events */ + x86_pmu_disable_all(); + x86_pmu_enable_all(0); +} +EXPORT_SYMBOL_GPL(amd_pmu_enable_virt); + +void amd_pmu_disable_virt(void) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + + /* + * We only mask out the Host-only bit so that host-only counting works + * when SVM is disabled. If someone sets up a guest-only counter when + * SVM is disabled the Guest-only bits still gets set and the counter + * will not count anything. + */ + cpuc->perf_ctr_virt_mask = AMD64_EVENTSEL_HOSTONLY; + + /* Reload all events */ + x86_pmu_disable_all(); + x86_pmu_enable_all(0); +} +EXPORT_SYMBOL_GPL(amd_pmu_disable_virt); diff --git a/arch/x86/kernel/cpu/perf_event_amd_ibs.c b/arch/x86/kernel/cpu/perf_event_amd_ibs.c index 3b8a2d30d14..cbb1be3ed9e 100644 --- a/arch/x86/kernel/cpu/perf_event_amd_ibs.c +++ b/arch/x86/kernel/cpu/perf_event_amd_ibs.c @@ -9,43 +9,652 @@ #include <linux/perf_event.h> #include <linux/module.h> #include <linux/pci.h> +#include <linux/ptrace.h> +#include <linux/syscore_ops.h> #include <asm/apic.h> +#include "perf_event.h" + static u32 ibs_caps; #if defined(CONFIG_PERF_EVENTS) && defined(CONFIG_CPU_SUP_AMD) -static struct pmu perf_ibs; +#include <linux/kprobes.h> +#include <linux/hardirq.h> + +#include <asm/nmi.h> + +#define IBS_FETCH_CONFIG_MASK (IBS_FETCH_RAND_EN | IBS_FETCH_MAX_CNT) +#define IBS_OP_CONFIG_MASK IBS_OP_MAX_CNT + +enum ibs_states { + IBS_ENABLED = 0, + IBS_STARTED = 1, + IBS_STOPPING = 2, + + IBS_MAX_STATES, +}; + +struct cpu_perf_ibs { + struct perf_event *event; + unsigned long state[BITS_TO_LONGS(IBS_MAX_STATES)]; +}; + +struct perf_ibs { + struct pmu pmu; + unsigned int msr; + u64 config_mask; + u64 cnt_mask; + u64 enable_mask; + u64 valid_mask; + u64 max_period; + unsigned long offset_mask[1]; + int offset_max; + struct cpu_perf_ibs __percpu *pcpu; + + struct attribute **format_attrs; + struct attribute_group format_group; + const struct attribute_group *attr_groups[2]; + + u64 (*get_count)(u64 config); +}; + +struct perf_ibs_data { + u32 size; + union { + u32 data[0]; /* data buffer starts here */ + u32 caps; + }; + u64 regs[MSR_AMD64_IBS_REG_COUNT_MAX]; +}; + +static int +perf_event_set_period(struct hw_perf_event *hwc, u64 min, u64 max, u64 *hw_period) +{ + s64 left = local64_read(&hwc->period_left); + s64 period = hwc->sample_period; + int overflow = 0; + + /* + * If we are way outside a reasonable range then just skip forward: + */ + if (unlikely(left <= -period)) { + left = period; + local64_set(&hwc->period_left, left); + hwc->last_period = period; + overflow = 1; + } + + if (unlikely(left < (s64)min)) { + left += period; + local64_set(&hwc->period_left, left); + hwc->last_period = period; + overflow = 1; + } + + /* + * If the hw period that triggers the sw overflow is too short + * we might hit the irq handler. This biases the results. + * Thus we shorten the next-to-last period and set the last + * period to the max period. + */ + if (left > max) { + left -= max; + if (left > max) + left = max; + else if (left < min) + left = min; + } + + *hw_period = (u64)left; + + return overflow; +} + +static int +perf_event_try_update(struct perf_event *event, u64 new_raw_count, int width) +{ + struct hw_perf_event *hwc = &event->hw; + int shift = 64 - width; + u64 prev_raw_count; + u64 delta; + + /* + * Careful: an NMI might modify the previous event value. + * + * Our tactic to handle this is to first atomically read and + * exchange a new raw count - then add that new-prev delta + * count to the generic event atomically: + */ + prev_raw_count = local64_read(&hwc->prev_count); + if (local64_cmpxchg(&hwc->prev_count, prev_raw_count, + new_raw_count) != prev_raw_count) + return 0; + + /* + * Now we have the new raw value and have updated the prev + * timestamp already. We can now calculate the elapsed delta + * (event-)time and add that to the generic event. + * + * Careful, not all hw sign-extends above the physical width + * of the count. + */ + delta = (new_raw_count << shift) - (prev_raw_count << shift); + delta >>= shift; + + local64_add(delta, &event->count); + local64_sub(delta, &hwc->period_left); + + return 1; +} + +static struct perf_ibs perf_ibs_fetch; +static struct perf_ibs perf_ibs_op; + +static struct perf_ibs *get_ibs_pmu(int type) +{ + if (perf_ibs_fetch.pmu.type == type) + return &perf_ibs_fetch; + if (perf_ibs_op.pmu.type == type) + return &perf_ibs_op; + return NULL; +} + +/* + * Use IBS for precise event sampling: + * + * perf record -a -e cpu-cycles:p ... # use ibs op counting cycle count + * perf record -a -e r076:p ... # same as -e cpu-cycles:p + * perf record -a -e r0C1:p ... # use ibs op counting micro-ops + * + * IbsOpCntCtl (bit 19) of IBS Execution Control Register (IbsOpCtl, + * MSRC001_1033) is used to select either cycle or micro-ops counting + * mode. + * + * The rip of IBS samples has skid 0. Thus, IBS supports precise + * levels 1 and 2 and the PERF_EFLAGS_EXACT is set. In rare cases the + * rip is invalid when IBS was not able to record the rip correctly. + * We clear PERF_EFLAGS_EXACT and take the rip from pt_regs then. + * + */ +static int perf_ibs_precise_event(struct perf_event *event, u64 *config) +{ + switch (event->attr.precise_ip) { + case 0: + return -ENOENT; + case 1: + case 2: + break; + default: + return -EOPNOTSUPP; + } + + switch (event->attr.type) { + case PERF_TYPE_HARDWARE: + switch (event->attr.config) { + case PERF_COUNT_HW_CPU_CYCLES: + *config = 0; + return 0; + } + break; + case PERF_TYPE_RAW: + switch (event->attr.config) { + case 0x0076: + *config = 0; + return 0; + case 0x00C1: + *config = IBS_OP_CNT_CTL; + return 0; + } + break; + default: + return -ENOENT; + } + + return -EOPNOTSUPP; +} + +static const struct perf_event_attr ibs_notsupp = { + .exclude_user = 1, + .exclude_kernel = 1, + .exclude_hv = 1, + .exclude_idle = 1, + .exclude_host = 1, + .exclude_guest = 1, +}; static int perf_ibs_init(struct perf_event *event) { - if (perf_ibs.type != event->attr.type) + struct hw_perf_event *hwc = &event->hw; + struct perf_ibs *perf_ibs; + u64 max_cnt, config; + int ret; + + perf_ibs = get_ibs_pmu(event->attr.type); + if (perf_ibs) { + config = event->attr.config; + } else { + perf_ibs = &perf_ibs_op; + ret = perf_ibs_precise_event(event, &config); + if (ret) + return ret; + } + + if (event->pmu != &perf_ibs->pmu) return -ENOENT; + + if (perf_flags(&event->attr) & perf_flags(&ibs_notsupp)) + return -EINVAL; + + if (config & ~perf_ibs->config_mask) + return -EINVAL; + + if (hwc->sample_period) { + if (config & perf_ibs->cnt_mask) + /* raw max_cnt may not be set */ + return -EINVAL; + if (!event->attr.sample_freq && hwc->sample_period & 0x0f) + /* + * lower 4 bits can not be set in ibs max cnt, + * but allowing it in case we adjust the + * sample period to set a frequency. + */ + return -EINVAL; + hwc->sample_period &= ~0x0FULL; + if (!hwc->sample_period) + hwc->sample_period = 0x10; + } else { + max_cnt = config & perf_ibs->cnt_mask; + config &= ~perf_ibs->cnt_mask; + event->attr.sample_period = max_cnt << 4; + hwc->sample_period = event->attr.sample_period; + } + + if (!hwc->sample_period) + return -EINVAL; + + /* + * If we modify hwc->sample_period, we also need to update + * hwc->last_period and hwc->period_left. + */ + hwc->last_period = hwc->sample_period; + local64_set(&hwc->period_left, hwc->sample_period); + + hwc->config_base = perf_ibs->msr; + hwc->config = config; + return 0; } +static int perf_ibs_set_period(struct perf_ibs *perf_ibs, + struct hw_perf_event *hwc, u64 *period) +{ + int overflow; + + /* ignore lower 4 bits in min count: */ + overflow = perf_event_set_period(hwc, 1<<4, perf_ibs->max_period, period); + local64_set(&hwc->prev_count, 0); + + return overflow; +} + +static u64 get_ibs_fetch_count(u64 config) +{ + return (config & IBS_FETCH_CNT) >> 12; +} + +static u64 get_ibs_op_count(u64 config) +{ + u64 count = 0; + + if (config & IBS_OP_VAL) + count += (config & IBS_OP_MAX_CNT) << 4; /* cnt rolled over */ + + if (ibs_caps & IBS_CAPS_RDWROPCNT) + count += (config & IBS_OP_CUR_CNT) >> 32; + + return count; +} + +static void +perf_ibs_event_update(struct perf_ibs *perf_ibs, struct perf_event *event, + u64 *config) +{ + u64 count = perf_ibs->get_count(*config); + + /* + * Set width to 64 since we do not overflow on max width but + * instead on max count. In perf_ibs_set_period() we clear + * prev count manually on overflow. + */ + while (!perf_event_try_update(event, count, 64)) { + rdmsrl(event->hw.config_base, *config); + count = perf_ibs->get_count(*config); + } +} + +static inline void perf_ibs_enable_event(struct perf_ibs *perf_ibs, + struct hw_perf_event *hwc, u64 config) +{ + wrmsrl(hwc->config_base, hwc->config | config | perf_ibs->enable_mask); +} + +/* + * Erratum #420 Instruction-Based Sampling Engine May Generate + * Interrupt that Cannot Be Cleared: + * + * Must clear counter mask first, then clear the enable bit. See + * Revision Guide for AMD Family 10h Processors, Publication #41322. + */ +static inline void perf_ibs_disable_event(struct perf_ibs *perf_ibs, + struct hw_perf_event *hwc, u64 config) +{ + config &= ~perf_ibs->cnt_mask; + wrmsrl(hwc->config_base, config); + config &= ~perf_ibs->enable_mask; + wrmsrl(hwc->config_base, config); +} + +/* + * We cannot restore the ibs pmu state, so we always needs to update + * the event while stopping it and then reset the state when starting + * again. Thus, ignoring PERF_EF_RELOAD and PERF_EF_UPDATE flags in + * perf_ibs_start()/perf_ibs_stop() and instead always do it. + */ +static void perf_ibs_start(struct perf_event *event, int flags) +{ + struct hw_perf_event *hwc = &event->hw; + struct perf_ibs *perf_ibs = container_of(event->pmu, struct perf_ibs, pmu); + struct cpu_perf_ibs *pcpu = this_cpu_ptr(perf_ibs->pcpu); + u64 period; + + if (WARN_ON_ONCE(!(hwc->state & PERF_HES_STOPPED))) + return; + + WARN_ON_ONCE(!(hwc->state & PERF_HES_UPTODATE)); + hwc->state = 0; + + perf_ibs_set_period(perf_ibs, hwc, &period); + set_bit(IBS_STARTED, pcpu->state); + perf_ibs_enable_event(perf_ibs, hwc, period >> 4); + + perf_event_update_userpage(event); +} + +static void perf_ibs_stop(struct perf_event *event, int flags) +{ + struct hw_perf_event *hwc = &event->hw; + struct perf_ibs *perf_ibs = container_of(event->pmu, struct perf_ibs, pmu); + struct cpu_perf_ibs *pcpu = this_cpu_ptr(perf_ibs->pcpu); + u64 config; + int stopping; + + stopping = test_and_clear_bit(IBS_STARTED, pcpu->state); + + if (!stopping && (hwc->state & PERF_HES_UPTODATE)) + return; + + rdmsrl(hwc->config_base, config); + + if (stopping) { + set_bit(IBS_STOPPING, pcpu->state); + perf_ibs_disable_event(perf_ibs, hwc, config); + WARN_ON_ONCE(hwc->state & PERF_HES_STOPPED); + hwc->state |= PERF_HES_STOPPED; + } + + if (hwc->state & PERF_HES_UPTODATE) + return; + + /* + * Clear valid bit to not count rollovers on update, rollovers + * are only updated in the irq handler. + */ + config &= ~perf_ibs->valid_mask; + + perf_ibs_event_update(perf_ibs, event, &config); + hwc->state |= PERF_HES_UPTODATE; +} + static int perf_ibs_add(struct perf_event *event, int flags) { + struct perf_ibs *perf_ibs = container_of(event->pmu, struct perf_ibs, pmu); + struct cpu_perf_ibs *pcpu = this_cpu_ptr(perf_ibs->pcpu); + + if (test_and_set_bit(IBS_ENABLED, pcpu->state)) + return -ENOSPC; + + event->hw.state = PERF_HES_UPTODATE | PERF_HES_STOPPED; + + pcpu->event = event; + + if (flags & PERF_EF_START) + perf_ibs_start(event, PERF_EF_RELOAD); + return 0; } static void perf_ibs_del(struct perf_event *event, int flags) { + struct perf_ibs *perf_ibs = container_of(event->pmu, struct perf_ibs, pmu); + struct cpu_perf_ibs *pcpu = this_cpu_ptr(perf_ibs->pcpu); + + if (!test_and_clear_bit(IBS_ENABLED, pcpu->state)) + return; + + perf_ibs_stop(event, PERF_EF_UPDATE); + + pcpu->event = NULL; + + perf_event_update_userpage(event); } -static struct pmu perf_ibs = { - .event_init= perf_ibs_init, - .add= perf_ibs_add, - .del= perf_ibs_del, +static void perf_ibs_read(struct perf_event *event) { } + +PMU_FORMAT_ATTR(rand_en, "config:57"); +PMU_FORMAT_ATTR(cnt_ctl, "config:19"); + +static struct attribute *ibs_fetch_format_attrs[] = { + &format_attr_rand_en.attr, + NULL, }; +static struct attribute *ibs_op_format_attrs[] = { + NULL, /* &format_attr_cnt_ctl.attr if IBS_CAPS_OPCNT */ + NULL, +}; + +static struct perf_ibs perf_ibs_fetch = { + .pmu = { + .task_ctx_nr = perf_invalid_context, + + .event_init = perf_ibs_init, + .add = perf_ibs_add, + .del = perf_ibs_del, + .start = perf_ibs_start, + .stop = perf_ibs_stop, + .read = perf_ibs_read, + }, + .msr = MSR_AMD64_IBSFETCHCTL, + .config_mask = IBS_FETCH_CONFIG_MASK, + .cnt_mask = IBS_FETCH_MAX_CNT, + .enable_mask = IBS_FETCH_ENABLE, + .valid_mask = IBS_FETCH_VAL, + .max_period = IBS_FETCH_MAX_CNT << 4, + .offset_mask = { MSR_AMD64_IBSFETCH_REG_MASK }, + .offset_max = MSR_AMD64_IBSFETCH_REG_COUNT, + .format_attrs = ibs_fetch_format_attrs, + + .get_count = get_ibs_fetch_count, +}; + +static struct perf_ibs perf_ibs_op = { + .pmu = { + .task_ctx_nr = perf_invalid_context, + + .event_init = perf_ibs_init, + .add = perf_ibs_add, + .del = perf_ibs_del, + .start = perf_ibs_start, + .stop = perf_ibs_stop, + .read = perf_ibs_read, + }, + .msr = MSR_AMD64_IBSOPCTL, + .config_mask = IBS_OP_CONFIG_MASK, + .cnt_mask = IBS_OP_MAX_CNT, + .enable_mask = IBS_OP_ENABLE, + .valid_mask = IBS_OP_VAL, + .max_period = IBS_OP_MAX_CNT << 4, + .offset_mask = { MSR_AMD64_IBSOP_REG_MASK }, + .offset_max = MSR_AMD64_IBSOP_REG_COUNT, + .format_attrs = ibs_op_format_attrs, + + .get_count = get_ibs_op_count, +}; + +static int perf_ibs_handle_irq(struct perf_ibs *perf_ibs, struct pt_regs *iregs) +{ + struct cpu_perf_ibs *pcpu = this_cpu_ptr(perf_ibs->pcpu); + struct perf_event *event = pcpu->event; + struct hw_perf_event *hwc = &event->hw; + struct perf_sample_data data; + struct perf_raw_record raw; + struct pt_regs regs; + struct perf_ibs_data ibs_data; + int offset, size, check_rip, offset_max, throttle = 0; + unsigned int msr; + u64 *buf, *config, period; + + if (!test_bit(IBS_STARTED, pcpu->state)) { + /* + * Catch spurious interrupts after stopping IBS: After + * disabling IBS there could be still incoming NMIs + * with samples that even have the valid bit cleared. + * Mark all this NMIs as handled. + */ + return test_and_clear_bit(IBS_STOPPING, pcpu->state) ? 1 : 0; + } + + msr = hwc->config_base; + buf = ibs_data.regs; + rdmsrl(msr, *buf); + if (!(*buf++ & perf_ibs->valid_mask)) + return 0; + + config = &ibs_data.regs[0]; + perf_ibs_event_update(perf_ibs, event, config); + perf_sample_data_init(&data, 0, hwc->last_period); + if (!perf_ibs_set_period(perf_ibs, hwc, &period)) + goto out; /* no sw counter overflow */ + + ibs_data.caps = ibs_caps; + size = 1; + offset = 1; + check_rip = (perf_ibs == &perf_ibs_op && (ibs_caps & IBS_CAPS_RIPINVALIDCHK)); + if (event->attr.sample_type & PERF_SAMPLE_RAW) + offset_max = perf_ibs->offset_max; + else if (check_rip) + offset_max = 2; + else + offset_max = 1; + do { + rdmsrl(msr + offset, *buf++); + size++; + offset = find_next_bit(perf_ibs->offset_mask, + perf_ibs->offset_max, + offset + 1); + } while (offset < offset_max); + ibs_data.size = sizeof(u64) * size; + + regs = *iregs; + if (check_rip && (ibs_data.regs[2] & IBS_RIP_INVALID)) { + regs.flags &= ~PERF_EFLAGS_EXACT; + } else { + set_linear_ip(®s, ibs_data.regs[1]); + regs.flags |= PERF_EFLAGS_EXACT; + } + + if (event->attr.sample_type & PERF_SAMPLE_RAW) { + raw.size = sizeof(u32) + ibs_data.size; + raw.data = ibs_data.data; + data.raw = &raw; + } + + throttle = perf_event_overflow(event, &data, ®s); +out: + if (throttle) + perf_ibs_disable_event(perf_ibs, hwc, *config); + else + perf_ibs_enable_event(perf_ibs, hwc, period >> 4); + + perf_event_update_userpage(event); + + return 1; +} + +static int +perf_ibs_nmi_handler(unsigned int cmd, struct pt_regs *regs) +{ + int handled = 0; + + handled += perf_ibs_handle_irq(&perf_ibs_fetch, regs); + handled += perf_ibs_handle_irq(&perf_ibs_op, regs); + + if (handled) + inc_irq_stat(apic_perf_irqs); + + return handled; +} +NOKPROBE_SYMBOL(perf_ibs_nmi_handler); + +static __init int perf_ibs_pmu_init(struct perf_ibs *perf_ibs, char *name) +{ + struct cpu_perf_ibs __percpu *pcpu; + int ret; + + pcpu = alloc_percpu(struct cpu_perf_ibs); + if (!pcpu) + return -ENOMEM; + + perf_ibs->pcpu = pcpu; + + /* register attributes */ + if (perf_ibs->format_attrs[0]) { + memset(&perf_ibs->format_group, 0, sizeof(perf_ibs->format_group)); + perf_ibs->format_group.name = "format"; + perf_ibs->format_group.attrs = perf_ibs->format_attrs; + + memset(&perf_ibs->attr_groups, 0, sizeof(perf_ibs->attr_groups)); + perf_ibs->attr_groups[0] = &perf_ibs->format_group; + perf_ibs->pmu.attr_groups = perf_ibs->attr_groups; + } + + ret = perf_pmu_register(&perf_ibs->pmu, name, -1); + if (ret) { + perf_ibs->pcpu = NULL; + free_percpu(pcpu); + } + + return ret; +} + static __init int perf_event_ibs_init(void) { + struct attribute **attr = ibs_op_format_attrs; + if (!ibs_caps) return -ENODEV; /* ibs not supported by the cpu */ - perf_pmu_register(&perf_ibs, "ibs", -1); + perf_ibs_pmu_init(&perf_ibs_fetch, "ibs_fetch"); + + if (ibs_caps & IBS_CAPS_OPCNT) { + perf_ibs_op.config_mask |= IBS_OP_CNT_CTL; + *attr++ = &format_attr_cnt_ctl.attr; + } + perf_ibs_pmu_init(&perf_ibs_op, "ibs_op"); + + register_nmi_handler(NMI_LOCAL, perf_ibs_nmi_handler, 0, "perf_ibs"); printk(KERN_INFO "perf: AMD IBS detected (0x%08x)\n", ibs_caps); return 0; @@ -209,6 +818,18 @@ out: return ret; } +static void ibs_eilvt_setup(void) +{ + /* + * Force LVT offset assignment for family 10h: The offsets are + * not assigned by the BIOS for this family, so the OS is + * responsible for doing it. If the OS assignment fails, fall + * back to BIOS settings and try to setup this. + */ + if (boot_cpu_data.x86 == 0x10) + force_ibs_eilvt_setup(); +} + static inline int get_ibs_lvt_offset(void) { u64 val; @@ -244,7 +865,37 @@ static void clear_APIC_ibs(void *dummy) setup_APIC_eilvt(offset, 0, APIC_EILVT_MSG_FIX, 1); } -static int __cpuinit +#ifdef CONFIG_PM + +static int perf_ibs_suspend(void) +{ + clear_APIC_ibs(NULL); + return 0; +} + +static void perf_ibs_resume(void) +{ + ibs_eilvt_setup(); + setup_APIC_ibs(NULL); +} + +static struct syscore_ops perf_ibs_syscore_ops = { + .resume = perf_ibs_resume, + .suspend = perf_ibs_suspend, +}; + +static void perf_ibs_pm_init(void) +{ + register_syscore_ops(&perf_ibs_syscore_ops); +} + +#else + +static inline void perf_ibs_pm_init(void) { } + +#endif + +static int perf_ibs_cpu_notifier(struct notifier_block *self, unsigned long action, void *hcpu) { switch (action & ~CPU_TASKS_FROZEN) { @@ -270,25 +921,19 @@ static __init int amd_ibs_init(void) if (!caps) return -ENODEV; /* ibs not supported by the cpu */ - /* - * Force LVT offset assignment for family 10h: The offsets are - * not assigned by the BIOS for this family, so the OS is - * responsible for doing it. If the OS assignment fails, fall - * back to BIOS settings and try to setup this. - */ - if (boot_cpu_data.x86 == 0x10) - force_ibs_eilvt_setup(); + ibs_eilvt_setup(); if (!ibs_eilvt_valid()) goto out; - get_online_cpus(); + perf_ibs_pm_init(); + cpu_notifier_register_begin(); ibs_caps = caps; /* make ibs_caps visible to other cpus: */ smp_mb(); - perf_cpu_notifier(perf_ibs_cpu_notifier); smp_call_function(setup_APIC_ibs, NULL, 1); - put_online_cpus(); + __perf_cpu_notifier(perf_ibs_cpu_notifier); + cpu_notifier_register_done(); ret = perf_event_ibs_init(); out: diff --git a/arch/x86/kernel/cpu/perf_event_amd_iommu.c b/arch/x86/kernel/cpu/perf_event_amd_iommu.c new file mode 100644 index 00000000000..639d1289b1b --- /dev/null +++ b/arch/x86/kernel/cpu/perf_event_amd_iommu.c @@ -0,0 +1,502 @@ +/* + * Copyright (C) 2013 Advanced Micro Devices, Inc. + * + * Author: Steven Kinney <Steven.Kinney@amd.com> + * Author: Suravee Suthikulpanit <Suraveee.Suthikulpanit@amd.com> + * + * Perf: amd_iommu - AMD IOMMU Performance Counter PMU implementation + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/perf_event.h> +#include <linux/module.h> +#include <linux/cpumask.h> +#include <linux/slab.h> + +#include "perf_event.h" +#include "perf_event_amd_iommu.h" + +#define COUNTER_SHIFT 16 + +#define _GET_BANK(ev) ((u8)(ev->hw.extra_reg.reg >> 8)) +#define _GET_CNTR(ev) ((u8)(ev->hw.extra_reg.reg)) + +/* iommu pmu config masks */ +#define _GET_CSOURCE(ev) ((ev->hw.config & 0xFFULL)) +#define _GET_DEVID(ev) ((ev->hw.config >> 8) & 0xFFFFULL) +#define _GET_PASID(ev) ((ev->hw.config >> 24) & 0xFFFFULL) +#define _GET_DOMID(ev) ((ev->hw.config >> 40) & 0xFFFFULL) +#define _GET_DEVID_MASK(ev) ((ev->hw.extra_reg.config) & 0xFFFFULL) +#define _GET_PASID_MASK(ev) ((ev->hw.extra_reg.config >> 16) & 0xFFFFULL) +#define _GET_DOMID_MASK(ev) ((ev->hw.extra_reg.config >> 32) & 0xFFFFULL) + +static struct perf_amd_iommu __perf_iommu; + +struct perf_amd_iommu { + struct pmu pmu; + u8 max_banks; + u8 max_counters; + u64 cntr_assign_mask; + raw_spinlock_t lock; + const struct attribute_group *attr_groups[4]; +}; + +#define format_group attr_groups[0] +#define cpumask_group attr_groups[1] +#define events_group attr_groups[2] +#define null_group attr_groups[3] + +/*--------------------------------------------- + * sysfs format attributes + *---------------------------------------------*/ +PMU_FORMAT_ATTR(csource, "config:0-7"); +PMU_FORMAT_ATTR(devid, "config:8-23"); +PMU_FORMAT_ATTR(pasid, "config:24-39"); +PMU_FORMAT_ATTR(domid, "config:40-55"); +PMU_FORMAT_ATTR(devid_mask, "config1:0-15"); +PMU_FORMAT_ATTR(pasid_mask, "config1:16-31"); +PMU_FORMAT_ATTR(domid_mask, "config1:32-47"); + +static struct attribute *iommu_format_attrs[] = { + &format_attr_csource.attr, + &format_attr_devid.attr, + &format_attr_pasid.attr, + &format_attr_domid.attr, + &format_attr_devid_mask.attr, + &format_attr_pasid_mask.attr, + &format_attr_domid_mask.attr, + NULL, +}; + +static struct attribute_group amd_iommu_format_group = { + .name = "format", + .attrs = iommu_format_attrs, +}; + +/*--------------------------------------------- + * sysfs events attributes + *---------------------------------------------*/ +struct amd_iommu_event_desc { + struct kobj_attribute attr; + const char *event; +}; + +static ssize_t _iommu_event_show(struct kobject *kobj, + struct kobj_attribute *attr, char *buf) +{ + struct amd_iommu_event_desc *event = + container_of(attr, struct amd_iommu_event_desc, attr); + return sprintf(buf, "%s\n", event->event); +} + +#define AMD_IOMMU_EVENT_DESC(_name, _event) \ +{ \ + .attr = __ATTR(_name, 0444, _iommu_event_show, NULL), \ + .event = _event, \ +} + +static struct amd_iommu_event_desc amd_iommu_v2_event_descs[] = { + AMD_IOMMU_EVENT_DESC(mem_pass_untrans, "csource=0x01"), + AMD_IOMMU_EVENT_DESC(mem_pass_pretrans, "csource=0x02"), + AMD_IOMMU_EVENT_DESC(mem_pass_excl, "csource=0x03"), + AMD_IOMMU_EVENT_DESC(mem_target_abort, "csource=0x04"), + AMD_IOMMU_EVENT_DESC(mem_trans_total, "csource=0x05"), + AMD_IOMMU_EVENT_DESC(mem_iommu_tlb_pte_hit, "csource=0x06"), + AMD_IOMMU_EVENT_DESC(mem_iommu_tlb_pte_mis, "csource=0x07"), + AMD_IOMMU_EVENT_DESC(mem_iommu_tlb_pde_hit, "csource=0x08"), + AMD_IOMMU_EVENT_DESC(mem_iommu_tlb_pde_mis, "csource=0x09"), + AMD_IOMMU_EVENT_DESC(mem_dte_hit, "csource=0x0a"), + AMD_IOMMU_EVENT_DESC(mem_dte_mis, "csource=0x0b"), + AMD_IOMMU_EVENT_DESC(page_tbl_read_tot, "csource=0x0c"), + AMD_IOMMU_EVENT_DESC(page_tbl_read_nst, "csource=0x0d"), + AMD_IOMMU_EVENT_DESC(page_tbl_read_gst, "csource=0x0e"), + AMD_IOMMU_EVENT_DESC(int_dte_hit, "csource=0x0f"), + AMD_IOMMU_EVENT_DESC(int_dte_mis, "csource=0x10"), + AMD_IOMMU_EVENT_DESC(cmd_processed, "csource=0x11"), + AMD_IOMMU_EVENT_DESC(cmd_processed_inv, "csource=0x12"), + AMD_IOMMU_EVENT_DESC(tlb_inv, "csource=0x13"), + { /* end: all zeroes */ }, +}; + +/*--------------------------------------------- + * sysfs cpumask attributes + *---------------------------------------------*/ +static cpumask_t iommu_cpumask; + +static ssize_t _iommu_cpumask_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + int n = cpulist_scnprintf(buf, PAGE_SIZE - 2, &iommu_cpumask); + buf[n++] = '\n'; + buf[n] = '\0'; + return n; +} +static DEVICE_ATTR(cpumask, S_IRUGO, _iommu_cpumask_show, NULL); + +static struct attribute *iommu_cpumask_attrs[] = { + &dev_attr_cpumask.attr, + NULL, +}; + +static struct attribute_group amd_iommu_cpumask_group = { + .attrs = iommu_cpumask_attrs, +}; + +/*---------------------------------------------*/ + +static int get_next_avail_iommu_bnk_cntr(struct perf_amd_iommu *perf_iommu) +{ + unsigned long flags; + int shift, bank, cntr, retval; + int max_banks = perf_iommu->max_banks; + int max_cntrs = perf_iommu->max_counters; + + raw_spin_lock_irqsave(&perf_iommu->lock, flags); + + for (bank = 0, shift = 0; bank < max_banks; bank++) { + for (cntr = 0; cntr < max_cntrs; cntr++) { + shift = bank + (bank*3) + cntr; + if (perf_iommu->cntr_assign_mask & (1ULL<<shift)) { + continue; + } else { + perf_iommu->cntr_assign_mask |= (1ULL<<shift); + retval = ((u16)((u16)bank<<8) | (u8)(cntr)); + goto out; + } + } + } + retval = -ENOSPC; +out: + raw_spin_unlock_irqrestore(&perf_iommu->lock, flags); + return retval; +} + +static int clear_avail_iommu_bnk_cntr(struct perf_amd_iommu *perf_iommu, + u8 bank, u8 cntr) +{ + unsigned long flags; + int max_banks, max_cntrs; + int shift = 0; + + max_banks = perf_iommu->max_banks; + max_cntrs = perf_iommu->max_counters; + + if ((bank > max_banks) || (cntr > max_cntrs)) + return -EINVAL; + + shift = bank + cntr + (bank*3); + + raw_spin_lock_irqsave(&perf_iommu->lock, flags); + perf_iommu->cntr_assign_mask &= ~(1ULL<<shift); + raw_spin_unlock_irqrestore(&perf_iommu->lock, flags); + + return 0; +} + +static int perf_iommu_event_init(struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + struct perf_amd_iommu *perf_iommu; + u64 config, config1; + + /* test the event attr type check for PMU enumeration */ + if (event->attr.type != event->pmu->type) + return -ENOENT; + + /* + * IOMMU counters are shared across all cores. + * Therefore, it does not support per-process mode. + * Also, it does not support event sampling mode. + */ + if (is_sampling_event(event) || event->attach_state & PERF_ATTACH_TASK) + return -EINVAL; + + /* IOMMU counters do not have usr/os/guest/host bits */ + if (event->attr.exclude_user || event->attr.exclude_kernel || + event->attr.exclude_host || event->attr.exclude_guest) + return -EINVAL; + + if (event->cpu < 0) + return -EINVAL; + + perf_iommu = &__perf_iommu; + + if (event->pmu != &perf_iommu->pmu) + return -ENOENT; + + if (perf_iommu) { + config = event->attr.config; + config1 = event->attr.config1; + } else { + return -EINVAL; + } + + /* integrate with iommu base devid (0000), assume one iommu */ + perf_iommu->max_banks = + amd_iommu_pc_get_max_banks(IOMMU_BASE_DEVID); + perf_iommu->max_counters = + amd_iommu_pc_get_max_counters(IOMMU_BASE_DEVID); + if ((perf_iommu->max_banks == 0) || (perf_iommu->max_counters == 0)) + return -EINVAL; + + /* update the hw_perf_event struct with the iommu config data */ + hwc->config = config; + hwc->extra_reg.config = config1; + + return 0; +} + +static void perf_iommu_enable_event(struct perf_event *ev) +{ + u8 csource = _GET_CSOURCE(ev); + u16 devid = _GET_DEVID(ev); + u64 reg = 0ULL; + + reg = csource; + amd_iommu_pc_get_set_reg_val(devid, + _GET_BANK(ev), _GET_CNTR(ev) , + IOMMU_PC_COUNTER_SRC_REG, ®, true); + + reg = 0ULL | devid | (_GET_DEVID_MASK(ev) << 32); + if (reg) + reg |= (1UL << 31); + amd_iommu_pc_get_set_reg_val(devid, + _GET_BANK(ev), _GET_CNTR(ev) , + IOMMU_PC_DEVID_MATCH_REG, ®, true); + + reg = 0ULL | _GET_PASID(ev) | (_GET_PASID_MASK(ev) << 32); + if (reg) + reg |= (1UL << 31); + amd_iommu_pc_get_set_reg_val(devid, + _GET_BANK(ev), _GET_CNTR(ev) , + IOMMU_PC_PASID_MATCH_REG, ®, true); + + reg = 0ULL | _GET_DOMID(ev) | (_GET_DOMID_MASK(ev) << 32); + if (reg) + reg |= (1UL << 31); + amd_iommu_pc_get_set_reg_val(devid, + _GET_BANK(ev), _GET_CNTR(ev) , + IOMMU_PC_DOMID_MATCH_REG, ®, true); +} + +static void perf_iommu_disable_event(struct perf_event *event) +{ + u64 reg = 0ULL; + + amd_iommu_pc_get_set_reg_val(_GET_DEVID(event), + _GET_BANK(event), _GET_CNTR(event), + IOMMU_PC_COUNTER_SRC_REG, ®, true); +} + +static void perf_iommu_start(struct perf_event *event, int flags) +{ + struct hw_perf_event *hwc = &event->hw; + + pr_debug("perf: amd_iommu:perf_iommu_start\n"); + if (WARN_ON_ONCE(!(hwc->state & PERF_HES_STOPPED))) + return; + + WARN_ON_ONCE(!(hwc->state & PERF_HES_UPTODATE)); + hwc->state = 0; + + if (flags & PERF_EF_RELOAD) { + u64 prev_raw_count = local64_read(&hwc->prev_count); + amd_iommu_pc_get_set_reg_val(_GET_DEVID(event), + _GET_BANK(event), _GET_CNTR(event), + IOMMU_PC_COUNTER_REG, &prev_raw_count, true); + } + + perf_iommu_enable_event(event); + perf_event_update_userpage(event); + +} + +static void perf_iommu_read(struct perf_event *event) +{ + u64 count = 0ULL; + u64 prev_raw_count = 0ULL; + u64 delta = 0ULL; + struct hw_perf_event *hwc = &event->hw; + pr_debug("perf: amd_iommu:perf_iommu_read\n"); + + amd_iommu_pc_get_set_reg_val(_GET_DEVID(event), + _GET_BANK(event), _GET_CNTR(event), + IOMMU_PC_COUNTER_REG, &count, false); + + /* IOMMU pc counter register is only 48 bits */ + count &= 0xFFFFFFFFFFFFULL; + + prev_raw_count = local64_read(&hwc->prev_count); + if (local64_cmpxchg(&hwc->prev_count, prev_raw_count, + count) != prev_raw_count) + return; + + /* Handling 48-bit counter overflowing */ + delta = (count << COUNTER_SHIFT) - (prev_raw_count << COUNTER_SHIFT); + delta >>= COUNTER_SHIFT; + local64_add(delta, &event->count); + +} + +static void perf_iommu_stop(struct perf_event *event, int flags) +{ + struct hw_perf_event *hwc = &event->hw; + u64 config; + + pr_debug("perf: amd_iommu:perf_iommu_stop\n"); + + if (hwc->state & PERF_HES_UPTODATE) + return; + + perf_iommu_disable_event(event); + WARN_ON_ONCE(hwc->state & PERF_HES_STOPPED); + hwc->state |= PERF_HES_STOPPED; + + if (hwc->state & PERF_HES_UPTODATE) + return; + + config = hwc->config; + perf_iommu_read(event); + hwc->state |= PERF_HES_UPTODATE; +} + +static int perf_iommu_add(struct perf_event *event, int flags) +{ + int retval; + struct perf_amd_iommu *perf_iommu = + container_of(event->pmu, struct perf_amd_iommu, pmu); + + pr_debug("perf: amd_iommu:perf_iommu_add\n"); + event->hw.state = PERF_HES_UPTODATE | PERF_HES_STOPPED; + + /* request an iommu bank/counter */ + retval = get_next_avail_iommu_bnk_cntr(perf_iommu); + if (retval != -ENOSPC) + event->hw.extra_reg.reg = (u16)retval; + else + return retval; + + if (flags & PERF_EF_START) + perf_iommu_start(event, PERF_EF_RELOAD); + + return 0; +} + +static void perf_iommu_del(struct perf_event *event, int flags) +{ + struct perf_amd_iommu *perf_iommu = + container_of(event->pmu, struct perf_amd_iommu, pmu); + + pr_debug("perf: amd_iommu:perf_iommu_del\n"); + perf_iommu_stop(event, PERF_EF_UPDATE); + + /* clear the assigned iommu bank/counter */ + clear_avail_iommu_bnk_cntr(perf_iommu, + _GET_BANK(event), + _GET_CNTR(event)); + + perf_event_update_userpage(event); +} + +static __init int _init_events_attrs(struct perf_amd_iommu *perf_iommu) +{ + struct attribute **attrs; + struct attribute_group *attr_group; + int i = 0, j; + + while (amd_iommu_v2_event_descs[i].attr.attr.name) + i++; + + attr_group = kzalloc(sizeof(struct attribute *) + * (i + 1) + sizeof(*attr_group), GFP_KERNEL); + if (!attr_group) + return -ENOMEM; + + attrs = (struct attribute **)(attr_group + 1); + for (j = 0; j < i; j++) + attrs[j] = &amd_iommu_v2_event_descs[j].attr.attr; + + attr_group->name = "events"; + attr_group->attrs = attrs; + perf_iommu->events_group = attr_group; + + return 0; +} + +static __init void amd_iommu_pc_exit(void) +{ + if (__perf_iommu.events_group != NULL) { + kfree(__perf_iommu.events_group); + __perf_iommu.events_group = NULL; + } +} + +static __init int _init_perf_amd_iommu( + struct perf_amd_iommu *perf_iommu, char *name) +{ + int ret; + + raw_spin_lock_init(&perf_iommu->lock); + + /* Init format attributes */ + perf_iommu->format_group = &amd_iommu_format_group; + + /* Init cpumask attributes to only core 0 */ + cpumask_set_cpu(0, &iommu_cpumask); + perf_iommu->cpumask_group = &amd_iommu_cpumask_group; + + /* Init events attributes */ + if (_init_events_attrs(perf_iommu) != 0) + pr_err("perf: amd_iommu: Only support raw events.\n"); + + /* Init null attributes */ + perf_iommu->null_group = NULL; + perf_iommu->pmu.attr_groups = perf_iommu->attr_groups; + + ret = perf_pmu_register(&perf_iommu->pmu, name, -1); + if (ret) { + pr_err("perf: amd_iommu: Failed to initialized.\n"); + amd_iommu_pc_exit(); + } else { + pr_info("perf: amd_iommu: Detected. (%d banks, %d counters/bank)\n", + amd_iommu_pc_get_max_banks(IOMMU_BASE_DEVID), + amd_iommu_pc_get_max_counters(IOMMU_BASE_DEVID)); + } + + return ret; +} + +static struct perf_amd_iommu __perf_iommu = { + .pmu = { + .event_init = perf_iommu_event_init, + .add = perf_iommu_add, + .del = perf_iommu_del, + .start = perf_iommu_start, + .stop = perf_iommu_stop, + .read = perf_iommu_read, + }, + .max_banks = 0x00, + .max_counters = 0x00, + .cntr_assign_mask = 0ULL, + .format_group = NULL, + .cpumask_group = NULL, + .events_group = NULL, + .null_group = NULL, +}; + +static __init int amd_iommu_pc_init(void) +{ + /* Make sure the IOMMU PC resource is available */ + if (!amd_iommu_pc_supported()) + return -ENODEV; + + _init_perf_amd_iommu(&__perf_iommu, "amd_iommu"); + + return 0; +} + +device_initcall(amd_iommu_pc_init); diff --git a/arch/x86/kernel/cpu/perf_event_amd_iommu.h b/arch/x86/kernel/cpu/perf_event_amd_iommu.h new file mode 100644 index 00000000000..845d173278e --- /dev/null +++ b/arch/x86/kernel/cpu/perf_event_amd_iommu.h @@ -0,0 +1,40 @@ +/* + * Copyright (C) 2013 Advanced Micro Devices, Inc. + * + * Author: Steven Kinney <Steven.Kinney@amd.com> + * Author: Suravee Suthikulpanit <Suraveee.Suthikulpanit@amd.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#ifndef _PERF_EVENT_AMD_IOMMU_H_ +#define _PERF_EVENT_AMD_IOMMU_H_ + +/* iommu pc mmio region register indexes */ +#define IOMMU_PC_COUNTER_REG 0x00 +#define IOMMU_PC_COUNTER_SRC_REG 0x08 +#define IOMMU_PC_PASID_MATCH_REG 0x10 +#define IOMMU_PC_DOMID_MATCH_REG 0x18 +#define IOMMU_PC_DEVID_MATCH_REG 0x20 +#define IOMMU_PC_COUNTER_REPORT_REG 0x28 + +/* maximun specified bank/counters */ +#define PC_MAX_SPEC_BNKS 64 +#define PC_MAX_SPEC_CNTRS 16 + +/* iommu pc reg masks*/ +#define IOMMU_BASE_DEVID 0x0000 + +/* amd_iommu_init.c external support functions */ +extern bool amd_iommu_pc_supported(void); + +extern u8 amd_iommu_pc_get_max_banks(u16 devid); + +extern u8 amd_iommu_pc_get_max_counters(u16 devid); + +extern int amd_iommu_pc_get_set_reg_val(u16 devid, u8 bank, u8 cntr, + u8 fxn, u64 *value, bool is_write); + +#endif /*_PERF_EVENT_AMD_IOMMU_H_*/ diff --git a/arch/x86/kernel/cpu/perf_event_amd_uncore.c b/arch/x86/kernel/cpu/perf_event_amd_uncore.c new file mode 100644 index 00000000000..3bbdf4cd38b --- /dev/null +++ b/arch/x86/kernel/cpu/perf_event_amd_uncore.c @@ -0,0 +1,547 @@ +/* + * Copyright (C) 2013 Advanced Micro Devices, Inc. + * + * Author: Jacob Shin <jacob.shin@amd.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/perf_event.h> +#include <linux/percpu.h> +#include <linux/types.h> +#include <linux/slab.h> +#include <linux/init.h> +#include <linux/cpu.h> +#include <linux/cpumask.h> + +#include <asm/cpufeature.h> +#include <asm/perf_event.h> +#include <asm/msr.h> + +#define NUM_COUNTERS_NB 4 +#define NUM_COUNTERS_L2 4 +#define MAX_COUNTERS NUM_COUNTERS_NB + +#define RDPMC_BASE_NB 6 +#define RDPMC_BASE_L2 10 + +#define COUNTER_SHIFT 16 + +struct amd_uncore { + int id; + int refcnt; + int cpu; + int num_counters; + int rdpmc_base; + u32 msr_base; + cpumask_t *active_mask; + struct pmu *pmu; + struct perf_event *events[MAX_COUNTERS]; + struct amd_uncore *free_when_cpu_online; +}; + +static struct amd_uncore * __percpu *amd_uncore_nb; +static struct amd_uncore * __percpu *amd_uncore_l2; + +static struct pmu amd_nb_pmu; +static struct pmu amd_l2_pmu; + +static cpumask_t amd_nb_active_mask; +static cpumask_t amd_l2_active_mask; + +static bool is_nb_event(struct perf_event *event) +{ + return event->pmu->type == amd_nb_pmu.type; +} + +static bool is_l2_event(struct perf_event *event) +{ + return event->pmu->type == amd_l2_pmu.type; +} + +static struct amd_uncore *event_to_amd_uncore(struct perf_event *event) +{ + if (is_nb_event(event) && amd_uncore_nb) + return *per_cpu_ptr(amd_uncore_nb, event->cpu); + else if (is_l2_event(event) && amd_uncore_l2) + return *per_cpu_ptr(amd_uncore_l2, event->cpu); + + return NULL; +} + +static void amd_uncore_read(struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + u64 prev, new; + s64 delta; + + /* + * since we do not enable counter overflow interrupts, + * we do not have to worry about prev_count changing on us + */ + + prev = local64_read(&hwc->prev_count); + rdpmcl(hwc->event_base_rdpmc, new); + local64_set(&hwc->prev_count, new); + delta = (new << COUNTER_SHIFT) - (prev << COUNTER_SHIFT); + delta >>= COUNTER_SHIFT; + local64_add(delta, &event->count); +} + +static void amd_uncore_start(struct perf_event *event, int flags) +{ + struct hw_perf_event *hwc = &event->hw; + + if (flags & PERF_EF_RELOAD) + wrmsrl(hwc->event_base, (u64)local64_read(&hwc->prev_count)); + + hwc->state = 0; + wrmsrl(hwc->config_base, (hwc->config | ARCH_PERFMON_EVENTSEL_ENABLE)); + perf_event_update_userpage(event); +} + +static void amd_uncore_stop(struct perf_event *event, int flags) +{ + struct hw_perf_event *hwc = &event->hw; + + wrmsrl(hwc->config_base, hwc->config); + hwc->state |= PERF_HES_STOPPED; + + if ((flags & PERF_EF_UPDATE) && !(hwc->state & PERF_HES_UPTODATE)) { + amd_uncore_read(event); + hwc->state |= PERF_HES_UPTODATE; + } +} + +static int amd_uncore_add(struct perf_event *event, int flags) +{ + int i; + struct amd_uncore *uncore = event_to_amd_uncore(event); + struct hw_perf_event *hwc = &event->hw; + + /* are we already assigned? */ + if (hwc->idx != -1 && uncore->events[hwc->idx] == event) + goto out; + + for (i = 0; i < uncore->num_counters; i++) { + if (uncore->events[i] == event) { + hwc->idx = i; + goto out; + } + } + + /* if not, take the first available counter */ + hwc->idx = -1; + for (i = 0; i < uncore->num_counters; i++) { + if (cmpxchg(&uncore->events[i], NULL, event) == NULL) { + hwc->idx = i; + break; + } + } + +out: + if (hwc->idx == -1) + return -EBUSY; + + hwc->config_base = uncore->msr_base + (2 * hwc->idx); + hwc->event_base = uncore->msr_base + 1 + (2 * hwc->idx); + hwc->event_base_rdpmc = uncore->rdpmc_base + hwc->idx; + hwc->state = PERF_HES_UPTODATE | PERF_HES_STOPPED; + + if (flags & PERF_EF_START) + amd_uncore_start(event, PERF_EF_RELOAD); + + return 0; +} + +static void amd_uncore_del(struct perf_event *event, int flags) +{ + int i; + struct amd_uncore *uncore = event_to_amd_uncore(event); + struct hw_perf_event *hwc = &event->hw; + + amd_uncore_stop(event, PERF_EF_UPDATE); + + for (i = 0; i < uncore->num_counters; i++) { + if (cmpxchg(&uncore->events[i], event, NULL) == event) + break; + } + + hwc->idx = -1; +} + +static int amd_uncore_event_init(struct perf_event *event) +{ + struct amd_uncore *uncore; + struct hw_perf_event *hwc = &event->hw; + + if (event->attr.type != event->pmu->type) + return -ENOENT; + + /* + * NB and L2 counters (MSRs) are shared across all cores that share the + * same NB / L2 cache. Interrupts can be directed to a single target + * core, however, event counts generated by processes running on other + * cores cannot be masked out. So we do not support sampling and + * per-thread events. + */ + if (is_sampling_event(event) || event->attach_state & PERF_ATTACH_TASK) + return -EINVAL; + + /* NB and L2 counters do not have usr/os/guest/host bits */ + if (event->attr.exclude_user || event->attr.exclude_kernel || + event->attr.exclude_host || event->attr.exclude_guest) + return -EINVAL; + + /* and we do not enable counter overflow interrupts */ + hwc->config = event->attr.config & AMD64_RAW_EVENT_MASK_NB; + hwc->idx = -1; + + if (event->cpu < 0) + return -EINVAL; + + uncore = event_to_amd_uncore(event); + if (!uncore) + return -ENODEV; + + /* + * since request can come in to any of the shared cores, we will remap + * to a single common cpu. + */ + event->cpu = uncore->cpu; + + return 0; +} + +static ssize_t amd_uncore_attr_show_cpumask(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + int n; + cpumask_t *active_mask; + struct pmu *pmu = dev_get_drvdata(dev); + + if (pmu->type == amd_nb_pmu.type) + active_mask = &amd_nb_active_mask; + else if (pmu->type == amd_l2_pmu.type) + active_mask = &amd_l2_active_mask; + else + return 0; + + n = cpulist_scnprintf(buf, PAGE_SIZE - 2, active_mask); + buf[n++] = '\n'; + buf[n] = '\0'; + return n; +} +static DEVICE_ATTR(cpumask, S_IRUGO, amd_uncore_attr_show_cpumask, NULL); + +static struct attribute *amd_uncore_attrs[] = { + &dev_attr_cpumask.attr, + NULL, +}; + +static struct attribute_group amd_uncore_attr_group = { + .attrs = amd_uncore_attrs, +}; + +PMU_FORMAT_ATTR(event, "config:0-7,32-35"); +PMU_FORMAT_ATTR(umask, "config:8-15"); + +static struct attribute *amd_uncore_format_attr[] = { + &format_attr_event.attr, + &format_attr_umask.attr, + NULL, +}; + +static struct attribute_group amd_uncore_format_group = { + .name = "format", + .attrs = amd_uncore_format_attr, +}; + +static const struct attribute_group *amd_uncore_attr_groups[] = { + &amd_uncore_attr_group, + &amd_uncore_format_group, + NULL, +}; + +static struct pmu amd_nb_pmu = { + .attr_groups = amd_uncore_attr_groups, + .name = "amd_nb", + .event_init = amd_uncore_event_init, + .add = amd_uncore_add, + .del = amd_uncore_del, + .start = amd_uncore_start, + .stop = amd_uncore_stop, + .read = amd_uncore_read, +}; + +static struct pmu amd_l2_pmu = { + .attr_groups = amd_uncore_attr_groups, + .name = "amd_l2", + .event_init = amd_uncore_event_init, + .add = amd_uncore_add, + .del = amd_uncore_del, + .start = amd_uncore_start, + .stop = amd_uncore_stop, + .read = amd_uncore_read, +}; + +static struct amd_uncore *amd_uncore_alloc(unsigned int cpu) +{ + return kzalloc_node(sizeof(struct amd_uncore), GFP_KERNEL, + cpu_to_node(cpu)); +} + +static void amd_uncore_cpu_up_prepare(unsigned int cpu) +{ + struct amd_uncore *uncore; + + if (amd_uncore_nb) { + uncore = amd_uncore_alloc(cpu); + uncore->cpu = cpu; + uncore->num_counters = NUM_COUNTERS_NB; + uncore->rdpmc_base = RDPMC_BASE_NB; + uncore->msr_base = MSR_F15H_NB_PERF_CTL; + uncore->active_mask = &amd_nb_active_mask; + uncore->pmu = &amd_nb_pmu; + *per_cpu_ptr(amd_uncore_nb, cpu) = uncore; + } + + if (amd_uncore_l2) { + uncore = amd_uncore_alloc(cpu); + uncore->cpu = cpu; + uncore->num_counters = NUM_COUNTERS_L2; + uncore->rdpmc_base = RDPMC_BASE_L2; + uncore->msr_base = MSR_F16H_L2I_PERF_CTL; + uncore->active_mask = &amd_l2_active_mask; + uncore->pmu = &amd_l2_pmu; + *per_cpu_ptr(amd_uncore_l2, cpu) = uncore; + } +} + +static struct amd_uncore * +amd_uncore_find_online_sibling(struct amd_uncore *this, + struct amd_uncore * __percpu *uncores) +{ + unsigned int cpu; + struct amd_uncore *that; + + for_each_online_cpu(cpu) { + that = *per_cpu_ptr(uncores, cpu); + + if (!that) + continue; + + if (this == that) + continue; + + if (this->id == that->id) { + that->free_when_cpu_online = this; + this = that; + break; + } + } + + this->refcnt++; + return this; +} + +static void amd_uncore_cpu_starting(unsigned int cpu) +{ + unsigned int eax, ebx, ecx, edx; + struct amd_uncore *uncore; + + if (amd_uncore_nb) { + uncore = *per_cpu_ptr(amd_uncore_nb, cpu); + cpuid(0x8000001e, &eax, &ebx, &ecx, &edx); + uncore->id = ecx & 0xff; + + uncore = amd_uncore_find_online_sibling(uncore, amd_uncore_nb); + *per_cpu_ptr(amd_uncore_nb, cpu) = uncore; + } + + if (amd_uncore_l2) { + unsigned int apicid = cpu_data(cpu).apicid; + unsigned int nshared; + + uncore = *per_cpu_ptr(amd_uncore_l2, cpu); + cpuid_count(0x8000001d, 2, &eax, &ebx, &ecx, &edx); + nshared = ((eax >> 14) & 0xfff) + 1; + uncore->id = apicid - (apicid % nshared); + + uncore = amd_uncore_find_online_sibling(uncore, amd_uncore_l2); + *per_cpu_ptr(amd_uncore_l2, cpu) = uncore; + } +} + +static void uncore_online(unsigned int cpu, + struct amd_uncore * __percpu *uncores) +{ + struct amd_uncore *uncore = *per_cpu_ptr(uncores, cpu); + + kfree(uncore->free_when_cpu_online); + uncore->free_when_cpu_online = NULL; + + if (cpu == uncore->cpu) + cpumask_set_cpu(cpu, uncore->active_mask); +} + +static void amd_uncore_cpu_online(unsigned int cpu) +{ + if (amd_uncore_nb) + uncore_online(cpu, amd_uncore_nb); + + if (amd_uncore_l2) + uncore_online(cpu, amd_uncore_l2); +} + +static void uncore_down_prepare(unsigned int cpu, + struct amd_uncore * __percpu *uncores) +{ + unsigned int i; + struct amd_uncore *this = *per_cpu_ptr(uncores, cpu); + + if (this->cpu != cpu) + return; + + /* this cpu is going down, migrate to a shared sibling if possible */ + for_each_online_cpu(i) { + struct amd_uncore *that = *per_cpu_ptr(uncores, i); + + if (cpu == i) + continue; + + if (this == that) { + perf_pmu_migrate_context(this->pmu, cpu, i); + cpumask_clear_cpu(cpu, that->active_mask); + cpumask_set_cpu(i, that->active_mask); + that->cpu = i; + break; + } + } +} + +static void amd_uncore_cpu_down_prepare(unsigned int cpu) +{ + if (amd_uncore_nb) + uncore_down_prepare(cpu, amd_uncore_nb); + + if (amd_uncore_l2) + uncore_down_prepare(cpu, amd_uncore_l2); +} + +static void uncore_dead(unsigned int cpu, struct amd_uncore * __percpu *uncores) +{ + struct amd_uncore *uncore = *per_cpu_ptr(uncores, cpu); + + if (cpu == uncore->cpu) + cpumask_clear_cpu(cpu, uncore->active_mask); + + if (!--uncore->refcnt) + kfree(uncore); + *per_cpu_ptr(amd_uncore_nb, cpu) = NULL; +} + +static void amd_uncore_cpu_dead(unsigned int cpu) +{ + if (amd_uncore_nb) + uncore_dead(cpu, amd_uncore_nb); + + if (amd_uncore_l2) + uncore_dead(cpu, amd_uncore_l2); +} + +static int +amd_uncore_cpu_notifier(struct notifier_block *self, unsigned long action, + void *hcpu) +{ + unsigned int cpu = (long)hcpu; + + switch (action & ~CPU_TASKS_FROZEN) { + case CPU_UP_PREPARE: + amd_uncore_cpu_up_prepare(cpu); + break; + + case CPU_STARTING: + amd_uncore_cpu_starting(cpu); + break; + + case CPU_ONLINE: + amd_uncore_cpu_online(cpu); + break; + + case CPU_DOWN_PREPARE: + amd_uncore_cpu_down_prepare(cpu); + break; + + case CPU_UP_CANCELED: + case CPU_DEAD: + amd_uncore_cpu_dead(cpu); + break; + + default: + break; + } + + return NOTIFY_OK; +} + +static struct notifier_block amd_uncore_cpu_notifier_block = { + .notifier_call = amd_uncore_cpu_notifier, + .priority = CPU_PRI_PERF + 1, +}; + +static void __init init_cpu_already_online(void *dummy) +{ + unsigned int cpu = smp_processor_id(); + + amd_uncore_cpu_starting(cpu); + amd_uncore_cpu_online(cpu); +} + +static int __init amd_uncore_init(void) +{ + unsigned int cpu; + int ret = -ENODEV; + + if (boot_cpu_data.x86_vendor != X86_VENDOR_AMD) + return -ENODEV; + + if (!cpu_has_topoext) + return -ENODEV; + + if (cpu_has_perfctr_nb) { + amd_uncore_nb = alloc_percpu(struct amd_uncore *); + perf_pmu_register(&amd_nb_pmu, amd_nb_pmu.name, -1); + + printk(KERN_INFO "perf: AMD NB counters detected\n"); + ret = 0; + } + + if (cpu_has_perfctr_l2) { + amd_uncore_l2 = alloc_percpu(struct amd_uncore *); + perf_pmu_register(&amd_l2_pmu, amd_l2_pmu.name, -1); + + printk(KERN_INFO "perf: AMD L2I counters detected\n"); + ret = 0; + } + + if (ret) + return -ENODEV; + + cpu_notifier_register_begin(); + + /* init cpus already online before registering for hotplug notifier */ + for_each_online_cpu(cpu) { + amd_uncore_cpu_up_prepare(cpu); + smp_call_function_single(cpu, init_cpu_already_online, NULL, 1); + } + + __register_cpu_notifier(&amd_uncore_cpu_notifier_block); + cpu_notifier_register_done(); + + return 0; +} +device_initcall(amd_uncore_init); diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 3bd37bdf1b8..2502d0d9d24 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -5,12 +5,15 @@ * among events on a single PMU. */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/stddef.h> #include <linux/types.h> #include <linux/init.h> #include <linux/slab.h> #include <linux/export.h> +#include <asm/cpufeature.h> #include <asm/hardirq.h> #include <asm/apic.h> @@ -21,14 +24,14 @@ */ static u64 intel_perfmon_event_map[PERF_COUNT_HW_MAX] __read_mostly = { - [PERF_COUNT_HW_CPU_CYCLES] = 0x003c, - [PERF_COUNT_HW_INSTRUCTIONS] = 0x00c0, - [PERF_COUNT_HW_CACHE_REFERENCES] = 0x4f2e, - [PERF_COUNT_HW_CACHE_MISSES] = 0x412e, - [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = 0x00c4, - [PERF_COUNT_HW_BRANCH_MISSES] = 0x00c5, - [PERF_COUNT_HW_BUS_CYCLES] = 0x013c, - [PERF_COUNT_HW_REF_CPU_CYCLES] = 0x0300, /* pseudo-encoding */ + [PERF_COUNT_HW_CPU_CYCLES] = 0x003c, + [PERF_COUNT_HW_INSTRUCTIONS] = 0x00c0, + [PERF_COUNT_HW_CACHE_REFERENCES] = 0x4f2e, + [PERF_COUNT_HW_CACHE_MISSES] = 0x412e, + [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = 0x00c4, + [PERF_COUNT_HW_BRANCH_MISSES] = 0x00c5, + [PERF_COUNT_HW_BUS_CYCLES] = 0x013c, + [PERF_COUNT_HW_REF_CPU_CYCLES] = 0x0300, /* pseudo-encoding */ }; static struct event_constraint intel_core_event_constraints[] __read_mostly = @@ -78,7 +81,9 @@ static struct event_constraint intel_nehalem_event_constraints[] __read_mostly = static struct extra_reg intel_nehalem_extra_regs[] __read_mostly = { - INTEL_EVENT_EXTRA_REG(0xb7, MSR_OFFCORE_RSP_0, 0xffff, RSP_0), + /* must define OFFCORE_RSP_X first, see intel_fixup_er() */ + INTEL_UEVENT_EXTRA_REG(0x01b7, MSR_OFFCORE_RSP_0, 0xffff, RSP_0), + INTEL_UEVENT_PEBS_LDLAT_EXTRA_REG(0x100b), EVENT_EXTRA_END }; @@ -99,16 +104,51 @@ static struct event_constraint intel_snb_event_constraints[] __read_mostly = FIXED_EVENT_CONSTRAINT(0x00c0, 0), /* INST_RETIRED.ANY */ FIXED_EVENT_CONSTRAINT(0x003c, 1), /* CPU_CLK_UNHALTED.CORE */ FIXED_EVENT_CONSTRAINT(0x0300, 2), /* CPU_CLK_UNHALTED.REF */ + INTEL_UEVENT_CONSTRAINT(0x04a3, 0xf), /* CYCLE_ACTIVITY.CYCLES_NO_DISPATCH */ + INTEL_UEVENT_CONSTRAINT(0x05a3, 0xf), /* CYCLE_ACTIVITY.STALLS_L2_PENDING */ + INTEL_UEVENT_CONSTRAINT(0x02a3, 0x4), /* CYCLE_ACTIVITY.CYCLES_L1D_PENDING */ + INTEL_UEVENT_CONSTRAINT(0x06a3, 0x4), /* CYCLE_ACTIVITY.STALLS_L1D_PENDING */ INTEL_EVENT_CONSTRAINT(0x48, 0x4), /* L1D_PEND_MISS.PENDING */ INTEL_UEVENT_CONSTRAINT(0x01c0, 0x2), /* INST_RETIRED.PREC_DIST */ INTEL_EVENT_CONSTRAINT(0xcd, 0x8), /* MEM_TRANS_RETIRED.LOAD_LATENCY */ + INTEL_UEVENT_CONSTRAINT(0x04a3, 0xf), /* CYCLE_ACTIVITY.CYCLES_NO_DISPATCH */ + INTEL_UEVENT_CONSTRAINT(0x02a3, 0x4), /* CYCLE_ACTIVITY.CYCLES_L1D_PENDING */ + EVENT_CONSTRAINT_END +}; + +static struct event_constraint intel_ivb_event_constraints[] __read_mostly = +{ + FIXED_EVENT_CONSTRAINT(0x00c0, 0), /* INST_RETIRED.ANY */ + FIXED_EVENT_CONSTRAINT(0x003c, 1), /* CPU_CLK_UNHALTED.CORE */ + FIXED_EVENT_CONSTRAINT(0x0300, 2), /* CPU_CLK_UNHALTED.REF */ + INTEL_UEVENT_CONSTRAINT(0x0148, 0x4), /* L1D_PEND_MISS.PENDING */ + INTEL_UEVENT_CONSTRAINT(0x0279, 0xf), /* IDQ.EMTPY */ + INTEL_UEVENT_CONSTRAINT(0x019c, 0xf), /* IDQ_UOPS_NOT_DELIVERED.CORE */ + INTEL_UEVENT_CONSTRAINT(0x02a3, 0xf), /* CYCLE_ACTIVITY.CYCLES_LDM_PENDING */ + INTEL_UEVENT_CONSTRAINT(0x04a3, 0xf), /* CYCLE_ACTIVITY.CYCLES_NO_EXECUTE */ + INTEL_UEVENT_CONSTRAINT(0x05a3, 0xf), /* CYCLE_ACTIVITY.STALLS_L2_PENDING */ + INTEL_UEVENT_CONSTRAINT(0x06a3, 0xf), /* CYCLE_ACTIVITY.STALLS_LDM_PENDING */ + INTEL_UEVENT_CONSTRAINT(0x08a3, 0x4), /* CYCLE_ACTIVITY.CYCLES_L1D_PENDING */ + INTEL_UEVENT_CONSTRAINT(0x0ca3, 0x4), /* CYCLE_ACTIVITY.STALLS_L1D_PENDING */ + INTEL_UEVENT_CONSTRAINT(0x01c0, 0x2), /* INST_RETIRED.PREC_DIST */ + /* + * Errata BV98 -- MEM_*_RETIRED events can leak between counters of SMT + * siblings; disable these events because they can corrupt unrelated + * counters. + */ + INTEL_EVENT_CONSTRAINT(0xd0, 0x0), /* MEM_UOPS_RETIRED.* */ + INTEL_EVENT_CONSTRAINT(0xd1, 0x0), /* MEM_LOAD_UOPS_RETIRED.* */ + INTEL_EVENT_CONSTRAINT(0xd2, 0x0), /* MEM_LOAD_UOPS_LLC_HIT_RETIRED.* */ + INTEL_EVENT_CONSTRAINT(0xd3, 0x0), /* MEM_LOAD_UOPS_LLC_MISS_RETIRED.* */ EVENT_CONSTRAINT_END }; static struct extra_reg intel_westmere_extra_regs[] __read_mostly = { - INTEL_EVENT_EXTRA_REG(0xb7, MSR_OFFCORE_RSP_0, 0xffff, RSP_0), - INTEL_EVENT_EXTRA_REG(0xbb, MSR_OFFCORE_RSP_1, 0xffff, RSP_1), + /* must define OFFCORE_RSP_X first, see intel_fixup_er() */ + INTEL_UEVENT_EXTRA_REG(0x01b7, MSR_OFFCORE_RSP_0, 0xffff, RSP_0), + INTEL_UEVENT_EXTRA_REG(0x01bb, MSR_OFFCORE_RSP_1, 0xffff, RSP_1), + INTEL_UEVENT_PEBS_LDLAT_EXTRA_REG(0x100b), EVENT_EXTRA_END }; @@ -125,17 +165,144 @@ static struct event_constraint intel_gen_event_constraints[] __read_mostly = EVENT_CONSTRAINT_END }; +static struct event_constraint intel_slm_event_constraints[] __read_mostly = +{ + FIXED_EVENT_CONSTRAINT(0x00c0, 0), /* INST_RETIRED.ANY */ + FIXED_EVENT_CONSTRAINT(0x003c, 1), /* CPU_CLK_UNHALTED.CORE */ + FIXED_EVENT_CONSTRAINT(0x0300, 2), /* pseudo CPU_CLK_UNHALTED.REF */ + EVENT_CONSTRAINT_END +}; + static struct extra_reg intel_snb_extra_regs[] __read_mostly = { - INTEL_EVENT_EXTRA_REG(0xb7, MSR_OFFCORE_RSP_0, 0x3fffffffffull, RSP_0), - INTEL_EVENT_EXTRA_REG(0xbb, MSR_OFFCORE_RSP_1, 0x3fffffffffull, RSP_1), + /* must define OFFCORE_RSP_X first, see intel_fixup_er() */ + INTEL_UEVENT_EXTRA_REG(0x01b7, MSR_OFFCORE_RSP_0, 0x3f807f8fffull, RSP_0), + INTEL_UEVENT_EXTRA_REG(0x01bb, MSR_OFFCORE_RSP_1, 0x3f807f8fffull, RSP_1), + INTEL_UEVENT_PEBS_LDLAT_EXTRA_REG(0x01cd), EVENT_EXTRA_END }; +static struct extra_reg intel_snbep_extra_regs[] __read_mostly = { + /* must define OFFCORE_RSP_X first, see intel_fixup_er() */ + INTEL_UEVENT_EXTRA_REG(0x01b7, MSR_OFFCORE_RSP_0, 0x3fffff8fffull, RSP_0), + INTEL_UEVENT_EXTRA_REG(0x01bb, MSR_OFFCORE_RSP_1, 0x3fffff8fffull, RSP_1), + INTEL_UEVENT_PEBS_LDLAT_EXTRA_REG(0x01cd), + EVENT_EXTRA_END +}; + +EVENT_ATTR_STR(mem-loads, mem_ld_nhm, "event=0x0b,umask=0x10,ldlat=3"); +EVENT_ATTR_STR(mem-loads, mem_ld_snb, "event=0xcd,umask=0x1,ldlat=3"); +EVENT_ATTR_STR(mem-stores, mem_st_snb, "event=0xcd,umask=0x2"); + +struct attribute *nhm_events_attrs[] = { + EVENT_PTR(mem_ld_nhm), + NULL, +}; + +struct attribute *snb_events_attrs[] = { + EVENT_PTR(mem_ld_snb), + EVENT_PTR(mem_st_snb), + NULL, +}; + +static struct event_constraint intel_hsw_event_constraints[] = { + FIXED_EVENT_CONSTRAINT(0x00c0, 0), /* INST_RETIRED.ANY */ + FIXED_EVENT_CONSTRAINT(0x003c, 1), /* CPU_CLK_UNHALTED.CORE */ + FIXED_EVENT_CONSTRAINT(0x0300, 2), /* CPU_CLK_UNHALTED.REF */ + INTEL_EVENT_CONSTRAINT(0x48, 0x4), /* L1D_PEND_MISS.* */ + INTEL_UEVENT_CONSTRAINT(0x01c0, 0x2), /* INST_RETIRED.PREC_DIST */ + INTEL_EVENT_CONSTRAINT(0xcd, 0x8), /* MEM_TRANS_RETIRED.LOAD_LATENCY */ + /* CYCLE_ACTIVITY.CYCLES_L1D_PENDING */ + INTEL_EVENT_CONSTRAINT(0x08a3, 0x4), + /* CYCLE_ACTIVITY.STALLS_L1D_PENDING */ + INTEL_EVENT_CONSTRAINT(0x0ca3, 0x4), + /* CYCLE_ACTIVITY.CYCLES_NO_EXECUTE */ + INTEL_EVENT_CONSTRAINT(0x04a3, 0xf), + EVENT_CONSTRAINT_END +}; + static u64 intel_pmu_event_map(int hw_event) { return intel_perfmon_event_map[hw_event]; } +#define SNB_DMND_DATA_RD (1ULL << 0) +#define SNB_DMND_RFO (1ULL << 1) +#define SNB_DMND_IFETCH (1ULL << 2) +#define SNB_DMND_WB (1ULL << 3) +#define SNB_PF_DATA_RD (1ULL << 4) +#define SNB_PF_RFO (1ULL << 5) +#define SNB_PF_IFETCH (1ULL << 6) +#define SNB_LLC_DATA_RD (1ULL << 7) +#define SNB_LLC_RFO (1ULL << 8) +#define SNB_LLC_IFETCH (1ULL << 9) +#define SNB_BUS_LOCKS (1ULL << 10) +#define SNB_STRM_ST (1ULL << 11) +#define SNB_OTHER (1ULL << 15) +#define SNB_RESP_ANY (1ULL << 16) +#define SNB_NO_SUPP (1ULL << 17) +#define SNB_LLC_HITM (1ULL << 18) +#define SNB_LLC_HITE (1ULL << 19) +#define SNB_LLC_HITS (1ULL << 20) +#define SNB_LLC_HITF (1ULL << 21) +#define SNB_LOCAL (1ULL << 22) +#define SNB_REMOTE (0xffULL << 23) +#define SNB_SNP_NONE (1ULL << 31) +#define SNB_SNP_NOT_NEEDED (1ULL << 32) +#define SNB_SNP_MISS (1ULL << 33) +#define SNB_NO_FWD (1ULL << 34) +#define SNB_SNP_FWD (1ULL << 35) +#define SNB_HITM (1ULL << 36) +#define SNB_NON_DRAM (1ULL << 37) + +#define SNB_DMND_READ (SNB_DMND_DATA_RD|SNB_LLC_DATA_RD) +#define SNB_DMND_WRITE (SNB_DMND_RFO|SNB_LLC_RFO) +#define SNB_DMND_PREFETCH (SNB_PF_DATA_RD|SNB_PF_RFO) + +#define SNB_SNP_ANY (SNB_SNP_NONE|SNB_SNP_NOT_NEEDED| \ + SNB_SNP_MISS|SNB_NO_FWD|SNB_SNP_FWD| \ + SNB_HITM) + +#define SNB_DRAM_ANY (SNB_LOCAL|SNB_REMOTE|SNB_SNP_ANY) +#define SNB_DRAM_REMOTE (SNB_REMOTE|SNB_SNP_ANY) + +#define SNB_L3_ACCESS SNB_RESP_ANY +#define SNB_L3_MISS (SNB_DRAM_ANY|SNB_NON_DRAM) + +static __initconst const u64 snb_hw_cache_extra_regs + [PERF_COUNT_HW_CACHE_MAX] + [PERF_COUNT_HW_CACHE_OP_MAX] + [PERF_COUNT_HW_CACHE_RESULT_MAX] = +{ + [ C(LL ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = SNB_DMND_READ|SNB_L3_ACCESS, + [ C(RESULT_MISS) ] = SNB_DMND_READ|SNB_L3_MISS, + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = SNB_DMND_WRITE|SNB_L3_ACCESS, + [ C(RESULT_MISS) ] = SNB_DMND_WRITE|SNB_L3_MISS, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = SNB_DMND_PREFETCH|SNB_L3_ACCESS, + [ C(RESULT_MISS) ] = SNB_DMND_PREFETCH|SNB_L3_MISS, + }, + }, + [ C(NODE) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = SNB_DMND_READ|SNB_DRAM_ANY, + [ C(RESULT_MISS) ] = SNB_DMND_READ|SNB_DRAM_REMOTE, + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = SNB_DMND_WRITE|SNB_DRAM_ANY, + [ C(RESULT_MISS) ] = SNB_DMND_WRITE|SNB_DRAM_REMOTE, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = SNB_DMND_PREFETCH|SNB_DRAM_ANY, + [ C(RESULT_MISS) ] = SNB_DMND_PREFETCH|SNB_DRAM_REMOTE, + }, + }, +}; + static __initconst const u64 snb_hw_cache_event_ids [PERF_COUNT_HW_CACHE_MAX] [PERF_COUNT_HW_CACHE_OP_MAX] @@ -233,16 +400,16 @@ static __initconst const u64 snb_hw_cache_event_ids }, [ C(NODE) ] = { [ C(OP_READ) ] = { - [ C(RESULT_ACCESS) ] = -1, - [ C(RESULT_MISS) ] = -1, + [ C(RESULT_ACCESS) ] = 0x01b7, + [ C(RESULT_MISS) ] = 0x01b7, }, [ C(OP_WRITE) ] = { - [ C(RESULT_ACCESS) ] = -1, - [ C(RESULT_MISS) ] = -1, + [ C(RESULT_ACCESS) ] = 0x01b7, + [ C(RESULT_MISS) ] = 0x01b7, }, [ C(OP_PREFETCH) ] = { - [ C(RESULT_ACCESS) ] = -1, - [ C(RESULT_MISS) ] = -1, + [ C(RESULT_ACCESS) ] = 0x01b7, + [ C(RESULT_MISS) ] = 0x01b7, }, }, @@ -385,14 +552,15 @@ static __initconst const u64 westmere_hw_cache_event_ids #define NHM_LOCAL_DRAM (1 << 14) #define NHM_NON_DRAM (1 << 15) -#define NHM_ALL_DRAM (NHM_REMOTE_DRAM|NHM_LOCAL_DRAM) +#define NHM_LOCAL (NHM_LOCAL_DRAM|NHM_REMOTE_CACHE_FWD) +#define NHM_REMOTE (NHM_REMOTE_DRAM) #define NHM_DMND_READ (NHM_DMND_DATA_RD) #define NHM_DMND_WRITE (NHM_DMND_RFO|NHM_DMND_WB) #define NHM_DMND_PREFETCH (NHM_PF_DATA_RD|NHM_PF_DATA_RFO) #define NHM_L3_HIT (NHM_UNCORE_HIT|NHM_OTHER_CORE_HIT_SNP|NHM_OTHER_CORE_HITM) -#define NHM_L3_MISS (NHM_NON_DRAM|NHM_ALL_DRAM|NHM_REMOTE_CACHE_FWD) +#define NHM_L3_MISS (NHM_NON_DRAM|NHM_LOCAL_DRAM|NHM_REMOTE_DRAM|NHM_REMOTE_CACHE_FWD) #define NHM_L3_ACCESS (NHM_L3_HIT|NHM_L3_MISS) static __initconst const u64 nehalem_hw_cache_extra_regs @@ -416,16 +584,16 @@ static __initconst const u64 nehalem_hw_cache_extra_regs }, [ C(NODE) ] = { [ C(OP_READ) ] = { - [ C(RESULT_ACCESS) ] = NHM_DMND_READ|NHM_ALL_DRAM, - [ C(RESULT_MISS) ] = NHM_DMND_READ|NHM_REMOTE_DRAM, + [ C(RESULT_ACCESS) ] = NHM_DMND_READ|NHM_LOCAL|NHM_REMOTE, + [ C(RESULT_MISS) ] = NHM_DMND_READ|NHM_REMOTE, }, [ C(OP_WRITE) ] = { - [ C(RESULT_ACCESS) ] = NHM_DMND_WRITE|NHM_ALL_DRAM, - [ C(RESULT_MISS) ] = NHM_DMND_WRITE|NHM_REMOTE_DRAM, + [ C(RESULT_ACCESS) ] = NHM_DMND_WRITE|NHM_LOCAL|NHM_REMOTE, + [ C(RESULT_MISS) ] = NHM_DMND_WRITE|NHM_REMOTE, }, [ C(OP_PREFETCH) ] = { - [ C(RESULT_ACCESS) ] = NHM_DMND_PREFETCH|NHM_ALL_DRAM, - [ C(RESULT_MISS) ] = NHM_DMND_PREFETCH|NHM_REMOTE_DRAM, + [ C(RESULT_ACCESS) ] = NHM_DMND_PREFETCH|NHM_LOCAL|NHM_REMOTE, + [ C(RESULT_MISS) ] = NHM_DMND_PREFETCH|NHM_REMOTE, }, }, }; @@ -727,13 +895,161 @@ static __initconst const u64 atom_hw_cache_event_ids }, }; +static struct extra_reg intel_slm_extra_regs[] __read_mostly = +{ + /* must define OFFCORE_RSP_X first, see intel_fixup_er() */ + INTEL_UEVENT_EXTRA_REG(0x01b7, MSR_OFFCORE_RSP_0, 0x768005ffffull, RSP_0), + INTEL_UEVENT_EXTRA_REG(0x02b7, MSR_OFFCORE_RSP_1, 0x768005ffffull, RSP_1), + EVENT_EXTRA_END +}; + +#define SLM_DMND_READ SNB_DMND_DATA_RD +#define SLM_DMND_WRITE SNB_DMND_RFO +#define SLM_DMND_PREFETCH (SNB_PF_DATA_RD|SNB_PF_RFO) + +#define SLM_SNP_ANY (SNB_SNP_NONE|SNB_SNP_MISS|SNB_NO_FWD|SNB_HITM) +#define SLM_LLC_ACCESS SNB_RESP_ANY +#define SLM_LLC_MISS (SLM_SNP_ANY|SNB_NON_DRAM) + +static __initconst const u64 slm_hw_cache_extra_regs + [PERF_COUNT_HW_CACHE_MAX] + [PERF_COUNT_HW_CACHE_OP_MAX] + [PERF_COUNT_HW_CACHE_RESULT_MAX] = +{ + [ C(LL ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = SLM_DMND_READ|SLM_LLC_ACCESS, + [ C(RESULT_MISS) ] = SLM_DMND_READ|SLM_LLC_MISS, + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = SLM_DMND_WRITE|SLM_LLC_ACCESS, + [ C(RESULT_MISS) ] = SLM_DMND_WRITE|SLM_LLC_MISS, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = SLM_DMND_PREFETCH|SLM_LLC_ACCESS, + [ C(RESULT_MISS) ] = SLM_DMND_PREFETCH|SLM_LLC_MISS, + }, + }, +}; + +static __initconst const u64 slm_hw_cache_event_ids + [PERF_COUNT_HW_CACHE_MAX] + [PERF_COUNT_HW_CACHE_OP_MAX] + [PERF_COUNT_HW_CACHE_RESULT_MAX] = +{ + [ C(L1D) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0x0104, /* LD_DCU_MISS */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(L1I ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0380, /* ICACHE.ACCESSES */ + [ C(RESULT_MISS) ] = 0x0280, /* ICACGE.MISSES */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(LL ) ] = { + [ C(OP_READ) ] = { + /* OFFCORE_RESPONSE.ANY_DATA.LOCAL_CACHE */ + [ C(RESULT_ACCESS) ] = 0x01b7, + /* OFFCORE_RESPONSE.ANY_DATA.ANY_LLC_MISS */ + [ C(RESULT_MISS) ] = 0x01b7, + }, + [ C(OP_WRITE) ] = { + /* OFFCORE_RESPONSE.ANY_RFO.LOCAL_CACHE */ + [ C(RESULT_ACCESS) ] = 0x01b7, + /* OFFCORE_RESPONSE.ANY_RFO.ANY_LLC_MISS */ + [ C(RESULT_MISS) ] = 0x01b7, + }, + [ C(OP_PREFETCH) ] = { + /* OFFCORE_RESPONSE.PREFETCH.LOCAL_CACHE */ + [ C(RESULT_ACCESS) ] = 0x01b7, + /* OFFCORE_RESPONSE.PREFETCH.ANY_LLC_MISS */ + [ C(RESULT_MISS) ] = 0x01b7, + }, + }, + [ C(DTLB) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0x0804, /* LD_DTLB_MISS */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(ITLB) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x00c0, /* INST_RETIRED.ANY_P */ + [ C(RESULT_MISS) ] = 0x0282, /* ITLB.MISSES */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + }, + [ C(BPU ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x00c4, /* BR_INST_RETIRED.ANY */ + [ C(RESULT_MISS) ] = 0x00c5, /* BP_INST_RETIRED.MISPRED */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + }, +}; + +static inline bool intel_pmu_needs_lbr_smpl(struct perf_event *event) +{ + /* user explicitly requested branch sampling */ + if (has_branch_stack(event)) + return true; + + /* implicit branch sampling to correct PEBS skid */ + if (x86_pmu.intel_cap.pebs_trap && event->attr.precise_ip > 1 && + x86_pmu.intel_cap.pebs_format < 2) + return true; + + return false; +} + static void intel_pmu_disable_all(void) { struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, 0); - if (test_bit(X86_PMC_IDX_FIXED_BTS, cpuc->active_mask)) + if (test_bit(INTEL_PMC_IDX_FIXED_BTS, cpuc->active_mask)) intel_pmu_disable_bts(); intel_pmu_pebs_disable_all(); @@ -749,9 +1065,9 @@ static void intel_pmu_enable_all(int added) wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, x86_pmu.intel_ctrl & ~cpuc->intel_ctrl_guest_mask); - if (test_bit(X86_PMC_IDX_FIXED_BTS, cpuc->active_mask)) { + if (test_bit(INTEL_PMC_IDX_FIXED_BTS, cpuc->active_mask)) { struct perf_event *event = - cpuc->events[X86_PMC_IDX_FIXED_BTS]; + cpuc->events[INTEL_PMC_IDX_FIXED_BTS]; if (WARN_ON_ONCE(!event)) return; @@ -857,7 +1173,7 @@ static inline void intel_pmu_ack_status(u64 ack) static void intel_pmu_disable_fixed(struct hw_perf_event *hwc) { - int idx = hwc->idx - X86_PMC_IDX_FIXED; + int idx = hwc->idx - INTEL_PMC_IDX_FIXED; u64 ctrl_val, mask; mask = 0xfULL << (idx * 4); @@ -867,12 +1183,17 @@ static void intel_pmu_disable_fixed(struct hw_perf_event *hwc) wrmsrl(hwc->config_base, ctrl_val); } +static inline bool event_is_checkpointed(struct perf_event *event) +{ + return (event->hw.config & HSW_IN_TX_CHECKPOINTED) != 0; +} + static void intel_pmu_disable_event(struct perf_event *event) { struct hw_perf_event *hwc = &event->hw; struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); - if (unlikely(hwc->idx == X86_PMC_IDX_FIXED_BTS)) { + if (unlikely(hwc->idx == INTEL_PMC_IDX_FIXED_BTS)) { intel_pmu_disable_bts(); intel_pmu_drain_bts_buffer(); return; @@ -880,6 +1201,14 @@ static void intel_pmu_disable_event(struct perf_event *event) cpuc->intel_ctrl_guest_mask &= ~(1ull << hwc->idx); cpuc->intel_ctrl_host_mask &= ~(1ull << hwc->idx); + cpuc->intel_cp_status &= ~(1ull << hwc->idx); + + /* + * must disable before any actual event + * because any event may be combined with LBR + */ + if (intel_pmu_needs_lbr_smpl(event)) + intel_pmu_lbr_disable(event); if (unlikely(hwc->config_base == MSR_ARCH_PERFMON_FIXED_CTR_CTRL)) { intel_pmu_disable_fixed(hwc); @@ -894,7 +1223,7 @@ static void intel_pmu_disable_event(struct perf_event *event) static void intel_pmu_enable_fixed(struct hw_perf_event *hwc) { - int idx = hwc->idx - X86_PMC_IDX_FIXED; + int idx = hwc->idx - INTEL_PMC_IDX_FIXED; u64 ctrl_val, bits, mask; /* @@ -928,19 +1257,28 @@ static void intel_pmu_enable_event(struct perf_event *event) struct hw_perf_event *hwc = &event->hw; struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); - if (unlikely(hwc->idx == X86_PMC_IDX_FIXED_BTS)) { + if (unlikely(hwc->idx == INTEL_PMC_IDX_FIXED_BTS)) { if (!__this_cpu_read(cpu_hw_events.enabled)) return; intel_pmu_enable_bts(hwc->config); return; } + /* + * must enabled before any actual event + * because any event may be combined with LBR + */ + if (intel_pmu_needs_lbr_smpl(event)) + intel_pmu_lbr_enable(event); if (event->attr.exclude_host) cpuc->intel_ctrl_guest_mask |= (1ull << hwc->idx); if (event->attr.exclude_guest) cpuc->intel_ctrl_host_mask |= (1ull << hwc->idx); + if (unlikely(event_is_checkpointed(event))) + cpuc->intel_cp_status |= (1ull << hwc->idx); + if (unlikely(hwc->config_base == MSR_ARCH_PERFMON_FIXED_CTR_CTRL)) { intel_pmu_enable_fixed(hwc); return; @@ -959,6 +1297,17 @@ static void intel_pmu_enable_event(struct perf_event *event) int intel_pmu_save_and_restart(struct perf_event *event) { x86_perf_event_update(event); + /* + * For a checkpointed counter always reset back to 0. This + * avoids a situation where the counter overflows, aborts the + * transaction and is then set back to shortly before the + * overflow, and overflows and aborts again. + */ + if (unlikely(event_is_checkpointed(event))) { + /* No race with NMIs because the counter should not be armed */ + wrmsrl(event->hw.event_base, 0); + local64_set(&event->hw.prev_count, 0); + } return x86_perf_event_set_period(event); } @@ -973,14 +1322,14 @@ static void intel_pmu_reset(void) local_irq_save(flags); - printk("clearing PMU state on CPU#%d\n", smp_processor_id()); + pr_info("clearing PMU state on CPU#%d\n", smp_processor_id()); for (idx = 0; idx < x86_pmu.num_counters; idx++) { - checking_wrmsrl(x86_pmu_config_addr(idx), 0ull); - checking_wrmsrl(x86_pmu_event_addr(idx), 0ull); + wrmsrl_safe(x86_pmu_config_addr(idx), 0ull); + wrmsrl_safe(x86_pmu_event_addr(idx), 0ull); } for (idx = 0; idx < x86_pmu.num_counters_fixed; idx++) - checking_wrmsrl(MSR_ARCH_PERFMON_FIXED_CTR0 + idx, 0ull); + wrmsrl_safe(MSR_ARCH_PERFMON_FIXED_CTR0 + idx, 0ull); if (ds) ds->bts_index = ds->bts_buffer_base; @@ -1000,34 +1349,30 @@ static int intel_pmu_handle_irq(struct pt_regs *regs) u64 status; int handled; - perf_sample_data_init(&data, 0); - cpuc = &__get_cpu_var(cpu_hw_events); /* - * Some chipsets need to unmask the LVTPC in a particular spot - * inside the nmi handler. As a result, the unmasking was pushed - * into all the nmi handlers. - * - * This handler doesn't seem to have any issues with the unmasking - * so it was left at the top. + * No known reason to not always do late ACK, + * but just in case do it opt-in. */ - apic_write(APIC_LVTPC, APIC_DM_NMI); - + if (!x86_pmu.late_ack) + apic_write(APIC_LVTPC, APIC_DM_NMI); intel_pmu_disable_all(); handled = intel_pmu_drain_bts_buffer(); status = intel_pmu_get_status(); - if (!status) { - intel_pmu_enable_all(0); - return handled; - } + if (!status) + goto done; loops = 0; again: intel_pmu_ack_status(status); if (++loops > 100) { - WARN_ONCE(1, "perfevents: irq loop stuck!\n"); - perf_event_print_debug(); + static bool warned = false; + if (!warned) { + WARN(1, "perfevents: irq loop stuck!\n"); + perf_event_print_debug(); + warned = true; + } intel_pmu_reset(); goto done; } @@ -1037,6 +1382,15 @@ again: intel_pmu_lbr_read(); /* + * CondChgd bit 63 doesn't mean any overflow status. Ignore + * and clear the bit. + */ + if (__test_and_clear_bit(63, (unsigned long *)&status)) { + if (!status) + goto done; + } + + /* * PEBS overflow sets bit 62 in the global status register */ if (__test_and_clear_bit(62, (unsigned long *)&status)) { @@ -1044,6 +1398,13 @@ again: x86_pmu.drain_pebs(regs); } + /* + * Checkpointed counters can lead to 'spurious' PMIs because the + * rollback caused by the PMI will have cleared the overflow status + * bit. Therefore always force probe these counters. + */ + status |= cpuc->intel_cp_status; + for_each_set_bit(bit, (unsigned long *)&status, X86_PMC_IDX_MAX) { struct perf_event *event = cpuc->events[bit]; @@ -1055,7 +1416,10 @@ again: if (!intel_pmu_save_and_restart(event)) continue; - data.period = event->hw.last_period; + perf_sample_data_init(&data, 0, event->hw.last_period); + + if (has_branch_stack(event)) + data.br_stack = &cpuc->lbr_stack; if (perf_event_overflow(event, &data, regs)) x86_pmu_stop(event, 0); @@ -1070,6 +1434,13 @@ again: done: intel_pmu_enable_all(0); + /* + * Only unmask the NMI after the overflow counters + * have been reset. This avoids spurious NMIs on + * Haswell CPUs. + */ + if (x86_pmu.late_ack) + apic_write(APIC_LVTPC, APIC_DM_NMI); return handled; } @@ -1091,27 +1462,33 @@ intel_bts_constraints(struct perf_event *event) return NULL; } -static bool intel_try_alt_er(struct perf_event *event, int orig_idx) +static int intel_alt_er(int idx) { if (!(x86_pmu.er_flags & ERF_HAS_RSP_1)) - return false; + return idx; - if (event->hw.extra_reg.idx == EXTRA_REG_RSP_0) { - event->hw.config &= ~INTEL_ARCH_EVENT_MASK; - event->hw.config |= 0x01bb; - event->hw.extra_reg.idx = EXTRA_REG_RSP_1; - event->hw.extra_reg.reg = MSR_OFFCORE_RSP_1; - } else if (event->hw.extra_reg.idx == EXTRA_REG_RSP_1) { + if (idx == EXTRA_REG_RSP_0) + return EXTRA_REG_RSP_1; + + if (idx == EXTRA_REG_RSP_1) + return EXTRA_REG_RSP_0; + + return idx; +} + +static void intel_fixup_er(struct perf_event *event, int idx) +{ + event->hw.extra_reg.idx = idx; + + if (idx == EXTRA_REG_RSP_0) { event->hw.config &= ~INTEL_ARCH_EVENT_MASK; - event->hw.config |= 0x01b7; - event->hw.extra_reg.idx = EXTRA_REG_RSP_0; + event->hw.config |= x86_pmu.extra_regs[EXTRA_REG_RSP_0].event; event->hw.extra_reg.reg = MSR_OFFCORE_RSP_0; + } else if (idx == EXTRA_REG_RSP_1) { + event->hw.config &= ~INTEL_ARCH_EVENT_MASK; + event->hw.config |= x86_pmu.extra_regs[EXTRA_REG_RSP_1].event; + event->hw.extra_reg.reg = MSR_OFFCORE_RSP_1; } - - if (event->hw.extra_reg.idx == orig_idx) - return false; - - return true; } /* @@ -1123,20 +1500,24 @@ static bool intel_try_alt_er(struct perf_event *event, int orig_idx) */ static struct event_constraint * __intel_shared_reg_get_constraints(struct cpu_hw_events *cpuc, - struct perf_event *event) + struct perf_event *event, + struct hw_perf_event_extra *reg) { struct event_constraint *c = &emptyconstraint; - struct hw_perf_event_extra *reg = &event->hw.extra_reg; struct er_account *era; unsigned long flags; - int orig_idx = reg->idx; + int idx = reg->idx; - /* already allocated shared msr */ - if (reg->alloc) - return &unconstrained; + /* + * reg->alloc can be set due to existing state, so for fake cpuc we + * need to ignore this, otherwise we might fail to allocate proper fake + * state for this extra reg constraint. Also see the comment below. + */ + if (reg->alloc && !cpuc->is_fake) + return NULL; /* call x86_get_event_constraint() */ again: - era = &cpuc->shared_regs->regs[reg->idx]; + era = &cpuc->shared_regs->regs[idx]; /* * we use spin_lock_irqsave() to avoid lockdep issues when * passing a fake cpuc @@ -1145,6 +1526,29 @@ again: if (!atomic_read(&era->ref) || era->config == reg->config) { + /* + * If its a fake cpuc -- as per validate_{group,event}() we + * shouldn't touch event state and we can avoid doing so + * since both will only call get_event_constraints() once + * on each event, this avoids the need for reg->alloc. + * + * Not doing the ER fixup will only result in era->reg being + * wrong, but since we won't actually try and program hardware + * this isn't a problem either. + */ + if (!cpuc->is_fake) { + if (idx != reg->idx) + intel_fixup_er(event, idx); + + /* + * x86_schedule_events() can call get_event_constraints() + * multiple times on events in the case of incremental + * scheduling(). reg->alloc ensures we only do the ER + * allocation once. + */ + reg->alloc = 1; + } + /* lock in msr value */ era->config = reg->config; era->reg = reg->reg; @@ -1152,21 +1556,17 @@ again: /* one more user */ atomic_inc(&era->ref); - /* no need to reallocate during incremental event scheduling */ - reg->alloc = 1; - /* - * All events using extra_reg are unconstrained. - * Avoids calling x86_get_event_constraints() - * - * Must revisit if extra_reg controlling events - * ever have constraints. Worst case we go through - * the regular event constraint table. + * need to call x86_get_event_constraint() + * to check if associated event has constraints */ - c = &unconstrained; - } else if (intel_try_alt_er(event, orig_idx)) { - raw_spin_unlock_irqrestore(&era->lock, flags); - goto again; + c = NULL; + } else { + idx = intel_alt_er(idx); + if (idx != reg->idx) { + raw_spin_unlock_irqrestore(&era->lock, flags); + goto again; + } } raw_spin_unlock_irqrestore(&era->lock, flags); @@ -1180,11 +1580,14 @@ __intel_shared_reg_put_constraints(struct cpu_hw_events *cpuc, struct er_account *era; /* - * only put constraint if extra reg was actually - * allocated. Also takes care of event which do - * not use an extra shared reg + * Only put constraint if extra reg was actually allocated. Also takes + * care of event which do not use an extra shared reg. + * + * Also, if this is a fake cpuc we shouldn't touch any event state + * (reg->alloc) and we don't care about leaving inconsistent cpuc state + * either since it'll be thrown out. */ - if (!reg->alloc) + if (!reg->alloc || cpuc->is_fake) return; era = &cpuc->shared_regs->regs[reg->idx]; @@ -1200,11 +1603,23 @@ static struct event_constraint * intel_shared_regs_constraints(struct cpu_hw_events *cpuc, struct perf_event *event) { - struct event_constraint *c = NULL; - - if (event->hw.extra_reg.idx != EXTRA_REG_NONE) - c = __intel_shared_reg_get_constraints(cpuc, event); - + struct event_constraint *c = NULL, *d; + struct hw_perf_event_extra *xreg, *breg; + + xreg = &event->hw.extra_reg; + if (xreg->idx != EXTRA_REG_NONE) { + c = __intel_shared_reg_get_constraints(cpuc, event, xreg); + if (c == &emptyconstraint) + return c; + } + breg = &event->hw.branch_reg; + if (breg->idx != EXTRA_REG_NONE) { + d = __intel_shared_reg_get_constraints(cpuc, event, breg); + if (d == &emptyconstraint) { + __intel_shared_reg_put_constraints(cpuc, xreg); + c = d; + } + } return c; } @@ -1215,8 +1630,10 @@ x86_get_event_constraints(struct cpu_hw_events *cpuc, struct perf_event *event) if (x86_pmu.event_constraints) { for_each_event_constraint(c, x86_pmu.event_constraints) { - if ((event->hw.config & c->cmask) == c->code) + if ((event->hw.config & c->cmask) == c->code) { + event->hw.flags |= c->flags; return c; + } } } @@ -1252,6 +1669,10 @@ intel_put_shared_regs_event_constraints(struct cpu_hw_events *cpuc, reg = &event->hw.extra_reg; if (reg->idx != EXTRA_REG_NONE) __intel_shared_reg_put_constraints(cpuc, reg); + + reg = &event->hw.branch_reg; + if (reg->idx != EXTRA_REG_NONE) + __intel_shared_reg_put_constraints(cpuc, reg); } static void intel_put_event_constraints(struct cpu_hw_events *cpuc, @@ -1260,15 +1681,9 @@ static void intel_put_event_constraints(struct cpu_hw_events *cpuc, intel_put_shared_regs_event_constraints(cpuc, event); } -static int intel_pmu_hw_config(struct perf_event *event) +static void intel_pebs_aliases_core2(struct perf_event *event) { - int ret = x86_pmu_hw_config(event); - - if (ret) - return ret; - - if (event->attr.precise_ip && - (event->hw.config & X86_RAW_EVENT_MASK) == 0x003c) { + if ((event->hw.config & X86_RAW_EVENT_MASK) == 0x003c) { /* * Use an alternative encoding for CPU_CLK_UNHALTED.THREAD_P * (0x003c) so that we can use it with PEBS. @@ -1287,11 +1702,56 @@ static int intel_pmu_hw_config(struct perf_event *event) * * Thereby we gain a PEBS capable cycle counter. */ - u64 alt_config = 0x108000c0; /* INST_RETIRED.TOTAL_CYCLES */ + u64 alt_config = X86_CONFIG(.event=0xc0, .inv=1, .cmask=16); + + alt_config |= (event->hw.config & ~X86_RAW_EVENT_MASK); + event->hw.config = alt_config; + } +} + +static void intel_pebs_aliases_snb(struct perf_event *event) +{ + if ((event->hw.config & X86_RAW_EVENT_MASK) == 0x003c) { + /* + * Use an alternative encoding for CPU_CLK_UNHALTED.THREAD_P + * (0x003c) so that we can use it with PEBS. + * + * The regular CPU_CLK_UNHALTED.THREAD_P event (0x003c) isn't + * PEBS capable. However we can use UOPS_RETIRED.ALL + * (0x01c2), which is a PEBS capable event, to get the same + * count. + * + * UOPS_RETIRED.ALL counts the number of cycles that retires + * CNTMASK micro-ops. By setting CNTMASK to a value (16) + * larger than the maximum number of micro-ops that can be + * retired per cycle (4) and then inverting the condition, we + * count all cycles that retire 16 or less micro-ops, which + * is every cycle. + * + * Thereby we gain a PEBS capable cycle counter. + */ + u64 alt_config = X86_CONFIG(.event=0xc2, .umask=0x01, .inv=1, .cmask=16); alt_config |= (event->hw.config & ~X86_RAW_EVENT_MASK); event->hw.config = alt_config; } +} + +static int intel_pmu_hw_config(struct perf_event *event) +{ + int ret = x86_pmu_hw_config(event); + + if (ret) + return ret; + + if (event->attr.precise_ip && x86_pmu.pebs_aliases) + x86_pmu.pebs_aliases(event); + + if (intel_pmu_needs_lbr_smpl(event)) { + ret = intel_pmu_setup_lbr_filter(event); + if (ret) + return ret; + } if (event->attr.type != PERF_TYPE_RAW) return 0; @@ -1327,8 +1787,16 @@ static struct perf_guest_switch_msr *intel_guest_get_msrs(int *nr) arr[0].msr = MSR_CORE_PERF_GLOBAL_CTRL; arr[0].host = x86_pmu.intel_ctrl & ~cpuc->intel_ctrl_guest_mask; arr[0].guest = x86_pmu.intel_ctrl & ~cpuc->intel_ctrl_host_mask; + /* + * If PMU counter has PEBS enabled it is not enough to disable counter + * on a guest entry since PEBS memory write can overshoot guest entry + * and corrupt guest memory. Disabling PEBS solves the problem. + */ + arr[1].msr = MSR_IA32_PEBS_ENABLE; + arr[1].host = cpuc->pebs_enabled; + arr[1].guest = 0; - *nr = 1; + *nr = 2; return arr; } @@ -1382,6 +1850,88 @@ static void core_pmu_enable_all(int added) } } +static int hsw_hw_config(struct perf_event *event) +{ + int ret = intel_pmu_hw_config(event); + + if (ret) + return ret; + if (!boot_cpu_has(X86_FEATURE_RTM) && !boot_cpu_has(X86_FEATURE_HLE)) + return 0; + event->hw.config |= event->attr.config & (HSW_IN_TX|HSW_IN_TX_CHECKPOINTED); + + /* + * IN_TX/IN_TX-CP filters are not supported by the Haswell PMU with + * PEBS or in ANY thread mode. Since the results are non-sensical forbid + * this combination. + */ + if ((event->hw.config & (HSW_IN_TX|HSW_IN_TX_CHECKPOINTED)) && + ((event->hw.config & ARCH_PERFMON_EVENTSEL_ANY) || + event->attr.precise_ip > 0)) + return -EOPNOTSUPP; + + if (event_is_checkpointed(event)) { + /* + * Sampling of checkpointed events can cause situations where + * the CPU constantly aborts because of a overflow, which is + * then checkpointed back and ignored. Forbid checkpointing + * for sampling. + * + * But still allow a long sampling period, so that perf stat + * from KVM works. + */ + if (event->attr.sample_period > 0 && + event->attr.sample_period < 0x7fffffff) + return -EOPNOTSUPP; + } + return 0; +} + +static struct event_constraint counter2_constraint = + EVENT_CONSTRAINT(0, 0x4, 0); + +static struct event_constraint * +hsw_get_event_constraints(struct cpu_hw_events *cpuc, struct perf_event *event) +{ + struct event_constraint *c = intel_get_event_constraints(cpuc, event); + + /* Handle special quirk on in_tx_checkpointed only in counter 2 */ + if (event->hw.config & HSW_IN_TX_CHECKPOINTED) { + if (c->idxmsk64 & (1U << 2)) + return &counter2_constraint; + return &emptyconstraint; + } + + return c; +} + +PMU_FORMAT_ATTR(event, "config:0-7" ); +PMU_FORMAT_ATTR(umask, "config:8-15" ); +PMU_FORMAT_ATTR(edge, "config:18" ); +PMU_FORMAT_ATTR(pc, "config:19" ); +PMU_FORMAT_ATTR(any, "config:21" ); /* v3 + */ +PMU_FORMAT_ATTR(inv, "config:23" ); +PMU_FORMAT_ATTR(cmask, "config:24-31" ); +PMU_FORMAT_ATTR(in_tx, "config:32"); +PMU_FORMAT_ATTR(in_tx_cp, "config:33"); + +static struct attribute *intel_arch_formats_attr[] = { + &format_attr_event.attr, + &format_attr_umask.attr, + &format_attr_edge.attr, + &format_attr_pc.attr, + &format_attr_inv.attr, + &format_attr_cmask.attr, + NULL, +}; + +ssize_t intel_event_sysfs_show(char *page, u64 config) +{ + u64 event = (config & ARCH_PERFMON_EVENTSEL_EVENT); + + return x86_event_sysfs_show(page, config, event); +} + static __initconst const struct x86_pmu core_pmu = { .name = "core", .handle_irq = x86_pmu_handle_irq, @@ -1406,6 +1956,8 @@ static __initconst const struct x86_pmu core_pmu = { .put_event_constraints = intel_put_event_constraints, .event_constraints = intel_core_event_constraints, .guest_get_msrs = core_guest_get_msrs, + .format_attrs = intel_arch_formats_attr, + .events_sysfs_show = intel_event_sysfs_show, }; struct intel_shared_regs *allocate_shared_regs(int cpu) @@ -1431,7 +1983,7 @@ static int intel_pmu_cpu_prepare(int cpu) { struct cpu_hw_events *cpuc = &per_cpu(cpu_hw_events, cpu); - if (!x86_pmu.extra_regs) + if (!(x86_pmu.extra_regs || x86_pmu.lbr_sel_map)) return NOTIFY_OK; cpuc->shared_regs = allocate_shared_regs(cpu); @@ -1453,22 +2005,28 @@ static void intel_pmu_cpu_starting(int cpu) */ intel_pmu_lbr_reset(); - if (!cpuc->shared_regs || (x86_pmu.er_flags & ERF_NO_HT_SHARING)) + cpuc->lbr_sel = NULL; + + if (!cpuc->shared_regs) return; - for_each_cpu(i, topology_thread_cpumask(cpu)) { - struct intel_shared_regs *pc; + if (!(x86_pmu.er_flags & ERF_NO_HT_SHARING)) { + for_each_cpu(i, topology_thread_cpumask(cpu)) { + struct intel_shared_regs *pc; - pc = per_cpu(cpu_hw_events, i).shared_regs; - if (pc && pc->core_id == core_id) { - cpuc->kfree_on_online = cpuc->shared_regs; - cpuc->shared_regs = pc; - break; + pc = per_cpu(cpu_hw_events, i).shared_regs; + if (pc && pc->core_id == core_id) { + cpuc->kfree_on_online = cpuc->shared_regs; + cpuc->shared_regs = pc; + break; + } } + cpuc->shared_regs->core_id = core_id; + cpuc->shared_regs->refcnt++; } - cpuc->shared_regs->core_id = core_id; - cpuc->shared_regs->refcnt++; + if (x86_pmu.lbr_sel_map) + cpuc->lbr_sel = &cpuc->shared_regs->regs[EXTRA_REG_LBR]; } static void intel_pmu_cpu_dying(int cpu) @@ -1486,6 +2044,38 @@ static void intel_pmu_cpu_dying(int cpu) fini_debug_store_on_cpu(cpu); } +static void intel_pmu_flush_branch_stack(void) +{ + /* + * Intel LBR does not tag entries with the + * PID of the current task, then we need to + * flush it on ctxsw + * For now, we simply reset it + */ + if (x86_pmu.lbr_nr) + intel_pmu_lbr_reset(); +} + +PMU_FORMAT_ATTR(offcore_rsp, "config1:0-63"); + +PMU_FORMAT_ATTR(ldlat, "config1:0-15"); + +static struct attribute *intel_arch3_formats_attr[] = { + &format_attr_event.attr, + &format_attr_umask.attr, + &format_attr_edge.attr, + &format_attr_pc.attr, + &format_attr_any.attr, + &format_attr_inv.attr, + &format_attr_cmask.attr, + &format_attr_in_tx.attr, + &format_attr_in_tx_cp.attr, + + &format_attr_offcore_rsp.attr, /* XXX do NHM/WSM + SNB breakout */ + &format_attr_ldlat.attr, /* PEBS load latency */ + NULL, +}; + static __initconst const struct x86_pmu intel_pmu = { .name = "Intel", .handle_irq = intel_pmu_handle_irq, @@ -1508,11 +2098,16 @@ static __initconst const struct x86_pmu intel_pmu = { .max_period = (1ULL << 31) - 1, .get_event_constraints = intel_get_event_constraints, .put_event_constraints = intel_put_event_constraints, + .pebs_aliases = intel_pebs_aliases_core2, + + .format_attrs = intel_arch3_formats_attr, + .events_sysfs_show = intel_event_sysfs_show, .cpu_prepare = intel_pmu_cpu_prepare, .cpu_starting = intel_pmu_cpu_starting, .cpu_dying = intel_pmu_cpu_dying, .guest_get_msrs = intel_guest_get_msrs, + .flush_branch_stack = intel_pmu_flush_branch_stack, }; static __init void intel_clovertown_quirk(void) @@ -1536,16 +2131,96 @@ static __init void intel_clovertown_quirk(void) * But taken together it might just make sense to not enable PEBS on * these chips. */ - printk(KERN_WARNING "PEBS disabled due to CPU errata.\n"); + pr_warn("PEBS disabled due to CPU errata\n"); x86_pmu.pebs = 0; x86_pmu.pebs_constraints = NULL; } +static int intel_snb_pebs_broken(int cpu) +{ + u32 rev = UINT_MAX; /* default to broken for unknown models */ + + switch (cpu_data(cpu).x86_model) { + case 42: /* SNB */ + rev = 0x28; + break; + + case 45: /* SNB-EP */ + switch (cpu_data(cpu).x86_mask) { + case 6: rev = 0x618; break; + case 7: rev = 0x70c; break; + } + } + + return (cpu_data(cpu).microcode < rev); +} + +static void intel_snb_check_microcode(void) +{ + int pebs_broken = 0; + int cpu; + + get_online_cpus(); + for_each_online_cpu(cpu) { + if ((pebs_broken = intel_snb_pebs_broken(cpu))) + break; + } + put_online_cpus(); + + if (pebs_broken == x86_pmu.pebs_broken) + return; + + /* + * Serialized by the microcode lock.. + */ + if (x86_pmu.pebs_broken) { + pr_info("PEBS enabled due to microcode update\n"); + x86_pmu.pebs_broken = 0; + } else { + pr_info("PEBS disabled due to CPU errata, please upgrade microcode\n"); + x86_pmu.pebs_broken = 1; + } +} + +/* + * Under certain circumstances, access certain MSR may cause #GP. + * The function tests if the input MSR can be safely accessed. + */ +static bool check_msr(unsigned long msr, u64 mask) +{ + u64 val_old, val_new, val_tmp; + + /* + * Read the current value, change it and read it back to see if it + * matches, this is needed to detect certain hardware emulators + * (qemu/kvm) that don't trap on the MSR access and always return 0s. + */ + if (rdmsrl_safe(msr, &val_old)) + return false; + + /* + * Only change the bits which can be updated by wrmsrl. + */ + val_tmp = val_old ^ mask; + if (wrmsrl_safe(msr, val_tmp) || + rdmsrl_safe(msr, &val_new)) + return false; + + if (val_new != val_tmp) + return false; + + /* Here it's sure that the MSR can be safely accessed. + * Restore the old value and return. + */ + wrmsrl(msr, val_old); + + return true; +} + static __init void intel_sandybridge_quirk(void) { - printk(KERN_WARNING "PEBS disabled due to CPU errata.\n"); - x86_pmu.pebs = 0; - x86_pmu.pebs_constraints = NULL; + x86_pmu.check_microcode = intel_snb_check_microcode; + intel_snb_check_microcode(); } static const struct { int id; char *name; } intel_arch_events_map[] __initconst = { @@ -1565,8 +2240,8 @@ static __init void intel_arch_events_quirk(void) /* disable event that reported as not presend by cpuid */ for_each_set_bit(bit, x86_pmu.events_mask, ARRAY_SIZE(intel_arch_events_map)) { intel_perfmon_event_map[intel_arch_events_map[bit].id] = 0; - printk(KERN_WARNING "CPUID marked event: \'%s\' unavailable\n", - intel_arch_events_map[bit].name); + pr_warn("CPUID marked event: \'%s\' unavailable\n", + intel_arch_events_map[bit].name); } } @@ -1585,22 +2260,61 @@ static __init void intel_nehalem_quirk(void) intel_perfmon_event_map[PERF_COUNT_HW_BRANCH_MISSES] = 0x7f89; ebx.split.no_branch_misses_retired = 0; x86_pmu.events_maskl = ebx.full; - printk(KERN_INFO "CPU erratum AAJ80 worked around\n"); + pr_info("CPU erratum AAJ80 worked around\n"); } } +EVENT_ATTR_STR(mem-loads, mem_ld_hsw, "event=0xcd,umask=0x1,ldlat=3"); +EVENT_ATTR_STR(mem-stores, mem_st_hsw, "event=0xd0,umask=0x82") + +/* Haswell special events */ +EVENT_ATTR_STR(tx-start, tx_start, "event=0xc9,umask=0x1"); +EVENT_ATTR_STR(tx-commit, tx_commit, "event=0xc9,umask=0x2"); +EVENT_ATTR_STR(tx-abort, tx_abort, "event=0xc9,umask=0x4"); +EVENT_ATTR_STR(tx-capacity, tx_capacity, "event=0x54,umask=0x2"); +EVENT_ATTR_STR(tx-conflict, tx_conflict, "event=0x54,umask=0x1"); +EVENT_ATTR_STR(el-start, el_start, "event=0xc8,umask=0x1"); +EVENT_ATTR_STR(el-commit, el_commit, "event=0xc8,umask=0x2"); +EVENT_ATTR_STR(el-abort, el_abort, "event=0xc8,umask=0x4"); +EVENT_ATTR_STR(el-capacity, el_capacity, "event=0x54,umask=0x2"); +EVENT_ATTR_STR(el-conflict, el_conflict, "event=0x54,umask=0x1"); +EVENT_ATTR_STR(cycles-t, cycles_t, "event=0x3c,in_tx=1"); +EVENT_ATTR_STR(cycles-ct, cycles_ct, "event=0x3c,in_tx=1,in_tx_cp=1"); + +static struct attribute *hsw_events_attrs[] = { + EVENT_PTR(tx_start), + EVENT_PTR(tx_commit), + EVENT_PTR(tx_abort), + EVENT_PTR(tx_capacity), + EVENT_PTR(tx_conflict), + EVENT_PTR(el_start), + EVENT_PTR(el_commit), + EVENT_PTR(el_abort), + EVENT_PTR(el_capacity), + EVENT_PTR(el_conflict), + EVENT_PTR(cycles_t), + EVENT_PTR(cycles_ct), + EVENT_PTR(mem_ld_hsw), + EVENT_PTR(mem_st_hsw), + NULL +}; + __init int intel_pmu_init(void) { union cpuid10_edx edx; union cpuid10_eax eax; union cpuid10_ebx ebx; + struct event_constraint *c; unsigned int unused; - int version; + struct extra_reg *er; + int version, i; if (!cpu_has(&boot_cpu_data, X86_FEATURE_ARCH_PERFMON)) { switch (boot_cpu_data.x86) { case 0x6: return p6_pmu_init(); + case 0xb: + return knc_pmu_init(); case 0xf: return p4_pmu_init(); } @@ -1629,6 +2343,8 @@ __init int intel_pmu_init(void) x86_pmu.events_maskl = ebx.full; x86_pmu.events_mask_len = eax.split.mask_length; + x86_pmu.max_pebs_events = min_t(unsigned, MAX_PEBS_EVENTS, x86_pmu.num_counters); + /* * Quirk: v2 perfmon does not report fixed-purpose events, so * assume at least 3 events: @@ -1636,10 +2352,7 @@ __init int intel_pmu_init(void) if (version > 1) x86_pmu.num_counters_fixed = max((int)edx.split.num_counters_fixed, 3); - /* - * v2 and above have a perf capabilities MSR - */ - if (version > 1) { + if (boot_cpu_has(X86_FEATURE_PDCM)) { u64 capabilities; rdmsrl(MSR_IA32_PERF_CAPABILITIES, capabilities); @@ -1688,10 +2401,14 @@ __init int intel_pmu_init(void) x86_pmu.enable_all = intel_pmu_nhm_enable_all; x86_pmu.extra_regs = intel_nehalem_extra_regs; + x86_pmu.cpu_events = nhm_events_attrs; + /* UOPS_ISSUED.STALLED_CYCLES */ - intel_perfmon_event_map[PERF_COUNT_HW_STALLED_CYCLES_FRONTEND] = 0x180010e; + intel_perfmon_event_map[PERF_COUNT_HW_STALLED_CYCLES_FRONTEND] = + X86_CONFIG(.event=0x0e, .umask=0x01, .inv=1, .cmask=1); /* UOPS_EXECUTED.CORE_ACTIVE_CYCLES,c=1,i=1 */ - intel_perfmon_event_map[PERF_COUNT_HW_STALLED_CYCLES_BACKEND] = 0x1803fb1; + intel_perfmon_event_map[PERF_COUNT_HW_STALLED_CYCLES_BACKEND] = + X86_CONFIG(.event=0xb1, .umask=0x3f, .inv=1, .cmask=1); x86_add_quirk(intel_nehalem_quirk); @@ -1699,6 +2416,10 @@ __init int intel_pmu_init(void) break; case 28: /* Atom */ + case 38: /* Lincroft */ + case 39: /* Penwell */ + case 53: /* Cloverview */ + case 54: /* Cedarview */ memcpy(hw_cache_event_ids, atom_hw_cache_event_ids, sizeof(hw_cache_event_ids)); @@ -1709,6 +2430,22 @@ __init int intel_pmu_init(void) pr_cont("Atom events, "); break; + case 55: /* Atom 22nm "Silvermont" */ + case 77: /* Avoton "Silvermont" */ + memcpy(hw_cache_event_ids, slm_hw_cache_event_ids, + sizeof(hw_cache_event_ids)); + memcpy(hw_cache_extra_regs, slm_hw_cache_extra_regs, + sizeof(hw_cache_extra_regs)); + + intel_pmu_lbr_init_atom(); + + x86_pmu.event_constraints = intel_slm_event_constraints; + x86_pmu.pebs_constraints = intel_slm_pebs_event_constraints; + x86_pmu.extra_regs = intel_slm_extra_regs; + x86_pmu.er_flags |= ERF_HAS_RSP_1; + pr_cont("Silvermont events, "); + break; + case 37: /* 32 nm nehalem, "Clarkdale" */ case 44: /* 32 nm nehalem, "Gulftown" */ case 47: /* 32 nm Xeon E7 */ @@ -1725,36 +2462,108 @@ __init int intel_pmu_init(void) x86_pmu.extra_regs = intel_westmere_extra_regs; x86_pmu.er_flags |= ERF_HAS_RSP_1; + x86_pmu.cpu_events = nhm_events_attrs; + /* UOPS_ISSUED.STALLED_CYCLES */ - intel_perfmon_event_map[PERF_COUNT_HW_STALLED_CYCLES_FRONTEND] = 0x180010e; + intel_perfmon_event_map[PERF_COUNT_HW_STALLED_CYCLES_FRONTEND] = + X86_CONFIG(.event=0x0e, .umask=0x01, .inv=1, .cmask=1); /* UOPS_EXECUTED.CORE_ACTIVE_CYCLES,c=1,i=1 */ - intel_perfmon_event_map[PERF_COUNT_HW_STALLED_CYCLES_BACKEND] = 0x1803fb1; + intel_perfmon_event_map[PERF_COUNT_HW_STALLED_CYCLES_BACKEND] = + X86_CONFIG(.event=0xb1, .umask=0x3f, .inv=1, .cmask=1); pr_cont("Westmere events, "); break; case 42: /* SandyBridge */ - x86_add_quirk(intel_sandybridge_quirk); case 45: /* SandyBridge, "Romely-EP" */ + x86_add_quirk(intel_sandybridge_quirk); memcpy(hw_cache_event_ids, snb_hw_cache_event_ids, sizeof(hw_cache_event_ids)); + memcpy(hw_cache_extra_regs, snb_hw_cache_extra_regs, + sizeof(hw_cache_extra_regs)); - intel_pmu_lbr_init_nhm(); + intel_pmu_lbr_init_snb(); x86_pmu.event_constraints = intel_snb_event_constraints; x86_pmu.pebs_constraints = intel_snb_pebs_event_constraints; - x86_pmu.extra_regs = intel_snb_extra_regs; + x86_pmu.pebs_aliases = intel_pebs_aliases_snb; + if (boot_cpu_data.x86_model == 45) + x86_pmu.extra_regs = intel_snbep_extra_regs; + else + x86_pmu.extra_regs = intel_snb_extra_regs; /* all extra regs are per-cpu when HT is on */ x86_pmu.er_flags |= ERF_HAS_RSP_1; x86_pmu.er_flags |= ERF_NO_HT_SHARING; + x86_pmu.cpu_events = snb_events_attrs; + /* UOPS_ISSUED.ANY,c=1,i=1 to count stall cycles */ - intel_perfmon_event_map[PERF_COUNT_HW_STALLED_CYCLES_FRONTEND] = 0x180010e; + intel_perfmon_event_map[PERF_COUNT_HW_STALLED_CYCLES_FRONTEND] = + X86_CONFIG(.event=0x0e, .umask=0x01, .inv=1, .cmask=1); /* UOPS_DISPATCHED.THREAD,c=1,i=1 to count stall cycles*/ - intel_perfmon_event_map[PERF_COUNT_HW_STALLED_CYCLES_BACKEND] = 0x18001b1; + intel_perfmon_event_map[PERF_COUNT_HW_STALLED_CYCLES_BACKEND] = + X86_CONFIG(.event=0xb1, .umask=0x01, .inv=1, .cmask=1); pr_cont("SandyBridge events, "); break; + case 58: /* IvyBridge */ + case 62: /* IvyBridge EP */ + memcpy(hw_cache_event_ids, snb_hw_cache_event_ids, + sizeof(hw_cache_event_ids)); + /* dTLB-load-misses on IVB is different than SNB */ + hw_cache_event_ids[C(DTLB)][C(OP_READ)][C(RESULT_MISS)] = 0x8108; /* DTLB_LOAD_MISSES.DEMAND_LD_MISS_CAUSES_A_WALK */ + + memcpy(hw_cache_extra_regs, snb_hw_cache_extra_regs, + sizeof(hw_cache_extra_regs)); + + intel_pmu_lbr_init_snb(); + + x86_pmu.event_constraints = intel_ivb_event_constraints; + x86_pmu.pebs_constraints = intel_ivb_pebs_event_constraints; + x86_pmu.pebs_aliases = intel_pebs_aliases_snb; + if (boot_cpu_data.x86_model == 62) + x86_pmu.extra_regs = intel_snbep_extra_regs; + else + x86_pmu.extra_regs = intel_snb_extra_regs; + /* all extra regs are per-cpu when HT is on */ + x86_pmu.er_flags |= ERF_HAS_RSP_1; + x86_pmu.er_flags |= ERF_NO_HT_SHARING; + + x86_pmu.cpu_events = snb_events_attrs; + + /* UOPS_ISSUED.ANY,c=1,i=1 to count stall cycles */ + intel_perfmon_event_map[PERF_COUNT_HW_STALLED_CYCLES_FRONTEND] = + X86_CONFIG(.event=0x0e, .umask=0x01, .inv=1, .cmask=1); + + pr_cont("IvyBridge events, "); + break; + + + case 60: /* Haswell Client */ + case 70: + case 71: + case 63: + case 69: + x86_pmu.late_ack = true; + memcpy(hw_cache_event_ids, snb_hw_cache_event_ids, sizeof(hw_cache_event_ids)); + memcpy(hw_cache_extra_regs, snb_hw_cache_extra_regs, sizeof(hw_cache_extra_regs)); + + intel_pmu_lbr_init_snb(); + + x86_pmu.event_constraints = intel_hsw_event_constraints; + x86_pmu.pebs_constraints = intel_hsw_pebs_event_constraints; + x86_pmu.extra_regs = intel_snb_extra_regs; + x86_pmu.pebs_aliases = intel_pebs_aliases_snb; + /* all extra regs are per-cpu when HT is on */ + x86_pmu.er_flags |= ERF_HAS_RSP_1; + x86_pmu.er_flags |= ERF_NO_HT_SHARING; + + x86_pmu.hw_config = hsw_hw_config; + x86_pmu.get_event_constraints = hsw_get_event_constraints; + x86_pmu.cpu_events = hsw_events_attrs; + x86_pmu.lbr_double_abort = true; + pr_cont("Haswell events, "); + break; default: switch (x86_pmu.version) { @@ -1772,5 +2581,72 @@ __init int intel_pmu_init(void) } } + if (x86_pmu.num_counters > INTEL_PMC_MAX_GENERIC) { + WARN(1, KERN_ERR "hw perf events %d > max(%d), clipping!", + x86_pmu.num_counters, INTEL_PMC_MAX_GENERIC); + x86_pmu.num_counters = INTEL_PMC_MAX_GENERIC; + } + x86_pmu.intel_ctrl = (1 << x86_pmu.num_counters) - 1; + + if (x86_pmu.num_counters_fixed > INTEL_PMC_MAX_FIXED) { + WARN(1, KERN_ERR "hw perf events fixed %d > max(%d), clipping!", + x86_pmu.num_counters_fixed, INTEL_PMC_MAX_FIXED); + x86_pmu.num_counters_fixed = INTEL_PMC_MAX_FIXED; + } + + x86_pmu.intel_ctrl |= + ((1LL << x86_pmu.num_counters_fixed)-1) << INTEL_PMC_IDX_FIXED; + + if (x86_pmu.event_constraints) { + /* + * event on fixed counter2 (REF_CYCLES) only works on this + * counter, so do not extend mask to generic counters + */ + for_each_event_constraint(c, x86_pmu.event_constraints) { + if (c->cmask != FIXED_EVENT_FLAGS + || c->idxmsk64 == INTEL_PMC_MSK_FIXED_REF_CYCLES) { + continue; + } + + c->idxmsk64 |= (1ULL << x86_pmu.num_counters) - 1; + c->weight += x86_pmu.num_counters; + } + } + + /* + * Access LBR MSR may cause #GP under certain circumstances. + * E.g. KVM doesn't support LBR MSR + * Check all LBT MSR here. + * Disable LBR access if any LBR MSRs can not be accessed. + */ + if (x86_pmu.lbr_nr && !check_msr(x86_pmu.lbr_tos, 0x3UL)) + x86_pmu.lbr_nr = 0; + for (i = 0; i < x86_pmu.lbr_nr; i++) { + if (!(check_msr(x86_pmu.lbr_from + i, 0xffffUL) && + check_msr(x86_pmu.lbr_to + i, 0xffffUL))) + x86_pmu.lbr_nr = 0; + } + + /* + * Access extra MSR may cause #GP under certain circumstances. + * E.g. KVM doesn't support offcore event + * Check all extra_regs here. + */ + if (x86_pmu.extra_regs) { + for (er = x86_pmu.extra_regs; er->msr; er++) { + er->extra_msr_access = check_msr(er->msr, 0x1ffUL); + /* Disable LBR select mapping */ + if ((er->idx == EXTRA_REG_LBR) && !er->extra_msr_access) + x86_pmu.lbr_sel_map = NULL; + } + } + + /* Support full width counters using alternative MSR range */ + if (x86_pmu.intel_cap.full_width_write) { + x86_pmu.max_period = x86_pmu.cntval_mask; + x86_pmu.perfctr = MSR_IA32_PMC0; + pr_cont("full-width counters, "); + } + return 0; } diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index d6bd49faa40..696ade311de 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -3,6 +3,7 @@ #include <linux/slab.h> #include <asm/perf_event.h> +#include <asm/insn.h> #include "perf_event.h" @@ -11,6 +12,7 @@ #define BTS_BUFFER_SIZE (PAGE_SIZE << 4) #define PEBS_BUFFER_SIZE PAGE_SIZE +#define PEBS_FIXUP_SIZE PAGE_SIZE /* * pebs_record_32 for p4 and core not supported @@ -23,6 +25,159 @@ struct pebs_record_32 { */ +union intel_x86_pebs_dse { + u64 val; + struct { + unsigned int ld_dse:4; + unsigned int ld_stlb_miss:1; + unsigned int ld_locked:1; + unsigned int ld_reserved:26; + }; + struct { + unsigned int st_l1d_hit:1; + unsigned int st_reserved1:3; + unsigned int st_stlb_miss:1; + unsigned int st_locked:1; + unsigned int st_reserved2:26; + }; +}; + + +/* + * Map PEBS Load Latency Data Source encodings to generic + * memory data source information + */ +#define P(a, b) PERF_MEM_S(a, b) +#define OP_LH (P(OP, LOAD) | P(LVL, HIT)) +#define SNOOP_NONE_MISS (P(SNOOP, NONE) | P(SNOOP, MISS)) + +static const u64 pebs_data_source[] = { + P(OP, LOAD) | P(LVL, MISS) | P(LVL, L3) | P(SNOOP, NA),/* 0x00:ukn L3 */ + OP_LH | P(LVL, L1) | P(SNOOP, NONE), /* 0x01: L1 local */ + OP_LH | P(LVL, LFB) | P(SNOOP, NONE), /* 0x02: LFB hit */ + OP_LH | P(LVL, L2) | P(SNOOP, NONE), /* 0x03: L2 hit */ + OP_LH | P(LVL, L3) | P(SNOOP, NONE), /* 0x04: L3 hit */ + OP_LH | P(LVL, L3) | P(SNOOP, MISS), /* 0x05: L3 hit, snoop miss */ + OP_LH | P(LVL, L3) | P(SNOOP, HIT), /* 0x06: L3 hit, snoop hit */ + OP_LH | P(LVL, L3) | P(SNOOP, HITM), /* 0x07: L3 hit, snoop hitm */ + OP_LH | P(LVL, REM_CCE1) | P(SNOOP, HIT), /* 0x08: L3 miss snoop hit */ + OP_LH | P(LVL, REM_CCE1) | P(SNOOP, HITM), /* 0x09: L3 miss snoop hitm*/ + OP_LH | P(LVL, LOC_RAM) | P(SNOOP, HIT), /* 0x0a: L3 miss, shared */ + OP_LH | P(LVL, REM_RAM1) | P(SNOOP, HIT), /* 0x0b: L3 miss, shared */ + OP_LH | P(LVL, LOC_RAM) | SNOOP_NONE_MISS,/* 0x0c: L3 miss, excl */ + OP_LH | P(LVL, REM_RAM1) | SNOOP_NONE_MISS,/* 0x0d: L3 miss, excl */ + OP_LH | P(LVL, IO) | P(SNOOP, NONE), /* 0x0e: I/O */ + OP_LH | P(LVL, UNC) | P(SNOOP, NONE), /* 0x0f: uncached */ +}; + +static u64 precise_store_data(u64 status) +{ + union intel_x86_pebs_dse dse; + u64 val = P(OP, STORE) | P(SNOOP, NA) | P(LVL, L1) | P(TLB, L2); + + dse.val = status; + + /* + * bit 4: TLB access + * 1 = stored missed 2nd level TLB + * + * so it either hit the walker or the OS + * otherwise hit 2nd level TLB + */ + if (dse.st_stlb_miss) + val |= P(TLB, MISS); + else + val |= P(TLB, HIT); + + /* + * bit 0: hit L1 data cache + * if not set, then all we know is that + * it missed L1D + */ + if (dse.st_l1d_hit) + val |= P(LVL, HIT); + else + val |= P(LVL, MISS); + + /* + * bit 5: Locked prefix + */ + if (dse.st_locked) + val |= P(LOCK, LOCKED); + + return val; +} + +static u64 precise_store_data_hsw(struct perf_event *event, u64 status) +{ + union perf_mem_data_src dse; + u64 cfg = event->hw.config & INTEL_ARCH_EVENT_MASK; + + dse.val = 0; + dse.mem_op = PERF_MEM_OP_STORE; + dse.mem_lvl = PERF_MEM_LVL_NA; + + /* + * L1 info only valid for following events: + * + * MEM_UOPS_RETIRED.STLB_MISS_STORES + * MEM_UOPS_RETIRED.LOCK_STORES + * MEM_UOPS_RETIRED.SPLIT_STORES + * MEM_UOPS_RETIRED.ALL_STORES + */ + if (cfg != 0x12d0 && cfg != 0x22d0 && cfg != 0x42d0 && cfg != 0x82d0) + return dse.mem_lvl; + + if (status & 1) + dse.mem_lvl = PERF_MEM_LVL_L1 | PERF_MEM_LVL_HIT; + else + dse.mem_lvl = PERF_MEM_LVL_L1 | PERF_MEM_LVL_MISS; + + /* Nothing else supported. Sorry. */ + return dse.val; +} + +static u64 load_latency_data(u64 status) +{ + union intel_x86_pebs_dse dse; + u64 val; + int model = boot_cpu_data.x86_model; + int fam = boot_cpu_data.x86; + + dse.val = status; + + /* + * use the mapping table for bit 0-3 + */ + val = pebs_data_source[dse.ld_dse]; + + /* + * Nehalem models do not support TLB, Lock infos + */ + if (fam == 0x6 && (model == 26 || model == 30 + || model == 31 || model == 46)) { + val |= P(TLB, NA) | P(LOCK, NA); + return val; + } + /* + * bit 4: TLB access + * 0 = did not miss 2nd level TLB + * 1 = missed 2nd level TLB + */ + if (dse.ld_stlb_miss) + val |= P(TLB, MISS) | P(TLB, L2); + else + val |= P(TLB, HIT) | P(TLB, L1) | P(TLB, L2); + + /* + * bit 5: locked prefix + */ + if (dse.ld_locked) + val |= P(LOCK, LOCKED); + + return val; +} + struct pebs_record_core { u64 flags, ip; u64 ax, bx, cx, dx; @@ -40,6 +195,36 @@ struct pebs_record_nhm { u64 status, dla, dse, lat; }; +/* + * Same as pebs_record_nhm, with two additional fields. + */ +struct pebs_record_hsw { + u64 flags, ip; + u64 ax, bx, cx, dx; + u64 si, di, bp, sp; + u64 r8, r9, r10, r11; + u64 r12, r13, r14, r15; + u64 status, dla, dse, lat; + u64 real_ip, tsx_tuning; +}; + +union hsw_tsx_tuning { + struct { + u32 cycles_last_block : 32, + hle_abort : 1, + rtm_abort : 1, + instruction_abort : 1, + non_instruction_abort : 1, + retry : 1, + data_conflict : 1, + capacity_writes : 1, + capacity_reads : 1; + }; + u64 value; +}; + +#define PEBS_HSW_TSX_FLAGS 0xff00000000ULL + void init_debug_store_on_cpu(int cpu) { struct debug_store *ds = per_cpu(cpu_hw_events, cpu).ds; @@ -60,20 +245,35 @@ void fini_debug_store_on_cpu(int cpu) wrmsr_on_cpu(cpu, MSR_IA32_DS_AREA, 0, 0); } +static DEFINE_PER_CPU(void *, insn_buffer); + static int alloc_pebs_buffer(int cpu) { struct debug_store *ds = per_cpu(cpu_hw_events, cpu).ds; int node = cpu_to_node(cpu); int max, thresh = 1; /* always use a single PEBS record */ - void *buffer; + void *buffer, *ibuffer; if (!x86_pmu.pebs) return 0; - buffer = kmalloc_node(PEBS_BUFFER_SIZE, GFP_KERNEL | __GFP_ZERO, node); + buffer = kzalloc_node(PEBS_BUFFER_SIZE, GFP_KERNEL, node); if (unlikely(!buffer)) return -ENOMEM; + /* + * HSW+ already provides us the eventing ip; no need to allocate this + * buffer then. + */ + if (x86_pmu.intel_cap.pebs_format < 2) { + ibuffer = kzalloc_node(PEBS_FIXUP_SIZE, GFP_KERNEL, node); + if (!ibuffer) { + kfree(buffer); + return -ENOMEM; + } + per_cpu(insn_buffer, cpu) = ibuffer; + } + max = PEBS_BUFFER_SIZE / x86_pmu.pebs_record_size; ds->pebs_buffer_base = (u64)(unsigned long)buffer; @@ -94,6 +294,9 @@ static void release_pebs_buffer(int cpu) if (!ds || !x86_pmu.pebs) return; + kfree(per_cpu(insn_buffer, cpu)); + per_cpu(insn_buffer, cpu) = NULL; + kfree((void *)(unsigned long)ds->pebs_buffer_base); ds->pebs_buffer_base = 0; } @@ -108,9 +311,11 @@ static int alloc_bts_buffer(int cpu) if (!x86_pmu.bts) return 0; - buffer = kmalloc_node(BTS_BUFFER_SIZE, GFP_KERNEL | __GFP_ZERO, node); - if (unlikely(!buffer)) + buffer = kzalloc_node(BTS_BUFFER_SIZE, GFP_KERNEL | __GFP_NOWARN, node); + if (unlikely(!buffer)) { + WARN_ONCE(1, "%s: BTS buffer allocation failure\n", __func__); return -ENOMEM; + } max = BTS_BUFFER_SIZE / BTS_RECORD_SIZE; thresh = max / 16; @@ -141,7 +346,7 @@ static int alloc_ds_buffer(int cpu) int node = cpu_to_node(cpu); struct debug_store *ds; - ds = kmalloc_node(sizeof(*ds), GFP_KERNEL | __GFP_ZERO, node); + ds = kzalloc_node(sizeof(*ds), GFP_KERNEL, node); if (unlikely(!ds)) return -ENOMEM; @@ -247,7 +452,7 @@ void reserve_ds_buffers(void) */ struct event_constraint bts_constraint = - EVENT_CONSTRAINT(0, 1ULL << X86_PMC_IDX_FIXED_BTS, 0); + EVENT_CONSTRAINT(0, 1ULL << INTEL_PMC_IDX_FIXED_BTS, 0); void intel_pmu_enable_bts(u64 config) { @@ -294,7 +499,7 @@ int intel_pmu_drain_bts_buffer(void) u64 to; u64 flags; }; - struct perf_event *event = cpuc->events[X86_PMC_IDX_FIXED_BTS]; + struct perf_event *event = cpuc->events[INTEL_PMC_IDX_FIXED_BTS]; struct bts_record *at, *top; struct perf_output_handle handle; struct perf_event_header header; @@ -313,11 +518,11 @@ int intel_pmu_drain_bts_buffer(void) if (top <= at) return 0; + memset(®s, 0, sizeof(regs)); + ds->bts_index = ds->bts_buffer_base; - perf_sample_data_init(&data, 0); - data.period = event->hw.last_period; - regs.ip = 0; + perf_sample_data_init(&data, 0, event->hw.last_period); /* * Prepare a generic sample, i.e. fill in the invariant fields. @@ -363,8 +568,34 @@ struct event_constraint intel_atom_pebs_event_constraints[] = { EVENT_CONSTRAINT_END }; +struct event_constraint intel_slm_pebs_event_constraints[] = { + INTEL_UEVENT_CONSTRAINT(0x0103, 0x1), /* REHABQ.LD_BLOCK_ST_FORWARD_PS */ + INTEL_UEVENT_CONSTRAINT(0x0803, 0x1), /* REHABQ.LD_SPLITS_PS */ + INTEL_UEVENT_CONSTRAINT(0x0204, 0x1), /* MEM_UOPS_RETIRED.L2_HIT_LOADS_PS */ + INTEL_UEVENT_CONSTRAINT(0x0404, 0x1), /* MEM_UOPS_RETIRED.L2_MISS_LOADS_PS */ + INTEL_UEVENT_CONSTRAINT(0x0804, 0x1), /* MEM_UOPS_RETIRED.DTLB_MISS_LOADS_PS */ + INTEL_UEVENT_CONSTRAINT(0x2004, 0x1), /* MEM_UOPS_RETIRED.HITM_PS */ + INTEL_UEVENT_CONSTRAINT(0x00c0, 0x1), /* INST_RETIRED.ANY_PS */ + INTEL_UEVENT_CONSTRAINT(0x00c4, 0x1), /* BR_INST_RETIRED.ALL_BRANCHES_PS */ + INTEL_UEVENT_CONSTRAINT(0x7ec4, 0x1), /* BR_INST_RETIRED.JCC_PS */ + INTEL_UEVENT_CONSTRAINT(0xbfc4, 0x1), /* BR_INST_RETIRED.FAR_BRANCH_PS */ + INTEL_UEVENT_CONSTRAINT(0xebc4, 0x1), /* BR_INST_RETIRED.NON_RETURN_IND_PS */ + INTEL_UEVENT_CONSTRAINT(0xf7c4, 0x1), /* BR_INST_RETIRED.RETURN_PS */ + INTEL_UEVENT_CONSTRAINT(0xf9c4, 0x1), /* BR_INST_RETIRED.CALL_PS */ + INTEL_UEVENT_CONSTRAINT(0xfbc4, 0x1), /* BR_INST_RETIRED.IND_CALL_PS */ + INTEL_UEVENT_CONSTRAINT(0xfdc4, 0x1), /* BR_INST_RETIRED.REL_CALL_PS */ + INTEL_UEVENT_CONSTRAINT(0xfec4, 0x1), /* BR_INST_RETIRED.TAKEN_JCC_PS */ + INTEL_UEVENT_CONSTRAINT(0x00c5, 0x1), /* BR_INST_MISP_RETIRED.ALL_BRANCHES_PS */ + INTEL_UEVENT_CONSTRAINT(0x7ec5, 0x1), /* BR_INST_MISP_RETIRED.JCC_PS */ + INTEL_UEVENT_CONSTRAINT(0xebc5, 0x1), /* BR_INST_MISP_RETIRED.NON_RETURN_IND_PS */ + INTEL_UEVENT_CONSTRAINT(0xf7c5, 0x1), /* BR_INST_MISP_RETIRED.RETURN_PS */ + INTEL_UEVENT_CONSTRAINT(0xfbc5, 0x1), /* BR_INST_MISP_RETIRED.IND_CALL_PS */ + INTEL_UEVENT_CONSTRAINT(0xfec5, 0x1), /* BR_INST_MISP_RETIRED.TAKEN_JCC_PS */ + EVENT_CONSTRAINT_END +}; + struct event_constraint intel_nehalem_pebs_event_constraints[] = { - INTEL_EVENT_CONSTRAINT(0x0b, 0xf), /* MEM_INST_RETIRED.* */ + INTEL_PLD_CONSTRAINT(0x100b, 0xf), /* MEM_INST_RETIRED.* */ INTEL_EVENT_CONSTRAINT(0x0f, 0xf), /* MEM_UNCORE_RETIRED.* */ INTEL_UEVENT_CONSTRAINT(0x010c, 0xf), /* MEM_STORE_RETIRED.DTLB_MISS */ INTEL_EVENT_CONSTRAINT(0xc0, 0xf), /* INST_RETIRED.ANY */ @@ -379,7 +610,7 @@ struct event_constraint intel_nehalem_pebs_event_constraints[] = { }; struct event_constraint intel_westmere_pebs_event_constraints[] = { - INTEL_EVENT_CONSTRAINT(0x0b, 0xf), /* MEM_INST_RETIRED.* */ + INTEL_PLD_CONSTRAINT(0x100b, 0xf), /* MEM_INST_RETIRED.* */ INTEL_EVENT_CONSTRAINT(0x0f, 0xf), /* MEM_UNCORE_RETIRED.* */ INTEL_UEVENT_CONSTRAINT(0x010c, 0xf), /* MEM_STORE_RETIRED.DTLB_MISS */ INTEL_EVENT_CONSTRAINT(0xc0, 0xf), /* INSTR_RETIRED.* */ @@ -399,21 +630,67 @@ struct event_constraint intel_snb_pebs_event_constraints[] = { INTEL_UEVENT_CONSTRAINT(0x02c2, 0xf), /* UOPS_RETIRED.RETIRE_SLOTS */ INTEL_EVENT_CONSTRAINT(0xc4, 0xf), /* BR_INST_RETIRED.* */ INTEL_EVENT_CONSTRAINT(0xc5, 0xf), /* BR_MISP_RETIRED.* */ - INTEL_EVENT_CONSTRAINT(0xcd, 0x8), /* MEM_TRANS_RETIRED.* */ - INTEL_UEVENT_CONSTRAINT(0x11d0, 0xf), /* MEM_UOP_RETIRED.STLB_MISS_LOADS */ - INTEL_UEVENT_CONSTRAINT(0x12d0, 0xf), /* MEM_UOP_RETIRED.STLB_MISS_STORES */ - INTEL_UEVENT_CONSTRAINT(0x21d0, 0xf), /* MEM_UOP_RETIRED.LOCK_LOADS */ - INTEL_UEVENT_CONSTRAINT(0x22d0, 0xf), /* MEM_UOP_RETIRED.LOCK_STORES */ - INTEL_UEVENT_CONSTRAINT(0x41d0, 0xf), /* MEM_UOP_RETIRED.SPLIT_LOADS */ - INTEL_UEVENT_CONSTRAINT(0x42d0, 0xf), /* MEM_UOP_RETIRED.SPLIT_STORES */ - INTEL_UEVENT_CONSTRAINT(0x81d0, 0xf), /* MEM_UOP_RETIRED.ANY_LOADS */ - INTEL_UEVENT_CONSTRAINT(0x82d0, 0xf), /* MEM_UOP_RETIRED.ANY_STORES */ + INTEL_PLD_CONSTRAINT(0x01cd, 0x8), /* MEM_TRANS_RETIRED.LAT_ABOVE_THR */ + INTEL_PST_CONSTRAINT(0x02cd, 0x8), /* MEM_TRANS_RETIRED.PRECISE_STORES */ + INTEL_EVENT_CONSTRAINT(0xd0, 0xf), /* MEM_UOP_RETIRED.* */ INTEL_EVENT_CONSTRAINT(0xd1, 0xf), /* MEM_LOAD_UOPS_RETIRED.* */ INTEL_EVENT_CONSTRAINT(0xd2, 0xf), /* MEM_LOAD_UOPS_LLC_HIT_RETIRED.* */ + INTEL_EVENT_CONSTRAINT(0xd3, 0xf), /* MEM_LOAD_UOPS_LLC_MISS_RETIRED.* */ INTEL_UEVENT_CONSTRAINT(0x02d4, 0xf), /* MEM_LOAD_UOPS_MISC_RETIRED.LLC_MISS */ EVENT_CONSTRAINT_END }; +struct event_constraint intel_ivb_pebs_event_constraints[] = { + INTEL_UEVENT_CONSTRAINT(0x01c0, 0x2), /* INST_RETIRED.PRECDIST */ + INTEL_UEVENT_CONSTRAINT(0x01c2, 0xf), /* UOPS_RETIRED.ALL */ + INTEL_UEVENT_CONSTRAINT(0x02c2, 0xf), /* UOPS_RETIRED.RETIRE_SLOTS */ + INTEL_EVENT_CONSTRAINT(0xc4, 0xf), /* BR_INST_RETIRED.* */ + INTEL_EVENT_CONSTRAINT(0xc5, 0xf), /* BR_MISP_RETIRED.* */ + INTEL_PLD_CONSTRAINT(0x01cd, 0x8), /* MEM_TRANS_RETIRED.LAT_ABOVE_THR */ + INTEL_PST_CONSTRAINT(0x02cd, 0x8), /* MEM_TRANS_RETIRED.PRECISE_STORES */ + INTEL_EVENT_CONSTRAINT(0xd0, 0xf), /* MEM_UOP_RETIRED.* */ + INTEL_EVENT_CONSTRAINT(0xd1, 0xf), /* MEM_LOAD_UOPS_RETIRED.* */ + INTEL_EVENT_CONSTRAINT(0xd2, 0xf), /* MEM_LOAD_UOPS_LLC_HIT_RETIRED.* */ + INTEL_EVENT_CONSTRAINT(0xd3, 0xf), /* MEM_LOAD_UOPS_LLC_MISS_RETIRED.* */ + EVENT_CONSTRAINT_END +}; + +struct event_constraint intel_hsw_pebs_event_constraints[] = { + INTEL_UEVENT_CONSTRAINT(0x01c0, 0x2), /* INST_RETIRED.PRECDIST */ + INTEL_PST_HSW_CONSTRAINT(0x01c2, 0xf), /* UOPS_RETIRED.ALL */ + INTEL_UEVENT_CONSTRAINT(0x02c2, 0xf), /* UOPS_RETIRED.RETIRE_SLOTS */ + INTEL_EVENT_CONSTRAINT(0xc4, 0xf), /* BR_INST_RETIRED.* */ + INTEL_UEVENT_CONSTRAINT(0x01c5, 0xf), /* BR_MISP_RETIRED.CONDITIONAL */ + INTEL_UEVENT_CONSTRAINT(0x04c5, 0xf), /* BR_MISP_RETIRED.ALL_BRANCHES */ + INTEL_UEVENT_CONSTRAINT(0x20c5, 0xf), /* BR_MISP_RETIRED.NEAR_TAKEN */ + INTEL_PLD_CONSTRAINT(0x01cd, 0x8), /* MEM_TRANS_RETIRED.* */ + /* MEM_UOPS_RETIRED.STLB_MISS_LOADS */ + INTEL_UEVENT_CONSTRAINT(0x11d0, 0xf), + /* MEM_UOPS_RETIRED.STLB_MISS_STORES */ + INTEL_UEVENT_CONSTRAINT(0x12d0, 0xf), + INTEL_UEVENT_CONSTRAINT(0x21d0, 0xf), /* MEM_UOPS_RETIRED.LOCK_LOADS */ + INTEL_UEVENT_CONSTRAINT(0x41d0, 0xf), /* MEM_UOPS_RETIRED.SPLIT_LOADS */ + /* MEM_UOPS_RETIRED.SPLIT_STORES */ + INTEL_UEVENT_CONSTRAINT(0x42d0, 0xf), + INTEL_UEVENT_CONSTRAINT(0x81d0, 0xf), /* MEM_UOPS_RETIRED.ALL_LOADS */ + INTEL_PST_HSW_CONSTRAINT(0x82d0, 0xf), /* MEM_UOPS_RETIRED.ALL_STORES */ + INTEL_UEVENT_CONSTRAINT(0x01d1, 0xf), /* MEM_LOAD_UOPS_RETIRED.L1_HIT */ + INTEL_UEVENT_CONSTRAINT(0x02d1, 0xf), /* MEM_LOAD_UOPS_RETIRED.L2_HIT */ + INTEL_UEVENT_CONSTRAINT(0x04d1, 0xf), /* MEM_LOAD_UOPS_RETIRED.L3_HIT */ + /* MEM_LOAD_UOPS_RETIRED.HIT_LFB */ + INTEL_UEVENT_CONSTRAINT(0x40d1, 0xf), + /* MEM_LOAD_UOPS_LLC_HIT_RETIRED.XSNP_MISS */ + INTEL_UEVENT_CONSTRAINT(0x01d2, 0xf), + /* MEM_LOAD_UOPS_LLC_HIT_RETIRED.XSNP_HIT */ + INTEL_UEVENT_CONSTRAINT(0x02d2, 0xf), + /* MEM_LOAD_UOPS_LLC_MISS_RETIRED.LOCAL_DRAM */ + INTEL_UEVENT_CONSTRAINT(0x01d3, 0xf), + INTEL_UEVENT_CONSTRAINT(0x04c8, 0xf), /* HLE_RETIRED.Abort */ + INTEL_UEVENT_CONSTRAINT(0x04c9, 0xf), /* RTM_RETIRED.Abort */ + + EVENT_CONSTRAINT_END +}; + struct event_constraint *intel_pebs_constraints(struct perf_event *event) { struct event_constraint *c; @@ -423,8 +700,10 @@ struct event_constraint *intel_pebs_constraints(struct perf_event *event) if (x86_pmu.pebs_constraints) { for_each_event_constraint(c, x86_pmu.pebs_constraints) { - if ((event->hw.config & c->cmask) == c->code) + if ((event->hw.config & c->cmask) == c->code) { + event->hw.flags |= c->flags; return c; + } } } @@ -440,8 +719,10 @@ void intel_pmu_pebs_enable(struct perf_event *event) cpuc->pebs_enabled |= 1ULL << hwc->idx; - if (x86_pmu.intel_cap.pebs_trap && event->attr.precise_ip > 1) - intel_pmu_lbr_enable(event); + if (event->hw.flags & PERF_X86_EVENT_PEBS_LDLAT) + cpuc->pebs_enabled |= 1ULL << (hwc->idx + 32); + else if (event->hw.flags & PERF_X86_EVENT_PEBS_ST) + cpuc->pebs_enabled |= 1ULL << 63; } void intel_pmu_pebs_disable(struct perf_event *event) @@ -450,13 +731,16 @@ void intel_pmu_pebs_disable(struct perf_event *event) struct hw_perf_event *hwc = &event->hw; cpuc->pebs_enabled &= ~(1ULL << hwc->idx); + + if (event->hw.constraint->flags & PERF_X86_EVENT_PEBS_LDLAT) + cpuc->pebs_enabled &= ~(1ULL << (hwc->idx + 32)); + else if (event->hw.constraint->flags & PERF_X86_EVENT_PEBS_ST) + cpuc->pebs_enabled &= ~(1ULL << 63); + if (cpuc->enabled) wrmsrl(MSR_IA32_PEBS_ENABLE, cpuc->pebs_enabled); hwc->config |= ARCH_PERFMON_EVENTSEL_INT; - - if (x86_pmu.intel_cap.pebs_trap && event->attr.precise_ip > 1) - intel_pmu_lbr_disable(event); } void intel_pmu_pebs_enable_all(void) @@ -475,17 +759,6 @@ void intel_pmu_pebs_disable_all(void) wrmsrl(MSR_IA32_PEBS_ENABLE, 0); } -#include <asm/insn.h> - -static inline bool kernel_ip(unsigned long ip) -{ -#ifdef CONFIG_X86_32 - return ip > PAGE_OFFSET; -#else - return (long)ip < 0; -#endif -} - static int intel_pmu_pebs_fixup_ip(struct pt_regs *regs) { struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); @@ -493,6 +766,7 @@ static int intel_pmu_pebs_fixup_ip(struct pt_regs *regs) unsigned long old_to, to = cpuc->lbr_entries[0].to; unsigned long ip = regs->ip; int is_64bit = 0; + void *kaddr; /* * We don't need to fixup if the PEBS assist is fault like @@ -516,44 +790,48 @@ static int intel_pmu_pebs_fixup_ip(struct pt_regs *regs) * unsigned math, either ip is before the start (impossible) or * the basic block is larger than 1 page (sanity) */ - if ((ip - to) > PAGE_SIZE) + if ((ip - to) > PEBS_FIXUP_SIZE) return 0; /* * We sampled a branch insn, rewind using the LBR stack */ if (ip == to) { - regs->ip = from; + set_linear_ip(regs, from); return 1; } + if (!kernel_ip(ip)) { + int size, bytes; + u8 *buf = this_cpu_read(insn_buffer); + + size = ip - to; /* Must fit our buffer, see above */ + bytes = copy_from_user_nmi(buf, (void __user *)to, size); + if (bytes != 0) + return 0; + + kaddr = buf; + } else { + kaddr = (void *)to; + } + do { struct insn insn; - u8 buf[MAX_INSN_SIZE]; - void *kaddr; old_to = to; - if (!kernel_ip(ip)) { - int bytes, size = MAX_INSN_SIZE; - - bytes = copy_from_user_nmi(buf, (void __user *)to, size); - if (bytes != size) - return 0; - - kaddr = buf; - } else - kaddr = (void *)to; #ifdef CONFIG_X86_64 is_64bit = kernel_ip(to) || !test_thread_flag(TIF_IA32); #endif insn_init(&insn, kaddr, is_64bit); insn_get_length(&insn); + to += insn.length; + kaddr += insn.length; } while (to < ip); if (to == ip) { - regs->ip = old_to; + set_linear_ip(regs, old_to); return 1; } @@ -564,23 +842,74 @@ static int intel_pmu_pebs_fixup_ip(struct pt_regs *regs) return 0; } +static inline u64 intel_hsw_weight(struct pebs_record_hsw *pebs) +{ + if (pebs->tsx_tuning) { + union hsw_tsx_tuning tsx = { .value = pebs->tsx_tuning }; + return tsx.cycles_last_block; + } + return 0; +} + +static inline u64 intel_hsw_transaction(struct pebs_record_hsw *pebs) +{ + u64 txn = (pebs->tsx_tuning & PEBS_HSW_TSX_FLAGS) >> 32; + + /* For RTM XABORTs also log the abort code from AX */ + if ((txn & PERF_TXN_TRANSACTION) && (pebs->ax & 1)) + txn |= ((pebs->ax >> 24) & 0xff) << PERF_TXN_ABORT_SHIFT; + return txn; +} + static void __intel_pmu_pebs_event(struct perf_event *event, struct pt_regs *iregs, void *__pebs) { /* - * We cast to pebs_record_core since that is a subset of - * both formats and we don't use the other fields in this - * routine. + * We cast to the biggest pebs_record but are careful not to + * unconditionally access the 'extra' entries. */ - struct pebs_record_core *pebs = __pebs; + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + struct pebs_record_hsw *pebs = __pebs; struct perf_sample_data data; struct pt_regs regs; + u64 sample_type; + int fll, fst; if (!intel_pmu_save_and_restart(event)) return; - perf_sample_data_init(&data, 0); + fll = event->hw.flags & PERF_X86_EVENT_PEBS_LDLAT; + fst = event->hw.flags & (PERF_X86_EVENT_PEBS_ST | + PERF_X86_EVENT_PEBS_ST_HSW); + + perf_sample_data_init(&data, 0, event->hw.last_period); + data.period = event->hw.last_period; + sample_type = event->attr.sample_type; + + /* + * if PEBS-LL or PreciseStore + */ + if (fll || fst) { + /* + * Use latency for weight (only avail with PEBS-LL) + */ + if (fll && (sample_type & PERF_SAMPLE_WEIGHT)) + data.weight = pebs->lat; + + /* + * data.data_src encodes the data source + */ + if (sample_type & PERF_SAMPLE_DATA_SRC) { + if (fll) + data.data_src.val = load_latency_data(pebs->dse); + else if (event->hw.flags & PERF_X86_EVENT_PEBS_ST_HSW) + data.data_src.val = + precise_store_data_hsw(event, pebs->dse); + else + data.data_src.val = precise_store_data(pebs->dse); + } + } /* * We use the interrupt regs as a base because the PEBS record @@ -593,15 +922,35 @@ static void __intel_pmu_pebs_event(struct perf_event *event, * A possible PERF_SAMPLE_REGS will have to transfer all regs. */ regs = *iregs; - regs.ip = pebs->ip; + regs.flags = pebs->flags; + set_linear_ip(®s, pebs->ip); regs.bp = pebs->bp; regs.sp = pebs->sp; - if (event->attr.precise_ip > 1 && intel_pmu_pebs_fixup_ip(®s)) + if (event->attr.precise_ip > 1 && x86_pmu.intel_cap.pebs_format >= 2) { + regs.ip = pebs->real_ip; + regs.flags |= PERF_EFLAGS_EXACT; + } else if (event->attr.precise_ip > 1 && intel_pmu_pebs_fixup_ip(®s)) regs.flags |= PERF_EFLAGS_EXACT; else regs.flags &= ~PERF_EFLAGS_EXACT; + if ((event->attr.sample_type & PERF_SAMPLE_ADDR) && + x86_pmu.intel_cap.pebs_format >= 1) + data.addr = pebs->dla; + + if (x86_pmu.intel_cap.pebs_format >= 2) { + /* Only set the TSX weight when no memory weight. */ + if ((event->attr.sample_type & PERF_SAMPLE_WEIGHT) && !fll) + data.weight = intel_hsw_weight(pebs); + + if (event->attr.sample_type & PERF_SAMPLE_TRANSACTION) + data.txn = intel_hsw_transaction(pebs); + } + + if (has_branch_stack(event)) + data.br_stack = &cpuc->lbr_stack; + if (perf_event_overflow(event, &data, ®s)) x86_pmu_stop(event, 0); } @@ -641,7 +990,7 @@ static void intel_pmu_drain_pebs_core(struct pt_regs *iregs) * Should not happen, we program the threshold at 1 and do not * set a reset value. */ - WARN_ON_ONCE(n > 1); + WARN_ONCE(n > 1, "bad leftover pebs %d\n", n); at += n - 1; __intel_pmu_pebs_event(event, iregs, at); @@ -651,10 +1000,10 @@ static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) { struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); struct debug_store *ds = cpuc->ds; - struct pebs_record_nhm *at, *top; struct perf_event *event = NULL; + void *at, *top; u64 status = 0; - int bit, n; + int bit; if (!x86_pmu.pebs_active) return; @@ -664,18 +1013,22 @@ static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) ds->pebs_index = ds->pebs_buffer_base; - n = top - at; - if (n <= 0) + if (unlikely(at > top)) return; /* * Should not happen, we program the threshold at 1 and do not * set a reset value. */ - WARN_ON_ONCE(n > MAX_PEBS_EVENTS); + WARN_ONCE(top - at > x86_pmu.max_pebs_events * x86_pmu.pebs_record_size, + "Unexpected number of pebs records %ld\n", + (long)(top - at) / x86_pmu.pebs_record_size); - for ( ; at < top; at++) { - for_each_set_bit(bit, (unsigned long *)&at->status, MAX_PEBS_EVENTS) { + for (; at < top; at += x86_pmu.pebs_record_size) { + struct pebs_record_nhm *p = at; + + for_each_set_bit(bit, (unsigned long *)&p->status, + x86_pmu.max_pebs_events) { event = cpuc->events[bit]; if (!test_bit(bit, cpuc->active_mask)) continue; @@ -691,7 +1044,7 @@ static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) break; } - if (!event || bit >= MAX_PEBS_EVENTS) + if (!event || bit >= x86_pmu.max_pebs_events) continue; __intel_pmu_pebs_event(event, iregs, at); @@ -729,9 +1082,25 @@ void intel_ds_init(void) x86_pmu.drain_pebs = intel_pmu_drain_pebs_nhm; break; + case 2: + pr_cont("PEBS fmt2%c, ", pebs_type); + x86_pmu.pebs_record_size = sizeof(struct pebs_record_hsw); + x86_pmu.drain_pebs = intel_pmu_drain_pebs_nhm; + break; + default: printk(KERN_CONT "no PEBS fmt%d%c, ", format, pebs_type); x86_pmu.pebs = 0; } } } + +void perf_restore_debug_store(void) +{ + struct debug_store *ds = __this_cpu_read(cpu_hw_events.ds); + + if (!x86_pmu.bts && !x86_pmu.pebs) + return; + + wrmsrl(MSR_IA32_DS_AREA, (unsigned long)ds); +} diff --git a/arch/x86/kernel/cpu/perf_event_intel_lbr.c b/arch/x86/kernel/cpu/perf_event_intel_lbr.c index 47a7e63bfe5..9dd2459a4c7 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_lbr.c +++ b/arch/x86/kernel/cpu/perf_event_intel_lbr.c @@ -3,6 +3,7 @@ #include <asm/perf_event.h> #include <asm/msr.h> +#include <asm/insn.h> #include "perf_event.h" @@ -11,8 +12,119 @@ enum { LBR_FORMAT_LIP = 0x01, LBR_FORMAT_EIP = 0x02, LBR_FORMAT_EIP_FLAGS = 0x03, + LBR_FORMAT_EIP_FLAGS2 = 0x04, + LBR_FORMAT_MAX_KNOWN = LBR_FORMAT_EIP_FLAGS2, }; +static enum { + LBR_EIP_FLAGS = 1, + LBR_TSX = 2, +} lbr_desc[LBR_FORMAT_MAX_KNOWN + 1] = { + [LBR_FORMAT_EIP_FLAGS] = LBR_EIP_FLAGS, + [LBR_FORMAT_EIP_FLAGS2] = LBR_EIP_FLAGS | LBR_TSX, +}; + +/* + * Intel LBR_SELECT bits + * Intel Vol3a, April 2011, Section 16.7 Table 16-10 + * + * Hardware branch filter (not available on all CPUs) + */ +#define LBR_KERNEL_BIT 0 /* do not capture at ring0 */ +#define LBR_USER_BIT 1 /* do not capture at ring > 0 */ +#define LBR_JCC_BIT 2 /* do not capture conditional branches */ +#define LBR_REL_CALL_BIT 3 /* do not capture relative calls */ +#define LBR_IND_CALL_BIT 4 /* do not capture indirect calls */ +#define LBR_RETURN_BIT 5 /* do not capture near returns */ +#define LBR_IND_JMP_BIT 6 /* do not capture indirect jumps */ +#define LBR_REL_JMP_BIT 7 /* do not capture relative jumps */ +#define LBR_FAR_BIT 8 /* do not capture far branches */ + +#define LBR_KERNEL (1 << LBR_KERNEL_BIT) +#define LBR_USER (1 << LBR_USER_BIT) +#define LBR_JCC (1 << LBR_JCC_BIT) +#define LBR_REL_CALL (1 << LBR_REL_CALL_BIT) +#define LBR_IND_CALL (1 << LBR_IND_CALL_BIT) +#define LBR_RETURN (1 << LBR_RETURN_BIT) +#define LBR_REL_JMP (1 << LBR_REL_JMP_BIT) +#define LBR_IND_JMP (1 << LBR_IND_JMP_BIT) +#define LBR_FAR (1 << LBR_FAR_BIT) + +#define LBR_PLM (LBR_KERNEL | LBR_USER) + +#define LBR_SEL_MASK 0x1ff /* valid bits in LBR_SELECT */ +#define LBR_NOT_SUPP -1 /* LBR filter not supported */ +#define LBR_IGN 0 /* ignored */ + +#define LBR_ANY \ + (LBR_JCC |\ + LBR_REL_CALL |\ + LBR_IND_CALL |\ + LBR_RETURN |\ + LBR_REL_JMP |\ + LBR_IND_JMP |\ + LBR_FAR) + +#define LBR_FROM_FLAG_MISPRED (1ULL << 63) +#define LBR_FROM_FLAG_IN_TX (1ULL << 62) +#define LBR_FROM_FLAG_ABORT (1ULL << 61) + +#define for_each_branch_sample_type(x) \ + for ((x) = PERF_SAMPLE_BRANCH_USER; \ + (x) < PERF_SAMPLE_BRANCH_MAX; (x) <<= 1) + +/* + * x86control flow change classification + * x86control flow changes include branches, interrupts, traps, faults + */ +enum { + X86_BR_NONE = 0, /* unknown */ + + X86_BR_USER = 1 << 0, /* branch target is user */ + X86_BR_KERNEL = 1 << 1, /* branch target is kernel */ + + X86_BR_CALL = 1 << 2, /* call */ + X86_BR_RET = 1 << 3, /* return */ + X86_BR_SYSCALL = 1 << 4, /* syscall */ + X86_BR_SYSRET = 1 << 5, /* syscall return */ + X86_BR_INT = 1 << 6, /* sw interrupt */ + X86_BR_IRET = 1 << 7, /* return from interrupt */ + X86_BR_JCC = 1 << 8, /* conditional */ + X86_BR_JMP = 1 << 9, /* jump */ + X86_BR_IRQ = 1 << 10,/* hw interrupt or trap or fault */ + X86_BR_IND_CALL = 1 << 11,/* indirect calls */ + X86_BR_ABORT = 1 << 12,/* transaction abort */ + X86_BR_IN_TX = 1 << 13,/* in transaction */ + X86_BR_NO_TX = 1 << 14,/* not in transaction */ +}; + +#define X86_BR_PLM (X86_BR_USER | X86_BR_KERNEL) +#define X86_BR_ANYTX (X86_BR_NO_TX | X86_BR_IN_TX) + +#define X86_BR_ANY \ + (X86_BR_CALL |\ + X86_BR_RET |\ + X86_BR_SYSCALL |\ + X86_BR_SYSRET |\ + X86_BR_INT |\ + X86_BR_IRET |\ + X86_BR_JCC |\ + X86_BR_JMP |\ + X86_BR_IRQ |\ + X86_BR_ABORT |\ + X86_BR_IND_CALL) + +#define X86_BR_ALL (X86_BR_PLM | X86_BR_ANY) + +#define X86_BR_ANY_CALL \ + (X86_BR_CALL |\ + X86_BR_IND_CALL |\ + X86_BR_SYSCALL |\ + X86_BR_IRQ |\ + X86_BR_INT) + +static void intel_pmu_lbr_filter(struct cpu_hw_events *cpuc); + /* * We only support LBR implementations that have FREEZE_LBRS_ON_PMI * otherwise it becomes near impossible to get a reliable stack. @@ -21,6 +133,10 @@ enum { static void __intel_pmu_lbr_enable(void) { u64 debugctl; + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + + if (cpuc->lbr_sel) + wrmsrl(MSR_LBR_SELECT, cpuc->lbr_sel->config); rdmsrl(MSR_IA32_DEBUGCTLMSR, debugctl); debugctl |= (DEBUGCTLMSR_LBR | DEBUGCTLMSR_FREEZE_LBRS_ON_PMI); @@ -76,11 +192,11 @@ void intel_pmu_lbr_enable(struct perf_event *event) * Reset the LBR stack if we changed task context to * avoid data leaks. */ - if (event->ctx->task && cpuc->lbr_context != event->ctx) { intel_pmu_lbr_reset(); cpuc->lbr_context = event->ctx; } + cpuc->br_sel = event->hw.branch_reg.reg; cpuc->lbr_users++; } @@ -95,8 +211,11 @@ void intel_pmu_lbr_disable(struct perf_event *event) cpuc->lbr_users--; WARN_ON_ONCE(cpuc->lbr_users < 0); - if (cpuc->enabled && !cpuc->lbr_users) + if (cpuc->enabled && !cpuc->lbr_users) { __intel_pmu_lbr_disable(); + /* avoid stale pointer */ + cpuc->lbr_context = NULL; + } } void intel_pmu_lbr_enable_all(void) @@ -115,6 +234,9 @@ void intel_pmu_lbr_disable_all(void) __intel_pmu_lbr_disable(); } +/* + * TOS = most recently recorded branch + */ static inline u64 intel_pmu_lbr_tos(void) { u64 tos; @@ -142,15 +264,15 @@ static void intel_pmu_lbr_read_32(struct cpu_hw_events *cpuc) rdmsrl(x86_pmu.lbr_from + lbr_idx, msr_lastbranch.lbr); - cpuc->lbr_entries[i].from = msr_lastbranch.from; - cpuc->lbr_entries[i].to = msr_lastbranch.to; - cpuc->lbr_entries[i].flags = 0; + cpuc->lbr_entries[i].from = msr_lastbranch.from; + cpuc->lbr_entries[i].to = msr_lastbranch.to; + cpuc->lbr_entries[i].mispred = 0; + cpuc->lbr_entries[i].predicted = 0; + cpuc->lbr_entries[i].reserved = 0; } cpuc->lbr_stack.nr = i; } -#define LBR_FROM_FLAG_MISPRED (1ULL << 63) - /* * Due to lack of segmentation in Linux the effective address (offset) * is the same as the linear address, allowing us to merge the LIP and EIP @@ -162,24 +284,50 @@ static void intel_pmu_lbr_read_64(struct cpu_hw_events *cpuc) int lbr_format = x86_pmu.intel_cap.lbr_format; u64 tos = intel_pmu_lbr_tos(); int i; + int out = 0; for (i = 0; i < x86_pmu.lbr_nr; i++) { unsigned long lbr_idx = (tos - i) & mask; - u64 from, to, flags = 0; + u64 from, to, mis = 0, pred = 0, in_tx = 0, abort = 0; + int skip = 0; + int lbr_flags = lbr_desc[lbr_format]; rdmsrl(x86_pmu.lbr_from + lbr_idx, from); rdmsrl(x86_pmu.lbr_to + lbr_idx, to); - if (lbr_format == LBR_FORMAT_EIP_FLAGS) { - flags = !!(from & LBR_FROM_FLAG_MISPRED); - from = (u64)((((s64)from) << 1) >> 1); + if (lbr_flags & LBR_EIP_FLAGS) { + mis = !!(from & LBR_FROM_FLAG_MISPRED); + pred = !mis; + skip = 1; } - - cpuc->lbr_entries[i].from = from; - cpuc->lbr_entries[i].to = to; - cpuc->lbr_entries[i].flags = flags; + if (lbr_flags & LBR_TSX) { + in_tx = !!(from & LBR_FROM_FLAG_IN_TX); + abort = !!(from & LBR_FROM_FLAG_ABORT); + skip = 3; + } + from = (u64)((((s64)from) << skip) >> skip); + + /* + * Some CPUs report duplicated abort records, + * with the second entry not having an abort bit set. + * Skip them here. This loop runs backwards, + * so we need to undo the previous record. + * If the abort just happened outside the window + * the extra entry cannot be removed. + */ + if (abort && x86_pmu.lbr_double_abort && out > 0) + out--; + + cpuc->lbr_entries[out].from = from; + cpuc->lbr_entries[out].to = to; + cpuc->lbr_entries[out].mispred = mis; + cpuc->lbr_entries[out].predicted = pred; + cpuc->lbr_entries[out].in_tx = in_tx; + cpuc->lbr_entries[out].abort = abort; + cpuc->lbr_entries[out].reserved = 0; + out++; } - cpuc->lbr_stack.nr = i; + cpuc->lbr_stack.nr = out; } void intel_pmu_lbr_read(void) @@ -193,28 +341,439 @@ void intel_pmu_lbr_read(void) intel_pmu_lbr_read_32(cpuc); else intel_pmu_lbr_read_64(cpuc); + + intel_pmu_lbr_filter(cpuc); +} + +/* + * SW filter is used: + * - in case there is no HW filter + * - in case the HW filter has errata or limitations + */ +static void intel_pmu_setup_sw_lbr_filter(struct perf_event *event) +{ + u64 br_type = event->attr.branch_sample_type; + int mask = 0; + + if (br_type & PERF_SAMPLE_BRANCH_USER) + mask |= X86_BR_USER; + + if (br_type & PERF_SAMPLE_BRANCH_KERNEL) + mask |= X86_BR_KERNEL; + + /* we ignore BRANCH_HV here */ + + if (br_type & PERF_SAMPLE_BRANCH_ANY) + mask |= X86_BR_ANY; + + if (br_type & PERF_SAMPLE_BRANCH_ANY_CALL) + mask |= X86_BR_ANY_CALL; + + if (br_type & PERF_SAMPLE_BRANCH_ANY_RETURN) + mask |= X86_BR_RET | X86_BR_IRET | X86_BR_SYSRET; + + if (br_type & PERF_SAMPLE_BRANCH_IND_CALL) + mask |= X86_BR_IND_CALL; + + if (br_type & PERF_SAMPLE_BRANCH_ABORT_TX) + mask |= X86_BR_ABORT; + + if (br_type & PERF_SAMPLE_BRANCH_IN_TX) + mask |= X86_BR_IN_TX; + + if (br_type & PERF_SAMPLE_BRANCH_NO_TX) + mask |= X86_BR_NO_TX; + + if (br_type & PERF_SAMPLE_BRANCH_COND) + mask |= X86_BR_JCC; + + /* + * stash actual user request into reg, it may + * be used by fixup code for some CPU + */ + event->hw.branch_reg.reg = mask; +} + +/* + * setup the HW LBR filter + * Used only when available, may not be enough to disambiguate + * all branches, may need the help of the SW filter + */ +static int intel_pmu_setup_hw_lbr_filter(struct perf_event *event) +{ + struct hw_perf_event_extra *reg; + u64 br_type = event->attr.branch_sample_type; + u64 mask = 0, m; + u64 v; + + for_each_branch_sample_type(m) { + if (!(br_type & m)) + continue; + + v = x86_pmu.lbr_sel_map[m]; + if (v == LBR_NOT_SUPP) + return -EOPNOTSUPP; + + if (v != LBR_IGN) + mask |= v; + } + reg = &event->hw.branch_reg; + reg->idx = EXTRA_REG_LBR; + + /* LBR_SELECT operates in suppress mode so invert mask */ + reg->config = ~mask & x86_pmu.lbr_sel_mask; + + return 0; +} + +int intel_pmu_setup_lbr_filter(struct perf_event *event) +{ + int ret = 0; + + /* + * no LBR on this PMU + */ + if (!x86_pmu.lbr_nr) + return -EOPNOTSUPP; + + /* + * setup SW LBR filter + */ + intel_pmu_setup_sw_lbr_filter(event); + + /* + * setup HW LBR filter, if any + */ + if (x86_pmu.lbr_sel_map) + ret = intel_pmu_setup_hw_lbr_filter(event); + + return ret; } +/* + * return the type of control flow change at address "from" + * intruction is not necessarily a branch (in case of interrupt). + * + * The branch type returned also includes the priv level of the + * target of the control flow change (X86_BR_USER, X86_BR_KERNEL). + * + * If a branch type is unknown OR the instruction cannot be + * decoded (e.g., text page not present), then X86_BR_NONE is + * returned. + */ +static int branch_type(unsigned long from, unsigned long to, int abort) +{ + struct insn insn; + void *addr; + int bytes, size = MAX_INSN_SIZE; + int ret = X86_BR_NONE; + int ext, to_plm, from_plm; + u8 buf[MAX_INSN_SIZE]; + int is64 = 0; + + to_plm = kernel_ip(to) ? X86_BR_KERNEL : X86_BR_USER; + from_plm = kernel_ip(from) ? X86_BR_KERNEL : X86_BR_USER; + + /* + * maybe zero if lbr did not fill up after a reset by the time + * we get a PMU interrupt + */ + if (from == 0 || to == 0) + return X86_BR_NONE; + + if (abort) + return X86_BR_ABORT | to_plm; + + if (from_plm == X86_BR_USER) { + /* + * can happen if measuring at the user level only + * and we interrupt in a kernel thread, e.g., idle. + */ + if (!current->mm) + return X86_BR_NONE; + + /* may fail if text not present */ + bytes = copy_from_user_nmi(buf, (void __user *)from, size); + if (bytes != 0) + return X86_BR_NONE; + + addr = buf; + } else { + /* + * The LBR logs any address in the IP, even if the IP just + * faulted. This means userspace can control the from address. + * Ensure we don't blindy read any address by validating it is + * a known text address. + */ + if (kernel_text_address(from)) + addr = (void *)from; + else + return X86_BR_NONE; + } + + /* + * decoder needs to know the ABI especially + * on 64-bit systems running 32-bit apps + */ +#ifdef CONFIG_X86_64 + is64 = kernel_ip((unsigned long)addr) || !test_thread_flag(TIF_IA32); +#endif + insn_init(&insn, addr, is64); + insn_get_opcode(&insn); + + switch (insn.opcode.bytes[0]) { + case 0xf: + switch (insn.opcode.bytes[1]) { + case 0x05: /* syscall */ + case 0x34: /* sysenter */ + ret = X86_BR_SYSCALL; + break; + case 0x07: /* sysret */ + case 0x35: /* sysexit */ + ret = X86_BR_SYSRET; + break; + case 0x80 ... 0x8f: /* conditional */ + ret = X86_BR_JCC; + break; + default: + ret = X86_BR_NONE; + } + break; + case 0x70 ... 0x7f: /* conditional */ + ret = X86_BR_JCC; + break; + case 0xc2: /* near ret */ + case 0xc3: /* near ret */ + case 0xca: /* far ret */ + case 0xcb: /* far ret */ + ret = X86_BR_RET; + break; + case 0xcf: /* iret */ + ret = X86_BR_IRET; + break; + case 0xcc ... 0xce: /* int */ + ret = X86_BR_INT; + break; + case 0xe8: /* call near rel */ + case 0x9a: /* call far absolute */ + ret = X86_BR_CALL; + break; + case 0xe0 ... 0xe3: /* loop jmp */ + ret = X86_BR_JCC; + break; + case 0xe9 ... 