diff options
Diffstat (limited to 'arch/x86/kernel')
109 files changed, 12725 insertions, 9085 deletions
diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile index 77807d4769c..55ff016e9f6 100644 --- a/arch/x86/kernel/Makefile +++ b/arch/x86/kernel/Makefile @@ -2,7 +2,7 @@ # Makefile for the linux kernel. # -extra-y := head_$(BITS).o head$(BITS).o init_task.o vmlinux.lds +extra-y := head_$(BITS).o head$(BITS).o head.o init_task.o vmlinux.lds CPPFLAGS_vmlinux.lds += -U$(UTS_MACHINE) @@ -13,20 +13,21 @@ CPPFLAGS_vmlinux.lds += -U$(UTS_MACHINE) nostackp := $(call cc-option, -fno-stack-protector) CFLAGS_vsyscall_64.o := $(PROFILING) -g0 $(nostackp) CFLAGS_hpet.o := $(nostackp) -CFLAGS_tsc_64.o := $(nostackp) +CFLAGS_tsc.o := $(nostackp) obj-y := process_$(BITS).o signal_$(BITS).o entry_$(BITS).o obj-y += traps_$(BITS).o irq_$(BITS).o obj-y += time_$(BITS).o ioport.o ldt.o -obj-y += setup_$(BITS).o i8259_$(BITS).o setup.o +obj-y += setup.o i8259.o irqinit_$(BITS).o setup_percpu.o +obj-$(CONFIG_X86_VISWS) += visws_quirks.o +obj-$(CONFIG_X86_32) += probe_roms_32.o obj-$(CONFIG_X86_32) += sys_i386_32.o i386_ksyms_32.o obj-$(CONFIG_X86_64) += sys_x86_64.o x8664_ksyms_64.o -obj-$(CONFIG_X86_64) += syscall_64.o vsyscall_64.o setup64.o -obj-y += bootflag.o e820_$(BITS).o +obj-$(CONFIG_X86_64) += syscall_64.o vsyscall_64.o +obj-y += bootflag.o e820.o obj-y += pci-dma.o quirks.o i8237.o topology.o kdebugfs.o obj-y += alternative.o i8253.o pci-nommu.o -obj-$(CONFIG_X86_64) += bugs_64.o -obj-y += tsc_$(BITS).o io_delay.o rtc.o +obj-y += tsc.o io_delay.o rtc.o obj-$(CONFIG_X86_TRAMPOLINE) += trampoline.o obj-y += process.o @@ -53,7 +54,7 @@ obj-$(CONFIG_X86_32_SMP) += smpcommon.o obj-$(CONFIG_X86_64_SMP) += tsc_sync.o smpcommon.o obj-$(CONFIG_X86_TRAMPOLINE) += trampoline_$(BITS).o obj-$(CONFIG_X86_MPPARSE) += mpparse.o -obj-$(CONFIG_X86_LOCAL_APIC) += apic_$(BITS).o nmi_$(BITS).o +obj-$(CONFIG_X86_LOCAL_APIC) += apic_$(BITS).o nmi.o obj-$(CONFIG_X86_IO_APIC) += io_apic_$(BITS).o obj-$(CONFIG_X86_REBOOTFIXUPS) += reboot_fixups_32.o obj-$(CONFIG_KEXEC) += machine_kexec_$(BITS).o @@ -64,7 +65,6 @@ obj-$(CONFIG_X86_SUMMIT_NUMA) += summit_32.o obj-y += vsmp_64.o obj-$(CONFIG_KPROBES) += kprobes.o obj-$(CONFIG_MODULES) += module_$(BITS).o -obj-$(CONFIG_ACPI_SRAT) += srat_32.o obj-$(CONFIG_EFI) += efi.o efi_$(BITS).o efi_stub_$(BITS).o obj-$(CONFIG_DOUBLEFAULT) += doublefault_32.o obj-$(CONFIG_KGDB) += kgdb.o @@ -94,12 +94,13 @@ obj-$(CONFIG_OLPC) += olpc.o ### # 64 bit specific files ifeq ($(CONFIG_X86_64),y) - obj-y += genapic_64.o genapic_flat_64.o genx2apic_uv_x.o + obj-y += genapic_64.o genapic_flat_64.o genx2apic_uv_x.o tlb_uv.o obj-$(CONFIG_X86_PM_TIMER) += pmtimer_64.o obj-$(CONFIG_AUDIT) += audit_64.o obj-$(CONFIG_GART_IOMMU) += pci-gart_64.o aperture_64.o obj-$(CONFIG_CALGARY_IOMMU) += pci-calgary_64.o tce_64.o + obj-$(CONFIG_AMD_IOMMU) += amd_iommu_init.o amd_iommu.o obj-$(CONFIG_SWIOTLB) += pci-swiotlb_64.o obj-$(CONFIG_PCI_MMCONFIG) += mmconf-fam10h_64.o diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c index 33c5216fd3e..f489d7a9be9 100644 --- a/arch/x86/kernel/acpi/boot.c +++ b/arch/x86/kernel/acpi/boot.c @@ -37,6 +37,7 @@ #include <asm/pgtable.h> #include <asm/io_apic.h> #include <asm/apic.h> +#include <asm/genapic.h> #include <asm/io.h> #include <asm/mpspec.h> #include <asm/smp.h> @@ -106,21 +107,6 @@ static u64 acpi_lapic_addr __initdata = APIC_DEFAULT_PHYS_BASE; */ enum acpi_irq_model_id acpi_irq_model = ACPI_IRQ_MODEL_PIC; -#ifdef CONFIG_X86_64 - -/* rely on all ACPI tables being in the direct mapping */ -char *__init __acpi_map_table(unsigned long phys_addr, unsigned long size) -{ - if (!phys_addr || !size) - return NULL; - - if (phys_addr+size <= (max_pfn_mapped << PAGE_SHIFT) + PAGE_SIZE) - return __va(phys_addr); - - return NULL; -} - -#else /* * Temporarily use the virtual area starting from FIX_IO_APIC_BASE_END, @@ -139,11 +125,15 @@ char *__init __acpi_map_table(unsigned long phys, unsigned long size) unsigned long base, offset, mapped_size; int idx; - if (phys + size < 8 * 1024 * 1024) + if (!phys || !size) + return NULL; + + if (phys+size <= (max_low_pfn_mapped << PAGE_SHIFT)) return __va(phys); offset = phys & (PAGE_SIZE - 1); mapped_size = PAGE_SIZE - offset; + clear_fixmap(FIX_ACPI_END); set_fixmap(FIX_ACPI_END, phys); base = fix_to_virt(FIX_ACPI_END); @@ -155,13 +145,13 @@ char *__init __acpi_map_table(unsigned long phys, unsigned long size) if (--idx < FIX_ACPI_BEGIN) return NULL; /* cannot handle this */ phys += PAGE_SIZE; + clear_fixmap(idx); set_fixmap(idx, phys); mapped_size += PAGE_SIZE; } return ((unsigned char *)base + offset); } -#endif #ifdef CONFIG_PCI_MMCONFIG /* The physical address of the MMCONFIG aperture. Set from ACPI tables. */ @@ -338,8 +328,6 @@ acpi_parse_lapic_nmi(struct acpi_subtable_header * header, const unsigned long e #ifdef CONFIG_X86_IO_APIC -struct mp_ioapic_routing mp_ioapic_routing[MAX_IO_APICS]; - static int __init acpi_parse_ioapic(struct acpi_subtable_header * header, const unsigned long end) { @@ -514,8 +502,6 @@ int acpi_register_gsi(u32 gsi, int triggering, int polarity) * Make sure all (legacy) PCI IRQs are set as level-triggered. */ if (acpi_irq_model == ACPI_IRQ_MODEL_PIC) { - extern void eisa_set_level_irq(unsigned int irq); - if (triggering == ACPI_LEVEL_SENSITIVE) eisa_set_level_irq(gsi); } @@ -860,6 +846,364 @@ static int __init acpi_parse_madt_lapic_entries(void) #endif /* CONFIG_X86_LOCAL_APIC */ #ifdef CONFIG_X86_IO_APIC +#define MP_ISA_BUS 0 + +#ifdef CONFIG_X86_ES7000 +extern int es7000_plat; +#endif + +static struct { + int apic_id; + int gsi_base; + int gsi_end; + DECLARE_BITMAP(pin_programmed, MP_MAX_IOAPIC_PIN + 1); +} mp_ioapic_routing[MAX_IO_APICS]; + +static int mp_find_ioapic(int gsi) +{ + int i = 0; + + /* Find the IOAPIC that manages this GSI. */ + for (i = 0; i < nr_ioapics; i++) { + if ((gsi >= mp_ioapic_routing[i].gsi_base) + && (gsi <= mp_ioapic_routing[i].gsi_end)) + return i; + } + + printk(KERN_ERR "ERROR: Unable to locate IOAPIC for GSI %d\n", gsi); + return -1; +} + +static u8 __init uniq_ioapic_id(u8 id) +{ +#ifdef CONFIG_X86_32 + if ((boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) && + !APIC_XAPIC(apic_version[boot_cpu_physical_apicid])) + return io_apic_get_unique_id(nr_ioapics, id); + else + return id; +#else + int i; + DECLARE_BITMAP(used, 256); + bitmap_zero(used, 256); + for (i = 0; i < nr_ioapics; i++) { + struct mp_config_ioapic *ia = &mp_ioapics[i]; + __set_bit(ia->mp_apicid, used); + } + if (!test_bit(id, used)) + return id; + return find_first_zero_bit(used, 256); +#endif +} + +static int bad_ioapic(unsigned long address) +{ + if (nr_ioapics >= MAX_IO_APICS) { + printk(KERN_ERR "ERROR: Max # of I/O APICs (%d) exceeded " + "(found %d)\n", MAX_IO_APICS, nr_ioapics); + panic("Recompile kernel with bigger MAX_IO_APICS!\n"); + } + if (!address) { + printk(KERN_ERR "WARNING: Bogus (zero) I/O APIC address" + " found in table, skipping!\n"); + return 1; + } + return 0; +} + +void __init mp_register_ioapic(int id, u32 address, u32 gsi_base) +{ + int idx = 0; + + if (bad_ioapic(address)) + return; + + idx = nr_ioapics; + + mp_ioapics[idx].mp_type = MP_IOAPIC; + mp_ioapics[idx].mp_flags = MPC_APIC_USABLE; + mp_ioapics[idx].mp_apicaddr = address; + + set_fixmap_nocache(FIX_IO_APIC_BASE_0 + idx, address); + mp_ioapics[idx].mp_apicid = uniq_ioapic_id(id); +#ifdef CONFIG_X86_32 + mp_ioapics[idx].mp_apicver = io_apic_get_version(idx); +#else + mp_ioapics[idx].mp_apicver = 0; +#endif + /* + * Build basic GSI lookup table to facilitate gsi->io_apic lookups + * and to prevent reprogramming of IOAPIC pins (PCI GSIs). + */ + mp_ioapic_routing[idx].apic_id = mp_ioapics[idx].mp_apicid; + mp_ioapic_routing[idx].gsi_base = gsi_base; + mp_ioapic_routing[idx].gsi_end = gsi_base + + io_apic_get_redir_entries(idx); + + printk(KERN_INFO "IOAPIC[%d]: apic_id %d, version %d, address 0x%lx, " + "GSI %d-%d\n", idx, mp_ioapics[idx].mp_apicid, + mp_ioapics[idx].mp_apicver, mp_ioapics[idx].mp_apicaddr, + mp_ioapic_routing[idx].gsi_base, mp_ioapic_routing[idx].gsi_end); + + nr_ioapics++; +} + +static void assign_to_mp_irq(struct mp_config_intsrc *m, + struct mp_config_intsrc *mp_irq) +{ + memcpy(mp_irq, m, sizeof(struct mp_config_intsrc)); +} + +static int mp_irq_cmp(struct mp_config_intsrc *mp_irq, + struct mp_config_intsrc *m) +{ + return memcmp(mp_irq, m, sizeof(struct mp_config_intsrc)); +} + +static void save_mp_irq(struct mp_config_intsrc *m) +{ + int i; + + for (i = 0; i < mp_irq_entries; i++) { + if (!mp_irq_cmp(&mp_irqs[i], m)) + return; + } + + assign_to_mp_irq(m, &mp_irqs[mp_irq_entries]); + if (++mp_irq_entries == MAX_IRQ_SOURCES) + panic("Max # of irq sources exceeded!!\n"); +} + +void __init mp_override_legacy_irq(u8 bus_irq, u8 polarity, u8 trigger, u32 gsi) +{ + int ioapic; + int pin; + struct mp_config_intsrc mp_irq; + + /* + * Convert 'gsi' to 'ioapic.pin'. + */ + ioapic = mp_find_ioapic(gsi); + if (ioapic < 0) + return; + pin = gsi - mp_ioapic_routing[ioapic].gsi_base; + + /* + * TBD: This check is for faulty timer entries, where the override + * erroneously sets the trigger to level, resulting in a HUGE + * increase of timer interrupts! + */ + if ((bus_irq == 0) && (trigger == 3)) + trigger = 1; + + mp_irq.mp_type = MP_INTSRC; + mp_irq.mp_irqtype = mp_INT; + mp_irq.mp_irqflag = (trigger << 2) | polarity; + mp_irq.mp_srcbus = MP_ISA_BUS; + mp_irq.mp_srcbusirq = bus_irq; /* IRQ */ + mp_irq.mp_dstapic = mp_ioapics[ioapic].mp_apicid; /* APIC ID */ + mp_irq.mp_dstirq = pin; /* INTIN# */ + + save_mp_irq(&mp_irq); +} + +void __init mp_config_acpi_legacy_irqs(void) +{ + int i; + int ioapic; + unsigned int dstapic; + struct mp_config_intsrc mp_irq; + +#if defined (CONFIG_MCA) || defined (CONFIG_EISA) + /* + * Fabricate the legacy ISA bus (bus #31). + */ + mp_bus_id_to_type[MP_ISA_BUS] = MP_BUS_ISA; +#endif + set_bit(MP_ISA_BUS, mp_bus_not_pci); + Dprintk("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 + + /* + * Locate the IOAPIC that manages the ISA IRQs (0-15). + */ + ioapic = mp_find_ioapic(0); + if (ioapic < 0) + return; + dstapic = mp_ioapics[ioapic].mp_apicid; + + /* + * Use the default configuration for the IRQs 0-15. Unless + * overridden by (MADT) interrupt source override entries. + */ + for (i = 0; i < 16; i++) { + int idx; + + for (idx = 0; idx < mp_irq_entries; idx++) { + struct mp_config_intsrc *irq = mp_irqs + idx; + + /* Do we already have a mapping for this ISA IRQ? */ + if (irq->mp_srcbus == MP_ISA_BUS + && irq->mp_srcbusirq == i) + break; + + /* Do we already have a mapping for this IOAPIC pin */ + if (irq->mp_dstapic == dstapic && + irq->mp_dstirq == i) + break; + } + + if (idx != mp_irq_entries) { + printk(KERN_DEBUG "ACPI: IRQ%d used by override.\n", i); + continue; /* IRQ already used */ + } + + mp_irq.mp_type = MP_INTSRC; + mp_irq.mp_irqflag = 0; /* Conforming */ + mp_irq.mp_srcbus = MP_ISA_BUS; + mp_irq.mp_dstapic = dstapic; + mp_irq.mp_irqtype = mp_INT; + mp_irq.mp_srcbusirq = i; /* Identity mapped */ + mp_irq.mp_dstirq = i; + + save_mp_irq(&mp_irq); + } +} + +int mp_register_gsi(u32 gsi, int triggering, int polarity) +{ + int ioapic; + int ioapic_pin; +#ifdef CONFIG_X86_32 +#define MAX_GSI_NUM 4096 +#define IRQ_COMPRESSION_START 64 + + static int pci_irq = IRQ_COMPRESSION_START; + /* + * Mapping between Global System Interrupts, which + * represent all possible interrupts, and IRQs + * assigned to actual devices. + */ + static int gsi_to_irq[MAX_GSI_NUM]; +#else + + if (acpi_irq_model != ACPI_IRQ_MODEL_IOAPIC) + return gsi; +#endif + + /* Don't set up the ACPI SCI because it's already set up */ + if (acpi_gbl_FADT.sci_interrupt == gsi) + return gsi; + + ioapic = mp_find_ioapic(gsi); + if (ioapic < 0) { + printk(KERN_WARNING "No IOAPIC for GSI %u\n", gsi); + return gsi; + } + + ioapic_pin = gsi - mp_ioapic_routing[ioapic].gsi_base; + +#ifdef CONFIG_X86_32 + if (ioapic_renumber_irq) + gsi = ioapic_renumber_irq(ioapic, gsi); +#endif + + /* + * Avoid pin reprogramming. PRTs typically include entries + * with redundant pin->gsi mappings (but unique PCI devices); + * we only program the IOAPIC on the first. + */ + if (ioapic_pin > MP_MAX_IOAPIC_PIN) { + printk(KERN_ERR "Invalid reference to IOAPIC pin " + "%d-%d\n", mp_ioapic_routing[ioapic].apic_id, + ioapic_pin); + return gsi; + } + if (test_bit(ioapic_pin, mp_ioapic_routing[ioapic].pin_programmed)) { + Dprintk(KERN_DEBUG "Pin %d-%d already programmed\n", + mp_ioapic_routing[ioapic].apic_id, ioapic_pin); +#ifdef CONFIG_X86_32 + return (gsi < IRQ_COMPRESSION_START ? gsi : gsi_to_irq[gsi]); +#else + return gsi; +#endif + } + + set_bit(ioapic_pin, mp_ioapic_routing[ioapic].pin_programmed); +#ifdef CONFIG_X86_32 + /* + * For GSI >= 64, use IRQ compression + */ + if ((gsi >= IRQ_COMPRESSION_START) + && (triggering == ACPI_LEVEL_SENSITIVE)) { + /* + * For PCI devices assign IRQs in order, avoiding gaps + * due to unused I/O APIC pins. + */ + int irq = gsi; + if (gsi < MAX_GSI_NUM) { + /* + * Retain the VIA chipset work-around (gsi > 15), but + * avoid a problem where the 8254 timer (IRQ0) is setup + * via an override (so it's not on pin 0 of the ioapic), + * and at the same time, the pin 0 interrupt is a PCI + * type. The gsi > 15 test could cause these two pins + * to be shared as IRQ0, and they are not shareable. + * So test for this condition, and if necessary, avoid + * the pin collision. + */ + gsi = pci_irq++; + /* + * Don't assign IRQ used by ACPI SCI + */ + if (gsi == acpi_gbl_FADT.sci_interrupt) + gsi = pci_irq++; + gsi_to_irq[irq] = gsi; + } else { + printk(KERN_ERR "GSI %u is too high\n", gsi); + return gsi; + } + } +#endif + io_apic_set_pci_routing(ioapic, ioapic_pin, gsi, + triggering == ACPI_EDGE_SENSITIVE ? 0 : 1, + polarity == ACPI_ACTIVE_HIGH ? 0 : 1); + return gsi; +} + +int mp_config_acpi_gsi(unsigned char number, unsigned int devfn, u8 pin, + u32 gsi, int triggering, int polarity) +{ +#ifdef CONFIG_X86_MPPARSE + struct mp_config_intsrc mp_irq; + int ioapic; + + if (!acpi_ioapic) + return 0; + + /* print the entry should happen on mptable identically */ + mp_irq.mp_type = MP_INTSRC; + mp_irq.mp_irqtype = mp_INT; + mp_irq.mp_irqflag = (triggering == ACPI_EDGE_SENSITIVE ? 4 : 0x0c) | + (polarity == ACPI_ACTIVE_HIGH ? 1 : 3); + mp_irq.mp_srcbus = number; + mp_irq.mp_srcbusirq = (((devfn >> 3) & 0x1f) << 2) | ((pin - 1) & 3); + ioapic = mp_find_ioapic(gsi); + mp_irq.mp_dstapic = mp_ioapic_routing[ioapic].apic_id; + mp_irq.mp_dstirq = gsi - mp_ioapic_routing[ioapic].gsi_base; + + save_mp_irq(&mp_irq); +#endif + return 0; +} + /* * Parse IOAPIC related entries in MADT * returns 0 on success, < 0 on error @@ -1009,8 +1353,6 @@ static void __init acpi_process_madt(void) return; } -#ifdef __i386__ - static int __init disable_acpi_irq(const struct dmi_system_id *d) { if (!acpi_force) { @@ -1061,6 +1403,16 @@ static int __init force_acpi_ht(const struct dmi_system_id *d) } /* + * Force ignoring BIOS IRQ0 pin2 override + */ +static int __init dmi_ignore_irq0_timer_override(const struct dmi_system_id *d) +{ + pr_notice("%s detected: Ignoring BIOS IRQ0 pin2 override\n", d->ident); + acpi_skip_timer_override = 1; + return 0; +} + +/* * If your system is blacklisted here, but you find that acpi=force * works for you, please contact acpi-devel@sourceforge.net */ @@ -1227,11 +1579,35 @@ static struct dmi_system_id __initdata acpi_dmi_table[] = { DMI_MATCH(DMI_PRODUCT_NAME, "TravelMate 360"), }, }, + /* + * HP laptops which use a DSDT reporting as HP/SB400/10000, + * which includes some code which overrides all temperature + * trip points to 16C if the INTIN2 input of the I/O APIC + * 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 + * override in that cases. + */ + { + .callback = dmi_ignore_irq0_timer_override, + .ident = "HP NX6125 laptop", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"), + DMI_MATCH(DMI_PRODUCT_NAME, "HP Compaq nx6125"), + }, + }, + { + .callback = dmi_ignore_irq0_timer_override, + .ident = "HP NX6325 laptop", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "Hewlett-Packard"), + DMI_MATCH(DMI_PRODUCT_NAME, "HP Compaq nx6325"), + }, + }, {} }; -#endif /* __i386__ */ - /* * acpi_boot_table_init() and acpi_boot_init() * called from setup_arch(), always. @@ -1259,9 +1635,7 @@ int __init acpi_boot_table_init(void) { int error; -#ifdef __i386__ dmi_check_system(acpi_dmi_table); -#endif /* * If acpi_disabled, bail out @@ -1386,6 +1760,20 @@ static int __init parse_pci(char *arg) } early_param("pci", parse_pci); +int __init acpi_mps_check(void) +{ +#if defined(CONFIG_X86_LOCAL_APIC) && !defined(CONFIG_X86_MPPARSE) +/* mptable code is not built-in*/ + if (acpi_disabled || acpi_noirq) { + printk(KERN_WARNING "MPS support code is not built-in.\n" + "Using acpi=off or acpi=noirq or pci=noacpi " + "may have problem\n"); + return 1; + } +#endif + return 0; +} + #ifdef CONFIG_X86_IO_APIC static int __init parse_acpi_skip_timer_override(char *arg) { diff --git a/arch/x86/kernel/acpi/sleep.c b/arch/x86/kernel/acpi/sleep.c index 36af01f029e..e6a4b564cca 100644 --- a/arch/x86/kernel/acpi/sleep.c +++ b/arch/x86/kernel/acpi/sleep.c @@ -86,7 +86,9 @@ int acpi_save_state_mem(void) saved_magic = 0x12345678; #else /* CONFIG_64BIT */ header->trampoline_segment = setup_trampoline() >> 4; - init_rsp = (unsigned long)temp_stack + 4096; +#ifdef CONFIG_SMP + stack_start.sp = temp_stack + 4096; +#endif initial_code = (unsigned long)wakeup_long64; saved_magic = 0x123456789abcdef0; #endif /* CONFIG_64BIT */ diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c new file mode 100644 index 00000000000..f2766d84c7a --- /dev/null +++ b/arch/x86/kernel/amd_iommu.c @@ -0,0 +1,962 @@ +/* + * Copyright (C) 2007-2008 Advanced Micro Devices, Inc. + * Author: Joerg Roedel <joerg.roedel@amd.com> + * Leo Duran <leo.duran@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. + * + * 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 + */ + +#include <linux/pci.h> +#include <linux/gfp.h> +#include <linux/bitops.h> +#include <linux/scatterlist.h> +#include <linux/iommu-helper.h> +#include <asm/proto.h> +#include <asm/gart.h> +#include <asm/amd_iommu_types.h> +#include <asm/amd_iommu.h> + +#define CMD_SET_TYPE(cmd, t) ((cmd)->data[1] |= ((t) << 28)) + +#define to_pages(addr, size) \ + (round_up(((addr) & ~PAGE_MASK) + (size), PAGE_SIZE) >> PAGE_SHIFT) + +static DEFINE_RWLOCK(amd_iommu_devtable_lock); + +struct command { + u32 data[4]; +}; + +static int dma_ops_unity_map(struct dma_ops_domain *dma_dom, + struct unity_map_entry *e); + +static int iommu_has_npcache(struct amd_iommu *iommu) +{ + return iommu->cap & IOMMU_CAP_NPCACHE; +} + +static int __iommu_queue_command(struct amd_iommu *iommu, struct command *cmd) +{ + u32 tail, head; + u8 *target; + + tail = readl(iommu->mmio_base + MMIO_CMD_TAIL_OFFSET); + target = (iommu->cmd_buf + tail); + memcpy_toio(target, cmd, sizeof(*cmd)); + tail = (tail + sizeof(*cmd)) % iommu->cmd_buf_size; + head = readl(iommu->mmio_base + MMIO_CMD_HEAD_OFFSET); + if (tail == head) + return -ENOMEM; + writel(tail, iommu->mmio_base + MMIO_CMD_TAIL_OFFSET); + + return 0; +} + +static int iommu_queue_command(struct amd_iommu *iommu, struct command *cmd) +{ + unsigned long flags; + int ret; + + spin_lock_irqsave(&iommu->lock, flags); + ret = __iommu_queue_command(iommu, cmd); + spin_unlock_irqrestore(&iommu->lock, flags); + + return ret; +} + +static int iommu_completion_wait(struct amd_iommu *iommu) +{ + int ret; + struct command cmd; + volatile u64 ready = 0; + unsigned long ready_phys = virt_to_phys(&ready); + + memset(&cmd, 0, sizeof(cmd)); + cmd.data[0] = LOW_U32(ready_phys) | CMD_COMPL_WAIT_STORE_MASK; + cmd.data[1] = HIGH_U32(ready_phys); + cmd.data[2] = 1; /* value written to 'ready' */ + CMD_SET_TYPE(&cmd, CMD_COMPL_WAIT); + + iommu->need_sync = 0; + + ret = iommu_queue_command(iommu, &cmd); + + if (ret) + return ret; + + while (!ready) + cpu_relax(); + + return 0; +} + +static int iommu_queue_inv_dev_entry(struct amd_iommu *iommu, u16 devid) +{ + struct command cmd; + + BUG_ON(iommu == NULL); + + memset(&cmd, 0, sizeof(cmd)); + CMD_SET_TYPE(&cmd, CMD_INV_DEV_ENTRY); + cmd.data[0] = devid; + + iommu->need_sync = 1; + + return iommu_queue_command(iommu, &cmd); +} + +static int iommu_queue_inv_iommu_pages(struct amd_iommu *iommu, + u64 address, u16 domid, int pde, int s) +{ + struct command cmd; + + memset(&cmd, 0, sizeof(cmd)); + address &= PAGE_MASK; + CMD_SET_TYPE(&cmd, CMD_INV_IOMMU_PAGES); + cmd.data[1] |= domid; + cmd.data[2] = LOW_U32(address); + cmd.data[3] = HIGH_U32(address); + if (s) + cmd.data[2] |= CMD_INV_IOMMU_PAGES_SIZE_MASK; + if (pde) + cmd.data[2] |= CMD_INV_IOMMU_PAGES_PDE_MASK; + + iommu->need_sync = 1; + + return iommu_queue_command(iommu, &cmd); +} + +static int iommu_flush_pages(struct amd_iommu *iommu, u16 domid, + u64 address, size_t size) +{ + int s = 0; + unsigned pages = to_pages(address, size); + + address &= PAGE_MASK; + + if (pages > 1) { + /* + * If we have to flush more than one page, flush all + * TLB entries for this domain + */ + address = CMD_INV_IOMMU_ALL_PAGES_ADDRESS; + s = 1; + } + + iommu_queue_inv_iommu_pages(iommu, address, domid, 0, s); + + return 0; +} + +static int iommu_map(struct protection_domain *dom, + unsigned long bus_addr, + unsigned long phys_addr, + int prot) +{ + u64 __pte, *pte, *page; + + bus_addr = PAGE_ALIGN(bus_addr); + phys_addr = PAGE_ALIGN(bus_addr); + + /* only support 512GB address spaces for now */ + if (bus_addr > IOMMU_MAP_SIZE_L3 || !(prot & IOMMU_PROT_MASK)) + return -EINVAL; + + pte = &dom->pt_root[IOMMU_PTE_L2_INDEX(bus_addr)]; + + if (!IOMMU_PTE_PRESENT(*pte)) { + page = (u64 *)get_zeroed_page(GFP_KERNEL); + if (!page) + return -ENOMEM; + *pte = IOMMU_L2_PDE(virt_to_phys(page)); + } + + pte = IOMMU_PTE_PAGE(*pte); + pte = &pte[IOMMU_PTE_L1_INDEX(bus_addr)]; + + if (!IOMMU_PTE_PRESENT(*pte)) { + page = (u64 *)get_zeroed_page(GFP_KERNEL); + if (!page) + return -ENOMEM; + *pte = IOMMU_L1_PDE(virt_to_phys(page)); + } + + pte = IOMMU_PTE_PAGE(*pte); + pte = &pte[IOMMU_PTE_L0_INDEX(bus_addr)]; + + if (IOMMU_PTE_PRESENT(*pte)) + return -EBUSY; + + __pte = phys_addr | IOMMU_PTE_P; + if (prot & IOMMU_PROT_IR) + __pte |= IOMMU_PTE_IR; + if (prot & IOMMU_PROT_IW) + __pte |= IOMMU_PTE_IW; + + *pte = __pte; + + return 0; +} + +static int iommu_for_unity_map(struct amd_iommu *iommu, + struct unity_map_entry *entry) +{ + u16 bdf, i; + + for (i = entry->devid_start; i <= entry->devid_end; ++i) { + bdf = amd_iommu_alias_table[i]; + if (amd_iommu_rlookup_table[bdf] == iommu) + return 1; + } + + return 0; +} + +static int iommu_init_unity_mappings(struct amd_iommu *iommu) +{ + struct unity_map_entry *entry; + int ret; + + list_for_each_entry(entry, &amd_iommu_unity_map, list) { + if (!iommu_for_unity_map(iommu, entry)) + continue; + ret = dma_ops_unity_map(iommu->default_dom, entry); + if (ret) + return ret; + } + + return 0; +} + +static int dma_ops_unity_map(struct dma_ops_domain *dma_dom, + struct unity_map_entry *e) +{ + u64 addr; + int ret; + + for (addr = e->address_start; addr < e->address_end; + addr += PAGE_SIZE) { + ret = iommu_map(&dma_dom->domain, addr, addr, e->prot); + if (ret) + return ret; + /* + * if unity mapping is in aperture range mark the page + * as allocated in the aperture + */ + if (addr < dma_dom->aperture_size) + __set_bit(addr >> PAGE_SHIFT, dma_dom->bitmap); + } + + return 0; +} + +static int init_unity_mappings_for_device(struct dma_ops_domain *dma_dom, + u16 devid) +{ + struct unity_map_entry *e; + int ret; + + list_for_each_entry(e, &amd_iommu_unity_map, list) { + if (!(devid >= e->devid_start && devid <= e->devid_end)) + continue; + ret = dma_ops_unity_map(dma_dom, e); + if (ret) + return ret; + } + + return 0; +} + +static unsigned long dma_mask_to_pages(unsigned long mask) +{ + return (mask >> PAGE_SHIFT) + + (PAGE_ALIGN(mask & ~PAGE_MASK) >> PAGE_SHIFT); +} + +static unsigned long dma_ops_alloc_addresses(struct device *dev, + struct dma_ops_domain *dom, + unsigned int pages) +{ + unsigned long limit = dma_mask_to_pages(*dev->dma_mask); + unsigned long address; + unsigned long size = dom->aperture_size >> PAGE_SHIFT; + unsigned long boundary_size; + + boundary_size = ALIGN(dma_get_seg_boundary(dev) + 1, + PAGE_SIZE) >> PAGE_SHIFT; + limit = limit < size ? limit : size; + + if (dom->next_bit >= limit) + dom->next_bit = 0; + + address = iommu_area_alloc(dom->bitmap, limit, dom->next_bit, pages, + 0 , boundary_size, 0); + if (address == -1) + address = iommu_area_alloc(dom->bitmap, limit, 0, pages, + 0, boundary_size, 0); + + if (likely(address != -1)) { + dom->next_bit = address + pages; + address <<= PAGE_SHIFT; + } else + address = bad_dma_address; + + WARN_ON((address + (PAGE_SIZE*pages)) > dom->aperture_size); + + return address; +} + +static void dma_ops_free_addresses(struct dma_ops_domain *dom, + unsigned long address, + unsigned int pages) +{ + address >>= PAGE_SHIFT; + iommu_area_free(dom->bitmap, address, pages); +} + +static u16 domain_id_alloc(void) +{ + unsigned long flags; + int id; + + write_lock_irqsave(&amd_iommu_devtable_lock, flags); + id = find_first_zero_bit(amd_iommu_pd_alloc_bitmap, MAX_DOMAIN_ID); + BUG_ON(id == 0); + if (id > 0 && id < MAX_DOMAIN_ID) + __set_bit(id, amd_iommu_pd_alloc_bitmap); + else + id = 0; + write_unlock_irqrestore(&amd_iommu_devtable_lock, flags); + + return id; +} + +static void dma_ops_reserve_addresses(struct dma_ops_domain *dom, + unsigned long start_page, + unsigned int pages) +{ + unsigned int last_page = dom->aperture_size >> PAGE_SHIFT; + + if (start_page + pages > last_page) + pages = last_page - start_page; + + set_bit_string(dom->bitmap, start_page, pages); +} + +static void dma_ops_free_pagetable(struct dma_ops_domain *dma_dom) +{ + int i, j; + u64 *p1, *p2, *p3; + + p1 = dma_dom->domain.pt_root; + + if (!p1) + return; + + for (i = 0; i < 512; ++i) { + if (!IOMMU_PTE_PRESENT(p1[i])) + continue; + + p2 = IOMMU_PTE_PAGE(p1[i]); + for (j = 0; j < 512; ++i) { + if (!IOMMU_PTE_PRESENT(p2[j])) + continue; + p3 = IOMMU_PTE_PAGE(p2[j]); + free_page((unsigned long)p3); + } + + free_page((unsigned long)p2); + } + + free_page((unsigned long)p1); +} + +static void dma_ops_domain_free(struct dma_ops_domain *dom) +{ + if (!dom) + return; + + dma_ops_free_pagetable(dom); + + kfree(dom->pte_pages); + + kfree(dom->bitmap); + + kfree(dom); +} + +static struct dma_ops_domain *dma_ops_domain_alloc(struct amd_iommu *iommu, + unsigned order) +{ + struct dma_ops_domain *dma_dom; + unsigned i, num_pte_pages; + u64 *l2_pde; + u64 address; + + /* + * Currently the DMA aperture must be between 32 MB and 1GB in size + */ + if ((order < 25) || (order > 30)) + return NULL; + + dma_dom = kzalloc(sizeof(struct dma_ops_domain), GFP_KERNEL); + if (!dma_dom) + return NULL; + + spin_lock_init(&dma_dom->domain.lock); + + dma_dom->domain.id = domain_id_alloc(); + if (dma_dom->domain.id == 0) + goto free_dma_dom; + dma_dom->domain.mode = PAGE_MODE_3_LEVEL; + dma_dom->domain.pt_root = (void *)get_zeroed_page(GFP_KERNEL); + dma_dom->domain.priv = dma_dom; + if (!dma_dom->domain.pt_root) + goto free_dma_dom; + dma_dom->aperture_size = (1ULL << order); + dma_dom->bitmap = kzalloc(dma_dom->aperture_size / (PAGE_SIZE * 8), + GFP_KERNEL); + if (!dma_dom->bitmap) + goto free_dma_dom; + /* + * mark the first page as allocated so we never return 0 as + * a valid dma-address. So we can use 0 as error value + */ + dma_dom->bitmap[0] = 1; + dma_dom->next_bit = 0; + + if (iommu->exclusion_start && + iommu->exclusion_start < dma_dom->aperture_size) { + unsigned long startpage = iommu->exclusion_start >> PAGE_SHIFT; + int pages = to_pages(iommu->exclusion_start, + iommu->exclusion_length); + dma_ops_reserve_addresses(dma_dom, startpage, pages); + } + + num_pte_pages = dma_dom->aperture_size / (PAGE_SIZE * 512); + dma_dom->pte_pages = kzalloc(num_pte_pages * sizeof(void *), + GFP_KERNEL); + if (!dma_dom->pte_pages) + goto free_dma_dom; + + l2_pde = (u64 *)get_zeroed_page(GFP_KERNEL); + if (l2_pde == NULL) + goto free_dma_dom; + + dma_dom->domain.pt_root[0] = IOMMU_L2_PDE(virt_to_phys(l2_pde)); + + for (i = 0; i < num_pte_pages; ++i) { + dma_dom->pte_pages[i] = (u64 *)get_zeroed_page(GFP_KERNEL); + if (!dma_dom->pte_pages[i]) + goto free_dma_dom; + address = virt_to_phys(dma_dom->pte_pages[i]); + l2_pde[i] = IOMMU_L1_PDE(address); + } + + return dma_dom; + +free_dma_dom: + dma_ops_domain_free(dma_dom); + + return NULL; +} + +static struct protection_domain *domain_for_device(u16 devid) +{ + struct protection_domain *dom; + unsigned long flags; + + read_lock_irqsave(&amd_iommu_devtable_lock, flags); + dom = amd_iommu_pd_table[devid]; + read_unlock_irqrestore(&amd_iommu_devtable_lock, flags); + + return dom; +} + +static void set_device_domain(struct amd_iommu *iommu, + struct protection_domain *domain, + u16 devid) +{ + unsigned long flags; + + u64 pte_root = virt_to_phys(domain->pt_root); + + pte_root |= (domain->mode & 0x07) << 9; + pte_root |= IOMMU_PTE_IR | IOMMU_PTE_IW | IOMMU_PTE_P | 2; + + write_lock_irqsave(&amd_iommu_devtable_lock, flags); + amd_iommu_dev_table[devid].data[0] = pte_root; + amd_iommu_dev_table[devid].data[1] = pte_root >> 32; + amd_iommu_dev_table[devid].data[2] = domain->id; + + amd_iommu_pd_table[devid] = domain; + write_unlock_irqrestore(&amd_iommu_devtable_lock, flags); + + iommu_queue_inv_dev_entry(iommu, devid); + + iommu->need_sync = 1; +} + +static int get_device_resources(struct device *dev, + struct amd_iommu **iommu, + struct protection_domain **domain, + u16 *bdf) +{ + struct dma_ops_domain *dma_dom; + struct pci_dev *pcidev; + u16 _bdf; + + BUG_ON(!dev || dev->bus != &pci_bus_type || !dev->dma_mask); + + pcidev = to_pci_dev(dev); + _bdf = (pcidev->bus->number << 8) | pcidev->devfn; + + if (_bdf >= amd_iommu_last_bdf) { + *iommu = NULL; + *domain = NULL; + *bdf = 0xffff; + return 0; + } + + *bdf = amd_iommu_alias_table[_bdf]; + + *iommu = amd_iommu_rlookup_table[*bdf]; + if (*iommu == NULL) + return 0; + dma_dom = (*iommu)->default_dom; + *domain = domain_for_device(*bdf); + if (*domain == NULL) { + *domain = &dma_dom->domain; + set_device_domain(*iommu, *domain, *bdf); + printk(KERN_INFO "AMD IOMMU: Using protection domain %d for " + "device ", (*domain)->id); + print_devid(_bdf, 1); + } + + return 1; +} + +static dma_addr_t dma_ops_domain_map(struct amd_iommu *iommu, + struct dma_ops_domain *dom, + unsigned long address, + phys_addr_t paddr, + int direction) +{ + u64 *pte, __pte; + + WARN_ON(address > dom->aperture_size); + + paddr &= PAGE_MASK; + + pte = dom->pte_pages[IOMMU_PTE_L1_INDEX(address)]; + pte += IOMMU_PTE_L0_INDEX(address); + + __pte = paddr | IOMMU_PTE_P | IOMMU_PTE_FC; + + if (direction == DMA_TO_DEVICE) + __pte |= IOMMU_PTE_IR; + else if (direction == DMA_FROM_DEVICE) + __pte |= IOMMU_PTE_IW; + else if (direction == DMA_BIDIRECTIONAL) + __pte |= IOMMU_PTE_IR | IOMMU_PTE_IW; + + WARN_ON(*pte); + + *pte = __pte; + + return (dma_addr_t)address; +} + +static void dma_ops_domain_unmap(struct amd_iommu *iommu, + struct dma_ops_domain *dom, + unsigned long address) +{ + u64 *pte; + + if (address >= dom->aperture_size) + return; + + WARN_ON(address & 0xfffULL || address > dom->aperture_size); + + pte = dom->pte_pages[IOMMU_PTE_L1_INDEX(address)]; + pte += IOMMU_PTE_L0_INDEX(address); + + WARN_ON(!*pte); + + *pte = 0ULL; +} + +static dma_addr_t __map_single(struct device *dev, + struct amd_iommu *iommu, + struct dma_ops_domain *dma_dom, + phys_addr_t paddr, + size_t size, + int dir) +{ + dma_addr_t offset = paddr & ~PAGE_MASK; + dma_addr_t address, start; + unsigned int pages; + int i; + + pages = to_pages(paddr, size); + paddr &= PAGE_MASK; + + address = dma_ops_alloc_addresses(dev, dma_dom, pages); + if (unlikely(address == bad_dma_address)) + goto out; + + start = address; + for (i = 0; i < pages; ++i) { + dma_ops_domain_map(iommu, dma_dom, start, paddr, dir); + paddr += PAGE_SIZE; + start += PAGE_SIZE; + } + address += offset; + +out: + return address; +} + +static void __unmap_single(struct amd_iommu *iommu, + struct dma_ops_domain *dma_dom, + dma_addr_t dma_addr, + size_t size, + int dir) +{ + dma_addr_t i, start; + unsigned int pages; + + if ((dma_addr == 0) || (dma_addr + size > dma_dom->aperture_size)) + return; + + pages = to_pages(dma_addr, size); + dma_addr &= PAGE_MASK; + start = dma_addr; + + for (i = 0; i < pages; ++i) { + dma_ops_domain_unmap(iommu, dma_dom, start); + start += PAGE_SIZE; + } + + dma_ops_free_addresses(dma_dom, dma_addr, pages); +} + +static dma_addr_t map_single(struct device *dev, phys_addr_t paddr, + size_t size, int dir) +{ + unsigned long flags; + struct amd_iommu *iommu; + struct protection_domain *domain; + u16 devid; + dma_addr_t addr; + + get_device_resources(dev, &iommu, &domain, &devid); + + if (iommu == NULL || domain == NULL) + return (dma_addr_t)paddr; + + spin_lock_irqsave(&domain->lock, flags); + addr = __map_single(dev, iommu, domain->priv, paddr, size, dir); + if (addr == bad_dma_address) + goto out; + + if (iommu_has_npcache(iommu)) + iommu_flush_pages(iommu, domain->id, addr, size); + + if (iommu->need_sync) + iommu_completion_wait(iommu); + +out: + spin_unlock_irqrestore(&domain->lock, flags); + + return addr; +} + +static void unmap_single(struct device *dev, dma_addr_t dma_addr, + size_t size, int dir) +{ + unsigned long flags; + struct amd_iommu *iommu; + struct protection_domain *domain; + u16 devid; + + if (!get_device_resources(dev, &iommu, &domain, &devid)) + return; + + spin_lock_irqsave(&domain->lock, flags); + + __unmap_single(iommu, domain->priv, dma_addr, size, dir); + + iommu_flush_pages(iommu, domain->id, dma_addr, size); + + if (iommu->need_sync) + iommu_completion_wait(iommu); + + spin_unlock_irqrestore(&domain->lock, flags); +} + +static int map_sg_no_iommu(struct device *dev, struct scatterlist *sglist, + int nelems, int dir) +{ + struct scatterlist *s; + int i; + + for_each_sg(sglist, s, nelems, i) { + s->dma_address = (dma_addr_t)sg_phys(s); + s->dma_length = s->length; + } + + return nelems; +} + +static int map_sg(struct device *dev, struct scatterlist *sglist, + int nelems, int dir) +{ + unsigned long flags; + struct amd_iommu *iommu; + struct protection_domain *domain; + u16 devid; + int i; + struct scatterlist *s; + phys_addr_t paddr; + int mapped_elems = 0; + + get_device_resources(dev, &iommu, &domain, &devid); + + if (!iommu || !domain) + return map_sg_no_iommu(dev, sglist, nelems, dir); + + spin_lock_irqsave(&domain->lock, flags); + + for_each_sg(sglist, s, nelems, i) { + paddr = sg_phys(s); + + s->dma_address = __map_single(dev, iommu, domain->priv, + paddr, s->length, dir); + + if (s->dma_address) { + s->dma_length = s->length; + mapped_elems++; + } else + goto unmap; + if (iommu_has_npcache(iommu)) + iommu_flush_pages(iommu, domain->id, s->dma_address, + s->dma_length); + } + + if (iommu->need_sync) + iommu_completion_wait(iommu); + +out: + spin_unlock_irqrestore(&domain->lock, flags); + + return mapped_elems; +unmap: + for_each_sg(sglist, s, mapped_elems, i) { + if (s->dma_address) + __unmap_single(iommu, domain->priv, s->dma_address, + s->dma_length, dir); + s->dma_address = s->dma_length = 0; + } + + mapped_elems = 0; + + goto out; +} + +static void unmap_sg(struct device *dev, struct scatterlist *sglist, + int nelems, int dir) +{ + unsigned long flags; + struct amd_iommu *iommu; + struct protection_domain *domain; + struct scatterlist *s; + u16 devid; + int i; + + if (!get_device_resources(dev, &iommu, &domain, &devid)) + return; + + spin_lock_irqsave(&domain->lock, flags); + + for_each_sg(sglist, s, nelems, i) { + __unmap_single(iommu, domain->priv, s->dma_address, + s->dma_length, dir); + iommu_flush_pages(iommu, domain->id, s->dma_address, + s->dma_length); + s->dma_address = s->dma_length = 0; + } + + if (iommu->need_sync) + iommu_completion_wait(iommu); + + spin_unlock_irqrestore(&domain->lock, flags); +} + +static void *alloc_coherent(struct device *dev, size_t size, + dma_addr_t *dma_addr, gfp_t flag) +{ + unsigned long flags; + void *virt_addr; + struct amd_iommu *iommu; + struct protection_domain *domain; + u16 devid; + phys_addr_t paddr; + + virt_addr = (void *)__get_free_pages(flag, get_order(size)); + if (!virt_addr) + return 0; + + memset(virt_addr, 0, size); + paddr = virt_to_phys(virt_addr); + + get_device_resources(dev, &iommu, &domain, &devid); + + if (!iommu || !domain) { + *dma_addr = (dma_addr_t)paddr; + return virt_addr; + } + + spin_lock_irqsave(&domain->lock, flags); + + *dma_addr = __map_single(dev, iommu, domain->priv, paddr, + size, DMA_BIDIRECTIONAL); + + if (*dma_addr == bad_dma_address) { + free_pages((unsigned long)virt_addr, get_order(size)); + virt_addr = NULL; + goto out; + } + + if (iommu_has_npcache(iommu)) + iommu_flush_pages(iommu, domain->id, *dma_addr, size); + + if (iommu->need_sync) + iommu_completion_wait(iommu); + +out: + spin_unlock_irqrestore(&domain->lock, flags); + + return virt_addr; +} + +static void free_coherent(struct device *dev, size_t size, + void *virt_addr, dma_addr_t dma_addr) +{ + unsigned long flags; + struct amd_iommu *iommu; + struct protection_domain *domain; + u16 devid; + + get_device_resources(dev, &iommu, &domain, &devid); + + if (!iommu || !domain) + goto free_mem; + + spin_lock_irqsave(&domain->lock, flags); + + __unmap_single(iommu, domain->priv, dma_addr, size, DMA_BIDIRECTIONAL); + iommu_flush_pages(iommu, domain->id, dma_addr, size); + + if (iommu->need_sync) + iommu_completion_wait(iommu); + + spin_unlock_irqrestore(&domain->lock, flags); + +free_mem: + free_pages((unsigned long)virt_addr, get_order(size)); +} + +/* + * If the driver core informs the DMA layer if a driver grabs a device + * we don't need to preallocate the protection domains anymore. + * For now we have to. + */ +void prealloc_protection_domains(void) +{ + struct pci_dev *dev = NULL; + struct dma_ops_domain *dma_dom; + struct amd_iommu *iommu; + int order = amd_iommu_aperture_order; + u16 devid; + + while ((dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) { + devid = (dev->bus->number << 8) | dev->devfn; + if (devid >= amd_iommu_last_bdf) + continue; + devid = amd_iommu_alias_table[devid]; + if (domain_for_device(devid)) + continue; + iommu = amd_iommu_rlookup_table[devid]; + if (!iommu) + continue; + dma_dom = dma_ops_domain_alloc(iommu, order); + if (!dma_dom) + continue; + init_unity_mappings_for_device(dma_dom, devid); + set_device_domain(iommu, &dma_dom->domain, devid); + printk(KERN_INFO "AMD IOMMU: Allocated domain %d for device ", + dma_dom->domain.id); + print_devid(devid, 1); + } +} + +static struct dma_mapping_ops amd_iommu_dma_ops = { + .alloc_coherent = alloc_coherent, + .free_coherent = free_coherent, + .map_single = map_single, + .unmap_single = unmap_single, + .map_sg = map_sg, + .unmap_sg = unmap_sg, +}; + +int __init amd_iommu_init_dma_ops(void) +{ + struct amd_iommu *iommu; + int order = amd_iommu_aperture_order; + int ret; + + list_for_each_entry(iommu, &amd_iommu_list, list) { + iommu->default_dom = dma_ops_domain_alloc(iommu, order); + if (iommu->default_dom == NULL) + return -ENOMEM; + ret = iommu_init_unity_mappings(iommu); + if (ret) + goto free_domains; + } + + if (amd_iommu_isolate) + prealloc_protection_domains(); + + iommu_detected = 1; + force_iommu = 1; + bad_dma_address = 0; +#ifdef CONFIG_GART_IOMMU + gart_iommu_aperture_disabled = 1; + gart_iommu_aperture = 0; +#endif + + dma_ops = &amd_iommu_dma_ops; + + return 0; + +free_domains: + + list_for_each_entry(iommu, &amd_iommu_list, list) { + if (iommu->default_dom) + dma_ops_domain_free(iommu->default_dom); + } + + return ret; +} diff --git a/arch/x86/kernel/amd_iommu_init.c b/arch/x86/kernel/amd_iommu_init.c new file mode 100644 index 00000000000..2a13e430437 --- /dev/null +++ b/arch/x86/kernel/amd_iommu_init.c @@ -0,0 +1,875 @@ +/* + * Copyright (C) 2007-2008 Advanced Micro Devices, Inc. + * Author: Joerg Roedel <joerg.roedel@amd.com> + * Leo Duran <leo.duran@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. + * + * 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 + */ + +#include <linux/pci.h> +#include <linux/acpi.h> +#include <linux/gfp.h> +#include <linux/list.h> +#include <linux/sysdev.h> +#include <asm/pci-direct.h> +#include <asm/amd_iommu_types.h> +#include <asm/amd_iommu.h> +#include <asm/gart.h> + +/* + * definitions for the ACPI scanning code + */ +#define UPDATE_LAST_BDF(x) do {\ + if ((x) > amd_iommu_last_bdf) \ + amd_iommu_last_bdf = (x); \ + } while (0); + +#define DEVID(bus, devfn) (((bus) << 8) | (devfn)) +#define PCI_BUS(x) (((x) >> 8) & 0xff) +#define IVRS_HEADER_LENGTH 48 +#define TBL_SIZE(x) (1 << (PAGE_SHIFT + get_order(amd_iommu_last_bdf * (x)))) + +#define ACPI_IVHD_TYPE 0x10 +#define ACPI_IVMD_TYPE_ALL 0x20 +#define ACPI_IVMD_TYPE 0x21 +#define ACPI_IVMD_TYPE_RANGE 0x22 + +#define IVHD_DEV_ALL 0x01 +#define IVHD_DEV_SELECT 0x02 +#define IVHD_DEV_SELECT_RANGE_START 0x03 +#define IVHD_DEV_RANGE_END 0x04 +#define IVHD_DEV_ALIAS 0x42 +#define IVHD_DEV_ALIAS_RANGE 0x43 +#define IVHD_DEV_EXT_SELECT 0x46 +#define IVHD_DEV_EXT_SELECT_RANGE 0x47 + +#define IVHD_FLAG_HT_TUN_EN 0x00 +#define IVHD_FLAG_PASSPW_EN 0x01 +#define IVHD_FLAG_RESPASSPW_EN 0x02 +#define IVHD_FLAG_ISOC_EN 0x03 + +#define IVMD_FLAG_EXCL_RANGE 0x08 +#define IVMD_FLAG_UNITY_MAP 0x01 + +#define ACPI_DEVFLAG_INITPASS 0x01 +#define ACPI_DEVFLAG_EXTINT 0x02 +#define ACPI_DEVFLAG_NMI 0x04 +#define ACPI_DEVFLAG_SYSMGT1 0x10 +#define ACPI_DEVFLAG_SYSMGT2 0x20 +#define ACPI_DEVFLAG_LINT0 0x40 +#define ACPI_DEVFLAG_LINT1 0x80 +#define ACPI_DEVFLAG_ATSDIS 0x10000000 + +struct ivhd_header { + u8 type; + u8 flags; + u16 length; + u16 devid; + u16 cap_ptr; + u64 mmio_phys; + u16 pci_seg; + u16 info; + u32 reserved; +} __attribute__((packed)); + +struct ivhd_entry { + u8 type; + u16 devid; + u8 flags; + u32 ext; +} __attribute__((packed)); + +struct ivmd_header { + u8 type; + u8 flags; + u16 length; + u16 devid; + u16 aux; + u64 resv; + u64 range_start; + u64 range_length; +} __attribute__((packed)); + +static int __initdata amd_iommu_detected; + +u16 amd_iommu_last_bdf; +struct list_head amd_iommu_unity_map; +unsigned amd_iommu_aperture_order = 26; +int amd_iommu_isolate; + +struct list_head amd_iommu_list; +struct dev_table_entry *amd_iommu_dev_table; +u16 *amd_iommu_alias_table; +struct amd_iommu **amd_iommu_rlookup_table; +struct protection_domain **amd_iommu_pd_table; +unsigned long *amd_iommu_pd_alloc_bitmap; + +static u32 dev_table_size; +static u32 alias_table_size; +static u32 rlookup_table_size; + +static void __init iommu_set_exclusion_range(struct amd_iommu *iommu) +{ + u64 start = iommu->exclusion_start & PAGE_MASK; + u64 limit = (start + iommu->exclusion_length) & PAGE_MASK; + u64 entry; + + if (!iommu->exclusion_start) + return; + + entry = start | MMIO_EXCL_ENABLE_MASK; + memcpy_toio(iommu->mmio_base + MMIO_EXCL_BASE_OFFSET, + &entry, sizeof(entry)); + + entry = limit; + memcpy_toio(iommu->mmio_base + MMIO_EXCL_LIMIT_OFFSET, + &entry, sizeof(entry)); +} + +static void __init iommu_set_device_table(struct amd_iommu *iommu) +{ + u32 entry; + + BUG_ON(iommu->mmio_base == NULL); + + entry = virt_to_phys(amd_iommu_dev_table); + entry |= (dev_table_size >> 12) - 1; + memcpy_toio(iommu->mmio_base + MMIO_DEV_TABLE_OFFSET, + &entry, sizeof(entry)); +} + +static void __init iommu_feature_enable(struct amd_iommu *iommu, u8 bit) +{ + u32 ctrl; + + ctrl = readl(iommu->mmio_base + MMIO_CONTROL_OFFSET); + ctrl |= (1 << bit); + writel(ctrl, iommu->mmio_base + MMIO_CONTROL_OFFSET); +} + +static void __init iommu_feature_disable(struct amd_iommu *iommu, u8 bit) +{ + u32 ctrl; + + ctrl = (u64)readl(iommu->mmio_base + MMIO_CONTROL_OFFSET); + ctrl &= ~(1 << bit); + writel(ctrl, iommu->mmio_base + MMIO_CONTROL_OFFSET); +} + +void __init iommu_enable(struct amd_iommu *iommu) +{ + printk(KERN_INFO "AMD IOMMU: Enabling IOMMU at "); + print_devid(iommu->devid, 0); + printk(" cap 0x%hx\n", iommu->cap_ptr); + + iommu_feature_enable(iommu, CONTROL_IOMMU_EN); +} + +static u8 * __init iommu_map_mmio_space(u64 address) +{ + u8 *ret; + + if (!request_mem_region(address, MMIO_REGION_LENGTH, "amd_iommu")) + return NULL; + + ret = ioremap_nocache(address, MMIO_REGION_LENGTH); + if (ret != NULL) + return ret; + + release_mem_region(address, MMIO_REGION_LENGTH); + + return NULL; +} + +static void __init iommu_unmap_mmio_space(struct amd_iommu *iommu) +{ + if (iommu->mmio_base) + iounmap(iommu->mmio_base); + release_mem_region(iommu->mmio_phys, MMIO_REGION_LENGTH); +} + +static int __init find_last_devid_on_pci(int bus, int dev, int fn, int cap_ptr) +{ + u32 cap; + + cap = read_pci_config(bus, dev, fn, cap_ptr+MMIO_RANGE_OFFSET); + UPDATE_LAST_BDF(DEVID(MMIO_GET_BUS(cap), MMIO_GET_LD(cap))); + + return 0; +} + +static int __init find_last_devid_from_ivhd(struct ivhd_header *h) +{ + u8 *p = (void *)h, *end = (void *)h; + struct ivhd_entry *dev; + + p += sizeof(*h); + end += h->length; + + find_last_devid_on_pci(PCI_BUS(h->devid), + PCI_SLOT(h->devid), + PCI_FUNC(h->devid), + h->cap_ptr); + + while (p < end) { + dev = (struct ivhd_entry *)p; + switch (dev->type) { + case IVHD_DEV_SELECT: + case IVHD_DEV_RANGE_END: + case IVHD_DEV_ALIAS: + case IVHD_DEV_EXT_SELECT: + UPDATE_LAST_BDF(dev->devid); + break; + default: + break; + } + p += 0x04 << (*p >> 6); + } + + WARN_ON(p != end); + + return 0; +} + +static int __init find_last_devid_acpi(struct acpi_table_header *table) +{ + int i; + u8 checksum = 0, *p = (u8 *)table, *end = (u8 *)table; + struct ivhd_header *h; + + /* + * Validate checksum here so we don't need to do it when + * we actually parse the table + */ + for (i = 0; i < table->length; ++i) + checksum += p[i]; + if (checksum != 0) + /* ACPI table corrupt */ + return -ENODEV; + + p += IVRS_HEADER_LENGTH; + + end += table->length; + while (p < end) { + h = (struct ivhd_header *)p; + switch (h->type) { + case ACPI_IVHD_TYPE: + find_last_devid_from_ivhd(h); + break; + default: + break; + } + p += h->length; + } + WARN_ON(p != end); + + return 0; +} + +static u8 * __init alloc_command_buffer(struct amd_iommu *iommu) +{ + u8 *cmd_buf = (u8 *)__get_free_pages(GFP_KERNEL, + get_order(CMD_BUFFER_SIZE)); + u64 entry = 0; + + if (cmd_buf == NULL) + return NULL; + + iommu->cmd_buf_size = CMD_BUFFER_SIZE; + + memset(cmd_buf, 0, CMD_BUFFER_SIZE); + + entry = (u64)virt_to_phys(cmd_buf); + entry |= MMIO_CMD_SIZE_512; + memcpy_toio(iommu->mmio_base + MMIO_CMD_BUF_OFFSET, + &entry, sizeof(entry)); + + iommu_feature_enable(iommu, CONTROL_CMDBUF_EN); + + return cmd_buf; +} + +static void __init free_command_buffer(struct amd_iommu *iommu) +{ + if (iommu->cmd_buf) + free_pages((unsigned long)iommu->cmd_buf, + get_order(CMD_BUFFER_SIZE)); +} + +static void set_dev_entry_bit(u16 devid, u8 bit) +{ + int i = (bit >> 5) & 0x07; + int _bit = bit & 0x1f; + + amd_iommu_dev_table[devid].data[i] |= (1 << _bit); +} + +static void __init set_dev_entry_from_acpi(u16 devid, u32 flags, u32 ext_flags) +{ + if (flags & ACPI_DEVFLAG_INITPASS) + set_dev_entry_bit(devid, DEV_ENTRY_INIT_PASS); + if (flags & ACPI_DEVFLAG_EXTINT) + set_dev_entry_bit(devid, DEV_ENTRY_EINT_PASS); + if (flags & ACPI_DEVFLAG_NMI) + set_dev_entry_bit(devid, DEV_ENTRY_NMI_PASS); + if (flags & ACPI_DEVFLAG_SYSMGT1) + set_dev_entry_bit(devid, DEV_ENTRY_SYSMGT1); + if (flags & ACPI_DEVFLAG_SYSMGT2) + set_dev_entry_bit(devid, DEV_ENTRY_SYSMGT2); + if (flags & ACPI_DEVFLAG_LINT0) + set_dev_entry_bit(devid, DEV_ENTRY_LINT0_PASS); + if (flags & ACPI_DEVFLAG_LINT1) + set_dev_entry_bit(devid, DEV_ENTRY_LINT1_PASS); +} + +static void __init set_iommu_for_device(struct amd_iommu *iommu, u16 devid) +{ + amd_iommu_rlookup_table[devid] = iommu; +} + +static void __init set_device_exclusion_range(u16 devid, struct ivmd_header *m) +{ + struct amd_iommu *iommu = amd_iommu_rlookup_table[devid]; + + if (!(m->flags & IVMD_FLAG_EXCL_RANGE)) + return; + + if (iommu) { + set_dev_entry_bit(m->devid, DEV_ENTRY_EX); + iommu->exclusion_start = m->range_start; + iommu->exclusion_length = m->range_length; + } +} + +static void __init init_iommu_from_pci(struct amd_iommu *iommu) +{ + int bus = PCI_BUS(iommu->devid); + int dev = PCI_SLOT(iommu->devid); + int fn = PCI_FUNC(iommu->devid); + int cap_ptr = iommu->cap_ptr; + u32 range; + + iommu->cap = read_pci_config(bus, dev, fn, cap_ptr+MMIO_CAP_HDR_OFFSET); + + range = read_pci_config(bus, dev, fn, cap_ptr+MMIO_RANGE_OFFSET); + iommu->first_device = DEVID(MMIO_GET_BUS(range), MMIO_GET_FD(range)); + iommu->last_device = DEVID(MMIO_GET_BUS(range), MMIO_GET_LD(range)); +} + +static void __init init_iommu_from_acpi(struct amd_iommu *iommu, + struct ivhd_header *h) +{ + u8 *p = (u8 *)h; + u8 *end = p, flags = 0; + u16 dev_i, devid = 0, devid_start = 0, devid_to = 0; + u32 ext_flags = 0; + bool alias = 0; + struct ivhd_entry *e; + + /* + * First set the recommended feature enable bits from ACPI + * into the IOMMU control registers + */ + h->flags & IVHD_FLAG_HT_TUN_EN ? + iommu_feature_enable(iommu, CONTROL_HT_TUN_EN) : + iommu_feature_disable(iommu, CONTROL_HT_TUN_EN); + + h->flags & IVHD_FLAG_PASSPW_EN ? + iommu_feature_enable(iommu, CONTROL_PASSPW_EN) : + iommu_feature_disable(iommu, CONTROL_PASSPW_EN); + + h->flags & IVHD_FLAG_RESPASSPW_EN ? + iommu_feature_enable(iommu, CONTROL_RESPASSPW_EN) : + iommu_feature_disable(iommu, CONTROL_RESPASSPW_EN); + + h->flags & IVHD_FLAG_ISOC_EN ? + iommu_feature_enable(iommu, CONTROL_ISOC_EN) : + iommu_feature_disable(iommu, CONTROL_ISOC_EN); + + /* + * make IOMMU memory accesses cache coherent + */ + iommu_feature_enable(iommu, CONTROL_COHERENT_EN); + + /* + * Done. Now parse the device entries + */ + p += sizeof(struct ivhd_header); + end += h->length; + + while (p < end) { + e = (struct ivhd_entry *)p; + switch (e->type) { + case IVHD_DEV_ALL: + for (dev_i = iommu->first_device; + dev_i <= iommu->last_device; ++dev_i) + set_dev_entry_from_acpi(dev_i, e->flags, 0); + break; + case IVHD_DEV_SELECT: + devid = e->devid; + set_dev_entry_from_acpi(devid, e->flags, 0); + break; + case IVHD_DEV_SELECT_RANGE_START: + devid_start = e->devid; + flags = e->flags; + ext_flags = 0; + alias = 0; + break; + case IVHD_DEV_ALIAS: + devid = e->devid; + devid_to = e->ext >> 8; + set_dev_entry_from_acpi(devid, e->flags, 0); + amd_iommu_alias_table[devid] = devid_to; + break; + case IVHD_DEV_ALIAS_RANGE: + devid_start = e->devid; + flags = e->flags; + devid_to = e->ext >> 8; + ext_flags = 0; + alias = 1; + break; + case IVHD_DEV_EXT_SELECT: + devid = e->devid; + set_dev_entry_from_acpi(devid, e->flags, e->ext); + break; + case IVHD_DEV_EXT_SELECT_RANGE: + devid_start = e->devid; + flags = e->flags; + ext_flags = e->ext; + alias = 0; + break; + case IVHD_DEV_RANGE_END: + devid = e->devid; + for (dev_i = devid_start; dev_i <= devid; ++dev_i) { + if (alias) + amd_iommu_alias_table[dev_i] = devid_to; + set_dev_entry_from_acpi( + amd_iommu_alias_table[dev_i], + flags, ext_flags); + } + break; + default: + break; + } + + p += 0x04 << (e->type >> 6); + } +} + +static int __init init_iommu_devices(struct amd_iommu *iommu) +{ + u16 i; + + for (i = iommu->first_device; i <= iommu->last_device; ++i) + set_iommu_for_device(iommu, i); + + return 0; +} + +static void __init free_iommu_one(struct amd_iommu *iommu) +{ + free_command_buffer(iommu); + iommu_unmap_mmio_space(iommu); +} + +static void __init free_iommu_all(void) +{ + struct amd_iommu *iommu, *next; + + list_for_each_entry_safe(iommu, next, &amd_iommu_list, list) { + list_del(&iommu->list); + free_iommu_one(iommu); + kfree(iommu); + } +} + +static int __init init_iommu_one(struct amd_iommu *iommu, struct ivhd_header *h) +{ + spin_lock_init(&iommu->lock); + list_add_tail(&iommu->list, &amd_iommu_list); + + /* + * Copy data from ACPI table entry to the iommu struct + */ + iommu->devid = h->devid; + iommu->cap_ptr = h->cap_ptr; + iommu->mmio_phys = h->mmio_phys; + iommu->mmio_base = iommu_map_mmio_space(h->mmio_phys); + if (!iommu->mmio_base) + return -ENOMEM; + + iommu_set_device_table(iommu); + iommu->cmd_buf = alloc_command_buffer(iommu); + if (!iommu->cmd_buf) + return -ENOMEM; + + init_iommu_from_pci(iommu); + init_iommu_from_acpi(iommu, h); + init_iommu_devices(iommu); + + return 0; +} + +static int __init init_iommu_all(struct acpi_table_header *table) +{ + u8 *p = (u8 *)table, *end = (u8 *)table; + struct ivhd_header *h; + struct amd_iommu *iommu; + int ret; + + INIT_LIST_HEAD(&amd_iommu_list); + + end += table->length; + p += IVRS_HEADER_LENGTH; + + while (p < end) { + h = (struct ivhd_header *)p; + switch (*p) { + case ACPI_IVHD_TYPE: + iommu = kzalloc(sizeof(struct amd_iommu), GFP_KERNEL); + if (iommu == NULL) + return -ENOMEM; + ret = init_iommu_one(iommu, h); + if (ret) + return ret; + break; + default: + break; + } + p += h->length; + + } + WARN_ON(p != end); + + return 0; +} + +static void __init free_unity_maps(void) +{ + struct unity_map_entry *entry, *next; + + list_for_each_entry_safe(entry, next, &amd_iommu_unity_map, list) { + list_del(&entry->list); + kfree(entry); + } +} + +static int __init init_exclusion_range(struct ivmd_header *m) +{ + int i; + + switch (m->type) { + case ACPI_IVMD_TYPE: + set_device_exclusion_range(m->devid, m); + break; + case ACPI_IVMD_TYPE_ALL: + for (i = 0; i < amd_iommu_last_bdf; ++i) + set_device_exclusion_range(i, m); + break; + case ACPI_IVMD_TYPE_RANGE: + for (i = m->devid; i <= m->aux; ++i) + set_device_exclusion_range(i, m); + break; + default: + break; + } + + return 0; +} + +static int __init init_unity_map_range(struct ivmd_header *m) +{ + struct unity_map_entry *e = 0; + + e = kzalloc(sizeof(*e), GFP_KERNEL); + if (e == NULL) + return -ENOMEM; + + switch (m->type) { + default: + case ACPI_IVMD_TYPE: + e->devid_start = e->devid_end = m->devid; + break; + case ACPI_IVMD_TYPE_ALL: + e->devid_start = 0; + e->devid_end = amd_iommu_last_bdf; + break; + case ACPI_IVMD_TYPE_RANGE: + e->devid_start = m->devid; + e->devid_end = m->aux; + break; + } + e->address_start = PAGE_ALIGN(m->range_start); + e->address_end = e->address_start + PAGE_ALIGN(m->range_length); + e->prot = m->flags >> 1; + + list_add_tail(&e->list, &amd_iommu_unity_map); + + return 0; +} + +static int __init init_memory_definitions(struct acpi_table_header *table) +{ + u8 *p = (u8 *)table, *end = (u8 *)table; + struct ivmd_header *m; + + INIT_LIST_HEAD(&amd_iommu_unity_map); + + end += table->length; + p += IVRS_HEADER_LENGTH; + + while (p < end) { + m = (struct ivmd_header *)p; + if (m->flags & IVMD_FLAG_EXCL_RANGE) + init_exclusion_range(m); + else if (m->flags & IVMD_FLAG_UNITY_MAP) + init_unity_map_range(m); + + p += m->length; + } + + return 0; +} + +static void __init enable_iommus(void) +{ + struct amd_iommu *iommu; + + list_for_each_entry(iommu, &amd_iommu_list, list) { + iommu_set_exclusion_range(iommu); + iommu_enable(iommu); + } +} + +/* + * Suspend/Resume support + * disable suspend until real resume implemented + */ + +static int amd_iommu_resume(struct sys_device *dev) +{ + return 0; +} + +static int amd_iommu_suspend(struct sys_device *dev, pm_message_t state) +{ + return -EINVAL; +} + +static struct sysdev_class amd_iommu_sysdev_class = { + .name = "amd_iommu", + .suspend = amd_iommu_suspend, + .resume = amd_iommu_resume, +}; + +static struct sys_device device_amd_iommu = { + .id = 0, + .cls = &amd_iommu_sysdev_class, +}; + +int __init amd_iommu_init(void) +{ + int i, ret = 0; + + + if (no_iommu) { + printk(KERN_INFO "AMD IOMMU disabled by kernel command line\n"); + return 0; + } + + if (!amd_iommu_detected) + return -ENODEV; + + /* + * First parse ACPI tables to find the largest Bus/Dev/Func + * we need to handle. Upon this information the shared data + * structures for the IOMMUs in the system will be allocated + */ + if (acpi_table_parse("IVRS", find_last_devid_acpi) != 0) + return -ENODEV; + + dev_table_size = TBL_SIZE(DEV_TABLE_ENTRY_SIZE); + alias_table_size = TBL_SIZE(ALIAS_TABLE_ENTRY_SIZE); + rlookup_table_size = TBL_SIZE(RLOOKUP_TABLE_ENTRY_SIZE); + + ret = -ENOMEM; + + /* Device table - directly used by all IOMMUs */ + amd_iommu_dev_table = (void *)__get_free_pages(GFP_KERNEL, + get_order(dev_table_size)); + if (amd_iommu_dev_table == NULL) + goto out; + + /* + * Alias table - map PCI Bus/Dev/Func to Bus/Dev/Func the + * IOMMU see for that device + */ + amd_iommu_alias_table = (void *)__get_free_pages(GFP_KERNEL, + get_order(alias_table_size)); + if (amd_iommu_alias_table == NULL) + goto free; + + /* IOMMU rlookup table - find the IOMMU for a specific device */ + amd_iommu_rlookup_table = (void *)__get_free_pages(GFP_KERNEL, + get_order(rlookup_table_size)); + if (amd_iommu_rlookup_table == NULL) + goto free; + + /* + * Protection Domain table - maps devices to protection domains + * This table has the same size as the rlookup_table + */ + amd_iommu_pd_table = (void *)__get_free_pages(GFP_KERNEL, + get_order(rlookup_table_size)); + if (amd_iommu_pd_table == NULL) + goto free; + + amd_iommu_pd_alloc_bitmap = (void *)__get_free_pages(GFP_KERNEL, + get_order(MAX_DOMAIN_ID/8)); + if (amd_iommu_pd_alloc_bitmap == NULL) + goto free; + + /* + * memory is allocated now; initialize the device table with all zeroes + * and let all alias entries point to itself + */ + memset(amd_iommu_dev_table, 0, dev_table_size); + for (i = 0; i < amd_iommu_last_bdf; ++i) + amd_iommu_alias_table[i] = i; + + memset(amd_iommu_pd_table, 0, rlookup_table_size); + memset(amd_iommu_pd_alloc_bitmap, 0, MAX_DOMAIN_ID / 8); + + /* + * never allocate domain 0 because its used as the non-allocated and + * error value placeholder + */ + amd_iommu_pd_alloc_bitmap[0] = 1; + + /* + * now the data structures are allocated and basically initialized + * start the real acpi table scan + */ + ret = -ENODEV; + if (acpi_table_parse("IVRS", init_iommu_all) != 0) + goto free; + + if (acpi_table_parse("IVRS", init_memory_definitions) != 0) + goto free; + + ret = amd_iommu_init_dma_ops(); + if (ret) + goto free; + + ret = sysdev_class_register(&amd_iommu_sysdev_class); + if (ret) + goto free; + + ret = sysdev_register(&device_amd_iommu); + if (ret) + goto free; + + enable_iommus(); + + printk(KERN_INFO "AMD IOMMU: aperture size is %d MB\n", + (1 << (amd_iommu_aperture_order-20))); + + printk(KERN_INFO "AMD IOMMU: device isolation "); + if (amd_iommu_isolate) + printk("enabled\n"); + else + printk("disabled\n"); + +out: + return ret; + +free: + if (amd_iommu_pd_alloc_bitmap) + free_pages((unsigned long)amd_iommu_pd_alloc_bitmap, 1); + + if (amd_iommu_pd_table) + free_pages((unsigned long)amd_iommu_pd_table, + get_order(rlookup_table_size)); + + if (amd_iommu_rlookup_table) + free_pages((unsigned long)amd_iommu_rlookup_table, + get_order(rlookup_table_size)); + + if (amd_iommu_alias_table) + free_pages((unsigned long)amd_iommu_alias_table, + get_order(alias_table_size)); + + if (amd_iommu_dev_table) + free_pages((unsigned long)amd_iommu_dev_table, + get_order(dev_table_size)); + + free_iommu_all(); + + free_unity_maps(); + + goto out; +} + +static int __init early_amd_iommu_detect(struct acpi_table_header *table) +{ + return 0; +} + +void __init amd_iommu_detect(void) +{ + if (swiotlb || no_iommu || iommu_detected) + return; + + if (acpi_table_parse("IVRS", early_amd_iommu_detect) == 0) { + iommu_detected = 1; + amd_iommu_detected = 1; +#ifdef CONFIG_GART_IOMMU + gart_iommu_aperture_disabled = 1; + gart_iommu_aperture = 0; +#endif + } +} + +static int __init parse_amd_iommu_options(char *str) +{ + for (; *str; ++str) { + if (strcmp(str, "isolate") == 0) + amd_iommu_isolate = 1; + } + + return 1; +} + +static int __init parse_amd_iommu_size_options(char *str) +{ + for (; *str; ++str) { + if (strcmp(str, "32M") == 0) + amd_iommu_aperture_order = 25; + if (strcmp(str, "64M") == 0) + amd_iommu_aperture_order = 26; + if (strcmp(str, "128M") == 0) + amd_iommu_aperture_order = 27; + if (strcmp(str, "256M") == 0) + amd_iommu_aperture_order = 28; + if (strcmp(str, "512M") == 0) + amd_iommu_aperture_order = 29; + if (strcmp(str, "1G") == 0) + amd_iommu_aperture_order = 30; + } + + return 1; +} + +__setup("amd_iommu=", parse_amd_iommu_options); +__setup("amd_iommu_size=", parse_amd_iommu_size_options); diff --git a/arch/x86/kernel/aperture_64.c b/arch/x86/kernel/aperture_64.c index 479926d9e00..9f907806c1a 100644 --- a/arch/x86/kernel/aperture_64.c +++ b/arch/x86/kernel/aperture_64.c @@ -35,6 +35,18 @@ int fallback_aper_force __initdata; int fix_aperture __initdata = 1; +struct bus_dev_range { + int bus; + int dev_base; + int dev_limit; +}; + +static struct bus_dev_range bus_dev_ranges[] __initdata = { + { 0x00, 0x18, 0x20}, + { 0xff, 0x00, 0x20}, + { 0xfe, 0x00, 0x20} +}; + static struct resource gart_resource = { .name = "GART", .flags = IORESOURCE_MEM, @@ -55,8 +67,9 @@ static u32 __init allocate_aperture(void) u32 aper_size; void *p; - if (fallback_aper_order > 7) - fallback_aper_order = 7; + /* aper_size should <= 1G */ + if (fallback_aper_order > 5) + fallback_aper_order = 5; aper_size = (32 * 1024 * 1024) << fallback_aper_order; /* @@ -65,7 +78,20 @@ static u32 __init allocate_aperture(void) * memory. Unfortunately we cannot move it up because that would * make the IOMMU useless. */ - p = __alloc_bootmem_nopanic(aper_size, aper_size, 0); + /* + * using 512M as goal, in case kexec will load kernel_big + * that will do the on position decompress, and could overlap with + * that positon with gart that is used. + * sequende: + * kernel_small + * ==> kexec (with kdump trigger path or previous doesn't shutdown gart) + * ==> kernel_small(gart area become e820_reserved) + * ==> kexec (with kdump trigger path or previous doesn't shutdown gart) + * ==> kerne_big (uncompressed size will be big than 64M or 128M) + * so don't use 512M below as gart iommu, leave the space for kernel + * code for safe + */ + p = __alloc_bootmem_nopanic(aper_size, aper_size, 512ULL<<20); if (!p || __pa(p)+aper_size > 0xffffffff) { printk(KERN_ERR "Cannot allocate aperture memory hole (%p,%uK)\n", @@ -83,69 +109,53 @@ static u32 __init allocate_aperture(void) return (u32)__pa(p); } -static int __init aperture_valid(u64 aper_base, u32 aper_size) -{ - if (!aper_base) - return 0; - - if (aper_base + aper_size > 0x100000000UL) { - printk(KERN_ERR "Aperture beyond 4GB. Ignoring.\n"); - return 0; - } - if (e820_any_mapped(aper_base, aper_base + aper_size, E820_RAM)) { - printk(KERN_ERR "Aperture pointing to e820 RAM. Ignoring.\n"); - return 0; - } - if (aper_size < 64*1024*1024) { - printk(KERN_ERR "Aperture too small (%d MB)\n", aper_size>>20); - return 0; - } - - return 1; -} /* Find a PCI capability */ -static __u32 __init find_cap(int num, int slot, int func, int cap) +static u32 __init find_cap(int bus, int slot, int func, int cap) { int bytes; u8 pos; - if (!(read_pci_config_16(num, slot, func, PCI_STATUS) & + if (!(read_pci_config_16(bus, slot, func, PCI_STATUS) & PCI_STATUS_CAP_LIST)) return 0; - pos = read_pci_config_byte(num, slot, func, PCI_CAPABILITY_LIST); + pos = read_pci_config_byte(bus, slot, func, PCI_CAPABILITY_LIST); for (bytes = 0; bytes < 48 && pos >= 0x40; bytes++) { u8 id; pos &= ~3; - id = read_pci_config_byte(num, slot, func, pos+PCI_CAP_LIST_ID); + id = read_pci_config_byte(bus, slot, func, pos+PCI_CAP_LIST_ID); if (id == 0xff) break; if (id == cap) return pos; - pos = read_pci_config_byte(num, slot, func, + pos = read_pci_config_byte(bus, slot, func, pos+PCI_CAP_LIST_NEXT); } return 0; } /* Read a standard AGPv3 bridge header */ -static __u32 __init read_agp(int num, int slot, int func, int cap, u32 *order) +static u32 __init read_agp(int bus, int slot, int func, int cap, u32 *order) { u32 apsize; u32 apsizereg; int nbits; u32 aper_low, aper_hi; u64 aper; + u32 old_order; - printk(KERN_INFO "AGP bridge at %02x:%02x:%02x\n", num, slot, func); - apsizereg = read_pci_config_16(num, slot, func, cap + 0x14); + printk(KERN_INFO "AGP bridge at %02x:%02x:%02x\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"); return 0; } + /* old_order could be the value from NB gart setting */ + old_order = *order; + apsize = apsizereg & 0xfff; /* Some BIOS use weird encodings not in the AGPv3 table. */ if (apsize & 0xff) @@ -155,14 +165,26 @@ static __u32 __init read_agp(int num, int slot, int func, int cap, u32 *order) if ((int)*order < 0) /* < 32MB */ *order = 0; - aper_low = read_pci_config(num, slot, func, 0x10); - aper_hi = read_pci_config(num, slot, func, 0x14); + aper_low = read_pci_config(bus, slot, func, 0x10); + aper_hi = read_pci_config(bus, slot, func, 0x14); aper = (aper_low & ~((1<<22)-1)) | ((u64)aper_hi << 32); + /* + * 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); + 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); + *order = old_order; + } + printk(KERN_INFO "Aperture from AGP @ %Lx size %u MB (APSIZE %x)\n", aper, 32 << *order, apsizereg); - if (!aperture_valid(aper, (32*1024*1024) << *order)) + if (!aperture_valid(aper, (32*1024*1024) << *order, 32<<20)) return 0; return (u32)aper; } @@ -180,17 +202,17 @@ static __u32 __init read_agp(int num, int slot, int func, int cap, u32 *order) * the AGP bridges should be always an own bus on the HT hierarchy, * but do it here for future safety. */ -static __u32 __init search_agp_bridge(u32 *order, int *valid_agp) +static u32 __init search_agp_bridge(u32 *order, int *valid_agp) { - int num, slot, func; + int bus, slot, func; /* Poor man's PCI discovery */ - for (num = 0; num < 256; num++) { + for (bus = 0; bus < 256; bus++) { for (slot = 0; slot < 32; slot++) { for (func = 0; func < 8; func++) { u32 class, cap; u8 type; - class = read_pci_config(num, slot, func, + class = read_pci_config(bus, slot, func, PCI_CLASS_REVISION); if (class == 0xffffffff) break; @@ -199,17 +221,17 @@ static __u32 __init search_agp_bridge(u32 *order, int *valid_agp) case PCI_CLASS_BRIDGE_HOST: case PCI_CLASS_BRIDGE_OTHER: /* needed? */ /* AGP bridge? */ - cap = find_cap(num, slot, func, + cap = find_cap(bus, slot, func, PCI_CAP_ID_AGP); if (!cap) break; *valid_agp = 1; - return read_agp(num, slot, func, cap, + return read_agp(bus, slot, func, cap, order); } /* No multi-function device? */ - type = read_pci_config_byte(num, slot, func, + type = read_pci_config_byte(bus, slot, func, PCI_HEADER_TYPE); if (!(type & 0x80)) break; @@ -249,36 +271,50 @@ void __init early_gart_iommu_check(void) * or BIOS forget to put that in reserved. * try to update e820 to make that region as reserved. */ - int fix, num; + int i, fix, slot; u32 ctl; u32 aper_size = 0, aper_order = 0, last_aper_order = 0; u64 aper_base = 0, last_aper_base = 0; - int aper_enabled = 0, last_aper_enabled = 0; + int aper_enabled = 0, last_aper_enabled = 0, last_valid = 0; if (!early_pci_allowed()) return; + /* This is mostly duplicate of iommu_hole_init */ fix = 0; - for (num = 24; num < 32; num++) { - if (!early_is_k8_nb(read_pci_config(0, num, 3, 0x00))) - continue; - - ctl = read_pci_config(0, num, 3, 0x90); - aper_enabled = ctl & 1; - aper_order = (ctl >> 1) & 7; - aper_size = (32 * 1024 * 1024) << aper_order; - aper_base = read_pci_config(0, num, 3, 0x94) & 0x7fff; - aper_base <<= 25; - - if ((last_aper_order && aper_order != last_aper_order) || - (last_aper_base && aper_base != last_aper_base) || - (last_aper_enabled && aper_enabled != last_aper_enabled)) { - fix = 1; - break; + for (i = 0; i < ARRAY_SIZE(bus_dev_ranges); i++) { + int bus; + int dev_base, dev_limit; + + bus = bus_dev_ranges[i].bus; + dev_base = bus_dev_ranges[i].dev_base; + dev_limit = bus_dev_ranges[i].dev_limit; + + for (slot = dev_base; slot < dev_limit; slot++) { + if (!early_is_k8_nb(read_pci_config(bus, slot, 3, 0x00))) + continue; + + ctl = read_pci_config(bus, slot, 3, AMD64_GARTAPERTURECTL); + aper_enabled = ctl & AMD64_GARTEN; + aper_order = (ctl >> 1) & 7; + aper_size = (32 * 1024 * 1024) << aper_order; + aper_base = read_pci_config(bus, slot, 3, AMD64_GARTAPERTUREBASE) & 0x7fff; + aper_base <<= 25; + + if (last_valid) { + if ((aper_order != last_aper_order) || + (aper_base != last_aper_base) || + (aper_enabled != last_aper_enabled)) { + fix = 1; + break; + } + } + + last_aper_order = aper_order; + last_aper_base = aper_base; + last_aper_enabled = aper_enabled; + last_valid = 1; } - last_aper_order = aper_order; - last_aper_base = aper_base; - last_aper_enabled = aper_enabled; } if (!fix && !aper_enabled) @@ -290,32 +326,46 @@ void __init early_gart_iommu_check(void) if (gart_fix_e820 && !fix && aper_enabled) { if (e820_any_mapped(aper_base, aper_base + aper_size, E820_RAM)) { - /* reserved it, so we can resuse it in second kernel */ + /* reserve it, so we can reuse it in second kernel */ printk(KERN_INFO "update e820 for GART\n"); - add_memory_region(aper_base, aper_size, E820_RESERVED); + e820_add_region(aper_base, aper_size, E820_RESERVED); update_e820(); } - return; } + if (!fix) + return; + /* different nodes have different setting, disable them all at first*/ - for (num = 24; num < 32; num++) { - if (!early_is_k8_nb(read_pci_config(0, num, 3, 0x00))) - continue; + for (i = 0; i < ARRAY_SIZE(bus_dev_ranges); i++) { + int bus; + int dev_base, dev_limit; + + bus = bus_dev_ranges[i].bus; + dev_base = bus_dev_ranges[i].dev_base; + dev_limit = bus_dev_ranges[i].dev_limit; + + for (slot = dev_base; slot < dev_limit; slot++) { + if (!early_is_k8_nb(read_pci_config(bus, slot, 3, 0x00))) + continue; - ctl = read_pci_config(0, num, 3, 0x90); - ctl &= ~1; - write_pci_config(0, num, 3, 0x90, ctl); + ctl = read_pci_config(bus, slot, 3, AMD64_GARTAPERTURECTL); + ctl &= ~AMD64_GARTEN; + write_pci_config(bus, slot, 3, AMD64_GARTAPERTURECTL, ctl); + } } } +static int __initdata printed_gart_size_msg; + void __init gart_iommu_hole_init(void) { + u32 agp_aper_base = 0, agp_aper_order = 0; u32 aper_size, aper_alloc = 0, aper_order = 0, last_aper_order = 0; u64 aper_base, last_aper_base = 0; - int fix, num, valid_agp = 0; - int node; + int fix, slot, valid_agp = 0; + int i, node; if (gart_iommu_aperture_disabled || !fix_aperture || !early_pci_allowed()) @@ -323,38 +373,65 @@ void __init gart_iommu_hole_init(void) printk(KERN_INFO "Checking aperture...\n"); + if (!fallback_aper_force) + agp_aper_base = search_agp_bridge(&agp_aper_order, &valid_agp); + fix = 0; node = 0; - for (num = 24; num < 32; num++) { - if (!early_is_k8_nb(read_pci_config(0, num, 3, 0x00))) - continue; - - iommu_detected = 1; - gart_iommu_aperture = 1; - - aper_order = (read_pci_config(0, num, 3, 0x90) >> 1) & 7; - aper_size = (32 * 1024 * 1024) << aper_order; - aper_base = read_pci_config(0, num, 3, 0x94) & 0x7fff; - aper_base <<= 25; - - printk(KERN_INFO "Node %d: aperture @ %Lx size %u MB\n", - node, aper_base, aper_size >> 20); - node++; - - if (!aperture_valid(aper_base, aper_size)) { - fix = 1; - break; - } + for (i = 0; i < ARRAY_SIZE(bus_dev_ranges); i++) { + int bus; + int dev_base, dev_limit; + + bus = bus_dev_ranges[i].bus; + dev_base = bus_dev_ranges[i].dev_base; + dev_limit = bus_dev_ranges[i].dev_limit; + + for (slot = dev_base; slot < dev_limit; slot++) { + if (!early_is_k8_nb(read_pci_config(bus, slot, 3, 0x00))) + continue; + + iommu_detected = 1; + gart_iommu_aperture = 1; + + aper_order = (read_pci_config(bus, slot, 3, AMD64_GARTAPERTURECTL) >> 1) & 7; + aper_size = (32 * 1024 * 1024) << aper_order; + 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); + node++; + + if (!aperture_valid(aper_base, aper_size, 64<<20)) { + if (valid_agp && agp_aper_base && + agp_aper_base == aper_base && + agp_aper_order == aper_order) { + /* the same between two setting from NB and agp */ + 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"); + printed_gart_size_msg = 1; + } + } else { + fix = 1; + goto out; + } + } - if ((last_aper_order && aper_order != last_aper_order) || - (last_aper_base && aper_base != last_aper_base)) { - fix = 1; - break; + if ((last_aper_order && aper_order != last_aper_order) || + (last_aper_base && aper_base != last_aper_base)) { + fix = 1; + goto out; + } + last_aper_order = aper_order; + last_aper_base = aper_base; } - last_aper_order = aper_order; - last_aper_base = aper_base; } +out: if (!fix && !fallback_aper_force) { if (last_aper_base) { unsigned long n = (32 * 1024 * 1024) << last_aper_order; @@ -364,14 +441,16 @@ void __init gart_iommu_hole_init(void) return; } - if (!fallback_aper_force) - aper_alloc = search_agp_bridge(&aper_order, &valid_agp); + if (!fallback_aper_force) { + aper_alloc = agp_aper_base; + aper_order = agp_aper_order; + } if (aper_alloc) { /* Got the aperture from the AGP bridge */ } else if (swiotlb && !valid_agp) { /* Do nothing */ - } else if ((!no_iommu && end_pfn > MAX_DMA32_PFN) || + } else if ((!no_iommu && max_pfn > MAX_DMA32_PFN) || force_iommu || valid_agp || fallback_aper_force) { @@ -401,16 +480,24 @@ void __init gart_iommu_hole_init(void) } /* Fix up the north bridges */ - for (num = 24; num < 32; num++) { - if (!early_is_k8_nb(read_pci_config(0, num, 3, 0x00))) - continue; - - /* - * Don't enable translation yet. That is done later. - * Assume this BIOS didn't initialise the GART so - * just overwrite all previous bits - */ - write_pci_config(0, num, 3, 0x90, aper_order<<1); - write_pci_config(0, num, 3, 0x94, aper_alloc>>25); + for (i = 0; i < ARRAY_SIZE(bus_dev_ranges); i++) { + int bus; + int dev_base, dev_limit; + + bus = bus_dev_ranges[i].bus; + dev_base = bus_dev_ranges[i].dev_base; + dev_limit = bus_dev_ranges[i].dev_limit; + for (slot = dev_base; slot < dev_limit; slot++) { + if (!early_is_k8_nb(read_pci_config(bus, slot, 3, 0x00))) + continue; + + /* Don't enable translation yet. That is done later. + Assume this BIOS didn't initialise the GART so + just overwrite all previous bits */ + write_pci_config(bus, slot, 3, AMD64_GARTAPERTURECTL, aper_order << 1); + write_pci_config(bus, slot, 3, AMD64_GARTAPERTUREBASE, aper_alloc >> 25); + } } + + set_up_gart_resume(aper_order, aper_alloc); } diff --git a/arch/x86/kernel/apic_32.c b/arch/x86/kernel/apic_32.c index 4b99b1bdeb6..3e58b676d23 100644 --- a/arch/x86/kernel/apic_32.c +++ b/arch/x86/kernel/apic_32.c @@ -52,30 +52,41 @@ unsigned long mp_lapic_addr; -DEFINE_PER_CPU(u16, x86_bios_cpu_apicid) = BAD_APICID; -EXPORT_PER_CPU_SYMBOL(x86_bios_cpu_apicid); - /* * Knob to control our willingness to enable the local APIC. * - * -1=force-disable, +1=force-enable + * +1=force-enable */ -static int enable_local_apic __initdata; +static int force_enable_local_apic; +int disable_apic; /* Local APIC timer verification ok */ static int local_apic_timer_verify_ok; -/* Disable local APIC timer from the kernel commandline or via dmi quirk - or using CPU MSR check */ -int local_apic_timer_disabled; +/* Disable local APIC timer from the kernel commandline or via dmi quirk */ +static int local_apic_timer_disabled; /* Local APIC timer works in C2 */ int local_apic_timer_c2_ok; EXPORT_SYMBOL_GPL(local_apic_timer_c2_ok); +int first_system_vector = 0xfe; + +char system_vectors[NR_VECTORS] = { [0 ... NR_VECTORS-1] = SYS_VECTOR_FREE}; + /* * Debug level, exported for io_apic.c */ int apic_verbosity; +int pic_mode; + +/* Have we found an MP table */ +int smp_found_config; + +static struct resource lapic_resource = { + .name = "Local APIC", + .flags = IORESOURCE_MEM | IORESOURCE_BUSY, +}; + static unsigned int calibration_result; static int lapic_next_event(unsigned long delta, @@ -545,7 +556,7 @@ void __init setup_boot_APIC_clock(void) lapic_clockevent.features &= ~CLOCK_EVT_FEAT_DUMMY; else printk(KERN_WARNING "APIC timer registered as dummy," - " due to nmi_watchdog=1!\n"); + " due to nmi_watchdog=%d!\n", nmi_watchdog); } /* Setup the lapic or request the broadcast */ @@ -963,7 +974,7 @@ void __cpuinit setup_local_APIC(void) * Double-check whether this APIC is really registered. */ if (!apic_id_registered()) - BUG(); + WARN_ON_ONCE(1); /* * Intel recommends to set DFR, LDR and TPR before enabling @@ -1094,7 +1105,7 @@ static int __init detect_init_APIC(void) u32 h, l, features; /* Disabled by kernel option? */ - if (enable_local_apic < 0) + if (disable_apic) return -1; switch (boot_cpu_data.x86_vendor) { @@ -1117,7 +1128,7 @@ static int __init detect_init_APIC(void) * Over-ride BIOS and try to enable the local APIC only if * "lapic" specified. */ - if (enable_local_apic <= 0) { + if (!force_enable_local_apic) { printk(KERN_INFO "Local APIC disabled by BIOS -- " "you can enable it with \"lapic\"\n"); return -1; @@ -1154,9 +1165,6 @@ static int __init detect_init_APIC(void) if (l & MSR_IA32_APICBASE_ENABLE) mp_lapic_addr = l & MSR_IA32_APICBASE_BASE; - if (nmi_watchdog != NMI_NONE && nmi_watchdog != NMI_DISABLED) - nmi_watchdog = NMI_LOCAL_APIC; - printk(KERN_INFO "Found and enabled local APIC!\n"); apic_pm_activate(); @@ -1195,36 +1203,6 @@ void __init init_apic_mappings(void) if (boot_cpu_physical_apicid == -1U) boot_cpu_physical_apicid = GET_APIC_ID(read_apic_id()); -#ifdef CONFIG_X86_IO_APIC - { - unsigned long ioapic_phys, idx = FIX_IO_APIC_BASE_0; - int i; - - for (i = 0; i < nr_ioapics; i++) { - if (smp_found_config) { - ioapic_phys = mp_ioapics[i].mpc_apicaddr; - if (!ioapic_phys) { - printk(KERN_ERR - "WARNING: bogus zero IO-APIC " - "address found in MPTABLE, " - "disabling IO/APIC support!\n"); - smp_found_config = 0; - skip_ioapic_setup = 1; - goto fake_ioapic_page; - } - } else { -fake_ioapic_page: - ioapic_phys = (unsigned long) - alloc_bootmem_pages(PAGE_SIZE); - ioapic_phys = __pa(ioapic_phys); - } - set_fixmap_nocache(idx, ioapic_phys); - printk(KERN_DEBUG "mapped IOAPIC to %08lx (%08lx)\n", - __fix_to_virt(idx), ioapic_phys); - idx++; - } - } -#endif } /* @@ -1236,7 +1214,7 @@ int apic_version[MAX_APICS]; int __init APIC_init_uniprocessor(void) { - if (enable_local_apic < 0) + if (disable_apic) clear_cpu_cap(&boot_cpu_data, X86_FEATURE_APIC); if (!smp_found_config && !cpu_has_apic) @@ -1265,10 +1243,14 @@ int __init APIC_init_uniprocessor(void) #ifdef CONFIG_CRASH_DUMP boot_cpu_physical_apicid = GET_APIC_ID(read_apic_id()); #endif - phys_cpu_present_map = physid_mask_of_physid(boot_cpu_physical_apicid); + physid_set_mask_of_physid(boot_cpu_physical_apicid, &phys_cpu_present_map); setup_local_APIC(); +#ifdef CONFIG_X86_IO_APIC + if (!smp_found_config || skip_ioapic_setup || !nr_ioapics) +#endif + localise_nmi_watchdog(); end_local_APIC_setup(); #ifdef CONFIG_X86_IO_APIC if (smp_found_config) @@ -1351,13 +1333,13 @@ void __init smp_intr_init(void) * The reschedule interrupt is a CPU-to-CPU reschedule-helper * IPI, driven by wakeup. */ - set_intr_gate(RESCHEDULE_VECTOR, reschedule_interrupt); + alloc_intr_gate(RESCHEDULE_VECTOR, reschedule_interrupt); /* IPI for invalidation */ - set_intr_gate(INVALIDATE_TLB_VECTOR, invalidate_interrupt); + alloc_intr_gate(INVALIDATE_TLB_VECTOR, invalidate_interrupt); /* IPI for generic function call */ - set_intr_gate(CALL_FUNCTION_VECTOR, call_function_interrupt); + alloc_intr_gate(CALL_FUNCTION_VECTOR, call_function_interrupt); } #endif @@ -1370,15 +1352,15 @@ void __init apic_intr_init(void) smp_intr_init(); #endif /* self generated IPI for local APIC timer */ - set_intr_gate(LOCAL_TIMER_VECTOR, apic_timer_interrupt); + alloc_intr_gate(LOCAL_TIMER_VECTOR, apic_timer_interrupt); /* IPI vectors for APIC spurious and error interrupts */ - set_intr_gate(SPURIOUS_APIC_VECTOR, spurious_interrupt); - set_intr_gate(ERROR_APIC_VECTOR, error_interrupt); + alloc_intr_gate(SPURIOUS_APIC_VECTOR, spurious_interrupt); + alloc_intr_gate(ERROR_APIC_VECTOR, error_interrupt); /* thermal monitor LVT interrupt */ #ifdef CONFIG_X86_MCE_P4THERMAL - set_intr_gate(THERMAL_APIC_VECTOR, thermal_interrupt); + alloc_intr_gate(THERMAL_APIC_VECTOR, thermal_interrupt); #endif } @@ -1513,6 +1495,9 @@ void __cpuinit generic_processor_info(int apicid, int version) */ cpu = 0; + if (apicid > max_physical_apicid) + max_physical_apicid = apicid; + /* * Would be preferable to switch to bigsmp when CONFIG_HOTPLUG_CPU=y * but we need to work other dependencies like SMP_SUSPEND etc @@ -1520,7 +1505,7 @@ void __cpuinit generic_processor_info(int apicid, int version) * if (CPU_HOTPLUG_ENABLED || num_processors > 8) * - Ashok Raj <ashok.raj@intel.com> */ - if (num_processors > 8) { + if (max_physical_apicid >= 8) { switch (boot_cpu_data.x86_vendor) { case X86_VENDOR_INTEL: if (!APIC_XAPIC(version)) { @@ -1534,9 +1519,9 @@ void __cpuinit generic_processor_info(int apicid, int version) } #ifdef CONFIG_SMP /* are we being called early in kernel startup? */ - if (x86_cpu_to_apicid_early_ptr) { - u16 *cpu_to_apicid = x86_cpu_to_apicid_early_ptr; - u16 *bios_cpu_apicid = x86_bios_cpu_apicid_early_ptr; + if (early_per_cpu_ptr(x86_cpu_to_apicid)) { + u16 *cpu_to_apicid = early_per_cpu_ptr(x86_cpu_to_apicid); + u16 *bios_cpu_apicid = early_per_cpu_ptr(x86_bios_cpu_apicid); cpu_to_apicid[cpu] = apicid; bios_cpu_apicid[cpu] = apicid; @@ -1703,14 +1688,14 @@ static void apic_pm_activate(void) { } */ static int __init parse_lapic(char *arg) { - enable_local_apic = 1; + force_enable_local_apic = 1; return 0; } early_param("lapic", parse_lapic); static int __init parse_nolapic(char *arg) { - enable_local_apic = -1; + disable_apic = 1; clear_cpu_cap(&boot_cpu_data, X86_FEATURE_APIC); return 0; } @@ -1740,3 +1725,21 @@ static int __init apic_set_verbosity(char *str) } __setup("apic=", apic_set_verbosity); +static int __init lapic_insert_resource(void) +{ + if (!apic_phys) + return -1; + + /* Put local APIC into the resource map. */ + lapic_resource.start = apic_phys; + lapic_resource.end = lapic_resource.start + PAGE_SIZE - 1; + insert_resource(&iomem_resource, &lapic_resource); + + return 0; +} + +/* + * need call insert after e820_reserve_resources() + * that is using request_resource + */ +late_initcall(lapic_insert_resource); diff --git a/arch/x86/kernel/apic_64.c b/arch/x86/kernel/apic_64.c index 0633cfd0dc2..1e3d32e27c1 100644 --- a/arch/x86/kernel/apic_64.c +++ b/arch/x86/kernel/apic_64.c @@ -43,7 +43,7 @@ #include <mach_ipi.h> #include <mach_apic.h> -int disable_apic_timer __cpuinitdata; +static int disable_apic_timer __cpuinitdata; static int apic_calibrate_pmtmr __initdata; int disable_apic; @@ -56,6 +56,9 @@ EXPORT_SYMBOL_GPL(local_apic_timer_c2_ok); */ int apic_verbosity; +/* Have we found an MP table */ +int smp_found_config; + static struct resource lapic_resource = { .name = "Local APIC", .flags = IORESOURCE_MEM | IORESOURCE_BUSY, @@ -87,9 +90,6 @@ static unsigned long apic_phys; unsigned long mp_lapic_addr; -DEFINE_PER_CPU(u16, x86_bios_cpu_apicid) = BAD_APICID; -EXPORT_PER_CPU_SYMBOL(x86_bios_cpu_apicid); - unsigned int __cpuinitdata maxcpus = NR_CPUS; /* * Get the LAPIC version @@ -417,37 +417,13 @@ void __init setup_boot_APIC_clock(void) lapic_clockevent.features &= ~CLOCK_EVT_FEAT_DUMMY; else printk(KERN_WARNING "APIC timer registered as dummy," - " due to nmi_watchdog=1!\n"); + " due to nmi_watchdog=%d!\n", nmi_watchdog); setup_APIC_timer(); } -/* - * AMD C1E enabled CPUs have a real nasty problem: Some BIOSes set the - * C1E flag only in the secondary CPU, so when we detect the wreckage - * we already have enabled the boot CPU local apic timer. Check, if - * disable_apic_timer is set and the DUMMY flag is cleared. If yes, - * set the DUMMY flag again and force the broadcast mode in the - * clockevents layer. - */ -static void __cpuinit check_boot_apic_timer_broadcast(void) -{ - if (!disable_apic_timer || - (lapic_clockevent.features & CLOCK_EVT_FEAT_DUMMY)) - return; - - printk(KERN_INFO "AMD C1E detected late. Force timer broadcast.\n"); - lapic_clockevent.features |= CLOCK_EVT_FEAT_DUMMY; - - local_irq_enable(); - clockevents_notify(CLOCK_EVT_NOTIFY_BROADCAST_FORCE, - &boot_cpu_physical_apicid); - local_irq_disable(); -} - void __cpuinit setup_secondary_APIC_clock(void) { - check_boot_apic_timer_broadcast(); setup_APIC_timer(); } @@ -850,7 +826,6 @@ static void __cpuinit lapic_setup_esr(void) void __cpuinit end_local_APIC_setup(void) { lapic_setup_esr(); - nmi_watchdog_default(); setup_apic_nmi_watchdog(NULL); apic_pm_activate(); } @@ -875,7 +850,7 @@ static int __init detect_init_APIC(void) void __init early_init_lapic_mapping(void) { - unsigned long apic_phys; + unsigned long phys_addr; /* * If no local APIC can be found then go out @@ -884,11 +859,11 @@ void __init early_init_lapic_mapping(void) if (!smp_found_config) return; - apic_phys = mp_lapic_addr; + phys_addr = mp_lapic_addr; - set_fixmap_nocache(FIX_APIC_BASE, apic_phys); + set_fixmap_nocache(FIX_APIC_BASE, phys_addr); apic_printk(APIC_VERBOSE, "mapped APIC to %16lx (%16lx)\n", - APIC_BASE, apic_phys); + APIC_BASE, phys_addr); /* * Fetch the APIC ID of the BSP in case we have a @@ -942,7 +917,9 @@ int __init APIC_init_uniprocessor(void) verify_local_APIC(); - phys_cpu_present_map = physid_mask_of_physid(boot_cpu_physical_apicid); + connect_bsp_APIC(); + + physid_set_mask_of_physid(boot_cpu_physical_apicid, &phys_cpu_present_map); apic_write(APIC_ID, SET_APIC_ID(boot_cpu_physical_apicid)); setup_local_APIC(); @@ -954,6 +931,8 @@ int __init APIC_init_uniprocessor(void) if (!skip_ioapic_setup && nr_ioapics) enable_IO_APIC(); + if (!smp_found_config || skip_ioapic_setup || !nr_ioapics) + localise_nmi_watchdog(); end_local_APIC_setup(); if (smp_found_config && !skip_ioapic_setup && nr_ioapics) @@ -1021,6 +1000,14 @@ asmlinkage void smp_error_interrupt(void) irq_exit(); } +/** + * * connect_bsp_APIC - attach the APIC to the interrupt system + * */ +void __init connect_bsp_APIC(void) +{ + enable_apic_mode(); +} + void disconnect_bsp_APIC(int virt_wire_setup) { /* Go back to Virtual Wire compatibility mode */ @@ -1090,10 +1077,13 @@ void __cpuinit generic_processor_info(int apicid, int version) */ cpu = 0; } + if (apicid > max_physical_apicid) + max_physical_apicid = apicid; + /* are we being called early in kernel startup? */ - if (x86_cpu_to_apicid_early_ptr) { - u16 *cpu_to_apicid = x86_cpu_to_apicid_early_ptr; - u16 *bios_cpu_apicid = x86_bios_cpu_apicid_early_ptr; + if (early_per_cpu_ptr(x86_cpu_to_apicid)) { + u16 *cpu_to_apicid = early_per_cpu_ptr(x86_cpu_to_apicid); + u16 *bios_cpu_apicid = early_per_cpu_ptr(x86_bios_cpu_apicid); cpu_to_apicid[cpu] = apicid; bios_cpu_apicid[cpu] = apicid; @@ -1269,7 +1259,7 @@ __cpuinit int apic_is_clustered_box(void) if ((boot_cpu_data.x86_vendor == X86_VENDOR_AMD) && !is_vsmp_box()) return 0; - bios_cpu_apicid = x86_bios_cpu_apicid_early_ptr; + bios_cpu_apicid = early_per_cpu_ptr(x86_bios_cpu_apicid); bitmap_zero(clustermap, NUM_APIC_CLUSTERS); for (i = 0; i < NR_CPUS; i++) { diff --git a/arch/x86/kernel/apm_32.c b/arch/x86/kernel/apm_32.c index bf9290e2901..00e6d137095 100644 --- a/arch/x86/kernel/apm_32.c +++ b/arch/x86/kernel/apm_32.c @@ -228,6 +228,7 @@ #include <linux/suspend.h> #include <linux/kthread.h> #include <linux/jiffies.h> +#include <linux/smp_lock.h> #include <asm/system.h> #include <asm/uaccess.h> @@ -1149,7 +1150,7 @@ static void queue_event(apm_event_t event, struct apm_user *sender) as->event_tail = 0; } as->events[as->event_head] = event; - if ((!as->suser) || (!as->writer)) + if (!as->suser || !as->writer) continue; switch (event) { case APM_SYS_SUSPEND: @@ -1396,7 +1397,7 @@ 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)) { + if (as == NULL || as->magic != APM_BIOS_MAGIC) { printk(KERN_ERR "apm: %s passed bad filp\n", func); return 1; } @@ -1459,18 +1460,19 @@ static unsigned int do_poll(struct file *fp, poll_table *wait) return 0; } -static int do_ioctl(struct inode *inode, struct file *filp, - u_int cmd, u_long arg) +static long do_ioctl(struct file *filp, u_int cmd, u_long arg) { struct apm_user *as; + int ret; as = filp->private_data; if (check_apm_user(as, "ioctl")) return -EIO; - if ((!as->suser) || (!as->writer)) + if (!as->suser || !as->writer) return -EPERM; switch (cmd) { case APM_IOC_STANDBY: + lock_kernel(); if (as->standbys_read > 0) { as->standbys_read--; as->standbys_pending--; @@ -1479,8 +1481,10 @@ static int do_ioctl(struct inode *inode, struct file *filp, queue_event(APM_USER_STANDBY, as); if (standbys_pending <= 0) standby(); + unlock_kernel(); break; case APM_IOC_SUSPEND: + lock_kernel(); if (as->suspends_read > 0) { as->suspends_read--; as->suspends_pending--; @@ -1488,16 +1492,17 @@ static int do_ioctl(struct inode *inode, struct file *filp, } else queue_event(APM_USER_SUSPEND, as); if (suspends_pending <= 0) { - return suspend(1); + ret = suspend(1); } else { as->suspend_wait = 1; wait_event_interruptible(apm_suspend_waitqueue, as->suspend_wait == 0); - return as->suspend_result; + ret = as->suspend_result; } - break; + unlock_kernel(); + return ret; default: - return -EINVAL; + return -ENOTTY; } return 0; } @@ -1860,7 +1865,7 @@ static const struct file_operations apm_bios_fops = { .owner = THIS_MODULE, .read = do_read, .poll = do_poll, - .ioctl = do_ioctl, + .unlocked_ioctl = do_ioctl, .open = do_open, .release = do_release, }; diff --git a/arch/x86/kernel/asm-offsets_32.c b/arch/x86/kernel/asm-offsets_32.c index 92588083950..6649d09ad88 100644 --- a/arch/x86/kernel/asm-offsets_32.c +++ b/arch/x86/kernel/asm-offsets_32.c @@ -111,7 +111,7 @@ void foo(void) OFFSET(PV_IRQ_irq_disable, pv_irq_ops, irq_disable); OFFSET(PV_IRQ_irq_enable, pv_irq_ops, irq_enable); OFFSET(PV_CPU_iret, pv_cpu_ops, iret); - OFFSET(PV_CPU_irq_enable_syscall_ret, pv_cpu_ops, irq_enable_syscall_ret); + OFFSET(PV_CPU_irq_enable_sysexit, pv_cpu_ops, irq_enable_sysexit); OFFSET(PV_CPU_read_cr0, pv_cpu_ops, read_cr0); #endif diff --git a/arch/x86/kernel/asm-offsets_64.c b/arch/x86/kernel/asm-offsets_64.c index f126c05d617..bacf5deeec2 100644 --- a/arch/x86/kernel/asm-offsets_64.c +++ b/arch/x86/kernel/asm-offsets_64.c @@ -34,7 +34,7 @@ int main(void) ENTRY(pid); BLANK(); #undef ENTRY -#define ENTRY(entry) DEFINE(threadinfo_ ## entry, offsetof(struct thread_info, entry)) +#define ENTRY(entry) DEFINE(TI_ ## entry, offsetof(struct thread_info, entry)) ENTRY(flags); ENTRY(addr_limit); ENTRY(preempt_count); @@ -61,8 +61,11 @@ int main(void) OFFSET(PARAVIRT_PATCH_pv_irq_ops, paravirt_patch_template, pv_irq_ops); OFFSET(PV_IRQ_irq_disable, pv_irq_ops, irq_disable); OFFSET(PV_IRQ_irq_enable, pv_irq_ops, irq_enable); + OFFSET(PV_IRQ_adjust_exception_frame, pv_irq_ops, adjust_exception_frame); OFFSET(PV_CPU_iret, pv_cpu_ops, iret); - OFFSET(PV_CPU_irq_enable_syscall_ret, pv_cpu_ops, irq_enable_syscall_ret); + OFFSET(PV_CPU_usergs_sysret32, pv_cpu_ops, usergs_sysret32); + OFFSET(PV_CPU_usergs_sysret64, pv_cpu_ops, usergs_sysret64); + OFFSET(PV_CPU_irq_enable_sysexit, pv_cpu_ops, irq_enable_sysexit); OFFSET(PV_CPU_swapgs, pv_cpu_ops, swapgs); OFFSET(PV_MMU_read_cr2, pv_mmu_ops, read_cr2); #endif diff --git a/arch/x86/kernel/cpu/Makefile b/arch/x86/kernel/cpu/Makefile index a0c6f819088..ee76eaad300 100644 --- a/arch/x86/kernel/cpu/Makefile +++ b/arch/x86/kernel/cpu/Makefile @@ -6,11 +6,15 @@ obj-y := intel_cacheinfo.o addon_cpuid_features.o obj-y += proc.o feature_names.o obj-$(CONFIG_X86_32) += common.o bugs.o +obj-$(CONFIG_X86_64) += common_64.o bugs_64.o obj-$(CONFIG_X86_32) += amd.o +obj-$(CONFIG_X86_64) += amd_64.o obj-$(CONFIG_X86_32) += cyrix.o obj-$(CONFIG_X86_32) += centaur.o +obj-$(CONFIG_X86_64) += centaur_64.o obj-$(CONFIG_X86_32) += transmeta.o obj-$(CONFIG_X86_32) += intel.o +obj-$(CONFIG_X86_64) += intel_64.o obj-$(CONFIG_X86_32) += umc.o obj-$(CONFIG_X86_MCE) += mcheck/ diff --git a/arch/x86/kernel/cpu/addon_cpuid_features.c b/arch/x86/kernel/cpu/addon_cpuid_features.c index c2e1ce33c7c..84a8220a607 100644 --- a/arch/x86/kernel/cpu/addon_cpuid_features.c +++ b/arch/x86/kernel/cpu/addon_cpuid_features.c @@ -1,9 +1,7 @@ - /* * Routines to indentify additional cpu features that are scattered in * cpuid space. */ - #include <linux/cpu.h> #include <asm/pat.h> @@ -53,19 +51,20 @@ void __cpuinit init_scattered_cpuid_features(struct cpuinfo_x86 *c) #ifdef CONFIG_X86_PAT void __cpuinit validate_pat_support(struct cpuinfo_x86 *c) { + if (!cpu_has_pat) + pat_disable("PAT not supported by CPU."); + switch (c->x86_vendor) { - case X86_VENDOR_AMD: - if (c->x86 >= 0xf && c->x86 <= 0x11) - return; - break; case X86_VENDOR_INTEL: if (c->x86 == 0xF || (c->x86 == 6 && c->x86_model >= 15)) return; break; + case X86_VENDOR_AMD: + case X86_VENDOR_CENTAUR: + case X86_VENDOR_TRANSMETA: + return; } - pat_disable(cpu_has_pat ? - "PAT disabled. Not yet verified on this CPU type." : - "PAT not supported by CPU."); + pat_disable("PAT disabled. Not yet verified on this CPU type."); } #endif diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c index 24586682829..81a07ca65d4 100644 --- a/arch/x86/kernel/cpu/amd.c +++ b/arch/x86/kernel/cpu/amd.c @@ -24,43 +24,6 @@ extern void vide(void); __asm__(".align 4\nvide: ret"); -#ifdef CONFIG_X86_LOCAL_APIC -#define ENABLE_C1E_MASK 0x18000000 -#define CPUID_PROCESSOR_SIGNATURE 1 -#define CPUID_XFAM 0x0ff00000 -#define CPUID_XFAM_K8 0x00000000 -#define CPUID_XFAM_10H 0x00100000 -#define CPUID_XFAM_11H 0x00200000 -#define CPUID_XMOD 0x000f0000 -#define CPUID_XMOD_REV_F 0x00040000 - -/* AMD systems with C1E don't have a working lAPIC timer. Check for that. */ -static __cpuinit int amd_apic_timer_broken(void) -{ - u32 lo, hi; - u32 eax = cpuid_eax(CPUID_PROCESSOR_SIGNATURE); - switch (eax & CPUID_XFAM) { - case CPUID_XFAM_K8: - if ((eax & CPUID_XMOD) < CPUID_XMOD_REV_F) - break; - case CPUID_XFAM_10H: - case CPUID_XFAM_11H: - rdmsr(MSR_K8_ENABLE_C1E, lo, hi); - if (lo & ENABLE_C1E_MASK) { - if (smp_processor_id() != boot_cpu_physical_apicid) - printk(KERN_INFO "AMD C1E detected late. " - " Force timer broadcast.\n"); - return 1; - } - break; - default: - /* err on the side of caution */ - return 1; - } - return 0; -} -#endif - int force_mwait __cpuinitdata; static void __cpuinit early_init_amd(struct cpuinfo_x86 *c) @@ -297,11 +260,6 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) num_cache_leaves = 3; } -#ifdef CONFIG_X86_LOCAL_APIC - if (amd_apic_timer_broken()) - local_apic_timer_disabled = 1; -#endif - /* K6s reports MCEs but don't actually have all the MSRs */ if (c->x86 < 6) clear_cpu_cap(c, X86_FEATURE_MCE); diff --git a/arch/x86/kernel/cpu/amd_64.c b/arch/x86/kernel/cpu/amd_64.c new file mode 100644 index 00000000000..7c36fb8a28d --- /dev/null +++ b/arch/x86/kernel/cpu/amd_64.c @@ -0,0 +1,222 @@ +#include <linux/init.h> +#include <linux/mm.h> + +#include <asm/numa_64.h> +#include <asm/mmconfig.h> +#include <asm/cacheflush.h> + +#include <mach_apic.h> + +#include "cpu.h" + +int force_mwait __cpuinitdata; + +#ifdef CONFIG_NUMA +static int __cpuinit nearby_node(int apicid) +{ + int i, node; + + for (i = apicid - 1; i >= 0; i--) { + node = apicid_to_node[i]; + if (node != NUMA_NO_NODE && node_online(node)) + return node; + } + for (i = apicid + 1; i < MAX_LOCAL_APIC; i++) { + node = apicid_to_node[i]; + if (node != NUMA_NO_NODE && node_online(node)) + return node; + } + return first_node(node_online_map); /* Shouldn't happen */ +} +#endif + +/* + * On a AMD dual core setup the lower bits of the APIC id distingush the cores. + * Assumes number of cores is a power of two. + */ +static void __cpuinit amd_detect_cmp(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_SMP + unsigned bits; +#ifdef CONFIG_NUMA + int cpu = smp_processor_id(); + int node = 0; + unsigned apicid = hard_smp_processor_id(); +#endif + bits = c->x86_coreid_bits; + + /* Low order bits define the core id (index of core in socket) */ + c->cpu_core_id = c->initial_apicid & ((1 << bits)-1); + /* Convert the initial APIC ID into the socket ID */ + c->phys_proc_id = c->initial_apicid >> bits; + +#ifdef CONFIG_NUMA + node = c->phys_proc_id; + if (apicid_to_node[apicid] != NUMA_NO_NODE) + node = apicid_to_node[apicid]; + if (!node_online(node)) { + /* Two possibilities here: + - The CPU is missing memory and no node was created. + In that case try picking one from a nearby CPU + - The APIC IDs differ from the HyperTransport node IDs + which the K8 northbridge parsing fills in. + Assume they are all increased by a constant offset, + but in the same order as the HT nodeids. + If that doesn't result in a usable node fall back to the + path for the previous case. */ + + int ht_nodeid = c->initial_apicid; + + if (ht_nodeid >= 0 && + apicid_to_node[ht_nodeid] != NUMA_NO_NODE) + node = apicid_to_node[ht_nodeid]; + /* Pick a nearby node */ + if (!node_online(node)) + node = nearby_node(apicid); + } + numa_set_node(cpu, node); + + printk(KERN_INFO "CPU %d/%x -> Node %d\n", cpu, apicid, node); +#endif +#endif +} + +static void __cpuinit early_init_amd_mc(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_SMP + unsigned bits, ecx; + + /* Multi core CPU? */ + if (c->extended_cpuid_level < 0x80000008) + return; + + ecx = cpuid_ecx(0x80000008); + + c->x86_max_cores = (ecx & 0xff) + 1; + + /* CPU telling us the core id bits shift? */ + bits = (ecx >> 12) & 0xF; + + /* Otherwise recompute */ + if (bits == 0) { + while ((1 << bits) < c->x86_max_cores) + bits++; + } + + c->x86_coreid_bits = bits; + +#endif +} + +static void __cpuinit early_init_amd(struct cpuinfo_x86 *c) +{ + early_init_amd_mc(c); + + /* c->x86_power is 8000_0007 edx. Bit 8 is constant TSC */ + if (c->x86_power & (1<<8)) + set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); +} + +static void __cpuinit init_amd(struct cpuinfo_x86 *c) +{ + unsigned level; + +#ifdef CONFIG_SMP + unsigned long value; + + /* + * Disable TLB flush filter by setting HWCR.FFDIS on K8 + * bit 6 of msr C001_0015 + * + * Errata 63 for SH-B3 steppings + * Errata 122 for all steppings (F+ have it disabled by default) + */ + if (c->x86 == 0xf) { + rdmsrl(MSR_K8_HWCR, value); + value |= 1 << 6; + wrmsrl(MSR_K8_HWCR, value); + } +#endif + + /* Bit 31 in normal CPUID used for nonstandard 3DNow ID; + 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway */ + clear_cpu_cap(c, 0*32+31); + + /* On C+ stepping K8 rep microcode works well for copy/memset */ + if (c->x86 == 0xf) { + level = cpuid_eax(1); + if((level >= 0x0f48 && level < 0x0f50) || level >= 0x0f58) + set_cpu_cap(c, X86_FEATURE_REP_GOOD); + } + if (c->x86 == 0x10 || c->x86 == 0x11) + set_cpu_cap(c, X86_FEATURE_REP_GOOD); + + /* Enable workaround for FXSAVE leak */ + if (c->x86 >= 6) + set_cpu_cap(c, X86_FEATURE_FXSAVE_LEAK); + + level = get_model_name(c); + if (!level) { + switch (c->x86) { + case 0xf: + /* Should distinguish Models here, but this is only + a fallback anyways. */ + strcpy(c->x86_model_id, "Hammer"); + break; + } + } + display_cacheinfo(c); + + /* Multi core CPU? */ + if (c->extended_cpuid_level >= 0x80000008) + amd_detect_cmp(c); + + if (c->extended_cpuid_level >= 0x80000006 && + (cpuid_edx(0x80000006) & 0xf000)) + num_cache_leaves = 4; + else + num_cache_leaves = 3; + + if (c->x86 >= 0xf && c->x86 <= 0x11) + set_cpu_cap(c, X86_FEATURE_K8); + + /* MFENCE stops RDTSC speculation */ + set_cpu_cap(c, X86_FEATURE_MFENCE_RDTSC); + + if (c->x86 == 0x10) { + /* do this for boot cpu */ + if (c == &boot_cpu_data) + check_enable_amd_mmconf_dmi(); + + fam10h_check_enable_mmcfg(); + } + + if (c == &boot_cpu_data && c->x86 >= 0xf && c->x86 <= 0x11) { + unsigned long long tseg; + + /* + * Split up direct mapping around the TSEG SMM area. + * Don't do it for gbpages because there seems very little + * benefit in doing so. + */ + if (!rdmsrl_safe(MSR_K8_TSEG_ADDR, &tseg)) { + 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)))) + set_memory_4k((unsigned long)__va(tseg), 1); + } + } +} + +static struct cpu_dev amd_cpu_dev __cpuinitdata = { + .c_vendor = "AMD", + .c_ident = { "AuthenticAMD" }, + .c_early_init = early_init_amd, + .c_init = init_amd, +}; + +cpu_vendor_dev_register(X86_VENDOR_AMD, &amd_cpu_dev); + diff --git a/arch/x86/kernel/cpu/bugs.c b/arch/x86/kernel/cpu/bugs.c index 170d2f5523b..1b1c56bb338 100644 --- a/arch/x86/kernel/cpu/bugs.c +++ b/arch/x86/kernel/cpu/bugs.c @@ -59,8 +59,12 @@ static void __init check_fpu(void) return; } -/* trap_init() enabled FXSR and company _before_ testing for FP problems here. */ - /* Test for the divl bug.. */ + /* + * trap_init() enabled FXSR and company _before_ testing for FP + * problems here. + * + * Test for the divl bug.. + */ __asm__("fninit\n\t" "fldl %1\n\t" "fdivl %2\n\t" @@ -108,10 +112,15 @@ static void __init check_popad(void) "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( "Buggy.