0xeb: /* jmp */ + ret = X86_BR_JMP; + break; + case 0xff: /* call near absolute, call far absolute ind */ + insn_get_modrm(&insn); + ext = (insn.modrm.bytes[0] >> 3) & 0x7; + switch (ext) { + case 2: /* near ind call */ + case 3: /* far ind call */ + ret = X86_BR_IND_CALL; + break; + case 4: + case 5: + ret = X86_BR_JMP; + break; + } + break; + default: + ret = X86_BR_NONE; + } + /* + * interrupts, traps, faults (and thus ring transition) may + * occur on any instructions. Thus, to classify them correctly, + * we need to first look at the from and to priv levels. If they + * are different and to is in the kernel, then it indicates + * a ring transition. If the from instruction is not a ring + * transition instr (syscall, systenter, int), then it means + * it was a irq, trap or fault. + * + * we have no way of detecting kernel to kernel faults. + */ + if (from_plm == X86_BR_USER && to_plm == X86_BR_KERNEL + && ret != X86_BR_SYSCALL && ret != X86_BR_INT) + ret = X86_BR_IRQ; + + /* + * branch priv level determined by target as + * is done by HW when LBR_SELECT is implemented + */ + if (ret != X86_BR_NONE) + ret |= to_plm; + + return ret; +} + +/* + * implement actual branch filter based on user demand. + * Hardware may not exactly satisfy that request, thus + * we need to inspect opcodes. Mismatched branches are + * discarded. Therefore, the number of branches returned + * in PERF_SAMPLE_BRANCH_STACK sample may vary. + */ +static void +intel_pmu_lbr_filter(struct cpu_hw_events *cpuc) +{ + u64 from, to; + int br_sel = cpuc->br_sel; + int i, j, type; + bool compress = false; + + /* if sampling all branches, then nothing to filter */ + if ((br_sel & X86_BR_ALL) == X86_BR_ALL) + return; + + for (i = 0; i < cpuc->lbr_stack.nr; i++) { + + from = cpuc->lbr_entries[i].from; + to = cpuc->lbr_entries[i].to; + + type = branch_type(from, to, cpuc->lbr_entries[i].abort); + if (type != X86_BR_NONE && (br_sel & X86_BR_ANYTX)) { + if (cpuc->lbr_entries[i].in_tx) + type |= X86_BR_IN_TX; + else + type |= X86_BR_NO_TX; + } + + /* if type does not correspond, then discard */ + if (type == X86_BR_NONE || (br_sel & type) != type) { + cpuc->lbr_entries[i].from = 0; + compress = true; + } + } + + if (!compress) + return; + + /* remove all entries with from=0 */ + for (i = 0; i < cpuc->lbr_stack.nr; ) { + if (!cpuc->lbr_entries[i].from) { + j = i; + while (++j < cpuc->lbr_stack.nr) + cpuc->lbr_entries[j-1] = cpuc->lbr_entries[j]; + cpuc->lbr_stack.nr--; + if (!cpuc->lbr_entries[i].from) + continue; + } + i++; + } +} + +/* + * Map interface branch filters onto LBR filters + */ +static const int nhm_lbr_sel_map[PERF_SAMPLE_BRANCH_MAX] = { + [PERF_SAMPLE_BRANCH_ANY] = LBR_ANY, + [PERF_SAMPLE_BRANCH_USER] = LBR_USER, + [PERF_SAMPLE_BRANCH_KERNEL] = LBR_KERNEL, + [PERF_SAMPLE_BRANCH_HV] = LBR_IGN, + [PERF_SAMPLE_BRANCH_ANY_RETURN] = LBR_RETURN | LBR_REL_JMP + | LBR_IND_JMP | LBR_FAR, + /* + * NHM/WSM erratum: must include REL_JMP+IND_JMP to get CALL branches + */ + [PERF_SAMPLE_BRANCH_ANY_CALL] = + LBR_REL_CALL | LBR_IND_CALL | LBR_REL_JMP | LBR_IND_JMP | LBR_FAR, + /* + * NHM/WSM erratum: must include IND_JMP to capture IND_CALL + */ + [PERF_SAMPLE_BRANCH_IND_CALL] = LBR_IND_CALL | LBR_IND_JMP, + [PERF_SAMPLE_BRANCH_COND] = LBR_JCC, +}; + +static const int snb_lbr_sel_map[PERF_SAMPLE_BRANCH_MAX] = { + [PERF_SAMPLE_BRANCH_ANY] = LBR_ANY, + [PERF_SAMPLE_BRANCH_USER] = LBR_USER, + [PERF_SAMPLE_BRANCH_KERNEL] = LBR_KERNEL, + [PERF_SAMPLE_BRANCH_HV] = LBR_IGN, + [PERF_SAMPLE_BRANCH_ANY_RETURN] = LBR_RETURN | LBR_FAR, + [PERF_SAMPLE_BRANCH_ANY_CALL] = LBR_REL_CALL | LBR_IND_CALL + | LBR_FAR, + [PERF_SAMPLE_BRANCH_IND_CALL] = LBR_IND_CALL, + [PERF_SAMPLE_BRANCH_COND] = LBR_JCC, +}; + +/* core */ void intel_pmu_lbr_init_core(void) { x86_pmu.lbr_nr = 4; - x86_pmu.lbr_tos = 0x01c9; - x86_pmu.lbr_from = 0x40; - x86_pmu.lbr_to = 0x60; + x86_pmu.lbr_tos = MSR_LBR_TOS; + x86_pmu.lbr_from = MSR_LBR_CORE_FROM; + x86_pmu.lbr_to = MSR_LBR_CORE_TO; + + /* + * SW branch filter usage: + * - compensate for lack of HW filter + */ + pr_cont("4-deep LBR, "); } +/* nehalem/westmere */ void intel_pmu_lbr_init_nhm(void) { x86_pmu.lbr_nr = 16; - x86_pmu.lbr_tos = 0x01c9; - x86_pmu.lbr_from = 0x680; - x86_pmu.lbr_to = 0x6c0; + x86_pmu.lbr_tos = MSR_LBR_TOS; + x86_pmu.lbr_from = MSR_LBR_NHM_FROM; + x86_pmu.lbr_to = MSR_LBR_NHM_TO; + + x86_pmu.lbr_sel_mask = LBR_SEL_MASK; + x86_pmu.lbr_sel_map = nhm_lbr_sel_map; + + /* + * SW branch filter usage: + * - workaround LBR_SEL errata (see above) + * - support syscall, sysret capture. + * That requires LBR_FAR but that means far + * jmp need to be filtered out + */ + pr_cont("16-deep LBR, "); +} + +/* sandy bridge */ +void intel_pmu_lbr_init_snb(void) +{ + x86_pmu.lbr_nr = 16; + x86_pmu.lbr_tos = MSR_LBR_TOS; + x86_pmu.lbr_from = MSR_LBR_NHM_FROM; + x86_pmu.lbr_to = MSR_LBR_NHM_TO; + + x86_pmu.lbr_sel_mask = LBR_SEL_MASK; + x86_pmu.lbr_sel_map = snb_lbr_sel_map; + + /* + * SW branch filter usage: + * - support syscall, sysret capture. + * That requires LBR_FAR but that means far + * jmp need to be filtered out + */ + pr_cont("16-deep LBR, "); } +/* atom */ void intel_pmu_lbr_init_atom(void) { + /* + * only models starting at stepping 10 seems + * to have an operational LBR which can freeze + * on PMU interrupt + */ + if (boot_cpu_data.x86_model == 28 + && boot_cpu_data.x86_mask < 10) { + pr_cont("LBR disabled due to erratum"); + return; + } + x86_pmu.lbr_nr = 8; - x86_pmu.lbr_tos = 0x01c9; - x86_pmu.lbr_from = 0x40; - x86_pmu.lbr_to = 0x60; + x86_pmu.lbr_tos = MSR_LBR_TOS; + x86_pmu.lbr_from = MSR_LBR_CORE_FROM; + x86_pmu.lbr_to = MSR_LBR_CORE_TO; + + /* + * SW branch filter usage: + * - compensate for lack of HW filter + */ + pr_cont("8-deep LBR, "); } diff --git a/arch/x86/kernel/cpu/perf_event_intel_rapl.c b/arch/x86/kernel/cpu/perf_event_intel_rapl.c new file mode 100644 index 00000000000..619f7699487 --- /dev/null +++ b/arch/x86/kernel/cpu/perf_event_intel_rapl.c @@ -0,0 +1,714 @@ +/* + * perf_event_intel_rapl.c: support Intel RAPL energy consumption counters + * Copyright (C) 2013 Google, Inc., Stephane Eranian + * + * Intel RAPL interface is specified in the IA-32 Manual Vol3b + * section 14.7.1 (September 2013) + * + * RAPL provides more controls than just reporting energy consumption + * however here we only expose the 3 energy consumption free running + * counters (pp0, pkg, dram). + * + * Each of those counters increments in a power unit defined by the + * RAPL_POWER_UNIT MSR. On SandyBridge, this unit is 1/(2^16) Joules + * but it can vary. + * + * Counter to rapl events mappings: + * + * pp0 counter: consumption of all physical cores (power plane 0) + * event: rapl_energy_cores + * perf code: 0x1 + * + * pkg counter: consumption of the whole processor package + * event: rapl_energy_pkg + * perf code: 0x2 + * + * dram counter: consumption of the dram domain (servers only) + * event: rapl_energy_dram + * perf code: 0x3 + * + * dram counter: consumption of the builtin-gpu domain (client only) + * event: rapl_energy_gpu + * perf code: 0x4 + * + * We manage those counters as free running (read-only). They may be + * use simultaneously by other tools, such as turbostat. + * + * The events only support system-wide mode counting. There is no + * sampling support because it does not make sense and is not + * supported by the RAPL hardware. + * + * Because we want to avoid floating-point operations in the kernel, + * the events are all reported in fixed point arithmetic (32.32). + * Tools must adjust the counts to convert them to Watts using + * the duration of the measurement. Tools may use a function such as + * ldexp(raw_count, -32); + */ +#include <linux/module.h> +#include <linux/slab.h> +#include <linux/perf_event.h> +#include <asm/cpu_device_id.h> +#include "perf_event.h" + +/* + * RAPL energy status counters + */ +#define RAPL_IDX_PP0_NRG_STAT 0 /* all cores */ +#define INTEL_RAPL_PP0 0x1 /* pseudo-encoding */ +#define RAPL_IDX_PKG_NRG_STAT 1 /* entire package */ +#define INTEL_RAPL_PKG 0x2 /* pseudo-encoding */ +#define RAPL_IDX_RAM_NRG_STAT 2 /* DRAM */ +#define INTEL_RAPL_RAM 0x3 /* pseudo-encoding */ +#define RAPL_IDX_PP1_NRG_STAT 3 /* gpu */ +#define INTEL_RAPL_PP1 0x4 /* pseudo-encoding */ + +/* Clients have PP0, PKG */ +#define RAPL_IDX_CLN (1<<RAPL_IDX_PP0_NRG_STAT|\ + 1<<RAPL_IDX_PKG_NRG_STAT|\ + 1<<RAPL_IDX_PP1_NRG_STAT) + +/* Servers have PP0, PKG, RAM */ +#define RAPL_IDX_SRV (1<<RAPL_IDX_PP0_NRG_STAT|\ + 1<<RAPL_IDX_PKG_NRG_STAT|\ + 1<<RAPL_IDX_RAM_NRG_STAT) + +/* Servers have PP0, PKG, RAM, PP1 */ +#define RAPL_IDX_HSW (1<<RAPL_IDX_PP0_NRG_STAT|\ + 1<<RAPL_IDX_PKG_NRG_STAT|\ + 1<<RAPL_IDX_RAM_NRG_STAT|\ + 1<<RAPL_IDX_PP1_NRG_STAT) + +/* + * event code: LSB 8 bits, passed in attr->config + * any other bit is reserved + */ +#define RAPL_EVENT_MASK 0xFFULL + +#define DEFINE_RAPL_FORMAT_ATTR(_var, _name, _format) \ +static ssize_t __rapl_##_var##_show(struct kobject *kobj, \ + struct kobj_attribute *attr, \ + char *page) \ +{ \ + BUILD_BUG_ON(sizeof(_format) >= PAGE_SIZE); \ + return sprintf(page, _format "\n"); \ +} \ +static struct kobj_attribute format_attr_##_var = \ + __ATTR(_name, 0444, __rapl_##_var##_show, NULL) + +#define RAPL_EVENT_DESC(_name, _config) \ +{ \ + .attr = __ATTR(_name, 0444, rapl_event_show, NULL), \ + .config = _config, \ +} + +#define RAPL_CNTR_WIDTH 32 /* 32-bit rapl counters */ + +struct rapl_pmu { + spinlock_t lock; + int hw_unit; /* 1/2^hw_unit Joule */ + int n_active; /* number of active events */ + struct list_head active_list; + struct pmu *pmu; /* pointer to rapl_pmu_class */ + ktime_t timer_interval; /* in ktime_t unit */ + struct hrtimer hrtimer; +}; + +static struct pmu rapl_pmu_class; +static cpumask_t rapl_cpu_mask; +static int rapl_cntr_mask; + +static DEFINE_PER_CPU(struct rapl_pmu *, rapl_pmu); +static DEFINE_PER_CPU(struct rapl_pmu *, rapl_pmu_to_free); + +static inline u64 rapl_read_counter(struct perf_event *event) +{ + u64 raw; + rdmsrl(event->hw.event_base, raw); + return raw; +} + +static inline u64 rapl_scale(u64 v) +{ + /* + * scale delta to smallest unit (1/2^32) + * users must then scale back: count * 1/(1e9*2^32) to get Joules + * or use ldexp(count, -32). + * Watts = Joules/Time delta + */ + return v << (32 - __get_cpu_var(rapl_pmu)->hw_unit); +} + +static u64 rapl_event_update(struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + u64 prev_raw_count, new_raw_count; + s64 delta, sdelta; + int shift = RAPL_CNTR_WIDTH; + +again: + prev_raw_count = local64_read(&hwc->prev_count); + rdmsrl(event->hw.event_base, new_raw_count); + + if (local64_cmpxchg(&hwc->prev_count, prev_raw_count, + new_raw_count) != prev_raw_count) { + cpu_relax(); + goto again; + } + + /* + * Now we have the new raw value and have updated the prev + * timestamp already. We can now calculate the elapsed delta + * (event-)time and add that to the generic event. + * + * Careful, not all hw sign-extends above the physical width + * of the count. + */ + delta = (new_raw_count << shift) - (prev_raw_count << shift); + delta >>= shift; + + sdelta = rapl_scale(delta); + + local64_add(sdelta, &event->count); + + return new_raw_count; +} + +static void rapl_start_hrtimer(struct rapl_pmu *pmu) +{ + __hrtimer_start_range_ns(&pmu->hrtimer, + pmu->timer_interval, 0, + HRTIMER_MODE_REL_PINNED, 0); +} + +static void rapl_stop_hrtimer(struct rapl_pmu *pmu) +{ + hrtimer_cancel(&pmu->hrtimer); +} + +static enum hrtimer_restart rapl_hrtimer_handle(struct hrtimer *hrtimer) +{ + struct rapl_pmu *pmu = __get_cpu_var(rapl_pmu); + struct perf_event *event; + unsigned long flags; + + if (!pmu->n_active) + return HRTIMER_NORESTART; + + spin_lock_irqsave(&pmu->lock, flags); + + list_for_each_entry(event, &pmu->active_list, active_entry) { + rapl_event_update(event); + } + + spin_unlock_irqrestore(&pmu->lock, flags); + + hrtimer_forward_now(hrtimer, pmu->timer_interval); + + return HRTIMER_RESTART; +} + +static void rapl_hrtimer_init(struct rapl_pmu *pmu) +{ + struct hrtimer *hr = &pmu->hrtimer; + + hrtimer_init(hr, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + hr->function = rapl_hrtimer_handle; +} + +static void __rapl_pmu_event_start(struct rapl_pmu *pmu, + struct perf_event *event) +{ + if (WARN_ON_ONCE(!(event->hw.state & PERF_HES_STOPPED))) + return; + + event->hw.state = 0; + + list_add_tail(&event->active_entry, &pmu->active_list); + + local64_set(&event->hw.prev_count, rapl_read_counter(event)); + + pmu->n_active++; + if (pmu->n_active == 1) + rapl_start_hrtimer(pmu); +} + +static void rapl_pmu_event_start(struct perf_event *event, int mode) +{ + struct rapl_pmu *pmu = __get_cpu_var(rapl_pmu); + unsigned long flags; + + spin_lock_irqsave(&pmu->lock, flags); + __rapl_pmu_event_start(pmu, event); + spin_unlock_irqrestore(&pmu->lock, flags); +} + +static void rapl_pmu_event_stop(struct perf_event *event, int mode) +{ + struct rapl_pmu *pmu = __get_cpu_var(rapl_pmu); + struct hw_perf_event *hwc = &event->hw; + unsigned long flags; + + spin_lock_irqsave(&pmu->lock, flags); + + /* mark event as deactivated and stopped */ + if (!(hwc->state & PERF_HES_STOPPED)) { + WARN_ON_ONCE(pmu->n_active <= 0); + pmu->n_active--; + if (pmu->n_active == 0) + rapl_stop_hrtimer(pmu); + + list_del(&event->active_entry); + + WARN_ON_ONCE(hwc->state & PERF_HES_STOPPED); + hwc->state |= PERF_HES_STOPPED; + } + + /* check if update of sw counter is necessary */ + if ((mode & PERF_EF_UPDATE) && !(hwc->state & PERF_HES_UPTODATE)) { + /* + * Drain the remaining delta count out of a event + * that we are disabling: + */ + rapl_event_update(event); + hwc->state |= PERF_HES_UPTODATE; + } + + spin_unlock_irqrestore(&pmu->lock, flags); +} + +static int rapl_pmu_event_add(struct perf_event *event, int mode) +{ + struct rapl_pmu *pmu = __get_cpu_var(rapl_pmu); + struct hw_perf_event *hwc = &event->hw; + unsigned long flags; + + spin_lock_irqsave(&pmu->lock, flags); + + hwc->state = PERF_HES_UPTODATE | PERF_HES_STOPPED; + + if (mode & PERF_EF_START) + __rapl_pmu_event_start(pmu, event); + + spin_unlock_irqrestore(&pmu->lock, flags); + + return 0; +} + +static void rapl_pmu_event_del(struct perf_event *event, int flags) +{ + rapl_pmu_event_stop(event, PERF_EF_UPDATE); +} + +static int rapl_pmu_event_init(struct perf_event *event) +{ + u64 cfg = event->attr.config & RAPL_EVENT_MASK; + int bit, msr, ret = 0; + + /* only look at RAPL events */ + if (event->attr.type != rapl_pmu_class.type) + return -ENOENT; + + /* check only supported bits are set */ + if (event->attr.config & ~RAPL_EVENT_MASK) + return -EINVAL; + + /* + * check event is known (determines counter) + */ + switch (cfg) { + case INTEL_RAPL_PP0: + bit = RAPL_IDX_PP0_NRG_STAT; + msr = MSR_PP0_ENERGY_STATUS; + break; + case INTEL_RAPL_PKG: + bit = RAPL_IDX_PKG_NRG_STAT; + msr = MSR_PKG_ENERGY_STATUS; + break; + case INTEL_RAPL_RAM: + bit = RAPL_IDX_RAM_NRG_STAT; + msr = MSR_DRAM_ENERGY_STATUS; + break; + case INTEL_RAPL_PP1: + bit = RAPL_IDX_PP1_NRG_STAT; + msr = MSR_PP1_ENERGY_STATUS; + break; + default: + return -EINVAL; + } + /* check event supported */ + if (!(rapl_cntr_mask & (1 << bit))) + return -EINVAL; + + /* unsupported modes and filters */ + if (event->attr.exclude_user || + event->attr.exclude_kernel || + event->attr.exclude_hv || + event->attr.exclude_idle || + event->attr.exclude_host || + event->attr.exclude_guest || + event->attr.sample_period) /* no sampling */ + return -EINVAL; + + /* must be done before validate_group */ + event->hw.event_base = msr; + event->hw.config = cfg; + event->hw.idx = bit; + + return ret; +} + +static void rapl_pmu_event_read(struct perf_event *event) +{ + rapl_event_update(event); +} + +static ssize_t rapl_get_attr_cpumask(struct device *dev, + struct device_attribute *attr, char *buf) +{ + int n = cpulist_scnprintf(buf, PAGE_SIZE - 2, &rapl_cpu_mask); + + buf[n++] = '\n'; + buf[n] = '\0'; + return n; +} + +static DEVICE_ATTR(cpumask, S_IRUGO, rapl_get_attr_cpumask, NULL); + +static struct attribute *rapl_pmu_attrs[] = { + &dev_attr_cpumask.attr, + NULL, +}; + +static struct attribute_group rapl_pmu_attr_group = { + .attrs = rapl_pmu_attrs, +}; + +EVENT_ATTR_STR(energy-cores, rapl_cores, "event=0x01"); +EVENT_ATTR_STR(energy-pkg , rapl_pkg, "event=0x02"); +EVENT_ATTR_STR(energy-ram , rapl_ram, "event=0x03"); +EVENT_ATTR_STR(energy-gpu , rapl_gpu, "event=0x04"); + +EVENT_ATTR_STR(energy-cores.unit, rapl_cores_unit, "Joules"); +EVENT_ATTR_STR(energy-pkg.unit , rapl_pkg_unit, "Joules"); +EVENT_ATTR_STR(energy-ram.unit , rapl_ram_unit, "Joules"); +EVENT_ATTR_STR(energy-gpu.unit , rapl_gpu_unit, "Joules"); + +/* + * we compute in 0.23 nJ increments regardless of MSR + */ +EVENT_ATTR_STR(energy-cores.scale, rapl_cores_scale, "2.3283064365386962890625e-10"); +EVENT_ATTR_STR(energy-pkg.scale, rapl_pkg_scale, "2.3283064365386962890625e-10"); +EVENT_ATTR_STR(energy-ram.scale, rapl_ram_scale, "2.3283064365386962890625e-10"); +EVENT_ATTR_STR(energy-gpu.scale, rapl_gpu_scale, "2.3283064365386962890625e-10"); + +static struct attribute *rapl_events_srv_attr[] = { + EVENT_PTR(rapl_cores), + EVENT_PTR(rapl_pkg), + EVENT_PTR(rapl_ram), + + EVENT_PTR(rapl_cores_unit), + EVENT_PTR(rapl_pkg_unit), + EVENT_PTR(rapl_ram_unit), + + EVENT_PTR(rapl_cores_scale), + EVENT_PTR(rapl_pkg_scale), + EVENT_PTR(rapl_ram_scale), + NULL, +}; + +static struct attribute *rapl_events_cln_attr[] = { + EVENT_PTR(rapl_cores), + EVENT_PTR(rapl_pkg), + EVENT_PTR(rapl_gpu), + + EVENT_PTR(rapl_cores_unit), + EVENT_PTR(rapl_pkg_unit), + EVENT_PTR(rapl_gpu_unit), + + EVENT_PTR(rapl_cores_scale), + EVENT_PTR(rapl_pkg_scale), + EVENT_PTR(rapl_gpu_scale), + NULL, +}; + +static struct attribute *rapl_events_hsw_attr[] = { + EVENT_PTR(rapl_cores), + EVENT_PTR(rapl_pkg), + EVENT_PTR(rapl_gpu), + EVENT_PTR(rapl_ram), + + EVENT_PTR(rapl_cores_unit), + EVENT_PTR(rapl_pkg_unit), + EVENT_PTR(rapl_gpu_unit), + EVENT_PTR(rapl_ram_unit), + + EVENT_PTR(rapl_cores_scale), + EVENT_PTR(rapl_pkg_scale), + EVENT_PTR(rapl_gpu_scale), + EVENT_PTR(rapl_ram_scale), + NULL, +}; + +static struct attribute_group rapl_pmu_events_group = { + .name = "events", + .attrs = NULL, /* patched at runtime */ +}; + +DEFINE_RAPL_FORMAT_ATTR(event, event, "config:0-7"); +static struct attribute *rapl_formats_attr[] = { + &format_attr_event.attr, + NULL, +}; + +static struct attribute_group rapl_pmu_format_group = { + .name = "format", + .attrs = rapl_formats_attr, +}; + +const struct attribute_group *rapl_attr_groups[] = { + &rapl_pmu_attr_group, + &rapl_pmu_format_group, + &rapl_pmu_events_group, + NULL, +}; + +static struct pmu rapl_pmu_class = { + .attr_groups = rapl_attr_groups, + .task_ctx_nr = perf_invalid_context, /* system-wide only */ + .event_init = rapl_pmu_event_init, + .add = rapl_pmu_event_add, /* must have */ + .del = rapl_pmu_event_del, /* must have */ + .start = rapl_pmu_event_start, + .stop = rapl_pmu_event_stop, + .read = rapl_pmu_event_read, +}; + +static void rapl_cpu_exit(int cpu) +{ + struct rapl_pmu *pmu = per_cpu(rapl_pmu, cpu); + int i, phys_id = topology_physical_package_id(cpu); + int target = -1; + + /* find a new cpu on same package */ + for_each_online_cpu(i) { + if (i == cpu) + continue; + if (phys_id == topology_physical_package_id(i)) { + target = i; + break; + } + } + /* + * clear cpu from cpumask + * if was set in cpumask and still some cpu on package, + * then move to new cpu + */ + if (cpumask_test_and_clear_cpu(cpu, &rapl_cpu_mask) && target >= 0) + cpumask_set_cpu(target, &rapl_cpu_mask); + + WARN_ON(cpumask_empty(&rapl_cpu_mask)); + /* + * migrate events and context to new cpu + */ + if (target >= 0) + perf_pmu_migrate_context(pmu->pmu, cpu, target); + + /* cancel overflow polling timer for CPU */ + rapl_stop_hrtimer(pmu); +} + +static void rapl_cpu_init(int cpu) +{ + int i, phys_id = topology_physical_package_id(cpu); + + /* check if phys_is is already covered */ + for_each_cpu(i, &rapl_cpu_mask) { + if (phys_id == topology_physical_package_id(i)) + return; + } + /* was not found, so add it */ + cpumask_set_cpu(cpu, &rapl_cpu_mask); +} + +static int rapl_cpu_prepare(int cpu) +{ + struct rapl_pmu *pmu = per_cpu(rapl_pmu, cpu); + int phys_id = topology_physical_package_id(cpu); + u64 ms; + u64 msr_rapl_power_unit_bits; + + if (pmu) + return 0; + + if (phys_id < 0) + return -1; + + /* protect rdmsrl() to handle virtualization */ + if (rdmsrl_safe(MSR_RAPL_POWER_UNIT, &msr_rapl_power_unit_bits)) + return -1; + + pmu = kzalloc_node(sizeof(*pmu), GFP_KERNEL, cpu_to_node(cpu)); + if (!pmu) + return -1; + + spin_lock_init(&pmu->lock); + + INIT_LIST_HEAD(&pmu->active_list); + + /* + * grab power unit as: 1/2^unit Joules + * + * we cache in local PMU instance + */ + pmu->hw_unit = (msr_rapl_power_unit_bits >> 8) & 0x1FULL; + pmu->pmu = &rapl_pmu_class; + + /* + * use reference of 200W for scaling the timeout + * to avoid missing counter overflows. + * 200W = 200 Joules/sec + * divide interval by 2 to avoid lockstep (2 * 100) + * if hw unit is 32, then we use 2 ms 1/200/2 + */ + if (pmu->hw_unit < 32) + ms = (1000 / (2 * 100)) * (1ULL << (32 - pmu->hw_unit - 1)); + else + ms = 2; + + pmu->timer_interval = ms_to_ktime(ms); + + rapl_hrtimer_init(pmu); + + /* set RAPL pmu for this cpu for now */ + per_cpu(rapl_pmu, cpu) = pmu; + per_cpu(rapl_pmu_to_free, cpu) = NULL; + + return 0; +} + +static void rapl_cpu_kfree(int cpu) +{ + struct rapl_pmu *pmu = per_cpu(rapl_pmu_to_free, cpu); + + kfree(pmu); + + per_cpu(rapl_pmu_to_free, cpu) = NULL; +} + +static int rapl_cpu_dying(int cpu) +{ + struct rapl_pmu *pmu = per_cpu(rapl_pmu, cpu); + + if (!pmu) + return 0; + + per_cpu(rapl_pmu, cpu) = NULL; + + per_cpu(rapl_pmu_to_free, cpu) = pmu; + + return 0; +} + +static int rapl_cpu_notifier(struct notifier_block *self, + unsigned long action, void *hcpu) +{ + unsigned int cpu = (long)hcpu; + + switch (action & ~CPU_TASKS_FROZEN) { + case CPU_UP_PREPARE: + rapl_cpu_prepare(cpu); + break; + case CPU_STARTING: + rapl_cpu_init(cpu); + break; + case CPU_UP_CANCELED: + case CPU_DYING: + rapl_cpu_dying(cpu); + break; + case CPU_ONLINE: + case CPU_DEAD: + rapl_cpu_kfree(cpu); + break; + case CPU_DOWN_PREPARE: + rapl_cpu_exit(cpu); + break; + default: + break; + } + + return NOTIFY_OK; +} + +static const struct x86_cpu_id rapl_cpu_match[] = { + [0] = { .vendor = X86_VENDOR_INTEL, .family = 6 }, + [1] = {}, +}; + +static int __init rapl_pmu_init(void) +{ + struct rapl_pmu *pmu; + int cpu, ret; + + /* + * check for Intel processor family 6 + */ + if (!x86_match_cpu(rapl_cpu_match)) + return 0; + + /* check supported CPU */ + switch (boot_cpu_data.x86_model) { + case 42: /* Sandy Bridge */ + case 58: /* Ivy Bridge */ + rapl_cntr_mask = RAPL_IDX_CLN; + rapl_pmu_events_group.attrs = rapl_events_cln_attr; + break; + case 60: /* Haswell */ + case 69: /* Haswell-Celeron */ + rapl_cntr_mask = RAPL_IDX_HSW; + rapl_pmu_events_group.attrs = rapl_events_hsw_attr; + break; + case 45: /* Sandy Bridge-EP */ + case 62: /* IvyTown */ + rapl_cntr_mask = RAPL_IDX_SRV; + rapl_pmu_events_group.attrs = rapl_events_srv_attr; + break; + + default: + /* unsupported */ + return 0; + } + + cpu_notifier_register_begin(); + + for_each_online_cpu(cpu) { + ret = rapl_cpu_prepare(cpu); + if (ret) + goto out; + rapl_cpu_init(cpu); + } + + __perf_cpu_notifier(rapl_cpu_notifier); + + ret = perf_pmu_register(&rapl_pmu_class, "power", -1); + if (WARN_ON(ret)) { + pr_info("RAPL PMU detected, registration failed (%d), RAPL PMU disabled\n", ret); + cpu_notifier_register_done(); + return -1; + } + + pmu = __get_cpu_var(rapl_pmu); + + pr_info("RAPL PMU detected, hw unit 2^-%d Joules," + " API unit is 2^-32 Joules," + " %d fixed counters" + " %llu ms ovfl timer\n", + pmu->hw_unit, + hweight32(rapl_cntr_mask), + ktime_to_ms(pmu->timer_interval)); + +out: + cpu_notifier_register_done(); + + return 0; +} +device_initcall(rapl_pmu_init); diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore.c b/arch/x86/kernel/cpu/perf_event_intel_uncore.c new file mode 100644 index 00000000000..ae6552a0701 --- /dev/null +++ b/arch/x86/kernel/cpu/perf_event_intel_uncore.c @@ -0,0 +1,4298 @@ +#include "perf_event_intel_uncore.h" + +static struct intel_uncore_type *empty_uncore[] = { NULL, }; +static struct intel_uncore_type **msr_uncores = empty_uncore; +static struct intel_uncore_type **pci_uncores = empty_uncore; +/* pci bus to socket mapping */ +static int pcibus_to_physid[256] = { [0 ... 255] = -1, }; + +static struct pci_dev *extra_pci_dev[UNCORE_SOCKET_MAX][UNCORE_EXTRA_PCI_DEV_MAX]; + +static DEFINE_RAW_SPINLOCK(uncore_box_lock); + +/* mask of cpus that collect uncore events */ +static cpumask_t uncore_cpu_mask; + +/* constraint for the fixed counter */ +static struct event_constraint constraint_fixed = + EVENT_CONSTRAINT(~0ULL, 1 << UNCORE_PMC_IDX_FIXED, ~0ULL); +static struct event_constraint constraint_empty = + EVENT_CONSTRAINT(0, 0, 0); + +#define __BITS_VALUE(x, i, n) ((typeof(x))(((x) >> ((i) * (n))) & \ + ((1ULL << (n)) - 1))) + +DEFINE_UNCORE_FORMAT_ATTR(event, event, "config:0-7"); +DEFINE_UNCORE_FORMAT_ATTR(event_ext, event, "config:0-7,21"); +DEFINE_UNCORE_FORMAT_ATTR(umask, umask, "config:8-15"); +DEFINE_UNCORE_FORMAT_ATTR(edge, edge, "config:18"); +DEFINE_UNCORE_FORMAT_ATTR(tid_en, tid_en, "config:19"); +DEFINE_UNCORE_FORMAT_ATTR(inv, inv, "config:23"); +DEFINE_UNCORE_FORMAT_ATTR(cmask5, cmask, "config:24-28"); +DEFINE_UNCORE_FORMAT_ATTR(cmask8, cmask, "config:24-31"); +DEFINE_UNCORE_FORMAT_ATTR(thresh8, thresh, "config:24-31"); +DEFINE_UNCORE_FORMAT_ATTR(thresh5, thresh, "config:24-28"); +DEFINE_UNCORE_FORMAT_ATTR(occ_sel, occ_sel, "config:14-15"); +DEFINE_UNCORE_FORMAT_ATTR(occ_invert, occ_invert, "config:30"); +DEFINE_UNCORE_FORMAT_ATTR(occ_edge, occ_edge, "config:14-51"); +DEFINE_UNCORE_FORMAT_ATTR(filter_tid, filter_tid, "config1:0-4"); +DEFINE_UNCORE_FORMAT_ATTR(filter_link, filter_link, "config1:5-8"); +DEFINE_UNCORE_FORMAT_ATTR(filter_nid, filter_nid, "config1:10-17"); +DEFINE_UNCORE_FORMAT_ATTR(filter_nid2, filter_nid, "config1:32-47"); +DEFINE_UNCORE_FORMAT_ATTR(filter_state, filter_state, "config1:18-22"); +DEFINE_UNCORE_FORMAT_ATTR(filter_state2, filter_state, "config1:17-22"); +DEFINE_UNCORE_FORMAT_ATTR(filter_opc, filter_opc, "config1:23-31"); +DEFINE_UNCORE_FORMAT_ATTR(filter_opc2, filter_opc, "config1:52-60"); +DEFINE_UNCORE_FORMAT_ATTR(filter_band0, filter_band0, "config1:0-7"); +DEFINE_UNCORE_FORMAT_ATTR(filter_band1, filter_band1, "config1:8-15"); +DEFINE_UNCORE_FORMAT_ATTR(filter_band2, filter_band2, "config1:16-23"); +DEFINE_UNCORE_FORMAT_ATTR(filter_band3, filter_band3, "config1:24-31"); +DEFINE_UNCORE_FORMAT_ATTR(match_rds, match_rds, "config1:48-51"); +DEFINE_UNCORE_FORMAT_ATTR(match_rnid30, match_rnid30, "config1:32-35"); +DEFINE_UNCORE_FORMAT_ATTR(match_rnid4, match_rnid4, "config1:31"); +DEFINE_UNCORE_FORMAT_ATTR(match_dnid, match_dnid, "config1:13-17"); +DEFINE_UNCORE_FORMAT_ATTR(match_mc, match_mc, "config1:9-12"); +DEFINE_UNCORE_FORMAT_ATTR(match_opc, match_opc, "config1:5-8"); +DEFINE_UNCORE_FORMAT_ATTR(match_vnw, match_vnw, "config1:3-4"); +DEFINE_UNCORE_FORMAT_ATTR(match0, match0, "config1:0-31"); +DEFINE_UNCORE_FORMAT_ATTR(match1, match1, "config1:32-63"); +DEFINE_UNCORE_FORMAT_ATTR(mask_rds, mask_rds, "config2:48-51"); +DEFINE_UNCORE_FORMAT_ATTR(mask_rnid30, mask_rnid30, "config2:32-35"); +DEFINE_UNCORE_FORMAT_ATTR(mask_rnid4, mask_rnid4, "config2:31"); +DEFINE_UNCORE_FORMAT_ATTR(mask_dnid, mask_dnid, "config2:13-17"); +DEFINE_UNCORE_FORMAT_ATTR(mask_mc, mask_mc, "config2:9-12"); +DEFINE_UNCORE_FORMAT_ATTR(mask_opc, mask_opc, "config2:5-8"); +DEFINE_UNCORE_FORMAT_ATTR(mask_vnw, mask_vnw, "config2:3-4"); +DEFINE_UNCORE_FORMAT_ATTR(mask0, mask0, "config2:0-31"); +DEFINE_UNCORE_FORMAT_ATTR(mask1, mask1, "config2:32-63"); + +static void uncore_pmu_start_hrtimer(struct intel_uncore_box *box); +static void uncore_pmu_cancel_hrtimer(struct intel_uncore_box *box); +static void uncore_perf_event_update(struct intel_uncore_box *box, struct perf_event *event); +static void uncore_pmu_event_read(struct perf_event *event); + +static struct intel_uncore_pmu *uncore_event_to_pmu(struct perf_event *event) +{ + return container_of(event->pmu, struct intel_uncore_pmu, pmu); +} + +static struct intel_uncore_box * +uncore_pmu_to_box(struct intel_uncore_pmu *pmu, int cpu) +{ + struct intel_uncore_box *box; + + box = *per_cpu_ptr(pmu->box, cpu); + if (box) + return box; + + raw_spin_lock(&uncore_box_lock); + list_for_each_entry(box, &pmu->box_list, list) { + if (box->phys_id == topology_physical_package_id(cpu)) { + atomic_inc(&box->refcnt); + *per_cpu_ptr(pmu->box, cpu) = box; + break; + } + } + raw_spin_unlock(&uncore_box_lock); + + return *per_cpu_ptr(pmu->box, cpu); +} + +static struct intel_uncore_box *uncore_event_to_box(struct perf_event *event) +{ + /* + * perf core schedules event on the basis of cpu, uncore events are + * collected by one of the cpus inside a physical package. + */ + return uncore_pmu_to_box(uncore_event_to_pmu(event), smp_processor_id()); +} + +static u64 uncore_msr_read_counter(struct intel_uncore_box *box, struct perf_event *event) +{ + u64 count; + + rdmsrl(event->hw.event_base, count); + + return count; +} + +/* + * generic get constraint function for shared match/mask registers. + */ +static struct event_constraint * +uncore_get_constraint(struct intel_uncore_box *box, struct perf_event *event) +{ + struct intel_uncore_extra_reg *er; + struct hw_perf_event_extra *reg1 = &event->hw.extra_reg; + struct hw_perf_event_extra *reg2 = &event->hw.branch_reg; + unsigned long flags; + bool ok = false; + + /* + * reg->alloc can be set due to existing state, so for fake box we + * need to ignore this, otherwise we might fail to allocate proper + * fake state for this extra reg constraint. + */ + if (reg1->idx == EXTRA_REG_NONE || + (!uncore_box_is_fake(box) && reg1->alloc)) + return NULL; + + er = &box->shared_regs[reg1->idx]; + raw_spin_lock_irqsave(&er->lock, flags); + if (!atomic_read(&er->ref) || + (er->config1 == reg1->config && er->config2 == reg2->config)) { + atomic_inc(&er->ref); + er->config1 = reg1->config; + er->config2 = reg2->config; + ok = true; + } + raw_spin_unlock_irqrestore(&er->lock, flags); + + if (ok) { + if (!uncore_box_is_fake(box)) + reg1->alloc = 1; + return NULL; + } + + return &constraint_empty; +} + +static void uncore_put_constraint(struct intel_uncore_box *box, struct perf_event *event) +{ + struct intel_uncore_extra_reg *er; + struct hw_perf_event_extra *reg1 = &event->hw.extra_reg; + + /* + * Only put constraint if extra reg was actually allocated. Also + * takes care of event which do not use an extra shared reg. + * + * Also, if this is a fake box we shouldn't touch any event state + * (reg->alloc) and we don't care about leaving inconsistent box + * state either since it will be thrown out. + */ + if (uncore_box_is_fake(box) || !reg1->alloc) + return; + + er = &box->shared_regs[reg1->idx]; + atomic_dec(&er->ref); + reg1->alloc = 0; +} + +static u64 uncore_shared_reg_config(struct intel_uncore_box *box, int idx) +{ + struct intel_uncore_extra_reg *er; + unsigned long flags; + u64 config; + + er = &box->shared_regs[idx]; + + raw_spin_lock_irqsave(&er->lock, flags); + config = er->config; + raw_spin_unlock_irqrestore(&er->lock, flags); + + return config; +} + +/* Sandy Bridge-EP uncore support */ +static struct intel_uncore_type snbep_uncore_cbox; +static struct intel_uncore_type snbep_uncore_pcu; + +static void snbep_uncore_pci_disable_box(struct intel_uncore_box *box) +{ + struct pci_dev *pdev = box->pci_dev; + int box_ctl = uncore_pci_box_ctl(box); + u32 config = 0; + + if (!pci_read_config_dword(pdev, box_ctl, &config)) { + config |= SNBEP_PMON_BOX_CTL_FRZ; + pci_write_config_dword(pdev, box_ctl, config); + } +} + +static void snbep_uncore_pci_enable_box(struct intel_uncore_box *box) +{ + struct pci_dev *pdev = box->pci_dev; + int box_ctl = uncore_pci_box_ctl(box); + u32 config = 0; + + if (!pci_read_config_dword(pdev, box_ctl, &config)) { + config &= ~SNBEP_PMON_BOX_CTL_FRZ; + pci_write_config_dword(pdev, box_ctl, config); + } +} + +static void snbep_uncore_pci_enable_event(struct intel_uncore_box *box, struct perf_event *event) +{ + struct pci_dev *pdev = box->pci_dev; + struct hw_perf_event *hwc = &event->hw; + + pci_write_config_dword(pdev, hwc->config_base, hwc->config | SNBEP_PMON_CTL_EN); +} + +static void snbep_uncore_pci_disable_event(struct intel_uncore_box *box, struct perf_event *event) +{ + struct pci_dev *pdev = box->pci_dev; + struct hw_perf_event *hwc = &event->hw; + + pci_write_config_dword(pdev, hwc->config_base, hwc->config); +} + +static u64 snbep_uncore_pci_read_counter(struct intel_uncore_box *box, struct perf_event *event) +{ + struct pci_dev *pdev = box->pci_dev; + struct hw_perf_event *hwc = &event->hw; + u64 count = 0; + + pci_read_config_dword(pdev, hwc->event_base, (u32 *)&count); + pci_read_config_dword(pdev, hwc->event_base + 4, (u32 *)&count + 1); + + return count; +} + +static void snbep_uncore_pci_init_box(struct intel_uncore_box *box) +{ + struct pci_dev *pdev = box->pci_dev; + + pci_write_config_dword(pdev, SNBEP_PCI_PMON_BOX_CTL, SNBEP_PMON_BOX_CTL_INT); +} + +static void snbep_uncore_msr_disable_box(struct intel_uncore_box *box) +{ + u64 config; + unsigned msr; + + msr = uncore_msr_box_ctl(box); + if (msr) { + rdmsrl(msr, config); + config |= SNBEP_PMON_BOX_CTL_FRZ; + wrmsrl(msr, config); + } +} + +static void snbep_uncore_msr_enable_box(struct intel_uncore_box *box) +{ + u64 config; + unsigned msr; + + msr = uncore_msr_box_ctl(box); + if (msr) { + rdmsrl(msr, config); + config &= ~SNBEP_PMON_BOX_CTL_FRZ; + wrmsrl(msr, config); + } +} + +static void snbep_uncore_msr_enable_event(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &hwc->extra_reg; + + if (reg1->idx != EXTRA_REG_NONE) + wrmsrl(reg1->reg, uncore_shared_reg_config(box, 0)); + + wrmsrl(hwc->config_base, hwc->config | SNBEP_PMON_CTL_EN); +} + +static void snbep_uncore_msr_disable_event(struct intel_uncore_box *box, + struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + + wrmsrl(hwc->config_base, hwc->config); +} + +static void snbep_uncore_msr_init_box(struct intel_uncore_box *box) +{ + unsigned msr = uncore_msr_box_ctl(box); + + if (msr) + wrmsrl(msr, SNBEP_PMON_BOX_CTL_INT); +} + +static struct attribute *snbep_uncore_formats_attr[] = { + &format_attr_event.attr, + &format_attr_umask.attr, + &format_attr_edge.attr, + &format_attr_inv.attr, + &format_attr_thresh8.attr, + NULL, +}; + +static struct attribute *snbep_uncore_ubox_formats_attr[] = { + &format_attr_event.attr, + &format_attr_umask.attr, + &format_attr_edge.attr, + &format_attr_inv.attr, + &format_attr_thresh5.attr, + NULL, +}; + +static struct attribute *snbep_uncore_cbox_formats_attr[] = { + &format_attr_event.attr, + &format_attr_umask.attr, + &format_attr_edge.attr, + &format_attr_tid_en.attr, + &format_attr_inv.attr, + &format_attr_thresh8.attr, + &format_attr_filter_tid.attr, + &format_attr_filter_nid.attr, + &format_attr_filter_state.attr, + &format_attr_filter_opc.attr, + NULL, +}; + +static struct attribute *snbep_uncore_pcu_formats_attr[] = { + &format_attr_event_ext.attr, + &format_attr_occ_sel.attr, + &format_attr_edge.attr, + &format_attr_inv.attr, + &format_attr_thresh5.attr, + &format_attr_occ_invert.attr, + &format_attr_occ_edge.attr, + &format_attr_filter_band0.attr, + &format_attr_filter_band1.attr, + &format_attr_filter_band2.attr, + &format_attr_filter_band3.attr, + NULL, +}; + +static struct attribute *snbep_uncore_qpi_formats_attr[] = { + &format_attr_event_ext.attr, + &format_attr_umask.attr, + &format_attr_edge.attr, + &format_attr_inv.attr, + &format_attr_thresh8.attr, + &format_attr_match_rds.attr, + &format_attr_match_rnid30.attr, + &format_attr_match_rnid4.attr, + &format_attr_match_dnid.attr, + &format_attr_match_mc.attr, + &format_attr_match_opc.attr, + &format_attr_match_vnw.attr, + &format_attr_match0.attr, + &format_attr_match1.attr, + &format_attr_mask_rds.attr, + &format_attr_mask_rnid30.attr, + &format_attr_mask_rnid4.attr, + &format_attr_mask_dnid.attr, + &format_attr_mask_mc.attr, + &format_attr_mask_opc.attr, + &format_attr_mask_vnw.attr, + &format_attr_mask0.attr, + &format_attr_mask1.attr, + NULL, +}; + +static struct uncore_event_desc snbep_uncore_imc_events[] = { + INTEL_UNCORE_EVENT_DESC(clockticks, "event=0xff,umask=0x00"), + INTEL_UNCORE_EVENT_DESC(cas_count_read, "event=0x04,umask=0x03"), + INTEL_UNCORE_EVENT_DESC(cas_count_write, "event=0x04,umask=0x0c"), + { /* end: all zeroes */ }, +}; + +static struct uncore_event_desc snbep_uncore_qpi_events[] = { + INTEL_UNCORE_EVENT_DESC(clockticks, "event=0x14"), + INTEL_UNCORE_EVENT_DESC(txl_flits_active, "event=0x00,umask=0x06"), + INTEL_UNCORE_EVENT_DESC(drs_data, "event=0x102,umask=0x08"), + INTEL_UNCORE_EVENT_DESC(ncb_data, "event=0x103,umask=0x04"), + { /* end: all zeroes */ }, +}; + +static struct attribute_group snbep_uncore_format_group = { + .name = "format", + .attrs = snbep_uncore_formats_attr, +}; + +static struct attribute_group snbep_uncore_ubox_format_group = { + .name = "format", + .attrs = snbep_uncore_ubox_formats_attr, +}; + +static struct attribute_group snbep_uncore_cbox_format_group = { + .name = "format", + .attrs = snbep_uncore_cbox_formats_attr, +}; + +static struct attribute_group snbep_uncore_pcu_format_group = { + .name = "format", + .attrs = snbep_uncore_pcu_formats_attr, +}; + +static struct attribute_group snbep_uncore_qpi_format_group = { + .name = "format", + .attrs = snbep_uncore_qpi_formats_attr, +}; + +#define SNBEP_UNCORE_MSR_OPS_COMMON_INIT() \ + .init_box = snbep_uncore_msr_init_box, \ + .disable_box = snbep_uncore_msr_disable_box, \ + .enable_box = snbep_uncore_msr_enable_box, \ + .disable_event = snbep_uncore_msr_disable_event, \ + .enable_event = snbep_uncore_msr_enable_event, \ + .read_counter = uncore_msr_read_counter + +static struct intel_uncore_ops snbep_uncore_msr_ops = { + SNBEP_UNCORE_MSR_OPS_COMMON_INIT(), +}; + +#define SNBEP_UNCORE_PCI_OPS_COMMON_INIT() \ + .init_box = snbep_uncore_pci_init_box, \ + .disable_box = snbep_uncore_pci_disable_box, \ + .enable_box = snbep_uncore_pci_enable_box, \ + .disable_event = snbep_uncore_pci_disable_event, \ + .read_counter = snbep_uncore_pci_read_counter + +static struct intel_uncore_ops snbep_uncore_pci_ops = { + SNBEP_UNCORE_PCI_OPS_COMMON_INIT(), + .enable_event = snbep_uncore_pci_enable_event, \ +}; + +static struct event_constraint snbep_uncore_cbox_constraints[] = { + UNCORE_EVENT_CONSTRAINT(0x01, 0x1), + UNCORE_EVENT_CONSTRAINT(0x02, 0x3), + UNCORE_EVENT_CONSTRAINT(0x04, 0x3), + UNCORE_EVENT_CONSTRAINT(0x05, 0x3), + UNCORE_EVENT_CONSTRAINT(0x07, 0x3), + UNCORE_EVENT_CONSTRAINT(0x09, 0x3), + UNCORE_EVENT_CONSTRAINT(0x11, 0x1), + UNCORE_EVENT_CONSTRAINT(0x12, 0x3), + UNCORE_EVENT_CONSTRAINT(0x13, 0x3), + UNCORE_EVENT_CONSTRAINT(0x1b, 0xc), + UNCORE_EVENT_CONSTRAINT(0x1c, 0xc), + UNCORE_EVENT_CONSTRAINT(0x1d, 0xc), + UNCORE_EVENT_CONSTRAINT(0x1e, 0xc), + EVENT_CONSTRAINT_OVERLAP(0x1f, 0xe, 0xff), + UNCORE_EVENT_CONSTRAINT(0x21, 0x3), + UNCORE_EVENT_CONSTRAINT(0x23, 0x3), + UNCORE_EVENT_CONSTRAINT(0x31, 0x3), + UNCORE_EVENT_CONSTRAINT(0x32, 0x3), + UNCORE_EVENT_CONSTRAINT(0x33, 0x3), + UNCORE_EVENT_CONSTRAINT(0x34, 0x3), + UNCORE_EVENT_CONSTRAINT(0x35, 0x3), + UNCORE_EVENT_CONSTRAINT(0x36, 0x1), + UNCORE_EVENT_CONSTRAINT(0x37, 0x3), + UNCORE_EVENT_CONSTRAINT(0x38, 0x3), + UNCORE_EVENT_CONSTRAINT(0x39, 0x3), + UNCORE_EVENT_CONSTRAINT(0x3b, 0x1), + EVENT_CONSTRAINT_END +}; + +static struct event_constraint snbep_uncore_r2pcie_constraints[] = { + UNCORE_EVENT_CONSTRAINT(0x10, 0x3), + UNCORE_EVENT_CONSTRAINT(0x11, 0x3), + UNCORE_EVENT_CONSTRAINT(0x12, 0x1), + UNCORE_EVENT_CONSTRAINT(0x23, 0x3), + UNCORE_EVENT_CONSTRAINT(0x24, 0x3), + UNCORE_EVENT_CONSTRAINT(0x25, 0x3), + UNCORE_EVENT_CONSTRAINT(0x26, 0x3), + UNCORE_EVENT_CONSTRAINT(0x32, 0x3), + UNCORE_EVENT_CONSTRAINT(0x33, 0x3), + UNCORE_EVENT_CONSTRAINT(0x34, 0x3), + EVENT_CONSTRAINT_END +}; + +static struct event_constraint snbep_uncore_r3qpi_constraints[] = { + UNCORE_EVENT_CONSTRAINT(0x10, 0x3), + UNCORE_EVENT_CONSTRAINT(0x11, 0x3), + UNCORE_EVENT_CONSTRAINT(0x12, 0x3), + UNCORE_EVENT_CONSTRAINT(0x13, 0x1), + UNCORE_EVENT_CONSTRAINT(0x20, 0x3), + UNCORE_EVENT_CONSTRAINT(0x21, 0x3), + UNCORE_EVENT_CONSTRAINT(0x22, 0x3), + UNCORE_EVENT_CONSTRAINT(0x23, 0x3), + UNCORE_EVENT_CONSTRAINT(0x24, 0x3), + UNCORE_EVENT_CONSTRAINT(0x25, 0x3), + UNCORE_EVENT_CONSTRAINT(0x26, 0x3), + UNCORE_EVENT_CONSTRAINT(0x28, 0x3), + UNCORE_EVENT_CONSTRAINT(0x29, 0x3), + UNCORE_EVENT_CONSTRAINT(0x2a, 0x3), + UNCORE_EVENT_CONSTRAINT(0x2b, 0x3), + UNCORE_EVENT_CONSTRAINT(0x2c, 0x3), + UNCORE_EVENT_CONSTRAINT(0x2d, 0x3), + UNCORE_EVENT_CONSTRAINT(0x2e, 0x3), + UNCORE_EVENT_CONSTRAINT(0x2f, 0x3), + UNCORE_EVENT_CONSTRAINT(0x30, 0x3), + UNCORE_EVENT_CONSTRAINT(0x31, 0x3), + UNCORE_EVENT_CONSTRAINT(0x32, 0x3), + UNCORE_EVENT_CONSTRAINT(0x33, 0x3), + UNCORE_EVENT_CONSTRAINT(0x34, 0x3), + UNCORE_EVENT_CONSTRAINT(0x36, 0x3), + UNCORE_EVENT_CONSTRAINT(0x37, 0x3), + UNCORE_EVENT_CONSTRAINT(0x38, 0x3), + UNCORE_EVENT_CONSTRAINT(0x39, 0x3), + EVENT_CONSTRAINT_END +}; + +static struct intel_uncore_type snbep_uncore_ubox = { + .name = "ubox", + .num_counters = 2, + .num_boxes = 1, + .perf_ctr_bits = 44, + .fixed_ctr_bits = 48, + .perf_ctr = SNBEP_U_MSR_PMON_CTR0, + .event_ctl = SNBEP_U_MSR_PMON_CTL0, + .event_mask = SNBEP_U_MSR_PMON_RAW_EVENT_MASK, + .fixed_ctr = SNBEP_U_MSR_PMON_UCLK_FIXED_CTR, + .fixed_ctl = SNBEP_U_MSR_PMON_UCLK_FIXED_CTL, + .ops = &snbep_uncore_msr_ops, + .format_group = &snbep_uncore_ubox_format_group, +}; + +static struct extra_reg snbep_uncore_cbox_extra_regs[] = { + SNBEP_CBO_EVENT_EXTRA_REG(SNBEP_CBO_PMON_CTL_TID_EN, + SNBEP_CBO_PMON_CTL_TID_EN, 0x1), + SNBEP_CBO_EVENT_EXTRA_REG(0x0334, 0xffff, 0x4), + SNBEP_CBO_EVENT_EXTRA_REG(0x4334, 0xffff, 0x6), + SNBEP_CBO_EVENT_EXTRA_REG(0x0534, 0xffff, 0x4), + SNBEP_CBO_EVENT_EXTRA_REG(0x4534, 0xffff, 0x6), + SNBEP_CBO_EVENT_EXTRA_REG(0x0934, 0xffff, 0x4), + SNBEP_CBO_EVENT_EXTRA_REG(0x4934, 0xffff, 0x6), + SNBEP_CBO_EVENT_EXTRA_REG(0x4134, 0xffff, 0x6), + SNBEP_CBO_EVENT_EXTRA_REG(0x0135, 0xffff, 0x8), + SNBEP_CBO_EVENT_EXTRA_REG(0x0335, 0xffff, 0x8), + SNBEP_CBO_EVENT_EXTRA_REG(0x4135, 0xffff, 0xa), + SNBEP_CBO_EVENT_EXTRA_REG(0x4335, 0xffff, 0xa), + SNBEP_CBO_EVENT_EXTRA_REG(0x4435, 0xffff, 0x2), + SNBEP_CBO_EVENT_EXTRA_REG(0x4835, 0xffff, 0x2), + SNBEP_CBO_EVENT_EXTRA_REG(0x4a35, 0xffff, 0x2), + SNBEP_CBO_EVENT_EXTRA_REG(0x5035, 0xffff, 0x2), + SNBEP_CBO_EVENT_EXTRA_REG(0x0136, 0xffff, 0x8), + SNBEP_CBO_EVENT_EXTRA_REG(0x0336, 0xffff, 0x8), + SNBEP_CBO_EVENT_EXTRA_REG(0x4136, 0xffff, 0xa), + SNBEP_CBO_EVENT_EXTRA_REG(0x4336, 0xffff, 0xa), + SNBEP_CBO_EVENT_EXTRA_REG(0x4436, 0xffff, 0x2), + SNBEP_CBO_EVENT_EXTRA_REG(0x4836, 0xffff, 0x2), + SNBEP_CBO_EVENT_EXTRA_REG(0x4a36, 0xffff, 0x2), + SNBEP_CBO_EVENT_EXTRA_REG(0x4037, 0x40ff, 0x2), + EVENT_EXTRA_END +}; + +static void snbep_cbox_put_constraint(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event_extra *reg1 = &event->hw.extra_reg; + struct intel_uncore_extra_reg *er = &box->shared_regs[0]; + int i; + + if (uncore_box_is_fake(box)) + return; + + for (i = 0; i < 5; i++) { + if (reg1->alloc & (0x1 << i)) + atomic_sub(1 << (i * 6), &er->ref); + } + reg1->alloc = 0; +} + +static struct event_constraint * +__snbep_cbox_get_constraint(struct intel_uncore_box *box, struct perf_event *event, + u64 (*cbox_filter_mask)(int fields)) +{ + struct hw_perf_event_extra *reg1 = &event->hw.extra_reg; + struct intel_uncore_extra_reg *er = &box->shared_regs[0]; + int i, alloc = 0; + unsigned long flags; + u64 mask; + + if (reg1->idx == EXTRA_REG_NONE) + return NULL; + + raw_spin_lock_irqsave(&er->lock, flags); + for (i = 0; i < 5; i++) { + if (!(reg1->idx & (0x1 << i))) + continue; + if (!uncore_box_is_fake(box) && (reg1->alloc & (0x1 << i))) + continue; + + mask = cbox_filter_mask(0x1 << i); + if (!__BITS_VALUE(atomic_read(&er->ref), i, 6) || + !((reg1->config ^ er->config) & mask)) { + atomic_add(1 << (i * 6), &er->ref); + er->config &= ~mask; + er->config |= reg1->config & mask; + alloc |= (0x1 << i); + } else { + break; + } + } + raw_spin_unlock_irqrestore(&er->lock, flags); + if (i < 5) + goto fail; + + if (!uncore_box_is_fake(box)) + reg1->alloc |= alloc; + + return NULL; +fail: + for (; i >= 0; i--) { + if (alloc & (0x1 << i)) + atomic_sub(1 << (i * 6), &er->ref); + } + return &constraint_empty; +} + +static u64 snbep_cbox_filter_mask(int fields) +{ + u64 mask = 0; + + if (fields & 0x1) + mask |= SNBEP_CB0_MSR_PMON_BOX_FILTER_TID; + if (fields & 0x2) + mask |= SNBEP_CB0_MSR_PMON_BOX_FILTER_NID; + if (fields & 0x4) + mask |= SNBEP_CB0_MSR_PMON_BOX_FILTER_STATE; + if (fields & 0x8) + mask |= SNBEP_CB0_MSR_PMON_BOX_FILTER_OPC; + + return mask; +} + +static struct event_constraint * +snbep_cbox_get_constraint(struct intel_uncore_box *box, struct perf_event *event) +{ + return __snbep_cbox_get_constraint(box, event, snbep_cbox_filter_mask); +} + +static int snbep_cbox_hw_config(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event_extra *reg1 = &event->hw.extra_reg; + struct extra_reg *er; + int idx = 0; + + for (er = snbep_uncore_cbox_extra_regs; er->msr; er++) { + if (er->event != (event->hw.config & er->config_mask)) + continue; + idx |= er->idx; + } + + if (idx) { + reg1->reg = SNBEP_C0_MSR_PMON_BOX_FILTER + + SNBEP_CBO_MSR_OFFSET * box->pmu->pmu_idx; + reg1->config = event->attr.config1 & snbep_cbox_filter_mask(idx); + reg1->idx = idx; + } + return 0; +} + +static struct intel_uncore_ops snbep_uncore_cbox_ops = { + SNBEP_UNCORE_MSR_OPS_COMMON_INIT(), + .hw_config = snbep_cbox_hw_config, + .get_constraint = snbep_cbox_get_constraint, + .put_constraint = snbep_cbox_put_constraint, +}; + +static struct intel_uncore_type snbep_uncore_cbox = { + .name = "cbox", + .num_counters = 4, + .num_boxes = 8, + .perf_ctr_bits = 44, + .event_ctl = SNBEP_C0_MSR_PMON_CTL0, + .perf_ctr = SNBEP_C0_MSR_PMON_CTR0, + .event_mask = SNBEP_CBO_MSR_PMON_RAW_EVENT_MASK, + .box_ctl = SNBEP_C0_MSR_PMON_BOX_CTL, + .msr_offset = SNBEP_CBO_MSR_OFFSET, + .num_shared_regs = 1, + .constraints = snbep_uncore_cbox_constraints, + .ops = &snbep_uncore_cbox_ops, + .format_group = &snbep_uncore_cbox_format_group, +}; + +static u64 snbep_pcu_alter_er(struct perf_event *event, int new_idx, bool modify) +{ + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &hwc->extra_reg; + u64 config = reg1->config; + + if (new_idx > reg1->idx) + config <<= 8 * (new_idx - reg1->idx); + else + config >>= 8 * (reg1->idx - new_idx); + + if (modify) { + hwc->config += new_idx - reg1->idx; + reg1->config = config; + reg1->idx = new_idx; + } + return config; +} + +static struct event_constraint * +snbep_pcu_get_constraint(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event_extra *reg1 = &event->hw.extra_reg; + struct intel_uncore_extra_reg *er = &box->shared_regs[0]; + unsigned long flags; + int idx = reg1->idx; + u64 mask, config1 = reg1->config; + bool ok = false; + + if (reg1->idx == EXTRA_REG_NONE || + (!uncore_box_is_fake(box) && reg1->alloc)) + return NULL; +again: + mask = 0xffULL << (idx * 8); + raw_spin_lock_irqsave(&er->lock, flags); + if (!__BITS_VALUE(atomic_read(&er->ref), idx, 8) || + !((config1 ^ er->config) & mask)) { + atomic_add(1 << (idx * 8), &er->ref); + er->config &= ~mask; + er->config |= config1 & mask; + ok = true; + } + raw_spin_unlock_irqrestore(&er->lock, flags); + + if (!ok) { + idx = (idx + 1) % 4; + if (idx != reg1->idx) { + config1 = snbep_pcu_alter_er(event, idx, false); + goto again; + } + return &constraint_empty; + } + + if (!uncore_box_is_fake(box)) { + if (idx != reg1->idx) + snbep_pcu_alter_er(event, idx, true); + reg1->alloc = 1; + } + return NULL; +} + +static void snbep_pcu_put_constraint(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event_extra *reg1 = &event->hw.extra_reg; + struct intel_uncore_extra_reg *er = &box->shared_regs[0]; + + if (uncore_box_is_fake(box) || !reg1->alloc) + return; + + atomic_sub(1 << (reg1->idx * 8), &er->ref); + reg1->alloc = 0; +} + +static int snbep_pcu_hw_config(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &hwc->extra_reg; + int ev_sel = hwc->config & SNBEP_PMON_CTL_EV_SEL_MASK; + + if (ev_sel >= 0xb && ev_sel <= 0xe) { + reg1->reg = SNBEP_PCU_MSR_PMON_BOX_FILTER; + reg1->idx = ev_sel - 0xb; + reg1->config = event->attr.config1 & (0xff << reg1->idx); + } + return 0; +} + +static struct intel_uncore_ops snbep_uncore_pcu_ops = { + SNBEP_UNCORE_MSR_OPS_COMMON_INIT(), + .hw_config = snbep_pcu_hw_config, + .get_constraint = snbep_pcu_get_constraint, + .put_constraint = snbep_pcu_put_constraint, +}; + +static struct intel_uncore_type snbep_uncore_pcu = { + .name = "pcu", + .num_counters = 4, + .num_boxes = 1, + .perf_ctr_bits = 48, + .perf_ctr = SNBEP_PCU_MSR_PMON_CTR0, + .event_ctl = SNBEP_PCU_MSR_PMON_CTL0, + .event_mask = SNBEP_PCU_MSR_PMON_RAW_EVENT_MASK, + .box_ctl = SNBEP_PCU_MSR_PMON_BOX_CTL, + .num_shared_regs = 1, + .ops = &snbep_uncore_pcu_ops, + .format_group = &snbep_uncore_pcu_format_group, +}; + +static struct intel_uncore_type *snbep_msr_uncores[] = { + &snbep_uncore_ubox, + &snbep_uncore_cbox, + &snbep_uncore_pcu, + NULL, +}; + +enum { + SNBEP_PCI_QPI_PORT0_FILTER, + SNBEP_PCI_QPI_PORT1_FILTER, +}; + +static int snbep_qpi_hw_config(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &hwc->extra_reg; + struct hw_perf_event_extra *reg2 = &hwc->branch_reg; + + if ((hwc->config & SNBEP_PMON_CTL_EV_SEL_MASK) == 0x38) { + reg1->idx = 0; + reg1->reg = SNBEP_Q_Py_PCI_PMON_PKT_MATCH0; + reg1->config = event->attr.config1; + reg2->reg = SNBEP_Q_Py_PCI_PMON_PKT_MASK0; + reg2->config = event->attr.config2; + } + return 0; +} + +static void snbep_qpi_enable_event(struct intel_uncore_box *box, struct perf_event *event) +{ + struct pci_dev *pdev = box->pci_dev; + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &hwc->extra_reg; + struct hw_perf_event_extra *reg2 = &hwc->branch_reg; + + if (reg1->idx != EXTRA_REG_NONE) { + int idx = box->pmu->pmu_idx + SNBEP_PCI_QPI_PORT0_FILTER; + struct pci_dev *filter_pdev = extra_pci_dev[box->phys_id][idx]; + WARN_ON_ONCE(!filter_pdev); + if (filter_pdev) { + pci_write_config_dword(filter_pdev, reg1->reg, + (u32)reg1->config); + pci_write_config_dword(filter_pdev, reg1->reg + 4, + (u32)(reg1->config >> 32)); + pci_write_config_dword(filter_pdev, reg2->reg, + (u32)reg2->config); + pci_write_config_dword(filter_pdev, reg2->reg + 4, + (u32)(reg2->config >> 32)); + } + } + + pci_write_config_dword(pdev, hwc->config_base, hwc->config | SNBEP_PMON_CTL_EN); +} + +static struct intel_uncore_ops snbep_uncore_qpi_ops = { + SNBEP_UNCORE_PCI_OPS_COMMON_INIT(), + .enable_event = snbep_qpi_enable_event, + .hw_config = snbep_qpi_hw_config, + .get_constraint = uncore_get_constraint, + .put_constraint = uncore_put_constraint, +}; + +#define SNBEP_UNCORE_PCI_COMMON_INIT() \ + .perf_ctr = SNBEP_PCI_PMON_CTR0, \ + .event_ctl = SNBEP_PCI_PMON_CTL0, \ + .event_mask = SNBEP_PMON_RAW_EVENT_MASK, \ + .box_ctl = SNBEP_PCI_PMON_BOX_CTL, \ + .ops = &snbep_uncore_pci_ops, \ + .format_group = &snbep_uncore_format_group + +static struct intel_uncore_type snbep_uncore_ha = { + .name = "ha", + .num_counters = 4, + .num_boxes = 1, + .perf_ctr_bits = 48, + SNBEP_UNCORE_PCI_COMMON_INIT(), +}; + +static struct intel_uncore_type snbep_uncore_imc = { + .name = "imc", + .num_counters = 4, + .num_boxes = 4, + .perf_ctr_bits = 48, + .fixed_ctr_bits = 48, + .fixed_ctr = SNBEP_MC_CHy_PCI_PMON_FIXED_CTR, + .fixed_ctl = SNBEP_MC_CHy_PCI_PMON_FIXED_CTL, + .event_descs = snbep_uncore_imc_events, + SNBEP_UNCORE_PCI_COMMON_INIT(), +}; + +static struct intel_uncore_type snbep_uncore_qpi = { + .name = "qpi", + .num_counters = 4, + .num_boxes = 2, + .perf_ctr_bits = 48, + .perf_ctr = SNBEP_PCI_PMON_CTR0, + .event_ctl = SNBEP_PCI_PMON_CTL0, + .event_mask = SNBEP_QPI_PCI_PMON_RAW_EVENT_MASK, + .box_ctl = SNBEP_PCI_PMON_BOX_CTL, + .num_shared_regs = 1, + .ops = &snbep_uncore_qpi_ops, + .event_descs = snbep_uncore_qpi_events, + .format_group = &snbep_uncore_qpi_format_group, +}; + + +static struct intel_uncore_type snbep_uncore_r2pcie = { + .name = "r2pcie", + .num_counters = 4, + .num_boxes = 1, + .perf_ctr_bits = 44, + .constraints = snbep_uncore_r2pcie_constraints, + SNBEP_UNCORE_PCI_COMMON_INIT(), +}; + +static struct intel_uncore_type snbep_uncore_r3qpi = { + .name = "r3qpi", + .num_counters = 3, + .num_boxes = 2, + .perf_ctr_bits = 44, + .constraints = snbep_uncore_r3qpi_constraints, + SNBEP_UNCORE_PCI_COMMON_INIT(), +}; + +enum { + SNBEP_PCI_UNCORE_HA, + SNBEP_PCI_UNCORE_IMC, + SNBEP_PCI_UNCORE_QPI, + SNBEP_PCI_UNCORE_R2PCIE, + SNBEP_PCI_UNCORE_R3QPI, +}; + +static struct intel_uncore_type *snbep_pci_uncores[] = { + [SNBEP_PCI_UNCORE_HA] = &snbep_uncore_ha, + [SNBEP_PCI_UNCORE_IMC] = &snbep_uncore_imc, + [SNBEP_PCI_UNCORE_QPI] = &snbep_uncore_qpi, + [SNBEP_PCI_UNCORE_R2PCIE] = &snbep_uncore_r2pcie, + [SNBEP_PCI_UNCORE_R3QPI] = &snbep_uncore_r3qpi, + NULL, +}; + +static DEFINE_PCI_DEVICE_TABLE(snbep_uncore_pci_ids) = { + { /* Home Agent */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_UNC_HA), + .driver_data = UNCORE_PCI_DEV_DATA(SNBEP_PCI_UNCORE_HA, 0), + }, + { /* MC Channel 0 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_UNC_IMC0), + .driver_data = UNCORE_PCI_DEV_DATA(SNBEP_PCI_UNCORE_IMC, 0), + }, + { /* MC Channel 1 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_UNC_IMC1), + .driver_data = UNCORE_PCI_DEV_DATA(SNBEP_PCI_UNCORE_IMC, 1), + }, + { /* MC Channel 2 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_UNC_IMC2), + .driver_data = UNCORE_PCI_DEV_DATA(SNBEP_PCI_UNCORE_IMC, 2), + }, + { /* MC Channel 3 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_UNC_IMC3), + .driver_data = UNCORE_PCI_DEV_DATA(SNBEP_PCI_UNCORE_IMC, 3), + }, + { /* QPI Port 0 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_UNC_QPI0), + .driver_data = UNCORE_PCI_DEV_DATA(SNBEP_PCI_UNCORE_QPI, 0), + }, + { /* QPI Port 1 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_UNC_QPI1), + .driver_data = UNCORE_PCI_DEV_DATA(SNBEP_PCI_UNCORE_QPI, 1), + }, + { /* R2PCIe */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_UNC_R2PCIE), + .driver_data = UNCORE_PCI_DEV_DATA(SNBEP_PCI_UNCORE_R2PCIE, 0), + }, + { /* R3QPI Link 0 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_UNC_R3QPI0), + .driver_data = UNCORE_PCI_DEV_DATA(SNBEP_PCI_UNCORE_R3QPI, 0), + }, + { /* R3QPI Link 1 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_UNC_R3QPI1), + .driver_data = UNCORE_PCI_DEV_DATA(SNBEP_PCI_UNCORE_R3QPI, 1), + }, + { /* QPI Port 0 filter */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x3c86), + .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, + SNBEP_PCI_QPI_PORT0_FILTER), + }, + { /* QPI Port 0 filter */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0x3c96), + .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, + SNBEP_PCI_QPI_PORT1_FILTER), + }, + { /* end: all zeroes */ } +}; + +static struct pci_driver snbep_uncore_pci_driver = { + .name = "snbep_uncore", + .id_table = snbep_uncore_pci_ids, +}; + +/* + * build pci bus to socket mapping + */ +static int snbep_pci2phy_map_init(int devid) +{ + struct pci_dev *ubox_dev = NULL; + int i, bus, nodeid; + int err = 0; + u32 config = 0; + + while (1) { + /* find the UBOX device */ + ubox_dev = pci_get_device(PCI_VENDOR_ID_INTEL, devid, ubox_dev); + if (!ubox_dev) + break; + bus = ubox_dev->bus->number; + /* get the Node ID of the local register */ + err = pci_read_config_dword(ubox_dev, 0x40, &config); + if (err) + break; + nodeid = config; + /* get the Node ID mapping */ + err = pci_read_config_dword(ubox_dev, 0x54, &config); + if (err) + break; + /* + * every three bits in the Node ID mapping register maps + * to a particular node. + */ + for (i = 0; i < 8; i++) { + if (nodeid == ((config >> (3 * i)) & 0x7)) { + pcibus_to_physid[bus] = i; + break; + } + } + } + + if (!err) { + /* + * For PCI bus with no UBOX device, find the next bus + * that has UBOX device and use its mapping. + */ + i = -1; + for (bus = 255; bus >= 0; bus--) { + if (pcibus_to_physid[bus] >= 0) + i = pcibus_to_physid[bus]; + else + pcibus_to_physid[bus] = i; + } + } + + if (ubox_dev) + pci_dev_put(ubox_dev); + + return err ? pcibios_err_to_errno(err) : 0; +} +/* end of Sandy Bridge-EP uncore support */ + +/* IvyTown uncore support */ +static void ivt_uncore_msr_init_box(struct intel_uncore_box *box) +{ + unsigned msr = uncore_msr_box_ctl(box); + if (msr) + wrmsrl(msr, IVT_PMON_BOX_CTL_INT); +} + +static void ivt_uncore_pci_init_box(struct intel_uncore_box *box) +{ + struct pci_dev *pdev = box->pci_dev; + + pci_write_config_dword(pdev, SNBEP_PCI_PMON_BOX_CTL, IVT_PMON_BOX_CTL_INT); +} + +#define IVT_UNCORE_MSR_OPS_COMMON_INIT() \ + .init_box = ivt_uncore_msr_init_box, \ + .disable_box = snbep_uncore_msr_disable_box, \ + .enable_box = snbep_uncore_msr_enable_box, \ + .disable_event = snbep_uncore_msr_disable_event, \ + .enable_event = snbep_uncore_msr_enable_event, \ + .read_counter = uncore_msr_read_counter + +static struct intel_uncore_ops ivt_uncore_msr_ops = { + IVT_UNCORE_MSR_OPS_COMMON_INIT(), +}; + +static struct intel_uncore_ops ivt_uncore_pci_ops = { + .init_box = ivt_uncore_pci_init_box, + .disable_box = snbep_uncore_pci_disable_box, + .enable_box = snbep_uncore_pci_enable_box, + .disable_event = snbep_uncore_pci_disable_event, + .enable_event = snbep_uncore_pci_enable_event, + .read_counter = snbep_uncore_pci_read_counter, +}; + +#define IVT_UNCORE_PCI_COMMON_INIT() \ + .perf_ctr = SNBEP_PCI_PMON_CTR0, \ + .event_ctl = SNBEP_PCI_PMON_CTL0, \ + .event_mask = IVT_PMON_RAW_EVENT_MASK, \ + .box_ctl = SNBEP_PCI_PMON_BOX_CTL, \ + .ops = &ivt_uncore_pci_ops, \ + .format_group = &ivt_uncore_format_group + +static struct attribute *ivt_uncore_formats_attr[] = { + &format_attr_event.attr, + &format_attr_umask.attr, + &format_attr_edge.attr, + &format_attr_inv.attr, + &format_attr_thresh8.attr, + NULL, +}; + +static struct attribute *ivt_uncore_ubox_formats_attr[] = { + &format_attr_event.attr, + &format_attr_umask.attr, + &format_attr_edge.attr, + &format_attr_inv.attr, + &format_attr_thresh5.attr, + NULL, +}; + +static struct attribute *ivt_uncore_cbox_formats_attr[] = { + &format_attr_event.attr, + &format_attr_umask.attr, + &format_attr_edge.attr, + &format_attr_tid_en.attr, + &format_attr_thresh8.attr, + &format_attr_filter_tid.attr, + &format_attr_filter_link.attr, + &format_attr_filter_state2.attr, + &format_attr_filter_nid2.attr, + &format_attr_filter_opc2.attr, + NULL, +}; + +static struct attribute *ivt_uncore_pcu_formats_attr[] = { + &format_attr_event_ext.attr, + &format_attr_occ_sel.attr, + &format_attr_edge.attr, + &format_attr_thresh5.attr, + &format_attr_occ_invert.attr, + &format_attr_occ_edge.attr, + &format_attr_filter_band0.attr, + &format_attr_filter_band1.attr, + &format_attr_filter_band2.attr, + &format_attr_filter_band3.attr, + NULL, +}; + +static struct attribute *ivt_uncore_qpi_formats_attr[] = { + &format_attr_event_ext.attr, + &format_attr_umask.attr, + &format_attr_edge.attr, + &format_attr_thresh8.attr, + &format_attr_match_rds.attr, + &format_attr_match_rnid30.attr, + &format_attr_match_rnid4.attr, + &format_attr_match_dnid.attr, + &format_attr_match_mc.attr, + &format_attr_match_opc.attr, + &format_attr_match_vnw.attr, + &format_attr_match0.attr, + &format_attr_match1.attr, + &format_attr_mask_rds.attr, + &format_attr_mask_rnid30.attr, + &format_attr_mask_rnid4.attr, + &format_attr_mask_dnid.attr, + &format_attr_mask_mc.attr, + &format_attr_mask_opc.attr, + &format_attr_mask_vnw.attr, + &format_attr_mask0.attr, + &format_attr_mask1.attr, + NULL, +}; + +static struct attribute_group ivt_uncore_format_group = { + .name = "format", + .attrs = ivt_uncore_formats_attr, +}; + +static struct attribute_group ivt_uncore_ubox_format_group = { + .name = "format", + .attrs = ivt_uncore_ubox_formats_attr, +}; + +static struct attribute_group ivt_uncore_cbox_format_group = { + .name = "format", + .attrs = ivt_uncore_cbox_formats_attr, +}; + +static struct attribute_group ivt_uncore_pcu_format_group = { + .name = "format", + .attrs = ivt_uncore_pcu_formats_attr, +}; + +static struct attribute_group ivt_uncore_qpi_format_group = { + .name = "format", + .attrs = ivt_uncore_qpi_formats_attr, +}; + +static struct intel_uncore_type ivt_uncore_ubox = { + .name = "ubox", + .num_counters = 2, + .num_boxes = 1, + .perf_ctr_bits = 44, + .fixed_ctr_bits = 48, + .perf_ctr = SNBEP_U_MSR_PMON_CTR0, + .event_ctl = SNBEP_U_MSR_PMON_CTL0, + .event_mask = IVT_U_MSR_PMON_RAW_EVENT_MASK, + .fixed_ctr = SNBEP_U_MSR_PMON_UCLK_FIXED_CTR, + .fixed_ctl = SNBEP_U_MSR_PMON_UCLK_FIXED_CTL, + .ops = &ivt_uncore_msr_ops, + .format_group = &ivt_uncore_ubox_format_group, +}; + +static struct extra_reg ivt_uncore_cbox_extra_regs[] = { + SNBEP_CBO_EVENT_EXTRA_REG(SNBEP_CBO_PMON_CTL_TID_EN, + SNBEP_CBO_PMON_CTL_TID_EN, 0x1), + SNBEP_CBO_EVENT_EXTRA_REG(0x1031, 0x10ff, 0x2), + + SNBEP_CBO_EVENT_EXTRA_REG(0x1134, 0xffff, 0x4), + SNBEP_CBO_EVENT_EXTRA_REG(0x4134, 0xffff, 0xc), + SNBEP_CBO_EVENT_EXTRA_REG(0x5134, 0xffff, 0xc), + SNBEP_CBO_EVENT_EXTRA_REG(0x0334, 0xffff, 0x4), + SNBEP_CBO_EVENT_EXTRA_REG(0x4334, 0xffff, 0xc), + SNBEP_CBO_EVENT_EXTRA_REG(0x0534, 0xffff, 0x4), + SNBEP_CBO_EVENT_EXTRA_REG(0x4534, 0xffff, 0xc), + SNBEP_CBO_EVENT_EXTRA_REG(0x0934, 0xffff, 0x4), + SNBEP_CBO_EVENT_EXTRA_REG(0x4934, 0xffff, 0xc), + SNBEP_CBO_EVENT_EXTRA_REG(0x0135, 0xffff, 0x10), + SNBEP_CBO_EVENT_EXTRA_REG(0x0335, 0xffff, 0x10), + SNBEP_CBO_EVENT_EXTRA_REG(0x2135, 0xffff, 0x10), + SNBEP_CBO_EVENT_EXTRA_REG(0x2335, 0xffff, 0x10), + SNBEP_CBO_EVENT_EXTRA_REG(0x4135, 0xffff, 0x18), + SNBEP_CBO_EVENT_EXTRA_REG(0x4335, 0xffff, 0x18), + SNBEP_CBO_EVENT_EXTRA_REG(0x4435, 0xffff, 0x8), + SNBEP_CBO_EVENT_EXTRA_REG(0x4835, 0xffff, 0x8), + SNBEP_CBO_EVENT_EXTRA_REG(0x4a35, 0xffff, 0x8), + SNBEP_CBO_EVENT_EXTRA_REG(0x5035, 0xffff, 0x8), + SNBEP_CBO_EVENT_EXTRA_REG(0x8135, 0xffff, 0x10), + SNBEP_CBO_EVENT_EXTRA_REG(0x8335, 0xffff, 0x10), + SNBEP_CBO_EVENT_EXTRA_REG(0x0136, 0xffff, 0x10), + SNBEP_CBO_EVENT_EXTRA_REG(0x0336, 0xffff, 0x10), + SNBEP_CBO_EVENT_EXTRA_REG(0x2136, 0xffff, 0x10), + SNBEP_CBO_EVENT_EXTRA_REG(0x2336, 0xffff, 0x10), + SNBEP_CBO_EVENT_EXTRA_REG(0x4136, 0xffff, 0x18), + SNBEP_CBO_EVENT_EXTRA_REG(0x4336, 0xffff, 0x18), + SNBEP_CBO_EVENT_EXTRA_REG(0x4436, 0xffff, 0x8), + SNBEP_CBO_EVENT_EXTRA_REG(0x4836, 0xffff, 0x8), + SNBEP_CBO_EVENT_EXTRA_REG(0x4a36, 0xffff, 0x8), + SNBEP_CBO_EVENT_EXTRA_REG(0x5036, 0xffff, 0x8), + SNBEP_CBO_EVENT_EXTRA_REG(0x8136, 0xffff, 0x10), + SNBEP_CBO_EVENT_EXTRA_REG(0x8336, 0xffff, 0x10), + SNBEP_CBO_EVENT_EXTRA_REG(0x4037, 0x40ff, 0x8), + EVENT_EXTRA_END +}; + +static u64 ivt_cbox_filter_mask(int fields) +{ + u64 mask = 0; + + if (fields & 0x1) + mask |= IVT_CB0_MSR_PMON_BOX_FILTER_TID; + if (fields & 0x2) + mask |= IVT_CB0_MSR_PMON_BOX_FILTER_LINK; + if (fields & 0x4) + mask |= IVT_CB0_MSR_PMON_BOX_FILTER_STATE; + if (fields & 0x8) + mask |= IVT_CB0_MSR_PMON_BOX_FILTER_NID; + if (fields & 0x10) + mask |= IVT_CB0_MSR_PMON_BOX_FILTER_OPC; + + return mask; +} + +static struct event_constraint * +ivt_cbox_get_constraint(struct intel_uncore_box *box, struct perf_event *event) +{ + return __snbep_cbox_get_constraint(box, event, ivt_cbox_filter_mask); +} + +static int ivt_cbox_hw_config(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event_extra *reg1 = &event->hw.extra_reg; + struct extra_reg *er; + int idx = 0; + + for (er = ivt_uncore_cbox_extra_regs; er->msr; er++) { + if (er->event != (event->hw.config & er->config_mask)) + continue; + idx |= er->idx; + } + + if (idx) { + reg1->reg = SNBEP_C0_MSR_PMON_BOX_FILTER + + SNBEP_CBO_MSR_OFFSET * box->pmu->pmu_idx; + reg1->config = event->attr.config1 & ivt_cbox_filter_mask(idx); + reg1->idx = idx; + } + return 0; +} + +static void ivt_cbox_enable_event(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &hwc->extra_reg; + + if (reg1->idx != EXTRA_REG_NONE) { + u64 filter = uncore_shared_reg_config(box, 0); + wrmsrl(reg1->reg, filter & 0xffffffff); + wrmsrl(reg1->reg + 6, filter >> 32); + } + + wrmsrl(hwc->config_base, hwc->config | SNBEP_PMON_CTL_EN); +} + +static struct intel_uncore_ops ivt_uncore_cbox_ops = { + .init_box = ivt_uncore_msr_init_box, + .disable_box = snbep_uncore_msr_disable_box, + .enable_box = snbep_uncore_msr_enable_box, + .disable_event = snbep_uncore_msr_disable_event, + .enable_event = ivt_cbox_enable_event, + .read_counter = uncore_msr_read_counter, + .hw_config = ivt_cbox_hw_config, + .get_constraint = ivt_cbox_get_constraint, + .put_constraint = snbep_cbox_put_constraint, +}; + +static struct intel_uncore_type ivt_uncore_cbox = { + .name = "cbox", + .num_counters = 4, + .num_boxes = 15, + .perf_ctr_bits = 44, + .event_ctl = SNBEP_C0_MSR_PMON_CTL0, + .perf_ctr = SNBEP_C0_MSR_PMON_CTR0, + .event_mask = IVT_CBO_MSR_PMON_RAW_EVENT_MASK, + .box_ctl = SNBEP_C0_MSR_PMON_BOX_CTL, + .msr_offset = SNBEP_CBO_MSR_OFFSET, + .num_shared_regs = 1, + .constraints = snbep_uncore_cbox_constraints, + .ops = &ivt_uncore_cbox_ops, + .format_group = &ivt_uncore_cbox_format_group, +}; + +static struct intel_uncore_ops ivt_uncore_pcu_ops = { + IVT_UNCORE_MSR_OPS_COMMON_INIT(), + .hw_config = snbep_pcu_hw_config, + .get_constraint = snbep_pcu_get_constraint, + .put_constraint = snbep_pcu_put_constraint, +}; + +static struct intel_uncore_type ivt_uncore_pcu = { + .name = "pcu", + .num_counters = 4, + .num_boxes = 1, + .perf_ctr_bits = 48, + .perf_ctr = SNBEP_PCU_MSR_PMON_CTR0, + .event_ctl = SNBEP_PCU_MSR_PMON_CTL0, + .event_mask = IVT_PCU_MSR_PMON_RAW_EVENT_MASK, + .box_ctl = SNBEP_PCU_MSR_PMON_BOX_CTL, + .num_shared_regs = 1, + .ops = &ivt_uncore_pcu_ops, + .format_group = &ivt_uncore_pcu_format_group, +}; + +static struct intel_uncore_type *ivt_msr_uncores[] = { + &ivt_uncore_ubox, + &ivt_uncore_cbox, + &ivt_uncore_pcu, + NULL, +}; + +static struct intel_uncore_type ivt_uncore_ha = { + .name = "ha", + .num_counters = 4, + .num_boxes = 2, + .perf_ctr_bits = 48, + IVT_UNCORE_PCI_COMMON_INIT(), +}; + +static struct intel_uncore_type ivt_uncore_imc = { + .name = "imc", + .num_counters = 4, + .num_boxes = 8, + .perf_ctr_bits = 48, + .fixed_ctr_bits = 48, + .fixed_ctr = SNBEP_MC_CHy_PCI_PMON_FIXED_CTR, + .fixed_ctl = SNBEP_MC_CHy_PCI_PMON_FIXED_CTL, + IVT_UNCORE_PCI_COMMON_INIT(), +}; + +/* registers in IRP boxes are not properly aligned */ +static unsigned ivt_uncore_irp_ctls[] = {0xd8, 0xdc, 0xe0, 0xe4}; +static unsigned ivt_uncore_irp_ctrs[] = {0xa0, 0xb0, 0xb8, 0xc0}; + +static void ivt_uncore_irp_enable_event(struct intel_uncore_box *box, struct perf_event *event) +{ + struct pci_dev *pdev = box->pci_dev; + struct hw_perf_event *hwc = &event->hw; + + pci_write_config_dword(pdev, ivt_uncore_irp_ctls[hwc->idx], + hwc->config | SNBEP_PMON_CTL_EN); +} + +static void ivt_uncore_irp_disable_event(struct intel_uncore_box *box, struct perf_event *event) +{ + struct pci_dev *pdev = box->pci_dev; + struct hw_perf_event *hwc = &event->hw; + + pci_write_config_dword(pdev, ivt_uncore_irp_ctls[hwc->idx], hwc->config); +} + +static u64 ivt_uncore_irp_read_counter(struct intel_uncore_box *box, struct perf_event *event) +{ + struct pci_dev *pdev = box->pci_dev; + struct hw_perf_event *hwc = &event->hw; + u64 count = 0; + + pci_read_config_dword(pdev, ivt_uncore_irp_ctrs[hwc->idx], (u32 *)&count); + pci_read_config_dword(pdev, ivt_uncore_irp_ctrs[hwc->idx] + 4, (u32 *)&count + 1); + + return count; +} + +static struct intel_uncore_ops ivt_uncore_irp_ops = { + .init_box = ivt_uncore_pci_init_box, + .disable_box = snbep_uncore_pci_disable_box, + .enable_box = snbep_uncore_pci_enable_box, + .disable_event = ivt_uncore_irp_disable_event, + .enable_event = ivt_uncore_irp_enable_event, + .read_counter = ivt_uncore_irp_read_counter, +}; + +static struct intel_uncore_type ivt_uncore_irp = { + .name = "irp", + .num_counters = 4, + .num_boxes = 1, + .perf_ctr_bits = 48, + .event_mask = IVT_PMON_RAW_EVENT_MASK, + .box_ctl = SNBEP_PCI_PMON_BOX_CTL, + .ops = &ivt_uncore_irp_ops, + .format_group = &ivt_uncore_format_group, +}; + +static struct intel_uncore_ops ivt_uncore_qpi_ops = { + .init_box = ivt_uncore_pci_init_box, + .disable_box = snbep_uncore_pci_disable_box, + .enable_box = snbep_uncore_pci_enable_box, + .disable_event = snbep_uncore_pci_disable_event, + .enable_event = snbep_qpi_enable_event, + .read_counter = snbep_uncore_pci_read_counter, + .hw_config = snbep_qpi_hw_config, + .get_constraint = uncore_get_constraint, + .put_constraint = uncore_put_constraint, +}; + +static struct intel_uncore_type ivt_uncore_qpi = { + .name = "qpi", + .num_counters = 4, + .num_boxes = 3, + .perf_ctr_bits = 48, + .perf_ctr = SNBEP_PCI_PMON_CTR0, + .event_ctl = SNBEP_PCI_PMON_CTL0, + .event_mask = IVT_QPI_PCI_PMON_RAW_EVENT_MASK, + .box_ctl = SNBEP_PCI_PMON_BOX_CTL, + .num_shared_regs = 1, + .ops = &ivt_uncore_qpi_ops, + .format_group = &ivt_uncore_qpi_format_group, +}; + +static struct intel_uncore_type ivt_uncore_r2pcie = { + .name = "r2pcie", + .num_counters = 4, + .num_boxes = 1, + .perf_ctr_bits = 44, + .constraints = snbep_uncore_r2pcie_constraints, + IVT_UNCORE_PCI_COMMON_INIT(), +}; + +static struct intel_uncore_type ivt_uncore_r3qpi = { + .name = "r3qpi", + .num_counters = 3, + .num_boxes = 2, + .perf_ctr_bits = 44, + .constraints = snbep_uncore_r3qpi_constraints, + IVT_UNCORE_PCI_COMMON_INIT(), +}; + +enum { + IVT_PCI_UNCORE_HA, + IVT_PCI_UNCORE_IMC, + IVT_PCI_UNCORE_IRP, + IVT_PCI_UNCORE_QPI, + IVT_PCI_UNCORE_R2PCIE, + IVT_PCI_UNCORE_R3QPI, +}; + +static struct intel_uncore_type *ivt_pci_uncores[] = { + [IVT_PCI_UNCORE_HA] = &ivt_uncore_ha, + [IVT_PCI_UNCORE_IMC] = &ivt_uncore_imc, + [IVT_PCI_UNCORE_IRP] = &ivt_uncore_irp, + [IVT_PCI_UNCORE_QPI] = &ivt_uncore_qpi, + [IVT_PCI_UNCORE_R2PCIE] = &ivt_uncore_r2pcie, + [IVT_PCI_UNCORE_R3QPI] = &ivt_uncore_r3qpi, + NULL, +}; + +static DEFINE_PCI_DEVICE_TABLE(ivt_uncore_pci_ids) = { + { /* Home Agent 0 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xe30), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_HA, 0), + }, + { /* Home Agent 1 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xe38), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_HA, 1), + }, + { /* MC0 Channel 0 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xeb4), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_IMC, 0), + }, + { /* MC0 Channel 1 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xeb5), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_IMC, 1), + }, + { /* MC0 Channel 3 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xeb0), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_IMC, 2), + }, + { /* MC0 Channel 4 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xeb1), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_IMC, 3), + }, + { /* MC1 Channel 0 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xef4), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_IMC, 4), + }, + { /* MC1 Channel 1 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xef5), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_IMC, 5), + }, + { /* MC1 Channel 3 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xef0), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_IMC, 6), + }, + { /* MC1 Channel 4 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xef1), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_IMC, 7), + }, + { /* IRP */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xe39), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_IRP, 0), + }, + { /* QPI0 Port 0 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xe32), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_QPI, 0), + }, + { /* QPI0 Port 1 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xe33), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_QPI, 1), + }, + { /* QPI1 Port 2 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xe3a), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_QPI, 2), + }, + { /* R2PCIe */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xe34), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_R2PCIE, 0), + }, + { /* R3QPI0 Link 0 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xe36), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_R3QPI, 0), + }, + { /* R3QPI0 Link 1 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xe37), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_R3QPI, 1), + }, + { /* R3QPI1 Link 2 */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xe3e), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_R3QPI, 2), + }, + { /* QPI Port 0 filter */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xe86), + .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, + SNBEP_PCI_QPI_PORT0_FILTER), + }, + { /* QPI Port 0 filter */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xe96), + .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, + SNBEP_PCI_QPI_PORT1_FILTER), + }, + { /* end: all zeroes */ } +}; + +static struct pci_driver ivt_uncore_pci_driver = { + .name = "ivt_uncore", + .id_table = ivt_uncore_pci_ids, +}; +/* end of IvyTown uncore support */ + +/* Sandy Bridge uncore support */ +static void snb_uncore_msr_enable_event(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + + if (hwc->idx < UNCORE_PMC_IDX_FIXED) + wrmsrl(hwc->config_base, hwc->config | SNB_UNC_CTL_EN); + else + wrmsrl(hwc->config_base, SNB_UNC_CTL_EN); +} + +static void snb_uncore_msr_disable_event(struct intel_uncore_box *box, struct perf_event *event) +{ + wrmsrl(event->hw.config_base, 0); +} + +static void snb_uncore_msr_init_box(struct intel_uncore_box *box) +{ + if (box->pmu->pmu_idx == 0) { + wrmsrl(SNB_UNC_PERF_GLOBAL_CTL, + SNB_UNC_GLOBAL_CTL_EN | SNB_UNC_GLOBAL_CTL_CORE_ALL); + } +} + +static struct uncore_event_desc snb_uncore_events[] = { + INTEL_UNCORE_EVENT_DESC(clockticks, "event=0xff,umask=0x00"), + { /* end: all zeroes */ }, +}; + +static struct attribute *snb_uncore_formats_attr[] = { + &format_attr_event.attr, + &format_attr_umask.attr, + &format_attr_edge.attr, + &format_attr_inv.attr, + &format_attr_cmask5.attr, + NULL, +}; + +static struct attribute_group snb_uncore_format_group = { + .name = "format", + .attrs = snb_uncore_formats_attr, +}; + +static struct intel_uncore_ops snb_uncore_msr_ops = { + .init_box = snb_uncore_msr_init_box, + .disable_event = snb_uncore_msr_disable_event, + .enable_event = snb_uncore_msr_enable_event, + .read_counter = uncore_msr_read_counter, +}; + +static struct event_constraint snb_uncore_cbox_constraints[] = { + UNCORE_EVENT_CONSTRAINT(0x80, 0x1), + UNCORE_EVENT_CONSTRAINT(0x83, 0x1), + EVENT_CONSTRAINT_END +}; + +static struct intel_uncore_type snb_uncore_cbox = { + .name = "cbox", + .num_counters = 2, + .num_boxes = 4, + .perf_ctr_bits = 44, + .fixed_ctr_bits = 48, + .perf_ctr = SNB_UNC_CBO_0_PER_CTR0, + .event_ctl = SNB_UNC_CBO_0_PERFEVTSEL0, + .fixed_ctr = SNB_UNC_FIXED_CTR, + .fixed_ctl = SNB_UNC_FIXED_CTR_CTRL, + .single_fixed = 1, + .event_mask = SNB_UNC_RAW_EVENT_MASK, + .msr_offset = SNB_UNC_CBO_MSR_OFFSET, + .constraints = snb_uncore_cbox_constraints, + .ops = &snb_uncore_msr_ops, + .format_group = &snb_uncore_format_group, + .event_descs = snb_uncore_events, +}; + +static struct intel_uncore_type *snb_msr_uncores[] = { + &snb_uncore_cbox, + NULL, +}; + +enum { + SNB_PCI_UNCORE_IMC, +}; + +static struct uncore_event_desc snb_uncore_imc_events[] = { + INTEL_UNCORE_EVENT_DESC(data_reads, "event=0x01"), + INTEL_UNCORE_EVENT_DESC(data_reads.scale, "6.103515625e-5"), + INTEL_UNCORE_EVENT_DESC(data_reads.unit, "MiB"), + + INTEL_UNCORE_EVENT_DESC(data_writes, "event=0x02"), + INTEL_UNCORE_EVENT_DESC(data_writes.scale, "6.103515625e-5"), + INTEL_UNCORE_EVENT_DESC(data_writes.unit, "MiB"), + + { /* end: all zeroes */ }, +}; + +#define SNB_UNCORE_PCI_IMC_EVENT_MASK 0xff +#define SNB_UNCORE_PCI_IMC_BAR_OFFSET 0x48 + +/* page size multiple covering all config regs */ +#define SNB_UNCORE_PCI_IMC_MAP_SIZE 0x6000 + +#define SNB_UNCORE_PCI_IMC_DATA_READS 0x1 +#define SNB_UNCORE_PCI_IMC_DATA_READS_BASE 0x5050 +#define SNB_UNCORE_PCI_IMC_DATA_WRITES 0x2 +#define SNB_UNCORE_PCI_IMC_DATA_WRITES_BASE 0x5054 +#define SNB_UNCORE_PCI_IMC_CTR_BASE SNB_UNCORE_PCI_IMC_DATA_READS_BASE + +static struct attribute *snb_uncore_imc_formats_attr[] = { + &format_attr_event.attr, + NULL, +}; + +static struct attribute_group snb_uncore_imc_format_group = { + .name = "format", + .attrs = snb_uncore_imc_formats_attr, +}; + +static void snb_uncore_imc_init_box(struct intel_uncore_box *box) +{ + struct pci_dev *pdev = box->pci_dev; + int where = SNB_UNCORE_PCI_IMC_BAR_OFFSET; + resource_size_t addr; + u32 pci_dword; + + pci_read_config_dword(pdev, where, &pci_dword); + addr = pci_dword; + +#ifdef CONFIG_PHYS_ADDR_T_64BIT + pci_read_config_dword(pdev, where + 4, &pci_dword); + addr |= ((resource_size_t)pci_dword << 32); +#endif + + addr &= ~(PAGE_SIZE - 1); + + box->io_addr = ioremap(addr, SNB_UNCORE_PCI_IMC_MAP_SIZE); + box->hrtimer_duration = UNCORE_SNB_IMC_HRTIMER_INTERVAL; +} + +static void snb_uncore_imc_enable_box(struct intel_uncore_box *box) +{} + +static void snb_uncore_imc_disable_box(struct intel_uncore_box *box) +{} + +static void snb_uncore_imc_enable_event(struct intel_uncore_box *box, struct perf_event *event) +{} + +static void snb_uncore_imc_disable_event(struct intel_uncore_box *box, struct perf_event *event) +{} + +static u64 snb_uncore_imc_read_counter(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + + return (u64)*(unsigned int *)(box->io_addr + hwc->event_base); +} + +/* + * custom event_init() function because we define our own fixed, free + * running counters, so we do not want to conflict with generic uncore + * logic. Also simplifies processing + */ +static int snb_uncore_imc_event_init(struct perf_event *event) +{ + struct intel_uncore_pmu *pmu; + struct intel_uncore_box *box; + struct hw_perf_event *hwc = &event->hw; + u64 cfg = event->attr.config & SNB_UNCORE_PCI_IMC_EVENT_MASK; + int idx, base; + + if (event->attr.type != event->pmu->type) + return -ENOENT; + + pmu = uncore_event_to_pmu(event); + /* no device found for this pmu */ + if (pmu->func_id < 0) + return -ENOENT; + + /* Sampling not supported yet */ + if (hwc->sample_period) + return -EINVAL; + + /* unsupported modes and filters */ + if (event->attr.exclude_user || + event->attr.exclude_kernel || + event->attr.exclude_hv || + event->attr.exclude_idle || + event->attr.exclude_host || + event->attr.exclude_guest || + event->attr.sample_period) /* no sampling */ + return -EINVAL; + + /* + * Place all uncore events for a particular physical package + * onto a single cpu + */ + if (event->cpu < 0) + return -EINVAL; + + /* check only supported bits are set */ + if (event->attr.config & ~SNB_UNCORE_PCI_IMC_EVENT_MASK) + return -EINVAL; + + box = uncore_pmu_to_box(pmu, event->cpu); + if (!box || box->cpu < 0) + return -EINVAL; + + event->cpu = box->cpu; + + event->hw.idx = -1; + event->hw.last_tag = ~0ULL; + event->hw.extra_reg.idx = EXTRA_REG_NONE; + event->hw.branch_reg.idx = EXTRA_REG_NONE; + /* + * check event is known (whitelist, determines counter) + */ + switch (cfg) { + case SNB_UNCORE_PCI_IMC_DATA_READS: + base = SNB_UNCORE_PCI_IMC_DATA_READS_BASE; + idx = UNCORE_PMC_IDX_FIXED; + break; + case SNB_UNCORE_PCI_IMC_DATA_WRITES: + base = SNB_UNCORE_PCI_IMC_DATA_WRITES_BASE; + idx = UNCORE_PMC_IDX_FIXED + 1; + break; + default: + return -EINVAL; + } + + /* must be done before validate_group */ + event->hw.event_base = base; + event->hw.config = cfg; + event->hw.idx = idx; + + /* no group validation needed, we have free running counters */ + + return 0; +} + +static int snb_uncore_imc_hw_config(struct intel_uncore_box *box, struct perf_event *event) +{ + return 0; +} + +static void snb_uncore_imc_event_start(struct perf_event *event, int flags) +{ + struct intel_uncore_box *box = uncore_event_to_box(event); + u64 count; + + if (WARN_ON_ONCE(!(event->hw.state & PERF_HES_STOPPED))) + return; + + event->hw.state = 0; + box->n_active++; + + list_add_tail(&event->active_entry, &box->active_list); + + count = snb_uncore_imc_read_counter(box, event); + local64_set(&event->hw.prev_count, count); + + if (box->n_active == 1) + uncore_pmu_start_hrtimer(box); +} + +static void snb_uncore_imc_event_stop(struct perf_event *event, int flags) +{ + struct intel_uncore_box *box = uncore_event_to_box(event); + struct hw_perf_event *hwc = &event->hw; + + if (!(hwc->state & PERF_HES_STOPPED)) { + box->n_active--; + + WARN_ON_ONCE(hwc->state & PERF_HES_STOPPED); + hwc->state |= PERF_HES_STOPPED; + + list_del(&event->active_entry); + + if (box->n_active == 0) + uncore_pmu_cancel_hrtimer(box); + } + + if ((flags & PERF_EF_UPDATE) && !(hwc->state & PERF_HES_UPTODATE)) { + /* + * Drain the remaining delta count out of a event + * that we are disabling: + */ + uncore_perf_event_update(box, event); + hwc->state |= PERF_HES_UPTODATE; + } +} + +static int snb_uncore_imc_event_add(struct perf_event *event, int flags) +{ + struct intel_uncore_box *box = uncore_event_to_box(event); + struct hw_perf_event *hwc = &event->hw; + + if (!box) + return -ENODEV; + + hwc->state = PERF_HES_UPTODATE | PERF_HES_STOPPED; + if (!(flags & PERF_EF_START)) + hwc->state |= PERF_HES_ARCH; + + snb_uncore_imc_event_start(event, 0); + + box->n_events++; + + return 0; +} + +static void snb_uncore_imc_event_del(struct perf_event *event, int flags) +{ + struct intel_uncore_box *box = uncore_event_to_box(event); + int i; + + snb_uncore_imc_event_stop(event, PERF_EF_UPDATE); + + for (i = 0; i < box->n_events; i++) { + if (event == box->event_list[i]) { + --box->n_events; + break; + } + } +} + +static int snb_pci2phy_map_init(int devid) +{ + struct pci_dev *dev = NULL; + int bus; + + dev = pci_get_device(PCI_VENDOR_ID_INTEL, devid, dev); + if (!dev) + return -ENOTTY; + + bus = dev->bus->number; + + pcibus_to_physid[bus] = 0; + + pci_dev_put(dev); + + return 0; +} + +static struct pmu snb_uncore_imc_pmu = { + .task_ctx_nr = perf_invalid_context, + .event_init = snb_uncore_imc_event_init, + .add = snb_uncore_imc_event_add, + .del = snb_uncore_imc_event_del, + .start = snb_uncore_imc_event_start, + .stop = snb_uncore_imc_event_stop, + .read = uncore_pmu_event_read, +}; + +static struct intel_uncore_ops snb_uncore_imc_ops = { + .init_box = snb_uncore_imc_init_box, + .enable_box = snb_uncore_imc_enable_box, + .disable_box = snb_uncore_imc_disable_box, + .disable_event = snb_uncore_imc_disable_event, + .enable_event = snb_uncore_imc_enable_event, + .hw_config = snb_uncore_imc_hw_config, + .read_counter = snb_uncore_imc_read_counter, +}; + +static struct intel_uncore_type snb_uncore_imc = { + .name = "imc", + .num_counters = 2, + .num_boxes = 1, + .fixed_ctr_bits = 32, + .fixed_ctr = SNB_UNCORE_PCI_IMC_CTR_BASE, + .event_descs = snb_uncore_imc_events, + .format_group = &snb_uncore_imc_format_group, + .perf_ctr = SNB_UNCORE_PCI_IMC_DATA_READS_BASE, + .event_mask = SNB_UNCORE_PCI_IMC_EVENT_MASK, + .ops = &snb_uncore_imc_ops, + .pmu = &snb_uncore_imc_pmu, +}; + +static struct intel_uncore_type *snb_pci_uncores[] = { + [SNB_PCI_UNCORE_IMC] = &snb_uncore_imc, + NULL, +}; + +static DEFINE_PCI_DEVICE_TABLE(snb_uncore_pci_ids) = { + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_SNB_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* end: all zeroes */ }, +}; + +static DEFINE_PCI_DEVICE_TABLE(ivb_uncore_pci_ids) = { + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_IVB_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* end: all zeroes */ }, +}; + +static DEFINE_PCI_DEVICE_TABLE(hsw_uncore_pci_ids) = { + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_HSW_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* end: all zeroes */ }, +}; + +static struct pci_driver snb_uncore_pci_driver = { + .name = "snb_uncore", + .id_table = snb_uncore_pci_ids, +}; + +static struct pci_driver ivb_uncore_pci_driver = { + .name = "ivb_uncore", + .id_table = ivb_uncore_pci_ids, +}; + +static struct pci_driver hsw_uncore_pci_driver = { + .name = "hsw_uncore", + .id_table = hsw_uncore_pci_ids, +}; + +/* end of Sandy Bridge uncore support */ + +/* Nehalem uncore support */ +static void nhm_uncore_msr_disable_box(struct intel_uncore_box *box) +{ + wrmsrl(NHM_UNC_PERF_GLOBAL_CTL, 0); +} + +static void nhm_uncore_msr_enable_box(struct intel_uncore_box *box) +{ + wrmsrl(NHM_UNC_PERF_GLOBAL_CTL, NHM_UNC_GLOBAL_CTL_EN_PC_ALL | NHM_UNC_GLOBAL_CTL_EN_FC); +} + +static void nhm_uncore_msr_enable_event(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + + if (hwc->idx < UNCORE_PMC_IDX_FIXED) + wrmsrl(hwc->config_base, hwc->config | SNB_UNC_CTL_EN); + else + wrmsrl(hwc->config_base, NHM_UNC_FIXED_CTR_CTL_EN); +} + +static struct attribute *nhm_uncore_formats_attr[] = { + &format_attr_event.attr, + &format_attr_umask.attr, + &format_attr_edge.attr, + &format_attr_inv.attr, + &format_attr_cmask8.attr, + NULL, +}; + +static struct attribute_group nhm_uncore_format_group = { + .name = "format", + .attrs = nhm_uncore_formats_attr, +}; + +static struct uncore_event_desc nhm_uncore_events[] = { + INTEL_UNCORE_EVENT_DESC(clockticks, "event=0xff,umask=0x00"), + INTEL_UNCORE_EVENT_DESC(qmc_writes_full_any, "event=0x2f,umask=0x0f"), + INTEL_UNCORE_EVENT_DESC(qmc_normal_reads_any, "event=0x2c,umask=0x0f"), + INTEL_UNCORE_EVENT_DESC(qhl_request_ioh_reads, "event=0x20,umask=0x01"), + INTEL_UNCORE_EVENT_DESC(qhl_request_ioh_writes, "event=0x20,umask=0x02"), + INTEL_UNCORE_EVENT_DESC(qhl_request_remote_reads, "event=0x20,umask=0x04"), + INTEL_UNCORE_EVENT_DESC(qhl_request_remote_writes, "event=0x20,umask=0x08"), + INTEL_UNCORE_EVENT_DESC(qhl_request_local_reads, "event=0x20,umask=0x10"), + INTEL_UNCORE_EVENT_DESC(qhl_request_local_writes, "event=0x20,umask=0x20"), + { /* end: all zeroes */ }, +}; + +static struct intel_uncore_ops nhm_uncore_msr_ops = { + .disable_box = nhm_uncore_msr_disable_box, + .enable_box = nhm_uncore_msr_enable_box, + .disable_event = snb_uncore_msr_disable_event, + .enable_event = nhm_uncore_msr_enable_event, + .read_counter = uncore_msr_read_counter, +}; + +static struct intel_uncore_type nhm_uncore = { + .name = "", + .num_counters = 8, + .num_boxes = 1, + .perf_ctr_bits = 48, + .fixed_ctr_bits = 48, + .event_ctl = NHM_UNC_PERFEVTSEL0, + .perf_ctr = NHM_UNC_UNCORE_PMC0, + .fixed_ctr = NHM_UNC_FIXED_CTR, + .fixed_ctl = NHM_UNC_FIXED_CTR_CTRL, + .event_mask = NHM_UNC_RAW_EVENT_MASK, + .event_descs = nhm_uncore_events, + .ops = &nhm_uncore_msr_ops, + .format_group = &nhm_uncore_format_group, +}; + +static struct intel_uncore_type *nhm_msr_uncores[] = { + &nhm_uncore, + NULL, +}; +/* end of Nehalem uncore support */ + +/* Nehalem-EX uncore support */ +DEFINE_UNCORE_FORMAT_ATTR(event5, event, "config:1-5"); +DEFINE_UNCORE_FORMAT_ATTR(counter, counter, "config:6-7"); +DEFINE_UNCORE_FORMAT_ATTR(match, match, "config1:0-63"); +DEFINE_UNCORE_FORMAT_ATTR(mask, mask, "config2:0-63"); + +static void nhmex_uncore_msr_init_box(struct intel_uncore_box *box) +{ + wrmsrl(NHMEX_U_MSR_PMON_GLOBAL_CTL, NHMEX_U_PMON_GLOBAL_EN_ALL); +} + +static void nhmex_uncore_msr_disable_box(struct intel_uncore_box *box) +{ + unsigned msr = uncore_msr_box_ctl(box); + u64 config; + + if (msr) { + rdmsrl(msr, config); + config &= ~((1ULL << uncore_num_counters(box)) - 1); + /* WBox has a fixed counter */ + if (uncore_msr_fixed_ctl(box)) + config &= ~NHMEX_W_PMON_GLOBAL_FIXED_EN; + wrmsrl(msr, config); + } +} + +static void nhmex_uncore_msr_enable_box(struct intel_uncore_box *box) +{ + unsigned msr = uncore_msr_box_ctl(box); + u64 config; + + if (msr) { + rdmsrl(msr, config); + config |= (1ULL << uncore_num_counters(box)) - 1; + /* WBox has a fixed counter */ + if (uncore_msr_fixed_ctl(box)) + config |= NHMEX_W_PMON_GLOBAL_FIXED_EN; + wrmsrl(msr, config); + } +} + +static void nhmex_uncore_msr_disable_event(struct intel_uncore_box *box, struct perf_event *event) +{ + wrmsrl(event->hw.config_base, 0); +} + +static void nhmex_uncore_msr_enable_event(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + + if (hwc->idx >= UNCORE_PMC_IDX_FIXED) + wrmsrl(hwc->config_base, NHMEX_PMON_CTL_EN_BIT0); + else if (box->pmu->type->event_mask & NHMEX_PMON_CTL_EN_BIT0) + wrmsrl(hwc->config_base, hwc->config | NHMEX_PMON_CTL_EN_BIT22); + else + wrmsrl(hwc->config_base, hwc->config | NHMEX_PMON_CTL_EN_BIT0); +} + +#define NHMEX_UNCORE_OPS_COMMON_INIT() \ + .init_box = nhmex_uncore_msr_init_box, \ + .disable_box = nhmex_uncore_msr_disable_box, \ + .enable_box = nhmex_uncore_msr_enable_box, \ + .disable_event = nhmex_uncore_msr_disable_event, \ + .read_counter = uncore_msr_read_counter + +static struct intel_uncore_ops nhmex_uncore_ops = { + NHMEX_UNCORE_OPS_COMMON_INIT(), + .enable_event = nhmex_uncore_msr_enable_event, +}; + +static struct attribute *nhmex_uncore_ubox_formats_attr[] = { + &format_attr_event.attr, + &format_attr_edge.attr, + NULL, +}; + +static struct attribute_group nhmex_uncore_ubox_format_group = { + .name = "format", + .attrs = nhmex_uncore_ubox_formats_attr, +}; + +static struct intel_uncore_type nhmex_uncore_ubox = { + .name = "ubox", + .num_counters = 1, + .num_boxes = 1, + .perf_ctr_bits = 48, + .event_ctl = NHMEX_U_MSR_PMON_EV_SEL, + .perf_ctr = NHMEX_U_MSR_PMON_CTR, + .event_mask = NHMEX_U_PMON_RAW_EVENT_MASK, + .box_ctl = NHMEX_U_MSR_PMON_GLOBAL_CTL, + .ops = &nhmex_uncore_ops, + .format_group = &nhmex_uncore_ubox_format_group +}; + +static struct attribute *nhmex_uncore_cbox_formats_attr[] = { + &format_attr_event.attr, + &format_attr_umask.attr, + &format_attr_edge.attr, + &format_attr_inv.attr, + &format_attr_thresh8.attr, + NULL, +}; + +static struct attribute_group nhmex_uncore_cbox_format_group = { + .name = "format", + .attrs = nhmex_uncore_cbox_formats_attr, +}; + +/* msr offset for each instance of cbox */ +static unsigned nhmex_cbox_msr_offsets[] = { + 0x0, 0x80, 0x40, 0xc0, 0x20, 0xa0, 0x60, 0xe0, 0x240, 0x2c0, +}; + +static struct intel_uncore_type nhmex_uncore_cbox = { + .name = "cbox", + .num_counters = 6, + .num_boxes = 10, + .perf_ctr_bits = 48, + .event_ctl = NHMEX_C0_MSR_PMON_EV_SEL0, + .perf_ctr = NHMEX_C0_MSR_PMON_CTR0, + .event_mask = NHMEX_PMON_RAW_EVENT_MASK, + .box_ctl = NHMEX_C0_MSR_PMON_GLOBAL_CTL, + .msr_offsets = nhmex_cbox_msr_offsets, + .pair_ctr_ctl = 1, + .ops = &nhmex_uncore_ops, + .format_group = &nhmex_uncore_cbox_format_group +}; + +static struct uncore_event_desc nhmex_uncore_wbox_events[] = { + INTEL_UNCORE_EVENT_DESC(clockticks, "event=0xff,umask=0"), + { /* end: all zeroes */ }, +}; + +static struct intel_uncore_type nhmex_uncore_wbox = { + .name = "wbox", + .num_counters = 4, + .num_boxes = 1, + .perf_ctr_bits = 48, + .event_ctl = NHMEX_W_MSR_PMON_CNT0, + .perf_ctr = NHMEX_W_MSR_PMON_EVT_SEL0, + .fixed_ctr = NHMEX_W_MSR_PMON_FIXED_CTR, + .fixed_ctl = NHMEX_W_MSR_PMON_FIXED_CTL, + .event_mask = NHMEX_PMON_RAW_EVENT_MASK, + .box_ctl = NHMEX_W_MSR_GLOBAL_CTL, + .pair_ctr_ctl = 1, + .event_descs = nhmex_uncore_wbox_events, + .ops = &nhmex_uncore_ops, + .format_group = &nhmex_uncore_cbox_format_group +}; + +static int nhmex_bbox_hw_config(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &hwc->extra_reg; + struct hw_perf_event_extra *reg2 = &hwc->branch_reg; + int ctr, ev_sel; + + ctr = (hwc->config & NHMEX_B_PMON_CTR_MASK) >> + NHMEX_B_PMON_CTR_SHIFT; + ev_sel = (hwc->config & NHMEX_B_PMON_CTL_EV_SEL_MASK) >> + NHMEX_B_PMON_CTL_EV_SEL_SHIFT; + + /* events that do not use the match/mask registers */ + if ((ctr == 0 && ev_sel > 0x3) || (ctr == 1 && ev_sel > 0x6) || + (ctr == 2 && ev_sel != 0x4) || ctr == 3) + return 0; + + if (box->pmu->pmu_idx == 0) + reg1->reg = NHMEX_B0_MSR_MATCH; + else + reg1->reg = NHMEX_B1_MSR_MATCH; + reg1->idx = 0; + reg1->config = event->attr.config1; + reg2->config = event->attr.config2; + return 0; +} + +static void nhmex_bbox_msr_enable_event(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &hwc->extra_reg; + struct hw_perf_event_extra *reg2 = &hwc->branch_reg; + + if (reg1->idx != EXTRA_REG_NONE) { + wrmsrl(reg1->reg, reg1->config); + wrmsrl(reg1->reg + 1, reg2->config); + } + wrmsrl(hwc->config_base, NHMEX_PMON_CTL_EN_BIT0 | + (hwc->config & NHMEX_B_PMON_CTL_EV_SEL_MASK)); +} + +/* + * The Bbox has 4 counters, but each counter monitors different events. + * Use bits 6-7 in the event config to select counter. + */ +static struct event_constraint nhmex_uncore_bbox_constraints[] = { + EVENT_CONSTRAINT(0 , 1, 0xc0), + EVENT_CONSTRAINT(0x40, 2, 0xc0), + EVENT_CONSTRAINT(0x80, 4, 0xc0), + EVENT_CONSTRAINT(0xc0, 8, 0xc0), + EVENT_CONSTRAINT_END, +}; + +static struct attribute *nhmex_uncore_bbox_formats_attr[] = { + &format_attr_event5.attr, + &format_attr_counter.attr, + &format_attr_match.attr, + &format_attr_mask.attr, + NULL, +}; + +static struct attribute_group nhmex_uncore_bbox_format_group = { + .name = "format", + .attrs = nhmex_uncore_bbox_formats_attr, +}; + +static struct intel_uncore_ops nhmex_uncore_bbox_ops = { + NHMEX_UNCORE_OPS_COMMON_INIT(), + .enable_event = nhmex_bbox_msr_enable_event, + .hw_config = nhmex_bbox_hw_config, + .get_constraint = uncore_get_constraint, + .put_constraint = uncore_put_constraint, +}; + +static struct intel_uncore_type nhmex_uncore_bbox = { + .name = "bbox", + .num_counters = 4, + .num_boxes = 2, + .perf_ctr_bits = 48, + .event_ctl = NHMEX_B0_MSR_PMON_CTL0, + .perf_ctr = NHMEX_B0_MSR_PMON_CTR0, + .event_mask = NHMEX_B_PMON_RAW_EVENT_MASK, + .box_ctl = NHMEX_B0_MSR_PMON_GLOBAL_CTL, + .msr_offset = NHMEX_B_MSR_OFFSET, + .pair_ctr_ctl = 1, + .num_shared_regs = 1, + .constraints = nhmex_uncore_bbox_constraints, + .ops = &nhmex_uncore_bbox_ops, + .format_group = &nhmex_uncore_bbox_format_group +}; + +static int nhmex_sbox_hw_config(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &hwc->extra_reg; + struct hw_perf_event_extra *reg2 = &hwc->branch_reg; + + /* only TO_R_PROG_EV event uses the match/mask register */ + if ((hwc->config & NHMEX_PMON_CTL_EV_SEL_MASK) != + NHMEX_S_EVENT_TO_R_PROG_EV) + return 0; + + if (box->pmu->pmu_idx == 0) + reg1->reg = NHMEX_S0_MSR_MM_CFG; + else + reg1->reg = NHMEX_S1_MSR_MM_CFG; + reg1->idx = 0; + reg1->config = event->attr.config1; + reg2->config = event->attr.config2; + return 0; +} + +static void nhmex_sbox_msr_enable_event(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &hwc->extra_reg; + struct hw_perf_event_extra *reg2 = &hwc->branch_reg; + + if (reg1->idx != EXTRA_REG_NONE) { + wrmsrl(reg1->reg, 0); + wrmsrl(reg1->reg + 1, reg1->config); + wrmsrl(reg1->reg + 2, reg2->config); + wrmsrl(reg1->reg, NHMEX_S_PMON_MM_CFG_EN); + } + wrmsrl(hwc->config_base, hwc->config | NHMEX_PMON_CTL_EN_BIT22); +} + +static struct attribute *nhmex_uncore_sbox_formats_attr[] = { + &format_attr_event.attr, + &format_attr_umask.attr, + &format_attr_edge.attr, + &format_attr_inv.attr, + &format_attr_thresh8.attr, + &format_attr_match.attr, + &format_attr_mask.attr, + NULL, +}; + +static struct attribute_group nhmex_uncore_sbox_format_group = { + .name = "format", + .attrs = nhmex_uncore_sbox_formats_attr, +}; + +static struct intel_uncore_ops nhmex_uncore_sbox_ops = { + NHMEX_UNCORE_OPS_COMMON_INIT(), + .enable_event = nhmex_sbox_msr_enable_event, + .hw_config = nhmex_sbox_hw_config, + .get_constraint = uncore_get_constraint, + .put_constraint = uncore_put_constraint, +}; + +static struct intel_uncore_type nhmex_uncore_sbox = { + .name = "sbox", + .num_counters = 4, + .num_boxes = 2, + .perf_ctr_bits = 48, + .event_ctl = NHMEX_S0_MSR_PMON_CTL0, + .perf_ctr = NHMEX_S0_MSR_PMON_CTR0, + .event_mask = NHMEX_PMON_RAW_EVENT_MASK, + .box_ctl = NHMEX_S0_MSR_PMON_GLOBAL_CTL, + .msr_offset = NHMEX_S_MSR_OFFSET, + .pair_ctr_ctl = 1, + .num_shared_regs = 1, + .ops = &nhmex_uncore_sbox_ops, + .format_group = &nhmex_uncore_sbox_format_group +}; + +enum { + EXTRA_REG_NHMEX_M_FILTER, + EXTRA_REG_NHMEX_M_DSP, + EXTRA_REG_NHMEX_M_ISS, + EXTRA_REG_NHMEX_M_MAP, + EXTRA_REG_NHMEX_M_MSC_THR, + EXTRA_REG_NHMEX_M_PGT, + EXTRA_REG_NHMEX_M_PLD, + EXTRA_REG_NHMEX_M_ZDP_CTL_FVC, +}; + +static struct extra_reg nhmex_uncore_mbox_extra_regs[] = { + MBOX_INC_SEL_EXTAR_REG(0x0, DSP), + MBOX_INC_SEL_EXTAR_REG(0x4, MSC_THR), + MBOX_INC_SEL_EXTAR_REG(0x5, MSC_THR), + MBOX_INC_SEL_EXTAR_REG(0x9, ISS), + /* event 0xa uses two extra registers */ + MBOX_INC_SEL_EXTAR_REG(0xa, ISS), + MBOX_INC_SEL_EXTAR_REG(0xa, PLD), + MBOX_INC_SEL_EXTAR_REG(0xb, PLD), + /* events 0xd ~ 0x10 use the same extra register */ + MBOX_INC_SEL_EXTAR_REG(0xd, ZDP_CTL_FVC), + MBOX_INC_SEL_EXTAR_REG(0xe, ZDP_CTL_FVC), + MBOX_INC_SEL_EXTAR_REG(0xf, ZDP_CTL_FVC), + MBOX_INC_SEL_EXTAR_REG(0x10, ZDP_CTL_FVC), + MBOX_INC_SEL_EXTAR_REG(0x16, PGT), + MBOX_SET_FLAG_SEL_EXTRA_REG(0x0, DSP), + MBOX_SET_FLAG_SEL_EXTRA_REG(0x1, ISS), + MBOX_SET_FLAG_SEL_EXTRA_REG(0x5, PGT), + MBOX_SET_FLAG_SEL_EXTRA_REG(0x6, MAP), + EVENT_EXTRA_END +}; + +/* Nehalem-EX or Westmere-EX ? */ +static bool uncore_nhmex; + +static bool nhmex_mbox_get_shared_reg(struct intel_uncore_box *box, int idx, u64 config) +{ + struct intel_uncore_extra_reg *er; + unsigned long flags; + bool ret = false; + u64 mask; + + if (idx < EXTRA_REG_NHMEX_M_ZDP_CTL_FVC) { + er = &box->shared_regs[idx]; + raw_spin_lock_irqsave(&er->lock, flags); + if (!atomic_read(&er->ref) || er->config == config) { + atomic_inc(&er->ref); + er->config = config; + ret = true; + } + raw_spin_unlock_irqrestore(&er->lock, flags); + + return ret; + } + /* + * The ZDP_CTL_FVC MSR has 4 fields which are used to control + * events 0xd ~ 0x10. Besides these 4 fields, there are additional + * fields which are shared. + */ + idx -= EXTRA_REG_NHMEX_M_ZDP_CTL_FVC; + if (WARN_ON_ONCE(idx >= 4)) + return false; + + /* mask of the shared fields */ + if (uncore_nhmex) + mask = NHMEX_M_PMON_ZDP_CTL_FVC_MASK; + else + mask = WSMEX_M_PMON_ZDP_CTL_FVC_MASK; + er = &box->shared_regs[EXTRA_REG_NHMEX_M_ZDP_CTL_FVC]; + + raw_spin_lock_irqsave(&er->lock, flags); + /* add mask of the non-shared field if it's in use */ + if (__BITS_VALUE(atomic_read(&er->ref), idx, 8)) { + if (uncore_nhmex) + mask |= NHMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(idx); + else + mask |= WSMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(idx); + } + + if (!atomic_read(&er->ref) || !((er->config ^ config) & mask)) { + atomic_add(1 << (idx * 8), &er->ref); + if (uncore_nhmex) + mask = NHMEX_M_PMON_ZDP_CTL_FVC_MASK | + NHMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(idx); + else + mask = WSMEX_M_PMON_ZDP_CTL_FVC_MASK | + WSMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(idx); + er->config &= ~mask; + er->config |= (config & mask); + ret = true; + } + raw_spin_unlock_irqrestore(&er->lock, flags); + + return ret; +} + +static void nhmex_mbox_put_shared_reg(struct intel_uncore_box *box, int idx) +{ + struct intel_uncore_extra_reg *er; + + if (idx < EXTRA_REG_NHMEX_M_ZDP_CTL_FVC) { + er = &box->shared_regs[idx]; + atomic_dec(&er->ref); + return; + } + + idx -= EXTRA_REG_NHMEX_M_ZDP_CTL_FVC; + er = &box->shared_regs[EXTRA_REG_NHMEX_M_ZDP_CTL_FVC]; + atomic_sub(1 << (idx * 8), &er->ref); +} + +static u64 nhmex_mbox_alter_er(struct perf_event *event, int new_idx, bool modify) +{ + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &hwc->extra_reg; + u64 idx, orig_idx = __BITS_VALUE(reg1->idx, 0, 8); + u64 config = reg1->config; + + /* get the non-shared control bits and shift them */ + idx = orig_idx - EXTRA_REG_NHMEX_M_ZDP_CTL_FVC; + if (uncore_nhmex) + config &= NHMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(idx); + else + config &= WSMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(idx); + if (new_idx > orig_idx) { + idx = new_idx - orig_idx; + config <<= 3 * idx; + } else { + idx = orig_idx - new_idx; + config >>= 3 * idx; + } + + /* add the shared control bits back */ + if (uncore_nhmex) + config |= NHMEX_M_PMON_ZDP_CTL_FVC_MASK & reg1->config; + else + config |= WSMEX_M_PMON_ZDP_CTL_FVC_MASK & reg1->config; + config |= NHMEX_M_PMON_ZDP_CTL_FVC_MASK & reg1->config; + if (modify) { + /* adjust the main event selector */ + if (new_idx > orig_idx) + hwc->config += idx << NHMEX_M_PMON_CTL_INC_SEL_SHIFT; + else + hwc->config -= idx << NHMEX_M_PMON_CTL_INC_SEL_SHIFT; + reg1->config = config; + reg1->idx = ~0xff | new_idx; + } + return config; +} + +static struct event_constraint * +nhmex_mbox_get_constraint(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event_extra *reg1 = &event->hw.extra_reg; + struct hw_perf_event_extra *reg2 = &event->hw.branch_reg; + int i, idx[2], alloc = 0; + u64 config1 = reg1->config; + + idx[0] = __BITS_VALUE(reg1->idx, 0, 8); + idx[1] = __BITS_VALUE(reg1->idx, 1, 8); +again: + for (i = 0; i < 2; i++) { + if (!uncore_box_is_fake(box) && (reg1->alloc & (0x1 << i))) + idx[i] = 0xff; + + if (idx[i] == 0xff) + continue; + + if (!nhmex_mbox_get_shared_reg(box, idx[i], + __BITS_VALUE(config1, i, 32))) + goto fail; + alloc |= (0x1 << i); + } + + /* for the match/mask registers */ + if (reg2->idx != EXTRA_REG_NONE && + (uncore_box_is_fake(box) || !reg2->alloc) && + !nhmex_mbox_get_shared_reg(box, reg2->idx, reg2->config)) + goto fail; + + /* + * If it's a fake box -- as per validate_{group,event}() we + * shouldn't touch event state and we can avoid doing so + * since both will only call get_event_constraints() once + * on each event, this avoids the need for reg->alloc. + */ + if (!uncore_box_is_fake(box)) { + if (idx[0] != 0xff && idx[0] != __BITS_VALUE(reg1->idx, 0, 8)) + nhmex_mbox_alter_er(event, idx[0], true); + reg1->alloc |= alloc; + if (reg2->idx != EXTRA_REG_NONE) + reg2->alloc = 1; + } + return NULL; +fail: + if (idx[0] != 0xff && !(alloc & 0x1) && + idx[0] >= EXTRA_REG_NHMEX_M_ZDP_CTL_FVC) { + /* + * events 0xd ~ 0x10 are functional identical, but are + * controlled by different fields in the ZDP_CTL_FVC + * register. If we failed to take one field, try the + * rest 3 choices. + */ + BUG_ON(__BITS_VALUE(reg1->idx, 1, 8) != 0xff); + idx[0] -= EXTRA_REG_NHMEX_M_ZDP_CTL_FVC; + idx[0] = (idx[0] + 1) % 4; + idx[0] += EXTRA_REG_NHMEX_M_ZDP_CTL_FVC; + if (idx[0] != __BITS_VALUE(reg1->idx, 0, 8)) { + config1 = nhmex_mbox_alter_er(event, idx[0], false); + goto again; + } + } + + if (alloc & 0x1) + nhmex_mbox_put_shared_reg(box, idx[0]); + if (alloc & 0x2) + nhmex_mbox_put_shared_reg(box, idx[1]); + return &constraint_empty; +} + +static void nhmex_mbox_put_constraint(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event_extra *reg1 = &event->hw.extra_reg; + struct hw_perf_event_extra *reg2 = &event->hw.branch_reg; + + if (uncore_box_is_fake(box)) + return; + + if (reg1->alloc & 0x1) + nhmex_mbox_put_shared_reg(box, __BITS_VALUE(reg1->idx, 0, 8)); + if (reg1->alloc & 0x2) + nhmex_mbox_put_shared_reg(box, __BITS_VALUE(reg1->idx, 1, 8)); + reg1->alloc = 0; + + if (reg2->alloc) { + nhmex_mbox_put_shared_reg(box, reg2->idx); + reg2->alloc = 0; + } +} + +static int nhmex_mbox_extra_reg_idx(struct extra_reg *er) +{ + if (er->idx < EXTRA_REG_NHMEX_M_ZDP_CTL_FVC) + return er->idx; + return er->idx + (er->event >> NHMEX_M_PMON_CTL_INC_SEL_SHIFT) - 0xd; +} + +static int nhmex_mbox_hw_config(struct intel_uncore_box *box, struct perf_event *event) +{ + struct intel_uncore_type *type = box->pmu->type; + struct hw_perf_event_extra *reg1 = &event->hw.extra_reg; + struct hw_perf_event_extra *reg2 = &event->hw.branch_reg; + struct extra_reg *er; + unsigned msr; + int reg_idx = 0; + /* + * The mbox events may require 2 extra MSRs at the most. But only + * the lower 32 bits in these MSRs are significant, so we can use + * config1 to pass two MSRs' config. + */ + for (er = nhmex_uncore_mbox_extra_regs; er->msr; er++) { + if (er->event != (event->hw.config & er->config_mask)) + continue; + if (event->attr.config1 & ~er->valid_mask) + return -EINVAL; + + msr = er->msr + type->msr_offset * box->pmu->pmu_idx; + if (WARN_ON_ONCE(msr >= 0xffff || er->idx >= 0xff)) + return -EINVAL; + + /* always use the 32~63 bits to pass the PLD config */ + if (er->idx == EXTRA_REG_NHMEX_M_PLD) + reg_idx = 1; + else if (WARN_ON_ONCE(reg_idx > 0)) + return -EINVAL; + + reg1->idx &= ~(0xff << (reg_idx * 8)); + reg1->reg &= ~(0xffff << (reg_idx * 16)); + reg1->idx |= nhmex_mbox_extra_reg_idx(er) << (reg_idx * 8); + reg1->reg |= msr << (reg_idx * 16); + reg1->config = event->attr.config1; + reg_idx++; + } + /* + * The mbox only provides ability to perform address matching + * for the PLD events. + */ + if (reg_idx == 2) { + reg2->idx = EXTRA_REG_NHMEX_M_FILTER; + if (event->attr.config2 & NHMEX_M_PMON_MM_CFG_EN) + reg2->config = event->attr.config2; + else + reg2->config = ~0ULL; + if (box->pmu->pmu_idx == 0) + reg2->reg = NHMEX_M0_MSR_PMU_MM_CFG; + else + reg2->reg = NHMEX_M1_MSR_PMU_MM_CFG; + } + return 0; +} + +static u64 nhmex_mbox_shared_reg_config(struct intel_uncore_box *box, int idx) +{ + struct intel_uncore_extra_reg *er; + unsigned long flags; + u64 config; + + if (idx < EXTRA_REG_NHMEX_M_ZDP_CTL_FVC) + return box->shared_regs[idx].config; + + er = &box->shared_regs[EXTRA_REG_NHMEX_M_ZDP_CTL_FVC]; + raw_spin_lock_irqsave(&er->lock, flags); + config = er->config; + raw_spin_unlock_irqrestore(&er->lock, flags); + return config; +} + +static void nhmex_mbox_msr_enable_event(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &hwc->extra_reg; + struct hw_perf_event_extra *reg2 = &hwc->branch_reg; + int idx; + + idx = __BITS_VALUE(reg1->idx, 0, 8); + if (idx != 0xff) + wrmsrl(__BITS_VALUE(reg1->reg, 0, 16), + nhmex_mbox_shared_reg_config(box, idx)); + idx = __BITS_VALUE(reg1->idx, 1, 8); + if (idx != 0xff) + wrmsrl(__BITS_VALUE(reg1->reg, 1, 16), + nhmex_mbox_shared_reg_config(box, idx)); + + if (reg2->idx != EXTRA_REG_NONE) { + wrmsrl(reg2->reg, 0); + if (reg2->config != ~0ULL) { + wrmsrl(reg2->reg + 1, + reg2->config & NHMEX_M_PMON_ADDR_MATCH_MASK); + wrmsrl(reg2->reg + 2, NHMEX_M_PMON_ADDR_MASK_MASK & + (reg2->config >> NHMEX_M_PMON_ADDR_MASK_SHIFT)); + wrmsrl(reg2->reg, NHMEX_M_PMON_MM_CFG_EN); + } + } + + wrmsrl(hwc->config_base, hwc->config | NHMEX_PMON_CTL_EN_BIT0); +} + +DEFINE_UNCORE_FORMAT_ATTR(count_mode, count_mode, "config:2-3"); +DEFINE_UNCORE_FORMAT_ATTR(storage_mode, storage_mode, "config:4-5"); +DEFINE_UNCORE_FORMAT_ATTR(wrap_mode, wrap_mode, "config:6"); +DEFINE_UNCORE_FORMAT_ATTR(flag_mode, flag_mode, "config:7"); +DEFINE_UNCORE_FORMAT_ATTR(inc_sel, inc_sel, "config:9-13"); +DEFINE_UNCORE_FORMAT_ATTR(set_flag_sel, set_flag_sel, "config:19-21"); +DEFINE_UNCORE_FORMAT_ATTR(filter_cfg_en, filter_cfg_en, "config2:63"); +DEFINE_UNCORE_FORMAT_ATTR(filter_match, filter_match, "config2:0-33"); +DEFINE_UNCORE_FORMAT_ATTR(filter_mask, filter_mask, "config2:34-61"); +DEFINE_UNCORE_FORMAT_ATTR(dsp, dsp, "config1:0-31"); +DEFINE_UNCORE_FORMAT_ATTR(thr, thr, "config1:0-31"); +DEFINE_UNCORE_FORMAT_ATTR(fvc, fvc, "config1:0-31"); +DEFINE_UNCORE_FORMAT_ATTR(pgt, pgt, "config1:0-31"); +DEFINE_UNCORE_FORMAT_ATTR(map, map, "config1:0-31"); +DEFINE_UNCORE_FORMAT_ATTR(iss, iss, "config1:0-31"); +DEFINE_UNCORE_FORMAT_ATTR(pld, pld, "config1:32-63"); + +static struct attribute *nhmex_uncore_mbox_formats_attr[] = { + &format_attr_count_mode.attr, + &format_attr_storage_mode.attr, + &format_attr_wrap_mode.attr, + &format_attr_flag_mode.attr, + &format_attr_inc_sel.attr, + &format_attr_set_flag_sel.attr, + &format_attr_filter_cfg_en.attr, + &format_attr_filter_match.attr, + &format_attr_filter_mask.attr, + &format_attr_dsp.attr, + &format_attr_thr.attr, + &format_attr_fvc.attr, + &format_attr_pgt.attr, + &format_attr_map.attr, + &format_attr_iss.attr, + &format_attr_pld.attr, + NULL, +}; + +static struct attribute_group nhmex_uncore_mbox_format_group = { + .name = "format", + .attrs = nhmex_uncore_mbox_formats_attr, +}; + +static struct uncore_event_desc nhmex_uncore_mbox_events[] = { + INTEL_UNCORE_EVENT_DESC(bbox_cmds_read, "inc_sel=0xd,fvc=0x2800"), + INTEL_UNCORE_EVENT_DESC(bbox_cmds_write, "inc_sel=0xd,fvc=0x2820"), + { /* end: all zeroes */ }, +}; + +static struct uncore_event_desc wsmex_uncore_mbox_events[] = { + INTEL_UNCORE_EVENT_DESC(bbox_cmds_read, "inc_sel=0xd,fvc=0x5000"), + INTEL_UNCORE_EVENT_DESC(bbox_cmds_write, "inc_sel=0xd,fvc=0x5040"), + { /* end: all zeroes */ }, +}; + +static struct intel_uncore_ops nhmex_uncore_mbox_ops = { + NHMEX_UNCORE_OPS_COMMON_INIT(), + .enable_event = nhmex_mbox_msr_enable_event, + .hw_config = nhmex_mbox_hw_config, + .get_constraint = nhmex_mbox_get_constraint, + .put_constraint = nhmex_mbox_put_constraint, +}; + +static struct intel_uncore_type nhmex_uncore_mbox = { + .name = "mbox", + .num_counters = 6, + .num_boxes = 2, + .perf_ctr_bits = 48, + .event_ctl = NHMEX_M0_MSR_PMU_CTL0, + .perf_ctr = NHMEX_M0_MSR_PMU_CNT0, + .event_mask = NHMEX_M_PMON_RAW_EVENT_MASK, + .box_ctl = NHMEX_M0_MSR_GLOBAL_CTL, + .msr_offset = NHMEX_M_MSR_OFFSET, + .pair_ctr_ctl = 1, + .num_shared_regs = 8, + .event_descs = nhmex_uncore_mbox_events, + .ops = &nhmex_uncore_mbox_ops, + .format_group = &nhmex_uncore_mbox_format_group, +}; + +static void nhmex_rbox_alter_er(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &hwc->extra_reg; + + /* adjust the main event selector and extra register index */ + if (reg1->idx % 2) { + reg1->idx--; + hwc->config -= 1 << NHMEX_R_PMON_CTL_EV_SEL_SHIFT; + } else { + reg1->idx++; + hwc->config += 1 << NHMEX_R_PMON_CTL_EV_SEL_SHIFT; + } + + /* adjust extra register config */ + switch (reg1->idx % 6) { + case 2: + /* shift the 8~15 bits to the 0~7 bits */ + reg1->config >>= 8; + break; + case 3: + /* shift the 0~7 bits to the 8~15 bits */ + reg1->config <<= 8; + break; + }; +} + +/* + * Each rbox has 4 event set which monitor PQI port 0~3 or 4~7. + * An event set consists of 6 events, the 3rd and 4th events in + * an event set use the same extra register. So an event set uses + * 5 extra registers. + */ +static struct event_constraint * +nhmex_rbox_get_constraint(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &hwc->extra_reg; + struct hw_perf_event_extra *reg2 = &hwc->branch_reg; + struct intel_uncore_extra_reg *er; + unsigned long flags; + int idx, er_idx; + u64 config1; + bool ok = false; + + if (!uncore_box_is_fake(box) && reg1->alloc) + return NULL; + + idx = reg1->idx % 6; + config1 = reg1->config; +again: + er_idx = idx; + /* the 3rd and 4th events use the same extra register */ + if (er_idx > 2) + er_idx--; + er_idx += (reg1->idx / 6) * 5; + + er = &box->shared_regs[er_idx]; + raw_spin_lock_irqsave(&er->lock, flags); + if (idx < 2) { + if (!atomic_read(&er->ref) || er->config == reg1->config) { + atomic_inc(&er->ref); + er->config = reg1->config; + ok = true; + } + } else if (idx == 2 || idx == 3) { + /* + * these two events use different fields in a extra register, + * the 0~7 bits and the 8~15 bits respectively. + */ + u64 mask = 0xff << ((idx - 2) * 8); + if (!__BITS_VALUE(atomic_read(&er->ref), idx - 2, 8) || + !((er->config ^ config1) & mask)) { + atomic_add(1 << ((idx - 2) * 8), &er->ref); + er->config &= ~mask; + er->config |= config1 & mask; + ok = true; + } + } else { + if (!atomic_read(&er->ref) || + (er->config == (hwc->config >> 32) && + er->config1 == reg1->config && + er->config2 == reg2->config)) { + atomic_inc(&er->ref); + er->config = (hwc->config >> 32); + er->config1 = reg1->config; + er->config2 = reg2->config; + ok = true; + } + } + raw_spin_unlock_irqrestore(&er->lock, flags); + + if (!ok) { + /* + * The Rbox events are always in pairs. The paired + * events are functional identical, but use different + * extra registers. If we failed to take an extra + * register, try the alternative. + */ + if (idx % 2) + idx--; + else + idx++; + if (idx != reg1->idx % 6) { + if (idx == 2) + config1 >>= 8; + else if (idx == 3) + config1 <<= 8; + goto again; + } + } else { + if (!uncore_box_is_fake(box)) { + if (idx != reg1->idx % 6) + nhmex_rbox_alter_er(box, event); + reg1->alloc = 1; + } + return NULL; + } + return &constraint_empty; +} + +static void nhmex_rbox_put_constraint(struct intel_uncore_box *box, struct perf_event *event) +{ + struct intel_uncore_extra_reg *er; + struct hw_perf_event_extra *reg1 = &event->hw.extra_reg; + int idx, er_idx; + + if (uncore_box_is_fake(box) || !reg1->alloc) + return; + + idx = reg1->idx % 6; + er_idx = idx; + if (er_idx > 2) + er_idx--; + er_idx += (reg1->idx / 6) * 5; + + er = &box->shared_regs[er_idx]; + if (idx == 2 || idx == 3) + atomic_sub(1 << ((idx - 2) * 8), &er->ref); + else + atomic_dec(&er->ref); + + reg1->alloc = 0; +} + +static int nhmex_rbox_hw_config(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &event->hw.extra_reg; + struct hw_perf_event_extra *reg2 = &event->hw.branch_reg; + int idx; + + idx = (event->hw.config & NHMEX_R_PMON_CTL_EV_SEL_MASK) >> + NHMEX_R_PMON_CTL_EV_SEL_SHIFT; + if (idx >= 0x18) + return -EINVAL; + + reg1->idx = idx; + reg1->config = event->attr.config1; + + switch (idx % 6) { + case 4: + case 5: + hwc->config |= event->attr.config & (~0ULL << 32); + reg2->config = event->attr.config2; + break; + }; + return 0; +} + +static void nhmex_rbox_msr_enable_event(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + struct hw_perf_event_extra *reg1 = &hwc->extra_reg; + struct hw_perf_event_extra *reg2 = &hwc->branch_reg; + int idx, port; + + idx = reg1->idx; + port = idx / 6 + box->pmu->pmu_idx * 4; + + switch (idx % 6) { + case 0: + wrmsrl(NHMEX_R_MSR_PORTN_IPERF_CFG0(port), reg1->config); + break; + case 1: + wrmsrl(NHMEX_R_MSR_PORTN_IPERF_CFG1(port), reg1->config); + break; + case 2: + case 3: + wrmsrl(NHMEX_R_MSR_PORTN_QLX_CFG(port), + uncore_shared_reg_config(box, 2 + (idx / 6) * 5)); + break; + case 4: + wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET1_MM_CFG(port), + hwc->config >> 32); + wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET1_MATCH(port), reg1->config); + wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET1_MASK(port), reg2->config); + break; + case 5: + wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET2_MM_CFG(port), + hwc->config >> 32); + wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET2_MATCH(port), reg1->config); + wrmsrl(NHMEX_R_MSR_PORTN_XBR_SET2_MASK(port), reg2->config); + break; + }; + + wrmsrl(hwc->config_base, NHMEX_PMON_CTL_EN_BIT0 | + (hwc->config & NHMEX_R_PMON_CTL_EV_SEL_MASK)); +} + +DEFINE_UNCORE_FORMAT_ATTR(xbr_mm_cfg, xbr_mm_cfg, "config:32-63"); +DEFINE_UNCORE_FORMAT_ATTR(xbr_match, xbr_match, "config1:0-63"); +DEFINE_UNCORE_FORMAT_ATTR(xbr_mask, xbr_mask, "config2:0-63"); +DEFINE_UNCORE_FORMAT_ATTR(qlx_cfg, qlx_cfg, "config1:0-15"); +DEFINE_UNCORE_FORMAT_ATTR(iperf_cfg, iperf_cfg, "config1:0-31"); + +static struct attribute *nhmex_uncore_rbox_formats_attr[] = { + &format_attr_event5.attr, + &format_attr_xbr_mm_cfg.attr, + &format_attr_xbr_match.attr, + &format_attr_xbr_mask.attr, + &format_attr_qlx_cfg.attr, + &format_attr_iperf_cfg.attr, + NULL, +}; + +static struct attribute_group nhmex_uncore_rbox_format_group = { + .name = "format", + .attrs = nhmex_uncore_rbox_formats_attr, +}; + +static struct uncore_event_desc nhmex_uncore_rbox_events[] = { + INTEL_UNCORE_EVENT_DESC(qpi0_flit_send, "event=0x0,iperf_cfg=0x80000000"), + INTEL_UNCORE_EVENT_DESC(qpi1_filt_send, "event=0x6,iperf_cfg=0x80000000"), + INTEL_UNCORE_EVENT_DESC(qpi0_idle_filt, "event=0x0,iperf_cfg=0x40000000"), + INTEL_UNCORE_EVENT_DESC(qpi1_idle_filt, "event=0x6,iperf_cfg=0x40000000"), + INTEL_UNCORE_EVENT_DESC(qpi0_date_response, "event=0x0,iperf_cfg=0xc4"), + INTEL_UNCORE_EVENT_DESC(qpi1_date_response, "event=0x6,iperf_cfg=0xc4"), + { /* end: all zeroes */ }, +}; + +static struct intel_uncore_ops nhmex_uncore_rbox_ops = { + NHMEX_UNCORE_OPS_COMMON_INIT(), + .enable_event = nhmex_rbox_msr_enable_event, + .hw_config = nhmex_rbox_hw_config, + .get_constraint = nhmex_rbox_get_constraint, + .put_constraint = nhmex_rbox_put_constraint, +}; + +static struct intel_uncore_type nhmex_uncore_rbox = { + .name = "rbox", + .num_counters = 8, + .num_boxes = 2, + .perf_ctr_bits = 48, + .event_ctl = NHMEX_R_MSR_PMON_CTL0, + .perf_ctr = NHMEX_R_MSR_PMON_CNT0, + .event_mask = NHMEX_R_PMON_RAW_EVENT_MASK, + .box_ctl = NHMEX_R_MSR_GLOBAL_CTL, + .msr_offset = NHMEX_R_MSR_OFFSET, + .pair_ctr_ctl = 1, + .num_shared_regs = 20, + .event_descs = nhmex_uncore_rbox_events, + .ops = &nhmex_uncore_rbox_ops, + .format_group = &nhmex_uncore_rbox_format_group +}; + +static struct intel_uncore_type *nhmex_msr_uncores[] = { + &nhmex_uncore_ubox, + &nhmex_uncore_cbox, + &nhmex_uncore_bbox, + &nhmex_uncore_sbox, + &nhmex_uncore_mbox, + &nhmex_uncore_rbox, + &nhmex_uncore_wbox, + NULL, +}; +/* end of Nehalem-EX uncore support */ + +static void uncore_assign_hw_event(struct intel_uncore_box *box, struct perf_event *event, int idx) +{ + struct hw_perf_event *hwc = &event->hw; + + hwc->idx = idx; + hwc->last_tag = ++box->tags[idx]; + + if (hwc->idx == UNCORE_PMC_IDX_FIXED) { + hwc->event_base = uncore_fixed_ctr(box); + hwc->config_base = uncore_fixed_ctl(box); + return; + } + + hwc->config_base = uncore_event_ctl(box, hwc->idx); + hwc->event_base = uncore_perf_ctr(box, hwc->idx); +} + +static void uncore_perf_event_update(struct intel_uncore_box *box, struct perf_event *event) +{ + u64 prev_count, new_count, delta; + int shift; + + if (event->hw.idx >= UNCORE_PMC_IDX_FIXED) + shift = 64 - uncore_fixed_ctr_bits(box); + else + shift = 64 - uncore_perf_ctr_bits(box); + + /* the hrtimer might modify the previous event value */ +again: + prev_count = local64_read(&event->hw.prev_count); + new_count = uncore_read_counter(box, event); + if (local64_xchg(&event->hw.prev_count, new_count) != prev_count) + goto again; + + delta = (new_count << shift) - (prev_count << shift); + delta >>= shift; + + local64_add(delta, &event->count); +} + +/* + * The overflow interrupt is unavailable for SandyBridge-EP, is broken + * for SandyBridge. So we use hrtimer to periodically poll the counter + * to avoid overflow. + */ +static enum hrtimer_restart uncore_pmu_hrtimer(struct hrtimer *hrtimer) +{ + struct intel_uncore_box *box; + struct perf_event *event; + unsigned long flags; + int bit; + + box = container_of(hrtimer, struct intel_uncore_box, hrtimer); + if (!box->n_active || box->cpu != smp_processor_id()) + return HRTIMER_NORESTART; + /* + * disable local interrupt to prevent uncore_pmu_event_start/stop + * to interrupt the update process + */ + local_irq_save(flags); + + /* + * handle boxes with an active event list as opposed to active + * counters + */ + list_for_each_entry(event, &box->active_list, active_entry) { + uncore_perf_event_update(box, event); + } + + for_each_set_bit(bit, box->active_mask, UNCORE_PMC_IDX_MAX) + uncore_perf_event_update(box, box->events[bit]); + + local_irq_restore(flags); + + hrtimer_forward_now(hrtimer, ns_to_ktime(box->hrtimer_duration)); + return HRTIMER_RESTART; +} + +static void uncore_pmu_start_hrtimer(struct intel_uncore_box *box) +{ + __hrtimer_start_range_ns(&box->hrtimer, + ns_to_ktime(box->hrtimer_duration), 0, + HRTIMER_MODE_REL_PINNED, 0); +} + +static void uncore_pmu_cancel_hrtimer(struct intel_uncore_box *box) +{ + hrtimer_cancel(&box->hrtimer); +} + +static void uncore_pmu_init_hrtimer(struct intel_uncore_box *box) +{ + hrtimer_init(&box->hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + box->hrtimer.function = uncore_pmu_hrtimer; +} + +static struct intel_uncore_box *uncore_alloc_box(struct intel_uncore_type *type, int node) +{ + struct intel_uncore_box *box; + int i, size; + + size = sizeof(*box) + type->num_shared_regs * sizeof(struct intel_uncore_extra_reg); + + box = kzalloc_node(size, GFP_KERNEL, node); + if (!box) + return NULL; + + for (i = 0; i < type->num_shared_regs; i++) + raw_spin_lock_init(&box->shared_regs[i].lock); + + uncore_pmu_init_hrtimer(box); + atomic_set(&box->refcnt, 1); + box->cpu = -1; + box->phys_id = -1; + + /* set default hrtimer timeout */ + box->hrtimer_duration = UNCORE_PMU_HRTIMER_INTERVAL; + + INIT_LIST_HEAD(&box->active_list); + + return box; +} + +static int +uncore_collect_events(struct intel_uncore_box *box, struct perf_event *leader, bool dogrp) +{ + struct perf_event *event; + int n, max_count; + + max_count = box->pmu->type->num_counters; + if (box->pmu->type->fixed_ctl) + max_count++; + + if (box->n_events >= max_count) + return -EINVAL; + + n = box->n_events; + box->event_list[n] = leader; + n++; + if (!dogrp) + return n; + + list_for_each_entry(event, &leader->sibling_list, group_entry) { + if (event->state <= PERF_EVENT_STATE_OFF) + continue; + + if (n >= max_count) + return -EINVAL; + + box->event_list[n] = event; + n++; + } + return n; +} + +static struct event_constraint * +uncore_get_event_constraint(struct intel_uncore_box *box, struct perf_event *event) +{ + struct intel_uncore_type *type = box->pmu->type; + struct event_constraint *c; + + if (type->ops->get_constraint) { + c = type->ops->get_constraint(box, event); + if (c) + return c; + } + + if (event->attr.config == UNCORE_FIXED_EVENT) + return &constraint_fixed; + + if (type->constraints) { + for_each_event_constraint(c, type->constraints) { + if ((event->hw.config & c->cmask) == c->code) + return c; + } + } + + return &type->unconstrainted; +} + +static void uncore_put_event_constraint(struct intel_uncore_box *box, struct perf_event *event) +{ + if (box->pmu->type->ops->put_constraint) + box->pmu->type->ops->put_constraint(box, event); +} + +static int uncore_assign_events(struct intel_uncore_box *box, int assign[], int n) +{ + unsigned long used_mask[BITS_TO_LONGS(UNCORE_PMC_IDX_MAX)]; + struct event_constraint *c; + int i, wmin, wmax, ret = 0; + struct hw_perf_event *hwc; + + bitmap_zero(used_mask, UNCORE_PMC_IDX_MAX); + + for (i = 0, wmin = UNCORE_PMC_IDX_MAX, wmax = 0; i < n; i++) { + hwc = &box->event_list[i]->hw; + c = uncore_get_event_constraint(box, box->event_list[i]); + hwc->constraint = c; + wmin = min(wmin, c->weight); + wmax = max(wmax, c->weight); + } + + /* fastpath, try to reuse previous register */ + for (i = 0; i < n; i++) { + hwc = &box->event_list[i]->hw; + c = hwc->constraint; + + /* never assigned */ + if (hwc->idx == -1) + break; + + /* constraint still honored */ + if (!test_bit(hwc->idx, c->idxmsk)) + break; + + /* not already used */ + if (test_bit(hwc->idx, used_mask)) + break; + + __set_bit(hwc->idx, used_mask); + if (assign) + assign[i] = hwc->idx; + } + /* slow path */ + if (i != n) + ret = perf_assign_events(box->event_list, n, + wmin, wmax, assign); + + if (!assign || ret) { + for (i = 0; i < n; i++) + uncore_put_event_constraint(box, box->event_list[i]); + } + return ret ? -EINVAL : 0; +} + +static void uncore_pmu_event_start(struct perf_event *event, int flags) +{ + struct intel_uncore_box *box = uncore_event_to_box(event); + int idx = event->hw.idx; + + if (WARN_ON_ONCE(!(event->hw.state & PERF_HES_STOPPED))) + return; + + if (WARN_ON_ONCE(idx == -1 || idx >= UNCORE_PMC_IDX_MAX)) + return; + + event->hw.state = 0; + box->events[idx] = event; + box->n_active++; + __set_bit(idx, box->active_mask); + + local64_set(&event->hw.prev_count, uncore_read_counter(box, event)); + uncore_enable_event(box, event); + + if (box->n_active == 1) { + uncore_enable_box(box); + uncore_pmu_start_hrtimer(box); + } +} + +static void uncore_pmu_event_stop(struct perf_event *event, int flags) +{ + struct intel_uncore_box *box = uncore_event_to_box(event); + struct hw_perf_event *hwc = &event->hw; + + if (__test_and_clear_bit(hwc->idx, box->active_mask)) { + uncore_disable_event(box, event); + box->n_active--; + box->events[hwc->idx] = NULL; + WARN_ON_ONCE(hwc->state & PERF_HES_STOPPED); + hwc->state |= PERF_HES_STOPPED; + + if (box->n_active == 0) { + uncore_disable_box(box); + uncore_pmu_cancel_hrtimer(box); + } + } + + if ((flags & PERF_EF_UPDATE) && !(hwc->state & PERF_HES_UPTODATE)) { + /* + * Drain the remaining delta count out of a event + * that we are disabling: + */ + uncore_perf_event_update(box, event); + hwc->state |= PERF_HES_UPTODATE; + } +} + +static int uncore_pmu_event_add(struct perf_event *event, int flags) +{ + struct intel_uncore_box *box = uncore_event_to_box(event); + struct hw_perf_event *hwc = &event->hw; + int assign[UNCORE_PMC_IDX_MAX]; + int i, n, ret; + + if (!box) + return -ENODEV; + + ret = n = uncore_collect_events(box, event, false); + if (ret < 0) + return ret; + + hwc->state = PERF_HES_UPTODATE | PERF_HES_STOPPED; + if (!