\n" ); - else printk( "OK.\n" ); + : "ecx", "edi"); + /* + * If this fails, it means that any user program may lock the + * CPU hard. Too bad. + */ + if (res != 12345678) + printk("Buggy.\n"); + else + printk("OK.\n"); #endif } @@ -137,7 +146,8 @@ static void __init check_config(void) * 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 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 @@ -170,6 +180,7 @@ void __init check_bugs(void) check_fpu(); check_hlt(); check_popad(); - init_utsname()->machine[1] = '0' + (boot_cpu_data.x86 > 6 ? 6 : boot_cpu_data.x86); + init_utsname()->machine[1] = + '0' + (boot_cpu_data.x86 > 6 ? 6 : boot_cpu_data.x86); alternative_instructions(); } diff --git a/arch/x86/kernel/bugs_64.c b/arch/x86/kernel/cpu/bugs_64.c index 9a3ed0649d4..9a3ed0649d4 100644 --- a/arch/x86/kernel/bugs_64.c +++ b/arch/x86/kernel/cpu/bugs_64.c diff --git a/arch/x86/kernel/cpu/centaur_64.c b/arch/x86/kernel/cpu/centaur_64.c new file mode 100644 index 00000000000..1d181c40e2e --- /dev/null +++ b/arch/x86/kernel/cpu/centaur_64.c @@ -0,0 +1,35 @@ +#include <linux/init.h> +#include <linux/smp.h> + +#include <asm/cpufeature.h> +#include <asm/processor.h> + +#include "cpu.h" + +static void __cpuinit early_init_centaur(struct cpuinfo_x86 *c) +{ + if (c->x86 == 0x6 && c->x86_model >= 0xf) + set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); + + set_cpu_cap(c, X86_FEATURE_SYSENTER32); +} + +static void __cpuinit init_centaur(struct cpuinfo_x86 *c) +{ + if (c->x86 == 0x6 && c->x86_model >= 0xf) { + c->x86_cache_alignment = c->x86_clflush_size * 2; + set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); + set_cpu_cap(c, X86_FEATURE_REP_GOOD); + } + set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC); +} + +static struct cpu_dev centaur_cpu_dev __cpuinitdata = { + .c_vendor = "Centaur", + .c_ident = { "CentaurHauls" }, + .c_early_init = early_init_centaur, + .c_init = init_centaur, +}; + +cpu_vendor_dev_register(X86_VENDOR_CENTAUR, ¢aur_cpu_dev); + diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index d0463a94624..80ab20d4fa3 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -427,7 +427,7 @@ __setup("serialnumber", x86_serial_nr_setup); /* * This does the hard work of actually picking apart the CPU stuff... */ -void __cpuinit identify_cpu(struct cpuinfo_x86 *c) +static void __cpuinit identify_cpu(struct cpuinfo_x86 *c) { int i; diff --git a/arch/x86/kernel/cpu/common_64.c b/arch/x86/kernel/cpu/common_64.c new file mode 100644 index 00000000000..7b8cc72feb4 --- /dev/null +++ b/arch/x86/kernel/cpu/common_64.c @@ -0,0 +1,681 @@ +#include <linux/init.h> +#include <linux/kernel.h> +#include <linux/sched.h> +#include <linux/string.h> +#include <linux/bootmem.h> +#include <linux/bitops.h> +#include <linux/module.h> +#include <linux/kgdb.h> +#include <linux/topology.h> +#include <linux/string.h> +#include <linux/delay.h> +#include <linux/smp.h> +#include <linux/module.h> +#include <linux/percpu.h> +#include <asm/processor.h> +#include <asm/i387.h> +#include <asm/msr.h> +#include <asm/io.h> +#include <asm/mmu_context.h> +#include <asm/mtrr.h> +#include <asm/mce.h> +#include <asm/pat.h> +#include <asm/numa.h> +#ifdef CONFIG_X86_LOCAL_APIC +#include <asm/mpspec.h> +#include <asm/apic.h> +#include <mach_apic.h> +#endif +#include <asm/pda.h> +#include <asm/pgtable.h> +#include <asm/processor.h> +#include <asm/desc.h> +#include <asm/atomic.h> +#include <asm/proto.h> +#include <asm/sections.h> +#include <asm/setup.h> +#include <asm/genapic.h> + +#include "cpu.h" + +/* We need valid kernel segments for data and code in long mode too + * IRET will check the segment types kkeil 2000/10/28 + * Also sysret mandates a special GDT layout + */ +/* The TLS descriptors are currently at a different place compared to i386. + Hopefully nobody expects them at a fixed place (Wine?) */ +DEFINE_PER_CPU(struct gdt_page, gdt_page) = { .gdt = { + [GDT_ENTRY_KERNEL32_CS] = { { { 0x0000ffff, 0x00cf9b00 } } }, + [GDT_ENTRY_KERNEL_CS] = { { { 0x0000ffff, 0x00af9b00 } } }, + [GDT_ENTRY_KERNEL_DS] = { { { 0x0000ffff, 0x00cf9300 } } }, + [GDT_ENTRY_DEFAULT_USER32_CS] = { { { 0x0000ffff, 0x00cffb00 } } }, + [GDT_ENTRY_DEFAULT_USER_DS] = { { { 0x0000ffff, 0x00cff300 } } }, + [GDT_ENTRY_DEFAULT_USER_CS] = { { { 0x0000ffff, 0x00affb00 } } }, +} }; +EXPORT_PER_CPU_SYMBOL_GPL(gdt_page); + +__u32 cleared_cpu_caps[NCAPINTS] __cpuinitdata; + +/* Current gdt points %fs at the "master" per-cpu area: after this, + * it's on the real one. */ +void switch_to_new_gdt(void) +{ + struct desc_ptr gdt_descr; + + gdt_descr.address = (long)get_cpu_gdt_table(smp_processor_id()); + gdt_descr.size = GDT_SIZE - 1; + load_gdt(&gdt_descr); +} + +struct cpu_dev *cpu_devs[X86_VENDOR_NUM] = {}; + +static void __cpuinit default_init(struct cpuinfo_x86 *c) +{ + display_cacheinfo(c); +} + +static struct cpu_dev __cpuinitdata default_cpu = { + .c_init = default_init, + .c_vendor = "Unknown", +}; +static struct cpu_dev *this_cpu __cpuinitdata = &default_cpu; + +int __cpuinit get_model_name(struct cpuinfo_x86 *c) +{ + unsigned int *v; + + if (c->extended_cpuid_level < 0x80000004) + return 0; + + v = (unsigned int *) c->x86_model_id; + cpuid(0x80000002, &v[0], &v[1], &v[2], &v[3]); + cpuid(0x80000003, &v[4], &v[5], &v[6], &v[7]); + cpuid(0x80000004, &v[8], &v[9], &v[10], &v[11]); + c->x86_model_id[48] = 0; + return 1; +} + + +void __cpuinit display_cacheinfo(struct cpuinfo_x86 *c) +{ + unsigned int n, dummy, ebx, ecx, edx; + + n = c->extended_cpuid_level; + + if (n >= 0x80000005) { + cpuid(0x80000005, &dummy, &ebx, &ecx, &edx); + printk(KERN_INFO "CPU: L1 I Cache: %dK (%d bytes/line), " + "D cache %dK (%d bytes/line)\n", + edx>>24, edx&0xFF, ecx>>24, ecx&0xFF); + c->x86_cache_size = (ecx>>24) + (edx>>24); + /* On K8 L1 TLB is inclusive, so don't count it */ + c->x86_tlbsize = 0; + } + + if (n >= 0x80000006) { + cpuid(0x80000006, &dummy, &ebx, &ecx, &edx); + ecx = cpuid_ecx(0x80000006); + c->x86_cache_size = ecx >> 16; + c->x86_tlbsize += ((ebx >> 16) & 0xfff) + (ebx & 0xfff); + + printk(KERN_INFO "CPU: L2 Cache: %dK (%d bytes/line)\n", + c->x86_cache_size, ecx & 0xFF); + } +} + +void __cpuinit detect_ht(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_SMP + u32 eax, ebx, ecx, edx; + int index_msb, core_bits; + + cpuid(1, &eax, &ebx, &ecx, &edx); + + + if (!cpu_has(c, X86_FEATURE_HT)) + return; + if (cpu_has(c, X86_FEATURE_CMP_LEGACY)) + goto out; + + smp_num_siblings = (ebx & 0xff0000) >> 16; + + if (smp_num_siblings == 1) { + printk(KERN_INFO "CPU: Hyper-Threading is disabled\n"); + } else if (smp_num_siblings > 1) { + + if (smp_num_siblings > NR_CPUS) { + printk(KERN_WARNING "CPU: Unsupported number of " + "siblings %d", smp_num_siblings); + smp_num_siblings = 1; + return; + } + + index_msb = get_count_order(smp_num_siblings); + c->phys_proc_id = phys_pkg_id(index_msb); + + smp_num_siblings = smp_num_siblings / c->x86_max_cores; + + index_msb = get_count_order(smp_num_siblings); + + core_bits = get_count_order(c->x86_max_cores); + + c->cpu_core_id = phys_pkg_id(index_msb) & + ((1 << core_bits) - 1); + } +out: + if ((c->x86_max_cores * smp_num_siblings) > 1) { + printk(KERN_INFO "CPU: Physical Processor ID: %d\n", + c->phys_proc_id); + printk(KERN_INFO "CPU: Processor Core ID: %d\n", + c->cpu_core_id); + } + +#endif +} + +static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c) +{ + char *v = c->x86_vendor_id; + int i; + static int printed; + + for (i = 0; i < X86_VENDOR_NUM; i++) { + if (cpu_devs[i]) { + if (!strcmp(v, cpu_devs[i]->c_ident[0]) || + (cpu_devs[i]->c_ident[1] && + !strcmp(v, cpu_devs[i]->c_ident[1]))) { + c->x86_vendor = i; + this_cpu = cpu_devs[i]; + return; + } + } + } + if (!printed) { + printed++; + printk(KERN_ERR "CPU: Vendor unknown, using generic init.\n"); + printk(KERN_ERR "CPU: Your system may be unstable.\n"); + } + c->x86_vendor = X86_VENDOR_UNKNOWN; +} + +static void __init early_cpu_support_print(void) +{ + int i,j; + struct cpu_dev *cpu_devx; + + printk("KERNEL supported cpus:\n"); + for (i = 0; i < X86_VENDOR_NUM; i++) { + cpu_devx = cpu_devs[i]; + if (!cpu_devx) + continue; + for (j = 0; j < 2; j++) { + if (!cpu_devx->c_ident[j]) + continue; + printk(" %s %s\n", cpu_devx->c_vendor, + cpu_devx->c_ident[j]); + } + } +} + +static void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c); + +void __init early_cpu_init(void) +{ + struct cpu_vendor_dev *cvdev; + + for (cvdev = __x86cpuvendor_start ; + cvdev < __x86cpuvendor_end ; + cvdev++) + cpu_devs[cvdev->vendor] = cvdev->cpu_dev; + early_cpu_support_print(); + early_identify_cpu(&boot_cpu_data); +} + +/* Do some early cpuid on the boot CPU to get some parameter that are + needed before check_bugs. Everything advanced is in identify_cpu + below. */ +static void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c) +{ + u32 tfms, xlvl; + + c->loops_per_jiffy = loops_per_jiffy; + c->x86_cache_size = -1; + c->x86_vendor = X86_VENDOR_UNKNOWN; + c->x86_model = c->x86_mask = 0; /* So far unknown... */ + c->x86_vendor_id[0] = '\0'; /* Unset */ + c->x86_model_id[0] = '\0'; /* Unset */ + c->x86_clflush_size = 64; + c->x86_cache_alignment = c->x86_clflush_size; + c->x86_max_cores = 1; + c->x86_coreid_bits = 0; + c->extended_cpuid_level = 0; + memset(&c->x86_capability, 0, sizeof c->x86_capability); + + /* Get vendor name */ + cpuid(0x00000000, (unsigned int *)&c->cpuid_level, + (unsigned int *)&c->x86_vendor_id[0], + (unsigned int *)&c->x86_vendor_id[8], + (unsigned int *)&c->x86_vendor_id[4]); + + get_cpu_vendor(c); + + /* Initialize the standard set of capabilities */ + /* Note that the vendor-specific code below might override */ + + /* Intel-defined flags: level 0x00000001 */ + if (c->cpuid_level >= 0x00000001) { + __u32 misc; + cpuid(0x00000001, &tfms, &misc, &c->x86_capability[4], + &c->x86_capability[0]); + c->x86 = (tfms >> 8) & 0xf; + c->x86_model = (tfms >> 4) & 0xf; + c->x86_mask = tfms & 0xf; + if (c->x86 == 0xf) + c->x86 += (tfms >> 20) & 0xff; + if (c->x86 >= 0x6) + c->x86_model += ((tfms >> 16) & 0xF) << 4; + if (test_cpu_cap(c, X86_FEATURE_CLFLSH)) + c->x86_clflush_size = ((misc >> 8) & 0xff) * 8; + } else { + /* Have CPUID level 0 only - unheard of */ + c->x86 = 4; + } + + c->initial_apicid = (cpuid_ebx(1) >> 24) & 0xff; +#ifdef CONFIG_SMP + c->phys_proc_id = c->initial_apicid; +#endif + /* AMD-defined flags: level 0x80000001 */ + xlvl = cpuid_eax(0x80000000); + c->extended_cpuid_level = xlvl; + if ((xlvl & 0xffff0000) == 0x80000000) { + if (xlvl >= 0x80000001) { + c->x86_capability[1] = cpuid_edx(0x80000001); + c->x86_capability[6] = cpuid_ecx(0x80000001); + } + if (xlvl >= 0x80000004) + get_model_name(c); /* Default name */ + } + + /* Transmeta-defined flags: level 0x80860001 */ + xlvl = cpuid_eax(0x80860000); + if ((xlvl & 0xffff0000) == 0x80860000) { + /* Don't set x86_cpuid_level here for now to not confuse. */ + if (xlvl >= 0x80860001) + c->x86_capability[2] = cpuid_edx(0x80860001); + } + + c->extended_cpuid_level = cpuid_eax(0x80000000); + if (c->extended_cpuid_level >= 0x80000007) + c->x86_power = cpuid_edx(0x80000007); + + if (c->extended_cpuid_level >= 0x80000008) { + u32 eax = cpuid_eax(0x80000008); + + c->x86_virt_bits = (eax >> 8) & 0xff; + c->x86_phys_bits = eax & 0xff; + } + + /* Assume all 64-bit CPUs support 32-bit syscall */ + set_cpu_cap(c, X86_FEATURE_SYSCALL32); + + if (c->x86_vendor != X86_VENDOR_UNKNOWN && + cpu_devs[c->x86_vendor]->c_early_init) + cpu_devs[c->x86_vendor]->c_early_init(c); + + validate_pat_support(c); + + /* early_param could clear that, but recall get it set again */ + if (disable_apic) + clear_cpu_cap(c, X86_FEATURE_APIC); +} + +/* + * This does the hard work of actually picking apart the CPU stuff... + */ +static void __cpuinit identify_cpu(struct cpuinfo_x86 *c) +{ + int i; + + early_identify_cpu(c); + + init_scattered_cpuid_features(c); + + c->apicid = phys_pkg_id(0); + + /* + * Vendor-specific initialization. In this section we + * canonicalize the feature flags, meaning if there are + * features a certain CPU supports which CPUID doesn't + * tell us, CPUID claiming incorrect flags, or other bugs, + * we handle them here. + * + * At the end of this section, c->x86_capability better + * indicate the features this CPU genuinely supports! + */ + if (this_cpu->c_init) + this_cpu->c_init(c); + + detect_ht(c); + + /* + * On SMP, boot_cpu_data holds the common feature set between + * all CPUs; so make sure that we indicate which features are + * common between the CPUs. The first time this routine gets + * executed, c == &boot_cpu_data. + */ + if (c != &boot_cpu_data) { + /* AND the already accumulated flags with these */ + for (i = 0; i < NCAPINTS; i++) + boot_cpu_data.x86_capability[i] &= c->x86_capability[i]; + } + + /* Clear all flags overriden by options */ + for (i = 0; i < NCAPINTS; i++) + c->x86_capability[i] &= ~cleared_cpu_caps[i]; + +#ifdef CONFIG_X86_MCE + mcheck_init(c); +#endif + select_idle_routine(c); + +#ifdef CONFIG_NUMA + numa_add_cpu(smp_processor_id()); +#endif + +} + +void __cpuinit identify_boot_cpu(void) +{ + identify_cpu(&boot_cpu_data); +} + +void __cpuinit identify_secondary_cpu(struct cpuinfo_x86 *c) +{ + BUG_ON(c == &boot_cpu_data); + identify_cpu(c); + mtrr_ap_init(); +} + +static __init int setup_noclflush(char *arg) +{ + setup_clear_cpu_cap(X86_FEATURE_CLFLSH); + return 1; +} +__setup("noclflush", setup_noclflush); + +void __cpuinit print_cpu_info(struct cpuinfo_x86 *c) +{ + if (c->x86_model_id[0]) + printk(KERN_CONT "%s", c->x86_model_id); + + if (c->x86_mask || c->cpuid_level >= 0) + printk(KERN_CONT " stepping %02x\n", c->x86_mask); + else + printk(KERN_CONT "\n"); +} + +static __init int setup_disablecpuid(char *arg) +{ + int bit; + if (get_option(&arg, &bit) && bit < NCAPINTS*32) + setup_clear_cpu_cap(bit); + else + return 0; + return 1; +} +__setup("clearcpuid=", setup_disablecpuid); + +cpumask_t cpu_initialized __cpuinitdata = CPU_MASK_NONE; + +struct x8664_pda **_cpu_pda __read_mostly; +EXPORT_SYMBOL(_cpu_pda); + +struct desc_ptr idt_descr = { 256 * 16 - 1, (unsigned long) idt_table }; + +char boot_cpu_stack[IRQSTACKSIZE] __page_aligned_bss; + +unsigned long __supported_pte_mask __read_mostly = ~0UL; +EXPORT_SYMBOL_GPL(__supported_pte_mask); + +static int do_not_nx __cpuinitdata; + +/* noexec=on|off +Control non executable mappings for 64bit processes. + +on Enable(default) +off Disable +*/ +static int __init nonx_setup(char *str) +{ + if (!str) + return -EINVAL; + if (!strncmp(str, "on", 2)) { + __supported_pte_mask |= _PAGE_NX; + do_not_nx = 0; + } else if (!strncmp(str, "off", 3)) { + do_not_nx = 1; + __supported_pte_mask &= ~_PAGE_NX; + } + return 0; +} +early_param("noexec", nonx_setup); + +int force_personality32; + +/* noexec32=on|off +Control non executable heap for 32bit processes. +To control the stack too use noexec=off + +on PROT_READ does not imply PROT_EXEC for 32bit processes (default) +off PROT_READ implies PROT_EXEC +*/ +static int __init nonx32_setup(char *str) +{ + if (!strcmp(str, "on")) + force_personality32 &= ~READ_IMPLIES_EXEC; + else if (!strcmp(str, "off")) + force_personality32 |= READ_IMPLIES_EXEC; + return 1; +} +__setup("noexec32=", nonx32_setup); + +void pda_init(int cpu) +{ + struct x8664_pda *pda = cpu_pda(cpu); + + /* Setup up data that may be needed in __get_free_pages early */ + loadsegment(fs, 0); + loadsegment(gs, 0); + /* Memory clobbers used to order PDA accessed */ + mb(); + wrmsrl(MSR_GS_BASE, pda); + mb(); + + pda->cpunumber = cpu; + pda->irqcount = -1; + pda->kernelstack = (unsigned long)stack_thread_info() - + PDA_STACKOFFSET + THREAD_SIZE; + pda->active_mm = &init_mm; + pda->mmu_state = 0; + + if (cpu == 0) { + /* others are initialized in smpboot.c */ + pda->pcurrent = &init_task; + pda->irqstackptr = boot_cpu_stack; + } else { + pda->irqstackptr = (char *) + __get_free_pages(GFP_ATOMIC, IRQSTACK_ORDER); + if (!pda->irqstackptr) + panic("cannot allocate irqstack for cpu %d", cpu); + + if (pda->nodenumber == 0 && cpu_to_node(cpu) != NUMA_NO_NODE) + pda->nodenumber = cpu_to_node(cpu); + } + + pda->irqstackptr += IRQSTACKSIZE-64; +} + +char boot_exception_stacks[(N_EXCEPTION_STACKS - 1) * EXCEPTION_STKSZ + + DEBUG_STKSZ] +__attribute__((section(".bss.page_aligned"))); + +extern asmlinkage void ignore_sysret(void); + +/* May not be marked __init: used by software suspend */ +void syscall_init(void) +{ + /* + * LSTAR and STAR live in a bit strange symbiosis. + * They both write to the same internal register. STAR allows to + * set CS/DS but only a 32bit target. LSTAR sets the 64bit rip. + */ + wrmsrl(MSR_STAR, ((u64)__USER32_CS)<<48 | ((u64)__KERNEL_CS)<<32); + wrmsrl(MSR_LSTAR, system_call); + wrmsrl(MSR_CSTAR, ignore_sysret); + +#ifdef CONFIG_IA32_EMULATION + syscall32_cpu_init(); +#endif + + /* Flags to clear on syscall */ + wrmsrl(MSR_SYSCALL_MASK, + X86_EFLAGS_TF|X86_EFLAGS_DF|X86_EFLAGS_IF|X86_EFLAGS_IOPL); +} + +void __cpuinit check_efer(void) +{ + unsigned long efer; + + rdmsrl(MSR_EFER, efer); + if (!(efer & EFER_NX) || do_not_nx) + __supported_pte_mask &= ~_PAGE_NX; +} + +unsigned long kernel_eflags; + +/* + * Copies of the original ist values from the tss are only accessed during + * debugging, no special alignment required. + */ +DEFINE_PER_CPU(struct orig_ist, orig_ist); + +/* + * 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 + * 'CPU state barrier', nothing should get across. + * A lot of state is already set up in PDA init. + */ +void __cpuinit cpu_init(void) +{ + int cpu = stack_smp_processor_id(); + struct tss_struct *t = &per_cpu(init_tss, cpu); + struct orig_ist *orig_ist = &per_cpu(orig_ist, cpu); + unsigned long v; + char *estacks = NULL; + struct task_struct *me; + int i; + + /* CPU 0 is initialised in head64.c */ + if (cpu != 0) + pda_init(cpu); + else + estacks = boot_exception_stacks; + + me = current; + + if (cpu_test_and_set(cpu, cpu_initialized)) + panic("CPU#%d already initialized!\n", cpu); + + printk(KERN_INFO "Initializing CPU#%d\n", cpu); + + clear_in_cr4(X86_CR4_VME|X86_CR4_PVI|X86_CR4_TSD|X86_CR4_DE); + + /* + * Initialize the per-CPU GDT with the boot GDT, + * and set up the GDT descriptor: + */ + + switch_to_new_gdt(); + load_idt((const struct desc_ptr *)&idt_descr); + + memset(me->thread.tls_array, 0, GDT_ENTRY_TLS_ENTRIES * 8); + syscall_init(); + + wrmsrl(MSR_FS_BASE, 0); + wrmsrl(MSR_KERNEL_GS_BASE, 0); + barrier(); + + check_efer(); + + /* + * set up and load the per-CPU TSS + */ + for (v = 0; v < N_EXCEPTION_STACKS; v++) { + static const unsigned int order[N_EXCEPTION_STACKS] = { + [0 ... N_EXCEPTION_STACKS - 1] = EXCEPTION_STACK_ORDER, + [DEBUG_STACK - 1] = DEBUG_STACK_ORDER + }; + if (cpu) { + estacks = (char *)__get_free_pages(GFP_ATOMIC, order[v]); + if (!estacks) + panic("Cannot allocate exception stack %ld %d\n", + v, cpu); + } + estacks += PAGE_SIZE << order[v]; + orig_ist->ist[v] = t->x86_tss.ist[v] = (unsigned long)estacks; + } + + t->x86_tss.io_bitmap_base = offsetof(struct tss_struct, io_bitmap); + /* + * <= is required because the CPU will access up to + * 8 bits beyond the end of the IO permission bitmap. + */ + for (i = 0; i <= IO_BITMAP_LONGS; i++) + t->io_bitmap[i] = ~0UL; + + atomic_inc(&init_mm.mm_count); + me->active_mm = &init_mm; + if (me->mm) + BUG(); + enter_lazy_tlb(&init_mm, me); + + load_sp0(t, ¤t->thread); + set_tss_desc(cpu, t); + load_TR_desc(); + load_LDT(&init_mm.context); + +#ifdef CONFIG_KGDB + /* + * If the kgdb is connected no debug regs should be altered. This + * is only applicable when KGDB and a KGDB I/O module are built + * into the kernel and you are using early debugging with + * kgdbwait. KGDB will control the kernel HW breakpoint registers. + */ + if (kgdb_connected && arch_kgdb_ops.correct_hw_break) + arch_kgdb_ops.correct_hw_break(); + else { +#endif + /* + * Clear all 6 debug registers: + */ + + set_debugreg(0UL, 0); + set_debugreg(0UL, 1); + set_debugreg(0UL, 2); + set_debugreg(0UL, 3); + set_debugreg(0UL, 6); + set_debugreg(0UL, 7); +#ifdef CONFIG_KGDB + /* If the kgdb is connected no debug regs should be altered. */ + } +#endif + + fpu_init(); + + raw_local_save_flags(kernel_eflags); + + if (is_uv_system()) + uv_cpu_init(); +} diff --git a/arch/x86/kernel/cpu/cpu.h b/arch/x86/kernel/cpu/cpu.h index 783691b2a73..4d894e8565f 100644 --- a/arch/x86/kernel/cpu/cpu.h +++ b/arch/x86/kernel/cpu/cpu.h @@ -1,3 +1,6 @@ +#ifndef ARCH_X86_CPU_H + +#define ARCH_X86_CPU_H struct cpu_model_info { int vendor; @@ -36,3 +39,5 @@ extern struct cpu_vendor_dev __x86cpuvendor_start[], __x86cpuvendor_end[]; extern int get_model_name(struct cpuinfo_x86 *c); extern void display_cacheinfo(struct cpuinfo_x86 *c); + +#endif diff --git a/arch/x86/kernel/cpu/cpufreq/cpufreq-nforce2.c b/arch/x86/kernel/cpu/cpufreq/cpufreq-nforce2.c index f03e9153618..965ea52767a 100644 --- a/arch/x86/kernel/cpu/cpufreq/cpufreq-nforce2.c +++ b/arch/x86/kernel/cpu/cpufreq/cpufreq-nforce2.c @@ -26,9 +26,10 @@ #define NFORCE2_SAFE_DISTANCE 50 /* Delay in ms between FSB changes */ -//#define NFORCE2_DELAY 10 +/* #define NFORCE2_DELAY 10 */ -/* nforce2_chipset: +/* + * nforce2_chipset: * FSB is changed using the chipset */ static struct pci_dev *nforce2_chipset_dev; @@ -36,13 +37,13 @@ static struct pci_dev *nforce2_chipset_dev; /* fid: * multiplier * 10 */ -static int fid = 0; +static int fid; /* min_fsb, max_fsb: * minimum and maximum FSB (= FSB at boot time) */ -static int min_fsb = 0; -static int max_fsb = 0; +static int min_fsb; +static int max_fsb; MODULE_AUTHOR("Sebastian Witt <se.witt@gmx.net>"); MODULE_DESCRIPTION("nForce2 FSB changing cpufreq driver"); @@ -53,7 +54,7 @@ module_param(min_fsb, int, 0444); MODULE_PARM_DESC(fid, "CPU multiplier to use (11.5 = 115)"); MODULE_PARM_DESC(min_fsb, - "Minimum FSB to use, if not defined: current FSB - 50"); + "Minimum FSB to use, if not defined: current FSB - 50"); #define dprintk(msg...) cpufreq_debug_printk(CPUFREQ_DEBUG_DRIVER, "cpufreq-nforce2", msg) @@ -139,7 +140,7 @@ static unsigned int nforce2_fsb_read(int bootfsb) /* Get chipset boot FSB from subdevice 5 (FSB at boot-time) */ nforce2_sub5 = pci_get_subsys(PCI_VENDOR_ID_NVIDIA, - 0x01EF,PCI_ANY_ID,PCI_ANY_ID,NULL); + 0x01EF, PCI_ANY_ID, PCI_ANY_ID, NULL); if (!nforce2_sub5) return 0; @@ -147,13 +148,13 @@ static unsigned int nforce2_fsb_read(int bootfsb) fsb /= 1000000; /* Check if PLL register is already set */ - pci_read_config_byte(nforce2_chipset_dev,NFORCE2_PLLENABLE, (u8 *)&temp); + pci_read_config_byte(nforce2_chipset_dev, NFORCE2_PLLENABLE, (u8 *)&temp); - if(bootfsb || !temp) + if (bootfsb || !temp) return fsb; - + /* Use PLL register FSB value */ - pci_read_config_dword(nforce2_chipset_dev,NFORCE2_PLLREG, &temp); + pci_read_config_dword(nforce2_chipset_dev, NFORCE2_PLLREG, &temp); fsb = nforce2_calc_fsb(temp); return fsb; @@ -184,7 +185,7 @@ static int nforce2_set_fsb(unsigned int fsb) } /* First write? Then set actual value */ - pci_read_config_byte(nforce2_chipset_dev,NFORCE2_PLLENABLE, (u8 *)&temp); + pci_read_config_byte(nforce2_chipset_dev, NFORCE2_PLLENABLE, (u8 *)&temp); if (!temp) { pll = nforce2_calc_pll(tfsb); @@ -210,7 +211,8 @@ static int nforce2_set_fsb(unsigned int fsb) tfsb--; /* Calculate the PLL reg. value */ - if ((pll = nforce2_calc_pll(tfsb)) == -1) + pll = nforce2_calc_pll(tfsb); + if (pll == -1) return -EINVAL; nforce2_write_pll(pll); @@ -249,7 +251,7 @@ static unsigned int nforce2_get(unsigned int cpu) static int nforce2_target(struct cpufreq_policy *policy, unsigned int target_freq, unsigned int relation) { -// unsigned long flags; +/* unsigned long flags; */ struct cpufreq_freqs freqs; unsigned int target_fsb; @@ -271,17 +273,17 @@ static int nforce2_target(struct cpufreq_policy *policy, cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE); /* Disable IRQs */ - //local_irq_save(flags); + /* local_irq_save(flags); */ if (nforce2_set_fsb(target_fsb) < 0) printk(KERN_ERR "cpufreq: Changing FSB to %d failed\n", - target_fsb); + target_fsb); else dprintk("Changed FSB successfully to %d\n", - target_fsb); + target_fsb); /* Enable IRQs */ - //local_irq_restore(flags); + /* local_irq_restore(flags); */ cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE); @@ -302,8 +304,8 @@ static int nforce2_verify(struct cpufreq_policy *policy) policy->max = (fsb_pol_max + 1) * fid * 100; cpufreq_verify_within_limits(policy, - policy->cpuinfo.min_freq, - policy->cpuinfo.max_freq); + policy->cpuinfo.min_freq, + policy->cpuinfo.max_freq); return 0; } @@ -347,7 +349,7 @@ static int nforce2_cpu_init(struct cpufreq_policy *policy) /* Set maximum FSB to FSB at boot time */ max_fsb = nforce2_fsb_read(1); - if(!max_fsb) + if (!max_fsb) return -EIO; if (!min_fsb) diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c index fe9224c51d3..70609efdf1d 100644 --- a/arch/x86/kernel/cpu/intel.c +++ b/arch/x86/kernel/cpu/intel.c @@ -226,6 +226,10 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c) if (cpu_has_bts) ds_init_intel(c); + +#ifdef CONFIG_X86_NUMAQ + numaq_tsc_disable(); +#endif } static unsigned int __cpuinit intel_size_cache(struct cpuinfo_x86 *c, unsigned int size) diff --git a/arch/x86/kernel/cpu/intel_64.c b/arch/x86/kernel/cpu/intel_64.c new file mode 100644 index 00000000000..1019c58d39f --- /dev/null +++ b/arch/x86/kernel/cpu/intel_64.c @@ -0,0 +1,95 @@ +#include <linux/init.h> +#include <linux/smp.h> +#include <asm/processor.h> +#include <asm/ptrace.h> +#include <asm/topology.h> +#include <asm/numa_64.h> + +#include "cpu.h" + +static void __cpuinit early_init_intel(struct cpuinfo_x86 *c) +{ + if ((c->x86 == 0xf && c->x86_model >= 0x03) || + (c->x86 == 0x6 && c->x86_model >= 0x0e)) + set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); + + set_cpu_cap(c, X86_FEATURE_SYSENTER32); +} + +/* + * find out the number of processor cores on the die + */ +static int __cpuinit intel_num_cpu_cores(struct cpuinfo_x86 *c) +{ + unsigned int eax, t; + + if (c->cpuid_level < 4) + return 1; + + cpuid_count(4, 0, &eax, &t, &t, &t); + + if (eax & 0x1f) + return ((eax >> 26) + 1); + else + return 1; +} + +static void __cpuinit srat_detect_node(void) +{ +#ifdef CONFIG_NUMA + unsigned node; + int cpu = smp_processor_id(); + int apicid = hard_smp_processor_id(); + + /* Don't do the funky fallback heuristics the AMD version employs + for now. */ + node = apicid_to_node[apicid]; + if (node == NUMA_NO_NODE || !node_online(node)) + node = first_node(node_online_map); + numa_set_node(cpu, node); + + printk(KERN_INFO "CPU %d/%x -> Node %d\n", cpu, apicid, node); +#endif +} + +static void __cpuinit init_intel(struct cpuinfo_x86 *c) +{ + init_intel_cacheinfo(c); + if (c->cpuid_level > 9) { + unsigned eax = cpuid_eax(10); + /* Check for version and the number of counters */ + if ((eax & 0xff) && (((eax>>8) & 0xff) > 1)) + set_cpu_cap(c, X86_FEATURE_ARCH_PERFMON); + } + + if (cpu_has_ds) { + unsigned int l1, l2; + rdmsr(MSR_IA32_MISC_ENABLE, l1, l2); + if (!(l1 & (1<<11))) + set_cpu_cap(c, X86_FEATURE_BTS); + if (!(l1 & (1<<12))) + set_cpu_cap(c, X86_FEATURE_PEBS); + } + + + if (cpu_has_bts) + ds_init_intel(c); + + if (c->x86 == 15) + c->x86_cache_alignment = c->x86_clflush_size * 2; + if (c->x86 == 6) + set_cpu_cap(c, X86_FEATURE_REP_GOOD); + set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC); + c->x86_max_cores = intel_num_cpu_cores(c); + + srat_detect_node(); +} + +static struct cpu_dev intel_cpu_dev __cpuinitdata = { + .c_vendor = "Intel", + .c_ident = { "GenuineIntel" }, + .c_early_init = early_init_intel, + .c_init = init_intel, +}; +cpu_vendor_dev_register(X86_VENDOR_INTEL, &intel_cpu_dev); + diff --git a/arch/x86/kernel/cpu/intel_cacheinfo.c b/arch/x86/kernel/cpu/intel_cacheinfo.c index 26d615dcb14..2c8afafa18e 100644 --- a/arch/x86/kernel/cpu/intel_cacheinfo.c +++ b/arch/x86/kernel/cpu/intel_cacheinfo.c @@ -62,6 +62,7 @@ static struct _cache_table cache_table[] __cpuinitdata = { 0x4b, LVL_3, 8192 }, /* 16-way set assoc, 64 byte line size */ { 0x4c, LVL_3, 12288 }, /* 12-way set assoc, 64 byte line size */ { 0x4d, LVL_3, 16384 }, /* 16-way set assoc, 64 byte line size */ + { 0x4e, LVL_2, 6144 }, /* 24-way set assoc, 64 byte line size */ { 0x60, LVL_1_DATA, 16 }, /* 8-way set assoc, sectored cache, 64 byte line size */ { 0x66, LVL_1_DATA, 8 }, /* 4-way set assoc, sectored cache, 64 byte line size */ { 0x67, LVL_1_DATA, 16 }, /* 4-way set assoc, sectored cache, 64 byte line size */ diff --git a/arch/x86/kernel/cpu/mcheck/k7.c b/arch/x86/kernel/cpu/mcheck/k7.c index e633c9c2b76..f390c9f6635 100644 --- a/arch/x86/kernel/cpu/mcheck/k7.c +++ b/arch/x86/kernel/cpu/mcheck/k7.c @@ -9,23 +9,23 @@ #include <linux/interrupt.h> #include <linux/smp.h> -#include <asm/processor.h> +#include <asm/processor.h> #include <asm/system.h> #include <asm/msr.h> #include "mce.h" /* Machine Check Handler For AMD Athlon/Duron */ -static void k7_machine_check(struct pt_regs * regs, long error_code) +static void k7_machine_check(struct pt_regs *regs, long error_code) { - int recover=1; + int recover = 1; u32 alow, ahigh, high, low; u32 mcgstl, mcgsth; int i; - rdmsr (MSR_IA32_MCG_STATUS, mcgstl, mcgsth); + rdmsr(MSR_IA32_MCG_STATUS, mcgstl, mcgsth); if (mcgstl & (1<<0)) /* Recoverable ? */ - recover=0; + recover = 0; printk(KERN_EMERG "CPU %d: Machine Check Exception: %08x%08x\n", smp_processor_id(), mcgsth, mcgstl); @@ -60,12 +60,12 @@ static void k7_machine_check(struct pt_regs * regs, long error_code) } if (recover&2) - panic ("CPU context corrupt"); + panic("CPU context corrupt"); if (recover&1) - panic ("Unable to continue"); - printk (KERN_EMERG "Attempting to continue.\n"); + panic("Unable to continue"); + printk(KERN_EMERG "Attempting to continue.\n"); mcgstl &= ~(1<<2); - wrmsr (MSR_IA32_MCG_STATUS,mcgstl, mcgsth); + wrmsr(MSR_IA32_MCG_STATUS, mcgstl, mcgsth); } @@ -81,25 +81,25 @@ void amd_mcheck_init(struct cpuinfo_x86 *c) machine_check_vector = k7_machine_check; wmb(); - printk (KERN_INFO "Intel machine check architecture supported.\n"); - rdmsr (MSR_IA32_MCG_CAP, l, h); + printk(KERN_INFO "Intel machine check architecture supported.\n"); + rdmsr(MSR_IA32_MCG_CAP, l, h); if (l & (1<<8)) /* Control register present ? */ - wrmsr (MSR_IA32_MCG_CTL, 0xffffffff, 0xffffffff); + wrmsr(MSR_IA32_MCG_CTL, 0xffffffff, 0xffffffff); nr_mce_banks = l & 0xff; /* Clear status for MC index 0 separately, we don't touch CTL, * as some K7 Athlons cause spurious MCEs when its enabled. */ if (boot_cpu_data.x86 == 6) { - wrmsr (MSR_IA32_MC0_STATUS, 0x0, 0x0); + wrmsr(MSR_IA32_MC0_STATUS, 0x0, 0x0); i = 1; } else i = 0; - for (; i<nr_mce_banks; i++) { - wrmsr (MSR_IA32_MC0_CTL+4*i, 0xffffffff, 0xffffffff); - wrmsr (MSR_IA32_MC0_STATUS+4*i, 0x0, 0x0); + for (; i < nr_mce_banks; i++) { + wrmsr(MSR_IA32_MC0_CTL+4*i, 0xffffffff, 0xffffffff); + wrmsr(MSR_IA32_MC0_STATUS+4*i, 0x0, 0x0); } - set_in_cr4 (X86_CR4_MCE); - printk (KERN_INFO "Intel machine check reporting enabled on CPU#%d.\n", + set_in_cr4(X86_CR4_MCE); + printk(KERN_INFO "Intel machine check reporting enabled on CPU#%d.\n", smp_processor_id()); } diff --git a/arch/x86/kernel/cpu/mcheck/mce_64.c b/arch/x86/kernel/cpu/mcheck/mce_64.c index e07e8c068ae..501ca1cea27 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_64.c +++ b/arch/x86/kernel/cpu/mcheck/mce_64.c @@ -31,7 +31,7 @@ #include <asm/idle.h> #define MISC_MCELOG_MINOR 227 -#define NR_BANKS 6 +#define NR_SYSFS_BANKS 6 atomic_t mce_entry; @@ -46,7 +46,7 @@ static int mce_dont_init; */ static int tolerant = 1; static int banks; -static unsigned long bank[NR_BANKS] = { [0 ... NR_BANKS-1] = ~0UL }; +static unsigned long bank[NR_SYSFS_BANKS] = { [0 ... NR_SYSFS_BANKS-1] = ~0UL }; static unsigned long notify_user; static int rip_msr; static int mce_bootlog = -1; @@ -209,7 +209,7 @@ void do_machine_check(struct pt_regs * regs, long error_code) barrier(); for (i = 0; i < banks; i++) { - if (!bank[i]) + if (i < NR_SYSFS_BANKS && !bank[i]) continue; m.misc = 0; @@ -444,9 +444,10 @@ static void mce_init(void *dummy) rdmsrl(MSR_IA32_MCG_CAP, cap); banks = cap & 0xff; - if (banks > NR_BANKS) { - printk(KERN_INFO "MCE: warning: using only %d banks\n", banks); - banks = NR_BANKS; + if (banks > MCE_EXTENDED_BANK) { + banks = MCE_EXTENDED_BANK; + printk(KERN_INFO "MCE: warning: using only %d banks\n", + MCE_EXTENDED_BANK); } /* Use accurate RIP reporting if available. */ if ((cap & (1<<9)) && ((cap >> 16) & 0xff) >= 9) @@ -462,7 +463,11 @@ static void mce_init(void *dummy) wrmsr(MSR_IA32_MCG_CTL, 0xffffffff, 0xffffffff); for (i = 0; i < banks; i++) { - wrmsrl(MSR_IA32_MC0_CTL+4*i, bank[i]); + if (i < NR_SYSFS_BANKS) + wrmsrl(MSR_IA32_MC0_CTL+4*i, bank[i]); + else + wrmsrl(MSR_IA32_MC0_CTL+4*i, ~0UL); + wrmsrl(MSR_IA32_MC0_STATUS+4*i, 0); } } @@ -766,7 +771,10 @@ DEFINE_PER_CPU(struct sys_device, device_mce); } \ static SYSDEV_ATTR(name, 0644, show_ ## name, set_ ## name); -/* TBD should generate these dynamically based on number of available banks */ +/* + * TBD should generate these dynamically based on number of available banks. + * Have only 6 contol banks in /sysfs until then. + */ ACCESSOR(bank0ctl,bank[0],mce_restart()) ACCESSOR(bank1ctl,bank[1],mce_restart()) ACCESSOR(bank2ctl,bank[2],mce_restart()) diff --git a/arch/x86/kernel/cpu/mcheck/p4.c b/arch/x86/kernel/cpu/mcheck/p4.c index cb03345554a..eef001ad3bd 100644 --- a/arch/x86/kernel/cpu/mcheck/p4.c +++ b/arch/x86/kernel/cpu/mcheck/p4.c @@ -8,7 +8,7 @@ #include <linux/interrupt.h> #include <linux/smp.h> -#include <asm/processor.h> +#include <asm/processor.h> #include <asm/system.h> #include <asm/msr.h> #include <asm/apic.h> @@ -32,12 +32,12 @@ struct intel_mce_extended_msrs { /* u32 *reserved[]; */ }; -static int mce_num_extended_msrs = 0; +static int mce_num_extended_msrs; #ifdef CONFIG_X86_MCE_P4THERMAL static void unexpected_thermal_interrupt(struct pt_regs *regs) -{ +{ printk(KERN_ERR "CPU%d: Unexpected LVT TMR interrupt!\n", smp_processor_id()); add_taint(TAINT_MACHINE_CHECK); @@ -83,7 +83,7 @@ static void intel_init_thermal(struct cpuinfo_x86 *c) * be some SMM goo which handles it, so we can't even put a handler * since it might be delivered via SMI already -zwanem. */ - rdmsr (MSR_IA32_MISC_ENABLE, l, h); + rdmsr(MSR_IA32_MISC_ENABLE, l, h); h = apic_read(APIC_LVTTHMR); if ((l & (1<<3)) && (h & APIC_DM_SMI)) { printk(KERN_DEBUG "CPU%d: Thermal monitoring handled by SMI\n", @@ -91,7 +91,7 @@ static void intel_init_thermal(struct cpuinfo_x86 *c) return; /* -EBUSY */ } - /* check whether a vector already exists, temporarily masked? */ + /* check whether a vector already exists, temporarily masked? */ if (h & APIC_VECTOR_MASK) { printk(KERN_DEBUG "CPU%d: Thermal LVT vector (%#x) already " "installed\n", @@ -104,18 +104,18 @@ static void intel_init_thermal(struct cpuinfo_x86 *c) h |= (APIC_DM_FIXED | APIC_LVT_MASKED); /* we'll mask till we're ready */ apic_write_around(APIC_LVTTHMR, h); - rdmsr (MSR_IA32_THERM_INTERRUPT, l, h); - wrmsr (MSR_IA32_THERM_INTERRUPT, l | 0x03 , h); + rdmsr(MSR_IA32_THERM_INTERRUPT, l, h); + wrmsr(MSR_IA32_THERM_INTERRUPT, l | 0x03 , h); /* ok we're good to go... */ vendor_thermal_interrupt = intel_thermal_interrupt; - - rdmsr (MSR_IA32_MISC_ENABLE, l, h); - wrmsr (MSR_IA32_MISC_ENABLE, l | (1<<3), h); - l = apic_read (APIC_LVTTHMR); - apic_write_around (APIC_LVTTHMR, l & ~APIC_LVT_MASKED); - printk (KERN_INFO "CPU%d: Thermal monitoring enabled\n", cpu); + rdmsr(MSR_IA32_MISC_ENABLE, l, h); + wrmsr(MSR_IA32_MISC_ENABLE, l | (1<<3), h); + + l = apic_read(APIC_LVTTHMR); + apic_write_around(APIC_LVTTHMR, l & ~APIC_LVT_MASKED); + printk(KERN_INFO "CPU%d: Thermal monitoring enabled\n", cpu); /* enable thermal throttle processing */ atomic_set(&therm_throt_en, 1); @@ -129,28 +129,28 @@ static inline void intel_get_extended_msrs(struct intel_mce_extended_msrs *r) { u32 h; - rdmsr (MSR_IA32_MCG_EAX, r->eax, h); - rdmsr (MSR_IA32_MCG_EBX, r->ebx, h); - rdmsr (MSR_IA32_MCG_ECX, r->ecx, h); - rdmsr (MSR_IA32_MCG_EDX, r->edx, h); - rdmsr (MSR_IA32_MCG_ESI, r->esi, h); - rdmsr (MSR_IA32_MCG_EDI, r->edi, h); - rdmsr (MSR_IA32_MCG_EBP, r->ebp, h); - rdmsr (MSR_IA32_MCG_ESP, r->esp, h); - rdmsr (MSR_IA32_MCG_EFLAGS, r->eflags, h); - rdmsr (MSR_IA32_MCG_EIP, r->eip, h); + rdmsr(MSR_IA32_MCG_EAX, r->eax, h); + rdmsr(MSR_IA32_MCG_EBX, r->ebx, h); + rdmsr(MSR_IA32_MCG_ECX, r->ecx, h); + rdmsr(MSR_IA32_MCG_EDX, r->edx, h); + rdmsr(MSR_IA32_MCG_ESI, r->esi, h); + rdmsr(MSR_IA32_MCG_EDI, r->edi, h); + rdmsr(MSR_IA32_MCG_EBP, r->ebp, h); + rdmsr(MSR_IA32_MCG_ESP, r->esp, h); + rdmsr(MSR_IA32_MCG_EFLAGS, r->eflags, h); + rdmsr(MSR_IA32_MCG_EIP, r->eip, h); } -static void intel_machine_check(struct pt_regs * regs, long error_code) +static void intel_machine_check(struct pt_regs *regs, long error_code) { - int recover=1; + int recover = 1; u32 alow, ahigh, high, low; u32 mcgstl, mcgsth; int i; - rdmsr (MSR_IA32_MCG_STATUS, mcgstl, mcgsth); + rdmsr(MSR_IA32_MCG_STATUS, mcgstl, mcgsth); if (mcgstl & (1<<0)) /* Recoverable ? */ - recover=0; + recover = 0; printk(KERN_EMERG "CPU %d: Machine Check Exception: %08x%08x\n", smp_processor_id(), mcgsth, mcgstl); @@ -191,20 +191,20 @@ static void intel_machine_check(struct pt_regs * regs, long error_code) } if (recover & 2) - panic ("CPU context corrupt"); + panic("CPU context corrupt"); if (recover & 1) - panic ("Unable to continue"); + panic("Unable to continue"); printk(KERN_EMERG "Attempting to continue.\n"); - /* - * Do not clear the MSR_IA32_MCi_STATUS if the error is not + /* + * Do not clear the MSR_IA32_MCi_STATUS if the error is not * recoverable/continuable.This will allow BIOS to look at the MSRs * for errors if the OS could not log the error. */ - for (i=0; i<nr_mce_banks; i++) { + for (i = 0; i < nr_mce_banks; i++) { u32 msr; msr = MSR_IA32_MC0_STATUS+i*4; - rdmsr (msr, low, high); + rdmsr(msr, low, high); if (high&(1<<31)) { /* Clear it */ wrmsr(msr, 0UL, 0UL); @@ -214,7 +214,7 @@ static void intel_machine_check(struct pt_regs * regs, long error_code) } } mcgstl &= ~(1<<2); - wrmsr (MSR_IA32_MCG_STATUS,mcgstl, mcgsth); + wrmsr(MSR_IA32_MCG_STATUS, mcgstl, mcgsth); } @@ -222,30 +222,30 @@ void intel_p4_mcheck_init(struct cpuinfo_x86 *c) { u32 l, h; int i; - + machine_check_vector = intel_machine_check; wmb(); - printk (KERN_INFO "Intel machine check architecture supported.\n"); - rdmsr (MSR_IA32_MCG_CAP, l, h); + printk(KERN_INFO "Intel machine check architecture supported.\n"); + rdmsr(MSR_IA32_MCG_CAP, l, h); if (l & (1<<8)) /* Control register present ? */ - wrmsr (MSR_IA32_MCG_CTL, 0xffffffff, 0xffffffff); + wrmsr(MSR_IA32_MCG_CTL, 0xffffffff, 0xffffffff); nr_mce_banks = l & 0xff; - for (i=0; i<nr_mce_banks; i++) { - wrmsr (MSR_IA32_MC0_CTL+4*i, 0xffffffff, 0xffffffff); - wrmsr (MSR_IA32_MC0_STATUS+4*i, 0x0, 0x0); + for (i = 0; i < nr_mce_banks; i++) { + wrmsr(MSR_IA32_MC0_CTL+4*i, 0xffffffff, 0xffffffff); + wrmsr(MSR_IA32_MC0_STATUS+4*i, 0x0, 0x0); } - set_in_cr4 (X86_CR4_MCE); - printk (KERN_INFO "Intel machine check reporting enabled on CPU#%d.\n", + set_in_cr4(X86_CR4_MCE); + printk(KERN_INFO "Intel machine check reporting enabled on CPU#%d.\n", smp_processor_id()); /* Check for P4/Xeon extended MCE MSRs */ - rdmsr (MSR_IA32_MCG_CAP, l, h); + rdmsr(MSR_IA32_MCG_CAP, l, h); if (l & (1<<9)) {/* MCG_EXT_P */ mce_num_extended_msrs = (l >> 16) & 0xff; - printk (KERN_INFO "CPU%d: Intel P4/Xeon Extended MCE MSRs (%d)" + printk(KERN_INFO "CPU%d: Intel P4/Xeon Extended MCE MSRs (%d)" " available\n", smp_processor_id(), mce_num_extended_msrs); diff --git a/arch/x86/kernel/cpu/mtrr/generic.c b/arch/x86/kernel/cpu/mtrr/generic.c index 5d241ce94a4..509bd3d9eac 100644 --- a/arch/x86/kernel/cpu/mtrr/generic.c +++ b/arch/x86/kernel/cpu/mtrr/generic.c @@ -37,7 +37,7 @@ static struct fixed_range_block fixed_range_blocks[] = { static unsigned long smp_changes_mask; static struct mtrr_state mtrr_state = {}; static int mtrr_state_set; -static u64 tom2; +u64 mtrr_tom2; #undef MODULE_PARAM_PREFIX #define MODULE_PARAM_PREFIX "mtrr." @@ -139,8 +139,8 @@ u8 mtrr_type_lookup(u64 start, u64 end) } } - if (tom2) { - if (start >= (1ULL<<32) && (end < tom2)) + if (mtrr_tom2) { + if (start >= (1ULL<<32) && (end < mtrr_tom2)) return MTRR_TYPE_WRBACK; } @@ -158,6 +158,20 @@ get_mtrr_var_range(unsigned int index, struct mtrr_var_range *vr) rdmsr(MTRRphysMask_MSR(index), vr->mask_lo, vr->mask_hi); } +/* fill the MSR pair relating to a var range */ +void fill_mtrr_var_range(unsigned int index, + u32 base_lo, u32 base_hi, u32 mask_lo, u32 mask_hi) +{ + struct mtrr_var_range *vr; + + vr = mtrr_state.var_ranges; + + vr[index].base_lo = base_lo; + vr[index].base_hi = base_hi; + vr[index].mask_lo = mask_lo; + vr[index].mask_hi = mask_hi; +} + static void get_fixed_ranges(mtrr_type * frs) { @@ -213,13 +227,13 @@ void __init get_mtrr_state(void) mtrr_state.enabled = (lo & 0xc00) >> 10; if (amd_special_default_mtrr()) { - unsigned lo, hi; + unsigned low, high; /* TOP_MEM2 */ - rdmsr(MSR_K8_TOP_MEM2, lo, hi); - tom2 = hi; - tom2 <<= 32; - tom2 |= lo; - tom2 &= 0xffffff8000000ULL; + rdmsr(MSR_K8_TOP_MEM2, low, high); + mtrr_tom2 = high; + mtrr_tom2 <<= 32; + mtrr_tom2 |= low; + mtrr_tom2 &= 0xffffff800000ULL; } if (mtrr_show) { int high_width; @@ -251,9 +265,9 @@ void __init get_mtrr_state(void) else printk(KERN_INFO "MTRR %u disabled\n", i); } - if (tom2) { + if (mtrr_tom2) { printk(KERN_INFO "TOM2: %016llx aka %lldM\n", - tom2, tom2>>20); + mtrr_tom2, mtrr_tom2>>20); } } mtrr_state_set = 1; @@ -328,7 +342,7 @@ static void set_fixed_range(int msr, bool *changed, unsigned int *msrwords) if (lo != msrwords[0] || hi != msrwords[1]) { if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD && - boot_cpu_data.x86 == 15 && + (boot_cpu_data.x86 >= 0x0f && boot_cpu_data.x86 <= 0x11) && ((msrwords[0] | msrwords[1]) & K8_MTRR_RDMEM_WRMEM_MASK)) k8_enable_fixed_iorrs(); mtrr_wrmsr(msr, msrwords[0], msrwords[1]); diff --git a/arch/x86/kernel/cpu/mtrr/main.c b/arch/x86/kernel/cpu/mtrr/main.c index 6a1e278d932..105afe12beb 100644 --- a/arch/x86/kernel/cpu/mtrr/main.c +++ b/arch/x86/kernel/cpu/mtrr/main.c @@ -37,6 +37,7 @@ #include <linux/smp.h> #include <linux/cpu.h> #include <linux/mutex.h> +#include <linux/sort.h> #include <asm/e820.h> #include <asm/mtrr.h> @@ -609,6 +610,787 @@ static struct sysdev_driver mtrr_sysdev_driver = { .resume = mtrr_restore, }; +/* should be related to MTRR_VAR_RANGES nums */ +#define RANGE_NUM 256 + +struct res_range { + unsigned long start; + unsigned long end; +}; + +static int __init +add_range(struct res_range *range, int nr_range, unsigned long start, + unsigned long end) +{ + /* out of slots */ + if (nr_range >= RANGE_NUM) + return nr_range; + + range[nr_range].start = start; + range[nr_range].end = end; + + nr_range++; + + return nr_range; +} + +static int __init +add_range_with_merge(struct res_range *range, int nr_range, unsigned long start, + unsigned long end) +{ + int i; + + /* try to merge it with old one */ + for (i = 0; i < nr_range; i++) { + unsigned long final_start, final_end; + unsigned long common_start, common_end; + + if (!range[i].end) + continue; + + common_start = max(range[i].start, start); + common_end = min(range[i].end, end); + if (common_start > common_end + 1) + continue; + + final_start = min(range[i].start, start); + final_end = max(range[i].end, end); + + range[i].start = final_start; + range[i].end = final_end; + return nr_range; + } + + /* need to add that */ + return add_range(range, nr_range, start, end); +} + +static void __init +subtract_range(struct res_range *range, unsigned long start, unsigned long end) +{ + int i, j; + + for (j = 0; j < RANGE_NUM; j++) { + if (!range[j].end) + continue; + + if (start <= range[j].start && end >= range[j].end) { + range[j].start = 0; + range[j].end = 0; + continue; + } + + if (start <= range[j].start && end < range[j].end && + range[j].start < end + 1) { + range[j].start = end + 1; + continue; + } + + + if (start > range[j].start && end >= range[j].end && + range[j].end > start - 1) { + range[j].end = start - 1; + continue; + } + + if (start > range[j].start && end < range[j].end) { + /* find the new spare */ + for (i = 0; i < RANGE_NUM; i++) { + if (range[i].end == 0) + break; + } + if (i < RANGE_NUM) { + range[i].end = range[j].end; + range[i].start = end + 1; + } else { + printk(KERN_ERR "run of slot in ranges\n"); + } + range[j].end = start - 1; + continue; + } + } +} + +static int __init cmp_range(const void *x1, const void *x2) +{ + const struct res_range *r1 = x1; + const struct res_range *r2 = x2; + long start1, start2; + + start1 = r1->start; + start2 = r2->start; + + return start1 - start2; +} + +struct var_mtrr_range_state { + unsigned long base_pfn; + unsigned long size_pfn; + mtrr_type type; +}; + +struct var_mtrr_range_state __initdata range_state[RANGE_NUM]; +static int __initdata debug_print; + +static int __init +x86_get_mtrr_mem_range(struct res_range *range, int nr_range, + unsigned long extra_remove_base, + unsigned long extra_remove_size) +{ + unsigned long i, base, size; + mtrr_type type; + + for (i = 0; i < num_var_ranges; i++) { + type = range_state[i].type; + if (type != MTRR_TYPE_WRBACK) + continue; + base = range_state[i].base_pfn; + size = range_state[i].size_pfn; + nr_range = add_range_with_merge(range, nr_range, base, + base + size - 1); + } + if (debug_print) { + printk(KERN_DEBUG "After WB checking\n"); + for (i = 0; i < nr_range; i++) + printk(KERN_DEBUG "MTRR MAP PFN: %016lx - %016lx\n", + range[i].start, range[i].end + 1); + } + + /* take out UC ranges */ + for (i = 0; i < num_var_ranges; i++) { + type = range_state[i].type; + if (type != MTRR_TYPE_UNCACHABLE) + continue; + size = range_state[i].size_pfn; + if (!size) + continue; + base = range_state[i].base_pfn; + subtract_range(range, base, base + size - 1); + } + if (extra_remove_size) + subtract_range(range, extra_remove_base, + extra_remove_base + extra_remove_size - 1); + + /* get new range num */ + nr_range = 0; + for (i = 0; i < RANGE_NUM; i++) { + if (!range[i].end) + continue; + nr_range++; + } + if (debug_print) { + printk(KERN_DEBUG "After UC checking\n"); + for (i = 0; i < nr_range; i++) + printk(KERN_DEBUG "MTRR MAP PFN: %016lx - %016lx\n", + range[i].start, range[i].end + 1); + } + + /* sort the ranges */ + sort(range, nr_range, sizeof(struct res_range), cmp_range, NULL); + if (debug_print) { + printk(KERN_DEBUG "After sorting\n"); + for (i = 0; i < nr_range; i++) + printk(KERN_DEBUG "MTRR MAP PFN: %016lx - %016lx\n", + range[i].start, range[i].end + 1); + } + + /* clear those is not used */ + for (i = nr_range; i < RANGE_NUM; i++) + memset(&range[i], 0, sizeof(range[i])); + + return nr_range; +} + +static struct res_range __initdata range[RANGE_NUM]; + +#ifdef CONFIG_MTRR_SANITIZER + +static unsigned long __init sum_ranges(struct res_range *range, int nr_range) +{ + unsigned long sum; + int i; + + sum = 0; + for (i = 0; i < nr_range; i++) + sum += range[i].end + 1 - range[i].start; + + return sum; +} + +static int enable_mtrr_cleanup __initdata = + CONFIG_MTRR_SANITIZER_ENABLE_DEFAULT; + +static int __init disable_mtrr_cleanup_setup(char *str) +{ + if (enable_mtrr_cleanup != -1) + enable_mtrr_cleanup = 0; + return 0; +} +early_param("disable_mtrr_cleanup", disable_mtrr_cleanup_setup); + +static int __init enable_mtrr_cleanup_setup(char *str) +{ + if (enable_mtrr_cleanup != -1) + enable_mtrr_cleanup = 1; + return 0; +} +early_param("enble_mtrr_cleanup", enable_mtrr_cleanup_setup); + +struct var_mtrr_state { + unsigned long range_startk; + unsigned long range_sizek; + unsigned long chunk_sizek; + unsigned long gran_sizek; + unsigned int reg; +}; + +static void __init +set_var_mtrr(unsigned int reg, unsigned long basek, unsigned long sizek, + unsigned char type, unsigned int address_bits) +{ + u32 base_lo, base_hi, mask_lo, mask_hi; + u64 base, mask; + + if (!sizek) { + fill_mtrr_var_range(reg, 0, 0, 0, 0); + return; + } + + mask = (1ULL << address_bits) - 1; + mask &= ~((((u64)sizek) << 10) - 1); + + base = ((u64)basek) << 10; + + base |= type; + mask |= 0x800; + + base_lo = base & ((1ULL<<32) - 1); + base_hi = base >> 32; + + mask_lo = mask & ((1ULL<<32) - 1); + mask_hi = mask >> 32; + + fill_mtrr_var_range(reg, base_lo, base_hi, mask_lo, mask_hi); +} + +static void __init +save_var_mtrr(unsigned int reg, unsigned long basek, unsigned long sizek, + unsigned char type) +{ + range_state[reg].base_pfn = basek >> (PAGE_SHIFT - 10); + range_state[reg].size_pfn = sizek >> (PAGE_SHIFT - 10); + range_state[reg].type = type; +} + +static void __init +set_var_mtrr_all(unsigned int address_bits) +{ + unsigned long basek, sizek; + unsigned char type; + unsigned int reg; + + for (reg = 0; reg < num_var_ranges; reg++) { + basek = range_state[reg].base_pfn << (PAGE_SHIFT - 10); + sizek = range_state[reg].size_pfn << (PAGE_SHIFT - 10); + type = range_state[reg].type; + + set_var_mtrr(reg, basek, sizek, type, address_bits); + } +} + +static unsigned int __init +range_to_mtrr(unsigned int reg, unsigned long range_startk, + unsigned long range_sizek, unsigned char type) +{ + if (!range_sizek || (reg >= num_var_ranges)) + return reg; + + while (range_sizek) { + unsigned long max_align, align; + unsigned long sizek; + + /* Compute the maximum size I can make a range */ + if (range_startk) + max_align = ffs(range_startk) - 1; + else + max_align = 32; + align = fls(range_sizek) - 1; + if (align > max_align) + align = max_align; + + sizek = 1 << align; + if (debug_print) + printk(KERN_DEBUG "Setting variable MTRR %d, " + "base: %ldMB, range: %ldMB, type %s\n", + reg, range_startk >> 10, sizek >> 10, + (type == MTRR_TYPE_UNCACHABLE)?"UC": + ((type == MTRR_TYPE_WRBACK)?"WB":"Other") + ); + save_var_mtrr(reg++, range_startk, sizek, type); + range_startk += sizek; + range_sizek -= sizek; + if (reg >= num_var_ranges) + break; + } + return reg; +} + +static unsigned __init +range_to_mtrr_with_hole(struct var_mtrr_state *state, unsigned long basek, + unsigned long sizek) +{ + unsigned long hole_basek, hole_sizek; + unsigned long second_basek, second_sizek; + unsigned long range0_basek, range0_sizek; + unsigned long range_basek, range_sizek; + unsigned long chunk_sizek; + unsigned long gran_sizek; + + hole_basek = 0; + hole_sizek = 0; + second_basek = 0; + second_sizek = 0; + chunk_sizek = state->chunk_sizek; + gran_sizek = state->gran_sizek; + + /* align with gran size, prevent small block used up MTRRs */ + range_basek = ALIGN(state->range_startk, gran_sizek); + if ((range_basek > basek) && basek) + return second_sizek; + state->range_sizek -= (range_basek - state->range_startk); + range_sizek = ALIGN(state->range_sizek, gran_sizek); + + while (range_sizek > state->range_sizek) { + range_sizek -= gran_sizek; + if (!range_sizek) + return 0; + } + state->range_sizek = range_sizek; + + /* try to append some small hole */ + range0_basek = state->range_startk; + range0_sizek = ALIGN(state->range_sizek, chunk_sizek); + if (range0_sizek == state->range_sizek) { + if (debug_print) + printk(KERN_DEBUG "rangeX: %016lx - %016lx\n", + range0_basek<<10, + (range0_basek + state->range_sizek)<<10); + state->reg = range_to_mtrr(state->reg, range0_basek, + state->range_sizek, MTRR_TYPE_WRBACK); + return 0; + } + + range0_sizek -= chunk_sizek; + if (range0_sizek && sizek) { + while (range0_basek + range0_sizek > (basek + sizek)) { + range0_sizek -= chunk_sizek; + if (!range0_sizek) + break; + } + } + + if (range0_sizek) { + if (debug_print) + printk(KERN_DEBUG "range0: %016lx - %016lx\n", + range0_basek<<10, + (range0_basek + range0_sizek)<<10); + state->reg = range_to_mtrr(state->reg, range0_basek, + range0_sizek, MTRR_TYPE_WRBACK); + + } + + range_basek = range0_basek + range0_sizek; + range_sizek = chunk_sizek; + + if (range_basek + range_sizek > basek && + range_basek + range_sizek <= (basek + sizek)) { + /* one hole */ + second_basek = basek; + second_sizek = range_basek + range_sizek - basek; + } + + /* if last piece, only could one hole near end */ + if ((second_basek || !basek) && + range_sizek - (state->range_sizek - range0_sizek) - second_sizek < + (chunk_sizek >> 1)) { + /* + * one hole in middle (second_sizek is 0) or at end + * (second_sizek is 0 ) + */ + hole_sizek = range_sizek - (state->range_sizek - range0_sizek) + - second_sizek; + hole_basek = range_basek + range_sizek - hole_sizek + - second_sizek; + } else { + /* fallback for big hole, or several holes */ + range_sizek = state->range_sizek - range0_sizek; + second_basek = 0; + second_sizek = 0; + } + + if (debug_print) + printk(KERN_DEBUG "range: %016lx - %016lx\n", range_basek<<10, + (range_basek + range_sizek)<<10); + state->reg = range_to_mtrr(state->reg, range_basek, range_sizek, + MTRR_TYPE_WRBACK); + if (hole_sizek) { + if (debug_print) + printk(KERN_DEBUG "hole: %016lx - %016lx\n", + hole_basek<<10, (hole_basek + hole_sizek)<<10); + state->reg = range_to_mtrr(state->reg, hole_basek, hole_sizek, + MTRR_TYPE_UNCACHABLE); + + } + + return second_sizek; +} + +static void __init +set_var_mtrr_range(struct var_mtrr_state *state, unsigned long base_pfn, + unsigned long size_pfn) +{ + unsigned long basek, sizek; + unsigned long second_sizek = 0; + + if (state->reg >= num_var_ranges) + return; + + basek = base_pfn << (PAGE_SHIFT - 10); + sizek = size_pfn << (PAGE_SHIFT - 10); + + /* See if I can merge with the last range */ + if ((basek <= 1024) || + (state->range_startk + state->range_sizek == basek)) { + unsigned long endk = basek + sizek; + state->range_sizek = endk - state->range_startk; + return; + } + /* Write the range mtrrs */ + if (state->range_sizek != 0) + second_sizek = range_to_mtrr_with_hole(state, basek, sizek); + + /* Allocate an msr */ + state->range_startk = basek + second_sizek; + state->range_sizek = sizek - second_sizek; +} + +/* mininum size of mtrr block that can take hole */ +static u64 mtrr_chunk_size __initdata = (256ULL<<20); + +static int __init parse_mtrr_chunk_size_opt(char *p) +{ + if (!p) + return -EINVAL; + mtrr_chunk_size = memparse(p, &p); + return 0; +} +early_param("mtrr_chunk_size", parse_mtrr_chunk_size_opt); + +/* granity of mtrr of block */ +static u64 mtrr_gran_size __initdata; + +static int __init parse_mtrr_gran_size_opt(char *p) +{ + if (!p) + return -EINVAL; + mtrr_gran_size = memparse(p, &p); + return 0; +} +early_param("mtrr_gran_size", parse_mtrr_gran_size_opt); + +static int nr_mtrr_spare_reg __initdata = + CONFIG_MTRR_SANITIZER_SPARE_REG_NR_DEFAULT; + +static int __init parse_mtrr_spare_reg(char *arg) +{ + if (arg) + nr_mtrr_spare_reg = simple_strtoul(arg, NULL, 0); + return 0; +} + +early_param("mtrr_spare_reg_nr", parse_mtrr_spare_reg); + +static int __init +x86_setup_var_mtrrs(struct res_range *range, int nr_range, + u64 chunk_size, u64 gran_size) +{ + struct var_mtrr_state var_state; + int i; + int num_reg; + + var_state.range_startk = 0; + var_state.range_sizek = 0; + var_state.reg = 0; + var_state.chunk_sizek = chunk_size >> 10; + var_state.gran_sizek = gran_size >> 10; + + memset(range_state, 0, sizeof(range_state)); + + /* Write the range etc */ + for (i = 0; i < nr_range; i++) + set_var_mtrr_range(&var_state, range[i].start, + range[i].end - range[i].start + 1); + + /* Write the last range */ + if (var_state.range_sizek != 0) + range_to_mtrr_with_hole(&var_state, 0, 0); + + num_reg = var_state.reg; + /* Clear out the extra MTRR's */ + while (var_state.reg < num_var_ranges) { + save_var_mtrr(var_state.reg, 0, 0, 0); + var_state.reg++; + } + + return num_reg; +} + +struct mtrr_cleanup_result { + unsigned long gran_sizek; + unsigned long chunk_sizek; + unsigned long lose_cover_sizek; + unsigned int num_reg; + int bad; +}; + +/* + * gran_size: 1M, 2M, ..., 2G + * chunk size: gran_size, ..., 4G + * so we need (2+13)*6 + */ +#define NUM_RESULT 90 +#define PSHIFT (PAGE_SHIFT - 10) + +static struct mtrr_cleanup_result __initdata result[NUM_RESULT]; +static struct res_range __initdata range_new[RANGE_NUM]; +static unsigned long __initdata min_loss_pfn[RANGE_NUM]; + +static int __init mtrr_cleanup(unsigned address_bits) +{ + unsigned long extra_remove_base, extra_remove_size; + unsigned long i, base, size, def, dummy; + mtrr_type type; + int nr_range, nr_range_new; + u64 chunk_size, gran_size; + unsigned long range_sums, range_sums_new; + int index_good; + int num_reg_good; + + /* extra one for all 0 */ + int num[MTRR_NUM_TYPES + 1]; + + if (!is_cpu(INTEL) || enable_mtrr_cleanup < 1) + return 0; + rdmsr(MTRRdefType_MSR, def, dummy); + def &= 0xff; + if (def != MTRR_TYPE_UNCACHABLE) + return 0; + + /* get it and store it aside */ + memset(range_state, 0, sizeof(range_state)); + for (i = 0; i < num_var_ranges; i++) { + mtrr_if->get(i, &base, &size, &type); + range_state[i].base_pfn = base; + range_state[i].size_pfn = size; + range_state[i].type = type; + } + + /* check entries number */ + memset(num, 0, sizeof(num)); + for (i = 0; i < num_var_ranges; i++) { + type = range_state[i].type; + size = range_state[i].size_pfn; + if (type >= MTRR_NUM_TYPES) + continue; + if (!size) + type = MTRR_NUM_TYPES; + num[type]++; + } + + /* check if we got UC entries */ + if (!num[MTRR_TYPE_UNCACHABLE]) + return 0; + + /* check if we only had WB and UC */ + if (num[MTRR_TYPE_WRBACK] + num[MTRR_TYPE_UNCACHABLE] != + num_var_ranges - num[MTRR_NUM_TYPES]) + return 0; + + memset(range, 0, sizeof(range)); + extra_remove_size = 0; + if (mtrr_tom2) { + extra_remove_base = 1 << (32 - PAGE_SHIFT); + extra_remove_size = + (mtrr_tom2 >> PAGE_SHIFT) - extra_remove_base; + } + nr_range = x86_get_mtrr_mem_range(range, 0, extra_remove_base, + extra_remove_size); + range_sums = sum_ranges(range, nr_range); + printk(KERN_INFO "total RAM coverred: %ldM\n", + range_sums >> (20 - PAGE_SHIFT)); + + if (mtrr_chunk_size && mtrr_gran_size) { + int num_reg; + + debug_print = 1; + /* convert ranges to var ranges state */ + num_reg = x86_setup_var_mtrrs(range, nr_range, mtrr_chunk_size, + mtrr_gran_size); + + /* we got new setting in range_state, check it */ + memset(range_new, 0, sizeof(range_new)); + nr_range_new = x86_get_mtrr_mem_range(range_new, 0, + extra_remove_base, + extra_remove_size); + range_sums_new = sum_ranges(range_new, nr_range_new); + + i = 0; + result[i].chunk_sizek = mtrr_chunk_size >> 10; + result[i].gran_sizek = mtrr_gran_size >> 10; + result[i].num_reg = num_reg; + if (range_sums < range_sums_new) { + result[i].lose_cover_sizek = + (range_sums_new - range_sums) << PSHIFT; + result[i].bad = 1; + } else + result[i].lose_cover_sizek = + (range_sums - range_sums_new) << PSHIFT; + + printk(KERN_INFO "%sgran_size: %ldM \tchunk_size: %ldM \t", + result[i].bad?"*BAD*":" ", result[i].gran_sizek >> 10, + result[i].chunk_sizek >> 10); + printk(KERN_CONT "num_reg: %d \tlose cover RAM: %s%ldM \n", + result[i].num_reg, result[i].bad?"-":"", + result[i].lose_cover_sizek >> 10); + if (!result[i].bad) { + set_var_mtrr_all(address_bits); + return 1; + } + printk(KERN_INFO "invalid mtrr_gran_size or mtrr_chunk_size, " + "will find optimal one\n"); + debug_print = 0; + memset(result, 0, sizeof(result[0])); + } + + i = 0; + memset(min_loss_pfn, 0xff, sizeof(min_loss_pfn)); + memset(result, 0, sizeof(result)); + for (gran_size = (1ULL<<20); gran_size < (1ULL<<32); gran_size <<= 1) { + for (chunk_size = gran_size; chunk_size < (1ULL<<33); + chunk_size <<= 1) { + int num_reg; + + if (debug_print) + printk(KERN_INFO + "\ngran_size: %lldM chunk_size_size: %lldM\n", + gran_size >> 20, chunk_size >> 20); + if (i >= NUM_RESULT) + continue; + + /* convert ranges to var ranges state */ + num_reg = x86_setup_var_mtrrs(range, nr_range, + chunk_size, gran_size); + + /* we got new setting in range_state, check it */ + memset(range_new, 0, sizeof(range_new)); + nr_range_new = x86_get_mtrr_mem_range(range_new, 0, + extra_remove_base, extra_remove_size); + range_sums_new = sum_ranges(range_new, nr_range_new); + + result[i].chunk_sizek = chunk_size >> 10; + result[i].gran_sizek = gran_size >> 10; + result[i].num_reg = num_reg; + if (range_sums < range_sums_new) { + result[i].lose_cover_sizek = + (range_sums_new - range_sums) << PSHIFT; + result[i].bad = 1; + } else + result[i].lose_cover_sizek = + (range_sums - range_sums_new) << PSHIFT; + + /* double check it */ + if (!result[i].bad && !result[i].lose_cover_sizek) { + if (nr_range_new != nr_range || + memcmp(range, range_new, sizeof(range))) + result[i].bad = 1; + } + + if (!result[i].bad && (range_sums - range_sums_new < + min_loss_pfn[num_reg])) { + min_loss_pfn[num_reg] = + range_sums - range_sums_new; + } + i++; + } + } + + /* print out all */ + for (i = 0; i < NUM_RESULT; i++) { + printk(KERN_INFO "%sgran_size: %ldM \tchunk_size: %ldM \t", + result[i].bad?"*BAD* ":" ", result[i].gran_sizek >> 10, + result[i].chunk_sizek >> 10); + printk(KERN_CONT "num_reg: %d \tlose RAM: %s%ldM\n", + result[i].num_reg, result[i].bad?"-":"", + result[i].lose_cover_sizek >> 10); + } + + /* try to find the optimal index */ + if (nr_mtrr_spare_reg >= num_var_ranges) + nr_mtrr_spare_reg = num_var_ranges - 1; + num_reg_good = -1; + for (i = num_var_ranges - nr_mtrr_spare_reg; i > 0; i--) { + if (!min_loss_pfn[i]) { + num_reg_good = i; + break; + } + } + + index_good = -1; + if (num_reg_good != -1) { + for (i = 0; i < NUM_RESULT; i++) { + if (!result[i].bad && + result[i].num_reg == num_reg_good && + !result[i].lose_cover_sizek) { + index_good = i; + break; + } + } + } + + if (index_good != -1) { + printk(KERN_INFO "Found optimal setting for mtrr clean up\n"); + i = index_good; + printk(KERN_INFO "gran_size: %ldM \tchunk_size: %ldM \t", + result[i].gran_sizek >> 10, + result[i].chunk_sizek >> 10); + printk(KERN_CONT "num_reg: %d \tlose RAM: %ldM\n", + result[i].num_reg, + result[i].lose_cover_sizek >> 10); + /* convert ranges to var ranges state */ + chunk_size = result[i].chunk_sizek; + chunk_size <<= 10; + gran_size = result[i].gran_sizek; + gran_size <<= 10; + debug_print = 1; + x86_setup_var_mtrrs(range, nr_range, chunk_size, gran_size); + set_var_mtrr_all(address_bits); + return 1; + } + + printk(KERN_INFO "mtrr_cleanup: can not find optimal value\n"); + printk(KERN_INFO "please specify mtrr_gran_size/mtrr_chunk_size\n"); + + return 0; +} +#else +static int __init mtrr_cleanup(unsigned address_bits) +{ + return 0; +} +#endif + +static int __initdata changed_by_mtrr_cleanup; + static int disable_mtrr_trim; static int __init disable_mtrr_trim_setup(char *str) @@ -648,6 +1430,19 @@ int __init amd_special_default_mtrr(void) return 0; } +static u64 __init real_trim_memory(unsigned long start_pfn, + unsigned long limit_pfn) +{ + u64 trim_start, trim_size; + trim_start = start_pfn; + trim_start <<= PAGE_SHIFT; + trim_size = limit_pfn; + trim_size <<= PAGE_SHIFT; + trim_size -= trim_start; + + return e820_update_range(trim_start, trim_size, E820_RAM, + E820_RESERVED); +} /** * mtrr_trim_uncached_memory - trim RAM not covered by MTRRs * @end_pfn: ending page frame number @@ -663,8 +1458,11 @@ int __init mtrr_trim_uncached_memory(unsigned long end_pfn) { unsigned long i, base, size, highest_pfn = 0, def, dummy; mtrr_type type; - u64 trim_start, trim_size; + int nr_range; + u64 total_trim_size; + /* extra one for all 0 */ + int num[MTRR_NUM_TYPES + 1]; /* * Make sure we only trim uncachable memory on machines that * support the Intel MTRR architecture: @@ -676,14 +1474,22 @@ int __init mtrr_trim_uncached_memory(unsigned long end_pfn) if (def != MTRR_TYPE_UNCACHABLE) return 0; - if (amd_special_default_mtrr()) - return 0; + /* get it and store it aside */ + memset(range_state, 0, sizeof(range_state)); + for (i = 0; i < num_var_ranges; i++) { + mtrr_if->get(i, &base, &size, &type); + range_state[i].base_pfn = base; + range_state[i].size_pfn = size; + range_state[i].type = type; + } /* Find highest cached pfn */ for (i = 0; i < num_var_ranges; i++) { - mtrr_if->get(i, &base, &size, &type); + type = range_state[i].type; if (type != MTRR_TYPE_WRBACK) continue; + base = range_state[i].base_pfn; + size = range_state[i].size_pfn; if (highest_pfn < base + size) highest_pfn = base + size; } @@ -698,22 +1504,65 @@ int __init mtrr_trim_uncached_memory(unsigned long end_pfn) return 0; } - if (highest_pfn < end_pfn) { + /* check entries number */ + memset(num, 0, sizeof(num)); + for (i = 0; i < num_var_ranges; i++) { + type = range_state[i].type; + if (type >= MTRR_NUM_TYPES) + continue; + size = range_state[i].size_pfn; + if (!size) + type = MTRR_NUM_TYPES; + num[type]++; + } + + /* no entry for WB? */ + if (!num[MTRR_TYPE_WRBACK]) + return 0; + + /* check if we only had WB and UC */ + if (num[MTRR_TYPE_WRBACK] + num[MTRR_TYPE_UNCACHABLE] != + num_var_ranges - num[MTRR_NUM_TYPES]) + return 0; + + memset(range, 0, sizeof(range)); + nr_range = 0; + if (mtrr_tom2) { + range[nr_range].start = (1ULL<<(32 - PAGE_SHIFT)); + range[nr_range].end = (mtrr_tom2 >> PAGE_SHIFT) - 1; + if (highest_pfn < range[nr_range].end + 1) + highest_pfn = range[nr_range].end + 1; + nr_range++; + } + nr_range = x86_get_mtrr_mem_range(range, nr_range, 0, 0); + + total_trim_size = 0; + /* check the head */ + if (range[0].start) + total_trim_size += real_trim_memory(0, range[0].start); + /* check the holes */ + for (i = 0; i < nr_range - 1; i++) { + if (range[i].end + 1 < range[i+1].start) + total_trim_size += real_trim_memory(range[i].end + 1, + range[i+1].start); + } + /* check the top */ + i = nr_range - 1; + if (range[i].end + 1 < end_pfn) + total_trim_size += real_trim_memory(range[i].end + 1, + end_pfn); + + if (total_trim_size) { printk(KERN_WARNING "WARNING: BIOS bug: CPU MTRRs don't cover" - " all of memory, losing %luMB of RAM.\n", - (end_pfn - highest_pfn) >> (20 - PAGE_SHIFT)); + " all of memory, losing %lluMB of RAM.\n", + total_trim_size >> 20); - WARN_ON(1); + if (!changed_by_mtrr_cleanup) + WARN_ON(1); printk(KERN_INFO "update e820 for mtrr\n"); - trim_start = highest_pfn; - trim_start <<= PAGE_SHIFT; - trim_size = end_pfn; - trim_size <<= PAGE_SHIFT; - trim_size -= trim_start; - update_memory_range(trim_start, trim_size, E820_RAM, - E820_RESERVED); update_e820(); + return 1; } @@ -729,18 +1578,21 @@ int __init mtrr_trim_uncached_memory(unsigned long end_pfn) */ void __init mtrr_bp_init(void) { + u32 phys_addr; init_ifs(); + phys_addr = 32; + if (cpu_has_mtrr) { mtrr_if = &generic_mtrr_ops; size_or_mask = 0xff000000; /* 36 bits */ 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 bus of the Xeon. */ if (cpuid_eax(0x80000000) >= 0x80000008) { - u32 phys_addr; phys_addr = cpuid_eax(0x80000008) & 0xff; /* CPUID workaround for Intel 0F33/0F34 CPU */ if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL && @@ -758,6 +1610,7 @@ void __init mtrr_bp_init(void) don't support PAE */ size_or_mask = 0xfff00000; /* 32 bits */ size_and_mask = 0; + phys_addr = 32; } } else { switch (boot_cpu_data.x86_vendor) { @@ -791,8 +1644,15 @@ void __init mtrr_bp_init(void) if (mtrr_if) { set_num_var_ranges(); init_table(); - if (use_intel()) + if (use_intel()) { get_mtrr_state(); + + if (mtrr_cleanup(phys_addr)) { + changed_by_mtrr_cleanup = 1; + mtrr_if->set_all(); + } + + } } } @@ -829,9 +1689,10 @@ static int __init mtrr_init_finialize(void) { if (!mtrr_if) return 0; - if (use_intel()) - mtrr_state_warn(); - else { + if (use_intel()) { + if (!changed_by_mtrr_cleanup) + mtrr_state_warn(); + } else { /* The CPUs haven't MTRR and seem to not support SMP. They have * specific drivers, we use a tricky method to support * suspend/resume for them. diff --git a/arch/x86/kernel/cpu/mtrr/mtrr.h b/arch/x86/kernel/cpu/mtrr/mtrr.h index 2cc77eb6fea..2dc4ec656b2 100644 --- a/arch/x86/kernel/cpu/mtrr/mtrr.h +++ b/arch/x86/kernel/cpu/mtrr/mtrr.h @@ -81,6 +81,8 @@ void set_mtrr_done(struct set_mtrr_context *ctxt); void set_mtrr_cache_disable(struct set_mtrr_context *ctxt); void set_mtrr_prepare_save(struct set_mtrr_context *ctxt); +void fill_mtrr_var_range(unsigned int index, + u32 base_lo, u32 base_hi, u32 mask_lo, u32 mask_hi); void get_mtrr_state(void); extern void set_mtrr_ops(struct mtrr_ops * ops); @@ -92,6 +94,7 @@ extern struct mtrr_ops * mtrr_if; #define use_intel() (mtrr_if && mtrr_if->use_intel_if == 1) extern unsigned int num_var_ranges; +extern u64 mtrr_tom2; void mtrr_state_warn(void); const char *mtrr_attrib_to_str(int x); diff --git a/arch/x86/kernel/cpu/perfctr-watchdog.c b/arch/x86/kernel/cpu/perfctr-watchdog.c index f9ae93adffe..2e9bef6e3aa 100644 --- a/arch/x86/kernel/cpu/perfctr-watchdog.c +++ b/arch/x86/kernel/cpu/perfctr-watchdog.c @@ -1,11 +1,15 @@ -/* local apic based NMI watchdog for various CPUs. - This file also handles reservation of performance counters for coordination - with other users (like oprofile). - - Note that these events normally don't tick when the CPU idles. This means - the frequency varies with CPU load. - - Original code for K7/P6 written by Keith Owens */ +/* + * local apic based NMI watchdog for various CPUs. + * + * This file also handles reservation of performance counters for coordination + * with other users (like oprofile). + * + * Note that these events normally don't tick when the CPU idles. This means + * the frequency varies with CPU load. + * + * Original code for K7/P6 written by Keith Owens + * + */ #include <linux/percpu.h> #include <linux/module.h> @@ -36,12 +40,16 @@ struct wd_ops { static const struct wd_ops *wd_ops; -/* this number is calculated from Intel's MSR_P4_CRU_ESCR5 register and it's - * offset from MSR_P4_BSU_ESCR0. It will be the max for all platforms (for now) +/* + * this number is calculated from Intel's MSR_P4_CRU_ESCR5 register and it's + * offset from MSR_P4_BSU_ESCR0. + * + * It will be the max for all platforms (for now) */ #define NMI_MAX_COUNTER_BITS 66 -/* perfctr_nmi_owner tracks the ownership of the perfctr registers: +/* + * perfctr_nmi_owner tracks the ownership of the perfctr registers: * evtsel_nmi_owner tracks the ownership of the event selection * - different performance counters/ event selection may be reserved for * different subsystems this reservation system just tries to coordinate @@ -73,8 +81,10 @@ static inline unsigned int nmi_perfctr_msr_to_bit(unsigned int msr) return 0; } -/* converts an msr to an appropriate reservation bit */ -/* returns the bit offset of the event selection register */ +/* + * converts an msr to an appropriate reservation bit + * returns the bit offset of the event selection register + */ static inline unsigned int nmi_evntsel_msr_to_bit(unsigned int msr) { /* returns the bit offset of the event selection register */ @@ -114,6 +124,7 @@ int avail_to_resrv_perfctr_nmi(unsigned int msr) return (!test_bit(counter, perfctr_nmi_owner)); } +EXPORT_SYMBOL(avail_to_resrv_perfctr_nmi_bit); int reserve_perfctr_nmi(unsigned int msr) { @@ -128,6 +139,7 @@ int reserve_perfctr_nmi(unsigned int msr) return 1; return 0; } +EXPORT_SYMBOL(reserve_perfctr_nmi); void release_perfctr_nmi(unsigned int msr) { @@ -140,6 +152,7 @@ void release_perfctr_nmi(unsigned int msr) clear_bit(counter, perfctr_nmi_owner); } +EXPORT_SYMBOL(release_perfctr_nmi); int reserve_evntsel_nmi(unsigned int msr) { @@ -154,6 +167,7 @@ int reserve_evntsel_nmi(unsigned int msr) return 1; return 0; } +EXPORT_SYMBOL(reserve_evntsel_nmi); void release_evntsel_nmi(unsigned int msr) { @@ -166,11 +180,6 @@ void release_evntsel_nmi(unsigned int msr) clear_bit(counter, evntsel_nmi_owner); } - -EXPORT_SYMBOL(avail_to_resrv_perfctr_nmi_bit); -EXPORT_SYMBOL(reserve_perfctr_nmi); -EXPORT_SYMBOL(release_perfctr_nmi); -EXPORT_SYMBOL(reserve_evntsel_nmi); EXPORT_SYMBOL(release_evntsel_nmi); void disable_lapic_nmi_watchdog(void) @@ -181,7 +190,9 @@ void disable_lapic_nmi_watchdog(void) return; on_each_cpu(stop_apic_nmi_watchdog, NULL, 0, 1); - wd_ops->unreserve(); + + if (wd_ops) + wd_ops->unreserve(); BUG_ON(atomic_read(&nmi_active) != 0); } @@ -232,8 +243,8 @@ static unsigned int adjust_for_32bit_ctr(unsigned int hz) return retval; } -static void -write_watchdog_counter(unsigned int perfctr_msr, const char *descr, unsigned nmi_hz) +static void write_watchdog_counter(unsigned int perfctr_msr, + const char *descr, unsigned nmi_hz) { u64 count = (u64)cpu_khz * 1000; @@ -244,7 +255,7 @@ write_watchdog_counter(unsigned int perfctr_msr, const char *descr, unsigned nmi } static void write_watchdog_counter32(unsigned int perfctr_msr, - const char *descr, unsigned nmi_hz) + const char *descr, unsigned nmi_hz) { u64 count = (u64)cpu_khz * 1000; @@ -254,9 +265,10 @@ static void write_watchdog_counter32(unsigned int perfctr_msr, wrmsr(perfctr_msr, (u32)(-count), 0); } -/* AMD K7/K8/Family10h/Family11h support. AMD keeps this interface - nicely stable so there is not much variety */ - +/* + * AMD K7/K8/Family10h/Family11h support. + * AMD keeps this interface nicely stable so there is not much variety + */ #define K7_EVNTSEL_ENABLE (1 << 22) #define K7_EVNTSEL_INT (1 << 20) #define K7_EVNTSEL_OS (1 << 17) @@ -289,7 +301,7 @@ static int setup_k7_watchdog(unsigned nmi_hz) wd->perfctr_msr = perfctr_msr; wd->evntsel_msr = evntsel_msr; - wd->cccr_msr = 0; //unused + wd->cccr_msr = 0; /* unused */ return 1; } @@ -325,18 +337,19 @@ static void single_msr_rearm(struct nmi_watchdog_ctlblk *wd, unsigned nmi_hz) } static const struct wd_ops k7_wd_ops = { - .reserve = single_msr_reserve, - .unreserve = single_msr_unreserve, - .setup = setup_k7_watchdog, - .rearm = single_msr_rearm, - .stop = single_msr_stop_watchdog, - .perfctr = MSR_K7_PERFCTR0, - .evntsel = MSR_K7_EVNTSEL0, - .checkbit = 1ULL<<47, + .reserve = single_msr_reserve, + .unreserve = single_msr_unreserve, + .setup = setup_k7_watchdog, + .rearm = single_msr_rearm, + .stop = single_msr_stop_watchdog, + .perfctr = MSR_K7_PERFCTR0, + .evntsel = MSR_K7_EVNTSEL0, + .checkbit = 1ULL << 47, }; -/* Intel Model 6 (PPro+,P2,P3,P-M,Core1) */ - +/* + * Intel Model 6 (PPro+,P2,P3,P-M,Core1) + */ #define P6_EVNTSEL0_ENABLE (1 << 22) #define P6_EVNTSEL_INT (1 << 20) #define P6_EVNTSEL_OS (1 << 17) @@ -372,52 +385,58 @@ static int setup_p6_watchdog(unsigned nmi_hz) wd->perfctr_msr = perfctr_msr; wd->evntsel_msr = evntsel_msr; - wd->cccr_msr = 0; //unused + wd->cccr_msr = 0; /* unused */ return 1; } static void p6_rearm(struct nmi_watchdog_ctlblk *wd, unsigned nmi_hz) { - /* P6 based Pentium M need to re-unmask + /* + * P6 based Pentium M need to re-unmask * the apic vector but it doesn't hurt * other P6 variant. - * ArchPerfom/Core Duo also needs this */ + * ArchPerfom/Core Duo also needs this + */ apic_write(APIC_LVTPC, APIC_DM_NMI); + /* P6/ARCH_PERFMON has 32 bit counter write */ write_watchdog_counter32(wd->perfctr_msr, NULL,nmi_hz); } static const struct wd_ops p6_wd_ops = { - .reserve = single_msr_reserve, - .unreserve = single_msr_unreserve, - .setup = setup_p6_watchdog, - .rearm = p6_rearm, - .stop = single_msr_stop_watchdog, - .perfctr = MSR_P6_PERFCTR0, - .evntsel = MSR_P6_EVNTSEL0, - .checkbit = 1ULL<<39, + .reserve = single_msr_reserve, + .unreserve = single_msr_unreserve, + .setup = setup_p6_watchdog, + .rearm = p6_rearm, + .stop = single_msr_stop_watchdog, + .perfctr = MSR_P6_PERFCTR0, + .evntsel = MSR_P6_EVNTSEL0, + .checkbit = 1ULL << 39, }; -/* Intel P4 performance counters. By far the most complicated of all. */ - -#define MSR_P4_MISC_ENABLE_PERF_AVAIL (1<<7) -#define P4_ESCR_EVENT_SELECT(N) ((N)<<25) -#define P4_ESCR_OS (1<<3) -#define P4_ESCR_USR (1<<2) -#define P4_CCCR_OVF_PMI0 (1<<26) -#define P4_CCCR_OVF_PMI1 (1<<27) -#define P4_CCCR_THRESHOLD(N) ((N)<<20) -#define P4_CCCR_COMPLEMENT (1<<19) -#define P4_CCCR_COMPARE (1<<18) -#define P4_CCCR_REQUIRED (3<<16) -#define P4_CCCR_ESCR_SELECT(N) ((N)<<13) -#define P4_CCCR_ENABLE (1<<12) -#define P4_CCCR_OVF (1<<31) - -/* Set up IQ_COUNTER0 to behave like a clock, by having IQ_CCCR0 filter - CRU_ESCR0 (with any non-null event selector) through a complemented - max threshold. [IA32-Vol3, Section 14.9.9] */ +/* + * Intel P4 performance counters. + * By far the most complicated of all. + */ +#define MSR_P4_MISC_ENABLE_PERF_AVAIL (1 << 7) +#define P4_ESCR_EVENT_SELECT(N) ((N) << 25) +#define P4_ESCR_OS (1 << 3) +#define P4_ESCR_USR (1 << 2) +#define P4_CCCR_OVF_PMI0 (1 << 26) +#define P4_CCCR_OVF_PMI1 (1 << 27) +#define P4_CCCR_THRESHOLD(N) ((N) << 20) +#define P4_CCCR_COMPLEMENT (1 << 19) +#define P4_CCCR_COMPARE (1 << 18) +#define P4_CCCR_REQUIRED (3 << 16) +#define P4_CCCR_ESCR_SELECT(N) ((N) << 13) +#define P4_CCCR_ENABLE (1 << 12) +#define P4_CCCR_OVF (1 << 31) +/* + * Set up IQ_COUNTER0 to behave like a clock, by having IQ_CCCR0 filter + * CRU_ESCR0 (with any non-null event selector) through a complemented + * max threshold. [IA32-Vol3, Section 14.9.9] + */ static int setup_p4_watchdog(unsigned nmi_hz) { unsigned int perfctr_msr, evntsel_msr, cccr_msr; @@ -442,7 +461,8 @@ static int setup_p4_watchdog(unsigned nmi_hz) #endif ht_num = 0; - /* performance counters are shared resources + /* + * performance counters are shared resources * assign each hyperthread its own set * (re-use the ESCR0 register, seems safe * and keeps the cccr_val the same) @@ -540,20 +560,21 @@ static void p4_rearm(struct nmi_watchdog_ctlblk *wd, unsigned nmi_hz) } static const struct wd_ops p4_wd_ops = { - .reserve = p4_reserve, - .unreserve = p4_unreserve, - .setup = setup_p4_watchdog, - .rearm = p4_rearm, - .stop = stop_p4_watchdog, + .reserve = p4_reserve, + .unreserve = p4_unreserve, + .setup = setup_p4_watchdog, + .rearm = p4_rearm, + .stop = stop_p4_watchdog, /* RED-PEN this is wrong for the other sibling */ - .perfctr = MSR_P4_BPU_PERFCTR0, - .evntsel = MSR_P4_BSU_ESCR0, - .checkbit = 1ULL<<39, + .perfctr = MSR_P4_BPU_PERFCTR0, + .evntsel = MSR_P4_BSU_ESCR0, + .checkbit = 1ULL << 39, }; -/* Watchdog using the Intel architected PerfMon. Used for Core2 and hopefully - all future Intel CPUs. */ - +/* + * Watchdog using the Intel architected PerfMon. + * Used for Core2 and hopefully all future Intel CPUs. + */ #define ARCH_PERFMON_NMI_EVENT_SEL ARCH_PERFMON_UNHALTED_CORE_CYCLES_SEL #define ARCH_PERFMON_NMI_EVENT_UMASK ARCH_PERFMON_UNHALTED_CORE_CYCLES_UMASK @@ -599,19 +620,19 @@ static int setup_intel_arch_watchdog(unsigned nmi_hz) wd->perfctr_msr = perfctr_msr; wd->evntsel_msr = evntsel_msr; - wd->cccr_msr = 0; //unused + wd->cccr_msr = 0; /* unused */ intel_arch_wd_ops.checkbit = 1ULL << (eax.split.bit_width - 1); return 1; } static struct wd_ops intel_arch_wd_ops __read_mostly = { - .reserve = single_msr_reserve, - .unreserve = single_msr_unreserve, - .setup = setup_intel_arch_watchdog, - .rearm = p6_rearm, - .stop = single_msr_stop_watchdog, - .perfctr = MSR_ARCH_PERFMON_PERFCTR1, - .evntsel = MSR_ARCH_PERFMON_EVENTSEL1, + .reserve = single_msr_reserve, + .unreserve = single_msr_unreserve, + .setup = setup_intel_arch_watchdog, + .rearm = p6_rearm, + .stop = single_msr_stop_watchdog, + .perfctr = MSR_ARCH_PERFMON_PERFCTR1, + .evntsel = MSR_ARCH_PERFMON_EVENTSEL1, }; static void probe_nmi_watchdog(void) @@ -624,8 +645,10 @@ static void probe_nmi_watchdog(void) wd_ops = &k7_wd_ops; break; case X86_VENDOR_INTEL: - /* Work around Core Duo (Yonah) errata AE49 where perfctr1 - doesn't have a working enable bit. */ + /* + * Work around Core Duo (Yonah) errata AE49 where perfctr1 + * doesn't have a working enable bit. + */ if (boot_cpu_data.x86 == 6 && boot_cpu_data.x86_model == 14) { intel_arch_wd_ops.perfctr = MSR_ARCH_PERFMON_PERFCTR0; intel_arch_wd_ops.evntsel = MSR_ARCH_PERFMON_EVENTSEL0; @@ -636,7 +659,7 @@ static void probe_nmi_watchdog(void) } switch (boot_cpu_data.x86) { case 6: - if (boot_cpu_data.x86_model > 0xd) + if (boot_cpu_data.x86_model > 13) return; wd_ops = &p6_wd_ops; @@ -697,10 +720,11 @@ int lapic_wd_event(unsigned nmi_hz) { struct nmi_watchdog_ctlblk *wd = &__get_cpu_var(nmi_watchdog_ctlblk); u64 ctr; + rdmsrl(wd->perfctr_msr, ctr); - if (ctr & wd_ops->checkbit) { /* perfctr still running? */ + if (ctr & wd_ops->checkbit) /* perfctr still running? */ return 0; - } + wd_ops->rearm(wd, nmi_hz); return 1; } diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c new file mode 100644 index 00000000000..28c29180b38 --- /dev/null +++ b/arch/x86/kernel/e820.c @@ -0,0 +1,1390 @@ +/* + * Handle the memory map. + * The functions here do the job until bootmem takes over. + * + * Getting sanitize_e820_map() in sync with i386 version by applying change: + * - Provisions for empty E820 memory regions (reported by certain BIOSes). + * Alex Achenbach <xela@slit.de>, December 2002. + * Venkatesh Pallipadi <venkatesh.pallipadi@intel.com> + * + */ +#include <linux/kernel.h> +#include <linux/types.h> +#include <linux/init.h> +#include <linux/bootmem.h> +#include <linux/ioport.h> +#include <linux/string.h> +#include <linux/kexec.h> +#include <linux/module.h> +#include <linux/mm.h> +#include <linux/pfn.h> +#include <linux/suspend.h> +#include <linux/firmware-map.h> + +#include <asm/pgtable.h> +#include <asm/page.h> +#include <asm/e820.h> +#include <asm/proto.h> +#include <asm/setup.h> +#include <asm/trampoline.h> + +/* + * The e820 map is the map that gets modified e.g. with command line parameters + * and that is also registered with modifications in the kernel resource tree + * with the iomem_resource as parent. + * + * The e820_saved is directly saved after the BIOS-provided memory map is + * copied. It doesn't get modified afterwards. It's registered for the + * /sys/firmware/memmap interface. + * + * That memory map is not modified and is used as base for kexec. The kexec'd + * kernel should get the same memory map as the firmware provides. Then the + * user can e.g. boot the original kernel with mem=1G while still booting the + * next kernel with full memory. + */ +struct e820map e820; +struct e820map e820_saved; + +/* For PCI or other memory-mapped resources */ +unsigned long pci_mem_start = 0xaeedbabe; +#ifdef CONFIG_PCI +EXPORT_SYMBOL(pci_mem_start); +#endif + +/* + * This function checks if any part of the range <start,end> is mapped + * with type. + */ +int +e820_any_mapped(u64 start, u64 end, unsigned type) +{ + int i; + + for (i = 0; i < e820.nr_map; i++) { + struct e820entry *ei = &e820.map[i]; + + if (type && ei->type != type) + continue; + if (ei->addr >= end || ei->addr + ei->size <= start) + continue; + return 1; + } + return 0; +} +EXPORT_SYMBOL_GPL(e820_any_mapped); + +/* + * This function checks if the entire range <start,end> is mapped with type. + * + * Note: this function only works correct if the e820 table is sorted and + * not-overlapping, which is the case + */ +int __init e820_all_mapped(u64 start, u64 end, unsigned type) +{ + int i; + + for (i = 0; i < e820.nr_map; i++) { + struct e820entry *ei = &e820.map[i]; + + if (type && ei->type != type) + continue; + /* is the region (part) in overlap with the current region ?*/ + if (ei->addr >= end || ei->addr + ei->size <= start) + continue; + + /* if the region is at the beginning of <start,end> we move + * start to the end of the region since it's ok until there + */ + if (ei->addr <= start) + start = ei->addr + ei->size; + /* + * if start is now at or beyond end, we're done, full + * coverage + */ + if (start >= end) + return 1; + } + return 0; +} + +/* + * Add a memory region to the kernel e820 map. + */ +void __init e820_add_region(u64 start, u64 size, int type) +{ + int x = e820.nr_map; + + if (x == ARRAY_SIZE(e820.map)) { + printk(KERN_ERR "Ooops! Too many entries in the memory map!\n"); + return; + } + + e820.map[x].addr = start; + e820.map[x].size = size; + e820.map[x].type = type; + e820.nr_map++; +} + +void __init e820_print_map(char *who) +{ + int i; + + for (i = 0; i < e820.nr_map; i++) { + printk(KERN_INFO " %s: %016Lx - %016Lx ", who, + (unsigned long long) e820.map[i].addr, + (unsigned long long) + (e820.map[i].addr + e820.map[i].size)); + switch (e820.map[i].type) { + case E820_RAM: + case E820_RESERVED_KERN: + printk(KERN_CONT "(usable)\n"); + break; + case E820_RESERVED: + printk(KERN_CONT "(reserved)\n"); + break; + case E820_ACPI: + printk(KERN_CONT "(ACPI data)\n"); + break; + case E820_NVS: + printk(KERN_CONT "(ACPI NVS)\n"); + break; + default: + printk(KERN_CONT "type %u\n", e820.map[i].type); + break; + } + } +} + +/* + * Sanitize the BIOS e820 map. + * + * Some e820 responses include overlapping entries. The following + * replaces the original e820 map with a new one, removing overlaps, + * and resolving conflicting memory types in favor of highest + * numbered type. + * + * The input parameter biosmap points to an array of 'struct + * e820entry' which on entry has elements in the range [0, *pnr_map) + * valid, and which has space for up to max_nr_map entries. + * On return, the resulting sanitized e820 map entries will be in + * overwritten in the same location, starting at biosmap. + * + * The integer pointed to by pnr_map must be valid on entry (the + * current number of valid entries located at biosmap) and will + * be updated on return, with the new number of valid entries + * (something no more than max_nr_map.) + * + * The return value from sanitize_e820_map() is zero if it + * successfully 'sanitized' the map entries passed in, and is -1 + * if it did nothing, which can happen if either of (1) it was + * only passed one map entry, or (2) any of the input map entries + * were invalid (start + size < start, meaning that the size was + * so big the described memory range wrapped around through zero.) + * + * Visually we're performing the following + * (1,2,3,4 = memory types)... + * + * Sample memory map (w/overlaps): + * ____22__________________ + * ______________________4_ + * ____1111________________ + * _44_____________________ + * 11111111________________ + * ____________________33__ + * ___________44___________ + * __________33333_________ + * ______________22________ + * ___________________2222_ + * _________111111111______ + * _____________________11_ + * _________________4______ + * + * Sanitized equivalent (no overlap): + * 1_______________________ + * _44_____________________ + * ___1____________________ + * ____22__________________ + * ______11________________ + * _________1______________ + * __________3_____________ + * ___________44___________ + * _____________33_________ + * _______________2________ + * ________________1_______ + * _________________4______ + * ___________________2____ + * ____________________33__ + * ______________________4_ + */ + +int __init sanitize_e820_map(struct e820entry *biosmap, int max_nr_map, + int *pnr_map) +{ + struct change_member { + struct e820entry *pbios; /* pointer to original bios entry */ + unsigned long long addr; /* address for this change point */ + }; + static struct change_member change_point_list[2*E820_X_MAX] __initdata; + static struct change_member *change_point[2*E820_X_MAX] __initdata; + static struct e820entry *overlap_list[E820_X_MAX] __initdata; + static struct e820entry new_bios[E820_X_MAX] __initdata; + struct change_member *change_tmp; + unsigned long current_type, last_type; + unsigned long long last_addr; + int chgidx, still_changing; + int overlap_entries; + int new_bios_entry; + int old_nr, new_nr, chg_nr; + int i; + + /* if there's only one memory region, don't bother */ + if (*pnr_map < 2) + return -1; + + old_nr = *pnr_map; + BUG_ON(old_nr > max_nr_map); + + /* bail out if we find any unreasonable addresses in bios map */ + for (i = 0; i < old_nr; i++) + if (biosmap[i].addr + biosmap[i].size < biosmap[i].addr) + return -1; + + /* create pointers for initial change-point information (for sorting) */ + for (i = 0; i < 2 * old_nr; i++) + change_point[i] = &change_point_list[i]; + + /* record all known change-points (starting and ending addresses), + omitting those that are for empty memory regions */ + chgidx = 0; + for (i = 0; i < old_nr; i++) { + if (biosmap[i].size != 0) { + change_point[chgidx]->addr = biosmap[i].addr; + change_point[chgidx++]->pbios = &biosmap[i]; + change_point[chgidx]->addr = biosmap[i].addr + + biosmap[i].size; + change_point[chgidx++]->pbios = &biosmap[i]; + } + } + chg_nr = chgidx; + + /* sort change-point list by memory addresses (low -> high) */ + still_changing = 1; + while (still_changing) { + still_changing = 0; + for (i = 1; i < chg_nr; i++) { + unsigned long long curaddr, lastaddr; + unsigned long long curpbaddr, lastpbaddr; + + curaddr = change_point[i]->addr; + lastaddr = change_point[i - 1]->addr; + curpbaddr = change_point[i]->pbios->addr; + lastpbaddr = change_point[i - 1]->pbios->addr; + + /* + * swap entries, when: + * + * curaddr > lastaddr or + * curaddr == lastaddr and curaddr == curpbaddr and + * lastaddr != lastpbaddr + */ + if (curaddr < lastaddr || + (curaddr == lastaddr && curaddr == curpbaddr && + lastaddr != lastpbaddr)) { + change_tmp = change_point[i]; + change_point[i] = change_point[i-1]; + change_point[i-1] = change_tmp; + still_changing = 1; + } + } + } + + /* create a new bios memory map, removing overlaps */ + overlap_entries = 0; /* number of entries in the overlap table */ + new_bios_entry = 0; /* index for creating new bios map entries */ + last_type = 0; /* start with undefined memory type */ + last_addr = 0; /* start with 0 as last starting address */ + + /* loop through change-points, determining affect on the new bios map */ + for (chgidx = 0; chgidx < chg_nr; chgidx++) { + /* keep track of all overlapping bios entries */ + if (change_point[chgidx]->addr == + change_point[chgidx]->pbios->addr) { + /* + * add map entry to overlap list (> 1 entry + * implies an overlap) + */ + overlap_list[overlap_entries++] = + change_point[chgidx]->pbios; + } else { + /* + * remove entry from list (order independent, + * so swap with last) + */ + for (i = 0; i < overlap_entries; i++) { + if (overlap_list[i] == + change_point[chgidx]->pbios) + overlap_list[i] = + overlap_list[overlap_entries-1]; + } + overlap_entries--; + } + /* + * if there are overlapping entries, decide which + * "type" to use (larger value takes precedence -- + * 1=usable, 2,3,4,4+=unusable) + */ + current_type = 0; + for (i = 0; i < overlap_entries; i++) + if (overlap_list[i]->type > current_type) + current_type = overlap_list[i]->type; + /* + * continue building up new bios map based on this + * information + */ + if (current_type != last_type) { + if (last_type != 0) { + new_bios[new_bios_entry].size = + change_point[chgidx]->addr - last_addr; + /* + * move forward only if the new size + * was non-zero + */ + if (new_bios[new_bios_entry].size != 0) + /* + * no more space left for new + * bios entries ? + */ + if (++new_bios_entry >= max_nr_map) + break; + } + if (current_type != 0) { + new_bios[new_bios_entry].addr = + change_point[chgidx]->addr; + new_bios[new_bios_entry].type = current_type; + last_addr = change_point[chgidx]->addr; + } + last_type = current_type; + } + } + /* retain count for new bios entries */ + new_nr = new_bios_entry; + + /* copy new bios mapping into original location */ + memcpy(biosmap, new_bios, new_nr * sizeof(struct e820entry)); + *pnr_map = new_nr; + + return 0; +} + +static int __init __append_e820_map(struct e820entry *biosmap, int nr_map) +{ + while (nr_map) { + u64 start = biosmap->addr; + u64 size = biosmap->size; + u64 end = start + size; + u32 type = biosmap->type; + + /* Overflow in 64 bits? Ignore the memory map. */ + if (start > end) + return -1; + + e820_add_region(start, size, type); + + biosmap++; + nr_map--; + } + return 0; +} + +/* + * Copy the BIOS e820 map into a safe place. + * + * Sanity-check it while we're at it.. + * + * If we're lucky and live on a modern system, the setup code + * will have given us a memory map that we can use to properly + * set up memory. If we aren't, we'll fake a memory map. + */ +static int __init append_e820_map(struct e820entry *biosmap, int nr_map) +{ + /* Only one memory region (or negative)? Ignore it */ + if (nr_map < 2) + return -1; + + return __append_e820_map(biosmap, nr_map); +} + +static u64 __init e820_update_range_map(struct e820map *e820x, u64 start, + u64 size, unsigned old_type, + unsigned new_type) +{ + int i; + u64 real_updated_size = 0; + + BUG_ON(old_type == new_type); + + if (size > (ULLONG_MAX - start)) + size = ULLONG_MAX - start; + + for (i = 0; i < e820.nr_map; i++) { + struct e820entry *ei = &e820x->map[i]; + u64 final_start, final_end; + if (ei->type != old_type) + continue; + /* totally covered? */ + if (ei->addr >= start && + (ei->addr + ei->size) <= (start + size)) { + ei->type = new_type; + real_updated_size += ei->size; + continue; + } + /* partially covered */ + final_start = max(start, ei->addr); + final_end = min(start + size, ei->addr + ei->size); + if (final_start >= final_end) + continue; + e820_add_region(final_start, final_end - final_start, + new_type); + real_updated_size += final_end - final_start; + + ei->size -= final_end - final_start; + if (ei->addr < final_start) + continue; + ei->addr = final_end; + } + return real_updated_size; +} + +u64 __init e820_update_range(u64 start, u64 size, unsigned old_type, + unsigned new_type) +{ + return e820_update_range_map(&e820, start, size, old_type, new_type); +} + +static u64 __init e820_update_range_saved(u64 start, u64 size, + unsigned old_type, unsigned new_type) +{ + return e820_update_range_map(&e820_saved, start, size, old_type, + new_type); +} + +/* make e820 not cover the range */ +u64 __init e820_remove_range(u64 start, u64 size, unsigned old_type, + int checktype) +{ + int i; + u64 real_removed_size = 0; + + if (size > (ULLONG_MAX - start)) + size = ULLONG_MAX - start; + + for (i = 0; i < e820.nr_map; i++) { + struct e820entry *ei = &e820.map[i]; + u64 final_start, final_end; + + if (checktype && ei->type != old_type) + continue; + /* totally covered? */ + if (ei->addr >= start && + (ei->addr + ei->size) <= (start + size)) { + real_removed_size += ei->size; + memset(ei, 0, sizeof(struct e820entry)); + continue; + } + /* partially covered */ + final_start = max(start, ei->addr); + final_end = min(start + size, ei->addr + ei->size); + if (final_start >= final_end) + continue; + real_removed_size += final_end - final_start; + + ei->size -= final_end - final_start; + if (ei->addr < final_start) + continue; + ei->addr = final_end; + } + return real_removed_size; +} + +void __init update_e820(void) +{ + int nr_map; + + nr_map = e820.nr_map; + 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"); + e820_print_map("modified"); +} +static void __init update_e820_saved(void) +{ + int nr_map; + + nr_map = e820_saved.nr_map; + if (sanitize_e820_map(e820_saved.map, ARRAY_SIZE(e820_saved.map), &nr_map)) + return; + e820_saved.nr_map = nr_map; +} +#define MAX_GAP_END 0x100000000ull +/* + * Search for a gap in the e820 memory space from start_addr to end_addr. + */ +__init int e820_search_gap(unsigned long *gapstart, unsigned long *gapsize, + unsigned long start_addr, unsigned long long end_addr) +{ + unsigned long long last; + int i = e820.nr_map; + int found = 0; + + last = (end_addr && end_addr < MAX_GAP_END) ? end_addr : MAX_GAP_END; + + while (--i >= 0) { + unsigned long long start = e820.map[i].addr; + unsigned long long end = start + e820.map[i].size; + + if (end < start_addr) + continue; + + /* + * Since "last" is at most 4GB, we know we'll + * fit in 32 bits if this condition is true + */ + if (last > end) { + unsigned long gap = last - end; + + if (gap >= *gapsize) { + *gapsize = gap; + *gapstart = end; + found = 1; + } + } + if (start < last) + last = start; + } + return found; +} + +/* + * Search for the biggest gap in the low 32 bits of the e820 + * memory space. We pass this space to PCI to assign MMIO resources + * for hotplug or unconfigured devices in. + * Hopefully the BIOS let enough space left. + */ +__init void e820_setup_gap(void) +{ + unsigned long gapstart, gapsize, round; + int found; + + gapstart = 0x10000000; + gapsize = 0x400000; + found = e820_search_gap(&gapstart, &gapsize, 0, MAX_GAP_END); + +#ifdef CONFIG_X86_64 + if (!found) { + gapstart = (max_pfn << PAGE_SHIFT) + 1024*1024; + printk(KERN_ERR "PCI: Warning: Cannot find a gap in the 32bit " + "address range\n" + KERN_ERR "PCI: Unassigned devices with 32bit resource " + "registers may break!\n"); + } +#endif + + /* + * See how much we want to round up: start off with + * rounding to the next 1MB area. + */ + round = 0x100000; + while ((gapsize >> 4) > round) + round += round; + /* Fun with two's complement */ + pci_mem_start = (gapstart + round) & -round; + + printk(KERN_INFO + "Allocating PCI resources starting at %lx (gap: %lx:%lx)\n", + pci_mem_start, gapstart, gapsize); +} + +/** + * Because of the size limitation of struct boot_params, only first + * 128 E820 memory entries are passed to kernel via + * 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, unsigned long pa_data) +{ + u32 map_len; + int entries; + struct e820entry *extmap; + + entries = sdata->len / sizeof(struct e820entry); + map_len = sdata->len + sizeof(struct setup_data); + if (map_len > PAGE_SIZE) + sdata = early_ioremap(pa_data, map_len); + extmap = (struct e820entry *)(sdata->data); + __append_e820_map(extmap, entries); + sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &e820.nr_map); + if (map_len > PAGE_SIZE) + early_iounmap(sdata, map_len); + printk(KERN_INFO "extended physical RAM map:\n"); + e820_print_map("extended"); +} + +#if defined(CONFIG_X86_64) || \ + (defined(CONFIG_X86_32) && defined(CONFIG_HIBERNATION)) +/** + * Find the ranges of physical addresses that do not correspond to + * e820 RAM areas and mark the corresponding pages as nosave for + * hibernation (32 bit) or software suspend and suspend to RAM (64 bit). + * + * This function requires the e820 map to be sorted and without any + * overlapping entries and assumes the first e820 area to be RAM. + */ +void __init e820_mark_nosave_regions(unsigned long limit_pfn) +{ + int i; + unsigned long pfn; + + pfn = PFN_DOWN(e820.map[0].addr + e820.map[0].size); + for (i = 1; i < e820.nr_map; i++) { + struct e820entry *ei = &e820.map[i]; + + if (pfn < PFN_UP(ei->addr)) + register_nosave_region(pfn, PFN_UP(ei->addr)); + + pfn = PFN_DOWN(ei->addr + ei->size); + if (ei->type != E820_RAM && ei->type != E820_RESERVED_KERN) + register_nosave_region(PFN_UP(ei->addr), pfn); + + if (pfn >= limit_pfn) + break; + } +} +#endif + +/* + * Early reserved memory areas. + */ +#define MAX_EARLY_RES 20 + +struct early_res { + u64 start, end; + char name[16]; + char overlap_ok; +}; +static struct early_res early_res[MAX_EARLY_RES] __initdata = { + { 0, PAGE_SIZE, "BIOS data page" }, /* BIOS data page */ +#if defined(CONFIG_X86_64) && defined(CONFIG_X86_TRAMPOLINE) + { TRAMPOLINE_BASE, TRAMPOLINE_BASE + 2 * PAGE_SIZE, "TRAMPOLINE" }, +#endif +#if defined(CONFIG_X86_32) && defined(CONFIG_SMP) + /* + * But first pinch a few for the stack/trampoline stuff + * FIXME: Don't need the extra page at 4K, but need to fix + * trampoline before removing it. (see the GDT stuff) + */ + { PAGE_SIZE, PAGE_SIZE + PAGE_SIZE, "EX TRAMPOLINE" }, + /* + * Has to be in very low memory so we can execute + * real-mode AP code. + */ + { TRAMPOLINE_BASE, TRAMPOLINE_BASE + PAGE_SIZE, "TRAMPOLINE" }, +#endif + {} +}; + +static int __init find_overlapped_early(u64 start, u64 end) +{ + int i; + struct early_res *r; + + for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) { + r = &early_res[i]; + if (end > r->start && start < r->end) + break; + } + + return i; +} + +/* + * Drop the i-th range from the early reservation map, + * by copying any higher ranges down one over it, and + * clearing what had been the last slot. + */ +static void __init drop_range(int i) +{ + int j; + + for (j = i + 1; j < MAX_EARLY_RES && early_res[j].end; j++) + ; + + memmove(&early_res[i], &early_res[i + 1], + (j - 1 - i) * sizeof(struct early_res)); + + early_res[j - 1].end = 0; +} + +/* + * Split any existing ranges that: + * 1) are marked 'overlap_ok', and + * 2) overlap with the stated range [start, end) + * into whatever portion (if any) of the existing range is entirely + * below or entirely above the stated range. Drop the portion + * of the existing range that overlaps with the stated range, + * which will allow the caller of this routine to then add that + * stated range without conflicting with any existing range. + */ +static void __init drop_overlaps_that_are_ok(u64 start, u64 end) +{ + int i; + struct early_res *r; + u64 lower_start, lower_end; + u64 upper_start, upper_end; + char name[16]; + + for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) { + r = &early_res[i]; + + /* Continue past non-overlapping ranges */ + if (end <= r->start || start >= r->end) + continue; + + /* + * Leave non-ok overlaps as is; let caller + * panic "Overlapping early reservations" + * when it hits this overlap. + */ + if (!r->overlap_ok) + return; + + /* + * We have an ok overlap. We will drop it from the early + * reservation map, and add back in any non-overlapping + * portions (lower or upper) as separate, overlap_ok, + * non-overlapping ranges. + */ + + /* 1. Note any non-overlapping (lower or upper) ranges. */ + strncpy(name, r->name, sizeof(name) - 1); + + lower_start = lower_end = 0; + upper_start = upper_end = 0; + if (r->start < start) { + lower_start = r->start; + lower_end = start; + } + if (r->end > end) { + upper_start = end; + upper_end = r->end; + } + + /* 2. Drop the original ok overlapping range */ + drop_range(i); + + i--; /* resume for-loop on copied down entry */ + + /* 3. Add back in any non-overlapping ranges. */ + if (lower_end) + reserve_early_overlap_ok(lower_start, lower_end, name); + if (upper_end) + reserve_early_overlap_ok(upper_start, upper_end, name); + } +} + +static void __init __reserve_early(u64 start, u64 end, char *name, + int overlap_ok) +{ + int i; + struct early_res *r; + + i = find_overlapped_early(start, end); + if (i >= MAX_EARLY_RES) + panic("Too many early reservations"); + r = &early_res[i]; + if (r->end) + panic("Overlapping early reservations " + "%llx-%llx %s to %llx-%llx %s\n", + start, end - 1, name?name:"", r->start, + r->end - 1, r->name); + r->start = start; + r->end = end; + r->overlap_ok = overlap_ok; + if (name) + strncpy(r->name, name, sizeof(r->name) - 1); +} + +/* + * A few early reservtations come here. + * + * The 'overlap_ok' in the name of this routine does -not- mean it + * is ok for these reservations to overlap an earlier reservation. + * Rather it means that it is ok for subsequent reservations to + * overlap this one. + * + * Use this entry point to reserve early ranges when you are doing + * so out of "Paranoia", reserving perhaps more memory than you need, + * just in case, and don't mind a subsequent overlapping reservation + * that is known to be needed. + * + * The drop_overlaps_that_are_ok() call here isn't really needed. + * It would be needed if we had two colliding 'overlap_ok' + * reservations, so that the second such would not panic on the + * overlap with the first. We don't have any such as of this + * writing, but might as well tolerate such if it happens in + * the future. + */ +void __init reserve_early_overlap_ok(u64 start, u64 end, char *name) +{ + drop_overlaps_that_are_ok(start, end); + __reserve_early(start, end, name, 1); +} + +/* + * Most early reservations come here. + * + * We first have drop_overlaps_that_are_ok() drop any pre-existing + * 'overlap_ok' ranges, so that we can then reserve this memory + * range without risk of panic'ing on an overlapping overlap_ok + * early reservation. + */ +void __init reserve_early(u64 start, u64 end, char *name) +{ + drop_overlaps_that_are_ok(start, end); + __reserve_early(start, end, name, 0); +} + +void __init free_early(u64 start, u64 end) +{ + struct early_res *r; + int i; + + i = find_overlapped_early(start, end); + r = &early_res[i]; + if (i >= MAX_EARLY_RES || r->end != end || r->start != start) + panic("free_early on not reserved area: %llx-%llx!", + start, end - 1); + + drop_range(i); +} + +void __init early_res_to_bootmem(u64 start, u64 end) +{ + int i, count; + u64 final_start, final_end; + + count = 0; + for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) + count++; + + printk(KERN_INFO "(%d early reservations) ==> bootmem\n", count); + for (i = 0; i < count; i++) { + struct early_res *r = &early_res[i]; + printk(KERN_INFO " #%d [%010llx - %010llx] %16s", i, + r->start, r->end, r->name); + final_start = max(start, r->start); + final_end = min(end, r->end); + if (final_start >= final_end) { + printk(KERN_CONT "\n"); + continue; + } + printk(KERN_CONT " ==> [%010llx - %010llx]\n", + final_start, final_end); + reserve_bootmem_generic(final_start, final_end - final_start, + BOOTMEM_DEFAULT); + } +} + +/* Check for already reserved areas */ +static inline int __init bad_addr(u64 *addrp, u64 size, u64 align) +{ + int i; + u64 addr = *addrp; + int changed = 0; + struct early_res *r; +again: + i = find_overlapped_early(addr, addr + size); + r = &early_res[i]; + if (i < MAX_EARLY_RES && r->end) { + *addrp = addr = round_up(r->end, align); + changed = 1; + goto again; + } + return changed; +} + +/* Check for already reserved areas */ +static inline int __init bad_addr_size(u64 *addrp, u64 *sizep, u64 align) +{ + int i; + u64 addr = *addrp, last; + u64 size = *sizep; + int changed = 0; +again: + last = addr + size; + for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) { + struct early_res *r = &early_res[i]; + if (last > r->start && addr < r->start) { + size = r->start - addr; + changed = 1; + goto again; + } + if (last > r->end && addr < r->end) { + addr = round_up(r->end, align); + size = last - addr; + changed = 1; + goto again; + } + if (last <= r->end && addr >= r->start) { + (*sizep)++; + return 0; + } + } + if (changed) { + *addrp = addr; + *sizep = size; + } + return changed; +} + +/* + * Find a free area with specified alignment in a specific range. + */ +u64 __init find_e820_area(u64 start, u64 end, u64 size, u64 align) +{ + int i; + + for (i = 0; i < e820.nr_map; i++) { + struct e820entry *ei = &e820.map[i]; + u64 addr, last; + u64 ei_last; + + if (ei->type != E820_RAM) + continue; + addr = round_up(ei->addr, align); + ei_last = ei->addr + ei->size; + if (addr < start) + addr = round_up(start, align); + if (addr >= ei_last) + continue; + while (bad_addr(&addr, size, align) && addr+size <= ei_last) + ; + last = addr + size; + if (last > ei_last) + continue; + if (last > end) + continue; + return addr; + } + return -1ULL; +} + +/* + * Find next free range after *start + */ +u64 __init find_e820_area_size(u64 start, u64 *sizep, u64 align) +{ + int i; + + for (i = 0; i < e820.nr_map; i++) { + struct e820entry *ei = &e820.map[i]; + u64 addr, last; + u64 ei_last; + + if (ei->type != E820_RAM) + continue; + addr = round_up(ei->addr, align); + ei_last = ei->addr + ei->size; + if (addr < start) + addr = round_up(start, align); + if (addr >= ei_last) + continue; + *sizep = ei_last - addr; + while (bad_addr_size(&addr, sizep, align) && + addr + *sizep <= ei_last) + ; + last = addr + *sizep; + if (last > ei_last) + continue; + return addr; + } + return -1UL; + +} + +/* + * pre allocated 4k and reserved it in e820 + */ +u64 __init early_reserve_e820(u64 startt, u64 sizet, u64 align) +{ + u64 size = 0; + u64 addr; + u64 start; + + start = startt; + while (size < sizet) + start = find_e820_area_size(start, &size, align); + + if (size < sizet) + return 0; + + addr = round_down(start + size - sizet, align); + e820_update_range(addr, sizet, E820_RAM, E820_RESERVED); + e820_update_range_saved(addr, sizet, E820_RAM, E820_RESERVED); + printk(KERN_INFO "update e820 for early_reserve_e820\n"); + update_e820(); + update_e820_saved(); + + return addr; +} + +#ifdef CONFIG_X86_32 +# ifdef CONFIG_X86_PAE +# define MAX_ARCH_PFN (1ULL<<(36-PAGE_SHIFT)) +# else +# define MAX_ARCH_PFN (1ULL<<(32-PAGE_SHIFT)) +# endif +#else /* CONFIG_X86_32 */ +# define MAX_ARCH_PFN MAXMEM>>PAGE_SHIFT +#endif + +/* + * Find the highest page frame number we have available + */ +static unsigned long __init e820_end_pfn(unsigned long limit_pfn, unsigned type) +{ + int i; + unsigned long last_pfn = 0; + unsigned long max_arch_pfn = MAX_ARCH_PFN; + + for (i = 0; i < e820.nr_map; i++) { + struct e820entry *ei = &e820.map[i]; + unsigned long start_pfn; + unsigned long end_pfn; + + if (ei->type != type) + continue; + + start_pfn = ei->addr >> PAGE_SHIFT; + end_pfn = (ei->addr + ei->size) >> PAGE_SHIFT; + + if (start_pfn >= limit_pfn) + continue; + if (end_pfn > limit_pfn) { + last_pfn = limit_pfn; + break; + } + if (end_pfn > last_pfn) + last_pfn = end_pfn; + } + + if (last_pfn > max_arch_pfn) + last_pfn = max_arch_pfn; + + printk(KERN_INFO "last_pfn = %#lx max_arch_pfn = %#lx\n", + last_pfn, max_arch_pfn); + return last_pfn; +} +unsigned long __init e820_end_of_ram_pfn(void) +{ + return e820_end_pfn(MAX_ARCH_PFN, E820_RAM); +} + +unsigned long __init e820_end_of_low_ram_pfn(void) +{ + return e820_end_pfn(1UL<<(32 - PAGE_SHIFT), E820_RAM); +} +/* + * Finds an active region in the address range from start_pfn to last_pfn and + * returns its range in ei_startpfn and ei_endpfn for the e820 entry. + */ +int __init e820_find_active_region(const struct e820entry *ei, + unsigned long start_pfn, + unsigned long last_pfn, + unsigned long *ei_startpfn, + unsigned long *ei_endpfn) +{ + u64 align = PAGE_SIZE; + + *ei_startpfn = round_up(ei->addr, align) >> PAGE_SHIFT; + *ei_endpfn = round_down(ei->addr + ei->size, align) >> PAGE_SHIFT; + + /* Skip map entries smaller than a page */ + if (*ei_startpfn >= *ei_endpfn) + return 0; + + /* Skip if map is outside the node */ + if (ei->type != E820_RAM || *ei_endpfn <= start_pfn || + *ei_startpfn >= last_pfn) + return 0; + + /* Check for overlaps */ + if (*ei_startpfn < start_pfn) + *ei_startpfn = start_pfn; + if (*ei_endpfn > last_pfn) + *ei_endpfn = last_pfn; + + return 1; +} + +/* Walk the e820 map and register active regions within a node */ +void __init e820_register_active_regions(int nid, unsigned long start_pfn, + unsigned long last_pfn) +{ + unsigned long ei_startpfn; + unsigned long ei_endpfn; + int i; + + for (i = 0; i < e820.nr_map; i++) + if (e820_find_active_region(&e820.map[i], + start_pfn, last_pfn, + &ei_startpfn, &ei_endpfn)) + add_active_range(nid, ei_startpfn, ei_endpfn); +} + +/* + * Find the hole size (in bytes) in the memory range. + * @start: starting address of the memory range to scan + * @end: ending address of the memory range to scan + */ +u64 __init e820_hole_size(u64 start, u64 end) +{ + unsigned long start_pfn = start >> PAGE_SHIFT; + unsigned long last_pfn = end >> PAGE_SHIFT; + unsigned long ei_startpfn, ei_endpfn, ram = 0; + int i; + + for (i = 0; i < e820.nr_map; i++) { + if (e820_find_active_region(&e820.map[i], + start_pfn, last_pfn, + &ei_startpfn, &ei_endpfn)) + ram += ei_endpfn - ei_startpfn; + } + return end - start - ((u64)ram << PAGE_SHIFT); +} + +static void early_panic(char *msg) +{ + early_printk(msg); + panic(msg); +} + +static int userdef __initdata; + +/* "mem=nopentium" disables the 4MB page tables. */ +static int __init parse_memopt(char *p) +{ + u64 mem_size; + + if (!p) + return -EINVAL; + +#ifdef CONFIG_X86_32 + if (!strcmp(p, "nopentium")) { + setup_clear_cpu_cap(X86_FEATURE_PSE); + return 0; + } +#endif + + userdef = 1; + mem_size = memparse(p, &p); + e820_remove_range(mem_size, ULLONG_MAX - mem_size, E820_RAM, 1); + + return 0; +} +early_param("mem", parse_memopt); + +static int __init parse_memmap_opt(char *p) +{ + char *oldp; + u64 start_at, mem_size; + + if (!p) + return -EINVAL; + + if (!strcmp(p, "exactmap")) { +#ifdef CONFIG_CRASH_DUMP + /* + * If we are doing a crash dump, we still need to know + * the real mem size before original memory map is + * reset. + */ + saved_max_pfn = e820_end_of_ram_pfn(); +#endif + e820.nr_map = 0; + userdef = 1; + return 0; + } + + oldp = p; + mem_size = memparse(p, &p); + if (p == oldp) + return -EINVAL; + + userdef = 1; + if (*p == '@') { + start_at = memparse(p+1, &p); + e820_add_region(start_at, mem_size, E820_RAM); + } else if (*p == '#') { + start_at = memparse(p+1, &p); + e820_add_region(start_at, mem_size, E820_ACPI); + } else if (*p == '$') { + start_at = memparse(p+1, &p); + e820_add_region(start_at, mem_size, E820_RESERVED); + } else + e820_remove_range(mem_size, ULLONG_MAX - mem_size, E820_RAM, 1); + + return *p == '\0' ? 0 : -EINVAL; +} +early_param("memmap", parse_memmap_opt); + +void __init finish_e820_parsing(void) +{ + if (userdef) { + int nr = e820.nr_map; + + if (sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &nr) < 0) + early_panic("Invalid user supplied memory map"); + e820.nr_map = nr; + + printk(KERN_INFO "user-defined physical RAM map:\n"); + e820_print_map("user"); + } +} + +static inline const char *e820_type_to_string(int e820_type) +{ + switch (e820_type) { + case E820_RESERVED_KERN: + case E820_RAM: return "System RAM"; + case E820_ACPI: return "ACPI Tables"; + case E820_NVS: return "ACPI Non-volatile Storage"; + default: return "reserved"; + } +} + +/* + * Mark e820 reserved areas as busy for the resource manager. + */ +void __init e820_reserve_resources(void) +{ + int i; + struct resource *res; + u64 end; + + res = alloc_bootmem_low(sizeof(struct resource) * e820.nr_map); + for (i = 0; i < e820.nr_map; i++) { + end = e820.map[i].addr + e820.map[i].size - 1; +#ifndef CONFIG_RESOURCES_64BIT + if (end > 0x100000000ULL) { + res++; + continue; + } +#endif + res->name = e820_type_to_string(e820.map[i].type); + res->start = e820.map[i].addr; + res->end = end; + + res->flags = IORESOURCE_MEM | IORESOURCE_BUSY; + insert_resource(&iomem_resource, res); + res++; + } + + 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, + e820_type_to_string(entry->type)); + } +} + +/* + * Non-standard memory setup can be specified via this quirk: + */ +char * (*arch_memory_setup_quirk)(void); + +char *__init default_machine_specific_memory_setup(void) +{ + char *who = "BIOS-e820"; + int new_nr; + /* + * Try to copy the BIOS-supplied E820-map. + * + * Otherwise fake a memory map; one section from 0k->640k, + * the next section from 1mb->appropriate_mem_k + */ + new_nr = boot_params.e820_entries; + sanitize_e820_map(boot_params.e820_map, + ARRAY_SIZE(boot_params.e820_map), + &new_nr); + boot_params.e820_entries = new_nr; + if (append_e820_map(boot_params.e820_map, boot_params.e820_entries) + < 0) { + u64 mem_size; + + /* compare results from other methods and take the greater */ + if (boot_params.alt_mem_k + < boot_params.screen_info.ext_mem_k) { + mem_size = boot_params.screen_info.ext_mem_k; + who = "BIOS-88"; + } else { + mem_size = boot_params.alt_mem_k; + who = "BIOS-e801"; + } + + e820.nr_map = 0; + e820_add_region(0, LOWMEMSIZE(), E820_RAM); + e820_add_region(HIGH_MEMORY, mem_size << 10, E820_RAM); + } + + /* In case someone cares... */ + return who; +} + +char *__init __attribute__((weak)) machine_specific_memory_setup(void) +{ + if (arch_memory_setup_quirk) { + char *who = arch_memory_setup_quirk(); + + if (who) + return who; + } + return default_machine_specific_memory_setup(); +} + +/* Overridden in paravirt.c if CONFIG_PARAVIRT */ +char * __init __attribute__((weak)) memory_setup(void) +{ + return machine_specific_memory_setup(); +} + +void __init setup_memory_map(void) +{ + char *who; + + who = memory_setup(); + memcpy(&e820_saved, &e820, sizeof(struct e820map)); + printk(KERN_INFO "BIOS-provided physical RAM map:\n"); + e820_print_map(who); +} + +#ifdef CONFIG_X86_64 +int __init arch_get_ram_range(int slot, u64 *addr, u64 *size) +{ + int i; + + if (slot < 0 || slot >= e820.nr_map) + return -1; + for (i = slot; i < e820.nr_map; i++) { + if (e820.map[i].type != E820_RAM) + continue; + break; + } + if (i == e820.nr_map || e820.map[i].addr > (max_pfn << PAGE_SHIFT)) + return -1; + *addr = e820.map[i].addr; + *size = min_t(u64, e820.map[i].size + e820.map[i].addr, + max_pfn << PAGE_SHIFT) - *addr; + return i + 1; +} +#endif diff --git a/arch/x86/kernel/e820_32.c b/arch/x86/kernel/e820_32.c deleted file mode 100644 index ed733e7cf4e..00000000000 --- a/arch/x86/kernel/e820_32.c +++ /dev/null @@ -1,775 +0,0 @@ -#include <linux/kernel.h> -#include <linux/types.h> -#include <linux/init.h> -#include <linux/bootmem.h> -#include <linux/ioport.h> -#include <linux/string.h> -#include <linux/kexec.h> -#include <linux/module.h> -#include <linux/mm.h> -#include <linux/pfn.h> -#include <linux/uaccess.h> -#include <linux/suspend.h> - -#include <asm/pgtable.h> -#include <asm/page.h> -#include <asm/e820.h> -#include <asm/setup.h> - -struct e820map e820; -struct change_member { - struct e820entry *pbios; /* pointer to original bios entry */ - unsigned long long addr; /* address for this change point */ -}; -static struct change_member change_point_list[2*E820MAX] __initdata; -static struct change_member *change_point[2*E820MAX] __initdata; -static struct e820entry *overlap_list[E820MAX] __initdata; -static struct e820entry new_bios[E820MAX] __initdata; -/* For PCI or other memory-mapped resources */ -unsigned long pci_mem_start = 0x10000000; -#ifdef CONFIG_PCI -EXPORT_SYMBOL(pci_mem_start); -#endif -extern int user_defined_memmap; - -static struct resource system_rom_resource = { - .name = "System ROM", - .start = 0xf0000, - .end = 0xfffff, - .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM -}; - -static struct resource extension_rom_resource = { - .name = "Extension ROM", - .start = 0xe0000, - .end = 0xeffff, - .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM -}; - -static struct resource adapter_rom_resources[] = { { - .name = "Adapter ROM", - .start = 0xc8000, - .end = 0, - .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM -}, { - .name = "Adapter ROM", - .start = 0, - .end = 0, - .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM -}, { - .name = "Adapter ROM", - .start = 0, - .end = 0, - .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM -}, { - .name = "Adapter ROM", - .start = 0, - .end = 0, - .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM -}, { - .name = "Adapter ROM", - .start = 0, - .end = 0, - .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM -}, { - .name = "Adapter ROM", - .start = 0, - .end = 0, - .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM -} }; - -static struct resource video_rom_resource = { - .name = "Video ROM", - .start = 0xc0000, - .end = 0xc7fff, - .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM -}; - -#define ROMSIGNATURE 0xaa55 - -static int __init romsignature(const unsigned char *rom) -{ - const unsigned short * const ptr = (const unsigned short *)rom; - unsigned short sig; - - return probe_kernel_address(ptr, sig) == 0 && sig == ROMSIGNATURE; -} - -static int __init romchecksum(const unsigned char *rom, unsigned long length) -{ - unsigned char sum, c; - - for (sum = 0; length && probe_kernel_address(rom++, c) == 0; length--) - sum += c; - return !length && !sum; -} - -static void __init probe_roms(void) -{ - const unsigned char *rom; - unsigned long start, length, upper; - unsigned char c; - int i; - - /* video rom */ - upper = adapter_rom_resources[0].start; - for (start = video_rom_resource.start; start < upper; start += 2048) { - rom = isa_bus_to_virt(start); - if (!romsignature(rom)) - continue; - - video_rom_resource.start = start; - - if (probe_kernel_address(rom + 2, c) != 0) - continue; - - /* 0 < length <= 0x7f * 512, historically */ - length = c * 512; - - /* if checksum okay, trust length byte */ - if (length && romchecksum(rom, length)) - video_rom_resource.end = start + length - 1; - - request_resource(&iomem_resource, &video_rom_resource); - break; - } - - start = (video_rom_resource.end + 1 + 2047) & ~2047UL; - if (start < upper) - start = upper; - - /* system rom */ - request_resource(&iomem_resource, &system_rom_resource); - upper = system_rom_resource.start; - - /* check for extension rom (ignore length byte!) */ - rom = isa_bus_to_virt(extension_rom_resource.start); - if (romsignature(rom)) { - length = extension_rom_resource.end - extension_rom_resource.start + 1; - if (romchecksum(rom, length)) { - request_resource(&iomem_resource, &extension_rom_resource); - upper = extension_rom_resource.start; - } - } - - /* check for adapter roms on 2k boundaries */ - for (i = 0; i < ARRAY_SIZE(adapter_rom_resources) && start < upper; start += 2048) { - rom = isa_bus_to_virt(start); - if (!romsignature(rom)) - continue; - - if (probe_kernel_address(rom + 2, c) != 0) - continue; - - /* 0 < length <= 0x7f * 512, historically */ - length = c * 512; - - /* but accept any length that fits if checksum okay */ - if (!length || start + length > upper || !romchecksum(rom, length)) - continue; - - adapter_rom_resources[i].start = start; - adapter_rom_resources[i].end = start + length - 1; - request_resource(&iomem_resource, &adapter_rom_resources[i]); - - start = adapter_rom_resources[i++].end & ~2047UL; - } -} - -/* - * Request address space for all standard RAM and ROM resources - * and also for regions reported as reserved by the e820. - */ -void __init init_iomem_resources(struct resource *code_resource, - struct resource *data_resource, - struct resource *bss_resource) -{ - int i; - - probe_roms(); - for (i = 0; i < e820.nr_map; i++) { - struct resource *res; -#ifndef CONFIG_RESOURCES_64BIT - if (e820.map[i].addr + e820.map[i].size > 0x100000000ULL) - continue; -#endif - res = kzalloc(sizeof(struct resource), GFP_ATOMIC); - switch (e820.map[i].type) { - case E820_RAM: res->name = "System RAM"; break; - case E820_ACPI: res->name = "ACPI Tables"; break; - case E820_NVS: res->name = "ACPI Non-volatile Storage"; break; - default: res->name = "reserved"; - } - res->start = e820.map[i].addr; - res->end = res->start + e820.map[i].size - 1; - res->flags = IORESOURCE_MEM | IORESOURCE_BUSY; - if (request_resource(&iomem_resource, res)) { - kfree(res); - continue; - } - if (e820.map[i].type == E820_RAM) { - /* - * We don't know which RAM region contains kernel data, - * so we try it repeatedly and let the resource manager - * test it. - */ - request_resource(res, code_resource); - request_resource(res, data_resource); - request_resource(res, bss_resource); -#ifdef CONFIG_KEXEC - if (crashk_res.start != crashk_res.end) - request_resource(res, &crashk_res); -#endif - } - } -} - -#if defined(CONFIG_PM) && defined(CONFIG_HIBERNATION) -/** - * e820_mark_nosave_regions - Find the ranges of physical addresses that do not - * correspond to e820 RAM areas and mark the corresponding pages as nosave for - * hibernation. - * - * This function requires the e820 map to be sorted and without any - * overlapping entries and assumes the first e820 area to be RAM. - */ -void __init e820_mark_nosave_regions(void) -{ - int i; - unsigned long pfn; - - pfn = PFN_DOWN(e820.map[0].addr + e820.map[0].size); - for (i = 1; i < e820.nr_map; i++) { - struct e820entry *ei = &e820.map[i]; - - if (pfn < PFN_UP(ei->addr)) - register_nosave_region(pfn, PFN_UP(ei->addr)); - - pfn = PFN_DOWN(ei->addr + ei->size); - if (ei->type != E820_RAM) - register_nosave_region(PFN_UP(ei->addr), pfn); - - if (pfn >= max_low_pfn) - break; - } -} -#endif - -void __init add_memory_region(unsigned long long start, - unsigned long long size, int type) -{ - int x; - - x = e820.nr_map; - - if (x == E820MAX) { - printk(KERN_ERR "Ooops! Too many entries in the memory map!\n"); - return; - } - - e820.map[x].addr = start; - e820.map[x].size = size; - e820.map[x].type = type; - e820.nr_map++; -} /* add_memory_region */ - -/* - * Sanitize the BIOS e820 map. - * - * Some e820 responses include overlapping entries. The following - * replaces the original e820 map with a new one, removing overlaps. - * - */ -int __init sanitize_e820_map(struct e820entry * biosmap, char * pnr_map) -{ - struct change_member *change_tmp; - unsigned long current_type, last_type; - unsigned long long last_addr; - int chgidx, still_changing; - int overlap_entries; - int new_bios_entry; - int old_nr, new_nr, chg_nr; - int i; - - /* - Visually we're performing the following (1,2,3,4 = memory types)... - - Sample memory map (w/overlaps): - ____22__________________ - ______________________4_ - ____1111________________ - _44_____________________ - 11111111________________ - ____________________33__ - ___________44___________ - __________33333_________ - ______________22________ - ___________________2222_ - _________111111111______ - _____________________11_ - _________________4______ - - Sanitized equivalent (no overlap): - 1_______________________ - _44_____________________ - ___1____________________ - ____22__________________ - ______11________________ - _________1______________ - __________3_____________ - ___________44___________ - _____________33_________ - _______________2________ - ________________1_______ - _________________4______ - ___________________2____ - ____________________33__ - ______________________4_ - */ - /* if there's only one memory region, don't bother */ - if (*pnr_map < 2) { - return -1; - } - - old_nr = *pnr_map; - - /* bail out if we find any unreasonable addresses in bios map */ - for (i=0; i<old_nr; i++) - if (biosmap[i].addr + biosmap[i].size < biosmap[i].addr) { - return -1; - } - - /* create pointers for initial change-point information (for sorting) */ - for (i=0; i < 2*old_nr; i++) - change_point[i] = &change_point_list[i]; - - /* record all known change-points (starting and ending addresses), - omitting those that are for empty memory regions */ - chgidx = 0; - for (i=0; i < old_nr; i++) { - if (biosmap[i].size != 0) { - change_point[chgidx]->addr = biosmap[i].addr; - change_point[chgidx++]->pbios = &biosmap[i]; - change_point[chgidx]->addr = biosmap[i].addr + biosmap[i].size; - change_point[chgidx++]->pbios = &biosmap[i]; - } - } - chg_nr = chgidx; /* true number of change-points */ - - /* sort change-point list by memory addresses (low -> high) */ - still_changing = 1; - while (still_changing) { - still_changing = 0; - for (i=1; i < chg_nr; i++) { - /* if <current_addr> > <last_addr>, swap */ - /* or, if current=<start_addr> & last=<end_addr>, swap */ - if ((change_point[i]->addr < change_point[i-1]->addr) || - ((change_point[i]->addr == change_point[i-1]->addr) && - (change_point[i]->addr == change_point[i]->pbios->addr) && - (change_point[i-1]->addr != change_point[i-1]->pbios->addr)) - ) - { - change_tmp = change_point[i]; - change_point[i] = change_point[i-1]; - change_point[i-1] = change_tmp; - still_changing=1; - } - } - } - - /* create a new bios memory map, removing overlaps */ - overlap_entries=0; /* number of entries in the overlap table */ - new_bios_entry=0; /* index for creating new bios map entries */ - last_type = 0; /* start with undefined memory type */ - last_addr = 0; /* start with 0 as last starting address */ - /* loop through change-points, determining affect on the new bios map */ - for (chgidx=0; chgidx < chg_nr; chgidx++) - { - /* keep track of all overlapping bios entries */ - if (change_point[chgidx]->addr == change_point[chgidx]->pbios->addr) - { - /* add map entry to overlap list (> 1 entry implies an overlap) */ - overlap_list[overlap_entries++]=change_point[chgidx]->pbios; - } - else - { - /* remove entry from list (order independent, so swap with last) */ - for (i=0; i<overlap_entries; i++) - { - if (overlap_list[i] == change_point[chgidx]->pbios) - overlap_list[i] = overlap_list[overlap_entries-1]; - } - overlap_entries--; - } - /* if there are overlapping entries, decide which "type" to use */ - /* (larger value takes precedence -- 1=usable, 2,3,4,4+=unusable) */ - current_type = 0; - for (i=0; i<overlap_entries; i++) - if (overlap_list[i]->type > current_type) - current_type = overlap_list[i]->type; - /* continue building up new bios map based on this information */ - if (current_type != last_type) { - if (last_type != 0) { - new_bios[new_bios_entry].size = - change_point[chgidx]->addr - last_addr; - /* move forward only if the new size was non-zero */ - if (new_bios[new_bios_entry].size != 0) - if (++new_bios_entry >= E820MAX) - break; /* no more space left for new bios entries */ - } - if (current_type != 0) { - new_bios[new_bios_entry].addr = change_point[chgidx]->addr; - new_bios[new_bios_entry].type = current_type; - last_addr=change_point[chgidx]->addr; - } - last_type = current_type; - } - } - new_nr = new_bios_entry; /* retain count for new bios entries */ - - /* copy new bios mapping into original location */ - memcpy(biosmap, new_bios, new_nr*sizeof(struct e820entry)); - *pnr_map = new_nr; - - return 0; -} - -/* - * Copy the BIOS e820 map into a safe place. - * - * Sanity-check it while we're at it.. - * - * If we're lucky and live on a modern system, the setup code - * will have given us a memory map that we can use to properly - * set up memory. If we aren't, we'll fake a memory map. - * - * We check to see that the memory map contains at least 2 elements - * before we'll use it, because the detection code in setup.S may - * not be perfect and most every PC known to man has two memory - * regions: one from 0 to 640k, and one from 1mb up. (The IBM - * thinkpad 560x, for example, does not cooperate with the memory - * detection code.) - */ -int __init copy_e820_map(struct e820entry *biosmap, int nr_map) -{ - /* Only one memory region (or negative)? Ignore it */ - if (nr_map < 2) - return -1; - - do { - u64 start = biosmap->addr; - u64 size = biosmap->size; - u64 end = start + size; - u32 type = biosmap->type; - - /* Overflow in 64 bits? Ignore the memory map. */ - if (start > end) - return -1; - - add_memory_region(start, size, type); - } while (biosmap++, --nr_map); - - return 0; -} - -/* - * Find the highest page frame number we have available - */ -void __init propagate_e820_map(void) -{ - int i; - - max_pfn = 0; - - for (i = 0; i < e820.nr_map; i++) { - unsigned long start, end; - /* RAM? */ - if (e820.map[i].type != E820_RAM) - continue; - start = PFN_UP(e820.map[i].addr); - end = PFN_DOWN(e820.map[i].addr + e820.map[i].size); - if (start >= end) - continue; - if (end > max_pfn) - max_pfn = end; - memory_present(0, start, end); - } -} - -/* - * Register fully available low RAM pages with the bootmem allocator. - */ -void __init register_bootmem_low_pages(unsigned long max_low_pfn) -{ - int i; - - for (i = 0; i < e820.nr_map; i++) { - unsigned long curr_pfn, last_pfn, size; - /* - * Reserve usable low memory - */ - if (e820.map[i].type != E820_RAM) - continue; - /* - * We are rounding up the start address of usable memory: - */ - curr_pfn = PFN_UP(e820.map[i].addr); - if (curr_pfn >= max_low_pfn) - continue; - /* - * ... and at the end of the usable range downwards: - */ - last_pfn = PFN_DOWN(e820.map[i].addr + e820.map[i].size); - - if (last_pfn > max_low_pfn) - last_pfn = max_low_pfn; - - /* - * .. finally, did all the rounding and playing - * around just make the area go away? - */ - if (last_pfn <= curr_pfn) - continue; - - size = last_pfn - curr_pfn; - free_bootmem(PFN_PHYS(curr_pfn), PFN_PHYS(size)); - } -} - -void __init e820_register_memory(void) -{ - unsigned long gapstart, gapsize, round; - unsigned long long last; - int i; - - /* - * Search for the biggest gap in the low 32 bits of the e820 - * memory space. - */ - last = 0x100000000ull; - gapstart = 0x10000000; - gapsize = 0x400000; - i = e820.nr_map; - while (--i >= 0) { - unsigned long long start = e820.map[i].addr; - unsigned long long end = start + e820.map[i].size; - - /* - * Since "last" is at most 4GB, we know we'll - * fit in 32 bits if this condition is true - */ - if (last > end) { - unsigned long gap = last - end; - - if (gap > gapsize) { - gapsize = gap; - gapstart = end; - } - } - if (start < last) - last = start; - } - - /* - * See how much we want to round up: start off with - * rounding to the next 1MB area. - */ - round = 0x100000; - while ((gapsize >> 4) > round) - round += round; - /* Fun with two's complement */ - pci_mem_start = (gapstart + round) & -round; - - printk("Allocating PCI resources starting at %08lx (gap: %08lx:%08lx)\n", - pci_mem_start, gapstart, gapsize); -} - -void __init print_memory_map(char *who) -{ - int i; - - for (i = 0; i < e820.nr_map; i++) { - printk(" %s: %016Lx - %016Lx ", who, - e820.map[i].addr, - e820.map[i].addr + e820.map[i].size); - switch (e820.map[i].type) { - case E820_RAM: printk("(usable)\n"); - break; - case E820_RESERVED: - printk("(reserved)\n"); - break; - case E820_ACPI: - printk("(ACPI data)\n"); - break; - case E820_NVS: - printk("(ACPI NVS)\n"); - break; - default: printk("type %u\n", e820.map[i].type); - break; - } - } -} - -void __init limit_regions(unsigned long long size) -{ - unsigned long long current_addr; - int i; - - print_memory_map("limit_regions start"); - for (i = 0; i < e820.nr_map; i++) { - current_addr = e820.map[i].addr + e820.map[i].size; - if (current_addr < size) - continue; - - if (e820.map[i].type != E820_RAM) - continue; - - if (e820.map[i].addr >= size) { - /* - * This region starts past the end of the - * requested size, skip it completely. - */ - e820.nr_map = i; - } else { - e820.nr_map = i + 1; - e820.map[i].size -= current_addr - size; - } - print_memory_map("limit_regions endfor"); - return; - } - print_memory_map("limit_regions endfunc"); -} - -/* - * This function checks if any part of the range <start,end> is mapped - * with type. - */ -int -e820_any_mapped(u64 start, u64 end, unsigned type) -{ - int i; - for (i = 0; i < e820.nr_map; i++) { - const struct e820entry *ei = &e820.map[i]; - if (type && ei->type != type) - continue; - if (ei->addr >= end || ei->addr + ei->size <= start) - continue; - return 1; - } - return 0; -} -EXPORT_SYMBOL_GPL(e820_any_mapped); - - /* - * This function checks if the entire range <start,end> is mapped with type. - * - * Note: this function only works correct if the e820 table is sorted and - * not-overlapping, which is the case - */ -int __init -e820_all_mapped(unsigned long s, unsigned long e, unsigned type) -{ - u64 start = s; - u64 end = e; - int i; - for (i = 0; i < e820.nr_map; i++) { - struct e820entry *ei = &e820.map[i]; - if (type && ei->type != type) - continue; - /* is the region (part) in overlap with the current region ?*/ - if (ei->addr >= end || ei->addr + ei->size <= start) - continue; - /* if the region is at the beginning of <start,end> we move - * start to the end of the region since it's ok until there - */ - if (ei->addr <= start) - start = ei->addr + ei->size; - /* if start is now at or beyond end, we're done, full - * coverage */ - if (start >= end) - return 1; /* we're done */ - } - return 0; -} - -static int __init parse_memmap(char *arg) -{ - if (!arg) - return -EINVAL; - - if (strcmp(arg, "exactmap") == 0) { -#ifdef CONFIG_CRASH_DUMP - /* If we are doing a crash dump, we - * still need to know the real mem - * size before original memory map is - * reset. - */ - propagate_e820_map(); - saved_max_pfn = max_pfn; -#endif - e820.nr_map = 0; - user_defined_memmap = 1; - } else { - /* If the user specifies memory size, we - * limit the BIOS-provided memory map to - * that size. exactmap can be used to specify - * the exact map. mem=number can be used to - * trim the existing memory map. - */ - unsigned long long start_at, mem_size; - - mem_size = memparse(arg, &arg); - if (*arg == '@') { - start_at = memparse(arg+1, &arg); - add_memory_region(start_at, mem_size, E820_RAM); - } else if (*arg == '#') { - start_at = memparse(arg+1, &arg); - add_memory_region(start_at, mem_size, E820_ACPI); - } else if (*arg == '$') { - start_at = memparse(arg+1, &arg); - add_memory_region(start_at, mem_size, E820_RESERVED); - } else { - limit_regions(mem_size); - user_defined_memmap = 1; - } - } - return 0; -} -early_param("memmap", parse_memmap); -void __init update_memory_range(u64 start, u64 size, unsigned old_type, - unsigned new_type) -{ - int i; - - BUG_ON(old_type == new_type); - - for (i = 0; i < e820.nr_map; i++) { - struct e820entry *ei = &e820.map[i]; - u64 final_start, final_end; - if (ei->type != old_type) - continue; - /* totally covered? */ - if (ei->addr >= start && ei->size <= size) { - ei->type = new_type; - continue; - } - /* partially covered */ - final_start = max(start, ei->addr); - final_end = min(start + size, ei->addr + ei->size); - if (final_start >= final_end) - continue; - add_memory_region(final_start, final_end - final_start, - new_type); - } -} -void __init update_e820(void) -{ - u8 nr_map; - - nr_map = e820.nr_map; - if (sanitize_e820_map(e820.map, &nr_map)) - return; - e820.nr_map = nr_map; - printk(KERN_INFO "modified physical RAM map:\n"); - print_memory_map("modified"); -} diff --git a/arch/x86/kernel/e820_64.c b/arch/x86/kernel/e820_64.c deleted file mode 100644 index 124480c0008..00000000000 --- a/arch/x86/kernel/e820_64.c +++ /dev/null @@ -1,952 +0,0 @@ -/* - * Handle the memory map. - * The functions here do the job until bootmem takes over. - * - * Getting sanitize_e820_map() in sync with i386 version by applying change: - * - Provisions for empty E820 memory regions (reported by certain BIOSes). - * Alex Achenbach <xela@slit.de>, December 2002. - * Venkatesh Pallipadi <venkatesh.pallipadi@intel.com> - * - */ -#include <linux/kernel.h> -#include <linux/types.h> -#include <linux/init.h> -#include <linux/bootmem.h> -#include <linux/ioport.h> -#include <linux/string.h> -#include <linux/kexec.h> -#include <linux/module.h> -#include <linux/mm.h> -#include <linux/suspend.h> -#include <linux/pfn.h> - -#include <asm/pgtable.h> -#include <asm/page.h> -#include <asm/e820.h> -#include <asm/proto.h> -#include <asm/setup.h> -#include <asm/sections.h> -#include <asm/kdebug.h> -#include <asm/trampoline.h> - -struct e820map e820; - -/* - * PFN of last memory page. - */ -unsigned long end_pfn; - -/* - * 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. - */ -unsigned long max_pfn_mapped; - -/* - * Last pfn which the user wants to use. - */ -static unsigned long __initdata end_user_pfn = MAXMEM>>PAGE_SHIFT; - -/* - * Early reserved memory areas. - */ -#define MAX_EARLY_RES 20 - -struct early_res { - unsigned long start, end; - char name[16]; -}; -static struct early_res early_res[MAX_EARLY_RES] __initdata = { - { 0, PAGE_SIZE, "BIOS data page" }, /* BIOS data page */ -#ifdef CONFIG_X86_TRAMPOLINE - { TRAMPOLINE_BASE, TRAMPOLINE_BASE + 2 * PAGE_SIZE, "TRAMPOLINE" }, -#endif - {} -}; - -void __init reserve_early(unsigned long start, unsigned long end, char *name) -{ - int i; - struct early_res *r; - for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) { - r = &early_res[i]; - if (end > r->start && start < r->end) - panic("Overlapping early reservations %lx-%lx %s to %lx-%lx %s\n", - start, end - 1, name?name:"", r->start, r->end - 1, r->name); - } - if (i >= MAX_EARLY_RES) - panic("Too many early reservations"); - r = &early_res[i]; - r->start = start; - r->end = end; - if (name) - strncpy(r->name, name, sizeof(r->name) - 1); -} - -void __init free_early(unsigned long start, unsigned long end) -{ - struct early_res *r; - int i, j; - - for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) { - r = &early_res[i]; - if (start == r->start && end == r->end) - break; - } - if (i >= MAX_EARLY_RES || !early_res[i].end) - panic("free_early on not reserved area: %lx-%lx!", start, end); - - for (j = i + 1; j < MAX_EARLY_RES && early_res[j].end; j++) - ; - - memmove(&early_res[i], &early_res[i + 1], - (j - 1 - i) * sizeof(struct early_res)); - - early_res[j - 1].end = 0; -} - -void __init early_res_to_bootmem(unsigned long start, unsigned long end) -{ - int i; - unsigned long final_start, final_end; - for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) { - struct early_res *r = &early_res[i]; - final_start = max(start, r->start); - final_end = min(end, r->end); - if (final_start >= final_end) - continue; - printk(KERN_INFO " early res: %d [%lx-%lx] %s\n", i, - final_start, final_end - 1, r->name); - reserve_bootmem_generic(final_start, final_end - final_start); - } -} - -/* Check for already reserved areas */ -static inline int __init -bad_addr(unsigned long *addrp, unsigned long size, unsigned long align) -{ - int i; - unsigned long addr = *addrp, last; - int changed = 0; -again: - last = addr + size; - for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) { - struct early_res *r = &early_res[i]; - if (last >= r->start && addr < r->end) { - *addrp = addr = round_up(r->end, align); - changed = 1; - goto again; - } - } - return changed; -} - -/* Check for already reserved areas */ -static inline int __init -bad_addr_size(unsigned long *addrp, unsigned long *sizep, unsigned long align) -{ - int i; - unsigned long addr = *addrp, last; - unsigned long size = *sizep; - int changed = 0; -again: - last = addr + size; - for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) { - struct early_res *r = &early_res[i]; - if (last > r->start && addr < r->start) { - size = r->start - addr; - changed = 1; - goto again; - } - if (last > r->end && addr < r->end) { - addr = round_up(r->end, align); - size = last - addr; - changed = 1; - goto again; - } - if (last <= r->end && addr >= r->start) { - (*sizep)++; - return 0; - } - } - if (changed) { - *addrp = addr; - *sizep = size; - } - return changed; -} -/* - * This function checks if any part of the range <start,end> is mapped - * with type. - */ -int -e820_any_mapped(unsigned long start, unsigned long end, unsigned type) -{ - int i; - - for (i = 0; i < e820.nr_map; i++) { - struct e820entry *ei = &e820.map[i]; - - if (type && ei->type != type) - continue; - if (ei->addr >= end || ei->addr + ei->size <= start) - continue; - return 1; - } - return 0; -} -EXPORT_SYMBOL_GPL(e820_any_mapped); - -/* - * This function checks if the entire range <start,end> is mapped with type. - * - * Note: this function only works correct if the e820 table is sorted and - * not-overlapping, which is the case - */ -int __init e820_all_mapped(unsigned long start, unsigned long end, - unsigned type) -{ - int i; - - for (i = 0; i < e820.nr_map; i++) { - struct e820entry *ei = &e820.map[i]; - - if (type && ei->type != type) - continue; - /* is the region (part) in overlap with the current region ?