(flags & PERF_EF_START)) + hwc->state |= PERF_HES_ARCH; + + ret = uncore_assign_events(box, assign, n); + if (ret) + return ret; + + /* save events moving to new counters */ + for (i = 0; i < box->n_events; i++) { + event = box->event_list[i]; + hwc = &event->hw; + + if (hwc->idx == assign[i] && + hwc->last_tag == box->tags[assign[i]]) + continue; + /* + * Ensure we don't accidentally enable a stopped + * counter simply because we rescheduled. + */ + if (hwc->state & PERF_HES_STOPPED) + hwc->state |= PERF_HES_ARCH; + + uncore_pmu_event_stop(event, PERF_EF_UPDATE); + } + + /* reprogram moved events into new counters */ + for (i = 0; i < n; i++) { + event = box->event_list[i]; + hwc = &event->hw; + + if (hwc->idx != assign[i] || + hwc->last_tag != box->tags[assign[i]]) + uncore_assign_hw_event(box, event, assign[i]); + else if (i < box->n_events) + continue; + + if (hwc->state & PERF_HES_ARCH) + continue; + + uncore_pmu_event_start(event, 0); + } + box->n_events = n; + + return 0; +} + +static void uncore_pmu_event_del(struct perf_event *event, int flags) +{ + struct intel_uncore_box *box = uncore_event_to_box(event); + int i; + + uncore_pmu_event_stop(event, PERF_EF_UPDATE); + + for (i = 0; i < box->n_events; i++) { + if (event == box->event_list[i]) { + uncore_put_event_constraint(box, event); + + while (++i < box->n_events) + box->event_list[i - 1] = box->event_list[i]; + + --box->n_events; + break; + } + } + + event->hw.idx = -1; + event->hw.last_tag = ~0ULL; +} + +static void uncore_pmu_event_read(struct perf_event *event) +{ + struct intel_uncore_box *box = uncore_event_to_box(event); + uncore_perf_event_update(box, event); +} + +/* + * validation ensures the group can be loaded onto the + * PMU if it was the only group available. + */ +static int uncore_validate_group(struct intel_uncore_pmu *pmu, + struct perf_event *event) +{ + struct perf_event *leader = event->group_leader; + struct intel_uncore_box *fake_box; + int ret = -EINVAL, n; + + fake_box = uncore_alloc_box(pmu->type, NUMA_NO_NODE); + if (!fake_box) + return -ENOMEM; + + fake_box->pmu = pmu; + /* + * the event is not yet connected with its + * siblings therefore we must first collect + * existing siblings, then add the new event + * before we can simulate the scheduling + */ + n = uncore_collect_events(fake_box, leader, true); + if (n < 0) + goto out; + + fake_box->n_events = n; + n = uncore_collect_events(fake_box, event, false); + if (n < 0) + goto out; + + fake_box->n_events = n; + + ret = uncore_assign_events(fake_box, NULL, n); +out: + kfree(fake_box); + return ret; +} + +static int uncore_pmu_event_init(struct perf_event *event) +{ + struct intel_uncore_pmu *pmu; + struct intel_uncore_box *box; + struct hw_perf_event *hwc = &event->hw; + int ret; + + if (event->attr.type != event->pmu->type) + return -ENOENT; + + pmu = uncore_event_to_pmu(event); + /* no device found for this pmu */ + if (pmu->func_id < 0) + return -ENOENT; + + /* + * Uncore PMU does measure at all privilege level all the time. + * So it doesn't make sense to specify any exclude bits. + */ + if (event->attr.exclude_user || event->attr.exclude_kernel || + event->attr.exclude_hv || event->attr.exclude_idle) + return -EINVAL; + + /* Sampling not supported yet */ + if (hwc->sample_period) + return -EINVAL; + + /* + * Place all uncore events for a particular physical package + * onto a single cpu + */ + if (event->cpu < 0) + return -EINVAL; + box = uncore_pmu_to_box(pmu, event->cpu); + if (!box || box->cpu < 0) + return -EINVAL; + event->cpu = box->cpu; + + event->hw.idx = -1; + event->hw.last_tag = ~0ULL; + event->hw.extra_reg.idx = EXTRA_REG_NONE; + event->hw.branch_reg.idx = EXTRA_REG_NONE; + + if (event->attr.config == UNCORE_FIXED_EVENT) { + /* no fixed counter */ + if (!pmu->type->fixed_ctl) + return -EINVAL; + /* + * if there is only one fixed counter, only the first pmu + * can access the fixed counter + */ + if (pmu->type->single_fixed && pmu->pmu_idx > 0) + return -EINVAL; + + /* fixed counters have event field hardcoded to zero */ + hwc->config = 0ULL; + } else { + hwc->config = event->attr.config & pmu->type->event_mask; + if (pmu->type->ops->hw_config) { + ret = pmu->type->ops->hw_config(box, event); + if (ret) + return ret; + } + } + + if (event->group_leader != event) + ret = uncore_validate_group(pmu, event); + else + ret = 0; + + return ret; +} + +static ssize_t uncore_get_attr_cpumask(struct device *dev, + struct device_attribute *attr, char *buf) +{ + int n = cpulist_scnprintf(buf, PAGE_SIZE - 2, &uncore_cpu_mask); + + buf[n++] = '\n'; + buf[n] = '\0'; + return n; +} + +static DEVICE_ATTR(cpumask, S_IRUGO, uncore_get_attr_cpumask, NULL); + +static struct attribute *uncore_pmu_attrs[] = { + &dev_attr_cpumask.attr, + NULL, +}; + +static struct attribute_group uncore_pmu_attr_group = { + .attrs = uncore_pmu_attrs, +}; + +static int __init uncore_pmu_register(struct intel_uncore_pmu *pmu) +{ + int ret; + + if (!pmu->type->pmu) { + pmu->pmu = (struct pmu) { + .attr_groups = pmu->type->attr_groups, + .task_ctx_nr = perf_invalid_context, + .event_init = uncore_pmu_event_init, + .add = uncore_pmu_event_add, + .del = uncore_pmu_event_del, + .start = uncore_pmu_event_start, + .stop = uncore_pmu_event_stop, + .read = uncore_pmu_event_read, + }; + } else { + pmu->pmu = *pmu->type->pmu; + pmu->pmu.attr_groups = pmu->type->attr_groups; + } + + if (pmu->type->num_boxes == 1) { + if (strlen(pmu->type->name) > 0) + sprintf(pmu->name, "uncore_%s", pmu->type->name); + else + sprintf(pmu->name, "uncore"); + } else { + sprintf(pmu->name, "uncore_%s_%d", pmu->type->name, + pmu->pmu_idx); + } + + ret = perf_pmu_register(&pmu->pmu, pmu->name, -1); + return ret; +} + +static void __init uncore_type_exit(struct intel_uncore_type *type) +{ + int i; + + for (i = 0; i < type->num_boxes; i++) + free_percpu(type->pmus[i].box); + kfree(type->pmus); + type->pmus = NULL; + kfree(type->events_group); + type->events_group = NULL; +} + +static void __init uncore_types_exit(struct intel_uncore_type **types) +{ + int i; + for (i = 0; types[i]; i++) + uncore_type_exit(types[i]); +} + +static int __init uncore_type_init(struct intel_uncore_type *type) +{ + struct intel_uncore_pmu *pmus; + struct attribute_group *attr_group; + struct attribute **attrs; + int i, j; + + pmus = kzalloc(sizeof(*pmus) * type->num_boxes, GFP_KERNEL); + if (!pmus) + return -ENOMEM; + + type->pmus = pmus; + + type->unconstrainted = (struct event_constraint) + __EVENT_CONSTRAINT(0, (1ULL << type->num_counters) - 1, + 0, type->num_counters, 0, 0); + + for (i = 0; i < type->num_boxes; i++) { + pmus[i].func_id = -1; + pmus[i].pmu_idx = i; + pmus[i].type = type; + INIT_LIST_HEAD(&pmus[i].box_list); + pmus[i].box = alloc_percpu(struct intel_uncore_box *); + if (!pmus[i].box) + goto fail; + } + + if (type->event_descs) { + i = 0; + while (type->event_descs[i].attr.attr.name) + i++; + + attr_group = kzalloc(sizeof(struct attribute *) * (i + 1) + + sizeof(*attr_group), GFP_KERNEL); + if (!attr_group) + goto fail; + + attrs = (struct attribute **)(attr_group + 1); + attr_group->name = "events"; + attr_group->attrs = attrs; + + for (j = 0; j < i; j++) + attrs[j] = &type->event_descs[j].attr.attr; + + type->events_group = attr_group; + } + + type->pmu_group = &uncore_pmu_attr_group; + return 0; +fail: + uncore_type_exit(type); + return -ENOMEM; +} + +static int __init uncore_types_init(struct intel_uncore_type **types) +{ + int i, ret; + + for (i = 0; types[i]; i++) { + ret = uncore_type_init(types[i]); + if (ret) + goto fail; + } + return 0; +fail: + while (--i >= 0) + uncore_type_exit(types[i]); + return ret; +} + +static struct pci_driver *uncore_pci_driver; +static bool pcidrv_registered; + +/* + * add a pci uncore device + */ +static int uncore_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id) +{ + struct intel_uncore_pmu *pmu; + struct intel_uncore_box *box; + struct intel_uncore_type *type; + int phys_id; + + phys_id = pcibus_to_physid[pdev->bus->number]; + if (phys_id < 0) + return -ENODEV; + + if (UNCORE_PCI_DEV_TYPE(id->driver_data) == UNCORE_EXTRA_PCI_DEV) { + extra_pci_dev[phys_id][UNCORE_PCI_DEV_IDX(id->driver_data)] = pdev; + pci_set_drvdata(pdev, NULL); + return 0; + } + + type = pci_uncores[UNCORE_PCI_DEV_TYPE(id->driver_data)]; + box = uncore_alloc_box(type, NUMA_NO_NODE); + if (!box) + return -ENOMEM; + + /* + * for performance monitoring unit with multiple boxes, + * each box has a different function id. + */ + pmu = &type->pmus[UNCORE_PCI_DEV_IDX(id->driver_data)]; + if (pmu->func_id < 0) + pmu->func_id = pdev->devfn; + else + WARN_ON_ONCE(pmu->func_id != pdev->devfn); + + box->phys_id = phys_id; + box->pci_dev = pdev; + box->pmu = pmu; + uncore_box_init(box); + pci_set_drvdata(pdev, box); + + raw_spin_lock(&uncore_box_lock); + list_add_tail(&box->list, &pmu->box_list); + raw_spin_unlock(&uncore_box_lock); + + return 0; +} + +static void uncore_pci_remove(struct pci_dev *pdev) +{ + struct intel_uncore_box *box = pci_get_drvdata(pdev); + struct intel_uncore_pmu *pmu; + int i, cpu, phys_id = pcibus_to_physid[pdev->bus->number]; + + box = pci_get_drvdata(pdev); + if (!box) { + for (i = 0; i < UNCORE_EXTRA_PCI_DEV_MAX; i++) { + if (extra_pci_dev[phys_id][i] == pdev) { + extra_pci_dev[phys_id][i] = NULL; + break; + } + } + WARN_ON_ONCE(i >= UNCORE_EXTRA_PCI_DEV_MAX); + return; + } + + pmu = box->pmu; + if (WARN_ON_ONCE(phys_id != box->phys_id)) + return; + + pci_set_drvdata(pdev, NULL); + + raw_spin_lock(&uncore_box_lock); + list_del(&box->list); + raw_spin_unlock(&uncore_box_lock); + + for_each_possible_cpu(cpu) { + if (*per_cpu_ptr(pmu->box, cpu) == box) { + *per_cpu_ptr(pmu->box, cpu) = NULL; + atomic_dec(&box->refcnt); + } + } + + WARN_ON_ONCE(atomic_read(&box->refcnt) != 1); + kfree(box); +} + +static int __init uncore_pci_init(void) +{ + int ret; + + switch (boot_cpu_data.x86_model) { + case 45: /* Sandy Bridge-EP */ + ret = snbep_pci2phy_map_init(0x3ce0); + if (ret) + return ret; + pci_uncores = snbep_pci_uncores; + uncore_pci_driver = &snbep_uncore_pci_driver; + break; + case 62: /* IvyTown */ + ret = snbep_pci2phy_map_init(0x0e1e); + if (ret) + return ret; + pci_uncores = ivt_pci_uncores; + uncore_pci_driver = &ivt_uncore_pci_driver; + break; + case 42: /* Sandy Bridge */ + ret = snb_pci2phy_map_init(PCI_DEVICE_ID_INTEL_SNB_IMC); + if (ret) + return ret; + pci_uncores = snb_pci_uncores; + uncore_pci_driver = &snb_uncore_pci_driver; + break; + case 58: /* Ivy Bridge */ + ret = snb_pci2phy_map_init(PCI_DEVICE_ID_INTEL_IVB_IMC); + if (ret) + return ret; + pci_uncores = snb_pci_uncores; + uncore_pci_driver = &ivb_uncore_pci_driver; + break; + case 60: /* Haswell */ + case 69: /* Haswell Celeron */ + ret = snb_pci2phy_map_init(PCI_DEVICE_ID_INTEL_HSW_IMC); + if (ret) + return ret; + pci_uncores = snb_pci_uncores; + uncore_pci_driver = &hsw_uncore_pci_driver; + break; + default: + return 0; + } + + ret = uncore_types_init(pci_uncores); + if (ret) + return ret; + + uncore_pci_driver->probe = uncore_pci_probe; + uncore_pci_driver->remove = uncore_pci_remove; + + ret = pci_register_driver(uncore_pci_driver); + if (ret == 0) + pcidrv_registered = true; + else + uncore_types_exit(pci_uncores); + + return ret; +} + +static void __init uncore_pci_exit(void) +{ + if (pcidrv_registered) { + pcidrv_registered = false; + pci_unregister_driver(uncore_pci_driver); + uncore_types_exit(pci_uncores); + } +} + +/* CPU hot plug/unplug are serialized by cpu_add_remove_lock mutex */ +static LIST_HEAD(boxes_to_free); + +static void uncore_kfree_boxes(void) +{ + struct intel_uncore_box *box; + + while (!list_empty(&boxes_to_free)) { + box = list_entry(boxes_to_free.next, + struct intel_uncore_box, list); + list_del(&box->list); + kfree(box); + } +} + +static void uncore_cpu_dying(int cpu) +{ + struct intel_uncore_type *type; + struct intel_uncore_pmu *pmu; + struct intel_uncore_box *box; + int i, j; + + for (i = 0; msr_uncores[i]; i++) { + type = msr_uncores[i]; + for (j = 0; j < type->num_boxes; j++) { + pmu = &type->pmus[j]; + box = *per_cpu_ptr(pmu->box, cpu); + *per_cpu_ptr(pmu->box, cpu) = NULL; + if (box && atomic_dec_and_test(&box->refcnt)) + list_add(&box->list, &boxes_to_free); + } + } +} + +static int uncore_cpu_starting(int cpu) +{ + struct intel_uncore_type *type; + struct intel_uncore_pmu *pmu; + struct intel_uncore_box *box, *exist; + int i, j, k, phys_id; + + phys_id = topology_physical_package_id(cpu); + + for (i = 0; msr_uncores[i]; i++) { + type = msr_uncores[i]; + for (j = 0; j < type->num_boxes; j++) { + pmu = &type->pmus[j]; + box = *per_cpu_ptr(pmu->box, cpu); + /* called by uncore_cpu_init? */ + if (box && box->phys_id >= 0) { + uncore_box_init(box); + continue; + } + + for_each_online_cpu(k) { + exist = *per_cpu_ptr(pmu->box, k); + if (exist && exist->phys_id == phys_id) { + atomic_inc(&exist->refcnt); + *per_cpu_ptr(pmu->box, cpu) = exist; + if (box) { + list_add(&box->list, + &boxes_to_free); + box = NULL; + } + break; + } + } + + if (box) { + box->phys_id = phys_id; + uncore_box_init(box); + } + } + } + return 0; +} + +static int uncore_cpu_prepare(int cpu, int phys_id) +{ + struct intel_uncore_type *type; + struct intel_uncore_pmu *pmu; + struct intel_uncore_box *box; + int i, j; + + for (i = 0; msr_uncores[i]; i++) { + type = msr_uncores[i]; + for (j = 0; j < type->num_boxes; j++) { + pmu = &type->pmus[j]; + if (pmu->func_id < 0) + pmu->func_id = j; + + box = uncore_alloc_box(type, cpu_to_node(cpu)); + if (!box) + return -ENOMEM; + + box->pmu = pmu; + box->phys_id = phys_id; + *per_cpu_ptr(pmu->box, cpu) = box; + } + } + return 0; +} + +static void +uncore_change_context(struct intel_uncore_type **uncores, int old_cpu, int new_cpu) +{ + struct intel_uncore_type *type; + struct intel_uncore_pmu *pmu; + struct intel_uncore_box *box; + int i, j; + + for (i = 0; uncores[i]; i++) { + type = uncores[i]; + for (j = 0; j < type->num_boxes; j++) { + pmu = &type->pmus[j]; + if (old_cpu < 0) + box = uncore_pmu_to_box(pmu, new_cpu); + else + box = uncore_pmu_to_box(pmu, old_cpu); + if (!box) + continue; + + if (old_cpu < 0) { + WARN_ON_ONCE(box->cpu != -1); + box->cpu = new_cpu; + continue; + } + + WARN_ON_ONCE(box->cpu != old_cpu); + if (new_cpu >= 0) { + uncore_pmu_cancel_hrtimer(box); + perf_pmu_migrate_context(&pmu->pmu, + old_cpu, new_cpu); + box->cpu = new_cpu; + } else { + box->cpu = -1; + } + } + } +} + +static void uncore_event_exit_cpu(int cpu) +{ + int i, phys_id, target; + + /* if exiting cpu is used for collecting uncore events */ + if (!cpumask_test_and_clear_cpu(cpu, &uncore_cpu_mask)) + return; + + /* find a new cpu to collect uncore events */ + phys_id = topology_physical_package_id(cpu); + target = -1; + for_each_online_cpu(i) { + if (i == cpu) + continue; + if (phys_id == topology_physical_package_id(i)) { + target = i; + break; + } + } + + /* migrate uncore events to the new cpu */ + if (target >= 0) + cpumask_set_cpu(target, &uncore_cpu_mask); + + uncore_change_context(msr_uncores, cpu, target); + uncore_change_context(pci_uncores, cpu, target); +} + +static void uncore_event_init_cpu(int cpu) +{ + int i, phys_id; + + phys_id = topology_physical_package_id(cpu); + for_each_cpu(i, &uncore_cpu_mask) { + if (phys_id == topology_physical_package_id(i)) + return; + } + + cpumask_set_cpu(cpu, &uncore_cpu_mask); + + uncore_change_context(msr_uncores, -1, cpu); + uncore_change_context(pci_uncores, -1, cpu); +} + +static int uncore_cpu_notifier(struct notifier_block *self, + unsigned long action, void *hcpu) +{ + unsigned int cpu = (long)hcpu; + + /* allocate/free data structure for uncore box */ + switch (action & ~CPU_TASKS_FROZEN) { + case CPU_UP_PREPARE: + uncore_cpu_prepare(cpu, -1); + break; + case CPU_STARTING: + uncore_cpu_starting(cpu); + break; + case CPU_UP_CANCELED: + case CPU_DYING: + uncore_cpu_dying(cpu); + break; + case CPU_ONLINE: + case CPU_DEAD: + uncore_kfree_boxes(); + break; + default: + break; + } + + /* select the cpu that collects uncore events */ + switch (action & ~CPU_TASKS_FROZEN) { + case CPU_DOWN_FAILED: + case CPU_STARTING: + uncore_event_init_cpu(cpu); + break; + case CPU_DOWN_PREPARE: + uncore_event_exit_cpu(cpu); + break; + default: + break; + } + + return NOTIFY_OK; +} + +static struct notifier_block uncore_cpu_nb = { + .notifier_call = uncore_cpu_notifier, + /* + * to migrate uncore events, our notifier should be executed + * before perf core's notifier. + */ + .priority = CPU_PRI_PERF + 1, +}; + +static void __init uncore_cpu_setup(void *dummy) +{ + uncore_cpu_starting(smp_processor_id()); +} + +static int __init uncore_cpu_init(void) +{ + int ret, max_cores; + + max_cores = boot_cpu_data.x86_max_cores; + switch (boot_cpu_data.x86_model) { + case 26: /* Nehalem */ + case 30: + case 37: /* Westmere */ + case 44: + msr_uncores = nhm_msr_uncores; + break; + case 42: /* Sandy Bridge */ + case 58: /* Ivy Bridge */ + if (snb_uncore_cbox.num_boxes > max_cores) + snb_uncore_cbox.num_boxes = max_cores; + msr_uncores = snb_msr_uncores; + break; + case 45: /* Sandy Bridge-EP */ + if (snbep_uncore_cbox.num_boxes > max_cores) + snbep_uncore_cbox.num_boxes = max_cores; + msr_uncores = snbep_msr_uncores; + break; + case 46: /* Nehalem-EX */ + uncore_nhmex = true; + case 47: /* Westmere-EX aka. Xeon E7 */ + if (!uncore_nhmex) + nhmex_uncore_mbox.event_descs = wsmex_uncore_mbox_events; + if (nhmex_uncore_cbox.num_boxes > max_cores) + nhmex_uncore_cbox.num_boxes = max_cores; + msr_uncores = nhmex_msr_uncores; + break; + case 62: /* IvyTown */ + if (ivt_uncore_cbox.num_boxes > max_cores) + ivt_uncore_cbox.num_boxes = max_cores; + msr_uncores = ivt_msr_uncores; + break; + + default: + return 0; + } + + ret = uncore_types_init(msr_uncores); + if (ret) + return ret; + + return 0; +} + +static int __init uncore_pmus_register(void) +{ + struct intel_uncore_pmu *pmu; + struct intel_uncore_type *type; + int i, j; + + for (i = 0; msr_uncores[i]; i++) { + type = msr_uncores[i]; + for (j = 0; j < type->num_boxes; j++) { + pmu = &type->pmus[j]; + uncore_pmu_register(pmu); + } + } + + for (i = 0; pci_uncores[i]; i++) { + type = pci_uncores[i]; + for (j = 0; j < type->num_boxes; j++) { + pmu = &type->pmus[j]; + uncore_pmu_register(pmu); + } + } + + return 0; +} + +static void __init uncore_cpumask_init(void) +{ + int cpu; + + /* + * ony invoke once from msr or pci init code + */ + if (!cpumask_empty(&uncore_cpu_mask)) + return; + + cpu_notifier_register_begin(); + + for_each_online_cpu(cpu) { + int i, phys_id = topology_physical_package_id(cpu); + + for_each_cpu(i, &uncore_cpu_mask) { + if (phys_id == topology_physical_package_id(i)) { + phys_id = -1; + break; + } + } + if (phys_id < 0) + continue; + + uncore_cpu_prepare(cpu, phys_id); + uncore_event_init_cpu(cpu); + } + on_each_cpu(uncore_cpu_setup, NULL, 1); + + __register_cpu_notifier(&uncore_cpu_nb); + + cpu_notifier_register_done(); +} + + +static int __init intel_uncore_init(void) +{ + int ret; + + if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL) + return -ENODEV; + + if (cpu_has_hypervisor) + return -ENODEV; + + ret = uncore_pci_init(); + if (ret) + goto fail; + ret = uncore_cpu_init(); + if (ret) { + uncore_pci_exit(); + goto fail; + } + uncore_cpumask_init(); + + uncore_pmus_register(); + return 0; +fail: + return ret; +} +device_initcall(intel_uncore_init); diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore.h b/arch/x86/kernel/cpu/perf_event_intel_uncore.h new file mode 100644 index 00000000000..90236f0c94a --- /dev/null +++ b/arch/x86/kernel/cpu/perf_event_intel_uncore.h @@ -0,0 +1,696 @@ +#include <linux/module.h> +#include <linux/slab.h> +#include <linux/pci.h> +#include <linux/perf_event.h> +#include "perf_event.h" + +#define UNCORE_PMU_NAME_LEN 32 +#define UNCORE_PMU_HRTIMER_INTERVAL (60LL * NSEC_PER_SEC) +#define UNCORE_SNB_IMC_HRTIMER_INTERVAL (5ULL * NSEC_PER_SEC) + +#define UNCORE_FIXED_EVENT 0xff +#define UNCORE_PMC_IDX_MAX_GENERIC 8 +#define UNCORE_PMC_IDX_FIXED UNCORE_PMC_IDX_MAX_GENERIC +#define UNCORE_PMC_IDX_MAX (UNCORE_PMC_IDX_FIXED + 1) + +#define UNCORE_PCI_DEV_DATA(type, idx) ((type << 8) | idx) +#define UNCORE_PCI_DEV_TYPE(data) ((data >> 8) & 0xff) +#define UNCORE_PCI_DEV_IDX(data) (data & 0xff) +#define UNCORE_EXTRA_PCI_DEV 0xff +#define UNCORE_EXTRA_PCI_DEV_MAX 2 + +/* support up to 8 sockets */ +#define UNCORE_SOCKET_MAX 8 + +#define UNCORE_EVENT_CONSTRAINT(c, n) EVENT_CONSTRAINT(c, n, 0xff) + +/* SNB event control */ +#define SNB_UNC_CTL_EV_SEL_MASK 0x000000ff +#define SNB_UNC_CTL_UMASK_MASK 0x0000ff00 +#define SNB_UNC_CTL_EDGE_DET (1 << 18) +#define SNB_UNC_CTL_EN (1 << 22) +#define SNB_UNC_CTL_INVERT (1 << 23) +#define SNB_UNC_CTL_CMASK_MASK 0x1f000000 +#define NHM_UNC_CTL_CMASK_MASK 0xff000000 +#define NHM_UNC_FIXED_CTR_CTL_EN (1 << 0) + +#define SNB_UNC_RAW_EVENT_MASK (SNB_UNC_CTL_EV_SEL_MASK | \ + SNB_UNC_CTL_UMASK_MASK | \ + SNB_UNC_CTL_EDGE_DET | \ + SNB_UNC_CTL_INVERT | \ + SNB_UNC_CTL_CMASK_MASK) + +#define NHM_UNC_RAW_EVENT_MASK (SNB_UNC_CTL_EV_SEL_MASK | \ + SNB_UNC_CTL_UMASK_MASK | \ + SNB_UNC_CTL_EDGE_DET | \ + SNB_UNC_CTL_INVERT | \ + NHM_UNC_CTL_CMASK_MASK) + +/* SNB global control register */ +#define SNB_UNC_PERF_GLOBAL_CTL 0x391 +#define SNB_UNC_FIXED_CTR_CTRL 0x394 +#define SNB_UNC_FIXED_CTR 0x395 + +/* SNB uncore global control */ +#define SNB_UNC_GLOBAL_CTL_CORE_ALL ((1 << 4) - 1) +#define SNB_UNC_GLOBAL_CTL_EN (1 << 29) + +/* SNB Cbo register */ +#define SNB_UNC_CBO_0_PERFEVTSEL0 0x700 +#define SNB_UNC_CBO_0_PER_CTR0 0x706 +#define SNB_UNC_CBO_MSR_OFFSET 0x10 + +/* NHM global control register */ +#define NHM_UNC_PERF_GLOBAL_CTL 0x391 +#define NHM_UNC_FIXED_CTR 0x394 +#define NHM_UNC_FIXED_CTR_CTRL 0x395 + +/* NHM uncore global control */ +#define NHM_UNC_GLOBAL_CTL_EN_PC_ALL ((1ULL << 8) - 1) +#define NHM_UNC_GLOBAL_CTL_EN_FC (1ULL << 32) + +/* NHM uncore register */ +#define NHM_UNC_PERFEVTSEL0 0x3c0 +#define NHM_UNC_UNCORE_PMC0 0x3b0 + +/* SNB-EP Box level control */ +#define SNBEP_PMON_BOX_CTL_RST_CTRL (1 << 0) +#define SNBEP_PMON_BOX_CTL_RST_CTRS (1 << 1) +#define SNBEP_PMON_BOX_CTL_FRZ (1 << 8) +#define SNBEP_PMON_BOX_CTL_FRZ_EN (1 << 16) +#define SNBEP_PMON_BOX_CTL_INT (SNBEP_PMON_BOX_CTL_RST_CTRL | \ + SNBEP_PMON_BOX_CTL_RST_CTRS | \ + SNBEP_PMON_BOX_CTL_FRZ_EN) +/* SNB-EP event control */ +#define SNBEP_PMON_CTL_EV_SEL_MASK 0x000000ff +#define SNBEP_PMON_CTL_UMASK_MASK 0x0000ff00 +#define SNBEP_PMON_CTL_RST (1 << 17) +#define SNBEP_PMON_CTL_EDGE_DET (1 << 18) +#define SNBEP_PMON_CTL_EV_SEL_EXT (1 << 21) +#define SNBEP_PMON_CTL_EN (1 << 22) +#define SNBEP_PMON_CTL_INVERT (1 << 23) +#define SNBEP_PMON_CTL_TRESH_MASK 0xff000000 +#define SNBEP_PMON_RAW_EVENT_MASK (SNBEP_PMON_CTL_EV_SEL_MASK | \ + SNBEP_PMON_CTL_UMASK_MASK | \ + SNBEP_PMON_CTL_EDGE_DET | \ + SNBEP_PMON_CTL_INVERT | \ + SNBEP_PMON_CTL_TRESH_MASK) + +/* SNB-EP Ubox event control */ +#define SNBEP_U_MSR_PMON_CTL_TRESH_MASK 0x1f000000 +#define SNBEP_U_MSR_PMON_RAW_EVENT_MASK \ + (SNBEP_PMON_CTL_EV_SEL_MASK | \ + SNBEP_PMON_CTL_UMASK_MASK | \ + SNBEP_PMON_CTL_EDGE_DET | \ + SNBEP_PMON_CTL_INVERT | \ + SNBEP_U_MSR_PMON_CTL_TRESH_MASK) + +#define SNBEP_CBO_PMON_CTL_TID_EN (1 << 19) +#define SNBEP_CBO_MSR_PMON_RAW_EVENT_MASK (SNBEP_PMON_RAW_EVENT_MASK | \ + SNBEP_CBO_PMON_CTL_TID_EN) + +/* SNB-EP PCU event control */ +#define SNBEP_PCU_MSR_PMON_CTL_OCC_SEL_MASK 0x0000c000 +#define SNBEP_PCU_MSR_PMON_CTL_TRESH_MASK 0x1f000000 +#define SNBEP_PCU_MSR_PMON_CTL_OCC_INVERT (1 << 30) +#define SNBEP_PCU_MSR_PMON_CTL_OCC_EDGE_DET (1 << 31) +#define SNBEP_PCU_MSR_PMON_RAW_EVENT_MASK \ + (SNBEP_PMON_CTL_EV_SEL_MASK | \ + SNBEP_PCU_MSR_PMON_CTL_OCC_SEL_MASK | \ + SNBEP_PMON_CTL_EDGE_DET | \ + SNBEP_PMON_CTL_EV_SEL_EXT | \ + SNBEP_PMON_CTL_INVERT | \ + SNBEP_PCU_MSR_PMON_CTL_TRESH_MASK | \ + SNBEP_PCU_MSR_PMON_CTL_OCC_INVERT | \ + SNBEP_PCU_MSR_PMON_CTL_OCC_EDGE_DET) + +#define SNBEP_QPI_PCI_PMON_RAW_EVENT_MASK \ + (SNBEP_PMON_RAW_EVENT_MASK | \ + SNBEP_PMON_CTL_EV_SEL_EXT) + +/* SNB-EP pci control register */ +#define SNBEP_PCI_PMON_BOX_CTL 0xf4 +#define SNBEP_PCI_PMON_CTL0 0xd8 +/* SNB-EP pci counter register */ +#define SNBEP_PCI_PMON_CTR0 0xa0 + +/* SNB-EP home agent register */ +#define SNBEP_HA_PCI_PMON_BOX_ADDRMATCH0 0x40 +#define SNBEP_HA_PCI_PMON_BOX_ADDRMATCH1 0x44 +#define SNBEP_HA_PCI_PMON_BOX_OPCODEMATCH 0x48 +/* SNB-EP memory controller register */ +#define SNBEP_MC_CHy_PCI_PMON_FIXED_CTL 0xf0 +#define SNBEP_MC_CHy_PCI_PMON_FIXED_CTR 0xd0 +/* SNB-EP QPI register */ +#define SNBEP_Q_Py_PCI_PMON_PKT_MATCH0 0x228 +#define SNBEP_Q_Py_PCI_PMON_PKT_MATCH1 0x22c +#define SNBEP_Q_Py_PCI_PMON_PKT_MASK0 0x238 +#define SNBEP_Q_Py_PCI_PMON_PKT_MASK1 0x23c + +/* SNB-EP Ubox register */ +#define SNBEP_U_MSR_PMON_CTR0 0xc16 +#define SNBEP_U_MSR_PMON_CTL0 0xc10 + +#define SNBEP_U_MSR_PMON_UCLK_FIXED_CTL 0xc08 +#define SNBEP_U_MSR_PMON_UCLK_FIXED_CTR 0xc09 + +/* SNB-EP Cbo register */ +#define SNBEP_C0_MSR_PMON_CTR0 0xd16 +#define SNBEP_C0_MSR_PMON_CTL0 0xd10 +#define SNBEP_C0_MSR_PMON_BOX_CTL 0xd04 +#define SNBEP_C0_MSR_PMON_BOX_FILTER 0xd14 +#define SNBEP_CBO_MSR_OFFSET 0x20 + +#define SNBEP_CB0_MSR_PMON_BOX_FILTER_TID 0x1f +#define SNBEP_CB0_MSR_PMON_BOX_FILTER_NID 0x3fc00 +#define SNBEP_CB0_MSR_PMON_BOX_FILTER_STATE 0x7c0000 +#define SNBEP_CB0_MSR_PMON_BOX_FILTER_OPC 0xff800000 + +#define SNBEP_CBO_EVENT_EXTRA_REG(e, m, i) { \ + .event = (e), \ + .msr = SNBEP_C0_MSR_PMON_BOX_FILTER, \ + .config_mask = (m), \ + .idx = (i) \ +} + +/* SNB-EP PCU register */ +#define SNBEP_PCU_MSR_PMON_CTR0 0xc36 +#define SNBEP_PCU_MSR_PMON_CTL0 0xc30 +#define SNBEP_PCU_MSR_PMON_BOX_CTL 0xc24 +#define SNBEP_PCU_MSR_PMON_BOX_FILTER 0xc34 +#define SNBEP_PCU_MSR_PMON_BOX_FILTER_MASK 0xffffffff +#define SNBEP_PCU_MSR_CORE_C3_CTR 0x3fc +#define SNBEP_PCU_MSR_CORE_C6_CTR 0x3fd + +/* IVT event control */ +#define IVT_PMON_BOX_CTL_INT (SNBEP_PMON_BOX_CTL_RST_CTRL | \ + SNBEP_PMON_BOX_CTL_RST_CTRS) +#define IVT_PMON_RAW_EVENT_MASK (SNBEP_PMON_CTL_EV_SEL_MASK | \ + SNBEP_PMON_CTL_UMASK_MASK | \ + SNBEP_PMON_CTL_EDGE_DET | \ + SNBEP_PMON_CTL_TRESH_MASK) +/* IVT Ubox */ +#define IVT_U_MSR_PMON_GLOBAL_CTL 0xc00 +#define IVT_U_PMON_GLOBAL_FRZ_ALL (1 << 31) +#define IVT_U_PMON_GLOBAL_UNFRZ_ALL (1 << 29) + +#define IVT_U_MSR_PMON_RAW_EVENT_MASK \ + (SNBEP_PMON_CTL_EV_SEL_MASK | \ + SNBEP_PMON_CTL_UMASK_MASK | \ + SNBEP_PMON_CTL_EDGE_DET | \ + SNBEP_U_MSR_PMON_CTL_TRESH_MASK) +/* IVT Cbo */ +#define IVT_CBO_MSR_PMON_RAW_EVENT_MASK (IVT_PMON_RAW_EVENT_MASK | \ + SNBEP_CBO_PMON_CTL_TID_EN) + +#define IVT_CB0_MSR_PMON_BOX_FILTER_TID (0x1fULL << 0) +#define IVT_CB0_MSR_PMON_BOX_FILTER_LINK (0xfULL << 5) +#define IVT_CB0_MSR_PMON_BOX_FILTER_STATE (0x3fULL << 17) +#define IVT_CB0_MSR_PMON_BOX_FILTER_NID (0xffffULL << 32) +#define IVT_CB0_MSR_PMON_BOX_FILTER_OPC (0x1ffULL << 52) +#define IVT_CB0_MSR_PMON_BOX_FILTER_C6 (0x1ULL << 61) +#define IVT_CB0_MSR_PMON_BOX_FILTER_NC (0x1ULL << 62) +#define IVT_CB0_MSR_PMON_BOX_FILTER_IOSC (0x1ULL << 63) + +/* IVT home agent */ +#define IVT_HA_PCI_PMON_CTL_Q_OCC_RST (1 << 16) +#define IVT_HA_PCI_PMON_RAW_EVENT_MASK \ + (IVT_PMON_RAW_EVENT_MASK | \ + IVT_HA_PCI_PMON_CTL_Q_OCC_RST) +/* IVT PCU */ +#define IVT_PCU_MSR_PMON_RAW_EVENT_MASK \ + (SNBEP_PMON_CTL_EV_SEL_MASK | \ + SNBEP_PMON_CTL_EV_SEL_EXT | \ + SNBEP_PCU_MSR_PMON_CTL_OCC_SEL_MASK | \ + SNBEP_PMON_CTL_EDGE_DET | \ + SNBEP_PCU_MSR_PMON_CTL_TRESH_MASK | \ + SNBEP_PCU_MSR_PMON_CTL_OCC_INVERT | \ + SNBEP_PCU_MSR_PMON_CTL_OCC_EDGE_DET) +/* IVT QPI */ +#define IVT_QPI_PCI_PMON_RAW_EVENT_MASK \ + (IVT_PMON_RAW_EVENT_MASK | \ + SNBEP_PMON_CTL_EV_SEL_EXT) + +/* NHM-EX event control */ +#define NHMEX_PMON_CTL_EV_SEL_MASK 0x000000ff +#define NHMEX_PMON_CTL_UMASK_MASK 0x0000ff00 +#define NHMEX_PMON_CTL_EN_BIT0 (1 << 0) +#define NHMEX_PMON_CTL_EDGE_DET (1 << 18) +#define NHMEX_PMON_CTL_PMI_EN (1 << 20) +#define NHMEX_PMON_CTL_EN_BIT22 (1 << 22) +#define NHMEX_PMON_CTL_INVERT (1 << 23) +#define NHMEX_PMON_CTL_TRESH_MASK 0xff000000 +#define NHMEX_PMON_RAW_EVENT_MASK (NHMEX_PMON_CTL_EV_SEL_MASK | \ + NHMEX_PMON_CTL_UMASK_MASK | \ + NHMEX_PMON_CTL_EDGE_DET | \ + NHMEX_PMON_CTL_INVERT | \ + NHMEX_PMON_CTL_TRESH_MASK) + +/* NHM-EX Ubox */ +#define NHMEX_U_MSR_PMON_GLOBAL_CTL 0xc00 +#define NHMEX_U_MSR_PMON_CTR 0xc11 +#define NHMEX_U_MSR_PMON_EV_SEL 0xc10 + +#define NHMEX_U_PMON_GLOBAL_EN (1 << 0) +#define NHMEX_U_PMON_GLOBAL_PMI_CORE_SEL 0x0000001e +#define NHMEX_U_PMON_GLOBAL_EN_ALL (1 << 28) +#define NHMEX_U_PMON_GLOBAL_RST_ALL (1 << 29) +#define NHMEX_U_PMON_GLOBAL_FRZ_ALL (1 << 31) + +#define NHMEX_U_PMON_RAW_EVENT_MASK \ + (NHMEX_PMON_CTL_EV_SEL_MASK | \ + NHMEX_PMON_CTL_EDGE_DET) + +/* NHM-EX Cbox */ +#define NHMEX_C0_MSR_PMON_GLOBAL_CTL 0xd00 +#define NHMEX_C0_MSR_PMON_CTR0 0xd11 +#define NHMEX_C0_MSR_PMON_EV_SEL0 0xd10 +#define NHMEX_C_MSR_OFFSET 0x20 + +/* NHM-EX Bbox */ +#define NHMEX_B0_MSR_PMON_GLOBAL_CTL 0xc20 +#define NHMEX_B0_MSR_PMON_CTR0 0xc31 +#define NHMEX_B0_MSR_PMON_CTL0 0xc30 +#define NHMEX_B_MSR_OFFSET 0x40 +#define NHMEX_B0_MSR_MATCH 0xe45 +#define NHMEX_B0_MSR_MASK 0xe46 +#define NHMEX_B1_MSR_MATCH 0xe4d +#define NHMEX_B1_MSR_MASK 0xe4e + +#define NHMEX_B_PMON_CTL_EN (1 << 0) +#define NHMEX_B_PMON_CTL_EV_SEL_SHIFT 1 +#define NHMEX_B_PMON_CTL_EV_SEL_MASK \ + (0x1f << NHMEX_B_PMON_CTL_EV_SEL_SHIFT) +#define NHMEX_B_PMON_CTR_SHIFT 6 +#define NHMEX_B_PMON_CTR_MASK \ + (0x3 << NHMEX_B_PMON_CTR_SHIFT) +#define NHMEX_B_PMON_RAW_EVENT_MASK \ + (NHMEX_B_PMON_CTL_EV_SEL_MASK | \ + NHMEX_B_PMON_CTR_MASK) + +/* NHM-EX Sbox */ +#define NHMEX_S0_MSR_PMON_GLOBAL_CTL 0xc40 +#define NHMEX_S0_MSR_PMON_CTR0 0xc51 +#define NHMEX_S0_MSR_PMON_CTL0 0xc50 +#define NHMEX_S_MSR_OFFSET 0x80 +#define NHMEX_S0_MSR_MM_CFG 0xe48 +#define NHMEX_S0_MSR_MATCH 0xe49 +#define NHMEX_S0_MSR_MASK 0xe4a +#define NHMEX_S1_MSR_MM_CFG 0xe58 +#define NHMEX_S1_MSR_MATCH 0xe59 +#define NHMEX_S1_MSR_MASK 0xe5a + +#define NHMEX_S_PMON_MM_CFG_EN (0x1ULL << 63) +#define NHMEX_S_EVENT_TO_R_PROG_EV 0 + +/* NHM-EX Mbox */ +#define NHMEX_M0_MSR_GLOBAL_CTL 0xca0 +#define NHMEX_M0_MSR_PMU_DSP 0xca5 +#define NHMEX_M0_MSR_PMU_ISS 0xca6 +#define NHMEX_M0_MSR_PMU_MAP 0xca7 +#define NHMEX_M0_MSR_PMU_MSC_THR 0xca8 +#define NHMEX_M0_MSR_PMU_PGT 0xca9 +#define NHMEX_M0_MSR_PMU_PLD 0xcaa +#define NHMEX_M0_MSR_PMU_ZDP_CTL_FVC 0xcab +#define NHMEX_M0_MSR_PMU_CTL0 0xcb0 +#define NHMEX_M0_MSR_PMU_CNT0 0xcb1 +#define NHMEX_M_MSR_OFFSET 0x40 +#define NHMEX_M0_MSR_PMU_MM_CFG 0xe54 +#define NHMEX_M1_MSR_PMU_MM_CFG 0xe5c + +#define NHMEX_M_PMON_MM_CFG_EN (1ULL << 63) +#define NHMEX_M_PMON_ADDR_MATCH_MASK 0x3ffffffffULL +#define NHMEX_M_PMON_ADDR_MASK_MASK 0x7ffffffULL +#define NHMEX_M_PMON_ADDR_MASK_SHIFT 34 + +#define NHMEX_M_PMON_CTL_EN (1 << 0) +#define NHMEX_M_PMON_CTL_PMI_EN (1 << 1) +#define NHMEX_M_PMON_CTL_COUNT_MODE_SHIFT 2 +#define NHMEX_M_PMON_CTL_COUNT_MODE_MASK \ + (0x3 << NHMEX_M_PMON_CTL_COUNT_MODE_SHIFT) +#define NHMEX_M_PMON_CTL_STORAGE_MODE_SHIFT 4 +#define NHMEX_M_PMON_CTL_STORAGE_MODE_MASK \ + (0x3 << NHMEX_M_PMON_CTL_STORAGE_MODE_SHIFT) +#define NHMEX_M_PMON_CTL_WRAP_MODE (1 << 6) +#define NHMEX_M_PMON_CTL_FLAG_MODE (1 << 7) +#define NHMEX_M_PMON_CTL_INC_SEL_SHIFT 9 +#define NHMEX_M_PMON_CTL_INC_SEL_MASK \ + (0x1f << NHMEX_M_PMON_CTL_INC_SEL_SHIFT) +#define NHMEX_M_PMON_CTL_SET_FLAG_SEL_SHIFT 19 +#define NHMEX_M_PMON_CTL_SET_FLAG_SEL_MASK \ + (0x7 << NHMEX_M_PMON_CTL_SET_FLAG_SEL_SHIFT) +#define NHMEX_M_PMON_RAW_EVENT_MASK \ + (NHMEX_M_PMON_CTL_COUNT_MODE_MASK | \ + NHMEX_M_PMON_CTL_STORAGE_MODE_MASK | \ + NHMEX_M_PMON_CTL_WRAP_MODE | \ + NHMEX_M_PMON_CTL_FLAG_MODE | \ + NHMEX_M_PMON_CTL_INC_SEL_MASK | \ + NHMEX_M_PMON_CTL_SET_FLAG_SEL_MASK) + +#define NHMEX_M_PMON_ZDP_CTL_FVC_MASK (((1 << 11) - 1) | (1 << 23)) +#define NHMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(n) (0x7ULL << (11 + 3 * (n))) + +#define WSMEX_M_PMON_ZDP_CTL_FVC_MASK (((1 << 12) - 1) | (1 << 24)) +#define WSMEX_M_PMON_ZDP_CTL_FVC_EVENT_MASK(n) (0x7ULL << (12 + 3 * (n))) + +/* + * use the 9~13 bits to select event If the 7th bit is not set, + * otherwise use the 19~21 bits to select event. + */ +#define MBOX_INC_SEL(x) ((x) << NHMEX_M_PMON_CTL_INC_SEL_SHIFT) +#define MBOX_SET_FLAG_SEL(x) (((x) << NHMEX_M_PMON_CTL_SET_FLAG_SEL_SHIFT) | \ + NHMEX_M_PMON_CTL_FLAG_MODE) +#define MBOX_INC_SEL_MASK (NHMEX_M_PMON_CTL_INC_SEL_MASK | \ + NHMEX_M_PMON_CTL_FLAG_MODE) +#define MBOX_SET_FLAG_SEL_MASK (NHMEX_M_PMON_CTL_SET_FLAG_SEL_MASK | \ + NHMEX_M_PMON_CTL_FLAG_MODE) +#define MBOX_INC_SEL_EXTAR_REG(c, r) \ + EVENT_EXTRA_REG(MBOX_INC_SEL(c), NHMEX_M0_MSR_PMU_##r, \ + MBOX_INC_SEL_MASK, (u64)-1, NHMEX_M_##r) +#define MBOX_SET_FLAG_SEL_EXTRA_REG(c, r) \ + EVENT_EXTRA_REG(MBOX_SET_FLAG_SEL(c), NHMEX_M0_MSR_PMU_##r, \ + MBOX_SET_FLAG_SEL_MASK, \ + (u64)-1, NHMEX_M_##r) + +/* NHM-EX Rbox */ +#define NHMEX_R_MSR_GLOBAL_CTL 0xe00 +#define NHMEX_R_MSR_PMON_CTL0 0xe10 +#define NHMEX_R_MSR_PMON_CNT0 0xe11 +#define NHMEX_R_MSR_OFFSET 0x20 + +#define NHMEX_R_MSR_PORTN_QLX_CFG(n) \ + ((n) < 4 ? (0xe0c + (n)) : (0xe2c + (n) - 4)) +#define NHMEX_R_MSR_PORTN_IPERF_CFG0(n) (0xe04 + (n)) +#define NHMEX_R_MSR_PORTN_IPERF_CFG1(n) (0xe24 + (n)) +#define NHMEX_R_MSR_PORTN_XBR_OFFSET(n) \ + (((n) < 4 ? 0 : 0x10) + (n) * 4) +#define NHMEX_R_MSR_PORTN_XBR_SET1_MM_CFG(n) \ + (0xe60 + NHMEX_R_MSR_PORTN_XBR_OFFSET(n)) +#define NHMEX_R_MSR_PORTN_XBR_SET1_MATCH(n) \ + (NHMEX_R_MSR_PORTN_XBR_SET1_MM_CFG(n) + 1) +#define NHMEX_R_MSR_PORTN_XBR_SET1_MASK(n) \ + (NHMEX_R_MSR_PORTN_XBR_SET1_MM_CFG(n) + 2) +#define NHMEX_R_MSR_PORTN_XBR_SET2_MM_CFG(n) \ + (0xe70 + NHMEX_R_MSR_PORTN_XBR_OFFSET(n)) +#define NHMEX_R_MSR_PORTN_XBR_SET2_MATCH(n) \ + (NHMEX_R_MSR_PORTN_XBR_SET2_MM_CFG(n) + 1) +#define NHMEX_R_MSR_PORTN_XBR_SET2_MASK(n) \ + (NHMEX_R_MSR_PORTN_XBR_SET2_MM_CFG(n) + 2) + +#define NHMEX_R_PMON_CTL_EN (1 << 0) +#define NHMEX_R_PMON_CTL_EV_SEL_SHIFT 1 +#define NHMEX_R_PMON_CTL_EV_SEL_MASK \ + (0x1f << NHMEX_R_PMON_CTL_EV_SEL_SHIFT) +#define NHMEX_R_PMON_CTL_PMI_EN (1 << 6) +#define NHMEX_R_PMON_RAW_EVENT_MASK NHMEX_R_PMON_CTL_EV_SEL_MASK + +/* NHM-EX Wbox */ +#define NHMEX_W_MSR_GLOBAL_CTL 0xc80 +#define NHMEX_W_MSR_PMON_CNT0 0xc90 +#define NHMEX_W_MSR_PMON_EVT_SEL0 0xc91 +#define NHMEX_W_MSR_PMON_FIXED_CTR 0x394 +#define NHMEX_W_MSR_PMON_FIXED_CTL 0x395 + +#define NHMEX_W_PMON_GLOBAL_FIXED_EN (1ULL << 31) + +struct intel_uncore_ops; +struct intel_uncore_pmu; +struct intel_uncore_box; +struct uncore_event_desc; + +struct intel_uncore_type { + const char *name; + int num_counters; + int num_boxes; + int perf_ctr_bits; + int fixed_ctr_bits; + unsigned perf_ctr; + unsigned event_ctl; + unsigned event_mask; + unsigned fixed_ctr; + unsigned fixed_ctl; + unsigned box_ctl; + unsigned msr_offset; + unsigned num_shared_regs:8; + unsigned single_fixed:1; + unsigned pair_ctr_ctl:1; + unsigned *msr_offsets; + struct event_constraint unconstrainted; + struct event_constraint *constraints; + struct intel_uncore_pmu *pmus; + struct intel_uncore_ops *ops; + struct uncore_event_desc *event_descs; + const struct attribute_group *attr_groups[4]; + struct pmu *pmu; /* for custom pmu ops */ +}; + +#define pmu_group attr_groups[0] +#define format_group attr_groups[1] +#define events_group attr_groups[2] + +struct intel_uncore_ops { + void (*init_box)(struct intel_uncore_box *); + void (*disable_box)(struct intel_uncore_box *); + void (*enable_box)(struct intel_uncore_box *); + void (*disable_event)(struct intel_uncore_box *, struct perf_event *); + void (*enable_event)(struct intel_uncore_box *, struct perf_event *); + u64 (*read_counter)(struct intel_uncore_box *, struct perf_event *); + int (*hw_config)(struct intel_uncore_box *, struct perf_event *); + struct event_constraint *(*get_constraint)(struct intel_uncore_box *, + struct perf_event *); + void (*put_constraint)(struct intel_uncore_box *, struct perf_event *); +}; + +struct intel_uncore_pmu { + struct pmu pmu; + char name[UNCORE_PMU_NAME_LEN]; + int pmu_idx; + int func_id; + struct intel_uncore_type *type; + struct intel_uncore_box ** __percpu box; + struct list_head box_list; +}; + +struct intel_uncore_extra_reg { + raw_spinlock_t lock; + u64 config, config1, config2; + atomic_t ref; +}; + +struct intel_uncore_box { + int phys_id; + int n_active; /* number of active events */ + int n_events; + int cpu; /* cpu to collect events */ + unsigned long flags; + atomic_t refcnt; + struct perf_event *events[UNCORE_PMC_IDX_MAX]; + struct perf_event *event_list[UNCORE_PMC_IDX_MAX]; + unsigned long active_mask[BITS_TO_LONGS(UNCORE_PMC_IDX_MAX)]; + u64 tags[UNCORE_PMC_IDX_MAX]; + struct pci_dev *pci_dev; + struct intel_uncore_pmu *pmu; + u64 hrtimer_duration; /* hrtimer timeout for this box */ + struct hrtimer hrtimer; + struct list_head list; + struct list_head active_list; + void *io_addr; + struct intel_uncore_extra_reg shared_regs[0]; +}; + +#define UNCORE_BOX_FLAG_INITIATED 0 + +struct uncore_event_desc { + struct kobj_attribute attr; + const char *config; +}; + +#define INTEL_UNCORE_EVENT_DESC(_name, _config) \ +{ \ + .attr = __ATTR(_name, 0444, uncore_event_show, NULL), \ + .config = _config, \ +} + +#define DEFINE_UNCORE_FORMAT_ATTR(_var, _name, _format) \ +static ssize_t __uncore_##_var##_show(struct kobject *kobj, \ + struct kobj_attribute *attr, \ + char *page) \ +{ \ + BUILD_BUG_ON(sizeof(_format) >= PAGE_SIZE); \ + return sprintf(page, _format "\n"); \ +} \ +static struct kobj_attribute format_attr_##_var = \ + __ATTR(_name, 0444, __uncore_##_var##_show, NULL) + + +static ssize_t uncore_event_show(struct kobject *kobj, + struct kobj_attribute *attr, char *buf) +{ + struct uncore_event_desc *event = + container_of(attr, struct uncore_event_desc, attr); + return sprintf(buf, "%s", event->config); +} + +static inline unsigned uncore_pci_box_ctl(struct intel_uncore_box *box) +{ + return box->pmu->type->box_ctl; +} + +static inline unsigned uncore_pci_fixed_ctl(struct intel_uncore_box *box) +{ + return box->pmu->type->fixed_ctl; +} + +static inline unsigned uncore_pci_fixed_ctr(struct intel_uncore_box *box) +{ + return box->pmu->type->fixed_ctr; +} + +static inline +unsigned uncore_pci_event_ctl(struct intel_uncore_box *box, int idx) +{ + return idx * 4 + box->pmu->type->event_ctl; +} + +static inline +unsigned uncore_pci_perf_ctr(struct intel_uncore_box *box, int idx) +{ + return idx * 8 + box->pmu->type->perf_ctr; +} + +static inline unsigned uncore_msr_box_offset(struct intel_uncore_box *box) +{ + struct intel_uncore_pmu *pmu = box->pmu; + return pmu->type->msr_offsets ? + pmu->type->msr_offsets[pmu->pmu_idx] : + pmu->type->msr_offset * pmu->pmu_idx; +} + +static inline unsigned uncore_msr_box_ctl(struct intel_uncore_box *box) +{ + if (!box->pmu->type->box_ctl) + return 0; + return box->pmu->type->box_ctl + uncore_msr_box_offset(box); +} + +static inline unsigned uncore_msr_fixed_ctl(struct intel_uncore_box *box) +{ + if (!box->pmu->type->fixed_ctl) + return 0; + return box->pmu->type->fixed_ctl + uncore_msr_box_offset(box); +} + +static inline unsigned uncore_msr_fixed_ctr(struct intel_uncore_box *box) +{ + return box->pmu->type->fixed_ctr + uncore_msr_box_offset(box); +} + +static inline +unsigned uncore_msr_event_ctl(struct intel_uncore_box *box, int idx) +{ + return box->pmu->type->event_ctl + + (box->pmu->type->pair_ctr_ctl ? 2 * idx : idx) + + uncore_msr_box_offset(box); +} + +static inline +unsigned uncore_msr_perf_ctr(struct intel_uncore_box *box, int idx) +{ + return box->pmu->type->perf_ctr + + (box->pmu->type->pair_ctr_ctl ? 2 * idx : idx) + + uncore_msr_box_offset(box); +} + +static inline +unsigned uncore_fixed_ctl(struct intel_uncore_box *box) +{ + if (box->pci_dev) + return uncore_pci_fixed_ctl(box); + else + return uncore_msr_fixed_ctl(box); +} + +static inline +unsigned uncore_fixed_ctr(struct intel_uncore_box *box) +{ + if (box->pci_dev) + return uncore_pci_fixed_ctr(box); + else + return uncore_msr_fixed_ctr(box); +} + +static inline +unsigned uncore_event_ctl(struct intel_uncore_box *box, int idx) +{ + if (box->pci_dev) + return uncore_pci_event_ctl(box, idx); + else + return uncore_msr_event_ctl(box, idx); +} + +static inline +unsigned uncore_perf_ctr(struct intel_uncore_box *box, int idx) +{ + if (box->pci_dev) + return uncore_pci_perf_ctr(box, idx); + else + return uncore_msr_perf_ctr(box, idx); +} + +static inline int uncore_perf_ctr_bits(struct intel_uncore_box *box) +{ + return box->pmu->type->perf_ctr_bits; +} + +static inline int uncore_fixed_ctr_bits(struct intel_uncore_box *box) +{ + return box->pmu->type->fixed_ctr_bits; +} + +static inline int uncore_num_counters(struct intel_uncore_box *box) +{ + return box->pmu->type->num_counters; +} + +static inline void uncore_disable_box(struct intel_uncore_box *box) +{ + if (box->pmu->type->ops->disable_box) + box->pmu->type->ops->disable_box(box); +} + +static inline void uncore_enable_box(struct intel_uncore_box *box) +{ + if (box->pmu->type->ops->enable_box) + box->pmu->type->ops->enable_box(box); +} + +static inline void uncore_disable_event(struct intel_uncore_box *box, + struct perf_event *event) +{ + box->pmu->type->ops->disable_event(box, event); +} + +static inline void uncore_enable_event(struct intel_uncore_box *box, + struct perf_event *event) +{ + box->pmu->type->ops->enable_event(box, event); +} + +static inline u64 uncore_read_counter(struct intel_uncore_box *box, + struct perf_event *event) +{ + return box->pmu->type->ops->read_counter(box, event); +} + +static inline void uncore_box_init(struct intel_uncore_box *box) +{ + if (!test_and_set_bit(UNCORE_BOX_FLAG_INITIATED, &box->flags)) { + if (box->pmu->type->ops->init_box) + box->pmu->type->ops->init_box(box); + } +} + +static inline bool uncore_box_is_fake(struct intel_uncore_box *box) +{ + return (box->phys_id < 0); +} diff --git a/arch/x86/kernel/cpu/perf_event_knc.c b/arch/x86/kernel/cpu/perf_event_knc.c new file mode 100644 index 00000000000..838fa8772c6 --- /dev/null +++ b/arch/x86/kernel/cpu/perf_event_knc.c @@ -0,0 +1,319 @@ +/* Driver for Intel Xeon Phi "Knights Corner" PMU */ + +#include <linux/perf_event.h> +#include <linux/types.h> + +#include <asm/hardirq.h> + +#include "perf_event.h" + +static const u64 knc_perfmon_event_map[] = +{ + [PERF_COUNT_HW_CPU_CYCLES] = 0x002a, + [PERF_COUNT_HW_INSTRUCTIONS] = 0x0016, + [PERF_COUNT_HW_CACHE_REFERENCES] = 0x0028, + [PERF_COUNT_HW_CACHE_MISSES] = 0x0029, + [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = 0x0012, + [PERF_COUNT_HW_BRANCH_MISSES] = 0x002b, +}; + +static const u64 __initconst knc_hw_cache_event_ids + [PERF_COUNT_HW_CACHE_MAX] + [PERF_COUNT_HW_CACHE_OP_MAX] + [PERF_COUNT_HW_CACHE_RESULT_MAX] = +{ + [ C(L1D) ] = { + [ C(OP_READ) ] = { + /* On Xeon Phi event "0" is a valid DATA_READ */ + /* (L1 Data Cache Reads) Instruction. */ + /* We code this as ARCH_PERFMON_EVENTSEL_INT as this */ + /* bit will always be set in x86_pmu_hw_config(). */ + [ C(RESULT_ACCESS) ] = ARCH_PERFMON_EVENTSEL_INT, + /* DATA_READ */ + [ C(RESULT_MISS) ] = 0x0003, /* DATA_READ_MISS */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0x0001, /* DATA_WRITE */ + [ C(RESULT_MISS) ] = 0x0004, /* DATA_WRITE_MISS */ + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0x0011, /* L1_DATA_PF1 */ + [ C(RESULT_MISS) ] = 0x001c, /* L1_DATA_PF1_MISS */ + }, + }, + [ C(L1I ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x000c, /* CODE_READ */ + [ C(RESULT_MISS) ] = 0x000e, /* CODE_CACHE_MISS */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0x0, + [ C(RESULT_MISS) ] = 0x0, + }, + }, + [ C(LL ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0x10cb, /* L2_READ_MISS */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0x10cc, /* L2_WRITE_HIT */ + [ C(RESULT_MISS) ] = 0, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0x10fc, /* L2_DATA_PF2 */ + [ C(RESULT_MISS) ] = 0x10fe, /* L2_DATA_PF2_MISS */ + }, + }, + [ C(DTLB) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = ARCH_PERFMON_EVENTSEL_INT, + /* DATA_READ */ + /* see note on L1 OP_READ */ + [ C(RESULT_MISS) ] = 0x0002, /* DATA_PAGE_WALK */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0x0001, /* DATA_WRITE */ + [ C(RESULT_MISS) ] = 0x0002, /* DATA_PAGE_WALK */ + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0x0, + [ C(RESULT_MISS) ] = 0x0, + }, + }, + [ C(ITLB) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x000c, /* CODE_READ */ + [ C(RESULT_MISS) ] = 0x000d, /* CODE_PAGE_WALK */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + }, + [ C(BPU ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0012, /* BRANCHES */ + [ C(RESULT_MISS) ] = 0x002b, /* BRANCHES_MISPREDICTED */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + }, +}; + + +static u64 knc_pmu_event_map(int hw_event) +{ + return knc_perfmon_event_map[hw_event]; +} + +static struct event_constraint knc_event_constraints[] = +{ + INTEL_EVENT_CONSTRAINT(0xc3, 0x1), /* HWP_L2HIT */ + INTEL_EVENT_CONSTRAINT(0xc4, 0x1), /* HWP_L2MISS */ + INTEL_EVENT_CONSTRAINT(0xc8, 0x1), /* L2_READ_HIT_E */ + INTEL_EVENT_CONSTRAINT(0xc9, 0x1), /* L2_READ_HIT_M */ + INTEL_EVENT_CONSTRAINT(0xca, 0x1), /* L2_READ_HIT_S */ + INTEL_EVENT_CONSTRAINT(0xcb, 0x1), /* L2_READ_MISS */ + INTEL_EVENT_CONSTRAINT(0xcc, 0x1), /* L2_WRITE_HIT */ + INTEL_EVENT_CONSTRAINT(0xce, 0x1), /* L2_STRONGLY_ORDERED_STREAMING_VSTORES_MISS */ + INTEL_EVENT_CONSTRAINT(0xcf, 0x1), /* L2_WEAKLY_ORDERED_STREAMING_VSTORE_MISS */ + INTEL_EVENT_CONSTRAINT(0xd7, 0x1), /* L2_VICTIM_REQ_WITH_DATA */ + INTEL_EVENT_CONSTRAINT(0xe3, 0x1), /* SNP_HITM_BUNIT */ + INTEL_EVENT_CONSTRAINT(0xe6, 0x1), /* SNP_HIT_L2 */ + INTEL_EVENT_CONSTRAINT(0xe7, 0x1), /* SNP_HITM_L2 */ + INTEL_EVENT_CONSTRAINT(0xf1, 0x1), /* L2_DATA_READ_MISS_CACHE_FILL */ + INTEL_EVENT_CONSTRAINT(0xf2, 0x1), /* L2_DATA_WRITE_MISS_CACHE_FILL */ + INTEL_EVENT_CONSTRAINT(0xf6, 0x1), /* L2_DATA_READ_MISS_MEM_FILL */ + INTEL_EVENT_CONSTRAINT(0xf7, 0x1), /* L2_DATA_WRITE_MISS_MEM_FILL */ + INTEL_EVENT_CONSTRAINT(0xfc, 0x1), /* L2_DATA_PF2 */ + INTEL_EVENT_CONSTRAINT(0xfd, 0x1), /* L2_DATA_PF2_DROP */ + INTEL_EVENT_CONSTRAINT(0xfe, 0x1), /* L2_DATA_PF2_MISS */ + INTEL_EVENT_CONSTRAINT(0xff, 0x1), /* L2_DATA_HIT_INFLIGHT_PF2 */ + EVENT_CONSTRAINT_END +}; + +#define MSR_KNC_IA32_PERF_GLOBAL_STATUS 0x0000002d +#define MSR_KNC_IA32_PERF_GLOBAL_OVF_CONTROL 0x0000002e +#define MSR_KNC_IA32_PERF_GLOBAL_CTRL 0x0000002f + +#define KNC_ENABLE_COUNTER0 0x00000001 +#define KNC_ENABLE_COUNTER1 0x00000002 + +static void knc_pmu_disable_all(void) +{ + u64 val; + + rdmsrl(MSR_KNC_IA32_PERF_GLOBAL_CTRL, val); + val &= ~(KNC_ENABLE_COUNTER0|KNC_ENABLE_COUNTER1); + wrmsrl(MSR_KNC_IA32_PERF_GLOBAL_CTRL, val); +} + +static void knc_pmu_enable_all(int added) +{ + u64 val; + + rdmsrl(MSR_KNC_IA32_PERF_GLOBAL_CTRL, val); + val |= (KNC_ENABLE_COUNTER0|KNC_ENABLE_COUNTER1); + wrmsrl(MSR_KNC_IA32_PERF_GLOBAL_CTRL, val); +} + +static inline void +knc_pmu_disable_event(struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + u64 val; + + val = hwc->config; + val &= ~ARCH_PERFMON_EVENTSEL_ENABLE; + + (void)wrmsrl_safe(hwc->config_base + hwc->idx, val); +} + +static void knc_pmu_enable_event(struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + u64 val; + + val = hwc->config; + val |= ARCH_PERFMON_EVENTSEL_ENABLE; + + (void)wrmsrl_safe(hwc->config_base + hwc->idx, val); +} + +static inline u64 knc_pmu_get_status(void) +{ + u64 status; + + rdmsrl(MSR_KNC_IA32_PERF_GLOBAL_STATUS, status); + + return status; +} + +static inline void knc_pmu_ack_status(u64 ack) +{ + wrmsrl(MSR_KNC_IA32_PERF_GLOBAL_OVF_CONTROL, ack); +} + +static int knc_pmu_handle_irq(struct pt_regs *regs) +{ + struct perf_sample_data data; + struct cpu_hw_events *cpuc; + int handled = 0; + int bit, loops; + u64 status; + + cpuc = &__get_cpu_var(cpu_hw_events); + + knc_pmu_disable_all(); + + status = knc_pmu_get_status(); + if (!status) { + knc_pmu_enable_all(0); + return handled; + } + + loops = 0; +again: + knc_pmu_ack_status(status); + if (++loops > 100) { + WARN_ONCE(1, "perf: irq loop stuck!\n"); + perf_event_print_debug(); + goto done; + } + + inc_irq_stat(apic_perf_irqs); + + for_each_set_bit(bit, (unsigned long *)&status, X86_PMC_IDX_MAX) { + struct perf_event *event = cpuc->events[bit]; + + handled++; + + if (!test_bit(bit, cpuc->active_mask)) + continue; + + if (!intel_pmu_save_and_restart(event)) + continue; + + perf_sample_data_init(&data, 0, event->hw.last_period); + + if (perf_event_overflow(event, &data, regs)) + x86_pmu_stop(event, 0); + } + + /* + * Repeat if there is more work to be done: + */ + status = knc_pmu_get_status(); + if (status) + goto again; + +done: + knc_pmu_enable_all(0); + + return handled; +} + + +PMU_FORMAT_ATTR(event, "config:0-7" ); +PMU_FORMAT_ATTR(umask, "config:8-15" ); +PMU_FORMAT_ATTR(edge, "config:18" ); +PMU_FORMAT_ATTR(inv, "config:23" ); +PMU_FORMAT_ATTR(cmask, "config:24-31" ); + +static struct attribute *intel_knc_formats_attr[] = { + &format_attr_event.attr, + &format_attr_umask.attr, + &format_attr_edge.attr, + &format_attr_inv.attr, + &format_attr_cmask.attr, + NULL, +}; + +static const struct x86_pmu knc_pmu __initconst = { + .name = "knc", + .handle_irq = knc_pmu_handle_irq, + .disable_all = knc_pmu_disable_all, + .enable_all = knc_pmu_enable_all, + .enable = knc_pmu_enable_event, + .disable = knc_pmu_disable_event, + .hw_config = x86_pmu_hw_config, + .schedule_events = x86_schedule_events, + .eventsel = MSR_KNC_EVNTSEL0, + .perfctr = MSR_KNC_PERFCTR0, + .event_map = knc_pmu_event_map, + .max_events = ARRAY_SIZE(knc_perfmon_event_map), + .apic = 1, + .max_period = (1ULL << 39) - 1, + .version = 0, + .num_counters = 2, + .cntval_bits = 40, + .cntval_mask = (1ULL << 40) - 1, + .get_event_constraints = x86_get_event_constraints, + .event_constraints = knc_event_constraints, + .format_attrs = intel_knc_formats_attr, +}; + +__init int knc_pmu_init(void) +{ + x86_pmu = knc_pmu; + + memcpy(hw_cache_event_ids, knc_hw_cache_event_ids, + sizeof(hw_cache_event_ids)); + + return 0; +} diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c index ef484d9d0a2..5d466b7d860 100644 --- a/arch/x86/kernel/cpu/perf_event_p4.c +++ b/arch/x86/kernel/cpu/perf_event_p4.c @@ -895,8 +895,8 @@ static void p4_pmu_disable_pebs(void) * So at moment let leave metrics turned on forever -- it's * ok for now but need to be revisited! * - * (void)checking_wrmsrl(MSR_IA32_PEBS_ENABLE, (u64)0); - * (void)checking_wrmsrl(MSR_P4_PEBS_MATRIX_VERT, (u64)0); + * (void)wrmsrl_safe(MSR_IA32_PEBS_ENABLE, 0); + * (void)wrmsrl_safe(MSR_P4_PEBS_MATRIX_VERT, 0); */ } @@ -909,9 +909,8 @@ static inline void p4_pmu_disable_event(struct perf_event *event) * state we need to clear P4_CCCR_OVF, otherwise interrupt get * asserted again and again */ - (void)checking_wrmsrl(hwc->config_base, - (u64)(p4_config_unpack_cccr(hwc->config)) & - ~P4_CCCR_ENABLE & ~P4_CCCR_OVF & ~P4_CCCR_RESERVED); + (void)wrmsrl_safe(hwc->config_base, + p4_config_unpack_cccr(hwc->config) & ~P4_CCCR_ENABLE & ~P4_CCCR_OVF & ~P4_CCCR_RESERVED); } static void p4_pmu_disable_all(void) @@ -943,8 +942,8 @@ static void p4_pmu_enable_pebs(u64 config) bind = &p4_pebs_bind_map[idx]; - (void)checking_wrmsrl(MSR_IA32_PEBS_ENABLE, (u64)bind->metric_pebs); - (void)checking_wrmsrl(MSR_P4_PEBS_MATRIX_VERT, (u64)bind->metric_vert); + (void)wrmsrl_safe(MSR_IA32_PEBS_ENABLE, (u64)bind->metric_pebs); + (void)wrmsrl_safe(MSR_P4_PEBS_MATRIX_VERT, (u64)bind->metric_vert); } static void p4_pmu_enable_event(struct perf_event *event) @@ -957,7 +956,7 @@ static void p4_pmu_enable_event(struct perf_event *event) u64 escr_addr, cccr; bind = &p4_event_bind_map[idx]; - escr_addr = (u64)bind->escr_msr[thread]; + escr_addr = bind->escr_msr[thread]; /* * - we dont support cascaded counters yet @@ -978,8 +977,8 @@ static void p4_pmu_enable_event(struct perf_event *event) */ p4_pmu_enable_pebs(hwc->config); - (void)checking_wrmsrl(escr_addr, escr_conf); - (void)checking_wrmsrl(hwc->config_base, + (void)wrmsrl_safe(escr_addr, escr_conf); + (void)wrmsrl_safe(hwc->config_base, (cccr & ~P4_CCCR_RESERVED) | P4_CCCR_ENABLE); } @@ -1005,8 +1004,6 @@ static int p4_pmu_handle_irq(struct pt_regs *regs) int idx, handled = 0; u64 val; - perf_sample_data_init(&data, 0); - cpuc = &__get_cpu_var(cpu_hw_events); for (idx = 0; idx < x86_pmu.num_counters; idx++) { @@ -1034,10 +1031,12 @@ static int p4_pmu_handle_irq(struct pt_regs *regs) handled += overflow; /* event overflow for sure */ - data.period = event->hw.last_period; + perf_sample_data_init(&data, 0, hwc->last_period); if (!x86_perf_event_set_period(event)) continue; + + if (perf_event_overflow(event, &data, regs)) x86_pmu_stop(event, 0); } @@ -1258,7 +1257,24 @@ again: pass++; goto again; } - + /* + * Perf does test runs to see if a whole group can be assigned + * together succesfully. There can be multiple rounds of this. + * Unfortunately, p4_pmu_swap_config_ts touches the hwc->config + * bits, such that the next round of group assignments will + * cause the above p4_should_swap_ts to pass instead of fail. + * This leads to counters exclusive to thread0 being used by + * thread1. + * + * Solve this with a cheap hack, reset the idx back to -1 to + * force a new lookup (p4_next_cntr) to get the right counter + * for the right thread. + * + * This probably doesn't comply with the general spirit of how + * perf wants to work, but P4 is special. :-( + */ + if (p4_should_swap_ts(hwc->config, cpu)) + hwc->idx = -1; p4_pmu_swap_config_ts(hwc, cpu); if (assign) assign[i] = cntr_idx; @@ -1271,6 +1287,17 @@ done: return num ? -EINVAL : 0; } +PMU_FORMAT_ATTR(cccr, "config:0-31" ); +PMU_FORMAT_ATTR(escr, "config:32-62"); +PMU_FORMAT_ATTR(ht, "config:63" ); + +static struct attribute *intel_p4_formats_attr[] = { + &format_attr_cccr.attr, + &format_attr_escr.attr, + &format_attr_ht.attr, + NULL, +}; + static __initconst const struct x86_pmu p4_pmu = { .name = "Netburst P4/Xeon", .handle_irq = p4_pmu_handle_irq, @@ -1305,14 +1332,17 @@ static __initconst const struct x86_pmu p4_pmu = { * the former idea is taken from OProfile code */ .perfctr_second_write = 1, + + .format_attrs = intel_p4_formats_attr, }; __init int p4_pmu_init(void) { unsigned int low, high; + int i, reg; /* If we get stripped -- indexing fails */ - BUILD_BUG_ON(ARCH_P4_MAX_CCCR > X86_PMC_MAX_GENERIC); + BUILD_BUG_ON(ARCH_P4_MAX_CCCR > INTEL_PMC_MAX_GENERIC); rdmsr(MSR_IA32_MISC_ENABLE, low, high); if (!(low & (1 << 7))) { @@ -1328,5 +1358,19 @@ __init int p4_pmu_init(void) x86_pmu = p4_pmu; + /* + * Even though the counters are configured to interrupt a particular + * logical processor when an overflow happens, testing has shown that + * on kdump kernels (which uses a single cpu), thread1's counter + * continues to run and will report an NMI on thread0. Due to the + * overflow bug, this leads to a stream of unknown NMIs. + * + * Solve this by zero'ing out the registers to mimic a reset. + */ + for (i = 0; i < x86_pmu.num_counters; i++) { + reg = x86_pmu_config_addr(i); + wrmsrl_safe(reg, 0ULL); + } + return 0; } diff --git a/arch/x86/kernel/cpu/perf_event_p6.c b/arch/x86/kernel/cpu/perf_event_p6.c index c7181befecd..7c1a0c07b60 100644 --- a/arch/x86/kernel/cpu/perf_event_p6.c +++ b/arch/x86/kernel/cpu/perf_event_p6.c @@ -8,13 +8,106 @@ */ static const u64 p6_perfmon_event_map[] = { - [PERF_COUNT_HW_CPU_CYCLES] = 0x0079, - [PERF_COUNT_HW_INSTRUCTIONS] = 0x00c0, - [PERF_COUNT_HW_CACHE_REFERENCES] = 0x0f2e, - [PERF_COUNT_HW_CACHE_MISSES] = 0x012e, - [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = 0x00c4, - [PERF_COUNT_HW_BRANCH_MISSES] = 0x00c5, - [PERF_COUNT_HW_BUS_CYCLES] = 0x0062, + [PERF_COUNT_HW_CPU_CYCLES] = 0x0079, /* CPU_CLK_UNHALTED */ + [PERF_COUNT_HW_INSTRUCTIONS] = 0x00c0, /* INST_RETIRED */ + [PERF_COUNT_HW_CACHE_REFERENCES] = 0x0f2e, /* L2_RQSTS:M:E:S:I */ + [PERF_COUNT_HW_CACHE_MISSES] = 0x012e, /* L2_RQSTS:I */ + [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = 0x00c4, /* BR_INST_RETIRED */ + [PERF_COUNT_HW_BRANCH_MISSES] = 0x00c5, /* BR_MISS_PRED_RETIRED */ + [PERF_COUNT_HW_BUS_CYCLES] = 0x0062, /* BUS_DRDY_CLOCKS */ + [PERF_COUNT_HW_STALLED_CYCLES_FRONTEND] = 0x00a2, /* RESOURCE_STALLS */ + +}; + +static const u64 __initconst p6_hw_cache_event_ids + [PERF_COUNT_HW_CACHE_MAX] + [PERF_COUNT_HW_CACHE_OP_MAX] + [PERF_COUNT_HW_CACHE_RESULT_MAX] = +{ + [ C(L1D) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0043, /* DATA_MEM_REFS */ + [ C(RESULT_MISS) ] = 0x0045, /* DCU_LINES_IN */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0x0f29, /* L2_LD:M:E:S:I */ + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(L1I ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0080, /* IFU_IFETCH */ + [ C(RESULT_MISS) ] = 0x0f28, /* L2_IFETCH:M:E:S:I */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(LL ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0x0025, /* L2_M_LINES_INM */ + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(DTLB) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0043, /* DATA_MEM_REFS */ + [ C(RESULT_MISS) ] = 0, + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = 0, + [ C(RESULT_MISS) ] = 0, + }, + }, + [ C(ITLB) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0080, /* IFU_IFETCH */ + [ C(RESULT_MISS) ] = 0x0085, /* ITLB_MISS */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + }, + [ C(BPU ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x00c4, /* BR_INST_RETIRED */ + [ C(RESULT_MISS) ] = 0x00c5, /* BR_MISS_PRED_RETIRED */ + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + }, }; static u64 p6_pmu_event_map(int hw_event) @@ -34,7 +127,7 @@ static struct event_constraint p6_event_constraints[] = { INTEL_EVENT_CONSTRAINT(0xc1, 0x1), /* FLOPS */ INTEL_EVENT_CONSTRAINT(0x10, 0x1), /* FP_COMP_OPS_EXE */ - INTEL_EVENT_CONSTRAINT(0x11, 0x1), /* FP_ASSIST */ + INTEL_EVENT_CONSTRAINT(0x11, 0x2), /* FP_ASSIST */ INTEL_EVENT_CONSTRAINT(0x12, 0x2), /* MUL */ INTEL_EVENT_CONSTRAINT(0x13, 0x2), /* DIV */ INTEL_EVENT_CONSTRAINT(0x14, 0x1), /* CYCLES_DIV_BUSY */ @@ -64,29 +157,46 @@ static void p6_pmu_enable_all(int added) static inline void p6_pmu_disable_event(struct perf_event *event) { - struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); struct hw_perf_event *hwc = &event->hw; u64 val = P6_NOP_EVENT; - if (cpuc->enabled) - val |= ARCH_PERFMON_EVENTSEL_ENABLE; - - (void)checking_wrmsrl(hwc->config_base, val); + (void)wrmsrl_safe(hwc->config_base, val); } static void p6_pmu_enable_event(struct perf_event *event) { - struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); struct hw_perf_event *hwc = &event->hw; u64 val; val = hwc->config; - if (cpuc->enabled) - val |= ARCH_PERFMON_EVENTSEL_ENABLE; - (void)checking_wrmsrl(hwc->config_base, val); + /* + * p6 only has a global event enable, set on PerfEvtSel0 + * We "disable" events by programming P6_NOP_EVENT + * and we rely on p6_pmu_enable_all() being called + * to actually enable the events. + */ + + (void)wrmsrl_safe(hwc->config_base, val); } +PMU_FORMAT_ATTR(event, "config:0-7" ); +PMU_FORMAT_ATTR(umask, "config:8-15" ); +PMU_FORMAT_ATTR(edge, "config:18" ); +PMU_FORMAT_ATTR(pc, "config:19" ); +PMU_FORMAT_ATTR(inv, "config:23" ); +PMU_FORMAT_ATTR(cmask, "config:24-31" ); + +static struct attribute *intel_p6_formats_attr[] = { + &format_attr_event.attr, + &format_attr_umask.attr, + &format_attr_edge.attr, + &format_attr_pc.attr, + &format_attr_inv.attr, + &format_attr_cmask.attr, + NULL, +}; + static __initconst const struct x86_pmu p6_pmu = { .name = "p6", .handle_irq = x86_pmu_handle_irq, @@ -115,29 +225,55 @@ static __initconst const struct x86_pmu p6_pmu = { .cntval_mask = (1ULL << 32) - 1, .get_event_constraints = x86_get_event_constraints, .event_constraints = p6_event_constraints, + + .format_attrs = intel_p6_formats_attr, + .events_sysfs_show = intel_event_sysfs_show, + }; +static __init void p6_pmu_rdpmc_quirk(void) +{ + if (boot_cpu_data.x86_mask < 9) { + /* + * PPro erratum 26; fixed in stepping 9 and above. + */ + pr_warn("Userspace RDPMC support disabled due to a CPU erratum\n"); + x86_pmu.attr_rdpmc_broken = 1; + x86_pmu.attr_rdpmc = 0; + } +} + __init int p6_pmu_init(void) { + x86_pmu = p6_pmu; + switch (boot_cpu_data.x86_model) { - case 1: - case 3: /* Pentium Pro */ - case 5: - case 6: /* Pentium II */ - case 7: - case 8: - case 11: /* Pentium III */ - case 9: - case 13: - /* Pentium M */ + case 1: /* Pentium Pro */ + x86_add_quirk(p6_pmu_rdpmc_quirk); + break; + + case 3: /* Pentium II - Klamath */ + case 5: /* Pentium II - Deschutes */ + case 6: /* Pentium II - Mendocino */ break; + + case 7: /* Pentium III - Katmai */ + case 8: /* Pentium III - Coppermine */ + case 10: /* Pentium III Xeon */ + case 11: /* Pentium III - Tualatin */ + break; + + case 9: /* Pentium M - Banias */ + case 13: /* Pentium M - Dothan */ + break; + default: - pr_cont("unsupported p6 CPU model %d ", - boot_cpu_data.x86_model); + pr_cont("unsupported p6 CPU model %d ", boot_cpu_data.x86_model); return -ENODEV; } - x86_pmu = p6_pmu; + memcpy(hw_cache_event_ids, p6_hw_cache_event_ids, + sizeof(hw_cache_event_ids)); return 0; } diff --git a/arch/x86/kernel/cpu/perfctr-watchdog.c b/arch/x86/kernel/cpu/perfctr-watchdog.c index 966512b2cac..2e8caf03f59 100644 --- a/arch/x86/kernel/cpu/perfctr-watchdog.c +++ b/arch/x86/kernel/cpu/perfctr-watchdog.c @@ -56,6 +56,8 @@ static inline unsigned int nmi_perfctr_msr_to_bit(unsigned int msr) switch (boot_cpu_data.x86) { case 6: return msr - MSR_P6_PERFCTR0; + case 11: + return msr - MSR_KNC_PERFCTR0; case 15: return msr - MSR_P4_BPU_PERFCTR0; } @@ -82,6 +84,8 @@ static inline unsigned int nmi_evntsel_msr_to_bit(unsigned int msr) switch (boot_cpu_data.x86) { case 6: return msr - MSR_P6_EVNTSEL0; + case 11: + return msr - MSR_KNC_EVNTSEL0; case 15: return msr - MSR_P4_BSU_ESCR0; } diff --git a/arch/x86/kernel/cpu/powerflags.c b/arch/x86/kernel/cpu/powerflags.c index 7b3fe56b1c2..31f0f335ed2 100644 --- a/arch/x86/kernel/cpu/powerflags.c +++ b/arch/x86/kernel/cpu/powerflags.c @@ -11,10 +11,10 @@ const char *const x86_power_flags[32] = { "fid", /* frequency id control */ "vid", /* voltage id control */ "ttp", /* thermal trip */ - "tm", - "stc", - "100mhzsteps", - "hwpstate", + "tm", /* hardware thermal control */ + "stc", /* software thermal control */ + "100mhzsteps", /* 100 MHz multiplier control */ + "hwpstate", /* hardware P-state control */ "", /* tsc invariant mapped to constant_tsc */ "cpb", /* core performance boost */ "eff_freq_ro", /* Readonly aperf/mperf */ diff --git a/arch/x86/kernel/cpu/proc.c b/arch/x86/kernel/cpu/proc.c index 8022c668148..06fe3ed8b85 100644 --- a/arch/x86/kernel/cpu/proc.c +++ b/arch/x86/kernel/cpu/proc.c @@ -11,41 +11,31 @@ static void show_cpuinfo_core(struct seq_file *m, struct cpuinfo_x86 *c, unsigned int cpu) { #ifdef CONFIG_SMP - if (c->x86_max_cores * smp_num_siblings > 1) { - seq_printf(m, "physical id\t: %d\n", c->phys_proc_id); - seq_printf(m, "siblings\t: %d\n", - cpumask_weight(cpu_core_mask(cpu))); - seq_printf(m, "core id\t\t: %d\n", c->cpu_core_id); - seq_printf(m, "cpu cores\t: %d\n", c->booted_cores); - seq_printf(m, "apicid\t\t: %d\n", c->apicid); - seq_printf(m, "initial apicid\t: %d\n", c->initial_apicid); - } + seq_printf(m, "physical id\t: %d\n", c->phys_proc_id); + seq_printf(m, "siblings\t: %d\n", cpumask_weight(cpu_core_mask(cpu))); + seq_printf(m, "core id\t\t: %d\n", c->cpu_core_id); + seq_printf(m, "cpu cores\t: %d\n", c->booted_cores); + seq_printf(m, "apicid\t\t: %d\n", c->apicid); + seq_printf(m, "initial apicid\t: %d\n", c->initial_apicid); #endif } #ifdef CONFIG_X86_32 static void show_cpuinfo_misc(struct seq_file *m, struct cpuinfo_x86 *c) { - /* - * We use exception 16 if we have hardware math and we've either seen - * it or the CPU claims it is internal - */ - int fpu_exception = c->hard_math && (ignore_fpu_irq || cpu_has_fpu); seq_printf(m, "fdiv_bug\t: %s\n" - "hlt_bug\t\t: %s\n" "f00f_bug\t: %s\n" "coma_bug\t: %s\n" "fpu\t\t: %s\n" "fpu_exception\t: %s\n" "cpuid level\t: %d\n" "wp\t\t: %s\n", - c->fdiv_bug ? "yes" : "no", - c->hlt_works_ok ? "no" : "yes", - c->f00f_bug ? "yes" : "no", - c->coma_bug ? "yes" : "no", - c->hard_math ? "yes" : "no", - fpu_exception ? "yes" : "no", + static_cpu_has_bug(X86_BUG_FDIV) ? "yes" : "no", + static_cpu_has_bug(X86_BUG_F00F) ? "yes" : "no", + static_cpu_has_bug(X86_BUG_COMA) ? "yes" : "no", + static_cpu_has(X86_FEATURE_FPU) ? "yes" : "no", + static_cpu_has(X86_FEATURE_FPU) ? "yes" : "no", c->cpuid_level, c->wp_works_ok ? "yes" : "no"); } @@ -140,10 +130,7 @@ static int show_cpuinfo(struct seq_file *m, void *v) static void *c_start(struct seq_file *m, loff_t *pos) { - if (*pos == 0) /* just in case, cpu 0 is not the first */ - *pos = cpumask_first(cpu_online_mask); - else - *pos = cpumask_next(*pos - 1, cpu_online_mask); + *pos = cpumask_next(*pos - 1, cpu_online_mask); if ((*pos) < nr_cpu_ids) return &cpu_data(*pos); return NULL; diff --git a/arch/x86/kernel/cpu/rdrand.c b/arch/x86/kernel/cpu/rdrand.c index feca286c2bb..136ac74dee8 100644 --- a/arch/x86/kernel/cpu/rdrand.c +++ b/arch/x86/kernel/cpu/rdrand.c @@ -27,24 +27,11 @@ static int __init x86_rdrand_setup(char *s) { setup_clear_cpu_cap(X86_FEATURE_RDRAND); + setup_clear_cpu_cap(X86_FEATURE_RDSEED); return 1; } __setup("nordrand", x86_rdrand_setup); -/* We can't use arch_get_random_long() here since alternatives haven't run */ -static inline int rdrand_long(unsigned long *v) -{ - int ok; - asm volatile("1: " RDRAND_LONG "\n\t" - "jc 2f\n\t" - "decl %0\n\t" - "jnz 1b\n\t" - "2:" - : "=r" (ok), "=a" (*v) - : "0" (RDRAND_RETRY_LOOPS)); - return ok; -} - /* * Force a reseed cycle; we are architecturally guaranteed a reseed * after no more than 512 128-bit chunks of random data. This also @@ -52,7 +39,7 @@ static inline int rdrand_long(unsigned long *v) */ #define RESEED_LOOP ((512*128)/sizeof(unsigned long)) -void __cpuinit x86_init_rdrand(struct cpuinfo_x86 *c) +void x86_init_rdrand(struct cpuinfo_x86 *c) { #ifdef CONFIG_ARCH_RANDOM unsigned long tmp; diff --git a/arch/x86/kernel/cpu/scattered.c b/arch/x86/kernel/cpu/scattered.c index c7f64e6f537..b6f794aa169 100644 --- a/arch/x86/kernel/cpu/scattered.c +++ b/arch/x86/kernel/cpu/scattered.c @@ -1,5 +1,5 @@ /* - * Routines to indentify additional cpu features that are scattered in + * Routines to identify additional cpu features that are scattered in * cpuid space. */ #include <linux/cpu.h> @@ -24,14 +24,14 @@ enum cpuid_regs { CR_EBX }; -void __cpuinit init_scattered_cpuid_features(struct cpuinfo_x86 *c) +void init_scattered_cpuid_features(struct cpuinfo_x86 *c) { u32 max_level; u32 regs[4]; const struct cpuid_bit *cb; - static const struct cpuid_bit __cpuinitconst cpuid_bits[] = { - { X86_FEATURE_DTS, CR_EAX, 0, 0x00000006, 0 }, + static const struct cpuid_bit cpuid_bits[] = { + { X86_FEATURE_DTHERM, CR_EAX, 0, 0x00000006, 0 }, { X86_FEATURE_IDA, CR_EAX, 1, 0x00000006, 0 }, { X86_FEATURE_ARAT, CR_EAX, 2, 0x00000006, 0 }, { X86_FEATURE_PLN, CR_EAX, 4, 0x00000006, 0 }, @@ -39,7 +39,9 @@ void __cpuinit init_scattered_cpuid_features(struct cpuinfo_x86 *c) { X86_FEATURE_APERFMPERF, CR_ECX, 0, 0x00000006, 0 }, { X86_FEATURE_EPB, CR_ECX, 3, 0x00000006, 0 }, { X86_FEATURE_XSAVEOPT, CR_EAX, 0, 0x0000000d, 1 }, + { X86_FEATURE_HW_PSTATE, CR_EDX, 7, 0x80000007, 0 }, { X86_FEATURE_CPB, CR_EDX, 9, 0x80000007, 0 }, + { X86_FEATURE_PROC_FEEDBACK, CR_EDX,11, 0x80000007, 0 }, { X86_FEATURE_NPT, CR_EDX, 0, 0x8000000a, 0 }, { X86_FEATURE_LBRV, CR_EDX, 1, 0x8000000a, 0 }, { X86_FEATURE_SVML, CR_EDX, 2, 0x8000000a, 0 }, diff --git a/arch/x86/kernel/cpu/sched.c b/arch/x86/kernel/cpu/sched.c deleted file mode 100644 index a640ae5ad20..00000000000 --- a/arch/x86/kernel/cpu/sched.c +++ /dev/null @@ -1,55 +0,0 @@ -#include <linux/sched.h> -#include <linux/math64.h> -#include <linux/percpu.h> -#include <linux/irqflags.h> - -#include <asm/cpufeature.h> -#include <asm/processor.h> - -#ifdef CONFIG_SMP - -static DEFINE_PER_CPU(struct aperfmperf, old_perf_sched); - -static unsigned long scale_aperfmperf(void) -{ - struct aperfmperf val, *old = &__get_cpu_var(old_perf_sched); - unsigned long ratio, flags; - - local_irq_save(flags); - get_aperfmperf(&val); - local_irq_restore(flags); - - ratio = calc_aperfmperf_ratio(old, &val); - *old = val; - - return ratio; -} - -unsigned long arch_scale_freq_power(struct sched_domain *sd, int cpu) -{ - /* - * do aperf/mperf on the cpu level because it includes things - * like turbo mode, which are relevant to full cores. - */ - if (boot_cpu_has(X86_FEATURE_APERFMPERF)) - return scale_aperfmperf(); - - /* - * maybe have something cpufreq here - */ - - return default_scale_freq_power(sd, cpu); -} - -unsigned long arch_scale_smt_power(struct sched_domain *sd, int cpu) -{ - /* - * aperf/mperf already includes the smt gain - */ - if (boot_cpu_has(X86_FEATURE_APERFMPERF)) - return SCHED_LOAD_SCALE; - - return default_scale_smt_power(sd, cpu); -} - -#endif diff --git a/arch/x86/kernel/cpu/topology.c b/arch/x86/kernel/cpu/topology.c index 4397e987a1c..4c60eaf0571 100644 --- a/arch/x86/kernel/cpu/topology.c +++ b/arch/x86/kernel/cpu/topology.c @@ -26,7 +26,7 @@ * exists, use it for populating initial_apicid and cpu topology * detection. */ -void __cpuinit detect_extended_topology(struct cpuinfo_x86 *c) +void detect_extended_topology(struct cpuinfo_x86 *c) { #ifdef CONFIG_SMP unsigned int eax, ebx, ecx, edx, sub_index; diff --git a/arch/x86/kernel/cpu/transmeta.c b/arch/x86/kernel/cpu/transmeta.c index 28000743bbb..3fa0e5ad86b 100644 --- a/arch/x86/kernel/cpu/transmeta.c +++ b/arch/x86/kernel/cpu/transmeta.c @@ -1,11 +1,10 @@ #include <linux/kernel.h> #include <linux/mm.h> -#include <linux/init.h> #include <asm/processor.h> #include <asm/msr.h> #include "cpu.h" -static void __cpuinit early_init_transmeta(struct cpuinfo_x86 *c) +static void early_init_transmeta(struct cpuinfo_x86 *c) { u32 xlvl; @@ -17,7 +16,7 @@ static void __cpuinit early_init_transmeta(struct cpuinfo_x86 *c) } } -static void __cpuinit init_transmeta(struct cpuinfo_x86 *c) +static void init_transmeta(struct cpuinfo_x86 *c) { unsigned int cap_mask, uk, max, dummy; unsigned int cms_rev1, cms_rev2; @@ -98,7 +97,7 @@ static void __cpuinit init_transmeta(struct cpuinfo_x86 *c) #endif } -static const struct cpu_dev __cpuinitconst transmeta_cpu_dev = { +static const struct cpu_dev transmeta_cpu_dev = { .c_vendor = "Transmeta", .c_ident = { "GenuineTMx86", "TransmetaCPU" }, .c_early_init = early_init_transmeta, diff --git a/arch/x86/kernel/cpu/umc.c b/arch/x86/kernel/cpu/umc.c index fd2c37bf7ac..ef9c2a0078b 100644 --- a/arch/x86/kernel/cpu/umc.c +++ b/arch/x86/kernel/cpu/umc.c @@ -1,5 +1,4 @@ #include <linux/kernel.h> -#include <linux/init.h> #include <asm/processor.h> #include "cpu.h" @@ -8,11 +7,11 @@ * so no special init takes place. */ -static const struct cpu_dev __cpuinitconst umc_cpu_dev = { +static const struct cpu_dev umc_cpu_dev = { .c_vendor = "UMC", .c_ident = { "UMC UMC UMC" }, - .c_models = { - { .vendor = X86_VENDOR_UMC, .family = 4, .model_names = + .legacy_models = { + { .family = 4, .model_names = { [1] = "U5D", [2] = "U5S", diff --git a/arch/x86/kernel/cpu/vmware.c b/arch/x86/kernel/cpu/vmware.c index d22d0c4edcf..628a059a9a0 100644 --- a/arch/x86/kernel/cpu/vmware.c +++ b/arch/x86/kernel/cpu/vmware.c @@ -33,6 +33,9 @@ #define VMWARE_PORT_CMD_GETVERSION 10 #define VMWARE_PORT_CMD_GETHZ 45 +#define VMWARE_PORT_CMD_GETVCPU_INFO 68 +#define VMWARE_PORT_CMD_LEGACY_X2APIC 3 +#define VMWARE_PORT_CMD_VCPU_RESERVED 31 #define VMWARE_PORT(cmd, eax, ebx, ecx, edx) \ __asm__("inl (%%dx)" : \ @@ -90,7 +93,7 @@ static void __init vmware_platform_setup(void) * serial key should be enough, as this will always have a VMware * specific string when running under VMware hypervisor. */ -static bool __init vmware_platform(void) +static uint32_t __init vmware_platform(void) { if (cpu_has_hypervisor) { unsigned int eax; @@ -99,12 +102,12 @@ static bool __init vmware_platform(void) cpuid(CPUID_VMWARE_INFO_LEAF, &eax, &hyper_vendor_id[0], &hyper_vendor_id[1], &hyper_vendor_id[2]); if (!memcmp(hyper_vendor_id, "VMwareVMware", 12)) - return true; + return CPUID_VMWARE_INFO_LEAF; } else if (dmi_available && dmi_name_in_serial("VMware") && __vmware_platform()) - return true; + return 1; - return false; + return 0; } /* @@ -119,16 +122,26 @@ static bool __init vmware_platform(void) * so that the kernel could just trust the hypervisor with providing a * reliable virtual TSC that is suitable for timekeeping. */ -static void __cpuinit vmware_set_cpu_features(struct cpuinfo_x86 *c) +static void vmware_set_cpu_features(struct cpuinfo_x86 *c) { set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); set_cpu_cap(c, X86_FEATURE_TSC_RELIABLE); } +/* Checks if hypervisor supports x2apic without VT-D interrupt remapping. */ +static bool __init vmware_legacy_x2apic_available(void) +{ + uint32_t eax, ebx, ecx, edx; + VMWARE_PORT(GETVCPU_INFO, eax, ebx, ecx, edx); + return (eax & (1 << VMWARE_PORT_CMD_VCPU_RESERVED)) == 0 && + (eax & (1 << VMWARE_PORT_CMD_LEGACY_X2APIC)) != 0; +} + const __refconst struct hypervisor_x86 x86_hyper_vmware = { .name = "VMware", .detect = vmware_platform, .set_cpu_features = vmware_set_cpu_features, .init_platform = vmware_platform_setup, + .x2apic_available = vmware_legacy_x2apic_available, }; EXPORT_SYMBOL(x86_hyper_vmware); diff --git a/arch/x86/kernel/cpuid.c b/arch/x86/kernel/cpuid.c index a524353d93f..3225ae6c518 100644 --- a/arch/x86/kernel/cpuid.c +++ b/arch/x86/kernel/cpuid.c @@ -43,7 +43,6 @@ #include <asm/processor.h> #include <asm/msr.h> -#include <asm/system.h> static struct class *cpuid_class; @@ -86,7 +85,7 @@ static ssize_t cpuid_read(struct file *file, char __user *buf, { char __user *tmp = buf; struct cpuid_regs cmd; - int cpu = iminor(file->f_path.dentry->d_inode); + int cpu = iminor(file_inode(file)); u64 pos = *ppos; ssize_t bytes = 0; int err = 0; @@ -117,7 +116,7 @@ static int cpuid_open(struct inode *inode, struct file *file) unsigned int cpu; struct cpuinfo_x86 *c; - cpu = iminor(file->f_path.dentry->d_inode); + cpu = iminor(file_inode(file)); if (cpu >= nr_cpu_ids || !cpu_online(cpu)) return -ENXIO; /* No such CPU */ @@ -138,7 +137,7 @@ static const struct file_operations cpuid_fops = { .open = cpuid_open, }; -static __cpuinit int cpuid_device_create(int cpu) +static int cpuid_device_create(int cpu) { struct device *dev; @@ -152,9 +151,8 @@ static void cpuid_device_destroy(int cpu) device_destroy(cpuid_class, MKDEV(CPUID_MAJOR, cpu)); } -static int __cpuinit cpuid_class_cpu_callback(struct notifier_block *nfb, - unsigned long action, - void *hcpu) +static int cpuid_class_cpu_callback(struct notifier_block *nfb, + unsigned long action, void *hcpu) { unsigned int cpu = (unsigned long)hcpu; int err = 0; @@ -200,12 +198,15 @@ static int __init cpuid_init(void) goto out_chrdev; } cpuid_class->devnode = cpuid_devnode; + + cpu_notifier_register_begin(); for_each_online_cpu(i) { err = cpuid_device_create(i); if (err != 0) goto out_class; } - register_hotcpu_notifier(&cpuid_class_cpu_notifier); + __register_hotcpu_notifier(&cpuid_class_cpu_notifier); + cpu_notifier_register_done(); err = 0; goto out; @@ -215,6 +216,7 @@ out_class: for_each_online_cpu(i) { cpuid_device_destroy(i); } + cpu_notifier_register_done(); class_destroy(cpuid_class); out_chrdev: __unregister_chrdev(CPUID_MAJOR, 0, NR_CPUS, "cpu/cpuid"); @@ -226,11 +228,13 @@ static void __exit cpuid_exit(void) { int cpu = 0; + cpu_notifier_register_begin(); for_each_online_cpu(cpu) cpuid_device_destroy(cpu); class_destroy(cpuid_class); __unregister_chrdev(CPUID_MAJOR, 0, NR_CPUS, "cpu/cpuid"); - unregister_hotcpu_notifier(&cpuid_class_cpu_notifier); + __unregister_hotcpu_notifier(&cpuid_class_cpu_notifier); + cpu_notifier_register_done(); } module_init(cpuid_init); diff --git a/arch/x86/kernel/crash.c b/arch/x86/kernel/crash.c index 13ad89971d4..507de806659 100644 --- a/arch/x86/kernel/crash.c +++ b/arch/x86/kernel/crash.c @@ -7,7 +7,6 @@ * */ -#include <linux/init.h> #include <linux/types.h> #include <linux/kernel.h> #include <linux/smp.h> @@ -16,6 +15,7 @@ #include <linux/delay.h> #include <linux/elf.h> #include <linux/elfcore.h> +#include <linux/module.h> #include <asm/processor.h> #include <asm/hardirq.h> @@ -30,15 +30,34 @@ int in_crash_kexec; +/* + * This is used to VMCLEAR all VMCSs loaded on the + * processor. And when loading kvm_intel module, the + * callback function pointer will be assigned. + * + * protected by rcu. + */ +crash_vmclear_fn __rcu *crash_vmclear_loaded_vmcss = NULL; +EXPORT_SYMBOL_GPL(crash_vmclear_loaded_vmcss); + +static inline void cpu_crash_vmclear_loaded_vmcss(void) +{ + crash_vmclear_fn *do_vmclear_operation = NULL; + + rcu_read_lock(); + do_vmclear_operation = rcu_dereference(crash_vmclear_loaded_vmcss); + if (do_vmclear_operation) + do_vmclear_operation(); + rcu_read_unlock(); +} + #if defined(CONFIG_SMP) && defined(CONFIG_X86_LOCAL_APIC) static void kdump_nmi_callback(int cpu, struct pt_regs *regs) { #ifdef CONFIG_X86_32 struct pt_regs fixed_regs; -#endif -#ifdef CONFIG_X86_32 if (!user_mode_vm(regs)) { crash_fixup_ss_esp(&fixed_regs, regs); regs = &fixed_regs; @@ -46,6 +65,11 @@ static void kdump_nmi_callback(int cpu, struct pt_regs *regs) #endif crash_save_cpu(regs, cpu); + /* + * VMCLEAR VMCSs loaded on all cpus if needed. + */ + cpu_crash_vmclear_loaded_vmcss(); + /* Disable VMX or SVM if needed. * * We need to disable virtualization on all CPUs. @@ -88,6 +112,11 @@ void native_machine_crash_shutdown(struct pt_regs *regs) kdump_nmi_shootdown_cpus(); + /* + * VMCLEAR VMCSs loaded on this cpu if needed. + */ + cpu_crash_vmclear_loaded_vmcss(); + /* Booting kdump kernel with VMX or SVM enabled won't work, * because (among other limitations) we can't disable paging * with the virt flags. @@ -95,10 +124,12 @@ void native_machine_crash_shutdown(struct pt_regs *regs) cpu_emergency_vmxoff(); cpu_emergency_svm_disable(); - lapic_shutdown(); -#if defined(CONFIG_X86_IO_APIC) +#ifdef CONFIG_X86_IO_APIC + /* Prevent crash_kexec() from deadlocking on ioapic_lock. */ + ioapic_zap_locks(); disable_IO_APIC(); #endif + lapic_shutdown(); #ifdef CONFIG_HPET_TIMER hpet_disable(); #endif diff --git a/arch/x86/kernel/crash_dump_32.c b/arch/x86/kernel/crash_dump_32.c index 642f75a68cd..11891ca7b71 100644 --- a/arch/x86/kernel/crash_dump_32.c +++ b/arch/x86/kernel/crash_dump_32.c @@ -62,16 +62,16 @@ ssize_t copy_oldmem_page(unsigned long pfn, char *buf, if (!userbuf) { memcpy(buf, (vaddr + offset), csize); - kunmap_atomic(vaddr, KM_PTE0); + kunmap_atomic(vaddr); } else { if (!kdump_buf_page) { printk(KERN_WARNING "Kdump: Kdump buffer page not" " allocated\n"); - kunmap_atomic(vaddr, KM_PTE0); + kunmap_atomic(vaddr); return -EFAULT; } copy_page(kdump_buf_page, vaddr); - kunmap_atomic(vaddr, KM_PTE0); + kunmap_atomic(vaddr); if (copy_to_user(buf, (kdump_buf_page + offset), csize)) return -EFAULT; } diff --git a/arch/x86/kernel/devicetree.c b/arch/x86/kernel/devicetree.c index 52821799a70..7db54b5d5f8 100644 --- a/arch/x86/kernel/devicetree.c +++ b/arch/x86/kernel/devicetree.c @@ -4,6 +4,7 @@ #include <linux/bootmem.h> #include <linux/export.h> #include <linux/io.h> +#include <linux/irqdomain.h> #include <linux/interrupt.h> #include <linux/list.h> #include <linux/of.h> @@ -17,74 +18,15 @@ #include <linux/initrd.h> #include <asm/hpet.h> -#include <asm/irq_controller.h> #include <asm/apic.h> #include <asm/pci_x86.h> +#include <asm/setup.h> __initdata u64 initial_dtb; char __initdata cmd_line[COMMAND_LINE_SIZE]; -static LIST_HEAD(irq_domains); -static DEFINE_RAW_SPINLOCK(big_irq_lock); int __initdata of_ioapic; -#ifdef CONFIG_X86_IO_APIC -static void add_interrupt_host(struct irq_domain *ih) -{ - unsigned long flags; - - raw_spin_lock_irqsave(&big_irq_lock, flags); - list_add(&ih->l, &irq_domains); - raw_spin_unlock_irqrestore(&big_irq_lock, flags); -} -#endif - -static struct irq_domain *get_ih_from_node(struct device_node *controller) -{ - struct irq_domain *ih, *found = NULL; - unsigned long flags; - - raw_spin_lock_irqsave(&big_irq_lock, flags); - list_for_each_entry(ih, &irq_domains, l) { - if (ih->controller == controller) { - found = ih; - break; - } - } - raw_spin_unlock_irqrestore(&big_irq_lock, flags); - return found; -} - -unsigned int irq_create_of_mapping(struct device_node *controller, - const u32 *intspec, unsigned int intsize) -{ - struct irq_domain *ih; - u32 virq, type; - int ret; - - ih = get_ih_from_node(controller); - if (!ih) - return 0; - ret = ih->xlate(ih, intspec, intsize, &virq, &type); - if (ret) - return 0; - if (type == IRQ_TYPE_NONE) - return virq; - irq_set_irq_type(virq, type); - return virq; -} -EXPORT_SYMBOL_GPL(irq_create_of_mapping); - -unsigned long pci_address_to_pio(phys_addr_t address) -{ - /* - * The ioport address can be directly used by inX / outX - */ - BUG_ON(address >= (1 << 16)); - return (unsigned long)address; -} -EXPORT_SYMBOL_GPL(pci_address_to_pio); - void __init early_init_dt_scan_chosen_arch(unsigned long node) { BUG(); @@ -100,16 +42,6 @@ void * __init early_init_dt_alloc_memory_arch(u64 size, u64 align) return __alloc_bootmem(size, align, __pa(MAX_DMA_ADDRESS)); } -#ifdef CONFIG_BLK_DEV_INITRD -void __init early_init_dt_setup_initrd_arch(unsigned long start, - unsigned long end) -{ - initrd_start = (unsigned long)__va(start); - initrd_end = (unsigned long)__va(end); - initrd_below_start_ok = 1; -} -#endif - void __init add_dtb(u64 data) { initial_dtb = data + offsetof(struct setup_data, data); @@ -155,7 +87,6 @@ struct device_node *pcibios_get_phb_of_node(struct pci_bus *bus) static int x86_of_pci_irq_enable(struct pci_dev *dev) { - struct of_irq oirq; u32 virq; int ret; u8 pin; @@ -166,12 +97,7 @@ static int x86_of_pci_irq_enable(struct pci_dev *dev) if (!pin) return 0; - ret = of_irq_map_pci(dev, &oirq); - if (ret) - return ret; - - virq = irq_create_of_mapping(oirq.controller, oirq.specifier, - oirq.size); + virq = of_irq_parse_and_map_pci(dev, 0, 0); if (virq == 0) return -EINVAL; dev->irq = virq; @@ -182,7 +108,7 @@ static void x86_of_pci_irq_disable(struct pci_dev *dev) { } -void __cpuinit x86_of_pci_init(void) +void x86_of_pci_init(void) { pcibios_enable_irq = x86_of_pci_irq_enable; pcibios_disable_irq = x86_of_pci_irq_disable; @@ -280,32 +206,23 @@ static void __init dtb_apic_setup(void) static void __init x86_flattree_get_config(void) { u32 size, map_len; - void *new_dtb; + void *dt; if (!initial_dtb) return; - map_len = max(PAGE_SIZE - (initial_dtb & ~PAGE_MASK), - (u64)sizeof(struct boot_param_header)); + map_len = max(PAGE_SIZE - (initial_dtb & ~PAGE_MASK), (u64)128); - initial_boot_params = early_memremap(initial_dtb, map_len); - size = be32_to_cpu(initial_boot_params->totalsize); + initial_boot_params = dt = early_memremap(initial_dtb, map_len); + size = of_get_flat_dt_size(); if (map_len < size) { - early_iounmap(initial_boot_params, map_len); - initial_boot_params = early_memremap(initial_dtb, size); + early_iounmap(dt, map_len); + initial_boot_params = dt = early_memremap(initial_dtb, size); map_len = size; } - new_dtb = alloc_bootmem(size); - memcpy(new_dtb, initial_boot_params, size); - early_iounmap(initial_boot_params, map_len); - - initial_boot_params = new_dtb; - - /* root level address cells */ - of_scan_flat_dt(early_init_dt_scan_root, NULL); - - unflatten_device_tree(); + unflatten_and_copy_device_tree(); + early_iounmap(dt, map_len); } #else static inline void x86_flattree_get_config(void) { } @@ -354,34 +271,80 @@ static struct of_ioapic_type of_ioapic_type[] = }, }; -static int ioapic_xlate(struct irq_domain *id, const u32 *intspec, u32 intsize, - u32 *out_hwirq, u32 *out_type) +static int ioapic_xlate(struct irq_domain *domain, + struct device_node *controller, + const u32 *intspec, u32 intsize, + irq_hw_number_t *out_hwirq, u32 *out_type) { - struct mp_ioapic_gsi *gsi_cfg; struct io_apic_irq_attr attr; struct of_ioapic_type *it; - u32 line, idx, type; + u32 line, idx; + int rc; - if (intsize < 2) + if (WARN_ON(intsize < 2)) return -EINVAL; - line = *intspec; - idx = (u32) id->priv; - gsi_cfg = mp_ioapic_gsi_routing(idx); - *out_hwirq = line + gsi_cfg->gsi_base; + line = intspec[0]; - intspec++; - type = *intspec; - - if (type >= ARRAY_SIZE(of_ioapic_type)) + if (intspec[1] >= ARRAY_SIZE(of_ioapic_type)) return -EINVAL; - it = of_ioapic_type + type; - *out_type = it->out_type; + it = &of_ioapic_type[intspec[1]]; + idx = (u32) domain->host_data; set_io_apic_irq_attr(&attr, idx, line, it->trigger, it->polarity); - return io_apic_setup_irq_pin_once(*out_hwirq, cpu_to_node(0), &attr); + rc = io_apic_setup_irq_pin_once(irq_find_mapping(domain, line), + cpu_to_node(0), &attr); + if (rc) + return rc; + + *out_hwirq = line; + *out_type = it->out_type; + return 0; +} + +const struct irq_domain_ops ioapic_irq_domain_ops = { + .xlate = ioapic_xlate, +}; + +static void dt_add_ioapic_domain(unsigned int ioapic_num, + struct device_node *np) +{ + struct irq_domain *id; + struct mp_ioapic_gsi *gsi_cfg; + int ret; + int num; + + gsi_cfg = mp_ioapic_gsi_routing(ioapic_num); + num = gsi_cfg->gsi_end - gsi_cfg->gsi_base + 1; + + id = irq_domain_add_linear(np, num, &ioapic_irq_domain_ops, + (void *)ioapic_num); + BUG_ON(!id); + if (gsi_cfg->gsi_base == 0) { + /* + * The first NR_IRQS_LEGACY irq descs are allocated in + * early_irq_init() and need just a mapping. The + * remaining irqs need both. All of them are preallocated + * and assigned so we can keep the 1:1 mapping which the ioapic + * is having. + */ + irq_domain_associate_many(id, 0, 0, NR_IRQS_LEGACY); + + if (num > NR_IRQS_LEGACY) { + ret = irq_create_strict_mappings(id, NR_IRQS_LEGACY, + NR_IRQS_LEGACY, num - NR_IRQS_LEGACY); + if (ret) + pr_err("Error creating mapping for the " + "remaining IRQs: %d\n", ret); + } + irq_set_default_host(id); + } else { + ret = irq_create_strict_mappings(id, gsi_cfg->gsi_base, 0, num); + if (ret) + pr_err("Error creating IRQ mapping: %d\n", ret); + } } static void __init ioapic_add_ofnode(struct device_node *np) @@ -398,14 +361,7 @@ static void __init ioapic_add_ofnode(struct device_node *np) for (i = 0; i < nr_ioapics; i++) { if (r.start == mpc_ioapic_addr(i)) { - struct irq_domain *id; - - id = kzalloc(sizeof(*id), GFP_KERNEL); - BUG_ON(!id); - id->controller = np; - id->xlate = ioapic_xlate; - id->priv = (void *)i; - add_interrupt_host(id); + dt_add_ioapic_domain(i, np); return; } } diff --git a/arch/x86/kernel/doublefault_32.c b/arch/x86/kernel/doublefault.c index 37250fe490b..f6dfd9334b6 100644 --- a/arch/x86/kernel/doublefault_32.c +++ b/arch/x86/kernel/doublefault.c @@ -1,6 +1,5 @@ #include <linux/mm.h> #include <linux/sched.h> -#include <linux/init.h> #include <linux/init_task.h> #include <linux/fs.h> @@ -9,6 +8,8 @@ #include <asm/processor.h> #include <asm/desc.h> +#ifdef CONFIG_X86_32 + #define DOUBLEFAULT_STACKSIZE (1024) static unsigned long doublefault_stack[DOUBLEFAULT_STACKSIZE]; #define STACK_START (unsigned long)(doublefault_stack+DOUBLEFAULT_STACKSIZE) @@ -20,7 +21,7 @@ static void doublefault_fn(void) struct desc_ptr gdt_desc = {0, 0}; unsigned long gdt, tss; - store_gdt(&gdt_desc); + native_store_gdt(&gdt_desc); gdt = gdt_desc.address; printk(KERN_EMERG "PANIC: double fault, gdt at %08lx [%d bytes]\n", gdt, gdt_desc.size); @@ -67,3 +68,16 @@ struct tss_struct doublefault_tss __cacheline_aligned = { .__cr3 = __pa_nodebug(swapper_pg_dir), } }; + +/* dummy for do_double_fault() call */ +void df_debug(struct pt_regs *regs, long error_code) {} + +#else /* !CONFIG_X86_32 */ + +void df_debug(struct pt_regs *regs, long error_code) +{ + pr_emerg("PANIC: double fault, error_code: 0x%lx\n", error_code); + show_regs(regs); + panic("Machine halted."); +} +#endif diff --git a/arch/x86/kernel/dumpstack.c b/arch/x86/kernel/dumpstack.c index 4025fe4f928..b74ebc7c440 100644 --- a/arch/x86/kernel/dumpstack.c +++ b/arch/x86/kernel/dumpstack.c @@ -25,10 +25,15 @@ unsigned int code_bytes = 64; int kstack_depth_to_print = 3 * STACKSLOTS_PER_LINE; static int die_counter; -void printk_address(unsigned long address, int reliable) +static void printk_stack_address(unsigned long address, int reliable) { - printk(" [<%p>] %s%pB\n", (void *) address, - reliable ? "" : "? ", (void *) address); + pr_cont(" [<%p>] %s%pB\n", + (void *)address, reliable ? "" : "? ", (void *)address); +} + +void printk_address(unsigned long address) +{ + pr_cont(" [<%p>] %pS\n", (void *)address, (void *)address); } #ifdef CONFIG_FUNCTION_GRAPH_TRACER @@ -37,13 +42,16 @@ print_ftrace_graph_addr(unsigned long addr, void *data, const struct stacktrace_ops *ops, struct thread_info *tinfo, int *graph) { - struct task_struct *task = tinfo->task; + struct task_struct *task; unsigned long ret_addr; - int index = task->curr_ret_stack; + int index; if (addr != (unsigned long)return_to_handler) return; + task = tinfo->task; + index = task->curr_ret_stack; + if (!task->ret_stack || index < *graph) return; @@ -148,7 +156,7 @@ static void print_trace_address(void *data, unsigned long addr, int reliable) { touch_nmi_watchdog(); printk(data); - printk_address(addr, reliable); + printk_stack_address(addr, reliable); } static const struct stacktrace_ops print_trace_ops = { @@ -173,32 +181,26 @@ void show_trace(struct task_struct *task, struct pt_regs *regs, void show_stack(struct task_struct *task, unsigned long *sp) { - show_stack_log_lvl(task, NULL, sp, 0, ""); -} - -/* - * The architecture-independent dump_stack generator - */ -void dump_stack(void) -{ - unsigned long bp; + unsigned long bp = 0; unsigned long stack; - bp = stack_frame(current, NULL); - printk("Pid: %d, comm: %.20s %s %s %.*s\n", - current->pid, current->comm, print_tainted(), - init_utsname()->release, - (int)strcspn(init_utsname()->version, " "), - init_utsname()->version); - show_trace(NULL, NULL, &stack, bp); + /* + * Stack frames below this one aren't interesting. Don't show them + * if we're printing for %current. + */ + if (!sp && (!task || task == current)) { + sp = &stack; + bp = stack_frame(current, NULL); + } + + show_stack_log_lvl(task, NULL, sp, bp, ""); } -EXPORT_SYMBOL(dump_stack); static arch_spinlock_t die_lock = __ARCH_SPIN_LOCK_UNLOCKED; static int die_owner = -1; static unsigned int die_nest_count; -unsigned __kprobes long oops_begin(void) +unsigned long oops_begin(void) { int cpu; unsigned long flags; @@ -221,15 +223,16 @@ unsigned __kprobes long oops_begin(void) return flags; } EXPORT_SYMBOL_GPL(oops_begin); +NOKPROBE_SYMBOL(oops_begin); -void __kprobes oops_end(unsigned long flags, struct pt_regs *regs, int signr) +void oops_end(unsigned long flags, struct pt_regs *regs, int signr) { if (regs && kexec_should_crash(current)) crash_kexec(regs); bust_spinlocks(0); die_owner = -1; - add_taint(TAINT_DIE); + add_taint(TAINT_DIE, LOCKDEP_NOW_UNRELIABLE); die_nest_count--; if (!die_nest_count) /* Nest count reaches zero, release the lock. */ @@ -245,8 +248,9 @@ void __kprobes oops_end(unsigned long flags, struct pt_regs *regs, int signr) panic("Fatal exception"); do_exit(signr); } +NOKPROBE_SYMBOL(oops_end); -int __kprobes __die(const char *str, struct pt_regs *regs, long err) +int __die(const char *str, struct pt_regs *regs, long err) { #ifdef CONFIG_X86_32 unsigned short ss; @@ -265,10 +269,11 @@ int __kprobes __die(const char *str, struct pt_regs *regs, long err) #endif printk("\n"); if (notify_die(DIE_OOPS, str, regs, err, - current->thread.trap_no, SIGSEGV) == NOTIFY_STOP) + current->thread.trap_nr, SIGSEGV) == NOTIFY_STOP) return 1; - show_registers(regs); + print_modules(); + show_regs(regs); #ifdef CONFIG_X86_32 if (user_mode_vm(regs)) { sp = regs->sp; @@ -283,11 +288,12 @@ int __kprobes __die(const char *str, struct pt_regs *regs, long err) #else /* Executive summary in case the oops scrolled away */ printk(KERN_ALERT "RIP "); - printk_address(regs->ip, 1); + printk_address(regs->ip); printk(" RSP <%016lx>\n", regs->sp); #endif return 0; } +NOKPROBE_SYMBOL(__die); /* * This is gone through when something in the kernel has done something bad @@ -308,16 +314,33 @@ void die(const char *str, struct pt_regs *regs, long err) static int __init kstack_setup(char *s) { + ssize_t ret; + unsigned long val; + if (!s) return -EINVAL; - kstack_depth_to_print = simple_strtoul(s, NULL, 0); + + ret = kstrtoul(s, 0, &val); + if (ret) + return ret; + kstack_depth_to_print = val; return 0; } early_param("kstack", kstack_setup); static int __init code_bytes_setup(char *s) { - code_bytes = simple_strtoul(s, NULL, 0); + ssize_t ret; + unsigned long val; + + if (!s) + return -EINVAL; + + ret = kstrtoul(s, 0, &val); + if (ret) + return ret; + + code_bytes = val; if (code_bytes > 8192) code_bytes = 8192; diff --git a/arch/x86/kernel/dumpstack_32.c b/arch/x86/kernel/dumpstack_32.c index c99f9ed013d..5abd4cd4230 100644 --- a/arch/x86/kernel/dumpstack_32.c +++ b/arch/x86/kernel/dumpstack_32.c @@ -16,12 +16,35 @@ #include <asm/stacktrace.h> +static void *is_irq_stack(void *p, void *irq) +{ + if (p < irq || p >= (irq + THREAD_SIZE)) + return NULL; + return irq + THREAD_SIZE; +} + + +static void *is_hardirq_stack(unsigned long *stack, int cpu) +{ + void *irq = per_cpu(hardirq_stack, cpu); + + return is_irq_stack(stack, irq); +} + +static void *is_softirq_stack(unsigned long *stack, int cpu) +{ + void *irq = per_cpu(softirq_stack, cpu); + + return is_irq_stack(stack, irq); +} void dump_trace(struct task_struct *task, struct pt_regs *regs, unsigned long *stack, unsigned long bp, const struct stacktrace_ops *ops, void *data) { + const unsigned cpu = get_cpu(); int graph = 0; + u32 *prev_esp; if (!task) task = current; @@ -30,7 +53,7 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, unsigned long dummy; stack = &dummy; - if (task && task != current) + if (task != current) stack = (unsigned long *)task->thread.sp; } @@ -39,18 +62,31 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, for (;;) { struct thread_info *context; + void *end_stack; + + end_stack = is_hardirq_stack(stack, cpu); + if (!end_stack) + end_stack = is_softirq_stack(stack, cpu); - context = (struct thread_info *) - ((unsigned long)stack & (~(THREAD_SIZE - 1))); - bp = ops->walk_stack(context, stack, bp, ops, data, NULL, &graph); + context = task_thread_info(task); + bp = ops->walk_stack(context, stack, bp, ops, data, + end_stack, &graph); - stack = (unsigned long *)context->previous_esp; + /* Stop if not on irq stack */ + if (!end_stack) + break; + + /* The previous esp is saved on the bottom of the stack */ + prev_esp = (u32 *)(end_stack - THREAD_SIZE); + stack = (unsigned long *)*prev_esp; if (!stack) break; + if (ops->stack(data, "IRQ") < 0) break; touch_nmi_watchdog(); } + put_cpu(); } EXPORT_SYMBOL(dump_trace); @@ -73,25 +109,22 @@ show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs, if (kstack_end(stack)) break; if (i && ((i % STACKSLOTS_PER_LINE) == 0)) - printk(KERN_CONT "\n"); - printk(KERN_CONT " %08lx", *stack++); + pr_cont("\n"); + pr_cont(" %08lx", *stack++); touch_nmi_watchdog(); } - printk(KERN_CONT "\n"); + pr_cont("\n"); show_trace_log_lvl(task, regs, sp, bp, log_lvl); } -void show_registers(struct pt_regs *regs) +void show_regs(struct pt_regs *regs) { int i; - print_modules(); - __show_regs(regs, 0); + show_regs_print_info(KERN_EMERG); + __show_regs(regs, !user_mode_vm(regs)); - printk(KERN_EMERG "Process %.*s (pid: %d, ti=%p task=%p task.ti=%p)\n", - TASK_COMM_LEN, current->comm, task_pid_nr(current), - current_thread_info(), current, task_thread_info(current)); /* * When in-kernel, we also print out the stack and code at the * time of the fault.. @@ -102,10 +135,10 @@ void show_registers(struct pt_regs *regs) unsigned char c; u8 *ip; - printk(KERN_EMERG "Stack:\n"); + pr_emerg("Stack:\n"); show_stack_log_lvl(NULL, regs, ®s->sp, 0, KERN_EMERG); - printk(KERN_EMERG "Code: "); + pr_emerg("Code:"); ip = (u8 *)regs->ip - code_prologue; if (ip < (u8 *)PAGE_OFFSET || probe_kernel_address(ip, c)) { @@ -116,16 +149,16 @@ void show_registers(struct pt_regs *regs) for (i = 0; i < code_len; i++, ip++) { if (ip < (u8 *)PAGE_OFFSET || probe_kernel_address(ip, c)) { - printk(KERN_CONT " Bad EIP value."); + pr_cont(" Bad EIP value."); break; } if (ip == (u8 *)regs->ip) - printk(KERN_CONT "<%02x> ", c); + pr_cont(" <%02x>", c); else - printk(KERN_CONT "%02x ", c); + pr_cont(" %02x", c); } } - printk(KERN_CONT "\n"); + pr_cont("\n"); } int is_valid_bugaddr(unsigned long ip) diff --git a/arch/x86/kernel/dumpstack_64.c b/arch/x86/kernel/dumpstack_64.c index 17107bd6e1f..1abcb50b48a 100644 --- a/arch/x86/kernel/dumpstack_64.c +++ b/arch/x86/kernel/dumpstack_64.c @@ -104,6 +104,44 @@ in_irq_stack(unsigned long *stack, unsigned long *irq_stack, return (stack >= irq_stack && stack < irq_stack_end); } +static const unsigned long irq_stack_size = + (IRQ_STACK_SIZE - 64) / sizeof(unsigned long); + +enum stack_type { + STACK_IS_UNKNOWN, + STACK_IS_NORMAL, + STACK_IS_EXCEPTION, + STACK_IS_IRQ, +}; + +static enum stack_type +analyze_stack(int cpu, struct task_struct *task, unsigned long *stack, + unsigned long **stack_end, unsigned long *irq_stack, + unsigned *used, char **id) +{ + unsigned long addr; + + addr = ((unsigned long)stack & (~(THREAD_SIZE - 1))); + if ((unsigned long)task_stack_page(task) == addr) + return STACK_IS_NORMAL; + + *stack_end = in_exception_stack(cpu, (unsigned long)stack, + used, id); + if (*stack_end) + return STACK_IS_EXCEPTION; + + if (!irq_stack) + return STACK_IS_NORMAL; + + *stack_end = irq_stack; + irq_stack = irq_stack - irq_stack_size; + + if (in_irq_stack(stack, irq_stack, *stack_end)) + return STACK_IS_IRQ; + + return STACK_IS_UNKNOWN; +} + /* * x86-64 can have up to three kernel stacks: * process stack @@ -116,12 +154,12 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, const struct stacktrace_ops *ops, void *data) { const unsigned cpu = get_cpu(); - unsigned long *irq_stack_end = - (unsigned long *)per_cpu(irq_stack_ptr, cpu); - unsigned used = 0; struct thread_info *tinfo; - int graph = 0; + unsigned long *irq_stack = (unsigned long *)per_cpu(irq_stack_ptr, cpu); unsigned long dummy; + unsigned used = 0; + int graph = 0; + int done = 0; if (!task) task = current; @@ -143,49 +181,61 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, * exceptions */ tinfo = task_thread_info(task); - for (;;) { + while (!done) { + unsigned long *stack_end; + enum stack_type stype; char *id; - unsigned long *estack_end; - estack_end = in_exception_stack(cpu, (unsigned long)stack, - &used, &id); - if (estack_end) { + stype = analyze_stack(cpu, task, stack, &stack_end, + irq_stack, &used, &id); + + /* Default finish unless specified to continue */ + done = 1; + + switch (stype) { + + /* Break out early if we are on the thread stack */ + case STACK_IS_NORMAL: + break; + + case STACK_IS_EXCEPTION: + if (ops->stack(data, id) < 0) break; bp = ops->walk_stack(tinfo, stack, bp, ops, - data, estack_end, &graph); + data, stack_end, &graph); ops->stack(data, "<EOE>"); /* * We link to the next stack via the * second-to-last pointer (index -2 to end) in the * exception stack: */ - stack = (unsigned long *) estack_end[-2]; - continue; - } - if (irq_stack_end) { - unsigned long *irq_stack; - irq_stack = irq_stack_end - - (IRQ_STACK_SIZE - 64) / sizeof(*irq_stack); - - if (in_irq_stack(stack, irq_stack, irq_stack_end)) { - if (ops->stack(data, "IRQ") < 0) - break; - bp = ops->walk_stack(tinfo, stack, bp, - ops, data, irq_stack_end, &graph); - /* - * We link to the next stack (which would be - * the process stack normally) the last - * pointer (index -1 to end) in the IRQ stack: - */ - stack = (unsigned long *) (irq_stack_end[-1]); - irq_stack_end = NULL; - ops->stack(data, "EOI"); - continue; - } + stack = (unsigned long *) stack_end[-2]; + done = 0; + break; + + case STACK_IS_IRQ: + + if (ops->stack(data, "IRQ") < 0) + break; + bp = ops->walk_stack(tinfo, stack, bp, + ops, data, stack_end, &graph); + /* + * We link to the next stack (which would be + * the process stack normally) the last + * pointer (index -1 to end) in the IRQ stack: + */ + stack = (unsigned long *) (stack_end[-1]); + irq_stack = NULL; + ops->stack(data, "EOI"); + done = 0; + break; + + case STACK_IS_UNKNOWN: + ops->stack(data, "UNK"); + break; } - break; } /* @@ -228,36 +278,31 @@ show_stack_log_lvl(struct task_struct *task, struct pt_regs *regs, if (stack >= irq_stack && stack <= irq_stack_end) { if (stack == irq_stack_end) { stack = (unsigned long *) (irq_stack_end[-1]); - printk(KERN_CONT " <EOI> "); + pr_cont(" <EOI> "); } } else { if (((long) stack & (THREAD_SIZE-1)) == 0) break; } if (i && ((i % STACKSLOTS_PER_LINE) == 0)) - printk(KERN_CONT "\n"); - printk(KERN_CONT " %016lx", *stack++); + pr_cont("\n"); + pr_cont(" %016lx", *stack++); touch_nmi_watchdog(); } preempt_enable(); - printk(KERN_CONT "\n"); + pr_cont("\n"); show_trace_log_lvl(task, regs, sp, bp, log_lvl); } -void show_registers(struct pt_regs *regs) +void show_regs(struct pt_regs *regs) { int i; unsigned long sp; - const int cpu = smp_processor_id(); - struct task_struct *cur = current; sp = regs->sp; - printk("CPU %d ", cpu); - print_modules(); + show_regs_print_info(KERN_DEFAULT); __show_regs(regs, 1); - printk("Process %s (pid: %d, threadinfo %p, task %p)\n", - cur->comm, cur->pid, task_thread_info(cur), cur); /* * When in-kernel, we also print out the stack and code at the @@ -284,16 +329,16 @@ void show_registers(struct pt_regs *regs) for (i = 0; i < code_len; i++, ip++) { if (ip < (u8 *)PAGE_OFFSET || probe_kernel_address(ip, c)) { - printk(KERN_CONT " Bad RIP value."); + pr_cont(" Bad RIP value."); break; } if (ip == (u8 *)regs->ip) - printk(KERN_CONT "<%02x> ", c); + pr_cont("<%02x> ", c); else - printk(KERN_CONT "%02x ", c); + pr_cont("%02x ", c); } } - printk(KERN_CONT "\n"); + pr_cont("\n"); } int is_valid_bugaddr(unsigned long ip) diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c index 62d61e9976e..988c00a1f60 100644 --- a/arch/x86/kernel/e820.c +++ b/arch/x86/kernel/e820.c @@ -113,7 +113,9 @@ static void __init __e820_add_region(struct e820map *e820x, u64 start, u64 size, int x = e820x->nr_map; if (x >= ARRAY_SIZE(e820x->map)) { - printk(KERN_ERR "Ooops! Too many entries in the memory map!\n"); + printk(KERN_ERR "e820: too many entries; ignoring [mem %#010llx-%#010llx]\n", + (unsigned long long) start, + (unsigned long long) (start + size - 1)); return; } @@ -133,19 +135,19 @@ static void __init e820_print_type(u32 type) switch (type) { case E820_RAM: case E820_RESERVED_KERN: - printk(KERN_CONT "(usable)"); + printk(KERN_CONT "usable"); break; case E820_RESERVED: - printk(KERN_CONT "(reserved)"); + printk(KERN_CONT "reserved"); break; case E820_ACPI: - printk(KERN_CONT "(ACPI data)"); + printk(KERN_CONT "ACPI data"); break; case E820_NVS: - printk(KERN_CONT "(ACPI NVS)"); + printk(KERN_CONT "ACPI NVS"); break; case E820_UNUSABLE: - printk(KERN_CONT "(unusable)"); + printk(KERN_CONT "unusable"); break; default: printk(KERN_CONT "type %u", type); @@ -158,10 +160,10 @@ void __init e820_print_map(char *who) int i; for (i = 0; i < e820.nr_map; i++) { - printk(KERN_INFO " %s: %016Lx - %016Lx ", who, + printk(KERN_INFO "%s: [mem %#018Lx-%#018Lx] ", who, (unsigned long long) e820.map[i].addr, (unsigned long long) - (e820.map[i].addr + e820.map[i].size)); + (e820.map[i].addr + e820.map[i].size - 1)); e820_print_type(e820.map[i].type); printk(KERN_CONT "\n"); } @@ -428,9 +430,8 @@ static u64 __init __e820_update_range(struct e820map *e820x, u64 start, size = ULLONG_MAX - start; end = start + size; - printk(KERN_DEBUG "e820 update range: %016Lx - %016Lx ", - (unsigned long long) start, - (unsigned long long) end); + printk(KERN_DEBUG "e820: update [mem %#010Lx-%#010Lx] ", + (unsigned long long) start, (unsigned long long) (end - 1)); e820_print_type(old_type); printk(KERN_CONT " ==> "); e820_print_type(new_type); @@ -509,9 +510,8 @@ u64 __init e820_remove_range(u64 start, u64 size, unsigned old_type, size = ULLONG_MAX - start; end = start + size; - printk(KERN_DEBUG "e820 remove range: %016Lx - %016Lx ", - (unsigned long long) start, - (unsigned long long) end); + printk(KERN_DEBUG "e820: remove [mem %#010Lx-%#010Lx] ", + (unsigned long long) start, (unsigned long long) (end - 1)); if (checktype) e820_print_type(old_type); printk(KERN_CONT "\n"); @@ -567,7 +567,7 @@ void __init update_e820(void) if (sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &nr_map)) return; e820.nr_map = nr_map; - printk(KERN_INFO "modified physical RAM map:\n"); + printk(KERN_INFO "e820: modified physical RAM map:\n"); e820_print_map("modified"); } static void __init update_e820_saved(void) @@ -637,8 +637,8 @@ __init void e820_setup_gap(void) if (!found) { gapstart = (max_pfn << PAGE_SHIFT) + 1024*1024; printk(KERN_ERR - "PCI: Warning: Cannot find a gap in the 32bit address range\n" - "PCI: Unassigned devices with 32bit resource registers may break!\n"); + "e820: cannot find a gap in the 32bit address range\n" + "e820: PCI devices with unassigned 32bit BARs may break!\n"); } #endif @@ -648,8 +648,8 @@ __init void e820_setup_gap(void) pci_mem_start = gapstart; printk(KERN_INFO - "Allocating PCI resources starting at %lx (gap: %lx:%lx)\n", - pci_mem_start, gapstart, gapsize); + "e820: [mem %#010lx-%#010lx] available for PCI devices\n", + gapstart, gapstart + gapsize - 1); } /** @@ -658,16 +658,19 @@ __init void e820_setup_gap(void) * boot_params.e820_map, others are passed via SETUP_E820_EXT node of * linked list of struct setup_data, which is parsed here. */ -void __init parse_e820_ext(struct setup_data *sdata) +void __init parse_e820_ext(u64 phys_addr, u32 data_len) { int entries; struct e820entry *extmap; + struct setup_data *sdata; + sdata = early_memremap(phys_addr, data_len); entries = sdata->len / sizeof(struct e820entry); extmap = (struct e820entry *)(sdata->data); __append_e820_map(extmap, entries); sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &e820.nr_map); - printk(KERN_INFO "extended physical RAM map:\n"); + early_iounmap(sdata, data_len); + printk(KERN_INFO "e820: extended physical RAM map:\n"); e820_print_map("extended"); } @@ -734,7 +737,7 @@ u64 __init early_reserve_e820(u64 size, u64 align) addr = __memblock_alloc_base(size, align, MEMBLOCK_ALLOC_ACCESSIBLE); if (addr) { e820_update_range_saved(addr, size, E820_RAM, E820_RESERVED); - printk(KERN_INFO "update e820_saved for early_reserve_e820\n"); + printk(KERN_INFO "e820: update e820_saved for early_reserve_e820\n"); update_e820_saved(); } @@ -784,7 +787,7 @@ static unsigned long __init e820_end_pfn(unsigned long limit_pfn, unsigned type) if (last_pfn > max_arch_pfn) last_pfn = max_arch_pfn; - printk(KERN_INFO "last_pfn = %#lx max_arch_pfn = %#lx\n", + printk(KERN_INFO "e820: last_pfn = %#lx max_arch_pfn = %#lx\n", last_pfn, max_arch_pfn); return last_pfn; } @@ -835,7 +838,7 @@ static int __init parse_memopt(char *p) } early_param("mem", parse_memopt); -static int __init parse_memmap_opt(char *p) +static int __init parse_memmap_one(char *p) { char *oldp; u64 start_at, mem_size; @@ -877,6 +880,20 @@ static int __init parse_memmap_opt(char *p) return *p == '\0' ? 0 : -EINVAL; } +static int __init parse_memmap_opt(char *str) +{ + while (str) { + char *k = strchr(str, ','); + + if (k) + *k++ = 0; + + parse_memmap_one(str); + str = k; + } + + return 0; +} early_param("memmap", parse_memmap_opt); void __init finish_e820_parsing(void) @@ -888,7 +905,7 @@ void __init finish_e820_parsing(void) early_panic("Invalid user supplied memory map"); e820.nr_map = nr; - printk(KERN_INFO "user-defined physical RAM map:\n"); + printk(KERN_INFO "e820: user-defined physical RAM map:\n"); e820_print_map("user"); } } @@ -944,7 +961,7 @@ void __init e820_reserve_resources(void) for (i = 0; i < e820_saved.nr_map; i++) { struct e820entry *entry = &e820_saved.map[i]; firmware_map_add_early(entry->addr, - entry->addr + entry->size - 1, + entry->addr + entry->size, e820_type_to_string(entry->type)); } } @@ -996,8 +1013,9 @@ void __init e820_reserve_resources_late(void) end = MAX_RESOURCE_SIZE; if (start >= end) continue; - printk(KERN_DEBUG "reserve RAM buffer: %016llx - %016llx ", - start, end); + printk(KERN_DEBUG + "e820: reserve RAM buffer [mem %#010llx-%#010llx]\n", + start, end); reserve_region_with_split(&iomem_resource, start, end, "RAM buffer"); } @@ -1047,7 +1065,7 @@ void __init setup_memory_map(void) who = x86_init.resources.memory_setup(); memcpy(&e820_saved, &e820, sizeof(struct e820map)); - printk(KERN_INFO "BIOS-provided physical RAM map:\n"); + printk(KERN_INFO "e820: BIOS-provided physical RAM map:\n"); e820_print_map(who); } @@ -1076,6 +1094,9 @@ void __init memblock_x86_fill(void) memblock_add(ei->addr, ei->size); } + /* throw away partial pages */ + memblock_trim_memory(PAGE_SIZE); + memblock_dump_all(); } @@ -1099,7 +1120,7 @@ void __init memblock_find_dma_reserve(void) nr_pages += end_pfn - start_pfn; } - for_each_free_mem_range(u, MAX_NUMNODES, &start, &end, NULL) { + for_each_free_mem_range(u, NUMA_NO_NODE, &start, &end, NULL) { start_pfn = min_t(unsigned long, PFN_UP(start), MAX_DMA_PFN); end_pfn = min_t(unsigned long, PFN_DOWN(end), MAX_DMA_PFN); if (start_pfn < end_pfn) diff --git a/arch/x86/kernel/early-quirks.c b/arch/x86/kernel/early-quirks.c index 3755ef49439..2e1a6853e00 100644 --- a/arch/x86/kernel/early-quirks.c +++ b/arch/x86/kernel/early-quirks.c @@ -12,12 +12,15 @@ #include <linux/pci.h> #include <linux/acpi.h> #include <linux/pci_ids.h> +#include <drm/i915_drm.h> #include <asm/pci-direct.h> #include <asm/dma.h> #include <asm/io_apic.h> #include <asm/apic.h> +#include <asm/hpet.h> #include <asm/iommu.h> #include <asm/gart.h> +#include <asm/irq_remapping.h> static void __init fix_hypertransport_config(int num, int slot, int func) { @@ -192,6 +195,377 @@ static void __init ati_bugs_contd(int num, int slot, int func) } #endif +static void __init intel_remapping_check(int num, int slot, int func) +{ + u8 revision; + u16 device; + + device = read_pci_config_16(num, slot, func, PCI_DEVICE_ID); + revision = read_pci_config_byte(num, slot, func, PCI_REVISION_ID); + + /* + * Revision <= 13 of all triggering devices id in this quirk + * have a problem draining interrupts when irq remapping is + * enabled, and should be flagged as broken. Additionally + * revision 0x22 of device id 0x3405 has this problem. + */ + if (revision <= 0x13) + set_irq_remapping_broken(); + else if (device == 0x3405 && revision == 0x22) + set_irq_remapping_broken(); +} + +/* + * Systems with Intel graphics controllers set aside memory exclusively + * for gfx driver use. This memory is not marked in the E820 as reserved + * or as RAM, and so is subject to overlap from E820 manipulation later + * in the boot process. On some systems, MMIO space is allocated on top, + * despite the efforts of the "RAM buffer" approach, which simply rounds + * memory boundaries up to 64M to try to catch space that may decode + * as RAM and so is not suitable for MMIO. + * + * And yes, so far on current devices the base addr is always under 4G. + */ +static u32 __init intel_stolen_base(int num, int slot, int func, size_t stolen_size) +{ + u32 base; + + /* + * For the PCI IDs in this quirk, the stolen base is always + * in 0x5c, aka the BDSM register (yes that's really what + * it's called). + */ + base = read_pci_config(num, slot, func, 0x5c); + base &= ~((1<<20) - 1); + + return base; +} + +#define KB(x) ((x) * 1024UL) +#define MB(x) (KB (KB (x))) +#define GB(x) (MB (KB (x))) + +static size_t __init i830_tseg_size(void) +{ + u8 tmp = read_pci_config_byte(0, 0, 0, I830_ESMRAMC); + + if (!(tmp & TSEG_ENABLE)) + return 0; + + if (tmp & I830_TSEG_SIZE_1M) + return MB(1); + else + return KB(512); +} + +static size_t __init i845_tseg_size(void) +{ + u8 tmp = read_pci_config_byte(0, 0, 0, I845_ESMRAMC); + + if (!(tmp & TSEG_ENABLE)) + return 0; + + switch (tmp & I845_TSEG_SIZE_MASK) { + case I845_TSEG_SIZE_512K: + return KB(512); + case I845_TSEG_SIZE_1M: + return MB(1); + default: + WARN_ON(1); + return 0; + } +} + +static size_t __init i85x_tseg_size(void) +{ + u8 tmp = read_pci_config_byte(0, 0, 0, I85X_ESMRAMC); + + if (!(tmp & TSEG_ENABLE)) + return 0; + + return MB(1); +} + +static size_t __init i830_mem_size(void) +{ + return read_pci_config_byte(0, 0, 0, I830_DRB3) * MB(32); +} + +static size_t __init i85x_mem_size(void) +{ + return read_pci_config_byte(0, 0, 1, I85X_DRB3) * MB(32); +} + +/* + * On 830/845/85x the stolen memory base isn't available in any + * register. We need to calculate it as TOM-TSEG_SIZE-stolen_size. + */ +static u32 __init i830_stolen_base(int num, int slot, int func, size_t stolen_size) +{ + return i830_mem_size() - i830_tseg_size() - stolen_size; +} + +static u32 __init i845_stolen_base(int num, int slot, int func, size_t stolen_size) +{ + return i830_mem_size() - i845_tseg_size() - stolen_size; +} + +static u32 __init i85x_stolen_base(int num, int slot, int func, size_t stolen_size) +{ + return i85x_mem_size() - i85x_tseg_size() - stolen_size; +} + +static u32 __init i865_stolen_base(int num, int slot, int func, size_t stolen_size) +{ + /* + * FIXME is the graphics stolen memory region + * always at TOUD? Ie. is it always the last + * one to be allocated by the BIOS? + */ + return read_pci_config_16(0, 0, 0, I865_TOUD) << 16; +} + +static size_t __init i830_stolen_size(int num, int slot, int func) +{ + size_t stolen_size; + u16 gmch_ctrl; + + gmch_ctrl = read_pci_config_16(0, 0, 0, I830_GMCH_CTRL); + + switch (gmch_ctrl & I830_GMCH_GMS_MASK) { + case I830_GMCH_GMS_STOLEN_512: + stolen_size = KB(512); + break; + case I830_GMCH_GMS_STOLEN_1024: + stolen_size = MB(1); + break; + case I830_GMCH_GMS_STOLEN_8192: + stolen_size = MB(8); + break; + case I830_GMCH_GMS_LOCAL: + /* local memory isn't part of the normal address space */ + stolen_size = 0; + break; + default: + return 0; + } + + return stolen_size; +} + +static size_t __init gen3_stolen_size(int num, int slot, int func) +{ + size_t stolen_size; + u16 gmch_ctrl; + + gmch_ctrl = read_pci_config_16(0, 0, 0, I830_GMCH_CTRL); + + switch (gmch_ctrl & I855_GMCH_GMS_MASK) { + case I855_GMCH_GMS_STOLEN_1M: + stolen_size = MB(1); + break; + case I855_GMCH_GMS_STOLEN_4M: + stolen_size = MB(4); + break; + case I855_GMCH_GMS_STOLEN_8M: + stolen_size = MB(8); + break; + case I855_GMCH_GMS_STOLEN_16M: + stolen_size = MB(16); + break; + case I855_GMCH_GMS_STOLEN_32M: + stolen_size = MB(32); + break; + case I915_GMCH_GMS_STOLEN_48M: + stolen_size = MB(48); + break; + case I915_GMCH_GMS_STOLEN_64M: + stolen_size = MB(64); + break; + case G33_GMCH_GMS_STOLEN_128M: + stolen_size = MB(128); + break; + case G33_GMCH_GMS_STOLEN_256M: + stolen_size = MB(256); + break; + case INTEL_GMCH_GMS_STOLEN_96M: + stolen_size = MB(96); + break; + case INTEL_GMCH_GMS_STOLEN_160M: + stolen_size = MB(160); + break; + case INTEL_GMCH_GMS_STOLEN_224M: + stolen_size = MB(224); + break; + case INTEL_GMCH_GMS_STOLEN_352M: + stolen_size = MB(352); + break; + default: + stolen_size = 0; + break; + } + + return stolen_size; +} + +static size_t __init gen6_stolen_size(int num, int slot, int func) +{ + u16 gmch_ctrl; + + gmch_ctrl = read_pci_config_16(num, slot, func, SNB_GMCH_CTRL); + gmch_ctrl >>= SNB_GMCH_GMS_SHIFT; + gmch_ctrl &= SNB_GMCH_GMS_MASK; + + return gmch_ctrl << 25; /* 32 MB units */ +} + +static size_t __init gen8_stolen_size(int num, int slot, int func) +{ + u16 gmch_ctrl; + + gmch_ctrl = read_pci_config_16(num, slot, func, SNB_GMCH_CTRL); + gmch_ctrl >>= BDW_GMCH_GMS_SHIFT; + gmch_ctrl &= BDW_GMCH_GMS_MASK; + return gmch_ctrl << 25; /* 32 MB units */ +} + +static size_t __init chv_stolen_size(int num, int slot, int func) +{ + u16 gmch_ctrl; + + gmch_ctrl = read_pci_config_16(num, slot, func, SNB_GMCH_CTRL); + gmch_ctrl >>= SNB_GMCH_GMS_SHIFT; + gmch_ctrl &= SNB_GMCH_GMS_MASK; + + /* + * 0x0 to 0x10: 32MB increments starting at 0MB + * 0x11 to 0x16: 4MB increments starting at 8MB + * 0x17 to 0x1d: 4MB increments start at 36MB + */ + if (gmch_ctrl < 0x11) + return gmch_ctrl << 25; + else if (gmch_ctrl < 0x17) + return (gmch_ctrl - 0x11 + 2) << 22; + else + return (gmch_ctrl - 0x17 + 9) << 22; +} + +struct intel_stolen_funcs { + size_t (*size)(int num, int slot, int func); + u32 (*base)(int num, int slot, int func, size_t size); +}; + +static const struct intel_stolen_funcs i830_stolen_funcs __initconst = { + .base = i830_stolen_base, + .size = i830_stolen_size, +}; + +static const struct intel_stolen_funcs i845_stolen_funcs __initconst = { + .base = i845_stolen_base, + .size = i830_stolen_size, +}; + +static const struct intel_stolen_funcs i85x_stolen_funcs __initconst = { + .base = i85x_stolen_base, + .size = gen3_stolen_size, +}; + +static const struct intel_stolen_funcs i865_stolen_funcs __initconst = { + .base = i865_stolen_base, + .size = gen3_stolen_size, +}; + +static const struct intel_stolen_funcs gen3_stolen_funcs __initconst = { + .base = intel_stolen_base, + .size = gen3_stolen_size, +}; + +static const struct intel_stolen_funcs gen6_stolen_funcs __initconst = { + .base = intel_stolen_base, + .size = gen6_stolen_size, +}; + +static const struct intel_stolen_funcs gen8_stolen_funcs __initconst = { + .base = intel_stolen_base, + .size = gen8_stolen_size, +}; + +static const struct intel_stolen_funcs chv_stolen_funcs __initconst = { + .base = intel_stolen_base, + .size = chv_stolen_size, +}; + +static const struct pci_device_id intel_stolen_ids[] __initconst = { + INTEL_I830_IDS(&i830_stolen_funcs), + INTEL_I845G_IDS(&i845_stolen_funcs), + INTEL_I85X_IDS(&i85x_stolen_funcs), + INTEL_I865G_IDS(&i865_stolen_funcs), + INTEL_I915G_IDS(&gen3_stolen_funcs), + INTEL_I915GM_IDS(&gen3_stolen_funcs), + INTEL_I945G_IDS(&gen3_stolen_funcs), + INTEL_I945GM_IDS(&gen3_stolen_funcs), + INTEL_VLV_M_IDS(&gen6_stolen_funcs), + INTEL_VLV_D_IDS(&gen6_stolen_funcs), + INTEL_PINEVIEW_IDS(&gen3_stolen_funcs), + INTEL_I965G_IDS(&gen3_stolen_funcs), + INTEL_G33_IDS(&gen3_stolen_funcs), + INTEL_I965GM_IDS(&gen3_stolen_funcs), + INTEL_GM45_IDS(&gen3_stolen_funcs), + INTEL_G45_IDS(&gen3_stolen_funcs), + INTEL_IRONLAKE_D_IDS(&gen3_stolen_funcs), + INTEL_IRONLAKE_M_IDS(&gen3_stolen_funcs), + INTEL_SNB_D_IDS(&gen6_stolen_funcs), + INTEL_SNB_M_IDS(&gen6_stolen_funcs), + INTEL_IVB_M_IDS(&gen6_stolen_funcs), + INTEL_IVB_D_IDS(&gen6_stolen_funcs), + INTEL_HSW_D_IDS(&gen6_stolen_funcs), + INTEL_HSW_M_IDS(&gen6_stolen_funcs), + INTEL_BDW_M_IDS(&gen8_stolen_funcs), + INTEL_BDW_D_IDS(&gen8_stolen_funcs), + INTEL_CHV_IDS(&chv_stolen_funcs), +}; + +static void __init intel_graphics_stolen(int num, int slot, int func) +{ + size_t size; + int i; + u32 start; + u16 device, subvendor, subdevice; + + device = read_pci_config_16(num, slot, func, PCI_DEVICE_ID); + subvendor = read_pci_config_16(num, slot, func, + PCI_SUBSYSTEM_VENDOR_ID); + subdevice = read_pci_config_16(num, slot, func, PCI_SUBSYSTEM_ID); + + for (i = 0; i < ARRAY_SIZE(intel_stolen_ids); i++) { + if (intel_stolen_ids[i].device == device) { + const struct intel_stolen_funcs *stolen_funcs = + (const struct intel_stolen_funcs *)intel_stolen_ids[i].driver_data; + size = stolen_funcs->size(num, slot, func); + start = stolen_funcs->base(num, slot, func, size); + if (size && start) { + printk(KERN_INFO "Reserving Intel graphics stolen memory at 0x%x-0x%x\n", + start, start + (u32)size - 1); + /* Mark this space as reserved */ + e820_add_region(start, size, E820_RESERVED); + sanitize_e820_map(e820.map, + ARRAY_SIZE(e820.map), + &e820.nr_map); + } + return; + } + } +} + +static void __init force_disable_hpet(int num, int slot, int func) +{ +#ifdef CONFIG_HPET_TIMER + boot_hpet_disable = 1; + pr_info("x86/hpet: Will disable the HPET for this platform because it's not reliable\n"); +#endif +} + + #define QFLAG_APPLY_ONCE 0x1 #define QFLAG_APPLIED 0x2 #define QFLAG_DONE (QFLAG_APPLY_ONCE|QFLAG_APPLIED) @@ -221,6 +595,20 @@ static struct chipset early_qrk[] __initdata = { PCI_CLASS_SERIAL_SMBUS, PCI_ANY_ID, 0, ati_bugs }, { PCI_VENDOR_ID_ATI, PCI_DEVICE_ID_ATI_SBX00_SMBUS, PCI_CLASS_SERIAL_SMBUS, PCI_ANY_ID, 0, ati_bugs_contd }, + { PCI_VENDOR_ID_INTEL, 0x3403, PCI_CLASS_BRIDGE_HOST, + PCI_BASE_CLASS_BRIDGE, 0, intel_remapping_check }, + { PCI_VENDOR_ID_INTEL, 0x3405, PCI_CLASS_BRIDGE_HOST, + PCI_BASE_CLASS_BRIDGE, 0, intel_remapping_check }, + { PCI_VENDOR_ID_INTEL, 0x3406, PCI_CLASS_BRIDGE_HOST, + PCI_BASE_CLASS_BRIDGE, 0, intel_remapping_check }, + { PCI_VENDOR_ID_INTEL, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA, PCI_ANY_ID, + QFLAG_APPLY_ONCE, intel_graphics_stolen }, + /* + * HPET on current version of Baytrail platform has accuracy + * problems, disable it for now: + */ + { PCI_VENDOR_ID_INTEL, 0x0f00, + PCI_CLASS_BRIDGE_HOST, PCI_ANY_ID, 0, force_disable_hpet}, {} }; diff --git a/arch/x86/kernel/early_printk.c b/arch/x86/kernel/early_printk.c index 9b9f18b4991..01d1c187c9f 100644 --- a/arch/x86/kernel/early_printk.c +++ b/arch/x86/kernel/early_printk.c @@ -14,9 +14,11 @@ #include <xen/hvc-console.h> #include <asm/pci-direct.h> #include <asm/fixmap.h> -#include <asm/mrst.h> +#include <asm/intel-mid.h> #include <asm/pgtable.h> #include <linux/usb/ehci_def.h> +#include <linux/efi.h> +#include <asm/efi.h> /* Simple VGA output */ #define VGABASE (__ISA_IO_base + 0xb8000) @@ -169,25 +171,9 @@ static struct console early_serial_console = { .index = -1, }; -/* Direct interface for emergencies */ -static struct console *early_console = &early_vga_console; -static int __initdata early_console_initialized; - -asmlinkage void early_printk(const char *fmt, ...) -{ - char buf[512]; - int n; - va_list ap; - - va_start(ap, fmt); - n = vscnprintf(buf, sizeof(buf), fmt, ap); - early_console->write(early_console, buf, n); - va_end(ap); -} - static inline void early_console_register(struct console *con, int keep_early) { - if (early_console->index != -1) { + if (con->index != -1) { printk(KERN_CRIT "ERROR: earlyprintk= %s already used\n", con->name); return; @@ -207,9 +193,8 @@ static int __init setup_early_printk(char *buf) if (!buf) return 0; - if (early_console_initialized) + if (early_console) return 0; - early_console_initialized = 1; keep = (strstr(buf, "keep") != NULL); @@ -251,6 +236,11 @@ static int __init setup_early_printk(char *buf) early_console_register(&early_hsu_console, keep); } #endif +#ifdef CONFIG_EARLY_PRINTK_EFI + if (!strncmp(buf, "efi", 3)) + early_console_register(&early_efi_console, keep); +#endif + buf++; } return 0; diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S index 79d97e68f04..0d0c9d4ab6d 100644 --- a/arch/x86/kernel/entry_32.S +++ b/arch/x86/kernel/entry_32.S @@ -56,6 +56,8 @@ #include <asm/irq_vectors.h> #include <asm/cpufeature.h> #include <asm/alternative-asm.h> +#include <asm/asm.h> +#include <asm/smap.h> /* Avoid __ASSEMBLER__'ifying <linux/audit.h> just for this. */ #include <linux/elf-em.h> @@ -98,12 +100,6 @@ #endif .endm -#ifdef CONFIG_VM86 -#define resume_userspace_sig check_userspace -#else -#define resume_userspace_sig resume_userspace -#endif - /* * User gs save/restore * @@ -157,10 +153,8 @@ .pushsection .fixup, "ax" 99: movl $0, (%esp) jmp 98b -.section __ex_table, "a" - .align 4 - .long 98b, 99b .popsection + _ASM_EXTABLE(98b,99b) .endm .macro PTGS_TO_GS @@ -170,10 +164,8 @@ .pushsection .fixup, "ax" 99: movl $0, PT_GS(%esp) jmp 98b -.section __ex_table, "a" - .align 4 - .long 98b, 99b .popsection + _ASM_EXTABLE(98b,99b) .endm .macro GS_TO_REG reg @@ -255,12 +247,10 @@ jmp 2b 6: movl $0, (%esp) jmp 3b -.section __ex_table, "a" - .align 4 - .long 1b, 4b - .long 2b, 5b - .long 3b, 6b .popsection + _ASM_EXTABLE(1b,4b) + _ASM_EXTABLE(2b,5b) + _ASM_EXTABLE(3b,6b) POP_GS_EX .endm @@ -309,10 +299,21 @@ ENTRY(ret_from_fork) CFI_ENDPROC END(ret_from_fork) -/* - * Interrupt exit functions should be protected against kprobes - */ - .pushsection .kprobes.text, "ax" +ENTRY(ret_from_kernel_thread) + CFI_STARTPROC + pushl_cfi %eax + call schedule_tail + GET_THREAD_INFO(%ebp) + popl_cfi %eax + pushl_cfi $0x0202 # Reset kernel eflags + popfl_cfi + movl PT_EBP(%esp),%eax + call *PT_EBX(%esp) + movl $0,PT_EAX(%esp) + jmp syscall_exit + CFI_ENDPROC +ENDPROC(ret_from_kernel_thread) + /* * Return to user mode is not as complex as all this looks, * but we want the default path for a system call return to @@ -327,10 +328,17 @@ ret_from_exception: preempt_stop(CLBR_ANY) ret_from_intr: GET_THREAD_INFO(%ebp) -check_userspace: +#ifdef CONFIG_VM86 movl PT_EFLAGS(%esp), %eax # mix EFLAGS and CS movb PT_CS(%esp), %al andl $(X86_EFLAGS_VM | SEGMENT_RPL_MASK), %eax +#else + /* + * We can be coming here from child spawned by kernel_thread(). + */ + movl PT_CS(%esp), %eax + andl $SEGMENT_RPL_MASK, %eax +#endif cmpl $USER_RPL, %eax jb resume_kernel # not returning to v8086 or userspace @@ -350,12 +358,9 @@ END(ret_from_exception) #ifdef CONFIG_PREEMPT ENTRY(resume_kernel) DISABLE_INTERRUPTS(CLBR_ANY) - cmpl $0,TI_preempt_count(%ebp) # non-zero preempt_count ? - jnz restore_all need_resched: - movl TI_flags(%ebp), %ecx # need_resched set ? - testb $_TIF_NEED_RESCHED, %cl - jz restore_all + cmpl $0,PER_CPU_VAR(__preempt_count) + jnz restore_all testl $X86_EFLAGS_IF,PT_EFLAGS(%esp) # interrupts off (exception path) ? jz restore_all call preempt_schedule_irq @@ -363,10 +368,6 @@ need_resched: END(resume_kernel) #endif CFI_ENDPROC -/* - * End of kprobes section - */ - .popsection /* SYSENTER_RETURN points to after the "sysenter" instruction in the vsyscall page. See vsyscall-sysentry.S, which defines the symbol. */ @@ -410,12 +411,11 @@ sysenter_past_esp: */ cmpl $__PAGE_OFFSET-3,%ebp jae syscall_fault + ASM_STAC 1: movl (%ebp),%ebp + ASM_CLAC movl %ebp,PT_EBP(%esp) -.section __ex_table,"a" - .align 4 - .long 1b,syscall_fault -.previous + _ASM_EXTABLE(1b,syscall_fault) GET_THREAD_INFO(%ebp) @@ -423,8 +423,9 @@ sysenter_past_esp: jnz sysenter_audit sysenter_do_call: cmpl $(NR_syscalls), %eax - jae syscall_badsys + jae sysenter_badsys call *sys_call_table(,%eax,4) +sysenter_after_call: movl %eax,PT_EAX(%esp) LOCKDEP_SYS_EXIT DISABLE_INTERRUPTS(CLBR_ANY) @@ -482,20 +483,15 @@ sysexit_audit: .pushsection .fixup,"ax" 2: movl $0,PT_FS(%esp) jmp 1b -.section __ex_table,"a" - .align 4 - .long 1b,2b .popsection + _ASM_EXTABLE(1b,2b) PTGS_TO_GS_EX ENDPROC(ia32_sysenter_target) -/* - * syscall stub including irq exit should be protected against kprobes - */ - .pushsection .kprobes.text, "ax" # system call handler stub ENTRY(system_call) RING0_INT_FRAME # can't unwind into user space anyway + ASM_CLAC pushl_cfi %eax # save orig_eax SAVE_ALL GET_THREAD_INFO(%ebp) @@ -506,6 +502,7 @@ ENTRY(system_call) jae syscall_badsys syscall_call: call *sys_call_table(,%eax,4) +syscall_after_call: movl %eax,PT_EAX(%esp) # store the return value syscall_exit: LOCKDEP_SYS_EXIT @@ -520,6 +517,7 @@ syscall_exit: restore_all: TRACE_IRQS_IRET restore_all_notrace: +#ifdef CONFIG_X86_ESPFIX32 movl PT_EFLAGS(%esp), %eax # mix EFLAGS, SS and CS # Warning: PT_OLDSS(%esp) contains the wrong/random values if we # are returning to the kernel. @@ -530,6 +528,7 @@ restore_all_notrace: cmpl $((SEGMENT_LDT << 8) | USER_RPL), %eax CFI_REMEMBER_STATE je ldt_ss # returning to user-space with LDT SS +#endif restore_nocheck: RESTORE_REGS 4 # skip orig_eax/error_code irq_return: @@ -540,18 +539,11 @@ ENTRY(iret_exc) pushl $do_iret_error jmp error_code .previous -.section __ex_table,"a" - .align 4 - .long irq_return,iret_exc -.previous + _ASM_EXTABLE(irq_return,iret_exc) +#ifdef CONFIG_X86_ESPFIX32 CFI_RESTORE_STATE ldt_ss: - larl PT_OLDSS(%esp), %eax - jnz restore_nocheck - testl $0x00400000, %eax # returning to 32bit stack? - jnz restore_nocheck # allright, normal return - #ifdef CONFIG_PARAVIRT /* * The kernel can't run on a non-flat stack if paravirt mode @@ -593,6 +585,7 @@ ldt_ss: lss (%esp), %esp /* switch to espfix segment */ CFI_ADJUST_CFA_OFFSET -8 jmp restore_nocheck +#endif CFI_ENDPROC ENDPROC(system_call) @@ -623,26 +616,29 @@ work_notifysig: # deal with pending signals and movl %esp, %eax jne work_notifysig_v86 # returning to kernel-space or # vm86-space +1: +#else + movl %esp, %eax +#endif TRACE_IRQS_ON ENABLE_INTERRUPTS(CLBR_NONE) + movb PT_CS(%esp), %bl + andb $SEGMENT_RPL_MASK, %bl + cmpb $USER_RPL, %bl + jb resume_kernel xorl %edx, %edx call do_notify_resume - jmp resume_userspace_sig + jmp resume_userspace +#ifdef CONFIG_VM86 ALIGN work_notifysig_v86: pushl_cfi %ecx # save ti_flags for do_notify_resume call save_v86_state # %eax contains pt_regs pointer popl_cfi %ecx movl %eax, %esp -#else - movl %esp, %eax + jmp 1b #endif - TRACE_IRQS_ON - ENABLE_INTERRUPTS(CLBR_NONE) - xorl %edx, %edx - call do_notify_resume - jmp resume_userspace_sig END(work_pending) # perform syscall exit tracing @@ -673,85 +669,22 @@ END(syscall_exit_work) RING0_INT_FRAME # can't unwind into user space anyway syscall_fault: + ASM_CLAC GET_THREAD_INFO(%ebp) movl $-EFAULT,PT_EAX(%esp) jmp resume_userspace END(syscall_fault) syscall_badsys: - movl $-ENOSYS,PT_EAX(%esp) - jmp resume_userspace + movl $-ENOSYS,%eax + jmp syscall_after_call END(syscall_badsys) - CFI_ENDPROC -/* - * End of kprobes section - */ - .popsection -/* - * System calls that need a pt_regs pointer. - */ -#define PTREGSCALL0(name) \ -ENTRY(ptregs_##name) ; \ - leal 4(%esp),%eax; \ - jmp sys_##name; \ -ENDPROC(ptregs_##name) - -#define PTREGSCALL1(name) \ -ENTRY(ptregs_##name) ; \ - leal 4(%esp),%edx; \ - movl (PT_EBX+4)(%esp),%eax; \ - jmp sys_##name; \ -ENDPROC(ptregs_##name) - -#define PTREGSCALL2(name) \ -ENTRY(ptregs_##name) ; \ - leal 4(%esp),%ecx; \ - movl (PT_ECX+4)(%esp),%edx; \ - movl (PT_EBX+4)(%esp),%eax; \ - jmp sys_##name; \ -ENDPROC(ptregs_##name) - -#define PTREGSCALL3(name) \ -ENTRY(ptregs_##name) ; \ - CFI_STARTPROC; \ - leal 4(%esp),%eax; \ - pushl_cfi %eax; \ - movl PT_EDX(%eax),%ecx; \ - movl PT_ECX(%eax),%edx; \ - movl PT_EBX(%eax),%eax; \ - call sys_##name; \ - addl $4,%esp; \ - CFI_ADJUST_CFA_OFFSET -4; \ - ret; \ - CFI_ENDPROC; \ -ENDPROC(ptregs_##name) - -PTREGSCALL1(iopl) -PTREGSCALL0(fork) -PTREGSCALL0(vfork) -PTREGSCALL3(execve) -PTREGSCALL2(sigaltstack) -PTREGSCALL0(sigreturn) -PTREGSCALL0(rt_sigreturn) -PTREGSCALL2(vm86) -PTREGSCALL1(vm86old) - -/* Clone is an oddball. The 4th arg is in %edi */ -ENTRY(ptregs_clone) - CFI_STARTPROC - leal 4(%esp),%eax - pushl_cfi %eax - pushl_cfi PT_EDI(%eax) - movl PT_EDX(%eax),%ecx - movl PT_ECX(%eax),%edx - movl PT_EBX(%eax),%eax - call sys_clone - addl $8,%esp - CFI_ADJUST_CFA_OFFSET -8 - ret +sysenter_badsys: + movl $-ENOSYS,%eax + jmp sysenter_after_call +END(syscall_badsys) CFI_ENDPROC -ENDPROC(ptregs_clone) .macro FIXUP_ESPFIX_STACK /* @@ -761,6 +694,7 @@ ENDPROC(ptregs_clone) * the high word of the segment base from the GDT and swiches to the * normal stack and adjusts ESP with the matching offset. */ +#ifdef CONFIG_X86_ESPFIX32 /* fixup the stack */ mov GDT_ESPFIX_SS + 4, %al /* bits 16..23 */ mov GDT_ESPFIX_SS + 7, %ah /* bits 24..31 */ @@ -770,8 +704,10 @@ ENDPROC(ptregs_clone) pushl_cfi %eax lss (%esp), %esp /* switch to the normal stack segment */ CFI_ADJUST_CFA_OFFSET -8 +#endif .endm .macro UNWIND_ESPFIX_STACK +#ifdef CONFIG_X86_ESPFIX32 movl %ss, %eax /* see if on espfix stack */ cmpw $__ESPFIX_SS, %ax @@ -782,6 +718,7 @@ ENDPROC(ptregs_clone) /* switch to normal stack */ FIXUP_ESPFIX_STACK 27: +#endif .endm /* @@ -828,6 +765,7 @@ END(interrupt) */ .p2align CONFIG_X86_L1_CACHE_SHIFT common_interrupt: + ASM_CLAC addl $-0x80,(%esp) /* Adjust vector into the [-256,-1] range */ SAVE_ALL TRACE_IRQS_OFF @@ -837,13 +775,10 @@ common_interrupt: ENDPROC(common_interrupt) CFI_ENDPROC -/* - * Irq entries should be protected against kprobes - */ - .pushsection .kprobes.text, "ax" #define BUILD_INTERRUPT3(name, nr, fn) \ ENTRY(name) \ RING0_INT_FRAME; \ + ASM_CLAC; \ pushl_cfi $~(nr); \ SAVE_ALL; \ TRACE_IRQS_OFF \ @@ -853,13 +788,24 @@ ENTRY(name) \ CFI_ENDPROC; \ ENDPROC(name) -#define BUILD_INTERRUPT(name, nr) BUILD_INTERRUPT3(name, nr, smp_##name) + +#ifdef CONFIG_TRACING +#define TRACE_BUILD_INTERRUPT(name, nr) \ + BUILD_INTERRUPT3(trace_##name, nr, smp_trace_##name) +#else +#define TRACE_BUILD_INTERRUPT(name, nr) +#endif + +#define BUILD_INTERRUPT(name, nr) \ + BUILD_INTERRUPT3(name, nr, smp_##name); \ + TRACE_BUILD_INTERRUPT(name, nr) /* The include is where all of the SMP etc. interrupts come from */ #include <asm/entry_arch.h> ENTRY(coprocessor_error) RING0_INT_FRAME + ASM_CLAC pushl_cfi $0 pushl_cfi $do_coprocessor_error jmp error_code @@ -868,6 +814,7 @@ END(coprocessor_error) ENTRY(simd_coprocessor_error) RING0_INT_FRAME + ASM_CLAC pushl_cfi $0 #ifdef CONFIG_X86_INVD_BUG /* AMD 486 bug: invd from userspace calls exception 19 instead of #GP */ @@ -889,6 +836,7 @@ END(simd_coprocessor_error) ENTRY(device_not_available) RING0_INT_FRAME + ASM_CLAC pushl_cfi $-1 # mark this as an int pushl_cfi $do_device_not_available jmp error_code @@ -898,10 +846,7 @@ END(device_not_available) #ifdef CONFIG_PARAVIRT ENTRY(native_iret) iret -.section __ex_table,"a" - .align 4 - .long native_iret, iret_exc -.previous + _ASM_EXTABLE(native_iret, iret_exc) END(native_iret) ENTRY(native_irq_enable_sysexit) @@ -912,6 +857,7 @@ END(native_irq_enable_sysexit) ENTRY(overflow) RING0_INT_FRAME + ASM_CLAC pushl_cfi $0 pushl_cfi $do_overflow jmp error_code @@ -920,6 +866,7 @@ END(overflow) ENTRY(bounds) RING0_INT_FRAME + ASM_CLAC pushl_cfi $0 pushl_cfi $do_bounds jmp error_code @@ -928,6 +875,7 @@ END(bounds) ENTRY(invalid_op) RING0_INT_FRAME + ASM_CLAC pushl_cfi $0 pushl_cfi $do_invalid_op jmp error_code @@ -936,6 +884,7 @@ END(invalid_op) ENTRY(coprocessor_segment_overrun) RING0_INT_FRAME + ASM_CLAC pushl_cfi $0 pushl_cfi $do_coprocessor_segment_overrun jmp error_code @@ -944,6 +893,7 @@ END(coprocessor_segment_overrun) ENTRY(invalid_TSS) RING0_EC_FRAME + ASM_CLAC pushl_cfi $do_invalid_TSS jmp error_code CFI_ENDPROC @@ -951,6 +901,7 @@ END(invalid_TSS) ENTRY(segment_not_present) RING0_EC_FRAME + ASM_CLAC pushl_cfi $do_segment_not_present jmp error_code CFI_ENDPROC @@ -958,6 +909,7 @@ END(segment_not_present) ENTRY(stack_segment) RING0_EC_FRAME + ASM_CLAC pushl_cfi $do_stack_segment jmp error_code CFI_ENDPROC @@ -965,6 +917,7 @@ END(stack_segment) ENTRY(alignment_check) RING0_EC_FRAME + ASM_CLAC pushl_cfi $do_alignment_check jmp error_code CFI_ENDPROC @@ -972,6 +925,7 @@ END(alignment_check) ENTRY(divide_error) RING0_INT_FRAME + ASM_CLAC pushl_cfi $0 # no error code pushl_cfi $do_divide_error jmp error_code @@ -981,6 +935,7 @@ END(divide_error) #ifdef CONFIG_X86_MCE ENTRY(machine_check) RING0_INT_FRAME + ASM_CLAC pushl_cfi $0 pushl_cfi machine_check_vector jmp error_code @@ -990,25 +945,12 @@ END(machine_check) ENTRY(spurious_interrupt_bug) RING0_INT_FRAME + ASM_CLAC pushl_cfi $0 pushl_cfi $do_spurious_interrupt_bug jmp error_code CFI_ENDPROC END(spurious_interrupt_bug) -/* - * End of kprobes section - */ - .popsection - -ENTRY(kernel_thread_helper) - pushl $0 # fake return address for unwinder - CFI_STARTPROC - movl %edi,%eax - call *%esi - call do_exit - ud2 # padding for call trace - CFI_ENDPROC -ENDPROC(kernel_thread_helper) #ifdef CONFIG_XEN /* Xen doesn't set %esp to be precisely what the normal sysenter @@ -1022,7 +964,7 @@ ENTRY(xen_sysenter_target) ENTRY(xen_hypervisor_callback) CFI_STARTPROC - pushl_cfi $0 + pushl_cfi $-1 /* orig_ax = -1 => not a system call */ SAVE_ALL TRACE_IRQS_OFF @@ -1064,14 +1006,15 @@ ENTRY(xen_failsafe_callback) 2: mov 8(%esp),%es 3: mov 12(%esp),%fs 4: mov 16(%esp),%gs + /* EAX == 0 => Category 1 (Bad segment) + EAX != 0 => Category 2 (Bad IRET) */ testl %eax,%eax popl_cfi %eax lea 16(%esp),%esp CFI_ADJUST_CFA_OFFSET -16 jz 5f - addl $16,%esp - jmp iret_exc # EAX != 0 => Category 2 (Bad IRET) -5: pushl_cfi $0 # EAX == 0 => Category 1 (Bad segment) + jmp iret_exc +5: pushl_cfi $-1 /* orig_ax = -1 => not a system call */ SAVE_ALL jmp ret_from_exception CFI_ENDPROC @@ -1090,20 +1033,24 @@ ENTRY(xen_failsafe_callback) movl %eax,16(%esp) jmp 4b .previous -.section __ex_table,"a" - .align 4 - .long 1b,6b - .long 2b,7b - .long 3b,8b - .long 4b,9b -.previous + _ASM_EXTABLE(1b,6b) + _ASM_EXTABLE(2b,7b) + _ASM_EXTABLE(3b,8b) + _ASM_EXTABLE(4b,9b) ENDPROC(xen_failsafe_callback) -BUILD_INTERRUPT3(xen_hvm_callback_vector, XEN_HVM_EVTCHN_CALLBACK, +BUILD_INTERRUPT3(xen_hvm_callback_vector, HYPERVISOR_CALLBACK_VECTOR, xen_evtchn_do_upcall) #endif /* CONFIG_XEN */ +#if IS_ENABLED(CONFIG_HYPERV) + +BUILD_INTERRUPT3(hyperv_callback_vector, HYPERVISOR_CALLBACK_VECTOR, + hyperv_vector_handler) + +#endif /* CONFIG_HYPERV */ + #ifdef CONFIG_FUNCTION_TRACER #ifdef CONFIG_DYNAMIC_FTRACE @@ -1118,17 +1065,21 @@ ENTRY(ftrace_caller) pushl %eax pushl %ecx pushl %edx - movl 0xc(%esp), %eax + pushl $0 /* Pass NULL as regs pointer */ + movl 4*4(%esp), %eax movl 0x4(%ebp), %edx + movl function_trace_op, %ecx subl $MCOUNT_INSN_SIZE, %eax .globl ftrace_call ftrace_call: call ftrace_stub + addl $4,%esp /* skip NULL pointer */ popl %edx popl %ecx popl %eax +ftrace_ret: #ifdef CONFIG_FUNCTION_GRAPH_TRACER .globl ftrace_graph_call ftrace_graph_call: @@ -1140,9 +1091,77 @@ ftrace_stub: ret END(ftrace_caller) +ENTRY(ftrace_regs_caller) + pushf /* push flags before compare (in cs location) */ + cmpl $0, function_trace_stop + jne ftrace_restore_flags + + /* + * i386 does not save SS and ESP when coming from kernel. + * Instead, to get sp, ®s->sp is used (see ptrace.h). + * Unfortunately, that means eflags must be at the same location + * as the current return ip is. We move the return ip into the + * ip location, and move flags into the return ip location. + */ + pushl 4(%esp) /* save return ip into ip slot */ + + pushl $0 /* Load 0 into orig_ax */ + pushl %gs + pushl %fs + pushl %es + pushl %ds + pushl %eax + pushl %ebp + pushl %edi + pushl %esi + pushl %edx + pushl %ecx + pushl %ebx + + movl 13*4(%esp), %eax /* Get the saved flags */ + movl %eax, 14*4(%esp) /* Move saved flags into regs->flags location */ + /* clobbering return ip */ + movl $__KERNEL_CS,13*4(%esp) + + movl 12*4(%esp), %eax /* Load ip (1st parameter) */ + subl $MCOUNT_INSN_SIZE, %eax /* Adjust ip */ + movl 0x4(%ebp), %edx /* Load parent ip (2nd parameter) */ + movl function_trace_op, %ecx /* Save ftrace_pos in 3rd parameter */ + pushl %esp /* Save pt_regs as 4th parameter */ + +GLOBAL(ftrace_regs_call) + call ftrace_stub + + addl $4, %esp /* Skip pt_regs */ + movl 14*4(%esp), %eax /* Move flags back into cs */ + movl %eax, 13*4(%esp) /* Needed to keep addl from modifying flags */ + movl 12*4(%esp), %eax /* Get return ip from regs->ip */ + movl %eax, 14*4(%esp) /* Put return ip back for ret */ + + popl %ebx + popl %ecx + popl %edx + popl %esi + popl %edi + popl %ebp + popl %eax + popl %ds + popl %es + popl %fs + popl %gs + addl $8, %esp /* Skip orig_ax and ip */ + popf /* Pop flags at end (no addl to corrupt flags) */ + jmp ftrace_ret + +ftrace_restore_flags: + popf + jmp ftrace_stub #else /* ! CONFIG_DYNAMIC_FTRACE */ ENTRY(mcount) + cmpl $__PAGE_OFFSET, %esp + jb ftrace_stub /* Paging not enabled yet? */ + cmpl $0, function_trace_stop jne ftrace_stub @@ -1180,9 +1199,6 @@ END(mcount) #ifdef CONFIG_FUNCTION_GRAPH_TRACER ENTRY(ftrace_graph_caller) - cmpl $0, function_trace_stop - jne ftrace_stub - pushl %eax pushl %ecx pushl %edx @@ -1209,13 +1225,19 @@ return_to_handler: jmp *%ecx #endif -/* - * Some functions should be protected against kprobes - */ - .pushsection .kprobes.text, "ax" +#ifdef CONFIG_TRACING +ENTRY(trace_page_fault) + RING0_EC_FRAME + ASM_CLAC + pushl_cfi $trace_do_page_fault + jmp error_code + CFI_ENDPROC +END(trace_page_fault) +#endif ENTRY(page_fault) RING0_EC_FRAME + ASM_CLAC pushl_cfi $do_page_fault ALIGN error_code: @@ -1288,6 +1310,7 @@ END(page_fault) ENTRY(debug) RING0_INT_FRAME + ASM_CLAC cmpl $ia32_sysenter_target,(%esp) jne debug_stack_correct FIX_STACK 12, debug_stack_correct, debug_esp_fix_insn @@ -1312,11 +1335,14 @@ END(debug) */ ENTRY(nmi) RING0_INT_FRAME + ASM_CLAC +#ifdef CONFIG_X86_ESPFIX32 pushl_cfi %eax movl %ss, %eax cmpw $__ESPFIX_SS, %ax popl_cfi %eax je nmi_espfix_stack +#endif cmpl $ia32_sysenter_target,(%esp) je nmi_stack_fixup pushl_cfi %eax @@ -1356,6 +1382,7 @@ nmi_debug_stack_check: FIX_STACK 24, nmi_stack_correct, 1 jmp nmi_stack_correct +#ifdef CONFIG_X86_ESPFIX32 nmi_espfix_stack: /* We have a RING0_INT_FRAME here. * @@ -1377,11 +1404,13 @@ nmi_espfix_stack: lss 12+4(%esp), %esp # back to espfix stack CFI_ADJUST_CFA_OFFSET -24 jmp irq_return +#endif CFI_ENDPROC END(nmi) ENTRY(int3) RING0_INT_FRAME + ASM_CLAC pushl_cfi $-1 # mark this as an int SAVE_ALL TRACE_IRQS_OFF @@ -1402,13 +1431,10 @@ END(general_protection) #ifdef CONFIG_KVM_GUEST ENTRY(async_page_fault) RING0_EC_FRAME + ASM_CLAC pushl_cfi $do_async_page_fault jmp error_code CFI_ENDPROC END(async_page_fault) #endif -/* - * End of kprobes section - */ - .popsection diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S index 3fe8239fd8f..c844f0816ab 100644 --- a/arch/x86/kernel/entry_64.S +++ b/arch/x86/kernel/entry_64.S @@ -36,7 +36,7 @@ * - FIXUP_TOP_OF_STACK/RESTORE_TOP_OF_STACK - Fix up the hardware stack * frame that is otherwise undefined after a SYSCALL * - TRACE_IRQ_* - Trace hard interrupt state for lock debugging. - * - errorentry/paranoidentry/zeroentry - Define exception entry points. + * - idtentry - Define exception entry points. */ #include <linux/linkage.h> @@ -53,8 +53,11 @@ #include <asm/page_types.h> #include <asm/irqflags.h> #include <asm/paravirt.h> -#include <asm/ftrace.h> #include <asm/percpu.h> +#include <asm/asm.h> +#include <asm/context_tracking.h> +#include <asm/smap.h> +#include <asm/pgtable_types.h> #include <linux/err.h> /* Avoid __ASSEMBLER__'ifying <linux/audit.h> just for this. */ @@ -66,107 +69,6 @@ .code64 .section .entry.text, "ax" -#ifdef CONFIG_FUNCTION_TRACER -#ifdef CONFIG_DYNAMIC_FTRACE -ENTRY(mcount) - retq -END(mcount) - -ENTRY(ftrace_caller) - cmpl $0, function_trace_stop - jne ftrace_stub - - MCOUNT_SAVE_FRAME - - movq 0x38(%rsp), %rdi - movq 8(%rbp), %rsi - subq $MCOUNT_INSN_SIZE, %rdi - -GLOBAL(ftrace_call) - call ftrace_stub - - MCOUNT_RESTORE_FRAME - -#ifdef CONFIG_FUNCTION_GRAPH_TRACER -GLOBAL(ftrace_graph_call) - jmp ftrace_stub -#endif - -GLOBAL(ftrace_stub) - retq -END(ftrace_caller) - -#else /* ! CONFIG_DYNAMIC_FTRACE */ -ENTRY(mcount) - cmpl $0, function_trace_stop - jne ftrace_stub - - cmpq $ftrace_stub, ftrace_trace_function - jnz trace - -#ifdef CONFIG_FUNCTION_GRAPH_TRACER - cmpq $ftrace_stub, ftrace_graph_return - jnz ftrace_graph_caller - - cmpq $ftrace_graph_entry_stub, ftrace_graph_entry - jnz ftrace_graph_caller -#endif - -GLOBAL(ftrace_stub) - retq - -trace: - MCOUNT_SAVE_FRAME - - movq 0x38(%rsp), %rdi - movq 8(%rbp), %rsi - subq $MCOUNT_INSN_SIZE, %rdi - - call *ftrace_trace_function - - MCOUNT_RESTORE_FRAME - - jmp ftrace_stub -END(mcount) -#endif /* CONFIG_DYNAMIC_FTRACE */ -#endif /* CONFIG_FUNCTION_TRACER */ - -#ifdef CONFIG_FUNCTION_GRAPH_TRACER -ENTRY(ftrace_graph_caller) - cmpl $0, function_trace_stop - jne ftrace_stub - - MCOUNT_SAVE_FRAME - - leaq 8(%rbp), %rdi - movq 0x38(%rsp), %rsi - movq (%rbp), %rdx - subq $MCOUNT_INSN_SIZE, %rsi - - call prepare_ftrace_return - - MCOUNT_RESTORE_FRAME - - retq -END(ftrace_graph_caller) - -GLOBAL(return_to_handler) - subq $24, %rsp - - /* Save the return values */ - movq %rax, (%rsp) - movq %rdx, 8(%rsp) - movq %rbp, %rdi - - call ftrace_return_to_handler - - movq %rax, %rdi - movq 8(%rsp), %rdx - movq (%rsp), %rax - addq $24, %rsp - jmp *%rdi -#endif - #ifndef CONFIG_PREEMPT #define retint_kernel retint_restore_args @@ -190,6 +92,44 @@ ENDPROC(native_usergs_sysret64) .endm /* + * When dynamic function tracer is enabled it will add a breakpoint + * to all locations that it is about to modify, sync CPUs, update + * all the code, sync CPUs, then remove the breakpoints. In this time + * if lockdep is enabled, it might jump back into the debug handler + * outside the updating of the IST protection. (TRACE_IRQS_ON/OFF). + * + * We need to change the IDT table before calling TRACE_IRQS_ON/OFF to + * make sure the stack pointer does not get reset back to the top + * of the debug stack, and instead just reuses the current stack. + */ +#if defined(CONFIG_DYNAMIC_FTRACE) && defined(CONFIG_TRACE_IRQFLAGS) + +.macro TRACE_IRQS_OFF_DEBUG + call debug_stack_set_zero + TRACE_IRQS_OFF + call debug_stack_reset +.endm + +.macro TRACE_IRQS_ON_DEBUG + call debug_stack_set_zero + TRACE_IRQS_ON + call debug_stack_reset +.endm + +.macro TRACE_IRQS_IRETQ_DEBUG offset=ARGOFFSET + bt $9,EFLAGS-\offset(%rsp) /* interrupts off? */ + jnc 1f + TRACE_IRQS_ON_DEBUG +1: +.endm + +#else +# define TRACE_IRQS_OFF_DEBUG TRACE_IRQS_OFF +# define TRACE_IRQS_ON_DEBUG TRACE_IRQS_ON +# define TRACE_IRQS_IRETQ_DEBUG TRACE_IRQS_IRETQ +#endif + +/* * C code is not supposed to know about undefined top of stack. Every time * a C function with an pt_regs argument is called from the SYSCALL based * fast path FIXUP_TOP_OF_STACK is needed. @@ -222,7 +162,7 @@ ENDPROC(native_usergs_sysret64) /*CFI_REL_OFFSET ss,0*/ pushq_cfi %rax /* rsp */ CFI_REL_OFFSET rsp,0 - pushq_cfi $(X86_EFLAGS_IF|X86_EFLAGS_BIT1) /* eflags - interrupts on */ + pushq_cfi $(X86_EFLAGS_IF|X86_EFLAGS_FIXED) /* eflags - interrupts on */ /*CFI_REL_OFFSET rflags,0*/ pushq_cfi $__KERNEL_CS /* cs */ /*CFI_REL_OFFSET cs,0*/ @@ -303,15 +243,15 @@ ENDPROC(native_usergs_sysret64) .macro SAVE_ARGS_IRQ cld /* start from rbp in pt_regs and jump over */ - movq_cfi rdi, RDI-RBP - movq_cfi rsi, RSI-RBP - movq_cfi rdx, RDX-RBP - movq_cfi rcx, RCX-RBP - movq_cfi rax, RAX-RBP - movq_cfi r8, R8-RBP - movq_cfi r9, R9-RBP - movq_cfi r10, R10-RBP - movq_cfi r11, R11-RBP + movq_cfi rdi, (RDI-RBP) + movq_cfi rsi, (RSI-RBP) + movq_cfi rdx, (RDX-RBP) + movq_cfi rcx, (RCX-RBP) + movq_cfi rax, (RAX-RBP) + movq_cfi r8, (R8-RBP) + movq_cfi r9, (R9-RBP) + movq_cfi r10, (R10-RBP) + movq_cfi r11, (R11-RBP) /* Save rbp so that we can unwind from get_irq_regs() */ movq_cfi rbp, 0 @@ -320,7 +260,7 @@ ENDPROC(native_usergs_sysret64) movq %rsp, %rsi leaq -RBP(%rsp),%rdi /* arg1 for handler */ - testl $3, CS(%rdi) + testl $3, CS-RBP(%rsi) je 1f SWAPGS /* @@ -330,11 +270,10 @@ ENDPROC(native_usergs_sysret64) * moving irq_enter into assembly, which would be too much work) */ 1: incl PER_CPU_VAR(irq_count) - jne 2f - mov PER_CPU_VAR(irq_stack_ptr),%rsp + cmovzq PER_CPU_VAR(irq_stack_ptr),%rsp CFI_DEF_CFA_REGISTER rsi -2: /* Store previous stack value */ + /* Store previous stack value */ pushq %rsi CFI_ESCAPE 0x0f /* DW_CFA_def_cfa_expression */, 6, \ 0x77 /* DW_OP_breg7 */, 0, \ @@ -345,23 +284,6 @@ ENDPROC(native_usergs_sysret64) TRACE_IRQS_OFF .endm -ENTRY(save_rest) - PARTIAL_FRAME 1 REST_SKIP+8 - movq 5*8+16(%rsp), %r11 /* save return address */ - movq_cfi rbx, RBX+16 - movq_cfi rbp, RBP+16 - movq_cfi r12, R12+16 - movq_cfi r13, R13+16 - movq_cfi r14, R14+16 - movq_cfi r15, R15+16 - movq %r11, 8(%rsp) /* return address */ - FIXUP_TOP_OF_STACK %r11, 16 - ret - CFI_ENDPROC -END(save_rest) - -/* save complete stack frame */ - .pushsection .kprobes.text, "ax" ENTRY(save_paranoid) XCPT_FRAME 1 RDI+8 cld @@ -390,7 +312,6 @@ ENTRY(save_paranoid) 1: ret CFI_ENDPROC END(save_paranoid) - .popsection /* * A newly forked process directly context switches into this address. @@ -402,7 +323,7 @@ ENTRY(ret_from_fork) LOCK ; btr $TIF_FORK,TI_flags(%r8) - pushq_cfi kernel_eflags(%rip) + pushq_cfi $0x0002 popfq_cfi # reset kernel eflags call schedule_tail # rdi: 'prev' task parameter @@ -412,7 +333,7 @@ ENTRY(ret_from_fork) RESTORE_REST testl $3, CS-ARGOFFSET(%rsp) # from kernel_thread? - jz retint_restore_args + jz 1f testl $_TIF_IA32, TI_flags(%rcx) # 32-bit compat task needs IRET jnz int_ret_from_sys_call @@ -420,6 +341,14 @@ ENTRY(ret_from_fork) RESTORE_TOP_OF_STACK %rdi, -ARGOFFSET jmp ret_from_sys_call # go to the SYSRET fastpath +1: + subq $REST_SKIP, %rsp # leave space for volatiles + CFI_ADJUST_CFA_OFFSET REST_SKIP + movq %rbp, %rdi + call *%rbx + movl $0, RAX(%rsp) + RESTORE_REST + jmp int_ret_from_sys_call CFI_ENDPROC END(ret_from_fork) @@ -427,7 +356,8 @@ END(ret_from_fork) * System call entry. Up to 6 arguments in registers are supported. * * SYSCALL does not save anything on the stack and does not change the - * stack pointer. + * stack pointer. However, it does mask the flags register for us, so + * CLD and CLAC are not needed. */ /* @@ -482,7 +412,12 @@ GLOBAL(system_call_after_swapgs) testl $_TIF_WORK_SYSCALL_ENTRY,TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET) jnz tracesys system_call_fastpath: +#if __SYSCALL_MASK == ~0 cmpq $__NR_syscall_max,%rax +#else + andl $__SYSCALL_MASK,%eax + cmpl $__NR_syscall_max,%eax +#endif ja badsys movq %r10,%rcx call *sys_call_table(,%rax,8) # XXX: rip relative @@ -522,7 +457,7 @@ sysret_careful: TRACE_IRQS_ON ENABLE_INTERRUPTS(CLBR_NONE) pushq_cfi %rdi - call schedule + SCHEDULE_USER popq_cfi %rdi jmp sysret_check @@ -596,7 +531,12 @@ tracesys: */ LOAD_ARGS ARGOFFSET, 1 RESTORE_REST +#if __SYSCALL_MASK == ~0 cmpq $__NR_syscall_max,%rax +#else + andl $__SYSCALL_MASK,%eax + cmpl $__NR_syscall_max,%eax +#endif ja int_ret_from_sys_call /* RAX(%rsp) set to -ENOSYS above */ movq %r10,%rcx /* fixup for C */ call *sys_call_table(,%rax,8) @@ -630,7 +570,7 @@ int_careful: TRACE_IRQS_ON ENABLE_INTERRUPTS(CLBR_NONE) pushq_cfi %rdi - call schedule + SCHEDULE_USER popq_cfi %rdi DISABLE_INTERRUPTS(CLBR_NONE) TRACE_IRQS_OFF @@ -667,28 +607,38 @@ int_restore_rest: CFI_ENDPROC END(system_call) -/* - * Certain special system calls that need to save a complete full stack frame. - */ - .macro PTREGSCALL label,func,arg -ENTRY(\label) - PARTIAL_FRAME 1 8 /* offset 8: return address */ - subq $REST_SKIP, %rsp - CFI_ADJUST_CFA_OFFSET REST_SKIP - call save_rest + .macro FORK_LIKE func +ENTRY(stub_\func) + CFI_STARTPROC + popq %r11 /* save return address */ + PARTIAL_FRAME 0 + SAVE_REST + pushq %r11 /* put it back on stack */ + FIXUP_TOP_OF_STACK %r11, 8 DEFAULT_FRAME 0 8 /* offset 8: return address */ - leaq 8(%rsp), \arg /* pt_regs pointer */ + call sys_\func + RESTORE_TOP_OF_STACK %r11, 8 + ret $REST_SKIP /* pop extended registers */ + CFI_ENDPROC +END(stub_\func) + .endm + + .macro FIXED_FRAME label,func +ENTRY(\label) + CFI_STARTPROC + PARTIAL_FRAME 0 8 /* offset 8: return address */ + FIXUP_TOP_OF_STACK %r11, 8-ARGOFFSET call \func - jmp ptregscall_common + RESTORE_TOP_OF_STACK %r11, 8-ARGOFFSET + ret CFI_ENDPROC END(\label) .endm - PTREGSCALL stub_clone, sys_clone, %r8 - PTREGSCALL stub_fork, sys_fork, %rdi - PTREGSCALL stub_vfork, sys_vfork, %rdi - PTREGSCALL stub_sigaltstack, sys_sigaltstack, %rdx - PTREGSCALL stub_iopl, sys_iopl, %rsi + FORK_LIKE clone + FORK_LIKE fork + FORK_LIKE vfork + FIXED_FRAME stub_iopl, sys_iopl ENTRY(ptregscall_common) DEFAULT_FRAME 1 8 /* offset 8: return address */ @@ -709,9 +659,7 @@ ENTRY(stub_execve) PARTIAL_FRAME 0 SAVE_REST FIXUP_TOP_OF_STACK %r11 - movq %rsp, %rcx call sys_execve - RESTORE_TOP_OF_STACK %r11 movq %rax,RAX(%rsp) RESTORE_REST jmp int_ret_from_sys_call @@ -727,7 +675,6 @@ ENTRY(stub_rt_sigreturn) addq $8, %rsp PARTIAL_FRAME 0 SAVE_REST - movq %rsp,%rdi FIXUP_TOP_OF_STACK %r11 call sys_rt_sigreturn movq %rax,RAX(%rsp) # fixme, this could be done at the higher layer @@ -736,6 +683,36 @@ ENTRY(stub_rt_sigreturn) CFI_ENDPROC END(stub_rt_sigreturn) +#ifdef CONFIG_X86_X32_ABI +ENTRY(stub_x32_rt_sigreturn) + CFI_STARTPROC + addq $8, %rsp + PARTIAL_FRAME 0 + SAVE_REST + FIXUP_TOP_OF_STACK %r11 + call sys32_x32_rt_sigreturn + movq %rax,RAX(%rsp) # fixme, this could be done at the higher layer + RESTORE_REST + jmp int_ret_from_sys_call + CFI_ENDPROC +END(stub_x32_rt_sigreturn) + +ENTRY(stub_x32_execve) + CFI_STARTPROC + addq $8, %rsp + PARTIAL_FRAME 0 + SAVE_REST + FIXUP_TOP_OF_STACK %r11 + call compat_sys_execve + RESTORE_TOP_OF_STACK %r11 + movq %rax,RAX(%rsp) + RESTORE_REST + jmp int_ret_from_sys_call + CFI_ENDPROC +END(stub_x32_execve) + +#endif + /* * Build the entry stubs and pointer table with some assembler magic. * We pack 7 stubs into a single 32-byte chunk, which will fit in a @@ -792,10 +769,6 @@ END(interrupt) call \func .endm -/* - * Interrupt entry/exit should be protected against kprobes - */ - .pushsection .kprobes.text, "ax" /* * The interrupt stubs push (~vector+0x80) onto the stack and * then jump to common_interrupt. @@ -803,6 +776,7 @@ END(interrupt) .p2align CONFIG_X86_L1_CACHE_SHIFT common_interrupt: XCPT_FRAME + ASM_CLAC addq $-0x80,(%rsp) /* Adjust vector to [-256,-1] range */ interrupt do_IRQ /* 0(%rsp): old_rsp-ARGOFFSET */ @@ -813,7 +787,7 @@ ret_from_intr: /* Restore saved previous stack */ popq %rsi - CFI_DEF_CFA_REGISTER rsi + CFI_DEF_CFA rsi,SS+8-RBP /* reg/off reset after def_cfa_expr */ leaq ARGOFFSET-RBP(%rsi), %rsp CFI_DEF_CFA_REGISTER rsp CFI_ADJUST_CFA_OFFSET RBP-ARGOFFSET @@ -858,17 +832,44 @@ restore_args: irq_return: INTERRUPT_RETURN - .section __ex_table, "a" - .quad irq_return, bad_iret - .previous - -#ifdef CONFIG_PARAVIRT ENTRY(native_iret) + /* + * Are we returning to a stack segment from the LDT? Note: in + * 64-bit mode SS:RSP on the exception stack is always valid. + */ +#ifdef CONFIG_X86_ESPFIX64 + testb $4,(SS-RIP)(%rsp) + jnz native_irq_return_ldt +#endif + +native_irq_return_iret: iretq + _ASM_EXTABLE(native_irq_return_iret, bad_iret) - .section __ex_table,"a" - .quad native_iret, bad_iret - .previous +#ifdef CONFIG_X86_ESPFIX64 +native_irq_return_ldt: + pushq_cfi %rax + pushq_cfi %rdi + SWAPGS + movq PER_CPU_VAR(espfix_waddr),%rdi + movq %rax,(0*8)(%rdi) /* RAX */ + movq (2*8)(%rsp),%rax /* RIP */ + movq %rax,(1*8)(%rdi) + movq (3*8)(%rsp),%rax /* CS */ + movq %rax,(2*8)(%rdi) + movq (4*8)(%rsp),%rax /* RFLAGS */ + movq %rax,(3*8)(%rdi) + movq (6*8)(%rsp),%rax /* SS */ + movq %rax,(5*8)(%rdi) + movq (5*8)(%rsp),%rax /* RSP */ + movq %rax,(4*8)(%rdi) + andl $0xffff0000,%eax + popq_cfi %rdi + orq PER_CPU_VAR(espfix_stack),%rax + SWAPGS + movq %rax,%rsp + popq_cfi %rax + jmp native_irq_return_iret #endif .section .fixup,"ax" @@ -898,7 +899,7 @@ retint_careful: TRACE_IRQS_ON ENABLE_INTERRUPTS(CLBR_NONE) pushq_cfi %rdi - call schedule + SCHEDULE_USER popq_cfi %rdi GET_THREAD_INFO(%rcx) DISABLE_INTERRUPTS(CLBR_NONE) @@ -925,29 +926,54 @@ retint_signal: /* Returning to kernel space. Check if we need preemption */ /* rcx: threadinfo. interrupts off. */ ENTRY(retint_kernel) - cmpl $0,TI_preempt_count(%rcx) + cmpl $0,PER_CPU_VAR(__preempt_count) jnz retint_restore_args - bt $TIF_NEED_RESCHED,TI_flags(%rcx) - jnc retint_restore_args bt $9,EFLAGS-ARGOFFSET(%rsp) /* interrupts off? */ jnc retint_restore_args call preempt_schedule_irq jmp exit_intr #endif - CFI_ENDPROC END(common_interrupt) -/* - * End of kprobes section - */ - .popsection + + /* + * If IRET takes a fault on the espfix stack, then we + * end up promoting it to a doublefault. In that case, + * modify the stack to make it look like we just entered + * the #GP handler from user space, similar to bad_iret. + */ +#ifdef CONFIG_X86_ESPFIX64 + ALIGN +__do_double_fault: + XCPT_FRAME 1 RDI+8 + movq RSP(%rdi),%rax /* Trap on the espfix stack? */ + sarq $PGDIR_SHIFT,%rax + cmpl $ESPFIX_PGD_ENTRY,%eax + jne do_double_fault /* No, just deliver the fault */ + cmpl $__KERNEL_CS,CS(%rdi) + jne do_double_fault + movq RIP(%rdi),%rax + cmpq $native_irq_return_iret,%rax + jne do_double_fault /* This shouldn't happen... */ + movq PER_CPU_VAR(kernel_stack),%rax + subq $(6*8-KERNEL_STACK_OFFSET),%rax /* Reset to original stack */ + movq %rax,RSP(%rdi) + movq $0,(%rax) /* Missing (lost) #GP error code */ + movq $general_protection,RIP(%rdi) + retq + CFI_ENDPROC +END(__do_double_fault) +#else +# define __do_double_fault do_double_fault +#endif /* * APIC interrupts. */ -.macro apicinterrupt num sym do_sym +.macro apicinterrupt3 num sym do_sym ENTRY(\sym) INTR_FRAME + ASM_CLAC pushq_cfi $~(\num) .Lcommon_\sym: interrupt \do_sym @@ -956,15 +982,32 @@ ENTRY(\sym) END(\sym) .endm +#ifdef CONFIG_TRACING +#define trace(sym) trace_##sym +#define smp_trace(sym) smp_trace_##sym + +.macro trace_apicinterrupt num sym +apicinterrupt3 \num trace(\sym) smp_trace(\sym) +.endm +#else +.macro trace_apicinterrupt num sym do_sym +.endm +#endif + +.macro apicinterrupt num sym do_sym +apicinterrupt3 \num \sym \do_sym +trace_apicinterrupt \num \sym +.endm + #ifdef CONFIG_SMP -apicinterrupt IRQ_MOVE_CLEANUP_VECTOR \ +apicinterrupt3 IRQ_MOVE_CLEANUP_VECTOR \ irq_move_cleanup_interrupt smp_irq_move_cleanup_interrupt -apicinterrupt REBOOT_VECTOR \ +apicinterrupt3 REBOOT_VECTOR \ reboot_interrupt smp_reboot_interrupt #endif #ifdef CONFIG_X86_UV -apicinterrupt UV_BAU_MESSAGE \ +apicinterrupt3 UV_BAU_MESSAGE \ uv_bau_message_intr1 uv_bau_message_interrupt #endif apicinterrupt LOCAL_TIMER_VECTOR \ @@ -972,28 +1015,20 @@ apicinterrupt LOCAL_TIMER_VECTOR \ apicinterrupt X86_PLATFORM_IPI_VECTOR \ x86_platform_ipi smp_x86_platform_ipi -#ifdef CONFIG_SMP - ALIGN - INTR_FRAME -.irp idx,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15, \ - 16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31 -.if NUM_INVALIDATE_TLB_VECTORS > \idx -ENTRY(invalidate_interrupt\idx) - pushq_cfi $~(INVALIDATE_TLB_VECTOR_START+\idx) - jmp .Lcommon_invalidate_interrupt0 - CFI_ADJUST_CFA_OFFSET -8 -END(invalidate_interrupt\idx) -.endif -.endr - CFI_ENDPROC -apicinterrupt INVALIDATE_TLB_VECTOR_START, \ - invalidate_interrupt0, smp_invalidate_interrupt +#ifdef CONFIG_HAVE_KVM +apicinterrupt3 POSTED_INTR_VECTOR \ + kvm_posted_intr_ipi smp_kvm_posted_intr_ipi #endif +#ifdef CONFIG_X86_MCE_THRESHOLD apicinterrupt THRESHOLD_APIC_VECTOR \ threshold_interrupt smp_threshold_interrupt +#endif + +#ifdef CONFIG_X86_THERMAL_VECTOR apicinterrupt THERMAL_APIC_VECTOR \ thermal_interrupt smp_thermal_interrupt +#endif #ifdef CONFIG_SMP apicinterrupt CALL_FUNCTION_SINGLE_VECTOR \ @@ -1017,109 +1052,100 @@ apicinterrupt IRQ_WORK_VECTOR \ /* * Exception entry points. */ -.macro zeroentry sym do_sym -ENTRY(\sym) - INTR_FRAME - PARAVIRT_ADJUST_EXCEPTION_FRAME - pushq_cfi $-1 /* ORIG_RAX: no syscall to restart */ - subq $ORIG_RAX-R15, %rsp - CFI_ADJUST_CFA_OFFSET ORIG_RAX-R15 - call error_entry - DEFAULT_FRAME 0 - movq %rsp,%rdi /* pt_regs pointer */ - xorl %esi,%esi /* no error code */ - call \do_sym - jmp error_exit /* %ebx: no swapgs flag */ - CFI_ENDPROC -END(\sym) -.endm +#define INIT_TSS_IST(x) PER_CPU_VAR(init_tss) + (TSS_ist + ((x) - 1) * 8) -.macro paranoidzeroentry sym do_sym +.macro idtentry sym do_sym has_error_code:req paranoid=0 shift_ist=-1 ENTRY(\sym) - INTR_FRAME - PARAVIRT_ADJUST_EXCEPTION_FRAME - pushq_cfi $-1 /* ORIG_RAX: no syscall to restart */ - subq $ORIG_RAX-R15, %rsp - CFI_ADJUST_CFA_OFFSET ORIG_RAX-R15 - call save_paranoid - TRACE_IRQS_OFF - movq %rsp,%rdi /* pt_regs pointer */ - xorl %esi,%esi /* no error code */ - call \do_sym - jmp paranoid_exit /* %ebx: no swapgs flag */ - CFI_ENDPROC -END(\sym) -.endm + /* Sanity check */ + .if \shift_ist != -1 && \paranoid == 0 + .error "using shift_ist requires paranoid=1" + .endif -#define INIT_TSS_IST(x) PER_CPU_VAR(init_tss) + (TSS_ist + ((x) - 1) * 8) -.macro paranoidzeroentry_ist sym do_sym ist -ENTRY(\sym) + .if \has_error_code + XCPT_FRAME + .else INTR_FRAME - PARAVIRT_ADJUST_EXCEPTION_FRAME - pushq_cfi $-1 /* ORIG_RAX: no syscall to restart */ - subq $ORIG_RAX-R15, %rsp - CFI_ADJUST_CFA_OFFSET ORIG_RAX-R15 - call save_paranoid - TRACE_IRQS_OFF - movq %rsp,%rdi /* pt_regs pointer */ - xorl %esi,%esi /* no error code */ - subq $EXCEPTION_STKSZ, INIT_TSS_IST(\ist) - call \do_sym - addq $EXCEPTION_STKSZ, INIT_TSS_IST(\ist) - jmp paranoid_exit /* %ebx: no swapgs flag */ - CFI_ENDPROC -END(\sym) -.endm + .endif -.macro errorentry sym do_sym -ENTRY(\sym) - XCPT_FRAME + ASM_CLAC PARAVIRT_ADJUST_EXCEPTION_FRAME - subq $ORIG_RAX-R15, %rsp - CFI_ADJUST_CFA_OFFSET ORIG_RAX-R15 - call error_entry - DEFAULT_FRAME 0 - movq %rsp,%rdi /* pt_regs pointer */ - movq ORIG_RAX(%rsp),%rsi /* get error code */ - movq $-1,ORIG_RAX(%rsp) /* no syscall to restart */ - call \do_sym - jmp error_exit /* %ebx: no swapgs flag */ - CFI_ENDPROC -END(\sym) -.endm - /* error code is on the stack already */ -.macro paranoiderrorentry sym do_sym -ENTRY(\sym) - XCPT_FRAME - PARAVIRT_ADJUST_EXCEPTION_FRAME + .ifeq \has_error_code + pushq_cfi $-1 /* ORIG_RAX: no syscall to restart */ + .endif + subq $ORIG_RAX-R15, %rsp CFI_ADJUST_CFA_OFFSET ORIG_RAX-R15 + + .if \paranoid call save_paranoid + .else + call error_entry + .endif + DEFAULT_FRAME 0 + + .if \paranoid + .if \shift_ist != -1 + TRACE_IRQS_OFF_DEBUG /* reload IDT in case of recursion */ + .else TRACE_IRQS_OFF + .endif + .endif + movq %rsp,%rdi /* pt_regs pointer */ + + .if \has_error_code movq ORIG_RAX(%rsp),%rsi /* get error code */ movq $-1,ORIG_RAX(%rsp) /* no syscall to restart */ + .else + xorl %esi,%esi /* no error code */ + .endif + + .if \shift_ist != -1 + subq $EXCEPTION_STKSZ, INIT_TSS_IST(\shift_ist) + .endif + call \do_sym + + .if \shift_ist != -1 + addq $EXCEPTION_STKSZ, INIT_TSS_IST(\shift_ist) + .endif + + .if \paranoid jmp paranoid_exit /* %ebx: no swapgs flag */ + .else + jmp error_exit /* %ebx: no swapgs flag */ + .endif + CFI_ENDPROC END(\sym) .endm -zeroentry divide_error do_divide_error -zeroentry overflow do_overflow -zeroentry bounds do_bounds -zeroentry invalid_op do_invalid_op -zeroentry device_not_available do_device_not_available -paranoiderrorentry double_fault do_double_fault -zeroentry coprocessor_segment_overrun do_coprocessor_segment_overrun -errorentry invalid_TSS do_invalid_TSS -errorentry segment_not_present do_segment_not_present -zeroentry spurious_interrupt_bug do_spurious_interrupt_bug -zeroentry coprocessor_error do_coprocessor_error -errorentry alignment_check do_alignment_check -zeroentry simd_coprocessor_error do_simd_coprocessor_error +#ifdef CONFIG_TRACING +.macro trace_idtentry sym do_sym has_error_code:req +idtentry trace(\sym) trace(\do_sym) has_error_code=\has_error_code +idtentry \sym \do_sym has_error_code=\has_error_code +.endm +#else +.macro trace_idtentry sym do_sym has_error_code:req +idtentry \sym \do_sym has_error_code=\has_error_code +.endm +#endif + +idtentry divide_error do_divide_error has_error_code=0 +idtentry overflow do_overflow has_error_code=0 +idtentry bounds do_bounds has_error_code=0 +idtentry invalid_op do_invalid_op has_error_code=0 +idtentry device_not_available do_device_not_available has_error_code=0 +idtentry double_fault __do_double_fault has_error_code=1 paranoid=1 +idtentry coprocessor_segment_overrun do_coprocessor_segment_overrun has_error_code=0 +idtentry invalid_TSS do_invalid_TSS has_error_code=1 +idtentry segment_not_present do_segment_not_present has_error_code=1 +idtentry spurious_interrupt_bug do_spurious_interrupt_bug has_error_code=0 +idtentry coprocessor_error do_coprocessor_error has_error_code=0 +idtentry alignment_check do_alignment_check has_error_code=1 +idtentry simd_coprocessor_error do_simd_coprocessor_error has_error_code=0 /* Reload gs selector with exception handling */ @@ -1138,10 +1164,7 @@ gs_change: CFI_ENDPROC END(native_load_gs_index) - .section __ex_table,"a" - .align 8 - .quad gs_change,bad_gs - .previous + _ASM_EXTABLE(gs_change,bad_gs) .section .fixup,"ax" /* running with kernelgs */ bad_gs: @@ -1151,54 +1174,8 @@ bad_gs: jmp 2b .previous -ENTRY(kernel_thread_helper) - pushq $0 # fake return address - CFI_STARTPROC - /* - * Here we are in the child and the registers are set as they were - * at kernel_thread() invocation in the parent. - */ - call *%rsi - # exit - mov %eax, %edi - call do_exit - ud2 # padding for call trace - CFI_ENDPROC -END(kernel_thread_helper) - -/* - * execve(). This function needs to use IRET, not SYSRET, to set up all state properly. - * - * C extern interface: - * extern long execve(const char *name, char **argv, char **envp) - * - * asm input arguments: - * rdi: name, rsi: argv, rdx: envp - * - * We want to fallback into: - * extern long sys_execve(const char *name, char **argv,char **envp, struct pt_regs *regs) - * - * do_sys_execve asm fallback arguments: - * rdi: name, rsi: argv, rdx: envp, rcx: fake frame on the stack - */ -ENTRY(kernel_execve) - CFI_STARTPROC - FAKE_STACK_FRAME $0 - SAVE_ALL - movq %rsp,%rcx - call sys_execve - movq %rax, RAX(%rsp) - RESTORE_REST - testq %rax,%rax - je int_ret_from_sys_call - RESTORE_ARGS - UNFAKE_STACK_FRAME - ret - CFI_ENDPROC -END(kernel_execve) - /* Call softirq on interrupt stack. Interrupts are off. */ -ENTRY(call_softirq) +ENTRY(do_softirq_own_stack) CFI_STARTPROC pushq_cfi %rbp CFI_REL_OFFSET rbp,0 @@ -1215,10 +1192,10 @@ ENTRY(call_softirq) decl PER_CPU_VAR(irq_count) ret CFI_ENDPROC -END(call_softirq) +END(do_softirq_own_stack) #ifdef CONFIG_XEN -zeroentry xen_hypervisor_callback xen_do_hypervisor_callback +idtentry xen_hypervisor_callback xen_do_hypervisor_callback has_error_code=0 /* * A note on the "critical region" in our callback handler. @@ -1308,37 +1285,37 @@ ENTRY(xen_failsafe_callback) CFI_RESTORE r11 addq $0x30,%rsp CFI_ADJUST_CFA_OFFSET -0x30 - pushq_cfi $0 + pushq_cfi $-1 /* orig_ax = -1 => not a system call */ SAVE_ALL jmp error_exit CFI_ENDPROC END(xen_failsafe_callback) -apicinterrupt XEN_HVM_EVTCHN_CALLBACK \ +apicinterrupt3 HYPERVISOR_CALLBACK_VECTOR \ xen_hvm_callback_vector xen_evtchn_do_upcall #endif /* CONFIG_XEN */ -/* - * Some functions should be protected against kprobes - */ - .pushsection .kprobes.text, "ax" +#if IS_ENABLED(CONFIG_HYPERV) +apicinterrupt3 HYPERVISOR_CALLBACK_VECTOR \ + hyperv_callback_vector hyperv_vector_handler +#endif /* CONFIG_HYPERV */ -paranoidzeroentry_ist debug do_debug DEBUG_STACK -paranoidzeroentry_ist int3 do_int3 DEBUG_STACK -paranoiderrorentry stack_segment do_stack_segment +idtentry debug do_debug has_error_code=0 paranoid=1 shift_ist=DEBUG_STACK +idtentry int3 do_int3 has_error_code=0 paranoid=1 shift_ist=DEBUG_STACK +idtentry stack_segment do_stack_segment has_error_code=1 paranoid=1 #ifdef CONFIG_XEN -zeroentry xen_debug do_debug -zeroentry xen_int3 do_int3 -errorentry xen_stack_segment do_stack_segment +idtentry xen_debug do_debug has_error_code=0 +idtentry xen_int3 do_int3 has_error_code=0 +idtentry xen_stack_segment do_stack_segment has_error_code=1 #endif -errorentry general_protection do_general_protection -errorentry page_fault do_page_fault +idtentry general_protection do_general_protection has_error_code=1 +trace_idtentry page_fault do_page_fault has_error_code=1 #ifdef CONFIG_KVM_GUEST -errorentry async_page_fault do_async_page_fault +idtentry async_page_fault do_async_page_fault has_error_code=1 #endif #ifdef CONFIG_X86_MCE -paranoidzeroentry machine_check *machine_check_vector(%rip) +idtentry machine_check has_error_code=0 paranoid=1 do_sym=*machine_check_vector(%rip) #endif /* @@ -1358,7 +1335,7 @@ paranoidzeroentry machine_check *machine_check_vector(%rip) ENTRY(paranoid_exit) DEFAULT_FRAME DISABLE_INTERRUPTS(CLBR_NONE) - TRACE_IRQS_OFF + TRACE_IRQS_OFF_DEBUG testl %ebx,%ebx /* swapgs needed? */ jnz paranoid_restore testl $3,CS(%rsp) @@ -1369,7 +1346,7 @@ paranoid_swapgs: RESTORE_ALL 8 jmp irq_return paranoid_restore: - TRACE_IRQS_IRETQ 0 + TRACE_IRQS_IRETQ_DEBUG 0 RESTORE_ALL 8 jmp irq_return paranoid_userspace: @@ -1394,7 +1371,7 @@ paranoid_userspace: paranoid_schedule: TRACE_IRQS_ON ENABLE_INTERRUPTS(CLBR_ANY) - call schedule + SCHEDULE_USER DISABLE_INTERRUPTS(CLBR_ANY) TRACE_IRQS_OFF jmp paranoid_userspace @@ -1443,7 +1420,7 @@ error_sti: */ error_kernelspace: incl %ebx - leaq irq_return(%rip),%rcx + leaq native_irq_return_iret(%rip),%rcx cmpq %rcx,RIP+8(%rsp) je error_swapgs movl %ecx,%eax /* zero extend */ @@ -1530,12 +1507,20 @@ ENTRY(nmi) /* Use %rdx as out temp variable throughout */ pushq_cfi %rdx + CFI_REL_OFFSET rdx, 0 + + /* + * If %cs was not the kernel segment, then the NMI triggered in user + * space, which means it is definitely not nested. + */ + cmpl $__KERNEL_CS, 16(%rsp) + jne first_nmi /* * Check the special variable on the stack to see if NMIs are * executing. */ - cmp $1, -8(%rsp) + cmpl $1, -8(%rsp) je nested_nmi /* @@ -1547,6 +1532,7 @@ ENTRY(nmi) */ lea 6*8(%rsp), %rdx test_in_nmi rdx, 4*8(%rsp), nested_nmi, first_nmi + CFI_REMEMBER_STATE nested_nmi: /* @@ -1563,9 +1549,10 @@ nested_nmi: 1: /* Set up the interrupted NMIs stack to jump to repeat_nmi */ - leaq -6*8(%rsp), %rdx + leaq -1*8(%rsp), %rdx movq %rdx, %rsp - CFI_ADJUST_CFA_OFFSET 6*8 + CFI_ADJUST_CFA_OFFSET 1*8 + leaq -10*8(%rsp), %rdx pushq_cfi $__KERNEL_DS pushq_cfi %rdx pushfq_cfi @@ -1573,15 +1560,17 @@ nested_nmi: pushq_cfi $repeat_nmi /* Put stack back */ - addq $(11*8), %rsp - CFI_ADJUST_CFA_OFFSET -11*8 + addq $(6*8), %rsp + CFI_ADJUST_CFA_OFFSET -6*8 nested_nmi_out: popq_cfi %rdx + CFI_RESTORE rdx /* No need to check faults here */ INTERRUPT_RETURN + CFI_RESTORE_STATE first_nmi: /* * Because nested NMIs will use the pushed location that we @@ -1598,48 +1587,79 @@ first_nmi: * +-------------------------+ * | NMI executing variable | * +-------------------------+ - * | Saved SS | - * | Saved Return RSP | - * | Saved RFLAGS | - * | Saved CS | - * | Saved RIP | - * +-------------------------+ * | copied SS | * | copied Return RSP | * | copied RFLAGS | * | copied CS | * | copied RIP | * +-------------------------+ + * | Saved SS | + * | Saved Return RSP | + * | Saved RFLAGS | + * | Saved CS | + * | Saved RIP | + * +-------------------------+ * | pt_regs | * +-------------------------+ * - * The saved RIP is used to fix up the copied RIP that a nested - * NMI may zero out. The original stack frame and the temp storage + * The saved stack frame is used to fix up the copied stack frame + * that a nested NMI may change to make the interrupted NMI iret jump + * to the repeat_nmi. The original stack frame and the temp storage * is also used by nested NMIs and can not be trusted on exit. */ + /* Do not pop rdx, nested NMIs will corrupt that part of the stack */ + movq (%rsp), %rdx + CFI_RESTORE rdx + /* Set the NMI executing variable on the stack. */ pushq_cfi $1 + /* + * Leave room for the "copied" frame + */ + subq $(5*8), %rsp + CFI_ADJUST_CFA_OFFSET 5*8 + /* Copy the stack frame to the Saved frame */ .rept 5 - pushq_cfi 6*8(%rsp) + pushq_cfi 11*8(%rsp) .endr + CFI_DEF_CFA_OFFSET SS+8-RIP + + /* Everything up to here is safe from nested NMIs */ + + /* + * If there was a nested NMI, the first NMI's iret will return + * here. But NMIs are still enabled and we can take another + * nested NMI. The nested NMI checks the interrupted RIP to see + * if it is between repeat_nmi and end_repeat_nmi, and if so + * it will just return, as we are about to repeat an NMI anyway. + * This makes it safe to copy to the stack frame that a nested + * NMI will update. + */ +repeat_nmi: + /* + * Update the stack variable to say we are still in NMI (the update + * is benign for the non-repeat case, where 1 was pushed just above + * to this very stack slot). + */ + movq $1, 10*8(%rsp) /* Make another copy, this one may be modified by nested NMIs */ + addq $(10*8), %rsp + CFI_ADJUST_CFA_OFFSET -10*8 .rept 5 - pushq_cfi 4*8(%rsp) + pushq_cfi -6*8(%rsp) .endr - - /* Do not pop rdx, nested NMIs will corrupt it */ - movq 11*8(%rsp), %rdx + subq $(5*8), %rsp + CFI_DEF_CFA_OFFSET SS+8-RIP +end_repeat_nmi: /* * Everything below this point can be preempted by a nested - * NMI if the first NMI took an exception. Repeated NMIs - * caused by an exception and nested NMI will start here, and - * can still be preempted by another NMI. + * NMI if the first NMI took an exception and reset our iret stack + * so that we repeat another NMI. */ -restart_nmi: pushq_cfi $-1 /* ORIG_RAX: no syscall to restart */ subq $ORIG_RAX-R15, %rsp CFI_ADJUST_CFA_OFFSET ORIG_RAX-R15 @@ -1652,42 +1672,44 @@ restart_nmi: */ call save_paranoid DEFAULT_FRAME 0 + + /* + * Save off the CR2 register. If we take a page fault in the NMI then + * it could corrupt the CR2 value. If the NMI preempts a page fault + * handler before it was able to read the CR2 register, and then the + * NMI itself takes a page fault, the page fault that was preempted + * will read the information from the NMI page fault and not the + * origin fault. Save it off and restore it if it changes. + * Use the r12 callee-saved register. + */ + movq %cr2, %r12 + /* paranoidentry do_nmi, 0; without TRACE_IRQS_OFF */ movq %rsp,%rdi movq $-1,%rsi call do_nmi + + /* Did the NMI take a page fault? Restore cr2 if it did */ + movq %cr2, %rcx + cmpq %rcx, %r12 + je 1f + movq %r12, %cr2 +1: + testl %ebx,%ebx /* swapgs needed? */ jnz nmi_restore nmi_swapgs: SWAPGS_UNSAFE_STACK nmi_restore: - RESTORE_ALL 8 + /* Pop the extra iret frame at once */ + RESTORE_ALL 6*8 + /* Clear the NMI executing stack variable */ - movq $0, 10*8(%rsp) + movq $0, 5*8(%rsp) jmp irq_return CFI_ENDPROC END(nmi) - /* - * If an NMI hit an iret because of an exception or breakpoint, - * it can lose its NMI context, and a nested NMI may come in. - * In that case, the nested NMI will change the preempted NMI's - * stack to jump to here when it does the final iret. - */ -repeat_nmi: - INTR_FRAME - /* Update the stack variable to say we are still in NMI */ - movq $1, 5*8(%rsp) - - /* copy the saved stack back to copy stack */ - .rept 5 - pushq_cfi 4*8(%rsp) - .endr - - jmp restart_nmi - CFI_ENDPROC -end_repeat_nmi: - ENTRY(ignore_sysret) CFI_STARTPROC mov $-ENOSYS,%eax @@ -1695,7 +1717,3 @@ ENTRY(ignore_sysret) CFI_ENDPROC END(ignore_sysret) -/* - * End of kprobes section - */ - .popsection diff --git a/arch/x86/kernel/espfix_64.c b/arch/x86/kernel/espfix_64.c new file mode 100644 index 00000000000..94d857fb103 --- /dev/null +++ b/arch/x86/kernel/espfix_64.c @@ -0,0 +1,208 @@ +/* ----------------------------------------------------------------------- * + * + * Copyright 2014 Intel Corporation; author: H. Peter Anvin + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * ----------------------------------------------------------------------- */ + +/* + * The IRET instruction, when returning to a 16-bit segment, only + * restores the bottom 16 bits of the user space stack pointer. This + * causes some 16-bit software to break, but it also leaks kernel state + * to user space. + * + * This works around this by creating percpu "ministacks", each of which + * is mapped 2^16 times 64K apart. When we detect that the return SS is + * on the LDT, we copy the IRET frame to the ministack and use the + * relevant alias to return to userspace. The ministacks are mapped + * readonly, so if the IRET fault we promote #GP to #DF which is an IST + * vector and thus has its own stack; we then do the fixup in the #DF + * handler. + * + * This file sets up the ministacks and the related page tables. The + * actual ministack invocation is in entry_64.S. + */ + +#include <linux/init.h> +#include <linux/init_task.h> +#include <linux/kernel.h> +#include <linux/percpu.h> +#include <linux/gfp.h> +#include <linux/random.h> +#include <asm/pgtable.h> +#include <asm/pgalloc.h> +#include <asm/setup.h> +#include <asm/espfix.h> + +/* + * Note: we only need 6*8 = 48 bytes for the espfix stack, but round + * it up to a cache line to avoid unnecessary sharing. + */ +#define ESPFIX_STACK_SIZE (8*8UL) +#define ESPFIX_STACKS_PER_PAGE (PAGE_SIZE/ESPFIX_STACK_SIZE) + +/* There is address space for how many espfix pages? */ +#define ESPFIX_PAGE_SPACE (1UL << (PGDIR_SHIFT-PAGE_SHIFT-16)) + +#define ESPFIX_MAX_CPUS (ESPFIX_STACKS_PER_PAGE * ESPFIX_PAGE_SPACE) +#if CONFIG_NR_CPUS > ESPFIX_MAX_CPUS +# error "Need more than one PGD for the ESPFIX hack" +#endif + +#define PGALLOC_GFP (GFP_KERNEL | __GFP_NOTRACK | __GFP_REPEAT | __GFP_ZERO) + +/* This contains the *bottom* address of the espfix stack */ +DEFINE_PER_CPU_READ_MOSTLY(unsigned long, espfix_stack); +DEFINE_PER_CPU_READ_MOSTLY(unsigned long, espfix_waddr); + +/* Initialization mutex - should this be a spinlock? */ +static DEFINE_MUTEX(espfix_init_mutex); + +/* Page allocation bitmap - each page serves ESPFIX_STACKS_PER_PAGE CPUs */ +#define ESPFIX_MAX_PAGES DIV_ROUND_UP(CONFIG_NR_CPUS, ESPFIX_STACKS_PER_PAGE) +static void *espfix_pages[ESPFIX_MAX_PAGES]; + +static __page_aligned_bss pud_t espfix_pud_page[PTRS_PER_PUD] + __aligned(PAGE_SIZE); + +static unsigned int page_random, slot_random; + +/* + * This returns the bottom address of the espfix stack for a specific CPU. + * The math allows for a non-power-of-two ESPFIX_STACK_SIZE, in which case + * we have to account for some amount of padding at the end of each page. + */ +static inline unsigned long espfix_base_addr(unsigned int cpu) +{ + unsigned long page, slot; + unsigned long addr; + + page = (cpu / ESPFIX_STACKS_PER_PAGE) ^ page_random; + slot = (cpu + slot_random) % ESPFIX_STACKS_PER_PAGE; + addr = (page << PAGE_SHIFT) + (slot * ESPFIX_STACK_SIZE); + addr = (addr & 0xffffUL) | ((addr & ~0xffffUL) << 16); + addr += ESPFIX_BASE_ADDR; + return addr; +} + +#define PTE_STRIDE (65536/PAGE_SIZE) +#define ESPFIX_PTE_CLONES (PTRS_PER_PTE/PTE_STRIDE) +#define ESPFIX_PMD_CLONES PTRS_PER_PMD +#define ESPFIX_PUD_CLONES (65536/(ESPFIX_PTE_CLONES*ESPFIX_PMD_CLONES)) + +#define PGTABLE_PROT ((_KERNPG_TABLE & ~_PAGE_RW) | _PAGE_NX) + +static void init_espfix_random(void) +{ + unsigned long rand; + + /* + * This is run before the entropy pools are initialized, + * but this is hopefully better than nothing. + */ + if (!arch_get_random_long(&rand)) { + /* The constant is an arbitrary large prime */ + rdtscll(rand); + rand *= 0xc345c6b72fd16123UL; + } + + slot_random = rand % ESPFIX_STACKS_PER_PAGE; + page_random = (rand / ESPFIX_STACKS_PER_PAGE) + & (ESPFIX_PAGE_SPACE - 1); +} + +void __init init_espfix_bsp(void) +{ + pgd_t *pgd_p; + pteval_t ptemask; + + ptemask = __supported_pte_mask; + + /* Install the espfix pud into the kernel page directory */ + pgd_p = &init_level4_pgt[pgd_index(ESPFIX_BASE_ADDR)]; + pgd_populate(&init_mm, pgd_p, (pud_t *)espfix_pud_page); + + /* Randomize the locations */ + init_espfix_random(); + + /* The rest is the same as for any other processor */ + init_espfix_ap(); +} + +void init_espfix_ap(void) +{ + unsigned int cpu, page; + unsigned long addr; + pud_t pud, *pud_p; + pmd_t pmd, *pmd_p; + pte_t pte, *pte_p; + int n; + void *stack_page; + pteval_t ptemask; + + /* We only have to do this once... */ + if (likely(this_cpu_read(espfix_stack))) + return; /* Already initialized */ + + cpu = smp_processor_id(); + addr = espfix_base_addr(cpu); + page = cpu/ESPFIX_STACKS_PER_PAGE; + + /* Did another CPU already set this up? */ + stack_page = ACCESS_ONCE(espfix_pages[page]); + if (likely(stack_page)) + goto done; + + mutex_lock(&espfix_init_mutex); + + /* Did we race on the lock? */ + stack_page = ACCESS_ONCE(espfix_pages[page]); + if (stack_page) + goto unlock_done; + + ptemask = __supported_pte_mask; + + pud_p = &espfix_pud_page[pud_index(addr)]; + pud = *pud_p; + if (!pud_present(pud)) { + pmd_p = (pmd_t *)__get_free_page(PGALLOC_GFP); + pud = __pud(__pa(pmd_p) | (PGTABLE_PROT & ptemask)); + paravirt_alloc_pmd(&init_mm, __pa(pmd_p) >> PAGE_SHIFT); + for (n = 0; n < ESPFIX_PUD_CLONES; n++) + set_pud(&pud_p[n], pud); + } + + pmd_p = pmd_offset(&pud, addr); + pmd = *pmd_p; + if (!pmd_present(pmd)) { + pte_p = (pte_t *)__get_free_page(PGALLOC_GFP); + pmd = __pmd(__pa(pte_p) | (PGTABLE_PROT & ptemask)); + paravirt_alloc_pte(&init_mm, __pa(pte_p) >> PAGE_SHIFT); + for (n = 0; n < ESPFIX_PMD_CLONES; n++) + set_pmd(&pmd_p[n], pmd); + } + + pte_p = pte_offset_kernel(&pmd, addr); + stack_page = (void *)__get_free_page(GFP_KERNEL); + pte = __pte(__pa(stack_page) | (__PAGE_KERNEL_RO & ptemask)); + for (n = 0; n < ESPFIX_PTE_CLONES; n++) + set_pte(&pte_p[n*PTE_STRIDE], pte); + + /* Job is done for this CPU and any CPU which shares this page */ + ACCESS_ONCE(espfix_pages[page]) = stack_page; + +unlock_done: + mutex_unlock(&espfix_init_mutex); +done: + this_cpu_write(espfix_stack, addr); + this_cpu_write(espfix_waddr, (unsigned long)stack_page + + (addr & ~PAGE_MASK)); +} diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index c9a281f272f..cbc4a91b131 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -24,40 +24,21 @@ #include <trace/syscall.h> #include <asm/cacheflush.h> +#include <asm/kprobes.h> #include <asm/ftrace.h> #include <asm/nops.h> -#include <asm/nmi.h> - #ifdef CONFIG_DYNAMIC_FTRACE -/* - * modifying_code is set to notify NMIs that they need to use - * memory barriers when entering or exiting. But we don't want - * to burden NMIs with unnecessary memory barriers when code - * modification is not being done (which is most of the time). - * - * A mutex is already held when ftrace_arch_code_modify_prepare - * and post_process are called. No locks need to be taken here. - * - * Stop machine will make sure currently running NMIs are done - * and new NMIs will see the updated variable before we need - * to worry about NMIs doing memory barriers. - */ -static int modifying_code __read_mostly; -static DEFINE_PER_CPU(int, save_modifying_code); - int ftrace_arch_code_modify_prepare(void) { set_kernel_text_rw(); set_all_modules_text_rw(); - modifying_code = 1; return 0; } int ftrace_arch_code_modify_post_process(void) { - modifying_code = 0; set_all_modules_text_ro(); set_kernel_text_ro(); return 0; @@ -90,255 +71,577 @@ static unsigned char *ftrace_call_replace(unsigned long ip, unsigned long addr) return calc.code; } +static inline int +within(unsigned long addr, unsigned long start, unsigned long end) +{ + return addr >= start && addr < end; +} + +static unsigned long text_ip_addr(unsigned long ip) +{ + /* + * On x86_64, kernel text mappings are mapped read-only with + * CONFIG_DEBUG_RODATA. So we use the kernel identity mapping instead + * of the kernel text mapping to modify the kernel text. + * + * For 32bit kernels, these mappings are same and we can use + * kernel identity mapping to modify code. + */ + if (within(ip, (unsigned long)_text, (unsigned long)_etext)) + ip = (unsigned long)__va(__pa_symbol(ip)); + + return ip; +} + +static const unsigned char *ftrace_nop_replace(void) +{ + return ideal_nops[NOP_ATOMIC5]; +} + +static int +ftrace_modify_code_direct(unsigned long ip, unsigned const char *old_code, + unsigned const char *new_code) +{ + unsigned char replaced[MCOUNT_INSN_SIZE]; + + /* + * Note: Due to modules and __init, code can + * disappear and change, we need to protect against faulting + * as well as code changing. We do this by using the + * probe_kernel_* functions. + * + * No real locking needed, this code is run through + * kstop_machine, or before SMP starts. + */ + + /* read the text we want to modify */ + if (probe_kernel_read(replaced, (void *)ip, MCOUNT_INSN_SIZE)) + return -EFAULT; + + /* Make sure it is what we expect it to be */ + if (memcmp(replaced, old_code, MCOUNT_INSN_SIZE) != 0) + return -EINVAL; + + ip = text_ip_addr(ip); + + /* replace the text with the new text */ + if (probe_kernel_write((void *)ip, new_code, MCOUNT_INSN_SIZE)) + return -EPERM; + + sync_core(); + + return 0; +} + +int ftrace_make_nop(struct module *mod, + struct dyn_ftrace *rec, unsigned long addr) +{ + unsigned const char *new, *old; + unsigned long ip = rec->ip; + + old = ftrace_call_replace(ip, addr); + new = ftrace_nop_replace(); + + /* + * On boot up, and when modules are loaded, the MCOUNT_ADDR + * is converted to a nop, and will never become MCOUNT_ADDR + * again. This code is either running before SMP (on boot up) + * or before the code will ever be executed (module load). + * We do not want to use the breakpoint version in this case, + * just modify the code directly. + */ + if (addr == MCOUNT_ADDR) + return ftrace_modify_code_direct(rec->ip, old, new); + + /* Normal cases use add_brk_on_nop */ + WARN_ONCE(1, "invalid use of ftrace_make_nop"); + return -EINVAL; +} + +int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr) +{ + unsigned const char *new, *old; + unsigned long ip = rec->ip; + + old = ftrace_nop_replace(); + new = ftrace_call_replace(ip, addr); + + /* Should only be called when module is loaded */ + return ftrace_modify_code_direct(rec->ip, old, new); +} + /* - * Modifying code must take extra care. On an SMP machine, if - * the code being modified is also being executed on another CPU - * that CPU will have undefined results and possibly take a GPF. - * We use kstop_machine to stop other CPUS from exectuing code. - * But this does not stop NMIs from happening. We still need - * to protect against that. We separate out the modification of - * the code to take care of this. + * The modifying_ftrace_code is used to tell the breakpoint + * handler to call ftrace_int3_handler(). If it fails to + * call this handler for a breakpoint added by ftrace, then + * the kernel may crash. + * + * As atomic_writes on x86 do not need a barrier, we do not + * need to add smp_mb()s for this to work. It is also considered + * that we can not read the modifying_ftrace_code before + * executing the breakpoint. That would be quite remarkable if + * it could do that. Here's the flow that is required: * - * Two buffers are added: An IP buffer and a "code" buffer. + * CPU-0 CPU-1 * - * 1) Put the instruction pointer into the IP buffer - * and the new code into the "code" buffer. - * 2) Wait for any running NMIs to finish and set a flag that says - * we are modifying code, it is done in an atomic operation. - * 3) Write the code - * 4) clear the flag. - * 5) Wait for any running NMIs to finish. + * atomic_inc(mfc); + * write int3s + * <trap-int3> // implicit (r)mb + * if (atomic_read(mfc)) + * call ftrace_int3_handler() * - * If an NMI is executed, the first thing it does is to call - * "ftrace_nmi_enter". This will check if the flag is set to write - * and if it is, it will write what is in the IP and "code" buffers. + * Then when we are finished: * - * The trick is, it does not matter if everyone is writing the same - * content to the code location. Also, if a CPU is executing code - * it is OK to write to that code location if the contents being written - * are the same as what exists. + * atomic_dec(mfc); + * + * If we hit a breakpoint that was not set by ftrace, it does not + * matter if ftrace_int3_handler() is called or not. It will + * simply be ignored. But it is crucial that a ftrace nop/caller + * breakpoint is handled. No other user should ever place a + * breakpoint on an ftrace nop/caller location. It must only + * be done by this code. */ +atomic_t modifying_ftrace_code __read_mostly; -#define MOD_CODE_WRITE_FLAG (1 << 31) /* set when NMI should do the write */ -static atomic_t nmi_running = ATOMIC_INIT(0); -static int mod_code_status; /* holds return value of text write */ -static void *mod_code_ip; /* holds the IP to write to */ -static const void *mod_code_newcode; /* holds the text to write to the IP */ +static int +ftrace_modify_code(unsigned long ip, unsigned const char *old_code, + unsigned const char *new_code); -static unsigned nmi_wait_count; -static atomic_t nmi_update_count = ATOMIC_INIT(0); +/* + * Should never be called: + * As it is only called by __ftrace_replace_code() which is called by + * ftrace_replace_code() that x86 overrides, and by ftrace_update_code() + * which is called to turn mcount into nops or nops into function calls + * but not to convert a function from not using regs to one that uses + * regs, which ftrace_modify_call() is for. + */ +int ftrace_modify_call(struct dyn_ftrace *rec, unsigned long old_addr, + unsigned long addr) +{ + WARN_ON(1); + return -EINVAL; +} -int ftrace_arch_read_dyn_info(char *buf, int size) +static unsigned long ftrace_update_func; + +static int update_ftrace_func(unsigned long ip, void *new) { - int r; + unsigned char old[MCOUNT_INSN_SIZE]; + int ret; + + memcpy(old, (void *)ip, MCOUNT_INSN_SIZE); + + ftrace_update_func = ip; + /* Make sure the breakpoints see the ftrace_update_func update */ + smp_wmb(); - r = snprintf(buf, size, "%u %u", - nmi_wait_count, - atomic_read(&nmi_update_count)); - return r; + /* See comment above by declaration of modifying_ftrace_code */ + atomic_inc(&modifying_ftrace_code); + + ret = ftrace_modify_code(ip, old, new); + + atomic_dec(&modifying_ftrace_code); + + return ret; } -static void clear_mod_flag(void) +int ftrace_update_ftrace_func(ftrace_func_t func) { - int old = atomic_read(&nmi_running); - - for (;;) { - int new = old & ~MOD_CODE_WRITE_FLAG; + unsigned long ip = (unsigned long)(&ftrace_call); + unsigned char *new; + int ret; - if (old == new) - break; + new = ftrace_call_replace(ip, (unsigned long)func); + ret = update_ftrace_func(ip, new); - old = atomic_cmpxchg(&nmi_running, old, new); + /* Also update the regs callback function */ + if (!ret) { + ip = (unsigned long)(&ftrace_regs_call); + new = ftrace_call_replace(ip, (unsigned long)func); + ret = update_ftrace_func(ip, new); } + + return ret; } -static void ftrace_mod_code(void) +static int is_ftrace_caller(unsigned long ip) { - /* - * Yes, more than one CPU process can be writing to mod_code_status. - * (and the code itself) - * But if one were to fail, then they all should, and if one were - * to succeed, then they all should. - */ - mod_code_status = probe_kernel_write(mod_code_ip, mod_code_newcode, - MCOUNT_INSN_SIZE); + if (ip == ftrace_update_func) + return 1; - /* if we fail, then kill any new writers */ - if (mod_code_status) - clear_mod_flag(); + return 0; } -void ftrace_nmi_enter(void) +/* + * A breakpoint was added to the code address we are about to + * modify, and this is the handle that will just skip over it. + * We are either changing a nop into a trace call, or a trace + * call to a nop. While the change is taking place, we treat + * it just like it was a nop. + */ +int ftrace_int3_handler(struct pt_regs *regs) { - __this_cpu_write(save_modifying_code, modifying_code); + unsigned long ip; - if (!__this_cpu_read(save_modifying_code)) - return; + if (WARN_ON_ONCE(!regs)) + return 0; - if (atomic_inc_return(&nmi_running) & MOD_CODE_WRITE_FLAG) { - smp_rmb(); - ftrace_mod_code(); - atomic_inc(&nmi_update_count); - } - /* Must have previous changes seen before executions */ - smp_mb(); + ip = regs->ip - 1; + if (!ftrace_location(ip) && !is_ftrace_caller(ip)) + return 0; + + regs->ip += MCOUNT_INSN_SIZE - 1; + + return 1; } -void ftrace_nmi_exit(void) +static int ftrace_write(unsigned long ip, const char *val, int size) { - if (!__this_cpu_read(save_modifying_code)) - return; + ip = text_ip_addr(ip); - /* Finish all executions before clearing nmi_running */ - smp_mb(); - atomic_dec(&nmi_running); + if (probe_kernel_write((void *)ip, val, size)) + return -EPERM; + + return 0; } -static void wait_for_nmi_and_set_mod_flag(void) +static int add_break(unsigned long ip, const char *old) { - if (!atomic_cmpxchg(&nmi_running, 0, MOD_CODE_WRITE_FLAG)) - return; + unsigned char replaced[MCOUNT_INSN_SIZE]; + unsigned char brk = BREAKPOINT_INSTRUCTION; - do { - cpu_relax(); - } while (atomic_cmpxchg(&nmi_running, 0, MOD_CODE_WRITE_FLAG)); + if (probe_kernel_read(replaced, (void *)ip, MCOUNT_INSN_SIZE)) + return -EFAULT; - nmi_wait_count++; + /* Make sure it is what we expect it to be */ + if (memcmp(replaced, old, MCOUNT_INSN_SIZE) != 0) + return -EINVAL; + + return ftrace_write(ip, &brk, 1); } -static void wait_for_nmi(void) +static int add_brk_on_call(struct dyn_ftrace *rec, unsigned long addr) { - if (!atomic_read(&nmi_running)) - return; + unsigned const char *old; + unsigned long ip = rec->ip; - do { - cpu_relax(); - } while (atomic_read(&nmi_running)); + old = ftrace_call_replace(ip, addr); - nmi_wait_count++; + return add_break(rec->ip, old); } -static inline int -within(unsigned long addr, unsigned long start, unsigned long end) + +static int add_brk_on_nop(struct dyn_ftrace *rec) { - return addr >= start && addr < end; + unsigned const char *old; + + old = ftrace_nop_replace(); + + return add_break(rec->ip, old); } -static int -do_ftrace_mod_code(unsigned long ip, const void *new_code) +static int add_breakpoints(struct dyn_ftrace *rec, int enable) { - /* - * On x86_64, kernel text mappings are mapped read-only with - * CONFIG_DEBUG_RODATA. So we use the kernel identity mapping instead - * of the kernel text mapping to modify the kernel text. - * - * For 32bit kernels, these mappings are same and we can use - * kernel identity mapping to modify code. - */ - if (within(ip, (unsigned long)_text, (unsigned long)_etext)) - ip = (unsigned long)__va(__pa(ip)); + unsigned long ftrace_addr; + int ret; + + ftrace_addr = ftrace_get_addr_curr(rec); + + ret = ftrace_test_record(rec, enable); - mod_code_ip = (void *)ip; - mod_code_newcode = new_code; + switch (ret) { + case FTRACE_UPDATE_IGNORE: + return 0; - /* The buffers need to be visible before we let NMIs write them */ - smp_mb(); + case FTRACE_UPDATE_MAKE_CALL: + /* converting nop to call */ + return add_brk_on_nop(rec); - wait_for_nmi_and_set_mod_flag(); + case FTRACE_UPDATE_MODIFY_CALL: + case FTRACE_UPDATE_MAKE_NOP: + /* converting a call to a nop */ + return add_brk_on_call(rec, ftrace_addr); + } + return 0; +} + +/* + * On error, we need to remove breakpoints. This needs to + * be done caefully. If the address does not currently have a + * breakpoint, we know we are done. Otherwise, we look at the + * remaining 4 bytes of the instruction. If it matches a nop + * we replace the breakpoint with the nop. Otherwise we replace + * it with the call instruction. + */ +static int remove_breakpoint(struct dyn_ftrace *rec) +{ + unsigned char ins[MCOUNT_INSN_SIZE]; + unsigned char brk = BREAKPOINT_INSTRUCTION; + const unsigned char *nop; + unsigned long ftrace_addr; + unsigned long ip = rec->ip; - /* Make sure all running NMIs have finished before we write the code */ - smp_mb(); + /* If we fail the read, just give up */ + if (probe_kernel_read(ins, (void *)ip, MCOUNT_INSN_SIZE)) + return -EFAULT; - ftrace_mod_code(); + /* If this does not have a breakpoint, we are done */ + if (ins[0] != brk) + return 0; - /* Make sure the write happens before clearing the bit */ - smp_mb(); + nop = ftrace_nop_replace(); - clear_mod_flag(); - wait_for_nmi(); + /* + * If the last 4 bytes of the instruction do not match + * a nop, then we assume that this is a call to ftrace_addr. + */ + if (memcmp(&ins[1], &nop[1], MCOUNT_INSN_SIZE - 1) != 0) { + /* + * For extra paranoidism, we check if the breakpoint is on + * a call that would actually jump to the ftrace_addr. + * If not, don't touch the breakpoint, we make just create + * a disaster. + */ + ftrace_addr = ftrace_get_addr_new(rec); + nop = ftrace_call_replace(ip, ftrace_addr); + + if (memcmp(&ins[1], &nop[1], MCOUNT_INSN_SIZE - 1) == 0) + goto update; + + /* Check both ftrace_addr and ftrace_old_addr */ + ftrace_addr = ftrace_get_addr_curr(rec); + nop = ftrace_call_replace(ip, ftrace_addr); + + if (memcmp(&ins[1], &nop[1], MCOUNT_INSN_SIZE - 1) != 0) + return -EINVAL; + } - return mod_code_status; + update: + return ftrace_write(ip, nop, 1); } -static const unsigned char *ftrace_nop_replace(void) +static int add_update_code(unsigned long ip, unsigned const char *new) { - return ideal_nops[NOP_ATOMIC5]; + /* skip breakpoint */ + ip++; + new++; + return ftrace_write(ip, new, MCOUNT_INSN_SIZE - 1); } -static int -ftrace_modify_code(unsigned long ip, unsigned const char *old_code, - unsigned const char *new_code) +static int add_update_call(struct dyn_ftrace *rec, unsigned long addr) { - unsigned char replaced[MCOUNT_INSN_SIZE]; + unsigned long ip = rec->ip; + unsigned const char *new; - /* - * Note: Due to modules and __init, code can - * disappear and change, we need to protect against faulting - * as well as code changing. We do this by using the - * probe_kernel_* functions. - * - * No real locking needed, this code is run through - * kstop_machine, or before SMP starts. - */ + new = ftrace_call_replace(ip, addr); + return add_update_code(ip, new); +} - /* read the text we want to modify */ - if (probe_kernel_read(replaced, (void *)ip, MCOUNT_INSN_SIZE)) - return -EFAULT; +static int add_update_nop(struct dyn_ftrace *rec) +{ + unsigned long ip = rec->ip; + unsigned const char *new; - /* Make sure it is what we expect it to be */ - if (memcmp(replaced, old_code, MCOUNT_INSN_SIZE) != 0) - return -EINVAL; + new = ftrace_nop_replace(); + return add_update_code(ip, new); +} - /* replace the text with the new text */ - if (do_ftrace_mod_code(ip, new_code)) - return -EPERM; +static int add_update(struct dyn_ftrace *rec, int enable) +{ + unsigned long ftrace_addr; + int ret; - sync_core(); + ret = ftrace_test_record(rec, enable); + + ftrace_addr = ftrace_get_addr_new(rec); + + switch (ret) { + case FTRACE_UPDATE_IGNORE: + return 0; + + case FTRACE_UPDATE_MODIFY_CALL: + case FTRACE_UPDATE_MAKE_CALL: + /* converting nop to call */ + return add_update_call(rec, ftrace_addr); + + case FTRACE_UPDATE_MAKE_NOP: + /* converting a call to a nop */ + return add_update_nop(rec); + } return 0; } -int ftrace_make_nop(struct module *mod, - struct dyn_ftrace *rec, unsigned long addr) +static int finish_update_call(struct dyn_ftrace *rec, unsigned long addr) { - unsigned const char *new, *old; unsigned long ip = rec->ip; + unsigned const char *new; - old = ftrace_call_replace(ip, addr); - new = ftrace_nop_replace(); + new = ftrace_call_replace(ip, addr); - return ftrace_modify_code(rec->ip, old, new); + return ftrace_write(ip, new, 1); } -int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr) +static int finish_update_nop(struct dyn_ftrace *rec) { - unsigned const char *new, *old; unsigned long ip = rec->ip; + unsigned const char *new; - old = ftrace_nop_replace(); - new = ftrace_call_replace(ip, addr); + new = ftrace_nop_replace(); - return ftrace_modify_code(rec->ip, old, new); + return ftrace_write(ip, new, 1); } -int ftrace_update_ftrace_func(ftrace_func_t func) +static int finish_update(struct dyn_ftrace *rec, int enable) { - unsigned long ip = (unsigned long)(&ftrace_call); - unsigned char old[MCOUNT_INSN_SIZE], *new; + unsigned long ftrace_addr; int ret; - memcpy(old, &ftrace_call, MCOUNT_INSN_SIZE); - new = ftrace_call_replace(ip, (unsigned long)func); - ret = ftrace_modify_code(ip, old, new); + ret = ftrace_update_record(rec, enable); + + ftrace_addr = ftrace_get_addr_new(rec); + + switch (ret) { + case FTRACE_UPDATE_IGNORE: + return 0; + + case FTRACE_UPDATE_MODIFY_CALL: + case FTRACE_UPDATE_MAKE_CALL: + /* converting nop to call */ + return finish_update_call(rec, ftrace_addr); + + case FTRACE_UPDATE_MAKE_NOP: + /* converting a call to a nop */ + return finish_update_nop(rec); + } + + return 0; +} + +static void do_sync_core(void *data) +{ + sync_core(); +} + +static void run_sync(void) +{ + int enable_irqs = irqs_disabled(); + + /* We may be called with interrupts disbled (on bootup). */ + if (enable_irqs) + local_irq_enable(); + on_each_cpu(do_sync_core, NULL, 1); + if (enable_irqs) + local_irq_disable(); +} + +void ftrace_replace_code(int enable) +{ + struct ftrace_rec_iter *iter; + struct dyn_ftrace *rec; + const char *report = "adding breakpoints"; + int count = 0; + int ret; + + for_ftrace_rec_iter(iter) { + rec = ftrace_rec_iter_record(iter); + + ret = add_breakpoints(rec, enable); + if (ret) + goto remove_breakpoints; + count++; + } + + run_sync(); + + report = "updating code"; + + for_ftrace_rec_iter(iter) { + rec = ftrace_rec_iter_record(iter); + + ret = add_update(rec, enable); + if (ret) + goto remove_breakpoints; + } + + run_sync(); + + report = "removing breakpoints"; + + for_ftrace_rec_iter(iter) { + rec = ftrace_rec_iter_record(iter); + ret = finish_update(rec, enable); + if (ret) + goto remove_breakpoints; + } + + run_sync(); + + return; + + remove_breakpoints: + pr_warn("Failed on %s (%d):\n", report, count); + ftrace_bug(ret, rec ? rec->ip : 0); + for_ftrace_rec_iter(iter) { + rec = ftrace_rec_iter_record(iter); + /* + * Breakpoints are handled only when this function is in + * progress. The system could not work with them. + */ + if (remove_breakpoint(rec)) + BUG(); + } + run_sync(); +} + +static int +ftrace_modify_code(unsigned long ip, unsigned const char *old_code, + unsigned const char *new_code) +{ + int ret; + + ret = add_break(ip, old_code); + if (ret) + goto out; + + run_sync(); + + ret = add_update_code(ip, new_code); + if (ret) + goto fail_update; + + run_sync(); + + ret = ftrace_write(ip, new_code, 1); + /* + * The breakpoint is handled only when this function is in progress. + * The system could not work if we could not remove it. + */ + BUG_ON(ret); + out: + run_sync(); return ret; + + fail_update: + /* Also here the system could not work with the breakpoint */ + if (ftrace_write(ip, old_code, 1)) + BUG(); + goto out; } -int __init ftrace_dyn_arch_init(void *data) +void arch_ftrace_update_code(int command) { - /* The return code is retured via data */ - *(unsigned long *)data = 0; + /* See comment above by declaration of modifying_ftrace_code */ + atomic_inc(&modifying_ftrace_code); + ftrace_modify_all_code(command); + + atomic_dec(&modifying_ftrace_code); +} + +int __init ftrace_dyn_arch_init(void) +{ return 0; } #endif @@ -348,45 +651,41 @@ int __init ftrace_dyn_arch_init(void *data) #ifdef CONFIG_DYNAMIC_FTRACE extern void ftrace_graph_call(void); -static int ftrace_mod_jmp(unsigned long ip, - int old_offset, int new_offset) +static unsigned char *ftrace_jmp_replace(unsigned long ip, unsigned long addr) { - unsigned char code[MCOUNT_INSN_SIZE]; + static union ftrace_code_union calc; - if (probe_kernel_read(code, (void *)ip, MCOUNT_INSN_SIZE)) - return -EFAULT; + /* Jmp not a call (ignore the .e8) */ + calc.e8 = 0xe9; + calc.offset = ftrace_calc_offset(ip + MCOUNT_INSN_SIZE, addr); - if (code[0] != 0xe9 || old_offset != *(int *)(&code[1])) - return -EINVAL; + /* + * ftrace external locks synchronize the access to the static variable. + */ + return calc.code; +} - *(int *)(&code[1]) = new_offset; +static int ftrace_mod_jmp(unsigned long ip, void *func) +{ + unsigned char *new; - if (do_ftrace_mod_code(ip, &code)) - return -EPERM; + new = ftrace_jmp_replace(ip, (unsigned long)func); - return 0; + return update_ftrace_func(ip, new); } int ftrace_enable_ftrace_graph_caller(void) { unsigned long ip = (unsigned long)(&ftrace_graph_call); - int old_offset, new_offset; - old_offset = (unsigned long)(&ftrace_stub) - (ip + MCOUNT_INSN_SIZE); - new_offset = (unsigned long)(&ftrace_graph_caller) - (ip + MCOUNT_INSN_SIZE); - - return ftrace_mod_jmp(ip, old_offset, new_offset); + return ftrace_mod_jmp(ip, &ftrace_graph_caller); } int ftrace_disable_ftrace_graph_caller(void) { unsigned long ip = (unsigned long)(&ftrace_graph_call); - int old_offset, new_offset; - - old_offset = (unsigned long)(&ftrace_graph_caller) - (ip + MCOUNT_INSN_SIZE); - new_offset = (unsigned long)(&ftrace_stub) - (ip + MCOUNT_INSN_SIZE); - return ftrace_mod_jmp(ip, old_offset, new_offset); + return ftrace_mod_jmp(ip, &ftrace_stub); } #endif /* !CONFIG_DYNAMIC_FTRACE */ diff --git a/arch/x86/kernel/head.c b/arch/x86/kernel/head.c index 48d9d4ea102..992f442ca15 100644 --- a/arch/x86/kernel/head.c +++ b/arch/x86/kernel/head.c @@ -5,8 +5,6 @@ #include <asm/setup.h> #include <asm/bios_ebda.h> -#define BIOS_LOWMEM_KILOBYTES 0x413 - /* * The BIOS places the EBDA/XBDA at the top of conventional * memory, and usually decreases the reported amount of @@ -16,17 +14,30 @@ * chipset: reserve a page before VGA to prevent PCI prefetch * into it (errata #56). Usually the page is reserved anyways, * unless you have no PS/2 mouse plugged in. + * + * This functions is deliberately very conservative. Losing + * memory in the bottom megabyte is rarely a problem, as long + * as we have enough memory to install the trampoline. Using + * memory that is in use by the BIOS or by some DMA device + * the BIOS didn't shut down *is* a big problem. */ + +#define BIOS_LOWMEM_KILOBYTES 0x413 +#define LOWMEM_CAP 0x9f000U /* Absolute maximum */ +#define INSANE_CUTOFF 0x20000U /* Less than this = insane */ + void __init reserve_ebda_region(void) { unsigned int lowmem, ebda_addr; - /* To determine the position of the EBDA and the */ - /* end of conventional memory, we need to look at */ - /* the BIOS data area. In a paravirtual environment */ - /* that area is absent. We'll just have to assume */ - /* that the paravirt case can handle memory setup */ - /* correctly, without our help. */ + /* + * To determine the position of the EBDA and the + * end of conventional memory, we need to look at + * the BIOS data area. In a paravirtual environment + * that area is absent. We'll just have to assume + * that the paravirt case can handle memory setup + * correctly, without our help. + */ if (paravirt_enabled()) return; @@ -37,19 +48,23 @@ void __init reserve_ebda_region(void) /* start of EBDA area */ ebda_addr = get_bios_ebda(); - /* Fixup: bios puts an EBDA in the top 64K segment */ - /* of conventional memory, but does not adjust lowmem. */ - if ((lowmem - ebda_addr) <= 0x10000) - lowmem = ebda_addr; + /* + * Note: some old Dells seem to need 4k EBDA without + * reporting so, so just consider the memory above 0x9f000 + * to be off limits (bugzilla 2990). + */ + + /* If the EBDA address is below 128K, assume it is bogus */ + if (ebda_addr < INSANE_CUTOFF) + ebda_addr = LOWMEM_CAP; - /* Fixup: bios does not report an EBDA at all. */ - /* Some old Dells seem to need 4k anyhow (bugzilla 2990) */ - if ((ebda_addr == 0) && (lowmem >= 0x9f000)) - lowmem = 0x9f000; + /* If lowmem is less than 128K, assume it is bogus */ + if (lowmem < INSANE_CUTOFF) + lowmem = LOWMEM_CAP; - /* Paranoia: should never happen, but... */ - if ((lowmem == 0) || (lowmem >= 0x100000)) - lowmem = 0x9f000; + /* Use the lower of the lowmem and EBDA markers as the cutoff */ + lowmem = min(lowmem, ebda_addr); + lowmem = min(lowmem, LOWMEM_CAP); /* Absolute cap */ /* reserve all memory between lowmem and the 1MB mark */ memblock_reserve(lowmem, 0x100000 - lowmem); diff --git a/arch/x86/kernel/head32.c b/arch/x86/kernel/head32.c index 51ff18616d5..d6c1b983699 100644 --- a/arch/x86/kernel/head32.c +++ b/arch/x86/kernel/head32.c @@ -14,11 +14,11 @@ #include <asm/sections.h> #include <asm/e820.h> #include <asm/page.h> -#include <asm/trampoline.h> #include <asm/apic.h> #include <asm/io_apic.h> #include <asm/bios_ebda.h> #include <asm/tlbflush.h> +#include <asm/bootparam_utils.h> static void __init i386_default_early_setup(void) { @@ -29,26 +29,14 @@ static void __init i386_default_early_setup(void) reserve_ebda_region(); } -void __init i386_start_kernel(void) +asmlinkage __visible void __init i386_start_kernel(void) { - memblock_reserve(__pa_symbol(&_text), - __pa_symbol(&__bss_stop) - __pa_symbol(&_text)); - -#ifdef CONFIG_BLK_DEV_INITRD - /* Reserve INITRD */ - if (boot_params.hdr.type_of_loader && boot_params.hdr.ramdisk_image) { - /* Assume only end is not page aligned */ - u64 ramdisk_image = boot_params.hdr.ramdisk_image; - u64 ramdisk_size = boot_params.hdr.ramdisk_size; - u64 ramdisk_end = PAGE_ALIGN(ramdisk_image + ramdisk_size); - memblock_reserve(ramdisk_image, ramdisk_end - ramdisk_image); - } -#endif + sanitize_boot_params(&boot_params); /* Call the subarch specific early setup function */ switch (boot_params.hdr.hardware_subarch) { - case X86_SUBARCH_MRST: - x86_mrst_early_setup(); + case X86_SUBARCH_INTEL_MID: + x86_intel_mid_early_setup(); break; case X86_SUBARCH_CE4100: x86_ce4100_early_setup(); @@ -58,11 +46,5 @@ void __init i386_start_kernel(void) break; } - /* - * At this point everything still needed from the boot loader - * or BIOS or kernel text should be early reserved or marked not - * RAM in e820. All other memory is free game. - */ - start_kernel(); } diff --git a/arch/x86/kernel/head64.c b/arch/x86/kernel/head64.c index 3a3b779f41d..eda1a865641 100644 --- a/arch/x86/kernel/head64.c +++ b/arch/x86/kernel/head64.c @@ -24,14 +24,86 @@ #include <asm/sections.h> #include <asm/kdebug.h> #include <asm/e820.h> -#include <asm/trampoline.h> #include <asm/bios_ebda.h> +#include <asm/bootparam_utils.h> +#include <asm/microcode.h> -static void __init zap_identity_mappings(void) +/* + * Manage page tables very early on. + */ +extern pgd_t early_level4_pgt[PTRS_PER_PGD]; +extern pmd_t early_dynamic_pgts[EARLY_DYNAMIC_PAGE_TABLES][PTRS_PER_PMD]; +static unsigned int __initdata next_early_pgt = 2; +pmdval_t early_pmd_flags = __PAGE_KERNEL_LARGE & ~(_PAGE_GLOBAL | _PAGE_NX); + +/* Wipe all early page tables except for the kernel symbol map */ +static void __init reset_early_page_tables(void) +{ + unsigned long i; + + for (i = 0; i < PTRS_PER_PGD-1; i++) + early_level4_pgt[i].pgd = 0; + + next_early_pgt = 0; + + write_cr3(__pa(early_level4_pgt)); +} + +/* Create a new PMD entry */ +int __init early_make_pgtable(unsigned long address) { - pgd_t *pgd = pgd_offset_k(0UL); - pgd_clear(pgd); - __flush_tlb_all(); + unsigned long physaddr = address - __PAGE_OFFSET; + unsigned long i; + pgdval_t pgd, *pgd_p; + pudval_t pud, *pud_p; + pmdval_t pmd, *pmd_p; + + /* Invalid address or early pgt is done ? */ + if (physaddr >= MAXMEM || read_cr3() != __pa(early_level4_pgt)) + return -1; + +again: + pgd_p = &early_level4_pgt[pgd_index(address)].pgd; + pgd = *pgd_p; + + /* + * The use of __START_KERNEL_map rather than __PAGE_OFFSET here is + * critical -- __PAGE_OFFSET would point us back into the dynamic + * range and we might end up looping forever... + */ + if (pgd) + pud_p = (pudval_t *)((pgd & PTE_PFN_MASK) + __START_KERNEL_map - phys_base); + else { + if (next_early_pgt >= EARLY_DYNAMIC_PAGE_TABLES) { + reset_early_page_tables(); + goto again; + } + + pud_p = (pudval_t *)early_dynamic_pgts[next_early_pgt++]; + for (i = 0; i < PTRS_PER_PUD; i++) + pud_p[i] = 0; + *pgd_p = (pgdval_t)pud_p - __START_KERNEL_map + phys_base + _KERNPG_TABLE; + } + pud_p += pud_index(address); + pud = *pud_p; + + if (pud) + pmd_p = (pmdval_t *)((pud & PTE_PFN_MASK) + __START_KERNEL_map - phys_base); + else { + if (next_early_pgt >= EARLY_DYNAMIC_PAGE_TABLES) { + reset_early_page_tables(); + goto again; + } + + pmd_p = (pmdval_t *)early_dynamic_pgts[next_early_pgt++]; + for (i = 0; i < PTRS_PER_PMD; i++) + pmd_p[i] = 0; + *pud_p = (pudval_t)pmd_p - __START_KERNEL_map + phys_base + _KERNPG_TABLE; + } + pmd = (physaddr & PMD_MASK) + early_pmd_flags; + pmd_p[pmd_index(address)] = pmd; + + return 0; } /* Don't add a printk in there. printk relies on the PDA which is not initialized @@ -42,18 +114,30 @@ static void __init clear_bss(void) (unsigned long) __bss_stop - (unsigned long) __bss_start); } +static unsigned long get_cmd_line_ptr(void) +{ + unsigned long cmd_line_ptr = boot_params.hdr.cmd_line_ptr; + + cmd_line_ptr |= (u64)boot_params.ext_cmd_line_ptr << 32; + + return cmd_line_ptr; +} + static void __init copy_bootdata(char *real_mode_data) { char * command_line; + unsigned long cmd_line_ptr; memcpy(&boot_params, real_mode_data, sizeof boot_params); - if (boot_params.hdr.cmd_line_ptr) { - command_line = __va(boot_params.hdr.cmd_line_ptr); + sanitize_boot_params(&boot_params); + cmd_line_ptr = get_cmd_line_ptr(); + if (cmd_line_ptr) { + command_line = __va(cmd_line_ptr); memcpy(boot_command_line, command_line, COMMAND_LINE_SIZE); } } -void __init x86_64_start_kernel(char * real_mode_data) +asmlinkage __visible void __init x86_64_start_kernel(char * real_mode_data) { int i; @@ -61,64 +145,50 @@ void __init x86_64_start_kernel(char * real_mode_data) * Build-time sanity checks on the kernel image and module * area mappings. (these are purely build-time and produce no code) */ - BUILD_BUG_ON(MODULES_VADDR < KERNEL_IMAGE_START); - BUILD_BUG_ON(MODULES_VADDR-KERNEL_IMAGE_START < KERNEL_IMAGE_SIZE); + BUILD_BUG_ON(MODULES_VADDR < __START_KERNEL_map); + BUILD_BUG_ON(MODULES_VADDR - __START_KERNEL_map < KERNEL_IMAGE_SIZE); BUILD_BUG_ON(MODULES_LEN + KERNEL_IMAGE_SIZE > 2*PUD_SIZE); - BUILD_BUG_ON((KERNEL_IMAGE_START & ~PMD_MASK) != 0); + BUILD_BUG_ON((__START_KERNEL_map & ~PMD_MASK) != 0); BUILD_BUG_ON((MODULES_VADDR & ~PMD_MASK) != 0); BUILD_BUG_ON(!(MODULES_VADDR > __START_KERNEL)); BUILD_BUG_ON(!(((MODULES_END - 1) & PGDIR_MASK) == (__START_KERNEL & PGDIR_MASK))); BUILD_BUG_ON(__fix_to_virt(__end_of_fixed_addresses) <= MODULES_END); + /* Kill off the identity-map trampoline */ + reset_early_page_tables(); + /* clear bss before set_intr_gate with early_idt_handler */ clear_bss(); - /* Make NULL pointers segfault */ - zap_identity_mappings(); + for (i = 0; i < NUM_EXCEPTION_VECTORS; i++) + set_intr_gate(i, early_idt_handlers[i]); + load_idt((const struct desc_ptr *)&idt_descr); - max_pfn_mapped = KERNEL_IMAGE_SIZE >> PAGE_SHIFT; + copy_bootdata(__va(real_mode_data)); - for (i = 0; i < NUM_EXCEPTION_VECTORS; i++) { -#ifdef CONFIG_EARLY_PRINTK - set_intr_gate(i, &early_idt_handlers[i]); -#else - set_intr_gate(i, early_idt_handler); -#endif - } - load_idt((const struct desc_ptr *)&idt_descr); + /* + * Load microcode early on BSP. + */ + load_ucode_bsp(); - if (console_loglevel == 10) + if (console_loglevel >= CONSOLE_LOGLEVEL_DEBUG) early_printk("Kernel alive\n"); + clear_page(init_level4_pgt); + /* set init_level4_pgt kernel high mapping*/ + init_level4_pgt[511] = early_level4_pgt[511]; + x86_64_start_reservations(real_mode_data); } void __init x86_64_start_reservations(char *real_mode_data) { - copy_bootdata(__va(real_mode_data)); - - memblock_reserve(__pa_symbol(&_text), - __pa_symbol(&__bss_stop) - __pa_symbol(&_text)); - -#ifdef CONFIG_BLK_DEV_INITRD - /* Reserve INITRD */ - if (boot_params.hdr.type_of_loader && boot_params.hdr.ramdisk_image) { - /* Assume only end is not page aligned */ - unsigned long ramdisk_image = boot_params.hdr.ramdisk_image; - unsigned long ramdisk_size = boot_params.hdr.ramdisk_size; - unsigned long ramdisk_end = PAGE_ALIGN(ramdisk_image + ramdisk_size); - memblock_reserve(ramdisk_image, ramdisk_end - ramdisk_image); - } -#endif + /* version is always not zero if it is copied */ + if (!boot_params.hdr.version) + copy_bootdata(__va(real_mode_data)); reserve_ebda_region(); - /* - * At this point everything still needed from the boot loader - * or BIOS or kernel text should be early reserved or marked not - * RAM in e820. All other memory is free game. - */ - start_kernel(); } diff --git a/arch/x86/kernel/head_32.S b/arch/x86/kernel/head_32.S index ce0be7cd085..f36bd42d6f0 100644 --- a/arch/x86/kernel/head_32.S +++ b/arch/x86/kernel/head_32.S @@ -21,6 +21,7 @@ #include <asm/msr-index.h> #include <asm/cpufeature.h> #include <asm/percpu.h> +#include <asm/nops.h> /* Physical address */ #define pa(X) ((X) - __PAGE_OFFSET) @@ -143,6 +144,11 @@ ENTRY(startup_32) movl %eax, pa(olpc_ofw_pgd) #endif +#ifdef CONFIG_MICROCODE_EARLY + /* Early load ucode on BSP. */ + call load_ucode_bsp +#endif + /* * Initialize page tables. This creates a PDE and a set of page * tables, which are located immediately beyond __brk_base. The variable @@ -265,6 +271,19 @@ num_subarch_entries = (. - subarch_entries) / 4 jmp default_entry #endif /* CONFIG_PARAVIRT */ +#ifdef CONFIG_HOTPLUG_CPU +/* + * Boot CPU0 entry point. It's called from play_dead(). Everything has been set + * up already except stack. We just set up stack here. Then call + * start_secondary(). + */ +ENTRY(start_cpu0) + movl stack_start, %ecx + movl %ecx, %esp + jmp *(initial_code) +ENDPROC(start_cpu0) +#endif + /* * Non-boot CPU entry point; entered from trampoline.S * We can't lgdt here, because lgdt itself uses a data segment, but @@ -273,10 +292,6 @@ num_subarch_entries = (. - subarch_entries) / 4 * If cpu hotplug is not supported then this code can go in init section * which will be freed later */ - -__CPUINIT - -#ifdef CONFIG_SMP ENTRY(startup_32_smp) cld movl $(__BOOT_DS),%eax @@ -287,33 +302,60 @@ ENTRY(startup_32_smp) movl pa(stack_start),%ecx movl %eax,%ss leal -__PAGE_OFFSET(%ecx),%esp -#endif /* CONFIG_SMP */ + +#ifdef CONFIG_MICROCODE_EARLY + /* Early load ucode on AP. */ + call load_ucode_ap +#endif + + default_entry: +#define CR0_STATE (X86_CR0_PE | X86_CR0_MP | X86_CR0_ET | \ + X86_CR0_NE | X86_CR0_WP | X86_CR0_AM | \ + X86_CR0_PG) + movl $(CR0_STATE & ~X86_CR0_PG),%eax + movl %eax,%cr0 /* - * New page tables may be in 4Mbyte page mode and may - * be using the global pages. - * - * NOTE! If we are on a 486 we may have no cr4 at all! - * So we do not try to touch it unless we really have - * some bits in it to set. This won't work if the BSP - * implements cr4 but this AP does not -- very unlikely - * but be warned! The same applies to the pse feature - * if not equally supported. --macro + * We want to start out with EFLAGS unambiguously cleared. Some BIOSes leave + * bits like NT set. This would confuse the debugger if this code is traced. So + * initialize them properly now before switching to protected mode. That means + * DF in particular (even though we have cleared it earlier after copying the + * command line) because GCC expects it. + */ + pushl $0 + popfl + +/* + * New page tables may be in 4Mbyte page mode and may be using the global pages. * - * NOTE! We have to correct for the fact that we're - * not yet offset PAGE_OFFSET.. + * NOTE! If we are on a 486 we may have no cr4 at all! Specifically, cr4 exists + * if and only if CPUID exists and has flags other than the FPU flag set. */ -#define cr4_bits pa(mmu_cr4_features) - movl cr4_bits,%edx - andl %edx,%edx - jz 6f - movl %cr4,%eax # Turn on paging options (PSE,PAE,..) - orl %edx,%eax + movl $-1,pa(X86_CPUID) # preset CPUID level + movl $X86_EFLAGS_ID,%ecx + pushl %ecx + popfl # set EFLAGS=ID + pushfl + popl %eax # get EFLAGS + testl $X86_EFLAGS_ID,%eax # did EFLAGS.ID remained set? + jz enable_paging # hw disallowed setting of ID bit + # which means no CPUID and no CR4 + + xorl %eax,%eax + cpuid + movl %eax,pa(X86_CPUID) # save largest std CPUID function + + movl $1,%eax + cpuid + andl $~1,%edx # Ignore CPUID.FPU + jz enable_paging # No flags or only CPUID.FPU = no CR4 + + movl pa(mmu_cr4_features),%eax movl %eax,%cr4 testb $X86_CR4_PAE, %al # check if PAE is enabled - jz 6f + jz enable_paging /* Check if extended functions are implemented */ movl $0x80000000, %eax @@ -321,7 +363,7 @@ default_entry: /* Value must be in the range 0x80000001 to 0x8000ffff */ subl $0x80000001, %eax cmpl $(0x8000ffff-0x80000001), %eax - ja 6f + ja enable_paging /* Clear bogus XD_DISABLE bits */ call verify_cpu @@ -330,7 +372,7 @@ default_entry: cpuid /* Execute Disable bit supported? */ btl $(X86_FEATURE_NX & 31), %edx - jnc 6f + jnc enable_paging /* Setup EFER (Extended Feature Enable Register) */ movl $MSR_EFER, %ecx @@ -340,15 +382,14 @@ default_entry: /* Make changes effective */ wrmsr -6: +enable_paging: /* * Enable paging */ movl $pa(initial_page_table), %eax movl %eax,%cr3 /* set the page table pointer.. */ - movl %cr0,%eax - orl $X86_CR0_PG,%eax + movl $CR0_STATE,%eax movl %eax,%cr0 /* ..and set paging (PG) bit */ ljmp $__BOOT_CS,$1f /* Clear prefetch and normalize %eip */ 1: @@ -356,52 +397,20 @@ default_entry: addl $__PAGE_OFFSET, %esp /* - * Initialize eflags. Some BIOS's leave bits like NT set. This would - * confuse the debugger if this code is traced. - * XXX - best to initialize before switching to protected mode. - */ - pushl $0 - popfl - -#ifdef CONFIG_SMP - cmpb $0, ready - jnz checkCPUtype -#endif /* CONFIG_SMP */ - -/* * start system 32-bit setup. We need to re-do some of the things done * in 16-bit mode for the "real" operations. */ - call setup_idt - -checkCPUtype: - - movl $-1,X86_CPUID # -1 for no CPUID initially + movl setup_once_ref,%eax + andl %eax,%eax + jz 1f # Did we do this already? + call *%eax +1: -/* check if it is 486 or 386. */ /* - * XXX - this does a lot of unnecessary setup. Alignment checks don't - * apply at our cpl of 0 and the stack ought to be aligned already, and - * we don't need to preserve eflags. + * Check if it is 486 */ - - movb $3,X86 # at least 386 - pushfl # push EFLAGS - popl %eax # get EFLAGS - movl %eax,%ecx # save original EFLAGS - xorl $0x240000,%eax # flip AC and ID bits in EFLAGS - pushl %eax # copy to EFLAGS - popfl # set EFLAGS - pushfl # get new EFLAGS - popl %eax # put it in eax - xorl %ecx,%eax # change in flags - pushl %ecx # restore original EFLAGS - popfl - testl $0x40000,%eax # check if AC bit changed - je is386 - - movb $4,X86 # at least 486 - testl $0x200000,%eax # check if ID bit changed + movb $4,X86 # at least 486 + cmpl $-1,X86_CPUID je is486 /* get vendor info */ @@ -427,16 +436,13 @@ checkCPUtype: movb %cl,X86_MASK movl %edx,X86_CAPABILITY -is486: movl $0x50022,%ecx # set AM, WP, NE and MP - jmp 2f - -is386: movl $2,%ecx # set MP -2: movl %cr0,%eax +is486: + movl $0x50022,%ecx # set AM, WP, NE and MP + movl %cr0,%eax andl $0x80000011,%eax # Save PG,PE,ET orl %ecx,%eax movl %eax,%cr0 - call check_x87 lgdt early_gdt_descr lidt idt_descr ljmp $(__KERNEL_CS),$1f @@ -450,126 +456,134 @@ is386: movl $2,%ecx # set MP movl $(__KERNEL_PERCPU), %eax movl %eax,%fs # set this cpu's percpu -#ifdef CONFIG_CC_STACKPROTECTOR - /* - * The linker can't handle this by relocation. Manually set - * base address in stack canary segment descriptor. - */ - cmpb $0,ready - jne 1f - movl $gdt_page,%eax - movl $stack_canary,%ecx - movw %cx, 8 * GDT_ENTRY_STACK_CANARY + 2(%eax) - shrl $16, %ecx - movb %cl, 8 * GDT_ENTRY_STACK_CANARY + 4(%eax) - movb %ch, 8 * GDT_ENTRY_STACK_CANARY + 7(%eax) -1: -#endif movl $(__KERNEL_STACK_CANARY),%eax movl %eax,%gs xorl %eax,%eax # Clear LDT lldt %ax - cld # gcc2 wants the direction flag cleared at all times pushl $0 # fake return address for unwinder - movb $1, ready jmp *(initial_code) -/* - * We depend on ET to be correct. This checks for 287/387. - */ -check_x87: - movb $0,X86_HARD_MATH - clts - fninit - fstsw %ax - cmpb $0,%al - je 1f - movl %cr0,%eax /* no coprocessor: have to set bits */ - xorl $4,%eax /* set EM */ - movl %eax,%cr0 - ret - ALIGN -1: movb $1,X86_HARD_MATH - .byte 0xDB,0xE4 /* fsetpm for 287, ignored by 387 */ - ret +#include "verify_cpu.S" /* - * setup_idt + * setup_once * - * sets up a idt with 256 entries pointing to - * ignore_int, interrupt gates. It doesn't actually load - * idt - that can be done only after paging has been enabled - * and the kernel moved to PAGE_OFFSET. Interrupts - * are enabled elsewhere, when we can be relatively - * sure everything is ok. + * The setup work we only want to run on the BSP. * * Warning: %esi is live across this function. */ -setup_idt: - lea ignore_int,%edx - movl $(__KERNEL_CS << 16),%eax - movw %dx,%ax /* selector = 0x0010 = cs */ - movw $0x8E00,%dx /* interrupt gate - dpl=0, present */ +__INIT +setup_once: + /* + * Set up a idt with 256 entries pointing to ignore_int, + * interrupt gates. It doesn't actually load idt - that needs + * to be done on each CPU. Interrupts are enabled elsewhere, + * when we can be relatively sure everything is ok. + */ - lea idt_table,%edi - mov $256,%ecx -rp_sidt: + movl $idt_table,%edi + movl $early_idt_handlers,%eax + movl $NUM_EXCEPTION_VECTORS,%ecx +1: movl %eax,(%edi) - movl %edx,4(%edi) + movl %eax,4(%edi) + /* interrupt gate, dpl=0, present */ + movl $(0x8E000000 + __KERNEL_CS),2(%edi) + addl $9,%eax addl $8,%edi - dec %ecx - jne rp_sidt + loop 1b -.macro set_early_handler handler,trapno - lea \handler,%edx + movl $256 - NUM_EXCEPTION_VECTORS,%ecx + movl $ignore_int,%edx movl $(__KERNEL_CS << 16),%eax - movw %dx,%ax + movw %dx,%ax /* selector = 0x0010 = cs */ movw $0x8E00,%dx /* interrupt gate - dpl=0, present */ - lea idt_table,%edi - movl %eax,8*\trapno(%edi) - movl %edx,8*\trapno+4(%edi) -.endm +2: + movl %eax,(%edi) + movl %edx,4(%edi) + addl $8,%edi + loop 2b - set_early_handler handler=early_divide_err,trapno=0 - set_early_handler handler=early_illegal_opcode,trapno=6 - set_early_handler handler=early_protection_fault,trapno=13 - set_early_handler handler=early_page_fault,trapno=14 +#ifdef CONFIG_CC_STACKPROTECTOR + /* + * Configure the stack canary. The linker can't handle this by + * relocation. Manually set base address in stack canary + * segment descriptor. + */ + movl $gdt_page,%eax + movl $stack_canary,%ecx + movw %cx, 8 * GDT_ENTRY_STACK_CANARY + 2(%eax) + shrl $16, %ecx + movb %cl, 8 * GDT_ENTRY_STACK_CANARY + 4(%eax) + movb %ch, 8 * GDT_ENTRY_STACK_CANARY + 7(%eax) +#endif + andl $0,setup_once_ref /* Once is enough, thanks */ ret -early_divide_err: - xor %edx,%edx - pushl $0 /* fake errcode */ - jmp early_fault - -early_illegal_opcode: - movl $6,%edx - pushl $0 /* fake errcode */ - jmp early_fault +ENTRY(early_idt_handlers) + # 36(%esp) %eflags + # 32(%esp) %cs + # 28(%esp) %eip + # 24(%rsp) error code + i = 0 + .rept NUM_EXCEPTION_VECTORS + .if (EXCEPTION_ERRCODE_MASK >> i) & 1 + ASM_NOP2 + .else + pushl $0 # Dummy error code, to make stack frame uniform + .endif + pushl $i # 20(%esp) Vector number + jmp early_idt_handler + i = i + 1 + .endr +ENDPROC(early_idt_handlers) + + /* This is global to keep gas from relaxing the jumps */ +ENTRY(early_idt_handler) + cld -early_protection_fault: - movl $13,%edx - jmp early_fault + cmpl $2,(%esp) # X86_TRAP_NMI + je is_nmi # Ignore NMI -early_page_fault: - movl $14,%edx - jmp early_fault + cmpl $2,%ss:early_recursion_flag + je hlt_loop + incl %ss:early_recursion_flag -early_fault: - cld -#ifdef CONFIG_PRINTK - pusha + push %eax # 16(%esp) + push %ecx # 12(%esp) + push %edx # 8(%esp) + push %ds # 4(%esp) + push %es # 0(%esp) movl $(__KERNEL_DS),%eax movl %eax,%ds movl %eax,%es - cmpl $2,early_recursion_flag - je hlt_loop - incl early_recursion_flag + + cmpl $(__KERNEL_CS),32(%esp) + jne 10f + + leal 28(%esp),%eax # Pointer to %eip + call early_fixup_exception + andl %eax,%eax + jnz ex_entry /* found an exception entry */ + +10: +#ifdef CONFIG_PRINTK + xorl %eax,%eax + movw %ax,2(%esp) /* clean up the segment values on some cpus */ + movw %ax,6(%esp) + movw %ax,34(%esp) + leal 40(%esp),%eax + pushl %eax /* %esp before the exception */ + pushl %ebx + pushl %ebp + pushl %esi + pushl %edi movl %cr2,%eax pushl %eax - pushl %edx /* trapno */ + pushl (20+6*4)(%esp) /* trapno */ pushl $fault_msg call printk #endif @@ -578,6 +592,18 @@ hlt_loop: hlt jmp hlt_loop +ex_entry: + pop %es + pop %ds + pop %edx + pop %ecx + pop %eax + decl %ss:early_recursion_flag +is_nmi: + addl $8,%esp /* drop vector number and error code */ + iret +ENDPROC(early_idt_handler) + /* This is the default interrupt "handler" :-) */ ALIGN ignore_int: @@ -611,13 +637,18 @@ ignore_int: popl %eax #endif iret +ENDPROC(ignore_int) +__INITDATA + .align 4 +early_recursion_flag: + .long 0 -#include "verify_cpu.S" - - __REFDATA -.align 4 +__REFDATA + .align 4 ENTRY(initial_code) .long i386_start_kernel +ENTRY(setup_once_ref) + .long setup_once /* * BSS section @@ -670,22 +701,19 @@ ENTRY(initial_page_table) ENTRY(stack_start) .long init_thread_union+THREAD_SIZE -early_recursion_flag: - .long 0 - -ready: .byte 0 - +__INITRODATA int_msg: .asciz "Unknown interrupt or fault at: %p %p %p\n" fault_msg: /* fault info: */ .ascii "BUG: Int %d: CR2 %p\n" -/* pusha regs: */ - .ascii " EDI %p ESI %p EBP %p ESP %p\n" - .ascii " EBX %p EDX %p ECX %p EAX %p\n" +/* regs pushed in early_idt_handler: */ + .ascii " EDI %p ESI %p EBP %p EBX %p\n" + .ascii " ESP %p ES %p DS %p\n" + .ascii " EDX %p ECX %p EAX %p\n" /* fault frame: */ - .ascii " err %p EIP %p CS %p flg %p\n" + .ascii " vec %p err %p EIP %p CS %p flg %p\n" .ascii "Stack: %p %p %p %p %p %p %p %p\n" .ascii " %p %p %p %p %p %p %p %p\n" .asciz " %p %p %p %p %p %p %p %p\n" @@ -699,6 +727,7 @@ fault_msg: * segment size, and 32-bit linear address value: */ + .data .globl boot_gdt_descr .globl idt_descr diff --git a/arch/x86/kernel/head_64.S b/arch/x86/kernel/head_64.S index 40f4eb3766d..a468c0a65c4 100644 --- a/arch/x86/kernel/head_64.S +++ b/arch/x86/kernel/head_64.S @@ -19,12 +19,15 @@ #include <asm/cache.h> #include <asm/processor-flags.h> #include <asm/percpu.h> +#include <asm/nops.h> #ifdef CONFIG_PARAVIRT #include <asm/asm-offsets.h> #include <asm/paravirt.h> +#define GET_CR2_INTO(reg) GET_CR2_INTO_RAX ; movq %rax, reg #else -#define GET_CR2_INTO_RCX movq %cr2, %rcx +#define GET_CR2_INTO(reg) movq %cr2, reg +#define INTERRUPT_RETURN iretq #endif /* we are not able to switch in one step to the final KERNEL ADDRESS SPACE @@ -44,14 +47,13 @@ L3_START_KERNEL = pud_index(__START_KERNEL_map) .code64 .globl startup_64 startup_64: - /* - * At this point the CPU runs in 64bit mode CS.L = 1 CS.D = 1, + * At this point the CPU runs in 64bit mode CS.L = 1 CS.D = 0, * and someone has loaded an identity mapped page table * for us. These identity mapped page tables map all of the * kernel pages and possibly all of memory. * - * %esi holds a physical pointer to real_mode_data. + * %rsi holds a physical pointer to real_mode_data. * * We come here either directly from a 64bit bootloader, or from * arch/x86_64/boot/compressed/head.S. @@ -63,7 +65,8 @@ startup_64: * tables and then reload them. */ - /* Compute the delta between the address I am compiled to run at and the + /* + * Compute the delta between the address I am compiled to run at and the * address I am actually running at. */ leaq _text(%rip), %rbp @@ -75,45 +78,64 @@ startup_64: testl %eax, %eax jnz bad_address - /* Is the address too large? */ - leaq _text(%rip), %rdx - movq $PGDIR_SIZE, %rax - cmpq %rax, %rdx - jae bad_address - - /* Fixup the physical addresses in the page table + /* + * Is the address too large? */ - addq %rbp, init_level4_pgt + 0(%rip) - addq %rbp, init_level4_pgt + (L4_PAGE_OFFSET*8)(%rip) - addq %rbp, init_level4_pgt + (L4_START_KERNEL*8)(%rip) + leaq _text(%rip), %rax + shrq $MAX_PHYSMEM_BITS, %rax + jnz bad_address - addq %rbp, level3_ident_pgt + 0(%rip) + /* + * Fixup the physical addresses in the page table + */ + addq %rbp, early_level4_pgt + (L4_START_KERNEL*8)(%rip) addq %rbp, level3_kernel_pgt + (510*8)(%rip) addq %rbp, level3_kernel_pgt + (511*8)(%rip) addq %rbp, level2_fixmap_pgt + (506*8)(%rip) - /* Add an Identity mapping if I am above 1G */ + /* + * Set up the identity mapping for the switchover. These + * entries should *NOT* have the global bit set! This also + * creates a bunch of nonsense entries but that is fine -- + * it avoids problems around wraparound. + */ leaq _text(%rip), %rdi - andq $PMD_PAGE_MASK, %rdi + leaq early_level4_pgt(%rip), %rbx movq %rdi, %rax - shrq $PUD_SHIFT, %rax - andq $(PTRS_PER_PUD - 1), %rax - jz ident_complete + shrq $PGDIR_SHIFT, %rax + + leaq (4096 + _KERNPG_TABLE)(%rbx), %rdx + movq %rdx, 0(%rbx,%rax,8) + movq %rdx, 8(%rbx,%rax,8) - leaq (level2_spare_pgt - __START_KERNEL_map + _KERNPG_TABLE)(%rbp), %rdx - leaq level3_ident_pgt(%rip), %rbx - movq %rdx, 0(%rbx, %rax, 8) + addq $4096, %rdx + movq %rdi, %rax + shrq $PUD_SHIFT, %rax + andl $(PTRS_PER_PUD-1), %eax + movq %rdx, 4096(%rbx,%rax,8) + incl %eax + andl $(PTRS_PER_PUD-1), %eax + movq %rdx, 4096(%rbx,%rax,8) + addq $8192, %rbx movq %rdi, %rax - shrq $PMD_SHIFT, %rax - andq $(PTRS_PER_PMD - 1), %rax - leaq __PAGE_KERNEL_IDENT_LARGE_EXEC(%rdi), %rdx - leaq level2_spare_pgt(%rip), %rbx - movq %rdx, 0(%rbx, %rax, 8) -ident_complete: + shrq $PMD_SHIFT, %rdi + addq $(__PAGE_KERNEL_LARGE_EXEC & ~_PAGE_GLOBAL), %rax + leaq (_end - 1)(%rip), %rcx + shrq $PMD_SHIFT, %rcx + subq %rdi, %rcx + incl %ecx + +1: + andq $(PTRS_PER_PMD - 1), %rdi + movq %rax, (%rbx,%rdi,8) + incq %rdi + addq $PMD_SIZE, %rax + decl %ecx + jnz 1b /* * Fixup the kernel text+data virtual addresses. Note that @@ -121,7 +143,6 @@ ident_complete: * cleanup_highmap() fixes this up along with the mappings * beyond _end. */ - leaq level2_kernel_pgt(%rip), %rdi leaq 4096(%rdi), %r8 /* See if it is a valid page table entry */ @@ -136,21 +157,14 @@ ident_complete: /* Fixup phys_base */ addq %rbp, phys_base(%rip) - /* Fixup trampoline */ - addq %rbp, trampoline_level4_pgt + 0(%rip) - addq %rbp, trampoline_level4_pgt + (511*8)(%rip) - - /* Due to ENTRY(), sometimes the empty space gets filled with - * zeros. Better take a jmp than relying on empty space being - * filled with 0x90 (nop) - */ - jmp secondary_startup_64 + movq $(early_level4_pgt - __START_KERNEL_map), %rax + jmp 1f ENTRY(secondary_startup_64) /* - * At this point the CPU runs in 64bit mode CS.L = 1 CS.D = 1, + * At this point the CPU runs in 64bit mode CS.L = 1 CS.D = 0, * and someone has loaded a mapped page table. * - * %esi holds a physical pointer to real_mode_data. + * %rsi holds a physical pointer to real_mode_data. * * We come here either from startup_64 (using physical addresses) * or from trampoline.S (using virtual addresses). @@ -160,12 +174,14 @@ ENTRY(secondary_startup_64) * after the boot processor executes this code. */ + movq $(init_level4_pgt - __START_KERNEL_map), %rax +1: + /* Enable PAE mode and PGE */ - movl $(X86_CR4_PAE | X86_CR4_PGE), %eax - movq %rax, %cr4 + movl $(X86_CR4_PAE | X86_CR4_PGE), %ecx + movq %rcx, %cr4 /* Setup early boot stage 4 level pagetables. */ - movq $(init_level4_pgt - __START_KERNEL_map), %rax addq phys_base(%rip), %rax movq %rax, %cr3 @@ -186,6 +202,7 @@ ENTRY(secondary_startup_64) btl $20,%edi /* No Execute supported? */ jnc 1f btsl $_EFER_NX, %eax + btsq $_PAGE_BIT_NX,early_pmd_flags(%rip) 1: wrmsr /* Make changes effective */ /* Setup cr0 */ @@ -197,7 +214,7 @@ ENTRY(secondary_startup_64) movq %rax, %cr0 /* Setup a boot time stack */ - movq stack_start(%rip),%rsp + movq stack_start(%rip), %rsp /* zero EFLAGS after setting rsp */ pushq $0 @@ -237,15 +254,33 @@ ENTRY(secondary_startup_64) movl initial_gs+4(%rip),%edx wrmsr - /* esi is pointer to real mode structure with interesting info. + /* rsi is pointer to real mode structure with interesting info. pass it to C */ - movl %esi, %edi + movq %rsi, %rdi /* Finally jump to run C code and to be on real kernel address * Since we are running on identity-mapped space we have to jump * to the full 64bit address, this is only possible as indirect * jump. In addition we need to ensure %cs is set so we make this * a far return. + * + * Note: do not change to far jump indirect with 64bit offset. + * + * AMD does not support far jump indirect with 64bit offset. + * AMD64 Architecture Programmer's Manual, Volume 3: states only + * JMP FAR mem16:16 FF /5 Far jump indirect, + * with the target specified by a far pointer in memory. + * JMP FAR mem16:32 FF /5 Far jump indirect, + * with the target specified by a far pointer in memory. + * + * Intel64 does support 64bit offset. + * Software Developer Manual Vol 2: states: + * FF /5 JMP m16:16 Jump far, absolute indirect, + * address given in m16:16 + * FF /5 JMP m16:32 Jump far, absolute indirect, + * address given in m16:32. + * REX.W + FF /5 JMP m16:64 Jump far, absolute indirect, + * address given in m16:64. */ movq initial_code(%rip),%rax pushq $0 # fake return address to stop unwinder @@ -253,15 +288,31 @@ ENTRY(secondary_startup_64) pushq %rax # target address in negative space lretq +#ifdef CONFIG_HOTPLUG_CPU +/* + * Boot CPU0 entry point. It's called from play_dead(). Everything has been set + * up already except stack. We just set up stack here. Then call + * start_secondary(). + */ +ENTRY(start_cpu0) + movq stack_start(%rip),%rsp + movq initial_code(%rip),%rax + pushq $0 # fake return address to stop unwinder + pushq $__KERNEL_CS # set correct cs + pushq %rax # target address in negative space + lretq +ENDPROC(start_cpu0) +#endif + /* SMP bootup changes these two */ __REFDATA - .align 8 - ENTRY(initial_code) + .balign 8 + GLOBAL(initial_code) .quad x86_64_start_kernel - ENTRY(initial_gs) + GLOBAL(initial_gs) .quad INIT_PER_CPU_VAR(irq_stack_union) - ENTRY(stack_start) + GLOBAL(stack_start) .quad init_thread_union+THREAD_SIZE-8 .word 0 __FINITDATA @@ -269,37 +320,69 @@ ENTRY(secondary_startup_64) bad_address: jmp bad_address - .section ".init.text","ax" -#ifdef CONFIG_EARLY_PRINTK + __INIT .globl early_idt_handlers early_idt_handlers: + # 104(%rsp) %rflags + # 96(%rsp) %cs + # 88(%rsp) %rip + # 80(%rsp) error code i = 0 .rept NUM_EXCEPTION_VECTORS - movl $i, %esi + .if (EXCEPTION_ERRCODE_MASK >> i) & 1 + ASM_NOP2 + .else + pushq $0 # Dummy error code, to make stack frame uniform + .endif + pushq $i # 72(%rsp) Vector number jmp early_idt_handler i = i + 1 .endr -#endif +/* This is global to keep gas from relaxing the jumps */ ENTRY(early_idt_handler) -#ifdef CONFIG_EARLY_PRINTK + cld + + cmpl $2,(%rsp) # X86_TRAP_NMI + je is_nmi # Ignore NMI + cmpl $2,early_recursion_flag(%rip) jz 1f incl early_recursion_flag(%rip) - GET_CR2_INTO_RCX - movq %rcx,%r9 - xorl %r8d,%r8d # zero for error code - movl %esi,%ecx # get vector number - # Test %ecx against mask of vectors that push error code. - cmpl $31,%ecx - ja 0f - movl $1,%eax - salq %cl,%rax - testl $0x27d00,%eax - je 0f - popq %r8 # get error code -0: movq 0(%rsp),%rcx # get ip - movq 8(%rsp),%rdx # get cs + + pushq %rax # 64(%rsp) + pushq %rcx # 56(%rsp) + pushq %rdx # 48(%rsp) + pushq %rsi # 40(%rsp) + pushq %rdi # 32(%rsp) + pushq %r8 # 24(%rsp) + pushq %r9 # 16(%rsp) + pushq %r10 # 8(%rsp) + pushq %r11 # 0(%rsp) + + cmpl $__KERNEL_CS,96(%rsp) + jne 11f + + cmpl $14,72(%rsp) # Page fault? + jnz 10f + GET_CR2_INTO(%rdi) # can clobber any volatile register if pv + call early_make_pgtable + andl %eax,%eax + jz 20f # All good + +10: + leaq 88(%rsp),%rdi # Pointer to %rip + call early_fixup_exception + andl %eax,%eax + jnz 20f # Found an exception entry + +11: +#ifdef CONFIG_EARLY_PRINTK + GET_CR2_INTO(%r9) # can clobber any volatile register if pv + movl 80(%rsp),%r8d # error code + movl 72(%rsp),%esi # vector number + movl 96(%rsp),%edx # %cs + movq 88(%rsp),%rcx # %rip xorl %eax,%eax leaq early_idt_msg(%rip),%rdi call early_printk @@ -308,27 +391,45 @@ ENTRY(early_idt_handler) call dump_stack #ifdef CONFIG_KALLSYMS leaq early_idt_ripmsg(%rip),%rdi - movq 0(%rsp),%rsi # get rip again + movq 40(%rsp),%rsi # %rip again call __print_symbol #endif #endif /* EARLY_PRINTK */ 1: hlt jmp 1b -#ifdef CONFIG_EARLY_PRINTK +20: # Exception table entry found or page table generated + popq %r11 + popq %r10 + popq %r9 + popq %r8 + popq %rdi + popq %rsi + popq %rdx + popq %rcx + popq %rax + decl early_recursion_flag(%rip) +is_nmi: + addq $16,%rsp # drop vector number and error code + INTERRUPT_RETURN +ENDPROC(early_idt_handler) + + __INITDATA + + .balign 4 early_recursion_flag: .long 0 +#ifdef CONFIG_EARLY_PRINTK early_idt_msg: .asciz "PANIC: early exception %02lx rip %lx:%lx error %lx cr2 %lx\n" early_idt_ripmsg: .asciz "RIP %s\n" #endif /* CONFIG_EARLY_PRINTK */ - .previous #define NEXT_PAGE(name) \ .balign PAGE_SIZE; \ -ENTRY(name) +GLOBAL(name) /* Automate the creation of 1 to 1 mapping pmd entries */ #define PMDS(START, PERM, COUNT) \ @@ -338,24 +439,37 @@ ENTRY(name) i = i + 1 ; \ .endr + __INITDATA +NEXT_PAGE(early_level4_pgt) + .fill 511,8,0 + .quad level3_kernel_pgt - __START_KERNEL_map + _PAGE_TABLE + +NEXT_PAGE(early_dynamic_pgts) + .fill 512*EARLY_DYNAMIC_PAGE_TABLES,8,0 + .data - /* - * This default setting generates an ident mapping at address 0x100000 - * and a mapping for the kernel that precisely maps virtual address - * 0xffffffff80000000 to physical address 0x000000. (always using - * 2Mbyte large pages provided by PAE mode) - */ + +#ifndef CONFIG_XEN +NEXT_PAGE(init_level4_pgt) + .fill 512,8,0 +#else NEXT_PAGE(init_level4_pgt) - .quad level3_ident_pgt - __START_KERNEL_map + _KERNPG_TABLE - .org init_level4_pgt + L4_PAGE_OFFSET*8, 0 - .quad level3_ident_pgt - __START_KERNEL_map + _KERNPG_TABLE - .org init_level4_pgt + L4_START_KERNEL*8, 0 + .quad level3_ident_pgt - __START_KERNEL_map + _KERNPG_TABLE + .org init_level4_pgt + L4_PAGE_OFFSET*8, 0 + .quad level3_ident_pgt - __START_KERNEL_map + _KERNPG_TABLE + .org init_level4_pgt + L4_START_KERNEL*8, 0 /* (2^48-(2*1024*1024*1024))/(2^39) = 511 */ - .quad level3_kernel_pgt - __START_KERNEL_map + _PAGE_TABLE + .quad level3_kernel_pgt - __START_KERNEL_map + _PAGE_TABLE NEXT_PAGE(level3_ident_pgt) .quad level2_ident_pgt - __START_KERNEL_map + _KERNPG_TABLE - .fill 511,8,0 + .fill 511, 8, 0 +NEXT_PAGE(level2_ident_pgt) + /* Since I easily can, map the first 1G. + * Don't set NX because code runs from these pages. + */ + PMDS(0, __PAGE_KERNEL_IDENT_LARGE_EXEC, PTRS_PER_PMD) +#endif NEXT_PAGE(level3_kernel_pgt) .fill L3_START_KERNEL,8,0 @@ -363,21 +477,6 @@ NEXT_PAGE(level3_kernel_pgt) .quad level2_kernel_pgt - __START_KERNEL_map + _KERNPG_TABLE .quad level2_fixmap_pgt - __START_KERNEL_map + _PAGE_TABLE -NEXT_PAGE(level2_fixmap_pgt) - .fill 506,8,0 - .quad level1_fixmap_pgt - __START_KERNEL_map + _PAGE_TABLE - /* 8MB reserved for vsyscalls + a 2MB hole = 4 + 1 entries */ - .fill 5,8,0 - -NEXT_PAGE(level1_fixmap_pgt) - .fill 512,8,0 - -NEXT_PAGE(level2_ident_pgt) - /* Since I easily can, map the first 1G. - * Don't set NX because code runs from these pages. - */ - PMDS(0, __PAGE_KERNEL_IDENT_LARGE_EXEC, PTRS_PER_PMD) - NEXT_PAGE(level2_kernel_pgt) /* * 512 MB kernel mapping. We spend a full page on this pagetable @@ -392,11 +491,16 @@ NEXT_PAGE(level2_kernel_pgt) PMDS(0, __PAGE_KERNEL_LARGE_EXEC, KERNEL_IMAGE_SIZE/PMD_SIZE) -NEXT_PAGE(level2_spare_pgt) - .fill 512, 8, 0 +NEXT_PAGE(level2_fixmap_pgt) + .fill 506,8,0 + .quad level1_fixmap_pgt - __START_KERNEL_map + _PAGE_TABLE + /* 8MB reserved for vsyscalls + a 2MB hole = 4 + 1 entries */ + .fill 5,8,0 + +NEXT_PAGE(level1_fixmap_pgt) + .fill 512,8,0 #undef PMDS -#undef NEXT_PAGE .data .align 16 @@ -412,16 +516,6 @@ ENTRY(phys_base) #include "../../x86/xen/xen-head.S" - .section .bss, "aw", @nobits - .align L1_CACHE_BYTES -ENTRY(idt_table) - .skip IDT_ENTRIES * 16 - - .align L1_CACHE_BYTES -ENTRY(nmi_idt_table) - .skip IDT_ENTRIES * 16 - __PAGE_ALIGNED_BSS - .align PAGE_SIZE -ENTRY(empty_zero_page) +NEXT_PAGE(empty_zero_page) .skip PAGE_SIZE diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c index ad0de0c2714..319bcb9372f 100644 --- a/arch/x86/kernel/hpet.c +++ b/arch/x86/kernel/hpet.c @@ -74,9 +74,6 @@ static inline void hpet_writel(unsigned int d, unsigned int a) static inline void hpet_set_mapping(void) { hpet_virt_address = ioremap_nocache(hpet_address, HPET_MMAP_SIZE); -#ifdef CONFIG_X86_64 - __set_fixmap(VSYSCALL_HPET, hpet_address, PAGE_KERNEL_VVAR_NOCACHE); -#endif } static inline void hpet_clear_mapping(void) @@ -88,19 +85,24 @@ static inline void hpet_clear_mapping(void) /* * HPET command line enable / disable */ -static int boot_hpet_disable; +int boot_hpet_disable; int hpet_force_user; static int hpet_verbose; static int __init hpet_setup(char *str) { - if (str) { + while (str) { + char *next = strchr(str, ','); + + if (next) + *next++ = 0; if (!strncmp("disable", str, 7)) boot_hpet_disable = 1; if (!strncmp("force", str, 5)) hpet_force_user = 1; if (!strncmp("verbose", str, 7)) hpet_verbose = 1; + str = next; } return 1; } @@ -319,8 +321,6 @@ static void hpet_set_mode(enum clock_event_mode mode, now = hpet_readl(HPET_COUNTER); cmp = now + (unsigned int) delta; cfg = hpet_readl(HPET_Tn_CFG(timer)); - /* Make sure we use edge triggered interrupts */ - cfg &= ~HPET_TN_LEVEL; cfg |= HPET_TN_ENABLE | HPET_TN_PERIODIC | HPET_TN_SETVAL | HPET_TN_32BIT; hpet_writel(cfg, HPET_Tn_CFG(timer)); @@ -431,7 +431,7 @@ void hpet_msi_unmask(struct irq_data *data) /* unmask it */ cfg = hpet_readl(HPET_Tn_CFG(hdev->num)); - cfg |= HPET_TN_FSB; + cfg |= HPET_TN_ENABLE | HPET_TN_FSB; hpet_writel(cfg, HPET_Tn_CFG(hdev->num)); } @@ -442,7 +442,7 @@ void hpet_msi_mask(struct irq_data *data) /* mask it */ cfg = hpet_readl(HPET_Tn_CFG(hdev->num)); - cfg &= ~HPET_TN_FSB; + cfg &= ~(HPET_TN_ENABLE | HPET_TN_FSB); hpet_writel(cfg, HPET_Tn_CFG(hdev->num)); } @@ -475,8 +475,8 @@ static int hpet_msi_next_event(unsigned long delta, static int hpet_setup_msi_irq(unsigned int irq) { - if (arch_setup_hpet_msi(irq, hpet_blockid)) { - destroy_irq(irq); + if (x86_msi.setup_hpet_msi(irq, hpet_blockid)) { + irq_free_hwirq(irq); return -EINVAL; } return 0; @@ -484,9 +484,8 @@ static int hpet_setup_msi_irq(unsigned int irq) static int hpet_assign_irq(struct hpet_dev *dev) { - unsigned int irq; + unsigned int irq = irq_alloc_hwirq(-1); - irq = create_irq_nr(0, -1); if (!irq) return -EINVAL; @@ -518,7 +517,7 @@ static int hpet_setup_irq(struct hpet_dev *dev) { if (request_irq(dev->irq, hpet_interrupt_handler, - IRQF_TIMER | IRQF_DISABLED | IRQF_NOBALANCING, + IRQF_TIMER | IRQF_NOBALANCING, dev->name, dev)) return -1; @@ -696,7 +695,7 @@ static int hpet_cpuhp_notify(struct notifier_block *n, /* FIXME: add schedule_work_on() */ schedule_delayed_work_on(cpu, &work.work, 0); wait_for_completion(&work.complete); - destroy_timer_on_stack(&work.work.timer); + destroy_delayed_work_on_stack(&work.work); break; case CPU_DEAD: if (hdev) { @@ -749,9 +748,7 @@ static struct clocksource clocksource_hpet = { .mask = HPET_MASK, .flags = CLOCK_SOURCE_IS_CONTINUOUS, .resume = hpet_resume_counter, -#ifdef CONFIG_X86_64 .archdata = { .vclock_mode = VCLOCK_HPET }, -#endif }; static int hpet_clocksource_register(void) @@ -787,15 +784,16 @@ static int hpet_clocksource_register(void) return 0; } +static u32 *hpet_boot_cfg; + /** * hpet_enable - Try to setup the HPET timer. Returns 1 on success. */ int __init hpet_enable(void) { - unsigned long hpet_period; - unsigned int id; + u32 hpet_period, cfg, id; u64 freq; - int i; + unsigned int i, last; if (!is_hpet_capable()) return 0; @@ -847,15 +845,45 @@ int __init hpet_enable(void) id = hpet_readl(HPET_ID); hpet_print_config(); + last = (id & HPET_ID_NUMBER) >> HPET_ID_NUMBER_SHIFT; + #ifdef CONFIG_HPET_EMULATE_RTC /* * The legacy routing mode needs at least two channels, tick timer * and the rtc emulation channel. */ - if (!(id & HPET_ID_NUMBER)) + if (!last) goto out_nohpet; #endif + cfg = hpet_readl(HPET_CFG); + hpet_boot_cfg = kmalloc((last + 2) * sizeof(*hpet_boot_cfg), + GFP_KERNEL); + if (hpet_boot_cfg) + *hpet_boot_cfg = cfg; + else + pr_warn("HPET initial state will not be saved\n"); + cfg &= ~(HPET_CFG_ENABLE | HPET_CFG_LEGACY); + hpet_writel(cfg, HPET_CFG); + if (cfg) + pr_warn("HPET: Unrecognized bits %#x set in global cfg\n", + cfg); + + for (i = 0; i <= last; ++i) { + cfg = hpet_readl(HPET_Tn_CFG(i)); + if (hpet_boot_cfg) + hpet_boot_cfg[i + 1] = cfg; + cfg &= ~(HPET_TN_ENABLE | HPET_TN_LEVEL | HPET_TN_FSB); + hpet_writel(cfg, HPET_Tn_CFG(i)); + cfg &= ~(HPET_TN_PERIODIC | HPET_TN_PERIODIC_CAP + | HPET_TN_64BIT_CAP | HPET_TN_32BIT | HPET_TN_ROUTE + | HPET_TN_FSB | HPET_TN_FSB_CAP); + if (cfg) + pr_warn("HPET: Unrecognized bits %#x set in cfg#%u\n", + cfg, i); + } + hpet_print_config(); + if (hpet_clocksource_register()) goto out_nohpet; @@ -909,12 +937,14 @@ static __init int hpet_late_init(void) if (boot_cpu_has(X86_FEATURE_ARAT)) return 0; + cpu_notifier_register_begin(); for_each_online_cpu(cpu) { hpet_cpuhp_notify(NULL, CPU_ONLINE, (void *)(long)cpu); } /* This notifier should be called after workqueue is ready */ - hotcpu_notifier(hpet_cpuhp_notify, -20); + __hotcpu_notifier(hpet_cpuhp_notify, -20); + cpu_notifier_register_done(); return 0; } @@ -923,14 +953,28 @@ fs_initcall(hpet_late_init); void hpet_disable(void) { if (is_hpet_capable() && hpet_virt_address) { - unsigned int cfg = hpet_readl(HPET_CFG); + unsigned int cfg = hpet_readl(HPET_CFG), id, last; - if (hpet_legacy_int_enabled) { + if (hpet_boot_cfg) + cfg = *hpet_boot_cfg; + else if (hpet_legacy_int_enabled) { cfg &= ~HPET_CFG_LEGACY; hpet_legacy_int_enabled = 0; } cfg &= ~HPET_CFG_ENABLE; hpet_writel(cfg, HPET_CFG); + + if (!hpet_boot_cfg) + return; + + id = hpet_readl(HPET_ID); + last = ((id & HPET_ID_NUMBER) >> HPET_ID_NUMBER_SHIFT); + + for (id = 0; id <= last; ++id) + hpet_writel(hpet_boot_cfg[id + 1], HPET_Tn_CFG(id)); + + if (*hpet_boot_cfg & HPET_CFG_ENABLE) + hpet_writel(*hpet_boot_cfg, HPET_CFG); } } diff --git a/arch/x86/kernel/hw_breakpoint.c b/arch/x86/kernel/hw_breakpoint.c index 02f07634d26..5f9cf20cdb6 100644 --- a/arch/x86/kernel/hw_breakpoint.c +++ b/arch/x86/kernel/hw_breakpoint.c @@ -32,13 +32,11 @@ #include <linux/irqflags.h> #include <linux/notifier.h> #include <linux/kallsyms.h> -#include <linux/kprobes.h> #include <linux/percpu.h> #include <linux/kdebug.h> #include <linux/kernel.h> #include <linux/module.h> #include <linux/sched.h> -#include <linux/init.h> #include <linux/smp.h> #include <asm/hw_breakpoint.h> @@ -393,6 +391,9 @@ void flush_ptrace_hw_breakpoint(struct task_struct *tsk) unregister_hw_breakpoint(t->ptrace_bps[i]); t->ptrace_bps[i] = NULL; } + + t->debugreg6 = 0; + t->ptrace_dr7 = 0; } void hw_breakpoint_restore(void) @@ -422,7 +423,7 @@ EXPORT_SYMBOL_GPL(hw_breakpoint_restore); * NOTIFY_STOP returned for all other cases * */ -static int __kprobes hw_breakpoint_handler(struct die_args *args) +static int hw_breakpoint_handler(struct die_args *args) { int i, cpu, rc = NOTIFY_STOP; struct perf_event *bp; @@ -509,7 +510,7 @@ static int __kprobes hw_breakpoint_handler(struct die_args *args) /* * Handle debug exception notifications. */ -int __kprobes hw_breakpoint_exceptions_notify( +int hw_breakpoint_exceptions_notify( struct notifier_block *unused, unsigned long val, void *data) { if (val != DIE_DEBUG) diff --git a/arch/x86/kernel/i386_ksyms_32.c b/arch/x86/kernel/i386_ksyms_32.c index 9c3bd4a2050..05fd74f537d 100644 --- a/arch/x86/kernel/i386_ksyms_32.c +++ b/arch/x86/kernel/i386_ksyms_32.c @@ -26,6 +26,7 @@ EXPORT_SYMBOL(csum_partial_copy_generic); EXPORT_SYMBOL(__get_user_1); EXPORT_SYMBOL(__get_user_2); EXPORT_SYMBOL(__get_user_4); +EXPORT_SYMBOL(__get_user_8); EXPORT_SYMBOL(__put_user_1); EXPORT_SYMBOL(__put_user_2); @@ -36,3 +37,10 @@ EXPORT_SYMBOL(strstr); EXPORT_SYMBOL(csum_partial); EXPORT_SYMBOL(empty_zero_page); + +#ifdef CONFIG_PREEMPT +EXPORT_SYMBOL(___preempt_schedule); +#ifdef CONFIG_CONTEXT_TRACKING +EXPORT_SYMBOL(___preempt_schedule_context); +#endif +#endif diff --git a/arch/x86/kernel/i387.c b/arch/x86/kernel/i387.c index 739d8598f78..d5dd8081441 100644 --- a/arch/x86/kernel/i387.c +++ b/arch/x86/kernel/i387.c @@ -16,58 +16,131 @@ #include <asm/uaccess.h> #include <asm/ptrace.h> #include <asm/i387.h> +#include <asm/fpu-internal.h> #include <asm/user.h> -#ifdef CONFIG_X86_64 -# include <asm/sigcontext32.h> -# include <asm/user32.h> -#else -# define save_i387_xstate_ia32 save_i387_xstate -# define restore_i387_xstate_ia32 restore_i387_xstate -# define _fpstate_ia32 _fpstate -# define _xstate_ia32 _xstate -# define sig_xstate_ia32_size sig_xstate_size -# define fx_sw_reserved_ia32 fx_sw_reserved -# define user_i387_ia32_struct user_i387_struct -# define user32_fxsr_struct user_fxsr_struct -#endif +/* + * Were we in an interrupt that interrupted kernel mode? + * + * On others, we can do a kernel_fpu_begin/end() pair *ONLY* if that + * pair does nothing at all: the thread must not have fpu (so + * that we don't try to save the FPU state), and TS must + * be set (so that the clts/stts pair does nothing that is + * visible in the interrupted kernel thread). + * + * Except for the eagerfpu case when we return 1 unless we've already + * been eager and saved the state in kernel_fpu_begin(). + */ +static inline bool interrupted_kernel_fpu_idle(void) +{ + if (use_eager_fpu()) + return __thread_has_fpu(current); -#ifdef CONFIG_MATH_EMULATION -# define HAVE_HWFP (boot_cpu_data.hard_math) -#else -# define HAVE_HWFP 1 -#endif + return !__thread_has_fpu(current) && + (read_cr0() & X86_CR0_TS); +} + +/* + * Were we in user mode (or vm86 mode) when we were + * interrupted? + * + * Doing kernel_fpu_begin/end() is ok if we are running + * in an interrupt context from user mode - we'll just + * save the FPU state as required. + */ +static inline bool interrupted_user_mode(void) +{ + struct pt_regs *regs = get_irq_regs(); + return regs && user_mode_vm(regs); +} + +/* + * Can we use the FPU in kernel mode with the + * whole "kernel_fpu_begin/end()" sequence? + * + * It's always ok in process context (ie "not interrupt") + * but it is sometimes ok even from an irq. + */ +bool irq_fpu_usable(void) +{ + return !in_interrupt() || + interrupted_user_mode() || + interrupted_kernel_fpu_idle(); +} +EXPORT_SYMBOL(irq_fpu_usable); + +void __kernel_fpu_begin(void) +{ + struct task_struct *me = current; + + if (__thread_has_fpu(me)) { + __thread_clear_has_fpu(me); + __save_init_fpu(me); + /* We do 'stts()' in __kernel_fpu_end() */ + } else if (!use_eager_fpu()) { + this_cpu_write(fpu_owner_task, NULL); + clts(); + } +} +EXPORT_SYMBOL(__kernel_fpu_begin); + +void __kernel_fpu_end(void) +{ + if (use_eager_fpu()) { + /* + * For eager fpu, most the time, tsk_used_math() is true. + * Restore the user math as we are done with the kernel usage. + * At few instances during thread exit, signal handling etc, + * tsk_used_math() is false. Those few places will take proper + * actions, so we don't need to restore the math here. + */ + if (likely(tsk_used_math(current))) + math_state_restore(); + } else { + stts(); + } +} +EXPORT_SYMBOL(__kernel_fpu_end); + +void unlazy_fpu(struct task_struct *tsk) +{ + preempt_disable(); + if (__thread_has_fpu(tsk)) { + __save_init_fpu(tsk); + __thread_fpu_end(tsk); + } else + tsk->thread.fpu_counter = 0; + preempt_enable(); +} +EXPORT_SYMBOL(unlazy_fpu); -static unsigned int mxcsr_feature_mask __read_mostly = 0xffffffffu; +unsigned int mxcsr_feature_mask __read_mostly = 0xffffffffu; unsigned int xstate_size; EXPORT_SYMBOL_GPL(xstate_size); -unsigned int sig_xstate_ia32_size = sizeof(struct _fpstate_ia32); -static struct i387_fxsave_struct fx_scratch __cpuinitdata; +static struct i387_fxsave_struct fx_scratch; -void __cpuinit mxcsr_feature_mask_init(void) +static void mxcsr_feature_mask_init(void) { unsigned long mask = 0; - clts(); if (cpu_has_fxsr) { memset(&fx_scratch, 0, sizeof(struct i387_fxsave_struct)); - asm volatile("fxsave %0" : : "m" (fx_scratch)); + asm volatile("fxsave %0" : "+m" (fx_scratch)); mask = fx_scratch.mxcsr_mask; if (mask == 0) mask = 0x0000ffbf; } mxcsr_feature_mask &= mask; - stts(); } -static void __cpuinit init_thread_xstate(void) +static void init_thread_xstate(void) { /* * Note that xstate_size might be overwriten later during * xsave_init(). */ - if (!HAVE_HWFP) { + if (!cpu_has_fpu) { /* * Disable xsave as we do not support it if i387 * emulation is enabled. @@ -89,11 +162,19 @@ static void __cpuinit init_thread_xstate(void) * into all processes. */ -void __cpuinit fpu_init(void) +void fpu_init(void) { unsigned long cr0; unsigned long cr4_mask = 0; +#ifndef CONFIG_MATH_EMULATION + if (!cpu_has_fpu) { + pr_emerg("No FPU found and no math emulation present\n"); + pr_emerg("Giving up\n"); + for (;;) + asm volatile("hlt"); + } +#endif if (cpu_has_fxsr) cr4_mask |= X86_CR4_OSFXSR; if (cpu_has_xmm) @@ -103,33 +184,31 @@ void __cpuinit fpu_init(void) cr0 = read_cr0(); cr0 &= ~(X86_CR0_TS|X86_CR0_EM); /* clear TS and EM */ - if (!HAVE_HWFP) + if (!cpu_has_fpu) cr0 |= X86_CR0_EM; write_cr0(cr0); - if (!smp_processor_id()) + /* + * init_thread_xstate is only called once to avoid overriding + * xstate_size during boot time or during CPU hotplug. + */ + if (xstate_size == 0) init_thread_xstate(); mxcsr_feature_mask_init(); - /* clean state in init */ - current_thread_info()->status = 0; - clear_used_math(); + xsave_init(); + eager_fpu_init(); } void fpu_finit(struct fpu *fpu) { - if (!HAVE_HWFP) { + if (!cpu_has_fpu) { finit_soft_fpu(&fpu->state->soft); return; } if (cpu_has_fxsr) { - struct i387_fxsave_struct *fx = &fpu->state->fxsave; - - memset(fx, 0, xstate_size); - fx->cwd = 0x37f; - if (cpu_has_xmm) - fx->mxcsr = MXCSR_DEFAULT; + fx_finit(&fpu->state->fxsave); } else { struct i387_fsave_struct *fp = &fpu->state->fsave; memset(fp, 0, xstate_size); @@ -152,8 +231,9 @@ int init_fpu(struct task_struct *tsk) int ret; if (tsk_used_math(tsk)) { - if (HAVE_HWFP && tsk == current) + if (cpu_has_fpu && tsk == current) unlazy_fpu(tsk); + tsk->thread.fpu.last_cpu = ~0; return 0; } @@ -372,7 +452,7 @@ static inline u32 twd_fxsr_to_i387(struct i387_fxsave_struct *fxsave) * FXSR floating point environment conversions. */ -static void +void convert_from_fxsr(struct user_i387_ia32_struct *env, struct task_struct *tsk) { struct i387_fxsave_struct *fxsave = &tsk->thread.fpu.state->fxsave; @@ -409,8 +489,8 @@ convert_from_fxsr(struct user_i387_ia32_struct *env, struct task_struct *tsk) memcpy(&to[i], &from[i], sizeof(to[0])); } -static void convert_to_fxsr(struct task_struct *tsk, - const struct user_i387_ia32_struct *env) +void convert_to_fxsr(struct task_struct *tsk, + const struct user_i387_ia32_struct *env) { struct i387_fxsave_struct *fxsave = &tsk->thread.fpu.state->fxsave; @@ -448,14 +528,13 @@ int fpregs_get(struct task_struct *target, const struct user_regset *regset, if (ret) return ret; - if (!HAVE_HWFP) + if (!static_cpu_has(X86_FEATURE_FPU)) return fpregs_soft_get(target, regset, pos, count, kbuf, ubuf); - if (!cpu_has_fxsr) { + if (!cpu_has_fxsr) return user_regset_copyout(&pos, &count, &kbuf, &ubuf, &target->thread.fpu.state->fsave, 0, -1); - } sanitize_i387_state(target); @@ -482,13 +561,13 @@ int fpregs_set(struct task_struct *target, const struct user_regset *regset, sanitize_i387_state(target); - if (!HAVE_HWFP) + if (!static_cpu_has(X86_FEATURE_FPU)) return fpregs_soft_set(target, regset, pos, count, kbuf, ubuf); - if (!cpu_has_fxsr) { + if (!cpu_has_fxsr) return user_regset_copyin(&pos, &count, &kbuf, &ubuf, - &target->thread.fpu.state->fsave, 0, -1); - } + &target->thread.fpu.state->fsave, 0, + -1); if (pos > 0 || count < sizeof(env)) convert_from_fxsr(&env, target); @@ -507,223 +586,6 @@ int fpregs_set(struct task_struct *target, const struct user_regset *regset, } /* - * Signal frame handlers. - */ - -static inline int save_i387_fsave(struct _fpstate_ia32 __user *buf) -{ - struct task_struct *tsk = current; - struct i387_fsave_struct *fp = &tsk->thread.fpu.state->fsave; - - fp->status = fp->swd; - if (__copy_to_user(buf, fp, sizeof(struct i387_fsave_struct))) - return -1; - return 1; -} - -static int save_i387_fxsave(struct _fpstate_ia32 __user *buf) -{ - struct task_struct *tsk = current; - struct i387_fxsave_struct *fx = &tsk->thread.fpu.state->fxsave; - struct user_i387_ia32_struct env; - int err = 0; - - convert_from_fxsr(&env, tsk); - if (__copy_to_user(buf, &env, sizeof(env))) - return -1; - - err |= __put_user(fx->swd, &buf->status); - err |= __put_user(X86_FXSR_MAGIC, &buf->magic); - if (err) - return -1; - - if (__copy_to_user(&buf->_fxsr_env[0], fx, xstate_size)) - return -1; - return 1; -} - -static int save_i387_xsave(void __user *buf) -{ - struct task_struct *tsk = current; - struct _fpstate_ia32 __user *fx = buf; - int err = 0; - - - sanitize_i387_state(tsk); - - /* - * For legacy compatible, we always set FP/SSE bits in the bit - * vector while saving the state to the user context. - * This will enable us capturing any changes(during sigreturn) to - * the FP/SSE bits by the legacy applications which don't touch - * xstate_bv in the xsave header. - * - * xsave aware applications can change the xstate_bv in the xsave - * header as well as change any contents in the memory layout. - * xrestore as part of sigreturn will capture all the changes. - */ - tsk->thread.fpu.state->xsave.xsave_hdr.xstate_bv |= XSTATE_FPSSE; - - if (save_i387_fxsave(fx) < 0) - return -1; - - err = __copy_to_user(&fx->sw_reserved, &fx_sw_reserved_ia32, - sizeof(struct _fpx_sw_bytes)); - err |= __put_user(FP_XSTATE_MAGIC2, - (__u32 __user *) (buf + sig_xstate_ia32_size - - FP_XSTATE_MAGIC2_SIZE)); - if (err) - return -1; - - return 1; -} - -int save_i387_xstate_ia32(void __user *buf) -{ - struct _fpstate_ia32 __user *fp = (struct _fpstate_ia32 __user *) buf; - struct task_struct *tsk = current; - - if (!used_math()) - return 0; - - if (!access_ok(VERIFY_WRITE, buf, sig_xstate_ia32_size)) - return -EACCES; - /* - * This will cause a "finit" to be triggered by the next - * attempted FPU operation by the 'current' process. - */ - clear_used_math(); - - if (!HAVE_HWFP) { - return fpregs_soft_get(current, NULL, - 0, sizeof(struct user_i387_ia32_struct), - NULL, fp) ? -1 : 1; - } - - unlazy_fpu(tsk); - - if (cpu_has_xsave) - return save_i387_xsave(fp); - if (cpu_has_fxsr) - return save_i387_fxsave(fp); - else - return save_i387_fsave(fp); -} - -static inline int restore_i387_fsave(struct _fpstate_ia32 __user *buf) -{ - struct task_struct *tsk = current; - - return __copy_from_user(&tsk->thread.fpu.state->fsave, buf, - sizeof(struct i387_fsave_struct)); -} - -static int restore_i387_fxsave(struct _fpstate_ia32 __user *buf, - unsigned int size) -{ - struct task_struct *tsk = current; - struct user_i387_ia32_struct env; - int err; - - err = __copy_from_user(&tsk->thread.fpu.state->fxsave, &buf->_fxsr_env[0], - size); - /* mxcsr reserved bits must be masked to zero for security reasons */ - tsk->thread.fpu.state->fxsave.mxcsr &= mxcsr_feature_mask; - if (err || __copy_from_user(&env, buf, sizeof(env))) - return 1; - convert_to_fxsr(tsk, &env); - - return 0; -} - -static int restore_i387_xsave(void __user *buf) -{ - struct _fpx_sw_bytes fx_sw_user; - struct _fpstate_ia32 __user *fx_user = - ((struct _fpstate_ia32 __user *) buf); - struct i387_fxsave_struct __user *fx = - (struct i387_fxsave_struct __user *) &fx_user->_fxsr_env[0]; - struct xsave_hdr_struct *xsave_hdr = - ¤t->thread.fpu.state->xsave.xsave_hdr; - u64 mask; - int err; - - if (check_for_xstate(fx, buf, &fx_sw_user)) - goto fx_only; - - mask = fx_sw_user.xstate_bv; - - err = restore_i387_fxsave(buf, fx_sw_user.xstate_size); - - xsave_hdr->xstate_bv &= pcntxt_mask; - /* - * These bits must be zero. - */ - xsave_hdr->reserved1[0] = xsave_hdr->reserved1[1] = 0; - - /* - * Init the state that is not present in the memory layout - * and enabled by the OS. - */ - mask = ~(pcntxt_mask & ~mask); - xsave_hdr->xstate_bv &= mask; - - return err; -fx_only: - /* - * Couldn't find the extended state information in the memory - * layout. Restore the FP/SSE and init the other extended state - * enabled by the OS. - */ - xsave_hdr->xstate_bv = XSTATE_FPSSE; - return restore_i387_fxsave(buf, sizeof(struct i387_fxsave_struct)); -} - -int restore_i387_xstate_ia32(void __user *buf) -{ - int err; - struct task_struct *tsk = current; - struct _fpstate_ia32 __user *fp = (struct _fpstate_ia32 __user *) buf; - - if (HAVE_HWFP) - clear_fpu(tsk); - - if (!buf) { - if (used_math()) { - clear_fpu(tsk); - clear_used_math(); - } - - return 0; - } else - if (!access_ok(VERIFY_READ, buf, sig_xstate_ia32_size)) - return -EACCES; - - if (!used_math()) { - err = init_fpu(tsk); - if (err) - return err; - } - - if (HAVE_HWFP) { - if (cpu_has_xsave) - err = restore_i387_xsave(buf); - else if (cpu_has_fxsr) - err = restore_i387_fxsave(fp, sizeof(struct - i387_fxsave_struct)); - else - err = restore_i387_fsave(fp); - } else { - err = fpregs_soft_set(current, NULL, - 0, sizeof(struct user_i387_ia32_struct), - NULL, fp) != 0; - } - set_used_math(); - - return err; -} - -/* * FPU state for core dumps. * This is only used for a.out dumps now. * It is declared generically using elf_fpregset_t (which is @@ -746,3 +608,33 @@ int dump_fpu(struct pt_regs *regs, struct user_i387_struct *fpu) EXPORT_SYMBOL(dump_fpu); #endif /* CONFIG_X86_32 || CONFIG_IA32_EMULATION */ + +static int __init no_387(char *s) +{ + setup_clear_cpu_cap(X86_FEATURE_FPU); + return 1; +} + +__setup("no387", no_387); + +void fpu_detect(struct cpuinfo_x86 *c) +{ + unsigned long cr0; + u16 fsw, fcw; + + fsw = fcw = 0xffff; + + cr0 = read_cr0(); + cr0 &= ~(X86_CR0_TS | X86_CR0_EM); + write_cr0(cr0); + + asm volatile("fninit ; fnstsw %0 ; fnstcw %1" + : "+m" (fsw), "+m" (fcw)); + + if (fsw == 0 && (fcw & 0x103f) == 0x003f) + set_cpu_cap(c, X86_FEATURE_FPU); + else + clear_cpu_cap(c, X86_FEATURE_FPU); + + /* The final cr0 value is set in fpu_init() */ +} diff --git a/arch/x86/kernel/i8259.c b/arch/x86/kernel/i8259.c index 610485223bd..8af817105e2 100644 --- a/arch/x86/kernel/i8259.c +++ b/arch/x86/kernel/i8259.c @@ -15,7 +15,6 @@ #include <linux/delay.h> #include <linux/atomic.h> -#include <asm/system.h> #include <asm/timer.h> #include <asm/hw_irq.h> #include <asm/pgtable.h> @@ -264,7 +263,7 @@ static void i8259A_shutdown(void) * out of. */ outb(0xff, PIC_MASTER_IMR); /* mask all of 8259A-1 */ - outb(0xff, PIC_SLAVE_IMR); /* mask all of 8259A-1 */ + outb(0xff, PIC_SLAVE_IMR); /* mask all of 8259A-2 */ } static struct syscore_ops i8259_syscore_ops = { @@ -300,21 +299,38 @@ static void unmask_8259A(void) static void init_8259A(int auto_eoi) { unsigned long flags; + unsigned char probe_val = ~(1 << PIC_CASCADE_IR); + unsigned char new_val; i8259A_auto_eoi = auto_eoi; raw_spin_lock_irqsave(&i8259A_lock, flags); - outb(0xff, PIC_MASTER_IMR); /* mask all of 8259A-1 */ + /* + * Check to see if we have a PIC. + * Mask all except the cascade and read + * back the value we just wrote. If we don't + * have a PIC, we will read 0xff as opposed to the + * value we wrote. + */ outb(0xff, PIC_SLAVE_IMR); /* mask all of 8259A-2 */ + outb(probe_val, PIC_MASTER_IMR); + new_val = inb(PIC_MASTER_IMR); + if (new_val != probe_val) { + printk(KERN_INFO "Using NULL legacy PIC\n"); + legacy_pic = &null_legacy_pic; + raw_spin_unlock_irqrestore(&i8259A_lock, flags); + return; + } + + outb(0xff, PIC_MASTER_IMR); /* mask all of 8259A-1 */ /* * outb_pic - this has to work on a wide range of PC hardware. */ outb_pic(0x11, PIC_MASTER_CMD); /* ICW1: select 8259A-1 init */ - /* ICW2: 8259A-1 IR0-7 mapped to 0x30-0x37 on x86-64, - to 0x20-0x27 on i386 */ + /* ICW2: 8259A-1 IR0-7 mapped to 0x30-0x37 */ outb_pic(IRQ0_VECTOR, PIC_MASTER_IMR); /* 8259A-1 (the master) has a slave on IR2 */ diff --git a/arch/x86/kernel/init_task.c b/arch/x86/kernel/init_task.c deleted file mode 100644 index 43e9ccf4494..00000000000 --- a/arch/x86/kernel/init_task.c +++ /dev/null @@ -1,42 +0,0 @@ -#include <linux/mm.h> -#include <linux/module.h> -#include <linux/sched.h> -#include <linux/init.h> -#include <linux/init_task.h> -#include <linux/fs.h> -#include <linux/mqueue.h> - -#include <asm/uaccess.h> -#include <asm/pgtable.h> -#include <asm/desc.h> - -static struct signal_struct init_signals = INIT_SIGNALS(init_signals); -static struct sighand_struct init_sighand = INIT_SIGHAND(init_sighand); - -/* - * Initial thread structure. - * - * We need to make sure that this is THREAD_SIZE aligned due to the - * way process stacks are handled. This is done by having a special - * "init_task" linker map entry.. - */ -union thread_union init_thread_union __init_task_data = - { INIT_THREAD_INFO(init_task) }; - -/* - * Initial task structure. - * - * All other task structs will be allocated on slabs in fork.c - */ -struct task_struct init_task = INIT_TASK(init_task); -EXPORT_SYMBOL(init_task); - -/* - * per-CPU TSS segments. Threads are completely 'soft' on Linux, - * no more per-task TSS's. The TSS size is kept cacheline-aligned - * so they are allowed to end up in the .data..cacheline_aligned - * section. Since TSS's are completely CPU-local, we want them - * on exact cacheline boundaries, to eliminate cacheline ping-pong. - */ -DEFINE_PER_CPU_SHARED_ALIGNED(struct tss_struct, init_tss) = INIT_TSS; - diff --git a/arch/x86/kernel/ioport.c b/arch/x86/kernel/ioport.c index 8c968974253..4ddaf66ea35 100644 --- a/arch/x86/kernel/ioport.c +++ b/arch/x86/kernel/ioport.c @@ -93,8 +93,9 @@ asmlinkage long sys_ioperm(unsigned long from, unsigned long num, int turn_on) * on system-call entry - see also fork() and the signal handling * code. */ -long sys_iopl(unsigned int level, struct pt_regs *regs) +SYSCALL_DEFINE1(iopl, unsigned int, level) { + struct pt_regs *regs = current_pt_regs(); unsigned int old = (regs->flags >> 12) & 3; struct thread_struct *t = ¤t->thread; diff --git a/arch/x86/kernel/iosf_mbi.c b/arch/x86/kernel/iosf_mbi.c new file mode 100644 index 00000000000..d30acdc1229 --- /dev/null +++ b/arch/x86/kernel/iosf_mbi.c @@ -0,0 +1,237 @@ +/* + * IOSF-SB MailBox Interface Driver + * Copyright (c) 2013, Intel Corporation. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * + * The IOSF-SB is a fabric bus available on Atom based SOC's that uses a + * mailbox interface (MBI) to communicate with mutiple devices. This + * driver implements access to this interface for those platforms that can + * enumerate the device using PCI. + */ + +#include <linux/module.h> +#include <linux/init.h> +#include <linux/spinlock.h> +#include <linux/pci.h> + +#include <asm/iosf_mbi.h> + +#define PCI_DEVICE_ID_BAYTRAIL 0x0F00 +#define PCI_DEVICE_ID_QUARK_X1000 0x0958 + +static DEFINE_SPINLOCK(iosf_mbi_lock); + +static inline u32 iosf_mbi_form_mcr(u8 op, u8 port, u8 offset) +{ + return (op << 24) | (port << 16) | (offset << 8) | MBI_ENABLE; +} + +static struct pci_dev *mbi_pdev; /* one mbi device */ + +static int iosf_mbi_pci_read_mdr(u32 mcrx, u32 mcr, u32 *mdr) +{ + int result; + + if (!mbi_pdev) + return -ENODEV; + + if (mcrx) { + result = pci_write_config_dword(mbi_pdev, MBI_MCRX_OFFSET, + mcrx); + if (result < 0) + goto fail_read; + } + + result = pci_write_config_dword(mbi_pdev, MBI_MCR_OFFSET, mcr); + if (result < 0) + goto fail_read; + + result = pci_read_config_dword(mbi_pdev, MBI_MDR_OFFSET, mdr); + if (result < 0) + goto fail_read; + + return 0; + +fail_read: + dev_err(&mbi_pdev->dev, "PCI config access failed with %d\n", result); + return result; +} + +static int iosf_mbi_pci_write_mdr(u32 mcrx, u32 mcr, u32 mdr) +{ + int result; + + if (!mbi_pdev) + return -ENODEV; + + result = pci_write_config_dword(mbi_pdev, MBI_MDR_OFFSET, mdr); + if (result < 0) + goto fail_write; + + if (mcrx) { + result = pci_write_config_dword(mbi_pdev, MBI_MCRX_OFFSET, + mcrx); + if (result < 0) + goto fail_write; + } + + result = pci_write_config_dword(mbi_pdev, MBI_MCR_OFFSET, mcr); + if (result < 0) + goto fail_write; + + return 0; + +fail_write: + dev_err(&mbi_pdev->dev, "PCI config access failed with %d\n", result); + return result; +} + +int iosf_mbi_read(u8 port, u8 opcode, u32 offset, u32 *mdr) +{ + u32 mcr, mcrx; + unsigned long flags; + int ret; + + /*Access to the GFX unit is handled by GPU code */ + if (port == BT_MBI_UNIT_GFX) { + WARN_ON(1); + return -EPERM; + } + + mcr = iosf_mbi_form_mcr(opcode, port, offset & MBI_MASK_LO); + mcrx = offset & MBI_MASK_HI; + + spin_lock_irqsave(&iosf_mbi_lock, flags); + ret = iosf_mbi_pci_read_mdr(mcrx, mcr, mdr); + spin_unlock_irqrestore(&iosf_mbi_lock, flags); + + return ret; +} +EXPORT_SYMBOL(iosf_mbi_read); + +int iosf_mbi_write(u8 port, u8 opcode, u32 offset, u32 mdr) +{ + u32 mcr, mcrx; + unsigned long flags; + int ret; + + /*Access to the GFX unit is handled by GPU code */ + if (port == BT_MBI_UNIT_GFX) { + WARN_ON(1); + return -EPERM; + } + + mcr = iosf_mbi_form_mcr(opcode, port, offset & MBI_MASK_LO); + mcrx = offset & MBI_MASK_HI; + + spin_lock_irqsave(&iosf_mbi_lock, flags); + ret = iosf_mbi_pci_write_mdr(mcrx, mcr, mdr); + spin_unlock_irqrestore(&iosf_mbi_lock, flags); + + return ret; +} +EXPORT_SYMBOL(iosf_mbi_write); + +int iosf_mbi_modify(u8 port, u8 opcode, u32 offset, u32 mdr, u32 mask) +{ + u32 mcr, mcrx; + u32 value; + unsigned long flags; + int ret; + + /*Access to the GFX unit is handled by GPU code */ + if (port == BT_MBI_UNIT_GFX) { + WARN_ON(1); + return -EPERM; + } + + mcr = iosf_mbi_form_mcr(opcode, port, offset & MBI_MASK_LO); + mcrx = offset & MBI_MASK_HI; + + spin_lock_irqsave(&iosf_mbi_lock, flags); + + /* Read current mdr value */ + ret = iosf_mbi_pci_read_mdr(mcrx, mcr & MBI_RD_MASK, &value); + if (ret < 0) { + spin_unlock_irqrestore(&iosf_mbi_lock, flags); + return ret; + } + + /* Apply mask */ + value &= ~mask; + mdr &= mask; + value |= mdr; + + /* Write back */ + ret = iosf_mbi_pci_write_mdr(mcrx, mcr | MBI_WR_MASK, value); + + spin_unlock_irqrestore(&iosf_mbi_lock, flags); + + return ret; +} +EXPORT_SYMBOL(iosf_mbi_modify); + +bool iosf_mbi_available(void) +{ + /* Mbi isn't hot-pluggable. No remove routine is provided */ + return mbi_pdev; +} +EXPORT_SYMBOL(iosf_mbi_available); + +static int iosf_mbi_probe(struct pci_dev *pdev, + const struct pci_device_id *unused) +{ + int ret; + + ret = pci_enable_device(pdev); + if (ret < 0) { + dev_err(&pdev->dev, "error: could not enable device\n"); + return ret; + } + + mbi_pdev = pci_dev_get(pdev); + return 0; +} + +static DEFINE_PCI_DEVICE_TABLE(iosf_mbi_pci_ids) = { + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_BAYTRAIL) }, + { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_QUARK_X1000) }, + { 0, }, +}; +MODULE_DEVICE_TABLE(pci, iosf_mbi_pci_ids); + +static struct pci_driver iosf_mbi_pci_driver = { + .name = "iosf_mbi_pci", + .probe = iosf_mbi_probe, + .id_table = iosf_mbi_pci_ids, +}; + +static int __init iosf_mbi_init(void) +{ + return pci_register_driver(&iosf_mbi_pci_driver); +} + +static void __exit iosf_mbi_exit(void) +{ + pci_unregister_driver(&iosf_mbi_pci_driver); + if (mbi_pdev) { + pci_dev_put(mbi_pdev); + mbi_pdev = NULL; + } +} + +module_init(iosf_mbi_init); +module_exit(iosf_mbi_exit); + +MODULE_AUTHOR("David E. Box <david.e.box@linux.intel.com>"); +MODULE_DESCRIPTION("IOSF Mailbox Interface accessor"); +MODULE_LICENSE("GPL v2"); diff --git a/arch/x86/kernel/irq.c b/arch/x86/kernel/irq.c index 7943e0c21bd..922d2858102 100644 --- a/arch/x86/kernel/irq.c +++ b/arch/x86/kernel/irq.c @@ -17,6 +17,10 @@ #include <asm/idle.h> #include <asm/mce.h> #include <asm/hw_irq.h> +#include <asm/desc.h> + +#define CREATE_TRACE_POINTS +#include <asm/trace/irq_vectors.h> atomic_t irq_err_count; @@ -92,7 +96,8 @@ int arch_show_interrupts(struct seq_file *p, int prec) seq_printf(p, " Rescheduling interrupts\n"); seq_printf(p, "%*s: ", prec, "CAL"); for_each_online_cpu(j) - seq_printf(p, "%10u ", irq_stats(j)->irq_call_count); + seq_printf(p, "%10u ", irq_stats(j)->irq_call_count - + irq_stats(j)->irq_tlb_count); seq_printf(p, " Function call interrupts\n"); seq_printf(p, "%*s: ", prec, "TLB"); for_each_online_cpu(j) @@ -121,6 +126,12 @@ int arch_show_interrupts(struct seq_file *p, int prec) seq_printf(p, "%10u ", per_cpu(mce_poll_count, j)); seq_printf(p, " Machine check polls\n"); #endif +#if IS_ENABLED(CONFIG_HYPERV) || defined(CONFIG_XEN) + seq_printf(p, "%*s: ", prec, "THR"); + for_each_online_cpu(j) + seq_printf(p, "%10u ", irq_stats(j)->irq_hv_callback_count); + seq_printf(p, " Hypervisor callback interrupts\n"); +#endif seq_printf(p, "%*s: %10u\n", prec, "ERR", atomic_read(&irq_err_count)); #if defined(CONFIG_X86_IO_APIC) seq_printf(p, "%*s: %10u\n", prec, "MIS", atomic_read(&irq_mis_count)); @@ -147,7 +158,6 @@ u64 arch_irq_stat_cpu(unsigned int cpu) #ifdef CONFIG_SMP sum += irq_stats(cpu)->irq_resched_count; sum += irq_stats(cpu)->irq_call_count; - sum += irq_stats(cpu)->irq_tlb_count; #endif #ifdef CONFIG_X86_THERMAL_VECTOR sum += irq_stats(cpu)->irq_thermal_count; @@ -165,10 +175,6 @@ u64 arch_irq_stat_cpu(unsigned int cpu) u64 arch_irq_stat(void) { u64 sum = atomic_read(&irq_err_count); - -#ifdef CONFIG_X86_IO_APIC - sum += atomic_read(&irq_mis_count); -#endif return sum; } @@ -178,7 +184,7 @@ u64 arch_irq_stat(void) * SMP cross-CPU interrupts have their own specific * handlers). */ -unsigned int __irq_entry do_IRQ(struct pt_regs *regs) +__visible unsigned int __irq_entry do_IRQ(struct pt_regs *regs) { struct pt_regs *old_regs = set_irq_regs(regs); @@ -194,9 +200,13 @@ unsigned int __irq_entry do_IRQ(struct pt_regs *regs) if (!handle_irq(irq, regs)) { ack_APIC_irq(); - if (printk_ratelimit()) - pr_emerg("%s: %d.%d No irq handler for vector (irq %d)\n", - __func__, smp_processor_id(), vector, irq); + if (irq != VECTOR_RETRIGGERED) { + pr_emerg_ratelimited("%s: %d.%d No irq handler for vector (irq %d)\n", + __func__, smp_processor_id(), + vector, irq); + } else { + __this_cpu_write(vector_irq[vector], VECTOR_UNDEFINED); + } } irq_exit(); @@ -208,7 +218,29 @@ unsigned int __irq_entry do_IRQ(struct pt_regs *regs) /* * Handler for X86_PLATFORM_IPI_VECTOR. */ -void smp_x86_platform_ipi(struct pt_regs *regs) +void __smp_x86_platform_ipi(void) +{ + inc_irq_stat(x86_platform_ipis); + + if (x86_platform_ipi_callback) + x86_platform_ipi_callback(); +} + +__visible void smp_x86_platform_ipi(struct pt_regs *regs) +{ + struct pt_regs *old_regs = set_irq_regs(regs); + + entering_ack_irq(); + __smp_x86_platform_ipi(); + exiting_irq(); + set_irq_regs(old_regs); +} + +#ifdef CONFIG_HAVE_KVM +/* + * Handler for POSTED_INTERRUPT_VECTOR. + */ +__visible void smp_kvm_posted_intr_ipi(struct pt_regs *regs) { struct pt_regs *old_regs = set_irq_regs(regs); @@ -218,19 +250,113 @@ void smp_x86_platform_ipi(struct pt_regs *regs) exit_idle(); - inc_irq_stat(x86_platform_ipis); - - if (x86_platform_ipi_callback) - x86_platform_ipi_callback(); + inc_irq_stat(kvm_posted_intr_ipis); irq_exit(); set_irq_regs(old_regs); } +#endif + +__visible void smp_trace_x86_platform_ipi(struct pt_regs *regs) +{ + struct pt_regs *old_regs = set_irq_regs(regs); + + entering_ack_irq(); + trace_x86_platform_ipi_entry(X86_PLATFORM_IPI_VECTOR); + __smp_x86_platform_ipi(); + trace_x86_platform_ipi_exit(X86_PLATFORM_IPI_VECTOR); + exiting_irq(); + set_irq_regs(old_regs); +} EXPORT_SYMBOL_GPL(vector_used_by_percpu_irq); #ifdef CONFIG_HOTPLUG_CPU + +/* These two declarations are only used in check_irq_vectors_for_cpu_disable() + * below, which is protected by stop_machine(). Putting them on the stack + * results in a stack frame overflow. Dynamically allocating could result in a + * failure so declare these two cpumasks as global. + */ +static struct cpumask affinity_new, online_new; + +/* + * This cpu is going to be removed and its vectors migrated to the remaining + * online cpus. Check to see if there are enough vectors in the remaining cpus. + * This function is protected by stop_machine(). + */ +int check_irq_vectors_for_cpu_disable(void) +{ + int irq, cpu; + unsigned int this_cpu, vector, this_count, count; + struct irq_desc *desc; + struct irq_data *data; + + this_cpu = smp_processor_id(); + cpumask_copy(&online_new, cpu_online_mask); + cpu_clear(this_cpu, online_new); + + this_count = 0; + for (vector = FIRST_EXTERNAL_VECTOR; vector < NR_VECTORS; vector++) { + irq = __this_cpu_read(vector_irq[vector]); + if (irq >= 0) { + desc = irq_to_desc(irq); + data = irq_desc_get_irq_data(desc); + cpumask_copy(&affinity_new, data->affinity); + cpu_clear(this_cpu, affinity_new); + + /* Do not count inactive or per-cpu irqs. */ + if (!irq_has_action(irq) || irqd_is_per_cpu(data)) + continue; + + /* + * A single irq may be mapped to multiple + * cpu's vector_irq[] (for example IOAPIC cluster + * mode). In this case we have two + * possibilities: + * + * 1) the resulting affinity mask is empty; that is + * this the down'd cpu is the last cpu in the irq's + * affinity mask, or + * + * 2) the resulting affinity mask is no longer + * a subset of the online cpus but the affinity + * mask is not zero; that is the down'd cpu is the + * last online cpu in a user set affinity mask. + */ + if (cpumask_empty(&affinity_new) || + !cpumask_subset(&affinity_new, &online_new)) + this_count++; + } + } + + count = 0; + for_each_online_cpu(cpu) { + if (cpu == this_cpu) + continue; + /* + * We scan from FIRST_EXTERNAL_VECTOR to first system + * vector. If the vector is marked in the used vectors + * bitmap or an irq is assigned to it, we don't count + * it as available. + */ + for (vector = FIRST_EXTERNAL_VECTOR; + vector < first_system_vector; vector++) { + if (!test_bit(vector, used_vectors) && + per_cpu(vector_irq, cpu)[vector] < 0) + count++; + } + } + + if (count < this_count) { + pr_warn("CPU %d disable failed: CPU has %u vectors assigned and there are only %u available.\n", + this_cpu, this_count, count); + return -ERANGE; + } + return 0; +} + /* A cpu has been removed from cpu_online_mask. Reset irq affinities. */ void fixup_irqs(void) { @@ -239,6 +365,7 @@ void fixup_irqs(void) struct irq_desc *desc; struct irq_data *data; struct irq_chip *chip; + int ret; for_each_irq_desc(irq, desc) { int break_affinity = 0; @@ -270,28 +397,37 @@ void fixup_irqs(void) if (cpumask_any_and(affinity, cpu_online_mask) >= nr_cpu_ids) { break_affinity = 1; - affinity = cpu_all_mask; + affinity = cpu_online_mask; } chip = irq_data_get_irq_chip(data); if (!irqd_can_move_in_process_context(data) && chip->irq_mask) chip->irq_mask(data); - if (chip->irq_set_affinity) - chip->irq_set_affinity(data, affinity, true); - else if (!(warned++)) - set_affinity = 0; + if (chip->irq_set_affinity) { + ret = chip->irq_set_affinity(data, affinity, true); + if (ret == -ENOSPC) + pr_crit("IRQ %d set affinity failed because there are no available vectors. The device assigned to this IRQ is unstable.\n", irq); + } else { + if (!