*/ - if (ei->addr >= end || ei->addr + ei->size <= start) - continue; - - /* if the region is at the beginning of <start,end> we move - * start to the end of the region since it's ok until there - */ - if (ei->addr <= start) - start = ei->addr + ei->size; - /* - * if start is now at or beyond end, we're done, full - * coverage - */ - if (start >= end) - return 1; - } - return 0; -} - -/* - * Find a free area with specified alignment in a specific range. - */ -unsigned long __init find_e820_area(unsigned long start, unsigned long end, - unsigned long size, unsigned long align) -{ - int i; - - for (i = 0; i < e820.nr_map; i++) { - struct e820entry *ei = &e820.map[i]; - unsigned long addr, last; - unsigned long ei_last; - - if (ei->type != E820_RAM) - continue; - addr = round_up(ei->addr, align); - ei_last = ei->addr + ei->size; - if (addr < start) - addr = round_up(start, align); - if (addr >= ei_last) - continue; - while (bad_addr(&addr, size, align) && addr+size <= ei_last) - ; - last = addr + size; - if (last > ei_last) - continue; - if (last > end) - continue; - return addr; - } - return -1UL; -} - -/* - * Find next free range after *start - */ -unsigned long __init find_e820_area_size(unsigned long start, - unsigned long *sizep, - unsigned long align) -{ - int i; - - for (i = 0; i < e820.nr_map; i++) { - struct e820entry *ei = &e820.map[i]; - unsigned long addr, last; - unsigned long ei_last; - - if (ei->type != E820_RAM) - continue; - addr = round_up(ei->addr, align); - ei_last = ei->addr + ei->size; - if (addr < start) - addr = round_up(start, align); - if (addr >= ei_last) - continue; - *sizep = ei_last - addr; - while (bad_addr_size(&addr, sizep, align) && - addr + *sizep <= ei_last) - ; - last = addr + *sizep; - if (last > ei_last) - continue; - return addr; - } - return -1UL; - -} -/* - * Find the highest page frame number we have available - */ -unsigned long __init e820_end_of_ram(void) -{ - unsigned long end_pfn; - - end_pfn = find_max_pfn_with_active_regions(); - - if (end_pfn > max_pfn_mapped) - max_pfn_mapped = end_pfn; - if (max_pfn_mapped > MAXMEM>>PAGE_SHIFT) - max_pfn_mapped = MAXMEM>>PAGE_SHIFT; - if (end_pfn > end_user_pfn) - end_pfn = end_user_pfn; - if (end_pfn > max_pfn_mapped) - end_pfn = max_pfn_mapped; - - printk(KERN_INFO "max_pfn_mapped = %lu\n", max_pfn_mapped); - return end_pfn; -} - -/* - * Mark e820 reserved areas as busy for the resource manager. - */ -void __init e820_reserve_resources(void) -{ - int i; - struct resource *res; - - res = alloc_bootmem_low(sizeof(struct resource) * e820.nr_map); - for (i = 0; i < e820.nr_map; i++) { - switch (e820.map[i].type) { - case E820_RAM: res->name = "System RAM"; break; - case E820_ACPI: res->name = "ACPI Tables"; break; - case E820_NVS: res->name = "ACPI Non-volatile Storage"; break; - default: res->name = "reserved"; - } - res->start = e820.map[i].addr; - res->end = res->start + e820.map[i].size - 1; - res->flags = IORESOURCE_MEM | IORESOURCE_BUSY; - insert_resource(&iomem_resource, res); - res++; - } -} - -/* - * Find the ranges of physical addresses that do not correspond to - * e820 RAM areas and mark the corresponding pages as nosave for software - * suspend and suspend to RAM. - * - * This function requires the e820 map to be sorted and without any - * overlapping entries and assumes the first e820 area to be RAM. - */ -void __init e820_mark_nosave_regions(void) -{ - int i; - unsigned long paddr; - - paddr = round_down(e820.map[0].addr + e820.map[0].size, PAGE_SIZE); - for (i = 1; i < e820.nr_map; i++) { - struct e820entry *ei = &e820.map[i]; - - if (paddr < ei->addr) - register_nosave_region(PFN_DOWN(paddr), - PFN_UP(ei->addr)); - - paddr = round_down(ei->addr + ei->size, PAGE_SIZE); - if (ei->type != E820_RAM) - register_nosave_region(PFN_UP(ei->addr), - PFN_DOWN(paddr)); - - if (paddr >= (end_pfn << PAGE_SHIFT)) - break; - } -} - -/* - * Finds an active region in the address range from start_pfn to end_pfn and - * returns its range in ei_startpfn and ei_endpfn for the e820 entry. - */ -static int __init e820_find_active_region(const struct e820entry *ei, - unsigned long start_pfn, - unsigned long end_pfn, - unsigned long *ei_startpfn, - unsigned long *ei_endpfn) -{ - *ei_startpfn = round_up(ei->addr, PAGE_SIZE) >> PAGE_SHIFT; - *ei_endpfn = round_down(ei->addr + ei->size, PAGE_SIZE) >> PAGE_SHIFT; - - /* Skip map entries smaller than a page */ - if (*ei_startpfn >= *ei_endpfn) - return 0; - - /* Check if max_pfn_mapped should be updated */ - if (ei->type != E820_RAM && *ei_endpfn > max_pfn_mapped) - max_pfn_mapped = *ei_endpfn; - - /* Skip if map is outside the node */ - if (ei->type != E820_RAM || *ei_endpfn <= start_pfn || - *ei_startpfn >= end_pfn) - return 0; - - /* Check for overlaps */ - if (*ei_startpfn < start_pfn) - *ei_startpfn = start_pfn; - if (*ei_endpfn > end_pfn) - *ei_endpfn = end_pfn; - - /* Obey end_user_pfn to save on memmap */ - if (*ei_startpfn >= end_user_pfn) - return 0; - if (*ei_endpfn > end_user_pfn) - *ei_endpfn = end_user_pfn; - - return 1; -} - -/* Walk the e820 map and register active regions within a node */ -void __init -e820_register_active_regions(int nid, unsigned long start_pfn, - unsigned long end_pfn) -{ - unsigned long ei_startpfn; - unsigned long ei_endpfn; - int i; - - for (i = 0; i < e820.nr_map; i++) - if (e820_find_active_region(&e820.map[i], - start_pfn, end_pfn, - &ei_startpfn, &ei_endpfn)) - add_active_range(nid, ei_startpfn, ei_endpfn); -} - -/* - * Add a memory region to the kernel e820 map. - */ -void __init add_memory_region(unsigned long start, unsigned long size, int type) -{ - int x = e820.nr_map; - - if (x == E820MAX) { - printk(KERN_ERR "Ooops! Too many entries in the memory map!\n"); - return; - } - - e820.map[x].addr = start; - e820.map[x].size = size; - e820.map[x].type = type; - e820.nr_map++; -} - -/* - * Find the hole size (in bytes) in the memory range. - * @start: starting address of the memory range to scan - * @end: ending address of the memory range to scan - */ -unsigned long __init e820_hole_size(unsigned long start, unsigned long end) -{ - unsigned long start_pfn = start >> PAGE_SHIFT; - unsigned long end_pfn = end >> PAGE_SHIFT; - unsigned long ei_startpfn, ei_endpfn, ram = 0; - int i; - - for (i = 0; i < e820.nr_map; i++) { - if (e820_find_active_region(&e820.map[i], - start_pfn, end_pfn, - &ei_startpfn, &ei_endpfn)) - ram += ei_endpfn - ei_startpfn; - } - return end - start - (ram << PAGE_SHIFT); -} - -static void __init e820_print_map(char *who) -{ - int i; - - for (i = 0; i < e820.nr_map; i++) { - printk(KERN_INFO " %s: %016Lx - %016Lx ", who, - (unsigned long long) e820.map[i].addr, - (unsigned long long) - (e820.map[i].addr + e820.map[i].size)); - switch (e820.map[i].type) { - case E820_RAM: - printk(KERN_CONT "(usable)\n"); - break; - case E820_RESERVED: - printk(KERN_CONT "(reserved)\n"); - break; - case E820_ACPI: - printk(KERN_CONT "(ACPI data)\n"); - break; - case E820_NVS: - printk(KERN_CONT "(ACPI NVS)\n"); - break; - default: - printk(KERN_CONT "type %u\n", e820.map[i].type); - break; - } - } -} - -/* - * Sanitize the BIOS e820 map. - * - * Some e820 responses include overlapping entries. The following - * replaces the original e820 map with a new one, removing overlaps. - * - */ -static int __init sanitize_e820_map(struct e820entry *biosmap, char *pnr_map) -{ - struct change_member { - struct e820entry *pbios; /* pointer to original bios entry */ - unsigned long long addr; /* address for this change point */ - }; - static struct change_member change_point_list[2*E820MAX] __initdata; - static struct change_member *change_point[2*E820MAX] __initdata; - static struct e820entry *overlap_list[E820MAX] __initdata; - static struct e820entry new_bios[E820MAX] __initdata; - struct change_member *change_tmp; - unsigned long current_type, last_type; - unsigned long long last_addr; - int chgidx, still_changing; - int overlap_entries; - int new_bios_entry; - int old_nr, new_nr, chg_nr; - int i; - - /* - Visually we're performing the following - (1,2,3,4 = memory types)... - - Sample memory map (w/overlaps): - ____22__________________ - ______________________4_ - ____1111________________ - _44_____________________ - 11111111________________ - ____________________33__ - ___________44___________ - __________33333_________ - ______________22________ - ___________________2222_ - _________111111111______ - _____________________11_ - _________________4______ - - Sanitized equivalent (no overlap): - 1_______________________ - _44_____________________ - ___1____________________ - ____22__________________ - ______11________________ - _________1______________ - __________3_____________ - ___________44___________ - _____________33_________ - _______________2________ - ________________1_______ - _________________4______ - ___________________2____ - ____________________33__ - ______________________4_ - */ - - /* if there's only one memory region, don't bother */ - if (*pnr_map < 2) - return -1; - - old_nr = *pnr_map; - - /* bail out if we find any unreasonable addresses in bios map */ - for (i = 0; i < old_nr; i++) - if (biosmap[i].addr + biosmap[i].size < biosmap[i].addr) - return -1; - - /* create pointers for initial change-point information (for sorting) */ - for (i = 0; i < 2 * old_nr; i++) - change_point[i] = &change_point_list[i]; - - /* record all known change-points (starting and ending addresses), - omitting those that are for empty memory regions */ - chgidx = 0; - for (i = 0; i < old_nr; i++) { - if (biosmap[i].size != 0) { - change_point[chgidx]->addr = biosmap[i].addr; - change_point[chgidx++]->pbios = &biosmap[i]; - change_point[chgidx]->addr = biosmap[i].addr + - biosmap[i].size; - change_point[chgidx++]->pbios = &biosmap[i]; - } - } - chg_nr = chgidx; - - /* sort change-point list by memory addresses (low -> high) */ - still_changing = 1; - while (still_changing) { - still_changing = 0; - for (i = 1; i < chg_nr; i++) { - unsigned long long curaddr, lastaddr; - unsigned long long curpbaddr, lastpbaddr; - - curaddr = change_point[i]->addr; - lastaddr = change_point[i - 1]->addr; - curpbaddr = change_point[i]->pbios->addr; - lastpbaddr = change_point[i - 1]->pbios->addr; - - /* - * swap entries, when: - * - * curaddr > lastaddr or - * curaddr == lastaddr and curaddr == curpbaddr and - * lastaddr != lastpbaddr - */ - if (curaddr < lastaddr || - (curaddr == lastaddr && curaddr == curpbaddr && - lastaddr != lastpbaddr)) { - change_tmp = change_point[i]; - change_point[i] = change_point[i-1]; - change_point[i-1] = change_tmp; - still_changing = 1; - } - } - } - - /* create a new bios memory map, removing overlaps */ - overlap_entries = 0; /* number of entries in the overlap table */ - new_bios_entry = 0; /* index for creating new bios map entries */ - last_type = 0; /* start with undefined memory type */ - last_addr = 0; /* start with 0 as last starting address */ - - /* loop through change-points, determining affect on the new bios map */ - for (chgidx = 0; chgidx < chg_nr; chgidx++) { - /* keep track of all overlapping bios entries */ - if (change_point[chgidx]->addr == - change_point[chgidx]->pbios->addr) { - /* - * add map entry to overlap list (> 1 entry - * implies an overlap) - */ - overlap_list[overlap_entries++] = - change_point[chgidx]->pbios; - } else { - /* - * remove entry from list (order independent, - * so swap with last) - */ - for (i = 0; i < overlap_entries; i++) { - if (overlap_list[i] == - change_point[chgidx]->pbios) - overlap_list[i] = - overlap_list[overlap_entries-1]; - } - overlap_entries--; - } - /* - * if there are overlapping entries, decide which - * "type" to use (larger value takes precedence -- - * 1=usable, 2,3,4,4+=unusable) - */ - current_type = 0; - for (i = 0; i < overlap_entries; i++) - if (overlap_list[i]->type > current_type) - current_type = overlap_list[i]->type; - /* - * continue building up new bios map based on this - * information - */ - if (current_type != last_type) { - if (last_type != 0) { - new_bios[new_bios_entry].size = - change_point[chgidx]->addr - last_addr; - /* - * move forward only if the new size - * was non-zero - */ - if (new_bios[new_bios_entry].size != 0) - /* - * no more space left for new - * bios entries ? - */ - if (++new_bios_entry >= E820MAX) - break; - } - if (current_type != 0) { - new_bios[new_bios_entry].addr = - change_point[chgidx]->addr; - new_bios[new_bios_entry].type = current_type; - last_addr = change_point[chgidx]->addr; - } - last_type = current_type; - } - } - /* retain count for new bios entries */ - new_nr = new_bios_entry; - - /* copy new bios mapping into original location */ - memcpy(biosmap, new_bios, new_nr * sizeof(struct e820entry)); - *pnr_map = new_nr; - - return 0; -} - -/* - * Copy the BIOS e820 map into a safe place. - * - * Sanity-check it while we're at it.. - * - * If we're lucky and live on a modern system, the setup code - * will have given us a memory map that we can use to properly - * set up memory. If we aren't, we'll fake a memory map. - */ -static int __init copy_e820_map(struct e820entry *biosmap, int nr_map) -{ - /* Only one memory region (or negative)? Ignore it */ - if (nr_map < 2) - return -1; - - do { - u64 start = biosmap->addr; - u64 size = biosmap->size; - u64 end = start + size; - u32 type = biosmap->type; - - /* Overflow in 64 bits? Ignore the memory map. */ - if (start > end) - return -1; - - add_memory_region(start, size, type); - } while (biosmap++, --nr_map); - return 0; -} - -static void early_panic(char *msg) -{ - early_printk(msg); - panic(msg); -} - -/* We're not void only for x86 32-bit compat */ -char * __init machine_specific_memory_setup(void) -{ - char *who = "BIOS-e820"; - /* - * Try to copy the BIOS-supplied E820-map. - * - * Otherwise fake a memory map; one section from 0k->640k, - * the next section from 1mb->appropriate_mem_k - */ - sanitize_e820_map(boot_params.e820_map, &boot_params.e820_entries); - if (copy_e820_map(boot_params.e820_map, boot_params.e820_entries) < 0) - early_panic("Cannot find a valid memory map"); - printk(KERN_INFO "BIOS-provided physical RAM map:\n"); - e820_print_map(who); - - /* In case someone cares... */ - return who; -} - -static int __init parse_memopt(char *p) -{ - if (!p) - return -EINVAL; - end_user_pfn = memparse(p, &p); - end_user_pfn >>= PAGE_SHIFT; - return 0; -} -early_param("mem", parse_memopt); - -static int userdef __initdata; - -static int __init parse_memmap_opt(char *p) -{ - char *oldp; - unsigned long long start_at, mem_size; - - if (!strcmp(p, "exactmap")) { -#ifdef CONFIG_CRASH_DUMP - /* - * If we are doing a crash dump, we still need to know - * the real mem size before original memory map is - * reset. - */ - e820_register_active_regions(0, 0, -1UL); - saved_max_pfn = e820_end_of_ram(); - remove_all_active_ranges(); -#endif - max_pfn_mapped = 0; - e820.nr_map = 0; - userdef = 1; - return 0; - } - - oldp = p; - mem_size = memparse(p, &p); - if (p == oldp) - return -EINVAL; - - userdef = 1; - if (*p == '@') { - start_at = memparse(p+1, &p); - add_memory_region(start_at, mem_size, E820_RAM); - } else if (*p == '#') { - start_at = memparse(p+1, &p); - add_memory_region(start_at, mem_size, E820_ACPI); - } else if (*p == '$') { - start_at = memparse(p+1, &p); - add_memory_region(start_at, mem_size, E820_RESERVED); - } else { - end_user_pfn = (mem_size >> PAGE_SHIFT); - } - return *p == '\0' ? 0 : -EINVAL; -} -early_param("memmap", parse_memmap_opt); - -void __init finish_e820_parsing(void) -{ - if (userdef) { - char nr = e820.nr_map; - - if (sanitize_e820_map(e820.map, &nr) < 0) - early_panic("Invalid user supplied memory map"); - e820.nr_map = nr; - - printk(KERN_INFO "user-defined physical RAM map:\n"); - e820_print_map("user"); - } -} - -void __init update_memory_range(u64 start, u64 size, unsigned old_type, - unsigned new_type) -{ - int i; - - BUG_ON(old_type == new_type); - - for (i = 0; i < e820.nr_map; i++) { - struct e820entry *ei = &e820.map[i]; - u64 final_start, final_end; - if (ei->type != old_type) - continue; - /* totally covered? */ - if (ei->addr >= start && ei->size <= size) { - ei->type = new_type; - continue; - } - /* partially covered */ - final_start = max(start, ei->addr); - final_end = min(start + size, ei->addr + ei->size); - if (final_start >= final_end) - continue; - add_memory_region(final_start, final_end - final_start, - new_type); - } -} - -void __init update_e820(void) -{ - u8 nr_map; - - nr_map = e820.nr_map; - if (sanitize_e820_map(e820.map, &nr_map)) - return; - e820.nr_map = nr_map; - printk(KERN_INFO "modified physical RAM map:\n"); - e820_print_map("modified"); -} - -unsigned long pci_mem_start = 0xaeedbabe; -EXPORT_SYMBOL(pci_mem_start); - -/* - * Search for the biggest gap in the low 32 bits of the e820 - * memory space. We pass this space to PCI to assign MMIO resources - * for hotplug or unconfigured devices in. - * Hopefully the BIOS let enough space left. - */ -__init void e820_setup_gap(void) -{ - unsigned long gapstart, gapsize, round; - unsigned long last; - int i; - int found = 0; - - last = 0x100000000ull; - gapstart = 0x10000000; - gapsize = 0x400000; - i = e820.nr_map; - while (--i >= 0) { - unsigned long long start = e820.map[i].addr; - unsigned long long end = start + e820.map[i].size; - - /* - * Since "last" is at most 4GB, we know we'll - * fit in 32 bits if this condition is true - */ - if (last > end) { - unsigned long gap = last - end; - - if (gap > gapsize) { - gapsize = gap; - gapstart = end; - found = 1; - } - } - if (start < last) - last = start; - } - - if (!found) { - gapstart = (end_pfn << PAGE_SHIFT) + 1024*1024; - printk(KERN_ERR "PCI: Warning: Cannot find a gap in the 32bit " - "address range\n" - KERN_ERR "PCI: Unassigned devices with 32bit resource " - "registers may break!\n"); - } - - /* - * See how much we want to round up: start off with - * rounding to the next 1MB area. - */ - round = 0x100000; - while ((gapsize >> 4) > round) - round += round; - /* Fun with two's complement */ - pci_mem_start = (gapstart + round) & -round; - - printk(KERN_INFO - "Allocating PCI resources starting at %lx (gap: %lx:%lx)\n", - pci_mem_start, gapstart, gapsize); -} - -int __init arch_get_ram_range(int slot, u64 *addr, u64 *size) -{ - int i; - - if (slot < 0 || slot >= e820.nr_map) - return -1; - for (i = slot; i < e820.nr_map; i++) { - if (e820.map[i].type != E820_RAM) - continue; - break; - } - if (i == e820.nr_map || e820.map[i].addr > (max_pfn << PAGE_SHIFT)) - return -1; - *addr = e820.map[i].addr; - *size = min_t(u64, e820.map[i].size + e820.map[i].addr, - max_pfn << PAGE_SHIFT) - *addr; - return i + 1; -} diff --git a/arch/x86/kernel/early-quirks.c b/arch/x86/kernel/early-quirks.c index 9f51e1ea9e8..a4665f37cfc 100644 --- a/arch/x86/kernel/early-quirks.c +++ b/arch/x86/kernel/early-quirks.c @@ -50,7 +50,7 @@ static void __init fix_hypertransport_config(int num, int slot, int func) static void __init via_bugs(int num, int slot, int func) { #ifdef CONFIG_GART_IOMMU - if ((end_pfn > MAX_DMA32_PFN || force_iommu) && + if ((max_pfn > MAX_DMA32_PFN || force_iommu) && !gart_iommu_aperture_allowed) { printk(KERN_INFO "Looks like a VIA chipset. Disabling IOMMU." @@ -98,17 +98,6 @@ static void __init nvidia_bugs(int num, int slot, int func) } -static void __init ati_bugs(int num, int slot, int func) -{ -#ifdef CONFIG_X86_IO_APIC - if (timer_over_8254 == 1) { - timer_over_8254 = 0; - printk(KERN_INFO - "ATI board detected. Disabling timer routing over 8254.\n"); - } -#endif -} - #define QFLAG_APPLY_ONCE 0x1 #define QFLAG_APPLIED 0x2 #define QFLAG_DONE (QFLAG_APPLY_ONCE|QFLAG_APPLIED) @@ -126,8 +115,6 @@ static struct chipset early_qrk[] __initdata = { PCI_CLASS_BRIDGE_PCI, PCI_ANY_ID, QFLAG_APPLY_ONCE, nvidia_bugs }, { PCI_VENDOR_ID_VIA, PCI_ANY_ID, PCI_CLASS_BRIDGE_PCI, PCI_ANY_ID, QFLAG_APPLY_ONCE, via_bugs }, - { PCI_VENDOR_ID_ATI, PCI_ANY_ID, - PCI_CLASS_BRIDGE_PCI, PCI_ANY_ID, QFLAG_APPLY_ONCE, ati_bugs }, { PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_K8_NB, PCI_CLASS_BRIDGE_HOST, PCI_ANY_ID, 0, fix_hypertransport_config }, {} diff --git a/arch/x86/kernel/efi.c b/arch/x86/kernel/efi.c index 77d424cf68b..06cc8d4254b 100644 --- a/arch/x86/kernel/efi.c +++ b/arch/x86/kernel/efi.c @@ -64,6 +64,17 @@ static int __init setup_noefi(char *arg) } early_param("noefi", setup_noefi); +int add_efi_memmap; +EXPORT_SYMBOL(add_efi_memmap); + +static int __init setup_add_efi_memmap(char *arg) +{ + add_efi_memmap = 1; + return 0; +} +early_param("add_efi_memmap", setup_add_efi_memmap); + + static efi_status_t virt_efi_get_time(efi_time_t *tm, efi_time_cap_t *tc) { return efi_call_virt2(get_time, tm, tc); @@ -213,6 +224,50 @@ unsigned long efi_get_time(void) eft.minute, eft.second); } +/* + * Tell the kernel about the EFI memory map. This might include + * more than the max 128 entries that can fit in the e820 legacy + * (zeropage) memory map. + */ + +static void __init do_add_efi_memmap(void) +{ + void *p; + + for (p = memmap.map; p < memmap.map_end; p += memmap.desc_size) { + efi_memory_desc_t *md = p; + unsigned long long start = md->phys_addr; + unsigned long long size = md->num_pages << EFI_PAGE_SHIFT; + int e820_type; + + if (md->attribute & EFI_MEMORY_WB) + e820_type = E820_RAM; + else + e820_type = E820_RESERVED; + e820_add_region(start, size, e820_type); + } + sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &e820.nr_map); +} + +void __init efi_reserve_early(void) +{ + unsigned long pmap; + +#ifdef CONFIG_X86_32 + pmap = boot_params.efi_info.efi_memmap; +#else + pmap = (boot_params.efi_info.efi_memmap | + ((__u64)boot_params.efi_info.efi_memmap_hi<<32)); +#endif + memmap.phys_map = (void *)pmap; + memmap.nr_map = boot_params.efi_info.efi_memmap_size / + boot_params.efi_info.efi_memdesc_size; + memmap.desc_version = boot_params.efi_info.efi_memdesc_version; + memmap.desc_size = boot_params.efi_info.efi_memdesc_size; + reserve_early(pmap, pmap + memmap.nr_map * memmap.desc_size, + "EFI memmap"); +} + #if EFI_DEBUG static void __init print_efi_memmap(void) { @@ -244,19 +299,11 @@ void __init efi_init(void) #ifdef CONFIG_X86_32 efi_phys.systab = (efi_system_table_t *)boot_params.efi_info.efi_systab; - memmap.phys_map = (void *)boot_params.efi_info.efi_memmap; #else efi_phys.systab = (efi_system_table_t *) (boot_params.efi_info.efi_systab | ((__u64)boot_params.efi_info.efi_systab_hi<<32)); - memmap.phys_map = (void *) - (boot_params.efi_info.efi_memmap | - ((__u64)boot_params.efi_info.efi_memmap_hi<<32)); #endif - memmap.nr_map = boot_params.efi_info.efi_memmap_size / - boot_params.efi_info.efi_memdesc_size; - memmap.desc_version = boot_params.efi_info.efi_memdesc_version; - memmap.desc_size = boot_params.efi_info.efi_memdesc_size; efi.systab = early_ioremap((unsigned long)efi_phys.systab, sizeof(efi_system_table_t)); @@ -370,6 +417,8 @@ void __init efi_init(void) if (memmap.desc_size != sizeof(efi_memory_desc_t)) printk(KERN_WARNING "Kernel-defined memdesc" "doesn't match the one from EFI!\n"); + if (add_efi_memmap) + do_add_efi_memmap(); /* Setup for EFI runtime service */ reboot_type = BOOT_EFI; @@ -424,7 +473,7 @@ void __init efi_enter_virtual_mode(void) size = md->num_pages << EFI_PAGE_SHIFT; end = md->phys_addr + size; - if (PFN_UP(end) <= max_pfn_mapped) + if (PFN_UP(end) <= max_low_pfn_mapped) va = __va(md->phys_addr); else va = efi_ioremap(md->phys_addr, size); diff --git a/arch/x86/kernel/efi_64.c b/arch/x86/kernel/efi_64.c index d0060fdccca..652c5287215 100644 --- a/arch/x86/kernel/efi_64.c +++ b/arch/x86/kernel/efi_64.c @@ -97,13 +97,7 @@ void __init efi_call_phys_epilog(void) early_runtime_code_mapping_set_exec(0); } -void __init efi_reserve_bootmem(void) -{ - reserve_bootmem_generic((unsigned long)memmap.phys_map, - memmap.nr_map * memmap.desc_size); -} - -void __iomem * __init efi_ioremap(unsigned long phys_addr, unsigned long size) +void __iomem *__init efi_ioremap(unsigned long phys_addr, unsigned long size) { static unsigned pages_mapped __initdata; unsigned i, pages; diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S index c778e4fa55a..cfe28a71543 100644 --- a/arch/x86/kernel/entry_32.S +++ b/arch/x86/kernel/entry_32.S @@ -51,14 +51,14 @@ #include <asm/percpu.h> #include <asm/dwarf2.h> #include <asm/processor-flags.h> -#include "irq_vectors.h" +#include <asm/irq_vectors.h> /* * We use macros for low-level operations which need to be overridden * for paravirtualization. The following will never clobber any registers: * INTERRUPT_RETURN (aka. "iret") * GET_CR0_INTO_EAX (aka. "movl %cr0, %eax") - * ENABLE_INTERRUPTS_SYSCALL_RET (aka "sti; sysexit"). + * ENABLE_INTERRUPTS_SYSEXIT (aka "sti; sysexit"). * * For DISABLE_INTERRUPTS/ENABLE_INTERRUPTS (aka "cli"/"sti"), you must * specify what registers can be overwritten (CLBR_NONE, CLBR_EAX/EDX/ECX/ANY). @@ -349,7 +349,7 @@ sysenter_past_esp: xorl %ebp,%ebp TRACE_IRQS_ON 1: mov PT_FS(%esp), %fs - ENABLE_INTERRUPTS_SYSCALL_RET + ENABLE_INTERRUPTS_SYSEXIT CFI_ENDPROC .pushsection .fixup,"ax" 2: movl $0,PT_FS(%esp) @@ -874,10 +874,10 @@ ENTRY(native_iret) .previous END(native_iret) -ENTRY(native_irq_enable_syscall_ret) +ENTRY(native_irq_enable_sysexit) sti sysexit -END(native_irq_enable_syscall_ret) +END(native_irq_enable_sysexit) #endif KPROBE_ENTRY(int3) @@ -1024,6 +1024,7 @@ ENTRY(xen_sysenter_target) RING0_INT_FRAME addl $5*4, %esp /* remove xen-provided frame */ jmp sysenter_past_esp + CFI_ENDPROC ENTRY(xen_hypervisor_callback) CFI_STARTPROC diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S index 556a8df522a..bb4e22f4892 100644 --- a/arch/x86/kernel/entry_64.S +++ b/arch/x86/kernel/entry_64.S @@ -59,8 +59,7 @@ #endif #ifdef CONFIG_PARAVIRT -ENTRY(native_irq_enable_syscall_ret) - movq %gs:pda_oldrsp,%rsp +ENTRY(native_usergs_sysret64) swapgs sysretq #endif /* CONFIG_PARAVIRT */ @@ -104,7 +103,7 @@ ENTRY(native_irq_enable_syscall_ret) .macro FAKE_STACK_FRAME child_rip /* push in order ss, rsp, eflags, cs, rip */ xorl %eax, %eax - pushq %rax /* ss */ + pushq $__KERNEL_DS /* ss */ CFI_ADJUST_CFA_OFFSET 8 /*CFI_REL_OFFSET ss,0*/ pushq %rax /* rsp */ @@ -169,13 +168,13 @@ ENTRY(ret_from_fork) CFI_ADJUST_CFA_OFFSET -4 call schedule_tail GET_THREAD_INFO(%rcx) - testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT),threadinfo_flags(%rcx) + testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT),TI_flags(%rcx) jnz rff_trace rff_action: RESTORE_REST testl $3,CS-ARGOFFSET(%rsp) # from kernel_thread? je int_ret_from_sys_call - testl $_TIF_IA32,threadinfo_flags(%rcx) + testl $_TIF_IA32,TI_flags(%rcx) jnz int_ret_from_sys_call RESTORE_TOP_OF_STACK %rdi,ARGOFFSET jmp ret_from_sys_call @@ -244,7 +243,8 @@ ENTRY(system_call_after_swapgs) movq %rcx,RIP-ARGOFFSET(%rsp) CFI_REL_OFFSET rip,RIP-ARGOFFSET GET_THREAD_INFO(%rcx) - testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP),threadinfo_flags(%rcx) + testl $(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT|_TIF_SECCOMP), \ + TI_flags(%rcx) jnz tracesys cmpq $__NR_syscall_max,%rax ja badsys @@ -263,7 +263,7 @@ sysret_check: GET_THREAD_INFO(%rcx) DISABLE_INTERRUPTS(CLBR_NONE) TRACE_IRQS_OFF - movl threadinfo_flags(%rcx),%edx + movl TI_flags(%rcx),%edx andl %edi,%edx jnz sysret_careful CFI_REMEMBER_STATE @@ -275,7 +275,8 @@ sysret_check: CFI_REGISTER rip,rcx RESTORE_ARGS 0,-ARG_SKIP,1 /*CFI_REGISTER rflags,r11*/ - ENABLE_INTERRUPTS_SYSCALL_RET + movq %gs:pda_oldrsp, %rsp + USERGS_SYSRET64 CFI_RESTORE_STATE /* Handle reschedules */ @@ -305,7 +306,7 @@ sysret_signal: leaq -ARGOFFSET(%rsp),%rdi # &pt_regs -> arg1 xorl %esi,%esi # oldset -> arg2 call ptregscall_common -1: movl $_TIF_NEED_RESCHED,%edi +1: movl $_TIF_WORK_MASK,%edi /* Use IRET because user could have changed frame. This works because ptregscall_common has called FIXUP_TOP_OF_STACK. */ DISABLE_INTERRUPTS(CLBR_NONE) @@ -347,10 +348,10 @@ int_ret_from_sys_call: int_with_check: LOCKDEP_SYS_EXIT_IRQ GET_THREAD_INFO(%rcx) - movl threadinfo_flags(%rcx),%edx + movl TI_flags(%rcx),%edx andl %edi,%edx jnz int_careful - andl $~TS_COMPAT,threadinfo_status(%rcx) + andl $~TS_COMPAT,TI_status(%rcx) jmp retint_swapgs /* Either reschedule or signal or syscall exit tracking needed. */ @@ -393,7 +394,7 @@ int_signal: movq %rsp,%rdi # &ptregs -> arg1 xorl %esi,%esi # oldset -> arg2 call do_notify_resume -1: movl $_TIF_NEED_RESCHED,%edi +1: movl $_TIF_WORK_MASK,%edi int_restore_rest: RESTORE_REST DISABLE_INTERRUPTS(CLBR_NONE) @@ -420,7 +421,6 @@ END(\label) PTREGSCALL stub_clone, sys_clone, %r8 PTREGSCALL stub_fork, sys_fork, %rdi PTREGSCALL stub_vfork, sys_vfork, %rdi - PTREGSCALL stub_rt_sigsuspend, sys_rt_sigsuspend, %rdx PTREGSCALL stub_sigaltstack, sys_sigaltstack, %rdx PTREGSCALL stub_iopl, sys_iopl, %rsi @@ -559,7 +559,7 @@ retint_with_reschedule: movl $_TIF_WORK_MASK,%edi retint_check: LOCKDEP_SYS_EXIT_IRQ - movl threadinfo_flags(%rcx),%edx + movl TI_flags(%rcx),%edx andl %edi,%edx CFI_REMEMBER_STATE jnz retint_careful @@ -647,17 +647,16 @@ retint_signal: RESTORE_REST DISABLE_INTERRUPTS(CLBR_NONE) TRACE_IRQS_OFF - movl $_TIF_NEED_RESCHED,%edi GET_THREAD_INFO(%rcx) - jmp retint_check + jmp retint_with_reschedule #ifdef CONFIG_PREEMPT /* Returning to kernel space. Check if we need preemption */ /* rcx: threadinfo. interrupts off. */ ENTRY(retint_kernel) - cmpl $0,threadinfo_preempt_count(%rcx) + cmpl $0,TI_preempt_count(%rcx) jnz retint_restore_args - bt $TIF_NEED_RESCHED,threadinfo_flags(%rcx) + bt $TIF_NEED_RESCHED,TI_flags(%rcx) jnc retint_restore_args bt $9,EFLAGS-ARGOFFSET(%rsp) /* interrupts off? */ jnc retint_restore_args @@ -720,6 +719,10 @@ ENTRY(apic_timer_interrupt) apicinterrupt LOCAL_TIMER_VECTOR,smp_apic_timer_interrupt END(apic_timer_interrupt) +ENTRY(uv_bau_message_intr1) + apicinterrupt 220,uv_bau_message_interrupt +END(uv_bau_message_intr1) + ENTRY(error_interrupt) apicinterrupt ERROR_APIC_VECTOR,smp_error_interrupt END(error_interrupt) @@ -733,6 +736,7 @@ END(spurious_interrupt) */ .macro zeroentry sym INTR_FRAME + PARAVIRT_ADJUST_EXCEPTION_FRAME pushq $0 /* push error code/oldrax */ CFI_ADJUST_CFA_OFFSET 8 pushq %rax /* push real oldrax to the rdi slot */ @@ -745,6 +749,7 @@ END(spurious_interrupt) .macro errorentry sym XCPT_FRAME + PARAVIRT_ADJUST_EXCEPTION_FRAME pushq %rax CFI_ADJUST_CFA_OFFSET 8 CFI_REL_OFFSET rax,0 @@ -814,7 +819,7 @@ paranoid_restore\trace: jmp irq_return paranoid_userspace\trace: GET_THREAD_INFO(%rcx) - movl threadinfo_flags(%rcx),%ebx + movl TI_flags(%rcx),%ebx andl $_TIF_WORK_MASK,%ebx jz paranoid_swapgs\trace movq %rsp,%rdi /* &pt_regs */ @@ -912,7 +917,7 @@ error_exit: testl %eax,%eax jne retint_kernel LOCKDEP_SYS_EXIT_IRQ - movl threadinfo_flags(%rcx),%edx + movl TI_flags(%rcx),%edx movl $_TIF_WORK_MASK,%edi andl %edi,%edx jnz retint_careful @@ -926,11 +931,11 @@ error_kernelspace: iret run with kernel gs again, so don't set the user space flag. B stepping K8s sometimes report an truncated RIP for IRET exceptions returning to compat mode. Check for these here too. */ - leaq irq_return(%rip),%rbp - cmpq %rbp,RIP(%rsp) + leaq irq_return(%rip),%rcx + cmpq %rcx,RIP(%rsp) je error_swapgs - movl %ebp,%ebp /* zero extend */ - cmpq %rbp,RIP(%rsp) + movl %ecx,%ecx /* zero extend */ + cmpq %rcx,RIP(%rsp) je error_swapgs cmpq $gs_change,RIP(%rsp) je error_swapgs @@ -939,7 +944,7 @@ KPROBE_END(error_entry) /* Reload gs selector with exception handling */ /* edi: new selector */ -ENTRY(load_gs_index) +ENTRY(native_load_gs_index) CFI_STARTPROC pushf CFI_ADJUST_CFA_OFFSET 8 @@ -953,7 +958,7 @@ gs_change: CFI_ADJUST_CFA_OFFSET -8 ret CFI_ENDPROC -ENDPROC(load_gs_index) +ENDPROC(native_load_gs_index) .section __ex_table,"a" .align 8 @@ -1120,10 +1125,6 @@ ENTRY(coprocessor_segment_overrun) zeroentry do_coprocessor_segment_overrun END(coprocessor_segment_overrun) -ENTRY(reserved) - zeroentry do_reserved -END(reserved) - /* runs on exception stack */ ENTRY(double_fault) XCPT_FRAME diff --git a/arch/x86/kernel/genapic_64.c b/arch/x86/kernel/genapic_64.c index cbaaf69bedb..1fa8be5bd21 100644 --- a/arch/x86/kernel/genapic_64.c +++ b/arch/x86/kernel/genapic_64.c @@ -51,7 +51,7 @@ void __init setup_apic_routing(void) else #endif - if (num_possible_cpus() <= 8) + if (max_physical_apicid < 8) genapic = &apic_flat; else genapic = &apic_physflat; diff --git a/arch/x86/kernel/genx2apic_uv_x.c b/arch/x86/kernel/genx2apic_uv_x.c index ebf13908a74..711f11c30b0 100644 --- a/arch/x86/kernel/genx2apic_uv_x.c +++ b/arch/x86/kernel/genx2apic_uv_x.c @@ -5,9 +5,10 @@ * * SGI UV APIC functions (note: not an Intel compatible APIC) * - * Copyright (C) 2007 Silicon Graphics, Inc. All rights reserved. + * Copyright (C) 2007-2008 Silicon Graphics, Inc. All rights reserved. */ +#include <linux/kernel.h> #include <linux/threads.h> #include <linux/cpumask.h> #include <linux/string.h> @@ -20,6 +21,7 @@ #include <asm/smp.h> #include <asm/ipi.h> #include <asm/genapic.h> +#include <asm/pgtable.h> #include <asm/uv/uv_mmrs.h> #include <asm/uv/uv_hub.h> @@ -55,37 +57,37 @@ static cpumask_t uv_vector_allocation_domain(int cpu) int uv_wakeup_secondary(int phys_apicid, unsigned int start_rip) { unsigned long val; - int nasid; + int pnode; - nasid = uv_apicid_to_nasid(phys_apicid); + pnode = uv_apicid_to_pnode(phys_apicid); val = (1UL << UVH_IPI_INT_SEND_SHFT) | (phys_apicid << UVH_IPI_INT_APIC_ID_SHFT) | (((long)start_rip << UVH_IPI_INT_VECTOR_SHFT) >> 12) | APIC_DM_INIT; - uv_write_global_mmr64(nasid, UVH_IPI_INT, val); + uv_write_global_mmr64(pnode, UVH_IPI_INT, val); mdelay(10); val = (1UL << UVH_IPI_INT_SEND_SHFT) | (phys_apicid << UVH_IPI_INT_APIC_ID_SHFT) | (((long)start_rip << UVH_IPI_INT_VECTOR_SHFT) >> 12) | APIC_DM_STARTUP; - uv_write_global_mmr64(nasid, UVH_IPI_INT, val); + uv_write_global_mmr64(pnode, UVH_IPI_INT, val); return 0; } static void uv_send_IPI_one(int cpu, int vector) { unsigned long val, apicid, lapicid; - int nasid; + int pnode; apicid = per_cpu(x86_cpu_to_apicid, cpu); /* ZZZ - cache node-local ? */ lapicid = apicid & 0x3f; /* ZZZ macro needed */ - nasid = uv_apicid_to_nasid(apicid); + pnode = uv_apicid_to_pnode(apicid); val = (1UL << UVH_IPI_INT_SEND_SHFT) | (lapicid << UVH_IPI_INT_APIC_ID_SHFT) | (vector << UVH_IPI_INT_VECTOR_SHFT); - uv_write_global_mmr64(nasid, UVH_IPI_INT, val); + uv_write_global_mmr64(pnode, UVH_IPI_INT, val); } static void uv_send_IPI_mask(cpumask_t mask, int vector) @@ -159,39 +161,146 @@ struct genapic apic_x2apic_uv_x = { .phys_pkg_id = phys_pkg_id, /* Fixme ZZZ */ }; -static __cpuinit void set_x2apic_extra_bits(int nasid) +static __cpuinit void set_x2apic_extra_bits(int pnode) { - __get_cpu_var(x2apic_extra_bits) = ((nasid >> 1) << 6); + __get_cpu_var(x2apic_extra_bits) = (pnode << 6); } /* * Called on boot cpu. */ +static __init int boot_pnode_to_blade(int pnode) +{ + int blade; + + for (blade = 0; blade < uv_num_possible_blades(); blade++) + if (pnode == uv_blade_info[blade].pnode) + return blade; + BUG(); +} + +struct redir_addr { + unsigned long redirect; + unsigned long alias; +}; + +#define DEST_SHIFT UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_0_MMR_DEST_BASE_SHFT + +static __initdata struct redir_addr redir_addrs[] = { + {UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_0_MMR, UVH_SI_ALIAS0_OVERLAY_CONFIG}, + {UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_1_MMR, UVH_SI_ALIAS1_OVERLAY_CONFIG}, + {UVH_RH_GAM_ALIAS210_REDIRECT_CONFIG_2_MMR, UVH_SI_ALIAS2_OVERLAY_CONFIG}, +}; + +static __init void get_lowmem_redirect(unsigned long *base, unsigned long *size) +{ + union uvh_si_alias0_overlay_config_u alias; + union uvh_rh_gam_alias210_redirect_config_2_mmr_u redirect; + int i; + + for (i = 0; i < ARRAY_SIZE(redir_addrs); i++) { + alias.v = uv_read_local_mmr(redir_addrs[i].alias); + if (alias.s.base == 0) { + *size = (1UL << alias.s.m_alias); + redirect.v = uv_read_local_mmr(redir_addrs[i].redirect); + *base = (unsigned long)redirect.s.dest_base << DEST_SHIFT; + return; + } + } + BUG(); +} + +static __init void map_low_mmrs(void) +{ + init_extra_mapping_uc(UV_GLOBAL_MMR32_BASE, UV_GLOBAL_MMR32_SIZE); + init_extra_mapping_uc(UV_LOCAL_MMR_BASE, UV_LOCAL_MMR_SIZE); +} + +enum map_type {map_wb, map_uc}; + +static void map_high(char *id, unsigned long base, int shift, enum map_type map_type) +{ + unsigned long bytes, paddr; + + paddr = base << shift; + bytes = (1UL << shift); + printk(KERN_INFO "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_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, map_wb); +} + +static __init void map_config_high(int max_pnode) +{ + union uvh_rh_gam_cfg_overlay_config_mmr_u cfg; + int shift = UVH_RH_GAM_CFG_OVERLAY_CONFIG_MMR_BASE_SHFT; + + cfg.v = uv_read_local_mmr(UVH_RH_GAM_CFG_OVERLAY_CONFIG_MMR); + if (cfg.s.enable) + map_high("CONFIG", cfg.s.base, shift, map_uc); +} + +static __init void map_mmr_high(int max_pnode) +{ + union uvh_rh_gam_mmr_overlay_config_mmr_u mmr; + int shift = UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR_BASE_SHFT; + + mmr.v = uv_read_local_mmr(UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR); + if (mmr.s.enable) + map_high("MMR", mmr.s.base, shift, map_uc); +} + +static __init void map_mmioh_high(int max_pnode) +{ + union uvh_rh_gam_mmioh_overlay_config_mmr_u mmioh; + int shift = UVH_RH_GAM_MMIOH_OVERLAY_CONFIG_MMR_BASE_SHFT; + + mmioh.v = uv_read_local_mmr(UVH_RH_GAM_MMIOH_OVERLAY_CONFIG_MMR); + if (mmioh.s.enable) + map_high("MMIOH", mmioh.s.base, shift, map_uc); +} + static __init void uv_system_init(void) { union uvh_si_addr_map_config_u m_n_config; - int bytes, nid, cpu, lcpu, nasid, last_nasid, blade; - unsigned long mmr_base; + 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; + int max_pnode = 0; + unsigned long mmr_base, present; + + map_low_mmrs(); m_n_config.v = uv_read_local_mmr(UVH_SI_ADDR_MAP_CONFIG); + m_val = m_n_config.s.m_skt; + n_val = m_n_config.s.n_skt; mmr_base = uv_read_local_mmr(UVH_RH_GAM_MMR_OVERLAY_CONFIG_MMR) & ~UV_MMR_ENABLE; printk(KERN_DEBUG "UV: global MMR base 0x%lx\n", mmr_base); - last_nasid = -1; - for_each_possible_cpu(cpu) { - nid = cpu_to_node(cpu); - nasid = uv_apicid_to_nasid(per_cpu(x86_cpu_to_apicid, cpu)); - if (nasid != last_nasid) - uv_possible_blades++; - last_nasid = nasid; - } + for(i = 0; i < UVH_NODE_PRESENT_TABLE_DEPTH; i++) + uv_possible_blades += + hweight64(uv_read_local_mmr( UVH_NODE_PRESENT_TABLE + i * 8)); printk(KERN_DEBUG "UV: Found %d blades\n", uv_num_possible_blades()); bytes = sizeof(struct uv_blade_info) * uv_num_possible_blades(); uv_blade_info = alloc_bootmem_pages(bytes); + get_lowmem_redirect(&lowmem_redir_base, &lowmem_redir_size); + bytes = sizeof(uv_node_to_blade[0]) * num_possible_nodes(); uv_node_to_blade = alloc_bootmem_pages(bytes); memset(uv_node_to_blade, 255, bytes); @@ -200,43 +309,62 @@ static __init void uv_system_init(void) uv_cpu_to_blade = alloc_bootmem_pages(bytes); memset(uv_cpu_to_blade, 255, bytes); - last_nasid = -1; - blade = -1; - lcpu = -1; - for_each_possible_cpu(cpu) { - nid = cpu_to_node(cpu); - nasid = uv_apicid_to_nasid(per_cpu(x86_cpu_to_apicid, cpu)); - if (nasid != last_nasid) { - blade++; - lcpu = -1; - uv_blade_info[blade].nr_posible_cpus = 0; + blade = 0; + for (i = 0; i < UVH_NODE_PRESENT_TABLE_DEPTH; i++) { + present = uv_read_local_mmr(UVH_NODE_PRESENT_TABLE + i * 8); + for (j = 0; j < 64; j++) { + if (!test_bit(j, &present)) + continue; + uv_blade_info[blade].pnode = (i * 64 + j); + uv_blade_info[blade].nr_possible_cpus = 0; uv_blade_info[blade].nr_online_cpus = 0; + blade++; } - last_nasid = nasid; - lcpu++; + } - uv_cpu_hub_info(cpu)->m_val = m_n_config.s.m_skt; - uv_cpu_hub_info(cpu)->n_val = m_n_config.s.n_skt; + node_id.v = uv_read_local_mmr(UVH_NODE_ID); + gnode_upper = (((unsigned long)node_id.s.node_id) & + ~((1 << n_val) - 1)) << m_val; + + for_each_present_cpu(cpu) { + nid = cpu_to_node(cpu); + pnode = uv_apicid_to_pnode(per_cpu(x86_cpu_to_apicid, cpu)); + blade = boot_pnode_to_blade(pnode); + lcpu = uv_blade_info[blade].nr_possible_cpus; + uv_blade_info[blade].nr_possible_cpus++; + + uv_cpu_hub_info(cpu)->lowmem_remap_base = lowmem_redir_base; + uv_cpu_hub_info(cpu)->lowmem_remap_top = + lowmem_redir_base + lowmem_redir_size; + uv_cpu_hub_info(cpu)->m_val = m_val; + uv_cpu_hub_info(cpu)->n_val = m_val; uv_cpu_hub_info(cpu)->numa_blade_id = blade; uv_cpu_hub_info(cpu)->blade_processor_id = lcpu; - uv_cpu_hub_info(cpu)->local_nasid = nasid; - uv_cpu_hub_info(cpu)->gnode_upper = - nasid & ~((1 << uv_hub_info->n_val) - 1); + uv_cpu_hub_info(cpu)->pnode = pnode; + uv_cpu_hub_info(cpu)->pnode_mask = (1 << n_val) - 1; + uv_cpu_hub_info(cpu)->gpa_mask = (1 << (m_val + n_val)) - 1; + uv_cpu_hub_info(cpu)->gnode_upper = gnode_upper; uv_cpu_hub_info(cpu)->global_mmr_base = mmr_base; uv_cpu_hub_info(cpu)->coherency_domain_number = 0;/* ZZZ */ - uv_blade_info[blade].nasid = nasid; - uv_blade_info[blade].nr_posible_cpus++; uv_node_to_blade[nid] = blade; uv_cpu_to_blade[cpu] = blade; + max_pnode = max(pnode, max_pnode); - printk(KERN_DEBUG "UV cpu %d, apicid 0x%x, nasid %d, nid %d\n", - cpu, per_cpu(x86_cpu_to_apicid, cpu), nasid, nid); - printk(KERN_DEBUG "UV lcpu %d, blade %d\n", lcpu, blade); + printk(KERN_DEBUG "UV: cpu %d, apicid 0x%x, pnode %d, nid %d, " + "lcpu %d, blade %d\n", + cpu, per_cpu(x86_cpu_to_apicid, cpu), pnode, nid, + lcpu, blade); } + + map_gru_high(max_pnode); + map_mmr_high(max_pnode); + map_config_high(max_pnode); + map_mmioh_high(max_pnode); } /* * Called on each cpu to initialize the per_cpu UV data area. + * ZZZ hotplug not supported yet */ void __cpuinit uv_cpu_init(void) { @@ -246,5 +374,5 @@ void __cpuinit uv_cpu_init(void) uv_blade_info[uv_numa_blade_id()].nr_online_cpus++; if (get_uv_system_type() == UV_NON_UNIQUE_APIC) - set_x2apic_extra_bits(uv_hub_info->local_nasid); + set_x2apic_extra_bits(uv_hub_info->pnode); } diff --git a/arch/x86/kernel/head.c b/arch/x86/kernel/head.c new file mode 100644 index 00000000000..3e66bd364a9 --- /dev/null +++ b/arch/x86/kernel/head.c @@ -0,0 +1,55 @@ +#include <linux/kernel.h> +#include <linux/init.h> + +#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 + * conventional memory (int 0x12) too. This also contains a + * workaround for Dell systems that neglect to reserve EBDA. + * The same workaround also avoids a problem with the AMD768MPX + * 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. + */ +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. */ + if (paravirt_enabled()) + return; + + /* end of low (conventional) memory */ + lowmem = *(unsigned short *)__va(BIOS_LOWMEM_KILOBYTES); + lowmem <<= 10; + + /* 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; + + /* 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; + + /* Paranoia: should never happen, but... */ + if ((lowmem == 0) || (lowmem >= 0x100000)) + lowmem = 0x9f000; + + /* reserve all memory between lowmem and the 1MB mark */ + reserve_early_overlap_ok(lowmem, 0x100000, "BIOS reserved"); +} diff --git a/arch/x86/kernel/head32.c b/arch/x86/kernel/head32.c index 3db05905892..fa1d25dd83e 100644 --- a/arch/x86/kernel/head32.c +++ b/arch/x86/kernel/head32.c @@ -8,7 +8,34 @@ #include <linux/init.h> #include <linux/start_kernel.h> +#include <asm/setup.h> +#include <asm/sections.h> +#include <asm/e820.h> +#include <asm/bios_ebda.h> + void __init i386_start_kernel(void) { + reserve_early(__pa_symbol(&_text), __pa_symbol(&_end), "TEXT DATA BSS"); + +#ifdef CONFIG_BLK_DEV_INITRD + /* Reserve INITRD */ + if (boot_params.hdr.type_of_loader && boot_params.hdr.ramdisk_image) { + u64 ramdisk_image = boot_params.hdr.ramdisk_image; + u64 ramdisk_size = boot_params.hdr.ramdisk_size; + u64 ramdisk_end = ramdisk_image + ramdisk_size; + reserve_early(ramdisk_image, ramdisk_end, "RAMDISK"); + } +#endif + reserve_early(init_pg_tables_start, init_pg_tables_end, + "INIT_PG_TABLE"); + + 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/head64.c b/arch/x86/kernel/head64.c index e25c57b8aa8..c9781982914 100644 --- a/arch/x86/kernel/head64.c +++ b/arch/x86/kernel/head64.c @@ -25,6 +25,20 @@ #include <asm/e820.h> #include <asm/bios_ebda.h> +/* boot cpu pda */ +static struct x8664_pda _boot_cpu_pda __read_mostly; + +#ifdef CONFIG_SMP +/* + * We install an empty cpu_pda pointer table to indicate to early users + * (numa_set_node) that the cpu_pda pointer table for cpus other than + * the boot cpu is not yet setup. + */ +static struct x8664_pda *__cpu_pda[NR_CPUS] __initdata; +#else +static struct x8664_pda *__cpu_pda[NR_CPUS] __read_mostly; +#endif + static void __init zap_identity_mappings(void) { pgd_t *pgd = pgd_offset_k(0UL); @@ -51,74 +65,6 @@ static void __init copy_bootdata(char *real_mode_data) } } -#define BIOS_LOWMEM_KILOBYTES 0x413 - -/* - * The BIOS places the EBDA/XBDA at the top of conventional - * memory, and usually decreases the reported amount of - * conventional memory (int 0x12) too. This also contains a - * workaround for Dell systems that neglect to reserve EBDA. - * The same workaround also avoids a problem with the AMD768MPX - * 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. - */ -static 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. */ - if (paravirt_enabled()) - return; - - /* end of low (conventional) memory */ - lowmem = *(unsigned short *)__va(BIOS_LOWMEM_KILOBYTES); - lowmem <<= 10; - - /* 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; - - /* 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; - - /* Paranoia: should never happen, but... */ - if ((lowmem == 0) || (lowmem >= 0x100000)) - lowmem = 0x9f000; - - /* reserve all memory between lowmem and the 1MB mark */ - reserve_early(lowmem, 0x100000, "BIOS reserved"); -} - -static void __init reserve_setup_data(void) -{ - struct setup_data *data; - unsigned long pa_data; - char buf[32]; - - if (boot_params.hdr.version < 0x0209) - return; - pa_data = boot_params.hdr.setup_data; - while (pa_data) { - data = early_ioremap(pa_data, sizeof(*data)); - sprintf(buf, "setup data %x", data->type); - reserve_early(pa_data, pa_data+sizeof(*data)+data->len, buf); - pa_data = data->next; - early_iounmap(data, sizeof(*data)); - } -} - void __init x86_64_start_kernel(char * real_mode_data) { int i; @@ -156,10 +102,17 @@ void __init x86_64_start_kernel(char * real_mode_data) early_printk("Kernel alive\n"); - for (i = 0; i < NR_CPUS; i++) - cpu_pda(i) = &boot_cpu_pda[i]; - + _cpu_pda = __cpu_pda; + cpu_pda(0) = &_boot_cpu_pda; pda_init(0); + + early_printk("Kernel really alive\n"); + + x86_64_start_reservations(real_mode_data); +} + +void __init x86_64_start_reservations(char *real_mode_data) +{ copy_bootdata(__va(real_mode_data)); reserve_early(__pa_symbol(&_text), __pa_symbol(&_end), "TEXT DATA BSS"); @@ -175,7 +128,6 @@ void __init x86_64_start_kernel(char * real_mode_data) #endif reserve_ebda_region(); - reserve_setup_data(); /* * At this point everything still needed from the boot loader diff --git a/arch/x86/kernel/head_32.S b/arch/x86/kernel/head_32.S index f7357cc0162..f67e93441ca 100644 --- a/arch/x86/kernel/head_32.S +++ b/arch/x86/kernel/head_32.S @@ -194,6 +194,7 @@ default_entry: xorl %ebx,%ebx /* %ebx is kept at zero */ movl $pa(pg0), %edi + movl %edi, pa(init_pg_tables_start) movl $pa(swapper_pg_pmd), %edx movl $PTE_ATTR, %eax 10: @@ -219,6 +220,8 @@ default_entry: jb 10b 1: movl %edi,pa(init_pg_tables_end) + shrl $12, %eax + movl %eax, pa(max_pfn_mapped) /* Do early initialization of the fixmap area */ movl $pa(swapper_pg_fixmap)+PDE_ATTR,%eax @@ -228,6 +231,7 @@ default_entry: page_pde_offset = (__PAGE_OFFSET >> 20); movl $pa(pg0), %edi + movl %edi, pa(init_pg_tables_start) movl $pa(swapper_pg_dir), %edx movl $PTE_ATTR, %eax 10: @@ -249,6 +253,8 @@ page_pde_offset = (__PAGE_OFFSET >> 20); cmpl %ebp,%eax jb 10b movl %edi,pa(init_pg_tables_end) + shrl $12, %eax + movl %eax, pa(max_pfn_mapped) /* Do early initialization of the fixmap area */ movl $pa(swapper_pg_fixmap)+PDE_ATTR,%eax @@ -446,10 +452,13 @@ is386: movl $2,%ecx # set MP je 1f movl $(__KERNEL_PERCPU), %eax movl %eax,%fs # set this cpu's percpu - jmp initialize_secondary # all other CPUs call initialize_secondary + movl (stack_start), %esp 1: #endif /* CONFIG_SMP */ - jmp i386_start_kernel + jmp *(initial_code) +.align 4 +ENTRY(initial_code) + .long i386_start_kernel /* * We depend on ET to be correct. This checks for 287/387. diff --git a/arch/x86/kernel/head_64.S b/arch/x86/kernel/head_64.S index b817974ef94..b07ac7b217c 100644 --- a/arch/x86/kernel/head_64.S +++ b/arch/x86/kernel/head_64.S @@ -18,6 +18,7 @@ #include <asm/page.h> #include <asm/msr.h> #include <asm/cache.h> +#include <asm/processor-flags.h> #ifdef CONFIG_PARAVIRT #include <asm/asm-offsets.h> @@ -31,6 +32,13 @@ * */ +#define pud_index(x) (((x) >> PUD_SHIFT) & (PTRS_PER_PUD-1)) + +L4_PAGE_OFFSET = pgd_index(__PAGE_OFFSET) +L3_PAGE_OFFSET = pud_index(__PAGE_OFFSET) +L4_START_KERNEL = pgd_index(__START_KERNEL_map) +L3_START_KERNEL = pud_index(__START_KERNEL_map) + .text .section .text.head .code64 @@ -76,8 +84,8 @@ startup_64: /* Fixup the physical addresses in the page table */ addq %rbp, init_level4_pgt + 0(%rip) - addq %rbp, init_level4_pgt + (258*8)(%rip) - addq %rbp, init_level4_pgt + (511*8)(%rip) + addq %rbp, init_level4_pgt + (L4_PAGE_OFFSET*8)(%rip) + addq %rbp, init_level4_pgt + (L4_START_KERNEL*8)(%rip) addq %rbp, level3_ident_pgt + 0(%rip) @@ -154,9 +162,7 @@ ENTRY(secondary_startup_64) */ /* Enable PAE mode and PGE */ - xorq %rax, %rax - btsq $5, %rax - btsq $7, %rax + movl $(X86_CR4_PAE | X86_CR4_PGE), %eax movq %rax, %cr4 /* Setup early boot stage 4 level pagetables. */ @@ -184,19 +190,15 @@ ENTRY(secondary_startup_64) 1: wrmsr /* Make changes effective */ /* Setup cr0 */ -#define CR0_PM 1 /* protected mode */ -#define CR0_MP (1<<1) -#define CR0_ET (1<<4) -#define CR0_NE (1<<5) -#define CR0_WP (1<<16) -#define CR0_AM (1<<18) -#define CR0_PAGING (1<<31) - movl $CR0_PM|CR0_MP|CR0_ET|CR0_NE|CR0_WP|CR0_AM|CR0_PAGING,%eax +#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, %eax /* Make changes effective */ movq %rax, %cr0 /* Setup a boot time stack */ - movq init_rsp(%rip),%rsp + movq stack_start(%rip),%rsp /* zero EFLAGS after setting rsp */ pushq $0 @@ -208,7 +210,7 @@ ENTRY(secondary_startup_64) * addresses where we're currently running on. We have to do that here * because in 32bit we couldn't load a 64bit linear address. */ - lgdt cpu_gdt_descr(%rip) + lgdt early_gdt_descr(%rip) /* set up data segments. actually 0 would do too */ movl $__KERNEL_DS,%eax @@ -257,8 +259,9 @@ ENTRY(secondary_startup_64) .quad x86_64_start_kernel __FINITDATA - ENTRY(init_rsp) + ENTRY(stack_start) .quad init_thread_union+THREAD_SIZE-8 + .word 0 bad_address: jmp bad_address @@ -327,11 +330,11 @@ early_idt_ripmsg: ENTRY(name) /* Automate the creation of 1 to 1 mapping pmd entries */ -#define PMDS(START, PERM, COUNT) \ - i = 0 ; \ - .rept (COUNT) ; \ - .quad (START) + (i << 21) + (PERM) ; \ - i = i + 1 ; \ +#define PMDS(START, PERM, COUNT) \ + i = 0 ; \ + .rept (COUNT) ; \ + .quad (START) + (i << PMD_SHIFT) + (PERM) ; \ + i = i + 1 ; \ .endr /* @@ -342,9 +345,9 @@ ENTRY(name) */ NEXT_PAGE(init_level4_pgt) .quad level3_ident_pgt - __START_KERNEL_map + _KERNPG_TABLE - .fill 257,8,0 + .org init_level4_pgt + L4_PAGE_OFFSET*8, 0 .quad level3_ident_pgt - __START_KERNEL_map + _KERNPG_TABLE - .fill 252,8,0 + .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 @@ -353,7 +356,7 @@ NEXT_PAGE(level3_ident_pgt) .fill 511,8,0 NEXT_PAGE(level3_kernel_pgt) - .fill 510,8,0 + .fill L3_START_KERNEL,8,0 /* (2^48-(2*1024*1024*1024)-((2^39)*511))/(2^30) = 510 */ .quad level2_kernel_pgt - __START_KERNEL_map + _KERNPG_TABLE .quad level2_fixmap_pgt - __START_KERNEL_map + _PAGE_TABLE @@ -384,7 +387,7 @@ NEXT_PAGE(level2_kernel_pgt) * If you want to increase this then increase MODULES_VADDR * too.) */ - PMDS(0, __PAGE_KERNEL_LARGE_EXEC|_PAGE_GLOBAL, + PMDS(0, __PAGE_KERNEL_LARGE_EXEC, KERNEL_IMAGE_SIZE/PMD_SIZE) NEXT_PAGE(level2_spare_pgt) @@ -395,54 +398,16 @@ NEXT_PAGE(level2_spare_pgt) .data .align 16 - .globl cpu_gdt_descr -cpu_gdt_descr: - .word gdt_end-cpu_gdt_table-1 -gdt: - .quad cpu_gdt_table -#ifdef CONFIG_SMP - .rept NR_CPUS-1 - .word 0 - .quad 0 - .endr -#endif + .globl early_gdt_descr +early_gdt_descr: + .word GDT_ENTRIES*8-1 + .quad per_cpu__gdt_page ENTRY(phys_base) /* This must match the first entry in level2_kernel_pgt */ .quad 0x0000000000000000 -/* We need valid kernel segments for data and code in long mode too - * IRET will check the segment types kkeil 2000/10/28 - * Also sysret mandates a special GDT layout - */ - - .section .data.page_aligned, "aw" - .align PAGE_SIZE - -/* The TLS descriptors are currently at a different place compared to i386. - Hopefully nobody expects them at a fixed place (Wine?) */ -ENTRY(cpu_gdt_table) - .quad 0x0000000000000000 /* NULL descriptor */ - .quad 0x00cf9b000000ffff /* __KERNEL32_CS */ - .quad 0x00af9b000000ffff /* __KERNEL_CS */ - .quad 0x00cf93000000ffff /* __KERNEL_DS */ - .quad 0x00cffb000000ffff /* __USER32_CS */ - .quad 0x00cff3000000ffff /* __USER_DS, __USER32_DS */ - .quad 0x00affb000000ffff /* __USER_CS */ - .quad 0x0 /* unused */ - .quad 0,0 /* TSS */ - .quad 0,0 /* LDT */ - .quad 0,0,0 /* three TLS descriptors */ - .quad 0x0000f40000000000 /* node/CPU stored in limit */ -gdt_end: - /* asm/segment.h:GDT_ENTRIES must match this */ - /* This should be a multiple of the cache line size */ - /* GDTs of other CPUs are now dynamically allocated */ - - /* zero the remaining page */ - .fill PAGE_SIZE / 8 - GDT_ENTRIES,8,0 - .section .bss, "aw", @nobits .align L1_CACHE_BYTES ENTRY(idt_table) diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c index 9b5cfcdfc42..0ea6a19bfdf 100644 --- a/arch/x86/kernel/hpet.c +++ b/arch/x86/kernel/hpet.c @@ -17,7 +17,7 @@ /* FSEC = 10^-15 NSEC = 10^-9 */ -#define FSEC_PER_NSEC 1000000 +#define FSEC_PER_NSEC 1000000L /* * HPET address is set in acpi/boot.c, when an ACPI entry exists @@ -36,26 +36,15 @@ static inline void hpet_writel(unsigned long d, unsigned long a) } #ifdef CONFIG_X86_64 - #include <asm/pgtable.h> - -static inline void hpet_set_mapping(void) -{ - set_fixmap_nocache(FIX_HPET_BASE, hpet_address); - __set_fixmap(VSYSCALL_HPET, hpet_address, PAGE_KERNEL_VSYSCALL_NOCACHE); - hpet_virt_address = (void __iomem *)fix_to_virt(FIX_HPET_BASE); -} - -static inline void hpet_clear_mapping(void) -{ - hpet_virt_address = NULL; -} - -#else +#endif 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_VSYSCALL_NOCACHE); +#endif } static inline void hpet_clear_mapping(void) @@ -63,7 +52,6 @@ static inline void hpet_clear_mapping(void) iounmap(hpet_virt_address); hpet_virt_address = NULL; } -#endif /* * HPET command line enable / disable @@ -206,20 +194,19 @@ static void hpet_enable_legacy_int(void) static void hpet_legacy_clockevent_register(void) { - uint64_t hpet_freq; - /* Start HPET legacy interrupts */ hpet_enable_legacy_int(); /* - * The period is a femto seconds value. We need to calculate the - * scaled math multiplication factor for nanosecond to hpet tick - * conversion. + * The mult factor is defined as (include/linux/clockchips.h) + * mult/2^shift = cyc/ns (in contrast to ns/cyc in clocksource.h) + * hpet_period is in units of femtoseconds (per cycle), so + * mult/2^shift = cyc/ns = 10^6/hpet_period + * mult = (10^6 * 2^shift)/hpet_period + * mult = (FSEC_PER_NSEC << hpet_clockevent.shift)/hpet_period */ - hpet_freq = 1000000000000000ULL; - do_div(hpet_freq, hpet_period); - hpet_clockevent.mult = div_sc((unsigned long) hpet_freq, - NSEC_PER_SEC, hpet_clockevent.shift); + hpet_clockevent.mult = div_sc((unsigned long) FSEC_PER_NSEC, + hpet_period, hpet_clockevent.shift); /* Calculate the min / max delta */ hpet_clockevent.max_delta_ns = clockevent_delta2ns(0x7FFFFFFF, &hpet_clockevent); @@ -324,7 +311,7 @@ static struct clocksource clocksource_hpet = { static int hpet_clocksource_register(void) { - u64 tmp, start, now; + u64 start, now; cycle_t t1; /* Start the counter */ @@ -351,21 +338,15 @@ static int hpet_clocksource_register(void) return -ENODEV; } - /* Initialize and register HPET clocksource - * - * hpet period is in femto seconds per cycle - * so we need to convert this to ns/cyc units - * approximated by mult/2^shift - * - * fsec/cyc * 1nsec/1000000fsec = nsec/cyc = mult/2^shift - * fsec/cyc * 1ns/1000000fsec * 2^shift = mult - * fsec/cyc * 2^shift * 1nsec/1000000fsec = mult - * (fsec/cyc << shift)/1000000 = mult - * (hpet_period << shift)/FSEC_PER_NSEC = mult + /* + * The definition of mult is (include/linux/clocksource.h) + * mult/2^shift = ns/cyc and hpet_period is in units of fsec/cyc + * so we first need to convert hpet_period to ns/cyc units: + * mult/2^shift = ns/cyc = hpet_period/10^6 + * mult = (hpet_period * 2^shift)/10^6 + * mult = (hpet_period << shift)/FSEC_PER_NSEC */ - tmp = (u64)hpet_period << HPET_SHIFT; - do_div(tmp, FSEC_PER_NSEC); - clocksource_hpet.mult = (u32)tmp; + clocksource_hpet.mult = div_sc(hpet_period, FSEC_PER_NSEC, HPET_SHIFT); clocksource_register(&clocksource_hpet); diff --git a/arch/x86/kernel/i387.c b/arch/x86/kernel/i387.c index 95e80e5033c..eb9ddd8efb8 100644 --- a/arch/x86/kernel/i387.c +++ b/arch/x86/kernel/i387.c @@ -162,7 +162,7 @@ int xfpregs_get(struct task_struct *target, const struct user_regset *regset, int ret; if (!cpu_has_fxsr) - return -EIO; + return -ENODEV; ret = init_fpu(target); if (ret) @@ -179,7 +179,7 @@ int xfpregs_set(struct task_struct *target, const struct user_regset *regset, int ret; if (!cpu_has_fxsr) - return -EIO; + return -ENODEV; ret = init_fpu(target); if (ret) diff --git a/arch/x86/kernel/i8259_32.c b/arch/x86/kernel/i8259.c index fe631967d62..dc92b49d920 100644 --- a/arch/x86/kernel/i8259_32.c +++ b/arch/x86/kernel/i8259.c @@ -1,8 +1,10 @@ +#include <linux/linkage.h> #include <linux/errno.h> #include <linux/signal.h> #include <linux/sched.h> #include <linux/ioport.h> #include <linux/interrupt.h> +#include <linux/timex.h> #include <linux/slab.h> #include <linux/random.h> #include <linux/init.h> @@ -10,10 +12,12 @@ #include <linux/sysdev.h> #include <linux/bitops.h> +#include <asm/acpi.h> #include <asm/atomic.h> #include <asm/system.h> #include <asm/io.h> #include <asm/timer.h> +#include <asm/hw_irq.h> #include <asm/pgtable.h> #include <asm/delay.h> #include <asm/desc.h> @@ -32,7 +36,7 @@ static int i8259A_auto_eoi; DEFINE_SPINLOCK(i8259A_lock); static void mask_and_ack_8259A(unsigned int); -static struct irq_chip i8259A_chip = { +struct irq_chip i8259A_chip = { .name = "XT-PIC", .mask = disable_8259A_irq, .disable = disable_8259A_irq, @@ -125,14 +129,14 @@ static inline int i8259A_irq_real(unsigned int irq) int irqmask = 1<<irq; if (irq < 8) { - outb(0x0B,PIC_MASTER_CMD); /* ISR register */ + outb(0x0B, PIC_MASTER_CMD); /* ISR register */ value = inb(PIC_MASTER_CMD) & irqmask; - outb(0x0A,PIC_MASTER_CMD); /* back to the IRR register */ + outb(0x0A, PIC_MASTER_CMD); /* back to the IRR register */ return value; } - outb(0x0B,PIC_SLAVE_CMD); /* ISR register */ + outb(0x0B, PIC_SLAVE_CMD); /* ISR register */ value = inb(PIC_SLAVE_CMD) & (irqmask >> 8); - outb(0x0A,PIC_SLAVE_CMD); /* back to the IRR register */ + outb(0x0A, PIC_SLAVE_CMD); /* back to the IRR register */ return value; } @@ -171,12 +175,14 @@ handle_real_irq: if (irq & 8) { inb(PIC_SLAVE_IMR); /* DUMMY - (do we need this?) */ outb(cached_slave_mask, PIC_SLAVE_IMR); - outb(0x60+(irq&7),PIC_SLAVE_CMD);/* 'Specific EOI' to slave */ - outb(0x60+PIC_CASCADE_IR,PIC_MASTER_CMD); /* 'Specific EOI' to master-IRQ2 */ + /* 'Specific EOI' to slave */ + outb(0x60+(irq&7), PIC_SLAVE_CMD); + /* 'Specific EOI' to master-IRQ2 */ + outb(0x60+PIC_CASCADE_IR, PIC_MASTER_CMD); } else { inb(PIC_MASTER_IMR); /* DUMMY - (do we need this?) */ outb(cached_master_mask, PIC_MASTER_IMR); - outb(0x60+irq,PIC_MASTER_CMD); /* 'Specific EOI to master */ + outb(0x60+irq, PIC_MASTER_CMD); /* 'Specific EOI to master */ } spin_unlock_irqrestore(&i8259A_lock, flags); return; @@ -199,7 +205,8 @@ spurious_8259A_irq: * lets ACK and report it. [once per IRQ] */ if (!(spurious_irq_mask & irqmask)) { - printk(KERN_DEBUG "spurious 8259A interrupt: IRQ%d.\n", irq); + printk(KERN_DEBUG + "spurious 8259A interrupt: IRQ%d.\n", irq); spurious_irq_mask |= irqmask; } atomic_inc(&irq_err_count); @@ -290,17 +297,28 @@ void init_8259A(int auto_eoi) * outb_pic - this has to work on a wide range of PC hardware. */ outb_pic(0x11, PIC_MASTER_CMD); /* ICW1: select 8259A-1 init */ - outb_pic(0x20 + 0, PIC_MASTER_IMR); /* ICW2: 8259A-1 IR0-7 mapped to 0x20-0x27 */ - outb_pic(1U << PIC_CASCADE_IR, PIC_MASTER_IMR); /* 8259A-1 (the master) has a slave on IR2 */ + + /* ICW2: 8259A-1 IR0-7 mapped to 0x30-0x37 on x86-64, + to 0x20-0x27 on i386 */ + outb_pic(IRQ0_VECTOR, PIC_MASTER_IMR); + + /* 8259A-1 (the master) has a slave on IR2 */ + outb_pic(1U << PIC_CASCADE_IR, PIC_MASTER_IMR); + if (auto_eoi) /* master does Auto EOI */ outb_pic(MASTER_ICW4_DEFAULT | PIC_ICW4_AEOI, PIC_MASTER_IMR); else /* master expects normal EOI */ outb_pic(MASTER_ICW4_DEFAULT, PIC_MASTER_IMR); outb_pic(0x11, PIC_SLAVE_CMD); /* ICW1: select 8259A-2 init */ - outb_pic(0x20 + 8, PIC_SLAVE_IMR); /* ICW2: 8259A-2 IR0-7 mapped to 0x28-0x2f */ - outb_pic(PIC_CASCADE_IR, PIC_SLAVE_IMR); /* 8259A-2 is a slave on master's IR2 */ - outb_pic(SLAVE_ICW4_DEFAULT, PIC_SLAVE_IMR); /* (slave's support for AEOI in flat mode is to be investigated) */ + + /* ICW2: 8259A-2 IR0-7 mapped to IRQ8_VECTOR */ + outb_pic(IRQ8_VECTOR, PIC_SLAVE_IMR); + /* 8259A-2 is a slave on master's IR2 */ + outb_pic(PIC_CASCADE_IR, PIC_SLAVE_IMR); + /* (slave's support for AEOI in flat mode is to be investigated) */ + outb_pic(SLAVE_ICW4_DEFAULT, PIC_SLAVE_IMR); + if (auto_eoi) /* * In AEOI mode we just have to mask the interrupt @@ -317,93 +335,3 @@ void init_8259A(int auto_eoi) spin_unlock_irqrestore(&i8259A_lock, flags); } - -/* - * 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) -{ - extern void math_error(void __user *); - outb(0,0xF0); - if (ignore_fpu_irq || !boot_cpu_data.hard_math) - return IRQ_NONE; - math_error((void __user *)get_irq_regs()->ip); - 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, - .mask = CPU_MASK_NONE, - .name = "fpu", -}; - -void __init init_ISA_irqs (void) -{ - int i; - -#ifdef CONFIG_X86_LOCAL_APIC - init_bsp_APIC(); -#endif - init_8259A(0); - - /* - * 16 old-style INTA-cycle interrupts: - */ - for (i = 0; i < 16; i++) { - set_irq_chip_and_handler_name(i, &i8259A_chip, - handle_level_irq, "XT"); - } -} - -/* Overridden in paravirt.c */ -void init_IRQ(void) __attribute__((weak, alias("native_init_IRQ"))); - -void __init native_init_IRQ(void) -{ - int i; - - /* all the set up before the call gates are initialised */ - pre_intr_init_hook(); - - /* - * Cover the whole vector space, no vector can escape - * us. (some of these will be overridden and become - * 'special' SMP interrupts) - */ - for (i = 0; i < (NR_VECTORS - FIRST_EXTERNAL_VECTOR); i++) { - int vector = FIRST_EXTERNAL_VECTOR + i; - if (i >= NR_IRQS) - break; - /* SYSCALL_VECTOR was reserved in trap_init. */ - if (!test_bit(vector, used_vectors)) - set_intr_gate(vector, interrupt[i]); - } - - /* setup after call gates are initialised (usually add in - * the architecture specific gates) - */ - intr_init_hook(); - - /* - * 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()); -} diff --git a/arch/x86/kernel/i8259_64.c b/arch/x86/kernel/i8259_64.c deleted file mode 100644 index fa57a156850..00000000000 --- a/arch/x86/kernel/i8259_64.c +++ /dev/null @@ -1,512 +0,0 @@ -#include <linux/linkage.h> -#include <linux/errno.h> -#include <linux/signal.h> -#include <linux/sched.h> -#include <linux/ioport.h> -#include <linux/interrupt.h> -#include <linux/timex.h> -#include <linux/slab.h> -#include <linux/random.h> -#include <linux/init.h> -#include <linux/kernel_stat.h> -#include <linux/sysdev.h> -#include <linux/bitops.h> - -#include <asm/acpi.h> -#include <asm/atomic.h> -#include <asm/system.h> -#include <asm/io.h> -#include <asm/hw_irq.h> -#include <asm/pgtable.h> -#include <asm/delay.h> -#include <asm/desc.h> -#include <asm/apic.h> -#include <asm/i8259.h> - -/* - * Common place to define all x86 IRQ vectors - * - * This builds up the IRQ handler stubs using some ugly macros in irq.h - * - * These macros create the low-level assembly IRQ routines that save - * register context and call do_IRQ(). do_IRQ() then does all the - * operations that are needed to keep the AT (or SMP IOAPIC) - * interrupt-controller happy. - */ - -#define BI(x,y) \ - BUILD_IRQ(x##y) - -#define BUILD_16_IRQS(x) \ - BI(x,0) BI(x,1) BI(x,2) BI(x,3) \ - BI(x,4) BI(x,5) BI(x,6) BI(x,7) \ - BI(x,8) BI(x,9) BI(x,a) BI(x,b) \ - BI(x,c) BI(x,d) BI(x,e) BI(x,f) - -/* - * ISA PIC or low IO-APIC triggered (INTA-cycle or APIC) interrupts: - * (these are usually mapped to vectors 0x30-0x3f) - */ - -/* - * The IO-APIC gives us many more interrupt sources. Most of these - * are unused but an SMP system is supposed to have enough memory ... - * sometimes (mostly wrt. hw bugs) we get corrupted vectors all - * across the spectrum, so we really want to be prepared to get all - * of these. Plus, more powerful systems might have more than 64 - * IO-APIC registers. - * - * (these are usually mapped into the 0x30-0xff vector range) - */ - BUILD_16_IRQS(0x2) BUILD_16_IRQS(0x3) -BUILD_16_IRQS(0x4) BUILD_16_IRQS(0x5) BUILD_16_IRQS(0x6) BUILD_16_IRQS(0x7) -BUILD_16_IRQS(0x8) BUILD_16_IRQS(0x9) BUILD_16_IRQS(0xa) BUILD_16_IRQS(0xb) -BUILD_16_IRQS(0xc) BUILD_16_IRQS(0xd) BUILD_16_IRQS(0xe) BUILD_16_IRQS(0xf) - -#undef BUILD_16_IRQS -#undef BI - - -#define IRQ(x,y) \ - IRQ##x##y##_interrupt - -#define IRQLIST_16(x) \ - IRQ(x,0), IRQ(x,1), IRQ(x,2), IRQ(x,3), \ - IRQ(x,4), IRQ(x,5), IRQ(x,6), IRQ(x,7), \ - IRQ(x,8), IRQ(x,9), IRQ(x,a), IRQ(x,b), \ - IRQ(x,c), IRQ(x,d), IRQ(x,e), IRQ(x,f) - -/* for the irq vectors */ -static void (*__initdata interrupt[NR_VECTORS - FIRST_EXTERNAL_VECTOR])(void) = { - IRQLIST_16(0x2), IRQLIST_16(0x3), - IRQLIST_16(0x4), IRQLIST_16(0x5), IRQLIST_16(0x6), IRQLIST_16(0x7), - IRQLIST_16(0x8), IRQLIST_16(0x9), IRQLIST_16(0xa), IRQLIST_16(0xb), - IRQLIST_16(0xc), IRQLIST_16(0xd), IRQLIST_16(0xe), IRQLIST_16(0xf) -}; - -#undef IRQ -#undef IRQLIST_16 - -/* - * This is the 'legacy' 8259A Programmable Interrupt Controller, - * present in the majority of PC/AT boxes. - * plus some generic x86 specific things if generic specifics makes - * any sense at all. - * this file should become arch/i386/kernel/irq.c when the old irq.c - * moves to arch independent land - */ - -static int i8259A_auto_eoi; -DEFINE_SPINLOCK(i8259A_lock); -static void mask_and_ack_8259A(unsigned int); - -static struct irq_chip i8259A_chip = { - .name = "XT-PIC", - .mask = disable_8259A_irq, - .disable = disable_8259A_irq, - .unmask = enable_8259A_irq, - .mask_ack = mask_and_ack_8259A, -}; - -/* - * 8259A PIC functions to handle ISA devices: - */ - -/* - * This contains the irq mask for both 8259A irq controllers, - */ -unsigned int cached_irq_mask = 0xffff; - -/* - * Not all IRQs can be routed through the IO-APIC, eg. on certain (older) - * boards the timer interrupt is not really connected to any IO-APIC pin, - * it's fed to the master 8259A's IR0 line only. - * - * Any '1' bit in this mask means the IRQ is routed through the IO-APIC. - * this 'mixed mode' IRQ handling costs nothing because it's only used - * at IRQ setup time. - */ -unsigned long io_apic_irqs; - -void disable_8259A_irq(unsigned int irq) -{ - unsigned int mask = 1 << irq; - unsigned long flags; - - spin_lock_irqsave(&i8259A_lock, flags); - cached_irq_mask |= mask; - if (irq & 8) - outb(cached_slave_mask, PIC_SLAVE_IMR); - else - outb(cached_master_mask, PIC_MASTER_IMR); - spin_unlock_irqrestore(&i8259A_lock, flags); -} - -void enable_8259A_irq(unsigned int irq) -{ - unsigned int mask = ~(1 << irq); - unsigned long flags; - - spin_lock_irqsave(&i8259A_lock, flags); - cached_irq_mask &= mask; - if (irq & 8) - outb(cached_slave_mask, PIC_SLAVE_IMR); - else - outb(cached_master_mask, PIC_MASTER_IMR); - spin_unlock_irqrestore(&i8259A_lock, flags); -} - -int i8259A_irq_pending(unsigned int irq) -{ - unsigned int mask = 1<<irq; - unsigned long flags; - int ret; - - spin_lock_irqsave(&i8259A_lock, flags); - if (irq < 8) - ret = inb(PIC_MASTER_CMD) & mask; - else - ret = inb(PIC_SLAVE_CMD) & (mask >> 8); - spin_unlock_irqrestore(&i8259A_lock, flags); - - return ret; -} - -void make_8259A_irq(unsigned int irq) -{ - disable_irq_nosync(irq); - io_apic_irqs &= ~(1<<irq); - set_irq_chip_and_handler_name(irq, &i8259A_chip, handle_level_irq, - "XT"); - enable_irq(irq); -} - -/* - * This function assumes to be called rarely. Switching between - * 8259A registers is slow. - * This has to be protected by the irq controller spinlock - * before being called. - */ -static inline int i8259A_irq_real(unsigned int irq) -{ - int value; - int irqmask = 1<<irq; - - if (irq < 8) { - outb(0x0B,PIC_MASTER_CMD); /* ISR register */ - value = inb(PIC_MASTER_CMD) & irqmask; - outb(0x0A,PIC_MASTER_CMD); /* back to the IRR register */ - return value; - } - outb(0x0B,PIC_SLAVE_CMD); /* ISR register */ - value = inb(PIC_SLAVE_CMD) & (irqmask >> 8); - outb(0x0A,PIC_SLAVE_CMD); /* back to the IRR register */ - return value; -} - -/* - * Careful! The 8259A is a fragile beast, it pretty - * much _has_ to be done exactly like this (mask it - * first, _then_ send the EOI, and the order of EOI - * to the two 8259s is important! - */ -static void mask_and_ack_8259A(unsigned int irq) -{ - unsigned int irqmask = 1 << irq; - unsigned long flags; - - spin_lock_irqsave(&i8259A_lock, flags); - /* - * Lightweight spurious IRQ detection. We do not want - * to overdo spurious IRQ handling - it's usually a sign - * of hardware problems, so we only do the checks we can - * do without slowing down good hardware unnecessarily. - * - * Note that IRQ7 and IRQ15 (the two spurious IRQs - * usually resulting from the 8259A-1|2 PICs) occur - * even if the IRQ is masked in the 8259A. Thus we - * can check spurious 8259A IRQs without doing the - * quite slow i8259A_irq_real() call for every IRQ. - * This does not cover 100% of spurious interrupts, - * but should be enough to warn the user that there - * is something bad going on ... - */ - if (cached_irq_mask & irqmask) - goto spurious_8259A_irq; - cached_irq_mask |= irqmask; - -handle_real_irq: - if (irq & 8) { - inb(PIC_SLAVE_IMR); /* DUMMY - (do we need this?) */ - outb(cached_slave_mask, PIC_SLAVE_IMR); - /* 'Specific EOI' to slave */ - outb(0x60+(irq&7),PIC_SLAVE_CMD); - /* 'Specific EOI' to master-IRQ2 */ - outb(0x60+PIC_CASCADE_IR,PIC_MASTER_CMD); - } else { - inb(PIC_MASTER_IMR); /* DUMMY - (do we need this?) */ - outb(cached_master_mask, PIC_MASTER_IMR); - /* 'Specific EOI' to master */ - outb(0x60+irq,PIC_MASTER_CMD); - } - spin_unlock_irqrestore(&i8259A_lock, flags); - return; - -spurious_8259A_irq: - /* - * this is the slow path - should happen rarely. - */ - if (i8259A_irq_real(irq)) - /* - * oops, the IRQ _is_ in service according to the - * 8259A - not spurious, go handle it. - */ - goto handle_real_irq; - - { - static int spurious_irq_mask; - /* - * At this point we can be sure the IRQ is spurious, - * lets ACK and report it. [once per IRQ] - */ - if (!(spurious_irq_mask & irqmask)) { - printk(KERN_DEBUG - "spurious 8259A interrupt: IRQ%d.\n", irq); - spurious_irq_mask |= irqmask; - } - atomic_inc(&irq_err_count); - /* - * Theoretically we do not have to handle this IRQ, - * but in Linux this does not cause problems and is - * simpler for us. - */ - goto handle_real_irq; - } -} - -static char irq_trigger[2]; -/** - * ELCR registers (0x4d0, 0x4d1) control edge/level of IRQ - */ -static void restore_ELCR(char *trigger) -{ - outb(trigger[0], 0x4d0); - outb(trigger[1], 0x4d1); -} - -static void save_ELCR(char *trigger) -{ - /* IRQ 0,1,2,8,13 are marked as reserved */ - trigger[0] = inb(0x4d0) & 0xF8; - trigger[1] = inb(0x4d1) & 0xDE; -} - -static int i8259A_resume(struct sys_device *dev) -{ - init_8259A(i8259A_auto_eoi); - restore_ELCR(irq_trigger); - return 0; -} - -static int i8259A_suspend(struct sys_device *dev, pm_message_t state) -{ - save_ELCR(irq_trigger); - return 0; -} - -static int i8259A_shutdown(struct sys_device *dev) -{ - /* Put the i8259A into a quiescent state that - * the kernel initialization code can get it - * out of. - */ - outb(0xff, PIC_MASTER_IMR); /* mask all of 8259A-1 */ - outb(0xff, PIC_SLAVE_IMR); /* mask all of 8259A-1 */ - return 0; -} - -static struct sysdev_class i8259_sysdev_class = { - .name = "i8259", - .suspend = i8259A_suspend, - .resume = i8259A_resume, - .shutdown = i8259A_shutdown, -}; - -static struct sys_device device_i8259A = { - .id = 0, - .cls = &i8259_sysdev_class, -}; - -static int __init i8259A_init_sysfs(void) -{ - int error = sysdev_class_register(&i8259_sysdev_class); - if (!error) - error = sysdev_register(&device_i8259A); - return error; -} - -device_initcall(i8259A_init_sysfs); - -void init_8259A(int auto_eoi) -{ - unsigned long flags; - - i8259A_auto_eoi = auto_eoi; - - spin_lock_irqsave(&i8259A_lock, flags); - - outb(0xff, PIC_MASTER_IMR); /* mask all of 8259A-1 */ - outb(0xff, PIC_SLAVE_IMR); /* mask all of 8259A-2 */ - - /* - * 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 */ - outb_pic(IRQ0_VECTOR, PIC_MASTER_IMR); - /* 8259A-1 (the master) has a slave on IR2 */ - outb_pic(0x04, PIC_MASTER_IMR); - if (auto_eoi) /* master does Auto EOI */ - outb_pic(MASTER_ICW4_DEFAULT | PIC_ICW4_AEOI, PIC_MASTER_IMR); - else /* master expects normal EOI */ - outb_pic(MASTER_ICW4_DEFAULT, PIC_MASTER_IMR); - - outb_pic(0x11, PIC_SLAVE_CMD); /* ICW1: select 8259A-2 init */ - /* ICW2: 8259A-2 IR0-7 mapped to 0x38-0x3f */ - outb_pic(IRQ8_VECTOR, PIC_SLAVE_IMR); - /* 8259A-2 is a slave on master's IR2 */ - outb_pic(PIC_CASCADE_IR, PIC_SLAVE_IMR); - /* (slave's support for AEOI in flat mode is to be investigated) */ - outb_pic(SLAVE_ICW4_DEFAULT, PIC_SLAVE_IMR); - - if (auto_eoi) - /* - * In AEOI mode we just have to mask the interrupt - * when acking. - */ - i8259A_chip.mask_ack = disable_8259A_irq; - else - i8259A_chip.mask_ack = mask_and_ack_8259A; - - udelay(100); /* wait for 8259A to initialize */ - - outb(cached_master_mask, PIC_MASTER_IMR); /* restore master IRQ mask */ - outb(cached_slave_mask, PIC_SLAVE_IMR); /* restore slave IRQ mask */ - - spin_unlock_irqrestore(&i8259A_lock, flags); -} - - - - -/* - * IRQ2 is cascade interrupt to second interrupt controller - */ - -static struct irqaction irq2 = { - .handler = no_action, - .mask = CPU_MASK_NONE, - .name = "cascade", -}; -DEFINE_PER_CPU(vector_irq_t, vector_irq) = { - [0 ... IRQ0_VECTOR - 1] = -1, - [IRQ0_VECTOR] = 0, - [IRQ1_VECTOR] = 1, - [IRQ2_VECTOR] = 2, - [IRQ3_VECTOR] = 3, - [IRQ4_VECTOR] = 4, - [IRQ5_VECTOR] = 5, - [IRQ6_VECTOR] = 6, - [IRQ7_VECTOR] = 7, - [IRQ8_VECTOR] = 8, - [IRQ9_VECTOR] = 9, - [IRQ10_VECTOR] = 10, - [IRQ11_VECTOR] = 11, - [IRQ12_VECTOR] = 12, - [IRQ13_VECTOR] = 13, - [IRQ14_VECTOR] = 14, - [IRQ15_VECTOR] = 15, - [IRQ15_VECTOR + 1 ... NR_VECTORS - 1] = -1 -}; - -void __init init_ISA_irqs (void) -{ - int i; - - init_bsp_APIC(); - init_8259A(0); - - for (i = 0; i < NR_IRQS; i++) { - irq_desc[i].status = IRQ_DISABLED; - irq_desc[i].action = NULL; - irq_desc[i].depth = 1; - - if (i < 16) { - /* - * 16 old-style INTA-cycle interrupts: - */ - set_irq_chip_and_handler_name(i, &i8259A_chip, - handle_level_irq, "XT"); - } else { - /* - * 'high' PCI IRQs filled in on demand - */ - irq_desc[i].chip = &no_irq_chip; - } - } -} - -void init_IRQ(void) __attribute__((weak, alias("native_init_IRQ"))); - -void __init native_init_IRQ(void) -{ - int i; - - init_ISA_irqs(); - /* - * Cover the whole vector space, no vector can escape - * us. (some of these will be overridden and become - * 'special' SMP interrupts) - */ - for (i = 0; i < (NR_VECTORS - FIRST_EXTERNAL_VECTOR); i++) { - int vector = FIRST_EXTERNAL_VECTOR + i; - if (vector != IA32_SYSCALL_VECTOR) - set_intr_gate(vector, interrupt[i]); - } - -#ifdef CONFIG_SMP - /* - * The reschedule interrupt is a CPU-to-CPU reschedule-helper - * IPI, driven by wakeup. - */ - set_intr_gate(RESCHEDULE_VECTOR, reschedule_interrupt); - - /* IPIs for invalidation */ - set_intr_gate(INVALIDATE_TLB_VECTOR_START+0, invalidate_interrupt0); - set_intr_gate(INVALIDATE_TLB_VECTOR_START+1, invalidate_interrupt1); - set_intr_gate(INVALIDATE_TLB_VECTOR_START+2, invalidate_interrupt2); - set_intr_gate(INVALIDATE_TLB_VECTOR_START+3, invalidate_interrupt3); - set_intr_gate(INVALIDATE_TLB_VECTOR_START+4, invalidate_interrupt4); - set_intr_gate(INVALIDATE_TLB_VECTOR_START+5, invalidate_interrupt5); - set_intr_gate(INVALIDATE_TLB_VECTOR_START+6, invalidate_interrupt6); - set_intr_gate(INVALIDATE_TLB_VECTOR_START+7, invalidate_interrupt7); - - /* IPI for generic function call */ - set_intr_gate(CALL_FUNCTION_VECTOR, call_function_interrupt); - - /* Low priority IPI to cleanup after moving an irq */ - set_intr_gate(IRQ_MOVE_CLEANUP_VECTOR, irq_move_cleanup_interrupt); -#endif - set_intr_gate(THERMAL_APIC_VECTOR, thermal_interrupt); - set_intr_gate(THRESHOLD_APIC_VECTOR, threshold_interrupt); - - /* self generated IPI for local APIC timer */ - set_intr_gate(LOCAL_TIMER_VECTOR, apic_timer_interrupt); - - /* IPI vectors for APIC spurious and error interrupts */ - set_intr_gate(SPURIOUS_APIC_VECTOR, spurious_interrupt); - set_intr_gate(ERROR_APIC_VECTOR, error_interrupt); - - if (!acpi_ioapic) - setup_irq(2, &irq2); -} diff --git a/arch/x86/kernel/io_apic_32.c b/arch/x86/kernel/io_apic_32.c index 4dc8600d9d2..603261a5885 100644 --- a/arch/x86/kernel/io_apic_32.c +++ b/arch/x86/kernel/io_apic_32.c @@ -25,6 +25,7 @@ #include <linux/init.h> #include <linux/delay.h> #include <linux/sched.h> +#include <linux/bootmem.h> #include <linux/mc146818rtc.h> #include <linux/compiler.h> #include <linux/acpi.h> @@ -58,7 +59,7 @@ static struct { int pin, apic; } ioapic_i8259 = { -1, -1 }; static DEFINE_SPINLOCK(ioapic_lock); static DEFINE_SPINLOCK(vector_lock); -int timer_over_8254 __initdata = 1; +int timer_through_8259 __initdata; /* * Is the SiS APIC rmw bug present ? @@ -72,15 +73,21 @@ int sis_apic_bug = -1; int nr_ioapic_registers[MAX_IO_APICS]; /* I/O APIC entries */ -struct mpc_config_ioapic mp_ioapics[MAX_IO_APICS]; +struct mp_config_ioapic mp_ioapics[MAX_IO_APICS]; int nr_ioapics; /* MP IRQ source entries */ -struct mpc_config_intsrc mp_irqs[MAX_IRQ_SOURCES]; +struct mp_config_intsrc mp_irqs[MAX_IRQ_SOURCES]; /* # of MP IRQ source entries */ int mp_irq_entries; +#if defined (CONFIG_MCA) || defined (CONFIG_EISA) +int mp_bus_id_to_type[MAX_MP_BUSSES]; +#endif + +DECLARE_BITMAP(mp_bus_not_pci, MAX_MP_BUSSES); + static int disable_timer_pin_1 __initdata; /* @@ -110,7 +117,7 @@ struct io_apic { static __attribute_const__ struct io_apic __iomem *io_apic_base(int idx) { return (void __iomem *) __fix_to_virt(FIX_IO_APIC_BASE_0 + idx) - + (mp_ioapics[idx].mpc_apicaddr & ~PAGE_MASK); + + (mp_ioapics[idx].mp_apicaddr & ~PAGE_MASK); } static inline unsigned int io_apic_read(unsigned int apic, unsigned int reg) @@ -239,7 +246,7 @@ static void __init replace_pin_at_irq(unsigned int irq, } } -static void __modify_IO_APIC_irq (unsigned int irq, unsigned long enable, unsigned long disable) +static void __modify_IO_APIC_irq(unsigned int irq, unsigned long enable, unsigned long disable) { struct irq_pin_list *entry = irq_2_pin + irq; unsigned int pin, reg; @@ -259,30 +266,32 @@ static void __modify_IO_APIC_irq (unsigned int irq, unsigned long enable, unsign } /* mask = 1 */ -static void __mask_IO_APIC_irq (unsigned int irq) +static void __mask_IO_APIC_irq(unsigned int irq) { - __modify_IO_APIC_irq(irq, 0x00010000, 0); + __modify_IO_APIC_irq(irq, IO_APIC_REDIR_MASKED, 0); } /* mask = 0 */ -static void __unmask_IO_APIC_irq (unsigned int irq) +static void __unmask_IO_APIC_irq(unsigned int irq) { - __modify_IO_APIC_irq(irq, 0, 0x00010000); + __modify_IO_APIC_irq(irq, 0, IO_APIC_REDIR_MASKED); } /* mask = 1, trigger = 0 */ -static void __mask_and_edge_IO_APIC_irq (unsigned int irq) +static void __mask_and_edge_IO_APIC_irq(unsigned int irq) { - __modify_IO_APIC_irq(irq, 0x00010000, 0x00008000); + __modify_IO_APIC_irq(irq, IO_APIC_REDIR_MASKED, + IO_APIC_REDIR_LEVEL_TRIGGER); } /* mask = 0, trigger = 1 */ -static void __unmask_and_level_IO_APIC_irq (unsigned int irq) +static void __unmask_and_level_IO_APIC_irq(unsigned int irq) { - __modify_IO_APIC_irq(irq, 0x00008000, 0x00010000); + __modify_IO_APIC_irq(irq, IO_APIC_REDIR_LEVEL_TRIGGER, + IO_APIC_REDIR_MASKED); } -static void mask_IO_APIC_irq (unsigned int irq) +static void mask_IO_APIC_irq(unsigned int irq) { unsigned long flags; @@ -291,7 +300,7 @@ static void mask_IO_APIC_irq (unsigned int irq) spin_unlock_irqrestore(&ioapic_lock, flags); } -static void unmask_IO_APIC_irq (unsigned int irq) +static void unmask_IO_APIC_irq(unsigned int irq) { unsigned long flags; @@ -303,7 +312,7 @@ static void unmask_IO_APIC_irq (unsigned int irq) static void clear_IO_APIC_pin(unsigned int apic, unsigned int pin) { struct IO_APIC_route_entry entry; - + /* Check delivery_mode to be sure we're not clearing an SMI pin */ entry = ioapic_read_entry(apic, pin); if (entry.delivery_mode == dest_SMI) @@ -315,7 +324,7 @@ static void clear_IO_APIC_pin(unsigned int apic, unsigned int pin) ioapic_mask_entry(apic, pin); } -static void clear_IO_APIC (void) +static void clear_IO_APIC(void) { int apic, pin; @@ -332,7 +341,7 @@ static void set_ioapic_affinity_irq(unsigned int irq, cpumask_t cpumask) struct irq_pin_list *entry = irq_2_pin + irq; unsigned int apicid_value; cpumask_t tmp; - + cpus_and(tmp, cpumask, cpu_online_map); if (cpus_empty(tmp)) tmp = TARGET_CPUS; @@ -361,7 +370,7 @@ static void set_ioapic_affinity_irq(unsigned int irq, cpumask_t cpumask) # include <linux/kernel_stat.h> /* kstat */ # include <linux/slab.h> /* kmalloc() */ # include <linux/timer.h> - + #define IRQBALANCE_CHECK_ARCH -999 #define MAX_BALANCED_IRQ_INTERVAL (5*HZ) #define MIN_BALANCED_IRQ_INTERVAL (HZ/2) @@ -373,14 +382,14 @@ static int physical_balance __read_mostly; static long balanced_irq_interval __read_mostly = MAX_BALANCED_IRQ_INTERVAL; static struct irq_cpu_info { - unsigned long * last_irq; - unsigned long * irq_delta; + unsigned long *last_irq; + unsigned long *irq_delta; unsigned long irq; } irq_cpu_data[NR_CPUS]; #define CPU_IRQ(cpu) (irq_cpu_data[cpu].irq) -#define LAST_CPU_IRQ(cpu,irq) (irq_cpu_data[cpu].last_irq[irq]) -#define IRQ_DELTA(cpu,irq) (irq_cpu_data[cpu].irq_delta[irq]) +#define LAST_CPU_IRQ(cpu, irq) (irq_cpu_data[cpu].last_irq[irq]) +#define IRQ_DELTA(cpu, irq) (irq_cpu_data[cpu].irq_delta[irq]) #define IDLE_ENOUGH(cpu,now) \ (idle_cpu(cpu) && ((now) - per_cpu(irq_stat, (cpu)).idle_timestamp > 1)) @@ -419,8 +428,8 @@ inside: if (cpu == -1) cpu = NR_CPUS-1; } - } while (!cpu_online(cpu) || !IRQ_ALLOWED(cpu,allowed_mask) || - (search_idle && !IDLE_ENOUGH(cpu,now))); + } while (!cpu_online(cpu) || !IRQ_ALLOWED(cpu, allowed_mask) || + (search_idle && !IDLE_ENOUGH(cpu, now))); return cpu; } @@ -430,15 +439,14 @@ static inline void balance_irq(int cpu, int irq) unsigned long now = jiffies; cpumask_t allowed_mask; unsigned int new_cpu; - + if (irqbalance_disabled) - return; + return; cpus_and(allowed_mask, cpu_online_map, balance_irq_affinity[irq]); new_cpu = move(cpu, allowed_mask, now, 1); - if (cpu != new_cpu) { + if (cpu != new_cpu) set_pending_irq(irq, cpumask_of_cpu(new_cpu)); - } } static inline void rotate_irqs_among_cpus(unsigned long useful_load_threshold) @@ -450,14 +458,14 @@ static inline void rotate_irqs_among_cpus(unsigned long useful_load_threshold) if (!irq_desc[j].action) continue; /* Is it a significant load ? */ - if (IRQ_DELTA(CPU_TO_PACKAGEINDEX(i),j) < + if (IRQ_DELTA(CPU_TO_PACKAGEINDEX(i), j) < useful_load_threshold) continue; balance_irq(i, j); } } balanced_irq_interval = max((long)MIN_BALANCED_IRQ_INTERVAL, - balanced_irq_interval - BALANCED_IRQ_LESS_DELTA); + balanced_irq_interval - BALANCED_IRQ_LESS_DELTA); return; } @@ -486,22 +494,22 @@ static void do_irq_balance(void) /* Is this an active IRQ or balancing disabled ? */ if (!irq_desc[j].action || irq_balancing_disabled(j)) continue; - if ( package_index == i ) - IRQ_DELTA(package_index,j) = 0; + if (package_index == i) + IRQ_DELTA(package_index, j) = 0; /* Determine the total count per processor per IRQ */ value_now = (unsigned long) kstat_cpu(i).irqs[j]; /* Determine the activity per processor per IRQ */ - delta = value_now - LAST_CPU_IRQ(i,j); + delta = value_now - LAST_CPU_IRQ(i, j); /* Update last_cpu_irq[][] for the next time */ - LAST_CPU_IRQ(i,j) = value_now; + LAST_CPU_IRQ(i, j) = value_now; /* Ignore IRQs whose rate is less than the clock */ if (delta < useful_load_threshold) continue; /* update the load for the processor or package total */ - IRQ_DELTA(package_index,j) += delta; + IRQ_DELTA(package_index, j) += delta; /* Keep track of the higher numbered sibling as well */ if (i != package_index) @@ -527,7 +535,8 @@ static void do_irq_balance(void) max_cpu_irq = ULONG_MAX; tryanothercpu: - /* Look for heaviest loaded processor. + /* + * Look for heaviest loaded processor. * We may come back to get the next heaviest loaded processor. * Skip processors with trivial loads. */ @@ -536,7 +545,7 @@ tryanothercpu: for_each_online_cpu(i) { if (i != CPU_TO_PACKAGEINDEX(i)) continue; - if (max_cpu_irq <= CPU_IRQ(i)) + if (max_cpu_irq <= CPU_IRQ(i)) continue; if (tmp_cpu_irq < CPU_IRQ(i)) { tmp_cpu_irq = CPU_IRQ(i); @@ -545,8 +554,9 @@ tryanothercpu: } if (tmp_loaded == -1) { - /* In the case of small number of heavy interrupt sources, - * loading some of the cpus too much. We use Ingo's original + /* + * In the case of small number of heavy interrupt sources, + * loading some of the cpus too much. We use Ingo's original * approach to rotate them around. */ if (!first_attempt && imbalance >= useful_load_threshold) { @@ -555,13 +565,14 @@ tryanothercpu: } goto not_worth_the_effort; } - + first_attempt = 0; /* heaviest search */ max_cpu_irq = tmp_cpu_irq; /* load */ max_loaded = tmp_loaded; /* processor */ imbalance = (max_cpu_irq - min_cpu_irq) / 2; - - /* if imbalance is less than approx 10% of max load, then + + /* + * if imbalance is less than approx 10% of max load, then * observe diminishing returns action. - quit */ if (imbalance < (max_cpu_irq >> 3)) @@ -577,26 +588,25 @@ tryanotherirq: /* Is this an active IRQ? */ if (!irq_desc[j].action) continue; - if (imbalance <= IRQ_DELTA(max_loaded,j)) + if (imbalance <= IRQ_DELTA(max_loaded, j)) continue; /* Try to find the IRQ that is closest to the imbalance * without going over. */ - if (move_this_load < IRQ_DELTA(max_loaded,j)) { - move_this_load = IRQ_DELTA(max_loaded,j); + if (move_this_load < IRQ_DELTA(max_loaded, j)) { + move_this_load = IRQ_DELTA(max_loaded, j); selected_irq = j; } } - if (selected_irq == -1) { + if (selected_irq == -1) goto tryanothercpu; - } imbalance = move_this_load; - + /* For physical_balance case, we accumulated both load * values in the one of the siblings cpu_irq[], * to use the same code for physical and logical processors - * as much as possible. + * as much as possible. * * NOTE: the cpu_irq[] array holds the sum of the load for * sibling A and sibling B in the slot for the lowest numbered @@ -625,11 +635,11 @@ tryanotherirq: /* mark for change destination */ set_pending_irq(selected_irq, cpumask_of_cpu(min_loaded)); - /* Since we made a change, come back sooner to + /* Since we made a change, come back sooner to * check for more variation. */ balanced_irq_interval = max((long)MIN_BALANCED_IRQ_INTERVAL, - balanced_irq_interval - BALANCED_IRQ_LESS_DELTA); + balanced_irq_interval - BALANCED_IRQ_LESS_DELTA); return; } goto tryanotherirq; @@ -640,7 +650,7 @@ not_worth_the_effort: * upward */ balanced_irq_interval = min((long)MAX_BALANCED_IRQ_INTERVAL, - balanced_irq_interval + BALANCED_IRQ_MORE_DELTA); + balanced_irq_interval + BALANCED_IRQ_MORE_DELTA); return; } @@ -679,13 +689,13 @@ static int __init balanced_irq_init(void) cpumask_t tmp; cpus_shift_right(tmp, cpu_online_map, 2); - c = &boot_cpu_data; + c = &boot_cpu_data; /* When not overwritten by the command line ask subarchitecture. */ if (irqbalance_disabled == IRQBALANCE_CHECK_ARCH) irqbalance_disabled = NO_BALANCE_IRQ; if (irqbalance_disabled) return 0; - + /* disable irqbalance completely if there is only one processor online */ if (num_online_cpus() < 2) { irqbalance_disabled = 1; @@ -699,16 +709,14 @@ static int __init balanced_irq_init(void) physical_balance = 1; for_each_online_cpu(i) { - irq_cpu_data[i].irq_delta = kmalloc(sizeof(unsigned long) * NR_IRQS, GFP_KERNEL); - irq_cpu_data[i].last_irq = kmalloc(sizeof(unsigned long) * NR_IRQS, GFP_KERNEL); + irq_cpu_data[i].irq_delta = kzalloc(sizeof(unsigned long) * NR_IRQS, GFP_KERNEL); + irq_cpu_data[i].last_irq = kzalloc(sizeof(unsigned long) * NR_IRQS, GFP_KERNEL); if (irq_cpu_data[i].irq_delta == NULL || irq_cpu_data[i].last_irq == NULL) { printk(KERN_ERR "balanced_irq_init: out of memory"); goto failed; } - memset(irq_cpu_data[i].irq_delta,0,sizeof(unsigned long) * NR_IRQS); - memset(irq_cpu_data[i].last_irq,0,sizeof(unsigned long) * NR_IRQS); } - + printk(KERN_INFO "Starting balanced_irq\n"); if (!IS_ERR(kthread_run(balanced_irq, NULL, "kirqd"))) return 0; @@ -801,10 +809,10 @@ static int find_irq_entry(int apic, int pin, int type) int i; for (i = 0; i < mp_irq_entries; i++) - if (mp_irqs[i].mpc_irqtype == type && - (mp_irqs[i].mpc_dstapic == mp_ioapics[apic].mpc_apicid || - mp_irqs[i].mpc_dstapic == MP_APIC_ALL) && - mp_irqs[i].mpc_dstirq == pin) + if (mp_irqs[i].mp_irqtype == type && + (mp_irqs[i].mp_dstapic == mp_ioapics[apic].mp_apicid || + mp_irqs[i].mp_dstapic == MP_APIC_ALL) && + mp_irqs[i].mp_dstirq == pin) return i; return -1; @@ -818,13 +826,13 @@ static int __init find_isa_irq_pin(int irq, int type) int i; for (i = 0; i < mp_irq_entries; i++) { - int lbus = mp_irqs[i].mpc_srcbus; + int lbus = mp_irqs[i].mp_srcbus; if (test_bit(lbus, mp_bus_not_pci) && - (mp_irqs[i].mpc_irqtype == type) && - (mp_irqs[i].mpc_srcbusirq == irq)) + (mp_irqs[i].mp_irqtype == type) && + (mp_irqs[i].mp_srcbusirq == irq)) - return mp_irqs[i].mpc_dstirq; + return mp_irqs[i].mp_dstirq; } return -1; } @@ -834,17 +842,17 @@ static int __init find_isa_irq_apic(int irq, int type) int i; for (i = 0; i < mp_irq_entries; i++) { - int lbus = mp_irqs[i].mpc_srcbus; + int lbus = mp_irqs[i].mp_srcbus; if (test_bit(lbus, mp_bus_not_pci) && - (mp_irqs[i].mpc_irqtype == type) && - (mp_irqs[i].mpc_srcbusirq == irq)) + (mp_irqs[i].mp_irqtype == type) && + (mp_irqs[i].mp_srcbusirq == irq)) break; } if (i < mp_irq_entries) { int apic; - for(apic = 0; apic < nr_ioapics; apic++) { - if (mp_ioapics[apic].mpc_apicid == mp_irqs[i].mpc_dstapic) + for (apic = 0; apic < nr_ioapics; apic++) { + if (mp_ioapics[apic].mp_apicid == mp_irqs[i].mp_dstapic) return apic; } } @@ -864,28 +872,28 @@ int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin) apic_printk(APIC_DEBUG, "querying PCI -> IRQ mapping bus:%d, " "slot:%d, pin:%d.\n", bus, slot, pin); - if (mp_bus_id_to_pci_bus[bus] == -1) { + if (test_bit(bus, mp_bus_not_pci)) { printk(KERN_WARNING "PCI BIOS passed nonexistent PCI bus %d!\n", bus); return -1; } for (i = 0; i < mp_irq_entries; i++) { - int lbus = mp_irqs[i].mpc_srcbus; + int lbus = mp_irqs[i].mp_srcbus; for (apic = 0; apic < nr_ioapics; apic++) - if (mp_ioapics[apic].mpc_apicid == mp_irqs[i].mpc_dstapic || - mp_irqs[i].mpc_dstapic == MP_APIC_ALL) + if (mp_ioapics[apic].mp_apicid == mp_irqs[i].mp_dstapic || + mp_irqs[i].mp_dstapic == MP_APIC_ALL) break; if (!test_bit(lbus, mp_bus_not_pci) && - !mp_irqs[i].mpc_irqtype && + !mp_irqs[i].mp_irqtype && (bus == lbus) && - (slot == ((mp_irqs[i].mpc_srcbusirq >> 2) & 0x1f))) { - int irq = pin_2_irq(i,apic,mp_irqs[i].mpc_dstirq); + (slot == ((mp_irqs[i].mp_srcbusirq >> 2) & 0x1f))) { + int irq = pin_2_irq(i, apic, mp_irqs[i].mp_dstirq); if (!(apic || IO_APIC_IRQ(irq))) continue; - if (pin == (mp_irqs[i].mpc_srcbusirq & 3)) + if (pin == (mp_irqs[i].mp_srcbusirq & 3)) return irq; /* * Use the first all-but-pin matching entry as a @@ -900,7 +908,7 @@ int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin) EXPORT_SYMBOL(IO_APIC_get_PCI_irq_vector); /* - * This function currently is only a helper for the i386 smp boot process where + * This function currently is only a helper for the i386 smp boot process where * we need to reprogram the ioredtbls to cater for the cpus which have come online * so mask in all cases should simply be TARGET_CPUS */ @@ -952,7 +960,7 @@ static int EISA_ELCR(unsigned int irq) * EISA conforming in the MP table, that means its trigger type must * be read in from the ELCR */ -#define default_EISA_trigger(idx) (EISA_ELCR(mp_irqs[idx].mpc_srcbusirq)) +#define default_EISA_trigger(idx) (EISA_ELCR(mp_irqs[idx].mp_srcbusirq)) #define default_EISA_polarity(idx) default_ISA_polarity(idx) /* PCI interrupts are always polarity one level triggered, @@ -969,118 +977,115 @@ static int EISA_ELCR(unsigned int irq) static int MPBIOS_polarity(int idx) { - int bus = mp_irqs[idx].mpc_srcbus; + int bus = mp_irqs[idx].mp_srcbus; int polarity; /* * Determine IRQ line polarity (high active or low active): */ - switch (mp_irqs[idx].mpc_irqflag & 3) + switch (mp_irqs[idx].mp_irqflag & 3) { + case 0: /* conforms, ie. bus-type dependent polarity */ { - case 0: /* conforms, ie. bus-type dependent polarity */ - { - polarity = test_bit(bus, mp_bus_not_pci)? - default_ISA_polarity(idx): - default_PCI_polarity(idx); - break; - } - case 1: /* high active */ - { - polarity = 0; - break; - } - case 2: /* reserved */ - { - printk(KERN_WARNING "broken BIOS!!\n"); - polarity = 1; - break; - } - case 3: /* low active */ - { - polarity = 1; - break; - } - default: /* invalid */ - { - printk(KERN_WARNING "broken BIOS!!\n"); - polarity = 1; - break; - } + polarity = test_bit(bus, mp_bus_not_pci)? + default_ISA_polarity(idx): + default_PCI_polarity(idx); + break; + } + case 1: /* high active */ + { + polarity = 0; + break; + } + case 2: /* reserved */ + { + printk(KERN_WARNING "broken BIOS!!\n"); + polarity = 1; + break; + } + case 3: /* low active */ + { + polarity = 1; + break; + } + default: /* invalid */ + { + printk(KERN_WARNING "broken BIOS!!\n"); + polarity = 1; + break; + } } return polarity; } static int MPBIOS_trigger(int idx) { - int bus = mp_irqs[idx].mpc_srcbus; + int bus = mp_irqs[idx].mp_srcbus; int trigger; /* * Determine IRQ trigger mode (edge or level sensitive): */ - switch ((mp_irqs[idx].mpc_irqflag>>2) & 3) + switch ((mp_irqs[idx].mp_irqflag>>2) & 3) { + case 0: /* conforms, ie. bus-type dependent */ { - case 0: /* conforms, ie. bus-type dependent */ - { - trigger = test_bit(bus, mp_bus_not_pci)? - default_ISA_trigger(idx): - default_PCI_trigger(idx); + trigger = test_bit(bus, mp_bus_not_pci)? + default_ISA_trigger(idx): + default_PCI_trigger(idx); #if defined(CONFIG_EISA) || defined(CONFIG_MCA) - switch (mp_bus_id_to_type[bus]) - { - case MP_BUS_ISA: /* ISA pin */ - { - /* set before the switch */ - break; - } - case MP_BUS_EISA: /* EISA pin */ - { - trigger = default_EISA_trigger(idx); - break; - } - case MP_BUS_PCI: /* PCI pin */ - { - /* set before the switch */ - break; - } - case MP_BUS_MCA: /* MCA pin */ - { - trigger = default_MCA_trigger(idx); - break; - } - default: - { - printk(KERN_WARNING "broken BIOS!!\n"); - trigger = 1; - break; - } - } -#endif + switch (mp_bus_id_to_type[bus]) { + case MP_BUS_ISA: /* ISA pin */ + { + /* set before the switch */ break; } - case 1: /* edge */ + case MP_BUS_EISA: /* EISA pin */ { - trigger = 0; + trigger = default_EISA_trigger(idx); break; } - case 2: /* reserved */ + case MP_BUS_PCI: /* PCI pin */ { - printk(KERN_WARNING "broken BIOS!!\n"); - trigger = 1; + /* set before the switch */ break; } - case 3: /* level */ + case MP_BUS_MCA: /* MCA pin */ { - trigger = 1; + trigger = default_MCA_trigger(idx); break; } - default: /* invalid */ + default: { printk(KERN_WARNING "broken BIOS!!\n"); - trigger = 0; + trigger = 1; break; } } +#endif + break; + } + case 1: /* edge */ + { + trigger = 0; + break; + } + case 2: /* reserved */ + { + printk(KERN_WARNING "broken BIOS!!\n"); + trigger = 1; + break; + } + case 3: /* level */ + { + trigger = 1; + break; + } + default: /* invalid */ + { + printk(KERN_WARNING "broken BIOS!!\n"); + trigger = 0; + break; + } + } return trigger; } @@ -1097,16 +1102,16 @@ static inline int irq_trigger(int idx) static int pin_2_irq(int idx, int apic, int pin) { int irq, i; - int bus = mp_irqs[idx].mpc_srcbus; + int bus = mp_irqs[idx].mp_srcbus; /* * Debugging check, we are in big trouble if this message pops up! */ - if (mp_irqs[idx].mpc_dstirq != pin) + if (mp_irqs[idx].mp_dstirq != pin) printk(KERN_ERR "broken BIOS or MPTABLE parser, ayiee!!\n"); if (test_bit(bus, mp_bus_not_pci)) - irq = mp_irqs[idx].mpc_srcbusirq; + irq = mp_irqs[idx].mp_srcbusirq; else { /* * PCI IRQs are mapped in order @@ -1148,8 +1153,8 @@ static inline int IO_APIC_irq_trigger(int irq) for (apic = 0; apic < nr_ioapics; apic++) { for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) { - idx = find_irq_entry(apic,pin,mp_INT); - if ((idx != -1) && (irq == pin_2_irq(idx,apic,pin))) + idx = find_irq_entry(apic, pin, mp_INT); + if ((idx != -1) && (irq == pin_2_irq(idx, apic, pin))) return irq_trigger(idx); } } @@ -1164,7 +1169,7 @@ static u8 irq_vector[NR_IRQ_VECTORS] __read_mostly = { FIRST_DEVICE_VECTOR , 0 } static int __assign_irq_vector(int irq) { - static int current_vector = FIRST_DEVICE_VECTOR, current_offset = 0; + static int current_vector = FIRST_DEVICE_VECTOR, current_offset; int vector, offset; BUG_ON((unsigned)irq >= NR_IRQ_VECTORS); @@ -1176,7 +1181,7 @@ static int __assign_irq_vector(int irq) offset = current_offset; next: vector += 8; - if (vector >= FIRST_SYSTEM_VECTOR) { + if (vector >= first_system_vector) { offset = (offset + 1) % 8; vector = FIRST_DEVICE_VECTOR + offset; } @@ -1203,6 +1208,11 @@ static int assign_irq_vector(int irq) return vector; } + +void setup_vector_irq(int cpu) +{ +} + static struct irq_chip ioapic_chip; #define IOAPIC_AUTO -1 @@ -1237,25 +1247,25 @@ static void __init setup_IO_APIC_irqs(void) /* * add it to the IO-APIC irq-routing table: */ - memset(&entry,0,sizeof(entry)); + memset(&entry, 0, sizeof(entry)); entry.delivery_mode = INT_DELIVERY_MODE; entry.dest_mode = INT_DEST_MODE; entry.mask = 0; /* enable IRQ */ - entry.dest.logical.logical_dest = + entry.dest.logical.logical_dest = cpu_mask_to_apicid(TARGET_CPUS); - idx = find_irq_entry(apic,pin,mp_INT); + idx = find_irq_entry(apic, pin, mp_INT); if (idx == -1) { if (first_notcon) { apic_printk(APIC_VERBOSE, KERN_DEBUG " IO-APIC (apicid-pin) %d-%d", - mp_ioapics[apic].mpc_apicid, + mp_ioapics[apic].mp_apicid, pin); first_notcon = 0; } else apic_printk(APIC_VERBOSE, ", %d-%d", - mp_ioapics[apic].mpc_apicid, pin); + mp_ioapics[apic].mp_apicid, pin); continue; } @@ -1289,7 +1299,7 @@ static void __init setup_IO_APIC_irqs(void) vector = assign_irq_vector(irq); entry.vector = vector; ioapic_register_intr(irq, vector, IOAPIC_AUTO); - + if (!apic && (irq < 16)) disable_8259A_irq(irq); } @@ -1302,25 +1312,21 @@ static void __init setup_IO_APIC_irqs(void) } /* - * Set up the 8259A-master output pin: + * Set up the timer pin, possibly with the 8259A-master behind. */ -static void __init setup_ExtINT_IRQ0_pin(unsigned int apic, unsigned int pin, int vector) +static void __init setup_timer_IRQ0_pin(unsigned int apic, unsigned int pin, + int vector) { struct IO_APIC_route_entry entry; - memset(&entry,0,sizeof(entry)); - - disable_8259A_irq(0); - - /* mask LVT0 */ - apic_write_around(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT); + memset(&entry, 0, sizeof(entry)); /* * We use logical delivery to get the timer IRQ * to the first CPU. */ entry.dest_mode = INT_DEST_MODE; - entry.mask = 0; /* unmask IRQ now */ + entry.mask = 1; /* mask IRQ now */ entry.dest.logical.logical_dest = cpu_mask_to_apicid(TARGET_CPUS); entry.delivery_mode = INT_DELIVERY_MODE; entry.polarity = 0; @@ -1329,17 +1335,14 @@ static void __init setup_ExtINT_IRQ0_pin(unsigned int apic, unsigned int pin, in /* * The timer IRQ doesn't have to know that behind the - * scene we have a 8259A-master in AEOI mode ... + * scene we may have a 8259A-master in AEOI mode ... */ - irq_desc[0].chip = &ioapic_chip; - set_irq_handler(0, handle_edge_irq); + ioapic_register_intr(0, vector, IOAPIC_EDGE); /* * Add it to the IO-APIC irq-routing table: */ ioapic_write_entry(apic, pin, entry); - - enable_8259A_irq(0); } void __init print_IO_APIC(void) @@ -1354,10 +1357,10 @@ void __init print_IO_APIC(void) if (apic_verbosity == APIC_QUIET) return; - printk(KERN_DEBUG "number of MP IRQ sources: %d.\n", mp_irq_entries); + printk(KERN_DEBUG "number of MP IRQ sources: %d.