(warned++)) + set_affinity = 0; + } + /* + * We unmask if the irq was not marked masked by the + * core code. That respects the lazy irq disable + * behaviour. + */ if (!irqd_can_move_in_process_context(data) && - !irqd_irq_disabled(data) && chip->irq_unmask) + !irqd_irq_masked(data) && chip->irq_unmask) chip->irq_unmask(data); raw_spin_unlock(&desc->lock); if (break_affinity && set_affinity) - printk("Broke affinity for irq %i\n", irq); + pr_notice("Broke affinity for irq %i\n", irq); else if (!set_affinity) - printk("Cannot set affinity for irq %i\n", irq); + pr_notice("Cannot set affinity for irq %i\n", irq); } /* @@ -308,7 +444,7 @@ void fixup_irqs(void) for (vector = FIRST_EXTERNAL_VECTOR; vector < NR_VECTORS; vector++) { unsigned int irr; - if (__this_cpu_read(vector_irq[vector]) < 0) + if (__this_cpu_read(vector_irq[vector]) <= VECTOR_UNDEFINED) continue; irr = apic_read(APIC_IRR + (vector / 32 * 0x10)); @@ -319,10 +455,14 @@ void fixup_irqs(void) data = irq_desc_get_irq_data(desc); chip = irq_data_get_irq_chip(data); raw_spin_lock(&desc->lock); - if (chip->irq_retrigger) + if (chip->irq_retrigger) { chip->irq_retrigger(data); + __this_cpu_write(vector_irq[vector], VECTOR_RETRIGGERED); + } raw_spin_unlock(&desc->lock); } + if (__this_cpu_read(vector_irq[vector]) != VECTOR_RETRIGGERED) + __this_cpu_write(vector_irq[vector], VECTOR_UNDEFINED); } } #endif diff --git a/arch/x86/kernel/irq_32.c b/arch/x86/kernel/irq_32.c index 40fc86161d9..63ce838e5a5 100644 --- a/arch/x86/kernel/irq_32.c +++ b/arch/x86/kernel/irq_32.c @@ -55,16 +55,8 @@ static inline int check_stack_overflow(void) { return 0; } static inline void print_stack_overflow(void) { } #endif -/* - * per-CPU IRQ handling contexts (thread information and stack) - */ -union irq_ctx { - struct thread_info tinfo; - u32 stack[THREAD_SIZE/sizeof(u32)]; -} __attribute__((aligned(THREAD_SIZE))); - -static DEFINE_PER_CPU(union irq_ctx *, hardirq_ctx); -static DEFINE_PER_CPU(union irq_ctx *, softirq_ctx); +DEFINE_PER_CPU(struct irq_stack *, hardirq_stack); +DEFINE_PER_CPU(struct irq_stack *, softirq_stack); static void call_on_stack(void *func, void *stack) { @@ -77,14 +69,26 @@ static void call_on_stack(void *func, void *stack) : "memory", "cc", "edx", "ecx", "eax"); } +/* how to get the current stack pointer from C */ +#define current_stack_pointer ({ \ + unsigned long sp; \ + asm("mov %%esp,%0" : "=g" (sp)); \ + sp; \ +}) + +static inline void *current_stack(void) +{ + return (void *)(current_stack_pointer & ~(THREAD_SIZE - 1)); +} + static inline int execute_on_irq_stack(int overflow, struct irq_desc *desc, int irq) { - union irq_ctx *curctx, *irqctx; - u32 *isp, arg1, arg2; + struct irq_stack *curstk, *irqstk; + u32 *isp, *prev_esp, arg1, arg2; - curctx = (union irq_ctx *) current_thread_info(); - irqctx = __this_cpu_read(hardirq_ctx); + curstk = (struct irq_stack *) current_stack(); + irqstk = __this_cpu_read(hardirq_stack); /* * this is where we switch to the IRQ stack. However, if we are @@ -92,21 +96,14 @@ execute_on_irq_stack(int overflow, struct irq_desc *desc, int irq) * handler) we can't do that and just have to keep using the * current stack (which is the irq stack already after all) */ - if (unlikely(curctx == irqctx)) + if (unlikely(curstk == irqstk)) return 0; - /* build the stack frame on the IRQ stack */ - isp = (u32 *) ((char *)irqctx + sizeof(*irqctx)); - irqctx->tinfo.task = curctx->tinfo.task; - irqctx->tinfo.previous_esp = current_stack_pointer; + isp = (u32 *) ((char *)irqstk + sizeof(*irqstk)); - /* - * Copy the softirq bits in preempt_count so that the - * softirq checks work in the hardirq context. - */ - irqctx->tinfo.preempt_count = - (irqctx->tinfo.preempt_count & ~SOFTIRQ_MASK) | - (curctx->tinfo.preempt_count & SOFTIRQ_MASK); + /* Save the next esp at the bottom of the stack */ + prev_esp = (u32 *)irqstk; + *prev_esp = current_stack_pointer; if (unlikely(overflow)) call_on_stack(print_stack_overflow, isp); @@ -124,65 +121,44 @@ execute_on_irq_stack(int overflow, struct irq_desc *desc, int irq) /* * allocate per-cpu stacks for hardirq and for softirq processing */ -void __cpuinit irq_ctx_init(int cpu) +void irq_ctx_init(int cpu) { - union irq_ctx *irqctx; + struct irq_stack *irqstk; - if (per_cpu(hardirq_ctx, cpu)) + if (per_cpu(hardirq_stack, cpu)) return; - irqctx = page_address(alloc_pages_node(cpu_to_node(cpu), - THREAD_FLAGS, - THREAD_ORDER)); - memset(&irqctx->tinfo, 0, sizeof(struct thread_info)); - irqctx->tinfo.cpu = cpu; - irqctx->tinfo.preempt_count = HARDIRQ_OFFSET; - irqctx->tinfo.addr_limit = MAKE_MM_SEG(0); - - per_cpu(hardirq_ctx, cpu) = irqctx; + irqstk = page_address(alloc_pages_node(cpu_to_node(cpu), + THREADINFO_GFP, + THREAD_SIZE_ORDER)); + per_cpu(hardirq_stack, cpu) = irqstk; - irqctx = page_address(alloc_pages_node(cpu_to_node(cpu), - THREAD_FLAGS, - THREAD_ORDER)); - memset(&irqctx->tinfo, 0, sizeof(struct thread_info)); - irqctx->tinfo.cpu = cpu; - irqctx->tinfo.addr_limit = MAKE_MM_SEG(0); - - per_cpu(softirq_ctx, cpu) = irqctx; + irqstk = page_address(alloc_pages_node(cpu_to_node(cpu), + THREADINFO_GFP, + THREAD_SIZE_ORDER)); + per_cpu(softirq_stack, cpu) = irqstk; printk(KERN_DEBUG "CPU %u irqstacks, hard=%p soft=%p\n", - cpu, per_cpu(hardirq_ctx, cpu), per_cpu(softirq_ctx, cpu)); + cpu, per_cpu(hardirq_stack, cpu), per_cpu(softirq_stack, cpu)); } -asmlinkage void do_softirq(void) +void do_softirq_own_stack(void) { - unsigned long flags; - struct thread_info *curctx; - union irq_ctx *irqctx; - u32 *isp; + struct thread_info *curstk; + struct irq_stack *irqstk; + u32 *isp, *prev_esp; - if (in_interrupt()) - return; + curstk = current_stack(); + irqstk = __this_cpu_read(softirq_stack); - local_irq_save(flags); + /* build the stack frame on the softirq stack */ + isp = (u32 *) ((char *)irqstk + sizeof(*irqstk)); - if (local_softirq_pending()) { - curctx = current_thread_info(); - irqctx = __this_cpu_read(softirq_ctx); - irqctx->tinfo.task = curctx->task; - irqctx->tinfo.previous_esp = current_stack_pointer; - - /* build the stack frame on the softirq stack */ - isp = (u32 *) ((char *)irqctx + sizeof(*irqctx)); - - call_on_stack(__do_softirq, isp); - /* - * Shouldn't happen, we returned above if in_interrupt(): - */ - WARN_ON_ONCE(softirq_count()); - } + /* Push the previous esp onto the stack */ + prev_esp = (u32 *)irqstk; + *prev_esp = current_stack_pointer; - local_irq_restore(flags); + call_on_stack(__do_softirq, isp); } bool handle_irq(unsigned irq, struct pt_regs *regs) @@ -196,7 +172,7 @@ bool handle_irq(unsigned irq, struct pt_regs *regs) if (unlikely(!desc)) return false; - if (!execute_on_irq_stack(overflow, desc, irq)) { + if (user_mode_vm(regs) || !execute_on_irq_stack(overflow, desc, irq)) { if (unlikely(overflow)) print_stack_overflow(); desc->handle_irq(irq, desc); diff --git a/arch/x86/kernel/irq_64.c b/arch/x86/kernel/irq_64.c index d04d3ecded6..4d1c746892e 100644 --- a/arch/x86/kernel/irq_64.c +++ b/arch/x86/kernel/irq_64.c @@ -87,24 +87,3 @@ bool handle_irq(unsigned irq, struct pt_regs *regs) generic_handle_irq_desc(irq, desc); return true; } - - -extern void call_softirq(void); - -asmlinkage void do_softirq(void) -{ - __u32 pending; - unsigned long flags; - - if (in_interrupt()) - return; - - local_irq_save(flags); - pending = local_softirq_pending(); - /* Switch to interrupt stack */ - if (pending) { - call_softirq(); - WARN_ON_ONCE(softirq_count()); - } - local_irq_restore(flags); -} diff --git a/arch/x86/kernel/irq_work.c b/arch/x86/kernel/irq_work.c index ca8f703a1e7..1de84e3ab4e 100644 --- a/arch/x86/kernel/irq_work.c +++ b/arch/x86/kernel/irq_work.c @@ -8,14 +8,34 @@ #include <linux/irq_work.h> #include <linux/hardirq.h> #include <asm/apic.h> +#include <asm/trace/irq_vectors.h> -void smp_irq_work_interrupt(struct pt_regs *regs) +static inline void irq_work_entering_irq(void) { irq_enter(); ack_APIC_irq(); +} + +static inline void __smp_irq_work_interrupt(void) +{ inc_irq_stat(apic_irq_work_irqs); irq_work_run(); - irq_exit(); +} + +__visible void smp_irq_work_interrupt(struct pt_regs *regs) +{ + irq_work_entering_irq(); + __smp_irq_work_interrupt(); + exiting_irq(); +} + +__visible void smp_trace_irq_work_interrupt(struct pt_regs *regs) +{ + irq_work_entering_irq(); + trace_irq_work_entry(IRQ_WORK_VECTOR); + __smp_irq_work_interrupt(); + trace_irq_work_exit(IRQ_WORK_VECTOR); + exiting_irq(); } void arch_irq_work_raise(void) diff --git a/arch/x86/kernel/irqinit.c b/arch/x86/kernel/irqinit.c index 313fb5cddbc..7f50156542f 100644 --- a/arch/x86/kernel/irqinit.c +++ b/arch/x86/kernel/irqinit.c @@ -16,7 +16,6 @@ #include <linux/delay.h> #include <linux/atomic.h> -#include <asm/system.h> #include <asm/timer.h> #include <asm/hw_irq.h> #include <asm/pgtable.h> @@ -43,39 +42,6 @@ * (these are usually mapped into the 0x30-0xff vector range) */ -#ifdef CONFIG_X86_32 -/* - * Note that on a 486, we don't want to do a SIGFPE on an irq13 - * as the irq is unreliable, and exception 16 works correctly - * (ie as explained in the intel literature). On a 386, you - * can't use exception 16 due to bad IBM design, so we have to - * rely on the less exact irq13. - * - * Careful.. Not only is IRQ13 unreliable, but it is also - * leads to races. IBM designers who came up with it should - * be shot. - */ - -static irqreturn_t math_error_irq(int cpl, void *dev_id) -{ - outb(0, 0xF0); - if (ignore_fpu_irq || !boot_cpu_data.hard_math) - return IRQ_NONE; - math_error(get_irq_regs(), 0, 16); - return IRQ_HANDLED; -} - -/* - * New motherboards sometimes make IRQ 13 be a PCI interrupt, - * so allow interrupt sharing. - */ -static struct irqaction fpu_irq = { - .handler = math_error_irq, - .name = "fpu", - .flags = IRQF_NO_THREAD, -}; -#endif - /* * IRQ2 is cascade interrupt to second interrupt controller */ @@ -86,7 +52,7 @@ static struct irqaction irq2 = { }; DEFINE_PER_CPU(vector_irq_t, vector_irq) = { - [0 ... NR_VECTORS - 1] = -1, + [0 ... NR_VECTORS - 1] = VECTOR_UNDEFINED, }; int vector_used_by_percpu_irq(unsigned int vector) @@ -94,7 +60,7 @@ int vector_used_by_percpu_irq(unsigned int vector) int cpu; for_each_online_cpu(cpu) { - if (per_cpu(vector_irq, cpu)[vector] != -1) + if (per_cpu(vector_irq, cpu)[vector] > VECTOR_UNDEFINED) return 1; } @@ -172,79 +138,6 @@ static void __init smp_intr_init(void) */ alloc_intr_gate(RESCHEDULE_VECTOR, reschedule_interrupt); - /* IPIs for invalidation */ -#define ALLOC_INVTLB_VEC(NR) \ - alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+NR, \ - invalidate_interrupt##NR) - - switch (NUM_INVALIDATE_TLB_VECTORS) { - default: - ALLOC_INVTLB_VEC(31); - case 31: - ALLOC_INVTLB_VEC(30); - case 30: - ALLOC_INVTLB_VEC(29); - case 29: - ALLOC_INVTLB_VEC(28); - case 28: - ALLOC_INVTLB_VEC(27); - case 27: - ALLOC_INVTLB_VEC(26); - case 26: - ALLOC_INVTLB_VEC(25); - case 25: - ALLOC_INVTLB_VEC(24); - case 24: - ALLOC_INVTLB_VEC(23); - case 23: - ALLOC_INVTLB_VEC(22); - case 22: - ALLOC_INVTLB_VEC(21); - case 21: - ALLOC_INVTLB_VEC(20); - case 20: - ALLOC_INVTLB_VEC(19); - case 19: - ALLOC_INVTLB_VEC(18); - case 18: - ALLOC_INVTLB_VEC(17); - case 17: - ALLOC_INVTLB_VEC(16); - case 16: - ALLOC_INVTLB_VEC(15); - case 15: - ALLOC_INVTLB_VEC(14); - case 14: - ALLOC_INVTLB_VEC(13); - case 13: - ALLOC_INVTLB_VEC(12); - case 12: - ALLOC_INVTLB_VEC(11); - case 11: - ALLOC_INVTLB_VEC(10); - case 10: - ALLOC_INVTLB_VEC(9); - case 9: - ALLOC_INVTLB_VEC(8); - case 8: - ALLOC_INVTLB_VEC(7); - case 7: - ALLOC_INVTLB_VEC(6); - case 6: - ALLOC_INVTLB_VEC(5); - case 5: - ALLOC_INVTLB_VEC(4); - case 4: - ALLOC_INVTLB_VEC(3); - case 3: - ALLOC_INVTLB_VEC(2); - case 2: - ALLOC_INVTLB_VEC(1); - case 1: - ALLOC_INVTLB_VEC(0); - break; - } - /* IPI for generic function call */ alloc_intr_gate(CALL_FUNCTION_VECTOR, call_function_interrupt); @@ -279,6 +172,10 @@ static void __init apic_intr_init(void) /* IPI for X86 platform specific use */ alloc_intr_gate(X86_PLATFORM_IPI_VECTOR, x86_platform_ipi); +#ifdef CONFIG_HAVE_KVM + /* IPI for KVM to deliver posted interrupt */ + alloc_intr_gate(POSTED_INTR_VECTOR, kvm_posted_intr_ipi); +#endif /* IPI vectors for APIC spurious and error interrupts */ alloc_intr_gate(SPURIOUS_APIC_VECTOR, spurious_interrupt); @@ -306,23 +203,16 @@ void __init native_init_IRQ(void) * us. (some of these will be overridden and become * 'special' SMP interrupts) */ - for (i = FIRST_EXTERNAL_VECTOR; i < NR_VECTORS; i++) { + i = FIRST_EXTERNAL_VECTOR; + for_each_clear_bit_from(i, used_vectors, NR_VECTORS) { /* IA32_SYSCALL_VECTOR could be used in trap_init already. */ - if (!test_bit(i, used_vectors)) - set_intr_gate(i, interrupt[i-FIRST_EXTERNAL_VECTOR]); + set_intr_gate(i, interrupt[i - FIRST_EXTERNAL_VECTOR]); } if (!acpi_ioapic && !of_ioapic) setup_irq(2, &irq2); #ifdef CONFIG_X86_32 - /* - * External FPU? Set up irq13 if so, for - * original braindamaged IBM FERR coupling. - */ - if (boot_cpu_data.hard_math && !cpu_has_fpu) - setup_irq(FPU_IRQ, &fpu_irq); - irq_ctx_init(smp_processor_id()); #endif } diff --git a/arch/x86/kernel/jump_label.c b/arch/x86/kernel/jump_label.c index 2889b3d4388..26d5a55a273 100644 --- a/arch/x86/kernel/jump_label.c +++ b/arch/x86/kernel/jump_label.c @@ -24,20 +24,82 @@ union jump_code_union { } __attribute__((packed)); }; +static void bug_at(unsigned char *ip, int line) +{ + /* + * The location is not an op that we were expecting. + * Something went wrong. Crash the box, as something could be + * corrupting the kernel. + */ + pr_warning("Unexpected op at %pS [%p] (%02x %02x %02x %02x %02x) %s:%d\n", + ip, ip, ip[0], ip[1], ip[2], ip[3], ip[4], __FILE__, line); + BUG(); +} + static void __jump_label_transform(struct jump_entry *entry, enum jump_label_type type, - void *(*poker)(void *, const void *, size_t)) + void *(*poker)(void *, const void *, size_t), + int init) { union jump_code_union code; + const unsigned char default_nop[] = { STATIC_KEY_INIT_NOP }; + const unsigned char *ideal_nop = ideal_nops[NOP_ATOMIC5]; if (type == JUMP_LABEL_ENABLE) { + if (init) { + /* + * Jump label is enabled for the first time. + * So we expect a default_nop... + */ + if (unlikely(memcmp((void *)entry->code, default_nop, 5) + != 0)) + bug_at((void *)entry->code, __LINE__); + } else { + /* + * ...otherwise expect an ideal_nop. Otherwise + * something went horribly wrong. + */ + if (unlikely(memcmp((void *)entry->code, ideal_nop, 5) + != 0)) + bug_at((void *)entry->code, __LINE__); + } + code.jump = 0xe9; code.offset = entry->target - (entry->code + JUMP_LABEL_NOP_SIZE); - } else + } else { + /* + * We are disabling this jump label. If it is not what + * we think it is, then something must have gone wrong. + * If this is the first initialization call, then we + * are converting the default nop to the ideal nop. + */ + if (init) { + if (unlikely(memcmp((void *)entry->code, default_nop, 5) != 0)) + bug_at((void *)entry->code, __LINE__); + } else { + code.jump = 0xe9; + code.offset = entry->target - + (entry->code + JUMP_LABEL_NOP_SIZE); + if (unlikely(memcmp((void *)entry->code, &code, 5) != 0)) + bug_at((void *)entry->code, __LINE__); + } memcpy(&code, ideal_nops[NOP_ATOMIC5], JUMP_LABEL_NOP_SIZE); + } - (*poker)((void *)entry->code, &code, JUMP_LABEL_NOP_SIZE); + /* + * Make text_poke_bp() a default fallback poker. + * + * At the time the change is being done, just ignore whether we + * are doing nop -> jump or jump -> nop transition, and assume + * always nop being the 'currently valid' instruction + * + */ + if (poker) + (*poker)((void *)entry->code, &code, JUMP_LABEL_NOP_SIZE); + else + text_poke_bp((void *)entry->code, &code, JUMP_LABEL_NOP_SIZE, + (void *)entry->code + JUMP_LABEL_NOP_SIZE); } void arch_jump_label_transform(struct jump_entry *entry, @@ -45,15 +107,38 @@ void arch_jump_label_transform(struct jump_entry *entry, { get_online_cpus(); mutex_lock(&text_mutex); - __jump_label_transform(entry, type, text_poke_smp); + __jump_label_transform(entry, type, NULL, 0); mutex_unlock(&text_mutex); put_online_cpus(); } +static enum { + JL_STATE_START, + JL_STATE_NO_UPDATE, + JL_STATE_UPDATE, +} jlstate __initdata_or_module = JL_STATE_START; + __init_or_module void arch_jump_label_transform_static(struct jump_entry *entry, enum jump_label_type type) { - __jump_label_transform(entry, type, text_poke_early); + /* + * This function is called at boot up and when modules are + * first loaded. Check if the default nop, the one that is + * inserted at compile time, is the ideal nop. If it is, then + * we do not need to update the nop, and we can leave it as is. + * If it is not, then we need to update the nop to the ideal nop. + */ + if (jlstate == JL_STATE_START) { + const unsigned char default_nop[] = { STATIC_KEY_INIT_NOP }; + const unsigned char *ideal_nop = ideal_nops[NOP_ATOMIC5]; + + if (memcmp(ideal_nop, default_nop, 5) != 0) + jlstate = JL_STATE_UPDATE; + else + jlstate = JL_STATE_NO_UPDATE; + } + if (jlstate == JL_STATE_UPDATE) + __jump_label_transform(entry, type, text_poke_early, 1); } #endif diff --git a/arch/x86/kernel/kdebugfs.c b/arch/x86/kernel/kdebugfs.c index 90fcf62854b..dc1404bf8e4 100644 --- a/arch/x86/kernel/kdebugfs.c +++ b/arch/x86/kernel/kdebugfs.c @@ -68,16 +68,9 @@ static ssize_t setup_data_read(struct file *file, char __user *user_buf, return count; } -static int setup_data_open(struct inode *inode, struct file *file) -{ - file->private_data = inode->i_private; - - return 0; -} - static const struct file_operations fops_setup_data = { .read = setup_data_read, - .open = setup_data_open, + .open = simple_open, .llseek = default_llseek, }; @@ -114,7 +107,7 @@ static int __init create_setup_data_nodes(struct dentry *parent) { struct setup_data_node *node; struct setup_data *data; - int error = -ENOMEM; + int error; struct dentry *d; struct page *pg; u64 pa_data; @@ -128,8 +121,10 @@ static int __init create_setup_data_nodes(struct dentry *parent) while (pa_data) { node = kmalloc(sizeof(*node), GFP_KERNEL); - if (!node) + if (!node) { + error = -ENOMEM; goto err_dir; + } pg = pfn_to_page((pa_data+sizeof(*data)-1) >> PAGE_SHIFT); if (PageHighMem(pg)) { diff --git a/arch/x86/kernel/kgdb.c b/arch/x86/kernel/kgdb.c index faba5771aca..7ec1d5f8d28 100644 --- a/arch/x86/kernel/kgdb.c +++ b/arch/x86/kernel/kgdb.c @@ -39,14 +39,14 @@ #include <linux/sched.h> #include <linux/delay.h> #include <linux/kgdb.h> -#include <linux/init.h> #include <linux/smp.h> #include <linux/nmi.h> #include <linux/hw_breakpoint.h> +#include <linux/uaccess.h> +#include <linux/memory.h> #include <asm/debugreg.h> #include <asm/apicdef.h> -#include <asm/system.h> #include <asm/apic.h> #include <asm/nmi.h> @@ -67,8 +67,6 @@ struct dbg_reg_def_t dbg_reg_def[DBG_MAX_REG_NUM] = { "ss", 4, offsetof(struct pt_regs, ss) }, { "ds", 4, offsetof(struct pt_regs, ds) }, { "es", 4, offsetof(struct pt_regs, es) }, - { "fs", 4, -1 }, - { "gs", 4, -1 }, #else { "ax", 8, offsetof(struct pt_regs, ax) }, { "bx", 8, offsetof(struct pt_regs, bx) }, @@ -90,7 +88,11 @@ struct dbg_reg_def_t dbg_reg_def[DBG_MAX_REG_NUM] = { "flags", 4, offsetof(struct pt_regs, flags) }, { "cs", 4, offsetof(struct pt_regs, cs) }, { "ss", 4, offsetof(struct pt_regs, ss) }, + { "ds", 4, -1 }, + { "es", 4, -1 }, #endif + { "fs", 4, -1 }, + { "gs", 4, -1 }, }; int dbg_set_reg(int regno, void *mem, struct pt_regs *regs) @@ -441,12 +443,12 @@ void kgdb_roundup_cpus(unsigned long flags) /** * kgdb_arch_handle_exception - Handle architecture specific GDB packets. - * @vector: The error vector of the exception that happened. + * @e_vector: The error vector of the exception that happened. * @signo: The signal number of the exception that happened. * @err_code: The error code of the exception that happened. - * @remcom_in_buffer: The buffer of the packet we have read. - * @remcom_out_buffer: The buffer of %BUFMAX bytes to write a packet into. - * @regs: The &struct pt_regs of the current process. + * @remcomInBuffer: The buffer of the packet we have read. + * @remcomOutBuffer: The buffer of %BUFMAX bytes to write a packet into. + * @linux_regs: The &struct pt_regs of the current process. * * This function MUST handle the 'c' and 's' command packets, * as well packets to set / remove a hardware breakpoint, if used. @@ -740,6 +742,66 @@ void kgdb_arch_set_pc(struct pt_regs *regs, unsigned long ip) regs->ip = ip; } +int kgdb_arch_set_breakpoint(struct kgdb_bkpt *bpt) +{ + int err; +#ifdef CONFIG_DEBUG_RODATA + char opc[BREAK_INSTR_SIZE]; +#endif /* CONFIG_DEBUG_RODATA */ + + bpt->type = BP_BREAKPOINT; + err = probe_kernel_read(bpt->saved_instr, (char *)bpt->bpt_addr, + BREAK_INSTR_SIZE); + if (err) + return err; + err = probe_kernel_write((char *)bpt->bpt_addr, + arch_kgdb_ops.gdb_bpt_instr, BREAK_INSTR_SIZE); +#ifdef CONFIG_DEBUG_RODATA + if (!err) + return err; + /* + * It is safe to call text_poke() because normal kernel execution + * is stopped on all cores, so long as the text_mutex is not locked. + */ + if (mutex_is_locked(&text_mutex)) + return -EBUSY; + text_poke((void *)bpt->bpt_addr, arch_kgdb_ops.gdb_bpt_instr, + BREAK_INSTR_SIZE); + err = probe_kernel_read(opc, (char *)bpt->bpt_addr, BREAK_INSTR_SIZE); + if (err) + return err; + if (memcmp(opc, arch_kgdb_ops.gdb_bpt_instr, BREAK_INSTR_SIZE)) + return -EINVAL; + bpt->type = BP_POKE_BREAKPOINT; +#endif /* CONFIG_DEBUG_RODATA */ + return err; +} + +int kgdb_arch_remove_breakpoint(struct kgdb_bkpt *bpt) +{ +#ifdef CONFIG_DEBUG_RODATA + int err; + char opc[BREAK_INSTR_SIZE]; + + if (bpt->type != BP_POKE_BREAKPOINT) + goto knl_write; + /* + * It is safe to call text_poke() because normal kernel execution + * is stopped on all cores, so long as the text_mutex is not locked. + */ + if (mutex_is_locked(&text_mutex)) + goto knl_write; + text_poke((void *)bpt->bpt_addr, bpt->saved_instr, BREAK_INSTR_SIZE); + err = probe_kernel_read(opc, (char *)bpt->bpt_addr, BREAK_INSTR_SIZE); + if (err || memcmp(opc, bpt->saved_instr, BREAK_INSTR_SIZE)) + goto knl_write; + return err; +knl_write: +#endif /* CONFIG_DEBUG_RODATA */ + return probe_kernel_write((char *)bpt->bpt_addr, + (char *)bpt->saved_instr, BREAK_INSTR_SIZE); +} + struct kgdb_arch arch_kgdb_ops = { /* Breakpoint instruction: */ .gdb_bpt_instr = { 0xcc }, diff --git a/arch/x86/kernel/kprobes/Makefile b/arch/x86/kernel/kprobes/Makefile new file mode 100644 index 00000000000..0d33169cc1a --- /dev/null +++ b/arch/x86/kernel/kprobes/Makefile @@ -0,0 +1,7 @@ +# +# Makefile for kernel probes +# + +obj-$(CONFIG_KPROBES) += core.o +obj-$(CONFIG_OPTPROBES) += opt.o +obj-$(CONFIG_KPROBES_ON_FTRACE) += ftrace.o diff --git a/arch/x86/kernel/kprobes/common.h b/arch/x86/kernel/kprobes/common.h new file mode 100644 index 00000000000..c6ee63f927a --- /dev/null +++ b/arch/x86/kernel/kprobes/common.h @@ -0,0 +1,108 @@ +#ifndef __X86_KERNEL_KPROBES_COMMON_H +#define __X86_KERNEL_KPROBES_COMMON_H + +/* Kprobes and Optprobes common header */ + +#ifdef CONFIG_X86_64 +#define SAVE_REGS_STRING \ + /* Skip cs, ip, orig_ax. */ \ + " subq $24, %rsp\n" \ + " pushq %rdi\n" \ + " pushq %rsi\n" \ + " pushq %rdx\n" \ + " pushq %rcx\n" \ + " pushq %rax\n" \ + " pushq %r8\n" \ + " pushq %r9\n" \ + " pushq %r10\n" \ + " pushq %r11\n" \ + " pushq %rbx\n" \ + " pushq %rbp\n" \ + " pushq %r12\n" \ + " pushq %r13\n" \ + " pushq %r14\n" \ + " pushq %r15\n" +#define RESTORE_REGS_STRING \ + " popq %r15\n" \ + " popq %r14\n" \ + " popq %r13\n" \ + " popq %r12\n" \ + " popq %rbp\n" \ + " popq %rbx\n" \ + " popq %r11\n" \ + " popq %r10\n" \ + " popq %r9\n" \ + " popq %r8\n" \ + " popq %rax\n" \ + " popq %rcx\n" \ + " popq %rdx\n" \ + " popq %rsi\n" \ + " popq %rdi\n" \ + /* Skip orig_ax, ip, cs */ \ + " addq $24, %rsp\n" +#else +#define SAVE_REGS_STRING \ + /* Skip cs, ip, orig_ax and gs. */ \ + " subl $16, %esp\n" \ + " pushl %fs\n" \ + " pushl %es\n" \ + " pushl %ds\n" \ + " pushl %eax\n" \ + " pushl %ebp\n" \ + " pushl %edi\n" \ + " pushl %esi\n" \ + " pushl %edx\n" \ + " pushl %ecx\n" \ + " pushl %ebx\n" +#define RESTORE_REGS_STRING \ + " popl %ebx\n" \ + " popl %ecx\n" \ + " popl %edx\n" \ + " popl %esi\n" \ + " popl %edi\n" \ + " popl %ebp\n" \ + " popl %eax\n" \ + /* Skip ds, es, fs, gs, orig_ax, and ip. Note: don't pop cs here*/\ + " addl $24, %esp\n" +#endif + +/* Ensure if the instruction can be boostable */ +extern int can_boost(kprobe_opcode_t *instruction); +/* Recover instruction if given address is probed */ +extern unsigned long recover_probed_instruction(kprobe_opcode_t *buf, + unsigned long addr); +/* + * Copy an instruction and adjust the displacement if the instruction + * uses the %rip-relative addressing mode. + */ +extern int __copy_instruction(u8 *dest, u8 *src); + +/* Generate a relative-jump/call instruction */ +extern void synthesize_reljump(void *from, void *to); +extern void synthesize_relcall(void *from, void *to); + +#ifdef CONFIG_OPTPROBES +extern int setup_detour_execution(struct kprobe *p, struct pt_regs *regs, int reenter); +extern unsigned long __recover_optprobed_insn(kprobe_opcode_t *buf, unsigned long addr); +#else /* !CONFIG_OPTPROBES */ +static inline int setup_detour_execution(struct kprobe *p, struct pt_regs *regs, int reenter) +{ + return 0; +} +static inline unsigned long __recover_optprobed_insn(kprobe_opcode_t *buf, unsigned long addr) +{ + return addr; +} +#endif + +#ifdef CONFIG_KPROBES_ON_FTRACE +extern int skip_singlestep(struct kprobe *p, struct pt_regs *regs, + struct kprobe_ctlblk *kcb); +#else +static inline int skip_singlestep(struct kprobe *p, struct pt_regs *regs, + struct kprobe_ctlblk *kcb) +{ + return 0; +} +#endif +#endif diff --git a/arch/x86/kernel/kprobes.c b/arch/x86/kernel/kprobes/core.c index 7da647d8b64..67e6d19ef1b 100644 --- a/arch/x86/kernel/kprobes.c +++ b/arch/x86/kernel/kprobes/core.c @@ -30,16 +30,15 @@ * <jkenisto@us.ibm.com> and Prasanna S Panchamukhi * <prasanna@in.ibm.com> added function-return probes. * 2005-May Rusty Lynch <rusty.lynch@intel.com> - * Added function return probes functionality + * Added function return probes functionality * 2006-Feb Masami Hiramatsu <hiramatu@sdl.hitachi.co.jp> added - * kprobe-booster and kretprobe-booster for i386. + * kprobe-booster and kretprobe-booster for i386. * 2007-Dec Masami Hiramatsu <mhiramat@redhat.com> added kprobe-booster - * and kretprobe-booster for x86-64 + * and kretprobe-booster for x86-64 * 2007-Dec Masami Hiramatsu <mhiramat@redhat.com>, Arjan van de Ven - * <arjan@infradead.org> and Jim Keniston <jkenisto@us.ibm.com> - * unified x86 kprobes code. + * <arjan@infradead.org> and Jim Keniston <jkenisto@us.ibm.com> + * unified x86 kprobes code. */ - #include <linux/kprobes.h> #include <linux/ptrace.h> #include <linux/string.h> @@ -59,6 +58,8 @@ #include <asm/insn.h> #include <asm/debugreg.h> +#include "common.h" + void jprobe_return_end(void); DEFINE_PER_CPU(struct kprobe *, current_kprobe) = NULL; @@ -77,7 +78,7 @@ DEFINE_PER_CPU(struct kprobe_ctlblk, kprobe_ctlblk); * Groups, and some special opcodes can not boost. * This is non-const and volatile to keep gcc from statically * optimizing it out, as variable_test_bit makes gcc think only - * *(unsigned long*) is used. + * *(unsigned long*) is used. */ static volatile u32 twobyte_is_boostable[256 / 32] = { /* 0 1 2 3 4 5 6 7 8 9 a b c d e f */ @@ -108,14 +109,16 @@ struct kretprobe_blackpoint kretprobe_blacklist[] = { doesn't switch kernel stack.*/ {NULL, NULL} /* Terminator */ }; + const int kretprobe_blacklist_size = ARRAY_SIZE(kretprobe_blacklist); -static void __kprobes __synthesize_relative_insn(void *from, void *to, u8 op) +static nokprobe_inline void +__synthesize_relative_insn(void *from, void *to, u8 op) { struct __arch_relative_insn { u8 op; s32 raddr; - } __attribute__((packed)) *insn; + } __packed *insn; insn = (struct __arch_relative_insn *)from; insn->raddr = (s32)((long)(to) - ((long)(from) + 5)); @@ -123,15 +126,23 @@ static void __kprobes __synthesize_relative_insn(void *from, void *to, u8 op) } /* Insert a jump instruction at address 'from', which jumps to address 'to'.*/ -static void __kprobes synthesize_reljump(void *from, void *to) +void synthesize_reljump(void *from, void *to) { __synthesize_relative_insn(from, to, RELATIVEJUMP_OPCODE); } +NOKPROBE_SYMBOL(synthesize_reljump); + +/* Insert a call instruction at address 'from', which calls address 'to'.*/ +void synthesize_relcall(void *from, void *to) +{ + __synthesize_relative_insn(from, to, RELATIVECALL_OPCODE); +} +NOKPROBE_SYMBOL(synthesize_relcall); /* * Skip the prefixes of the instruction. */ -static kprobe_opcode_t *__kprobes skip_prefixes(kprobe_opcode_t *insn) +static kprobe_opcode_t *skip_prefixes(kprobe_opcode_t *insn) { insn_attr_t attr; @@ -146,12 +157,13 @@ static kprobe_opcode_t *__kprobes skip_prefixes(kprobe_opcode_t *insn) #endif return insn; } +NOKPROBE_SYMBOL(skip_prefixes); /* * Returns non-zero if opcode is boostable. * RIP relative instructions are adjusted at copying time in 64 bits mode */ -static int __kprobes can_boost(kprobe_opcode_t *opcodes) +int can_boost(kprobe_opcode_t *opcodes) { kprobe_opcode_t opcode; kprobe_opcode_t *orig_opcodes = opcodes; @@ -207,13 +219,15 @@ retry: } } -/* Recover the probed instruction at addr for further analysis. */ -static int recover_probed_instruction(kprobe_opcode_t *buf, unsigned long addr) +static unsigned long +__recover_probed_insn(kprobe_opcode_t *buf, unsigned long addr) { struct kprobe *kp; + kp = get_kprobe((void *)addr); + /* There is no probe, return original address */ if (!kp) - return -EINVAL; + return addr; /* * Basically, kp->ainsn.insn has an original instruction. @@ -230,14 +244,29 @@ static int recover_probed_instruction(kprobe_opcode_t *buf, unsigned long addr) */ memcpy(buf, kp->addr, MAX_INSN_SIZE * sizeof(kprobe_opcode_t)); buf[0] = kp->opcode; - return 0; + return (unsigned long)buf; +} + +/* + * Recover the probed instruction at addr for further analysis. + * Caller must lock kprobes by kprobe_mutex, or disable preemption + * for preventing to release referencing kprobes. + */ +unsigned long recover_probed_instruction(kprobe_opcode_t *buf, unsigned long addr) +{ + unsigned long __addr; + + __addr = __recover_optprobed_insn(buf, addr); + if (__addr != addr) + return __addr; + + return __recover_probed_insn(buf, addr); } /* Check if paddr is at an instruction boundary */ -static int __kprobes can_probe(unsigned long paddr) +static int can_probe(unsigned long paddr) { - int ret; - unsigned long addr, offset = 0; + unsigned long addr, __addr, offset = 0; struct insn insn; kprobe_opcode_t buf[MAX_INSN_SIZE]; @@ -247,26 +276,24 @@ static int __kprobes can_probe(unsigned long paddr) /* Decode instructions */ addr = paddr - offset; while (addr < paddr) { - kernel_insn_init(&insn, (void *)addr); - insn_get_opcode(&insn); - /* * Check if the instruction has been modified by another * kprobe, in which case we replace the breakpoint by the * original instruction in our buffer. + * Also, jump optimization will change the breakpoint to + * relative-jump. Since the relative-jump itself is + * normally used, we just go through if there is no kprobe. */ - if (insn.opcode.bytes[0] == BREAKPOINT_INSTRUCTION) { - ret = recover_probed_instruction(buf, addr); - if (ret) - /* - * Another debugging subsystem might insert - * this breakpoint. In that case, we can't - * recover it. - */ - return 0; - kernel_insn_init(&insn, buf); - } + __addr = recover_probed_instruction(buf, addr); + kernel_insn_init(&insn, (void *)__addr); insn_get_length(&insn); + + /* + * Another debugging subsystem might insert this breakpoint. + * In that case, we can't recover it. + */ + if (insn.opcode.bytes[0] == BREAKPOINT_INSTRUCTION) + return 0; addr += insn.length; } @@ -276,7 +303,7 @@ static int __kprobes can_probe(unsigned long paddr) /* * Returns non-zero if opcode modifies the interrupt flag. */ -static int __kprobes is_IF_modifier(kprobe_opcode_t *insn) +static int is_IF_modifier(kprobe_opcode_t *insn) { /* Skip prefixes */ insn = skip_prefixes(insn); @@ -299,24 +326,16 @@ static int __kprobes is_IF_modifier(kprobe_opcode_t *insn) * If not, return null. * Only applicable to 64-bit x86. */ -static int __kprobes __copy_instruction(u8 *dest, u8 *src, int recover) +int __copy_instruction(u8 *dest, u8 *src) { struct insn insn; - int ret; kprobe_opcode_t buf[MAX_INSN_SIZE]; - kernel_insn_init(&insn, src); - if (recover) { - insn_get_opcode(&insn); - if (insn.opcode.bytes[0] == BREAKPOINT_INSTRUCTION) { - ret = recover_probed_instruction(buf, - (unsigned long)src); - if (ret) - return 0; - kernel_insn_init(&insn, buf); - } - } + kernel_insn_init(&insn, (void *)recover_probed_instruction(buf, (unsigned long)src)); insn_get_length(&insn); + /* Another subsystem puts a breakpoint, failed to recover */ + if (insn.opcode.bytes[0] == BREAKPOINT_INSTRUCTION) + return 0; memcpy(dest, insn.kaddr, insn.length); #ifdef CONFIG_X86_64 @@ -337,9 +356,12 @@ static int __kprobes __copy_instruction(u8 *dest, u8 *src, int recover) * extension of the original signed 32-bit displacement would * have given. */ - newdisp = (u8 *) src + (s64) insn.displacement.value - - (u8 *) dest; - BUG_ON((s64) (s32) newdisp != newdisp); /* Sanity check. */ + newdisp = (u8 *) src + (s64) insn.displacement.value - (u8 *) dest; + if ((s64) (s32) newdisp != newdisp) { + pr_err("Kprobes error: new displacement does not fit into s32 (%llx)\n", newdisp); + pr_err("\tSrc: %p, Dest: %p, old disp: %x\n", src, dest, insn.displacement.value); + return 0; + } disp = (u8 *) dest + insn_offset_displacement(&insn); *(s32 *) disp = (s32) newdisp; } @@ -347,23 +369,34 @@ static int __kprobes __copy_instruction(u8 *dest, u8 *src, int recover) return insn.length; } -static void __kprobes arch_copy_kprobe(struct kprobe *p) +static int arch_copy_kprobe(struct kprobe *p) { + int ret; + + /* Copy an instruction with recovering if other optprobe modifies it.*/ + ret = __copy_instruction(p->ainsn.insn, p->addr); + if (!ret) + return -EINVAL; + /* - * Copy an instruction without recovering int3, because it will be - * put by another subsystem. + * __copy_instruction can modify the displacement of the instruction, + * but it doesn't affect boostable check. */ - __copy_instruction(p->ainsn.insn, p->addr, 0); - - if (can_boost(p->addr)) + if (can_boost(p->ainsn.insn)) p->ainsn.boostable = 0; else p->ainsn.boostable = -1; - p->opcode = *p->addr; + /* Check whether the instruction modifies Interrupt Flag or not */ + p->ainsn.if_modifier = is_IF_modifier(p->ainsn.insn); + + /* Also, displacement change doesn't affect the first byte */ + p->opcode = p->ainsn.insn[0]; + + return 0; } -int __kprobes arch_prepare_kprobe(struct kprobe *p) +int arch_prepare_kprobe(struct kprobe *p) { if (alternatives_text_reserved(p->addr, p->addr)) return -EINVAL; @@ -374,21 +407,21 @@ int __kprobes arch_prepare_kprobe(struct kprobe *p) p->ainsn.insn = get_insn_slot(); if (!p->ainsn.insn) return -ENOMEM; - arch_copy_kprobe(p); - return 0; + + return arch_copy_kprobe(p); } -void __kprobes arch_arm_kprobe(struct kprobe *p) +void arch_arm_kprobe(struct kprobe *p) { text_poke(p->addr, ((unsigned char []){BREAKPOINT_INSTRUCTION}), 1); } -void __kprobes arch_disarm_kprobe(struct kprobe *p) +void arch_disarm_kprobe(struct kprobe *p) { text_poke(p->addr, &p->opcode, 1); } -void __kprobes arch_remove_kprobe(struct kprobe *p) +void arch_remove_kprobe(struct kprobe *p) { if (p->ainsn.insn) { free_insn_slot(p->ainsn.insn, (p->ainsn.boostable == 1)); @@ -396,7 +429,8 @@ void __kprobes arch_remove_kprobe(struct kprobe *p) } } -static void __kprobes save_previous_kprobe(struct kprobe_ctlblk *kcb) +static nokprobe_inline void +save_previous_kprobe(struct kprobe_ctlblk *kcb) { kcb->prev_kprobe.kp = kprobe_running(); kcb->prev_kprobe.status = kcb->kprobe_status; @@ -404,7 +438,8 @@ static void __kprobes save_previous_kprobe(struct kprobe_ctlblk *kcb) kcb->prev_kprobe.saved_flags = kcb->kprobe_saved_flags; } -static void __kprobes restore_previous_kprobe(struct kprobe_ctlblk *kcb) +static nokprobe_inline void +restore_previous_kprobe(struct kprobe_ctlblk *kcb) { __this_cpu_write(current_kprobe, kcb->prev_kprobe.kp); kcb->kprobe_status = kcb->prev_kprobe.status; @@ -412,17 +447,18 @@ static void __kprobes restore_previous_kprobe(struct kprobe_ctlblk *kcb) kcb->kprobe_saved_flags = kcb->prev_kprobe.saved_flags; } -static void __kprobes set_current_kprobe(struct kprobe *p, struct pt_regs *regs, - struct kprobe_ctlblk *kcb) +static nokprobe_inline void +set_current_kprobe(struct kprobe *p, struct pt_regs *regs, + struct kprobe_ctlblk *kcb) { __this_cpu_write(current_kprobe, p); kcb->kprobe_saved_flags = kcb->kprobe_old_flags = (regs->flags & (X86_EFLAGS_TF | X86_EFLAGS_IF)); - if (is_IF_modifier(p->ainsn.insn)) + if (p->ainsn.if_modifier) kcb->kprobe_saved_flags &= ~X86_EFLAGS_IF; } -static void __kprobes clear_btf(void) +static nokprobe_inline void clear_btf(void) { if (test_thread_flag(TIF_BLOCKSTEP)) { unsigned long debugctl = get_debugctlmsr(); @@ -432,7 +468,7 @@ static void __kprobes clear_btf(void) } } -static void __kprobes restore_btf(void) +static nokprobe_inline void restore_btf(void) { if (test_thread_flag(TIF_BLOCKSTEP)) { unsigned long debugctl = get_debugctlmsr(); @@ -442,8 +478,7 @@ static void __kprobes restore_btf(void) } } -void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri, - struct pt_regs *regs) +void arch_prepare_kretprobe(struct kretprobe_instance *ri, struct pt_regs *regs) { unsigned long *sara = stack_addr(regs); @@ -452,17 +487,10 @@ void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri, /* Replace the return addr with trampoline addr */ *sara = (unsigned long) &kretprobe_trampoline; } +NOKPROBE_SYMBOL(arch_prepare_kretprobe); -#ifdef CONFIG_OPTPROBES -static int __kprobes setup_detour_execution(struct kprobe *p, - struct pt_regs *regs, - int reenter); -#else -#define setup_detour_execution(p, regs, reenter) (0) -#endif - -static void __kprobes setup_singlestep(struct kprobe *p, struct pt_regs *regs, - struct kprobe_ctlblk *kcb, int reenter) +static void setup_singlestep(struct kprobe *p, struct pt_regs *regs, + struct kprobe_ctlblk *kcb, int reenter) { if (setup_detour_execution(p, regs, reenter)) return; @@ -498,22 +526,24 @@ static void __kprobes setup_singlestep(struct kprobe *p, struct pt_regs *regs, else regs->ip = (unsigned long)p->ainsn.insn; } +NOKPROBE_SYMBOL(setup_singlestep); /* * We have reentered the kprobe_handler(), since another probe was hit while * within the handler. We save the original kprobes variables and just single * step on the instruction of the new probe without calling any user handlers. */ -static int __kprobes reenter_kprobe(struct kprobe *p, struct pt_regs *regs, - struct kprobe_ctlblk *kcb) +static int reenter_kprobe(struct kprobe *p, struct pt_regs *regs, + struct kprobe_ctlblk *kcb) { switch (kcb->kprobe_status) { case KPROBE_HIT_SSDONE: case KPROBE_HIT_ACTIVE: + case KPROBE_HIT_SS: kprobes_inc_nmissed_count(p); setup_singlestep(p, regs, kcb, 1); break; - case KPROBE_HIT_SS: + case KPROBE_REENTER: /* A probe has been hit in the codepath leading up to, or just * after, single-stepping of a probed instruction. This entire * codepath should strictly reside in .kprobes.text section. @@ -532,17 +562,21 @@ static int __kprobes reenter_kprobe(struct kprobe *p, struct pt_regs *regs, return 1; } +NOKPROBE_SYMBOL(reenter_kprobe); /* * Interrupts are disabled on entry as trap3 is an interrupt gate and they * remain disabled throughout this function. */ -static int __kprobes kprobe_handler(struct pt_regs *regs) +int kprobe_int3_handler(struct pt_regs *regs) { kprobe_opcode_t *addr; struct kprobe *p; struct kprobe_ctlblk *kcb; + if (user_mode_vm(regs)) + return 0; + addr = (kprobe_opcode_t *)(regs->ip - sizeof(kprobe_opcode_t)); /* * We don't want to be preempted for the entire @@ -591,7 +625,8 @@ static int __kprobes kprobe_handler(struct pt_regs *regs) } else if (kprobe_running()) { p = __this_cpu_read(current_kprobe); if (p->break_handler && p->break_handler(p, regs)) { - setup_singlestep(p, regs, kcb, 0); + if (!skip_singlestep(p, regs, kcb)) + setup_singlestep(p, regs, kcb, 0); return 1; } } /* else: not a kprobe fault; let the kernel handle it */ @@ -599,75 +634,13 @@ static int __kprobes kprobe_handler(struct pt_regs *regs) preempt_enable_no_resched(); return 0; } - -#ifdef CONFIG_X86_64 -#define SAVE_REGS_STRING \ - /* Skip cs, ip, orig_ax. */ \ - " subq $24, %rsp\n" \ - " pushq %rdi\n" \ - " pushq %rsi\n" \ - " pushq %rdx\n" \ - " pushq %rcx\n" \ - " pushq %rax\n" \ - " pushq %r8\n" \ - " pushq %r9\n" \ - " pushq %r10\n" \ - " pushq %r11\n" \ - " pushq %rbx\n" \ - " pushq %rbp\n" \ - " pushq %r12\n" \ - " pushq %r13\n" \ - " pushq %r14\n" \ - " pushq %r15\n" -#define RESTORE_REGS_STRING \ - " popq %r15\n" \ - " popq %r14\n" \ - " popq %r13\n" \ - " popq %r12\n" \ - " popq %rbp\n" \ - " popq %rbx\n" \ - " popq %r11\n" \ - " popq %r10\n" \ - " popq %r9\n" \ - " popq %r8\n" \ - " popq %rax\n" \ - " popq %rcx\n" \ - " popq %rdx\n" \ - " popq %rsi\n" \ - " popq %rdi\n" \ - /* Skip orig_ax, ip, cs */ \ - " addq $24, %rsp\n" -#else -#define SAVE_REGS_STRING \ - /* Skip cs, ip, orig_ax and gs. */ \ - " subl $16, %esp\n" \ - " pushl %fs\n" \ - " pushl %es\n" \ - " pushl %ds\n" \ - " pushl %eax\n" \ - " pushl %ebp\n" \ - " pushl %edi\n" \ - " pushl %esi\n" \ - " pushl %edx\n" \ - " pushl %ecx\n" \ - " pushl %ebx\n" -#define RESTORE_REGS_STRING \ - " popl %ebx\n" \ - " popl %ecx\n" \ - " popl %edx\n" \ - " popl %esi\n" \ - " popl %edi\n" \ - " popl %ebp\n" \ - " popl %eax\n" \ - /* Skip ds, es, fs, gs, orig_ax, and ip. Note: don't pop cs here*/\ - " addl $24, %esp\n" -#endif +NOKPROBE_SYMBOL(kprobe_int3_handler); /* * When a retprobed function returns, this code saves registers and * calls trampoline_handler() runs, which calls the kretprobe's handler. */ -static void __used __kprobes kretprobe_trampoline_holder(void) +static void __used kretprobe_trampoline_holder(void) { asm volatile ( ".global kretprobe_trampoline\n" @@ -698,15 +671,17 @@ static void __used __kprobes kretprobe_trampoline_holder(void) #endif " ret\n"); } +NOKPROBE_SYMBOL(kretprobe_trampoline_holder); +NOKPROBE_SYMBOL(kretprobe_trampoline); /* * Called from kretprobe_trampoline */ -static __used __kprobes void *trampoline_handler(struct pt_regs *regs) +__visible __used void *trampoline_handler(struct pt_regs *regs) { struct kretprobe_instance *ri = NULL; struct hlist_head *head, empty_rp; - struct hlist_node *node, *tmp; + struct hlist_node *tmp; unsigned long flags, orig_ret_address = 0; unsigned long trampoline_address = (unsigned long)&kretprobe_trampoline; kprobe_opcode_t *correct_ret_addr = NULL; @@ -736,7 +711,7 @@ static __used __kprobes void *trampoline_handler(struct pt_regs *regs) * will be the real return address, and all the rest will * point to kretprobe_trampoline. */ - hlist_for_each_entry_safe(ri, node, tmp, head, hlist) { + hlist_for_each_entry_safe(ri, tmp, head, hlist) { if (ri->task != current) /* another task is sharing our hash bucket */ continue; @@ -755,7 +730,7 @@ static __used __kprobes void *trampoline_handler(struct pt_regs *regs) kretprobe_assert(ri, orig_ret_address, trampoline_address); correct_ret_addr = ri->ret_addr; - hlist_for_each_entry_safe(ri, node, tmp, head, hlist) { + hlist_for_each_entry_safe(ri, tmp, head, hlist) { if (ri->task != current) /* another task is sharing our hash bucket */ continue; @@ -782,12 +757,13 @@ static __used __kprobes void *trampoline_handler(struct pt_regs *regs) kretprobe_hash_unlock(current, &flags); - hlist_for_each_entry_safe(ri, node, tmp, &empty_rp, hlist) { + hlist_for_each_entry_safe(ri, tmp, &empty_rp, hlist) { hlist_del(&ri->hlist); kfree(ri); } return (void *)orig_ret_address; } +NOKPROBE_SYMBOL(trampoline_handler); /* * Called after single-stepping. p->addr is the address of the @@ -816,8 +792,8 @@ static __used __kprobes void *trampoline_handler(struct pt_regs *regs) * jump instruction after the copied instruction, that jumps to the next * instruction after the probepoint. */ -static void __kprobes resume_execution(struct kprobe *p, - struct pt_regs *regs, struct kprobe_ctlblk *kcb) +static void resume_execution(struct kprobe *p, struct pt_regs *regs, + struct kprobe_ctlblk *kcb) { unsigned long *tos = stack_addr(regs); unsigned long copy_ip = (unsigned long)p->ainsn.insn; @@ -892,12 +868,13 @@ static void __kprobes resume_execution(struct kprobe *p, no_change: restore_btf(); } +NOKPROBE_SYMBOL(resume_execution); /* * Interrupts are disabled on entry as trap1 is an interrupt gate and they * remain disabled throughout this function. */ -static int __kprobes post_kprobe_handler(struct pt_regs *regs) +int kprobe_debug_handler(struct pt_regs *regs) { struct kprobe *cur = kprobe_running(); struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); @@ -932,15 +909,17 @@ out: return 1; } +NOKPROBE_SYMBOL(kprobe_debug_handler); -int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr) +int kprobe_fault_handler(struct pt_regs *regs, int trapnr) { struct kprobe *cur = kprobe_running(); struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); - switch (kcb->kprobe_status) { - case KPROBE_HIT_SS: - case KPROBE_REENTER: + if (unlikely(regs->ip == (unsigned long)cur->ainsn.insn)) { + /* This must happen on single-stepping */ + WARN_ON(kcb->kprobe_status != KPROBE_HIT_SS && + kcb->kprobe_status != KPROBE_REENTER); /* * We are here because the instruction being single * stepped caused a page fault. We reset the current @@ -955,9 +934,8 @@ int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr) else reset_current_kprobe(); preempt_enable_no_resched(); - break; - case KPROBE_HIT_ACTIVE: - case KPROBE_HIT_SSDONE: + } else if (kcb->kprobe_status == KPROBE_HIT_ACTIVE || + kcb->kprobe_status == KPROBE_HIT_SSDONE) { /* * We increment the nmissed count for accounting, * we can also use npre/npostfault count for accounting @@ -986,18 +964,17 @@ int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr) * fixup routine could not handle it, * Let do_page_fault() fix it. */ - break; - default: - break; } + return 0; } +NOKPROBE_SYMBOL(kprobe_fault_handler); /* * Wrapper routine for handling exceptions. */ -int __kprobes kprobe_exceptions_notify(struct notifier_block *self, - unsigned long val, void *data) +int kprobe_exceptions_notify(struct notifier_block *self, unsigned long val, + void *data) { struct die_args *args = data; int ret = NOTIFY_DONE; @@ -1005,22 +982,7 @@ int __kprobes kprobe_exceptions_notify(struct notifier_block *self, if (args->regs && user_mode_vm(args->regs)) return ret; - switch (val) { - case DIE_INT3: - if (kprobe_handler(args->regs)) - ret = NOTIFY_STOP; - break; - case DIE_DEBUG: - if (post_kprobe_handler(args->regs)) { - /* - * Reset the BS bit in dr6 (pointed by args->err) to - * denote completion of processing - */ - (*(unsigned long *)ERR_PTR(args->err)) &= ~DR_STEP; - ret = NOTIFY_STOP; - } - break; - case DIE_GPF: + if (val == DIE_GPF) { /* * To be potentially processing a kprobe fault and to * trust the result from kprobe_running(), we have @@ -1029,14 +991,12 @@ int __kprobes kprobe_exceptions_notify(struct notifier_block *self, if (!preemptible() && kprobe_running() && kprobe_fault_handler(args->regs, args->trapnr)) ret = NOTIFY_STOP; - break; - default: - break; } return ret; } +NOKPROBE_SYMBOL(kprobe_exceptions_notify); -int __kprobes setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs) +int setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs) { struct jprobe *jp = container_of(p, struct jprobe, kp); unsigned long addr; @@ -1060,8 +1020,9 @@ int __kprobes setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs) regs->ip = (unsigned long)(jp->entry); return 1; } +NOKPROBE_SYMBOL(setjmp_pre_handler); -void __kprobes jprobe_return(void) +void jprobe_return(void) { struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); @@ -1077,8 +1038,10 @@ void __kprobes jprobe_return(void) " nop \n"::"b" (kcb->jprobe_saved_sp):"memory"); } +NOKPROBE_SYMBOL(jprobe_return); +NOKPROBE_SYMBOL(jprobe_return_end); -int __kprobes longjmp_break_handler(struct kprobe *p, struct pt_regs *regs) +int longjmp_break_handler(struct kprobe *p, struct pt_regs *regs) { struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); u8 *addr = (u8 *) (regs->ip - 1); @@ -1092,9 +1055,9 @@ int __kprobes longjmp_break_handler(struct kprobe *p, struct pt_regs *regs) "current sp %p does not match saved sp %p\n", stack_addr(regs), kcb->jprobe_saved_sp); printk(KERN_ERR "Saved registers for jprobe %p\n", jp); - show_registers(saved_regs); + show_regs(saved_regs); printk(KERN_ERR "Current registers\n"); - show_registers(regs); + show_regs(regs); BUG(); } *regs = kcb->jprobe_saved_regs; @@ -1106,470 +1069,22 @@ int __kprobes longjmp_break_handler(struct kprobe *p, struct pt_regs *regs) } return 0; } +NOKPROBE_SYMBOL(longjmp_break_handler); - -#ifdef CONFIG_OPTPROBES - -/* Insert a call instruction at address 'from', which calls address 'to'.*/ -static void __kprobes synthesize_relcall(void *from, void *to) +bool arch_within_kprobe_blacklist(unsigned long addr) { - __synthesize_relative_insn(from, to, RELATIVECALL_OPCODE); + return (addr >= (unsigned long)__kprobes_text_start && + addr < (unsigned long)__kprobes_text_end) || + (addr >= (unsigned long)__entry_text_start && + addr < (unsigned long)__entry_text_end); } -/* Insert a move instruction which sets a pointer to eax/rdi (1st arg). */ -static void __kprobes synthesize_set_arg1(kprobe_opcode_t *addr, - unsigned long val) -{ -#ifdef CONFIG_X86_64 - *addr++ = 0x48; - *addr++ = 0xbf; -#else - *addr++ = 0xb8; -#endif - *(unsigned long *)addr = val; -} - -static void __used __kprobes kprobes_optinsn_template_holder(void) -{ - asm volatile ( - ".global optprobe_template_entry\n" - "optprobe_template_entry: \n" -#ifdef CONFIG_X86_64 - /* We don't bother saving the ss register */ - " pushq %rsp\n" - " pushfq\n" - SAVE_REGS_STRING - " movq %rsp, %rsi\n" - ".global optprobe_template_val\n" - "optprobe_template_val: \n" - ASM_NOP5 - ASM_NOP5 - ".global optprobe_template_call\n" - "optprobe_template_call: \n" - ASM_NOP5 - /* Move flags to rsp */ - " movq 144(%rsp), %rdx\n" - " movq %rdx, 152(%rsp)\n" - RESTORE_REGS_STRING - /* Skip flags entry */ - " addq $8, %rsp\n" - " popfq\n" -#else /* CONFIG_X86_32 */ - " pushf\n" - SAVE_REGS_STRING - " movl %esp, %edx\n" - ".global optprobe_template_val\n" - "optprobe_template_val: \n" - ASM_NOP5 - ".global optprobe_template_call\n" - "optprobe_template_call: \n" - ASM_NOP5 - RESTORE_REGS_STRING - " addl $4, %esp\n" /* skip cs */ - " popf\n" -#endif - ".global optprobe_template_end\n" - "optprobe_template_end: \n"); -} - -#define TMPL_MOVE_IDX \ - ((long)&optprobe_template_val - (long)&optprobe_template_entry) -#define TMPL_CALL_IDX \ - ((long)&optprobe_template_call - (long)&optprobe_template_entry) -#define TMPL_END_IDX \ - ((long)&optprobe_template_end - (long)&optprobe_template_entry) - -#define INT3_SIZE sizeof(kprobe_opcode_t) - -/* Optimized kprobe call back function: called from optinsn */ -static void __kprobes optimized_callback(struct optimized_kprobe *op, - struct pt_regs *regs) -{ - struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); - unsigned long flags; - - /* This is possible if op is under delayed unoptimizing */ - if (kprobe_disabled(&op->kp)) - return; - - local_irq_save(flags); - if (kprobe_running()) { - kprobes_inc_nmissed_count(&op->kp); - } else { - /* Save skipped registers */ -#ifdef CONFIG_X86_64 - regs->cs = __KERNEL_CS; -#else - regs->cs = __KERNEL_CS | get_kernel_rpl(); - regs->gs = 0; -#endif - regs->ip = (unsigned long)op->kp.addr + INT3_SIZE; - regs->orig_ax = ~0UL; - - __this_cpu_write(current_kprobe, &op->kp); - kcb->kprobe_status = KPROBE_HIT_ACTIVE; - opt_pre_handler(&op->kp, regs); - __this_cpu_write(current_kprobe, NULL); - } - local_irq_restore(flags); -} - -static int __kprobes copy_optimized_instructions(u8 *dest, u8 *src) -{ - int len = 0, ret; - - while (len < RELATIVEJUMP_SIZE) { - ret = __copy_instruction(dest + len, src + len, 1); - if (!ret || !can_boost(dest + len)) - return -EINVAL; - len += ret; - } - /* Check whether the address range is reserved */ - if (ftrace_text_reserved(src, src + len - 1) || - alternatives_text_reserved(src, src + len - 1) || - jump_label_text_reserved(src, src + len - 1)) - return -EBUSY; - - return len; -} - -/* Check whether insn is indirect jump */ -static int __kprobes insn_is_indirect_jump(struct insn *insn) -{ - return ((insn->opcode.bytes[0] == 0xff && - (X86_MODRM_REG(insn->modrm.value) & 6) == 4) || /* Jump */ - insn->opcode.bytes[0] == 0xea); /* Segment based jump */ -} - -/* Check whether insn jumps into specified address range */ -static int insn_jump_into_range(struct insn *insn, unsigned long start, int len) -{ - unsigned long target = 0; - - switch (insn->opcode.bytes[0]) { - case 0xe0: /* loopne */ - case 0xe1: /* loope */ - case 0xe2: /* loop */ - case 0xe3: /* jcxz */ - case 0xe9: /* near relative jump */ - case 0xeb: /* short relative jump */ - break; - case 0x0f: - if ((insn->opcode.bytes[1] & 0xf0) == 0x80) /* jcc near */ - break; - return 0; - default: - if ((insn->opcode.bytes[0] & 0xf0) == 0x70) /* jcc short */ - break; - return 0; - } - target = (unsigned long)insn->next_byte + insn->immediate.value; - - return (start <= target && target <= start + len); -} - -/* Decode whole function to ensure any instructions don't jump into target */ -static int __kprobes can_optimize(unsigned long paddr) -{ - int ret; - unsigned long addr, size = 0, offset = 0; - struct insn insn; - kprobe_opcode_t buf[MAX_INSN_SIZE]; - - /* Lookup symbol including addr */ - if (!kallsyms_lookup_size_offset(paddr, &size, &offset)) - return 0; - - /* - * Do not optimize in the entry code due to the unstable - * stack handling. - */ - if ((paddr >= (unsigned long )__entry_text_start) && - (paddr < (unsigned long )__entry_text_end)) - return 0; - - /* Check there is enough space for a relative jump. */ - if (size - offset < RELATIVEJUMP_SIZE) - return 0; - - /* Decode instructions */ - addr = paddr - offset; - while (addr < paddr - offset + size) { /* Decode until function end */ - if (search_exception_tables(addr)) - /* - * Since some fixup code will jumps into this function, - * we can't optimize kprobe in this function. - */ - return 0; - kernel_insn_init(&insn, (void *)addr); - insn_get_opcode(&insn); - if (insn.opcode.bytes[0] == BREAKPOINT_INSTRUCTION) { - ret = recover_probed_instruction(buf, addr); - if (ret) - return 0; - kernel_insn_init(&insn, buf); - } - insn_get_length(&insn); - /* Recover address */ - insn.kaddr = (void *)addr; - insn.next_byte = (void *)(addr + insn.length); - /* Check any instructions don't jump into target */ - if (insn_is_indirect_jump(&insn) || - insn_jump_into_range(&insn, paddr + INT3_SIZE, - RELATIVE_ADDR_SIZE)) - return 0; - addr += insn.length; - } - - return 1; -} - -/* Check optimized_kprobe can actually be optimized. */ -int __kprobes arch_check_optimized_kprobe(struct optimized_kprobe *op) -{ - int i; - struct kprobe *p; - - for (i = 1; i < op->optinsn.size; i++) { - p = get_kprobe(op->kp.addr + i); - if (p && !kprobe_disabled(p)) - return -EEXIST; - } - - return 0; -} - -/* Check the addr is within the optimized instructions. */ -int __kprobes arch_within_optimized_kprobe(struct optimized_kprobe *op, - unsigned long addr) -{ - return ((unsigned long)op->kp.addr <= addr && - (unsigned long)op->kp.addr + op->optinsn.size > addr); -} - -/* Free optimized instruction slot */ -static __kprobes -void __arch_remove_optimized_kprobe(struct optimized_kprobe *op, int dirty) -{ - if (op->optinsn.insn) { - free_optinsn_slot(op->optinsn.insn, dirty); - op->optinsn.insn = NULL; - op->optinsn.size = 0; - } -} - -void __kprobes arch_remove_optimized_kprobe(struct optimized_kprobe *op) -{ - __arch_remove_optimized_kprobe(op, 1); -} - -/* - * Copy replacing target instructions - * Target instructions MUST be relocatable (checked inside) - */ -int __kprobes arch_prepare_optimized_kprobe(struct optimized_kprobe *op) -{ - u8 *buf; - int ret; - long rel; - - if (!can_optimize((unsigned long)op->kp.addr)) - return -EILSEQ; - - op->optinsn.insn = get_optinsn_slot(); - if (!op->optinsn.insn) - return -ENOMEM; - - /* - * Verify if the address gap is in 2GB range, because this uses - * a relative jump. - */ - rel = (long)op->optinsn.insn - (long)op->kp.addr + RELATIVEJUMP_SIZE; - if (abs(rel) > 0x7fffffff) - return -ERANGE; - - buf = (u8 *)op->optinsn.insn; - - /* Copy instructions into the out-of-line buffer */ - ret = copy_optimized_instructions(buf + TMPL_END_IDX, op->kp.addr); - if (ret < 0) { - __arch_remove_optimized_kprobe(op, 0); - return ret; - } - op->optinsn.size = ret; - - /* Copy arch-dep-instance from template */ - memcpy(buf, &optprobe_template_entry, TMPL_END_IDX); - - /* Set probe information */ - synthesize_set_arg1(buf + TMPL_MOVE_IDX, (unsigned long)op); - - /* Set probe function call */ - synthesize_relcall(buf + TMPL_CALL_IDX, optimized_callback); - - /* Set returning jmp instruction at the tail of out-of-line buffer */ - synthesize_reljump(buf + TMPL_END_IDX + op->optinsn.size, - (u8 *)op->kp.addr + op->optinsn.size); - - flush_icache_range((unsigned long) buf, - (unsigned long) buf + TMPL_END_IDX + - op->optinsn.size + RELATIVEJUMP_SIZE); - return 0; -} - -#define MAX_OPTIMIZE_PROBES 256 -static struct text_poke_param *jump_poke_params; -static struct jump_poke_buffer { - u8 buf[RELATIVEJUMP_SIZE]; -} *jump_poke_bufs; - -static void __kprobes setup_optimize_kprobe(struct text_poke_param *tprm, - u8 *insn_buf, - struct optimized_kprobe *op) -{ - s32 rel = (s32)((long)op->optinsn.insn - - ((long)op->kp.addr + RELATIVEJUMP_SIZE)); - - /* Backup instructions which will be replaced by jump address */ - memcpy(op->optinsn.copied_insn, op->kp.addr + INT3_SIZE, - RELATIVE_ADDR_SIZE); - - insn_buf[0] = RELATIVEJUMP_OPCODE; - *(s32 *)(&insn_buf[1]) = rel; - - tprm->addr = op->kp.addr; - tprm->opcode = insn_buf; - tprm->len = RELATIVEJUMP_SIZE; -} - -/* - * Replace breakpoints (int3) with relative jumps. - * Caller must call with locking kprobe_mutex and text_mutex. - */ -void __kprobes arch_optimize_kprobes(struct list_head *oplist) -{ - struct optimized_kprobe *op, *tmp; - int c = 0; - - list_for_each_entry_safe(op, tmp, oplist, list) { - WARN_ON(kprobe_disabled(&op->kp)); - /* Setup param */ - setup_optimize_kprobe(&jump_poke_params[c], - jump_poke_bufs[c].buf, op); - list_del_init(&op->list); - if (++c >= MAX_OPTIMIZE_PROBES) - break; - } - - /* - * text_poke_smp doesn't support NMI/MCE code modifying. - * However, since kprobes itself also doesn't support NMI/MCE - * code probing, it's not a problem. - */ - text_poke_smp_batch(jump_poke_params, c); -} - -static void __kprobes setup_unoptimize_kprobe(struct text_poke_param *tprm, - u8 *insn_buf, - struct optimized_kprobe *op) -{ - /* Set int3 to first byte for kprobes */ - insn_buf[0] = BREAKPOINT_INSTRUCTION; - memcpy(insn_buf + 1, op->optinsn.copied_insn, RELATIVE_ADDR_SIZE); - - tprm->addr = op->kp.addr; - tprm->opcode = insn_buf; - tprm->len = RELATIVEJUMP_SIZE; -} - -/* - * Recover original instructions and breakpoints from relative jumps. - * Caller must call with locking kprobe_mutex. - */ -extern void arch_unoptimize_kprobes(struct list_head *oplist, - struct list_head *done_list) -{ - struct optimized_kprobe *op, *tmp; - int c = 0; - - list_for_each_entry_safe(op, tmp, oplist, list) { - /* Setup param */ - setup_unoptimize_kprobe(&jump_poke_params[c], - jump_poke_bufs[c].buf, op); - list_move(&op->list, done_list); - if (++c >= MAX_OPTIMIZE_PROBES) - break; - } - - /* - * text_poke_smp doesn't support NMI/MCE code modifying. - * However, since kprobes itself also doesn't support NMI/MCE - * code probing, it's not a problem. - */ - text_poke_smp_batch(jump_poke_params, c); -} - -/* Replace a relative jump with a breakpoint (int3). */ -void __kprobes arch_unoptimize_kprobe(struct optimized_kprobe *op) -{ - u8 buf[RELATIVEJUMP_SIZE]; - - /* Set int3 to first byte for kprobes */ - buf[0] = BREAKPOINT_INSTRUCTION; - memcpy(buf + 1, op->optinsn.copied_insn, RELATIVE_ADDR_SIZE); - text_poke_smp(op->kp.addr, buf, RELATIVEJUMP_SIZE); -} - -static int __kprobes setup_detour_execution(struct kprobe *p, - struct pt_regs *regs, - int reenter) -{ - struct optimized_kprobe *op; - - if (p->flags & KPROBE_FLAG_OPTIMIZED) { - /* This kprobe is really able to run optimized path. */ - op = container_of(p, struct optimized_kprobe, kp); - /* Detour through copied instructions */ - regs->ip = (unsigned long)op->optinsn.insn + TMPL_END_IDX; - if (!reenter) - reset_current_kprobe(); - preempt_enable_no_resched(); - return 1; - } - return 0; -} - -static int __kprobes init_poke_params(void) -{ - /* Allocate code buffer and parameter array */ - jump_poke_bufs = kmalloc(sizeof(struct jump_poke_buffer) * - MAX_OPTIMIZE_PROBES, GFP_KERNEL); - if (!jump_poke_bufs) - return -ENOMEM; - - jump_poke_params = kmalloc(sizeof(struct text_poke_param) * - MAX_OPTIMIZE_PROBES, GFP_KERNEL); - if (!jump_poke_params) { - kfree(jump_poke_bufs); - jump_poke_bufs = NULL; - return -ENOMEM; - } - - return 0; -} -#else /* !CONFIG_OPTPROBES */ -static int __kprobes init_poke_params(void) -{ - return 0; -} -#endif - int __init arch_init_kprobes(void) { - return init_poke_params(); + return 0; } -int __kprobes arch_trampoline_kprobe(struct kprobe *p) +int arch_trampoline_kprobe(struct kprobe *p) { return 0; } diff --git a/arch/x86/kernel/kprobes/ftrace.c b/arch/x86/kernel/kprobes/ftrace.c new file mode 100644 index 00000000000..717b02a22e6 --- /dev/null +++ b/arch/x86/kernel/kprobes/ftrace.c @@ -0,0 +1,96 @@ +/* + * Dynamic Ftrace based Kprobes Optimization + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * + * Copyright (C) Hitachi Ltd., 2012 + */ +#include <linux/kprobes.h> +#include <linux/ptrace.h> +#include <linux/hardirq.h> +#include <linux/preempt.h> +#include <linux/ftrace.h> + +#include "common.h" + +static nokprobe_inline +int __skip_singlestep(struct kprobe *p, struct pt_regs *regs, + struct kprobe_ctlblk *kcb) +{ + /* + * Emulate singlestep (and also recover regs->ip) + * as if there is a 5byte nop + */ + regs->ip = (unsigned long)p->addr + MCOUNT_INSN_SIZE; + if (unlikely(p->post_handler)) { + kcb->kprobe_status = KPROBE_HIT_SSDONE; + p->post_handler(p, regs, 0); + } + __this_cpu_write(current_kprobe, NULL); + return 1; +} + +int skip_singlestep(struct kprobe *p, struct pt_regs *regs, + struct kprobe_ctlblk *kcb) +{ + if (kprobe_ftrace(p)) + return __skip_singlestep(p, regs, kcb); + else + return 0; +} +NOKPROBE_SYMBOL(skip_singlestep); + +/* Ftrace callback handler for kprobes */ +void kprobe_ftrace_handler(unsigned long ip, unsigned long parent_ip, + struct ftrace_ops *ops, struct pt_regs *regs) +{ + struct kprobe *p; + struct kprobe_ctlblk *kcb; + unsigned long flags; + + /* Disable irq for emulating a breakpoint and avoiding preempt */ + local_irq_save(flags); + + p = get_kprobe((kprobe_opcode_t *)ip); + if (unlikely(!p) || kprobe_disabled(p)) + goto end; + + kcb = get_kprobe_ctlblk(); + if (kprobe_running()) { + kprobes_inc_nmissed_count(p); + } else { + /* Kprobe handler expects regs->ip = ip + 1 as breakpoint hit */ + regs->ip = ip + sizeof(kprobe_opcode_t); + + __this_cpu_write(current_kprobe, p); + kcb->kprobe_status = KPROBE_HIT_ACTIVE; + if (!p->pre_handler || !p->pre_handler(p, regs)) + __skip_singlestep(p, regs, kcb); + /* + * If pre_handler returns !0, it sets regs->ip and + * resets current kprobe. + */ + } +end: + local_irq_restore(flags); +} +NOKPROBE_SYMBOL(kprobe_ftrace_handler); + +int arch_prepare_kprobe_ftrace(struct kprobe *p) +{ + p->ainsn.insn = NULL; + p->ainsn.boostable = -1; + return 0; +} diff --git a/arch/x86/kernel/kprobes/opt.c b/arch/x86/kernel/kprobes/opt.c new file mode 100644 index 00000000000..f304773285a --- /dev/null +++ b/arch/x86/kernel/kprobes/opt.c @@ -0,0 +1,445 @@ +/* + * Kernel Probes Jump Optimization (Optprobes) + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * + * Copyright (C) IBM Corporation, 2002, 2004 + * Copyright (C) Hitachi Ltd., 2012 + */ +#include <linux/kprobes.h> +#include <linux/ptrace.h> +#include <linux/string.h> +#include <linux/slab.h> +#include <linux/hardirq.h> +#include <linux/preempt.h> +#include <linux/module.h> +#include <linux/kdebug.h> +#include <linux/kallsyms.h> +#include <linux/ftrace.h> + +#include <asm/cacheflush.h> +#include <asm/desc.h> +#include <asm/pgtable.h> +#include <asm/uaccess.h> +#include <asm/alternative.h> +#include <asm/insn.h> +#include <asm/debugreg.h> + +#include "common.h" + +unsigned long __recover_optprobed_insn(kprobe_opcode_t *buf, unsigned long addr) +{ + struct optimized_kprobe *op; + struct kprobe *kp; + long offs; + int i; + + for (i = 0; i < RELATIVEJUMP_SIZE; i++) { + kp = get_kprobe((void *)addr - i); + /* This function only handles jump-optimized kprobe */ + if (kp && kprobe_optimized(kp)) { + op = container_of(kp, struct optimized_kprobe, kp); + /* If op->list is not empty, op is under optimizing */ + if (list_empty(&op->list)) + goto found; + } + } + + return addr; +found: + /* + * If the kprobe can be optimized, original bytes which can be + * overwritten by jump destination address. In this case, original + * bytes must be recovered from op->optinsn.copied_insn buffer. + */ + memcpy(buf, (void *)addr, MAX_INSN_SIZE * sizeof(kprobe_opcode_t)); + if (addr == (unsigned long)kp->addr) { + buf[0] = kp->opcode; + memcpy(buf + 1, op->optinsn.copied_insn, RELATIVE_ADDR_SIZE); + } else { + offs = addr - (unsigned long)kp->addr - 1; + memcpy(buf, op->optinsn.copied_insn + offs, RELATIVE_ADDR_SIZE - offs); + } + + return (unsigned long)buf; +} + +/* Insert a move instruction which sets a pointer to eax/rdi (1st arg). */ +static void synthesize_set_arg1(kprobe_opcode_t *addr, unsigned long val) +{ +#ifdef CONFIG_X86_64 + *addr++ = 0x48; + *addr++ = 0xbf; +#else + *addr++ = 0xb8; +#endif + *(unsigned long *)addr = val; +} + +asm ( + ".global optprobe_template_entry\n" + "optprobe_template_entry:\n" +#ifdef CONFIG_X86_64 + /* We don't bother saving the ss register */ + " pushq %rsp\n" + " pushfq\n" + SAVE_REGS_STRING + " movq %rsp, %rsi\n" + ".global optprobe_template_val\n" + "optprobe_template_val:\n" + ASM_NOP5 + ASM_NOP5 + ".global optprobe_template_call\n" + "optprobe_template_call:\n" + ASM_NOP5 + /* Move flags to rsp */ + " movq 144(%rsp), %rdx\n" + " movq %rdx, 152(%rsp)\n" + RESTORE_REGS_STRING + /* Skip flags entry */ + " addq $8, %rsp\n" + " popfq\n" +#else /* CONFIG_X86_32 */ + " pushf\n" + SAVE_REGS_STRING + " movl %esp, %edx\n" + ".global optprobe_template_val\n" + "optprobe_template_val:\n" + ASM_NOP5 + ".global optprobe_template_call\n" + "optprobe_template_call:\n" + ASM_NOP5 + RESTORE_REGS_STRING + " addl $4, %esp\n" /* skip cs */ + " popf\n" +#endif + ".global optprobe_template_end\n" + "optprobe_template_end:\n"); + +#define TMPL_MOVE_IDX \ + ((long)&optprobe_template_val - (long)&optprobe_template_entry) +#define TMPL_CALL_IDX \ + ((long)&optprobe_template_call - (long)&optprobe_template_entry) +#define TMPL_END_IDX \ + ((long)&optprobe_template_end - (long)&optprobe_template_entry) + +#define INT3_SIZE sizeof(kprobe_opcode_t) + +/* Optimized kprobe call back function: called from optinsn */ +static void +optimized_callback(struct optimized_kprobe *op, struct pt_regs *regs) +{ + struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); + unsigned long flags; + + /* This is possible if op is under delayed unoptimizing */ + if (kprobe_disabled(&op->kp)) + return; + + local_irq_save(flags); + if (kprobe_running()) { + kprobes_inc_nmissed_count(&op->kp); + } else { + /* Save skipped registers */ +#ifdef CONFIG_X86_64 + regs->cs = __KERNEL_CS; +#else + regs->cs = __KERNEL_CS | get_kernel_rpl(); + regs->gs = 0; +#endif + regs->ip = (unsigned long)op->kp.addr + INT3_SIZE; + regs->orig_ax = ~0UL; + + __this_cpu_write(current_kprobe, &op->kp); + kcb->kprobe_status = KPROBE_HIT_ACTIVE; + opt_pre_handler(&op->kp, regs); + __this_cpu_write(current_kprobe, NULL); + } + local_irq_restore(flags); +} +NOKPROBE_SYMBOL(optimized_callback); + +static int copy_optimized_instructions(u8 *dest, u8 *src) +{ + int len = 0, ret; + + while (len < RELATIVEJUMP_SIZE) { + ret = __copy_instruction(dest + len, src + len); + if (!ret || !can_boost(dest + len)) + return -EINVAL; + len += ret; + } + /* Check whether the address range is reserved */ + if (ftrace_text_reserved(src, src + len - 1) || + alternatives_text_reserved(src, src + len - 1) || + jump_label_text_reserved(src, src + len - 1)) + return -EBUSY; + + return len; +} + +/* Check whether insn is indirect jump */ +static int insn_is_indirect_jump(struct insn *insn) +{ + return ((insn->opcode.bytes[0] == 0xff && + (X86_MODRM_REG(insn->modrm.value) & 6) == 4) || /* Jump */ + insn->opcode.bytes[0] == 0xea); /* Segment based jump */ +} + +/* Check whether insn jumps into specified address range */ +static int insn_jump_into_range(struct insn *insn, unsigned long start, int len) +{ + unsigned long target = 0; + + switch (insn->opcode.bytes[0]) { + case 0xe0: /* loopne */ + case 0xe1: /* loope */ + case 0xe2: /* loop */ + case 0xe3: /* jcxz */ + case 0xe9: /* near relative jump */ + case 0xeb: /* short relative jump */ + break; + case 0x0f: + if ((insn->opcode.bytes[1] & 0xf0) == 0x80) /* jcc near */ + break; + return 0; + default: + if ((insn->opcode.bytes[0] & 0xf0) == 0x70) /* jcc short */ + break; + return 0; + } + target = (unsigned long)insn->next_byte + insn->immediate.value; + + return (start <= target && target <= start + len); +} + +/* Decode whole function to ensure any instructions don't jump into target */ +static int can_optimize(unsigned long paddr) +{ + unsigned long addr, size = 0, offset = 0; + struct insn insn; + kprobe_opcode_t buf[MAX_INSN_SIZE]; + + /* Lookup symbol including addr */ + if (!kallsyms_lookup_size_offset(paddr, &size, &offset)) + return 0; + + /* + * Do not optimize in the entry code due to the unstable + * stack handling. + */ + if ((paddr >= (unsigned long)__entry_text_start) && + (paddr < (unsigned long)__entry_text_end)) + return 0; + + /* Check there is enough space for a relative jump. */ + if (size - offset < RELATIVEJUMP_SIZE) + return 0; + + /* Decode instructions */ + addr = paddr - offset; + while (addr < paddr - offset + size) { /* Decode until function end */ + if (search_exception_tables(addr)) + /* + * Since some fixup code will jumps into this function, + * we can't optimize kprobe in this function. + */ + return 0; + kernel_insn_init(&insn, (void *)recover_probed_instruction(buf, addr)); + insn_get_length(&insn); + /* Another subsystem puts a breakpoint */ + if (insn.opcode.bytes[0] == BREAKPOINT_INSTRUCTION) + return 0; + /* Recover address */ + insn.kaddr = (void *)addr; + insn.next_byte = (void *)(addr + insn.length); + /* Check any instructions don't jump into target */ + if (insn_is_indirect_jump(&insn) || + insn_jump_into_range(&insn, paddr + INT3_SIZE, + RELATIVE_ADDR_SIZE)) + return 0; + addr += insn.length; + } + + return 1; +} + +/* Check optimized_kprobe can actually be optimized. */ +int arch_check_optimized_kprobe(struct optimized_kprobe *op) +{ + int i; + struct kprobe *p; + + for (i = 1; i < op->optinsn.size; i++) { + p = get_kprobe(op->kp.addr + i); + if (p && !kprobe_disabled(p)) + return -EEXIST; + } + + return 0; +} + +/* Check the addr is within the optimized instructions. */ +int arch_within_optimized_kprobe(struct optimized_kprobe *op, + unsigned long addr) +{ + return ((unsigned long)op->kp.addr <= addr && + (unsigned long)op->kp.addr + op->optinsn.size > addr); +} + +/* Free optimized instruction slot */ +static +void __arch_remove_optimized_kprobe(struct optimized_kprobe *op, int dirty) +{ + if (op->optinsn.insn) { + free_optinsn_slot(op->optinsn.insn, dirty); + op->optinsn.insn = NULL; + op->optinsn.size = 0; + } +} + +void arch_remove_optimized_kprobe(struct optimized_kprobe *op) +{ + __arch_remove_optimized_kprobe(op, 1); +} + +/* + * Copy replacing target instructions + * Target instructions MUST be relocatable (checked inside) + * This is called when new aggr(opt)probe is allocated or reused. + */ +int arch_prepare_optimized_kprobe(struct optimized_kprobe *op) +{ + u8 *buf; + int ret; + long rel; + + if (!can_optimize((unsigned long)op->kp.addr)) + return -EILSEQ; + + op->optinsn.insn = get_optinsn_slot(); + if (!op->optinsn.insn) + return -ENOMEM; + + /* + * Verify if the address gap is in 2GB range, because this uses + * a relative jump. + */ + rel = (long)op->optinsn.insn - (long)op->kp.addr + RELATIVEJUMP_SIZE; + if (abs(rel) > 0x7fffffff) + return -ERANGE; + + buf = (u8 *)op->optinsn.insn; + + /* Copy instructions into the out-of-line buffer */ + ret = copy_optimized_instructions(buf + TMPL_END_IDX, op->kp.addr); + if (ret < 0) { + __arch_remove_optimized_kprobe(op, 0); + return ret; + } + op->optinsn.size = ret; + + /* Copy arch-dep-instance from template */ + memcpy(buf, &optprobe_template_entry, TMPL_END_IDX); + + /* Set probe information */ + synthesize_set_arg1(buf + TMPL_MOVE_IDX, (unsigned long)op); + + /* Set probe function call */ + synthesize_relcall(buf + TMPL_CALL_IDX, optimized_callback); + + /* Set returning jmp instruction at the tail of out-of-line buffer */ + synthesize_reljump(buf + TMPL_END_IDX + op->optinsn.size, + (u8 *)op->kp.addr + op->optinsn.size); + + flush_icache_range((unsigned long) buf, + (unsigned long) buf + TMPL_END_IDX + + op->optinsn.size + RELATIVEJUMP_SIZE); + return 0; +} + +/* + * Replace breakpoints (int3) with relative jumps. + * Caller must call with locking kprobe_mutex and text_mutex. + */ +void arch_optimize_kprobes(struct list_head *oplist) +{ + struct optimized_kprobe *op, *tmp; + u8 insn_buf[RELATIVEJUMP_SIZE]; + + list_for_each_entry_safe(op, tmp, oplist, list) { + s32 rel = (s32)((long)op->optinsn.insn - + ((long)op->kp.addr + RELATIVEJUMP_SIZE)); + + WARN_ON(kprobe_disabled(&op->kp)); + + /* Backup instructions which will be replaced by jump address */ + memcpy(op->optinsn.copied_insn, op->kp.addr + INT3_SIZE, + RELATIVE_ADDR_SIZE); + + insn_buf[0] = RELATIVEJUMP_OPCODE; + *(s32 *)(&insn_buf[1]) = rel; + + text_poke_bp(op->kp.addr, insn_buf, RELATIVEJUMP_SIZE, + op->optinsn.insn); + + list_del_init(&op->list); + } +} + +/* Replace a relative jump with a breakpoint (int3). */ +void arch_unoptimize_kprobe(struct optimized_kprobe *op) +{ + u8 insn_buf[RELATIVEJUMP_SIZE]; + + /* Set int3 to first byte for kprobes */ + insn_buf[0] = BREAKPOINT_INSTRUCTION; + memcpy(insn_buf + 1, op->optinsn.copied_insn, RELATIVE_ADDR_SIZE); + text_poke_bp(op->kp.addr, insn_buf, RELATIVEJUMP_SIZE, + op->optinsn.insn); +} + +/* + * Recover original instructions and breakpoints from relative jumps. + * Caller must call with locking kprobe_mutex. + */ +extern void arch_unoptimize_kprobes(struct list_head *oplist, + struct list_head *done_list) +{ + struct optimized_kprobe *op, *tmp; + + list_for_each_entry_safe(op, tmp, oplist, list) { + arch_unoptimize_kprobe(op); + list_move(&op->list, done_list); + } +} + +int setup_detour_execution(struct kprobe *p, struct pt_regs *regs, int reenter) +{ + struct optimized_kprobe *op; + + if (p->flags & KPROBE_FLAG_OPTIMIZED) { + /* This kprobe is really able to run optimized path. */ + op = container_of(p, struct optimized_kprobe, kp); + /* Detour through copied instructions */ + regs->ip = (unsigned long)op->optinsn.insn + TMPL_END_IDX; + if (!reenter) + reset_current_kprobe(); + preempt_enable_no_resched(); + return 1; + } + return 0; +} +NOKPROBE_SYMBOL(setup_detour_execution); diff --git a/arch/x86/kernel/ksysfs.c b/arch/x86/kernel/ksysfs.c new file mode 100644 index 00000000000..c2bedaea11f --- /dev/null +++ b/arch/x86/kernel/ksysfs.c @@ -0,0 +1,340 @@ +/* + * Architecture specific sysfs attributes in /sys/kernel + * + * Copyright (C) 2007, Intel Corp. + * Huang Ying <ying.huang@intel.com> + * Copyright (C) 2013, 2013 Red Hat, Inc. + * Dave Young <dyoung@redhat.com> + * + * This file is released under the GPLv2 + */ + +#include <linux/kobject.h> +#include <linux/string.h> +#include <linux/sysfs.h> +#include <linux/init.h> +#include <linux/stat.h> +#include <linux/slab.h> +#include <linux/mm.h> + +#include <asm/io.h> +#include <asm/setup.h> + +static ssize_t version_show(struct kobject *kobj, + struct kobj_attribute *attr, char *buf) +{ + return sprintf(buf, "0x%04x\n", boot_params.hdr.version); +} + +static struct kobj_attribute boot_params_version_attr = __ATTR_RO(version); + +static ssize_t boot_params_data_read(struct file *fp, struct kobject *kobj, + struct bin_attribute *bin_attr, + char *buf, loff_t off, size_t count) +{ + memcpy(buf, (void *)&boot_params + off, count); + return count; +} + +static struct bin_attribute boot_params_data_attr = { + .attr = { + .name = "data", + .mode = S_IRUGO, + }, + .read = boot_params_data_read, + .size = sizeof(boot_params), +}; + +static struct attribute *boot_params_version_attrs[] = { + &boot_params_version_attr.attr, + NULL, +}; + +static struct bin_attribute *boot_params_data_attrs[] = { + &boot_params_data_attr, + NULL, +}; + +static struct attribute_group boot_params_attr_group = { + .attrs = boot_params_version_attrs, + .bin_attrs = boot_params_data_attrs, +}; + +static int kobj_to_setup_data_nr(struct kobject *kobj, int *nr) +{ + const char *name; + + name = kobject_name(kobj); + return kstrtoint(name, 10, nr); +} + +static int get_setup_data_paddr(int nr, u64 *paddr) +{ + int i = 0; + struct setup_data *data; + u64 pa_data = boot_params.hdr.setup_data; + + while (pa_data) { + if (nr == i) { + *paddr = pa_data; + return 0; + } + data = ioremap_cache(pa_data, sizeof(*data)); + if (!data) + return -ENOMEM; + + pa_data = data->next; + iounmap(data); + i++; + } + return -EINVAL; +} + +static int __init get_setup_data_size(int nr, size_t *size) +{ + int i = 0; + struct setup_data *data; + u64 pa_data = boot_params.hdr.setup_data; + + while (pa_data) { + data = ioremap_cache(pa_data, sizeof(*data)); + if (!data) + return -ENOMEM; + if (nr == i) { + *size = data->len; + iounmap(data); + return 0; + } + + pa_data = data->next; + iounmap(data); + i++; + } + return -EINVAL; +} + +static ssize_t type_show(struct kobject *kobj, + struct kobj_attribute *attr, char *buf) +{ + int nr, ret; + u64 paddr; + struct setup_data *data; + + ret = kobj_to_setup_data_nr(kobj, &nr); + if (ret) + return ret; + + ret = get_setup_data_paddr(nr, &paddr); + if (ret) + return ret; + data = ioremap_cache(paddr, sizeof(*data)); + if (!data) + return -ENOMEM; + + ret = sprintf(buf, "0x%x\n", data->type); + iounmap(data); + return ret; +} + +static ssize_t setup_data_data_read(struct file *fp, + struct kobject *kobj, + struct bin_attribute *bin_attr, + char *buf, + loff_t off, size_t count) +{ + int nr, ret = 0; + u64 paddr; + struct setup_data *data; + void *p; + + ret = kobj_to_setup_data_nr(kobj, &nr); + if (ret) + return ret; + + ret = get_setup_data_paddr(nr, &paddr); + if (ret) + return ret; + data = ioremap_cache(paddr, sizeof(*data)); + if (!data) + return -ENOMEM; + + if (off > data->len) { + ret = -EINVAL; + goto out; + } + + if (count > data->len - off) + count = data->len - off; + + if (!count) + goto out; + + ret = count; + p = ioremap_cache(paddr + sizeof(*data), data->len); + if (!p) { + ret = -ENOMEM; + goto out; + } + memcpy(buf, p + off, count); + iounmap(p); +out: + iounmap(data); + return ret; +} + +static struct kobj_attribute type_attr = __ATTR_RO(type); + +static struct bin_attribute data_attr = { + .attr = { + .name = "data", + .mode = S_IRUGO, + }, + .read = setup_data_data_read, +}; + +static struct attribute *setup_data_type_attrs[] = { + &type_attr.attr, + NULL, +}; + +static struct bin_attribute *setup_data_data_attrs[] = { + &data_attr, + NULL, +}; + +static struct attribute_group setup_data_attr_group = { + .attrs = setup_data_type_attrs, + .bin_attrs = setup_data_data_attrs, +}; + +static int __init create_setup_data_node(struct kobject *parent, + struct kobject **kobjp, int nr) +{ + int ret = 0; + size_t size; + struct kobject *kobj; + char name[16]; /* should be enough for setup_data nodes numbers */ + snprintf(name, 16, "%d", nr); + + kobj = kobject_create_and_add(name, parent); + if (!kobj) + return -ENOMEM; + + ret = get_setup_data_size(nr, &size); + if (ret) + goto out_kobj; + + data_attr.size = size; + ret = sysfs_create_group(kobj, &setup_data_attr_group); + if (ret) + goto out_kobj; + *kobjp = kobj; + + return 0; +out_kobj: + kobject_put(kobj); + return ret; +} + +static void __init cleanup_setup_data_node(struct kobject *kobj) +{ + sysfs_remove_group(kobj, &setup_data_attr_group); + kobject_put(kobj); +} + +static int __init get_setup_data_total_num(u64 pa_data, int *nr) +{ + int ret = 0; + struct setup_data *data; + + *nr = 0; + while (pa_data) { + *nr += 1; + data = ioremap_cache(pa_data, sizeof(*data)); + if (!data) { + ret = -ENOMEM; + goto out; + } + pa_data = data->next; + iounmap(data); + } + +out: + return ret; +} + +static int __init create_setup_data_nodes(struct kobject *parent) +{ + struct kobject *setup_data_kobj, **kobjp; + u64 pa_data; + int i, j, nr, ret = 0; + + pa_data = boot_params.hdr.setup_data; + if (!pa_data) + return 0; + + setup_data_kobj = kobject_create_and_add("setup_data", parent); + if (!setup_data_kobj) { + ret = -ENOMEM; + goto out; + } + + ret = get_setup_data_total_num(pa_data, &nr); + if (ret) + goto out_setup_data_kobj; + + kobjp = kmalloc(sizeof(*kobjp) * nr, GFP_KERNEL); + if (!kobjp) { + ret = -ENOMEM; + goto out_setup_data_kobj; + } + + for (i = 0; i < nr; i++) { + ret = create_setup_data_node(setup_data_kobj, kobjp + i, i); + if (ret) + goto out_clean_nodes; + } + + kfree(kobjp); + return 0; + +out_clean_nodes: + for (j = i - 1; j > 0; j--) + cleanup_setup_data_node(*(kobjp + j)); + kfree(kobjp); +out_setup_data_kobj: + kobject_put(setup_data_kobj); +out: + return ret; +} + +static int __init boot_params_ksysfs_init(void) +{ + int ret; + struct kobject *boot_params_kobj; + + boot_params_kobj = kobject_create_and_add("boot_params", + kernel_kobj); + if (!boot_params_kobj) { + ret = -ENOMEM; + goto out; + } + + ret = sysfs_create_group(boot_params_kobj, &boot_params_attr_group); + if (ret) + goto out_boot_params_kobj; + + ret = create_setup_data_nodes(boot_params_kobj); + if (ret) + goto out_create_group; + + return 0; +out_create_group: + sysfs_remove_group(boot_params_kobj, &boot_params_attr_group); +out_boot_params_kobj: + kobject_put(boot_params_kobj); +out: + return ret; +} + +arch_initcall(boot_params_ksysfs_init); diff --git a/arch/x86/kernel/kvm.c b/arch/x86/kernel/kvm.c index f0c6fd6f176..3dd8e2c4d74 100644 --- a/arch/x86/kernel/kvm.c +++ b/arch/x86/kernel/kvm.c @@ -20,6 +20,7 @@ * Authors: Anthony Liguori <aliguori@us.ibm.com> */ +#include <linux/context_tracking.h> #include <linux/module.h> #include <linux/kernel.h> #include <linux/kvm_para.h> @@ -33,11 +34,17 @@ #include <linux/sched.h> #include <linux/slab.h> #include <linux/kprobes.h> +#include <linux/debugfs.h> #include <asm/timer.h> #include <asm/cpu.h> #include <asm/traps.h> #include <asm/desc.h> #include <asm/tlbflush.h> +#include <asm/idle.h> +#include <asm/apic.h> +#include <asm/apicdef.h> +#include <asm/hypervisor.h> +#include <asm/kvm_guest.h> static int kvmapf = 1; @@ -58,6 +65,15 @@ static int parse_no_stealacc(char *arg) early_param("no-steal-acc", parse_no_stealacc); +static int kvmclock_vsyscall = 1; +static int parse_no_kvmclock_vsyscall(char *arg) +{ + kvmclock_vsyscall = 0; + return 0; +} + +early_param("no-kvmclock-vsyscall", parse_no_kvmclock_vsyscall); + static DEFINE_PER_CPU(struct kvm_vcpu_pv_apf_data, apf_reason) __aligned(64); static DEFINE_PER_CPU(struct kvm_steal_time, steal_time) __aligned(64); static int has_steal_clock = 0; @@ -78,7 +94,6 @@ struct kvm_task_sleep_node { u32 token; int cpu; bool halted; - struct mm_struct *mm; }; static struct kvm_task_sleep_head { @@ -107,11 +122,8 @@ void kvm_async_pf_task_wait(u32 token) struct kvm_task_sleep_head *b = &async_pf_sleepers[key]; struct kvm_task_sleep_node n, *e; DEFINE_WAIT(wait); - int cpu, idle; - cpu = get_cpu(); - idle = idle_cpu(cpu); - put_cpu(); + rcu_irq_enter(); spin_lock(&b->lock); e = _find_apf_task(b, token); @@ -120,14 +132,14 @@ void kvm_async_pf_task_wait(u32 token) hlist_del(&e->link); kfree(e); spin_unlock(&b->lock); + + rcu_irq_exit(); return; } n.token = token; n.cpu = smp_processor_id(); - n.mm = current->active_mm; - n.halted = idle || preempt_count() > 1; - atomic_inc(&n.mm->mm_count); + n.halted = is_idle_task(current) || preempt_count() > 1; init_waitqueue_head(&n.wq); hlist_add_head(&n.link, &b->list); spin_unlock(&b->lock); @@ -146,13 +158,16 @@ void kvm_async_pf_task_wait(u32 token) /* * We cannot reschedule. So halt. */ + rcu_irq_exit(); native_safe_halt(); + rcu_irq_enter(); local_irq_disable(); } } if (!n.halted) finish_wait(&n.wq, &wait); + rcu_irq_exit(); return; } EXPORT_SYMBOL_GPL(kvm_async_pf_task_wait); @@ -160,9 +175,6 @@ EXPORT_SYMBOL_GPL(kvm_async_pf_task_wait); static void apf_task_wake_one(struct kvm_task_sleep_node *n) { hlist_del_init(&n->link); - if (!n->mm) - return; - mmdrop(n->mm); if (n->halted) smp_send_reschedule(n->cpu); else if (waitqueue_active(&n->wq)) @@ -206,7 +218,7 @@ again: * async PF was not yet handled. * Add dummy entry for the token. */ - n = kmalloc(sizeof(*n), GFP_ATOMIC); + n = kzalloc(sizeof(*n), GFP_ATOMIC); if (!n) { /* * Allocation failed! Busy wait while other cpu @@ -218,7 +230,6 @@ again: } n->token = token; n->cpu = smp_processor_id(); - n->mm = NULL; init_waitqueue_head(&n->wq); hlist_add_head(&n->link, &b->list); } else @@ -240,23 +251,33 @@ u32 kvm_read_and_reset_pf_reason(void) return reason; } EXPORT_SYMBOL_GPL(kvm_read_and_reset_pf_reason); +NOKPROBE_SYMBOL(kvm_read_and_reset_pf_reason); -dotraplinkage void __kprobes +dotraplinkage void do_async_page_fault(struct pt_regs *regs, unsigned long error_code) { + enum ctx_state prev_state; + switch (kvm_read_and_reset_pf_reason()) { default: - do_page_fault(regs, error_code); + trace_do_page_fault(regs, error_code); break; case KVM_PV_REASON_PAGE_NOT_PRESENT: /* page is swapped out by the host. */ + prev_state = exception_enter(); + exit_idle(); kvm_async_pf_task_wait((u32)read_cr2()); + exception_exit(prev_state); break; case KVM_PV_REASON_PAGE_READY: + rcu_irq_enter(); + exit_idle(); kvm_async_pf_task_wake((u32)read_cr2()); + rcu_irq_exit(); break; } } +NOKPROBE_SYMBOL(do_async_page_fault); static void __init paravirt_ops_setup(void) { @@ -281,18 +302,34 @@ static void kvm_register_steal_time(void) memset(st, 0, sizeof(*st)); - wrmsrl(MSR_KVM_STEAL_TIME, (__pa(st) | KVM_MSR_ENABLED)); - printk(KERN_INFO "kvm-stealtime: cpu %d, msr %lx\n", - cpu, __pa(st)); + wrmsrl(MSR_KVM_STEAL_TIME, (slow_virt_to_phys(st) | KVM_MSR_ENABLED)); + pr_info("kvm-stealtime: cpu %d, msr %llx\n", + cpu, (unsigned long long) slow_virt_to_phys(st)); +} + +static DEFINE_PER_CPU(unsigned long, kvm_apic_eoi) = KVM_PV_EOI_DISABLED; + +static void kvm_guest_apic_eoi_write(u32 reg, u32 val) +{ + /** + * This relies on __test_and_clear_bit to modify the memory + * in a way that is atomic with respect to the local CPU. + * The hypervisor only accesses this memory from the local CPU so + * there's no need for lock or memory barriers. + * An optimization barrier is implied in apic write. + */ + if (__test_and_clear_bit(KVM_PV_EOI_BIT, &__get_cpu_var(kvm_apic_eoi))) + return; + apic_write(APIC_EOI, APIC_EOI_ACK); } -void __cpuinit kvm_guest_cpu_init(void) +void kvm_guest_cpu_init(void) { if (!kvm_para_available()) return; if (kvm_para_has_feature(KVM_FEATURE_ASYNC_PF) && kvmapf) { - u64 pa = __pa(&__get_cpu_var(apf_reason)); + u64 pa = slow_virt_to_phys(&__get_cpu_var(apf_reason)); #ifdef CONFIG_PREEMPT pa |= KVM_ASYNC_PF_SEND_ALWAYS; @@ -303,11 +340,21 @@ void __cpuinit kvm_guest_cpu_init(void) smp_processor_id()); } + if (kvm_para_has_feature(KVM_FEATURE_PV_EOI)) { + unsigned long pa; + /* Size alignment is implied but just to make it explicit. */ + BUILD_BUG_ON(__alignof__(kvm_apic_eoi) < 4); + __get_cpu_var(kvm_apic_eoi) = 0; + pa = slow_virt_to_phys(&__get_cpu_var(kvm_apic_eoi)) + | KVM_MSR_ENABLED; + wrmsrl(MSR_KVM_PV_EOI_EN, pa); + } + if (has_steal_clock) kvm_register_steal_time(); } -static void kvm_pv_disable_apf(void *unused) +static void kvm_pv_disable_apf(void) { if (!__get_cpu_var(apf_reason).enabled) return; @@ -319,11 +366,24 @@ static void kvm_pv_disable_apf(void *unused) smp_processor_id()); } +static void kvm_pv_guest_cpu_reboot(void *unused) +{ + /* + * We disable PV EOI before we load a new kernel by kexec, + * since MSR_KVM_PV_EOI_EN stores a pointer into old kernel's memory. + * New kernel can re-enable when it boots. + */ + if (kvm_para_has_feature(KVM_FEATURE_PV_EOI)) + wrmsrl(MSR_KVM_PV_EOI_EN, 0); + kvm_pv_disable_apf(); + kvm_disable_steal_time(); +} + static int kvm_pv_reboot_notify(struct notifier_block *nb, unsigned long code, void *unused) { if (code == SYS_RESTART) - on_each_cpu(kvm_pv_disable_apf, NULL, 1); + on_each_cpu(kvm_pv_guest_cpu_reboot, NULL, 1); return NOTIFY_DONE; } @@ -359,14 +419,12 @@ void kvm_disable_steal_time(void) #ifdef CONFIG_SMP static void __init kvm_smp_prepare_boot_cpu(void) { -#ifdef CONFIG_KVM_CLOCK - WARN_ON(kvm_register_clock("primary cpu clock")); -#endif kvm_guest_cpu_init(); native_smp_prepare_boot_cpu(); + kvm_spinlock_init(); } -static void __cpuinit kvm_guest_cpu_online(void *dummy) +static void kvm_guest_cpu_online(void *dummy) { kvm_guest_cpu_init(); } @@ -374,12 +432,14 @@ static void __cpuinit kvm_guest_cpu_online(void *dummy) static void kvm_guest_cpu_offline(void *dummy) { kvm_disable_steal_time(); - kvm_pv_disable_apf(NULL); + if (kvm_para_has_feature(KVM_FEATURE_PV_EOI)) + wrmsrl(MSR_KVM_PV_EOI_EN, 0); + kvm_pv_disable_apf(); apf_task_wake_all(); } -static int __cpuinit kvm_cpu_notify(struct notifier_block *self, - unsigned long action, void *hcpu) +static int kvm_cpu_notify(struct notifier_block *self, unsigned long action, + void *hcpu) { int cpu = (unsigned long)hcpu; switch (action) { @@ -398,14 +458,14 @@ static int __cpuinit kvm_cpu_notify(struct notifier_block *self, return NOTIFY_OK; } -static struct notifier_block __cpuinitdata kvm_cpu_notifier = { +static struct notifier_block kvm_cpu_notifier = { .notifier_call = kvm_cpu_notify, }; #endif static void __init kvm_apf_trap_init(void) { - set_intr_gate(14, &async_page_fault); + set_intr_gate(14, async_page_fault); } void __init kvm_guest_init(void) @@ -427,6 +487,12 @@ void __init kvm_guest_init(void) pv_time_ops.steal_clock = kvm_steal_clock; } + if (kvm_para_has_feature(KVM_FEATURE_PV_EOI)) + apic_set_eoi_write(kvm_guest_apic_eoi_write); + + if (kvmclock_vsyscall) + kvm_setup_vsyscall_timeinfo(); + #ifdef CONFIG_SMP smp_ops.smp_prepare_boot_cpu = kvm_smp_prepare_boot_cpu; register_cpu_notifier(&kvm_cpu_notifier); @@ -435,14 +501,329 @@ void __init kvm_guest_init(void) #endif } +static noinline uint32_t __kvm_cpuid_base(void) +{ + if (boot_cpu_data.cpuid_level < 0) + return 0; /* So we don't blow up on old processors */ + + if (cpu_has_hypervisor) + return hypervisor_cpuid_base("KVMKVMKVM\0\0\0", 0); + + return 0; +} + +static inline uint32_t kvm_cpuid_base(void) +{ + static int kvm_cpuid_base = -1; + + if (kvm_cpuid_base == -1) + kvm_cpuid_base = __kvm_cpuid_base(); + + return kvm_cpuid_base; +} + +bool kvm_para_available(void) +{ + return kvm_cpuid_base() != 0; +} +EXPORT_SYMBOL_GPL(kvm_para_available); + +unsigned int kvm_arch_para_features(void) +{ + return cpuid_eax(kvm_cpuid_base() | KVM_CPUID_FEATURES); +} + +static uint32_t __init kvm_detect(void) +{ + return kvm_cpuid_base(); +} + +const struct hypervisor_x86 x86_hyper_kvm __refconst = { + .name = "KVM", + .detect = kvm_detect, + .x2apic_available = kvm_para_available, +}; +EXPORT_SYMBOL_GPL(x86_hyper_kvm); + static __init int activate_jump_labels(void) { if (has_steal_clock) { - jump_label_inc(¶virt_steal_enabled); + static_key_slow_inc(¶virt_steal_enabled); if (steal_acc) - jump_label_inc(¶virt_steal_rq_enabled); + static_key_slow_inc(¶virt_steal_rq_enabled); } return 0; } arch_initcall(activate_jump_labels); + +#ifdef CONFIG_PARAVIRT_SPINLOCKS + +/* Kick a cpu by its apicid. Used to wake up a halted vcpu */ +static void kvm_kick_cpu(int cpu) +{ + int apicid; + unsigned long flags = 0; + + apicid = per_cpu(x86_cpu_to_apicid, cpu); + kvm_hypercall2(KVM_HC_KICK_CPU, flags, apicid); +} + +enum kvm_contention_stat { + TAKEN_SLOW, + TAKEN_SLOW_PICKUP, + RELEASED_SLOW, + RELEASED_SLOW_KICKED, + NR_CONTENTION_STATS +}; + +#ifdef CONFIG_KVM_DEBUG_FS +#define HISTO_BUCKETS 30 + +static struct kvm_spinlock_stats +{ + u32 contention_stats[NR_CONTENTION_STATS]; + u32 histo_spin_blocked[HISTO_BUCKETS+1]; + u64 time_blocked; +} spinlock_stats; + +static u8 zero_stats; + +static inline void check_zero(void) +{ + u8 ret; + u8 old; + + old = ACCESS_ONCE(zero_stats); + if (unlikely(old)) { + ret = cmpxchg(&zero_stats, old, 0); + /* This ensures only one fellow resets the stat */ + if (ret == old) + memset(&spinlock_stats, 0, sizeof(spinlock_stats)); + } +} + +static inline void add_stats(enum kvm_contention_stat var, u32 val) +{ + check_zero(); + spinlock_stats.contention_stats[var] += val; +} + + +static inline u64 spin_time_start(void) +{ + return sched_clock(); +} + +static void __spin_time_accum(u64 delta, u32 *array) +{ + unsigned index; + + index = ilog2(delta); + check_zero(); + + if (index < HISTO_BUCKETS) + array[index]++; + else + array[HISTO_BUCKETS]++; +} + +static inline void spin_time_accum_blocked(u64 start) +{ + u32 delta; + + delta = sched_clock() - start; + __spin_time_accum(delta, spinlock_stats.histo_spin_blocked); + spinlock_stats.time_blocked += delta; +} + +static struct dentry *d_spin_debug; +static struct dentry *d_kvm_debug; + +struct dentry *kvm_init_debugfs(void) +{ + d_kvm_debug = debugfs_create_dir("kvm-guest", NULL); + if (!d_kvm_debug) + printk(KERN_WARNING "Could not create 'kvm' debugfs directory\n"); + + return d_kvm_debug; +} + +static int __init kvm_spinlock_debugfs(void) +{ + struct dentry *d_kvm; + + d_kvm = kvm_init_debugfs(); + if (d_kvm == NULL) + return -ENOMEM; + + d_spin_debug = debugfs_create_dir("spinlocks", d_kvm); + + debugfs_create_u8("zero_stats", 0644, d_spin_debug, &zero_stats); + + debugfs_create_u32("taken_slow", 0444, d_spin_debug, + &spinlock_stats.contention_stats[TAKEN_SLOW]); + debugfs_create_u32("taken_slow_pickup", 0444, d_spin_debug, + &spinlock_stats.contention_stats[TAKEN_SLOW_PICKUP]); + + debugfs_create_u32("released_slow", 0444, d_spin_debug, + &spinlock_stats.contention_stats[RELEASED_SLOW]); + debugfs_create_u32("released_slow_kicked", 0444, d_spin_debug, + &spinlock_stats.contention_stats[RELEASED_SLOW_KICKED]); + + debugfs_create_u64("time_blocked", 0444, d_spin_debug, + &spinlock_stats.time_blocked); + + debugfs_create_u32_array("histo_blocked", 0444, d_spin_debug, + spinlock_stats.histo_spin_blocked, HISTO_BUCKETS + 1); + + return 0; +} +fs_initcall(kvm_spinlock_debugfs); +#else /* !CONFIG_KVM_DEBUG_FS */ +static inline void add_stats(enum kvm_contention_stat var, u32 val) +{ +} + +static inline u64 spin_time_start(void) +{ + return 0; +} + +static inline void spin_time_accum_blocked(u64 start) +{ +} +#endif /* CONFIG_KVM_DEBUG_FS */ + +struct kvm_lock_waiting { + struct arch_spinlock *lock; + __ticket_t want; +}; + +/* cpus 'waiting' on a spinlock to become available */ +static cpumask_t waiting_cpus; + +/* Track spinlock on which a cpu is waiting */ +static DEFINE_PER_CPU(struct kvm_lock_waiting, klock_waiting); + +__visible void kvm_lock_spinning(struct arch_spinlock *lock, __ticket_t want) +{ + struct kvm_lock_waiting *w; + int cpu; + u64 start; + unsigned long flags; + + if (in_nmi()) + return; + + w = &__get_cpu_var(klock_waiting); + cpu = smp_processor_id(); + start = spin_time_start(); + + /* + * Make sure an interrupt handler can't upset things in a + * partially setup state. + */ + local_irq_save(flags); + + /* + * The ordering protocol on this is that the "lock" pointer + * may only be set non-NULL if the "want" ticket is correct. + * If we're updating "want", we must first clear "lock". + */ + w->lock = NULL; + smp_wmb(); + w->want = want; + smp_wmb(); + w->lock = lock; + + add_stats(TAKEN_SLOW, 1); + + /* + * This uses set_bit, which is atomic but we should not rely on its + * reordering gurantees. So barrier is needed after this call. + */ + cpumask_set_cpu(cpu, &waiting_cpus); + + barrier(); + + /* + * Mark entry to slowpath before doing the pickup test to make + * sure we don't deadlock with an unlocker. + */ + __ticket_enter_slowpath(lock); + + /* + * check again make sure it didn't become free while + * we weren't looking. + */ + if (ACCESS_ONCE(lock->tickets.head) == want) { + add_stats(TAKEN_SLOW_PICKUP, 1); + goto out; + } + + /* + * halt until it's our turn and kicked. Note that we do safe halt + * for irq enabled case to avoid hang when lock info is overwritten + * in irq spinlock slowpath and no spurious interrupt occur to save us. + */ + if (arch_irqs_disabled_flags(flags)) + halt(); + else + safe_halt(); + +out: + cpumask_clear_cpu(cpu, &waiting_cpus); + w->lock = NULL; + local_irq_restore(flags); + spin_time_accum_blocked(start); +} +PV_CALLEE_SAVE_REGS_THUNK(kvm_lock_spinning); + +/* Kick vcpu waiting on @lock->head to reach value @ticket */ +static void kvm_unlock_kick(struct arch_spinlock *lock, __ticket_t ticket) +{ + int cpu; + + add_stats(RELEASED_SLOW, 1); + for_each_cpu(cpu, &waiting_cpus) { + const struct kvm_lock_waiting *w = &per_cpu(klock_waiting, cpu); + if (ACCESS_ONCE(w->lock) == lock && + ACCESS_ONCE(w->want) == ticket) { + add_stats(RELEASED_SLOW_KICKED, 1); + kvm_kick_cpu(cpu); + break; + } + } +} + +/* + * Setup pv_lock_ops to exploit KVM_FEATURE_PV_UNHALT if present. + */ +void __init kvm_spinlock_init(void) +{ + if (!kvm_para_available()) + return; + /* Does host kernel support KVM_FEATURE_PV_UNHALT? */ + if (!kvm_para_has_feature(KVM_FEATURE_PV_UNHALT)) + return; + + pv_lock_ops.lock_spinning = PV_CALLEE_SAVE(kvm_lock_spinning); + pv_lock_ops.unlock_kick = kvm_unlock_kick; +} + +static __init int kvm_spinlock_init_jump(void) +{ + if (!kvm_para_available()) + return 0; + if (!kvm_para_has_feature(KVM_FEATURE_PV_UNHALT)) + return 0; + + static_key_slow_inc(¶virt_ticketlocks_enabled); + printk(KERN_INFO "KVM setup paravirtual spinlock\n"); + + return 0; +} +early_initcall(kvm_spinlock_init_jump); + +#endif /* CONFIG_PARAVIRT_SPINLOCKS */ diff --git a/arch/x86/kernel/kvmclock.c b/arch/x86/kernel/kvmclock.c index 44842d756b2..d9156ceecdf 100644 --- a/arch/x86/kernel/kvmclock.c +++ b/arch/x86/kernel/kvmclock.c @@ -22,6 +22,8 @@ #include <asm/msr.h> #include <asm/apic.h> #include <linux/percpu.h> +#include <linux/hardirq.h> +#include <linux/memblock.h> #include <asm/x86_init.h> #include <asm/reboot.h> @@ -38,7 +40,7 @@ static int parse_no_kvmclock(char *arg) early_param("no-kvmclock", parse_no_kvmclock); /* The hypervisor will put information about time periodically here */ -static DEFINE_PER_CPU_SHARED_ALIGNED(struct pvclock_vcpu_time_info, hv_clock); +static struct pvclock_vsyscall_time_info *hv_clock; static struct pvclock_wall_clock wall_clock; /* @@ -46,25 +48,27 @@ static struct pvclock_wall_clock wall_clock; * have elapsed since the hypervisor wrote the data. So we try to account for * that with system time */ -static unsigned long kvm_get_wallclock(void) +static void kvm_get_wallclock(struct timespec *now) { struct pvclock_vcpu_time_info *vcpu_time; - struct timespec ts; int low, high; + int cpu; low = (int)__pa_symbol(&wall_clock); high = ((u64)__pa_symbol(&wall_clock) >> 32); native_write_msr(msr_kvm_wall_clock, low, high); - vcpu_time = &get_cpu_var(hv_clock); - pvclock_read_wallclock(&wall_clock, vcpu_time, &ts); - put_cpu_var(hv_clock); + preempt_disable(); + cpu = smp_processor_id(); - return ts.tv_sec; + vcpu_time = &hv_clock[cpu].pvti; + pvclock_read_wallclock(&wall_clock, vcpu_time, now); + + preempt_enable(); } -static int kvm_set_wallclock(unsigned long now) +static int kvm_set_wallclock(const struct timespec *now) { return -1; } @@ -73,9 +77,11 @@ static cycle_t kvm_clock_read(void) { struct pvclock_vcpu_time_info *src; cycle_t ret; + int cpu; preempt_disable_notrace(); - src = &__get_cpu_var(hv_clock); + cpu = smp_processor_id(); + src = &hv_clock[cpu].pvti; ret = pvclock_clocksource_read(src); preempt_enable_notrace(); return ret; @@ -98,8 +104,15 @@ static cycle_t kvm_clock_get_cycles(struct clocksource *cs) static unsigned long kvm_get_tsc_khz(void) { struct pvclock_vcpu_time_info *src; - src = &per_cpu(hv_clock, 0); - return pvclock_tsc_khz(src); + int cpu; + unsigned long tsc_khz; + + preempt_disable(); + cpu = smp_processor_id(); + src = &hv_clock[cpu].pvti; + tsc_khz = pvclock_tsc_khz(src); + preempt_enable(); + return tsc_khz; } static void kvm_get_preset_lpj(void) @@ -114,6 +127,25 @@ static void kvm_get_preset_lpj(void) preset_lpj = lpj; } +bool kvm_check_and_clear_guest_paused(void) +{ + bool ret = false; + struct pvclock_vcpu_time_info *src; + int cpu = smp_processor_id(); + + if (!hv_clock) + return ret; + + src = &hv_clock[cpu].pvti; + if ((src->flags & PVCLOCK_GUEST_STOPPED) != 0) { + src->flags &= ~PVCLOCK_GUEST_STOPPED; + pvclock_touch_watchdogs(); + ret = true; + } + + return ret; +} + static struct clocksource kvm_clock = { .name = "kvm-clock", .read = kvm_clock_get_cycles, @@ -126,9 +158,14 @@ int kvm_register_clock(char *txt) { int cpu = smp_processor_id(); int low, high, ret; + struct pvclock_vcpu_time_info *src; + + if (!hv_clock) + return 0; - low = (int)__pa(&per_cpu(hv_clock, cpu)) | 1; - high = ((u64)__pa(&per_cpu(hv_clock, cpu)) >> 32); + src = &hv_clock[cpu].pvti; + low = (int)slow_virt_to_phys(src) | 1; + high = ((u64)slow_virt_to_phys(src) >> 32); ret = native_write_msr_safe(msr_kvm_system_time, low, high); printk(KERN_INFO "kvm-clock: cpu %d, msr %x:%x, %s\n", cpu, high, low, txt); @@ -136,16 +173,23 @@ int kvm_register_clock(char *txt) return ret; } +static void kvm_save_sched_clock_state(void) +{ +} + +static void kvm_restore_sched_clock_state(void) +{ + kvm_register_clock("primary cpu clock, resume"); +} + #ifdef CONFIG_X86_LOCAL_APIC -static void __cpuinit kvm_setup_secondary_clock(void) +static void kvm_setup_secondary_clock(void) { /* * Now that the first cpu already had this clocksource initialized, * we shouldn't fail. */ WARN_ON(kvm_register_clock("secondary cpu clock")); - /* ok, done with our trickery, call native */ - setup_secondary_APIC_clock(); } #endif @@ -175,6 +219,11 @@ static void kvm_shutdown(void) void __init kvmclock_init(void) { + unsigned long mem; + int size; + + size = PAGE_ALIGN(sizeof(struct pvclock_vsyscall_time_info)*NR_CPUS); + if (!kvm_para_available()) return; @@ -187,16 +236,27 @@ void __init kvmclock_init(void) printk(KERN_INFO "kvm-clock: Using msrs %x and %x", msr_kvm_system_time, msr_kvm_wall_clock); - if (kvm_register_clock("boot clock")) + mem = memblock_alloc(size, PAGE_SIZE); + if (!mem) return; + hv_clock = __va(mem); + memset(hv_clock, 0, size); + + if (kvm_register_clock("primary cpu clock")) { + hv_clock = NULL; + memblock_free(mem, size); + return; + } pv_time_ops.sched_clock = kvm_clock_read; x86_platform.calibrate_tsc = kvm_get_tsc_khz; x86_platform.get_wallclock = kvm_get_wallclock; x86_platform.set_wallclock = kvm_set_wallclock; #ifdef CONFIG_X86_LOCAL_APIC - x86_cpuinit.setup_percpu_clockev = + x86_cpuinit.early_percpu_clock_init = kvm_setup_secondary_clock; #endif + x86_platform.save_sched_clock_state = kvm_save_sched_clock_state; + x86_platform.restore_sched_clock_state = kvm_restore_sched_clock_state; machine_ops.shutdown = kvm_shutdown; #ifdef CONFIG_KEXEC machine_ops.crash_shutdown = kvm_crash_shutdown; @@ -209,3 +269,40 @@ void __init kvmclock_init(void) if (kvm_para_has_feature(KVM_FEATURE_CLOCKSOURCE_STABLE_BIT)) pvclock_set_flags(PVCLOCK_TSC_STABLE_BIT); } + +int __init kvm_setup_vsyscall_timeinfo(void) +{ +#ifdef CONFIG_X86_64 + int cpu; + int ret; + u8 flags; + struct pvclock_vcpu_time_info *vcpu_time; + unsigned int size; + + if (!hv_clock) + return 0; + + size = PAGE_ALIGN(sizeof(struct pvclock_vsyscall_time_info)*NR_CPUS); + + preempt_disable(); + cpu = smp_processor_id(); + + vcpu_time = &hv_clock[cpu].pvti; + flags = pvclock_read_flags(vcpu_time); + + if (!(flags & PVCLOCK_TSC_STABLE_BIT)) { + preempt_enable(); + return 1; + } + + if ((ret = pvclock_init_vsyscall(hv_clock, size))) { + preempt_enable(); + return ret; + } + + preempt_enable(); + + kvm_clock.archdata.vclock_mode = VCLOCK_PVCLOCK; +#endif + return 0; +} diff --git a/arch/x86/kernel/ldt.c b/arch/x86/kernel/ldt.c index ea697263b37..c37886d759c 100644 --- a/arch/x86/kernel/ldt.c +++ b/arch/x86/kernel/ldt.c @@ -15,7 +15,6 @@ #include <linux/vmalloc.h> #include <linux/uaccess.h> -#include <asm/system.h> #include <asm/ldt.h> #include <asm/desc.h> #include <asm/mmu_context.h> @@ -230,6 +229,11 @@ static int write_ldt(void __user *ptr, unsigned long bytecount, int oldmode) } } + if (!IS_ENABLED(CONFIG_X86_16BIT) && !ldt_info.seg_32bit) { + error = -EINVAL; + goto out_unlock; + } + fill_ldt(&ldt, &ldt_info); if (oldmode) ldt.avl = 0; diff --git a/arch/x86/kernel/machine_kexec_32.c b/arch/x86/kernel/machine_kexec_32.c index a3fa43ba5d3..1667b1de8d5 100644 --- a/arch/x86/kernel/machine_kexec_32.c +++ b/arch/x86/kernel/machine_kexec_32.c @@ -9,7 +9,6 @@ #include <linux/mm.h> #include <linux/kexec.h> #include <linux/delay.h> -#include <linux/init.h> #include <linux/numa.h> #include <linux/ftrace.h> #include <linux/suspend.h> @@ -23,7 +22,6 @@ #include <asm/apic.h> #include <asm/cpufeature.h> #include <asm/desc.h> -#include <asm/system.h> #include <asm/cacheflush.h> #include <asm/debugreg.h> diff --git a/arch/x86/kernel/machine_kexec_64.c b/arch/x86/kernel/machine_kexec_64.c index b3ea9db39db..679cef0791c 100644 --- a/arch/x86/kernel/machine_kexec_64.c +++ b/arch/x86/kernel/machine_kexec_64.c @@ -16,125 +16,12 @@ #include <linux/io.h> #include <linux/suspend.h> +#include <asm/init.h> #include <asm/pgtable.h> #include <asm/tlbflush.h> #include <asm/mmu_context.h> #include <asm/debugreg.h> -static int init_one_level2_page(struct kimage *image, pgd_t *pgd, - unsigned long addr) -{ - pud_t *pud; - pmd_t *pmd; - struct page *page; - int result = -ENOMEM; - - addr &= PMD_MASK; - pgd += pgd_index(addr); - if (!pgd_present(*pgd)) { - page = kimage_alloc_control_pages(image, 0); - if (!page) - goto out; - pud = (pud_t *)page_address(page); - clear_page(pud); - set_pgd(pgd, __pgd(__pa(pud) | _KERNPG_TABLE)); - } - pud = pud_offset(pgd, addr); - if (!pud_present(*pud)) { - page = kimage_alloc_control_pages(image, 0); - if (!page) - goto out; - pmd = (pmd_t *)page_address(page); - clear_page(pmd); - set_pud(pud, __pud(__pa(pmd) | _KERNPG_TABLE)); - } - pmd = pmd_offset(pud, addr); - if (!pmd_present(*pmd)) - set_pmd(pmd, __pmd(addr | __PAGE_KERNEL_LARGE_EXEC)); - result = 0; -out: - return result; -} - -static void init_level2_page(pmd_t *level2p, unsigned long addr) -{ - unsigned long end_addr; - - addr &= PAGE_MASK; - end_addr = addr + PUD_SIZE; - while (addr < end_addr) { - set_pmd(level2p++, __pmd(addr | __PAGE_KERNEL_LARGE_EXEC)); - addr += PMD_SIZE; - } -} - -static int init_level3_page(struct kimage *image, pud_t *level3p, - unsigned long addr, unsigned long last_addr) -{ - unsigned long end_addr; - int result; - - result = 0; - addr &= PAGE_MASK; - end_addr = addr + PGDIR_SIZE; - while ((addr < last_addr) && (addr < end_addr)) { - struct page *page; - pmd_t *level2p; - - page = kimage_alloc_control_pages(image, 0); - if (!page) { - result = -ENOMEM; - goto out; - } - level2p = (pmd_t *)page_address(page); - init_level2_page(level2p, addr); - set_pud(level3p++, __pud(__pa(level2p) | _KERNPG_TABLE)); - addr += PUD_SIZE; - } - /* clear the unused entries */ - while (addr < end_addr) { - pud_clear(level3p++); - addr += PUD_SIZE; - } -out: - return result; -} - - -static int init_level4_page(struct kimage *image, pgd_t *level4p, - unsigned long addr, unsigned long last_addr) -{ - unsigned long end_addr; - int result; - - result = 0; - addr &= PAGE_MASK; - end_addr = addr + (PTRS_PER_PGD * PGDIR_SIZE); - while ((addr < last_addr) && (addr < end_addr)) { - struct page *page; - pud_t *level3p; - - page = kimage_alloc_control_pages(image, 0); - if (!page) { - result = -ENOMEM; - goto out; - } - level3p = (pud_t *)page_address(page); - result = init_level3_page(image, level3p, addr, last_addr); - if (result) - goto out; - set_pgd(level4p++, __pgd(__pa(level3p) | _KERNPG_TABLE)); - addr += PGDIR_SIZE; - } - /* clear the unused entries */ - while (addr < end_addr) { - pgd_clear(level4p++); - addr += PGDIR_SIZE; - } -out: - return result; -} - static void free_transition_pgtable(struct kimage *image) { free_page((unsigned long)image->arch.pud); @@ -184,22 +71,62 @@ err: return result; } +static void *alloc_pgt_page(void *data) +{ + struct kimage *image = (struct kimage *)data; + struct page *page; + void *p = NULL; + + page = kimage_alloc_control_pages(image, 0); + if (page) { + p = page_address(page); + clear_page(p); + } + + return p; +} static int init_pgtable(struct kimage *image, unsigned long start_pgtable) { + struct x86_mapping_info info = { + .alloc_pgt_page = alloc_pgt_page, + .context = image, + .pmd_flag = __PAGE_KERNEL_LARGE_EXEC, + }; + unsigned long mstart, mend; pgd_t *level4p; int result; + int i; + level4p = (pgd_t *)__va(start_pgtable); - result = init_level4_page(image, level4p, 0, max_pfn << PAGE_SHIFT); - if (result) - return result; + clear_page(level4p); + for (i = 0; i < nr_pfn_mapped; i++) { + mstart = pfn_mapped[i].start << PAGE_SHIFT; + mend = pfn_mapped[i].end << PAGE_SHIFT; + + result = kernel_ident_mapping_init(&info, + level4p, mstart, mend); + if (result) + return result; + } + /* - * image->start may be outside 0 ~ max_pfn, for example when - * jump back to original kernel from kexeced kernel + * segments's mem ranges could be outside 0 ~ max_pfn, + * for example when jump back to original kernel from kexeced kernel. + * or first kernel is booted with user mem map, and second kernel + * could be loaded out of that range. */ - result = init_one_level2_page(image, level4p, image->start); - if (result) - return result; + for (i = 0; i < image->nr_segments; i++) { + mstart = image->segment[i].mem; + mend = mstart + image->segment[i].memsz; + + result = kernel_ident_mapping_init(&info, + level4p, mstart, mend); + + if (result) + return result; + } + return init_transition_pgtable(image, level4p); } @@ -352,5 +279,7 @@ void arch_crash_save_vmcoreinfo(void) VMCOREINFO_SYMBOL(node_data); VMCOREINFO_LENGTH(node_data, MAX_NUMNODES); #endif + vmcoreinfo_append_str("KERNELOFFSET=%lx\n", + (unsigned long)&_text - __START_KERNEL); } diff --git a/arch/x86/kernel/mca_32.c b/arch/x86/kernel/mca_32.c deleted file mode 100644 index 177183cbb6a..00000000000 --- a/arch/x86/kernel/mca_32.c +++ /dev/null @@ -1,477 +0,0 @@ -/* - * Written by Martin Kolinek, February 1996 - * - * Changes: - * - * Chris Beauregard July 28th, 1996 - * - Fixed up integrated SCSI detection - * - * Chris Beauregard August 3rd, 1996 - * - Made mca_info local - * - Made integrated registers accessible through standard function calls - * - Added name field - * - More sanity checking - * - * Chris Beauregard August 9th, 1996 - * - Rewrote /proc/mca - * - * Chris Beauregard January 7th, 1997 - * - Added basic NMI-processing - * - Added more information to mca_info structure - * - * David Weinehall October 12th, 1998 - * - Made a lot of cleaning up in the source - * - Added use of save_flags / restore_flags - * - Added the 'driver_loaded' flag in MCA_adapter - * - Added an alternative implemention of ZP Gu's mca_find_unused_adapter - * - * David Weinehall March 24th, 1999 - * - Fixed the output of 'Driver Installed' in /proc/mca/pos - * - Made the Integrated Video & SCSI show up even if they have id 0000 - * - * Alexander Viro November 9th, 1999 - * - Switched to regular procfs methods - * - * Alfred Arnold & David Weinehall August 23rd, 2000 - * - Added support for Planar POS-registers - */ - -#include <linux/module.h> -#include <linux/types.h> -#include <linux/errno.h> -#include <linux/kernel.h> -#include <linux/mca.h> -#include <linux/kprobes.h> -#include <linux/slab.h> -#include <asm/system.h> -#include <asm/io.h> -#include <linux/proc_fs.h> -#include <linux/mman.h> -#include <linux/mm.h> -#include <linux/pagemap.h> -#include <linux/ioport.h> -#include <asm/uaccess.h> -#include <linux/init.h> - -static unsigned char which_scsi; - -int MCA_bus; -EXPORT_SYMBOL(MCA_bus); - -/* - * Motherboard register spinlock. Untested on SMP at the moment, but - * are there any MCA SMP boxes? - * - * Yes - Alan - */ -static DEFINE_SPINLOCK(mca_lock); - -/* Build the status info for the adapter */ - -static void mca_configure_adapter_status(struct mca_device *mca_dev) -{ - mca_dev->status = MCA_ADAPTER_NONE; - - mca_dev->pos_id = mca_dev->pos[0] - + (mca_dev->pos[1] << 8); - - if (!mca_dev->pos_id && mca_dev->slot < MCA_MAX_SLOT_NR) { - - /* - * id = 0x0000 usually indicates hardware failure, - * however, ZP Gu (zpg@castle.net> reports that his 9556 - * has 0x0000 as id and everything still works. There - * also seem to be an adapter with id = 0x0000; the - * NCR Parallel Bus Memory Card. Until this is confirmed, - * however, this code will stay. - */ - - mca_dev->status = MCA_ADAPTER_ERROR; - - return; - } else if (mca_dev->pos_id != 0xffff) { - - /* - * 0xffff usually indicates that there's no adapter, - * however, some integrated adapters may have 0xffff as - * their id and still be valid. Examples are on-board - * VGA of the 55sx, the integrated SCSI of the 56 & 57, - * and possibly also the 95 ULTIMEDIA. - */ - - mca_dev->status = MCA_ADAPTER_NORMAL; - } - - if ((mca_dev->pos_id == 0xffff || - mca_dev->pos_id == 0x0000) && mca_dev->slot >= MCA_MAX_SLOT_NR) { - int j; - - for (j = 2; j < 8; j++) { - if (mca_dev->pos[j] != 0xff) { - mca_dev->status = MCA_ADAPTER_NORMAL; - break; - } - } - } - - if (!(mca_dev->pos[2] & MCA_ENABLED)) { - - /* enabled bit is in POS 2 */ - - mca_dev->status = MCA_ADAPTER_DISABLED; - } -} /* mca_configure_adapter_status */ - -/*--------------------------------------------------------------------*/ - -static struct resource mca_standard_resources[] = { - { .start = 0x60, .end = 0x60, .name = "system control port B (MCA)" }, - { .start = 0x90, .end = 0x90, .name = "arbitration (MCA)" }, - { .start = 0x91, .end = 0x91, .name = "card Select Feedback (MCA)" }, - { .start = 0x92, .end = 0x92, .name = "system Control port A (MCA)" }, - { .start = 0x94, .end = 0x94, .name = "system board setup (MCA)" }, - { .start = 0x96, .end = 0x97, .name = "POS (MCA)" }, - { .start = 0x100, .end = 0x107, .name = "POS (MCA)" } -}; - -#define MCA_STANDARD_RESOURCES ARRAY_SIZE(mca_standard_resources) - -/* - * mca_read_and_store_pos - read the POS registers into a memory buffer - * @pos: a char pointer to 8 bytes, contains the POS register value on - * successful return - * - * Returns 1 if a card actually exists (i.e. the pos isn't - * all 0xff) or 0 otherwise - */ -static int mca_read_and_store_pos(unsigned char *pos) -{ - int j; - int found = 0; - - for (j = 0; j < 8; j++) { - pos[j] = inb_p(MCA_POS_REG(j)); - if (pos[j] != 0xff) { - /* 0xff all across means no device. 0x00 means - * something's broken, but a device is - * probably there. However, if you get 0x00 - * from a motherboard register it won't matter - * what we find. For the record, on the - * 57SLC, the integrated SCSI adapter has - * 0xffff for the adapter ID, but nonzero for - * other registers. */ - - found = 1; - } - } - return found; -} - -static unsigned char mca_pc_read_pos(struct mca_device *mca_dev, int reg) -{ - unsigned char byte; - unsigned long flags; - - if (reg < 0 || reg >= 8) - return 0; - - spin_lock_irqsave(&mca_lock, flags); - if (mca_dev->pos_register) { - /* Disable adapter setup, enable motherboard setup */ - - outb_p(0, MCA_ADAPTER_SETUP_REG); - outb_p(mca_dev->pos_register, MCA_MOTHERBOARD_SETUP_REG); - - byte = inb_p(MCA_POS_REG(reg)); - outb_p(0xff, MCA_MOTHERBOARD_SETUP_REG); - } else { - - /* Make sure motherboard setup is off */ - - outb_p(0xff, MCA_MOTHERBOARD_SETUP_REG); - - /* Read the appropriate register */ - - outb_p(0x8|(mca_dev->slot & 0xf), MCA_ADAPTER_SETUP_REG); - byte = inb_p(MCA_POS_REG(reg)); - outb_p(0, MCA_ADAPTER_SETUP_REG); - } - spin_unlock_irqrestore(&mca_lock, flags); - - mca_dev->pos[reg] = byte; - - return byte; -} - -static void mca_pc_write_pos(struct mca_device *mca_dev, int reg, - unsigned char byte) -{ - unsigned long flags; - - if (reg < 0 || reg >= 8) - return; - - spin_lock_irqsave(&mca_lock, flags); - - /* Make sure motherboard setup is off */ - - outb_p(0xff, MCA_MOTHERBOARD_SETUP_REG); - - /* Read in the appropriate register */ - - outb_p(0x8|(mca_dev->slot&0xf), MCA_ADAPTER_SETUP_REG); - outb_p(byte, MCA_POS_REG(reg)); - outb_p(0, MCA_ADAPTER_SETUP_REG); - - spin_unlock_irqrestore(&mca_lock, flags); - - /* Update the global register list, while we have the byte */ - - mca_dev->pos[reg] = byte; - -} - -/* for the primary MCA bus, we have identity transforms */ -static int mca_dummy_transform_irq(struct mca_device *mca_dev, int irq) -{ - return irq; -} - -static int mca_dummy_transform_ioport(struct mca_device *mca_dev, int port) -{ - return port; -} - -static void *mca_dummy_transform_memory(struct mca_device *mca_dev, void *mem) -{ - return mem; -} - - -static int __init mca_init(void) -{ - unsigned int i, j; - struct mca_device *mca_dev; - unsigned char pos[8]; - short mca_builtin_scsi_ports[] = {0xf7, 0xfd, 0x00}; - struct mca_bus *bus; - - /* - * WARNING: Be careful when making changes here. Putting an adapter - * and the motherboard simultaneously into setup mode may result in - * damage to chips (according to The Indispensable PC Hardware Book - * by Hans-Peter Messmer). Also, we disable system interrupts (so - * that we are not disturbed in the middle of this). - */ - - /* Make sure the MCA bus is present */ - - if (mca_system_init()) { - printk(KERN_ERR "MCA bus system initialisation failed\n"); - return -ENODEV; - } - - if (!MCA_bus) - return -ENODEV; - - printk(KERN_INFO "Micro Channel bus detected.\n"); - - /* All MCA systems have at least a primary bus */ - bus = mca_attach_bus(MCA_PRIMARY_BUS); - if (!bus) - goto out_nomem; - bus->default_dma_mask = 0xffffffffLL; - bus->f.mca_write_pos = mca_pc_write_pos; - bus->f.mca_read_pos = mca_pc_read_pos; - bus->f.mca_transform_irq = mca_dummy_transform_irq; - bus->f.mca_transform_ioport = mca_dummy_transform_ioport; - bus->f.mca_transform_memory = mca_dummy_transform_memory; - - /* get the motherboard device */ - mca_dev = kzalloc(sizeof(struct mca_device), GFP_KERNEL); - if (unlikely(!mca_dev)) - goto out_nomem; - - /* - * We do not expect many MCA interrupts during initialization, - * but let us be safe: - */ - spin_lock_irq(&mca_lock); - - /* Make sure adapter setup is off */ - - outb_p(0, MCA_ADAPTER_SETUP_REG); - - /* Read motherboard POS registers */ - - mca_dev->pos_register = 0x7f; - outb_p(mca_dev->pos_register, MCA_MOTHERBOARD_SETUP_REG); - mca_dev->name[0] = 0; - mca_read_and_store_pos(mca_dev->pos); - mca_configure_adapter_status(mca_dev); - /* fake POS and slot for a motherboard */ - mca_dev->pos_id = MCA_MOTHERBOARD_POS; - mca_dev->slot = MCA_MOTHERBOARD; - mca_register_device(MCA_PRIMARY_BUS, mca_dev); - - mca_dev = kzalloc(sizeof(struct mca_device), GFP_ATOMIC); - if (unlikely(!mca_dev)) - goto out_unlock_nomem; - - /* Put motherboard into video setup mode, read integrated video - * POS registers, and turn motherboard setup off. - */ - - mca_dev->pos_register = 0xdf; - outb_p(mca_dev->pos_register, MCA_MOTHERBOARD_SETUP_REG); - mca_dev->name[0] = 0; - mca_read_and_store_pos(mca_dev->pos); - mca_configure_adapter_status(mca_dev); - /* fake POS and slot for the integrated video */ - mca_dev->pos_id = MCA_INTEGVIDEO_POS; - mca_dev->slot = MCA_INTEGVIDEO; - mca_register_device(MCA_PRIMARY_BUS, mca_dev); - - /* - * Put motherboard into scsi setup mode, read integrated scsi - * POS registers, and turn motherboard setup off. - * - * It seems there are two possible SCSI registers. Martin says that - * for the 56,57, 0xf7 is the one, but fails on the 76. - * Alfredo (apena@vnet.ibm.com) says - * 0xfd works on his machine. We'll try both of them. I figure it's - * a good bet that only one could be valid at a time. This could - * screw up though if one is used for something else on the other - * machine. - */ - - for (i = 0; (which_scsi = mca_builtin_scsi_ports[i]) != 0; i++) { - outb_p(which_scsi, MCA_MOTHERBOARD_SETUP_REG); - if (mca_read_and_store_pos(pos)) - break; - } - if (which_scsi) { - /* found a scsi card */ - mca_dev = kzalloc(sizeof(struct mca_device), GFP_ATOMIC); - if (unlikely(!mca_dev)) - goto out_unlock_nomem; - - for (j = 0; j < 8; j++) - mca_dev->pos[j] = pos[j]; - - mca_configure_adapter_status(mca_dev); - /* fake POS and slot for integrated SCSI controller */ - mca_dev->pos_id = MCA_INTEGSCSI_POS; - mca_dev->slot = MCA_INTEGSCSI; - mca_dev->pos_register = which_scsi; - mca_register_device(MCA_PRIMARY_BUS, mca_dev); - } - - /* Turn off motherboard setup */ - - outb_p(0xff, MCA_MOTHERBOARD_SETUP_REG); - - /* - * Now loop over MCA slots: put each adapter into setup mode, and - * read its POS registers. Then put adapter setup off. - */ - - for (i = 0; i < MCA_MAX_SLOT_NR; i++) { - outb_p(0x8|(i&0xf), MCA_ADAPTER_SETUP_REG); - if (!mca_read_and_store_pos(pos)) - continue; - - mca_dev = kzalloc(sizeof(struct mca_device), GFP_ATOMIC); - if (unlikely(!mca_dev)) - goto out_unlock_nomem; - - for (j = 0; j < 8; j++) - mca_dev->pos[j] = pos[j]; - - mca_dev->driver_loaded = 0; - mca_dev->slot = i; - mca_dev->pos_register = 0; - mca_configure_adapter_status(mca_dev); - mca_register_device(MCA_PRIMARY_BUS, mca_dev); - } - outb_p(0, MCA_ADAPTER_SETUP_REG); - - /* Enable interrupts and return memory start */ - spin_unlock_irq(&mca_lock); - - for (i = 0; i < MCA_STANDARD_RESOURCES; i++) - request_resource(&ioport_resource, mca_standard_resources + i); - - mca_do_proc_init(); - - return 0; - - out_unlock_nomem: - spin_unlock_irq(&mca_lock); - out_nomem: - printk(KERN_EMERG "Failed memory allocation in MCA setup!\n"); - return -ENOMEM; -} - -subsys_initcall(mca_init); - -/*--------------------------------------------------------------------*/ - -static __kprobes void -mca_handle_nmi_device(struct mca_device *mca_dev, int check_flag) -{ - int slot = mca_dev->slot; - - if (slot == MCA_INTEGSCSI) { - printk(KERN_CRIT "NMI: caused by MCA integrated SCSI adapter (%s)\n", - mca_dev->name); - } else if (slot == MCA_INTEGVIDEO) { - printk(KERN_CRIT "NMI: caused by MCA integrated video adapter (%s)\n", - mca_dev->name); - } else if (slot == MCA_MOTHERBOARD) { - printk(KERN_CRIT "NMI: caused by motherboard (%s)\n", - mca_dev->name); - } - - /* More info available in POS 6 and 7? */ - - if (check_flag) { - unsigned char pos6, pos7; - - pos6 = mca_device_read_pos(mca_dev, 6); - pos7 = mca_device_read_pos(mca_dev, 7); - - printk(KERN_CRIT "NMI: POS 6 = 0x%x, POS 7 = 0x%x\n", pos6, pos7); - } - -} /* mca_handle_nmi_slot */ - -/*--------------------------------------------------------------------*/ - -static int __kprobes mca_handle_nmi_callback(struct device *dev, void *data) -{ - struct mca_device *mca_dev = to_mca_device(dev); - unsigned char pos5; - - pos5 = mca_device_read_pos(mca_dev, 5); - - if (!(pos5 & 0x80)) { - /* - * Bit 7 of POS 5 is reset when this adapter has a hardware - * error. Bit 7 it reset if there's error information - * available in POS 6 and 7. - */ - mca_handle_nmi_device(mca_dev, !(pos5 & 0x40)); - return 1; - } - return 0; -} - -void __kprobes mca_handle_nmi(void) -{ - /* - * First try - scan the various adapters and see if a specific - * adapter was responsible for the error. - */ - bus_for_each_dev(&mca_bus_type, NULL, NULL, mca_handle_nmi_callback); -} diff --git a/arch/x86/kernel/mcount_64.S b/arch/x86/kernel/mcount_64.S new file mode 100644 index 00000000000..c050a015316 --- /dev/null +++ b/arch/x86/kernel/mcount_64.S @@ -0,0 +1,217 @@ +/* + * linux/arch/x86_64/mcount_64.S + * + * Copyright (C) 2014 Steven Rostedt, Red Hat Inc + */ + +#include <linux/linkage.h> +#include <asm/ptrace.h> +#include <asm/ftrace.h> + + + .code64 + .section .entry.text, "ax" + + +#ifdef CONFIG_FUNCTION_TRACER + +#ifdef CC_USING_FENTRY +# define function_hook __fentry__ +#else +# define function_hook mcount +#endif + +#ifdef CONFIG_DYNAMIC_FTRACE + +ENTRY(function_hook) + retq +END(function_hook) + +/* skip is set if stack has been adjusted */ +.macro ftrace_caller_setup skip=0 + MCOUNT_SAVE_FRAME \skip + + /* Load the ftrace_ops into the 3rd parameter */ + movq function_trace_op(%rip), %rdx + + /* Load ip into the first parameter */ + movq RIP(%rsp), %rdi + subq $MCOUNT_INSN_SIZE, %rdi + /* Load the parent_ip into the second parameter */ +#ifdef CC_USING_FENTRY + movq SS+16(%rsp), %rsi +#else + movq 8(%rbp), %rsi +#endif +.endm + +ENTRY(ftrace_caller) + /* Check if tracing was disabled (quick check) */ + cmpl $0, function_trace_stop + jne ftrace_stub + + ftrace_caller_setup + /* regs go into 4th parameter (but make it NULL) */ + movq $0, %rcx + +GLOBAL(ftrace_call) + call ftrace_stub + + MCOUNT_RESTORE_FRAME +ftrace_return: + +#ifdef CONFIG_FUNCTION_GRAPH_TRACER +GLOBAL(ftrace_graph_call) + jmp ftrace_stub +#endif + +GLOBAL(ftrace_stub) + retq +END(ftrace_caller) + +ENTRY(ftrace_regs_caller) + /* Save the current flags before compare (in SS location)*/ + pushfq + + /* Check if tracing was disabled (quick check) */ + cmpl $0, function_trace_stop + jne ftrace_restore_flags + + /* skip=8 to skip flags saved in SS */ + ftrace_caller_setup 8 + + /* Save the rest of pt_regs */ + movq %r15, R15(%rsp) + movq %r14, R14(%rsp) + movq %r13, R13(%rsp) + movq %r12, R12(%rsp) + movq %r11, R11(%rsp) + movq %r10, R10(%rsp) + movq %rbp, RBP(%rsp) + movq %rbx, RBX(%rsp) + /* Copy saved flags */ + movq SS(%rsp), %rcx + movq %rcx, EFLAGS(%rsp) + /* Kernel segments */ + movq $__KERNEL_DS, %rcx + movq %rcx, SS(%rsp) + movq $__KERNEL_CS, %rcx + movq %rcx, CS(%rsp) + /* Stack - skipping return address */ + leaq SS+16(%rsp), %rcx + movq %rcx, RSP(%rsp) + + /* regs go into 4th parameter */ + leaq (%rsp), %rcx + +GLOBAL(ftrace_regs_call) + call ftrace_stub + + /* Copy flags back to SS, to restore them */ + movq EFLAGS(%rsp), %rax + movq %rax, SS(%rsp) + + /* Handlers can change the RIP */ + movq RIP(%rsp), %rax + movq %rax, SS+8(%rsp) + + /* restore the rest of pt_regs */ + movq R15(%rsp), %r15 + movq R14(%rsp), %r14 + movq R13(%rsp), %r13 + movq R12(%rsp), %r12 + movq R10(%rsp), %r10 + movq RBP(%rsp), %rbp + movq RBX(%rsp), %rbx + + /* skip=8 to skip flags saved in SS */ + MCOUNT_RESTORE_FRAME 8 + + /* Restore flags */ + popfq + + jmp ftrace_return +ftrace_restore_flags: + popfq + jmp ftrace_stub + +END(ftrace_regs_caller) + + +#else /* ! CONFIG_DYNAMIC_FTRACE */ + +ENTRY(function_hook) + cmpl $0, function_trace_stop + jne ftrace_stub + + cmpq $ftrace_stub, ftrace_trace_function + jnz trace + +#ifdef CONFIG_FUNCTION_GRAPH_TRACER + cmpq $ftrace_stub, ftrace_graph_return + jnz ftrace_graph_caller + + cmpq $ftrace_graph_entry_stub, ftrace_graph_entry + jnz ftrace_graph_caller +#endif + +GLOBAL(ftrace_stub) + retq + +trace: + MCOUNT_SAVE_FRAME + + movq RIP(%rsp), %rdi +#ifdef CC_USING_FENTRY + movq SS+16(%rsp), %rsi +#else + movq 8(%rbp), %rsi +#endif + subq $MCOUNT_INSN_SIZE, %rdi + + call *ftrace_trace_function + + MCOUNT_RESTORE_FRAME + + jmp ftrace_stub +END(function_hook) +#endif /* CONFIG_DYNAMIC_FTRACE */ +#endif /* CONFIG_FUNCTION_TRACER */ + +#ifdef CONFIG_FUNCTION_GRAPH_TRACER +ENTRY(ftrace_graph_caller) + MCOUNT_SAVE_FRAME + +#ifdef CC_USING_FENTRY + leaq SS+16(%rsp), %rdi + movq $0, %rdx /* No framepointers needed */ +#else + leaq 8(%rbp), %rdi + movq (%rbp), %rdx +#endif + movq RIP(%rsp), %rsi + subq $MCOUNT_INSN_SIZE, %rsi + + call prepare_ftrace_return + + MCOUNT_RESTORE_FRAME + + retq +END(ftrace_graph_caller) + +GLOBAL(return_to_handler) + subq $24, %rsp + + /* Save the return values */ + movq %rax, (%rsp) + movq %rdx, 8(%rsp) + movq %rbp, %rdi + + call ftrace_return_to_handler + + movq %rax, %rdi + movq 8(%rsp), %rdx + movq (%rsp), %rax + addq $24, %rsp + jmp *%rdi +#endif diff --git a/arch/x86/kernel/microcode_amd.c b/arch/x86/kernel/microcode_amd.c deleted file mode 100644 index ac0417be913..00000000000 --- a/arch/x86/kernel/microcode_amd.c +++ /dev/null @@ -1,394 +0,0 @@ -/* - * AMD CPU Microcode Update Driver for Linux - * Copyright (C) 2008-2011 Advanced Micro Devices Inc. - * - * Author: Peter Oruba <peter.oruba@amd.com> - * - * Based on work by: - * Tigran Aivazian <tigran@aivazian.fsnet.co.uk> - * - * Maintainers: - * Andreas Herrmann <andreas.herrmann3@amd.com> - * Borislav Petkov <borislav.petkov@amd.com> - * - * This driver allows to upgrade microcode on F10h AMD - * CPUs and later. - * - * Licensed under the terms of the GNU General Public - * License version 2. See file COPYING for details. - */ - -#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt - -#include <linux/firmware.h> -#include <linux/pci_ids.h> -#include <linux/uaccess.h> -#include <linux/vmalloc.h> -#include <linux/kernel.h> -#include <linux/module.h> -#include <linux/pci.h> - -#include <asm/microcode.h> -#include <asm/processor.h> -#include <asm/msr.h> - -MODULE_DESCRIPTION("AMD Microcode Update Driver"); -MODULE_AUTHOR("Peter Oruba"); -MODULE_LICENSE("GPL v2"); - -#define UCODE_MAGIC 0x00414d44 -#define UCODE_EQUIV_CPU_TABLE_TYPE 0x00000000 -#define UCODE_UCODE_TYPE 0x00000001 - -struct equiv_cpu_entry { - u32 installed_cpu; - u32 fixed_errata_mask; - u32 fixed_errata_compare; - u16 equiv_cpu; - u16 res; -} __attribute__((packed)); - -struct microcode_header_amd { - u32 data_code; - u32 patch_id; - u16 mc_patch_data_id; - u8 mc_patch_data_len; - u8 init_flag; - u32 mc_patch_data_checksum; - u32 nb_dev_id; - u32 sb_dev_id; - u16 processor_rev_id; - u8 nb_rev_id; - u8 sb_rev_id; - u8 bios_api_rev; - u8 reserved1[3]; - u32 match_reg[8]; -} __attribute__((packed)); - -struct microcode_amd { - struct microcode_header_amd hdr; - unsigned int mpb[0]; -}; - -#define SECTION_HDR_SIZE 8 -#define CONTAINER_HDR_SZ 12 - -static struct equiv_cpu_entry *equiv_cpu_table; - -/* page-sized ucode patch buffer */ -void *patch; - -static int collect_cpu_info_amd(int cpu, struct cpu_signature *csig) -{ - struct cpuinfo_x86 *c = &cpu_data(cpu); - - if (c->x86_vendor != X86_VENDOR_AMD || c->x86 < 0x10) { - pr_warning("CPU%d: family %d not supported\n", cpu, c->x86); - return -1; - } - - csig->rev = c->microcode; - pr_info("CPU%d: patch_level=0x%08x\n", cpu, csig->rev); - - return 0; -} - -static unsigned int verify_ucode_size(int cpu, u32 patch_size, - unsigned int size) -{ - struct cpuinfo_x86 *c = &cpu_data(cpu); - u32 max_size; - -#define F1XH_MPB_MAX_SIZE 2048 -#define F14H_MPB_MAX_SIZE 1824 -#define F15H_MPB_MAX_SIZE 4096 - - switch (c->x86) { - case 0x14: - max_size = F14H_MPB_MAX_SIZE; - break; - case 0x15: - max_size = F15H_MPB_MAX_SIZE; - break; - default: - max_size = F1XH_MPB_MAX_SIZE; - break; - } - - if (patch_size > min_t(u32, size, max_size)) { - pr_err("patch size mismatch\n"); - return 0; - } - - return patch_size; -} - -static u16 find_equiv_id(void) -{ - unsigned int current_cpu_id, i = 0; - - BUG_ON(equiv_cpu_table == NULL); - - current_cpu_id = cpuid_eax(0x00000001); - - while (equiv_cpu_table[i].installed_cpu != 0) { - if (current_cpu_id == equiv_cpu_table[i].installed_cpu) - return equiv_cpu_table[i].equiv_cpu; - - i++; - } - return 0; -} - -/* - * we signal a good patch is found by returning its size > 0 - */ -static int get_matching_microcode(int cpu, const u8 *ucode_ptr, - unsigned int leftover_size, int rev, - unsigned int *current_size) -{ - struct microcode_header_amd *mc_hdr; - unsigned int actual_size; - u16 equiv_cpu_id; - - /* size of the current patch we're staring at */ - *current_size = *(u32 *)(ucode_ptr + 4) + SECTION_HDR_SIZE; - - equiv_cpu_id = find_equiv_id(); - if (!equiv_cpu_id) - return 0; - - /* - * let's look at the patch header itself now - */ - mc_hdr = (struct microcode_header_amd *)(ucode_ptr + SECTION_HDR_SIZE); - - if (mc_hdr->processor_rev_id != equiv_cpu_id) - return 0; - - /* ucode might be chipset specific -- currently we don't support this */ - if (mc_hdr->nb_dev_id || mc_hdr->sb_dev_id) { - pr_err("CPU%d: chipset specific code not yet supported\n", - cpu); - return 0; - } - - if (mc_hdr->patch_id <= rev) - return 0; - - /* - * now that the header looks sane, verify its size - */ - actual_size = verify_ucode_size(cpu, *current_size, leftover_size); - if (!actual_size) - return 0; - - /* clear the patch buffer */ - memset(patch, 0, PAGE_SIZE); - - /* all looks ok, get the binary patch */ - get_ucode_data(patch, ucode_ptr + SECTION_HDR_SIZE, actual_size); - - return actual_size; -} - -static int apply_microcode_amd(int cpu) -{ - u32 rev, dummy; - int cpu_num = raw_smp_processor_id(); - struct ucode_cpu_info *uci = ucode_cpu_info + cpu_num; - struct microcode_amd *mc_amd = uci->mc; - struct cpuinfo_x86 *c = &cpu_data(cpu); - - /* We should bind the task to the CPU */ - BUG_ON(cpu_num != cpu); - - if (mc_amd == NULL) - return 0; - - wrmsrl(MSR_AMD64_PATCH_LOADER, (u64)(long)&mc_amd->hdr.data_code); - /* get patch id after patching */ - rdmsr(MSR_AMD64_PATCH_LEVEL, rev, dummy); - - /* check current patch id and patch's id for match */ - if (rev != mc_amd->hdr.patch_id) { - pr_err("CPU%d: update failed for patch_level=0x%08x\n", - cpu, mc_amd->hdr.patch_id); - return -1; - } - - pr_info("CPU%d: new patch_level=0x%08x\n", cpu, rev); - uci->cpu_sig.rev = rev; - c->microcode = rev; - - return 0; -} - -static int install_equiv_cpu_table(const u8 *buf) -{ - unsigned int *ibuf = (unsigned int *)buf; - unsigned int type = ibuf[1]; - unsigned int size = ibuf[2]; - - if (type != UCODE_EQUIV_CPU_TABLE_TYPE || !size) { - pr_err("empty section/" - "invalid type field in container file section header\n"); - return -EINVAL; - } - - equiv_cpu_table = vmalloc(size); - if (!equiv_cpu_table) { - pr_err("failed to allocate equivalent CPU table\n"); - return -ENOMEM; - } - - get_ucode_data(equiv_cpu_table, buf + CONTAINER_HDR_SZ, size); - - /* add header length */ - return size + CONTAINER_HDR_SZ; -} - -static void free_equiv_cpu_table(void) -{ - vfree(equiv_cpu_table); - equiv_cpu_table = NULL; -} - -static enum ucode_state -generic_load_microcode(int cpu, const u8 *data, size_t size) -{ - struct ucode_cpu_info *uci = ucode_cpu_info + cpu; - struct microcode_header_amd *mc_hdr = NULL; - unsigned int mc_size, leftover, current_size = 0; - int offset; - const u8 *ucode_ptr = data; - void *new_mc = NULL; - unsigned int new_rev = uci->cpu_sig.rev; - enum ucode_state state = UCODE_ERROR; - - offset = install_equiv_cpu_table(ucode_ptr); - if (offset < 0) { - pr_err("failed to create equivalent cpu table\n"); - goto out; - } - ucode_ptr += offset; - leftover = size - offset; - - if (*(u32 *)ucode_ptr != UCODE_UCODE_TYPE) { - pr_err("invalid type field in container file section header\n"); - goto free_table; - } - - while (leftover) { - mc_size = get_matching_microcode(cpu, ucode_ptr, leftover, - new_rev, ¤t_size); - if (mc_size) { - mc_hdr = patch; - new_mc = patch; - new_rev = mc_hdr->patch_id; - goto out_ok; - } - - ucode_ptr += current_size; - leftover -= current_size; - } - - if (!new_mc) { - state = UCODE_NFOUND; - goto free_table; - } - -out_ok: - uci->mc = new_mc; - state = UCODE_OK; - pr_debug("CPU%d update ucode (0x%08x -> 0x%08x)\n", - cpu, uci->cpu_sig.rev, new_rev); - -free_table: - free_equiv_cpu_table(); - -out: - return state; -} - -/* - * AMD microcode firmware naming convention, up to family 15h they are in - * the legacy file: - * - * amd-ucode/microcode_amd.bin - * - * This legacy file is always smaller than 2K in size. - * - * Starting at family 15h they are in family specific firmware files: - * - * amd-ucode/microcode_amd_fam15h.bin - * amd-ucode/microcode_amd_fam16h.bin - * ... - * - * These might be larger than 2K. - */ -static enum ucode_state request_microcode_amd(int cpu, struct device *device) -{ - char fw_name[36] = "amd-ucode/microcode_amd.bin"; - const struct firmware *fw; - enum ucode_state ret = UCODE_NFOUND; - struct cpuinfo_x86 *c = &cpu_data(cpu); - - if (c->x86 >= 0x15) - snprintf(fw_name, sizeof(fw_name), "amd-ucode/microcode_amd_fam%.2xh.bin", c->x86); - - if (request_firmware(&fw, (const char *)fw_name, device)) { - pr_err("failed to load file %s\n", fw_name); - goto out; - } - - ret = UCODE_ERROR; - if (*(u32 *)fw->data != UCODE_MAGIC) { - pr_err("invalid magic value (0x%08x)\n", *(u32 *)fw->data); - goto fw_release; - } - - ret = generic_load_microcode(cpu, fw->data, fw->size); - -fw_release: - release_firmware(fw); - -out: - return ret; -} - -static enum ucode_state -request_microcode_user(int cpu, const void __user *buf, size_t size) -{ - pr_info("AMD microcode update via /dev/cpu/microcode not supported\n"); - return UCODE_ERROR; -} - -static void microcode_fini_cpu_amd(int cpu) -{ - struct ucode_cpu_info *uci = ucode_cpu_info + cpu; - - uci->mc = NULL; -} - -static struct microcode_ops microcode_amd_ops = { - .request_microcode_user = request_microcode_user, - .request_microcode_fw = request_microcode_amd, - .collect_cpu_info = collect_cpu_info_amd, - .apply_microcode = apply_microcode_amd, - .microcode_fini_cpu = microcode_fini_cpu_amd, -}; - -struct microcode_ops * __init init_amd_microcode(void) -{ - patch = (void *)get_zeroed_page(GFP_KERNEL); - if (!patch) - return NULL; - - return µcode_amd_ops; -} - -void __exit exit_amd_microcode(void) -{ - free_page((unsigned long)patch); -} diff --git a/arch/x86/kernel/mmconf-fam10h_64.c b/arch/x86/kernel/mmconf-fam10h_64.c index ac861b8348e..f4c886d9165 100644 --- a/arch/x86/kernel/mmconf-fam10h_64.c +++ b/arch/x86/kernel/mmconf-fam10h_64.c @@ -24,14 +24,14 @@ struct pci_hostbridge_probe { u32 device; }; -static u64 __cpuinitdata fam10h_pci_mmconf_base; +static u64 fam10h_pci_mmconf_base; -static struct pci_hostbridge_probe pci_probes[] __cpuinitdata = { +static struct pci_hostbridge_probe pci_probes[] = { { 0, 0x18, PCI_VENDOR_ID_AMD, 0x1200 }, { 0xff, 0, PCI_VENDOR_ID_AMD, 0x1200 }, }; -static int __cpuinit cmp_range(const void *x1, const void *x2) +static int cmp_range(const void *x1, const void *x2) { const struct range *r1 = x1; const struct range *r2 = x2; @@ -49,7 +49,7 @@ static int __cpuinit cmp_range(const void *x1, const void *x2) /* need to avoid (0xfd<<32), (0xfe<<32), and (0xff<<32), ht used space */ #define FAM10H_PCI_MMCONF_BASE (0xfcULL<<32) #define BASE_VALID(b) ((b) + MMCONF_SIZE <= (0xfdULL<<32) || (b) >= (1ULL<<40)) -static void __cpuinit get_fam10h_pci_mmconf_base(void) +static void get_fam10h_pci_mmconf_base(void) { int i; unsigned bus; @@ -166,7 +166,7 @@ out: fam10h_pci_mmconf_base = base; } -void __cpuinit fam10h_check_enable_mmcfg(void) +void fam10h_check_enable_mmcfg(void) { u64 val; u32 address; @@ -230,7 +230,7 @@ static const struct dmi_system_id __initconst mmconf_dmi_table[] = { {} }; -/* Called from a __cpuinit function, but only on the BSP. */ +/* Called from a non __init function, but only on the BSP. */ void __ref check_enable_amd_mmconf_dmi(void) { dmi_check_system(mmconf_dmi_table); diff --git a/arch/x86/kernel/module.c b/arch/x86/kernel/module.c index 925179f871d..e69f9882bf9 100644 --- a/arch/x86/kernel/module.c +++ b/arch/x86/kernel/module.c @@ -15,6 +15,9 @@ along with this program; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/moduleloader.h> #include <linux/elf.h> #include <linux/vmalloc.h> @@ -25,24 +28,68 @@ #include <linux/mm.h> #include <linux/gfp.h> #include <linux/jump_label.h> +#include <linux/random.h> -#include <asm/system.h> #include <asm/page.h> #include <asm/pgtable.h> #if 0 -#define DEBUGP printk +#define DEBUGP(fmt, ...) \ + printk(KERN_DEBUG fmt, ##__VA_ARGS__) #else -#define DEBUGP(fmt...) +#define DEBUGP(fmt, ...) \ +do { \ + if (0) \ + printk(KERN_DEBUG fmt, ##__VA_ARGS__); \ +} while (0) +#endif + +#ifdef CONFIG_RANDOMIZE_BASE +static unsigned long module_load_offset; +static int randomize_modules = 1; + +/* Mutex protects the module_load_offset. */ +static DEFINE_MUTEX(module_kaslr_mutex); + +static int __init parse_nokaslr(char *p) +{ + randomize_modules = 0; + return 0; +} +early_param("nokaslr", parse_nokaslr); + +static unsigned long int get_module_load_offset(void) +{ + if (randomize_modules) { + mutex_lock(&module_kaslr_mutex); + /* + * Calculate the module_load_offset the first time this + * code is called. Once calculated it stays the same until + * reboot. + */ + if (module_load_offset == 0) + module_load_offset = + (get_random_int() % 1024 + 1) * PAGE_SIZE; + mutex_unlock(&module_kaslr_mutex); + } + return module_load_offset; +} +#else +static unsigned long int get_module_load_offset(void) +{ + return 0; +} #endif void *module_alloc(unsigned long size) { if (PAGE_ALIGN(size) > MODULES_LEN) return NULL; - return __vmalloc_node_range(size, 1, MODULES_VADDR, MODULES_END, - GFP_KERNEL | __GFP_HIGHMEM, PAGE_KERNEL_EXEC, - -1, __builtin_return_address(0)); + return __vmalloc_node_range(size, 1, + MODULES_VADDR + get_module_load_offset(), + MODULES_END, GFP_KERNEL | __GFP_HIGHMEM, + PAGE_KERNEL_EXEC, NUMA_NO_NODE, + __builtin_return_address(0)); } #ifdef CONFIG_X86_32 @@ -57,8 +104,8 @@ int apply_relocate(Elf32_Shdr *sechdrs, Elf32_Sym *sym; uint32_t *location; - DEBUGP("Applying relocate section %u to %u\n", relsec, - sechdrs[relsec].sh_info); + DEBUGP("Applying relocate section %u to %u\n", + relsec, sechdrs[relsec].sh_info); for (i = 0; i < sechdrs[relsec].sh_size / sizeof(*rel); i++) { /* This is where to make the change */ location = (void *)sechdrs[sechdrs[relsec].sh_info].sh_addr @@ -74,11 +121,11 @@ int apply_relocate(Elf32_Shdr *sechdrs, *location += sym->st_value; break; case R_386_PC32: - /* Add the value, subtract its postition */ + /* Add the value, subtract its position */ *location += sym->st_value - (uint32_t)location; break; default: - printk(KERN_ERR "module %s: Unknown relocation: %u\n", + pr_err("%s: Unknown relocation: %u\n", me->name, ELF32_R_TYPE(rel[i].r_info)); return -ENOEXEC; } @@ -98,8 +145,8 @@ int apply_relocate_add(Elf64_Shdr *sechdrs, void *loc; u64 val; - DEBUGP("Applying relocate section %u to %u\n", relsec, - sechdrs[relsec].sh_info); + DEBUGP("Applying relocate section %u to %u\n", + relsec, sechdrs[relsec].sh_info); for (i = 0; i < sechdrs[relsec].sh_size / sizeof(*rel); i++) { /* This is where to make the change */ loc = (void *)sechdrs[sechdrs[relsec].sh_info].sh_addr @@ -111,8 +158,8 @@ int apply_relocate_add(Elf64_Shdr *sechdrs, + ELF64_R_SYM(rel[i].r_info); DEBUGP("type %d st_value %Lx r_addend %Lx loc %Lx\n", - (int)ELF64_R_TYPE(rel[i].r_info), - sym->st_value, rel[i].r_addend, (u64)loc); + (int)ELF64_R_TYPE(rel[i].r_info), + sym->st_value, rel[i].r_addend, (u64)loc); val = sym->st_value + rel[i].r_addend; @@ -141,7 +188,7 @@ int apply_relocate_add(Elf64_Shdr *sechdrs, #endif break; default: - printk(KERN_ERR "module %s: Unknown rela relocation: %llu\n", + pr_err("%s: Unknown rela relocation: %llu\n", me->name, ELF64_R_TYPE(rel[i].r_info)); return -ENOEXEC; } @@ -149,9 +196,9 @@ int apply_relocate_add(Elf64_Shdr *sechdrs, return 0; overflow: - printk(KERN_ERR "overflow in relocation type %d val %Lx\n", + pr_err("overflow in relocation type %d val %Lx\n", (int)ELF64_R_TYPE(rel[i].r_info), val); - printk(KERN_ERR "`%s' likely not compiled with -mcmodel=kernel\n", + pr_err("`%s' likely not compiled with -mcmodel=kernel\n", me->name); return -ENOEXEC; } diff --git a/arch/x86/kernel/mpparse.c b/arch/x86/kernel/mpparse.c index ca470e4c92d..d2b56489d70 100644 --- a/arch/x86/kernel/mpparse.c +++ b/arch/x86/kernel/mpparse.c @@ -27,7 +27,6 @@ #include <asm/proto.h> #include <asm/bios_ebda.h> #include <asm/e820.h> -#include <asm/trampoline.h> #include <asm/setup.h> #include <asm/smp.h> @@ -97,7 +96,7 @@ static void __init MP_bus_info(struct mpc_bus *m) set_bit(m->busid, mp_bus_not_pci); if (strncmp(str, BUSTYPE_ISA, sizeof(BUSTYPE_ISA) - 1) == 0) { -#if defined(CONFIG_EISA) || defined(CONFIG_MCA) +#ifdef CONFIG_EISA mp_bus_id_to_type[m->busid] = MP_BUS_ISA; #endif } else if (strncmp(str, BUSTYPE_PCI, sizeof(BUSTYPE_PCI) - 1) == 0) { @@ -105,12 +104,10 @@ static void __init MP_bus_info(struct mpc_bus *m) x86_init.mpparse.mpc_oem_pci_bus(m); clear_bit(m->busid, mp_bus_not_pci); -#if defined(CONFIG_EISA) || defined(CONFIG_MCA) +#ifdef CONFIG_EISA mp_bus_id_to_type[m->busid] = MP_BUS_PCI; } else if (strncmp(str, BUSTYPE_EISA, sizeof(BUSTYPE_EISA) - 1) == 0) { mp_bus_id_to_type[m->busid] = MP_BUS_EISA; - } else if (strncmp(str, BUSTYPE_MCA, sizeof(BUSTYPE_MCA) - 1) == 0) { - mp_bus_id_to_type[m->busid] = MP_BUS_MCA; #endif } else printk(KERN_WARNING "Unknown bustype %s - ignoring\n", str); @@ -368,9 +365,6 @@ static void __init construct_ioapic_table(int mpc_default_type) case 3: memcpy(bus.bustype, "EISA ", 6); break; - case 4: - case 7: - memcpy(bus.bustype, "MCA ", 6); } MP_bus_info(&bus); if (mpc_default_type > 4) { @@ -573,8 +567,8 @@ static int __init smp_scan_config(unsigned long base, unsigned long length) struct mpf_intel *mpf; unsigned long mem; - apic_printk(APIC_VERBOSE, "Scan SMP from %p for %ld bytes.\n", - bp, length); + apic_printk(APIC_VERBOSE, "Scan for SMP in [mem %#010lx-%#010lx]\n", + base, base + length - 1); BUILD_BUG_ON(sizeof(*mpf) != 16); while (length > 0) { @@ -589,8 +583,10 @@ static int __init smp_scan_config(unsigned long base, unsigned long length) #endif mpf_found = mpf; - printk(KERN_INFO "found SMP MP-table at [%p] %llx\n", - mpf, (u64)virt_to_phys(mpf)); + printk(KERN_INFO "found SMP MP-table at [mem %#010llx-%#010llx] mapped at [%p]\n", + (unsigned long long) virt_to_phys(mpf), + (unsigned long long) virt_to_phys(mpf) + + sizeof(*mpf) - 1, mpf); mem = virt_to_phys(mpf); memblock_reserve(mem, sizeof(*mpf)); @@ -623,7 +619,7 @@ void __init default_find_smp_config(void) return; /* * If it is an SMP machine we should know now, unless the - * configuration is in an EISA/MCA bus machine with an + * configuration is in an EISA bus machine with an * extended bios data area. * * there is a real-mode segmented pointer pointing to the diff --git a/arch/x86/kernel/msr.c b/arch/x86/kernel/msr.c index 96356762a51..c9603ac80de 100644 --- a/arch/x86/kernel/msr.c +++ b/arch/x86/kernel/msr.c @@ -40,14 +40,13 @@ #include <asm/processor.h> #include <asm/msr.h> -#include <asm/system.h> static struct class *msr_class; static loff_t msr_seek(struct file *file, loff_t offset, int orig) { loff_t ret; - struct inode *inode = file->f_mapping->host; + struct inode *inode = file_inode(file); mutex_lock(&inode->i_mutex); switch (orig) { @@ -72,7 +71,7 @@ static ssize_t msr_read(struct file *file, char __user *buf, u32 __user *tmp = (u32 __user *) buf; u32 data[2]; u32 reg = *ppos; - int cpu = iminor(file->f_path.dentry->d_inode); + int cpu = iminor(file_inode(file)); int err = 0; ssize_t bytes = 0; @@ -100,7 +99,7 @@ static ssize_t msr_write(struct file *file, const char __user *buf, const u32 __user *tmp = (const u32 __user *)buf; u32 data[2]; u32 reg = *ppos; - int cpu = iminor(file->f_path.dentry->d_inode); + int cpu = iminor(file_inode(file)); int err = 0; ssize_t bytes = 0; @@ -126,7 +125,7 @@ static long msr_ioctl(struct file *file, unsigned int ioc, unsigned long arg) { u32 __user *uregs = (u32 __user *)arg; u32 regs[8]; - int cpu = iminor(file->f_path.dentry->d_inode); + int cpu = iminor(file_inode(file)); int err; switch (ioc) { @@ -172,10 +171,12 @@ static long msr_ioctl(struct file *file, unsigned int ioc, unsigned long arg) static int msr_open(struct inode *inode, struct file *file) { - unsigned int cpu; + unsigned int cpu = iminor(file_inode(file)); struct cpuinfo_x86 *c; - cpu = iminor(file->f_path.dentry->d_inode); + if (!capable(CAP_SYS_RAWIO)) + return -EPERM; + if (cpu >= nr_cpu_ids || !cpu_online(cpu)) return -ENXIO; /* No such CPU */ @@ -199,7 +200,7 @@ static const struct file_operations msr_fops = { .compat_ioctl = msr_ioctl, }; -static int __cpuinit msr_device_create(int cpu) +static int msr_device_create(int cpu) { struct device *dev; @@ -213,8 +214,8 @@ static void msr_device_destroy(int cpu) device_destroy(msr_class, MKDEV(MSR_MAJOR, cpu)); } -static int __cpuinit msr_class_cpu_callback(struct notifier_block *nfb, - unsigned long action, void *hcpu) +static int msr_class_cpu_callback(struct notifier_block *nfb, + unsigned long action, void *hcpu) { unsigned int cpu = (unsigned long)hcpu; int err = 0; @@ -258,12 +259,15 @@ static int __init msr_init(void) goto out_chrdev; } msr_class->devnode = msr_devnode; + + cpu_notifier_register_begin(); for_each_online_cpu(i) { err = msr_device_create(i); if (err != 0) goto out_class; } - register_hotcpu_notifier(&msr_class_cpu_notifier); + __register_hotcpu_notifier(&msr_class_cpu_notifier); + cpu_notifier_register_done(); err = 0; goto out; @@ -272,6 +276,7 @@ out_class: i = 0; for_each_online_cpu(i) msr_device_destroy(i); + cpu_notifier_register_done(); class_destroy(msr_class); out_chrdev: __unregister_chrdev(MSR_MAJOR, 0, NR_CPUS, "cpu/msr"); @@ -282,11 +287,14 @@ out: static void __exit msr_exit(void) { int cpu = 0; + + cpu_notifier_register_begin(); for_each_online_cpu(cpu) msr_device_destroy(cpu); class_destroy(msr_class); __unregister_chrdev(MSR_MAJOR, 0, NR_CPUS, "cpu/msr"); - unregister_hotcpu_notifier(&msr_class_cpu_notifier); + __unregister_hotcpu_notifier(&msr_class_cpu_notifier); + cpu_notifier_register_done(); } module_init(msr_init); diff --git a/arch/x86/kernel/nmi.c b/arch/x86/kernel/nmi.c index 47acaf31916..c3e985d1751 100644 --- a/arch/x86/kernel/nmi.c +++ b/arch/x86/kernel/nmi.c @@ -14,13 +14,12 @@ #include <linux/kprobes.h> #include <linux/kdebug.h> #include <linux/nmi.h> +#include <linux/debugfs.h> #include <linux/delay.h> #include <linux/hardirq.h> #include <linux/slab.h> #include <linux/export.h> -#include <linux/mca.h> - #if defined(CONFIG_EDAC) #include <linux/edac.h> #endif @@ -31,13 +30,8 @@ #include <asm/nmi.h> #include <asm/x86_init.h> -#define NMI_MAX_NAMELEN 16 -struct nmiaction { - struct list_head list; - nmi_handler_t handler; - unsigned int flags; - char *name; -}; +#define CREATE_TRACE_POINTS +#include <trace/events/nmi.h> struct nmi_desc { spinlock_t lock; @@ -54,6 +48,14 @@ static struct nmi_desc nmi_desc[NMI_MAX] = .lock = __SPIN_LOCK_UNLOCKED(&nmi_desc[1].lock), .head = LIST_HEAD_INIT(nmi_desc[1].head), }, + { + .lock = __SPIN_LOCK_UNLOCKED(&nmi_desc[2].lock), + .head = LIST_HEAD_INIT(nmi_desc[2].head), + }, + { + .lock = __SPIN_LOCK_UNLOCKED(&nmi_desc[3].lock), + .head = LIST_HEAD_INIT(nmi_desc[3].head), + }, }; @@ -84,7 +86,31 @@ __setup("unknown_nmi_panic", setup_unknown_nmi_panic); #define nmi_to_desc(type) (&nmi_desc[type]) -static int notrace __kprobes nmi_handle(unsigned int type, struct pt_regs *regs, bool b2b) +static u64 nmi_longest_ns = 1 * NSEC_PER_MSEC; + +static int __init nmi_warning_debugfs(void) +{ + debugfs_create_u64("nmi_longest_ns", 0644, + arch_debugfs_dir, &nmi_longest_ns); + return 0; +} +fs_initcall(nmi_warning_debugfs); + +static void nmi_max_handler(struct irq_work *w) +{ + struct nmiaction *a = container_of(w, struct nmiaction, irq_work); + int remainder_ns, decimal_msecs; + u64 whole_msecs = ACCESS_ONCE(a->max_duration); + + remainder_ns = do_div(whole_msecs, (1000 * 1000)); + decimal_msecs = remainder_ns / 1000; + + printk_ratelimited(KERN_INFO + "INFO: NMI handler (%ps) took too long to run: %lld.%03d msecs\n", + a->handler, whole_msecs, decimal_msecs); +} + +static int nmi_handle(unsigned int type, struct pt_regs *regs, bool b2b) { struct nmi_desc *desc = nmi_to_desc(type); struct nmiaction *a; @@ -98,20 +124,40 @@ static int notrace __kprobes nmi_handle(unsigned int type, struct pt_regs *regs, * can be latched at any given time. Walk the whole list * to handle those situations. */ - list_for_each_entry_rcu(a, &desc->head, list) - handled += a->handler(type, regs); + list_for_each_entry_rcu(a, &desc->head, list) { + int thishandled; + u64 delta; + + delta = sched_clock(); + thishandled = a->handler(type, regs); + handled += thishandled; + delta = sched_clock() - delta; + trace_nmi_handler(a->handler, (int)delta, thishandled); + + if (delta < nmi_longest_ns || delta < a->max_duration) + continue; + + a->max_duration = delta; + irq_work_queue(&a->irq_work); + } rcu_read_unlock(); /* return total number of NMI events handled */ return handled; } +NOKPROBE_SYMBOL(nmi_handle); -static int __setup_nmi(unsigned int type, struct nmiaction *action) +int __register_nmi_handler(unsigned int type, struct nmiaction *action) { struct nmi_desc *desc = nmi_to_desc(type); unsigned long flags; + if (!action->handler) + return -EINVAL; + + init_irq_work(&action->irq_work, nmi_max_handler); + spin_lock_irqsave(&desc->lock, flags); /* @@ -120,6 +166,8 @@ static int __setup_nmi(unsigned int type, struct nmiaction *action) * to manage expectations */ WARN_ON_ONCE(type == NMI_UNKNOWN && !list_empty(&desc->head)); + WARN_ON_ONCE(type == NMI_SERR && !list_empty(&desc->head)); + WARN_ON_ONCE(type == NMI_IO_CHECK && !list_empty(&desc->head)); /* * some handlers need to be executed first otherwise a fake @@ -133,8 +181,9 @@ static int __setup_nmi(unsigned int type, struct nmiaction *action) spin_unlock_irqrestore(&desc->lock, flags); return 0; } +EXPORT_SYMBOL(__register_nmi_handler); -static struct nmiaction *__free_nmi(unsigned int type, const char *name) +void unregister_nmi_handler(unsigned int type, const char *name) { struct nmi_desc *desc = nmi_to_desc(type); struct nmiaction *n; @@ -157,61 +206,16 @@ static struct nmiaction *__free_nmi(unsigned int type, const char *name) spin_unlock_irqrestore(&desc->lock, flags); synchronize_rcu(); - return (n); } - -int register_nmi_handler(unsigned int type, nmi_handler_t handler, - unsigned long nmiflags, const char *devname) -{ - struct nmiaction *action; - int retval = -ENOMEM; - - if (!handler) - return -EINVAL; - - action = kzalloc(sizeof(struct nmiaction), GFP_KERNEL); - if (!action) - goto fail_action; - - action->handler = handler; - action->flags = nmiflags; - action->name = kstrndup(devname, NMI_MAX_NAMELEN, GFP_KERNEL); - if (!action->name) - goto fail_action_name; - - retval = __setup_nmi(type, action); - - if (retval) - goto fail_setup_nmi; - - return retval; - -fail_setup_nmi: - kfree(action->name); -fail_action_name: - kfree(action); -fail_action: - - return retval; -} -EXPORT_SYMBOL_GPL(register_nmi_handler); - -void unregister_nmi_handler(unsigned int type, const char *name) -{ - struct nmiaction *a; - - a = __free_nmi(type, name); - if (a) { - kfree(a->name); - kfree(a); - } -} - EXPORT_SYMBOL_GPL(unregister_nmi_handler); -static notrace __kprobes void +static void pci_serr_error(unsigned char reason, struct pt_regs *regs) { + /* check to see if anyone registered against these types of errors */ + if (nmi_handle(NMI_SERR, regs, false)) + return; + pr_emerg("NMI: PCI system error (SERR) for reason %02x on CPU %d.\n", reason, smp_processor_id()); @@ -235,16 +239,21 @@ pci_serr_error(unsigned char reason, struct pt_regs *regs) reason = (reason & NMI_REASON_CLEAR_MASK) | NMI_REASON_CLEAR_SERR; outb(reason, NMI_REASON_PORT); } +NOKPROBE_SYMBOL(pci_serr_error); -static notrace __kprobes void +static void io_check_error(unsigned char reason, struct pt_regs *regs) { unsigned long i; + /* check to see if anyone registered against these types of errors */ + if (nmi_handle(NMI_IO_CHECK, regs, false)) + return; + pr_emerg( "NMI: IOCK error (debug interrupt?) for reason %02x on CPU %d.\n", reason, smp_processor_id()); - show_registers(regs); + show_regs(regs); if (panic_on_io_nmi) panic("NMI IOCK error: Not continuing"); @@ -262,8 +271,9 @@ io_check_error(unsigned char reason, struct pt_regs *regs) reason &= ~NMI_REASON_CLEAR_IOCHK; outb(reason, NMI_REASON_PORT); } +NOKPROBE_SYMBOL(io_check_error); -static notrace __kprobes void +static void unknown_nmi_error(unsigned char reason, struct pt_regs *regs) { int handled; @@ -282,16 +292,6 @@ unknown_nmi_error(unsigned char reason, struct pt_regs *regs) __this_cpu_add(nmi_stats.unknown, 1); -#ifdef CONFIG_MCA - /* - * Might actually be able to figure out what the guilty party - * is: - */ - if (MCA_bus) { - mca_handle_nmi(); - return; - } -#endif pr_emerg("Uhhuh. NMI received for unknown reason %02x on CPU %d.\n", reason, smp_processor_id()); @@ -301,11 +301,12 @@ unknown_nmi_error(unsigned char reason, struct pt_regs *regs) pr_emerg("Dazed and confused, but trying to continue\n"); } +NOKPROBE_SYMBOL(unknown_nmi_error); static DEFINE_PER_CPU(bool, swallow_nmi); static DEFINE_PER_CPU(unsigned long, last_nmi_rip); -static notrace __kprobes void default_do_nmi(struct pt_regs *regs) +static void default_do_nmi(struct pt_regs *regs) { unsigned char reason = 0; int handled; @@ -404,6 +405,7 @@ static notrace __kprobes void default_do_nmi(struct pt_regs *regs) else unknown_nmi_error(reason, regs); } +NOKPROBE_SYMBOL(default_do_nmi); /* * NMIs can hit breakpoints which will cause it to lose its @@ -412,8 +414,9 @@ static notrace __kprobes void default_do_nmi(struct pt_regs *regs) #ifdef CONFIG_X86_32 /* * For i386, NMIs use the same stack as the kernel, and we can - * add a workaround to the iret problem in C. Simply have 3 states - * the NMI can be in. + * add a workaround to the iret problem in C (preventing nested + * NMIs if an NMI takes a trap). Simply have 3 states the NMI + * can be in: * * 1) not running * 2) executing @@ -430,32 +433,50 @@ static notrace __kprobes void default_do_nmi(struct pt_regs *regs) * If an NMI hits a breakpoint that executes an iret, another * NMI can preempt it. We do not want to allow this new NMI * to run, but we want to execute it when the first one finishes. - * We set the state to "latched", and the first NMI will perform - * an cmpxchg on the state, and if it doesn't successfully - * reset the state to "not running" it will restart the next - * NMI. + * We set the state to "latched", and the exit of the first NMI will + * perform a dec_return, if the result is zero (NOT_RUNNING), then + * it will simply exit the NMI handler. If not, the dec_return + * would have set the state to NMI_EXECUTING (what we want it to + * be when we are running). In this case, we simply jump back + * to rerun the NMI handler again, and restart the 'latched' NMI. + * + * No trap (breakpoint or page fault) should be hit before nmi_restart, + * thus there is no race between the first check of state for NOT_RUNNING + * and setting it to NMI_EXECUTING. The HW will prevent nested NMIs + * at this point. + * + * In case the NMI takes a page fault, we need to save off the CR2 + * because the NMI could have preempted another page fault and corrupt + * the CR2 that is about to be read. As nested NMIs must be restarted + * and they can not take breakpoints or page faults, the update of the + * CR2 must be done before converting the nmi state back to NOT_RUNNING. + * Otherwise, there would be a race of another nested NMI coming in + * after setting state to NOT_RUNNING but before updating the nmi_cr2. */ enum nmi_states { - NMI_NOT_RUNNING, + NMI_NOT_RUNNING = 0, NMI_EXECUTING, NMI_LATCHED, }; static DEFINE_PER_CPU(enum nmi_states, nmi_state); +static DEFINE_PER_CPU(unsigned long, nmi_cr2); #define nmi_nesting_preprocess(regs) \ do { \ - if (__get_cpu_var(nmi_state) != NMI_NOT_RUNNING) { \ - __get_cpu_var(nmi_state) = NMI_LATCHED; \ + if (this_cpu_read(nmi_state) != NMI_NOT_RUNNING) { \ + this_cpu_write(nmi_state, NMI_LATCHED); \ return; \ } \ - nmi_restart: \ - __get_cpu_var(nmi_state) = NMI_EXECUTING; \ - } while (0) + this_cpu_write(nmi_state, NMI_EXECUTING); \ + this_cpu_write(nmi_cr2, read_cr2()); \ + } while (0); \ + nmi_restart: #define nmi_nesting_postprocess() \ do { \ - if (cmpxchg(&__get_cpu_var(nmi_state), \ - NMI_EXECUTING, NMI_NOT_RUNNING) != NMI_EXECUTING) \ + if (unlikely(this_cpu_read(nmi_cr2) != read_cr2())) \ + write_cr2(this_cpu_read(nmi_cr2)); \ + if (this_cpu_dec_return(nmi_state)) \ goto nmi_restart; \ } while (0) #else /* x86_64 */ @@ -491,18 +512,20 @@ static inline void nmi_nesting_preprocess(struct pt_regs *regs) */ if (unlikely(is_debug_stack(regs->sp))) { debug_stack_set_zero(); - __get_cpu_var(update_debug_stack) = 1; + this_cpu_write(update_debug_stack, 1); } } static inline void nmi_nesting_postprocess(void) { - if (unlikely(__get_cpu_var(update_debug_stack))) + if (unlikely(this_cpu_read(update_debug_stack))) { debug_stack_reset(); + this_cpu_write(update_debug_stack, 0); + } } #endif -dotraplinkage notrace __kprobes void +dotraplinkage notrace void do_nmi(struct pt_regs *regs, long error_code) { nmi_nesting_preprocess(regs); @@ -519,6 +542,7 @@ do_nmi(struct pt_regs *regs, long error_code) /* On i386, may loop back to preprocess */ nmi_nesting_postprocess(); } +NOKPROBE_SYMBOL(do_nmi); void stop_nmi(void) { @@ -535,3 +559,4 @@ void local_touch_nmi(void) { __this_cpu_write(last_nmi_rip, 0); } +EXPORT_SYMBOL_GPL(local_touch_nmi); diff --git a/arch/x86/kernel/nmi_selftest.c b/arch/x86/kernel/nmi_selftest.c index 0d01a8ea4e1..6d9582ec032 100644 --- a/arch/x86/kernel/nmi_selftest.c +++ b/arch/x86/kernel/nmi_selftest.c @@ -12,6 +12,8 @@ #include <linux/smp.h> #include <linux/cpumask.h> #include <linux/delay.h> +#include <linux/init.h> +#include <linux/percpu.h> #include <asm/apic.h> #include <asm/nmi.h> @@ -20,35 +22,36 @@ #define FAILURE 1 #define TIMEOUT 2 -static int nmi_fail; +static int __initdata nmi_fail; /* check to see if NMI IPIs work on this machine */ -static DECLARE_BITMAP(nmi_ipi_mask, NR_CPUS) __read_mostly; +static DECLARE_BITMAP(nmi_ipi_mask, NR_CPUS) __initdata; -static int testcase_total; -static int testcase_successes; -static int expected_testcase_failures; -static int unexpected_testcase_failures; -static int unexpected_testcase_unknowns; +static int __initdata testcase_total; +static int __initdata testcase_successes; +static int __initdata expected_testcase_failures; +static int __initdata unexpected_testcase_failures; +static int __initdata unexpected_testcase_unknowns; -static int nmi_unk_cb(unsigned int val, struct pt_regs *regs) +static int __init nmi_unk_cb(unsigned int val, struct pt_regs *regs) { unexpected_testcase_unknowns++; return NMI_HANDLED; } -static void init_nmi_testsuite(void) +static void __init init_nmi_testsuite(void) { /* trap all the unknown NMIs we may generate */ - register_nmi_handler(NMI_UNKNOWN, nmi_unk_cb, 0, "nmi_selftest_unk"); + register_nmi_handler(NMI_UNKNOWN, nmi_unk_cb, 0, "nmi_selftest_unk", + __initdata); } -static void cleanup_nmi_testsuite(void) +static void __init cleanup_nmi_testsuite(void) { unregister_nmi_handler(NMI_UNKNOWN, "nmi_selftest_unk"); } -static int test_nmi_ipi_callback(unsigned int val, struct pt_regs *regs) +static int __init test_nmi_ipi_callback(unsigned int val, struct pt_regs *regs) { int cpu = raw_smp_processor_id(); @@ -58,12 +61,12 @@ static int test_nmi_ipi_callback(unsigned int val, struct pt_regs *regs) return NMI_DONE; } -static void test_nmi_ipi(struct cpumask *mask) +static void __init test_nmi_ipi(struct cpumask *mask) { unsigned long timeout; if (register_nmi_handler(NMI_LOCAL, test_nmi_ipi_callback, - NMI_FLAG_FIRST, "nmi_selftest")) { + NMI_FLAG_FIRST, "nmi_selftest", __initdata)) { nmi_fail = FAILURE; return; } @@ -86,7 +89,7 @@ static void test_nmi_ipi(struct cpumask *mask) return; } -static void remote_ipi(void) +static void __init remote_ipi(void) { cpumask_copy(to_cpumask(nmi_ipi_mask), cpu_online_mask); cpumask_clear_cpu(smp_processor_id(), to_cpumask(nmi_ipi_mask)); @@ -94,19 +97,19 @@ static void remote_ipi(void) test_nmi_ipi(to_cpumask(nmi_ipi_mask)); } -static void local_ipi(void) +static void __init local_ipi(void) { cpumask_clear(to_cpumask(nmi_ipi_mask)); cpumask_set_cpu(smp_processor_id(), to_cpumask(nmi_ipi_mask)); test_nmi_ipi(to_cpumask(nmi_ipi_mask)); } -static void reset_nmi(void) +static void __init reset_nmi(void) { nmi_fail = 0; } -static void dotest(void (*testcase_fn)(void), int expected) +static void __init dotest(void (*testcase_fn)(void), int expected) { testcase_fn(); /* @@ -116,27 +119,27 @@ static void dotest(void (*testcase_fn)(void), int expected) unexpected_testcase_failures++; if (nmi_fail == FAILURE) - printk("FAILED |"); + printk(KERN_CONT "FAILED |"); else if (nmi_fail == TIMEOUT) - printk("TIMEOUT|"); + printk(KERN_CONT "TIMEOUT|"); else - printk("ERROR |"); + printk(KERN_CONT "ERROR |"); dump_stack(); } else { testcase_successes++; - printk(" ok |"); + printk(KERN_CONT " ok |"); } testcase_total++; reset_nmi(); } -static inline void print_testname(const char *testname) +static inline void __init print_testname(const char *testname) { printk("%12s:", testname); } -void nmi_selftest(void) +void __init nmi_selftest(void) { init_nmi_testsuite(); @@ -149,10 +152,10 @@ void nmi_selftest(void) print_testname("remote IPI"); dotest(remote_ipi, SUCCESS); - printk("\n"); + printk(KERN_CONT "\n"); print_testname("local IPI"); dotest(local_ipi, SUCCESS); - printk("\n"); + printk(KERN_CONT "\n"); cleanup_nmi_testsuite(); diff --git a/arch/x86/kernel/paravirt-spinlocks.c b/arch/x86/kernel/paravirt-spinlocks.c index 676b8c77a97..bbb6c731634 100644 --- a/arch/x86/kernel/paravirt-spinlocks.c +++ b/arch/x86/kernel/paravirt-spinlocks.c @@ -4,25 +4,17 @@ */ #include <linux/spinlock.h> #include <linux/module.h> +#include <linux/jump_label.h> #include <asm/paravirt.h> -static inline void -default_spin_lock_flags(arch_spinlock_t *lock, unsigned long flags) -{ - arch_spin_lock(lock); -} - struct pv_lock_ops pv_lock_ops = { #ifdef CONFIG_SMP - .spin_is_locked = __ticket_spin_is_locked, - .spin_is_contended = __ticket_spin_is_contended, - - .spin_lock = __ticket_spin_lock, - .spin_lock_flags = default_spin_lock_flags, - .spin_trylock = __ticket_spin_trylock, - .spin_unlock = __ticket_spin_unlock, + .lock_spinning = __PV_IS_CALLEE_SAVE(paravirt_nop), + .unlock_kick = paravirt_nop, #endif }; EXPORT_SYMBOL(pv_lock_ops); +struct static_key paravirt_ticketlocks_enabled = STATIC_KEY_INIT_FALSE; +EXPORT_SYMBOL(paravirt_ticketlocks_enabled); diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c index d90272e6bc4..548d25f00c9 100644 --- a/arch/x86/kernel/paravirt.c +++ b/arch/x86/kernel/paravirt.c @@ -23,9 +23,11 @@ #include <linux/efi.h> #include <linux/bcd.h> #include <linux/highmem.h> +#include <linux/kprobes.h> #include <asm/bug.h> #include <asm/paravirt.h> +#include <asm/debugreg.h> #include <asm/desc.h> #include <asm/setup.h> #include <asm/pgtable.h> @@ -37,6 +39,7 @@ #include <asm/apic.h> #include <asm/tlbflush.h> #include <asm/timer.h> +#include <asm/special_insns.h> /* nop stub */ void _paravirt_nop(void) @@ -60,11 +63,6 @@ void __init default_banner(void) pv_info.name); } -/* Simple instruction patching code. */ -#define DEF_NATIVE(ops, name, code) \ - extern const char start_##ops##_##name[], end_##ops##_##name[]; \ - asm("start_" #ops "_" #name ": " code "; end_" #ops "_" #name ":") - /* Undefined instruction for dealing with missing ops pointers. */ static const unsigned char ud2a[] = { 0x0f, 0x0b }; @@ -202,8 +200,8 @@ static void native_flush_tlb_single(unsigned long addr) __native_flush_tlb_single(addr); } -struct jump_label_key paravirt_steal_enabled; -struct jump_label_key paravirt_steal_rq_enabled; +struct static_key paravirt_steal_enabled; +struct static_key paravirt_steal_rq_enabled; static u64 native_steal_clock(int cpu) { @@ -239,16 +237,16 @@ static DEFINE_PER_CPU(enum paravirt_lazy_mode, paravirt_lazy_mode) = PARAVIRT_LA static inline void enter_lazy(enum paravirt_lazy_mode mode) { - BUG_ON(percpu_read(paravirt_lazy_mode) != PARAVIRT_LAZY_NONE); + BUG_ON(this_cpu_read(paravirt_lazy_mode) != PARAVIRT_LAZY_NONE); - percpu_write(paravirt_lazy_mode, mode); + this_cpu_write(paravirt_lazy_mode, mode); } static void leave_lazy(enum paravirt_lazy_mode mode) { - BUG_ON(percpu_read(paravirt_lazy_mode) != mode); + BUG_ON(this_cpu_read(paravirt_lazy_mode) != mode); - percpu_write(paravirt_lazy_mode, PARAVIRT_LAZY_NONE); + this_cpu_write(paravirt_lazy_mode, PARAVIRT_LAZY_NONE); } void paravirt_enter_lazy_mmu(void) @@ -261,11 +259,23 @@ void paravirt_leave_lazy_mmu(void) leave_lazy(PARAVIRT_LAZY_MMU); } +void paravirt_flush_lazy_mmu(void) +{ + preempt_disable(); + + if (paravirt_get_lazy_mode() == PARAVIRT_LAZY_MMU) { + arch_leave_lazy_mmu_mode(); + arch_enter_lazy_mmu_mode(); + } + + preempt_enable(); +} + void paravirt_start_context_switch(struct task_struct *prev) { BUG_ON(preemptible()); - if (percpu_read(paravirt_lazy_mode) == PARAVIRT_LAZY_MMU) { + if (this_cpu_read(paravirt_lazy_mode) == PARAVIRT_LAZY_MMU) { arch_leave_lazy_mmu_mode(); set_ti_thread_flag(task_thread_info(prev), TIF_LAZY_MMU_UPDATES); } @@ -287,19 +297,7 @@ enum paravirt_lazy_mode paravirt_get_lazy_mode(void) if (in_interrupt()) return PARAVIRT_LAZY_NONE; - return percpu_read(paravirt_lazy_mode); -} - -void arch_flush_lazy_mmu_mode(void) -{ - preempt_disable(); - - if (paravirt_get_lazy_mode() == PARAVIRT_LAZY_MMU) { - arch_leave_lazy_mmu_mode(); - arch_enter_lazy_mmu_mode(); - } - - preempt_enable(); + return this_cpu_read(paravirt_lazy_mode); } struct pv_info pv_info = { @@ -322,7 +320,7 @@ struct pv_time_ops pv_time_ops = { .steal_clock = native_steal_clock, }; -struct pv_irq_ops pv_irq_ops = { +__visible struct pv_irq_ops pv_irq_ops = { .save_fl = __PV_IS_CALLEE_SAVE(native_save_fl), .restore_fl = __PV_IS_CALLEE_SAVE(native_restore_fl), .irq_disable = __PV_IS_CALLEE_SAVE(native_irq_disable), @@ -334,7 +332,7 @@ struct pv_irq_ops pv_irq_ops = { #endif }; -struct pv_cpu_ops pv_cpu_ops = { +__visible struct pv_cpu_ops pv_cpu_ops = { .cpuid = native_cpuid, .get_debugreg = native_get_debugreg, .set_debugreg = native_set_debugreg, @@ -350,9 +348,7 @@ struct pv_cpu_ops pv_cpu_ops = { #endif .wbinvd = native_wbinvd, .read_msr = native_read_msr_safe, - .rdmsr_regs = native_rdmsr_safe_regs, .write_msr = native_write_msr_safe, - .wrmsr_regs = native_wrmsr_safe_regs, .read_tsc = native_read_tsc, .read_pmc = native_read_pmc, .read_tscp = native_read_tscp, @@ -360,7 +356,6 @@ struct pv_cpu_ops pv_cpu_ops = { .set_ldt = native_set_ldt, .load_gdt = native_load_gdt, .load_idt = native_load_idt, - .store_gdt = native_store_gdt, .store_idt = native_store_idt, .store_tr = native_store_tr, .load_tls = native_load_tls, @@ -395,6 +390,11 @@ struct pv_cpu_ops pv_cpu_ops = { .end_context_switch = paravirt_nop, }; +/* At this point, native_get/set_debugreg has real function entries */ +NOKPROBE_SYMBOL(native_get_debugreg); +NOKPROBE_SYMBOL(native_set_debugreg); +NOKPROBE_SYMBOL(native_load_idt); + struct pv_apic_ops pv_apic_ops = { #ifdef CONFIG_X86_LOCAL_APIC .startup_ipi_hook = paravirt_nop, @@ -475,6 +475,7 @@ struct pv_mmu_ops pv_mmu_ops = { .lazy_mode = { .enter = paravirt_nop, .leave = paravirt_nop, + .flush = paravirt_nop, }, .set_fixmap = native_set_fixmap, diff --git a/arch/x86/kernel/paravirt_patch_64.c b/arch/x86/kernel/paravirt_patch_64.c index 3f08f34f93e..a1da6737ba5 100644 --- a/arch/x86/kernel/paravirt_patch_64.c +++ b/arch/x86/kernel/paravirt_patch_64.c @@ -6,7 +6,6 @@ DEF_NATIVE(pv_irq_ops, irq_disable, "cli"); DEF_NATIVE(pv_irq_ops, irq_enable, "sti"); DEF_NATIVE(pv_irq_ops, restore_fl, "pushq %rdi; popfq"); DEF_NATIVE(pv_irq_ops, save_fl, "pushfq; popq %rax"); -DEF_NATIVE(pv_cpu_ops, iret, "iretq"); DEF_NATIVE(pv_mmu_ops, read_cr2, "movq %cr2, %rax"); DEF_NATIVE(pv_mmu_ops, read_cr3, "movq %cr3, %rax"); DEF_NATIVE(pv_mmu_ops, write_cr3, "movq %rdi, %cr3"); @@ -50,7 +49,6 @@ unsigned native_patch(u8 type, u16 clobbers, void *ibuf, PATCH_SITE(pv_irq_ops, save_fl); PATCH_SITE(pv_irq_ops, irq_enable); PATCH_SITE(pv_irq_ops, irq_disable); - PATCH_SITE(pv_cpu_ops, iret); PATCH_SITE(pv_cpu_ops, irq_enable_sysexit); PATCH_SITE(pv_cpu_ops, usergs_sysret32); PATCH_SITE(pv_cpu_ops, usergs_sysret64); diff --git a/arch/x86/kernel/pci-calgary_64.c b/arch/x86/kernel/pci-calgary_64.c index 726494b5834..0497f719977 100644 --- a/arch/x86/kernel/pci-calgary_64.c +++ b/arch/x86/kernel/pci-calgary_64.c @@ -22,6 +22,8 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ +#define pr_fmt(fmt) "Calgary: " fmt + #include <linux/kernel.h> #include <linux/init.h> #include <linux/types.h> @@ -42,7 +44,6 @@ #include <asm/calgary.h> #include <asm/tce.h> #include <asm/pci-direct.h> -#include <asm/system.h> #include <asm/dma.h> #include <asm/rio.h> #include <asm/bios_ebda.h> @@ -246,7 +247,7 @@ static unsigned long iommu_range_alloc(struct device *dev, offset = iommu_area_alloc(tbl->it_map, tbl->it_size, 0, npages, 0, boundary_size, 0); if (offset == ~0UL) { - printk(KERN_WARNING "Calgary: IOMMU full.\n"); + pr_warn("IOMMU full\n"); spin_unlock_irqrestore(&tbl->it_lock, flags); if (panic_on_overflow) panic("Calgary: fix the allocator.\n"); @@ -272,8 +273,8 @@ static dma_addr_t iommu_alloc(struct device *dev, struct iommu_table *tbl, entry = iommu_range_alloc(dev, tbl, npages); if (unlikely(entry == DMA_ERROR_CODE)) { - printk(KERN_WARNING "Calgary: failed to allocate %u pages in " - "iommu %p\n", npages, tbl); + pr_warn("failed to allocate %u pages in iommu %p\n", + npages, tbl); return DMA_ERROR_CODE; } @@ -431,7 +432,7 @@ static void calgary_unmap_page(struct device *dev, dma_addr_t dma_addr, } static void* calgary_alloc_coherent(struct device *dev, size_t size, - dma_addr_t *dma_handle, gfp_t flag) + dma_addr_t *dma_handle, gfp_t flag, struct dma_attrs *attrs) { void *ret = NULL; dma_addr_t mapping; @@ -464,7 +465,8 @@ error: } static void calgary_free_coherent(struct device *dev, size_t size, - void *vaddr, dma_addr_t dma_handle) + void *vaddr, dma_addr_t dma_handle, + struct dma_attrs *attrs) { unsigned int npages; struct iommu_table *tbl = find_iommu_table(dev); @@ -477,8 +479,8 @@ static void calgary_free_coherent(struct device *dev, size_t size, } static struct dma_map_ops calgary_dma_ops = { - .alloc_coherent = calgary_alloc_coherent, - .free_coherent = calgary_free_coherent, + .alloc = calgary_alloc_coherent, + .free = calgary_free_coherent, .map_sg = calgary_map_sg, .unmap_sg = calgary_unmap_sg, .map_page = calgary_map_page, @@ -561,8 +563,7 @@ static void calgary_tce_cache_blast(struct iommu_table *tbl) i++; } while ((val & 0xff) != 0xff && i < 100); if (i == 100) - printk(KERN_WARNING "Calgary: PCI bus not quiesced, " - "continuing anyway\n"); + pr_warn("PCI bus not quiesced, continuing anyway\n"); /* invalidate TCE cache */ target = calgary_reg(bbar, tar_offset(tbl->it_busno)); @@ -604,8 +605,7 @@ begin: i++; } while ((val64 & 0xff) != 0xff && i < 100); if (i == 100) - printk(KERN_WARNING "CalIOC2: PCI bus not quiesced, " - "continuing anyway\n"); + pr_warn("CalIOC2: PCI bus not quiesced, continuing anyway\n"); /* 3. poll Page Migration DEBUG for SoftStopFault */ target = calgary_reg(bbar, phb_offset(bus) | PHB_PAGE_MIG_DEBUG); @@ -617,8 +617,7 @@ begin: if (++count < 100) goto begin; else { - printk(KERN_WARNING "CalIOC2: too many SoftStopFaults, " - "aborting TCE cache flush sequence!\n"); + pr_warn("CalIOC2: too many SoftStopFaults, aborting TCE cache flush sequence!\n"); return; /* pray for the best */ } } @@ -840,8 +839,8 @@ static void calgary_dump_error_regs(struct iommu_table *tbl) plssr = be32_to_cpu(readl(target)); /* If no error, the agent ID in the CSR is not valid */ - printk(KERN_EMERG "Calgary: DMA error on Calgary PHB 0x%x, " - "0x%08x@CSR 0x%08x@PLSSR\n", tbl->it_busno, csr, plssr); + pr_emerg("DMA error on Calgary PHB 0x%x, 0x%08x@CSR 0x%08x@PLSSR\n", + tbl->it_busno, csr, plssr); } static void calioc2_dump_error_regs(struct iommu_table *tbl) @@ -867,22 +866,21 @@ static void calioc2_dump_error_regs(struct iommu_table *tbl) target = calgary_reg(bbar, phboff | 0x800); mck = be32_to_cpu(readl(target)); - printk(KERN_EMERG "Calgary: DMA error on CalIOC2 PHB 0x%x\n", - tbl->it_busno); + pr_emerg("DMA error on CalIOC2 PHB 0x%x\n", tbl->it_busno); - printk(KERN_EMERG "Calgary: 0x%08x@CSR 0x%08x@PLSSR 0x%08x@CSMR 0x%08x@MCK\n", - csr, plssr, csmr, mck); + pr_emerg("0x%08x@CSR 0x%08x@PLSSR 0x%08x@CSMR 0x%08x@MCK\n", + csr, plssr, csmr, mck); /* dump rest of error regs */ - printk(KERN_EMERG "Calgary: "); + pr_emerg(""); for (i = 0; i < ARRAY_SIZE(errregs); i++) { /* err regs are at 0x810 - 0x870 */ erroff = (0x810 + (i * 0x10)); target = calgary_reg(bbar, phboff | erroff); errregs[i] = be32_to_cpu(readl(target)); - printk("0x%08x@0x%lx ", errregs[i], erroff); + pr_cont("0x%08x@0x%lx ", errregs[i], erroff); } - printk("\n"); + pr_cont("\n"); /* root complex status */ target = calgary_reg(bbar, phboff | PHB_ROOT_COMPLEX_STATUS); @@ -1209,23 +1207,31 @@ error: return ret; } -static inline int __init determine_tce_table_size(u64 ram) +static inline int __init determine_tce_table_size(void) { int ret; if (specified_table_size != TCE_TABLE_SIZE_UNSPECIFIED) return specified_table_size; - /* - * Table sizes are from 0 to 7 (TCE_TABLE_SIZE_64K to - * TCE_TABLE_SIZE_8M). Table size 0 has 8K entries and each - * larger table size has twice as many entries, so shift the - * max ram address by 13 to divide by 8K and then look at the - * order of the result to choose between 0-7. - */ - ret = get_order(ram >> 13); - if (ret > TCE_TABLE_SIZE_8M) + if (is_kdump_kernel() && saved_max_pfn) { + /* + * Table sizes are from 0 to 7 (TCE_TABLE_SIZE_64K to + * TCE_TABLE_SIZE_8M). Table size 0 has 8K entries and each + * larger table size has twice as many entries, so shift the + * max ram address by 13 to divide by 8K and then look at the + * order of the result to choose between 0-7. + */ + ret = get_order((saved_max_pfn * PAGE_SIZE) >> 13); + if (ret > TCE_TABLE_SIZE_8M) + ret = TCE_TABLE_SIZE_8M; + } else { + /* + * Use 8M by default (suggested by Muli) if it's not + * kdump kernel and saved_max_pfn isn't set. + */ ret = TCE_TABLE_SIZE_8M; + } return ret; } @@ -1420,8 +1426,7 @@ int __init detect_calgary(void) return -ENOMEM; } - specified_table_size = determine_tce_table_size((is_kdump_kernel() ? - saved_max_pfn : max_pfn) * PAGE_SIZE); + specified_table_size = determine_tce_table_size(); for (bus = 0; bus < MAX_PHB_BUS_NUM; bus++) { struct calgary_bus_info *info = &bus_info[bus]; @@ -1480,8 +1485,9 @@ cleanup: static int __init calgary_parse_options(char *p) { unsigned int bridge; + unsigned long val; size_t len; - char* endp; + ssize_t ret; while (*p) { if (!strncmp(p, "64k", 3)) @@ -1512,10 +1518,11 @@ static int __init calgary_parse_options(char *p) ++p; if (*p == '\0') break; - bridge = simple_strtoul(p, &endp, 0); - if (p == endp) + ret = kstrtoul(p, 0, &val); + if (ret) break; + bridge = val; if (bridge < MAX_PHB_BUS_NUM) { printk(KERN_INFO "Calgary: disabling " "translation for PHB %#x\n", bridge); diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c index 1c4d769e21e..a25e202bb31 100644 --- a/arch/x86/kernel/pci-dma.c +++ b/arch/x86/kernel/pci-dma.c @@ -45,15 +45,6 @@ int iommu_detected __read_mostly = 0; */ int iommu_pass_through __read_mostly; -/* - * Group multi-function PCI devices into a single device-group for the - * iommu_device_group interface. This tells the iommu driver to pretend - * it cannot distinguish between functions of a device, exposing only one - * group for the device. Useful for disallowing use of individual PCI - * functions from userspace drivers. - */ -int iommu_group_mf __read_mostly; - extern struct iommu_table_entry __iommu_table[], __iommu_table_end[]; /* Dummy device used for NULL arguments (normally ISA). */ @@ -65,7 +56,7 @@ struct device x86_dma_fallback_dev = { EXPORT_SYMBOL(x86_dma_fallback_dev); /* Number of entries preallocated for DMA-API debugging */ -#define PREALLOC_DMA_DEBUG_ENTRIES 32768 +#define PREALLOC_DMA_DEBUG_ENTRIES 65536 int dma_set_mask(struct device *dev, u64 mask) { @@ -96,17 +87,30 @@ void __init pci_iommu_alloc(void) } } void *dma_generic_alloc_coherent(struct device *dev, size_t size, - dma_addr_t *dma_addr, gfp_t flag) + dma_addr_t *dma_addr, gfp_t flag, + struct dma_attrs *attrs) { unsigned long dma_mask; struct page *page; + unsigned int count = PAGE_ALIGN(size) >> PAGE_SHIFT; dma_addr_t addr; dma_mask = dma_alloc_coherent_mask(dev, flag); - flag |= __GFP_ZERO; + flag &= ~__GFP_ZERO; again: - page = alloc_pages_node(dev_to_node(dev), flag, get_order(size)); + page = NULL; + /* CMA can be used only in the context which permits sleeping */ + if (flag & __GFP_WAIT) { + page = dma_alloc_from_contiguous(dev, count, get_order(size)); + if (page && page_to_phys(page) + size > dma_mask) { + dma_release_from_contiguous(dev, page, count); + page = NULL; + } + } + /* fallback */ + if (!page) + page = alloc_pages_node(dev_to_node(dev), flag, get_order(size)); if (!page) return NULL; @@ -121,11 +125,21 @@ again: return NULL; } - + memset(page_address(page), 0, size); *dma_addr = addr; return page_address(page); } +void dma_generic_free_coherent(struct device *dev, size_t size, void *vaddr, + dma_addr_t dma_addr, struct dma_attrs *attrs) +{ + unsigned int count = PAGE_ALIGN(size) >> PAGE_SHIFT; + struct page *page = virt_to_page(vaddr); + + if (!dma_release_from_contiguous(dev, page, count)) + free_pages((unsigned long)vaddr, get_order(size)); +} + /* * See <Documentation/x86/x86_64/boot-options.txt> for the iommu kernel * parameter documentation. @@ -178,8 +192,6 @@ static __init int iommu_setup(char *p) #endif if (!strncmp(p, "pt", 2)) iommu_pass_through = 1; - if (!strncmp(p, "group_mf", 8)) - iommu_group_mf = 1; gart_parse_options(p); @@ -260,12 +272,13 @@ rootfs_initcall(pci_iommu_init); #ifdef CONFIG_PCI /* Many VIA bridges seem to corrupt data for DAC. Disable it here */ -static __devinit void via_no_dac(struct pci_dev *dev) +static void via_no_dac(struct pci_dev *dev) { - if ((dev->class >> 8) == PCI_CLASS_BRIDGE_PCI && forbid_dac == 0) { + if (forbid_dac == 0) { dev_info(&dev->dev, "disabling DAC on VIA PCI bridge\n"); forbid_dac = 1; } } -DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_VIA, PCI_ANY_ID, via_no_dac); +DECLARE_PCI_FIXUP_CLASS_FINAL(PCI_VENDOR_ID_VIA, PCI_ANY_ID, + PCI_CLASS_BRIDGE_PCI, 8, via_no_dac); #endif diff --git a/arch/x86/kernel/pci-nommu.c b/arch/x86/kernel/pci-nommu.c index 3af4af810c0..da15918d1c8 100644 --- a/arch/x86/kernel/pci-nommu.c +++ b/arch/x86/kernel/pci-nommu.c @@ -3,7 +3,6 @@ #include <linux/dma-mapping.h> #include <linux/scatterlist.h> #include <linux/string.h> -#include <linux/init.h> #include <linux/gfp.h> #include <linux/pci.h> #include <linux/mm.h> @@ -74,12 +73,6 @@ static int nommu_map_sg(struct device *hwdev, struct scatterlist *sg, return nents; } -static void nommu_free_coherent(struct device *dev, size_t size, void *vaddr, - dma_addr_t dma_addr) -{ - free_pages((unsigned long)vaddr, get_order(size)); -} - static void nommu_sync_single_for_device(struct device *dev, dma_addr_t addr, size_t size, enum dma_data_direction dir) @@ -96,8 +89,8 @@ static void nommu_sync_sg_for_device(struct device *dev, } struct dma_map_ops nommu_dma_ops = { - .alloc_coherent = dma_generic_alloc_coherent, - .free_coherent = nommu_free_coherent, + .alloc = dma_generic_alloc_coherent, + .free = dma_generic_free_coherent, .map_sg = nommu_map_sg, .map_page = nommu_map_page, .sync_single_for_device = nommu_sync_single_for_device, diff --git a/arch/x86/kernel/pci-swiotlb.c b/arch/x86/kernel/pci-swiotlb.c index 8f972cbddef..77dd0ad58be 100644 --- a/arch/x86/kernel/pci-swiotlb.c +++ b/arch/x86/kernel/pci-swiotlb.c @@ -14,22 +14,34 @@ #include <asm/iommu_table.h> int swiotlb __read_mostly; -static void *x86_swiotlb_alloc_coherent(struct device *hwdev, size_t size, - dma_addr_t *dma_handle, gfp_t flags) +void *x86_swiotlb_alloc_coherent(struct device *hwdev, size_t size, + dma_addr_t *dma_handle, gfp_t flags, + struct dma_attrs *attrs) { void *vaddr; - vaddr = dma_generic_alloc_coherent(hwdev, size, dma_handle, flags); + vaddr = dma_generic_alloc_coherent(hwdev, size, dma_handle, flags, + attrs); if (vaddr) return vaddr; return swiotlb_alloc_coherent(hwdev, size, dma_handle, flags); } +void x86_swiotlb_free_coherent(struct device *dev, size_t size, + void *vaddr, dma_addr_t dma_addr, + struct dma_attrs *attrs) +{ + if (is_swiotlb_buffer(dma_to_phys(dev, dma_addr))) + swiotlb_free_coherent(dev, size, vaddr, dma_addr); + else + dma_generic_free_coherent(dev, size, vaddr, dma_addr, attrs); +} + static struct dma_map_ops swiotlb_dma_ops = { .mapping_error = swiotlb_dma_mapping_error, - .alloc_coherent = x86_swiotlb_alloc_coherent, - .free_coherent = swiotlb_free_coherent, + .alloc = x86_swiotlb_alloc_coherent, + .free = x86_swiotlb_free_coherent, .sync_single_for_cpu = swiotlb_sync_single_for_cpu, .sync_single_for_device = swiotlb_sync_single_for_device, .sync_sg_for_cpu = swiotlb_sync_sg_for_cpu, diff --git a/arch/x86/kernel/perf_regs.c b/arch/x86/kernel/perf_regs.c new file mode 100644 index 00000000000..e309cc5c276 --- /dev/null +++ b/arch/x86/kernel/perf_regs.c @@ -0,0 +1,105 @@ +#include <linux/errno.h> +#include <linux/kernel.h> +#include <linux/sched.h> +#include <linux/perf_event.h> +#include <linux/bug.h> +#include <linux/stddef.h> +#include <asm/perf_regs.h> +#include <asm/ptrace.h> + +#ifdef CONFIG_X86_32 +#define PERF_REG_X86_MAX PERF_REG_X86_32_MAX +#else +#define PERF_REG_X86_MAX PERF_REG_X86_64_MAX +#endif + +#define PT_REGS_OFFSET(id, r) [id] = offsetof(struct pt_regs, r) + +static unsigned int pt_regs_offset[PERF_REG_X86_MAX] = { + PT_REGS_OFFSET(PERF_REG_X86_AX, ax), + PT_REGS_OFFSET(PERF_REG_X86_BX, bx), + PT_REGS_OFFSET(PERF_REG_X86_CX, cx), + PT_REGS_OFFSET(PERF_REG_X86_DX, dx), + PT_REGS_OFFSET(PERF_REG_X86_SI, si), + PT_REGS_OFFSET(PERF_REG_X86_DI, di), + PT_REGS_OFFSET(PERF_REG_X86_BP, bp), + PT_REGS_OFFSET(PERF_REG_X86_SP, sp), + PT_REGS_OFFSET(PERF_REG_X86_IP, ip), + PT_REGS_OFFSET(PERF_REG_X86_FLAGS, flags), + PT_REGS_OFFSET(PERF_REG_X86_CS, cs), + PT_REGS_OFFSET(PERF_REG_X86_SS, ss), +#ifdef CONFIG_X86_32 + PT_REGS_OFFSET(PERF_REG_X86_DS, ds), + PT_REGS_OFFSET(PERF_REG_X86_ES, es), + PT_REGS_OFFSET(PERF_REG_X86_FS, fs), + PT_REGS_OFFSET(PERF_REG_X86_GS, gs), +#else + /* + * The pt_regs struct does not store + * ds, es, fs, gs in 64 bit mode. + */ + (unsigned int) -1, + (unsigned int) -1, + (unsigned int) -1, + (unsigned int) -1, +#endif +#ifdef CONFIG_X86_64 + PT_REGS_OFFSET(PERF_REG_X86_R8, r8), + PT_REGS_OFFSET(PERF_REG_X86_R9, r9), + PT_REGS_OFFSET(PERF_REG_X86_R10, r10), + PT_REGS_OFFSET(PERF_REG_X86_R11, r11), + PT_REGS_OFFSET(PERF_REG_X86_R12, r12), + PT_REGS_OFFSET(PERF_REG_X86_R13, r13), + PT_REGS_OFFSET(PERF_REG_X86_R14, r14), + PT_REGS_OFFSET(PERF_REG_X86_R15, r15), +#endif +}; + +u64 perf_reg_value(struct pt_regs *regs, int idx) +{ + if (WARN_ON_ONCE(idx >= ARRAY_SIZE(pt_regs_offset))) + return 0; + + return regs_get_register(regs, pt_regs_offset[idx]); +} + +#define REG_RESERVED (~((1ULL << PERF_REG_X86_MAX) - 1ULL)) + +#ifdef CONFIG_X86_32 +int perf_reg_validate(u64 mask) +{ + if (!mask || mask & REG_RESERVED) + return -EINVAL; + + return 0; +} + +u64 perf_reg_abi(struct task_struct *task) +{ + return PERF_SAMPLE_REGS_ABI_32; +} +#else /* CONFIG_X86_64 */ +#define REG_NOSUPPORT ((1ULL << PERF_REG_X86_DS) | \ + (1ULL << PERF_REG_X86_ES) | \ + (1ULL << PERF_REG_X86_FS) | \ + (1ULL << PERF_REG_X86_GS)) + +int perf_reg_validate(u64 mask) +{ + if (!mask || mask & REG_RESERVED) + return -EINVAL; + + if (mask & REG_NOSUPPORT) + return -EINVAL; + + return 0; +} + +u64 perf_reg_abi(struct task_struct *task) +{ + if (test_tsk_thread_flag(task, TIF_IA32)) + return PERF_SAMPLE_REGS_ABI_32; + else + return PERF_SAMPLE_REGS_ABI_64; +} +#endif /* CONFIG_X86_32 */ diff --git a/arch/x86/kernel/preempt.S b/arch/x86/kernel/preempt.S new file mode 100644 index 00000000000..ca7f0d58a87 --- /dev/null +++ b/arch/x86/kernel/preempt.S @@ -0,0 +1,25 @@ + +#include <linux/linkage.h> +#include <asm/dwarf2.h> +#include <asm/asm.h> +#include <asm/calling.h> + +ENTRY(___preempt_schedule) + CFI_STARTPROC + SAVE_ALL + call preempt_schedule + RESTORE_ALL + ret + CFI_ENDPROC + +#ifdef CONFIG_CONTEXT_TRACKING + +ENTRY(___preempt_schedule_context) + CFI_STARTPROC + SAVE_ALL + call preempt_schedule_context + RESTORE_ALL + ret + CFI_ENDPROC + +#endif diff --git a/arch/x86/kernel/probe_roms.c b/arch/x86/kernel/probe_roms.c index 34e06e84ce3..d5f15c3f7b2 100644 --- a/arch/x86/kernel/probe_roms.c +++ b/arch/x86/kernel/probe_roms.c @@ -12,6 +12,7 @@ #include <linux/pci.h> #include <linux/export.h> +#include <asm/probe_roms.h> #include <asm/pci-direct.h> #include <asm/e820.h> #include <asm/mmzone.h> @@ -149,7 +150,7 @@ static struct resource *find_oprom(struct pci_dev *pdev) return oprom; } -void *pci_map_biosrom(struct pci_dev *pdev) +void __iomem *pci_map_biosrom(struct pci_dev *pdev) { struct resource *oprom = find_oprom(pdev); diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c index 15763af7bfe..4505e2a950d 100644 --- a/arch/x86/kernel/process.c +++ b/arch/x86/kernel/process.c @@ -1,3 +1,5 @@ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/errno.h> #include <linux/kernel.h> #include <linux/mm.h> @@ -12,20 +14,54 @@ #include <linux/user-return-notifier.h> #include <linux/dmi.h> #include <linux/utsname.h> +#include <linux/stackprotector.h> +#include <linux/tick.h> +#include <linux/cpuidle.h> #include <trace/events/power.h> #include <linux/hw_breakpoint.h> #include <asm/cpu.h> -#include <asm/system.h> #include <asm/apic.h> #include <asm/syscalls.h> #include <asm/idle.h> #include <asm/uaccess.h> #include <asm/i387.h> +#include <asm/fpu-internal.h> #include <asm/debugreg.h> +#include <asm/nmi.h> + +/* + * per-CPU TSS segments. Threads are completely 'soft' on Linux, + * no more per-task TSS's. The TSS size is kept cacheline-aligned + * so they are allowed to end up in the .data..cacheline_aligned + * section. Since TSS's are completely CPU-local, we want them + * on exact cacheline boundaries, to eliminate cacheline ping-pong. + */ +__visible DEFINE_PER_CPU_SHARED_ALIGNED(struct tss_struct, init_tss) = INIT_TSS; + +#ifdef CONFIG_X86_64 +static DEFINE_PER_CPU(unsigned char, is_idle); +static ATOMIC_NOTIFIER_HEAD(idle_notifier); + +void idle_notifier_register(struct notifier_block *n) +{ + atomic_notifier_chain_register(&idle_notifier, n); +} +EXPORT_SYMBOL_GPL(idle_notifier_register); + +void idle_notifier_unregister(struct notifier_block *n) +{ + atomic_notifier_chain_unregister(&idle_notifier, n); +} +EXPORT_SYMBOL_GPL(idle_notifier_unregister); +#endif struct kmem_cache *task_xstate_cachep; EXPORT_SYMBOL_GPL(task_xstate_cachep); +/* + * this gets called so that we can store lazy state into memory and copy the + * current task into the new thread. + */ int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src) { int ret; @@ -36,7 +72,7 @@ int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src) ret = fpu_alloc(&dst->thread.fpu); if (ret) return ret; - fpu_copy(&dst->thread.fpu, &src->thread.fpu); + fpu_copy(dst, src); } return 0; } @@ -46,10 +82,9 @@ void free_thread_xstate(struct task_struct *tsk) fpu_free(&tsk->thread.fpu); } -void free_thread_info(struct thread_info *ti) +void arch_release_task_struct(struct task_struct *tsk) { - free_thread_xstate(ti->task); - free_pages((unsigned long)ti, THREAD_ORDER); + free_thread_xstate(tsk); } void arch_task_cache_init(void) @@ -82,38 +117,8 @@ void exit_thread(void) put_cpu(); kfree(bp); } -} - -void show_regs(struct pt_regs *regs) -{ - show_registers(regs); - show_trace(NULL, regs, (unsigned long *)kernel_stack_pointer(regs), 0); -} - -void show_regs_common(void) -{ - const char *vendor, *product, *board; - - vendor = dmi_get_system_info(DMI_SYS_VENDOR); - if (!vendor) - vendor = ""; - product = dmi_get_system_info(DMI_PRODUCT_NAME); - if (!product) - product = ""; - /* Board Name is optional */ - board = dmi_get_system_info(DMI_BOARD_NAME); - - printk(KERN_CONT "\n"); - printk(KERN_DEFAULT "Pid: %d, comm: %.20s %s %s %.*s", - current->pid, current->comm, print_tainted(), - init_utsname()->release, - (int)strcspn(init_utsname()->version, " "), - init_utsname()->version); - printk(KERN_CONT " %s %s", vendor, product); - if (board) - printk(KERN_CONT "/%s", board); - printk(KERN_CONT "\n"); + drop_fpu(me); } void flush_thread(void) @@ -122,12 +127,13 @@ void flush_thread(void) flush_ptrace_hw_breakpoint(tsk); memset(tsk->thread.tls_array, 0, sizeof(tsk->thread.tls_array)); + drop_init_fpu(tsk); /* - * Forget coprocessor state.. + * Free the FPU state for non xsave platforms. They get reallocated + * lazily at the first use. */ - tsk->fpu_counter = 0; - clear_fpu(tsk); - clear_used_math(); + if (!use_eager_fpu()) + free_thread_xstate(tsk); } static void hard_disable_TSC(void) @@ -232,185 +238,92 @@ void __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p, propagate_user_return_notify(prev_p, next_p); } -int sys_fork(struct pt_regs *regs) -{ - return do_fork(SIGCHLD, regs->sp, regs, 0, NULL, NULL); -} - /* - * This is trivial, and on the face of it looks like it - * could equally well be done in user mode. - * - * Not so, for quite unobvious reasons - register pressure. - * In user mode vfork() cannot have a stack frame, and if - * done by calling the "clone()" system call directly, you - * do not have enough call-clobbered registers to hold all - * the information you need. + * Idle related variables and functions */ -int sys_vfork(struct pt_regs *regs) +unsigned long boot_option_idle_override = IDLE_NO_OVERRIDE; +EXPORT_SYMBOL(boot_option_idle_override); + +static void (*x86_idle)(void); + +#ifndef CONFIG_SMP +static inline void play_dead(void) { - return do_fork(CLONE_VFORK | CLONE_VM | SIGCHLD, regs->sp, regs, 0, - NULL, NULL); + BUG(); } +#endif -long -sys_clone(unsigned long clone_flags, unsigned long newsp, - void __user *parent_tid, void __user *child_tid, struct pt_regs *regs) +#ifdef CONFIG_X86_64 +void enter_idle(void) { - if (!newsp) - newsp = regs->sp; - return do_fork(clone_flags, newsp, regs, 0, parent_tid, child_tid); + this_cpu_write(is_idle, 1); + atomic_notifier_call_chain(&idle_notifier, IDLE_START, NULL); } -/* - * This gets run with %si containing the - * function to call, and %di containing - * the "args". - */ -extern void kernel_thread_helper(void); - -/* - * Create a kernel thread - */ -int kernel_thread(int (*fn)(void *), void *arg, unsigned long flags) +static void __exit_idle(void) { - struct pt_regs regs; - - memset(®s, 0, sizeof(regs)); - - regs.si = (unsigned long) fn; - regs.di = (unsigned long) arg; - -#ifdef CONFIG_X86_32 - regs.ds = __USER_DS; - regs.es = __USER_DS; - regs.fs = __KERNEL_PERCPU; - regs.gs = __KERNEL_STACK_CANARY; -#else - regs.ss = __KERNEL_DS; -#endif - - regs.orig_ax = -1; - regs.ip = (unsigned long) kernel_thread_helper; - regs.cs = __KERNEL_CS | get_kernel_rpl(); - regs.flags = X86_EFLAGS_IF | X86_EFLAGS_BIT1; - - /* Ok, create the new process.. */ - return do_fork(flags | CLONE_VM | CLONE_UNTRACED, 0, ®s, 0, NULL, NULL); + if (x86_test_and_clear_bit_percpu(0, is_idle) == 0) + return; + atomic_notifier_call_chain(&idle_notifier, IDLE_END, NULL); } -EXPORT_SYMBOL(kernel_thread); -/* - * sys_execve() executes a new program. - */ -long sys_execve(const char __user *name, - const char __user *const __user *argv, - const char __user *const __user *envp, struct pt_regs *regs) -{ - long error; - char *filename; - - filename = getname(name); - error = PTR_ERR(filename); - if (IS_ERR(filename)) - return error; - error = do_execve(filename, argv, envp, regs); - -#ifdef CONFIG_X86_32 - if (error == 0) { - /* Make sure we don't return using sysenter.. */ - set_thread_flag(TIF_IRET); - } -#endif - - putname(filename); - return error; +/* Called from interrupts to signify idle end */ +void exit_idle(void) +{ + /* idle loop has pid 0 */ + if (current->pid) + return; + __exit_idle(); } - -/* - * Idle related variables and functions - */ -unsigned long boot_option_idle_override = IDLE_NO_OVERRIDE; -EXPORT_SYMBOL(boot_option_idle_override); - -/* - * Powermanagement idle function, if any.. - */ -void (*pm_idle)(void); -#ifdef CONFIG_APM_MODULE -EXPORT_SYMBOL(pm_idle); #endif -#ifdef CONFIG_X86_32 -/* - * This halt magic was a workaround for ancient floppy DMA - * wreckage. It should be safe to remove. - */ -static int hlt_counter; -void disable_hlt(void) +void arch_cpu_idle_enter(void) { - hlt_counter++; + local_touch_nmi(); + enter_idle(); } -EXPORT_SYMBOL(disable_hlt); -void enable_hlt(void) +void arch_cpu_idle_exit(void) { - hlt_counter--; + __exit_idle(); } -EXPORT_SYMBOL(enable_hlt); -static inline int hlt_use_halt(void) +void arch_cpu_idle_dead(void) { - return (!hlt_counter && boot_cpu_data.hlt_works_ok); + play_dead(); } -#else -static inline int hlt_use_halt(void) + +/* + * Called from the generic idle code. + */ +void arch_cpu_idle(void) { - return 1; + x86_idle(); } -#endif /* - * We use this if we don't have any better - * idle routine.. + * We use this if we don't have any better idle routine.. */ void default_idle(void) { - if (hlt_use_halt()) { - trace_power_start(POWER_CSTATE, 1, smp_processor_id()); - trace_cpu_idle(1, smp_processor_id()); - current_thread_info()->status &= ~TS_POLLING; - /* - * TS_POLLING-cleared state must be visible before we - * test NEED_RESCHED: - */ - smp_mb(); - - if (!need_resched()) - safe_halt(); /* enables interrupts racelessly */ - else - local_irq_enable(); - current_thread_info()->status |= TS_POLLING; - trace_power_end(smp_processor_id()); - trace_cpu_idle(PWR_EVENT_EXIT, smp_processor_id()); - } else { - local_irq_enable(); - /* loop is done by the caller */ - cpu_relax(); - } + trace_cpu_idle_rcuidle(1, smp_processor_id()); + safe_halt(); + trace_cpu_idle_rcuidle(PWR_EVENT_EXIT, smp_processor_id()); } #ifdef CONFIG_APM_MODULE EXPORT_SYMBOL(default_idle); #endif -bool set_pm_idle_to_default(void) +#ifdef CONFIG_XEN +bool xen_set_default_idle(void) { - bool ret = !!pm_idle; + bool ret = !!x86_idle; - pm_idle = default_idle; + x86_idle = default_idle; return ret; } +#endif void stop_this_cpu(void *dummy) { local_irq_disable(); @@ -420,106 +333,8 @@ void stop_this_cpu(void *dummy) set_cpu_online(smp_processor_id(), false); disable_local_APIC(); - for (;;) { - if (hlt_works(smp_processor_id())) - halt(); - } -} - -static void do_nothing(void *unused) -{ -} - -/* - * cpu_idle_wait - Used to ensure that all the CPUs discard old value of - * pm_idle and update to new pm_idle value. Required while changing pm_idle - * handler on SMP systems. - * - * Caller must have changed pm_idle to the new value before the call. Old - * pm_idle value will not be used by any CPU after the return of this function. - */ -void cpu_idle_wait(void) -{ - smp_mb(); - /* kick all the CPUs so that they exit out of pm_idle */ - smp_call_function(do_nothing, NULL, 1); -} -EXPORT_SYMBOL_GPL(cpu_idle_wait); - -/* Default MONITOR/MWAIT with no hints, used for default C1 state */ -static void mwait_idle(void) -{ - if (!need_resched()) { - trace_power_start(POWER_CSTATE, 1, smp_processor_id()); - trace_cpu_idle(1, smp_processor_id()); - if (this_cpu_has(X86_FEATURE_CLFLUSH_MONITOR)) - clflush((void *)¤t_thread_info()->flags); - - __monitor((void *)¤t_thread_info()->flags, 0, 0); - smp_mb(); - if (!need_resched()) - __sti_mwait(0, 0); - else - local_irq_enable(); - trace_power_end(smp_processor_id()); - trace_cpu_idle(PWR_EVENT_EXIT, smp_processor_id()); - } else - local_irq_enable(); -} - -/* - * On SMP it's slightly faster (but much more power-consuming!) - * to poll the ->work.need_resched flag instead of waiting for the - * cross-CPU IPI to arrive. Use this option with caution. - */ -static void poll_idle(void) -{ - trace_power_start(POWER_CSTATE, 0, smp_processor_id()); - trace_cpu_idle(0, smp_processor_id()); - local_irq_enable(); - while (!need_resched()) - cpu_relax(); - trace_power_end(smp_processor_id()); - trace_cpu_idle(PWR_EVENT_EXIT, smp_processor_id()); -} - -/* - * mwait selection logic: - * - * It depends on the CPU. For AMD CPUs that support MWAIT this is - * wrong. Family 0x10 and 0x11 CPUs will enter C1 on HLT. Powersavings - * then depend on a clock divisor and current Pstate of the core. If - * all cores of a processor are in halt state (C1) the processor can - * enter the C1E (C1 enhanced) state. If mwait is used this will never - * happen. - * - * idle=mwait overrides this decision and forces the usage of mwait. - */ - -#define MWAIT_INFO 0x05 -#define MWAIT_ECX_EXTENDED_INFO 0x01 -#define MWAIT_EDX_C1 0xf0 - -int mwait_usable(const struct cpuinfo_x86 *c) -{ - u32 eax, ebx, ecx, edx; - - if (boot_option_idle_override == IDLE_FORCE_MWAIT) - return 1; - - if (c->cpuid_level < MWAIT_INFO) - return 0; - - cpuid(MWAIT_INFO, &eax, &ebx, &ecx, &edx); - /* Check, whether EDX has extended info about MWAIT */ - if (!(ecx & MWAIT_ECX_EXTENDED_INFO)) - return 1; - - /* - * edx enumeratios MONITOR/MWAIT extensions. Check, whether - * C1 supports MWAIT - */ - return (edx & MWAIT_EDX_C1); + for (;;) + halt(); } bool amd_e400_c1e_detected; @@ -540,9 +355,6 @@ void amd_e400_remove_cpu(int cpu) */ static void amd_e400_idle(void) { - if (need_resched()) - return; - if (!amd_e400_c1e_detected) { u32 lo, hi; @@ -552,7 +364,7 @@ static void amd_e400_idle(void) amd_e400_c1e_detected = true; if (!boot_cpu_has(X86_FEATURE_NONSTOP_TSC)) mark_tsc_unstable("TSC halt in AMD C1E"); - printk(KERN_INFO "System has AMD C1E enabled\n"); + pr_info("System has AMD C1E enabled\n"); } } @@ -566,8 +378,7 @@ static void amd_e400_idle(void) */ clockevents_notify(CLOCK_EVT_NOTIFY_BROADCAST_FORCE, &cpu); - printk(KERN_INFO "Switch to broadcast mode on CPU%d\n", - cpu); + pr_info("Switch to broadcast mode on CPU%d\n", cpu); } clockevents_notify(CLOCK_EVT_NOTIFY_BROADCAST_ENTER, &cpu); @@ -577,42 +388,34 @@ static void amd_e400_idle(void) * The switch back from broadcast mode needs to be * called with interrupts disabled. */ - local_irq_disable(); - clockevents_notify(CLOCK_EVT_NOTIFY_BROADCAST_EXIT, &cpu); - local_irq_enable(); + local_irq_disable(); + clockevents_notify(CLOCK_EVT_NOTIFY_BROADCAST_EXIT, &cpu); + local_irq_enable(); } else default_idle(); } -void __cpuinit select_idle_routine(const struct cpuinfo_x86 *c) +void select_idle_routine(const struct cpuinfo_x86 *c) { #ifdef CONFIG_SMP - if (pm_idle == poll_idle && smp_num_siblings > 1) { - printk_once(KERN_WARNING "WARNING: polling idle and HT enabled," - " performance may degrade.\n"); - } + if (boot_option_idle_override == IDLE_POLL && smp_num_siblings > 1) + pr_warn_once("WARNING: polling idle and HT enabled, performance may degrade\n"); #endif - if (pm_idle) + if (x86_idle || boot_option_idle_override == IDLE_POLL) return; - if (cpu_has(c, X86_FEATURE_MWAIT) && mwait_usable(c)) { - /* - * One CPU supports mwait => All CPUs supports mwait - */ - printk(KERN_INFO "using mwait in idle threads.\n"); - pm_idle = mwait_idle; - } else if (cpu_has_amd_erratum(amd_erratum_400)) { + if (cpu_has_bug(c, X86_BUG_AMD_APIC_C1E)) { /* E400: APIC timer interrupt does not wake up CPU from C1e */ - printk(KERN_INFO "using AMD E400 aware idle routine\n"); - pm_idle = amd_e400_idle; + pr_info("using AMD E400 aware idle routine\n"); + x86_idle = amd_e400_idle; } else - pm_idle = default_idle; + x86_idle = default_idle; } void __init init_amd_e400_c1e_mask(void) { /* If we're using amd_e400_idle, we need to allocate amd_e400_c1e_mask. */ - if (pm_idle == amd_e400_idle) + if (x86_idle == amd_e400_idle) zalloc_cpumask_var(&amd_e400_c1e_mask, GFP_KERNEL); } @@ -622,12 +425,9 @@ static int __init idle_setup(char *str) return -EINVAL; if (!strcmp(str, "poll")) { - printk("using polling idle threads.\n"); - pm_idle = poll_idle; + pr_info("using polling idle threads\n"); boot_option_idle_override = IDLE_POLL; - } else if (!strcmp(str, "mwait")) { - boot_option_idle_override = IDLE_FORCE_MWAIT; - WARN_ONCE(1, "\"idle=mwait\" will be removed in 2012\n"); + cpu_idle_poll_ctrl(true); } else if (!strcmp(str, "halt")) { /* * When the boot option of idle=halt is added, halt is @@ -636,7 +436,7 @@ static int __init idle_setup(char *str) * To continue to load the CPU idle driver, don't touch * the boot_option_idle_override. */ - pm_idle = default_idle; + x86_idle = default_idle; boot_option_idle_override = IDLE_HALT; } else if (!strcmp(str, "nomwait")) { /* diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c index 80bfe1ab003..7bc86bbe748 100644 --- a/arch/x86/kernel/process_32.c +++ b/arch/x86/kernel/process_32.c @@ -9,7 +9,6 @@ * This file handles the architecture-dependent parts of process handling.. */ -#include <linux/stackprotector.h> #include <linux/cpu.h> #include <linux/errno.h> #include <linux/sched.h> @@ -25,26 +24,23 @@ #include <linux/interrupt.h> #include <linux/delay.h> #include <linux/reboot.h> -#include <linux/init.h> #include <linux/mc146818rtc.h> #include <linux/module.h> #include <linux/kallsyms.h> #include <linux/ptrace.h> #include <linux/personality.h> -#include <linux/tick.h> #include <linux/percpu.h> #include <linux/prctl.h> #include <linux/ftrace.h> #include <linux/uaccess.h> #include <linux/io.h> #include <linux/kdebug.h> -#include <linux/cpuidle.h> #include <asm/pgtable.h> -#include <asm/system.h> #include <asm/ldt.h> #include <asm/processor.h> #include <asm/i387.h> +#include <asm/fpu-internal.h> #include <asm/desc.h> #ifdef CONFIG_MATH_EMULATION #include <asm/math_emu.h> @@ -57,9 +53,10 @@ #include <asm/idle.h> #include <asm/syscalls.h> #include <asm/debugreg.h> -#include <asm/nmi.h> +#include <asm/switch_to.h> asmlinkage void ret_from_fork(void) __asm__("ret_from_fork"); +asmlinkage void ret_from_kernel_thread(void) __asm__("ret_from_kernel_thread"); /* * Return saved PC of a blocked thread. @@ -69,62 +66,6 @@ unsigned long thread_saved_pc(struct task_struct *tsk) return ((unsigned long *)tsk->thread.sp)[3]; } -#ifndef CONFIG_SMP -static inline void play_dead(void) -{ - BUG(); -} -#endif - -/* - * The idle thread. There's no useful work to be - * done, so just try to conserve power and have a - * low exit latency (ie sit in a loop waiting for - * somebody to say that they'd like to reschedule) - */ -void cpu_idle(void) -{ - int cpu = smp_processor_id(); - - /* - * If we're the non-boot CPU, nothing set the stack canary up - * for us. CPU0 already has it initialized but no harm in - * doing it again. This is a good place for updating it, as - * we wont ever return from this function (so the invalid - * canaries already on the stack wont ever trigger). - */ - boot_init_stack_canary(); - - current_thread_info()->status |= TS_POLLING; - - /* endless idle loop with no priority at all */ - while (1) { - tick_nohz_idle_enter(); - rcu_idle_enter(); - while (!need_resched()) { - - check_pgt_cache(); - rmb(); - - if (cpu_is_offline(cpu)) - play_dead(); - - local_touch_nmi(); - local_irq_disable(); - /* Don't trace irqs off for idle */ - stop_critical_timings(); - if (cpuidle_idle_call()) - pm_idle(); - start_critical_timings(); - } - rcu_idle_exit(); - tick_nohz_idle_exit(); - preempt_enable_no_resched(); - schedule(); - preempt_disable(); - } -} - void __show_regs(struct pt_regs *regs, int all) { unsigned long cr0 = 0L, cr2 = 0L, cr3 = 0L, cr4 = 0L; @@ -142,8 +83,6 @@ void __show_regs(struct pt_regs *regs, int all) savesegment(gs, gs); } - show_regs_common(); - printk(KERN_DEFAULT "EIP: %04x:[<%08lx>] EFLAGS: %08lx CPU: %d\n", (u16)regs->cs, regs->ip, regs->flags, smp_processor_id()); @@ -170,11 +109,16 @@ void __show_regs(struct pt_regs *regs, int all) get_debugreg(d1, 1); get_debugreg(d2, 2); get_debugreg(d3, 3); - printk(KERN_DEFAULT "DR0: %08lx DR1: %08lx DR2: %08lx DR3: %08lx\n", - d0, d1, d2, d3); - get_debugreg(d6, 6); get_debugreg(d7, 7); + + /* Only print out debug registers if they are in their non-default state. */ + if ((d0 == 0) && (d1 == 0) && (d2 == 0) && (d3 == 0) && + (d6 == DR6_RESERVED) && (d7 == 0x400)) + return; + + printk(KERN_DEFAULT "DR0: %08lx DR1: %08lx DR2: %08lx DR3: %08lx\n", + d0, d1, d2, d3); printk(KERN_DEFAULT "DR6: %08lx DR7: %08lx\n", d6, d7); } @@ -185,35 +129,43 @@ void release_thread(struct task_struct *dead_task) release_vm86_irqs(dead_task); } -/* - * This gets called before we allocate a new thread and copy - * the current task into it. - */ -void prepare_to_copy(struct task_struct *tsk) -{ - unlazy_fpu(tsk); -} - int copy_thread(unsigned long clone_flags, unsigned long sp, - unsigned long unused, - struct task_struct *p, struct pt_regs *regs) + unsigned long arg, struct task_struct *p) { - struct pt_regs *childregs; + struct pt_regs *childregs = task_pt_regs(p); struct task_struct *tsk; int err; - childregs = task_pt_regs(p); - *childregs = *regs; - childregs->ax = 0; - childregs->sp = sp; - p->thread.sp = (unsigned long) childregs; p->thread.sp0 = (unsigned long) (childregs+1); - p->thread.ip = (unsigned long) ret_from_fork; + if (unlikely(p->flags & PF_KTHREAD)) { + /* kernel thread */ + memset(childregs, 0, sizeof(struct pt_regs)); + p->thread.ip = (unsigned long) ret_from_kernel_thread; + task_user_gs(p) = __KERNEL_STACK_CANARY; + childregs->ds = __USER_DS; + childregs->es = __USER_DS; + childregs->fs = __KERNEL_PERCPU; + childregs->bx = sp; /* function */ + childregs->bp = arg; + childregs->orig_ax = -1; + childregs->cs = __KERNEL_CS | get_kernel_rpl(); + childregs->flags = X86_EFLAGS_IF | X86_EFLAGS_FIXED; + p->thread.fpu_counter = 0; + p->thread.io_bitmap_ptr = NULL; + memset(p->thread.ptrace_bps, 0, sizeof(p->thread.ptrace_bps)); + return 0; + } + *childregs = *current_pt_regs(); + childregs->ax = 0; + if (sp) + childregs->sp = sp; - task_user_gs(p) = get_user_gs(regs); + p->thread.ip = (unsigned long) ret_from_fork; + task_user_gs(p) = get_user_gs(current_pt_regs()); + p->thread.fpu_counter = 0; p->thread.io_bitmap_ptr = NULL; tsk = current; err = -ENOMEM; @@ -257,10 +209,12 @@ start_thread(struct pt_regs *regs, unsigned long new_ip, unsigned long new_sp) regs->cs = __USER_CS; regs->ip = new_ip; regs->sp = new_sp; + regs->flags = X86_EFLAGS_IF; /* - * Free the old FP and other extended state + * force it to the iret return path by making it look as if there was + * some work pending. */ - free_thread_xstate(current); + set_thread_flag(TIF_NOTIFY_RESUME); } EXPORT_SYMBOL_GPL(start_thread); @@ -292,7 +246,7 @@ EXPORT_SYMBOL_GPL(start_thread); * the task-switch, and shows up in ret_from_fork in entry.S, * for example. */ -__notrace_funcgraph struct task_struct * +__visible __notrace_funcgraph struct task_struct * __switch_to(struct task_struct *prev_p, struct task_struct *next_p) { struct thread_struct *prev = &prev_p->thread, @@ -303,7 +257,7 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) /* never put a printk in __switch_to... printk() calls wake_up*() indirectly */ - fpu = switch_fpu_prepare(prev_p, next_p); + fpu = switch_fpu_prepare(prev_p, next_p, cpu); /* * Reload esp0. @@ -337,6 +291,14 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) set_iopl_mask(next->iopl); /* + * If it were not for PREEMPT_ACTIVE we could guarantee that the + * preempt_count of all tasks was equal here and this would not be + * needed. + */ + task_thread_info(prev_p)->saved_preempt_count = this_cpu_read(__preempt_count); + this_cpu_write(__preempt_count, task_thread_info(next_p)->saved_preempt_count); + + /* * Now maybe handle debug registers and/or IO bitmaps */ if (unlikely(task_thread_info(prev_p)->flags & _TIF_WORK_CTXSW_PREV || @@ -352,6 +314,10 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) */ arch_end_context_switch(next_p); + this_cpu_write(kernel_stack, + (unsigned long)task_stack_page(next_p) + + THREAD_SIZE - KERNEL_STACK_OFFSET); + /* * Restore %gs if needed (which is common) */ @@ -360,7 +326,7 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) switch_fpu_finish(next_p, fpu); - percpu_write(current_task, next_p); + this_cpu_write(current_task, next_p); return prev_p; } diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c index 1fd94bc4279..ca5b02d405c 100644 --- a/arch/x86/kernel/process_64.c +++ b/arch/x86/kernel/process_64.c @@ -14,7 +14,6 @@ * This file handles the architecture-dependent parts of process handling.. */ -#include <linux/stackprotector.h> #include <linux/cpu.h> #include <linux/errno.h> #include <linux/sched.h> @@ -32,17 +31,15 @@ #include <linux/notifier.h> #include <linux/kprobes.h> #include <linux/kdebug.h> -#include <linux/tick.h> #include <linux/prctl.h> #include <linux/uaccess.h> #include <linux/io.h> #include <linux/ftrace.h> -#include <linux/cpuidle.h> #include <asm/pgtable.h> -#include <asm/system.h> #include <asm/processor.h> #include <asm/i387.h> +#include <asm/fpu-internal.h> #include <asm/mmu_context.h> #include <asm/prctl.h> #include <asm/desc.h> @@ -51,116 +48,11 @@ #include <asm/idle.h> #include <asm/syscalls.h> #include <asm/debugreg.h> -#include <asm/nmi.h> +#include <asm/switch_to.h> asmlinkage extern void ret_from_fork(void); -DEFINE_PER_CPU(unsigned long, old_rsp); -static DEFINE_PER_CPU(unsigned char, is_idle); - -static ATOMIC_NOTIFIER_HEAD(idle_notifier); - -void idle_notifier_register(struct notifier_block *n) -{ - atomic_notifier_chain_register(&idle_notifier, n); -} -EXPORT_SYMBOL_GPL(idle_notifier_register); - -void idle_notifier_unregister(struct notifier_block *n) -{ - atomic_notifier_chain_unregister(&idle_notifier, n); -} -EXPORT_SYMBOL_GPL(idle_notifier_unregister); - -void enter_idle(void) -{ - percpu_write(is_idle, 1); - atomic_notifier_call_chain(&idle_notifier, IDLE_START, NULL); -} - -static void __exit_idle(void) -{ - if (x86_test_and_clear_bit_percpu(0, is_idle) == 0) - return; - atomic_notifier_call_chain(&idle_notifier, IDLE_END, NULL); -} - -/* Called from interrupts to signify idle end */ -void exit_idle(void) -{ - /* idle loop has pid 0 */ - if (current->pid) - return; - __exit_idle(); -} - -#ifndef CONFIG_SMP -static inline void play_dead(void) -{ - BUG(); -} -#endif - -/* - * The idle thread. There's no useful work to be - * done, so just try to conserve power and have a - * low exit latency (ie sit in a loop waiting for - * somebody to say that they'd like to reschedule) - */ -void cpu_idle(void) -{ - current_thread_info()->status |= TS_POLLING; - - /* - * If we're the non-boot CPU, nothing set the stack canary up - * for us. CPU0 already has it initialized but no harm in - * doing it again. This is a good place for updating it, as - * we wont ever return from this function (so the invalid - * canaries already on the stack wont ever trigger). - */ - boot_init_stack_canary(); - - /* endless idle loop with no priority at all */ - while (1) { - tick_nohz_idle_enter(); - while (!need_resched()) { - - rmb(); - - if (cpu_is_offline(smp_processor_id())) - play_dead(); - /* - * Idle routines should keep interrupts disabled - * from here on, until they go to idle. - * Otherwise, idle callbacks can misfire. - */ - local_touch_nmi(); - local_irq_disable(); - enter_idle(); - /* Don't trace irqs off for idle */ - stop_critical_timings(); - - /* enter_idle() needs rcu for notifiers */ - rcu_idle_enter(); - - if (cpuidle_idle_call()) - pm_idle(); - - rcu_idle_exit(); - start_critical_timings(); - - /* In many cases the interrupt that ended idle - has already called exit_idle. But some idle - loops can be woken up without interrupt. */ - __exit_idle(); - } - - tick_nohz_idle_exit(); - preempt_enable_no_resched(); - schedule(); - preempt_disable(); - } -} +__visible DEFINE_PER_CPU(unsigned long, old_rsp); /* Prints also some state that isn't saved in the pt_regs */ void __show_regs(struct pt_regs *regs, int all) @@ -170,9 +62,8 @@ void __show_regs(struct pt_regs *regs, int all) unsigned int fsindex, gsindex; unsigned int ds, cs, es; - show_regs_common(); printk(KERN_DEFAULT "RIP: %04lx:[<%016lx>] ", regs->cs & 0xffff, regs->ip); - printk_address(regs->ip, 1); + printk_address(regs->ip); printk(KERN_DEFAULT "RSP: %04lx:%016lx EFLAGS: %08lx\n", regs->ss, regs->sp, regs->flags); printk(KERN_DEFAULT "RAX: %016lx RBX: %016lx RCX: %016lx\n", @@ -214,21 +105,28 @@ void __show_regs(struct pt_regs *regs, int all) get_debugreg(d0, 0); get_debugreg(d1, 1); get_debugreg(d2, 2); - printk(KERN_DEFAULT "DR0: %016lx DR1: %016lx DR2: %016lx\n", d0, d1, d2); get_debugreg(d3, 3); get_debugreg(d6, 6); get_debugreg(d7, 7); + + /* Only print out debug registers if they are in their non-default state. */ + if ((d0 == 0) && (d1 == 0) && (d2 == 0) && (d3 == 0) && + (d6 == DR6_RESERVED) && (d7 == 0x400)) + return; + + printk(KERN_DEFAULT "DR0: %016lx DR1: %016lx DR2: %016lx\n", d0, d1, d2); printk(KERN_DEFAULT "DR3: %016lx DR6: %016lx DR7: %016lx\n", d3, d6, d7); + } void release_thread(struct task_struct *dead_task) { if (dead_task->mm) { if (dead_task->mm->context.size) { - printk("WARNING: dead process %8s still has LDT? <%p/%d>\n", - dead_task->comm, - dead_task->mm->context.ldt, - dead_task->mm->context.size); + pr_warn("WARNING: dead process %s still has LDT? <%p/%d>\n", + dead_task->comm, + dead_task->mm->context.ldt, + dead_task->mm->context.size); BUG(); } } @@ -253,39 +151,19 @@ static inline u32 read_32bit_tls(struct task_struct *t, int tls) return get_desc_base(&t->thread.tls_array[tls]); } -/* - * This gets called before we allocate a new thread and copy - * the current task into it. - */ -void prepare_to_copy(struct task_struct *tsk) -{ - unlazy_fpu(tsk); -} - int copy_thread(unsigned long clone_flags, unsigned long sp, - unsigned long unused, - struct task_struct *p, struct pt_regs *regs) + unsigned long arg, struct task_struct *p) { int err; struct pt_regs *childregs; struct task_struct *me = current; - childregs = ((struct pt_regs *) - (THREAD_SIZE + task_stack_page(p))) - 1; - *childregs = *regs; - - childregs->ax = 0; - if (user_mode(regs)) - childregs->sp = sp; - else - childregs->sp = (unsigned long)childregs; - + p->thread.sp0 = (unsigned long)task_stack_page(p) + THREAD_SIZE; + childregs = task_pt_regs(p); p->thread.sp = (unsigned long) childregs; - p->thread.sp0 = (unsigned long) (childregs+1); p->thread.usersp = me->thread.usersp; - set_tsk_thread_flag(p, TIF_FORK); - + p->thread.fpu_counter = 0; p->thread.io_bitmap_ptr = NULL; savesegment(gs, p->thread.gsindex); @@ -294,6 +172,25 @@ int copy_thread(unsigned long clone_flags, unsigned long sp, p->thread.fs = p->thread.fsindex ? 0 : me->thread.fs; savesegment(es, p->thread.es); savesegment(ds, p->thread.ds); + memset(p->thread.ptrace_bps, 0, sizeof(p->thread.ptrace_bps)); + + if (unlikely(p->flags & PF_KTHREAD)) { + /* kernel thread */ + memset(childregs, 0, sizeof(struct pt_regs)); + childregs->sp = (unsigned long)childregs; + childregs->ss = __KERNEL_DS; + childregs->bx = sp; /* function */ + childregs->bp = arg; + childregs->orig_ax = -1; + childregs->cs = __KERNEL_CS | get_kernel_rpl(); + childregs->flags = X86_EFLAGS_IF | X86_EFLAGS_FIXED; + return 0; + } + *childregs = *current_pt_regs(); + + childregs->ax = 0; + if (sp) + childregs->sp = sp; err = -ENOMEM; memset(p->thread.ptrace_bps, 0, sizeof(p->thread.ptrace_bps)); @@ -341,16 +238,13 @@ start_thread_common(struct pt_regs *regs, unsigned long new_ip, loadsegment(es, _ds); loadsegment(ds, _ds); load_gs_index(0); + current->thread.usersp = new_sp; regs->ip = new_ip; regs->sp = new_sp; - percpu_write(old_rsp, new_sp); + this_cpu_write(old_rsp, new_sp); regs->cs = _cs; regs->ss = _ss; regs->flags = X86_EFLAGS_IF; - /* - * Free the old FP and other extended state - */ - free_thread_xstate(current); } void @@ -364,7 +258,9 @@ start_thread(struct pt_regs *regs, unsigned long new_ip, unsigned long new_sp) void start_thread_ia32(struct pt_regs *regs, u32 new_ip, u32 new_sp) { start_thread_common(regs, new_ip, new_sp, - __USER32_CS, __USER32_DS, __USER32_DS); + test_thread_flag(TIF_X32) + ? __USER_CS : __USER32_CS, + __USER_DS, __USER_DS); } #endif @@ -378,7 +274,7 @@ void start_thread_ia32(struct pt_regs *regs, u32 new_ip, u32 new_sp) * Kprobes not supported here. Set the probe on schedule instead. * Function graph tracer not supported too. */ -__notrace_funcgraph struct task_struct * +__visible __notrace_funcgraph struct task_struct * __switch_to(struct task_struct *prev_p, struct task_struct *next_p) { struct thread_struct *prev = &prev_p->thread; @@ -388,7 +284,7 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) unsigned fsindex, gsindex; fpu_switch_t fpu; - fpu = switch_fpu_prepare(prev_p, next_p); + fpu = switch_fpu_prepare(prev_p, next_p, cpu); /* * Reload esp0, LDT and the page table pointer: @@ -463,11 +359,19 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) /* * Switch the PDA and FPU contexts. */ - prev->usersp = percpu_read(old_rsp); - percpu_write(old_rsp, next->usersp); - percpu_write(current_task, next_p); + prev->usersp = this_cpu_read(old_rsp); + this_cpu_write(old_rsp, next->usersp); + this_cpu_write(current_task, next_p); + + /* + * If it were not for PREEMPT_ACTIVE we could guarantee that the + * preempt_count of all tasks was equal here and this would not be + * needed. + */ + task_thread_info(prev_p)->saved_preempt_count = this_cpu_read(__preempt_count); + this_cpu_write(__preempt_count, task_thread_info(next_p)->saved_preempt_count); - percpu_write(kernel_stack, + this_cpu_write(kernel_stack, (unsigned long)task_stack_page(next_p) + THREAD_SIZE - KERNEL_STACK_OFFSET); @@ -487,6 +391,8 @@ void set_personality_64bit(void) /* Make sure to be in 64bit mode */ clear_thread_flag(TIF_IA32); + clear_thread_flag(TIF_ADDR32); + clear_thread_flag(TIF_X32); /* Ensure the corresponding mm is not marked. */ if (current->mm) @@ -499,21 +405,34 @@ void set_personality_64bit(void) current->personality &= ~READ_IMPLIES_EXEC; } -void set_personality_ia32(void) +void set_personality_ia32(bool x32) { /* inherit personality from parent */ /* Make sure to be in 32bit mode */ - set_thread_flag(TIF_IA32); - current->personality |= force_personality32; + set_thread_flag(TIF_ADDR32); /* Mark the associated mm as containing 32-bit tasks. */ - if (current->mm) - current->mm->context.ia32_compat = 1; - - /* Prepare the first "return" to user space */ - current_thread_info()->status |= TS_COMPAT; + if (x32) { + clear_thread_flag(TIF_IA32); + set_thread_flag(TIF_X32); + if (current->mm) + current->mm->context.ia32_compat = TIF_X32; + current->personality &= ~READ_IMPLIES_EXEC; + /* is_compat_task() uses the presence of the x32 + syscall bit flag to determine compat status */ + current_thread_info()->status &= ~TS_COMPAT; + } else { + set_thread_flag(TIF_IA32); + clear_thread_flag(TIF_X32); + if (current->mm) + current->mm->context.ia32_compat = TIF_IA32; + current->personality |= force_personality32; + /* Prepare the first "return" to user space */ + current_thread_info()->status |= TS_COMPAT; + } } +EXPORT_SYMBOL_GPL(set_personality_ia32); unsigned long get_wchan(struct task_struct *p) { @@ -565,7 +484,7 @@ long do_arch_prctl(struct task_struct *task, int code, unsigned long addr) task->thread.gs = addr; if (doit) { load_gs_index(0); - ret = checking_wrmsrl(MSR_KERNEL_GS_BASE, addr); + ret = wrmsrl_safe(MSR_KERNEL_GS_BASE, addr); } } put_cpu(); @@ -593,7 +512,7 @@ long do_arch_prctl(struct task_struct *task, int code, unsigned long addr) /* set the selector to 0 to not confuse __switch_to */ loadsegment(fs, 0); - ret = checking_wrmsrl(MSR_FS_BASE, addr); + ret = wrmsrl_safe(MSR_FS_BASE, addr); } } put_cpu(); diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c index 50267386b76..678c0ada3b3 100644 --- a/arch/x86/kernel/ptrace.c +++ b/arch/x86/kernel/ptrace.c @@ -21,18 +21,22 @@ #include <linux/signal.h> #include <linux/perf_event.h> #include <linux/hw_breakpoint.h> +#include <linux/rcupdate.h> +#include <linux/export.h> +#include <linux/context_tracking.h> #include <asm/uaccess.h> #include <asm/pgtable.h> -#include <asm/system.h> #include <asm/processor.h> #include <asm/i387.h> +#include <asm/fpu-internal.h> #include <asm/debugreg.h> #include <asm/ldt.h> #include <asm/desc.h> #include <asm/prctl.h> #include <asm/proto.h> #include <asm/hw_breakpoint.h> +#include <asm/traps.h> #include "tls.h" @@ -164,6 +168,35 @@ static inline bool invalid_selector(u16 value) #define FLAG_MASK FLAG_MASK_32 +/* + * X86_32 CPUs don't save ss and esp if the CPU is already in kernel mode + * when it traps. The previous stack will be directly underneath the saved + * registers, and 'sp/ss' won't even have been saved. Thus the '®s->sp'. + * + * Now, if the stack is empty, '®s->sp' is out of range. In this + * case we try to take the previous stack. To always return a non-null + * stack pointer we fall back to regs as stack if no previous stack + * exists. + * + * This is valid only for kernel mode traps. + */ +unsigned long kernel_stack_pointer(struct pt_regs *regs) +{ + unsigned long context = (unsigned long)regs & ~(THREAD_SIZE - 1); + unsigned long sp = (unsigned long)®s->sp; + u32 *prev_esp; + + if (context == (sp & ~(THREAD_SIZE - 1))) + return sp; + + prev_esp = (u32 *)(context); + if (prev_esp) + return (unsigned long)prev_esp; + + return (unsigned long)regs; +} +EXPORT_SYMBOL_GPL(kernel_stack_pointer); + static unsigned long *pt_regs_access(struct pt_regs *regs, unsigned long regno) { BUILD_BUG_ON(offsetof(struct pt_regs, bx) != 0); @@ -568,30 +601,48 @@ static unsigned long ptrace_get_dr7(struct perf_event *bp[]) return dr7; } -static int -ptrace_modify_breakpoint(struct perf_event *bp, int len, int type, - struct task_struct *tsk, int disabled) +static int ptrace_fill_bp_fields(struct perf_event_attr *attr, + int len, int type, bool disabled) +{ + int err, bp_len, bp_type; + + err = arch_bp_generic_fields(len, type, &bp_len, &bp_type); + if (!err) { + attr->bp_len = bp_len; + attr->bp_type = bp_type; + attr->disabled = disabled; + } + + return err; +} + +static struct perf_event * +ptrace_register_breakpoint(struct task_struct *tsk, int len, int type, + unsigned long addr, bool disabled) { - int err; - int gen_len, gen_type; struct perf_event_attr attr; + int err; - /* - * We should have at least an inactive breakpoint at this - * slot. It means the user is writing dr7 without having - * written the address register first - */ - if (!bp) - return -EINVAL; + ptrace_breakpoint_init(&attr); + attr.bp_addr = addr; - err = arch_bp_generic_fields(len, type, &gen_len, &gen_type); + err = ptrace_fill_bp_fields(&attr, len, type, disabled); if (err) - return err; + return ERR_PTR(err); + + return register_user_hw_breakpoint(&attr, ptrace_triggered, + NULL, tsk); +} + +static int ptrace_modify_breakpoint(struct perf_event *bp, int len, int type, + int disabled) +{ + struct perf_event_attr attr = bp->attr; + int err; - attr = bp->attr; - attr.bp_len = gen_len; - attr.bp_type = gen_type; - attr.disabled = disabled; + err = ptrace_fill_bp_fields(&attr, len, type, disabled); + if (err) + return err; return modify_user_hw_breakpoint(bp, &attr); } @@ -601,67 +652,50 @@ ptrace_modify_breakpoint(struct perf_event *bp, int len, int type, */ static int ptrace_write_dr7(struct task_struct *tsk, unsigned long data) { - struct thread_struct *thread = &(tsk->thread); + struct thread_struct *thread = &tsk->thread; unsigned long old_dr7; - int i, orig_ret = 0, rc = 0; - int enabled, second_pass = 0; - unsigned len, type; - struct perf_event *bp; - - if (ptrace_get_breakpoints(tsk) < 0) - return -ESRCH; + bool second_pass = false; + int i, rc, ret = 0; data &= ~DR_CONTROL_RESERVED; old_dr7 = ptrace_get_dr7(thread->ptrace_bps); + restore: - /* - * Loop through all the hardware breakpoints, making the - * appropriate changes to each. - */ + rc = 0; for (i = 0; i < HBP_NUM; i++) { - enabled = decode_dr7(data, i, &len, &type); - bp = thread->ptrace_bps[i]; - - if (!enabled) { - if (bp) { - /* - * Don't unregister the breakpoints right-away, - * unless all register_user_hw_breakpoint() - * requests have succeeded. This prevents - * any window of opportunity for debug - * register grabbing by other users. - */ - if (!second_pass) - continue; - - rc = ptrace_modify_breakpoint(bp, len, type, - tsk, 1); - if (rc) - break; + unsigned len, type; + bool disabled = !decode_dr7(data, i, &len, &type); + struct perf_event *bp = thread->ptrace_bps[i]; + + if (!bp) { + if (disabled) + continue; + + bp = ptrace_register_breakpoint(tsk, + len, type, 0, disabled); + if (IS_ERR(bp)) { + rc = PTR_ERR(bp); + break; } + + thread->ptrace_bps[i] = bp; continue; } - rc = ptrace_modify_breakpoint(bp, len, type, tsk, 0); + rc = ptrace_modify_breakpoint(bp, len, type, disabled); if (rc) break; } - /* - * Make a second pass to free the remaining unused breakpoints - * or to restore the original breakpoints if an error occurred. - */ - if (!second_pass) { - second_pass = 1; - if (rc < 0) { - orig_ret = rc; - data = old_dr7; - } + + /* Restore if the first pass failed, second_pass shouldn't fail. */ + if (rc && !WARN_ON(second_pass)) { + ret = rc; + data = old_dr7; + second_pass = true; goto restore; } - ptrace_put_breakpoints(tsk); - - return ((orig_ret < 0) ? orig_ret : rc); + return ret; } /* @@ -669,25 +703,17 @@ restore: */ static unsigned long ptrace_get_debugreg(struct task_struct *tsk, int n) { - struct thread_struct *thread = &(tsk->thread); + struct thread_struct *thread = &tsk->thread; unsigned long val = 0; if (n < HBP_NUM) { - struct perf_event *bp; + struct perf_event *bp = thread->ptrace_bps[n]; - if (ptrace_get_breakpoints(tsk) < 0) - return -ESRCH; - - bp = thread->ptrace_bps[n]; - if (!bp) - val = 0; - else + if (bp) val = bp->hw.info.address; - - ptrace_put_breakpoints(tsk); } else if (n == 6) { val = thread->debugreg6; - } else if (n == 7) { + } else if (n == 7) { val = thread->ptrace_dr7; } return val; @@ -696,29 +722,14 @@ static unsigned long ptrace_get_debugreg(struct task_struct *tsk, int n) static int ptrace_set_breakpoint_addr(struct task_struct *tsk, int nr, unsigned long addr) { - struct perf_event *bp; struct thread_struct *t = &tsk->thread; - struct perf_event_attr attr; + struct perf_event *bp = t->ptrace_bps[nr]; int err = 0; - if (ptrace_get_breakpoints(tsk) < 0) - return -ESRCH; - - if (!t->ptrace_bps[nr]) { - ptrace_breakpoint_init(&attr); - /* - * Put stub len and type to register (reserve) an inactive but - * correct bp - */ - attr.bp_addr = addr; - attr.bp_len = HW_BREAKPOINT_LEN_1; - attr.bp_type = HW_BREAKPOINT_W; - attr.disabled = 1; - - bp = register_user_hw_breakpoint(&attr, ptrace_triggered, - NULL, tsk); - + if (!bp) { /* + * Put stub len and type to create an inactive but correct bp. + * * CHECKME: the previous code returned -EIO if the addr wasn't * a valid task virtual addr. The new one will return -EINVAL in * this case. @@ -727,22 +738,20 @@ static int ptrace_set_breakpoint_addr(struct task_struct *tsk, int nr, * writing for the user. And anyway this is the previous * behaviour. */ - if (IS_ERR(bp)) { + bp = ptrace_register_breakpoint(tsk, + X86_BREAKPOINT_LEN_1, X86_BREAKPOINT_WRITE, + addr, true); + if (IS_ERR(bp)) err = PTR_ERR(bp); - goto put; - } - - t->ptrace_bps[nr] = bp; + else + t->ptrace_bps[nr] = bp; } else { - bp = t->ptrace_bps[nr]; + struct perf_event_attr attr = bp->attr; - attr = bp->attr; attr.bp_addr = addr; err = modify_user_hw_breakpoint(bp, &attr); } -put: - ptrace_put_breakpoints(tsk); return err; } @@ -752,30 +761,20 @@ put: static int ptrace_set_debugreg(struct task_struct *tsk, int n, unsigned long val) { - struct thread_struct *thread = &(tsk->thread); - int rc = 0; - + struct thread_struct *thread = &tsk->thread; /* There are no DR4 or DR5 registers */ - if (n == 4 || n == 5) - return -EIO; + int rc = -EIO; - if (n == 6) { - thread->debugreg6 = val; - goto ret_path; - } if (n < HBP_NUM) { rc = ptrace_set_breakpoint_addr(tsk, n, val); - if (rc) - return rc; - } - /* All that's left is DR7 */ - if (n == 7) { + } else if (n == 6) { + thread->debugreg6 = val; + rc = 0; + } else if (n == 7) { rc = ptrace_write_dr7(tsk, val); if (!rc) thread->ptrace_dr7 = val; } - -ret_path: return rc; } @@ -1130,6 +1129,94 @@ static int genregs32_set(struct task_struct *target, return ret; } +#ifdef CONFIG_X86_X32_ABI +static long x32_arch_ptrace(struct task_struct *child, + compat_long_t request, compat_ulong_t caddr, + compat_ulong_t cdata) +{ + unsigned long addr = caddr; + unsigned long data = cdata; + void __user *datap = compat_ptr(data); + int ret; + + switch (request) { + /* Read 32bits at location addr in the USER area. Only allow + to return the lower 32bits of segment and debug registers. */ + case PTRACE_PEEKUSR: { + u32 tmp; + + ret = -EIO; + if ((addr & (sizeof(data) - 1)) || addr >= sizeof(struct user) || + addr < offsetof(struct user_regs_struct, cs)) + break; + + tmp = 0; /* Default return condition */ + if (addr < sizeof(struct user_regs_struct)) + tmp = getreg(child, addr); + else if (addr >= offsetof(struct user, u_debugreg[0]) && + addr <= offsetof(struct user, u_debugreg[7])) { + addr -= offsetof(struct user, u_debugreg[0]); + tmp = ptrace_get_debugreg(child, addr / sizeof(data)); + } + ret = put_user(tmp, (__u32 __user *)datap); + break; + } + + /* Write the word at location addr in the USER area. Only allow + to update segment and debug registers with the upper 32bits + zero-extended. */ + case PTRACE_POKEUSR: + ret = -EIO; + if ((addr & (sizeof(data) - 1)) || addr >= sizeof(struct user) || + addr < offsetof(struct user_regs_struct, cs)) + break; + + if (addr < sizeof(struct user_regs_struct)) + ret = putreg(child, addr, data); + else if (addr >= offsetof(struct user, u_debugreg[0]) && + addr <= offsetof(struct user, u_debugreg[7])) { + addr -= offsetof(struct user, u_debugreg[0]); + ret = ptrace_set_debugreg(child, + addr / sizeof(data), data); + } + break; + + case PTRACE_GETREGS: /* Get all gp regs from the child. */ + return copy_regset_to_user(child, + task_user_regset_view(current), + REGSET_GENERAL, + 0, sizeof(struct user_regs_struct), + datap); + + case PTRACE_SETREGS: /* Set all gp regs in the child. */ + return copy_regset_from_user(child, + task_user_regset_view(current), + REGSET_GENERAL, + 0, sizeof(struct user_regs_struct), + datap); + + case PTRACE_GETFPREGS: /* Get the child FPU state. */ + return copy_regset_to_user(child, + task_user_regset_view(current), + REGSET_FP, + 0, sizeof(struct user_i387_struct), + datap); + + case PTRACE_SETFPREGS: /* Set the child FPU state. */ + return copy_regset_from_user(child, + task_user_regset_view(current), + REGSET_FP, + 0, sizeof(struct user_i387_struct), + datap); + + default: + return compat_ptrace_request(child, request, addr, data); + } + + return ret; +} +#endif + long compat_arch_ptrace(struct task_struct *child, compat_long_t request, compat_ulong_t caddr, compat_ulong_t cdata) { @@ -1139,6 +1226,11 @@ long compat_arch_ptrace(struct task_struct *child, compat_long_t request, int ret; __u32 val; +#ifdef CONFIG_X86_X32_ABI + if (!is_ia32_task()) + return x32_arch_ptrace(child, request, caddr, cdata); +#endif + switch (request) { case PTRACE_PEEKUSR: ret = getreg32(child, addr, &val); @@ -1238,9 +1330,6 @@ static const struct user_regset_view user_x86_64_view = { #define genregs32_get genregs_get #define genregs32_set genregs_set -#define user_i387_ia32_struct user_i387_struct -#define user32_fxsr_struct user_fxsr_struct - #endif /* CONFIG_X86_64 */ #if defined CONFIG_X86_32 || defined CONFIG_IA32_EMULATION @@ -1326,7 +1415,7 @@ static void fill_sigtrap_info(struct task_struct *tsk, int error_code, int si_code, struct siginfo *info) { - tsk->thread.trap_no = 1; + tsk->thread.trap_nr = X86_TRAP_DB; tsk->thread.error_code = error_code; memset(info, 0, sizeof(*info)); @@ -1369,6 +1458,8 @@ long syscall_trace_enter(struct pt_regs *regs) { long ret = 0; + user_exit(); + /* * If we stepped into a sysenter/syscall insn, it trapped in * kernel mode; do_debug() cleared TF and set TIF_SINGLESTEP. @@ -1380,7 +1471,11 @@ long syscall_trace_enter(struct pt_regs *regs) regs->flags |= X86_EFLAGS_TF; /* do the secure computing check first */ - secure_computing(regs->orig_ax); + if (secure_computing(regs->orig_ax)) { + /* seccomp failures shouldn't expose any additional code. */ + ret = -1L; + goto out; + } if (unlikely(test_thread_flag(TIF_SYSCALL_EMU))) ret = -1L; @@ -1405,6 +1500,7 @@ long syscall_trace_enter(struct pt_regs *regs) regs->dx, regs->r10); #endif +out: return ret ?: regs->orig_ax; } @@ -1412,6 +1508,13 @@ void syscall_trace_leave(struct pt_regs *regs) { bool step; + /* + * We may come here right after calling schedule_user() + * or do_notify_resume(), in which case we can be in RCU + * user mode. + */ + user_exit(); + audit_syscall_exit(regs); if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) @@ -1427,4 +1530,6 @@ void syscall_trace_leave(struct pt_regs *regs) !test_thread_flag(TIF_SYSCALL_EMU); if (step || test_thread_flag(TIF_SYSCALL_TRACE)) tracehook_report_syscall_exit(regs, step); + + user_enter(); } diff --git a/arch/x86/kernel/pvclock.c b/arch/x86/kernel/pvclock.c index 42eb3300dfc..2f355d229a5 100644 --- a/arch/x86/kernel/pvclock.c +++ b/arch/x86/kernel/pvclock.c @@ -17,23 +17,13 @@ #include <linux/kernel.h> #include <linux/percpu.h> +#include <linux/notifier.h> +#include <linux/sched.h> +#include <linux/gfp.h> +#include <linux/bootmem.h> +#include <asm/fixmap.h> #include <asm/pvclock.h> -/* - * These are perodically updated - * xen: magic shared_info page - * kvm: gpa registered via msr - * and then copied here. - */ -struct pvclock_shadow_time { - u64 tsc_timestamp; /* TSC at last update of time vals. */ - u64 system_timestamp; /* Time, in nanosecs, since boot. */ - u32 tsc_to_nsec_mul; - int tsc_shift; - u32 version; - u8 flags; -}; - static u8 valid_flags __read_mostly = 0; void pvclock_set_flags(u8 flags) @@ -41,34 +31,6 @@ void pvclock_set_flags(u8 flags) valid_flags = flags; } -static u64 pvclock_get_nsec_offset(struct pvclock_shadow_time *shadow) -{ - u64 delta = native_read_tsc() - shadow->tsc_timestamp; - return pvclock_scale_delta(delta, shadow->tsc_to_nsec_mul, - shadow->tsc_shift); -} - -/* - * Reads a consistent set of time-base values from hypervisor, - * into a shadow data area. - */ -static unsigned pvclock_get_time_values(struct pvclock_shadow_time *dst, - struct pvclock_vcpu_time_info *src) -{ - do { - dst->version = src->version; - rmb(); /* fetch version before data */ - dst->tsc_timestamp = src->tsc_timestamp; - dst->system_timestamp = src->system_time; - dst->tsc_to_nsec_mul = src->tsc_to_system_mul; - dst->tsc_shift = src->tsc_shift; - dst->flags = src->flags; - rmb(); /* test version after fetching data */ - } while ((src->version & 1) || (dst->version != src->version)); - - return dst->version; -} - unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src) { u64 pv_tsc_khz = 1000000ULL << 32; @@ -81,6 +43,14 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src) return pv_tsc_khz; } +void pvclock_touch_watchdogs(void) +{ + touch_softlockup_watchdog_sync(); + clocksource_touch_watchdog(); + rcu_cpu_stall_reset(); + reset_hung_task_detector(); +} + static atomic64_t last_value = ATOMIC64_INIT(0); void pvclock_resume(void) @@ -88,23 +58,37 @@ void pvclock_resume(void) atomic64_set(&last_value, 0); } +u8 pvclock_read_flags(struct pvclock_vcpu_time_info *src) +{ + unsigned version; + cycle_t ret; + u8 flags; + + do { + version = __pvclock_read_cycles(src, &ret, &flags); + } while ((src->version & 1) || version != src->version); + + return flags & valid_flags; +} + cycle_t pvclock_clocksource_read(struct pvclock_vcpu_time_info *src) { - struct pvclock_shadow_time shadow; unsigned version; - cycle_t ret, offset; + cycle_t ret; u64 last; + u8 flags; do { - version = pvclock_get_time_values(&shadow, src); - barrier(); - offset = pvclock_get_nsec_offset(&shadow); - ret = shadow.system_timestamp + offset; - barrier(); - } while (version != src->version); + version = __pvclock_read_cycles(src, &ret, &flags); + } while ((src->version & 1) || version != src->version); + + if (unlikely((flags & PVCLOCK_GUEST_STOPPED) != 0)) { + src->flags &= ~PVCLOCK_GUEST_STOPPED; + pvclock_touch_watchdogs(); + } if ((valid_flags & PVCLOCK_TSC_STABLE_BIT) && - (shadow.flags & PVCLOCK_TSC_STABLE_BIT)) + (flags & PVCLOCK_TSC_STABLE_BIT)) return ret; /* @@ -156,3 +140,27 @@ void pvclock_read_wallclock(struct pvclock_wall_clock *wall_clock, set_normalized_timespec(ts, now.tv_sec, now.tv_nsec); } + +#ifdef CONFIG_X86_64 +/* + * Initialize the generic pvclock vsyscall state. This will allocate + * a/some page(s) for the per-vcpu pvclock information, set up a + * fixmap mapping for the page(s) + */ + +int __init pvclock_init_vsyscall(struct pvclock_vsyscall_time_info *i, + int size) +{ + int idx; + + WARN_ON (size != PVCLOCK_VSYSCALL_NR_PAGES*PAGE_SIZE); + + for (idx = 0; idx <= (PVCLOCK_FIXMAP_END-PVCLOCK_FIXMAP_BEGIN); idx++) { + __set_fixmap(PVCLOCK_FIXMAP_BEGIN + idx, + __pa(i) + (idx*PAGE_SIZE), + PAGE_KERNEL_VVAR); + } + + return 0; +} +#endif diff --git a/arch/x86/kernel/quirks.c b/arch/x86/kernel/quirks.c index 03920a15a63..ff898bbf579 100644 --- a/arch/x86/kernel/quirks.c +++ b/arch/x86/kernel/quirks.c @@ -8,7 +8,7 @@ #if defined(CONFIG_X86_IO_APIC) && defined(CONFIG_SMP) && defined(CONFIG_PCI) -static void __devinit quirk_intel_irqbalance(struct pci_dev *dev) +static void quirk_intel_irqbalance(struct pci_dev *dev) { u8 config; u16 word; @@ -354,18 +354,22 @@ static void ati_force_hpet_resume(void) static u32 ati_ixp4x0_rev(struct pci_dev *dev) { - u32 d; - u8 b; + int err = 0; + u32 d = 0; + u8 b = 0; - pci_read_config_byte(dev, 0xac, &b); + err = pci_read_config_byte(dev, 0xac, &b); b &= ~(1<<5); - pci_write_config_byte(dev, 0xac, b); - pci_read_config_dword(dev, 0x70, &d); + err |= pci_write_config_byte(dev, 0xac, b); + err |= pci_read_config_dword(dev, 0x70, &d); d |= 1<<8; - pci_write_config_dword(dev, 0x70, d); - pci_read_config_dword(dev, 0x8, &d); + err |= pci_write_config_dword(dev, 0x70, d); + err |= pci_read_config_dword(dev, 0x8, &d); d &= 0xff; dev_printk(KERN_DEBUG, &dev->dev, "SB4X0 revision 0x%x\n", d); + + WARN_ON_ONCE(err); + return d; } @@ -512,7 +516,7 @@ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_ATI, PCI_DEVICE_ID_ATI_SBX00_SMBUS, #if defined(CONFIG_PCI) && defined(CONFIG_NUMA) /* Set correct numa_node information for AMD NB functions */ -static void __init quirk_amd_nb_node(struct pci_dev *dev) +static void quirk_amd_nb_node(struct pci_dev *dev) { struct pci_dev *nb_ht; unsigned int devfn; @@ -525,7 +529,7 @@ static void __init quirk_amd_nb_node(struct pci_dev *dev) return; pci_read_config_dword(nb_ht, 0x60, &val); - node = val & 7; + node = pcibus_to_node(dev->bus) | (val & 7); /* * Some hardware may return an invalid node ID, * so check it first: @@ -567,3 +571,40 @@ DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_15H_NB_F5, quirk_amd_nb_node); #endif + +#ifdef CONFIG_PCI +/* + * Processor does not ensure DRAM scrub read/write sequence + * is atomic wrt accesses to CC6 save state area. Therefore + * if a concurrent scrub read/write access is to same address + * the entry may appear as if it is not written. This quirk + * applies to Fam16h models 00h-0Fh + * + * See "Revision Guide" for AMD F16h models 00h-0fh, + * document 51810 rev. 3.04, Nov 2013 + */ +static void amd_disable_seq_and_redirect_scrub(struct pci_dev *dev) +{ + u32 val; + + /* + * Suggested workaround: + * set D18F3x58[4:0] = 00h and set D18F3x5C[0] = 0b + */ + pci_read_config_dword(dev, 0x58, &val); + if (val & 0x1F) { + val &= ~(0x1F); + pci_write_config_dword(dev, 0x58, val); + } + + pci_read_config_dword(dev, 0x5C, &val); + if (val & BIT(0)) { + val &= ~BIT(0); + pci_write_config_dword(dev, 0x5c, val); + } +} + +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_16H_NB_F3, + amd_disable_seq_and_redirect_scrub); + +#endif diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c index d840e69a853..52b1157c53e 100644 --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -1,3 +1,5 @@ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/module.h> #include <linux/reboot.h> #include <linux/init.h> @@ -20,13 +22,12 @@ #include <asm/virtext.h> #include <asm/cpu.h> #include <asm/nmi.h> +#include <asm/smp.h> -#ifdef CONFIG_X86_32 -# include <linux/ctype.h> -# include <linux/mc146818rtc.h> -#else -# include <asm/x86_init.h> -#endif +#include <linux/ctype.h> +#include <linux/mc146818rtc.h> +#include <asm/realmode.h> +#include <asm/x86_init.h> /* * Power off function, if any @@ -35,23 +36,9 @@ void (*pm_power_off)(void); EXPORT_SYMBOL(pm_power_off); static const struct desc_ptr no_idt = {}; -static int reboot_mode; -enum reboot_type reboot_type = BOOT_ACPI; -int reboot_force; - -/* This variable is used privately to keep track of whether or not - * reboot_type is still set to its default value (i.e., reboot= hasn't - * been set on the command line). This is needed so that we can - * suppress DMI scanning for reboot quirks. Without it, it's - * impossible to override a faulty reboot quirk without recompiling. - */ -static int reboot_default = 1; - -#if defined(CONFIG_X86_32) && defined(CONFIG_SMP) -static int reboot_cpu = -1; -#endif -/* This is set if we need to go through the 'emergency' path. +/* + * This is set if we need to go through the 'emergency' path. * When machine_emergency_restart() is called, we may be on * an inconsistent state and won't be able to do a clean cleanup */ @@ -60,78 +47,6 @@ static int reboot_emergency; /* This is set by the PCI code if either type 1 or type 2 PCI is detected */ bool port_cf9_safe = false; -/* reboot=b[ios] | s[mp] | t[riple] | k[bd] | e[fi] [, [w]arm | [c]old] | p[ci] - warm Don't set the cold reboot flag - cold Set the cold reboot flag - bios Reboot by jumping through the BIOS (only for X86_32) - smp Reboot by executing reset on BSP or other CPU (only for X86_32) - triple Force a triple fault (init) - kbd Use the keyboard controller. cold reset (default) - acpi Use the RESET_REG in the FADT - efi Use efi reset_system runtime service - pci Use the so-called "PCI reset register", CF9 - force Avoid anything that could hang. - */ -static int __init reboot_setup(char *str) -{ - for (;;) { - /* Having anything passed on the command line via - * reboot= will cause us to disable DMI checking - * below. - */ - reboot_default = 0; - - switch (*str) { - case 'w': - reboot_mode = 0x1234; - break; - - case 'c': - reboot_mode = 0; - break; - -#ifdef CONFIG_X86_32 -#ifdef CONFIG_SMP - case 's': - if (isdigit(*(str+1))) { - reboot_cpu = (int) (*(str+1) - '0'); - if (isdigit(*(str+2))) - reboot_cpu = reboot_cpu*10 + (int)(*(str+2) - '0'); - } - /* we will leave sorting out the final value - when we are ready to reboot, since we might not - have detected BSP APIC ID or smp_num_cpu */ - break; -#endif /* CONFIG_SMP */ - - case 'b': -#endif - case 'a': - case 'k': - case 't': - case 'e': - case 'p': - reboot_type = *str; - break; - - case 'f': - reboot_force = 1; - break; - } - - str = strchr(str, ','); - if (str) - str++; - else - break; - } - return 1; -} - -__setup("reboot=", reboot_setup); - - -#ifdef CONFIG_X86_32 /* * Reboot options and system auto-detection code provided by * Dell Inc. so their systems "just work". :-) @@ -145,7 +60,64 @@ static int __init set_bios_reboot(const struct dmi_system_id *d) { if (reboot_type != BOOT_BIOS) { reboot_type = BOOT_BIOS; - printk(KERN_INFO "%s series board detected. Selecting BIOS-method for reboots.\n", d->ident); + pr_info("%s series board detected. Selecting %s-method for reboots.\n", + d->ident, "BIOS"); + } + return 0; +} + +void __noreturn machine_real_restart(unsigned int type) +{ + local_irq_disable(); + + /* + * Write zero to CMOS register number 0x0f, which the BIOS POST + * routine will recognize as telling it to do a proper reboot. (Well + * that's what this book in front of me says -- it may only apply to + * the Phoenix BIOS though, it's not clear). At the same time, + * disable NMIs by setting the top bit in the CMOS address register, + * as we're about to do peculiar things to the CPU. I'm not sure if + * `outb_p' is needed instead of just `outb'. Use it to be on the + * safe side. (Yes, CMOS_WRITE does outb_p's. - Paul G.) + */ + spin_lock(&rtc_lock); + CMOS_WRITE(0x00, 0x8f); + spin_unlock(&rtc_lock); + + /* + * Switch back to the initial page table. + */ +#ifdef CONFIG_X86_32 + load_cr3(initial_page_table); +#else + write_cr3(real_mode_header->trampoline_pgd); +#endif + + /* Jump to the identity-mapped low memory code */ +#ifdef CONFIG_X86_32 + asm volatile("jmpl *%0" : : + "rm" (real_mode_header->machine_real_restart_asm), + "a" (type)); +#else + asm volatile("ljmpl *%0" : : + "m" (real_mode_header->machine_real_restart_asm), + "D" (type)); +#endif + unreachable(); +} +#ifdef CONFIG_APM_MODULE +EXPORT_SYMBOL(machine_real_restart); +#endif + +/* + * Some Apple MacBook and MacBookPro's needs reboot=p to be able to reboot + */ +static int __init set_pci_reboot(const struct dmi_system_id *d) +{ + if (reboot_type != BOOT_CF9_FORCE) { + reboot_type = BOOT_CF9_FORCE; + pr_info("%s series board detected. Selecting %s-method for reboots.\n", + d->ident, "PCI"); } return 0; } @@ -154,337 +126,290 @@ static int __init set_kbd_reboot(const struct dmi_system_id *d) { if (reboot_type != BOOT_KBD) { reboot_type = BOOT_KBD; - printk(KERN_INFO "%s series board detected. Selecting KBD-method for reboot.\n", d->ident); + pr_info("%s series board detected. Selecting %s-method for reboot.\n", + d->ident, "KBD"); } return 0; } +/* + * This is a single dmi_table handling all reboot quirks. + */ static struct dmi_system_id __initdata reboot_dmi_table[] = { - { /* Handle problems with rebooting on Dell E520's */ - .callback = set_bios_reboot, - .ident = "Dell E520", + + /* Acer */ + { /* Handle reboot issue on Acer Aspire one */ + .callback = set_kbd_reboot, + .ident = "Acer Aspire One A110", .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "Dell DM061"), + DMI_MATCH(DMI_SYS_VENDOR, "Acer"), + DMI_MATCH(DMI_PRODUCT_NAME, "AOA110"), }, }, - { /* Handle problems with rebooting on Dell 1300's */ - .callback = set_bios_reboot, - .ident = "Dell PowerEdge 1300", + + /* Apple */ + { /* Handle problems with rebooting on Apple MacBook5 */ + .callback = set_pci_reboot, + .ident = "Apple MacBook5", .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "Dell Computer Corporation"), - DMI_MATCH(DMI_PRODUCT_NAME, "PowerEdge 1300/"), + DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."), + DMI_MATCH(DMI_PRODUCT_NAME, "MacBook5"), }, }, - { /* Handle problems with rebooting on Dell 300's */ - .callback = set_bios_reboot, - .ident = "Dell PowerEdge 300", + { /* Handle problems with rebooting on Apple MacBookPro5 */ + .callback = set_pci_reboot, + .ident = "Apple MacBookPro5", .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "Dell Computer Corporation"), - DMI_MATCH(DMI_PRODUCT_NAME, "PowerEdge 300/"), + DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."), + DMI_MATCH(DMI_PRODUCT_NAME, "MacBookPro5"), }, }, - { /* Handle problems with rebooting on Dell Optiplex 745's SFF*/ + { /* Handle problems with rebooting on Apple Macmini3,1 */ + .callback = set_pci_reboot, + .ident = "Apple Macmini3,1", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."), + DMI_MATCH(DMI_PRODUCT_NAME, "Macmini3,1"), + }, + }, + { /* Handle problems with rebooting on the iMac9,1. */ + .callback = set_pci_reboot, + .ident = "Apple iMac9,1", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."), + DMI_MATCH(DMI_PRODUCT_NAME, "iMac9,1"), + }, + }, + + /* ASUS */ + { /* Handle problems with rebooting on ASUS P4S800 */ .callback = set_bios_reboot, - .ident = "Dell OptiPlex 745", + .ident = "ASUS P4S800", .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 745"), + DMI_MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC."), + DMI_MATCH(DMI_BOARD_NAME, "P4S800"), }, }, - { /* Handle problems with rebooting on Dell Optiplex 745's DFF*/ + + /* Certec */ + { /* Handle problems with rebooting on Certec BPC600 */ + .callback = set_pci_reboot, + .ident = "Certec BPC600", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "Certec"), + DMI_MATCH(DMI_PRODUCT_NAME, "BPC600"), + }, + }, + + /* Dell */ + { /* Handle problems with rebooting on Dell DXP061 */ .callback = set_bios_reboot, - .ident = "Dell OptiPlex 745", + .ident = "Dell DXP061", .matches = { DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 745"), - DMI_MATCH(DMI_BOARD_NAME, "0MM599"), + DMI_MATCH(DMI_PRODUCT_NAME, "Dell DXP061"), }, }, - { /* Handle problems with rebooting on Dell Optiplex 745 with 0KW626 */ + { /* Handle problems with rebooting on Dell E520's */ .callback = set_bios_reboot, - .ident = "Dell OptiPlex 745", + .ident = "Dell E520", .matches = { DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 745"), - DMI_MATCH(DMI_BOARD_NAME, "0KW626"), + DMI_MATCH(DMI_PRODUCT_NAME, "Dell DM061"), }, }, - { /* Handle problems with rebooting on Dell Optiplex 330 with 0KP561 */ - .callback = set_bios_reboot, - .ident = "Dell OptiPlex 330", + { /* Handle problems with rebooting on the Latitude E5410. */ + .callback = set_pci_reboot, + .ident = "Dell Latitude E5410", .matches = { DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 330"), - DMI_MATCH(DMI_BOARD_NAME, "0KP561"), + DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E5410"), }, }, - { /* Handle problems with rebooting on Dell Optiplex 360 with 0T656F */ - .callback = set_bios_reboot, - .ident = "Dell OptiPlex 360", + { /* Handle problems with rebooting on the Latitude E5420. */ + .callback = set_pci_reboot, + .ident = "Dell Latitude E5420", .matches = { DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 360"), - DMI_MATCH(DMI_BOARD_NAME, "0T656F"), + DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E5420"), }, }, - { /* Handle problems with rebooting on Dell OptiPlex 760 with 0G919G*/ - .callback = set_bios_reboot, - .ident = "Dell OptiPlex 760", + { /* Handle problems with rebooting on the Latitude E6320. */ + .callback = set_pci_reboot, + .ident = "Dell Latitude E6320", .matches = { DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 760"), - DMI_MATCH(DMI_BOARD_NAME, "0G919G"), + DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E6320"), }, }, - { /* Handle problems with rebooting on Dell 2400's */ - .callback = set_bios_reboot, - .ident = "Dell PowerEdge 2400", + { /* Handle problems with rebooting on the Latitude E6420. */ + .callback = set_pci_reboot, + .ident = "Dell Latitude E6420", .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "Dell Computer Corporation"), - DMI_MATCH(DMI_PRODUCT_NAME, "PowerEdge 2400"), + DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), + DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E6420"), }, }, - { /* Handle problems with rebooting on Dell T5400's */ + { /* Handle problems with rebooting on Dell Optiplex 330 with 0KP561 */ .callback = set_bios_reboot, - .ident = "Dell Precision T5400", + .ident = "Dell OptiPlex 330", .matches = { DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "Precision WorkStation T5400"), + DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 330"), + DMI_MATCH(DMI_BOARD_NAME, "0KP561"), }, }, - { /* Handle problems with rebooting on Dell T7400's */ + { /* Handle problems with rebooting on Dell Optiplex 360 with 0T656F */ .callback = set_bios_reboot, - .ident = "Dell Precision T7400", + .ident = "Dell OptiPlex 360", .matches = { DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "Precision WorkStation T7400"), + DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 360"), + DMI_MATCH(DMI_BOARD_NAME, "0T656F"), }, }, - { /* Handle problems with rebooting on HP laptops */ + { /* Handle problems with rebooting on Dell Optiplex 745's SFF */ .callback = set_bios_reboot, - .ident = "HP Compaq Laptop", + .ident = "Dell OptiPlex 745", .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"), - DMI_MATCH(DMI_PRODUCT_NAME, "HP Compaq"), + DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), + DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 745"), }, }, - { /* Handle problems with rebooting on Dell XPS710 */ + { /* Handle problems with rebooting on Dell Optiplex 745's DFF */ .callback = set_bios_reboot, - .ident = "Dell XPS710", + .ident = "Dell OptiPlex 745", .matches = { DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "Dell XPS710"), + DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 745"), + DMI_MATCH(DMI_BOARD_NAME, "0MM599"), }, }, - { /* Handle problems with rebooting on Dell DXP061 */ + { /* Handle problems with rebooting on Dell Optiplex 745 with 0KW626 */ .callback = set_bios_reboot, - .ident = "Dell DXP061", + .ident = "Dell OptiPlex 745", .matches = { DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "Dell DXP061"), + DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 745"), + DMI_MATCH(DMI_BOARD_NAME, "0KW626"), }, }, - { /* Handle problems with rebooting on Sony VGN-Z540N */ + { /* Handle problems with rebooting on Dell OptiPlex 760 with 0G919G */ .callback = set_bios_reboot, - .ident = "Sony VGN-Z540N", + .ident = "Dell OptiPlex 760", .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "Sony Corporation"), - DMI_MATCH(DMI_PRODUCT_NAME, "VGN-Z540N"), + DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), + DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 760"), + DMI_MATCH(DMI_BOARD_NAME, "0G919G"), }, }, - { /* Handle problems with rebooting on CompuLab SBC-FITPC2 */ - .callback = set_bios_reboot, - .ident = "CompuLab SBC-FITPC2", + { /* Handle problems with rebooting on the OptiPlex 990. */ + .callback = set_pci_reboot, + .ident = "Dell OptiPlex 990", .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "CompuLab"), - DMI_MATCH(DMI_PRODUCT_NAME, "SBC-FITPC2"), + DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), + DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 990"), }, }, - { /* Handle problems with rebooting on ASUS P4S800 */ + { /* Handle problems with rebooting on Dell 300's */ .callback = set_bios_reboot, - .ident = "ASUS P4S800", + .ident = "Dell PowerEdge 300", .matches = { - DMI_MATCH(DMI_BOARD_VENDOR, "ASUSTeK Computer INC."), - DMI_MATCH(DMI_BOARD_NAME, "P4S800"), + DMI_MATCH(DMI_SYS_VENDOR, "Dell Computer Corporation"), + DMI_MATCH(DMI_PRODUCT_NAME, "PowerEdge 300/"), }, }, - { /* Handle reboot issue on Acer Aspire one */ - .callback = set_kbd_reboot, - .ident = "Acer Aspire One A110", + { /* Handle problems with rebooting on Dell 1300's */ + .callback = set_bios_reboot, + .ident = "Dell PowerEdge 1300", .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "Acer"), - DMI_MATCH(DMI_PRODUCT_NAME, "AOA110"), + DMI_MATCH(DMI_SYS_VENDOR, "Dell Computer Corporation"), + DMI_MATCH(DMI_PRODUCT_NAME, "PowerEdge 1300/"), }, }, - { } -}; - -static int __init reboot_init(void) -{ - /* Only do the DMI check if reboot_type hasn't been overridden - * on the command line - */ - if (reboot_default) { - dmi_check_system(reboot_dmi_table); - } - return 0; -} -core_initcall(reboot_init); - -extern const unsigned char machine_real_restart_asm[]; -extern const u64 machine_real_restart_gdt[3]; - -void machine_real_restart(unsigned int type) -{ - void *restart_va; - unsigned long restart_pa; - void (*restart_lowmem)(unsigned int); - u64 *lowmem_gdt; - - local_irq_disable(); - - /* Write zero to CMOS register number 0x0f, which the BIOS POST - routine will recognize as telling it to do a proper reboot. (Well - that's what this book in front of me says -- it may only apply to - the Phoenix BIOS though, it's not clear). At the same time, - disable NMIs by setting the top bit in the CMOS address register, - as we're about to do peculiar things to the CPU. I'm not sure if - `outb_p' is needed instead of just `outb'. Use it to be on the - safe side. (Yes, CMOS_WRITE does outb_p's. - Paul G.) - */ - spin_lock(&rtc_lock); - CMOS_WRITE(0x00, 0x8f); - spin_unlock(&rtc_lock); - - /* - * Switch back to the initial page table. - */ - load_cr3(initial_page_table); - - /* Write 0x1234 to absolute memory location 0x472. The BIOS reads - this on booting to tell it to "Bypass memory test (also warm - boot)". This seems like a fairly standard thing that gets set by - REBOOT.COM programs, and the previous reset routine did this - too. */ - *((unsigned short *)0x472) = reboot_mode; - - /* Patch the GDT in the low memory trampoline */ - lowmem_gdt = TRAMPOLINE_SYM(machine_real_restart_gdt); - - restart_va = TRAMPOLINE_SYM(machine_real_restart_asm); - restart_pa = virt_to_phys(restart_va); - restart_lowmem = (void (*)(unsigned int))restart_pa; - - /* GDT[0]: GDT self-pointer */ - lowmem_gdt[0] = - (u64)(sizeof(machine_real_restart_gdt) - 1) + - ((u64)virt_to_phys(lowmem_gdt) << 16); - /* GDT[1]: 64K real mode code segment */ - lowmem_gdt[1] = - GDT_ENTRY(0x009b, restart_pa, 0xffff); - - /* Jump to the identity-mapped low memory code */ - restart_lowmem(type); -} -#ifdef CONFIG_APM_MODULE -EXPORT_SYMBOL(machine_real_restart); -#endif - -#endif /* CONFIG_X86_32 */ - -/* - * Some Apple MacBook and MacBookPro's needs reboot=p to be able to reboot - */ -static int __init set_pci_reboot(const struct dmi_system_id *d) -{ - if (reboot_type != BOOT_CF9) { - reboot_type = BOOT_CF9; - printk(KERN_INFO "%s series board detected. " - "Selecting PCI-method for reboots.\n", d->ident); - } - return 0; -} - -static struct dmi_system_id __initdata pci_reboot_dmi_table[] = { - { /* Handle problems with rebooting on Apple MacBook5 */ - .callback = set_pci_reboot, - .ident = "Apple MacBook5", + { /* Handle problems with rebooting on Dell 2400's */ + .callback = set_bios_reboot, + .ident = "Dell PowerEdge 2400", .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "MacBook5"), + DMI_MATCH(DMI_SYS_VENDOR, "Dell Computer Corporation"), + DMI_MATCH(DMI_PRODUCT_NAME, "PowerEdge 2400"), }, }, - { /* Handle problems with rebooting on Apple MacBookPro5 */ + { /* Handle problems with rebooting on the Dell PowerEdge C6100. */ .callback = set_pci_reboot, - .ident = "Apple MacBookPro5", + .ident = "Dell PowerEdge C6100", .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "MacBookPro5"), + DMI_MATCH(DMI_SYS_VENDOR, "Dell"), + DMI_MATCH(DMI_PRODUCT_NAME, "C6100"), }, }, - { /* Handle problems with rebooting on Apple Macmini3,1 */ + { /* Handle problems with rebooting on the Precision M6600. */ .callback = set_pci_reboot, - .ident = "Apple Macmini3,1", + .ident = "Dell Precision M6600", .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "Macmini3,1"), + DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), + DMI_MATCH(DMI_PRODUCT_NAME, "Precision M6600"), }, }, - { /* Handle problems with rebooting on the iMac9,1. */ - .callback = set_pci_reboot, - .ident = "Apple iMac9,1", + { /* Handle problems with rebooting on Dell T5400's */ + .callback = set_bios_reboot, + .ident = "Dell Precision T5400", .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "Apple Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "iMac9,1"), + DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), + DMI_MATCH(DMI_PRODUCT_NAME, "Precision WorkStation T5400"), }, }, - { /* Handle problems with rebooting on the Latitude E6320. */ - .callback = set_pci_reboot, - .ident = "Dell Latitude E6320", + { /* Handle problems with rebooting on Dell T7400's */ + .callback = set_bios_reboot, + .ident = "Dell Precision T7400", .matches = { DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E6320"), + DMI_MATCH(DMI_PRODUCT_NAME, "Precision WorkStation T7400"), }, }, - { /* Handle problems with rebooting on the Latitude E5420. */ - .callback = set_pci_reboot, - .ident = "Dell Latitude E5420", + { /* Handle problems with rebooting on Dell XPS710 */ + .callback = set_bios_reboot, + .ident = "Dell XPS710", .matches = { DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E5420"), + DMI_MATCH(DMI_PRODUCT_NAME, "Dell XPS710"), }, }, - { /* Handle problems with rebooting on the Latitude E6420. */ - .callback = set_pci_reboot, - .ident = "Dell Latitude E6420", + + /* Hewlett-Packard */ + { /* Handle problems with rebooting on HP laptops */ + .callback = set_bios_reboot, + .ident = "HP Compaq Laptop", .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "Latitude E6420"), + DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"), + DMI_MATCH(DMI_PRODUCT_NAME, "HP Compaq"), }, }, - { /* Handle problems with rebooting on the OptiPlex 990. */ - .callback = set_pci_reboot, - .ident = "Dell OptiPlex 990", + + /* Sony */ + { /* Handle problems with rebooting on Sony VGN-Z540N */ + .callback = set_bios_reboot, + .ident = "Sony VGN-Z540N", .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), - DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex 990"), + DMI_MATCH(DMI_SYS_VENDOR, "Sony Corporation"), + DMI_MATCH(DMI_PRODUCT_NAME, "VGN-Z540N"), }, }, + { } }; -static int __init pci_reboot_init(void) +static int __init reboot_init(void) { - /* Only do the DMI check if reboot_type hasn't been overridden + /* + * Only do the DMI check if reboot_type hasn't been overridden * on the command line */ - if (reboot_default) { - dmi_check_system(pci_reboot_dmi_table); - } + if (reboot_default) + dmi_check_system(reboot_dmi_table); return 0; } -core_initcall(pci_reboot_init); +core_initcall(reboot_init); static inline void kb_wait(void) { @@ -502,14 +427,14 @@ static void vmxoff_nmi(int cpu, struct pt_regs *regs) cpu_emergency_vmxoff(); } -/* Use NMIs as IPIs to tell all CPUs to disable virtualization - */ +/* Use NMIs as IPIs to tell all CPUs to disable virtualization */ static void emergency_vmx_disable_all(void) { /* Just make sure we won't change CPUs while doing this */ local_irq_disable(); - /* We need to disable VMX on all CPUs before rebooting, otherwise + /* + * We need to disable VMX on all CPUs before rebooting, otherwise * we risk hanging up the machine, because the CPU ignore INIT * signals when VMX is enabled. * @@ -528,8 +453,7 @@ static void emergency_vmx_disable_all(void) * is still enabling VMX. */ if (cpu_has_vmx() && cpu_vmx_enabled()) { - /* Disable VMX on this CPU. - */ + /* Disable VMX on this CPU. */ cpu_vmxoff(); /* Halt and disable VMX on the other CPUs */ @@ -544,23 +468,30 @@ void __attribute__((weak)) mach_reboot_fixups(void) } /* - * Windows compatible x86 hardware expects the following on reboot: + * To the best of our knowledge Windows compatible x86 hardware expects + * the following on reboot: * * 1) If the FADT has the ACPI reboot register flag set, try it * 2) If still alive, write to the keyboard controller * 3) If still alive, write to the ACPI reboot register again * 4) If still alive, write to the keyboard controller again + * 5) If still alive, call the EFI runtime service to reboot + * 6) If no EFI runtime service, call the BIOS to do a reboot * - * If the machine is still alive at this stage, it gives up. We default to - * following the same pattern, except that if we're still alive after (4) we'll - * try to force a triple fault and then cycle between hitting the keyboard - * controller and doing that + * We default to following the same pattern. We also have + * two other reboot methods: 'triple fault' and 'PCI', which + * can be triggered via the reboot= kernel boot option or + * via quirks. + * + * This means that this function can never return, it can misbehave + * by not rebooting properly and hanging. */ static void native_machine_emergency_restart(void) { int i; int attempt = 0; int orig_reboot_type = reboot_type; + unsigned short mode; if (reboot_emergency) emergency_vmx_disable_all(); @@ -568,69 +499,72 @@ static void native_machine_emergency_restart(void) tboot_shutdown(TB_SHUTDOWN_REBOOT); /* Tell the BIOS if we want cold or warm reboot */ - *((unsigned short *)__va(0x472)) = reboot_mode; + mode = reboot_mode == REBOOT_WARM ? 0x1234 : 0; + *((unsigned short *)__va(0x472)) = mode; for (;;) { /* Could also try the reset bit in the Hammer NB */ switch (reboot_type) { + case BOOT_ACPI: + acpi_reboot(); + reboot_type = BOOT_KBD; + break; + case BOOT_KBD: - mach_reboot_fixups(); /* for board specific fixups */ + mach_reboot_fixups(); /* For board specific fixups */ for (i = 0; i < 10; i++) { kb_wait(); udelay(50); - outb(0xfe, 0x64); /* pulse reset low */ + outb(0xfe, 0x64); /* Pulse reset low */ udelay(50); } if (attempt == 0 && orig_reboot_type == BOOT_ACPI) { attempt = 1; reboot_type = BOOT_ACPI; } else { - reboot_type = BOOT_TRIPLE; + reboot_type = BOOT_EFI; } break; - case BOOT_TRIPLE: - load_idt(&no_idt); - __asm__ __volatile__("int3"); - - reboot_type = BOOT_KBD; + case BOOT_EFI: + if (efi_enabled(EFI_RUNTIME_SERVICES)) + efi.reset_system(reboot_mode == REBOOT_WARM ? + EFI_RESET_WARM : + EFI_RESET_COLD, + EFI_SUCCESS, 0, NULL); + reboot_type = BOOT_BIOS; break; -#ifdef CONFIG_X86_32 case BOOT_BIOS: machine_real_restart(MRR_BIOS); - reboot_type = BOOT_KBD; - break; -#endif - - case BOOT_ACPI: - acpi_reboot(); - reboot_type = BOOT_KBD; + /* We're probably dead after this, but... */ + reboot_type = BOOT_CF9_SAFE; break; - case BOOT_EFI: - if (efi_enabled) - efi.reset_system(reboot_mode ? - EFI_RESET_WARM : - EFI_RESET_COLD, - EFI_SUCCESS, 0, NULL); - reboot_type = BOOT_KBD; - break; - - case BOOT_CF9: + case BOOT_CF9_FORCE: port_cf9_safe = true; - /* fall through */ + /* Fall through */ - case BOOT_CF9_COND: + case BOOT_CF9_SAFE: if (port_cf9_safe) { - u8 cf9 = inb(0xcf9) & ~6; + u8 reboot_code = reboot_mode == REBOOT_WARM ? 0x06 : 0x0E; + u8 cf9 = inb(0xcf9) & ~reboot_code; outb(cf9|2, 0xcf9); /* Request hard reset */ udelay(50); - outb(cf9|6, 0xcf9); /* Actually do the reset */ + /* Actually do the reset */ + outb(cf9|reboot_code, 0xcf9); udelay(50); } + reboot_type = BOOT_TRIPLE; + break; + + case BOOT_TRIPLE: + load_idt(&no_idt); + __asm__ __volatile__("int3"); + + /* We're probably dead after this, but... */ reboot_type = BOOT_KBD; break; } @@ -640,37 +574,33 @@ static void native_machine_emergency_restart(void) void native_machine_shutdown(void) { /* Stop the cpus and apics */ -#ifdef CONFIG_SMP - - /* The boot cpu is always logical cpu 0 */ - int reboot_cpu_id = 0; - -#ifdef CONFIG_X86_32 - /* See if there has been given a command line override */ - if ((reboot_cpu != -1) && (reboot_cpu < nr_cpu_ids) && - cpu_online(reboot_cpu)) - reboot_cpu_id = reboot_cpu; +#ifdef CONFIG_X86_IO_APIC + /* + * Disabling IO APIC before local APIC is a workaround for + * erratum AVR31 in "Intel Atom Processor C2000 Product Family + * Specification Update". In this situation, interrupts that target + * a Logical Processor whose Local APIC is either in the process of + * being hardware disabled or software disabled are neither delivered + * nor discarded. When this erratum occurs, the processor may hang. + * + * Even without the erratum, it still makes sense to quiet IO APIC + * before disabling Local APIC. + */ + disable_IO_APIC(); #endif - /* Make certain the cpu I'm about to reboot on is online */ - if (!cpu_online(reboot_cpu_id)) - reboot_cpu_id = smp_processor_id(); - - /* Make certain I only run on the appropriate processor */ - set_cpus_allowed_ptr(current, cpumask_of(reboot_cpu_id)); - - /* O.K Now that I'm on the appropriate processor, - * stop all of the others. +#ifdef CONFIG_SMP + /* + * Stop all of the others. Also disable the local irq to + * not receive the per-cpu timer interrupt which may trigger + * scheduler's load balance. */ + local_irq_disable(); stop_other_cpus(); #endif lapic_shutdown(); -#ifdef CONFIG_X86_IO_APIC - disable_IO_APIC(); -#endif - #ifdef CONFIG_HPET_TIMER hpet_disable(); #endif @@ -688,7 +618,7 @@ static void __machine_emergency_restart(int emergency) static void native_machine_restart(char *__unused) { - printk("machine restart\n"); + pr_notice("machine restart\n"); if (!reboot_force) machine_shutdown(); @@ -697,12 +627,11 @@ static void native_machine_restart(char *__unused) static void native_machine_halt(void) { - /* stop other cpus and apics */ + /* Stop other cpus and apics */ machine_shutdown(); tboot_shutdown(TB_SHUTDOWN_HALT); - /* stop this cpu */ stop_this_cpu(NULL); } @@ -713,7 +642,7 @@ static void native_machine_power_off(void) machine_shutdown(); pm_power_off(); } - /* a fallback in case there is no PM info available */ + /* A fallback in case there is no PM info available */ tboot_shutdown(TB_SHUTDOWN_HALT); } @@ -775,7 +704,8 @@ static int crash_nmi_callback(unsigned int val, struct pt_regs *regs) cpu = raw_smp_processor_id(); - /* Don't do anything if this handler is invoked on crashing cpu. + /* + * Don't do anything if this handler is invoked on crashing cpu. * Otherwise, system will completely hang. Crashing cpu can get * an NMI if system was initially booted with nmi_watchdog parameter. */ @@ -799,7 +729,8 @@ static void smp_send_nmi_allbutself(void) apic->send_IPI_allbutself(NMI_VECTOR); } -/* Halt all other CPUs, calling the specified function on each of them +/* + * Halt all other CPUs, calling the specified function on each of them * * This function can be used to halt all other CPUs on crash * or emergency reboot time. The function passed as parameter @@ -810,7 +741,7 @@ void nmi_shootdown_cpus(nmi_shootdown_cb callback) unsigned long msecs; local_irq_disable(); - /* Make a note of crashing cpu. Will be used in NMI callback.*/ + /* Make a note of crashing cpu. Will be used in NMI callback. */ crashing_cpu = safe_smp_processor_id(); shootdown_callback = callback; @@ -819,8 +750,9 @@ void nmi_shootdown_cpus(nmi_shootdown_cb callback) /* Would it be better to replace the trap vector here? */ if (register_nmi_handler(NMI_LOCAL, crash_nmi_callback, NMI_FLAG_FIRST, "crash")) - return; /* return what? */ - /* Ensure the new callback function is set before sending + return; /* Return what? */ + /* + * Ensure the new callback function is set before sending * out the NMI */ wmb(); diff --git a/arch/x86/kernel/reboot_32.S b/arch/x86/kernel/reboot_32.S deleted file mode 100644 index 1d5c46df0d7..00000000000 --- a/arch/x86/kernel/reboot_32.S +++ /dev/null @@ -1,135 +0,0 @@ -#include <linux/linkage.h> -#include <linux/init.h> -#include <asm/segment.h> -#include <asm/page_types.h> - -/* - * The following code and data reboots the machine by switching to real - * mode and jumping to the BIOS reset entry point, as if the CPU has - * really been reset. The previous version asked the keyboard - * controller to pulse the CPU reset line, which is more thorough, but - * doesn't work with at least one type of 486 motherboard. It is easy - * to stop this code working; hence the copious comments. - * - * This code is called with the restart type (0 = BIOS, 1 = APM) in %eax. - */ - .section ".x86_trampoline","a" - .balign 16 - .code32 -ENTRY(machine_real_restart_asm) -r_base = . - /* Get our own relocated address */ - call 1f -1: popl %ebx - subl $(1b - r_base), %ebx - - /* Compute the equivalent real-mode segment */ - movl %ebx, %ecx - shrl $4, %ecx - - /* Patch post-real-mode segment jump */ - movw (dispatch_table - r_base)(%ebx,%eax,2),%ax - movw %ax, (101f - r_base)(%ebx) - movw %cx, (102f - r_base)(%ebx) - - /* Set up the IDT for real mode. */ - lidtl (machine_real_restart_idt - r_base)(%ebx) - - /* - * Set up a GDT from which we can load segment descriptors for real - * mode. The GDT is not used in real mode; it is just needed here to - * prepare the descriptors. - */ - lgdtl (machine_real_restart_gdt - r_base)(%ebx) - - /* - * Load the data segment registers with 16-bit compatible values - */ - movl $16, %ecx - movl %ecx, %ds - movl %ecx, %es - movl %ecx, %fs - movl %ecx, %gs - movl %ecx, %ss - ljmpl $8, $1f - r_base - -/* - * This is 16-bit protected mode code to disable paging and the cache, - * switch to real mode and jump to the BIOS reset code. - * - * The instruction that switches to real mode by writing to CR0 must be - * followed immediately by a far jump instruction, which set CS to a - * valid value for real mode, and flushes the prefetch queue to avoid - * running instructions that have already been decoded in protected - * mode. - * - * Clears all the flags except ET, especially PG (paging), PE - * (protected-mode enable) and TS (task switch for coprocessor state - * save). Flushes the TLB after paging has been disabled. Sets CD and - * NW, to disable the cache on a 486, and invalidates the cache. This - * is more like the state of a 486 after reset. I don't know if - * something else should be done for other chips. - * - * More could be done here to set up the registers as if a CPU reset had - * occurred; hopefully real BIOSs don't assume much. This is not the - * actual BIOS entry point, anyway (that is at 0xfffffff0). - * - * Most of this work is probably excessive, but it is what is tested. - */ - .code16 -1: - xorl %ecx, %ecx - movl %cr0, %eax - andl $0x00000011, %eax - orl $0x60000000, %eax - movl %eax, %cr0 - movl %ecx, %cr3 - movl %cr0, %edx - andl $0x60000000, %edx /* If no cache bits -> no wbinvd */ - jz 2f - wbinvd -2: - andb $0x10, %al - movl %eax, %cr0 - .byte 0xea /* ljmpw */ -101: .word 0 /* Offset */ -102: .word 0 /* Segment */ - -bios: - ljmpw $0xf000, $0xfff0 - -apm: - movw $0x1000, %ax - movw %ax, %ss - movw $0xf000, %sp - movw $0x5307, %ax - movw $0x0001, %bx - movw $0x0003, %cx - int $0x15 - -END(machine_real_restart_asm) - - .balign 16 - /* These must match <asm/reboot.h */ -dispatch_table: - .word bios - r_base - .word apm - r_base -END(dispatch_table) - - .balign 16 -machine_real_restart_idt: - .word 0xffff /* Length - real mode default value */ - .long 0 /* Base - real mode default value */ -END(machine_real_restart_idt) - - .balign 16 -ENTRY(machine_real_restart_gdt) - .quad 0 /* Self-pointer, filled in by PM code */ - .quad 0 /* 16-bit code segment, filled in by PM code */ - /* - * 16-bit data segment with the selector value 16 = 0x10 and - * base value 0x100; since this is consistent with real mode - * semantics we don't have to reload the segments once CR0.PE = 0. - */ - .quad GDT_ENTRY(0x0093, 0x100, 0xffff) -END(machine_real_restart_gdt) diff --git a/arch/x86/kernel/relocate_kernel_32.S b/arch/x86/kernel/relocate_kernel_32.S index 36818f8ec2b..e13f8e7c22a 100644 --- a/arch/x86/kernel/relocate_kernel_32.S +++ b/arch/x86/kernel/relocate_kernel_32.S @@ -186,7 +186,7 @@ identity_mapped: movl CP_PA_PGD(%ebx), %eax movl %eax, %cr3 movl %cr0, %eax - orl $(1<<31), %eax + orl $X86_CR0_PG, %eax movl %eax, %cr0 lea PAGE_SIZE(%edi), %esp movl %edi, %eax diff --git a/arch/x86/kernel/relocate_kernel_64.S b/arch/x86/kernel/relocate_kernel_64.S index 7a6f3b3be3c..3fd2c693e47 100644 --- a/arch/x86/kernel/relocate_kernel_64.S +++ b/arch/x86/kernel/relocate_kernel_64.S @@ -151,21 +151,21 @@ identity_mapped: testq %r11, %r11 jnz 1f - xorq %rax, %rax - xorq %rbx, %rbx - xorq %rcx, %rcx - xorq %rdx, %rdx - xorq %rsi, %rsi - xorq %rdi, %rdi - xorq %rbp, %rbp - xorq %r8, %r8 - xorq %r9, %r9 - xorq %r10, %r9 - xorq %r11, %r11 - xorq %r12, %r12 - xorq %r13, %r13 - xorq %r14, %r14 - xorq %r15, %r15 + xorl %eax, %eax + xorl %ebx, %ebx + xorl %ecx, %ecx + xorl %edx, %edx + xorl %esi, %esi + xorl %edi, %edi + xorl %ebp, %ebp + xorl %r8d, %r8d + xorl %r9d, %r9d + xorl %r10d, %r10d + xorl %r11d, %r11d + xorl %r12d, %r12d + xorl %r13d, %r13d + xorl %r14d, %r14d + xorl %r15d, %r15d ret @@ -212,8 +212,8 @@ virtual_mapped: /* Do the copies */ swap_pages: movq %rdi, %rcx /* Put the page_list in %rcx */ - xorq %rdi, %rdi - xorq %rsi, %rsi + xorl %edi, %edi + xorl %esi, %esi jmp 1f 0: /* top, read another word for the indirection page */ diff --git a/arch/x86/kernel/rtc.c b/arch/x86/kernel/rtc.c index af6db6ec5b2..ca9622a25e9 100644 --- a/arch/x86/kernel/rtc.c +++ b/arch/x86/kernel/rtc.c @@ -12,7 +12,8 @@ #include <asm/vsyscall.h> #include <asm/x86_init.h> #include <asm/time.h> -#include <asm/mrst.h> +#include <asm/intel-mid.h> +#include <asm/rtc.h> #ifdef CONFIG_X86_32 /* @@ -36,74 +37,29 @@ EXPORT_SYMBOL(rtc_lock); * nowtime is written into the registers of the CMOS clock, it will * jump to the next second precisely 500 ms later. Check the Motorola * MC146818A or Dallas DS12887 data sheet for details. - * - * BUG: This routine does not handle hour overflow properly; it just - * sets the minutes. Usually you'll only notice that after reboot! */ -int mach_set_rtc_mmss(unsigned long nowtime) +int mach_set_rtc_mmss(const struct timespec *now) { - int real_seconds, real_minutes, cmos_minutes; - unsigned char save_control, save_freq_select; - unsigned long flags; + unsigned long nowtime = now->tv_sec; + struct rtc_time tm; int retval = 0; - spin_lock_irqsave(&rtc_lock, flags); - - /* tell the clock it's being set */ - save_control = CMOS_READ(RTC_CONTROL); - CMOS_WRITE((save_control|RTC_SET), RTC_CONTROL); - - /* stop and reset prescaler */ - save_freq_select = CMOS_READ(RTC_FREQ_SELECT); - CMOS_WRITE((save_freq_select|RTC_DIV_RESET2), RTC_FREQ_SELECT); - - cmos_minutes = CMOS_READ(RTC_MINUTES); - if (!(save_control & RTC_DM_BINARY) || RTC_ALWAYS_BCD) - cmos_minutes = bcd2bin(cmos_minutes); - - /* - * since we're only adjusting minutes and seconds, - * don't interfere with hour overflow. This avoids - * messing with unknown time zones but requires your - * RTC not to be off by more than 15 minutes - */ - real_seconds = nowtime % 60; - real_minutes = nowtime / 60; - /* correct for half hour time zone */ - if (((abs(real_minutes - cmos_minutes) + 15)/30) & 1) - real_minutes += 30; - real_minutes %= 60; - - if (abs(real_minutes - cmos_minutes) < 30) { - if (!(save_control & RTC_DM_BINARY) || RTC_ALWAYS_BCD) { - real_seconds = bin2bcd(real_seconds); - real_minutes = bin2bcd(real_minutes); - } - CMOS_WRITE(real_seconds, RTC_SECONDS); - CMOS_WRITE(real_minutes, RTC_MINUTES); + rtc_time_to_tm(nowtime, &tm); + if (!rtc_valid_tm(&tm)) { + retval = set_rtc_time(&tm); + if (retval) + printk(KERN_ERR "%s: RTC write failed with error %d\n", + __FUNCTION__, retval); } else { - printk_once(KERN_NOTICE - "set_rtc_mmss: can't update from %d to %d\n", - cmos_minutes, real_minutes); - retval = -1; + printk(KERN_ERR + "%s: Invalid RTC value: write of %lx to RTC failed\n", + __FUNCTION__, nowtime); + retval = -EINVAL; } - - /* The following flags have to be released exactly in this order, - * otherwise the DS12887 (popular MC146818A clone with integrated - * battery and quartz) will not reset the oscillator and will not - * update precisely 500 ms later. You won't find this mentioned in - * the Dallas Semiconductor data sheets, but who believes data - * sheets anyway ... -- Markus Kuhn - */ - CMOS_WRITE(save_control, RTC_CONTROL); - CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT); - - spin_unlock_irqrestore(&rtc_lock, flags); - return retval; } -unsigned long mach_get_cmos_time(void) +void mach_get_cmos_time(struct timespec *now) { unsigned int status, year, mon, day, hour, min, sec, century = 0; unsigned long flags; @@ -149,11 +105,11 @@ unsigned long mach_get_cmos_time(void) if (century) { century = bcd2bin(century); year += century * 100; - printk(KERN_INFO "Extended CMOS year: %d\n", century * 100); } else year += CMOS_YEARS_OFFS; - return mktime(year, mon, day, hour, min, sec); + now->tv_sec = mktime(year, mon, day, hour, min, sec); + now->tv_nsec = 0; } /* Routines for accessing the CMOS RAM/RTC. */ @@ -181,25 +137,14 @@ EXPORT_SYMBOL(rtc_cmos_write); int update_persistent_clock(struct timespec now) { - return x86_platform.set_wallclock(now.tv_sec); + return x86_platform.set_wallclock(&now); } /* not static: needed by APM */ void read_persistent_clock(struct timespec *ts) { - unsigned long retval; - - retval = x86_platform.get_wallclock(); - - ts->tv_sec = retval; - ts->tv_nsec = 0; -} - -unsigned long long native_read_tsc(void) -{ - return __native_read_tsc(); + x86_platform.get_wallclock(ts); } -EXPORT_SYMBOL(native_read_tsc); static struct resource rtc_resources[] = { @@ -225,7 +170,7 @@ static struct platform_device rtc_device = { static __init int add_rtc_cmos(void) { #ifdef CONFIG_PNP - static const char *ids[] __initconst = + static const char * const const ids[] __initconst = { "PNP0b00", "PNP0b01", "PNP0b02", }; struct pnp_dev *dev; struct pnp_id *id; @@ -244,8 +189,16 @@ static __init int add_rtc_cmos(void) return 0; /* Intel MID platforms don't have ioport rtc */ - if (mrst_identify_cpu()) + if (intel_mid_identify_cpu()) + return -ENODEV; + +#ifdef CONFIG_ACPI + if (acpi_gbl_FADT.boot_flags & ACPI_FADT_NO_CMOS_RTC) { + /* This warning can likely go away again in a year or two. */ + pr_info("ACPI: not registering RTC platform device\n"); return -ENODEV; + } +#endif platform_device_register(&rtc_device); dev_info(&rtc_device.dev, diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c index d7d5099fe87..78a0e629892 100644 --- a/arch/x86/kernel/setup.c +++ b/arch/x86/kernel/setup.c @@ -34,7 +34,6 @@ #include <linux/memblock.h> #include <linux/seq_file.h> #include <linux/console.h> -#include <linux/mca.h> #include <linux/root_dev.h> #include <linux/highmem.h> #include <linux/module.h> @@ -50,6 +49,7 @@ #include <asm/pci-direct.h> #include <linux/init_ohci1394_dma.h> #include <linux/kvm_para.h> +#include <linux/dma-contiguous.h> #include <linux/errno.h> #include <linux/kernel.h> @@ -68,12 +68,13 @@ #include <linux/percpu.h> #include <linux/crash_dump.h> #include <linux/tboot.h> +#include <linux/jiffies.h> #include <video/edid.h> #include <asm/mtrr.h> #include <asm/apic.h> -#include <asm/trampoline.h> +#include <asm/realmode.h> #include <asm/e820.h> #include <asm/mpspec.h> #include <asm/setup.h> @@ -81,7 +82,6 @@ #include <asm/timer.h> #include <asm/i8259.h> #include <asm/sections.h> -#include <asm/dmi.h> #include <asm/io_apic.h> #include <asm/ist.h> #include <asm/setup_arch.h> @@ -90,7 +90,6 @@ #include <asm/processor.h> #include <asm/bugs.h> -#include <asm/system.h> #include <asm/vsyscall.h> #include <asm/cpu.h> #include <asm/desc.h> @@ -108,17 +107,16 @@ #include <asm/topology.h> #include <asm/apicdef.h> #include <asm/amd_nb.h> -#ifdef CONFIG_X86_64 -#include <asm/numa_64.h> -#endif #include <asm/mce.h> #include <asm/alternative.h> #include <asm/prom.h> /* - * end_pfn only includes RAM, while max_pfn_mapped includes all e820 entries. - * The direct mapping extends to max_pfn_mapped, so that we can directly access - * apertures, ACPI and other tables without having to play with fixmaps. + * max_low_pfn_mapped: highest direct mapped pfn under 4GB + * max_pfn_mapped: highest direct mapped pfn over 4GB + * + * The direct mapping only covers E820_RAM regions, so the ranges and gaps are + * represented by pfn_mapped */ unsigned long max_low_pfn_mapped; unsigned long max_pfn_mapped; @@ -143,11 +141,7 @@ int default_check_phys_apicid_present(int phys_apicid) } #endif -#ifndef CONFIG_DEBUG_BOOT_PARAMS -struct boot_params __initdata boot_params; -#else struct boot_params boot_params; -#endif /* * Machine setup.. @@ -176,16 +170,14 @@ static struct resource bss_resource = { #ifdef CONFIG_X86_32 /* cpu data as detected by the assembly code in head.S */ -struct cpuinfo_x86 new_cpu_data __cpuinitdata = {0, 0, 0, 0, -1, 1, 0, 0, -1}; +struct cpuinfo_x86 new_cpu_data = { + .wp_works_ok = -1, +}; /* common cpu data for all cpus */ -struct cpuinfo_x86 boot_cpu_data __read_mostly = {0, 0, 0, 0, -1, 1, 0, 0, -1}; +struct cpuinfo_x86 boot_cpu_data __read_mostly = { + .wp_works_ok = -1, +}; EXPORT_SYMBOL(boot_cpu_data); -static void set_mca_bus(int x) -{ -#ifdef CONFIG_MCA - MCA_bus = x; -#endif -} unsigned int def_to_bigsmp; @@ -214,9 +206,9 @@ EXPORT_SYMBOL(boot_cpu_data); #if !defined(CONFIG_X86_PAE) || defined(CONFIG_X86_64) -unsigned long mmu_cr4_features; +__visible unsigned long mmu_cr4_features; #else -unsigned long mmu_cr4_features = X86_CR4_PAE; +__visible unsigned long mmu_cr4_features = X86_CR4_PAE; #endif /* Boot loader ID and version as integers, for the benefit of proc_dointvec */ @@ -286,18 +278,7 @@ void * __init extend_brk(size_t size, size_t align) return ret; } -#ifdef CONFIG_X86_64 -static void __init init_gbpages(void) -{ - if (direct_gbpages && cpu_has_gbpages) - printk(KERN_INFO "Using GB pages for direct mapping\n"); - else - direct_gbpages = 0; -} -#else -static inline void init_gbpages(void) -{ -} +#ifdef CONFIG_X86_32 static void __init cleanup_highmap(void) { } @@ -306,57 +287,64 @@ static void __init cleanup_highmap(void) static void __init reserve_brk(void) { if (_brk_end > _brk_start) - memblock_reserve(__pa(_brk_start), - __pa(_brk_end) - __pa(_brk_start)); + memblock_reserve(__pa_symbol(_brk_start), + _brk_end - _brk_start); /* Mark brk area as locked down and no longer taking any new allocations */ _brk_start = 0; } +u64 relocated_ramdisk; + #ifdef CONFIG_BLK_DEV_INITRD +static u64 __init get_ramdisk_image(void) +{ + u64 ramdisk_image = boot_params.hdr.ramdisk_image; + + ramdisk_image |= (u64)boot_params.ext_ramdisk_image << 32; + + return ramdisk_image; +} +static u64 __init get_ramdisk_size(void) +{ + u64 ramdisk_size = boot_params.hdr.ramdisk_size; + + ramdisk_size |= (u64)boot_params.ext_ramdisk_size << 32; + + return ramdisk_size; +} + #define MAX_MAP_CHUNK (NR_FIX_BTMAPS << PAGE_SHIFT) static void __init relocate_initrd(void) { /* Assume only end is not page aligned */ - u64 ramdisk_image = boot_params.hdr.ramdisk_image; - u64 ramdisk_size = boot_params.hdr.ramdisk_size; + u64 ramdisk_image = get_ramdisk_image(); + u64 ramdisk_size = get_ramdisk_size(); u64 area_size = PAGE_ALIGN(ramdisk_size); - u64 end_of_lowmem = max_low_pfn_mapped << PAGE_SHIFT; - u64 ramdisk_here; unsigned long slop, clen, mapaddr; char *p, *q; - /* We need to move the initrd down into lowmem */ - ramdisk_here = memblock_find_in_range(0, end_of_lowmem, area_size, - PAGE_SIZE); + /* We need to move the initrd down into directly mapped mem */ + relocated_ramdisk = memblock_find_in_range(0, PFN_PHYS(max_pfn_mapped), + area_size, PAGE_SIZE); - if (!ramdisk_here) + if (!relocated_ramdisk) panic("Cannot find place for new RAMDISK of size %lld\n", - ramdisk_size); + ramdisk_size); - /* Note: this includes all the lowmem currently occupied by + /* Note: this includes all the mem currently occupied by the initrd, we rely on that fact to keep the data intact. */ - memblock_reserve(ramdisk_here, area_size); - initrd_start = ramdisk_here + PAGE_OFFSET; + memblock_reserve(relocated_ramdisk, area_size); + initrd_start = relocated_ramdisk + PAGE_OFFSET; initrd_end = initrd_start + ramdisk_size; - printk(KERN_INFO "Allocated new RAMDISK: %08llx - %08llx\n", - ramdisk_here, ramdisk_here + ramdisk_size); + printk(KERN_INFO "Allocated new RAMDISK: [mem %#010llx-%#010llx]\n", + relocated_ramdisk, relocated_ramdisk + ramdisk_size - 1); q = (char *)initrd_start; - /* Copy any lowmem portion of the initrd */ - if (ramdisk_image < end_of_lowmem) { - clen = end_of_lowmem - ramdisk_image; - p = (char *)__va(ramdisk_image); - memcpy(q, p, clen); - q += clen; - ramdisk_image += clen; - ramdisk_size -= clen; - } - - /* Copy the highmem portion of the initrd */ + /* Copy the initrd */ while (ramdisk_size) { slop = ramdisk_image & ~PAGE_MASK; clen = ramdisk_size; @@ -370,22 +358,35 @@ static void __init relocate_initrd(void) ramdisk_image += clen; ramdisk_size -= clen; } - /* high pages is not converted by early_res_to_bootmem */ - ramdisk_image = boot_params.hdr.ramdisk_image; - ramdisk_size = boot_params.hdr.ramdisk_size; - printk(KERN_INFO "Move RAMDISK from %016llx - %016llx to" - " %08llx - %08llx\n", + + ramdisk_image = get_ramdisk_image(); + ramdisk_size = get_ramdisk_size(); + printk(KERN_INFO "Move RAMDISK from [mem %#010llx-%#010llx] to" + " [mem %#010llx-%#010llx]\n", ramdisk_image, ramdisk_image + ramdisk_size - 1, - ramdisk_here, ramdisk_here + ramdisk_size - 1); + relocated_ramdisk, relocated_ramdisk + ramdisk_size - 1); } +static void __init early_reserve_initrd(void) +{ + /* Assume only end is not page aligned */ + u64 ramdisk_image = get_ramdisk_image(); + u64 ramdisk_size = get_ramdisk_size(); + u64 ramdisk_end = PAGE_ALIGN(ramdisk_image + ramdisk_size); + + if (!boot_params.hdr.type_of_loader || + !ramdisk_image || !ramdisk_size) + return; /* No initrd provided by bootloader */ + + memblock_reserve(ramdisk_image, ramdisk_end - ramdisk_image); +} static void __init reserve_initrd(void) { /* Assume only end is not page aligned */ - u64 ramdisk_image = boot_params.hdr.ramdisk_image; - u64 ramdisk_size = boot_params.hdr.ramdisk_size; + u64 ramdisk_image = get_ramdisk_image(); + u64 ramdisk_size = get_ramdisk_size(); u64 ramdisk_end = PAGE_ALIGN(ramdisk_image + ramdisk_size); - u64 end_of_lowmem = max_low_pfn_mapped << PAGE_SHIFT; + u64 mapped_size; if (!boot_params.hdr.type_of_loader || !ramdisk_image || !ramdisk_size) @@ -393,23 +394,18 @@ static void __init reserve_initrd(void) initrd_start = 0; - if (ramdisk_size >= (end_of_lowmem>>1)) { - memblock_free(ramdisk_image, ramdisk_end - ramdisk_image); - printk(KERN_ERR "initrd too large to handle, " - "disabling initrd\n"); - return; - } - - printk(KERN_INFO "RAMDISK: %08llx - %08llx\n", ramdisk_image, - ramdisk_end); + mapped_size = memblock_mem_size(max_pfn_mapped); + if (ramdisk_size >= (mapped_size>>1)) + panic("initrd too large to handle, " + "disabling initrd (%lld needed, %lld available)\n", + ramdisk_size, mapped_size>>1); + printk(KERN_INFO "RAMDISK: [mem %#010llx-%#010llx]\n", ramdisk_image, + ramdisk_end - 1); - if (ramdisk_end <= end_of_lowmem) { - /* All in lowmem, easy case */ - /* - * don't need to reserve again, already reserved early - * in i386_start_kernel - */ + if (pfn_range_is_mapped(PFN_DOWN(ramdisk_image), + PFN_DOWN(ramdisk_end))) { + /* All are mapped, easy case */ initrd_start = ramdisk_image + PAGE_OFFSET; initrd_end = initrd_start + ramdisk_size; return; @@ -420,6 +416,9 @@ static void __init reserve_initrd(void) memblock_free(ramdisk_image, ramdisk_end - ramdisk_image); } #else +static void __init early_reserve_initrd(void) +{ +} static void __init reserve_initrd(void) { } @@ -428,36 +427,34 @@ static void __init reserve_initrd(void) static void __init parse_setup_data(void) { struct setup_data *data; - u64 pa_data; + u64 pa_data, pa_next; - if (boot_params.hdr.version < 0x0209) - return; pa_data = boot_params.hdr.setup_data; while (pa_data) { - u32 data_len, map_len; + u32 data_len, map_len, data_type; map_len = max(PAGE_SIZE - (pa_data & ~PAGE_MASK), (u64)sizeof(struct setup_data)); data = early_memremap(pa_data, map_len); data_len = data->len + sizeof(struct setup_data); - if (data_len > map_len) { - early_iounmap(data, map_len); - data = early_memremap(pa_data, data_len); - map_len = data_len; - } + data_type = data->type; + pa_next = data->next; + early_iounmap(data, map_len); - switch (data->type) { + switch (data_type) { case SETUP_E820_EXT: - parse_e820_ext(data); + parse_e820_ext(pa_data, data_len); break; case SETUP_DTB: add_dtb(pa_data); break; + case SETUP_EFI: + parse_efi_setup(pa_data, data_len); + break; default: break; } - pa_data = data->next; - early_iounmap(data, map_len); + pa_data = pa_next; } } @@ -467,8 +464,6 @@ static void __init e820_reserve_setup_data(void) u64 pa_data; int found = 0; - if (boot_params.hdr.version < 0x0209) - return; pa_data = boot_params.hdr.setup_data; while (pa_data) { data = early_memremap(pa_data, sizeof(*data)); @@ -492,8 +487,6 @@ static void __init memblock_x86_reserve_range_setup_data(void) struct setup_data *data; u64 pa_data; - if (boot_params.hdr.version < 0x0209) - return; pa_data = boot_params.hdr.setup_data; while (pa_data) { data = early_memremap(pa_data, sizeof(*data)); @@ -509,54 +502,107 @@ static void __init memblock_x86_reserve_range_setup_data(void) #ifdef CONFIG_KEXEC -static inline unsigned long long get_total_mem(void) -{ - unsigned long long total; - - total = max_pfn - min_low_pfn; - - return total << PAGE_SHIFT; -} - /* * Keep the crash kernel below this limit. On 32 bits earlier kernels * would limit the kernel to the low 512 MiB due to mapping restrictions. - * On 64 bits, kexec-tools currently limits us to 896 MiB; increase this - * limit once kexec-tools are fixed. + * On 64bit, old kexec-tools need to under 896MiB. */ #ifdef CONFIG_X86_32 -# define CRASH_KERNEL_ADDR_MAX (512 << 20) +# define CRASH_KERNEL_ADDR_LOW_MAX (512 << 20) +# define CRASH_KERNEL_ADDR_HIGH_MAX (512 << 20) #else -# define CRASH_KERNEL_ADDR_MAX (896 << 20) +# define CRASH_KERNEL_ADDR_LOW_MAX (896UL<<20) +# define CRASH_KERNEL_ADDR_HIGH_MAX MAXMEM +#endif + +static void __init reserve_crashkernel_low(void) +{ +#ifdef CONFIG_X86_64 + const unsigned long long alignment = 16<<20; /* 16M */ + unsigned long long low_base = 0, low_size = 0; + unsigned long total_low_mem; + unsigned long long base; + bool auto_set = false; + int ret; + + total_low_mem = memblock_mem_size(1UL<<(32-PAGE_SHIFT)); + /* crashkernel=Y,low */ + ret = parse_crashkernel_low(boot_command_line, total_low_mem, + &low_size, &base); + if (ret != 0) { + /* + * two parts from lib/swiotlb.c: + * swiotlb size: user specified with swiotlb= or default. + * swiotlb overflow buffer: now is hardcoded to 32k. + * We round it to 8M for other buffers that + * may need to stay low too. + */ + low_size = swiotlb_size_or_default() + (8UL<<20); + auto_set = true; + } else { + /* passed with crashkernel=0,low ? */ + if (!low_size) + return; + } + + low_base = memblock_find_in_range(low_size, (1ULL<<32), + low_size, alignment); + + if (!low_base) { + if (!auto_set) + pr_info("crashkernel low reservation failed - No suitable area found.\n"); + + return; + } + + memblock_reserve(low_base, low_size); + pr_info("Reserving %ldMB of low memory at %ldMB for crashkernel (System low RAM: %ldMB)\n", + (unsigned long)(low_size >> 20), + (unsigned long)(low_base >> 20), + (unsigned long)(total_low_mem >> 20)); + crashk_low_res.start = low_base; + crashk_low_res.end = low_base + low_size - 1; + insert_resource(&iomem_resource, &crashk_low_res); #endif +} static void __init reserve_crashkernel(void) { + const unsigned long long alignment = 16<<20; /* 16M */ unsigned long long total_mem; unsigned long long crash_size, crash_base; + bool high = false; int ret; - total_mem = get_total_mem(); + total_mem = memblock_phys_mem_size(); + /* crashkernel=XM */ ret = parse_crashkernel(boot_command_line, total_mem, &crash_size, &crash_base); - if (ret != 0 || crash_size <= 0) - return; + if (ret != 0 || crash_size <= 0) { + /* crashkernel=X,high */ + ret = parse_crashkernel_high(boot_command_line, total_mem, + &crash_size, &crash_base); + if (ret != 0 || crash_size <= 0) + return; + high = true; + } /* 0 means: find the address automatically */ if (crash_base <= 0) { - const unsigned long long alignment = 16<<20; /* 16M */ - /* * kexec want bzImage is below CRASH_KERNEL_ADDR_MAX */ crash_base = memblock_find_in_range(alignment, - CRASH_KERNEL_ADDR_MAX, crash_size, alignment); + high ? CRASH_KERNEL_ADDR_HIGH_MAX : + CRASH_KERNEL_ADDR_LOW_MAX, + crash_size, alignment); if (!crash_base) { pr_info("crashkernel reservation failed - No suitable area found.\n"); return; } + } else { unsigned long long start; @@ -578,6 +624,9 @@ static void __init reserve_crashkernel(void) crashk_res.start = crash_base; crashk_res.end = crash_base + crash_size - 1; insert_resource(&iomem_resource, &crashk_res); + + if (crash_base >= (1ULL<<32)) + reserve_crashkernel_low(); } #else static void __init reserve_crashkernel(void) @@ -628,7 +677,82 @@ static __init void reserve_ibft_region(void) memblock_reserve(addr, size); } -static unsigned reserve_low = CONFIG_X86_RESERVE_LOW << 10; +static bool __init snb_gfx_workaround_needed(void) +{ +#ifdef CONFIG_PCI + int i; + u16 vendor, devid; + static const __initconst u16 snb_ids[] = { + 0x0102, + 0x0112, + 0x0122, + 0x0106, + 0x0116, + 0x0126, + 0x010a, + }; + + /* Assume no if something weird is going on with PCI */ + if (!early_pci_allowed()) + return false; + + vendor = read_pci_config_16(0, 2, 0, PCI_VENDOR_ID); + if (vendor != 0x8086) + return false; + + devid = read_pci_config_16(0, 2, 0, PCI_DEVICE_ID); + for (i = 0; i < ARRAY_SIZE(snb_ids); i++) + if (devid == snb_ids[i]) + return true; +#endif + + return false; +} + +/* + * Sandy Bridge graphics has trouble with certain ranges, exclude + * them from allocation. + */ +static void __init trim_snb_memory(void) +{ + static const __initconst unsigned long bad_pages[] = { + 0x20050000, + 0x20110000, + 0x20130000, + 0x20138000, + 0x40004000, + }; + int i; + + if (!snb_gfx_workaround_needed()) + return; + + printk(KERN_DEBUG "reserving inaccessible SNB gfx pages\n"); + + /* + * Reserve all memory below the 1 MB mark that has not + * already been reserved. + */ + memblock_reserve(0, 1<<20); + + for (i = 0; i < ARRAY_SIZE(bad_pages); i++) { + if (memblock_reserve(bad_pages[i], PAGE_SIZE)) + printk(KERN_WARNING "failed to reserve 0x%08lx\n", + bad_pages[i]); + } +} + +/* + * Here we put platform-specific memory range workarounds, i.e. + * memory known to be corrupt or otherwise in need to be reserved on + * specific platforms. + * + * If this gets used more widely it could use a real dispatch mechanism. + */ +static void __init trim_platform_memory_ranges(void) +{ + trim_snb_memory(); +} static void __init trim_bios_range(void) { @@ -641,8 +765,7 @@ static void __init trim_bios_range(void) * since some BIOSes are known to corrupt low memory. See the * Kconfig help text for X86_RESERVE_LOW. */ - e820_update_range(0, ALIGN(reserve_low, PAGE_SIZE), - E820_RAM, E820_RESERVED); + e820_update_range(0, PAGE_SIZE, E820_RAM, E820_RESERVED); /* * special case: Some BIOSen report the PC BIOS @@ -650,9 +773,33 @@ static void __init trim_bios_range(void) * take them out. */ e820_remove_range(BIOS_BEGIN, BIOS_END - BIOS_BEGIN, E820_RAM, 1); + sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &e820.nr_map); } +/* called before trim_bios_range() to spare extra sanitize */ +static void __init e820_add_kernel_range(void) +{ + u64 start = __pa_symbol(_text); + u64 size = __pa_symbol(_end) - start; + + /* + * Complain if .text .data and .bss are not marked as E820_RAM and + * attempt to fix it by adding the range. We may have a confused BIOS, + * or the user may have used memmap=exactmap or memmap=xxM$yyM to + * exclude kernel range. If we really are running on top non-RAM, + * we will crash later anyways. + */ + if (e820_all_mapped(start, start + size, E820_RAM)) + return; + + pr_warn(".text .data .bss are not marked as E820_RAM!\n"); + e820_remove_range(start, size, E820_RAM, 0); + e820_add_region(start, size, E820_RAM); +} + +static unsigned reserve_low = CONFIG_X86_RESERVE_LOW << 10; + static int __init parse_reservelow(char *p) { unsigned long long size; @@ -675,6 +822,25 @@ static int __init parse_reservelow(char *p) early_param("reservelow", parse_reservelow); +static void __init trim_low_memory_range(void) +{ + memblock_reserve(0, ALIGN(reserve_low, PAGE_SIZE)); +} + +/* + * Dump out kernel offset information on panic. + */ +static int +dump_kernel_offset(struct notifier_block *self, unsigned long v, void *p) +{ + pr_emerg("Kernel Offset: 0x%lx from 0x%lx " + "(relocation range: 0x%lx-0x%lx)\n", + (unsigned long)&_text - __START_KERNEL, __START_KERNEL, + __START_KERNEL_map, MODULES_VADDR-1); + + return 0; +} + /* * Determine if we were loaded by an EFI loader. If so, then we have also been * passed the efi memmap, systab, etc., so we should use these data structures @@ -690,9 +856,19 @@ early_param("reservelow", parse_reservelow); void __init setup_arch(char **cmdline_p) { + memblock_reserve(__pa_symbol(_text), + (unsigned long)__bss_stop - (unsigned long)_text); + + early_reserve_initrd(); + + /* + * At this point everything still needed from the boot loader + * or BIOS or kernel text should be early reserved or marked not + * RAM in e820. All other memory is free game. + */ + #ifdef CONFIG_X86_32 memcpy(&boot_cpu_data, &new_cpu_data, sizeof(new_cpu_data)); - visws_early_detect(); /* * copy kernel address range established so far and switch @@ -727,7 +903,6 @@ void __init setup_arch(char **cmdline_p) apm_info.bios = boot_params.apm_bios_info; ist_info = boot_params.ist_info; if (boot_params.sys_desc_table.length != 0) { - set_mca_bus(boot_params.sys_desc_table.table[3] & 0x2); machine_id = boot_params.sys_desc_table.table[0]; machine_submodel_id = boot_params.sys_desc_table.table[1]; BIOS_revision = boot_params.sys_desc_table.table[2]; @@ -749,10 +924,16 @@ void __init setup_arch(char **cmdline_p) #endif #ifdef CONFIG_EFI if (!strncmp((char *)&boot_params.efi_info.efi_loader_signature, - EFI_LOADER_SIGNATURE, 4)) { - efi_enabled = 1; - efi_memblock_x86_reserve_range(); + "EL32", 4)) { + set_bit(EFI_BOOT, &efi.flags); + } else if (!strncmp((char *)&boot_params.efi_info.efi_loader_signature, + "EL64", 4)) { + set_bit(EFI_BOOT, &efi.flags); + set_bit(EFI_64BIT, &efi.flags); } + + if (efi_enabled(EFI_BOOT)) + efi_memblock_x86_reserve_range(); #endif x86_init.oem.arch_setup(); @@ -760,8 +941,6 @@ void __init setup_arch(char **cmdline_p) iomem_resource.end = (1ULL << boot_cpu_data.x86_phys_bits) - 1; setup_memory_map(); parse_setup_data(); - /* update the e820_saved too */ - e820_reserve_setup_data(); copy_edd(); @@ -772,12 +951,12 @@ void __init setup_arch(char **cmdline_p) init_mm.end_data = (unsigned long) _edata; init_mm.brk = _brk_end; - code_resource.start = virt_to_phys(_text); - code_resource.end = virt_to_phys(_etext)-1; - data_resource.start = virt_to_phys(_etext); - data_resource.end = virt_to_phys(_edata)-1; - bss_resource.start = virt_to_phys(&__bss_start); - bss_resource.end = virt_to_phys(&__bss_stop)-1; + code_resource.start = __pa_symbol(_text); + code_resource.end = __pa_symbol(_etext)-1; + data_resource.start = __pa_symbol(_etext); + data_resource.end = __pa_symbol(_edata)-1; + bss_resource.start = __pa_symbol(__bss_start); + bss_resource.end = __pa_symbol(__bss_stop)-1; #ifdef CONFIG_CMDLINE_BOOL #ifdef CONFIG_CMDLINE_OVERRIDE @@ -823,12 +1002,16 @@ void __init setup_arch(char **cmdline_p) early_dump_pci_devices(); #endif + /* update the e820_saved too */ + e820_reserve_setup_data(); finish_e820_parsing(); - if (efi_enabled) + if (efi_enabled(EFI_BOOT)) efi_init(); dmi_scan_machine(); + dmi_memdev_walk(); + dmi_set_dump_stack_arch_desc(); /* * VMware detection requires dmi to be available, so this @@ -843,6 +1026,7 @@ void __init setup_arch(char **cmdline_p) insert_resource(&iomem_resource, &data_resource); insert_resource(&iomem_resource, &bss_resource); + e820_add_kernel_range(); trim_bios_range(); #ifdef CONFIG_X86_32 if (ppro_with_ram_bug()) { @@ -871,8 +1055,6 @@ void __init setup_arch(char **cmdline_p) /* max_low_pfn get updated here */ find_low_pfn_range(); #else - num_physpages = max_pfn; - check_x2apic(); /* How many end-of-memory variables you have, grandma! */ @@ -892,6 +1074,8 @@ void __init setup_arch(char **cmdline_p) reserve_ibft_region(); + early_alloc_pgt_buf(); + /* * Need to conclude brk, before memblock_x86_fill() * it could use memblock_find_in_range, could overlap with @@ -901,14 +1085,14 @@ void __init setup_arch(char **cmdline_p) cleanup_highmap(); - memblock.current_limit = get_max_mapped(); + memblock_set_current_limit(ISA_END_ADDRESS); memblock_x86_fill(); /* * The EFI specification says that boot service code won't be called * after ExitBootServices(). This is, in fact, a lie. */ - if (efi_enabled) + if (efi_enabled(EFI_MEMMAP)) efi_reserve_boot_services(); /* preallocate 4k for mptable mpc */ @@ -918,26 +1102,24 @@ void __init setup_arch(char **cmdline_p) setup_bios_corruption_check(); #endif - printk(KERN_DEBUG "initial memory mapped : 0 - %08lx\n", - max_pfn_mapped<<PAGE_SHIFT); +#ifdef CONFIG_X86_32 + printk(KERN_DEBUG "initial memory mapped: [mem 0x00000000-%#010lx]\n", + (max_pfn_mapped<<PAGE_SHIFT) - 1); +#endif - setup_trampolines(); + reserve_real_mode(); - init_gbpages(); + trim_platform_memory_ranges(); + trim_low_memory_range(); - /* max_pfn_mapped is updated here */ - max_low_pfn_mapped = init_memory_mapping(0, max_low_pfn<<PAGE_SHIFT); - max_pfn_mapped = max_low_pfn_mapped; + init_mem_mapping(); -#ifdef CONFIG_X86_64 - if (max_pfn > max_low_pfn) { - max_pfn_mapped = init_memory_mapping(1UL<<32, - max_pfn<<PAGE_SHIFT); - /* can we preseve max_low_pfn ?*/ - max_low_pfn = max_pfn; - } -#endif - memblock.current_limit = get_max_mapped(); + early_trap_pf_init(); + + setup_real_mode(); + + memblock_set_current_limit(get_max_mapped()); + dma_contiguous_reserve(max_pfn_mapped << PAGE_SHIFT); /* * NOTE: On x86-32, only from this point on, fixmaps are ready for use. @@ -952,7 +1134,9 @@ void __init setup_arch(char **cmdline_p) reserve_initrd(); - reserve_crashkernel(); +#if defined(CONFIG_ACPI) && defined(CONFIG_BLK_DEV_INITRD) + acpi_initrd_override((void *)initrd_start, initrd_end - initrd_start); +#endif vsmp_init(); @@ -966,19 +1150,26 @@ void __init setup_arch(char **cmdline_p) early_acpi_boot_init(); initmem_init(); + + /* + * Reserve memory for crash kernel after SRAT is parsed so that it + * won't consume hotpluggable memory. + */ + reserve_crashkernel(); + memblock_find_dma_reserve(); -#ifdef CONFIG_KVM_CLOCK +#ifdef CONFIG_KVM_GUEST kvmclock_init(); #endif - x86_init.paging.pagetable_setup_start(swapper_pg_dir); - paging_init(); - x86_init.paging.pagetable_setup_done(swapper_pg_dir); + x86_init.paging.pagetable_init(); if (boot_cpu_data.cpuid_level >= 0) { /* A CPU has %cr4 if and only if it has CPUID */ mmu_cr4_features = read_cr4(); + if (trampoline_cr4_features) + *trampoline_cr4_features = mmu_cr4_features; } #ifdef CONFIG_X86_32 @@ -1016,7 +1207,8 @@ void __init setup_arch(char **cmdline_p) init_cpu_to_node(); init_apic_mappings(); - ioapic_and_gsi_init(); + if (x86_io_apic_ops.init) + x86_io_apic_ops.init(); kvm_guest_init(); @@ -1029,7 +1221,7 @@ void __init setup_arch(char **cmdline_p) #ifdef CONFIG_VT #if defined(CONFIG_VGA_CONSOLE) - if (!efi_enabled || (efi_mem_type(0xa0000) != EFI_CONVENTIONAL_MEMORY)) + if (!efi_enabled(EFI_BOOT) || (efi_mem_type(0xa0000) != EFI_CONVENTIONAL_MEMORY)) conswitchp = &vga_con; #elif defined(CONFIG_DUMMY_CONSOLE) conswitchp = &dummy_con; @@ -1039,11 +1231,16 @@ void __init setup_arch(char **cmdline_p) x86_init.timers.wallclock_init(); - x86_platform.wallclock_init(); - mcheck_init(); arch_init_ideal_nops(); + + register_refined_jiffies(CLOCK_TICK_RATE); + +#ifdef CONFIG_EFI + if (efi_enabled(EFI_BOOT)) + efi_apply_memmap_quirks(); +#endif } #ifdef CONFIG_X86_32 @@ -1062,3 +1259,15 @@ void __init i386_reserve_resources(void) } #endif /* CONFIG_X86_32 */ + +static struct notifier_block kernel_offset_notifier = { + .notifier_call = dump_kernel_offset +}; + +static int __init register_kernel_offset_dumper(void) +{ + atomic_notifier_chain_register(&panic_notifier_list, + &kernel_offset_notifier); + return 0; +} +__initcall(register_kernel_offset_dumper); diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index 71f4727da37..5cdff035774 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -21,7 +21,7 @@ #include <asm/cpu.h> #include <asm/stackprotector.h> -DEFINE_PER_CPU(int, cpu_number); +DEFINE_PER_CPU_READ_MOSTLY(int, cpu_number); EXPORT_PER_CPU_SYMBOL(cpu_number); #ifdef CONFIG_X86_64 @@ -185,10 +185,22 @@ void __init setup_per_cpu_areas(void) #endif rc = -EINVAL; if (pcpu_chosen_fc != PCPU_FC_PAGE) { - const size_t atom_size = cpu_has_pse ? PMD_SIZE : PAGE_SIZE; const size_t dyn_size = PERCPU_MODULE_RESERVE + PERCPU_DYNAMIC_RESERVE - PERCPU_FIRST_CHUNK_RESERVE; + size_t atom_size; + /* + * On 64bit, use PMD_SIZE for atom_size so that embedded + * percpu areas are aligned to PMD. This, in the future, + * can also allow using PMD mappings in vmalloc area. Use + * PAGE_SIZE on 32bit as vmalloc space is highly contended + * and large vmalloc area allocs can easily fail. + */ +#ifdef CONFIG_X86_64 + atom_size = PMD_SIZE; +#else + atom_size = PAGE_SIZE; +#endif rc = pcpu_embed_first_chunk(PERCPU_FIRST_CHUNK_RESERVE, dyn_size, atom_size, pcpu_cpu_distance, diff --git a/arch/x86/kernel/signal.c b/arch/x86/kernel/signal.c index 46a01bdc27e..2851d63c120 100644 --- a/arch/x86/kernel/signal.c +++ b/arch/x86/kernel/signal.c @@ -6,30 +6,36 @@ * 2000-06-20 Pentium III FXSR, SSE support by Gareth Hughes * 2000-2002 x86-64 support by Andi Kleen */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/sched.h> #include <linux/mm.h> #include <linux/smp.h> #include <linux/kernel.h> -#include <linux/signal.h> #include <linux/errno.h> #include <linux/wait.h> -#include <linux/ptrace.h> #include <linux/tracehook.h> #include <linux/unistd.h> #include <linux/stddef.h> #include <linux/personality.h> #include <linux/uaccess.h> #include <linux/user-return-notifier.h> +#include <linux/uprobes.h> +#include <linux/context_tracking.h> #include <asm/processor.h> #include <asm/ucontext.h> #include <asm/i387.h> +#include <asm/fpu-internal.h> #include <asm/vdso.h> #include <asm/mce.h> +#include <asm/sighandling.h> #ifdef CONFIG_X86_64 #include <asm/proto.h> #include <asm/ia32_unistd.h> +#include <asm/sys_ia32.h> #endif /* CONFIG_X86_64 */ #include <asm/syscall.h> @@ -37,19 +43,6 @@ #include <asm/sigframe.h> -#define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP))) - -#define __FIX_EFLAGS (X86_EFLAGS_AC | X86_EFLAGS_OF | \ - X86_EFLAGS_DF | X86_EFLAGS_TF | X86_EFLAGS_SF | \ - X86_EFLAGS_ZF | X86_EFLAGS_AF | X86_EFLAGS_PF | \ - X86_EFLAGS_CF) - -#ifdef CONFIG_X86_32 -# define FIX_EFLAGS (__FIX_EFLAGS | X86_EFLAGS_RF) -#else -# define FIX_EFLAGS __FIX_EFLAGS -#endif - #define COPY(x) do { \ get_user_ex(regs->x, &sc->x); \ } while (0) @@ -68,9 +61,8 @@ regs->seg = GET_SEG(seg) | 3; \ } while (0) -static int -restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc, - unsigned long *pax) +int restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc, + unsigned long *pax) { void __user *buf; unsigned int tmpflags; @@ -117,17 +109,17 @@ restore_sigcontext(struct pt_regs *regs, struct sigcontext __user *sc, regs->orig_ax = -1; /* disable syscall checks */ get_user_ex(buf, &sc->fpstate); - err |= restore_i387_xstate(buf); get_user_ex(*pax, &sc->ax); } get_user_catch(err); + err |= restore_xstate_sig(buf, config_enabled(CONFIG_X86_32)); + return err; } -static int -setup_sigcontext(struct sigcontext __user *sc, void __user *fpstate, - struct pt_regs *regs, unsigned long mask) +int setup_sigcontext(struct sigcontext __user *sc, void __user *fpstate, + struct pt_regs *regs, unsigned long mask) { int err = 0; @@ -159,7 +151,7 @@ setup_sigcontext(struct sigcontext __user *sc, void __user *fpstate, put_user_ex(regs->r15, &sc->r15); #endif /* CONFIG_X86_64 */ - put_user_ex(current->thread.trap_no, &sc->trapno); + put_user_ex(current->thread.trap_nr, &sc->trapno); put_user_ex(current->thread.error_code, &sc->err); put_user_ex(regs->ip, &sc->ip); #ifdef CONFIG_X86_32 @@ -210,35 +202,32 @@ get_sigframe(struct k_sigaction *ka, struct pt_regs *regs, size_t frame_size, void __user **fpstate) { /* Default to using normal stack */ + unsigned long math_size = 0; unsigned long sp = regs->sp; + unsigned long buf_fx = 0; int onsigstack = on_sig_stack(sp); -#ifdef CONFIG_X86_64 /* redzone */ - sp -= 128; -#endif /* CONFIG_X86_64 */ + if (config_enabled(CONFIG_X86_64)) + sp -= 128; if (!onsigstack) { /* This is the X/Open sanctioned signal stack switching. */ if (ka->sa.sa_flags & SA_ONSTACK) { if (current->sas_ss_size) sp = current->sas_ss_sp + current->sas_ss_size; - } else { -#ifdef CONFIG_X86_32 - /* This is the legacy signal stack switching. */ - if ((regs->ss & 0xffff) != __USER_DS && - !(ka->sa.sa_flags & SA_RESTORER) && - ka->sa.sa_restorer) + } else if (config_enabled(CONFIG_X86_32) && + (regs->ss & 0xffff) != __USER_DS && + !(ka->sa.sa_flags & SA_RESTORER) && + ka->sa.sa_restorer) { + /* This is the legacy signal stack switching. */ sp = (unsigned long) ka->sa.sa_restorer; -#endif /* CONFIG_X86_32 */ } } if (used_math()) { - sp -= sig_xstate_size; -#ifdef CONFIG_X86_64 - sp = round_down(sp, 64); -#endif /* CONFIG_X86_64 */ + sp = alloc_mathframe(sp, config_enabled(CONFIG_X86_32), + &buf_fx, &math_size); *fpstate = (void __user *)sp; } @@ -251,8 +240,9 @@ get_sigframe(struct k_sigaction *ka, struct pt_regs *regs, size_t frame_size, if (onsigstack && !likely(on_sig_stack(sp))) return (void __user *)-1L; - /* save i387 state */ - if (used_math() && save_i387_xstate(*fpstate) < 0) + /* save i387 and extended state */ + if (used_math() && + save_xstate_sig(*fpstate, (void __user *)buf_fx, math_size) < 0) return (void __user *)-1L; return (void __user *)sp; @@ -282,7 +272,7 @@ static const struct { }; static int -__setup_frame(int sig, struct k_sigaction *ka, sigset_t *set, +__setup_frame(int sig, struct ksignal *ksig, sigset_t *set, struct pt_regs *regs) { struct sigframe __user *frame; @@ -290,7 +280,7 @@ __setup_frame(int sig, struct k_sigaction *ka, sigset_t *set, int err = 0; void __user *fpstate = NULL; - frame = get_sigframe(ka, regs, sizeof(*frame), &fpstate); + frame = get_sigframe(&ksig->ka, regs, sizeof(*frame), &fpstate); if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame))) return -EFAULT; @@ -308,11 +298,12 @@ __setup_frame(int sig, struct k_sigaction *ka, sigset_t *set, } if (current->mm->context.vdso) - restorer = VDSO32_SYMBOL(current->mm->context.vdso, sigreturn); + restorer = current->mm->context.vdso + + selected_vdso32->sym___kernel_sigreturn; else restorer = &frame->retcode; - if (ka->sa.sa_flags & SA_RESTORER) - restorer = ka->sa.sa_restorer; + if (ksig->ka.sa.sa_flags & SA_RESTORER) + restorer = ksig->ka.sa.sa_restorer; /* Set up to return from userspace. */ err |= __put_user(restorer, &frame->pretcode); @@ -331,7 +322,7 @@ __setup_frame(int sig, struct k_sigaction *ka, sigset_t *set, /* Set up registers for signal handler */ regs->sp = (unsigned long)frame; - regs->ip = (unsigned long)ka->sa.sa_handler; + regs->ip = (unsigned long)ksig->ka.sa.sa_handler; regs->ax = (unsigned long)sig; regs->dx = 0; regs->cx = 0; @@ -344,7 +335,7 @@ __setup_frame(int sig, struct k_sigaction *ka, sigset_t *set, return 0; } -static int __setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, +static int __setup_rt_frame(int sig, struct ksignal *ksig, sigset_t *set, struct pt_regs *regs) { struct rt_sigframe __user *frame; @@ -352,7 +343,7 @@ static int __setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, int err = 0; void __user *fpstate = NULL; - frame = get_sigframe(ka, regs, sizeof(*frame), &fpstate); + frame = get_sigframe(&ksig->ka, regs, sizeof(*frame), &fpstate); if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame))) return -EFAULT; @@ -361,7 +352,6 @@ static int __setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, put_user_ex(sig, &frame->sig); put_user_ex(&frame->info, &frame->pinfo); put_user_ex(&frame->uc, &frame->puc); - err |= copy_siginfo_to_user(&frame->info, info); /* Create the ucontext. */ if (cpu_has_xsave) @@ -369,18 +359,13 @@ static int __setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, else put_user_ex(0, &frame->uc.uc_flags); put_user_ex(0, &frame->uc.uc_link); - put_user_ex(current->sas_ss_sp, &frame->uc.uc_stack.ss_sp); - put_user_ex(sas_ss_flags(regs->sp), - &frame->uc.uc_stack.ss_flags); - put_user_ex(current->sas_ss_size, &frame->uc.uc_stack.ss_size); - err |= setup_sigcontext(&frame->uc.uc_mcontext, fpstate, - regs, set->sig[0]); - err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set)); + save_altstack_ex(&frame->uc.uc_stack, regs->sp); /* Set up to return from userspace. */ - restorer = VDSO32_SYMBOL(current->mm->context.vdso, rt_sigreturn); - if (ka->sa.sa_flags & SA_RESTORER) - restorer = ka->sa.sa_restorer; + restorer = current->mm->context.vdso + + selected_vdso32->sym___kernel_rt_sigreturn; + if (ksig->ka.sa.sa_flags & SA_RESTORER) + restorer = ksig->ka.sa.sa_restorer; put_user_ex(restorer, &frame->pretcode); /* @@ -392,13 +377,18 @@ static int __setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, */ put_user_ex(*((u64 *)&rt_retcode), (u64 *)frame->retcode); } put_user_catch(err); + + err |= copy_siginfo_to_user(&frame->info, &ksig->info); + err |= setup_sigcontext(&frame->uc.uc_mcontext, fpstate, + regs, set->sig[0]); + err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set)); if (err) return -EFAULT; /* Set up registers for signal handler */ regs->sp = (unsigned long)frame; - regs->ip = (unsigned long)ka->sa.sa_handler; + regs->ip = (unsigned long)ksig->ka.sa.sa_handler; regs->ax = (unsigned long)sig; regs->dx = (unsigned long)&frame->info; regs->cx = (unsigned long)&frame->uc; @@ -411,21 +401,20 @@ static int __setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, return 0; } #else /* !CONFIG_X86_32 */ -static int __setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, +static int __setup_rt_frame(int sig, struct ksignal *ksig, sigset_t *set, struct pt_regs *regs) { struct rt_sigframe __user *frame; void __user *fp = NULL; int err = 0; - struct task_struct *me = current; - frame = get_sigframe(ka, regs, sizeof(struct rt_sigframe), &fp); + frame = get_sigframe(&ksig->ka, regs, sizeof(struct rt_sigframe), &fp); if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame))) return -EFAULT; - if (ka->sa.sa_flags & SA_SIGINFO) { - if (copy_siginfo_to_user(&frame->info, info)) + if (ksig->ka.sa.sa_flags & SA_SIGINFO) { + if (copy_siginfo_to_user(&frame->info, &ksig->info)) return -EFAULT; } @@ -436,24 +425,22 @@ static int __setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, else put_user_ex(0, &frame->uc.uc_flags); put_user_ex(0, &frame->uc.uc_link); - put_user_ex(me->sas_ss_sp, &frame->uc.uc_stack.ss_sp); - put_user_ex(sas_ss_flags(regs->sp), - &frame->uc.uc_stack.ss_flags); - put_user_ex(me->sas_ss_size, &frame->uc.uc_stack.ss_size); - err |= setup_sigcontext(&frame->uc.uc_mcontext, fp, regs, set->sig[0]); - err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set)); + save_altstack_ex(&frame->uc.uc_stack, regs->sp); /* Set up to return from userspace. If provided, use a stub already in userspace. */ /* x86-64 should always use SA_RESTORER. */ - if (ka->sa.sa_flags & SA_RESTORER) { - put_user_ex(ka->sa.sa_restorer, &frame->pretcode); + if (ksig->ka.sa.sa_flags & SA_RESTORER) { + put_user_ex(ksig->ka.sa.sa_restorer, &frame->pretcode); } else { /* could use a vstub here */ err |= -EFAULT; } } put_user_catch(err); + err |= setup_sigcontext(&frame->uc.uc_mcontext, fp, regs, set->sig[0]); + err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set)); + if (err) return -EFAULT; @@ -466,7 +453,7 @@ static int __setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, next argument after the signal number on the stack. */ regs->si = (unsigned long)&frame->info; regs->dx = (unsigned long)&frame->uc; - regs->ip = (unsigned long) ka->sa.sa_handler; + regs->ip = (unsigned long) ksig->ka.sa.sa_handler; regs->sp = (unsigned long)frame; @@ -478,87 +465,79 @@ static int __setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, } #endif /* CONFIG_X86_32 */ -#ifdef CONFIG_X86_32 -/* - * Atomically swap in the new signal mask, and wait for a signal. - */ -asmlinkage int -sys_sigsuspend(int history0, int history1, old_sigset_t mask) +static int x32_setup_rt_frame(struct ksignal *ksig, + compat_sigset_t *set, + struct pt_regs *regs) { - sigset_t blocked; - - current->saved_sigmask = current->blocked; - - mask &= _BLOCKABLE; - siginitset(&blocked, mask); - set_current_blocked(&blocked); - - current->state = TASK_INTERRUPTIBLE; - schedule(); - - set_restore_sigmask(); - return -ERESTARTNOHAND; -} +#ifdef CONFIG_X86_X32_ABI + struct rt_sigframe_x32 __user *frame; + void __user *restorer; + int err = 0; + void __user *fpstate = NULL; -asmlinkage int -sys_sigaction(int sig, const struct old_sigaction __user *act, - struct old_sigaction __user *oact) -{ - struct k_sigaction new_ka, old_ka; - int ret = 0; + frame = get_sigframe(&ksig->ka, regs, sizeof(*frame), &fpstate); - if (act) { - old_sigset_t mask; + if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame))) + return -EFAULT; - if (!access_ok(VERIFY_READ, act, sizeof(*act))) + if (ksig->ka.sa.sa_flags & SA_SIGINFO) { + if (copy_siginfo_to_user32(&frame->info, &ksig->info)) return -EFAULT; + } - get_user_try { - get_user_ex(new_ka.sa.sa_handler, &act->sa_handler); - get_user_ex(new_ka.sa.sa_flags, &act->sa_flags); - get_user_ex(mask, &act->sa_mask); - get_user_ex(new_ka.sa.sa_restorer, &act->sa_restorer); - } get_user_catch(ret); + put_user_try { + /* Create the ucontext. */ + if (cpu_has_xsave) + put_user_ex(UC_FP_XSTATE, &frame->uc.uc_flags); + else + put_user_ex(0, &frame->uc.uc_flags); + put_user_ex(0, &frame->uc.uc_link); + compat_save_altstack_ex(&frame->uc.uc_stack, regs->sp); + put_user_ex(0, &frame->uc.uc__pad0); - if (ret) - return -EFAULT; - siginitset(&new_ka.sa.sa_mask, mask); - } + if (ksig->ka.sa.sa_flags & SA_RESTORER) { + restorer = ksig->ka.sa.sa_restorer; + } else { + /* could use a vstub here */ + restorer = NULL; + err |= -EFAULT; + } + put_user_ex(restorer, &frame->pretcode); + } put_user_catch(err); - ret = do_sigaction(sig, act ? &new_ka : NULL, oact ? &old_ka : NULL); + err |= setup_sigcontext(&frame->uc.uc_mcontext, fpstate, + regs, set->sig[0]); + err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set)); - if (!ret && oact) { - if (!access_ok(VERIFY_WRITE, oact, sizeof(*oact))) - return -EFAULT; + if (err) + return -EFAULT; - put_user_try { - put_user_ex(old_ka.sa.sa_handler, &oact->sa_handler); - put_user_ex(old_ka.sa.sa_flags, &oact->sa_flags); - put_user_ex(old_ka.sa.sa_mask.sig[0], &oact->sa_mask); - put_user_ex(old_ka.sa.sa_restorer, &oact->sa_restorer); - } put_user_catch(ret); + /* Set up registers for signal handler */ + regs->sp = (unsigned long) frame; + regs->ip = (unsigned long) ksig->ka.sa.sa_handler; - if (ret) - return -EFAULT; - } + /* We use the x32 calling convention here... */ + regs->di = ksig->sig; + regs->si = (unsigned long) &frame->info; + regs->dx = (unsigned long) &frame->uc; - return ret; -} -#endif /* CONFIG_X86_32 */ + loadsegment(ds, __USER_DS); + loadsegment(es, __USER_DS); -long -sys_sigaltstack(const stack_t __user *uss, stack_t __user *uoss, - struct pt_regs *regs) -{ - return do_sigaltstack(uss, uoss, regs->sp); + regs->cs = __USER_CS; + regs->ss = __USER_DS; +#endif /* CONFIG_X86_X32_ABI */ + + return 0; } /* * Do a signal return; undo the signal stack. */ #ifdef CONFIG_X86_32 -unsigned long sys_sigreturn(struct pt_regs *regs) +asmlinkage unsigned long sys_sigreturn(void) { + struct pt_regs *regs = current_pt_regs(); struct sigframe __user *frame; unsigned long ax; sigset_t set; @@ -572,7 +551,6 @@ unsigned long sys_sigreturn(struct pt_regs *regs) sizeof(frame->extramask)))) goto badframe; - sigdelsetmask(&set, ~_BLOCKABLE); set_current_blocked(&set); if (restore_sigcontext(regs, &frame->sc, &ax)) @@ -586,8 +564,9 @@ badframe: } #endif /* CONFIG_X86_32 */ -long sys_rt_sigreturn(struct pt_regs *regs) +asmlinkage long sys_rt_sigreturn(void) { + struct pt_regs *regs = current_pt_regs(); struct rt_sigframe __user *frame; unsigned long ax; sigset_t set; @@ -598,13 +577,12 @@ long sys_rt_sigreturn(struct pt_regs *regs) if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set))) goto badframe; - sigdelsetmask(&set, ~_BLOCKABLE); set_current_blocked(&set); if (restore_sigcontext(regs, &frame->uc.uc_mcontext, &ax)) goto badframe; - if (do_sigaltstack(&frame->uc.uc_stack, NULL, regs->sp) == -EFAULT) + if (restore_altstack(&frame->uc.uc_stack)) goto badframe; return ax; @@ -628,62 +606,30 @@ static int signr_convert(int sig) return sig; } -#ifdef CONFIG_X86_32 - -#define is_ia32 1 -#define ia32_setup_frame __setup_frame -#define ia32_setup_rt_frame __setup_rt_frame - -#else /* !CONFIG_X86_32 */ - -#ifdef CONFIG_IA32_EMULATION -#define is_ia32 test_thread_flag(TIF_IA32) -#else /* !CONFIG_IA32_EMULATION */ -#define is_ia32 0 -#endif /* CONFIG_IA32_EMULATION */ - -int ia32_setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, - sigset_t *set, struct pt_regs *regs); -int ia32_setup_frame(int sig, struct k_sigaction *ka, - sigset_t *set, struct pt_regs *regs); - -#endif /* CONFIG_X86_32 */ - static int -setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info, - struct pt_regs *regs) +setup_rt_frame(struct ksignal *ksig, struct pt_regs *regs) { - int usig = signr_convert(sig); - sigset_t *set = ¤t->blocked; - int ret; - - if (current_thread_info()->status & TS_RESTORE_SIGMASK) - set = ¤t->saved_sigmask; + int usig = signr_convert(ksig->sig); + sigset_t *set = sigmask_to_save(); + compat_sigset_t *cset = (compat_sigset_t *) set; /* Set up the stack frame */ - if (is_ia32) { - if (ka->sa.sa_flags & SA_SIGINFO) - ret = ia32_setup_rt_frame(usig, ka, info, set, regs); + if (is_ia32_frame()) { + if (ksig->ka.sa.sa_flags & SA_SIGINFO) + return ia32_setup_rt_frame(usig, ksig, cset, regs); else - ret = ia32_setup_frame(usig, ka, set, regs); - } else - ret = __setup_rt_frame(sig, ka, info, set, regs); - - if (ret) { - force_sigsegv(sig, current); - return -EFAULT; + return ia32_setup_frame(usig, ksig, cset, regs); + } else if (is_x32_frame()) { + return x32_setup_rt_frame(ksig, cset, regs); + } else { + return __setup_rt_frame(ksig->sig, ksig, set, regs); } - - current_thread_info()->status &= ~TS_RESTORE_SIGMASK; - return ret; } -static int -handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka, - struct pt_regs *regs) +static void +handle_signal(struct ksignal *ksig, struct pt_regs *regs) { - int ret; - + bool failed; /* Are we from a system call? */ if (syscall_get_nr(current, regs) >= 0) { /* If so, check system call restarting.. */ @@ -694,7 +640,7 @@ handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka, break; case -ERESTARTSYS: - if (!(ka->sa.sa_flags & SA_RESTART)) { + if (!(ksig->ka.sa.sa_flags & SA_RESTART)) { regs->ax = -EINTR; break; } @@ -714,30 +660,23 @@ handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka, likely(test_and_clear_thread_flag(TIF_FORCED_TF))) regs->flags &= ~X86_EFLAGS_TF; - ret = setup_rt_frame(sig, ka, info, regs); - - if (ret) - return ret; - - /* - * Clear the direction flag as per the ABI for function entry. - */ - regs->flags &= ~X86_EFLAGS_DF; - - /* - * Clear TF when entering the signal handler, but - * notify any tracer that was single-stepping it. - * The tracer may want to single-step inside the - * handler too. - */ - regs->flags &= ~X86_EFLAGS_TF; - - block_sigmask(ka, sig); - - tracehook_signal_handler(sig, info, ka, regs, - test_thread_flag(TIF_SINGLESTEP)); - - return 0; + failed = (setup_rt_frame(ksig, regs) < 0); + if (!failed) { + /* + * Clear the direction flag as per the ABI for function entry. + * + * Clear RF when entering the signal handler, because + * it might disable possible debug exception from the + * signal handler. + * + * Clear TF when entering the signal handler, but + * notify any tracer that was single-stepping it. + * The tracer may want to single-step inside the + * handler too. + */ + regs->flags &= ~(X86_EFLAGS_DF|X86_EFLAGS_RF|X86_EFLAGS_TF); + } + signal_setup_done(failed, ksig, test_thread_flag(TIF_SINGLESTEP)); } #ifdef CONFIG_X86_32 @@ -754,24 +693,11 @@ handle_signal(unsigned long sig, siginfo_t *info, struct k_sigaction *ka, */ static void do_signal(struct pt_regs *regs) { - struct k_sigaction ka; - siginfo_t info; - int signr; + struct ksignal ksig; - /* - * We want the common case to go fast, which is why we may in certain - * cases get here from kernel mode. Just return without doing anything - * if so. - * X86_32: vm86 regs switched out by assembly code before reaching - * here, so testing against kernel CS suffices. - */ - if (!user_mode(regs)) - return; - - signr = get_signal_to_deliver(&info, &ka, regs, NULL); - if (signr > 0) { + if (get_signal(&ksig)) { /* Whee! Actually deliver the signal. */ - handle_signal(signr, &info, &ka, regs); + handle_signal(&ksig, regs); return; } @@ -797,25 +723,27 @@ static void do_signal(struct pt_regs *regs) * If there's no signal to deliver, we just put the saved sigmask * back. */ - if (current_thread_info()->status & TS_RESTORE_SIGMASK) { - current_thread_info()->status &= ~TS_RESTORE_SIGMASK; - set_current_blocked(¤t->saved_sigmask); - } + restore_saved_sigmask(); } /* * notification of userspace execution resumption * - triggered by the TIF_WORK_MASK flags */ -void +__visible void do_notify_resume(struct pt_regs *regs, void *unused, __u32 thread_info_flags) { + user_exit(); + #ifdef CONFIG_X86_MCE /* notify userspace of pending MCEs */ if (thread_info_flags & _TIF_MCE_NOTIFY) mce_notify_process(); #endif /* CONFIG_X86_64 && CONFIG_X86_MCE */ + if (thread_info_flags & _TIF_UPROBE) + uprobe_notify_resume(regs); + /* deal with pending signal delivery */ if (thread_info_flags & _TIF_SIGPENDING) do_signal(regs); @@ -823,15 +751,11 @@ do_notify_resume(struct pt_regs *regs, void *unused, __u32 thread_info_flags) if (thread_info_flags & _TIF_NOTIFY_RESUME) { clear_thread_flag(TIF_NOTIFY_RESUME); tracehook_notify_resume(regs); - if (current->replacement_session_keyring) - key_replace_session_keyring(); } if (thread_info_flags & _TIF_USER_RETURN_NOTIFY) fire_user_return_notifiers(); -#ifdef CONFIG_X86_32 - clear_thread_flag(TIF_IRET); -#endif /* CONFIG_X86_32 */ + user_enter(); } void signal_fault(struct pt_regs *regs, void __user *frame, char *where) @@ -845,8 +769,39 @@ void signal_fault(struct pt_regs *regs, void __user *frame, char *where) me->comm, me->pid, where, frame, regs->ip, regs->sp, regs->orig_ax); print_vma_addr(" in ", regs->ip); - printk(KERN_CONT "\n"); + pr_cont("\n"); } force_sig(SIGSEGV, me); } + +#ifdef CONFIG_X86_X32_ABI +asmlinkage long sys32_x32_rt_sigreturn(void) +{ + struct pt_regs *regs = current_pt_regs(); + struct rt_sigframe_x32 __user *frame; + sigset_t set; + unsigned long ax; + + frame = (struct rt_sigframe_x32 __user *)(regs->sp - 8); + + if (!access_ok(VERIFY_READ, frame, sizeof(*frame))) + goto badframe; + if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set))) + goto badframe; + + set_current_blocked(&set); + + if (restore_sigcontext(regs, &frame->uc.uc_mcontext, &ax)) + goto badframe; + + if (compat_restore_altstack(&frame->uc.uc_stack)) + goto badframe; + + return ax; + +badframe: + signal_fault(regs, frame, "x32 rt_sigreturn"); + return 0; +} +#endif diff --git a/arch/x86/kernel/smp.c b/arch/x86/kernel/smp.c index 66c74f481ca..be8e1bde07a 100644 --- a/arch/x86/kernel/smp.c +++ b/arch/x86/kernel/smp.c @@ -30,6 +30,7 @@ #include <asm/proto.h> #include <asm/apic.h> #include <asm/nmi.h> +#include <asm/trace/irq_vectors.h> /* * Some notes on x86 processor bugs affecting SMP operation: * @@ -109,6 +110,9 @@ * about nothing of note with C stepping upwards. */ +static atomic_t stopping_cpu = ATOMIC_INIT(-1); +static bool smp_no_nmi_ipi = false; + /* * this function sends a 'reschedule' IPI to another CPU. * it goes straight through and wastes no time serializing @@ -149,8 +153,6 @@ void native_send_call_func_ipi(const struct cpumask *mask) free_cpumask_var(allbutself); } -static atomic_t stopping_cpu = ATOMIC_INIT(-1); - static int smp_stop_nmi_callback(unsigned int val, struct pt_regs *regs) { /* We are registered on stopping cpu too, avoid spurious NMI */ @@ -162,7 +164,19 @@ static int smp_stop_nmi_callback(unsigned int val, struct pt_regs *regs) return NMI_HANDLED; } -static void native_nmi_stop_other_cpus(int wait) +/* + * this function calls the 'stop' function on all other CPUs in the system. + */ + +asmlinkage __visible void smp_reboot_interrupt(void) +{ + ack_APIC_irq(); + irq_enter(); + stop_this_cpu(NULL); + irq_exit(); +} + +static void native_stop_other_cpus(int wait) { unsigned long flags; unsigned long timeout; @@ -174,20 +188,25 @@ static void native_nmi_stop_other_cpus(int wait) * Use an own vector here because smp_call_function * does lots of things not suitable in a panic situation. */ + + /* + * We start by using the REBOOT_VECTOR irq. + * The irq is treated as a sync point to allow critical + * regions of code on other cpus to release their spin locks + * and re-enable irqs. Jumping straight to an NMI might + * accidentally cause deadlocks with further shutdown/panic + * code. By syncing, we give the cpus up to one second to + * finish their work before we force them off with the NMI. + */ if (num_online_cpus() > 1) { /* did someone beat us here? */ if (atomic_cmpxchg(&stopping_cpu, -1, safe_smp_processor_id()) != -1) return; - if (register_nmi_handler(NMI_LOCAL, smp_stop_nmi_callback, - NMI_FLAG_FIRST, "smp_stop")) - /* Note: we ignore failures here */ - return; - - /* sync above data before sending NMI */ + /* sync above data before sending IRQ */ wmb(); - apic->send_IPI_allbutself(NMI_VECTOR); + apic->send_IPI_allbutself(REBOOT_VECTOR); /* * Don't wait longer than a second if the caller @@ -197,98 +216,127 @@ static void native_nmi_stop_other_cpus(int wait) while (num_online_cpus() > 1 && (wait || timeout--)) udelay(1); } + + /* if the REBOOT_VECTOR didn't work, try with the NMI */ + if ((num_online_cpus() > 1) && (!smp_no_nmi_ipi)) { + if (register_nmi_handler(NMI_LOCAL, smp_stop_nmi_callback, + NMI_FLAG_FIRST, "smp_stop")) + /* Note: we ignore failures here */ + /* Hope the REBOOT_IRQ is good enough */ + goto finish; - local_irq_save(flags); - disable_local_APIC(); - local_irq_restore(flags); -} - -/* - * this function calls the 'stop' function on all other CPUs in the system. - */ - -asmlinkage void smp_reboot_interrupt(void) -{ - ack_APIC_irq(); - irq_enter(); - stop_this_cpu(NULL); - irq_exit(); -} - -static void native_irq_stop_other_cpus(int wait) -{ - unsigned long flags; - unsigned long timeout; + /* sync above data before sending IRQ */ + wmb(); - if (reboot_force) - return; + pr_emerg("Shutting down cpus with NMI\n"); - /* - * Use an own vector here because smp_call_function - * does lots of things not suitable in a panic situation. - * On most systems we could also use an NMI here, - * but there are a few systems around where NMI - * is problematic so stay with an non NMI for now - * (this implies we cannot stop CPUs spinning with irq off - * currently) - */ - if (num_online_cpus() > 1) { - apic->send_IPI_allbutself(REBOOT_VECTOR); + apic->send_IPI_allbutself(NMI_VECTOR); /* - * Don't wait longer than a second if the caller + * Don't wait longer than a 10 ms if the caller * didn't ask us to wait. */ - timeout = USEC_PER_SEC; + timeout = USEC_PER_MSEC * 10; while (num_online_cpus() > 1 && (wait || timeout--)) udelay(1); } +finish: local_irq_save(flags); disable_local_APIC(); local_irq_restore(flags); } -static void native_smp_disable_nmi_ipi(void) -{ - smp_ops.stop_other_cpus = native_irq_stop_other_cpus; -} - /* * Reschedule call back. */ -void smp_reschedule_interrupt(struct pt_regs *regs) +static inline void __smp_reschedule_interrupt(void) { - ack_APIC_irq(); inc_irq_stat(irq_resched_count); scheduler_ipi(); +} + +__visible void smp_reschedule_interrupt(struct pt_regs *regs) +{ + ack_APIC_irq(); + __smp_reschedule_interrupt(); /* * KVM uses this interrupt to force a cpu out of guest mode */ } -void smp_call_function_interrupt(struct pt_regs *regs) +static inline void smp_entering_irq(void) { ack_APIC_irq(); irq_enter(); +} + +__visible void smp_trace_reschedule_interrupt(struct pt_regs *regs) +{ + /* + * Need to call irq_enter() before calling the trace point. + * __smp_reschedule_interrupt() calls irq_enter/exit() too (in + * scheduler_ipi(). This is OK, since those functions are allowed + * to nest. + */ + smp_entering_irq(); + trace_reschedule_entry(RESCHEDULE_VECTOR); + __smp_reschedule_interrupt(); + trace_reschedule_exit(RESCHEDULE_VECTOR); + exiting_irq(); + /* + * KVM uses this interrupt to force a cpu out of guest mode + */ +} + +static inline void __smp_call_function_interrupt(void) +{ generic_smp_call_function_interrupt(); inc_irq_stat(irq_call_count); - irq_exit(); } -void smp_call_function_single_interrupt(struct pt_regs *regs) +__visible void smp_call_function_interrupt(struct pt_regs *regs) +{ + smp_entering_irq(); + __smp_call_function_interrupt(); + exiting_irq(); +} + +__visible void smp_trace_call_function_interrupt(struct pt_regs *regs) +{ + smp_entering_irq(); + trace_call_function_entry(CALL_FUNCTION_VECTOR); + __smp_call_function_interrupt(); + trace_call_function_exit(CALL_FUNCTION_VECTOR); + exiting_irq(); +} + +static inline void __smp_call_function_single_interrupt(void) { - ack_APIC_irq(); - irq_enter(); generic_smp_call_function_single_interrupt(); inc_irq_stat(irq_call_count); - irq_exit(); +} + +__visible void smp_call_function_single_interrupt(struct pt_regs *regs) +{ + smp_entering_irq(); + __smp_call_function_single_interrupt(); + exiting_irq(); +} + +__visible void smp_trace_call_function_single_interrupt(struct pt_regs *regs) +{ + smp_entering_irq(); + trace_call_function_single_entry(CALL_FUNCTION_SINGLE_VECTOR); + __smp_call_function_single_interrupt(); + trace_call_function_single_exit(CALL_FUNCTION_SINGLE_VECTOR); + exiting_irq(); } static int __init nonmi_ipi_setup(char *str) { - native_smp_disable_nmi_ipi(); - return 1; + smp_no_nmi_ipi = true; + return 1; } __setup("nonmi_ipi", nonmi_ipi_setup); @@ -298,7 +346,7 @@ struct smp_ops smp_ops = { .smp_prepare_cpus = native_smp_prepare_cpus, .smp_cpus_done = native_smp_cpus_done, - .stop_other_cpus = native_nmi_stop_other_cpus, + .stop_other_cpus = native_stop_other_cpus, .smp_send_reschedule = native_smp_send_reschedule, .cpu_up = native_cpu_up, diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 66d250c00d1..5492798930e 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -1,4 +1,4 @@ -/* + /* * x86 SMP booting functions * * (c) 1995 Alan Cox, Building #3 <alan@lxorguk.ukuu.org.uk> @@ -39,6 +39,8 @@ * Glauber Costa : i386 and x86_64 integration */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/init.h> #include <linux/smp.h> #include <linux/module.h> @@ -50,13 +52,14 @@ #include <linux/tboot.h> #include <linux/stackprotector.h> #include <linux/gfp.h> +#include <linux/cpuidle.h> #include <asm/acpi.h> #include <asm/desc.h> #include <asm/nmi.h> #include <asm/irq.h> #include <asm/idle.h> -#include <asm/trampoline.h> +#include <asm/realmode.h> #include <asm/cpu.h> #include <asm/numa.h> #include <asm/pgtable.h> @@ -65,69 +68,35 @@ #include <asm/mwait.h> #include <asm/apic.h> #include <asm/io_apic.h> +#include <asm/i387.h> +#include <asm/fpu-internal.h> #include <asm/setup.h> #include <asm/uv/uv.h> #include <linux/mc146818rtc.h> - #include <asm/smpboot_hooks.h> #include <asm/i8259.h> +#include <asm/realmode.h> +#include <asm/misc.h> /* State of each CPU */ DEFINE_PER_CPU(int, cpu_state) = { 0 }; -/* Store all idle threads, this can be reused instead of creating -* a new thread. Also avoids complicated thread destroy functionality -* for idle threads. -*/ -#ifdef CONFIG_HOTPLUG_CPU -/* - * Needed only for CONFIG_HOTPLUG_CPU because __cpuinitdata is - * removed after init for !CONFIG_HOTPLUG_CPU. - */ -static DEFINE_PER_CPU(struct task_struct *, idle_thread_array); -#define get_idle_for_cpu(x) (per_cpu(idle_thread_array, x)) -#define set_idle_for_cpu(x, p) (per_cpu(idle_thread_array, x) = (p)) - -/* - * We need this for trampoline_base protection from concurrent accesses when - * off- and onlining cores wildly. - */ -static DEFINE_MUTEX(x86_cpu_hotplug_driver_mutex); - -void cpu_hotplug_driver_lock(void) -{ - mutex_lock(&x86_cpu_hotplug_driver_mutex); -} - -void cpu_hotplug_driver_unlock(void) -{ - mutex_unlock(&x86_cpu_hotplug_driver_mutex); -} - -ssize_t arch_cpu_probe(const char *buf, size_t count) { return -1; } -ssize_t arch_cpu_release(const char *buf, size_t count) { return -1; } -#else -static struct task_struct *idle_thread_array[NR_CPUS] __cpuinitdata ; -#define get_idle_for_cpu(x) (idle_thread_array[(x)]) -#define set_idle_for_cpu(x, p) (idle_thread_array[(x)] = (p)) -#endif - /* Number of siblings per CPU package */ int smp_num_siblings = 1; EXPORT_SYMBOL(smp_num_siblings); /* Last level cache ID of each logical CPU */ -DEFINE_PER_CPU(u16, cpu_llc_id) = BAD_APICID; +DEFINE_PER_CPU_READ_MOSTLY(u16, cpu_llc_id) = BAD_APICID; /* representing HT siblings of each logical CPU */ -DEFINE_PER_CPU(cpumask_var_t, cpu_sibling_map); +DEFINE_PER_CPU_READ_MOSTLY(cpumask_var_t, cpu_sibling_map); EXPORT_PER_CPU_SYMBOL(cpu_sibling_map); /* representing HT and core siblings of each logical CPU */ -DEFINE_PER_CPU(cpumask_var_t, cpu_core_map); +DEFINE_PER_CPU_READ_MOSTLY(cpumask_var_t, cpu_core_map); EXPORT_PER_CPU_SYMBOL(cpu_core_map); -DEFINE_PER_CPU(cpumask_var_t, cpu_llc_shared_map); +DEFINE_PER_CPU_READ_MOSTLY(cpumask_var_t, cpu_llc_shared_map); /* Per CPU bogomips and other parameters */ DEFINE_PER_CPU_SHARED_ALIGNED(struct cpuinfo_x86, cpu_info); @@ -136,10 +105,10 @@ EXPORT_PER_CPU_SYMBOL(cpu_info); atomic_t init_deasserted; /* - * Report back to the Boot Processor. - * Running on AP. + * Report back to the Boot Processor during boot time or to the caller processor + * during CPU online. */ -static void __cpuinit smp_callin(void) +static void smp_callin(void) { int cpuid, phys_id; unsigned long timeout; @@ -149,15 +118,18 @@ static void __cpuinit smp_callin(void) * we may get here before an INIT-deassert IPI reaches * our local APIC. We have to wait for the IPI or we'll * lock up on an APIC access. + * + * Since CPU0 is not wakened up by INIT, it doesn't wait for the IPI. */ - if (apic->wait_for_init_deassert) - apic->wait_for_init_deassert(&init_deasserted); + cpuid = smp_processor_id(); + if (apic->wait_for_init_deassert && cpuid) + while (!atomic_read(&init_deasserted)) + cpu_relax(); /* * (This works even if the APIC is not enabled.) */ phys_id = read_apic_id(); - cpuid = smp_processor_id(); if (cpumask_test_cpu(cpuid, cpu_callin_mask)) { panic("%s: phys CPU#%d, CPU#%d already present??\n", __func__, phys_id, cpuid); @@ -197,7 +169,7 @@ static void __cpuinit smp_callin(void) * boards) */ - pr_debug("CALLIN, before setup_local_APIC().\n"); + pr_debug("CALLIN, before setup_local_APIC()\n"); if (apic->smp_callin_clear_local_apic) apic->smp_callin_clear_local_apic(); setup_local_APIC(); @@ -219,14 +191,9 @@ static void __cpuinit smp_callin(void) * Update loops_per_jiffy in cpu_data. Previous call to * smp_store_cpu_info() stored a value that is close but not as * accurate as the value just calculated. - * - * Need to enable IRQs because it can take longer and then - * the NMI watchdog might kill us. */ - local_irq_enable(); calibrate_delay(); cpu_data(cpuid).loops_per_jiffy = loops_per_jiffy; - local_irq_disable(); pr_debug("Stack at about %p\n", &cpuid); /* @@ -244,10 +211,12 @@ static void __cpuinit smp_callin(void) cpumask_set_cpu(cpuid, cpu_callin_mask); } +static int cpu0_logical_apicid; +static int enable_start_cpu0; /* * Activate a secondary processor. */ -notrace static void __cpuinit start_secondary(void *unused) +static void notrace start_secondary(void *unused) { /* * Don't put *anything* before cpu_init(), SMP booting is too @@ -255,9 +224,12 @@ notrace static void __cpuinit start_secondary(void *unused) * most necessary things. */ cpu_init(); + x86_cpuinit.early_percpu_clock_init(); preempt_disable(); smp_callin(); + enable_start_cpu0 = 0; + #ifdef CONFIG_X86_32 /* switch away from the initial page table */ load_cr3(swapper_pg_dir); @@ -272,38 +244,23 @@ notrace static void __cpuinit start_secondary(void *unused) check_tsc_sync_target(); /* - * We need to hold call_lock, so there is no inconsistency - * between the time smp_call_function() determines number of - * IPI recipients, and the time when the determination is made - * for which cpus receive the IPI. Holding this - * lock helps us to not include this cpu in a currently in progress - * smp_call_function(). - * + * Enable the espfix hack for this CPU + */ +#ifdef CONFIG_X86_ESPFIX64 + init_espfix_ap(); +#endif + + /* * We need to hold vector_lock so there the set of online cpus * does not change while we are assigning vectors to cpus. Holding * this lock ensures we don't half assign or remove an irq from a cpu. */ - ipi_call_lock(); lock_vector_lock(); set_cpu_online(smp_processor_id(), true); unlock_vector_lock(); - ipi_call_unlock(); per_cpu(cpu_state, smp_processor_id()) = CPU_ONLINE; x86_platform.nmi_init(); - /* - * Wait until the cpu which brought this one up marked it - * online before enabling interrupts. If we don't do that then - * we can end up waking up the softirq thread before this cpu - * reached the active state, which makes the scheduler unhappy - * and schedule the softirq thread on the wrong cpu. This is - * only observable with forced threaded interrupts, but in - * theory it could also happen w/o them. It's just way harder - * to achieve. - */ - while (!cpumask_test_cpu(smp_processor_id(), cpu_active_mask)) - cpu_relax(); - /* enable local interrupts */ local_irq_enable(); @@ -313,77 +270,131 @@ notrace static void __cpuinit start_secondary(void *unused) x86_cpuinit.setup_percpu_clockev(); wmb(); - cpu_idle(); + cpu_startup_entry(CPUHP_ONLINE); +} + +void __init smp_store_boot_cpu_info(void) +{ + int id = 0; /* CPU 0 */ + struct cpuinfo_x86 *c = &cpu_data(id); + + *c = boot_cpu_data; + c->cpu_index = id; } /* * The bootstrap kernel entry code has set these up. Save them for * a given CPU */ - -void __cpuinit smp_store_cpu_info(int id) +void smp_store_cpu_info(int id) { struct cpuinfo_x86 *c = &cpu_data(id); *c = boot_cpu_data; c->cpu_index = id; - if (id != 0) - identify_secondary_cpu(c); + /* + * During boot time, CPU0 has this setup already. Save the info when + * bringing up AP or offlined CPU0. + */ + identify_secondary_cpu(c); } -static void __cpuinit link_thread_siblings(int cpu1, int cpu2) +static bool +topology_sane(struct cpuinfo_x86 *c, struct cpuinfo_x86 *o, const char *name) { - cpumask_set_cpu(cpu1, cpu_sibling_mask(cpu2)); - cpumask_set_cpu(cpu2, cpu_sibling_mask(cpu1)); - cpumask_set_cpu(cpu1, cpu_core_mask(cpu2)); - cpumask_set_cpu(cpu2, cpu_core_mask(cpu1)); - cpumask_set_cpu(cpu1, cpu_llc_shared_mask(cpu2)); - cpumask_set_cpu(cpu2, cpu_llc_shared_mask(cpu1)); + int cpu1 = c->cpu_index, cpu2 = o->cpu_index; + + return !WARN_ONCE(cpu_to_node(cpu1) != cpu_to_node(cpu2), + "sched: CPU #%d's %s-sibling CPU #%d is not on the same node! " + "[node: %d != %d]. Ignoring dependency.\n", + cpu1, name, cpu2, cpu_to_node(cpu1), cpu_to_node(cpu2)); } +#define link_mask(_m, c1, c2) \ +do { \ + cpumask_set_cpu((c1), cpu_##_m##_mask(c2)); \ + cpumask_set_cpu((c2), cpu_##_m##_mask(c1)); \ +} while (0) -void __cpuinit set_cpu_sibling_map(int cpu) +static bool match_smt(struct cpuinfo_x86 *c, struct cpuinfo_x86 *o) { - int i; + if (cpu_has_topoext) { + int cpu1 = c->cpu_index, cpu2 = o->cpu_index; + + if (c->phys_proc_id == o->phys_proc_id && + per_cpu(cpu_llc_id, cpu1) == per_cpu(cpu_llc_id, cpu2) && + c->compute_unit_id == o->compute_unit_id) + return topology_sane(c, o, "smt"); + + } else if (c->phys_proc_id == o->phys_proc_id && + c->cpu_core_id == o->cpu_core_id) { + return topology_sane(c, o, "smt"); + } + + return false; +} + +static bool match_llc(struct cpuinfo_x86 *c, struct cpuinfo_x86 *o) +{ + int cpu1 = c->cpu_index, cpu2 = o->cpu_index; + + if (per_cpu(cpu_llc_id, cpu1) != BAD_APICID && + per_cpu(cpu_llc_id, cpu1) == per_cpu(cpu_llc_id, cpu2)) + return topology_sane(c, o, "llc"); + + return false; +} + +static bool match_mc(struct cpuinfo_x86 *c, struct cpuinfo_x86 *o) +{ + if (c->phys_proc_id == o->phys_proc_id) { + if (cpu_has(c, X86_FEATURE_AMD_DCM)) + return true; + + return topology_sane(c, o, "mc"); + } + return false; +} + +void set_cpu_sibling_map(int cpu) +{ + bool has_smt = smp_num_siblings > 1; + bool has_mp = has_smt || boot_cpu_data.x86_max_cores > 1; struct cpuinfo_x86 *c = &cpu_data(cpu); + struct cpuinfo_x86 *o; + int i; cpumask_set_cpu(cpu, cpu_sibling_setup_mask); - if (smp_num_siblings > 1) { - for_each_cpu(i, cpu_sibling_setup_mask) { - struct cpuinfo_x86 *o = &cpu_data(i); - - if (cpu_has(c, X86_FEATURE_TOPOEXT)) { - if (c->phys_proc_id == o->phys_proc_id && - per_cpu(cpu_llc_id, cpu) == per_cpu(cpu_llc_id, i) && - c->compute_unit_id == o->compute_unit_id) - link_thread_siblings(cpu, i); - } else if (c->phys_proc_id == o->phys_proc_id && - c->cpu_core_id == o->cpu_core_id) { - link_thread_siblings(cpu, i); - } - } - } else { + if (!has_mp) { cpumask_set_cpu(cpu, cpu_sibling_mask(cpu)); + cpumask_set_cpu(cpu, cpu_llc_shared_mask(cpu)); + cpumask_set_cpu(cpu, cpu_core_mask(cpu)); + c->booted_cores = 1; + return; } - cpumask_set_cpu(cpu, cpu_llc_shared_mask(cpu)); + for_each_cpu(i, cpu_sibling_setup_mask) { + o = &cpu_data(i); + + if ((i == cpu) || (has_smt && match_smt(c, o))) + link_mask(sibling, cpu, i); + + if ((i == cpu) || (has_mp && match_llc(c, o))) + link_mask(llc_shared, cpu, i); - if (__this_cpu_read(cpu_info.x86_max_cores) == 1) { - cpumask_copy(cpu_core_mask(cpu), cpu_sibling_mask(cpu)); - c->booted_cores = 1; - return; } + /* + * This needs a separate iteration over the cpus because we rely on all + * cpu_sibling_mask links to be set-up. + */ for_each_cpu(i, cpu_sibling_setup_mask) { - if (per_cpu(cpu_llc_id, cpu) != BAD_APICID && - per_cpu(cpu_llc_id, cpu) == per_cpu(cpu_llc_id, i)) { - cpumask_set_cpu(i, cpu_llc_shared_mask(cpu)); - cpumask_set_cpu(cpu, cpu_llc_shared_mask(i)); - } - if (c->phys_proc_id == cpu_data(i).phys_proc_id) { - cpumask_set_cpu(i, cpu_core_mask(cpu)); - cpumask_set_cpu(cpu, cpu_core_mask(i)); + o = &cpu_data(i); + + if ((i == cpu) || (has_mp && match_mc(c, o))) { + link_mask(core, cpu, i); + /* * Does this new cpu bringup a new core? */ @@ -409,16 +420,7 @@ void __cpuinit set_cpu_sibling_map(int cpu) /* maps the cpu to the sched domain representing multi-core */ const struct cpumask *cpu_coregroup_mask(int cpu) { - struct cpuinfo_x86 *c = &cpu_data(cpu); - /* - * For perf, we return last level cache shared map. - * And for power savings, we return cpu_core_map - */ - if ((sched_mc_power_savings || sched_smt_power_savings) && - !(cpu_has(c, X86_FEATURE_AMD_DCM))) - return cpu_core_mask(cpu); - else - return cpu_llc_shared_mask(cpu); + return cpu_llc_shared_mask(cpu); } static void impress_friends(void) @@ -428,17 +430,16 @@ static void impress_friends(void) /* * Allow the user to impress friends. */ - pr_debug("Before bogomips.\n"); + pr_debug("Before bogomips\n"); for_each_possible_cpu(cpu) if (cpumask_test_cpu(cpu, cpu_callout_mask)) bogosum += cpu_data(cpu).loops_per_jiffy; - printk(KERN_INFO - "Total of %d processors activated (%lu.%02lu BogoMIPS).\n", + pr_info("Total of %d processors activated (%lu.%02lu BogoMIPS)\n", num_online_cpus(), bogosum/(500000/HZ), (bogosum/(5000/HZ))%100); - pr_debug("Before bogocount - setting activated=1.\n"); + pr_debug("Before bogocount - setting activated=1\n"); } void __inquire_remote_apic(int apicid) @@ -448,18 +449,17 @@ void __inquire_remote_apic(int apicid) int timeout; u32 status; - printk(KERN_INFO "Inquiring remote APIC 0x%x...\n", apicid); + pr_info("Inquiring remote APIC 0x%x...\n", apicid); for (i = 0; i < ARRAY_SIZE(regs); i++) { - printk(KERN_INFO "... APIC 0x%x %s: ", apicid, names[i]); + pr_info("... APIC 0x%x %s: ", apicid, names[i]); /* * Wait for idle. */ status = safe_apic_wait_icr_idle(); if (status) - printk(KERN_CONT - "a previous APIC delivery may have failed\n"); + pr_cont("a previous APIC delivery may have failed\n"); apic_icr_write(APIC_DM_REMRD | regs[i], apicid); @@ -472,10 +472,10 @@ void __inquire_remote_apic(int apicid) switch (status) { case APIC_ICR_RR_VALID: status = apic_read(APIC_RRR); - printk(KERN_CONT "%08x\n", status); + pr_cont("%08x\n", status); break; default: - printk(KERN_CONT "failed\n"); + pr_cont("failed\n"); } } } @@ -485,8 +485,8 @@ void __inquire_remote_apic(int apicid) * INIT, INIT, STARTUP sequence will reset the chip hard for us, and this * won't ... remember to clear down the APIC, etc later. */ -int __cpuinit -wakeup_secondary_cpu_via_nmi(int logical_apicid, unsigned long start_eip) +int +wakeup_secondary_cpu_via_nmi(int apicid, unsigned long start_eip) { unsigned long send_status, accept_status = 0; int maxlvt; @@ -494,7 +494,7 @@ wakeup_secondary_cpu_via_nmi(int logical_apicid, unsigned long start_eip) /* Target chip */ /* Boot on the stack */ /* Kick the second */ - apic_icr_write(APIC_DM_NMI | apic->dest_logical, logical_apicid); + apic_icr_write(APIC_DM_NMI | apic->dest_logical, apicid); pr_debug("Waiting for send to finish...\n"); send_status = safe_apic_wait_icr_idle(); @@ -509,17 +509,17 @@ wakeup_secondary_cpu_via_nmi(int logical_apicid, unsigned long start_eip) apic_write(APIC_ESR, 0); accept_status = (apic_read(APIC_ESR) & 0xEF); } - pr_debug("NMI sent.\n"); + pr_debug("NMI sent\n"); if (send_status) - printk(KERN_ERR "APIC never delivered???\n"); + pr_err("APIC never delivered???\n"); if (accept_status) - printk(KERN_ERR "APIC delivery error (%lx).\n", accept_status); + pr_err("APIC delivery error (%lx)\n", accept_status); return (send_status | accept_status); } -static int __cpuinit +static int wakeup_secondary_cpu_via_init(int phys_apicid, unsigned long start_eip) { unsigned long send_status, accept_status = 0; @@ -536,7 +536,7 @@ wakeup_secondary_cpu_via_init(int phys_apicid, unsigned long start_eip) apic_read(APIC_ESR); } - pr_debug("Asserting INIT.\n"); + pr_debug("Asserting INIT\n"); /* * Turn INIT on target chip @@ -552,7 +552,7 @@ wakeup_secondary_cpu_via_init(int phys_apicid, unsigned long start_eip) mdelay(10); - pr_debug("Deasserting INIT.\n"); + pr_debug("Deasserting INIT\n"); /* Target chip */ /* Send IPI */ @@ -585,14 +585,14 @@ wakeup_secondary_cpu_via_init(int phys_apicid, unsigned long start_eip) /* * Run STARTUP IPI loop. */ - pr_debug("#startup loops: %d.\n", num_starts); + pr_debug("#startup loops: %d\n", num_starts); for (j = 1; j <= num_starts; j++) { - pr_debug("Sending STARTUP #%d.\n", j); + pr_debug("Sending STARTUP #%d\n", j); if (maxlvt > 3) /* Due to the Pentium erratum 3AP. */ apic_write(APIC_ESR, 0); apic_read(APIC_ESR); - pr_debug("After apic_write.\n"); + pr_debug("After apic_write\n"); /* * STARTUP IPI @@ -609,7 +609,7 @@ wakeup_secondary_cpu_via_init(int phys_apicid, unsigned long start_eip) */ udelay(300); - pr_debug("Startup point 1.\n"); + pr_debug("Startup point 1\n"); pr_debug("Waiting for send to finish...\n"); send_status = safe_apic_wait_icr_idle(); @@ -624,113 +624,162 @@ wakeup_secondary_cpu_via_init(int phys_apicid, unsigned long start_eip) if (send_status || accept_status) break; } - pr_debug("After Startup.\n"); + pr_debug("After Startup\n"); if (send_status) - printk(KERN_ERR "APIC never delivered???\n"); + pr_err("APIC never delivered???\n"); if (accept_status) - printk(KERN_ERR "APIC delivery error (%lx).\n", accept_status); + pr_err("APIC delivery error (%lx)\n", accept_status); return (send_status | accept_status); } -struct create_idle { - struct work_struct work; - struct task_struct *idle; - struct completion done; - int cpu; -}; - -static void __cpuinit do_fork_idle(struct work_struct *work) +void smp_announce(void) { - struct create_idle *c_idle = - container_of(work, struct create_idle, work); + int num_nodes = num_online_nodes(); - c_idle->idle = fork_idle(c_idle->cpu); - complete(&c_idle->done); + printk(KERN_INFO "x86: Booted up %d node%s, %d CPUs\n", + num_nodes, (num_nodes > 1 ? "s" : ""), num_online_cpus()); } /* reduce the number of lines printed when booting a large cpu count system */ -static void __cpuinit announce_cpu(int cpu, int apicid) +static void announce_cpu(int cpu, int apicid) { static int current_node = -1; int node = early_cpu_to_node(cpu); + static int width, node_width; + + if (!width) + width = num_digits(num_possible_cpus()) + 1; /* + '#' sign */ + + if (!node_width) + node_width = num_digits(num_possible_nodes()) + 1; /* + '#' */ + + if (cpu == 1) + printk(KERN_INFO "x86: Booting SMP configuration:\n"); if (system_state == SYSTEM_BOOTING) { if (node != current_node) { if (current_node > (-1)) - pr_cont(" Ok.\n"); + pr_cont("\n"); current_node = node; - pr_info("Booting Node %3d, Processors ", node); + + printk(KERN_INFO ".... node %*s#%d, CPUs: ", + node_width - num_digits(node), " ", node); } - pr_cont(" #%d%s", cpu, cpu == (nr_cpu_ids - 1) ? " Ok.\n" : ""); - return; + + /* Add padding for the BSP */ + if (cpu == 1) + pr_cont("%*s", width + 1, " "); + + pr_cont("%*s#%d", width - num_digits(cpu), " ", cpu); + } else pr_info("Booting Node %d Processor %d APIC 0x%x\n", node, cpu, apicid); } -/* - * NOTE - on most systems this is a PHYSICAL apic ID, but on multiquad - * (ie clustered apic addressing mode), this is a LOGICAL apic ID. - * Returns zero if CPU booted OK, else error code from - * ->wakeup_secondary_cpu. - */ -static int __cpuinit do_boot_cpu(int apicid, int cpu) +static int wakeup_cpu0_nmi(unsigned int cmd, struct pt_regs *regs) { - unsigned long boot_error = 0; - unsigned long start_ip; - int timeout; - struct create_idle c_idle = { - .cpu = cpu, - .done = COMPLETION_INITIALIZER_ONSTACK(c_idle.done), - }; + int cpu; - INIT_WORK_ONSTACK(&c_idle.work, do_fork_idle); + cpu = smp_processor_id(); + if (cpu == 0 && !cpu_online(cpu) && enable_start_cpu0) + return NMI_HANDLED; - alternatives_smp_switch(1); + return NMI_DONE; +} + +/* + * Wake up AP by INIT, INIT, STARTUP sequence. + * + * Instead of waiting for STARTUP after INITs, BSP will execute the BIOS + * boot-strap code which is not a desired behavior for waking up BSP. To + * void the boot-strap code, wake up CPU0 by NMI instead. + * + * This works to wake up soft offlined CPU0 only. If CPU0 is hard offlined + * (i.e. physically hot removed and then hot added), NMI won't wake it up. + * We'll change this code in the future to wake up hard offlined CPU0 if + * real platform and request are available. + */ +static int +wakeup_cpu_via_init_nmi(int cpu, unsigned long start_ip, int apicid, + int *cpu0_nmi_registered) +{ + int id; + int boot_error; - c_idle.idle = get_idle_for_cpu(cpu); + preempt_disable(); /* - * We can't use kernel_thread since we must avoid to - * reschedule the child. + * Wake up AP by INIT, INIT, STARTUP sequence. */ - if (c_idle.idle) { - c_idle.idle->thread.sp = (unsigned long) (((struct pt_regs *) - (THREAD_SIZE + task_stack_page(c_idle.idle))) - 1); - init_idle(c_idle.idle, cpu); - goto do_rest; + if (cpu) { + boot_error = wakeup_secondary_cpu_via_init(apicid, start_ip); + goto out; } - schedule_work(&c_idle.work); - wait_for_completion(&c_idle.done); + /* + * Wake up BSP by nmi. + * + * Register a NMI handler to help wake up CPU0. + */ + boot_error = register_nmi_handler(NMI_LOCAL, + wakeup_cpu0_nmi, 0, "wake_cpu0"); - if (IS_ERR(c_idle.idle)) { - printk("failed fork for CPU %d\n", cpu); - destroy_work_on_stack(&c_idle.work); - return PTR_ERR(c_idle.idle); + if (!boot_error) { + enable_start_cpu0 = 1; + *cpu0_nmi_registered = 1; + if (apic->dest_logical == APIC_DEST_LOGICAL) + id = cpu0_logical_apicid; + else + id = apicid; + boot_error = wakeup_secondary_cpu_via_nmi(id, start_ip); } - set_idle_for_cpu(cpu, c_idle.idle); -do_rest: - per_cpu(current_task, cpu) = c_idle.idle; +out: + preempt_enable(); + + return boot_error; +} + +/* + * NOTE - on most systems this is a PHYSICAL apic ID, but on multiquad + * (ie clustered apic addressing mode), this is a LOGICAL apic ID. + * Returns zero if CPU booted OK, else error code from + * ->wakeup_secondary_cpu. + */ +static int do_boot_cpu(int apicid, int cpu, struct task_struct *idle) +{ + volatile u32 *trampoline_status = + (volatile u32 *) __va(real_mode_header->trampoline_status); + /* start_ip had better be page-aligned! */ + unsigned long start_ip = real_mode_header->trampoline_start; + + unsigned long boot_error = 0; + int timeout; + int cpu0_nmi_registered = 0; + + /* Just in case we booted with a single CPU. */ + alternatives_enable_smp(); + + idle->thread.sp = (unsigned long) (((struct pt_regs *) + (THREAD_SIZE + task_stack_page(idle))) - 1); + per_cpu(current_task, cpu) = idle; + #ifdef CONFIG_X86_32 /* Stack for startup_32 can be just as for start_secondary onwards */ irq_ctx_init(cpu); #else - clear_tsk_thread_flag(c_idle.idle, TIF_FORK); + clear_tsk_thread_flag(idle, TIF_FORK); initial_gs = per_cpu_offset(cpu); +#endif per_cpu(kernel_stack, cpu) = - (unsigned long)task_stack_page(c_idle.idle) - + (unsigned long)task_stack_page(idle) - KERNEL_STACK_OFFSET + THREAD_SIZE; -#endif early_gdt_descr.address = (unsigned long)get_cpu_gdt_table(cpu); initial_code = (unsigned long)start_secondary; - stack_start = c_idle.idle->thread.sp; - - /* start_ip had better be page-aligned! */ - start_ip = trampoline_address(); + stack_start = idle->thread.sp; /* So we see what's up */ announce_cpu(cpu, apicid); @@ -740,8 +789,6 @@ do_rest: * the targeted processor. */ - printk(KERN_DEBUG "smpboot cpu %d: start_ip = %lx\n", cpu, start_ip); - atomic_set(&init_deasserted, 0); if (get_uv_system_type() != UV_NON_UNIQUE_APIC) { @@ -759,21 +806,24 @@ do_rest: } /* - * Kick the secondary CPU. Use the method in the APIC driver - * if it's defined - or use an INIT boot APIC message otherwise: + * Wake up a CPU in difference cases: + * - Use the method in the APIC driver if it's defined + * Otherwise, + * - Use an INIT boot APIC message for APs or NMI for BSP. */ if (apic->wakeup_secondary_cpu) boot_error = apic->wakeup_secondary_cpu(apicid, start_ip); else - boot_error = wakeup_secondary_cpu_via_init(apicid, start_ip); + boot_error = wakeup_cpu_via_init_nmi(cpu, start_ip, apicid, + &cpu0_nmi_registered); if (!boot_error) { /* * allow APs to start initializing. */ - pr_debug("Before Callout %d.\n", cpu); + pr_debug("Before Callout %d\n", cpu); cpumask_set_cpu(cpu, cpu_callout_mask); - pr_debug("After Callout %d.\n", cpu); + pr_debug("After Callout %d\n", cpu); /* * Wait 5s total for a response @@ -791,17 +841,17 @@ do_rest: schedule(); } - if (cpumask_test_cpu(cpu, cpu_callin_mask)) + if (cpumask_test_cpu(cpu, cpu_callin_mask)) { + print_cpu_msr(&cpu_data(cpu)); pr_debug("CPU%d: has booted.\n", cpu); - else { + } else { boot_error = 1; - if (*(volatile u32 *)TRAMPOLINE_SYM(trampoline_status) - == 0xA5A5A5A5) + if (*trampoline_status == 0xA5A5A5A5) /* trampoline started but...? */ pr_err("CPU%d: Stuck ??\n", cpu); else /* trampoline code not run */ - pr_err("CPU%d: Not responding.\n", cpu); + pr_err("CPU%d: Not responding\n", cpu); if (apic->inquire_remote_apic) apic->inquire_remote_apic(apicid); } @@ -816,13 +866,10 @@ do_rest: /* was set by cpu_init() */ cpumask_clear_cpu(cpu, cpu_initialized_mask); - - set_cpu_present(cpu, false); - per_cpu(x86_cpu_to_apicid, cpu) = BAD_APICID; } /* mark "stuck" area as not stuck */ - *(volatile u32 *)TRAMPOLINE_SYM(trampoline_status) = 0; + *trampoline_status = 0; if (get_uv_system_type() != UV_NON_UNIQUE_APIC) { /* @@ -830,12 +877,17 @@ do_rest: */ smpboot_restore_warm_reset_vector(); } + /* + * Clean up the nmi handler. Do this after the callin and callout sync + * to avoid impact of possible long unregister time. + */ + if (cpu0_nmi_registered) + unregister_nmi_handler(NMI_LOCAL, "wake_cpu0"); - destroy_work_on_stack(&c_idle.work); return boot_error; } -int __cpuinit native_cpu_up(unsigned int cpu) +int native_cpu_up(unsigned int cpu, struct task_struct *tidle) { int apicid = apic->cpu_present_to_apicid(cpu); unsigned long flags; @@ -845,10 +897,10 @@ int __cpuinit native_cpu_up(unsigned int cpu) pr_debug("++++++++++++++++++++=_---CPU UP %u\n", cpu); - if (apicid == BAD_APICID || apicid == boot_cpu_physical_apicid || + if (apicid == BAD_APICID || !physid_isset(apicid, phys_cpu_present_map) || - (!x2apic_mode && apicid >= 255)) { - printk(KERN_ERR "%s: bad cpu %d\n", __func__, cpu); + !apic->apic_id_valid(apicid)) { + pr_err("%s: bad cpu %d\n", __func__, cpu); return -EINVAL; } @@ -868,9 +920,12 @@ int __cpuinit native_cpu_up(unsigned int cpu) per_cpu(cpu_state, cpu) = CPU_UP_PREPARE; - err = do_boot_cpu(apicid, cpu); + /* the FPU context is blank, nobody can own it */ + __cpu_disable_lazy_restore(cpu); + + err = do_boot_cpu(apicid, cpu, tidle); if (err) { - pr_debug("do_boot_cpu failed %d\n", err); + pr_err("do_boot_cpu failed(%d) to wakeup CPU#%u\n", err, cpu); return -EIO; } @@ -929,9 +984,8 @@ static int __init smp_sanity_check(unsigned max_cpus) unsigned int cpu; unsigned nr; - printk(KERN_WARNING - "More than 8 CPUs detected - skipping them.\n" - "Use CONFIG_X86_BIGSMP.\n"); + pr_warn("More than 8 CPUs detected - skipping them\n" + "Use CONFIG_X86_BIGSMP\n"); nr = 0; for_each_present_cpu(cpu) { @@ -952,8 +1006,7 @@ static int __init smp_sanity_check(unsigned max_cpus) #endif if (!physid_isset(hard_smp_processor_id(), phys_cpu_present_map)) { - printk(KERN_WARNING - "weird, boot CPU (#%d) not listed by the BIOS.\n", + pr_warn("weird, boot CPU (#%d) not listed by the BIOS\n", hard_smp_processor_id()); physid_set(hard_smp_processor_id(), phys_cpu_present_map); @@ -965,11 +1018,10 @@ static int __init smp_sanity_check(unsigned max_cpus) */ if (!smp_found_config && !acpi_lapic) { preempt_enable(); - printk(KERN_NOTICE "SMP motherboard not detected.\n"); + pr_notice("SMP motherboard not detected\n"); disable_smp(); if (APIC_init_uniprocessor()) - printk(KERN_NOTICE "Local APIC not detected." - " Using dummy APIC emulation.\n"); + pr_notice("Local APIC not detected. Using dummy APIC emulation.\n"); return -1; } @@ -978,9 +1030,8 @@ static int __init smp_sanity_check(unsigned max_cpus) * CPU too, but we do it for the sake of robustness anyway. */ if (!apic->check_phys_apicid_present(boot_cpu_physical_apicid)) { - printk(KERN_NOTICE - "weird, boot CPU (#%d) not listed by the BIOS.\n", - boot_cpu_physical_apicid); + pr_notice("weird, boot CPU (#%d) not listed by the BIOS\n", + boot_cpu_physical_apicid); physid_set(hard_smp_processor_id(), phys_cpu_present_map); } preempt_enable(); @@ -993,8 +1044,7 @@ static int __init smp_sanity_check(unsigned max_cpus) if (!disable_apic) { pr_err("BIOS bug, local APIC #%d not detected!...\n", boot_cpu_physical_apicid); - pr_err("... forcing use of dummy APIC emulation." - "(tell your hw vendor)\n"); + pr_err("... forcing use of dummy APIC emulation (tell your hw vendor)\n"); } smpboot_clear_io_apic(); disable_ioapic_support(); @@ -1007,7 +1057,7 @@ static int __init smp_sanity_check(unsigned max_cpus) * If SMP should be disabled, then really disable it! */ if (!max_cpus) { - printk(KERN_INFO "SMP mode deactivated.\n"); + pr_info("SMP mode deactivated\n"); smpboot_clear_io_apic(); connect_bsp_APIC(); @@ -1045,7 +1095,7 @@ void __init native_smp_prepare_cpus(unsigned int max_cpus) /* * Setup boot CPU information */ - smp_store_cpu_info(0); /* Final full version of the data */ + smp_store_boot_cpu_info(); /* Final full version of the data */ cpumask_copy(cpu_callin_mask, cpumask_of(0)); mb(); @@ -1059,7 +1109,7 @@ void __init native_smp_prepare_cpus(unsigned int max_cpus) if (smp_sanity_check(max_cpus) < 0) { - printk(KERN_INFO "SMP disabled\n"); + pr_info("SMP disabled\n"); disable_smp(); goto out; } @@ -1081,6 +1131,11 @@ void __init native_smp_prepare_cpus(unsigned int max_cpus) */ setup_local_APIC(); + if (x2apic_mode) + cpu0_logical_apicid = apic_read(APIC_LDR); + else + cpu0_logical_apicid = GET_APIC_LOGICAL_ID(apic_read(APIC_LDR)); + /* * Enable IO APIC before setting up error vector */ @@ -1097,7 +1152,7 @@ void __init native_smp_prepare_cpus(unsigned int max_cpus) * Set up local APIC timer on boot CPU. */ - printk(KERN_INFO "CPU%d: ", 0); + pr_info("CPU%d: ", 0); print_cpu_info(&cpu_data(0)); x86_init.timers.setup_percpu_clockev(); @@ -1109,20 +1164,6 @@ out: preempt_enable(); } -void arch_disable_nonboot_cpus_begin(void) -{ - /* - * Avoid the smp alternatives switch during the disable_nonboot_cpus(). - * In the suspend path, we will be back in the SMP mode shortly anyways. - */ - skip_smp_alternatives = true; -} - -void arch_disable_nonboot_cpus_end(void) -{ - skip_smp_alternatives = false; -} - void arch_enable_nonboot_cpus_begin(void) { set_mtrr_aps_delayed_init(); @@ -1147,7 +1188,7 @@ void __init native_smp_prepare_boot_cpu(void) void __init native_smp_cpus_done(unsigned int max_cpus) { - pr_debug("Boot done.\n"); + pr_debug("Boot done\n"); nmi_selftest(); impress_friends(); @@ -1208,8 +1249,7 @@ __init void prefill_possible_map(void) /* nr_cpu_ids could be reduced via nr_cpus= */ if (possible > nr_cpu_ids) { - printk(KERN_WARNING - "%d Processors exceeds NR_CPUS limit of %d\n", + pr_warn("%d Processors exceeds NR_CPUS limit of %d\n", possible, nr_cpu_ids); possible = nr_cpu_ids; } @@ -1218,13 +1258,12 @@ __init void prefill_possible_map(void) if (!setup_max_cpus) #endif if (possible > i) { - printk(KERN_WARNING - "%d Processors exceeds max_cpus limit of %u\n", + pr_warn("%d Processors exceeds max_cpus limit of %u\n", possible, setup_max_cpus); possible = i; } - printk(KERN_INFO "SMP: Allowing %d CPUs, %d hotplug CPUs\n", + pr_info("Allowing %d CPUs, %d hotplug CPUs\n", possible, max_t(int, possible - num_processors, 0)); for (i = 0; i < possible; i++) @@ -1285,18 +1324,11 @@ void cpu_disable_common(void) int native_cpu_disable(void) { - int cpu = smp_processor_id(); + int ret; - /* - * Perhaps use cpufreq to drop frequency, but that could go - * into generic code. - * - * We won't take down the boot processor on i386 due to some - * interrupts only being able to be serviced by the BSP. - * Especially so if we're not using an IOAPIC -zwane - */ - if (cpu == 0) - return -EBUSY; + ret = check_irq_vectors_for_cpu_disable(); + if (ret) + return ret; clear_local_APIC(); @@ -1314,9 +1346,6 @@ void native_cpu_die(unsigned int cpu) if (per_cpu(cpu_state, cpu) == CPU_DEAD) { if (system_state == SYSTEM_RUNNING) pr_info("CPU %u is now offline\n", cpu); - - if (1 == num_online_cpus()) - alternatives_smp_switch(0); return; } msleep(100); @@ -1340,6 +1369,14 @@ void play_dead_common(void) local_irq_disable(); } +static bool wakeup_cpu0(void) +{ + if (smp_processor_id() == 0 && enable_start_cpu0) + return true; + + return false; +} + /* * We need to flush the caches before going to sleep, lest we have * dirty data in our caches when we come back up. @@ -1349,13 +1386,12 @@ static inline void mwait_play_dead(void) unsigned int eax, ebx, ecx, edx; unsigned int highest_cstate = 0; unsigned int highest_subcstate = 0; - int i; void *mwait_ptr; - struct cpuinfo_x86 *c = __this_cpu_ptr(&cpu_info); + int i; - if (!(this_cpu_has(X86_FEATURE_MWAIT) && mwait_usable(c))) + if (!this_cpu_has(X86_FEATURE_MWAIT)) return; - if (!this_cpu_has(X86_FEATURE_CLFLSH)) + if (!this_cpu_has(X86_FEATURE_CLFLUSH)) return; if (__this_cpu_read(cpu_info.cpuid_level) < CPUID_MWAIT_LEAF) return; @@ -1399,10 +1435,17 @@ static inline void mwait_play_dead(void) * The WBINVD is insufficient due to the spurious-wakeup * case where we return around the loop. */ + mb(); clflush(mwait_ptr); + mb(); __monitor(mwait_ptr, 0, 0); mb(); __mwait(eax, 0); + /* + * If NMI wants to wake up CPU0, start CPU0. + */ + if (wakeup_cpu0()) + start_cpu0(); } } @@ -1413,6 +1456,11 @@ static inline void hlt_play_dead(void) while (1) { native_halt(); + /* + * If NMI wants to wake up CPU0, start CPU0. + */ + if (wakeup_cpu0()) + start_cpu0(); } } @@ -1422,7 +1470,8 @@ void native_play_dead(void) tboot_shutdown(TB_SHUTDOWN_WFS); mwait_play_dead(); /* Only returns on failure */ - hlt_play_dead(); + if (cpuidle_play_dead()) + hlt_play_dead(); } #else /* ... !CONFIG_HOTPLUG_CPU */ diff --git a/arch/x86/kernel/step.c b/arch/x86/kernel/step.c index c346d116148..9b4d51d0c0d 100644 --- a/arch/x86/kernel/step.c +++ b/arch/x86/kernel/step.c @@ -157,6 +157,34 @@ static int enable_single_step(struct task_struct *child) return 1; } +void set_task_blockstep(struct task_struct *task, bool on) +{ + unsigned long debugctl; + + /* + * Ensure irq/preemption can't change debugctl in between. + * Note also that both TIF_BLOCKSTEP and debugctl should + * be changed atomically wrt preemption. + * + * NOTE: this means that set/clear TIF_BLOCKSTEP is only safe if + * task is current or it can't be running, otherwise we can race + * with __switch_to_xtra(). We rely on ptrace_freeze_traced() but + * PTRACE_KILL is not safe. + */ + local_irq_disable(); + debugctl = get_debugctlmsr(); + if (on) { + debugctl |= DEBUGCTLMSR_BTF; + set_tsk_thread_flag(task, TIF_BLOCKSTEP); + } else { + debugctl &= ~DEBUGCTLMSR_BTF; + clear_tsk_thread_flag(task, TIF_BLOCKSTEP); + } + if (task == current) + update_debugctlmsr(debugctl); + local_irq_enable(); +} + /* * Enable single or block step. */ @@ -169,19 +197,10 @@ static void enable_step(struct task_struct *child, bool block) * So no one should try to use debugger block stepping in a program * that uses user-mode single stepping itself. */ - if (enable_single_step(child) && block) { - unsigned long debugctl = get_debugctlmsr(); - - debugctl |= DEBUGCTLMSR_BTF; - update_debugctlmsr(debugctl); - set_tsk_thread_flag(child, TIF_BLOCKSTEP); - } else if (test_tsk_thread_flag(child, TIF_BLOCKSTEP)) { - unsigned long debugctl = get_debugctlmsr(); - - debugctl &= ~DEBUGCTLMSR_BTF; - update_debugctlmsr(debugctl); - clear_tsk_thread_flag(child, TIF_BLOCKSTEP); - } + if (enable_single_step(child) && block) + set_task_blockstep(child, true); + else if (test_tsk_thread_flag(child, TIF_BLOCKSTEP)) + set_task_blockstep(child, false); } void user_enable_single_step(struct task_struct *child) @@ -199,13 +218,8 @@ void user_disable_single_step(struct task_struct *child) /* * Make sure block stepping (BTF) is disabled. */ - if (test_tsk_thread_flag(child, TIF_BLOCKSTEP)) { - unsigned long debugctl = get_debugctlmsr(); - - debugctl &= ~DEBUGCTLMSR_BTF; - update_debugctlmsr(debugctl); - clear_tsk_thread_flag(child, TIF_BLOCKSTEP); - } + if (test_tsk_thread_flag(child, TIF_BLOCKSTEP)) + set_task_blockstep(child, false); /* Always clear TIF_SINGLESTEP... */ clear_tsk_thread_flag(child, TIF_SINGLESTEP); diff --git a/arch/x86/kernel/sys_i386_32.c b/arch/x86/kernel/sys_i386_32.c deleted file mode 100644 index 0b0cb5fede1..00000000000 --- a/arch/x86/kernel/sys_i386_32.c +++ /dev/null @@ -1,40 +0,0 @@ -/* - * This file contains various random system calls that - * have a non-standard calling sequence on the Linux/i386 - * platform. - */ - -#include <linux/errno.h> -#include <linux/sched.h> -#include <linux/mm.h> -#include <linux/fs.h> -#include <linux/smp.h> -#include <linux/sem.h> -#include <linux/msg.h> -#include <linux/shm.h> -#include <linux/stat.h> -#include <linux/syscalls.h> -#include <linux/mman.h> -#include <linux/file.h> -#include <linux/utsname.h> -#include <linux/ipc.h> - -#include <linux/uaccess.h> -#include <linux/unistd.h> - -#include <asm/syscalls.h> - -/* - * Do a system call from kernel instead of calling sys_execve so we - * end up with proper pt_regs. - */ -int kernel_execve(const char *filename, - const char *const argv[], - const char *const envp[]) -{ - long __res; - asm volatile ("int $0x80" - : "=a" (__res) - : "0" (__NR_execve), "b" (filename), "c" (argv), "d" (envp) : "memory"); - return __res; -} diff --git a/arch/x86/kernel/sys_x86_64.c b/arch/x86/kernel/sys_x86_64.c index 051489082d5..30277e27431 100644 --- a/arch/x86/kernel/sys_x86_64.c +++ b/arch/x86/kernel/sys_x86_64.c @@ -21,37 +21,23 @@ /* * Align a virtual address to avoid aliasing in the I$ on AMD F15h. - * - * @flags denotes the allocation direction - bottomup or topdown - - * or vDSO; see call sites below. */ -unsigned long align_addr(unsigned long addr, struct file *filp, - enum align_flags flags) +static unsigned long get_align_mask(void) { - unsigned long tmp_addr; - /* handle 32- and 64-bit case with a single conditional */ if (va_align.flags < 0 || !(va_align.flags & (2 - mmap_is_ia32()))) - return addr; + return 0; if (!(current->flags & PF_RANDOMIZE)) - return addr; - - if (!((flags & ALIGN_VDSO) || filp)) - return addr; - - tmp_addr = addr; - - /* - * We need an address which is <= than the original - * one only when in topdown direction. - */ - if (!(flags & ALIGN_TOPDOWN)) - tmp_addr += va_align.mask; + return 0; - tmp_addr &= ~va_align.mask; + return va_align.mask; +} - return tmp_addr; +unsigned long align_vdso_addr(unsigned long addr) +{ + unsigned long align_mask = get_align_mask(); + return (addr + align_mask) & ~align_mask; } static int __init control_va_addr_alignment(char *str) @@ -98,7 +84,7 @@ out: static void find_start_end(unsigned long flags, unsigned long *begin, unsigned long *end) { - if (!test_thread_flag(TIF_IA32) && (flags & MAP_32BIT)) { + if (!test_thread_flag(TIF_ADDR32) && (flags & MAP_32BIT)) { unsigned long new_begin; /* This is usually used needed to map code in small model, so it needs to be in the first 31bit. Limit @@ -115,7 +101,7 @@ static void find_start_end(unsigned long flags, unsigned long *begin, *begin = new_begin; } } else { - *begin = TASK_UNMAPPED_BASE; + *begin = current->mm->mmap_legacy_base; *end = TASK_SIZE; } } @@ -126,7 +112,7 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr, { struct mm_struct *mm = current->mm; struct vm_area_struct *vma; - unsigned long start_addr; + struct vm_unmapped_area_info info; unsigned long begin, end; if (flags & MAP_FIXED) @@ -144,50 +130,16 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr, (!vma || addr + len <= vma->vm_start)) return addr; } - if (((flags & MAP_32BIT) || test_thread_flag(TIF_IA32)) - && len <= mm->cached_hole_size) { - mm->cached_hole_size = 0; - mm->free_area_cache = begin; - } - addr = mm->free_area_cache; - if (addr < begin) - addr = begin; - start_addr = addr; - -full_search: - - addr = align_addr(addr, filp, 0); - - for (vma = find_vma(mm, addr); ; vma = vma->vm_next) { - /* At this point: (!vma || addr < vma->vm_end). */ - if (end - len < addr) { - /* - * Start a new search - just in case we missed - * some holes. - */ - if (start_addr != begin) { - start_addr = addr = begin; - mm->cached_hole_size = 0; - goto full_search; - } - return -ENOMEM; - } - if (!vma || addr + len <= vma->vm_start) { - /* - * Remember the place where we stopped the search: - */ - mm->free_area_cache = addr + len; - return addr; - } - if (addr + mm->cached_hole_size < vma->vm_start) - mm->cached_hole_size = vma->vm_start - addr; - addr = vma->vm_end; - addr = align_addr(addr, filp, 0); - } + info.flags = 0; + info.length = len; + info.low_limit = begin; + info.high_limit = end; + info.align_mask = filp ? get_align_mask() : 0; + info.align_offset = pgoff << PAGE_SHIFT; + return vm_unmapped_area(&info); } - unsigned long arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, const unsigned long len, const unsigned long pgoff, @@ -196,6 +148,7 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, struct vm_area_struct *vma; struct mm_struct *mm = current->mm; unsigned long addr = addr0; + struct vm_unmapped_area_info info; /* requested length too big for entire address space */ if (len > TASK_SIZE) @@ -204,8 +157,8 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, if (flags & MAP_FIXED) return addr; - /* for MAP_32BIT mappings we force the legact mmap base */ - if (!test_thread_flag(TIF_IA32) && (flags & MAP_32BIT)) + /* for MAP_32BIT mappings we force the legacy mmap base */ + if (!test_thread_flag(TIF_ADDR32) && (flags & MAP_32BIT)) goto bottomup; /* requesting a specific address */ @@ -217,51 +170,16 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, return addr; } - /* check if free_area_cache is useful for us */ - if (len <= mm->cached_hole_size) { - mm->cached_hole_size = 0; - mm->free_area_cache = mm->mmap_base; - } - - /* either no address requested or can't fit in requested address hole */ - addr = mm->free_area_cache; - - /* make sure it can fit in the remaining address space */ - if (addr > len) { - unsigned long tmp_addr = align_addr(addr - len, filp, - ALIGN_TOPDOWN); - - vma = find_vma(mm, tmp_addr); - if (!vma || tmp_addr + len <= vma->vm_start) - /* remember the address as a hint for next time */ - return mm->free_area_cache = tmp_addr; - } - - if (mm->mmap_base < len) - goto bottomup; - - addr = mm->mmap_base-len; - - do { - addr = align_addr(addr, filp, ALIGN_TOPDOWN); - - /* - * Lookup failure means no vma is above this address, - * else if new region fits below vma->vm_start, - * return with success: - */ - vma = find_vma(mm, addr); - if (!vma || addr+len <= vma->vm_start) - /* remember the address as a hint for next time */ - return mm->free_area_cache = addr; - - /* remember the largest hole we saw so far */ - if (addr + mm->cached_hole_size < vma->vm_start) - mm->cached_hole_size = vma->vm_start - addr; - - /* try just below the current vma->vm_start */ - addr = vma->vm_start-len; - } while (len < vma->vm_start); + info.flags = VM_UNMAPPED_AREA_TOPDOWN; + info.length = len; + info.low_limit = PAGE_SIZE; + info.high_limit = mm->mmap_base; + info.align_mask = filp ? get_align_mask() : 0; + info.align_offset = pgoff << PAGE_SHIFT; + addr = vm_unmapped_area(&info); + if (!(addr & ~PAGE_MASK)) + return addr; + VM_BUG_ON(addr != -ENOMEM); bottomup: /* @@ -270,14 +188,5 @@ bottomup: * can happen with large stack limits and large mmap() * allocations. */ - mm->cached_hole_size = ~0UL; - mm->free_area_cache = TASK_UNMAPPED_BASE; - addr = arch_get_unmapped_area(filp, addr0, len, pgoff, flags); - /* - * Restore the topdown base: - */ - mm->free_area_cache = mm->mmap_base; - mm->cached_hole_size = ~0UL; - - return addr; + return arch_get_unmapped_area(filp, addr0, len, pgoff, flags); } diff --git a/arch/x86/kernel/syscall_32.c b/arch/x86/kernel/syscall_32.c index 147fcd4941c..e9bcd57d8a9 100644 --- a/arch/x86/kernel/syscall_32.c +++ b/arch/x86/kernel/syscall_32.c @@ -15,7 +15,7 @@ typedef asmlinkage void (*sys_call_ptr_t)(void); extern asmlinkage void sys_ni_syscall(void); -const sys_call_ptr_t sys_call_table[__NR_syscall_max+1] = { +__visible const sys_call_ptr_t sys_call_table[__NR_syscall_max+1] = { /* * Smells like a compiler bug -- it doesn't work * when the & below is removed. diff --git a/arch/x86/kernel/syscall_64.c b/arch/x86/kernel/syscall_64.c index 7ac7943be02..4ac730b37f0 100644 --- a/arch/x86/kernel/syscall_64.c +++ b/arch/x86/kernel/syscall_64.c @@ -4,6 +4,15 @@ #include <linux/sys.h> #include <linux/cache.h> #include <asm/asm-offsets.h> +#include <asm/syscall.h> + +#define __SYSCALL_COMMON(nr, sym, compat) __SYSCALL_64(nr, sym, compat) + +#ifdef CONFIG_X86_X32_ABI +# define __SYSCALL_X32(nr, sym, compat) __SYSCALL_64(nr, sym, compat) +#else +# define __SYSCALL_X32(nr, sym, compat) /* nothing */ +#endif #define __SYSCALL_64(nr, sym, compat) extern asmlinkage void sym(void) ; #include <asm/syscalls_64.h> @@ -11,11 +20,9 @@ #define __SYSCALL_64(nr, sym, compat) [nr] = sym, -typedef void (*sys_call_ptr_t)(void); - extern void sys_ni_syscall(void); -const sys_call_ptr_t sys_call_table[__NR_syscall_max+1] = { +asmlinkage const sys_call_ptr_t sys_call_table[__NR_syscall_max+1] = { /* * Smells like a compiler bug -- it doesn't work * when the & below is removed. diff --git a/arch/x86/kernel/sysfb.c b/arch/x86/kernel/sysfb.c new file mode 100644 index 00000000000..193ec2ce46c --- /dev/null +++ b/arch/x86/kernel/sysfb.c @@ -0,0 +1,74 @@ +/* + * Generic System Framebuffers on x86 + * Copyright (c) 2012-2013 David Herrmann <dh.herrmann@gmail.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the Free + * Software Foundation; either version 2 of the License, or (at your option) + * any later version. + */ + +/* + * Simple-Framebuffer support for x86 systems + * Create a platform-device for any available boot framebuffer. The + * simple-framebuffer platform device is already available on DT systems, so + * this module parses the global "screen_info" object and creates a suitable + * platform device compatible with the "simple-framebuffer" DT object. If + * the framebuffer is incompatible, we instead create a legacy + * "vesa-framebuffer", "efi-framebuffer" or "platform-framebuffer" device and + * pass the screen_info as platform_data. This allows legacy drivers + * to pick these devices up without messing with simple-framebuffer drivers. + * The global "screen_info" is still valid at all times. + * + * If CONFIG_X86_SYSFB is not selected, we never register "simple-framebuffer" + * platform devices, but only use legacy framebuffer devices for + * backwards compatibility. + * + * TODO: We set the dev_id field of all platform-devices to 0. This allows + * other x86 OF/DT parsers to create such devices, too. However, they must + * start at offset 1 for this to work. + */ + +#include <linux/err.h> +#include <linux/init.h> +#include <linux/kernel.h> +#include <linux/mm.h> +#include <linux/platform_data/simplefb.h> +#include <linux/platform_device.h> +#include <linux/screen_info.h> +#include <asm/sysfb.h> + +static __init int sysfb_init(void) +{ + struct screen_info *si = &screen_info; + struct simplefb_platform_data mode; + struct platform_device *pd; + const char *name; + bool compatible; + int ret; + + sysfb_apply_efi_quirks(); + + /* try to create a simple-framebuffer device */ + compatible = parse_mode(si, &mode); + if (compatible) { + ret = create_simplefb(si, &mode); + if (!ret) + return 0; + } + + /* if the FB is incompatible, create a legacy framebuffer device */ + if (si->orig_video_isVGA == VIDEO_TYPE_EFI) + name = "efi-framebuffer"; + else if (si->orig_video_isVGA == VIDEO_TYPE_VLFB) + name = "vesa-framebuffer"; + else + name = "platform-framebuffer"; + + pd = platform_device_register_resndata(NULL, name, 0, + NULL, 0, si, sizeof(*si)); + return IS_ERR(pd) ? PTR_ERR(pd) : 0; +} + +/* must execute after PCI subsystem for EFI quirks */ +device_initcall(sysfb_init); diff --git a/arch/x86/kernel/sysfb_efi.c b/arch/x86/kernel/sysfb_efi.c new file mode 100644 index 00000000000..b285d4e8c68 --- /dev/null +++ b/arch/x86/kernel/sysfb_efi.c @@ -0,0 +1,214 @@ +/* + * Generic System Framebuffers on x86 + * Copyright (c) 2012-2013 David Herrmann <dh.herrmann@gmail.com> + * + * EFI Quirks Copyright (c) 2006 Edgar Hucek <gimli@dark-green.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the Free + * Software Foundation; either version 2 of the License, or (at your option) + * any later version. + */ + +/* + * EFI Quirks + * Several EFI systems do not correctly advertise their boot framebuffers. + * Hence, we use this static table of known broken machines and fix up the + * information so framebuffer drivers can load corectly. + */ + +#include <linux/dmi.h> +#include <linux/err.h> +#include <linux/init.h> +#include <linux/kernel.h> +#include <linux/mm.h> +#include <linux/pci.h> +#include <linux/screen_info.h> +#include <video/vga.h> +#include <asm/sysfb.h> + +enum { + OVERRIDE_NONE = 0x0, + OVERRIDE_BASE = 0x1, + OVERRIDE_STRIDE = 0x2, + OVERRIDE_HEIGHT = 0x4, + OVERRIDE_WIDTH = 0x8, +}; + +struct efifb_dmi_info efifb_dmi_list[] = { + [M_I17] = { "i17", 0x80010000, 1472 * 4, 1440, 900, OVERRIDE_NONE }, + [M_I20] = { "i20", 0x80010000, 1728 * 4, 1680, 1050, OVERRIDE_NONE }, /* guess */ + [M_I20_SR] = { "imac7", 0x40010000, 1728 * 4, 1680, 1050, OVERRIDE_NONE }, + [M_I24] = { "i24", 0x80010000, 2048 * 4, 1920, 1200, OVERRIDE_NONE }, /* guess */ + [M_I24_8_1] = { "imac8", 0xc0060000, 2048 * 4, 1920, 1200, OVERRIDE_NONE }, + [M_I24_10_1] = { "imac10", 0xc0010000, 2048 * 4, 1920, 1080, OVERRIDE_NONE }, + [M_I27_11_1] = { "imac11", 0xc0010000, 2560 * 4, 2560, 1440, OVERRIDE_NONE }, + [M_MINI]= { "mini", 0x80000000, 2048 * 4, 1024, 768, OVERRIDE_NONE }, + [M_MINI_3_1] = { "mini31", 0x40010000, 1024 * 4, 1024, 768, OVERRIDE_NONE }, + [M_MINI_4_1] = { "mini41", 0xc0010000, 2048 * 4, 1920, 1200, OVERRIDE_NONE }, + [M_MB] = { "macbook", 0x80000000, 2048 * 4, 1280, 800, OVERRIDE_NONE }, + [M_MB_5_1] = { "macbook51", 0x80010000, 2048 * 4, 1280, 800, OVERRIDE_NONE }, + [M_MB_6_1] = { "macbook61", 0x80010000, 2048 * 4, 1280, 800, OVERRIDE_NONE }, + [M_MB_7_1] = { "macbook71", 0x80010000, 2048 * 4, 1280, 800, OVERRIDE_NONE }, + [M_MBA] = { "mba", 0x80000000, 2048 * 4, 1280, 800, OVERRIDE_NONE }, + /* 11" Macbook Air 3,1 passes the wrong stride */ + [M_MBA_3] = { "mba3", 0, 2048 * 4, 0, 0, OVERRIDE_STRIDE }, + [M_MBP] = { "mbp", 0x80010000, 1472 * 4, 1440, 900, OVERRIDE_NONE }, + [M_MBP_2] = { "mbp2", 0, 0, 0, 0, OVERRIDE_NONE }, /* placeholder */ + [M_MBP_2_2] = { "mbp22", 0x80010000, 1472 * 4, 1440, 900, OVERRIDE_NONE }, + [M_MBP_SR] = { "mbp3", 0x80030000, 2048 * 4, 1440, 900, OVERRIDE_NONE }, + [M_MBP_4] = { "mbp4", 0xc0060000, 2048 * 4, 1920, 1200, OVERRIDE_NONE }, + [M_MBP_5_1] = { "mbp51", 0xc0010000, 2048 * 4, 1440, 900, OVERRIDE_NONE }, + [M_MBP_5_2] = { "mbp52", 0xc0010000, 2048 * 4, 1920, 1200, OVERRIDE_NONE }, + [M_MBP_5_3] = { "mbp53", 0xd0010000, 2048 * 4, 1440, 900, OVERRIDE_NONE }, + [M_MBP_6_1] = { "mbp61", 0x90030000, 2048 * 4, 1920, 1200, OVERRIDE_NONE }, + [M_MBP_6_2] = { "mbp62", 0x90030000, 2048 * 4, 1680, 1050, OVERRIDE_NONE }, + [M_MBP_7_1] = { "mbp71", 0xc0010000, 2048 * 4, 1280, 800, OVERRIDE_NONE }, + [M_MBP_8_2] = { "mbp82", 0x90010000, 1472 * 4, 1440, 900, OVERRIDE_NONE }, + [M_UNKNOWN] = { NULL, 0, 0, 0, 0, OVERRIDE_NONE } +}; + +#define choose_value(dmivalue, fwvalue, field, flags) ({ \ + typeof(fwvalue) _ret_ = fwvalue; \ + if ((flags) & (field)) \ + _ret_ = dmivalue; \ + else if ((fwvalue) == 0) \ + _ret_ = dmivalue; \ + _ret_; \ + }) + +static int __init efifb_set_system(const struct dmi_system_id *id) +{ + struct efifb_dmi_info *info = id->driver_data; + + if (info->base == 0 && info->height == 0 && info->width == 0 && + info->stride == 0) + return 0; + + /* Trust the bootloader over the DMI tables */ + if (screen_info.lfb_base == 0) { +#if defined(CONFIG_PCI) + struct pci_dev *dev = NULL; + int found_bar = 0; +#endif + if (info->base) { + screen_info.lfb_base = choose_value(info->base, + screen_info.lfb_base, OVERRIDE_BASE, + info->flags); + +#if defined(CONFIG_PCI) + /* make sure that the address in the table is actually + * on a VGA device's PCI BAR */ + + for_each_pci_dev(dev) { + int i; + if ((dev->class >> 8) != PCI_CLASS_DISPLAY_VGA) + continue; + for (i = 0; i < DEVICE_COUNT_RESOURCE; i++) { + resource_size_t start, end; + + start = pci_resource_start(dev, i); + if (start == 0) + break; + end = pci_resource_end(dev, i); + if (screen_info.lfb_base >= start && + screen_info.lfb_base < end) { + found_bar = 1; + } + } + } + if (!found_bar) + screen_info.lfb_base = 0; +#endif + } + } + if (screen_info.lfb_base) { + screen_info.lfb_linelength = choose_value(info->stride, + screen_info.lfb_linelength, OVERRIDE_STRIDE, + info->flags); + screen_info.lfb_width = choose_value(info->width, + screen_info.lfb_width, OVERRIDE_WIDTH, + info->flags); + screen_info.lfb_height = choose_value(info->height, + screen_info.lfb_height, OVERRIDE_HEIGHT, + info->flags); + if (screen_info.orig_video_isVGA == 0) + screen_info.orig_video_isVGA = VIDEO_TYPE_EFI; + } else { + screen_info.lfb_linelength = 0; + screen_info.lfb_width = 0; + screen_info.lfb_height = 0; + screen_info.orig_video_isVGA = 0; + return 0; + } + + printk(KERN_INFO "efifb: dmi detected %s - framebuffer at 0x%08x " + "(%dx%d, stride %d)\n", id->ident, + screen_info.lfb_base, screen_info.lfb_width, + screen_info.lfb_height, screen_info.lfb_linelength); + + return 1; +} + +#define EFIFB_DMI_SYSTEM_ID(vendor, name, enumid) \ + { \ + efifb_set_system, \ + name, \ + { \ + DMI_MATCH(DMI_BIOS_VENDOR, vendor), \ + DMI_MATCH(DMI_PRODUCT_NAME, name) \ + }, \ + &efifb_dmi_list[enumid] \ + } + +static const struct dmi_system_id efifb_dmi_system_table[] __initconst = { + EFIFB_DMI_SYSTEM_ID("Apple Computer, Inc.", "iMac4,1", M_I17), + /* At least one of these two will be right; maybe both? */ + EFIFB_DMI_SYSTEM_ID("Apple Computer, Inc.", "iMac5,1", M_I20), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "iMac5,1", M_I20), + /* At least one of these two will be right; maybe both? */ + EFIFB_DMI_SYSTEM_ID("Apple Computer, Inc.", "iMac6,1", M_I24), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "iMac6,1", M_I24), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "iMac7,1", M_I20_SR), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "iMac8,1", M_I24_8_1), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "iMac10,1", M_I24_10_1), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "iMac11,1", M_I27_11_1), + EFIFB_DMI_SYSTEM_ID("Apple Computer, Inc.", "Macmini1,1", M_MINI), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "Macmini3,1", M_MINI_3_1), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "Macmini4,1", M_MINI_4_1), + EFIFB_DMI_SYSTEM_ID("Apple Computer, Inc.", "MacBook1,1", M_MB), + /* At least one of these two will be right; maybe both? */ + EFIFB_DMI_SYSTEM_ID("Apple Computer, Inc.", "MacBook2,1", M_MB), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "MacBook2,1", M_MB), + /* At least one of these two will be right; maybe both? */ + EFIFB_DMI_SYSTEM_ID("Apple Computer, Inc.", "MacBook3,1", M_MB), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "MacBook3,1", M_MB), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "MacBook4,1", M_MB), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "MacBook5,1", M_MB_5_1), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "MacBook6,1", M_MB_6_1), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "MacBook7,1", M_MB_7_1), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "MacBookAir1,1", M_MBA), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "MacBookAir3,1", M_MBA_3), + EFIFB_DMI_SYSTEM_ID("Apple Computer, Inc.", "MacBookPro1,1", M_MBP), + EFIFB_DMI_SYSTEM_ID("Apple Computer, Inc.", "MacBookPro2,1", M_MBP_2), + EFIFB_DMI_SYSTEM_ID("Apple Computer, Inc.", "MacBookPro2,2", M_MBP_2_2), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "MacBookPro2,1", M_MBP_2), + EFIFB_DMI_SYSTEM_ID("Apple Computer, Inc.", "MacBookPro3,1", M_MBP_SR), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "MacBookPro3,1", M_MBP_SR), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "MacBookPro4,1", M_MBP_4), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "MacBookPro5,1", M_MBP_5_1), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "MacBookPro5,2", M_MBP_5_2), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "MacBookPro5,3", M_MBP_5_3), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "MacBookPro6,1", M_MBP_6_1), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "MacBookPro6,2", M_MBP_6_2), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "MacBookPro7,1", M_MBP_7_1), + EFIFB_DMI_SYSTEM_ID("Apple Inc.", "MacBookPro8,2", M_MBP_8_2), + {}, +}; + +__init void sysfb_apply_efi_quirks(void) +{ + if (screen_info.orig_video_isVGA != VIDEO_TYPE_EFI || + !(screen_info.capabilities & VIDEO_CAPABILITY_SKIP_QUIRKS)) + dmi_check_system(efifb_dmi_system_table); +} diff --git a/arch/x86/kernel/sysfb_simplefb.c b/arch/x86/kernel/sysfb_simplefb.c new file mode 100644 index 00000000000..86179d40989 --- /dev/null +++ b/arch/x86/kernel/sysfb_simplefb.c @@ -0,0 +1,95 @@ +/* + * Generic System Framebuffers on x86 + * Copyright (c) 2012-2013 David Herrmann <dh.herrmann@gmail.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the Free + * Software Foundation; either version 2 of the License, or (at your option) + * any later version. + */ + +/* + * simple-framebuffer probing + * Try to convert "screen_info" into a "simple-framebuffer" compatible mode. + * If the mode is incompatible, we return "false" and let the caller create + * legacy nodes instead. + */ + +#include <linux/err.h> +#include <linux/init.h> +#include <linux/kernel.h> +#include <linux/mm.h> +#include <linux/platform_data/simplefb.h> +#include <linux/platform_device.h> +#include <linux/screen_info.h> +#include <asm/sysfb.h> + +static const char simplefb_resname[] = "BOOTFB"; +static const struct simplefb_format formats[] = SIMPLEFB_FORMATS; + +/* try parsing x86 screen_info into a simple-framebuffer mode struct */ +__init bool parse_mode(const struct screen_info *si, + struct simplefb_platform_data *mode) +{ + const struct simplefb_format *f; + __u8 type; + unsigned int i; + + type = si->orig_video_isVGA; + if (type != VIDEO_TYPE_VLFB && type != VIDEO_TYPE_EFI) + return false; + + for (i = 0; i < ARRAY_SIZE(formats); ++i) { + f = &formats[i]; + if (si->lfb_depth == f->bits_per_pixel && + si->red_size == f->red.length && + si->red_pos == f->red.offset && + si->green_size == f->green.length && + si->green_pos == f->green.offset && + si->blue_size == f->blue.length && + si->blue_pos == f->blue.offset && + si->rsvd_size == f->transp.length && + si->rsvd_pos == f->transp.offset) { + mode->format = f->name; + mode->width = si->lfb_width; + mode->height = si->lfb_height; + mode->stride = si->lfb_linelength; + return true; + } + } + + return false; +} + +__init int create_simplefb(const struct screen_info *si, + const struct simplefb_platform_data *mode) +{ + struct platform_device *pd; + struct resource res; + unsigned long len; + + /* don't use lfb_size as it may contain the whole VMEM instead of only + * the part that is occupied by the framebuffer */ + len = mode->height * mode->stride; + len = PAGE_ALIGN(len); + if (len > (u64)si->lfb_size << 16) { + printk(KERN_WARNING "sysfb: VRAM smaller than advertised\n"); + return -EINVAL; + } + + /* setup IORESOURCE_MEM as framebuffer memory */ + memset(&res, 0, sizeof(res)); + res.flags = IORESOURCE_MEM | IORESOURCE_BUSY; + res.name = simplefb_resname; + res.start = si->lfb_base; + res.end = si->lfb_base + len - 1; + if (res.end <= res.start) + return -EINVAL; + + pd = platform_device_register_resndata(NULL, "simple-framebuffer", 0, + &res, 1, mode, sizeof(*mode)); + if (IS_ERR(pd)) + return PTR_ERR(pd); + + return 0; +} diff --git a/arch/x86/kernel/tboot.c b/arch/x86/kernel/tboot.c index e2410e27f97..91a4496db43 100644 --- a/arch/x86/kernel/tboot.c +++ b/arch/x86/kernel/tboot.c @@ -31,8 +31,9 @@ #include <linux/pfn.h> #include <linux/mm.h> #include <linux/tboot.h> +#include <linux/debugfs.h> -#include <asm/trampoline.h> +#include <asm/realmode.h> #include <asm/processor.h> #include <asm/bootparam.h> #include <asm/pgtable.h> @@ -44,7 +45,7 @@ #include <asm/e820.h> #include <asm/io.h> -#include "acpi/realmode/wakeup.h" +#include "../realmode/rm/wakeup.h" /* Global pointer to shared data; NULL means no measured launch. */ struct tboot *tboot __read_mostly; @@ -201,7 +202,8 @@ static int tboot_setup_sleep(void) add_mac_region(e820.map[i].addr, e820.map[i].size); } - tboot->acpi_sinfo.kernel_s3_resume_vector = acpi_wakeup_address; + tboot->acpi_sinfo.kernel_s3_resume_vector = + real_mode_header->wakeup_start; return 0; } @@ -272,7 +274,7 @@ static void tboot_copy_fadt(const struct acpi_table_fadt *fadt) offsetof(struct acpi_table_facs, firmware_waking_vector); } -void tboot_sleep(u8 sleep_state, u32 pm1a_control, u32 pm1b_control) +static int tboot_sleep(u8 sleep_state, u32 pm1a_control, u32 pm1b_control) { static u32 acpi_shutdown_map[ACPI_S_STATE_COUNT] = { /* S0,1,2: */ -1, -1, -1, @@ -281,7 +283,7 @@ void tboot_sleep(u8 sleep_state, u32 pm1a_control, u32 pm1b_control) /* S5: */ TB_SHUTDOWN_S5 }; if (!tboot_enabled()) - return; + return 0; tboot_copy_fadt(&acpi_gbl_FADT); tboot->acpi_sinfo.pm1a_cnt_val = pm1a_control; @@ -292,10 +294,20 @@ void tboot_sleep(u8 sleep_state, u32 pm1a_control, u32 pm1b_control) if (sleep_state >= ACPI_S_STATE_COUNT || acpi_shutdown_map[sleep_state] == -1) { pr_warning("unsupported sleep state 0x%x\n", sleep_state); - return; + return -1; } tboot_shutdown(acpi_shutdown_map[sleep_state]); + return 0; +} + +static int tboot_extended_sleep(u8 sleep_state, u32 val_a, u32 val_b) +{ + if (!tboot_enabled()) + return 0; + + pr_warning("tboot is not able to suspend on platforms with reduced hardware sleep (ACPIv5)"); + return -ENODEV; } static atomic_t ap_wfs_count; @@ -317,8 +329,8 @@ static int tboot_wait_for_aps(int num_aps) return !(atomic_read((atomic_t *)&tboot->num_in_wfs) == num_aps); } -static int __cpuinit tboot_cpu_callback(struct notifier_block *nfb, - unsigned long action, void *hcpu) +static int tboot_cpu_callback(struct notifier_block *nfb, unsigned long action, + void *hcpu) { switch (action) { case CPU_DYING: @@ -331,11 +343,78 @@ static int __cpuinit tboot_cpu_callback(struct notifier_block *nfb, return NOTIFY_OK; } -static struct notifier_block tboot_cpu_notifier __cpuinitdata = +static struct notifier_block tboot_cpu_notifier = { .notifier_call = tboot_cpu_callback, }; +#ifdef CONFIG_DEBUG_FS + +#define TBOOT_LOG_UUID { 0x26, 0x25, 0x19, 0xc0, 0x30, 0x6b, 0xb4, 0x4d, \ + 0x4c, 0x84, 0xa3, 0xe9, 0x53, 0xb8, 0x81, 0x74 } + +#define TBOOT_SERIAL_LOG_ADDR 0x60000 +#define TBOOT_SERIAL_LOG_SIZE 0x08000 +#define LOG_MAX_SIZE_OFF 16 +#define LOG_BUF_OFF 24 + +static uint8_t tboot_log_uuid[16] = TBOOT_LOG_UUID; + +static ssize_t tboot_log_read(struct file *file, char __user *user_buf, size_t count, loff_t *ppos) +{ + void __iomem *log_base; + u8 log_uuid[16]; + u32 max_size; + void *kbuf; + int ret = -EFAULT; + + log_base = ioremap_nocache(TBOOT_SERIAL_LOG_ADDR, TBOOT_SERIAL_LOG_SIZE); + if (!log_base) + return ret; + + memcpy_fromio(log_uuid, log_base, sizeof(log_uuid)); + if (memcmp(&tboot_log_uuid, log_uuid, sizeof(log_uuid))) + goto err_iounmap; + + max_size = readl(log_base + LOG_MAX_SIZE_OFF); + if (*ppos >= max_size) { + ret = 0; + goto err_iounmap; + } + + if (*ppos + count > max_size) + count = max_size - *ppos; + + kbuf = kmalloc(count, GFP_KERNEL); + if (!kbuf) { + ret = -ENOMEM; + goto err_iounmap; + } + + memcpy_fromio(kbuf, log_base + LOG_BUF_OFF + *ppos, count); + if (copy_to_user(user_buf, kbuf, count)) + goto err_kfree; + + *ppos += count; + + ret = count; + +err_kfree: + kfree(kbuf); + +err_iounmap: + iounmap(log_base); + + return ret; +} + +static const struct file_operations tboot_log_fops = { + .read = tboot_log_read, + .llseek = default_llseek, +}; + +#endif /* CONFIG_DEBUG_FS */ + static __init int tboot_late_init(void) { if (!tboot_enabled()) @@ -345,6 +424,14 @@ static __init int tboot_late_init(void) atomic_set(&ap_wfs_count, 0); register_hotcpu_notifier(&tboot_cpu_notifier); + +#ifdef CONFIG_DEBUG_FS + debugfs_create_file("tboot_log", S_IRUSR, + arch_debugfs_dir, NULL, &tboot_log_fops); +#endif + + acpi_os_set_prepare_sleep(&tboot_sleep); + acpi_os_set_prepare_extended_sleep(&tboot_extended_sleep); return 0; } diff --git a/arch/x86/kernel/tce_64.c b/arch/x86/kernel/tce_64.c index 9e540fee700..ab40954e113 100644 --- a/arch/x86/kernel/tce_64.c +++ b/arch/x86/kernel/tce_64.c @@ -34,6 +34,7 @@ #include <asm/tce.h> #include <asm/calgary.h> #include <asm/proto.h> +#include <asm/cacheflush.h> /* flush a tce at 'tceaddr' to main memory */ static inline void flush_tce(void* tceaddr) diff --git a/arch/x86/kernel/test_rodata.c b/arch/x86/kernel/test_rodata.c index c29e235792a..b79133abda4 100644 --- a/arch/x86/kernel/test_rodata.c +++ b/arch/x86/kernel/test_rodata.c @@ -12,6 +12,7 @@ #include <linux/module.h> #include <asm/cacheflush.h> #include <asm/sections.h> +#include <asm/asm.h> int rodata_test(void) { @@ -42,14 +43,7 @@ int rodata_test(void) ".section .fixup,\"ax\"\n" "2: jmp 1b\n" ".previous\n" - ".section __ex_table,\"a\"\n" - " .align 16\n" -#ifdef CONFIG_X86_32 - " .long 0b,2b\n" -#else - " .quad 0b,2b\n" -#endif - ".previous" + _ASM_EXTABLE(0b,2b) : [rslt] "=r" (result) : [rodata_test] "r" (&rodata_test_data), [zero] "r" (0UL) ); diff --git a/arch/x86/kernel/time.c b/arch/x86/kernel/time.c index dd5fbf4101f..bf7ef5ce29d 100644 --- a/arch/x86/kernel/time.c +++ b/arch/x86/kernel/time.c @@ -14,7 +14,6 @@ #include <linux/i8253.h> #include <linux/time.h> #include <linux/export.h> -#include <linux/mca.h> #include <asm/vsyscall.h> #include <asm/x86_init.h> @@ -24,7 +23,7 @@ #include <asm/time.h> #ifdef CONFIG_X86_64 -DEFINE_VVAR(volatile unsigned long, jiffies) = INITIAL_JIFFIES; +__visible DEFINE_VVAR(volatile unsigned long, jiffies) = INITIAL_JIFFIES; #endif unsigned long profile_pc(struct pt_regs *regs) @@ -57,21 +56,13 @@ EXPORT_SYMBOL(profile_pc); */ static irqreturn_t timer_interrupt(int irq, void *dev_id) { - /* Keep nmi watchdog up to date */ - inc_irq_stat(irq0_irqs); - global_clock_event->event_handler(global_clock_event); - - /* MCA bus quirk: Acknowledge irq0 by setting bit 7 in port 0x61 */ - if (MCA_bus) - outb_p(inb_p(0x61)| 0x80, 0x61); - return IRQ_HANDLED; } static struct irqaction irq0 = { .handler = timer_interrupt, - .flags = IRQF_DISABLED | IRQF_NOBALANCING | IRQF_IRQPOLL | IRQF_TIMER, + .flags = IRQF_NOBALANCING | IRQF_IRQPOLL | IRQF_TIMER, .name = "timer" }; diff --git a/arch/x86/kernel/tls.c b/arch/x86/kernel/tls.c index 6bb7b8579e7..f7fec09e3e3 100644 --- a/arch/x86/kernel/tls.c +++ b/arch/x86/kernel/tls.c @@ -3,14 +3,13 @@ #include <linux/sched.h> #include <linux/user.h> #include <linux/regset.h> +#include <linux/syscalls.h> #include <asm/uaccess.h> #include <asm/desc.h> -#include <asm/system.h> #include <asm/ldt.h> #include <asm/processor.h> #include <asm/proto.h> -#include <asm/syscalls.h> #include "tls.h" @@ -90,11 +89,9 @@ int do_set_thread_area(struct task_struct *p, int idx, return 0; } -asmlinkage int sys_set_thread_area(struct user_desc __user *u_info) +SYSCALL_DEFINE1(set_thread_area, struct user_desc __user *, u_info) { - int ret = do_set_thread_area(current, -1, u_info, 1); - asmlinkage_protect(1, ret, u_info); - return ret; + return do_set_thread_area(current, -1, u_info, 1); } @@ -140,11 +137,9 @@ int do_get_thread_area(struct task_struct *p, int idx, return 0; } -asmlinkage int sys_get_thread_area(struct user_desc __user *u_info) +SYSCALL_DEFINE1(get_thread_area, struct user_desc __user *, u_info) { - int ret = do_get_thread_area(current, -1, u_info); - asmlinkage_protect(1, ret, u_info); - return ret; + return do_get_thread_area(current, -1, u_info); } int regset_tls_active(struct task_struct *target, @@ -163,7 +158,7 @@ int regset_tls_get(struct task_struct *target, const struct user_regset *regset, { const struct desc_struct *tls; - if (pos > GDT_ENTRY_TLS_ENTRIES * sizeof(struct user_desc) || + if (pos >= GDT_ENTRY_TLS_ENTRIES * sizeof(struct user_desc) || (pos % sizeof(struct user_desc)) != 0 || (count % sizeof(struct user_desc)) != 0) return -EINVAL; @@ -198,7 +193,7 @@ int regset_tls_set(struct task_struct *target, const struct user_regset *regset, struct user_desc infobuf[GDT_ENTRY_TLS_ENTRIES]; const struct user_desc *info; - if (pos > GDT_ENTRY_TLS_ENTRIES * sizeof(struct user_desc) || + if (pos >= GDT_ENTRY_TLS_ENTRIES * sizeof(struct user_desc) || (pos % sizeof(struct user_desc)) != 0 || (count % sizeof(struct user_desc)) != 0) return -EINVAL; diff --git a/arch/x86/kernel/topology.c b/arch/x86/kernel/topology.c index 76ee97709a0..649b010da00 100644 --- a/arch/x86/kernel/topology.c +++ b/arch/x86/kernel/topology.c @@ -30,23 +30,113 @@ #include <linux/mmzone.h> #include <linux/init.h> #include <linux/smp.h> +#include <linux/irq.h> #include <asm/cpu.h> static DEFINE_PER_CPU(struct x86_cpu, cpu_devices); #ifdef CONFIG_HOTPLUG_CPU + +#ifdef CONFIG_BOOTPARAM_HOTPLUG_CPU0 +static int cpu0_hotpluggable = 1; +#else +static int cpu0_hotpluggable; +static int __init enable_cpu0_hotplug(char *str) +{ + cpu0_hotpluggable = 1; + return 1; +} + +__setup("cpu0_hotplug", enable_cpu0_hotplug); +#endif + +#ifdef CONFIG_DEBUG_HOTPLUG_CPU0 +/* + * This function offlines a CPU as early as possible and allows userspace to + * boot up without the CPU. The CPU can be onlined back by user after boot. + * + * This is only called for debugging CPU offline/online feature. + */ +int __ref _debug_hotplug_cpu(int cpu, int action) +{ + struct device *dev = get_cpu_device(cpu); + int ret; + + if (!cpu_is_hotpluggable(cpu)) + return -EINVAL; + + lock_device_hotplug(); + + switch (action) { + case 0: + ret = cpu_down(cpu); + if (!ret) { + pr_info("CPU %u is now offline\n", cpu); + dev->offline = true; + kobject_uevent(&dev->kobj, KOBJ_OFFLINE); + } else + pr_debug("Can't offline CPU%d.\n", cpu); + break; + case 1: + ret = cpu_up(cpu); + if (!ret) { + dev->offline = false; + kobject_uevent(&dev->kobj, KOBJ_ONLINE); + } else { + pr_debug("Can't online CPU%d.\n", cpu); + } + break; + default: + ret = -EINVAL; + } + + unlock_device_hotplug(); + + return ret; +} + +static int __init debug_hotplug_cpu(void) +{ + _debug_hotplug_cpu(0, 0); + return 0; +} + +late_initcall_sync(debug_hotplug_cpu); +#endif /* CONFIG_DEBUG_HOTPLUG_CPU0 */ + int __ref arch_register_cpu(int num) { + struct cpuinfo_x86 *c = &cpu_data(num); + + /* + * Currently CPU0 is only hotpluggable on Intel platforms. Other + * vendors can add hotplug support later. + */ + if (c->x86_vendor != X86_VENDOR_INTEL) + cpu0_hotpluggable = 0; + /* - * CPU0 cannot be offlined due to several - * restrictions and assumptions in kernel. This basically - * doesn't add a control file, one cannot attempt to offline - * BSP. + * Two known BSP/CPU0 dependencies: Resume from suspend/hibernate + * depends on BSP. PIC interrupts depend on BSP. * - * Also certain PCI quirks require not to enable hotplug control - * for all CPU's. + * If the BSP depencies are under control, one can tell kernel to + * enable BSP hotplug. This basically adds a control file and + * one can attempt to offline BSP. */ - if (num) + if (num == 0 && cpu0_hotpluggable) { + unsigned int irq; + /* + * We won't take down the boot processor on i386 if some + * interrupts only are able to be serviced by the BSP in PIC. + */ + for_each_active_irq(irq) { + if (!IO_APIC_IRQ(irq) && irq_has_action(irq)) { + cpu0_hotpluggable = 0; + break; + } + } + } + if (num || cpu0_hotpluggable) per_cpu(cpu_devices, num).cpu.hotpluggable = 1; return register_cpu(&per_cpu(cpu_devices, num).cpu, num); diff --git a/arch/x86/kernel/trace_clock.c b/arch/x86/kernel/trace_clock.c new file mode 100644 index 00000000000..25b993729f9 --- /dev/null +++ b/arch/x86/kernel/trace_clock.c @@ -0,0 +1,21 @@ +/* + * X86 trace clocks + */ +#include <asm/trace_clock.h> +#include <asm/barrier.h> +#include <asm/msr.h> + +/* + * trace_clock_x86_tsc(): A clock that is just the cycle counter. + * + * Unlike the other clocks, this is not in nanoseconds. + */ +u64 notrace trace_clock_x86_tsc(void) +{ + u64 ret; + + rdtsc_barrier(); + rdtscll(ret); + + return ret; +} diff --git a/arch/x86/kernel/tracepoint.c b/arch/x86/kernel/tracepoint.c new file mode 100644 index 00000000000..1c113db9ed5 --- /dev/null +++ b/arch/x86/kernel/tracepoint.c @@ -0,0 +1,59 @@ +/* + * Code for supporting irq vector tracepoints. + * + * Copyright (C) 2013 Seiji Aguchi <seiji.aguchi@hds.com> + * + */ +#include <asm/hw_irq.h> +#include <asm/desc.h> +#include <linux/atomic.h> + +atomic_t trace_idt_ctr = ATOMIC_INIT(0); +struct desc_ptr trace_idt_descr = { NR_VECTORS * 16 - 1, + (unsigned long) trace_idt_table }; + +/* No need to be aligned, but done to keep all IDTs defined the same way. */ +gate_desc trace_idt_table[NR_VECTORS] __page_aligned_bss; + +static int trace_irq_vector_refcount; +static DEFINE_MUTEX(irq_vector_mutex); + +static void set_trace_idt_ctr(int val) +{ + atomic_set(&trace_idt_ctr, val); + /* Ensure the trace_idt_ctr is set before sending IPI */ + wmb(); +} + +static void switch_idt(void *arg) +{ + unsigned long flags; + + local_irq_save(flags); + load_current_idt(); + local_irq_restore(flags); +} + +void trace_irq_vector_regfunc(void) +{ + mutex_lock(&irq_vector_mutex); + if (!trace_irq_vector_refcount) { + set_trace_idt_ctr(1); + smp_call_function(switch_idt, NULL, 0); + switch_idt(NULL); + } + trace_irq_vector_refcount++; + mutex_unlock(&irq_vector_mutex); +} + +void trace_irq_vector_unregfunc(void) +{ + mutex_lock(&irq_vector_mutex); + trace_irq_vector_refcount--; + if (!trace_irq_vector_refcount) { + set_trace_idt_ctr(0); + smp_call_function(switch_idt, NULL, 0); + switch_idt(NULL); + } + mutex_unlock(&irq_vector_mutex); +} diff --git a/arch/x86/kernel/trampoline.c b/arch/x86/kernel/trampoline.c deleted file mode 100644 index a73b61055ad..00000000000 --- a/arch/x86/kernel/trampoline.c +++ /dev/null @@ -1,42 +0,0 @@ -#include <linux/io.h> -#include <linux/memblock.h> - -#include <asm/trampoline.h> -#include <asm/cacheflush.h> -#include <asm/pgtable.h> - -unsigned char *x86_trampoline_base; - -void __init setup_trampolines(void) -{ - phys_addr_t mem; - size_t size = PAGE_ALIGN(x86_trampoline_end - x86_trampoline_start); - - /* Has to be in very low memory so we can execute real-mode AP code. */ - mem = memblock_find_in_range(0, 1<<20, size, PAGE_SIZE); - if (!mem) - panic("Cannot allocate trampoline\n"); - - x86_trampoline_base = __va(mem); - memblock_reserve(mem, size); - - printk(KERN_DEBUG "Base memory trampoline at [%p] %llx size %zu\n", - x86_trampoline_base, (unsigned long long)mem, size); - - memcpy(x86_trampoline_base, x86_trampoline_start, size); -} - -/* - * setup_trampolines() gets called very early, to guarantee the - * availability of low memory. This is before the proper kernel page - * tables are set up, so we cannot set page permissions in that - * function. Thus, we use an arch_initcall instead. - */ -static int __init configure_trampolines(void) -{ - size_t size = PAGE_ALIGN(x86_trampoline_end - x86_trampoline_start); - - set_memory_x((unsigned long)x86_trampoline_base, size >> PAGE_SHIFT); - return 0; -} -arch_initcall(configure_trampolines); diff --git a/arch/x86/kernel/trampoline_32.S b/arch/x86/kernel/trampoline_32.S deleted file mode 100644 index 451c0a7ef7f..00000000000 --- a/arch/x86/kernel/trampoline_32.S +++ /dev/null @@ -1,83 +0,0 @@ -/* - * - * Trampoline.S Derived from Setup.S by Linus Torvalds - * - * 4 Jan 1997 Michael Chastain: changed to gnu as. - * - * This is only used for booting secondary CPUs in SMP machine - * - * Entry: CS:IP point to the start of our code, we are - * in real mode with no stack, but the rest of the - * trampoline page to make our stack and everything else - * is a mystery. - * - * We jump into arch/x86/kernel/head_32.S. - * - * On entry to trampoline_data, the processor is in real mode - * with 16-bit addressing and 16-bit data. CS has some value - * and IP is zero. Thus, data addresses need to be absolute - * (no relocation) and are taken with regard to r_base. - * - * If you work on this file, check the object module with - * objdump --reloc to make sure there are no relocation - * entries except for: - * - * TYPE VALUE - * R_386_32 startup_32_smp - * R_386_32 boot_gdt - */ - -#include <linux/linkage.h> -#include <linux/init.h> -#include <asm/segment.h> -#include <asm/page_types.h> - -#ifdef CONFIG_SMP - - .section ".x86_trampoline","a" - .balign PAGE_SIZE - .code16 - -ENTRY(trampoline_data) -r_base = . - wbinvd # Needed for NUMA-Q should be harmless for others - mov %cs, %ax # Code and data in the same place - mov %ax, %ds - - cli # We should be safe anyway - - movl $0xA5A5A5A5, trampoline_status - r_base - # write marker for master knows we're running - - /* GDT tables in non default location kernel can be beyond 16MB and - * lgdt will not be able to load the address as in real mode default - * operand size is 16bit. Use lgdtl instead to force operand size - * to 32 bit. - */ - - lidtl boot_idt_descr - r_base # load idt with 0, 0 - lgdtl boot_gdt_descr - r_base # load gdt with whatever is appropriate - - xor %ax, %ax - inc %ax # protected mode (PE) bit - lmsw %ax # into protected mode - # flush prefetch and jump to startup_32_smp in arch/i386/kernel/head.S - ljmpl $__BOOT_CS, $(startup_32_smp-__PAGE_OFFSET) - - # These need to be in the same 64K segment as the above; - # hence we don't use the boot_gdt_descr defined in head.S -boot_gdt_descr: - .word __BOOT_DS + 7 # gdt limit - .long boot_gdt - __PAGE_OFFSET # gdt base - -boot_idt_descr: - .word 0 # idt limit = 0 - .long 0 # idt base = 0L - -ENTRY(trampoline_status) - .long 0 - -.globl trampoline_end -trampoline_end: - -#endif /* CONFIG_SMP */ diff --git a/arch/x86/kernel/trampoline_64.S b/arch/x86/kernel/trampoline_64.S deleted file mode 100644 index 09ff51799e9..00000000000 --- a/arch/x86/kernel/trampoline_64.S +++ /dev/null @@ -1,171 +0,0 @@ -/* - * - * Trampoline.S Derived from Setup.S by Linus Torvalds - * - * 4 Jan 1997 Michael Chastain: changed to gnu as. - * 15 Sept 2005 Eric Biederman: 64bit PIC support - * - * Entry: CS:IP point to the start of our code, we are - * in real mode with no stack, but the rest of the - * trampoline page to make our stack and everything else - * is a mystery. - * - * On entry to trampoline_data, the processor is in real mode - * with 16-bit addressing and 16-bit data. CS has some value - * and IP is zero. Thus, data addresses need to be absolute - * (no relocation) and are taken with regard to r_base. - * - * With the addition of trampoline_level4_pgt this code can - * now enter a 64bit kernel that lives at arbitrary 64bit - * physical addresses. - * - * If you work on this file, check the object module with objdump - * --full-contents --reloc to make sure there are no relocation - * entries. - */ - -#include <linux/linkage.h> -#include <linux/init.h> -#include <asm/pgtable_types.h> -#include <asm/page_types.h> -#include <asm/msr.h> -#include <asm/segment.h> -#include <asm/processor-flags.h> - - .section ".x86_trampoline","a" - .balign PAGE_SIZE - .code16 - -ENTRY(trampoline_data) -r_base = . - cli # We should be safe anyway - wbinvd - mov %cs, %ax # Code and data in the same place - mov %ax, %ds - mov %ax, %es - mov %ax, %ss - - - movl $0xA5A5A5A5, trampoline_status - r_base - # write marker for master knows we're running - - # Setup stack - movw $(trampoline_stack_end - r_base), %sp - - call verify_cpu # Verify the cpu supports long mode - testl %eax, %eax # Check for return code - jnz no_longmode - - mov %cs, %ax - movzx %ax, %esi # Find the 32bit trampoline location - shll $4, %esi - - # Fixup the absolute vectors - leal (startup_32 - r_base)(%esi), %eax - movl %eax, startup_32_vector - r_base - leal (startup_64 - r_base)(%esi), %eax - movl %eax, startup_64_vector - r_base - leal (tgdt - r_base)(%esi), %eax - movl %eax, (tgdt + 2 - r_base) - - /* - * GDT tables in non default location kernel can be beyond 16MB and - * lgdt will not be able to load the address as in real mode default - * operand size is 16bit. Use lgdtl instead to force operand size - * to 32 bit. - */ - - lidtl tidt - r_base # load idt with 0, 0 - lgdtl tgdt - r_base # load gdt with whatever is appropriate - - mov $X86_CR0_PE, %ax # protected mode (PE) bit - lmsw %ax # into protected mode - - # flush prefetch and jump to startup_32 - ljmpl *(startup_32_vector - r_base) - - .code32 - .balign 4 -startup_32: - movl $__KERNEL_DS, %eax # Initialize the %ds segment register - movl %eax, %ds - - movl $X86_CR4_PAE, %eax - movl %eax, %cr4 # Enable PAE mode - - # Setup trampoline 4 level pagetables - leal (trampoline_level4_pgt - r_base)(%esi), %eax - movl %eax, %cr3 - - movl $MSR_EFER, %ecx - movl $(1 << _EFER_LME), %eax # Enable Long Mode - xorl %edx, %edx - wrmsr - - # Enable paging and in turn activate Long Mode - # Enable protected mode - movl $(X86_CR0_PG | X86_CR0_PE), %eax - movl %eax, %cr0 - - /* - * At this point we're in long mode but in 32bit compatibility mode - * with EFER.LME = 1, CS.L = 0, CS.D = 1 (and in turn - * EFER.LMA = 1). Now we want to jump in 64bit mode, to do that we use - * the new gdt/idt that has __KERNEL_CS with CS.L = 1. - */ - ljmp *(startup_64_vector - r_base)(%esi) - - .code64 - .balign 4 -startup_64: - # Now jump into the kernel using virtual addresses - movq $secondary_startup_64, %rax - jmp *%rax - - .code16 -no_longmode: - hlt - jmp no_longmode -#include "verify_cpu.S" - - .balign 4 - # Careful these need to be in the same 64K segment as the above; -tidt: - .word 0 # idt limit = 0 - .word 0, 0 # idt base = 0L - - # Duplicate the global descriptor table - # so the kernel can live anywhere - .balign 4 -tgdt: - .short tgdt_end - tgdt # gdt limit - .long tgdt - r_base - .short 0 - .quad 0x00cf9b000000ffff # __KERNEL32_CS - .quad 0x00af9b000000ffff # __KERNEL_CS - .quad 0x00cf93000000ffff # __KERNEL_DS -tgdt_end: - - .balign 4 -startup_32_vector: - .long startup_32 - r_base - .word __KERNEL32_CS, 0 - - .balign 4 -startup_64_vector: - .long startup_64 - r_base - .word __KERNEL_CS, 0 - - .balign 4 -ENTRY(trampoline_status) - .long 0 - -trampoline_stack: - .org 0x1000 -trampoline_stack_end: -ENTRY(trampoline_level4_pgt) - .quad level3_ident_pgt - __START_KERNEL_map + _KERNPG_TABLE - .fill 510,8,0 - .quad level3_kernel_pgt - __START_KERNEL_map + _KERNPG_TABLE - -ENTRY(trampoline_end) diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index 77da5b475ad..0d0e922fafc 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -9,6 +9,10 @@ /* * Handle hardware traps and faults. */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/context_tracking.h> #include <linux/interrupt.h> #include <linux/kallsyms.h> #include <linux/spinlock.h> @@ -19,6 +23,7 @@ #include <linux/kernel.h> #include <linux/module.h> #include <linux/ptrace.h> +#include <linux/uprobes.h> #include <linux/string.h> #include <linux/delay.h> #include <linux/errno.h> @@ -37,10 +42,6 @@ #include <linux/eisa.h> #endif -#ifdef CONFIG_MCA -#include <linux/mca.h> -#endif - #if defined(CONFIG_EDAC) #include <linux/edac.h> #endif @@ -50,34 +51,33 @@ #include <asm/processor.h> #include <asm/debugreg.h> #include <linux/atomic.h> -#include <asm/system.h> +#include <asm/ftrace.h> #include <asm/traps.h> #include <asm/desc.h> #include <asm/i387.h> +#include <asm/fpu-internal.h> #include <asm/mce.h> - +#include <asm/fixmap.h> #include <asm/mach_traps.h> +#include <asm/alternative.h> #ifdef CONFIG_X86_64 #include <asm/x86_init.h> #include <asm/pgalloc.h> #include <asm/proto.h> + +/* No need to be aligned, but done to keep all IDTs defined the same way. */ +gate_desc debug_idt_table[NR_VECTORS] __page_aligned_bss; #else #include <asm/processor-flags.h> #include <asm/setup.h> asmlinkage int system_call(void); - -/* Do we ignore FPU interrupts ? */ -char ignore_fpu_irq; - -/* - * The IDT has to be page-aligned to simplify the Pentium - * F0 0F bug workaround. - */ -gate_desc idt_table[NR_VECTORS] __page_aligned_data = { { { { 0, 0 } } }, }; #endif +/* Must be page-aligned because the real IDT is used in a fixmap. */ +gate_desc idt_table[NR_VECTORS] __page_aligned_bss; + DECLARE_BITMAP(used_vectors, NR_VECTORS); EXPORT_SYMBOL_GPL(used_vectors); @@ -89,7 +89,7 @@ static inline void conditional_sti(struct pt_regs *regs) static inline void preempt_conditional_sti(struct pt_regs *regs) { - inc_preempt_count(); + preempt_count_inc(); if (regs->flags & X86_EFLAGS_IF) local_irq_enable(); } @@ -104,35 +104,81 @@ static inline void preempt_conditional_cli(struct pt_regs *regs) { if (regs->flags & X86_EFLAGS_IF) local_irq_disable(); - dec_preempt_count(); + preempt_count_dec(); } -static void __kprobes -do_trap(int trapnr, int signr, char *str, struct pt_regs *regs, - long error_code, siginfo_t *info) +static nokprobe_inline int +do_trap_no_signal(struct task_struct *tsk, int trapnr, char *str, + struct pt_regs *regs, long error_code) { - struct task_struct *tsk = current; - #ifdef CONFIG_X86_32 if (regs->flags & X86_VM_MASK) { /* - * traps 0, 1, 3, 4, and 5 should be forwarded to vm86. + * Traps 0, 1, 3, 4, and 5 should be forwarded to vm86. * On nmi (interrupt 2), do_trap should not be called. */ - if (trapnr < 6) - goto vm86_trap; - goto trap_signal; + if (trapnr < X86_TRAP_UD) { + if (!handle_vm86_trap((struct kernel_vm86_regs *) regs, + error_code, trapnr)) + return 0; + } + return -1; } #endif + if (!user_mode(regs)) { + if (!fixup_exception(regs)) { + tsk->thread.error_code = error_code; + tsk->thread.trap_nr = trapnr; + die(str, regs, error_code); + } + return 0; + } - if (!user_mode(regs)) - goto kernel_trap; + return -1; +} -#ifdef CONFIG_X86_32 -trap_signal: -#endif +static siginfo_t *fill_trap_info(struct pt_regs *regs, int signr, int trapnr, + siginfo_t *info) +{ + unsigned long siaddr; + int sicode; + + switch (trapnr) { + default: + return SEND_SIG_PRIV; + + case X86_TRAP_DE: + sicode = FPE_INTDIV; + siaddr = uprobe_get_trap_addr(regs); + break; + case X86_TRAP_UD: + sicode = ILL_ILLOPN; + siaddr = uprobe_get_trap_addr(regs); + break; + case X86_TRAP_AC: + sicode = BUS_ADRALN; + siaddr = 0; + break; + } + + info->si_signo = signr; + info->si_errno = 0; + info->si_code = sicode; + info->si_addr = (void __user *)siaddr; + return info; +} + +static void +do_trap(int trapnr, int signr, char *str, struct pt_regs *regs, + long error_code, siginfo_t *info) +{ + struct task_struct *tsk = current; + + + if (!do_trap_no_signal(tsk, trapnr, str, regs, error_code)) + return; /* - * We want error_code and trap_no set for userspace faults and + * We want error_code and trap_nr set for userspace faults and * kernelspace faults which result in die(), but not * kernelspace faults which are fixed up. die() gives the * process no chance to handle the signal and notice the @@ -141,90 +187,71 @@ trap_signal: * delivered, faults. See also do_general_protection below. */ tsk->thread.error_code = error_code; - tsk->thread.trap_no = trapnr; + tsk->thread.trap_nr = trapnr; #ifdef CONFIG_X86_64 if (show_unhandled_signals && unhandled_signal(tsk, signr) && printk_ratelimit()) { - printk(KERN_INFO - "%s[%d] trap %s ip:%lx sp:%lx error:%lx", - tsk->comm, tsk->pid, str, - regs->ip, regs->sp, error_code); + pr_info("%s[%d] trap %s ip:%lx sp:%lx error:%lx", + tsk->comm, tsk->pid, str, + regs->ip, regs->sp, error_code); print_vma_addr(" in ", regs->ip); - printk("\n"); + pr_cont("\n"); } #endif - if (info) - force_sig_info(signr, info, tsk); - else - force_sig(signr, tsk); - return; + force_sig_info(signr, info ?: SEND_SIG_PRIV, tsk); +} +NOKPROBE_SYMBOL(do_trap); -kernel_trap: - if (!fixup_exception(regs)) { - tsk->thread.error_code = error_code; - tsk->thread.trap_no = trapnr; - die(str, regs, error_code); +static void do_error_trap(struct pt_regs *regs, long error_code, char *str, + unsigned long trapnr, int signr) +{ + enum ctx_state prev_state = exception_enter(); + siginfo_t info; + + if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) != + NOTIFY_STOP) { + conditional_sti(regs); + do_trap(trapnr, signr, str, regs, error_code, + fill_trap_info(regs, signr, trapnr, &info)); } - return; -#ifdef CONFIG_X86_32 -vm86_trap: - if (handle_vm86_trap((struct kernel_vm86_regs *) regs, - error_code, trapnr)) - goto trap_signal; - return; -#endif + exception_exit(prev_state); } #define DO_ERROR(trapnr, signr, str, name) \ dotraplinkage void do_##name(struct pt_regs *regs, long error_code) \ { \ - if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \ - == NOTIFY_STOP) \ - return; \ - conditional_sti(regs); \ - do_trap(trapnr, signr, str, regs, error_code, NULL); \ -} - -#define DO_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr) \ -dotraplinkage void do_##name(struct pt_regs *regs, long error_code) \ -{ \ - siginfo_t info; \ - info.si_signo = signr; \ - info.si_errno = 0; \ - info.si_code = sicode; \ - info.si_addr = (void __user *)siaddr; \ - if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \ - == NOTIFY_STOP) \ - return; \ - conditional_sti(regs); \ - do_trap(trapnr, signr, str, regs, error_code, &info); \ + do_error_trap(regs, error_code, str, trapnr, signr); \ } -DO_ERROR_INFO(0, SIGFPE, "divide error", divide_error, FPE_INTDIV, regs->ip) -DO_ERROR(4, SIGSEGV, "overflow", overflow) -DO_ERROR(5, SIGSEGV, "bounds", bounds) -DO_ERROR_INFO(6, SIGILL, "invalid opcode", invalid_op, ILL_ILLOPN, regs->ip) -DO_ERROR(9, SIGFPE, "coprocessor segment overrun", coprocessor_segment_overrun) -DO_ERROR(10, SIGSEGV, "invalid TSS", invalid_TSS) -DO_ERROR(11, SIGBUS, "segment not present", segment_not_present) +DO_ERROR(X86_TRAP_DE, SIGFPE, "divide error", divide_error) +DO_ERROR(X86_TRAP_OF, SIGSEGV, "overflow", overflow) +DO_ERROR(X86_TRAP_BR, SIGSEGV, "bounds", bounds) +DO_ERROR(X86_TRAP_UD, SIGILL, "invalid opcode", invalid_op) +DO_ERROR(X86_TRAP_OLD_MF, SIGFPE, "coprocessor segment overrun",coprocessor_segment_overrun) +DO_ERROR(X86_TRAP_TS, SIGSEGV, "invalid TSS", invalid_TSS) +DO_ERROR(X86_TRAP_NP, SIGBUS, "segment not present", segment_not_present) #ifdef CONFIG_X86_32 -DO_ERROR(12, SIGBUS, "stack segment", stack_segment) +DO_ERROR(X86_TRAP_SS, SIGBUS, "stack segment", stack_segment) #endif -DO_ERROR_INFO(17, SIGBUS, "alignment check", alignment_check, BUS_ADRALN, 0) +DO_ERROR(X86_TRAP_AC, SIGBUS, "alignment check", alignment_check) #ifdef CONFIG_X86_64 /* Runs on IST stack */ dotraplinkage void do_stack_segment(struct pt_regs *regs, long error_code) { + enum ctx_state prev_state; + + prev_state = exception_enter(); if (notify_die(DIE_TRAP, "stack segment", regs, error_code, - 12, SIGBUS) == NOTIFY_STOP) - return; - preempt_conditional_sti(regs); - do_trap(12, SIGBUS, "stack segment", regs, error_code, NULL); - preempt_conditional_cli(regs); + X86_TRAP_SS, SIGBUS) != NOTIFY_STOP) { + preempt_conditional_sti(regs); + do_trap(X86_TRAP_SS, SIGBUS, "stack segment", regs, error_code, NULL); + preempt_conditional_cli(regs); + } + exception_exit(prev_state); } dotraplinkage void do_double_fault(struct pt_regs *regs, long error_code) @@ -232,12 +259,16 @@ dotraplinkage void do_double_fault(struct pt_regs *regs, long error_code) static const char str[] = "double fault"; struct task_struct *tsk = current; + exception_enter(); /* Return not checked because double check cannot be ignored */ - notify_die(DIE_TRAP, str, regs, error_code, 8, SIGSEGV); + notify_die(DIE_TRAP, str, regs, error_code, X86_TRAP_DF, SIGSEGV); tsk->thread.error_code = error_code; - tsk->thread.trap_no = 8; + tsk->thread.trap_nr = X86_TRAP_DF; +#ifdef CONFIG_DOUBLEFAULT + df_debug(regs, error_code); +#endif /* * This is always a kernel trap and never fixable (and thus must * never return). @@ -247,69 +278,86 @@ dotraplinkage void do_double_fault(struct pt_regs *regs, long error_code) } #endif -dotraplinkage void __kprobes +dotraplinkage void do_general_protection(struct pt_regs *regs, long error_code) { struct task_struct *tsk; + enum ctx_state prev_state; + prev_state = exception_enter(); conditional_sti(regs); #ifdef CONFIG_X86_32 - if (regs->flags & X86_VM_MASK) - goto gp_in_vm86; + if (regs->flags & X86_VM_MASK) { + local_irq_enable(); + handle_vm86_fault((struct kernel_vm86_regs *) regs, error_code); + goto exit; + } #endif tsk = current; - if (!user_mode(regs)) - goto gp_in_kernel; + if (!user_mode(regs)) { + if (fixup_exception(regs)) + goto exit; + + tsk->thread.error_code = error_code; + tsk->thread.trap_nr = X86_TRAP_GP; + if (notify_die(DIE_GPF, "general protection fault", regs, error_code, + X86_TRAP_GP, SIGSEGV) != NOTIFY_STOP) + die("general protection fault", regs, error_code); + goto exit; + } tsk->thread.error_code = error_code; - tsk->thread.trap_no = 13; + tsk->thread.trap_nr = X86_TRAP_GP; if (show_unhandled_signals && unhandled_signal(tsk, SIGSEGV) && printk_ratelimit()) { - printk(KERN_INFO - "%s[%d] general protection ip:%lx sp:%lx error:%lx", + pr_info("%s[%d] general protection ip:%lx sp:%lx error:%lx", tsk->comm, task_pid_nr(tsk), regs->ip, regs->sp, error_code); print_vma_addr(" in ", regs->ip); - printk("\n"); + pr_cont("\n"); } - force_sig(SIGSEGV, tsk); - return; + force_sig_info(SIGSEGV, SEND_SIG_PRIV, tsk); +exit: + exception_exit(prev_state); +} +NOKPROBE_SYMBOL(do_general_protection); -#ifdef CONFIG_X86_32 -gp_in_vm86: - local_irq_enable(); - handle_vm86_fault((struct kernel_vm86_regs *) regs, error_code); - return; -#endif +/* May run on IST stack. */ +dotraplinkage void notrace do_int3(struct pt_regs *regs, long error_code) +{ + enum ctx_state prev_state; -gp_in_kernel: - if (fixup_exception(regs)) +#ifdef CONFIG_DYNAMIC_FTRACE + /* + * ftrace must be first, everything else may cause a recursive crash. + * See note by declaration of modifying_ftrace_code in ftrace.c + */ + if (unlikely(atomic_read(&modifying_ftrace_code)) && + ftrace_int3_handler(regs)) return; - - tsk->thread.error_code = error_code; - tsk->thread.trap_no = 13; - if (notify_die(DIE_GPF, "general protection fault", regs, - error_code, 13, SIGSEGV) == NOTIFY_STOP) +#endif + if (poke_int3_handler(regs)) return; - die("general protection fault", regs, error_code); -} -/* May run on IST stack. */ -dotraplinkage void __kprobes do_int3(struct pt_regs *regs, long error_code) -{ + prev_state = exception_enter(); #ifdef CONFIG_KGDB_LOW_LEVEL_TRAP - if (kgdb_ll_trap(DIE_INT3, "int3", regs, error_code, 3, SIGTRAP) - == NOTIFY_STOP) - return; + if (kgdb_ll_trap(DIE_INT3, "int3", regs, error_code, X86_TRAP_BP, + SIGTRAP) == NOTIFY_STOP) + goto exit; #endif /* CONFIG_KGDB_LOW_LEVEL_TRAP */ - if (notify_die(DIE_INT3, "int3", regs, error_code, 3, SIGTRAP) - == NOTIFY_STOP) - return; +#ifdef CONFIG_KPROBES + if (kprobe_int3_handler(regs)) + goto exit; +#endif + + if (notify_die(DIE_INT3, "int3", regs, error_code, X86_TRAP_BP, + SIGTRAP) == NOTIFY_STOP) + goto exit; /* * Let others (NMI) know that the debug stack is in use @@ -317,10 +365,13 @@ dotraplinkage void __kprobes do_int3(struct pt_regs *regs, long error_code) */ debug_stack_usage_inc(); preempt_conditional_sti(regs); - do_trap(3, SIGTRAP, "int3", regs, error_code, NULL); + do_trap(X86_TRAP_BP, SIGTRAP, "int3", regs, error_code, NULL); preempt_conditional_cli(regs); debug_stack_usage_dec(); +exit: + exception_exit(prev_state); } +NOKPROBE_SYMBOL(do_int3); #ifdef CONFIG_X86_64 /* @@ -328,7 +379,7 @@ dotraplinkage void __kprobes do_int3(struct pt_regs *regs, long error_code) * for scheduling or signal handling. The actual stack switch is done in * entry.S */ -asmlinkage __kprobes struct pt_regs *sync_regs(struct pt_regs *eregs) +asmlinkage __visible struct pt_regs *sync_regs(struct pt_regs *eregs) { struct pt_regs *regs = eregs; /* Did already sync */ @@ -347,6 +398,7 @@ asmlinkage __kprobes struct pt_regs *sync_regs(struct pt_regs *eregs) *regs = *eregs; return regs; } +NOKPROBE_SYMBOL(sync_regs); #endif /* @@ -373,13 +425,16 @@ asmlinkage __kprobes struct pt_regs *sync_regs(struct pt_regs *eregs) * * May run on IST stack. */ -dotraplinkage void __kprobes do_debug(struct pt_regs *regs, long error_code) +dotraplinkage void do_debug(struct pt_regs *regs, long error_code) { struct task_struct *tsk = current; + enum ctx_state prev_state; int user_icebp = 0; unsigned long dr6; int si_code; + prev_state = exception_enter(); + get_debugreg(dr6, 6); /* Filter out all the reserved bits which are preset to 1 */ @@ -395,7 +450,7 @@ dotraplinkage void __kprobes do_debug(struct pt_regs *regs, long error_code) /* Catch kmemcheck conditions first of all! */ if ((dr6 & DR_STEP) && kmemcheck_trap(regs)) - return; + goto exit; /* DR6 may or may not be cleared by the CPU */ set_debugreg(0, 6); @@ -408,9 +463,14 @@ dotraplinkage void __kprobes do_debug(struct pt_regs *regs, long error_code) /* Store the virtualized DR6 value */ tsk->thread.debugreg6 = dr6; - if (notify_die(DIE_DEBUG, "debug", regs, PTR_ERR(&dr6), error_code, +#ifdef CONFIG_KPROBES + if (kprobe_debug_handler(regs)) + goto exit; +#endif + + if (notify_die(DIE_DEBUG, "debug", regs, (long)&dr6, error_code, SIGTRAP) == NOTIFY_STOP) - return; + goto exit; /* * Let others (NMI) know that the debug stack is in use @@ -422,11 +482,11 @@ dotraplinkage void __kprobes do_debug(struct pt_regs *regs, long error_code) preempt_conditional_sti(regs); if (regs->flags & X86_VM_MASK) { - handle_vm86_trap((struct kernel_vm86_regs *) regs, - error_code, 1); + handle_vm86_trap((struct kernel_vm86_regs *) regs, error_code, + X86_TRAP_DB); preempt_conditional_cli(regs); debug_stack_usage_dec(); - return; + goto exit; } /* @@ -447,20 +507,23 @@ dotraplinkage void __kprobes do_debug(struct pt_regs *regs, long error_code) preempt_conditional_cli(regs); debug_stack_usage_dec(); - return; +exit: + exception_exit(prev_state); } +NOKPROBE_SYMBOL(do_debug); /* * Note that we play around with the 'TS' bit in an attempt to get * the correct behaviour even in the presence of the asynchronous * IRQ13 behaviour */ -void math_error(struct pt_regs *regs, int error_code, int trapnr) +static void math_error(struct pt_regs *regs, int error_code, int trapnr) { struct task_struct *task = current; siginfo_t info; unsigned short err; - char *str = (trapnr == 16) ? "fpu exception" : "simd exception"; + char *str = (trapnr == X86_TRAP_MF) ? "fpu exception" : + "simd exception"; if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, SIGFPE) == NOTIFY_STOP) return; @@ -470,7 +533,7 @@ void math_error(struct pt_regs *regs, int error_code, int trapnr) { if (!fixup_exception(regs)) { task->thread.error_code = error_code; - task->thread.trap_no = trapnr; + task->thread.trap_nr = trapnr; die(str, regs, error_code); } return; @@ -480,12 +543,12 @@ void math_error(struct pt_regs *regs, int error_code, int trapnr) * Save the info for the exception handler and clear the error. */ save_init_fpu(task); - task->thread.trap_no = trapnr; + task->thread.trap_nr = trapnr; task->thread.error_code = error_code; info.si_signo = SIGFPE; info.si_errno = 0; - info.si_addr = (void __user *)regs->ip; - if (trapnr == 16) { + info.si_addr = (void __user *)uprobe_get_trap_addr(regs); + if (trapnr == X86_TRAP_MF) { unsigned short cwd, swd; /* * (~cwd & swd) will mask out exceptions that are not set to unmasked @@ -529,27 +592,32 @@ void math_error(struct pt_regs *regs, int error_code, int trapnr) info.si_code = FPE_FLTRES; } else { /* - * If we're using IRQ 13, or supposedly even some trap 16 - * implementations, it's possible we get a spurious trap... + * If we're using IRQ 13, or supposedly even some trap + * X86_TRAP_MF implementations, it's possible + * we get a spurious trap, which is not an error. */ - return; /* Spurious trap, no error */ + return; } force_sig_info(SIGFPE, &info, task); } dotraplinkage void do_coprocessor_error(struct pt_regs *regs, long error_code) { -#ifdef CONFIG_X86_32 - ignore_fpu_irq = 1; -#endif + enum ctx_state prev_state; - math_error(regs, error_code, 16); + prev_state = exception_enter(); + math_error(regs, error_code, X86_TRAP_MF); + exception_exit(prev_state); } dotraplinkage void do_simd_coprocessor_error(struct pt_regs *regs, long error_code) { - math_error(regs, error_code, 19); + enum ctx_state prev_state; + + prev_state = exception_enter(); + math_error(regs, error_code, X86_TRAP_XF); + exception_exit(prev_state); } dotraplinkage void @@ -558,50 +626,19 @@ do_spurious_interrupt_bug(struct pt_regs *regs, long error_code) conditional_sti(regs); #if 0 /* No need to warn about this any longer. */ - printk(KERN_INFO "Ignoring P6 Local APIC Spurious Interrupt Bug...\n"); + pr_info("Ignoring P6 Local APIC Spurious Interrupt Bug...\n"); #endif } -asmlinkage void __attribute__((weak)) smp_thermal_interrupt(void) +asmlinkage __visible void __attribute__((weak)) smp_thermal_interrupt(void) { } -asmlinkage void __attribute__((weak)) smp_threshold_interrupt(void) +asmlinkage __visible void __attribute__((weak)) smp_threshold_interrupt(void) { } /* - * This gets called with the process already owning the - * FPU state, and with CR0.TS cleared. It just needs to - * restore the FPU register state. - */ -void __math_state_restore(struct task_struct *tsk) -{ - /* We need a safe address that is cheap to find and that is already - in L1. We've just brought in "tsk->thread.has_fpu", so use that */ -#define safe_address (tsk->thread.has_fpu) - - /* AMD K7/K8 CPUs don't save/restore FDP/FIP/FOP unless an exception - is pending. Clear the x87 state here by setting it to fixed - values. safe_address is a random variable that should be in L1 */ - alternative_input( - ASM_NOP8 ASM_NOP2, - "emms\n\t" /* clear stack tags */ - "fildl %P[addr]", /* set F?P to defined value */ - X86_FEATURE_FXSAVE_LEAK, - [addr] "m" (safe_address)); - - /* - * Paranoid restore. send a SIGSEGV if we fail to restore the state. - */ - if (unlikely(restore_fpu_checking(tsk))) { - __thread_fpu_end(tsk); - force_sig(SIGSEGV, tsk); - return; - } -} - -/* * 'math_state_restore()' saves the current math information in the * old math state array, and gets the new ones from the current task * @@ -631,15 +668,28 @@ void math_state_restore(void) } __thread_fpu_begin(tsk); - __math_state_restore(tsk); - tsk->fpu_counter++; + /* + * Paranoid restore. send a SIGSEGV if we fail to restore the state. + */ + if (unlikely(restore_fpu_checking(tsk))) { + drop_init_fpu(tsk); + force_sig_info(SIGSEGV, SEND_SIG_PRIV, tsk); + return; + } + + tsk->thread.fpu_counter++; } EXPORT_SYMBOL_GPL(math_state_restore); -dotraplinkage void __kprobes +dotraplinkage void do_device_not_available(struct pt_regs *regs, long error_code) { + enum ctx_state prev_state; + + prev_state = exception_enter(); + BUG_ON(use_eager_fpu()); + #ifdef CONFIG_MATH_EMULATION if (read_cr0() & X86_CR0_EM) { struct math_emu_info info = { }; @@ -648,6 +698,7 @@ do_device_not_available(struct pt_regs *regs, long error_code) info.regs = regs; math_emulate(&info); + exception_exit(prev_state); return; } #endif @@ -655,35 +706,51 @@ do_device_not_available(struct pt_regs *regs, long error_code) #ifdef CONFIG_X86_32 conditional_sti(regs); #endif + exception_exit(prev_state); } +NOKPROBE_SYMBOL(do_device_not_available); #ifdef CONFIG_X86_32 dotraplinkage void do_iret_error(struct pt_regs *regs, long error_code) { siginfo_t info; + enum ctx_state prev_state; + + prev_state = exception_enter(); local_irq_enable(); info.si_signo = SIGILL; info.si_errno = 0; info.si_code = ILL_BADSTK; info.si_addr = NULL; - if (notify_die(DIE_TRAP, "iret exception", - regs, error_code, 32, SIGILL) == NOTIFY_STOP) - return; - do_trap(32, SIGILL, "iret exception", regs, error_code, &info); + if (notify_die(DIE_TRAP, "iret exception", regs, error_code, + X86_TRAP_IRET, SIGILL) != NOTIFY_STOP) { + do_trap(X86_TRAP_IRET, SIGILL, "iret exception", regs, error_code, + &info); + } + exception_exit(prev_state); } #endif /* Set of traps needed for early debugging. */ void __init early_trap_init(void) { - set_intr_gate_ist(1, &debug, DEBUG_STACK); + set_intr_gate_ist(X86_TRAP_DB, &debug, DEBUG_STACK); /* int3 can be called from all */ - set_system_intr_gate_ist(3, &int3, DEBUG_STACK); - set_intr_gate(14, &page_fault); + set_system_intr_gate_ist(X86_TRAP_BP, &int3, DEBUG_STACK); +#ifdef CONFIG_X86_32 + set_intr_gate(X86_TRAP_PF, page_fault); +#endif load_idt(&idt_descr); } +void __init early_trap_pf_init(void) +{ +#ifdef CONFIG_X86_64 + set_intr_gate(X86_TRAP_PF, page_fault); +#endif +} + void __init trap_init(void) { int i; @@ -696,30 +763,30 @@ void __init trap_init(void) early_iounmap(p, 4); #endif - set_intr_gate(0, ÷_error); - set_intr_gate_ist(2, &nmi, NMI_STACK); + set_intr_gate(X86_TRAP_DE, divide_error); + set_intr_gate_ist(X86_TRAP_NMI, &nmi, NMI_STACK); /* int4 can be called from all */ - set_system_intr_gate(4, &overflow); - set_intr_gate(5, &bounds); - set_intr_gate(6, &invalid_op); - set_intr_gate(7, &device_not_available); + set_system_intr_gate(X86_TRAP_OF, &overflow); + set_intr_gate(X86_TRAP_BR, bounds); + set_intr_gate(X86_TRAP_UD, invalid_op); + set_intr_gate(X86_TRAP_NM, device_not_available); #ifdef CONFIG_X86_32 - set_task_gate(8, GDT_ENTRY_DOUBLEFAULT_TSS); + set_task_gate(X86_TRAP_DF, GDT_ENTRY_DOUBLEFAULT_TSS); #else - set_intr_gate_ist(8, &double_fault, DOUBLEFAULT_STACK); + set_intr_gate_ist(X86_TRAP_DF, &double_fault, DOUBLEFAULT_STACK); #endif - set_intr_gate(9, &coprocessor_segment_overrun); - set_intr_gate(10, &invalid_TSS); - set_intr_gate(11, &segment_not_present); - set_intr_gate_ist(12, &stack_segment, STACKFAULT_STACK); - set_intr_gate(13, &general_protection); - set_intr_gate(15, &spurious_interrupt_bug); - set_intr_gate(16, &coprocessor_error); - set_intr_gate(17, &alignment_check); + set_intr_gate(X86_TRAP_OLD_MF, coprocessor_segment_overrun); + set_intr_gate(X86_TRAP_TS, invalid_TSS); + set_intr_gate(X86_TRAP_NP, segment_not_present); + set_intr_gate_ist(X86_TRAP_SS, &stack_segment, STACKFAULT_STACK); + set_intr_gate(X86_TRAP_GP, general_protection); + set_intr_gate(X86_TRAP_SPURIOUS, spurious_interrupt_bug); + set_intr_gate(X86_TRAP_MF, coprocessor_error); + set_intr_gate(X86_TRAP_AC, alignment_check); #ifdef CONFIG_X86_MCE - set_intr_gate_ist(18, &machine_check, MCE_STACK); + set_intr_gate_ist(X86_TRAP_MC, &machine_check, MCE_STACK); #endif - set_intr_gate(19, &simd_coprocessor_error); + set_intr_gate(X86_TRAP_XF, simd_coprocessor_error); /* Reserve all the builtin and the syscall vector: */ for (i = 0; i < FIRST_EXTERNAL_VECTOR; i++) @@ -736,6 +803,14 @@ void __init trap_init(void) #endif /* + * Set the IDT descriptor to a fixed read-only location, so that the + * "sidt" instruction will not leak the location of the kernel, and + * to defend the IDT against arbitrary memory write vulnerabilities. + * It will be reloaded in cpu_init() */ + __set_fixmap(FIX_RO_IDT, __pa_symbol(idt_table), PAGE_KERNEL_RO); + idt_descr.address = fix_to_virt(FIX_RO_IDT); + + /* * Should be a barrier for any external CPU state: */ cpu_init(); @@ -743,8 +818,8 @@ void __init trap_init(void) x86_init.irqs.trap_init(); #ifdef CONFIG_X86_64 - memcpy(&nmi_idt_table, &idt_table, IDT_ENTRIES * 16); - set_nmi_gate(1, &debug); - set_nmi_gate(3, &int3); + memcpy(&debug_idt_table, &idt_table, IDT_ENTRIES * 16); + set_nmi_gate(X86_TRAP_DB, &debug); + set_nmi_gate(X86_TRAP_BP, &int3); #endif } diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c index a62c201c97e..ea030319b32 100644 --- a/arch/x86/kernel/tsc.c +++ b/arch/x86/kernel/tsc.c @@ -1,3 +1,5 @@ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/kernel.h> #include <linux/sched.h> #include <linux/init.h> @@ -9,6 +11,7 @@ #include <linux/clocksource.h> #include <linux/percpu.h> #include <linux/timex.h> +#include <linux/static_key.h> #include <asm/hpet.h> #include <asm/timer.h> @@ -35,13 +38,244 @@ static int __read_mostly tsc_unstable; erroneous rdtsc usage on !cpu_has_tsc processors */ static int __read_mostly tsc_disabled = -1; +static struct static_key __use_tsc = STATIC_KEY_INIT; + int tsc_clocksource_reliable; + +/* + * Use a ring-buffer like data structure, where a writer advances the head by + * writing a new data entry and a reader advances the tail when it observes a + * new entry. + * + * Writers are made to wait on readers until there's space to write a new + * entry. + * + * This means that we can always use an {offset, mul} pair to compute a ns + * value that is 'roughly' in the right direction, even if we're writing a new + * {offset, mul} pair during the clock read. + * + * The down-side is that we can no longer guarantee strict monotonicity anymore + * (assuming the TSC was that to begin with), because while we compute the + * intersection point of the two clock slopes and make sure the time is + * continuous at the point of switching; we can no longer guarantee a reader is + * strictly before or after the switch point. + * + * It does mean a reader no longer needs to disable IRQs in order to avoid + * CPU-Freq updates messing with his times, and similarly an NMI reader will + * no longer run the risk of hitting half-written state. + */ + +struct cyc2ns { + struct cyc2ns_data data[2]; /* 0 + 2*24 = 48 */ + struct cyc2ns_data *head; /* 48 + 8 = 56 */ + struct cyc2ns_data *tail; /* 56 + 8 = 64 */ +}; /* exactly fits one cacheline */ + +static DEFINE_PER_CPU_ALIGNED(struct cyc2ns, cyc2ns); + +struct cyc2ns_data *cyc2ns_read_begin(void) +{ + struct cyc2ns_data *head; + + preempt_disable(); + + head = this_cpu_read(cyc2ns.head); + /* + * Ensure we observe the entry when we observe the pointer to it. + * matches the wmb from cyc2ns_write_end(). + */ + smp_read_barrier_depends(); + head->__count++; + barrier(); + + return head; +} + +void cyc2ns_read_end(struct cyc2ns_data *head) +{ + barrier(); + /* + * If we're the outer most nested read; update the tail pointer + * when we're done. This notifies possible pending writers + * that we've observed the head pointer and that the other + * entry is now free. + */ + if (!--head->__count) { + /* + * x86-TSO does not reorder writes with older reads; + * therefore once this write becomes visible to another + * cpu, we must be finished reading the cyc2ns_data. + * + * matches with cyc2ns_write_begin(). + */ + this_cpu_write(cyc2ns.tail, head); + } + preempt_enable(); +} + +/* + * Begin writing a new @data entry for @cpu. + * + * Assumes some sort of write side lock; currently 'provided' by the assumption + * that cpufreq will call its notifiers sequentially. + */ +static struct cyc2ns_data *cyc2ns_write_begin(int cpu) +{ + struct cyc2ns *c2n = &per_cpu(cyc2ns, cpu); + struct cyc2ns_data *data = c2n->data; + + if (data == c2n->head) + data++; + + /* XXX send an IPI to @cpu in order to guarantee a read? */ + + /* + * When we observe the tail write from cyc2ns_read_end(), + * the cpu must be done with that entry and its safe + * to start writing to it. + */ + while (c2n->tail == data) + cpu_relax(); + + return data; +} + +static void cyc2ns_write_end(int cpu, struct cyc2ns_data *data) +{ + struct cyc2ns *c2n = &per_cpu(cyc2ns, cpu); + + /* + * Ensure the @data writes are visible before we publish the + * entry. Matches the data-depencency in cyc2ns_read_begin(). + */ + smp_wmb(); + + ACCESS_ONCE(c2n->head) = data; +} + +/* + * Accelerators for sched_clock() + * convert from cycles(64bits) => nanoseconds (64bits) + * basic equation: + * ns = cycles / (freq / ns_per_sec) + * ns = cycles * (ns_per_sec / freq) + * ns = cycles * (10^9 / (cpu_khz * 10^3)) + * ns = cycles * (10^6 / cpu_khz) + * + * Then we use scaling math (suggested by george@mvista.com) to get: + * ns = cycles * (10^6 * SC / cpu_khz) / SC + * ns = cycles * cyc2ns_scale / SC + * + * And since SC is a constant power of two, we can convert the div + * into a shift. + * + * We can use khz divisor instead of mhz to keep a better precision, since + * cyc2ns_scale is limited to 10^6 * 2^10, which fits in 32 bits. + * (mathieu.desnoyers@polymtl.ca) + * + * -johnstul@us.ibm.com "math is hard, lets go shopping!" + */ + +#define CYC2NS_SCALE_FACTOR 10 /* 2^10, carefully chosen */ + +static void cyc2ns_data_init(struct cyc2ns_data *data) +{ + data->cyc2ns_mul = 0; + data->cyc2ns_shift = CYC2NS_SCALE_FACTOR; + data->cyc2ns_offset = 0; + data->__count = 0; +} + +static void cyc2ns_init(int cpu) +{ + struct cyc2ns *c2n = &per_cpu(cyc2ns, cpu); + + cyc2ns_data_init(&c2n->data[0]); + cyc2ns_data_init(&c2n->data[1]); + + c2n->head = c2n->data; + c2n->tail = c2n->data; +} + +static inline unsigned long long cycles_2_ns(unsigned long long cyc) +{ + struct cyc2ns_data *data, *tail; + unsigned long long ns; + + /* + * See cyc2ns_read_*() for details; replicated in order to avoid + * an extra few instructions that came with the abstraction. + * Notable, it allows us to only do the __count and tail update + * dance when its actually needed. + */ + + preempt_disable_notrace(); + data = this_cpu_read(cyc2ns.head); + tail = this_cpu_read(cyc2ns.tail); + + if (likely(data == tail)) { + ns = data->cyc2ns_offset; + ns += mul_u64_u32_shr(cyc, data->cyc2ns_mul, CYC2NS_SCALE_FACTOR); + } else { + data->__count++; + + barrier(); + + ns = data->cyc2ns_offset; + ns += mul_u64_u32_shr(cyc, data->cyc2ns_mul, CYC2NS_SCALE_FACTOR); + + barrier(); + + if (!--data->__count) + this_cpu_write(cyc2ns.tail, data); + } + preempt_enable_notrace(); + + return ns; +} + +/* XXX surely we already have this someplace in the kernel?! */ +#define DIV_ROUND(n, d) (((n) + ((d) / 2)) / (d)) + +static void set_cyc2ns_scale(unsigned long cpu_khz, int cpu) +{ + unsigned long long tsc_now, ns_now; + struct cyc2ns_data *data; + unsigned long flags; + + local_irq_save(flags); + sched_clock_idle_sleep_event(); + + if (!cpu_khz) + goto done; + + data = cyc2ns_write_begin(cpu); + + rdtscll(tsc_now); + ns_now = cycles_2_ns(tsc_now); + + /* + * Compute a new multiplier as per the above comment and ensure our + * time function is continuous; see the comment near struct + * cyc2ns_data. + */ + data->cyc2ns_mul = DIV_ROUND(NSEC_PER_MSEC << CYC2NS_SCALE_FACTOR, cpu_khz); + data->cyc2ns_shift = CYC2NS_SCALE_FACTOR; + data->cyc2ns_offset = ns_now - + mul_u64_u32_shr(tsc_now, data->cyc2ns_mul, CYC2NS_SCALE_FACTOR); + + cyc2ns_write_end(cpu, data); + +done: + sched_clock_idle_wakeup_event(0); + local_irq_restore(flags); +} /* * Scheduler clock - returns current time in nanosec units. */ u64 native_sched_clock(void) { - u64 this_offset; + u64 tsc_now; /* * Fall back to jiffies if there's no TSC available: @@ -51,16 +285,16 @@ u64 native_sched_clock(void) * very important for it to be as fast as the platform * can achieve it. ) */ - if (unlikely(tsc_disabled)) { + if (!static_key_false(&__use_tsc)) { /* No locking but a rare wrong value is not a big deal: */ return (jiffies_64 - INITIAL_JIFFIES) * (1000000000 / HZ); } /* read the Time Stamp Counter: */ - rdtscll(this_offset); + rdtscll(tsc_now); /* return the value in ns */ - return __cycles_2_ns(this_offset); + return cycles_2_ns(tsc_now); } /* We need to define a real function for sched_clock, to override the @@ -75,17 +309,28 @@ unsigned long long sched_clock(void) __attribute__((alias("native_sched_clock"))); #endif +unsigned long long native_read_tsc(void) +{ + return __native_read_tsc(); +} +EXPORT_SYMBOL(native_read_tsc); + int check_tsc_unstable(void) { return tsc_unstable; } EXPORT_SYMBOL_GPL(check_tsc_unstable); +int check_tsc_disabled(void) +{ + return tsc_disabled; +} +EXPORT_SYMBOL_GPL(check_tsc_disabled); + #ifdef CONFIG_X86_TSC int __init notsc_setup(char *str) { - printk(KERN_WARNING "notsc: Kernel compiled with CONFIG_X86_TSC, " - "cannot disable TSC completely.\n"); + pr_warn("Kernel compiled with CONFIG_X86_TSC, cannot disable TSC completely\n"); tsc_disabled = 1; return 1; } @@ -373,7 +618,7 @@ static unsigned long quick_pit_calibrate(void) goto success; } } - printk("Fast TSC calibration failed\n"); + pr_err("Fast TSC calibration failed\n"); return 0; success: @@ -392,7 +637,7 @@ success: */ delta *= PIT_TICK_RATE; do_div(delta, i*256*1000); - printk("Fast TSC calibration using PIT\n"); + pr_info("Fast TSC calibration using PIT\n"); return delta; } @@ -406,6 +651,13 @@ unsigned long native_calibrate_tsc(void) unsigned long flags, latch, ms, fast_calibrate; int hpet = is_hpet_enabled(), i, loopmin; + /* Calibrate TSC using MSR for Intel Atom SoCs */ + local_irq_save(flags); + fast_calibrate = try_msr_calibrate_tsc(); + local_irq_restore(flags); + if (fast_calibrate) + return fast_calibrate; + local_irq_save(flags); fast_calibrate = quick_pit_calibrate(); local_irq_restore(flags); @@ -487,9 +739,8 @@ unsigned long native_calibrate_tsc(void) * use the reference value, as it is more precise. */ if (delta >= 90 && delta <= 110) { - printk(KERN_INFO - "TSC: PIT calibration matches %s. %d loops\n", - hpet ? "HPET" : "PMTIMER", i + 1); + pr_info("PIT calibration matches %s. %d loops\n", + hpet ? "HPET" : "PMTIMER", i + 1); return tsc_ref_min; } @@ -511,38 +762,36 @@ unsigned long native_calibrate_tsc(void) */ if (tsc_pit_min == ULONG_MAX) { /* PIT gave no useful value */ - printk(KERN_WARNING "TSC: Unable to calibrate against PIT\n"); + pr_warn("Unable to calibrate against PIT\n"); /* We don't have an alternative source, disable TSC */ if (!hpet && !ref1 && !ref2) { - printk("TSC: No reference (HPET/PMTIMER) available\n"); + pr_notice("No reference (HPET/PMTIMER) available\n"); return 0; } /* The alternative source failed as well, disable TSC */ if (tsc_ref_min == ULONG_MAX) { - printk(KERN_WARNING "TSC: HPET/PMTIMER calibration " - "failed.\n"); + pr_warn("HPET/PMTIMER calibration failed\n"); return 0; } /* Use the alternative source */ - printk(KERN_INFO "TSC: using %s reference calibration\n", - hpet ? "HPET" : "PMTIMER"); + pr_info("using %s reference calibration\n", + hpet ? "HPET" : "PMTIMER"); return tsc_ref_min; } /* We don't have an alternative source, use the PIT calibration value */ if (!hpet && !ref1 && !ref2) { - printk(KERN_INFO "TSC: Using PIT calibration value\n"); + pr_info("Using PIT calibration value\n"); return tsc_pit_min; } /* The alternative source failed, use the PIT calibration value */ if (tsc_ref_min == ULONG_MAX) { - printk(KERN_WARNING "TSC: HPET/PMTIMER calibration failed. " - "Using PIT calibration\n"); + pr_warn("HPET/PMTIMER calibration failed. Using PIT calibration.\n"); return tsc_pit_min; } @@ -551,9 +800,9 @@ unsigned long native_calibrate_tsc(void) * the PIT value as we know that there are PMTIMERs around * running at double speed. At least we let the user know: */ - printk(KERN_WARNING "TSC: PIT calibration deviates from %s: %lu %lu.\n", - hpet ? "HPET" : "PMTIMER", tsc_pit_min, tsc_ref_min); - printk(KERN_INFO "TSC: Using PIT calibration value\n"); + pr_warn("PIT calibration deviates from %s: %lu %lu\n", + hpet ? "HPET" : "PMTIMER", tsc_pit_min, tsc_ref_min); + pr_info("Using PIT calibration value\n"); return tsc_pit_min; } @@ -579,59 +828,11 @@ int recalibrate_cpu_khz(void) EXPORT_SYMBOL(recalibrate_cpu_khz); -/* Accelerators for sched_clock() - * convert from cycles(64bits) => nanoseconds (64bits) - * basic equation: - * ns = cycles / (freq / ns_per_sec) - * ns = cycles * (ns_per_sec / freq) - * ns = cycles * (10^9 / (cpu_khz * 10^3)) - * ns = cycles * (10^6 / cpu_khz) - * - * Then we use scaling math (suggested by george@mvista.com) to get: - * ns = cycles * (10^6 * SC / cpu_khz) / SC - * ns = cycles * cyc2ns_scale / SC - * - * And since SC is a constant power of two, we can convert the div - * into a shift. - * - * We can use khz divisor instead of mhz to keep a better precision, since - * cyc2ns_scale is limited to 10^6 * 2^10, which fits in 32 bits. - * (mathieu.desnoyers@polymtl.ca) - * - * -johnstul@us.ibm.com "math is hard, lets go shopping!" - */ - -DEFINE_PER_CPU(unsigned long, cyc2ns); -DEFINE_PER_CPU(unsigned long long, cyc2ns_offset); - -static void set_cyc2ns_scale(unsigned long cpu_khz, int cpu) -{ - unsigned long long tsc_now, ns_now, *offset; - unsigned long flags, *scale; - - local_irq_save(flags); - sched_clock_idle_sleep_event(); - - scale = &per_cpu(cyc2ns, cpu); - offset = &per_cpu(cyc2ns_offset, cpu); - - rdtscll(tsc_now); - ns_now = __cycles_2_ns(tsc_now); - - if (cpu_khz) { - *scale = (NSEC_PER_MSEC << CYC2NS_SCALE_FACTOR)/cpu_khz; - *offset = ns_now - (tsc_now * *scale >> CYC2NS_SCALE_FACTOR); - } - - sched_clock_idle_wakeup_event(0); - local_irq_restore(flags); -} - static unsigned long long cyc2ns_suspend; -void save_sched_clock_state(void) +void tsc_save_sched_clock_state(void) { - if (!sched_clock_stable) + if (!sched_clock_stable()) return; cyc2ns_suspend = sched_clock(); @@ -645,22 +846,32 @@ void save_sched_clock_state(void) * that sched_clock() continues from the point where it was left off during * suspend. */ -void restore_sched_clock_state(void) +void tsc_restore_sched_clock_state(void) { unsigned long long offset; unsigned long flags; int cpu; - if (!sched_clock_stable) + if (!sched_clock_stable()) return; local_irq_save(flags); - __this_cpu_write(cyc2ns_offset, 0); + /* + * We're comming out of suspend, there's no concurrency yet; don't + * bother being nice about the RCU stuff, just write to both + * data fields. + */ + + this_cpu_write(cyc2ns.data[0].cyc2ns_offset, 0); + this_cpu_write(cyc2ns.data[1].cyc2ns_offset, 0); + offset = cyc2ns_suspend - sched_clock(); - for_each_possible_cpu(cpu) - per_cpu(cyc2ns_offset, cpu) = offset; + for_each_possible_cpu(cpu) { + per_cpu(cyc2ns.data[0].cyc2ns_offset, cpu) = offset; + per_cpu(cyc2ns.data[1].cyc2ns_offset, cpu) = offset; + } local_irq_restore(flags); } @@ -703,16 +914,15 @@ static int time_cpufreq_notifier(struct notifier_block *nb, unsigned long val, tsc_khz_ref = tsc_khz; } if ((val == CPUFREQ_PRECHANGE && freq->old < freq->new) || - (val == CPUFREQ_POSTCHANGE && freq->old > freq->new) || - (val == CPUFREQ_RESUMECHANGE)) { + (val == CPUFREQ_POSTCHANGE && freq->old > freq->new)) { *lpj = cpufreq_scale(loops_per_jiffy_ref, ref_freq, freq->new); tsc_khz = cpufreq_scale(tsc_khz_ref, ref_freq, freq->new); if (!(freq->flags & CPUFREQ_CONST_LOOPS)) mark_tsc_unstable("cpufreq changes"); - } - set_cyc2ns_scale(tsc_khz, freq->cpu); + set_cyc2ns_scale(tsc_khz, freq->cpu); + } return 0; } @@ -762,7 +972,8 @@ static cycle_t read_tsc(struct clocksource *cs) static void resume_tsc(struct clocksource *cs) { - clocksource_tsc.cycle_last = 0; + if (!boot_cpu_has(X86_FEATURE_NONSTOP_TSC_S3)) + clocksource_tsc.cycle_last = 0; } static struct clocksource clocksource_tsc = { @@ -773,18 +984,16 @@ static struct clocksource clocksource_tsc = { .mask = CLOCKSOURCE_MASK(64), .flags = CLOCK_SOURCE_IS_CONTINUOUS | CLOCK_SOURCE_MUST_VERIFY, -#ifdef CONFIG_X86_64 .archdata = { .vclock_mode = VCLOCK_TSC }, -#endif }; void mark_tsc_unstable(char *reason) { if (!tsc_unstable) { tsc_unstable = 1; - sched_clock_stable = 0; + clear_sched_clock_stable(); disable_sched_clock_irqtime(); - printk(KERN_INFO "Marking TSC unstable due to %s\n", reason); + pr_info("Marking TSC unstable due to %s\n", reason); /* Change only the rating, when not registered */ if (clocksource_tsc.mult) clocksource_mark_unstable(&clocksource_tsc); @@ -817,7 +1026,7 @@ static void __init check_system_tsc_reliable(void) * Make an educated guess if the TSC is trustworthy and synchronized * over all CPUs. */ -__cpuinit int unsynchronized_tsc(void) +int unsynchronized_tsc(void) { if (!cpu_has_tsc || tsc_unstable) return 1; @@ -911,9 +1120,9 @@ static void tsc_refine_calibration_work(struct work_struct *work) goto out; tsc_khz = freq; - printk(KERN_INFO "Refined TSC clocksource calibration: " - "%lu.%03lu MHz.\n", (unsigned long)tsc_khz / 1000, - (unsigned long)tsc_khz % 1000); + pr_info("Refined TSC clocksource calibration: %lu.%03lu MHz\n", + (unsigned long)tsc_khz / 1000, + (unsigned long)tsc_khz % 1000); out: clocksource_register_khz(&clocksource_tsc, tsc_khz); @@ -932,6 +1141,19 @@ static int __init init_tsc_clocksource(void) clocksource_tsc.rating = 0; clocksource_tsc.flags &= ~CLOCK_SOURCE_IS_CONTINUOUS; } + + if (boot_cpu_has(X86_FEATURE_NONSTOP_TSC_S3)) + clocksource_tsc.flags |= CLOCK_SOURCE_SUSPEND_NONSTOP; + + /* + * Trust the results of the earlier calibration on systems + * exporting a reliable TSC. + */ + if (boot_cpu_has(X86_FEATURE_TSC_RELIABLE)) { + clocksource_register_khz(&clocksource_tsc, tsc_khz); + return 0; + } + schedule_delayed_work(&tsc_irqwork, 0); return 0; } @@ -959,9 +1181,9 @@ void __init tsc_init(void) return; } - printk("Detected %lu.%03lu MHz processor.\n", - (unsigned long)cpu_khz / 1000, - (unsigned long)cpu_khz % 1000); + pr_info("Detected %lu.%03lu MHz processor\n", + (unsigned long)cpu_khz / 1000, + (unsigned long)cpu_khz % 1000); /* * Secondary CPUs do not run through tsc_init(), so set up @@ -969,14 +1191,18 @@ void __init tsc_init(void) * speed as the bootup CPU. (cpufreq notifiers will fix this * up if their speed diverges) */ - for_each_possible_cpu(cpu) + for_each_possible_cpu(cpu) { + cyc2ns_init(cpu); set_cyc2ns_scale(cpu_khz, cpu); + } if (tsc_disabled > 0) return; /* now allow native_sched_clock() to use rdtsc */ + tsc_disabled = 0; + static_key_slow_inc(&__use_tsc); if (!no_sched_irq_time) enable_sched_clock_irqtime(); @@ -1000,7 +1226,7 @@ void __init tsc_init(void) * been calibrated. This assumes that CONSTANT_TSC applies to all * cpus in the socket - this should be a safe assumption. */ -unsigned long __cpuinit calibrate_delay_is_known(void) +unsigned long calibrate_delay_is_known(void) { int i, cpu = smp_processor_id(); diff --git a/arch/x86/kernel/tsc_msr.c b/arch/x86/kernel/tsc_msr.c new file mode 100644 index 00000000000..92ae6acac8a --- /dev/null +++ b/arch/x86/kernel/tsc_msr.c @@ -0,0 +1,127 @@ +/* + * tsc_msr.c - MSR based TSC calibration on Intel Atom SoC platforms. + * + * TSC in Intel Atom SoC runs at a constant rate which can be figured + * by this formula: + * <maximum core-clock to bus-clock ratio> * <maximum resolved frequency> + * See Intel 64 and IA-32 System Programming Guid section 16.12 and 30.11.5 + * for details. + * Especially some Intel Atom SoCs don't have PIT(i8254) or HPET, so MSR + * based calibration is the only option. + * + * + * Copyright (C) 2013 Intel Corporation + * Author: Bin Gao <bin.gao@intel.com> + * + * This file is released under the GPLv2. + */ + +#include <linux/kernel.h> +#include <asm/processor.h> +#include <asm/setup.h> +#include <asm/apic.h> +#include <asm/param.h> + +/* CPU reference clock frequency: in KHz */ +#define FREQ_83 83200 +#define FREQ_100 99840 +#define FREQ_133 133200 +#define FREQ_166 166400 + +#define MAX_NUM_FREQS 8 + +/* + * According to Intel 64 and IA-32 System Programming Guide, + * if MSR_PERF_STAT[31] is set, the maximum resolved bus ratio can be + * read in MSR_PLATFORM_ID[12:8], otherwise in MSR_PERF_STAT[44:40]. + * Unfortunately some Intel Atom SoCs aren't quite compliant to this, + * so we need manually differentiate SoC families. This is what the + * field msr_plat does. + */ +struct freq_desc { + u8 x86_family; /* CPU family */ + u8 x86_model; /* model */ + u8 msr_plat; /* 1: use MSR_PLATFORM_INFO, 0: MSR_IA32_PERF_STATUS */ + u32 freqs[MAX_NUM_FREQS]; +}; + +static struct freq_desc freq_desc_tables[] = { + /* PNW */ + { 6, 0x27, 0, { 0, 0, 0, 0, 0, FREQ_100, 0, FREQ_83 } }, + /* CLV+ */ + { 6, 0x35, 0, { 0, FREQ_133, 0, 0, 0, FREQ_100, 0, FREQ_83 } }, + /* TNG */ + { 6, 0x4a, 1, { 0, FREQ_100, FREQ_133, 0, 0, 0, 0, 0 } }, + /* VLV2 */ + { 6, 0x37, 1, { FREQ_83, FREQ_100, FREQ_133, FREQ_166, 0, 0, 0, 0 } }, + /* ANN */ + { 6, 0x5a, 1, { FREQ_83, FREQ_100, FREQ_133, FREQ_100, 0, 0, 0, 0 } }, +}; + +static int match_cpu(u8 family, u8 model) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(freq_desc_tables); i++) { + if ((family == freq_desc_tables[i].x86_family) && + (model == freq_desc_tables[i].x86_model)) + return i; + } + + return -1; +} + +/* Map CPU reference clock freq ID(0-7) to CPU reference clock freq(KHz) */ +#define id_to_freq(cpu_index, freq_id) \ + (freq_desc_tables[cpu_index].freqs[freq_id]) + +/* + * Do MSR calibration only for known/supported CPUs. + * + * Returns the calibration value or 0 if MSR calibration failed. + */ +unsigned long try_msr_calibrate_tsc(void) +{ + u32 lo, hi, ratio, freq_id, freq; + unsigned long res; + int cpu_index; + + cpu_index = match_cpu(boot_cpu_data.x86, boot_cpu_data.x86_model); + if (cpu_index < 0) + return 0; + + if (freq_desc_tables[cpu_index].msr_plat) { + rdmsr(MSR_PLATFORM_INFO, lo, hi); + ratio = (lo >> 8) & 0x1f; + } else { + rdmsr(MSR_IA32_PERF_STATUS, lo, hi); + ratio = (hi >> 8) & 0x1f; + } + pr_info("Maximum core-clock to bus-clock ratio: 0x%x\n", ratio); + + if (!ratio) + goto fail; + + /* Get FSB FREQ ID */ + rdmsr(MSR_FSB_FREQ, lo, hi); + freq_id = lo & 0x7; + freq = id_to_freq(cpu_index, freq_id); + pr_info("Resolved frequency ID: %u, frequency: %u KHz\n", + freq_id, freq); + if (!freq) + goto fail; + + /* TSC frequency = maximum resolved freq * maximum resolved bus ratio */ + res = freq * ratio; + pr_info("TSC runs at %lu KHz\n", res); + +#ifdef CONFIG_X86_LOCAL_APIC + lapic_timer_frequency = (freq * 1000) / HZ; + pr_info("lapic_timer_frequency = %d\n", lapic_timer_frequency); +#endif + return res; + +fail: + pr_warn("Fast TSC calibration using MSR failed\n"); + return 0; +} diff --git a/arch/x86/kernel/tsc_sync.c b/arch/x86/kernel/tsc_sync.c index 9eba29b46cb..26488487bc6 100644 --- a/arch/x86/kernel/tsc_sync.c +++ b/arch/x86/kernel/tsc_sync.c @@ -16,7 +16,6 @@ */ #include <linux/spinlock.h> #include <linux/kernel.h> -#include <linux/init.h> #include <linux/smp.h> #include <linux/nmi.h> #include <asm/tsc.h> @@ -25,24 +24,24 @@ * Entry/exit counters that make sure that both CPUs * run the measurement code at once: */ -static __cpuinitdata atomic_t start_count; -static __cpuinitdata atomic_t stop_count; +static atomic_t start_count; +static atomic_t stop_count; /* * We use a raw spinlock in this exceptional case, because * we want to have the fastest, inlined, non-debug version * of a critical section, to be able to prove TSC time-warps: */ -static __cpuinitdata arch_spinlock_t sync_lock = __ARCH_SPIN_LOCK_UNLOCKED; +static arch_spinlock_t sync_lock = __ARCH_SPIN_LOCK_UNLOCKED; -static __cpuinitdata cycles_t last_tsc; -static __cpuinitdata cycles_t max_warp; -static __cpuinitdata int nr_warps; +static cycles_t last_tsc; +static cycles_t max_warp; +static int nr_warps; /* * TSC-warp measurement loop running on both CPUs: */ -static __cpuinit void check_tsc_warp(void) +static void check_tsc_warp(unsigned int timeout) { cycles_t start, now, prev, end; int i; @@ -51,9 +50,9 @@ static __cpuinit void check_tsc_warp(void) start = get_cycles(); rdtsc_barrier(); /* - * The measurement runs for 20 msecs: + * The measurement runs for 'timeout' msecs: */ - end = start + tsc_khz * 20ULL; + end = start + (cycles_t) tsc_khz * timeout; now = start; for (i = 0; ; i++) { @@ -99,10 +98,29 @@ static __cpuinit void check_tsc_warp(void) } /* + * If the target CPU coming online doesn't have any of its core-siblings + * online, a timeout of 20msec will be used for the TSC-warp measurement + * loop. Otherwise a smaller timeout of 2msec will be used, as we have some + * information about this socket already (and this information grows as we + * have more and more logical-siblings in that socket). + * + * Ideally we should be able to skip the TSC sync check on the other + * core-siblings, if the first logical CPU in a socket passed the sync test. + * But as the TSC is per-logical CPU and can potentially be modified wrongly + * by the bios, TSC sync test for smaller duration should be able + * to catch such errors. Also this will catch the condition where all the + * cores in the socket doesn't get reset at the same time. + */ +static inline unsigned int loop_timeout(int cpu) +{ + return (cpumask_weight(cpu_core_mask(cpu)) > 1) ? 2 : 20; +} + +/* * Source CPU calls into this - it waits for the freshly booted * target CPU to arrive and then starts the measurement: */ -void __cpuinit check_tsc_sync_source(int cpu) +void check_tsc_sync_source(int cpu) { int cpus = 2; @@ -135,7 +153,7 @@ void __cpuinit check_tsc_sync_source(int cpu) */ atomic_inc(&start_count); - check_tsc_warp(); + check_tsc_warp(loop_timeout(cpu)); while (atomic_read(&stop_count) != cpus-1) cpu_relax(); @@ -168,7 +186,7 @@ void __cpuinit check_tsc_sync_source(int cpu) /* * Freshly booted CPUs call into this: */ -void __cpuinit check_tsc_sync_target(void) +void check_tsc_sync_target(void) { int cpus = 2; @@ -183,7 +201,7 @@ void __cpuinit check_tsc_sync_target(void) while (atomic_read(&start_count) != cpus) cpu_relax(); - check_tsc_warp(); + check_tsc_warp(loop_timeout(smp_processor_id())); /* * Ok, we are done: diff --git a/arch/x86/kernel/uprobes.c b/arch/x86/kernel/uprobes.c new file mode 100644 index 00000000000..5d1cbfe4ae5 --- /dev/null +++ b/arch/x86/kernel/uprobes.c @@ -0,0 +1,928 @@ +/* + * User-space Probes (UProbes) for x86 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * + * Copyright (C) IBM Corporation, 2008-2011 + * Authors: + * Srikar Dronamraju + * Jim Keniston + */ +#include <linux/kernel.h> +#include <linux/sched.h> +#include <linux/ptrace.h> +#include <linux/uprobes.h> +#include <linux/uaccess.h> + +#include <linux/kdebug.h> +#include <asm/processor.h> +#include <asm/insn.h> + +/* Post-execution fixups. */ + +/* Adjust IP back to vicinity of actual insn */ +#define UPROBE_FIX_IP 0x01 + +/* Adjust the return address of a call insn */ +#define UPROBE_FIX_CALL 0x02 + +/* Instruction will modify TF, don't change it */ +#define UPROBE_FIX_SETF 0x04 + +#define UPROBE_FIX_RIP_SI 0x08 +#define UPROBE_FIX_RIP_DI 0x10 +#define UPROBE_FIX_RIP_BX 0x20 +#define UPROBE_FIX_RIP_MASK \ + (UPROBE_FIX_RIP_SI | UPROBE_FIX_RIP_DI | UPROBE_FIX_RIP_BX) + +#define UPROBE_TRAP_NR UINT_MAX + +/* Adaptations for mhiramat x86 decoder v14. */ +#define OPCODE1(insn) ((insn)->opcode.bytes[0]) +#define OPCODE2(insn) ((insn)->opcode.bytes[1]) +#define OPCODE3(insn) ((insn)->opcode.bytes[2]) +#define MODRM_REG(insn) X86_MODRM_REG((insn)->modrm.value) + +#define W(row, b0, b1, b2, b3, b4, b5, b6, b7, b8, b9, ba, bb, bc, bd, be, bf)\ + (((b0##UL << 0x0)|(b1##UL << 0x1)|(b2##UL << 0x2)|(b3##UL << 0x3) | \ + (b4##UL << 0x4)|(b5##UL << 0x5)|(b6##UL << 0x6)|(b7##UL << 0x7) | \ + (b8##UL << 0x8)|(b9##UL << 0x9)|(ba##UL << 0xa)|(bb##UL << 0xb) | \ + (bc##UL << 0xc)|(bd##UL << 0xd)|(be##UL << 0xe)|(bf##UL << 0xf)) \ + << (row % 32)) + +/* + * Good-instruction tables for 32-bit apps. This is non-const and volatile + * to keep gcc from statically optimizing it out, as variable_test_bit makes + * some versions of gcc to think only *(unsigned long*) is used. + */ +#if defined(CONFIG_X86_32) || defined(CONFIG_IA32_EMULATION) +static volatile u32 good_insns_32[256 / 32] = { + /* 0 1 2 3 4 5 6 7 8 9 a b c d e f */ + /* ---------------------------------------------- */ + W(0x00, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 0) | /* 00 */ + W(0x10, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 0) , /* 10 */ + W(0x20, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 0, 1) | /* 20 */ + W(0x30, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 0, 1) , /* 30 */ + W(0x40, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* 40 */ + W(0x50, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* 50 */ + W(0x60, 1, 1, 1, 0, 1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0) | /* 60 */ + W(0x70, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* 70 */ + W(0x80, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* 80 */ + W(0x90, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* 90 */ + W(0xa0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* a0 */ + W(0xb0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* b0 */ + W(0xc0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0) | /* c0 */ + W(0xd0, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* d0 */ + W(0xe0, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0) | /* e0 */ + W(0xf0, 0, 0, 1, 1, 0, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1) /* f0 */ + /* ---------------------------------------------- */ + /* 0 1 2 3 4 5 6 7 8 9 a b c d e f */ +}; +#else +#define good_insns_32 NULL +#endif + +/* Good-instruction tables for 64-bit apps */ +#if defined(CONFIG_X86_64) +static volatile u32 good_insns_64[256 / 32] = { + /* 0 1 2 3 4 5 6 7 8 9 a b c d e f */ + /* ---------------------------------------------- */ + W(0x00, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 0, 0) | /* 00 */ + W(0x10, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 0, 0) , /* 10 */ + W(0x20, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 0, 0) | /* 20 */ + W(0x30, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 0, 0) , /* 30 */ + W(0x40, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) | /* 40 */ + W(0x50, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* 50 */ + W(0x60, 0, 0, 0, 1, 1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0) | /* 60 */ + W(0x70, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* 70 */ + W(0x80, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* 80 */ + W(0x90, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* 90 */ + W(0xa0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* a0 */ + W(0xb0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* b0 */ + W(0xc0, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0) | /* c0 */ + W(0xd0, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* d0 */ + W(0xe0, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0) | /* e0 */ + W(0xf0, 0, 0, 1, 1, 0, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1) /* f0 */ + /* ---------------------------------------------- */ + /* 0 1 2 3 4 5 6 7 8 9 a b c d e f */ +}; +#else +#define good_insns_64 NULL +#endif + +/* Using this for both 64-bit and 32-bit apps */ +static volatile u32 good_2byte_insns[256 / 32] = { + /* 0 1 2 3 4 5 6 7 8 9 a b c d e f */ + /* ---------------------------------------------- */ + W(0x00, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1) | /* 00 */ + W(0x10, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1) , /* 10 */ + W(0x20, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1) | /* 20 */ + W(0x30, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* 30 */ + W(0x40, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* 40 */ + W(0x50, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* 50 */ + W(0x60, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* 60 */ + W(0x70, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1) , /* 70 */ + W(0x80, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* 80 */ + W(0x90, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* 90 */ + W(0xa0, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 0, 1) | /* a0 */ + W(0xb0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1) , /* b0 */ + W(0xc0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* c0 */ + W(0xd0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* d0 */ + W(0xe0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* e0 */ + W(0xf0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0) /* f0 */ + /* ---------------------------------------------- */ + /* 0 1 2 3 4 5 6 7 8 9 a b c d e f */ +}; +#undef W + +/* + * opcodes we'll probably never support: + * + * 6c-6d, e4-e5, ec-ed - in + * 6e-6f, e6-e7, ee-ef - out + * cc, cd - int3, int + * cf - iret + * d6 - illegal instruction + * f1 - int1/icebp + * f4 - hlt + * fa, fb - cli, sti + * 0f - lar, lsl, syscall, clts, sysret, sysenter, sysexit, invd, wbinvd, ud2 + * + * invalid opcodes in 64-bit mode: + * + * 06, 0e, 16, 1e, 27, 2f, 37, 3f, 60-62, 82, c4-c5, d4-d5 + * 63 - we support this opcode in x86_64 but not in i386. + * + * opcodes we may need to refine support for: + * + * 0f - 2-byte instructions: For many of these instructions, the validity + * depends on the prefix and/or the reg field. On such instructions, we + * just consider the opcode combination valid if it corresponds to any + * valid instruction. + * + * 8f - Group 1 - only reg = 0 is OK + * c6-c7 - Group 11 - only reg = 0 is OK + * d9-df - fpu insns with some illegal encodings + * f2, f3 - repnz, repz prefixes. These are also the first byte for + * certain floating-point instructions, such as addsd. + * + * fe - Group 4 - only reg = 0 or 1 is OK + * ff - Group 5 - only reg = 0-6 is OK + * + * others -- Do we need to support these? + * + * 0f - (floating-point?) prefetch instructions + * 07, 17, 1f - pop es, pop ss, pop ds + * 26, 2e, 36, 3e - es:, cs:, ss:, ds: segment prefixes -- + * but 64 and 65 (fs: and gs:) seem to be used, so we support them + * 67 - addr16 prefix + * ce - into + * f0 - lock prefix + */ + +/* + * TODO: + * - Where necessary, examine the modrm byte and allow only valid instructions + * in the different Groups and fpu instructions. + */ + +static bool is_prefix_bad(struct insn *insn) +{ + int i; + + for (i = 0; i < insn->prefixes.nbytes; i++) { + switch (insn->prefixes.bytes[i]) { + case 0x26: /* INAT_PFX_ES */ + case 0x2E: /* INAT_PFX_CS */ + case 0x36: /* INAT_PFX_DS */ + case 0x3E: /* INAT_PFX_SS */ + case 0xF0: /* INAT_PFX_LOCK */ + return true; + } + } + return false; +} + +static int uprobe_init_insn(struct arch_uprobe *auprobe, struct insn *insn, bool x86_64) +{ + u32 volatile *good_insns; + + insn_init(insn, auprobe->insn, x86_64); + /* has the side-effect of processing the entire instruction */ + insn_get_length(insn); + if (WARN_ON_ONCE(!insn_complete(insn))) + return -ENOEXEC; + + if (is_prefix_bad(insn)) + return -ENOTSUPP; + + if (x86_64) + good_insns = good_insns_64; + else + good_insns = good_insns_32; + + if (test_bit(OPCODE1(insn), (unsigned long *)good_insns)) + return 0; + + if (insn->opcode.nbytes == 2) { + if (test_bit(OPCODE2(insn), (unsigned long *)good_2byte_insns)) + return 0; + } + + return -ENOTSUPP; +} + +#ifdef CONFIG_X86_64 +static inline bool is_64bit_mm(struct mm_struct *mm) +{ + return !config_enabled(CONFIG_IA32_EMULATION) || + !(mm->context.ia32_compat == TIF_IA32); +} +/* + * If arch_uprobe->insn doesn't use rip-relative addressing, return + * immediately. Otherwise, rewrite the instruction so that it accesses + * its memory operand indirectly through a scratch register. Set + * defparam->fixups accordingly. (The contents of the scratch register + * will be saved before we single-step the modified instruction, + * and restored afterward). + * + * We do this because a rip-relative instruction can access only a + * relatively small area (+/- 2 GB from the instruction), and the XOL + * area typically lies beyond that area. At least for instructions + * that store to memory, we can't execute the original instruction + * and "fix things up" later, because the misdirected store could be + * disastrous. + * + * Some useful facts about rip-relative instructions: + * + * - There's always a modrm byte with bit layout "00 reg 101". + * - There's never a SIB byte. + * - The displacement is always 4 bytes. + * - REX.B=1 bit in REX prefix, which normally extends r/m field, + * has no effect on rip-relative mode. It doesn't make modrm byte + * with r/m=101 refer to register 1101 = R13. + */ +static void riprel_analyze(struct arch_uprobe *auprobe, struct insn *insn) +{ + u8 *cursor; + u8 reg; + u8 reg2; + + if (!insn_rip_relative(insn)) + return; + + /* + * insn_rip_relative() would have decoded rex_prefix, vex_prefix, modrm. + * Clear REX.b bit (extension of MODRM.rm field): + * we want to encode low numbered reg, not r8+. + */ + if (insn->rex_prefix.nbytes) { + cursor = auprobe->insn + insn_offset_rex_prefix(insn); + /* REX byte has 0100wrxb layout, clearing REX.b bit */ + *cursor &= 0xfe; + } + /* + * Similar treatment for VEX3 prefix. + * TODO: add XOP/EVEX treatment when insn decoder supports them + */ + if (insn->vex_prefix.nbytes == 3) { + /* + * vex2: c5 rvvvvLpp (has no b bit) + * vex3/xop: c4/8f rxbmmmmm wvvvvLpp + * evex: 62 rxbR00mm wvvvv1pp zllBVaaa + * (evex will need setting of both b and x since + * in non-sib encoding evex.x is 4th bit of MODRM.rm) + * Setting VEX3.b (setting because it has inverted meaning): + */ + cursor = auprobe->insn + insn_offset_vex_prefix(insn) + 1; + *cursor |= 0x20; + } + + /* + * Convert from rip-relative addressing to register-relative addressing + * via a scratch register. + * + * This is tricky since there are insns with modrm byte + * which also use registers not encoded in modrm byte: + * [i]div/[i]mul: implicitly use dx:ax + * shift ops: implicitly use cx + * cmpxchg: implicitly uses ax + * cmpxchg8/16b: implicitly uses dx:ax and bx:cx + * Encoding: 0f c7/1 modrm + * The code below thinks that reg=1 (cx), chooses si as scratch. + * mulx: implicitly uses dx: mulx r/m,r1,r2 does r1:r2 = dx * r/m. + * First appeared in Haswell (BMI2 insn). It is vex-encoded. + * Example where none of bx,cx,dx can be used as scratch reg: + * c4 e2 63 f6 0d disp32 mulx disp32(%rip),%ebx,%ecx + * [v]pcmpistri: implicitly uses cx, xmm0 + * [v]pcmpistrm: implicitly uses xmm0 + * [v]pcmpestri: implicitly uses ax, dx, cx, xmm0 + * [v]pcmpestrm: implicitly uses ax, dx, xmm0 + * Evil SSE4.2 string comparison ops from hell. + * maskmovq/[v]maskmovdqu: implicitly uses (ds:rdi) as destination. + * Encoding: 0f f7 modrm, 66 0f f7 modrm, vex-encoded: c5 f9 f7 modrm. + * Store op1, byte-masked by op2 msb's in each byte, to (ds:rdi). + * AMD says it has no 3-operand form (vex.vvvv must be 1111) + * and that it can have only register operands, not mem + * (its modrm byte must have mode=11). + * If these restrictions will ever be lifted, + * we'll need code to prevent selection of di as scratch reg! + * + * Summary: I don't know any insns with modrm byte which + * use SI register implicitly. DI register is used only + * by one insn (maskmovq) and BX register is used + * only by one too (cmpxchg8b). + * BP is stack-segment based (may be a problem?). + * AX, DX, CX are off-limits (many implicit users). + * SP is unusable (it's stack pointer - think about "pop mem"; + * also, rsp+disp32 needs sib encoding -> insn length change). + */ + + reg = MODRM_REG(insn); /* Fetch modrm.reg */ + reg2 = 0xff; /* Fetch vex.vvvv */ + if (insn->vex_prefix.nbytes == 2) + reg2 = insn->vex_prefix.bytes[1]; + else if (insn->vex_prefix.nbytes == 3) + reg2 = insn->vex_prefix.bytes[2]; + /* + * TODO: add XOP, EXEV vvvv reading. + * + * vex.vvvv field is in bits 6-3, bits are inverted. + * But in 32-bit mode, high-order bit may be ignored. + * Therefore, let's consider only 3 low-order bits. + */ + reg2 = ((reg2 >> 3) & 0x7) ^ 0x7; + /* + * Register numbering is ax,cx,dx,bx, sp,bp,si,di, r8..r15. + * + * Choose scratch reg. Order is important: must not select bx + * if we can use si (cmpxchg8b case!) + */ + if (reg != 6 && reg2 != 6) { + reg2 = 6; + auprobe->defparam.fixups |= UPROBE_FIX_RIP_SI; + } else if (reg != 7 && reg2 != 7) { + reg2 = 7; + auprobe->defparam.fixups |= UPROBE_FIX_RIP_DI; + /* TODO (paranoia): force maskmovq to not use di */ + } else { + reg2 = 3; + auprobe->defparam.fixups |= UPROBE_FIX_RIP_BX; + } + /* + * Point cursor at the modrm byte. The next 4 bytes are the + * displacement. Beyond the displacement, for some instructions, + * is the immediate operand. + */ + cursor = auprobe->insn + insn_offset_modrm(insn); + /* + * Change modrm from "00 reg 101" to "10 reg reg2". Example: + * 89 05 disp32 mov %eax,disp32(%rip) becomes + * 89 86 disp32 mov %eax,disp32(%rsi) + */ + *cursor = 0x80 | (reg << 3) | reg2; +} + +static inline unsigned long * +scratch_reg(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + if (auprobe->defparam.fixups & UPROBE_FIX_RIP_SI) + return ®s->si; + if (auprobe->defparam.fixups & UPROBE_FIX_RIP_DI) + return ®s->di; + return ®s->bx; +} + +/* + * If we're emulating a rip-relative instruction, save the contents + * of the scratch register and store the target address in that register. + */ +static void riprel_pre_xol(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + if (auprobe->defparam.fixups & UPROBE_FIX_RIP_MASK) { + struct uprobe_task *utask = current->utask; + unsigned long *sr = scratch_reg(auprobe, regs); + + utask->autask.saved_scratch_register = *sr; + *sr = utask->vaddr + auprobe->defparam.ilen; + } +} + +static void riprel_post_xol(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + if (auprobe->defparam.fixups & UPROBE_FIX_RIP_MASK) { + struct uprobe_task *utask = current->utask; + unsigned long *sr = scratch_reg(auprobe, regs); + + *sr = utask->autask.saved_scratch_register; + } +} +#else /* 32-bit: */ +static inline bool is_64bit_mm(struct mm_struct *mm) +{ + return false; +} +/* + * No RIP-relative addressing on 32-bit + */ +static void riprel_analyze(struct arch_uprobe *auprobe, struct insn *insn) +{ +} +static void riprel_pre_xol(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ +} +static void riprel_post_xol(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ +} +#endif /* CONFIG_X86_64 */ + +struct uprobe_xol_ops { + bool (*emulate)(struct arch_uprobe *, struct pt_regs *); + int (*pre_xol)(struct arch_uprobe *, struct pt_regs *); + int (*post_xol)(struct arch_uprobe *, struct pt_regs *); + void (*abort)(struct arch_uprobe *, struct pt_regs *); +}; + +static inline int sizeof_long(void) +{ + return is_ia32_task() ? 4 : 8; +} + +static int default_pre_xol_op(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + riprel_pre_xol(auprobe, regs); + return 0; +} + +static int push_ret_address(struct pt_regs *regs, unsigned long ip) +{ + unsigned long new_sp = regs->sp - sizeof_long(); + + if (copy_to_user((void __user *)new_sp, &ip, sizeof_long())) + return -EFAULT; + + regs->sp = new_sp; + return 0; +} + +/* + * We have to fix things up as follows: + * + * Typically, the new ip is relative to the copied instruction. We need + * to make it relative to the original instruction (FIX_IP). Exceptions + * are return instructions and absolute or indirect jump or call instructions. + * + * If the single-stepped instruction was a call, the return address that + * is atop the stack is the address following the copied instruction. We + * need to make it the address following the original instruction (FIX_CALL). + * + * If the original instruction was a rip-relative instruction such as + * "movl %edx,0xnnnn(%rip)", we have instead executed an equivalent + * instruction using a scratch register -- e.g., "movl %edx,0xnnnn(%rsi)". + * We need to restore the contents of the scratch register + * (FIX_RIP_reg). + */ +static int default_post_xol_op(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + struct uprobe_task *utask = current->utask; + + riprel_post_xol(auprobe, regs); + if (auprobe->defparam.fixups & UPROBE_FIX_IP) { + long correction = utask->vaddr - utask->xol_vaddr; + regs->ip += correction; + } else if (auprobe->defparam.fixups & UPROBE_FIX_CALL) { + regs->sp += sizeof_long(); /* Pop incorrect return address */ + if (push_ret_address(regs, utask->vaddr + auprobe->defparam.ilen)) + return -ERESTART; + } + /* popf; tell the caller to not touch TF */ + if (auprobe->defparam.fixups & UPROBE_FIX_SETF) + utask->autask.saved_tf = true; + + return 0; +} + +static void default_abort_op(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + riprel_post_xol(auprobe, regs); +} + +static struct uprobe_xol_ops default_xol_ops = { + .pre_xol = default_pre_xol_op, + .post_xol = default_post_xol_op, + .abort = default_abort_op, +}; + +static bool branch_is_call(struct arch_uprobe *auprobe) +{ + return auprobe->branch.opc1 == 0xe8; +} + +#define CASE_COND \ + COND(70, 71, XF(OF)) \ + COND(72, 73, XF(CF)) \ + COND(74, 75, XF(ZF)) \ + COND(78, 79, XF(SF)) \ + COND(7a, 7b, XF(PF)) \ + COND(76, 77, XF(CF) || XF(ZF)) \ + COND(7c, 7d, XF(SF) != XF(OF)) \ + COND(7e, 7f, XF(ZF) || XF(SF) != XF(OF)) + +#define COND(op_y, op_n, expr) \ + case 0x ## op_y: DO((expr) != 0) \ + case 0x ## op_n: DO((expr) == 0) + +#define XF(xf) (!!(flags & X86_EFLAGS_ ## xf)) + +static bool is_cond_jmp_opcode(u8 opcode) +{ + switch (opcode) { + #define DO(expr) \ + return true; + CASE_COND + #undef DO + + default: + return false; + } +} + +static bool check_jmp_cond(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + unsigned long flags = regs->flags; + + switch (auprobe->branch.opc1) { + #define DO(expr) \ + return expr; + CASE_COND + #undef DO + + default: /* not a conditional jmp */ + return true; + } +} + +#undef XF +#undef COND +#undef CASE_COND + +static bool branch_emulate_op(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + unsigned long new_ip = regs->ip += auprobe->branch.ilen; + unsigned long offs = (long)auprobe->branch.offs; + + if (branch_is_call(auprobe)) { + /* + * If it fails we execute this (mangled, see the comment in + * branch_clear_offset) insn out-of-line. In the likely case + * this should trigger the trap, and the probed application + * should die or restart the same insn after it handles the + * signal, arch_uprobe_post_xol() won't be even called. + * + * But there is corner case, see the comment in ->post_xol(). + */ + if (push_ret_address(regs, new_ip)) + return false; + } else if (!check_jmp_cond(auprobe, regs)) { + offs = 0; + } + + regs->ip = new_ip + offs; + return true; +} + +static int branch_post_xol_op(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + BUG_ON(!branch_is_call(auprobe)); + /* + * We can only get here if branch_emulate_op() failed to push the ret + * address _and_ another thread expanded our stack before the (mangled) + * "call" insn was executed out-of-line. Just restore ->sp and restart. + * We could also restore ->ip and try to call branch_emulate_op() again. + */ + regs->sp += sizeof_long(); + return -ERESTART; +} + +static void branch_clear_offset(struct arch_uprobe *auprobe, struct insn *insn) +{ + /* + * Turn this insn into "call 1f; 1:", this is what we will execute + * out-of-line if ->emulate() fails. We only need this to generate + * a trap, so that the probed task receives the correct signal with + * the properly filled siginfo. + * + * But see the comment in ->post_xol(), in the unlikely case it can + * succeed. So we need to ensure that the new ->ip can not fall into + * the non-canonical area and trigger #GP. + * + * We could turn it into (say) "pushf", but then we would need to + * divorce ->insn[] and ->ixol[]. We need to preserve the 1st byte + * of ->insn[] for set_orig_insn(). + */ + memset(auprobe->insn + insn_offset_immediate(insn), + 0, insn->immediate.nbytes); +} + +static struct uprobe_xol_ops branch_xol_ops = { + .emulate = branch_emulate_op, + .post_xol = branch_post_xol_op, +}; + +/* Returns -ENOSYS if branch_xol_ops doesn't handle this insn */ +static int branch_setup_xol_ops(struct arch_uprobe *auprobe, struct insn *insn) +{ + u8 opc1 = OPCODE1(insn); + int i; + + switch (opc1) { + case 0xeb: /* jmp 8 */ + case 0xe9: /* jmp 32 */ + case 0x90: /* prefix* + nop; same as jmp with .offs = 0 */ + break; + + case 0xe8: /* call relative */ + branch_clear_offset(auprobe, insn); + break; + + case 0x0f: + if (insn->opcode.nbytes != 2) + return -ENOSYS; + /* + * If it is a "near" conditional jmp, OPCODE2() - 0x10 matches + * OPCODE1() of the "short" jmp which checks the same condition. + */ + opc1 = OPCODE2(insn) - 0x10; + default: + if (!is_cond_jmp_opcode(opc1)) + return -ENOSYS; + } + + /* + * 16-bit overrides such as CALLW (66 e8 nn nn) are not supported. + * Intel and AMD behavior differ in 64-bit mode: Intel ignores 66 prefix. + * No one uses these insns, reject any branch insns with such prefix. + */ + for (i = 0; i < insn->prefixes.nbytes; i++) { + if (insn->prefixes.bytes[i] == 0x66) + return -ENOTSUPP; + } + + auprobe->branch.opc1 = opc1; + auprobe->branch.ilen = insn->length; + auprobe->branch.offs = insn->immediate.value; + + auprobe->ops = &branch_xol_ops; + return 0; +} + +/** + * arch_uprobe_analyze_insn - instruction analysis including validity and fixups. + * @mm: the probed address space. + * @arch_uprobe: the probepoint information. + * @addr: virtual address at which to install the probepoint + * Return 0 on success or a -ve number on error. + */ +int arch_uprobe_analyze_insn(struct arch_uprobe *auprobe, struct mm_struct *mm, unsigned long addr) +{ + struct insn insn; + u8 fix_ip_or_call = UPROBE_FIX_IP; + int ret; + + ret = uprobe_init_insn(auprobe, &insn, is_64bit_mm(mm)); + if (ret) + return ret; + + ret = branch_setup_xol_ops(auprobe, &insn); + if (ret != -ENOSYS) + return ret; + + /* + * Figure out which fixups default_post_xol_op() will need to perform, + * and annotate defparam->fixups accordingly. + */ + switch (OPCODE1(&insn)) { + case 0x9d: /* popf */ + auprobe->defparam.fixups |= UPROBE_FIX_SETF; + break; + case 0xc3: /* ret or lret -- ip is correct */ + case 0xcb: + case 0xc2: + case 0xca: + case 0xea: /* jmp absolute -- ip is correct */ + fix_ip_or_call = 0; + break; + case 0x9a: /* call absolute - Fix return addr, not ip */ + fix_ip_or_call = UPROBE_FIX_CALL; + break; + case 0xff: + switch (MODRM_REG(&insn)) { + case 2: case 3: /* call or lcall, indirect */ + fix_ip_or_call = UPROBE_FIX_CALL; + break; + case 4: case 5: /* jmp or ljmp, indirect */ + fix_ip_or_call = 0; + break; + } + /* fall through */ + default: + riprel_analyze(auprobe, &insn); + } + + auprobe->defparam.ilen = insn.length; + auprobe->defparam.fixups |= fix_ip_or_call; + + auprobe->ops = &default_xol_ops; + return 0; +} + +/* + * arch_uprobe_pre_xol - prepare to execute out of line. + * @auprobe: the probepoint information. + * @regs: reflects the saved user state of current task. + */ +int arch_uprobe_pre_xol(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + struct uprobe_task *utask = current->utask; + + if (auprobe->ops->pre_xol) { + int err = auprobe->ops->pre_xol(auprobe, regs); + if (err) + return err; + } + + regs->ip = utask->xol_vaddr; + utask->autask.saved_trap_nr = current->thread.trap_nr; + current->thread.trap_nr = UPROBE_TRAP_NR; + + utask->autask.saved_tf = !!(regs->flags & X86_EFLAGS_TF); + regs->flags |= X86_EFLAGS_TF; + if (test_tsk_thread_flag(current, TIF_BLOCKSTEP)) + set_task_blockstep(current, false); + + return 0; +} + +/* + * If xol insn itself traps and generates a signal(Say, + * SIGILL/SIGSEGV/etc), then detect the case where a singlestepped + * instruction jumps back to its own address. It is assumed that anything + * like do_page_fault/do_trap/etc sets thread.trap_nr != -1. + * + * arch_uprobe_pre_xol/arch_uprobe_post_xol save/restore thread.trap_nr, + * arch_uprobe_xol_was_trapped() simply checks that ->trap_nr is not equal to + * UPROBE_TRAP_NR == -1 set by arch_uprobe_pre_xol(). + */ +bool arch_uprobe_xol_was_trapped(struct task_struct *t) +{ + if (t->thread.trap_nr != UPROBE_TRAP_NR) + return true; + + return false; +} + +/* + * Called after single-stepping. To avoid the SMP problems that can + * occur when we temporarily put back the original opcode to + * single-step, we single-stepped a copy of the instruction. + * + * This function prepares to resume execution after the single-step. + */ +int arch_uprobe_post_xol(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + struct uprobe_task *utask = current->utask; + bool send_sigtrap = utask->autask.saved_tf; + int err = 0; + + WARN_ON_ONCE(current->thread.trap_nr != UPROBE_TRAP_NR); + current->thread.trap_nr = utask->autask.saved_trap_nr; + + if (auprobe->ops->post_xol) { + err = auprobe->ops->post_xol(auprobe, regs); + if (err) { + /* + * Restore ->ip for restart or post mortem analysis. + * ->post_xol() must not return -ERESTART unless this + * is really possible. + */ + regs->ip = utask->vaddr; + if (err == -ERESTART) + err = 0; + send_sigtrap = false; + } + } + /* + * arch_uprobe_pre_xol() doesn't save the state of TIF_BLOCKSTEP + * so we can get an extra SIGTRAP if we do not clear TF. We need + * to examine the opcode to make it right. + */ + if (send_sigtrap) + send_sig(SIGTRAP, current, 0); + + if (!utask->autask.saved_tf) + regs->flags &= ~X86_EFLAGS_TF; + + return err; +} + +/* callback routine for handling exceptions. */ +int arch_uprobe_exception_notify(struct notifier_block *self, unsigned long val, void *data) +{ + struct die_args *args = data; + struct pt_regs *regs = args->regs; + int ret = NOTIFY_DONE; + + /* We are only interested in userspace traps */ + if (regs && !user_mode_vm(regs)) + return NOTIFY_DONE; + + switch (val) { + case DIE_INT3: + if (uprobe_pre_sstep_notifier(regs)) + ret = NOTIFY_STOP; + + break; + + case DIE_DEBUG: + if (uprobe_post_sstep_notifier(regs)) + ret = NOTIFY_STOP; + + default: + break; + } + + return ret; +} + +/* + * This function gets called when XOL instruction either gets trapped or + * the thread has a fatal signal. Reset the instruction pointer to its + * probed address for the potential restart or for post mortem analysis. + */ +void arch_uprobe_abort_xol(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + struct uprobe_task *utask = current->utask; + + if (auprobe->ops->abort) + auprobe->ops->abort(auprobe, regs); + + current->thread.trap_nr = utask->autask.saved_trap_nr; + regs->ip = utask->vaddr; + /* clear TF if it was set by us in arch_uprobe_pre_xol() */ + if (!utask->autask.saved_tf) + regs->flags &= ~X86_EFLAGS_TF; +} + +static bool __skip_sstep(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + if (auprobe->ops->emulate) + return auprobe->ops->emulate(auprobe, regs); + return false; +} + +bool arch_uprobe_skip_sstep(struct arch_uprobe *auprobe, struct pt_regs *regs) +{ + bool ret = __skip_sstep(auprobe, regs); + if (ret && (regs->flags & X86_EFLAGS_TF)) + send_sig(SIGTRAP, current, 0); + return ret; +} + +unsigned long +arch_uretprobe_hijack_return_addr(unsigned long trampoline_vaddr, struct pt_regs *regs) +{ + int rasize = sizeof_long(), nleft; + unsigned long orig_ret_vaddr = 0; /* clear high bits for 32-bit apps */ + + if (copy_from_user(&orig_ret_vaddr, (void __user *)regs->sp, rasize)) + return -1; + + /* check whether address has been already hijacked */ + if (orig_ret_vaddr == trampoline_vaddr) + return orig_ret_vaddr; + + nleft = copy_to_user((void __user *)regs->sp, &trampoline_vaddr, rasize); + if (likely(!nleft)) + return orig_ret_vaddr; + + if (nleft != rasize) { + pr_err("uprobe: return address clobbered: pid=%d, %%sp=%#lx, " + "%%ip=%#lx\n", current->pid, regs->sp, regs->ip); + + force_sig_info(SIGSEGV, SEND_SIG_FORCED, current); + } + + return -1; +} diff --git a/arch/x86/kernel/vm86_32.c b/arch/x86/kernel/vm86_32.c index b466cab5ba1..e8edcf52e06 100644 --- a/arch/x86/kernel/vm86_32.c +++ b/arch/x86/kernel/vm86_32.c @@ -28,9 +28,12 @@ * */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/capability.h> #include <linux/errno.h> #include <linux/interrupt.h> +#include <linux/syscalls.h> #include <linux/sched.h> #include <linux/kernel.h> #include <linux/signal.h> @@ -46,7 +49,6 @@ #include <asm/io.h> #include <asm/tlbflush.h> #include <asm/irq.h> -#include <asm/syscalls.h> /* * Known problems: @@ -137,14 +139,14 @@ struct pt_regs *save_v86_state(struct kernel_vm86_regs *regs) local_irq_enable(); if (!current->thread.vm86_info) { - printk("no vm86_info: BAD\n"); + pr_alert("no vm86_info: BAD\n"); do_exit(SIGSEGV); } set_flags(regs->pt.flags, VEFLAGS, X86_EFLAGS_VIF | current->thread.v86mask); tmp = copy_vm86_regs_to_user(¤t->thread.vm86_info->regs, regs); tmp += put_user(current->thread.screen_bitmap, ¤t->thread.vm86_info->screen_bitmap); if (tmp) { - printk("vm86: could not access userspace vm86_info\n"); + pr_alert("could not access userspace vm86_info\n"); do_exit(SIGSEGV); } @@ -172,6 +174,7 @@ static void mark_screen_rdonly(struct mm_struct *mm) spinlock_t *ptl; int i; + down_write(&mm->mmap_sem); pgd = pgd_offset(mm, 0xA0000); if (pgd_none_or_clear_bad(pgd)) goto out; @@ -179,7 +182,7 @@ static void mark_screen_rdonly(struct mm_struct *mm) if (pud_none_or_clear_bad(pud)) goto out; pmd = pmd_offset(pud, 0xA0000); - split_huge_page_pmd(mm, pmd); + split_huge_page_pmd_mm(mm, 0xA0000, pmd); if (pmd_none_or_clear_bad(pmd)) goto out; pte = pte_offset_map_lock(mm, pmd, 0xA0000, &ptl); @@ -190,6 +193,7 @@ static void mark_screen_rdonly(struct mm_struct *mm) } pte_unmap_unlock(pte, ptl); out: + up_write(&mm->mmap_sem); flush_tlb(); } @@ -198,36 +202,32 @@ out: static int do_vm86_irq_handling(int subfunction, int irqnumber); static void do_sys_vm86(struct kernel_vm86_struct *info, struct task_struct *tsk); -int sys_vm86old(struct vm86_struct __user *v86, struct pt_regs *regs) +SYSCALL_DEFINE1(vm86old, struct vm86_struct __user *, v86) { struct kernel_vm86_struct info; /* declare this _on top_, * this avoids wasting of stack space. * This remains on the stack until we * return to 32 bit user space. */ - struct task_struct *tsk; - int tmp, ret = -EPERM; + struct task_struct *tsk = current; + int tmp; - tsk = current; if (tsk->thread.saved_sp0) - goto out; + return -EPERM; tmp = copy_vm86_regs_from_user(&info.regs, &v86->regs, offsetof(struct kernel_vm86_struct, vm86plus) - sizeof(info.regs)); - ret = -EFAULT; if (tmp) - goto out; + return -EFAULT; memset(&info.vm86plus, 0, (int)&info.regs32 - (int)&info.vm86plus); - info.regs32 = regs; + info.regs32 = current_pt_regs(); tsk->thread.vm86_info = v86; do_sys_vm86(&info, tsk); - ret = 0; /* we never return here */ -out: - return ret; + return 0; /* we never return here */ } -int sys_vm86(unsigned long cmd, unsigned long arg, struct pt_regs *regs) +SYSCALL_DEFINE2(vm86, unsigned long, cmd, unsigned long, arg) { struct kernel_vm86_struct info; /* declare this _on top_, * this avoids wasting of stack space. @@ -235,7 +235,7 @@ int sys_vm86(unsigned long cmd, unsigned long arg, struct pt_regs *regs) * return to 32 bit user space. */ struct task_struct *tsk; - int tmp, ret; + int tmp; struct vm86plus_struct __user *v86; tsk = current; @@ -244,8 +244,7 @@ int sys_vm86(unsigned long cmd, unsigned long arg, struct pt_regs *regs) case VM86_FREE_IRQ: case VM86_GET_IRQ_BITS: case VM86_GET_AND_RESET_IRQ: - ret = do_vm86_irq_handling(cmd, (int)arg); - goto out; + return do_vm86_irq_handling(cmd, (int)arg); case VM86_PLUS_INSTALL_CHECK: /* * NOTE: on old vm86 stuff this will return the error @@ -253,28 +252,23 @@ int sys_vm86(unsigned long cmd, unsigned long arg, struct pt_regs *regs) * interpreted as (invalid) address to vm86_struct. * So the installation check works. */ - ret = 0; - goto out; + return 0; } /* we come here only for functions VM86_ENTER, VM86_ENTER_NO_BYPASS */ - ret = -EPERM; if (tsk->thread.saved_sp0) - goto out; + return -EPERM; v86 = (struct vm86plus_struct __user *)arg; tmp = copy_vm86_regs_from_user(&info.regs, &v86->regs, offsetof(struct kernel_vm86_struct, regs32) - sizeof(info.regs)); - ret = -EFAULT; if (tmp) - goto out; - info.regs32 = regs; + return -EFAULT; + info.regs32 = current_pt_regs(); info.vm86plus.is_vm86pus = 1; tsk->thread.vm86_info = (struct vm86_struct __user *)v86; do_sys_vm86(&info, tsk); - ret = 0; /* we never return here */ -out: - return ret; + return 0; /* we never return here */ } @@ -557,9 +551,9 @@ int handle_vm86_trap(struct kernel_vm86_regs *regs, long error_code, int trapno) if ((trapno == 3) || (trapno == 1)) { KVM86->regs32->ax = VM86_TRAP + (trapno << 8); /* setting this flag forces the code in entry_32.S to - call save_v86_state() and change the stack pointer - to KVM86->regs32 */ - set_thread_flag(TIF_IRET); + the path where we call save_v86_state() and change + the stack pointer to KVM86->regs32 */ + set_thread_flag(TIF_NOTIFY_RESUME); return 0; } do_int(regs, trapno, (unsigned char __user *) (regs->pt.ss << 4), SP(regs)); @@ -567,7 +561,7 @@ int handle_vm86_trap(struct kernel_vm86_regs *regs, long error_code, int trapno) } if (trapno != 1) return 1; /* we let this handle by the calling routine */ - current->thread.trap_no = trapno; + current->thread.trap_nr = trapno; current->thread.error_code = error_code; force_sig(SIGTRAP, current); return 0; diff --git a/arch/x86/kernel/vmlinux.lds.S b/arch/x86/kernel/vmlinux.lds.S index 0f703f10901..49edf2dd361 100644 --- a/arch/x86/kernel/vmlinux.lds.S +++ b/arch/x86/kernel/vmlinux.lds.S @@ -94,10 +94,6 @@ SECTIONS _text = .; /* bootstrapping code */ HEAD_TEXT -#ifdef CONFIG_X86_32 - . = ALIGN(PAGE_SIZE); - *(.text..page_aligned) -#endif . = ALIGN(8); _stext = .; TEXT_TEXT @@ -151,7 +147,6 @@ SECTIONS _edata = .; } :data -#ifdef CONFIG_X86_64 . = ALIGN(PAGE_SIZE); __vvar_page = .; @@ -169,12 +164,15 @@ SECTIONS #undef __VVAR_KERNEL_LDS #undef EMIT_VVAR + /* + * Pad the rest of the page with zeros. Otherwise the loader + * can leave garbage here. + */ + . = __vvar_beginning_hack + PAGE_SIZE; } :data . = ALIGN(__vvar_page + PAGE_SIZE, PAGE_SIZE); -#endif /* CONFIG_X86_64 */ - /* Init code and data - will be freed after init */ . = ALIGN(PAGE_SIZE); .init.begin : AT(ADDR(.init.begin) - LOAD_OFFSET) { @@ -197,24 +195,21 @@ SECTIONS INIT_DATA_SECTION(16) - /* - * Code and data for a variety of lowlevel trampolines, to be - * copied into base memory (< 1 MiB) during initialization. - * Since it is copied early, the main copy can be discarded - * afterwards. - */ - .x86_trampoline : AT(ADDR(.x86_trampoline) - LOAD_OFFSET) { - x86_trampoline_start = .; - *(.x86_trampoline) - x86_trampoline_end = .; - } - .x86_cpu_dev.init : AT(ADDR(.x86_cpu_dev.init) - LOAD_OFFSET) { __x86_cpu_dev_start = .; *(.x86_cpu_dev.init) __x86_cpu_dev_end = .; } +#ifdef CONFIG_X86_INTEL_MID + .x86_intel_mid_dev.init : AT(ADDR(.x86_intel_mid_dev.init) - \ + LOAD_OFFSET) { + __x86_intel_mid_dev_start = .; + *(.x86_intel_mid_dev.init) + __x86_intel_mid_dev_end = .; + } +#endif + /* * start address and size of operations which during runtime * can be patched with virtualization friendly instructions or diff --git a/arch/x86/kernel/vsmp_64.c b/arch/x86/kernel/vsmp_64.c index a1d804bcd48..b99b9ad8540 100644 --- a/arch/x86/kernel/vsmp_64.c +++ b/arch/x86/kernel/vsmp_64.c @@ -15,6 +15,8 @@ #include <linux/init.h> #include <linux/pci_ids.h> #include <linux/pci_regs.h> +#include <linux/smp.h> +#include <linux/irq.h> #include <asm/apic.h> #include <asm/pci-direct.h> @@ -22,6 +24,11 @@ #include <asm/paravirt.h> #include <asm/setup.h> +#define TOPOLOGY_REGISTER_OFFSET 0x10 + +/* Flag below is initialized once during vSMP PCI initialization. */ +static int irq_routing_comply = 1; + #if defined CONFIG_PCI && defined CONFIG_PARAVIRT /* * Interrupt control on vSMPowered systems: @@ -29,7 +36,7 @@ * and vice versa. */ -static unsigned long vsmp_save_fl(void) +asmlinkage __visible unsigned long vsmp_save_fl(void) { unsigned long flags = native_save_fl(); @@ -39,7 +46,7 @@ static unsigned long vsmp_save_fl(void) } PV_CALLEE_SAVE_REGS_THUNK(vsmp_save_fl); -static void vsmp_restore_fl(unsigned long flags) +__visible void vsmp_restore_fl(unsigned long flags) { if (flags & X86_EFLAGS_IF) flags &= ~X86_EFLAGS_AC; @@ -49,7 +56,7 @@ static void vsmp_restore_fl(unsigned long flags) } PV_CALLEE_SAVE_REGS_THUNK(vsmp_restore_fl); -static void vsmp_irq_disable(void) +asmlinkage __visible void vsmp_irq_disable(void) { unsigned long flags = native_save_fl(); @@ -57,7 +64,7 @@ static void vsmp_irq_disable(void) } PV_CALLEE_SAVE_REGS_THUNK(vsmp_irq_disable); -static void vsmp_irq_enable(void) +asmlinkage __visible void vsmp_irq_enable(void) { unsigned long flags = native_save_fl(); @@ -92,6 +99,22 @@ static void __init set_vsmp_pv_ops(void) ctl = readl(address + 4); printk(KERN_INFO "vSMP CTL: capabilities:0x%08x control:0x%08x\n", cap, ctl); + + /* If possible, let the vSMP foundation route the interrupt optimally */ +#ifdef CONFIG_SMP + if (cap & ctl & BIT(8)) { + ctl &= ~BIT(8); + + /* Interrupt routing set to ignore */ + irq_routing_comply = 0; + +#ifdef CONFIG_PROC_FS + /* Don't let users change irq affinity via procfs */ + no_irq_affinity = 1; +#endif + } +#endif + if (cap & ctl & (1 << 4)) { /* Setup irq ops and turn on vSMP IRQ fastpath handling */ pv_irq_ops.irq_disable = PV_CALLEE_SAVE(vsmp_irq_disable); @@ -99,12 +122,11 @@ static void __init set_vsmp_pv_ops(void) pv_irq_ops.save_fl = PV_CALLEE_SAVE(vsmp_save_fl); pv_irq_ops.restore_fl = PV_CALLEE_SAVE(vsmp_restore_fl); pv_init_ops.patch = vsmp_patch; - ctl &= ~(1 << 4); - writel(ctl, address + 4); - ctl = readl(address + 4); - printk(KERN_INFO "vSMP CTL: control set to:0x%08x\n", ctl); } + writel(ctl, address + 4); + ctl = readl(address + 4); + pr_info("vSMP CTL: control set to:0x%08x\n", ctl); early_iounmap(address, 8); } @@ -149,12 +171,75 @@ int is_vsmp_box(void) return 0; } #endif + +static void __init vsmp_cap_cpus(void) +{ +#if !defined(CONFIG_X86_VSMP) && defined(CONFIG_SMP) + void __iomem *address; + unsigned int cfg, topology, node_shift, maxcpus; + + /* + * CONFIG_X86_VSMP is not configured, so limit the number CPUs to the + * ones present in the first board, unless explicitly overridden by + * setup_max_cpus + */ + if (setup_max_cpus != NR_CPUS) + return; + + /* Read the vSMP Foundation topology register */ + cfg = read_pci_config(0, 0x1f, 0, PCI_BASE_ADDRESS_0); + address = early_ioremap(cfg + TOPOLOGY_REGISTER_OFFSET, 4); + if (WARN_ON(!address)) + return; + + topology = readl(address); + node_shift = (topology >> 16) & 0x7; + if (!node_shift) + /* The value 0 should be decoded as 8 */ + node_shift = 8; + maxcpus = (topology & ((1 << node_shift) - 1)) + 1; + + pr_info("vSMP CTL: Capping CPUs to %d (CONFIG_X86_VSMP is unset)\n", + maxcpus); + setup_max_cpus = maxcpus; + early_iounmap(address, 4); +#endif +} + +static int apicid_phys_pkg_id(int initial_apic_id, int index_msb) +{ + return hard_smp_processor_id() >> index_msb; +} + +/* + * In vSMP, all cpus should be capable of handling interrupts, regardless of + * the APIC used. + */ +static void fill_vector_allocation_domain(int cpu, struct cpumask *retmask, + const struct cpumask *mask) +{ + cpumask_setall(retmask); +} + +static void vsmp_apic_post_init(void) +{ + /* need to update phys_pkg_id */ + apic->phys_pkg_id = apicid_phys_pkg_id; + + if (!irq_routing_comply) + apic->vector_allocation_domain = fill_vector_allocation_domain; +} + void __init vsmp_init(void) { detect_vsmp_box(); if (!is_vsmp_box()) return; + x86_platform.apic_post_init = vsmp_apic_post_init; + + vsmp_cap_cpus(); + set_vsmp_pv_ops(); return; } diff --git a/arch/x86/kernel/vsyscall_64.c b/arch/x86/kernel/vsyscall_64.c index b07ba939356..ea5b5709aa7 100644 --- a/arch/x86/kernel/vsyscall_64.c +++ b/arch/x86/kernel/vsyscall_64.c @@ -18,6 +18,8 @@ * use the vDSO. */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/time.h> #include <linux/init.h> #include <linux/kernel.h> @@ -26,7 +28,7 @@ #include <linux/jiffies.h> #include <linux/sysctl.h> #include <linux/topology.h> -#include <linux/clocksource.h> +#include <linux/timekeeper_internal.h> #include <linux/getcpu.h> #include <linux/cpu.h> #include <linux/smp.h> @@ -45,17 +47,12 @@ #include <asm/segment.h> #include <asm/desc.h> #include <asm/topology.h> -#include <asm/vgtod.h> #include <asm/traps.h> #define CREATE_TRACE_POINTS #include "vsyscall_trace.h" DEFINE_VVAR(int, vgetcpu_mode); -DEFINE_VVAR(struct vsyscall_gtod_data, vsyscall_gtod_data) = -{ - .lock = __SEQLOCK_UNLOCKED(__vsyscall_gtod_data.lock), -}; static enum { EMULATE, NATIVE, NONE } vsyscall_mode = EMULATE; @@ -78,59 +75,23 @@ static int __init vsyscall_setup(char *str) } early_param("vsyscall", vsyscall_setup); -void update_vsyscall_tz(void) -{ - unsigned long flags; - - write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags); - /* sys_tz has changed */ - vsyscall_gtod_data.sys_tz = sys_tz; - write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags); -} - -void update_vsyscall(struct timespec *wall_time, struct timespec *wtm, - struct clocksource *clock, u32 mult) -{ - unsigned long flags; - - write_seqlock_irqsave(&vsyscall_gtod_data.lock, flags); - - /* copy vsyscall data */ - vsyscall_gtod_data.clock.vclock_mode = clock->archdata.vclock_mode; - vsyscall_gtod_data.clock.cycle_last = clock->cycle_last; - vsyscall_gtod_data.clock.mask = clock->mask; - vsyscall_gtod_data.clock.mult = mult; - vsyscall_gtod_data.clock.shift = clock->shift; - vsyscall_gtod_data.wall_time_sec = wall_time->tv_sec; - vsyscall_gtod_data.wall_time_nsec = wall_time->tv_nsec; - vsyscall_gtod_data.wall_to_monotonic = *wtm; - vsyscall_gtod_data.wall_time_coarse = __current_kernel_time(); - - write_sequnlock_irqrestore(&vsyscall_gtod_data.lock, flags); -} - static void warn_bad_vsyscall(const char *level, struct pt_regs *regs, const char *message) { - static DEFINE_RATELIMIT_STATE(rs, DEFAULT_RATELIMIT_INTERVAL, DEFAULT_RATELIMIT_BURST); - struct task_struct *tsk; - - if (!show_unhandled_signals || !__ratelimit(&rs)) + if (!show_unhandled_signals) return; - tsk = current; - - printk("%s%s[%d] %s ip:%lx cs:%lx sp:%lx ax:%lx si:%lx di:%lx\n", - level, tsk->comm, task_pid_nr(tsk), - message, regs->ip, regs->cs, - regs->sp, regs->ax, regs->si, regs->di); + pr_notice_ratelimited("%s%s[%d] %s ip:%lx cs:%lx sp:%lx ax:%lx si:%lx di:%lx\n", + level, current->comm, task_pid_nr(current), + message, regs->ip, regs->cs, + regs->sp, regs->ax, regs->si, regs->di); } static int addr_to_vsyscall_nr(unsigned long addr) { int nr; - if ((addr & ~0xC00UL) != VSYSCALL_START) + if ((addr & ~0xC00UL) != VSYSCALL_ADDR) return -EINVAL; nr = (addr & 0xC00UL) >> 10; @@ -153,7 +114,7 @@ static bool write_ok_or_segv(unsigned long ptr, size_t size) thread->error_code = 6; /* user fault, no page, write */ thread->cr2 = ptr; - thread->trap_no = 14; + thread->trap_nr = X86_TRAP_PF; memset(&info, 0, sizeof(info)); info.si_signo = SIGSEGV; @@ -172,7 +133,7 @@ bool emulate_vsyscall(struct pt_regs *regs, unsigned long address) { struct task_struct *tsk; unsigned long caller; - int vsyscall_nr; + int vsyscall_nr, syscall_nr, tmp; int prev_sig_on_uaccess_error; long ret; @@ -206,8 +167,63 @@ bool emulate_vsyscall(struct pt_regs *regs, unsigned long address) } tsk = current; - if (seccomp_mode(&tsk->seccomp)) - do_exit(SIGKILL); + + /* + * Check for access_ok violations and find the syscall nr. + * + * NULL is a valid user pointer (in the access_ok sense) on 32-bit and + * 64-bit, so we don't need to special-case it here. For all the + * vsyscalls, NULL means "don't write anything" not "write it at + * address 0". + */ + switch (vsyscall_nr) { + case 0: + if (!write_ok_or_segv(regs->di, sizeof(struct timeval)) || + !write_ok_or_segv(regs->si, sizeof(struct timezone))) { + ret = -EFAULT; + goto check_fault; + } + + syscall_nr = __NR_gettimeofday; + break; + + case 1: + if (!write_ok_or_segv(regs->di, sizeof(time_t))) { + ret = -EFAULT; + goto check_fault; + } + + syscall_nr = __NR_time; + break; + + case 2: + if (!write_ok_or_segv(regs->di, sizeof(unsigned)) || + !write_ok_or_segv(regs->si, sizeof(unsigned))) { + ret = -EFAULT; + goto check_fault; + } + + syscall_nr = __NR_getcpu; + break; + } + + /* + * Handle seccomp. regs->ip must be the original value. + * See seccomp_send_sigsys and Documentation/prctl/seccomp_filter.txt. + * + * We could optimize the seccomp disabled case, but performance + * here doesn't matter. + */ + regs->orig_ax = syscall_nr; + regs->ax = -ENOSYS; + tmp = secure_computing(syscall_nr); + if ((!tmp && regs->orig_ax != syscall_nr) || regs->ip != address) { + warn_bad_vsyscall(KERN_DEBUG, regs, + "seccomp tried to change syscall nr or ip"); + do_exit(SIGSYS); + } + if (tmp) + goto do_ret; /* skip requested */ /* * With a real vsyscall, page faults cause SIGSEGV. We want to @@ -216,44 +232,28 @@ bool emulate_vsyscall(struct pt_regs *regs, unsigned long address) prev_sig_on_uaccess_error = current_thread_info()->sig_on_uaccess_error; current_thread_info()->sig_on_uaccess_error = 1; - /* - * 0 is a valid user pointer (in the access_ok sense) on 32-bit and - * 64-bit, so we don't need to special-case it here. For all the - * vsyscalls, 0 means "don't write anything" not "write it at - * address 0". - */ ret = -EFAULT; switch (vsyscall_nr) { case 0: - if (!write_ok_or_segv(regs->di, sizeof(struct timeval)) || - !write_ok_or_segv(regs->si, sizeof(struct timezone))) - break; - ret = sys_gettimeofday( (struct timeval __user *)regs->di, (struct timezone __user *)regs->si); break; case 1: - if (!write_ok_or_segv(regs->di, sizeof(time_t))) - break; - ret = sys_time((time_t __user *)regs->di); break; case 2: - if (!write_ok_or_segv(regs->di, sizeof(unsigned)) || - !write_ok_or_segv(regs->si, sizeof(unsigned))) - break; - ret = sys_getcpu((unsigned __user *)regs->di, (unsigned __user *)regs->si, - 0); + NULL); break; } current_thread_info()->sig_on_uaccess_error = prev_sig_on_uaccess_error; +check_fault: if (ret == -EFAULT) { /* Bad news -- userspace fed a bad pointer to a vsyscall. */ warn_bad_vsyscall(KERN_INFO, regs, @@ -272,10 +272,10 @@ bool emulate_vsyscall(struct pt_regs *regs, unsigned long address) regs->ax = ret; +do_ret: /* Emulate a ret instruction. */ regs->ip = caller; regs->sp += 8; - return true; sigsegv: @@ -287,7 +287,7 @@ sigsegv: * Assume __initcall executes before all user space. Hopefully kmod * doesn't violate that. We'll find out if it does. */ -static void __cpuinit vsyscall_set_cpu(int cpu) +static void vsyscall_set_cpu(int cpu) { unsigned long d; unsigned long node = 0; @@ -309,13 +309,13 @@ static void __cpuinit vsyscall_set_cpu(int cpu) write_gdt_entry(get_cpu_gdt_table(cpu), GDT_ENTRY_PER_CPU, &d, DESCTYPE_S); } -static void __cpuinit cpu_vsyscall_init(void *arg) +static void cpu_vsyscall_init(void *arg) { /* preemption should be already off */ vsyscall_set_cpu(raw_smp_processor_id()); } -static int __cpuinit +static int cpu_vsyscall_notifier(struct notifier_block *n, unsigned long action, void *arg) { long cpu = (long)arg; @@ -330,28 +330,24 @@ void __init map_vsyscall(void) { extern char __vsyscall_page; unsigned long physaddr_vsyscall = __pa_symbol(&__vsyscall_page); - extern char __vvar_page; - unsigned long physaddr_vvar_page = __pa_symbol(&__vvar_page); - __set_fixmap(VSYSCALL_FIRST_PAGE, physaddr_vsyscall, + __set_fixmap(VSYSCALL_PAGE, physaddr_vsyscall, vsyscall_mode == NATIVE ? PAGE_KERNEL_VSYSCALL : PAGE_KERNEL_VVAR); - BUILD_BUG_ON((unsigned long)__fix_to_virt(VSYSCALL_FIRST_PAGE) != - (unsigned long)VSYSCALL_START); - - __set_fixmap(VVAR_PAGE, physaddr_vvar_page, PAGE_KERNEL_VVAR); - BUILD_BUG_ON((unsigned long)__fix_to_virt(VVAR_PAGE) != - (unsigned long)VVAR_ADDRESS); + BUILD_BUG_ON((unsigned long)__fix_to_virt(VSYSCALL_PAGE) != + (unsigned long)VSYSCALL_ADDR); } static int __init vsyscall_init(void) { - BUG_ON(VSYSCALL_ADDR(0) != __fix_to_virt(VSYSCALL_FIRST_PAGE)); + cpu_notifier_register_begin(); on_each_cpu(cpu_vsyscall_init, NULL, 1); /* notifier priority > KVM */ - hotcpu_notifier(cpu_vsyscall_notifier, 30); + __hotcpu_notifier(cpu_vsyscall_notifier, 30); + + cpu_notifier_register_done(); return 0; } diff --git a/arch/x86/kernel/vsyscall_gtod.c b/arch/x86/kernel/vsyscall_gtod.c new file mode 100644 index 00000000000..9531fbb123b --- /dev/null +++ b/arch/x86/kernel/vsyscall_gtod.c @@ -0,0 +1,69 @@ +/* + * Copyright (C) 2001 Andrea Arcangeli <andrea@suse.de> SuSE + * Copyright 2003 Andi Kleen, SuSE Labs. + * + * Modified for x86 32 bit architecture by + * Stefani Seibold <stefani@seibold.net> + * sponsored by Rohde & Schwarz GmbH & Co. KG Munich/Germany + * + * Thanks to hpa@transmeta.com for some useful hint. + * Special thanks to Ingo Molnar for his early experience with + * a different vsyscall implementation for Linux/IA32 and for the name. + * + */ + +#include <linux/timekeeper_internal.h> +#include <asm/vgtod.h> +#include <asm/vvar.h> + +DEFINE_VVAR(struct vsyscall_gtod_data, vsyscall_gtod_data); + +void update_vsyscall_tz(void) +{ + vsyscall_gtod_data.tz_minuteswest = sys_tz.tz_minuteswest; + vsyscall_gtod_data.tz_dsttime = sys_tz.tz_dsttime; +} + +void update_vsyscall(struct timekeeper *tk) +{ + struct vsyscall_gtod_data *vdata = &vsyscall_gtod_data; + + gtod_write_begin(vdata); + + /* copy vsyscall data */ + vdata->vclock_mode = tk->clock->archdata.vclock_mode; + vdata->cycle_last = tk->clock->cycle_last; + vdata->mask = tk->clock->mask; + vdata->mult = tk->mult; + vdata->shift = tk->shift; + + vdata->wall_time_sec = tk->xtime_sec; + vdata->wall_time_snsec = tk->xtime_nsec; + + vdata->monotonic_time_sec = tk->xtime_sec + + tk->wall_to_monotonic.tv_sec; + vdata->monotonic_time_snsec = tk->xtime_nsec + + ((u64)tk->wall_to_monotonic.tv_nsec + << tk->shift); + while (vdata->monotonic_time_snsec >= + (((u64)NSEC_PER_SEC) << tk->shift)) { + vdata->monotonic_time_snsec -= + ((u64)NSEC_PER_SEC) << tk->shift; + vdata->monotonic_time_sec++; + } + + vdata->wall_time_coarse_sec = tk->xtime_sec; + vdata->wall_time_coarse_nsec = (long)(tk->xtime_nsec >> tk->shift); + + vdata->monotonic_time_coarse_sec = + vdata->wall_time_coarse_sec + tk->wall_to_monotonic.tv_sec; + vdata->monotonic_time_coarse_nsec = + vdata->wall_time_coarse_nsec + tk->wall_to_monotonic.tv_nsec; + + while (vdata->monotonic_time_coarse_nsec >= NSEC_PER_SEC) { + vdata->monotonic_time_coarse_nsec -= NSEC_PER_SEC; + vdata->monotonic_time_coarse_sec++; + } + + gtod_write_end(vdata); +} diff --git a/arch/x86/kernel/x8664_ksyms_64.c b/arch/x86/kernel/x8664_ksyms_64.c index 9796c2f3d07..040681928e9 100644 --- a/arch/x86/kernel/x8664_ksyms_64.c +++ b/arch/x86/kernel/x8664_ksyms_64.c @@ -13,9 +13,13 @@ #include <asm/ftrace.h> #ifdef CONFIG_FUNCTION_TRACER -/* mcount is defined in assembly */ +/* mcount and __fentry__ are defined in assembly */ +#ifdef CC_USING_FENTRY +EXPORT_SYMBOL(__fentry__); +#else EXPORT_SYMBOL(mcount); #endif +#endif EXPORT_SYMBOL(__get_user_1); EXPORT_SYMBOL(__get_user_2); @@ -28,6 +32,7 @@ EXPORT_SYMBOL(__put_user_8); EXPORT_SYMBOL(copy_user_generic_string); EXPORT_SYMBOL(copy_user_generic_unrolled); +EXPORT_SYMBOL(copy_user_enhanced_fast_string); EXPORT_SYMBOL(__copy_user_nocache); EXPORT_SYMBOL(_copy_from_user); EXPORT_SYMBOL(_copy_to_user); @@ -54,7 +59,17 @@ EXPORT_SYMBOL(memcpy); EXPORT_SYMBOL(__memcpy); EXPORT_SYMBOL(memmove); +#ifndef CONFIG_DEBUG_VIRTUAL +EXPORT_SYMBOL(phys_base); +#endif EXPORT_SYMBOL(empty_zero_page); #ifndef CONFIG_PARAVIRT EXPORT_SYMBOL(native_load_gs_index); #endif + +#ifdef CONFIG_PREEMPT +EXPORT_SYMBOL(___preempt_schedule); +#ifdef CONFIG_CONTEXT_TRACKING +EXPORT_SYMBOL(___preempt_schedule_context); +#endif +#endif diff --git a/arch/x86/kernel/x86_init.c b/arch/x86/kernel/x86_init.c index 947a06ccc67..e48b674639c 100644 --- a/arch/x86/kernel/x86_init.c +++ b/arch/x86/kernel/x86_init.c @@ -18,17 +18,17 @@ #include <asm/e820.h> #include <asm/time.h> #include <asm/irq.h> +#include <asm/io_apic.h> +#include <asm/hpet.h> #include <asm/pat.h> #include <asm/tsc.h> #include <asm/iommu.h> #include <asm/mach_traps.h> -void __cpuinit x86_init_noop(void) { } +void x86_init_noop(void) { } void __init x86_init_uint_noop(unsigned int unused) { } -void __init x86_init_pgd_noop(pgd_t *unused) { } int __init iommu_init_noop(void) { return 0; } void iommu_shutdown_noop(void) { } -void wallclock_init_noop(void) { } /* * The platform setup functions are preset with the default functions @@ -63,13 +63,8 @@ struct x86_init_ops x86_init __initdata = { .banner = default_banner, }, - .mapping = { - .pagetable_reserve = native_pagetable_reserve, - }, - .paging = { - .pagetable_setup_start = native_pagetable_setup_start, - .pagetable_setup_done = native_pagetable_setup_done, + .pagetable_init = native_pagetable_init, }, .timers = { @@ -90,9 +85,9 @@ struct x86_init_ops x86_init __initdata = { }, }; -struct x86_cpuinit_ops x86_cpuinit __cpuinitdata = { +struct x86_cpuinit_ops x86_cpuinit = { + .early_percpu_clock_init = x86_init_noop, .setup_percpu_clockev = setup_secondary_APIC_clock, - .fixup_cpu_id = x86_default_fixup_cpu_id, }; static void default_nmi_init(void) { }; @@ -100,20 +95,69 @@ static int default_i8042_detect(void) { return 1; }; struct x86_platform_ops x86_platform = { .calibrate_tsc = native_calibrate_tsc, - .wallclock_init = wallclock_init_noop, .get_wallclock = mach_get_cmos_time, .set_wallclock = mach_set_rtc_mmss, .iommu_shutdown = iommu_shutdown_noop, .is_untracked_pat_range = is_ISA_range, .nmi_init = default_nmi_init, .get_nmi_reason = default_get_nmi_reason, - .i8042_detect = default_i8042_detect + .i8042_detect = default_i8042_detect, + .save_sched_clock_state = tsc_save_sched_clock_state, + .restore_sched_clock_state = tsc_restore_sched_clock_state, }; EXPORT_SYMBOL_GPL(x86_platform); + +#if defined(CONFIG_PCI_MSI) struct x86_msi_ops x86_msi = { - .setup_msi_irqs = native_setup_msi_irqs, - .teardown_msi_irq = native_teardown_msi_irq, - .teardown_msi_irqs = default_teardown_msi_irqs, - .restore_msi_irqs = default_restore_msi_irqs, + .setup_msi_irqs = native_setup_msi_irqs, + .compose_msi_msg = native_compose_msi_msg, + .teardown_msi_irq = native_teardown_msi_irq, + .teardown_msi_irqs = default_teardown_msi_irqs, + .restore_msi_irqs = default_restore_msi_irqs, + .setup_hpet_msi = default_setup_hpet_msi, + .msi_mask_irq = default_msi_mask_irq, + .msix_mask_irq = default_msix_mask_irq, +}; + +/* MSI arch specific hooks */ +int arch_setup_msi_irqs(struct pci_dev *dev, int nvec, int type) +{ + return x86_msi.setup_msi_irqs(dev, nvec, type); +} + +void arch_teardown_msi_irqs(struct pci_dev *dev) +{ + x86_msi.teardown_msi_irqs(dev); +} + +void arch_teardown_msi_irq(unsigned int irq) +{ + x86_msi.teardown_msi_irq(irq); +} + +void arch_restore_msi_irqs(struct pci_dev *dev) +{ + x86_msi.restore_msi_irqs(dev); +} +u32 arch_msi_mask_irq(struct msi_desc *desc, u32 mask, u32 flag) +{ + return x86_msi.msi_mask_irq(desc, mask, flag); +} +u32 arch_msix_mask_irq(struct msi_desc *desc, u32 flag) +{ + return x86_msi.msix_mask_irq(desc, flag); +} +#endif + +struct x86_io_apic_ops x86_io_apic_ops = { + .init = native_io_apic_init_mappings, + .read = native_io_apic_read, + .write = native_io_apic_write, + .modify = native_io_apic_modify, + .disable = native_disable_io_apic, + .print_entries = native_io_apic_print_entries, + .set_affinity = native_ioapic_set_affinity, + .setup_entry = native_setup_ioapic_entry, + .eoi_ioapic_pin = native_eoi_ioapic_pin, }; diff --git a/arch/x86/kernel/xsave.c b/arch/x86/kernel/xsave.c index 71109111411..a4b451c6add 100644 --- a/arch/x86/kernel/xsave.c +++ b/arch/x86/kernel/xsave.c @@ -3,12 +3,14 @@ * * Author: Suresh Siddha <suresh.b.siddha@intel.com> */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/bootmem.h> #include <linux/compat.h> #include <asm/i387.h> -#ifdef CONFIG_IA32_EMULATION -#include <asm/sigcontext32.h> -#endif +#include <asm/fpu-internal.h> +#include <asm/sigframe.h> #include <asm/xcr.h> /* @@ -19,13 +21,9 @@ u64 pcntxt_mask; /* * Represents init state for the supported extended state. */ -static struct xsave_struct *init_xstate_buf; - -struct _fpx_sw_bytes fx_sw_reserved; -#ifdef CONFIG_IA32_EMULATION -struct _fpx_sw_bytes fx_sw_reserved_ia32; -#endif +struct xsave_struct *init_xstate_buf; +static struct _fpx_sw_bytes fx_sw_reserved, fx_sw_reserved_ia32; static unsigned int *xstate_offsets, *xstate_sizes, xstate_features; /* @@ -40,15 +38,13 @@ static unsigned int *xstate_offsets, *xstate_sizes, xstate_features; */ void __sanitize_i387_state(struct task_struct *tsk) { - u64 xstate_bv; - int feature_bit = 0x2; struct i387_fxsave_struct *fx = &tsk->thread.fpu.state->fxsave; + int feature_bit = 0x2; + u64 xstate_bv; if (!fx) return; - BUG_ON(__thread_has_fpu(tsk)); - xstate_bv = tsk->thread.fpu.state->xsave.xsave_hdr.xstate_bv; /* @@ -102,213 +98,325 @@ void __sanitize_i387_state(struct task_struct *tsk) * Check for the presence of extended state information in the * user fpstate pointer in the sigcontext. */ -int check_for_xstate(struct i387_fxsave_struct __user *buf, - void __user *fpstate, - struct _fpx_sw_bytes *fx_sw_user) +static inline int check_for_xstate(struct i387_fxsave_struct __user *buf, + void __user *fpstate, + struct _fpx_sw_bytes *fx_sw) { int min_xstate_size = sizeof(struct i387_fxsave_struct) + sizeof(struct xsave_hdr_struct); unsigned int magic2; - int err; - err = __copy_from_user(fx_sw_user, &buf->sw_reserved[0], - sizeof(struct _fpx_sw_bytes)); - if (err) - return -EFAULT; + if (__copy_from_user(fx_sw, &buf->sw_reserved[0], sizeof(*fx_sw))) + return -1; - /* - * First Magic check failed. - */ - if (fx_sw_user->magic1 != FP_XSTATE_MAGIC1) - return -EINVAL; + /* Check for the first magic field and other error scenarios. */ + if (fx_sw->magic1 != FP_XSTATE_MAGIC1 || + fx_sw->xstate_size < min_xstate_size || + fx_sw->xstate_size > xstate_size || + fx_sw->xstate_size > fx_sw->extended_size) + return -1; /* - * Check for error scenarios. - */ - if (fx_sw_user->xstate_size < min_xstate_size || - fx_sw_user->xstate_size > xstate_size || - fx_sw_user->xstate_size > fx_sw_user->extended_size) - return -EINVAL; - - err = __get_user(magic2, (__u32 *) (((void *)fpstate) + - fx_sw_user->extended_size - - FP_XSTATE_MAGIC2_SIZE)); - if (err) - return err; - /* * Check for the presence of second magic word at the end of memory * layout. This detects the case where the user just copied the legacy * fpstate layout with out copying the extended state information * in the memory layout. */ - if (magic2 != FP_XSTATE_MAGIC2) - return -EFAULT; + if (__get_user(magic2, (__u32 __user *)(fpstate + fx_sw->xstate_size)) + || magic2 != FP_XSTATE_MAGIC2) + return -1; return 0; } -#ifdef CONFIG_X86_64 /* * Signal frame handlers. */ - -int save_i387_xstate(void __user *buf) +static inline int save_fsave_header(struct task_struct *tsk, void __user *buf) { - struct task_struct *tsk = current; - int err = 0; - - if (!access_ok(VERIFY_WRITE, buf, sig_xstate_size)) - return -EACCES; - - BUG_ON(sig_xstate_size < xstate_size); - - if ((unsigned long)buf % 64) - printk("save_i387_xstate: bad fpstate %p\n", buf); + if (use_fxsr()) { + struct xsave_struct *xsave = &tsk->thread.fpu.state->xsave; + struct user_i387_ia32_struct env; + struct _fpstate_ia32 __user *fp = buf; - if (!used_math()) - return 0; + convert_from_fxsr(&env, tsk); - if (user_has_fpu()) { - if (use_xsave()) - err = xsave_user(buf); - else - err = fxsave_user(buf); - - if (err) - return err; - user_fpu_end(); + if (__copy_to_user(buf, &env, sizeof(env)) || + __put_user(xsave->i387.swd, &fp->status) || + __put_user(X86_FXSR_MAGIC, &fp->magic)) + return -1; } else { - sanitize_i387_state(tsk); - if (__copy_to_user(buf, &tsk->thread.fpu.state->fxsave, - xstate_size)) + struct i387_fsave_struct __user *fp = buf; + u32 swd; + if (__get_user(swd, &fp->swd) || __put_user(swd, &fp->status)) return -1; } - clear_used_math(); /* trigger finit */ + return 0; +} - if (use_xsave()) { - struct _fpstate __user *fx = buf; - struct _xstate __user *x = buf; - u64 xstate_bv; +static inline int save_xstate_epilog(void __user *buf, int ia32_frame) +{ + struct xsave_struct __user *x = buf; + struct _fpx_sw_bytes *sw_bytes; + u32 xstate_bv; + int err; - err = __copy_to_user(&fx->sw_reserved, &fx_sw_reserved, - sizeof(struct _fpx_sw_bytes)); + /* Setup the bytes not touched by the [f]xsave and reserved for SW. */ + sw_bytes = ia32_frame ? &fx_sw_reserved_ia32 : &fx_sw_reserved; + err = __copy_to_user(&x->i387.sw_reserved, sw_bytes, sizeof(*sw_bytes)); - err |= __put_user(FP_XSTATE_MAGIC2, - (__u32 __user *) (buf + sig_xstate_size - - FP_XSTATE_MAGIC2_SIZE)); + if (!use_xsave()) + return err; - /* - * Read the xstate_bv which we copied (directly from the cpu or - * from the state in task struct) to the user buffers and - * set the FP/SSE bits. - */ - err |= __get_user(xstate_bv, &x->xstate_hdr.xstate_bv); + err |= __put_user(FP_XSTATE_MAGIC2, (__u32 *)(buf + xstate_size)); - /* - * For legacy compatible, we always set FP/SSE bits in the bit - * vector while saving the state to the user context. This will - * enable us capturing any changes(during sigreturn) to - * the FP/SSE bits by the legacy applications which don't touch - * xstate_bv in the xsave header. - * - * xsave aware apps can change the xstate_bv in the xsave - * header as well as change any contents in the memory layout. - * xrestore as part of sigreturn will capture all the changes. - */ - xstate_bv |= XSTATE_FPSSE; + /* + * Read the xstate_bv which we copied (directly from the cpu or + * from the state in task struct) to the user buffers. + */ + err |= __get_user(xstate_bv, (__u32 *)&x->xsave_hdr.xstate_bv); - err |= __put_user(xstate_bv, &x->xstate_hdr.xstate_bv); + /* + * For legacy compatible, we always set FP/SSE bits in the bit + * vector while saving the state to the user context. This will + * enable us capturing any changes(during sigreturn) to + * the FP/SSE bits by the legacy applications which don't touch + * xstate_bv in the xsave header. + * + * xsave aware apps can change the xstate_bv in the xsave + * header as well as change any contents in the memory layout. + * xrestore as part of sigreturn will capture all the changes. + */ + xstate_bv |= XSTATE_FPSSE; - if (err) - return err; - } + err |= __put_user(xstate_bv, (__u32 *)&x->xsave_hdr.xstate_bv); - return 1; + return err; +} + +static inline int save_user_xstate(struct xsave_struct __user *buf) +{ + int err; + + if (use_xsave()) + err = xsave_user(buf); + else if (use_fxsr()) + err = fxsave_user((struct i387_fxsave_struct __user *) buf); + else + err = fsave_user((struct i387_fsave_struct __user *) buf); + + if (unlikely(err) && __clear_user(buf, xstate_size)) + err = -EFAULT; + return err; } /* - * Restore the extended state if present. Otherwise, restore the FP/SSE - * state. + * Save the fpu, extended register state to the user signal frame. + * + * 'buf_fx' is the 64-byte aligned pointer at which the [f|fx|x]save + * state is copied. + * 'buf' points to the 'buf_fx' or to the fsave header followed by 'buf_fx'. + * + * buf == buf_fx for 64-bit frames and 32-bit fsave frame. + * buf != buf_fx for 32-bit frames with fxstate. + * + * If the fpu, extended register state is live, save the state directly + * to the user frame pointed by the aligned pointer 'buf_fx'. Otherwise, + * copy the thread's fpu state to the user frame starting at 'buf_fx'. + * + * If this is a 32-bit frame with fxstate, put a fsave header before + * the aligned state at 'buf_fx'. + * + * For [f]xsave state, update the SW reserved fields in the [f]xsave frame + * indicating the absence/presence of the extended state to the user. */ -static int restore_user_xstate(void __user *buf) +int save_xstate_sig(void __user *buf, void __user *buf_fx, int size) { - struct _fpx_sw_bytes fx_sw_user; - u64 mask; - int err; + struct xsave_struct *xsave = ¤t->thread.fpu.state->xsave; + struct task_struct *tsk = current; + int ia32_fxstate = (buf != buf_fx); - if (((unsigned long)buf % 64) || - check_for_xstate(buf, buf, &fx_sw_user)) - goto fx_only; + ia32_fxstate &= (config_enabled(CONFIG_X86_32) || + config_enabled(CONFIG_IA32_EMULATION)); - mask = fx_sw_user.xstate_bv; + if (!access_ok(VERIFY_WRITE, buf, size)) + return -EACCES; - /* - * restore the state passed by the user. - */ - err = xrestore_user(buf, mask); - if (err) - return err; + if (!static_cpu_has(X86_FEATURE_FPU)) + return fpregs_soft_get(current, NULL, 0, + sizeof(struct user_i387_ia32_struct), NULL, + (struct _fpstate_ia32 __user *) buf) ? -1 : 1; - /* - * init the state skipped by the user. - */ - mask = pcntxt_mask & ~mask; - if (unlikely(mask)) - xrstor_state(init_xstate_buf, mask); + if (user_has_fpu()) { + /* Save the live register state to the user directly. */ + if (save_user_xstate(buf_fx)) + return -1; + /* Update the thread's fxstate to save the fsave header. */ + if (ia32_fxstate) + fpu_fxsave(&tsk->thread.fpu); + } else { + sanitize_i387_state(tsk); + if (__copy_to_user(buf_fx, xsave, xstate_size)) + return -1; + } + + /* Save the fsave header for the 32-bit frames. */ + if ((ia32_fxstate || !use_fxsr()) && save_fsave_header(tsk, buf)) + return -1; + + if (use_fxsr() && save_xstate_epilog(buf_fx, ia32_fxstate)) + return -1; + + drop_init_fpu(tsk); /* trigger finit */ return 0; +} -fx_only: - /* - * couldn't find the extended state information in the - * memory layout. Restore just the FP/SSE and init all - * the other extended state. - */ - xrstor_state(init_xstate_buf, pcntxt_mask & ~XSTATE_FPSSE); - return fxrstor_checking((__force struct i387_fxsave_struct *)buf); +static inline void +sanitize_restored_xstate(struct task_struct *tsk, + struct user_i387_ia32_struct *ia32_env, + u64 xstate_bv, int fx_only) +{ + struct xsave_struct *xsave = &tsk->thread.fpu.state->xsave; + struct xsave_hdr_struct *xsave_hdr = &xsave->xsave_hdr; + + if (use_xsave()) { + /* These bits must be zero. */ + xsave_hdr->reserved1[0] = xsave_hdr->reserved1[1] = 0; + + /* + * Init the state that is not present in the memory + * layout and not enabled by the OS. + */ + if (fx_only) + xsave_hdr->xstate_bv = XSTATE_FPSSE; + else + xsave_hdr->xstate_bv &= (pcntxt_mask & xstate_bv); + } + + if (use_fxsr()) { + /* + * mscsr reserved bits must be masked to zero for security + * reasons. + */ + xsave->i387.mxcsr &= mxcsr_feature_mask; + + convert_to_fxsr(tsk, ia32_env); + } } /* - * This restores directly out of user space. Exceptions are handled. + * Restore the extended state if present. Otherwise, restore the FP/SSE state. */ -int restore_i387_xstate(void __user *buf) +static inline int restore_user_xstate(void __user *buf, u64 xbv, int fx_only) +{ + if (use_xsave()) { + if ((unsigned long)buf % 64 || fx_only) { + u64 init_bv = pcntxt_mask & ~XSTATE_FPSSE; + xrstor_state(init_xstate_buf, init_bv); + return fxrstor_user(buf); + } else { + u64 init_bv = pcntxt_mask & ~xbv; + if (unlikely(init_bv)) + xrstor_state(init_xstate_buf, init_bv); + return xrestore_user(buf, xbv); + } + } else if (use_fxsr()) { + return fxrstor_user(buf); + } else + return frstor_user(buf); +} + +int __restore_xstate_sig(void __user *buf, void __user *buf_fx, int size) { + int ia32_fxstate = (buf != buf_fx); struct task_struct *tsk = current; - int err = 0; + int state_size = xstate_size; + u64 xstate_bv = 0; + int fx_only = 0; + + ia32_fxstate &= (config_enabled(CONFIG_X86_32) || + config_enabled(CONFIG_IA32_EMULATION)); if (!buf) { - if (used_math()) - goto clear; + drop_init_fpu(tsk); return 0; - } else - if (!access_ok(VERIFY_READ, buf, sig_xstate_size)) - return -EACCES; + } - if (!used_math()) { - err = init_fpu(tsk); - if (err) - return err; + if (!access_ok(VERIFY_READ, buf, size)) + return -EACCES; + + if (!used_math() && init_fpu(tsk)) + return -1; + + if (!static_cpu_has(X86_FEATURE_FPU)) + return fpregs_soft_set(current, NULL, + 0, sizeof(struct user_i387_ia32_struct), + NULL, buf) != 0; + + if (use_xsave()) { + struct _fpx_sw_bytes fx_sw_user; + if (unlikely(check_for_xstate(buf_fx, buf_fx, &fx_sw_user))) { + /* + * Couldn't find the extended state information in the + * memory layout. Restore just the FP/SSE and init all + * the other extended state. + */ + state_size = sizeof(struct i387_fxsave_struct); + fx_only = 1; + } else { + state_size = fx_sw_user.xstate_size; + xstate_bv = fx_sw_user.xstate_bv; + } } - user_fpu_begin(); - if (use_xsave()) - err = restore_user_xstate(buf); - else - err = fxrstor_checking((__force struct i387_fxsave_struct *) - buf); - if (unlikely(err)) { + if (ia32_fxstate) { + /* + * For 32-bit frames with fxstate, copy the user state to the + * thread's fpu state, reconstruct fxstate from the fsave + * header. Sanitize the copied state etc. + */ + struct xsave_struct *xsave = &tsk->thread.fpu.state->xsave; + struct user_i387_ia32_struct env; + int err = 0; + + /* + * Drop the current fpu which clears used_math(). This ensures + * that any context-switch during the copy of the new state, + * avoids the intermediate state from getting restored/saved. + * Thus avoiding the new restored state from getting corrupted. + * We will be ready to restore/save the state only after + * set_used_math() is again set. + */ + drop_fpu(tsk); + + if (__copy_from_user(xsave, buf_fx, state_size) || + __copy_from_user(&env, buf, sizeof(env))) { + err = -1; + } else { + sanitize_restored_xstate(tsk, &env, xstate_bv, fx_only); + set_used_math(); + } + + if (use_eager_fpu()) + math_state_restore(); + + return err; + } else { /* - * Encountered an error while doing the restore from the - * user buffer, clear the fpu state. + * For 64-bit frames and 32-bit fsave frames, restore the user + * state to the registers directly (with exceptions handled). */ -clear: - clear_fpu(tsk); - clear_used_math(); + user_fpu_begin(); + if (restore_user_xstate(buf_fx, xstate_bv, fx_only)) { + drop_init_fpu(tsk); + return -1; + } } - return err; + + return 0; } -#endif /* * Prepare the SW reserved portion of the fxsave memory layout, indicating @@ -319,31 +427,22 @@ clear: */ static void prepare_fx_sw_frame(void) { - int size_extended = (xstate_size - sizeof(struct i387_fxsave_struct)) + - FP_XSTATE_MAGIC2_SIZE; - - sig_xstate_size = sizeof(struct _fpstate) + size_extended; - -#ifdef CONFIG_IA32_EMULATION - sig_xstate_ia32_size = sizeof(struct _fpstate_ia32) + size_extended; -#endif + int fsave_header_size = sizeof(struct i387_fsave_struct); + int size = xstate_size + FP_XSTATE_MAGIC2_SIZE; - memset(&fx_sw_reserved, 0, sizeof(fx_sw_reserved)); + if (config_enabled(CONFIG_X86_32)) + size += fsave_header_size; fx_sw_reserved.magic1 = FP_XSTATE_MAGIC1; - fx_sw_reserved.extended_size = sig_xstate_size; + fx_sw_reserved.extended_size = size; fx_sw_reserved.xstate_bv = pcntxt_mask; fx_sw_reserved.xstate_size = xstate_size; -#ifdef CONFIG_IA32_EMULATION - memcpy(&fx_sw_reserved_ia32, &fx_sw_reserved, - sizeof(struct _fpx_sw_bytes)); - fx_sw_reserved_ia32.extended_size = sig_xstate_ia32_size; -#endif -} -#ifdef CONFIG_X86_64 -unsigned int sig_xstate_size = sizeof(struct _fpstate); -#endif + if (config_enabled(CONFIG_IA32_EMULATION)) { + fx_sw_reserved_ia32 = fx_sw_reserved; + fx_sw_reserved_ia32.extended_size += fsave_header_size; + } +} /* * Enable the extended processor state save/restore feature @@ -382,19 +481,21 @@ static void __init setup_xstate_features(void) /* * setup the xstate image representing the init state */ -static void __init setup_xstate_init(void) +static void __init setup_init_fpu_buf(void) { - setup_xstate_features(); - /* * Setup init_xstate_buf to represent the init state of * all the features managed by the xsave */ init_xstate_buf = alloc_bootmem_align(xstate_size, __alignof__(struct xsave_struct)); - init_xstate_buf->i387.mxcsr = MXCSR_DEFAULT; + fx_finit(&init_xstate_buf->i387); + + if (!cpu_has_xsave) + return; + + setup_xstate_features(); - clts(); /* * Init all the features state with header_bv being 0x0 */ @@ -404,9 +505,21 @@ static void __init setup_xstate_init(void) * of any feature which is not represented by all zero's. */ xsave_state(init_xstate_buf, -1); - stts(); } +static enum { AUTO, ENABLE, DISABLE } eagerfpu = AUTO; +static int __init eager_fpu_setup(char *s) +{ + if (!strcmp(s, "on")) + eagerfpu = ENABLE; + else if (!strcmp(s, "off")) + eagerfpu = DISABLE; + else if (!strcmp(s, "auto")) + eagerfpu = AUTO; + return 1; +} +__setup("eagerfpu=", eager_fpu_setup); + /* * Enable and initialize the xsave feature. */ @@ -423,7 +536,7 @@ static void __init xstate_enable_boot_cpu(void) pcntxt_mask = eax + ((u64)edx << 32); if ((pcntxt_mask & XSTATE_FPSSE) != XSTATE_FPSSE) { - printk(KERN_ERR "FP/SSE not shown under xsave features 0x%llx\n", + pr_err("FP/SSE not shown under xsave features 0x%llx\n", pcntxt_mask); BUG(); } @@ -443,12 +556,24 @@ static void __init xstate_enable_boot_cpu(void) update_regset_xstate_info(xstate_size, pcntxt_mask); prepare_fx_sw_frame(); + setup_init_fpu_buf(); + + /* Auto enable eagerfpu for xsaveopt */ + if (cpu_has_xsaveopt && eagerfpu != DISABLE) + eagerfpu = ENABLE; + + if (pcntxt_mask & XSTATE_EAGER) { + if (eagerfpu == DISABLE) { + pr_err("eagerfpu not present, disabling some xstate features: 0x%llx\n", + pcntxt_mask & XSTATE_EAGER); + pcntxt_mask &= ~XSTATE_EAGER; + } else { + eagerfpu = ENABLE; + } + } - setup_xstate_init(); - - printk(KERN_INFO "xsave/xrstor: enabled xstate_bv 0x%llx, " - "cntxt size 0x%x\n", - pcntxt_mask, xstate_size); + pr_info("enabled xstate_bv 0x%llx, cntxt size 0x%x\n", + pcntxt_mask, xstate_size); } /* @@ -458,7 +583,7 @@ static void __init xstate_enable_boot_cpu(void) * This is somewhat obfuscated due to the lack of powerful enough * overrides for the section checks. */ -void __cpuinit xsave_init(void) +void xsave_init(void) { static __refdata void (*next_func)(void) = xstate_enable_boot_cpu; void (*this_func)(void); @@ -470,3 +595,43 @@ void __cpuinit xsave_init(void) next_func = xstate_enable; this_func(); } + +static inline void __init eager_fpu_init_bp(void) +{ + current->thread.fpu.state = + alloc_bootmem_align(xstate_size, __alignof__(struct xsave_struct)); + if (!init_xstate_buf) + setup_init_fpu_buf(); +} + +void eager_fpu_init(void) +{ + static __refdata void (*boot_func)(void) = eager_fpu_init_bp; + + clear_used_math(); + current_thread_info()->status = 0; + + if (eagerfpu == ENABLE) + setup_force_cpu_cap(X86_FEATURE_EAGER_FPU); + + if (!cpu_has_eager_fpu) { + stts(); + return; + } + + if (boot_func) { + boot_func(); + boot_func = NULL; + } + + /* + * This is same as math_state_restore(). But use_xsave() is + * not yet patched to use math_state_restore(). + */ + init_fpu(current); + __thread_fpu_begin(current); + if (cpu_has_xsave) + xrstor_state(init_xstate_buf, -1); + else + fxrstor_checking(&init_xstate_buf->i387); +} |