\n", mp_irq_entries); for (i = 0; i < nr_ioapics; i++) printk(KERN_DEBUG "number of IO-APIC #%d registers: %d.\n", - mp_ioapics[i].mpc_apicid, nr_ioapic_registers[i]); + mp_ioapics[i].mp_apicid, nr_ioapic_registers[i]); /* * We are a bit conservative about what we expect. We have to @@ -1376,7 +1379,7 @@ void __init print_IO_APIC(void) reg_03.raw = io_apic_read(apic, 3); spin_unlock_irqrestore(&ioapic_lock, flags); - printk(KERN_DEBUG "IO APIC #%d......\n", mp_ioapics[apic].mpc_apicid); + printk(KERN_DEBUG "IO APIC #%d......\n", mp_ioapics[apic].mp_apicid); printk(KERN_DEBUG ".... register #00: %08X\n", reg_00.raw); printk(KERN_DEBUG "....... : physical APIC id: %02X\n", reg_00.bits.ID); printk(KERN_DEBUG "....... : Delivery Type: %X\n", reg_00.bits.delivery_type); @@ -1459,7 +1462,7 @@ void __init print_IO_APIC(void) #if 0 -static void print_APIC_bitfield (int base) +static void print_APIC_bitfield(int base) { unsigned int v; int i, j; @@ -1480,7 +1483,7 @@ static void print_APIC_bitfield (int base) } } -void /*__init*/ print_local_APIC(void * dummy) +void /*__init*/ print_local_APIC(void *dummy) { unsigned int v, ver, maxlvt; @@ -1489,6 +1492,7 @@ void /*__init*/ print_local_APIC(void * dummy) printk("\n" KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n", smp_processor_id(), hard_smp_processor_id()); + v = apic_read(APIC_ID); printk(KERN_INFO "... APIC ID: %08x (%01x)\n", v, GET_APIC_ID(read_apic_id())); v = apic_read(APIC_LVR); @@ -1563,7 +1567,7 @@ void /*__init*/ print_local_APIC(void * dummy) printk("\n"); } -void print_all_local_APICs (void) +void print_all_local_APICs(void) { on_each_cpu(print_local_APIC, NULL, 1, 1); } @@ -1586,11 +1590,11 @@ void /*__init*/ print_PIC(void) v = inb(0xa0) << 8 | inb(0x20); printk(KERN_DEBUG "... PIC IRR: %04x\n", v); - outb(0x0b,0xa0); - outb(0x0b,0x20); + outb(0x0b, 0xa0); + outb(0x0b, 0x20); v = inb(0xa0) << 8 | inb(0x20); - outb(0x0a,0xa0); - outb(0x0a,0x20); + outb(0x0a, 0xa0); + outb(0x0a, 0x20); spin_unlock_irqrestore(&i8259A_lock, flags); @@ -1626,7 +1630,7 @@ static void __init enable_IO_APIC(void) spin_unlock_irqrestore(&ioapic_lock, flags); nr_ioapic_registers[apic] = reg_01.bits.entries+1; } - for(apic = 0; apic < nr_ioapics; apic++) { + for (apic = 0; apic < nr_ioapics; apic++) { int pin; /* See if any of the pins is in ExtINT mode */ for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) { @@ -1716,7 +1720,6 @@ void disable_IO_APIC(void) * by Matt Domsch <Matt_Domsch@dell.com> Tue Dec 21 12:25:05 CST 1999 */ -#ifndef CONFIG_X86_NUMAQ static void __init setup_ioapic_ids_from_mpc(void) { union IO_APIC_reg_00 reg_00; @@ -1726,6 +1729,11 @@ static void __init setup_ioapic_ids_from_mpc(void) unsigned char old_id; unsigned long flags; +#ifdef CONFIG_X86_NUMAQ + if (found_numaq) + return; +#endif + /* * Don't check I/O APIC IDs for xAPIC systems. They have * no meaning without the serial APIC bus. @@ -1748,15 +1756,15 @@ static void __init setup_ioapic_ids_from_mpc(void) spin_lock_irqsave(&ioapic_lock, flags); reg_00.raw = io_apic_read(apic, 0); spin_unlock_irqrestore(&ioapic_lock, flags); - - old_id = mp_ioapics[apic].mpc_apicid; - if (mp_ioapics[apic].mpc_apicid >= get_physical_broadcast()) { + old_id = mp_ioapics[apic].mp_apicid; + + if (mp_ioapics[apic].mp_apicid >= get_physical_broadcast()) { printk(KERN_ERR "BIOS bug, IO-APIC#%d ID is %d in the MPC table!...\n", - apic, mp_ioapics[apic].mpc_apicid); + apic, mp_ioapics[apic].mp_apicid); printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n", reg_00.bits.ID); - mp_ioapics[apic].mpc_apicid = reg_00.bits.ID; + mp_ioapics[apic].mp_apicid = reg_00.bits.ID; } /* @@ -1765,9 +1773,9 @@ static void __init setup_ioapic_ids_from_mpc(void) * 'stuck on smp_invalidate_needed IPI wait' messages. */ if (check_apicid_used(phys_id_present_map, - mp_ioapics[apic].mpc_apicid)) { + mp_ioapics[apic].mp_apicid)) { printk(KERN_ERR "BIOS bug, IO-APIC#%d ID %d is already used!...\n", - apic, mp_ioapics[apic].mpc_apicid); + apic, mp_ioapics[apic].mp_apicid); for (i = 0; i < get_physical_broadcast(); i++) if (!physid_isset(i, phys_id_present_map)) break; @@ -1776,13 +1784,13 @@ static void __init setup_ioapic_ids_from_mpc(void) printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n", i); physid_set(i, phys_id_present_map); - mp_ioapics[apic].mpc_apicid = i; + mp_ioapics[apic].mp_apicid = i; } else { physid_mask_t tmp; - tmp = apicid_to_cpu_present(mp_ioapics[apic].mpc_apicid); + tmp = apicid_to_cpu_present(mp_ioapics[apic].mp_apicid); apic_printk(APIC_VERBOSE, "Setting %d in the " "phys_id_present_map\n", - mp_ioapics[apic].mpc_apicid); + mp_ioapics[apic].mp_apicid); physids_or(phys_id_present_map, phys_id_present_map, tmp); } @@ -1791,21 +1799,21 @@ static void __init setup_ioapic_ids_from_mpc(void) * We need to adjust the IRQ routing table * if the ID changed. */ - if (old_id != mp_ioapics[apic].mpc_apicid) + if (old_id != mp_ioapics[apic].mp_apicid) for (i = 0; i < mp_irq_entries; i++) - if (mp_irqs[i].mpc_dstapic == old_id) - mp_irqs[i].mpc_dstapic - = mp_ioapics[apic].mpc_apicid; + if (mp_irqs[i].mp_dstapic == old_id) + mp_irqs[i].mp_dstapic + = mp_ioapics[apic].mp_apicid; /* * Read the right value from the MPC table and * write it into the ID register. - */ + */ apic_printk(APIC_VERBOSE, KERN_INFO "...changing IO-APIC physical APIC ID to %d ...", - mp_ioapics[apic].mpc_apicid); + mp_ioapics[apic].mp_apicid); - reg_00.bits.ID = mp_ioapics[apic].mpc_apicid; + reg_00.bits.ID = mp_ioapics[apic].mp_apicid; spin_lock_irqsave(&ioapic_lock, flags); io_apic_write(apic, 0, reg_00.raw); spin_unlock_irqrestore(&ioapic_lock, flags); @@ -1816,15 +1824,12 @@ static void __init setup_ioapic_ids_from_mpc(void) spin_lock_irqsave(&ioapic_lock, flags); reg_00.raw = io_apic_read(apic, 0); spin_unlock_irqrestore(&ioapic_lock, flags); - if (reg_00.bits.ID != mp_ioapics[apic].mpc_apicid) + if (reg_00.bits.ID != mp_ioapics[apic].mp_apicid) printk("could not set ID!\n"); else apic_printk(APIC_VERBOSE, " ok.\n"); } } -#else -static void __init setup_ioapic_ids_from_mpc(void) { } -#endif int no_timer_check __initdata; @@ -2015,12 +2020,12 @@ static inline void init_IO_APIC_traps(void) * The local APIC irq-chip implementation: */ -static void ack_apic(unsigned int irq) +static void ack_lapic_irq(unsigned int irq) { ack_APIC_irq(); } -static void mask_lapic_irq (unsigned int irq) +static void mask_lapic_irq(unsigned int irq) { unsigned long v; @@ -2028,7 +2033,7 @@ static void mask_lapic_irq (unsigned int irq) apic_write_around(APIC_LVT0, v | APIC_LVT_MASKED); } -static void unmask_lapic_irq (unsigned int irq) +static void unmask_lapic_irq(unsigned int irq) { unsigned long v; @@ -2037,23 +2042,31 @@ static void unmask_lapic_irq (unsigned int irq) } static struct irq_chip lapic_chip __read_mostly = { - .name = "local-APIC-edge", + .name = "local-APIC", .mask = mask_lapic_irq, .unmask = unmask_lapic_irq, - .eoi = ack_apic, + .ack = ack_lapic_irq, }; +static void lapic_register_intr(int irq, int vector) +{ + irq_desc[irq].status &= ~IRQ_LEVEL; + set_irq_chip_and_handler_name(irq, &lapic_chip, handle_edge_irq, + "edge"); + set_intr_gate(vector, interrupt[irq]); +} + static void __init setup_nmi(void) { /* - * Dirty trick to enable the NMI watchdog ... + * Dirty trick to enable the NMI watchdog ... * We put the 8259A master into AEOI mode and * unmask on all local APICs LVT0 as NMI. * * The idea to use the 8259A in AEOI mode ('8259A Virtual Wire') * is from Maciej W. Rozycki - so we do not have to EOI from * the NMI handler or the timer interrupt. - */ + */ apic_printk(APIC_VERBOSE, KERN_INFO "activating NMI Watchdog ..."); enable_NMI_through_LVT0(); @@ -2129,11 +2142,16 @@ static inline void __init unlock_ExtINT_logic(void) static inline void __init check_timer(void) { int apic1, pin1, apic2, pin2; + int no_pin1 = 0; int vector; + unsigned int ver; unsigned long flags; local_irq_save(flags); + ver = apic_read(APIC_LVR); + ver = GET_APIC_VERSION(ver); + /* * get/set the timer IRQ vector: */ @@ -2142,17 +2160,17 @@ static inline void __init check_timer(void) set_intr_gate(vector, interrupt[0]); /* - * Subtle, code in do_timer_interrupt() expects an AEOI - * mode for the 8259A whenever interrupts are routed - * through I/O APICs. Also IRQ0 has to be enabled in - * the 8259A which implies the virtual wire has to be - * disabled in the local APIC. + * As IRQ0 is to be enabled in the 8259A, the virtual + * wire has to be disabled in the local APIC. Also + * timer interrupts need to be acknowledged manually in + * the 8259A for the i82489DX when using the NMI + * watchdog as that APIC treats NMIs as level-triggered. + * The AEOI mode will finish them in the 8259A + * automatically. */ apic_write_around(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT); init_8259A(1); - timer_ack = 1; - if (timer_over_8254 > 0) - enable_8259A_irq(0); + timer_ack = (nmi_watchdog == NMI_IO_APIC && !APIC_INTEGRATED(ver)); pin1 = find_isa_irq_pin(0, mp_INT); apic1 = find_isa_irq_apic(0, mp_INT); @@ -2162,14 +2180,33 @@ static inline void __init check_timer(void) printk(KERN_INFO "..TIMER: vector=0x%02X apic1=%d pin1=%d apic2=%d pin2=%d\n", vector, apic1, pin1, apic2, pin2); + /* + * Some BIOS writers are clueless and report the ExtINTA + * I/O APIC input from the cascaded 8259A as the timer + * interrupt input. So just in case, if only one pin + * was found above, try it both directly and through the + * 8259A. + */ + if (pin1 == -1) { + pin1 = pin2; + apic1 = apic2; + no_pin1 = 1; + } else if (pin2 == -1) { + pin2 = pin1; + apic2 = apic1; + } + if (pin1 != -1) { /* * Ok, does IRQ0 through the IOAPIC work? */ + if (no_pin1) { + add_pin_to_irq(0, apic1, pin1); + setup_timer_IRQ0_pin(apic1, pin1, vector); + } unmask_IO_APIC_irq(0); if (timer_irq_works()) { if (nmi_watchdog == NMI_IO_APIC) { - disable_8259A_irq(0); setup_nmi(); enable_8259A_irq(0); } @@ -2178,45 +2215,47 @@ static inline void __init check_timer(void) goto out; } clear_IO_APIC_pin(apic1, pin1); - printk(KERN_ERR "..MP-BIOS bug: 8254 timer not connected to " - "IO-APIC\n"); - } + if (!no_pin1) + printk(KERN_ERR "..MP-BIOS bug: " + "8254 timer not connected to IO-APIC\n"); - printk(KERN_INFO "...trying to set up timer (IRQ0) through the 8259A ... "); - if (pin2 != -1) { + printk(KERN_INFO "...trying to set up timer (IRQ0) " + "through the 8259A ... "); printk("\n..... (found pin %d) ...", pin2); /* * legacy devices should be connected to IO APIC #0 */ - setup_ExtINT_IRQ0_pin(apic2, pin2, vector); + replace_pin_at_irq(0, apic1, pin1, apic2, pin2); + setup_timer_IRQ0_pin(apic2, pin2, vector); + unmask_IO_APIC_irq(0); + enable_8259A_irq(0); if (timer_irq_works()) { printk("works.\n"); - if (pin1 != -1) - replace_pin_at_irq(0, apic1, pin1, apic2, pin2); - else - add_pin_to_irq(0, apic2, pin2); + timer_through_8259 = 1; if (nmi_watchdog == NMI_IO_APIC) { + disable_8259A_irq(0); setup_nmi(); + enable_8259A_irq(0); } goto out; } /* * Cleanup, just in case ... */ + disable_8259A_irq(0); clear_IO_APIC_pin(apic2, pin2); + printk(" failed.\n"); } - printk(" failed.\n"); if (nmi_watchdog == NMI_IO_APIC) { printk(KERN_WARNING "timer doesn't work through the IO-APIC - disabling NMI Watchdog!\n"); - nmi_watchdog = 0; + nmi_watchdog = NMI_NONE; } + timer_ack = 0; printk(KERN_INFO "...trying to set up timer as Virtual Wire IRQ..."); - disable_8259A_irq(0); - set_irq_chip_and_handler_name(0, &lapic_chip, handle_fasteoi_irq, - "fasteoi"); + lapic_register_intr(0, vector); apic_write_around(APIC_LVT0, APIC_DM_FIXED | vector); /* Fixed mode */ enable_8259A_irq(0); @@ -2224,12 +2263,12 @@ static inline void __init check_timer(void) printk(" works.\n"); goto out; } + disable_8259A_irq(0); apic_write_around(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | vector); printk(" failed.\n"); printk(KERN_INFO "...trying to set up timer as ExtINT IRQ..."); - timer_ack = 0; init_8259A(0); make_8259A_irq(0); apic_write_around(APIC_LVT0, APIC_DM_EXTINT); @@ -2248,11 +2287,21 @@ out: } /* - * - * IRQ's that are handled by the PIC in the MPS IOAPIC case. - * - IRQ2 is the cascade IRQ, and cannot be a io-apic IRQ. - * Linux doesn't really care, as it's not actually used - * for any interrupt handling anyway. + * Traditionally ISA IRQ2 is the cascade IRQ, and is not available + * to devices. However there may be an I/O APIC pin available for + * this interrupt regardless. The pin may be left unconnected, but + * typically it will be reused as an ExtINT cascade interrupt for + * the master 8259A. In the MPS case such a pin will normally be + * reported as an ExtINT interrupt in the MP table. With ACPI + * there is no provision for ExtINT interrupts, and in the absence + * of an override it would be treated as an ordinary ISA I/O APIC + * interrupt, that is edge-triggered and unmasked by default. We + * used to do this, but it caused problems on some systems because + * of the NMI watchdog and sometimes IRQ0 of the 8254 timer using + * the same ExtINT cascade interrupt to drive the local APIC of the + * bootstrap processor. Therefore we refrain from routing IRQ2 to + * the I/O APIC in all cases now. No actual device should request + * it anyway. --macro */ #define PIC_IRQS (1 << PIC_CASCADE_IR) @@ -2261,15 +2310,12 @@ void __init setup_IO_APIC(void) int i; /* Reserve all the system vectors. */ - for (i = FIRST_SYSTEM_VECTOR; i < NR_VECTORS; i++) + for (i = first_system_vector; i < NR_VECTORS; i++) set_bit(i, used_vectors); enable_IO_APIC(); - if (acpi_ioapic) - io_apic_irqs = ~0; /* all IRQs go through IOAPIC */ - else - io_apic_irqs = ~PIC_IRQS; + io_apic_irqs = ~PIC_IRQS; printk("ENABLING IO-APIC IRQs\n"); @@ -2286,28 +2332,14 @@ void __init setup_IO_APIC(void) print_IO_APIC(); } -static int __init setup_disable_8254_timer(char *s) -{ - timer_over_8254 = -1; - return 1; -} -static int __init setup_enable_8254_timer(char *s) -{ - timer_over_8254 = 2; - return 1; -} - -__setup("disable_8254_timer", setup_disable_8254_timer); -__setup("enable_8254_timer", setup_enable_8254_timer); - /* * Called after all the initialization is done. If we didnt find any * APIC bugs then we can allow the modify fast path */ - + static int __init io_apic_bug_finalize(void) { - if(sis_apic_bug == -1) + if (sis_apic_bug == -1) sis_apic_bug = 0; return 0; } @@ -2318,17 +2350,17 @@ struct sysfs_ioapic_data { struct sys_device dev; struct IO_APIC_route_entry entry[0]; }; -static struct sysfs_ioapic_data * mp_ioapic_data[MAX_IO_APICS]; +static struct sysfs_ioapic_data *mp_ioapic_data[MAX_IO_APICS]; static int ioapic_suspend(struct sys_device *dev, pm_message_t state) { struct IO_APIC_route_entry *entry; struct sysfs_ioapic_data *data; int i; - + data = container_of(dev, struct sysfs_ioapic_data, dev); entry = data->entry; - for (i = 0; i < nr_ioapic_registers[dev->id]; i ++) + for (i = 0; i < nr_ioapic_registers[dev->id]; i++) entry[i] = ioapic_read_entry(dev->id, i); return 0; @@ -2341,18 +2373,18 @@ static int ioapic_resume(struct sys_device *dev) unsigned long flags; union IO_APIC_reg_00 reg_00; int i; - + data = container_of(dev, struct sysfs_ioapic_data, dev); entry = data->entry; spin_lock_irqsave(&ioapic_lock, flags); reg_00.raw = io_apic_read(dev->id, 0); - if (reg_00.bits.ID != mp_ioapics[dev->id].mpc_apicid) { - reg_00.bits.ID = mp_ioapics[dev->id].mpc_apicid; + if (reg_00.bits.ID != mp_ioapics[dev->id].mp_apicid) { + reg_00.bits.ID = mp_ioapics[dev->id].mp_apicid; io_apic_write(dev->id, 0, reg_00.raw); } spin_unlock_irqrestore(&ioapic_lock, flags); - for (i = 0; i < nr_ioapic_registers[dev->id]; i ++) + for (i = 0; i < nr_ioapic_registers[dev->id]; i++) ioapic_write_entry(dev->id, i, entry[i]); return 0; @@ -2366,24 +2398,23 @@ static struct sysdev_class ioapic_sysdev_class = { static int __init ioapic_init_sysfs(void) { - struct sys_device * dev; + struct sys_device *dev; int i, size, error = 0; error = sysdev_class_register(&ioapic_sysdev_class); if (error) return error; - for (i = 0; i < nr_ioapics; i++ ) { - size = sizeof(struct sys_device) + nr_ioapic_registers[i] + for (i = 0; i < nr_ioapics; i++) { + size = sizeof(struct sys_device) + nr_ioapic_registers[i] * sizeof(struct IO_APIC_route_entry); - mp_ioapic_data[i] = kmalloc(size, GFP_KERNEL); + mp_ioapic_data[i] = kzalloc(size, GFP_KERNEL); if (!mp_ioapic_data[i]) { printk(KERN_ERR "Can't suspend/resume IOAPIC %d\n", i); continue; } - memset(mp_ioapic_data[i], 0, size); dev = &mp_ioapic_data[i]->dev; - dev->id = i; + dev->id = i; dev->cls = &ioapic_sysdev_class; error = sysdev_register(dev); if (error) { @@ -2458,7 +2489,7 @@ static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_ms msg->address_lo = MSI_ADDR_BASE_LO | ((INT_DEST_MODE == 0) ? - MSI_ADDR_DEST_MODE_PHYSICAL: +MSI_ADDR_DEST_MODE_PHYSICAL: MSI_ADDR_DEST_MODE_LOGICAL) | ((INT_DELIVERY_MODE != dest_LowestPrio) ? MSI_ADDR_REDIRECTION_CPU: @@ -2469,7 +2500,7 @@ static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_ms MSI_DATA_TRIGGER_EDGE | MSI_DATA_LEVEL_ASSERT | ((INT_DELIVERY_MODE != dest_LowestPrio) ? - MSI_DATA_DELIVERY_FIXED: +MSI_DATA_DELIVERY_FIXED: MSI_DATA_DELIVERY_LOWPRI) | MSI_DATA_VECTOR(vector); } @@ -2640,12 +2671,12 @@ int arch_setup_ht_irq(unsigned int irq, struct pci_dev *dev) #endif /* CONFIG_HT_IRQ */ /* -------------------------------------------------------------------------- - ACPI-based IOAPIC Configuration + ACPI-based IOAPIC Configuration -------------------------------------------------------------------------- */ #ifdef CONFIG_ACPI -int __init io_apic_get_unique_id (int ioapic, int apic_id) +int __init io_apic_get_unique_id(int ioapic, int apic_id) { union IO_APIC_reg_00 reg_00; static physid_mask_t apic_id_map = PHYSID_MASK_NONE; @@ -2654,10 +2685,10 @@ int __init io_apic_get_unique_id (int ioapic, int apic_id) int i = 0; /* - * The P4 platform supports up to 256 APIC IDs on two separate APIC - * buses (one for LAPICs, one for IOAPICs), where predecessors only + * The P4 platform supports up to 256 APIC IDs on two separate APIC + * buses (one for LAPICs, one for IOAPICs), where predecessors only * supports up to 16 on one shared APIC bus. - * + * * TBD: Expand LAPIC/IOAPIC support on P4-class systems to take full * advantage of new APIC bus architecture. */ @@ -2676,7 +2707,7 @@ int __init io_apic_get_unique_id (int ioapic, int apic_id) } /* - * Every APIC in a system must have a unique ID or we get lots of nice + * Every APIC in a system must have a unique ID or we get lots of nice * 'stuck on smp_invalidate_needed IPI wait' messages. */ if (check_apicid_used(apic_id_map, apic_id)) { @@ -2693,7 +2724,7 @@ int __init io_apic_get_unique_id (int ioapic, int apic_id) "trying %d\n", ioapic, apic_id, i); apic_id = i; - } + } tmp = apicid_to_cpu_present(apic_id); physids_or(apic_id_map, apic_id_map, tmp); @@ -2720,7 +2751,7 @@ int __init io_apic_get_unique_id (int ioapic, int apic_id) } -int __init io_apic_get_version (int ioapic) +int __init io_apic_get_version(int ioapic) { union IO_APIC_reg_01 reg_01; unsigned long flags; @@ -2733,7 +2764,7 @@ int __init io_apic_get_version (int ioapic) } -int __init io_apic_get_redir_entries (int ioapic) +int __init io_apic_get_redir_entries(int ioapic) { union IO_APIC_reg_01 reg_01; unsigned long flags; @@ -2746,7 +2777,7 @@ int __init io_apic_get_redir_entries (int ioapic) } -int io_apic_set_pci_routing (int ioapic, int pin, int irq, int edge_level, int active_high_low) +int io_apic_set_pci_routing(int ioapic, int pin, int irq, int edge_level, int active_high_low) { struct IO_APIC_route_entry entry; @@ -2762,7 +2793,7 @@ int io_apic_set_pci_routing (int ioapic, int pin, int irq, int edge_level, int a * corresponding device driver registers for this IRQ. */ - memset(&entry,0,sizeof(entry)); + memset(&entry, 0, sizeof(entry)); entry.delivery_mode = INT_DELIVERY_MODE; entry.dest_mode = INT_DEST_MODE; @@ -2781,7 +2812,7 @@ int io_apic_set_pci_routing (int ioapic, int pin, int irq, int edge_level, int a apic_printk(APIC_DEBUG, KERN_DEBUG "IOAPIC[%d]: Set PCI routing entry " "(%d-%d -> 0x%x -> IRQ %d Mode:%i Active:%i)\n", ioapic, - mp_ioapics[ioapic].mpc_apicid, pin, entry.vector, irq, + mp_ioapics[ioapic].mp_apicid, pin, entry.vector, irq, edge_level, active_high_low); ioapic_register_intr(irq, entry.vector, edge_level); @@ -2802,8 +2833,8 @@ int acpi_get_override_irq(int bus_irq, int *trigger, int *polarity) return -1; for (i = 0; i < mp_irq_entries; i++) - if (mp_irqs[i].mpc_irqtype == mp_INT && - mp_irqs[i].mpc_srcbusirq == bus_irq) + if (mp_irqs[i].mp_irqtype == mp_INT && + mp_irqs[i].mp_srcbusirq == bus_irq) break; if (i >= mp_irq_entries) return -1; @@ -2836,3 +2867,34 @@ static int __init parse_noapic(char *arg) return 0; } early_param("noapic", parse_noapic); + +void __init ioapic_init_mappings(void) +{ + unsigned long ioapic_phys, idx = FIX_IO_APIC_BASE_0; + int i; + + for (i = 0; i < nr_ioapics; i++) { + if (smp_found_config) { + ioapic_phys = mp_ioapics[i].mp_apicaddr; + if (!ioapic_phys) { + printk(KERN_ERR + "WARNING: bogus zero IO-APIC " + "address found in MPTABLE, " + "disabling IO/APIC support!\n"); + smp_found_config = 0; + skip_ioapic_setup = 1; + goto fake_ioapic_page; + } + } else { +fake_ioapic_page: + ioapic_phys = (unsigned long) + alloc_bootmem_pages(PAGE_SIZE); + ioapic_phys = __pa(ioapic_phys); + } + set_fixmap_nocache(idx, ioapic_phys); + printk(KERN_DEBUG "mapped IOAPIC to %08lx (%08lx)\n", + __fix_to_virt(idx), ioapic_phys); + idx++; + } +} + diff --git a/arch/x86/kernel/io_apic_64.c b/arch/x86/kernel/io_apic_64.c index ef1a8dfcc52..b16ef029cf8 100644 --- a/arch/x86/kernel/io_apic_64.c +++ b/arch/x86/kernel/io_apic_64.c @@ -61,7 +61,7 @@ struct irq_cfg { }; /* irq_cfg is indexed by the sum of all RTEs in all I/O APICs. */ -struct irq_cfg irq_cfg[NR_IRQS] __read_mostly = { +static struct irq_cfg irq_cfg[NR_IRQS] __read_mostly = { [0] = { .domain = CPU_MASK_ALL, .vector = IRQ0_VECTOR, }, [1] = { .domain = CPU_MASK_ALL, .vector = IRQ1_VECTOR, }, [2] = { .domain = CPU_MASK_ALL, .vector = IRQ2_VECTOR, }, @@ -82,6 +82,10 @@ struct irq_cfg irq_cfg[NR_IRQS] __read_mostly = { static int assign_irq_vector(int irq, cpumask_t mask); +int first_system_vector = 0xfe; + +char system_vectors[NR_VECTORS] = { [0 ... NR_VECTORS-1] = SYS_VECTOR_FREE}; + #define __apicdebuginit __init int sis_apic_bug; /* not actually supported, dummy for compile */ @@ -90,7 +94,7 @@ static int no_timer_check; static int disable_timer_pin_1 __initdata; -int timer_over_8254 __initdata = 1; +int timer_through_8259 __initdata; /* Where if anywhere is the i8259 connect in external int mode */ static struct { int pin, apic; } ioapic_i8259 = { -1, -1 }; @@ -104,15 +108,17 @@ DEFINE_SPINLOCK(vector_lock); int nr_ioapic_registers[MAX_IO_APICS]; /* I/O APIC entries */ -struct mpc_config_ioapic mp_ioapics[MAX_IO_APICS]; +struct mp_config_ioapic mp_ioapics[MAX_IO_APICS]; int nr_ioapics; /* MP IRQ source entries */ -struct mpc_config_intsrc mp_irqs[MAX_IRQ_SOURCES]; +struct mp_config_intsrc mp_irqs[MAX_IRQ_SOURCES]; /* # of MP IRQ source entries */ int mp_irq_entries; +DECLARE_BITMAP(mp_bus_not_pci, MAX_MP_BUSSES); + /* * Rough estimation of how many shared IRQs there are, can * be changed anytime. @@ -140,7 +146,7 @@ struct io_apic { static __attribute_const__ struct io_apic __iomem *io_apic_base(int idx) { return (void __iomem *) __fix_to_virt(FIX_IO_APIC_BASE_0 + idx) - + (mp_ioapics[idx].mpc_apicaddr & ~PAGE_MASK); + + (mp_ioapics[idx].mp_apicaddr & ~PAGE_MASK); } static inline unsigned int io_apic_read(unsigned int apic, unsigned int reg) @@ -183,7 +189,7 @@ static bool io_apic_level_ack_pending(unsigned int irq) break; reg = io_apic_read(entry->apic, 0x10 + pin*2); /* Is the remote IRR bit set? */ - if ((reg >> 14) & 1) { + if (reg & IO_APIC_REDIR_REMOTE_IRR) { spin_unlock_irqrestore(&ioapic_lock, flags); return true; } @@ -298,7 +304,7 @@ static void __target_IO_APIC_irq(unsigned int irq, unsigned int dest, u8 vector) break; io_apic_write(apic, 0x11 + pin*2, dest); reg = io_apic_read(apic, 0x10 + pin*2); - reg &= ~0x000000ff; + reg &= ~IO_APIC_REDIR_VECTOR_MASK; reg |= vector; io_apic_modify(apic, reg); if (!entry->next) @@ -360,16 +366,37 @@ static void add_pin_to_irq(unsigned int irq, int apic, int pin) entry->pin = pin; } +/* + * Reroute an IRQ to a different pin. + */ +static void __init replace_pin_at_irq(unsigned int irq, + int oldapic, int oldpin, + int newapic, int newpin) +{ + struct irq_pin_list *entry = irq_2_pin + irq; + + while (1) { + if (entry->apic == oldapic && entry->pin == oldpin) { + entry->apic = newapic; + entry->pin = newpin; + } + if (!entry->next) + break; + entry = irq_2_pin + entry->next; + } +} + #define DO_ACTION(name,R,ACTION, FINAL) \ \ static void name##_IO_APIC_irq (unsigned int irq) \ __DO_ACTION(R, ACTION, FINAL) -DO_ACTION( __mask, 0, |= 0x00010000, io_apic_sync(entry->apic) ) - /* mask = 1 */ -DO_ACTION( __unmask, 0, &= 0xfffeffff, ) - /* mask = 0 */ +/* mask = 1 */ +DO_ACTION(__mask, 0, |= IO_APIC_REDIR_MASKED, io_apic_sync(entry->apic)) + +/* mask = 0 */ +DO_ACTION(__unmask, 0, &= ~IO_APIC_REDIR_MASKED, ) static void mask_IO_APIC_irq (unsigned int irq) { @@ -430,20 +457,6 @@ static int __init disable_timer_pin_setup(char *arg) } __setup("disable_timer_pin_1", disable_timer_pin_setup); -static int __init setup_disable_8254_timer(char *s) -{ - timer_over_8254 = -1; - return 1; -} -static int __init setup_enable_8254_timer(char *s) -{ - timer_over_8254 = 2; - return 1; -} - -__setup("disable_8254_timer", setup_disable_8254_timer); -__setup("enable_8254_timer", setup_enable_8254_timer); - /* * Find the IRQ entry number of a certain pin. @@ -453,10 +466,10 @@ static int find_irq_entry(int apic, int pin, int type) int i; for (i = 0; i < mp_irq_entries; i++) - if (mp_irqs[i].mpc_irqtype == type && - (mp_irqs[i].mpc_dstapic == mp_ioapics[apic].mpc_apicid || - mp_irqs[i].mpc_dstapic == MP_APIC_ALL) && - mp_irqs[i].mpc_dstirq == pin) + if (mp_irqs[i].mp_irqtype == type && + (mp_irqs[i].mp_dstapic == mp_ioapics[apic].mp_apicid || + mp_irqs[i].mp_dstapic == MP_APIC_ALL) && + mp_irqs[i].mp_dstirq == pin) return i; return -1; @@ -470,13 +483,13 @@ static int __init find_isa_irq_pin(int irq, int type) int i; for (i = 0; i < mp_irq_entries; i++) { - int lbus = mp_irqs[i].mpc_srcbus; + int lbus = mp_irqs[i].mp_srcbus; if (test_bit(lbus, mp_bus_not_pci) && - (mp_irqs[i].mpc_irqtype == type) && - (mp_irqs[i].mpc_srcbusirq == irq)) + (mp_irqs[i].mp_irqtype == type) && + (mp_irqs[i].mp_srcbusirq == irq)) - return mp_irqs[i].mpc_dstirq; + return mp_irqs[i].mp_dstirq; } return -1; } @@ -486,17 +499,17 @@ static int __init find_isa_irq_apic(int irq, int type) int i; for (i = 0; i < mp_irq_entries; i++) { - int lbus = mp_irqs[i].mpc_srcbus; + int lbus = mp_irqs[i].mp_srcbus; if (test_bit(lbus, mp_bus_not_pci) && - (mp_irqs[i].mpc_irqtype == type) && - (mp_irqs[i].mpc_srcbusirq == irq)) + (mp_irqs[i].mp_irqtype == type) && + (mp_irqs[i].mp_srcbusirq == irq)) break; } if (i < mp_irq_entries) { int apic; for(apic = 0; apic < nr_ioapics; apic++) { - if (mp_ioapics[apic].mpc_apicid == mp_irqs[i].mpc_dstapic) + if (mp_ioapics[apic].mp_apicid == mp_irqs[i].mp_dstapic) return apic; } } @@ -516,28 +529,28 @@ int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin) apic_printk(APIC_DEBUG, "querying PCI -> IRQ mapping bus:%d, slot:%d, pin:%d.\n", bus, slot, pin); - if (mp_bus_id_to_pci_bus[bus] == -1) { + if (test_bit(bus, mp_bus_not_pci)) { apic_printk(APIC_VERBOSE, "PCI BIOS passed nonexistent PCI bus %d!\n", bus); return -1; } for (i = 0; i < mp_irq_entries; i++) { - int lbus = mp_irqs[i].mpc_srcbus; + int lbus = mp_irqs[i].mp_srcbus; for (apic = 0; apic < nr_ioapics; apic++) - if (mp_ioapics[apic].mpc_apicid == mp_irqs[i].mpc_dstapic || - mp_irqs[i].mpc_dstapic == MP_APIC_ALL) + if (mp_ioapics[apic].mp_apicid == mp_irqs[i].mp_dstapic || + mp_irqs[i].mp_dstapic == MP_APIC_ALL) break; if (!test_bit(lbus, mp_bus_not_pci) && - !mp_irqs[i].mpc_irqtype && + !mp_irqs[i].mp_irqtype && (bus == lbus) && - (slot == ((mp_irqs[i].mpc_srcbusirq >> 2) & 0x1f))) { - int irq = pin_2_irq(i,apic,mp_irqs[i].mpc_dstirq); + (slot == ((mp_irqs[i].mp_srcbusirq >> 2) & 0x1f))) { + int irq = pin_2_irq(i,apic,mp_irqs[i].mp_dstirq); if (!(apic || IO_APIC_IRQ(irq))) continue; - if (pin == (mp_irqs[i].mpc_srcbusirq & 3)) + if (pin == (mp_irqs[i].mp_srcbusirq & 3)) return irq; /* * Use the first all-but-pin matching entry as a @@ -565,13 +578,13 @@ int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin) static int MPBIOS_polarity(int idx) { - int bus = mp_irqs[idx].mpc_srcbus; + int bus = mp_irqs[idx].mp_srcbus; int polarity; /* * Determine IRQ line polarity (high active or low active): */ - switch (mp_irqs[idx].mpc_irqflag & 3) + switch (mp_irqs[idx].mp_irqflag & 3) { case 0: /* conforms, ie. bus-type dependent polarity */ if (test_bit(bus, mp_bus_not_pci)) @@ -607,13 +620,13 @@ static int MPBIOS_polarity(int idx) static int MPBIOS_trigger(int idx) { - int bus = mp_irqs[idx].mpc_srcbus; + int bus = mp_irqs[idx].mp_srcbus; int trigger; /* * Determine IRQ trigger mode (edge or level sensitive): */ - switch ((mp_irqs[idx].mpc_irqflag>>2) & 3) + switch ((mp_irqs[idx].mp_irqflag>>2) & 3) { case 0: /* conforms, ie. bus-type dependent */ if (test_bit(bus, mp_bus_not_pci)) @@ -660,16 +673,16 @@ static inline int irq_trigger(int idx) static int pin_2_irq(int idx, int apic, int pin) { int irq, i; - int bus = mp_irqs[idx].mpc_srcbus; + int bus = mp_irqs[idx].mp_srcbus; /* * Debugging check, we are in big trouble if this message pops up! */ - if (mp_irqs[idx].mpc_dstirq != pin) + if (mp_irqs[idx].mp_dstirq != pin) printk(KERN_ERR "broken BIOS or MPTABLE parser, ayiee!!\n"); if (test_bit(bus, mp_bus_not_pci)) { - irq = mp_irqs[idx].mpc_srcbusirq; + irq = mp_irqs[idx].mp_srcbusirq; } else { /* * PCI IRQs are mapped in order @@ -730,7 +743,7 @@ static int __assign_irq_vector(int irq, cpumask_t mask) offset = current_offset; next: vector += 8; - if (vector >= FIRST_SYSTEM_VECTOR) { + if (vector >= first_system_vector) { /* If we run out of vectors on large boxen, must share them. */ offset = (offset + 1) % 8; vector = FIRST_DEVICE_VECTOR + offset; @@ -788,7 +801,7 @@ static void __clear_irq_vector(int irq) cpus_clear(cfg->domain); } -void __setup_vector_irq(int cpu) +static void __setup_vector_irq(int cpu) { /* Initialize vector_irq on a new cpu */ /* This function must be called with vector_lock held */ @@ -811,6 +824,13 @@ void __setup_vector_irq(int cpu) } } +void setup_vector_irq(int cpu) +{ + spin_lock(&vector_lock); + __setup_vector_irq(smp_processor_id()); + spin_unlock(&vector_lock); +} + static struct irq_chip ioapic_chip; @@ -846,7 +866,7 @@ static void setup_IO_APIC_irq(int apic, int pin, unsigned int irq, apic_printk(APIC_VERBOSE,KERN_DEBUG "IOAPIC[%d]: Set routing entry (%d-%d -> 0x%x -> " "IRQ %d Mode:%i Active:%i)\n", - apic, mp_ioapics[apic].mpc_apicid, pin, cfg->vector, + apic, mp_ioapics[apic].mp_apicid, pin, cfg->vector, irq, trigger, polarity); /* @@ -887,10 +907,10 @@ static void __init setup_IO_APIC_irqs(void) idx = find_irq_entry(apic,pin,mp_INT); if (idx == -1) { if (first_notcon) { - apic_printk(APIC_VERBOSE, KERN_DEBUG " IO-APIC (apicid-pin) %d-%d", mp_ioapics[apic].mpc_apicid, pin); + apic_printk(APIC_VERBOSE, KERN_DEBUG " IO-APIC (apicid-pin) %d-%d", mp_ioapics[apic].mp_apicid, pin); first_notcon = 0; } else - apic_printk(APIC_VERBOSE, ", %d-%d", mp_ioapics[apic].mpc_apicid, pin); + apic_printk(APIC_VERBOSE, ", %d-%d", mp_ioapics[apic].mp_apicid, pin); continue; } if (!first_notcon) { @@ -911,26 +931,21 @@ static void __init setup_IO_APIC_irqs(void) } /* - * Set up the 8259A-master output pin as broadcast to all - * CPUs. + * Set up the timer pin, possibly with the 8259A-master behind. */ -static void __init setup_ExtINT_IRQ0_pin(unsigned int apic, unsigned int pin, int vector) +static void __init setup_timer_IRQ0_pin(unsigned int apic, unsigned int pin, + int vector) { struct IO_APIC_route_entry entry; memset(&entry, 0, sizeof(entry)); - disable_8259A_irq(0); - - /* mask LVT0 */ - apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT); - /* * We use logical delivery to get the timer IRQ * to the first CPU. */ entry.dest_mode = INT_DEST_MODE; - entry.mask = 0; /* unmask IRQ now */ + entry.mask = 1; /* mask IRQ now */ entry.dest = cpu_mask_to_apicid(TARGET_CPUS); entry.delivery_mode = INT_DELIVERY_MODE; entry.polarity = 0; @@ -939,7 +954,7 @@ static void __init setup_ExtINT_IRQ0_pin(unsigned int apic, unsigned int pin, in /* * The timer IRQ doesn't have to know that behind the - * scene we have a 8259A-master in AEOI mode ... + * scene we may have a 8259A-master in AEOI mode ... */ set_irq_chip_and_handler_name(0, &ioapic_chip, handle_edge_irq, "edge"); @@ -947,8 +962,6 @@ static void __init setup_ExtINT_IRQ0_pin(unsigned int apic, unsigned int pin, in * Add it to the IO-APIC irq-routing table: */ ioapic_write_entry(apic, pin, entry); - - enable_8259A_irq(0); } void __apicdebuginit print_IO_APIC(void) @@ -965,7 +978,7 @@ void __apicdebuginit print_IO_APIC(void) printk(KERN_DEBUG "number of MP IRQ sources: %d.\n", mp_irq_entries); for (i = 0; i < nr_ioapics; i++) printk(KERN_DEBUG "number of IO-APIC #%d registers: %d.\n", - mp_ioapics[i].mpc_apicid, nr_ioapic_registers[i]); + mp_ioapics[i].mp_apicid, nr_ioapic_registers[i]); /* * We are a bit conservative about what we expect. We have to @@ -983,7 +996,7 @@ void __apicdebuginit print_IO_APIC(void) spin_unlock_irqrestore(&ioapic_lock, flags); printk("\n"); - printk(KERN_DEBUG "IO APIC #%d......\n", mp_ioapics[apic].mpc_apicid); + printk(KERN_DEBUG "IO APIC #%d......\n", mp_ioapics[apic].mp_apicid); printk(KERN_DEBUG ".... register #00: %08X\n", reg_00.raw); printk(KERN_DEBUG "....... : physical APIC id: %02X\n", reg_00.bits.ID); @@ -1077,6 +1090,7 @@ void __apicdebuginit print_local_APIC(void * dummy) printk("\n" KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n", smp_processor_id(), hard_smp_processor_id()); + v = apic_read(APIC_ID); printk(KERN_INFO "... APIC ID: %08x (%01x)\n", v, GET_APIC_ID(read_apic_id())); v = apic_read(APIC_LVR); printk(KERN_INFO "... APIC VERSION: %08x\n", v); @@ -1540,7 +1554,7 @@ static inline void init_IO_APIC_traps(void) } } -static void enable_lapic_irq (unsigned int irq) +static void unmask_lapic_irq(unsigned int irq) { unsigned long v; @@ -1548,7 +1562,7 @@ static void enable_lapic_irq (unsigned int irq) apic_write(APIC_LVT0, v & ~APIC_LVT_MASKED); } -static void disable_lapic_irq (unsigned int irq) +static void mask_lapic_irq(unsigned int irq) { unsigned long v; @@ -1561,19 +1575,20 @@ static void ack_lapic_irq (unsigned int irq) ack_APIC_irq(); } -static void end_lapic_irq (unsigned int i) { /* nothing */ } - -static struct hw_interrupt_type lapic_irq_type __read_mostly = { - .name = "local-APIC", - .typename = "local-APIC-edge", - .startup = NULL, /* startup_irq() not used for IRQ0 */ - .shutdown = NULL, /* shutdown_irq() not used for IRQ0 */ - .enable = enable_lapic_irq, - .disable = disable_lapic_irq, - .ack = ack_lapic_irq, - .end = end_lapic_irq, +static struct irq_chip lapic_chip __read_mostly = { + .name = "local-APIC", + .mask = mask_lapic_irq, + .unmask = unmask_lapic_irq, + .ack = ack_lapic_irq, }; +static void lapic_register_intr(int irq) +{ + irq_desc[irq].status &= ~IRQ_LEVEL; + set_irq_chip_and_handler_name(irq, &lapic_chip, handle_edge_irq, + "edge"); +} + static void __init setup_nmi(void) { /* @@ -1659,6 +1674,7 @@ static inline void __init check_timer(void) struct irq_cfg *cfg = irq_cfg + 0; int apic1, pin1, apic2, pin2; unsigned long flags; + int no_pin1 = 0; local_irq_save(flags); @@ -1669,16 +1685,11 @@ static inline void __init check_timer(void) assign_irq_vector(0, TARGET_CPUS); /* - * Subtle, code in do_timer_interrupt() expects an AEOI - * mode for the 8259A whenever interrupts are routed - * through I/O APICs. Also IRQ0 has to be enabled in - * the 8259A which implies the virtual wire has to be - * disabled in the local APIC. + * As IRQ0 is to be enabled in the 8259A, the virtual + * wire has to be disabled in the local APIC. */ apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT); init_8259A(1); - if (timer_over_8254 > 0) - enable_8259A_irq(0); pin1 = find_isa_irq_pin(0, mp_INT); apic1 = find_isa_irq_apic(0, mp_INT); @@ -1688,15 +1699,33 @@ static inline void __init check_timer(void) apic_printk(APIC_VERBOSE,KERN_INFO "..TIMER: vector=0x%02X apic1=%d pin1=%d apic2=%d pin2=%d\n", cfg->vector, apic1, pin1, apic2, pin2); + /* + * Some BIOS writers are clueless and report the ExtINTA + * I/O APIC input from the cascaded 8259A as the timer + * interrupt input. So just in case, if only one pin + * was found above, try it both directly and through the + * 8259A. + */ + if (pin1 == -1) { + pin1 = pin2; + apic1 = apic2; + no_pin1 = 1; + } else if (pin2 == -1) { + pin2 = pin1; + apic2 = apic1; + } + if (pin1 != -1) { /* * Ok, does IRQ0 through the IOAPIC work? */ + if (no_pin1) { + add_pin_to_irq(0, apic1, pin1); + setup_timer_IRQ0_pin(apic1, pin1, cfg->vector); + } unmask_IO_APIC_irq(0); if (!no_timer_check && timer_irq_works()) { - nmi_watchdog_default(); if (nmi_watchdog == NMI_IO_APIC) { - disable_8259A_irq(0); setup_nmi(); enable_8259A_irq(0); } @@ -1705,43 +1734,48 @@ static inline void __init check_timer(void) goto out; } clear_IO_APIC_pin(apic1, pin1); - apic_printk(APIC_QUIET,KERN_ERR "..MP-BIOS bug: 8254 timer not " - "connected to IO-APIC\n"); - } + if (!no_pin1) + apic_printk(APIC_QUIET,KERN_ERR "..MP-BIOS bug: " + "8254 timer not connected to IO-APIC\n"); - apic_printk(APIC_VERBOSE,KERN_INFO "...trying to set up timer (IRQ0) " - "through the 8259A ... "); - if (pin2 != -1) { + apic_printk(APIC_VERBOSE,KERN_INFO + "...trying to set up timer (IRQ0) " + "through the 8259A ... "); apic_printk(APIC_VERBOSE,"\n..... (found apic %d pin %d) ...", apic2, pin2); /* * legacy devices should be connected to IO APIC #0 */ - setup_ExtINT_IRQ0_pin(apic2, pin2, cfg->vector); + replace_pin_at_irq(0, apic1, pin1, apic2, pin2); + setup_timer_IRQ0_pin(apic2, pin2, cfg->vector); + unmask_IO_APIC_irq(0); + enable_8259A_irq(0); if (timer_irq_works()) { apic_printk(APIC_VERBOSE," works.\n"); - nmi_watchdog_default(); + timer_through_8259 = 1; if (nmi_watchdog == NMI_IO_APIC) { + disable_8259A_irq(0); setup_nmi(); + enable_8259A_irq(0); } goto out; } /* * Cleanup, just in case ... */ + disable_8259A_irq(0); clear_IO_APIC_pin(apic2, pin2); + apic_printk(APIC_VERBOSE," failed.\n"); } - apic_printk(APIC_VERBOSE," failed.\n"); if (nmi_watchdog == NMI_IO_APIC) { printk(KERN_WARNING "timer doesn't work through the IO-APIC - disabling NMI Watchdog!\n"); - nmi_watchdog = 0; + nmi_watchdog = NMI_NONE; } apic_printk(APIC_VERBOSE, KERN_INFO "...trying to set up timer as Virtual Wire IRQ..."); - disable_8259A_irq(0); - irq_desc[0].chip = &lapic_irq_type; + lapic_register_intr(0); apic_write(APIC_LVT0, APIC_DM_FIXED | cfg->vector); /* Fixed mode */ enable_8259A_irq(0); @@ -1749,6 +1783,7 @@ static inline void __init check_timer(void) apic_printk(APIC_VERBOSE," works.\n"); goto out; } + disable_8259A_irq(0); apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | cfg->vector); apic_printk(APIC_VERBOSE," failed.\n"); @@ -1778,11 +1813,21 @@ static int __init notimercheck(char *s) __setup("no_timer_check", notimercheck); /* - * - * IRQs that are handled by the PIC in the MPS IOAPIC case. - * - IRQ2 is the cascade IRQ, and cannot be a io-apic IRQ. - * Linux doesn't really care, as it's not actually used - * for any interrupt handling anyway. + * Traditionally ISA IRQ2 is the cascade IRQ, and is not available + * to devices. However there may be an I/O APIC pin available for + * this interrupt regardless. The pin may be left unconnected, but + * typically it will be reused as an ExtINT cascade interrupt for + * the master 8259A. In the MPS case such a pin will normally be + * reported as an ExtINT interrupt in the MP table. With ACPI + * there is no provision for ExtINT interrupts, and in the absence + * of an override it would be treated as an ordinary ISA I/O APIC + * interrupt, that is edge-triggered and unmasked by default. We + * used to do this, but it caused problems on some systems because + * of the NMI watchdog and sometimes IRQ0 of the 8254 timer using + * the same ExtINT cascade interrupt to drive the local APIC of the + * bootstrap processor. Therefore we refrain from routing IRQ2 to + * the I/O APIC in all cases now. No actual device should request + * it anyway. --macro */ #define PIC_IRQS (1<<2) @@ -1793,10 +1838,7 @@ void __init setup_IO_APIC(void) * calling enable_IO_APIC() is moved to setup_local_APIC for BP */ - if (acpi_ioapic) - io_apic_irqs = ~0; /* all IRQs go through IOAPIC */ - else - io_apic_irqs = ~PIC_IRQS; + io_apic_irqs = ~PIC_IRQS; apic_printk(APIC_VERBOSE, "ENABLING IO-APIC IRQs\n"); @@ -1841,8 +1883,8 @@ static int ioapic_resume(struct sys_device *dev) spin_lock_irqsave(&ioapic_lock, flags); reg_00.raw = io_apic_read(dev->id, 0); - if (reg_00.bits.ID != mp_ioapics[dev->id].mpc_apicid) { - reg_00.bits.ID = mp_ioapics[dev->id].mpc_apicid; + if (reg_00.bits.ID != mp_ioapics[dev->id].mp_apicid) { + reg_00.bits.ID = mp_ioapics[dev->id].mp_apicid; io_apic_write(dev->id, 0, reg_00.raw); } spin_unlock_irqrestore(&ioapic_lock, flags); @@ -2242,8 +2284,8 @@ int acpi_get_override_irq(int bus_irq, int *trigger, int *polarity) return -1; for (i = 0; i < mp_irq_entries; i++) - if (mp_irqs[i].mpc_irqtype == mp_INT && - mp_irqs[i].mpc_srcbusirq == bus_irq) + if (mp_irqs[i].mp_irqtype == mp_INT && + mp_irqs[i].mp_srcbusirq == bus_irq) break; if (i >= mp_irq_entries) return -1; @@ -2336,7 +2378,7 @@ void __init ioapic_init_mappings(void) ioapic_res = ioapic_setup_resources(); for (i = 0; i < nr_ioapics; i++) { if (smp_found_config) { - ioapic_phys = mp_ioapics[i].mpc_apicaddr; + ioapic_phys = mp_ioapics[i].mp_apicaddr; } else { ioapic_phys = (unsigned long) alloc_bootmem_pages(PAGE_SIZE); diff --git a/arch/x86/kernel/ipi.c b/arch/x86/kernel/ipi.c index c0df7b89ca2..9d98cda39ad 100644 --- a/arch/x86/kernel/ipi.c +++ b/arch/x86/kernel/ipi.c @@ -8,7 +8,6 @@ #include <linux/kernel_stat.h> #include <linux/mc146818rtc.h> #include <linux/cache.h> -#include <linux/interrupt.h> #include <linux/cpu.h> #include <linux/module.h> diff --git a/arch/x86/kernel/irq_32.c b/arch/x86/kernel/irq_32.c index 147352df28b..47a6f6f1247 100644 --- a/arch/x86/kernel/irq_32.c +++ b/arch/x86/kernel/irq_32.c @@ -48,6 +48,29 @@ void ack_bad_irq(unsigned int irq) #endif } +#ifdef CONFIG_DEBUG_STACKOVERFLOW +/* Debugging check for stack overflow: is there less than 1KB free? */ +static int check_stack_overflow(void) +{ + long sp; + + __asm__ __volatile__("andl %%esp,%0" : + "=r" (sp) : "0" (THREAD_SIZE - 1)); + + return sp < (sizeof(struct thread_info) + STACK_WARN); +} + +static void print_stack_overflow(void) +{ + printk(KERN_WARNING "low stack detected by irq handler\n"); + dump_stack(); +} + +#else +static inline int check_stack_overflow(void) { return 0; } +static inline void print_stack_overflow(void) { } +#endif + #ifdef CONFIG_4KSTACKS /* * per-CPU IRQ handling contexts (thread information and stack) @@ -59,48 +82,29 @@ union irq_ctx { static union irq_ctx *hardirq_ctx[NR_CPUS] __read_mostly; static union irq_ctx *softirq_ctx[NR_CPUS] __read_mostly; -#endif -/* - * do_IRQ handles all normal device IRQ's (the special - * SMP cross-CPU interrupts have their own specific - * handlers). - */ -unsigned int do_IRQ(struct pt_regs *regs) -{ - struct pt_regs *old_regs; - /* high bit used in ret_from_ code */ - int irq = ~regs->orig_ax; - struct irq_desc *desc = irq_desc + irq; -#ifdef CONFIG_4KSTACKS - union irq_ctx *curctx, *irqctx; - u32 *isp; -#endif +static char softirq_stack[NR_CPUS * THREAD_SIZE] + __attribute__((__section__(".bss.page_aligned"))); - if (unlikely((unsigned)irq >= NR_IRQS)) { - printk(KERN_EMERG "%s: cannot handle IRQ %d\n", - __func__, irq); - BUG(); - } +static char hardirq_stack[NR_CPUS * THREAD_SIZE] + __attribute__((__section__(".bss.page_aligned"))); - old_regs = set_irq_regs(regs); - irq_enter(); -#ifdef CONFIG_DEBUG_STACKOVERFLOW - /* Debugging check for stack overflow: is there less than 1KB free? */ - { - long sp; - - __asm__ __volatile__("andl %%esp,%0" : - "=r" (sp) : "0" (THREAD_SIZE - 1)); - if (unlikely(sp < (sizeof(struct thread_info) + STACK_WARN))) { - printk("do_IRQ: stack overflow: %ld\n", - sp - sizeof(struct thread_info)); - dump_stack(); - } - } -#endif +static void call_on_stack(void *func, void *stack) +{ + asm volatile("xchgl %%ebx,%%esp \n" + "call *%%edi \n" + "movl %%ebx,%%esp \n" + : "=b" (stack) + : "0" (stack), + "D"(func) + : "memory", "cc", "edx", "ecx", "eax"); +} -#ifdef CONFIG_4KSTACKS +static inline int +execute_on_irq_stack(int overflow, struct irq_desc *desc, int irq) +{ + union irq_ctx *curctx, *irqctx; + u32 *isp, arg1, arg2; curctx = (union irq_ctx *) current_thread_info(); irqctx = hardirq_ctx[smp_processor_id()]; @@ -111,52 +115,39 @@ unsigned int do_IRQ(struct pt_regs *regs) * handler) we can't do that and just have to keep using the * current stack (which is the irq stack already after all) */ - if (curctx != irqctx) { - int arg1, arg2, bx; - - /* 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; + if (unlikely(curctx == irqctx)) + return 0; - /* - * 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); - - asm volatile( - " xchgl %%ebx,%%esp \n" - " call *%%edi \n" - " movl %%ebx,%%esp \n" - : "=a" (arg1), "=d" (arg2), "=b" (bx) - : "0" (irq), "1" (desc), "2" (isp), - "D" (desc->handle_irq) - : "memory", "cc", "ecx" - ); - } else -#endif - desc->handle_irq(irq, desc); + /* 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; - irq_exit(); - set_irq_regs(old_regs); + /* + * 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); + + if (unlikely(overflow)) + call_on_stack(print_stack_overflow, isp); + + asm volatile("xchgl %%ebx,%%esp \n" + "call *%%edi \n" + "movl %%ebx,%%esp \n" + : "=a" (arg1), "=d" (arg2), "=b" (isp) + : "0" (irq), "1" (desc), "2" (isp), + "D" (desc->handle_irq) + : "memory", "cc", "ecx"); return 1; } -#ifdef CONFIG_4KSTACKS - -static char softirq_stack[NR_CPUS * THREAD_SIZE] - __attribute__((__section__(".bss.page_aligned"))); - -static char hardirq_stack[NR_CPUS * THREAD_SIZE] - __attribute__((__section__(".bss.page_aligned"))); - /* * allocate per-cpu stacks for hardirq and for softirq processing */ -void irq_ctx_init(int cpu) +void __cpuinit irq_ctx_init(int cpu) { union irq_ctx *irqctx; @@ -164,25 +155,25 @@ void irq_ctx_init(int cpu) return; irqctx = (union irq_ctx*) &hardirq_stack[cpu*THREAD_SIZE]; - irqctx->tinfo.task = NULL; - irqctx->tinfo.exec_domain = NULL; - irqctx->tinfo.cpu = cpu; - irqctx->tinfo.preempt_count = HARDIRQ_OFFSET; - irqctx->tinfo.addr_limit = MAKE_MM_SEG(0); + irqctx->tinfo.task = NULL; + irqctx->tinfo.exec_domain = NULL; + irqctx->tinfo.cpu = cpu; + irqctx->tinfo.preempt_count = HARDIRQ_OFFSET; + irqctx->tinfo.addr_limit = MAKE_MM_SEG(0); hardirq_ctx[cpu] = irqctx; irqctx = (union irq_ctx*) &softirq_stack[cpu*THREAD_SIZE]; - irqctx->tinfo.task = NULL; - irqctx->tinfo.exec_domain = NULL; - irqctx->tinfo.cpu = cpu; - irqctx->tinfo.preempt_count = 0; - irqctx->tinfo.addr_limit = MAKE_MM_SEG(0); + irqctx->tinfo.task = NULL; + irqctx->tinfo.exec_domain = NULL; + irqctx->tinfo.cpu = cpu; + irqctx->tinfo.preempt_count = 0; + irqctx->tinfo.addr_limit = MAKE_MM_SEG(0); softirq_ctx[cpu] = irqctx; - printk("CPU %u irqstacks, hard=%p soft=%p\n", - cpu,hardirq_ctx[cpu],softirq_ctx[cpu]); + printk(KERN_DEBUG "CPU %u irqstacks, hard=%p soft=%p\n", + cpu,hardirq_ctx[cpu],softirq_ctx[cpu]); } void irq_ctx_exit(int cpu) @@ -211,25 +202,56 @@ asmlinkage void do_softirq(void) /* build the stack frame on the softirq stack */ isp = (u32*) ((char*)irqctx + sizeof(*irqctx)); - asm volatile( - " xchgl %%ebx,%%esp \n" - " call __do_softirq \n" - " movl %%ebx,%%esp \n" - : "=b"(isp) - : "0"(isp) - : "memory", "cc", "edx", "ecx", "eax" - ); + call_on_stack(__do_softirq, isp); /* * Shouldnt happen, we returned above if in_interrupt(): - */ + */ WARN_ON_ONCE(softirq_count()); } local_irq_restore(flags); } + +#else +static inline int +execute_on_irq_stack(int overflow, struct irq_desc *desc, int irq) { return 0; } #endif /* + * do_IRQ handles all normal device IRQ's (the special + * SMP cross-CPU interrupts have their own specific + * handlers). + */ +unsigned int do_IRQ(struct pt_regs *regs) +{ + struct pt_regs *old_regs; + /* high bit used in ret_from_ code */ + int overflow, irq = ~regs->orig_ax; + struct irq_desc *desc = irq_desc + irq; + + if (unlikely((unsigned)irq >= NR_IRQS)) { + printk(KERN_EMERG "%s: cannot handle IRQ %d\n", + __func__, irq); + BUG(); + } + + old_regs = set_irq_regs(regs); + irq_enter(); + + overflow = check_stack_overflow(); + + if (!execute_on_irq_stack(overflow, desc, irq)) { + if (unlikely(overflow)) + print_stack_overflow(); + desc->handle_irq(irq, desc); + } + + irq_exit(); + set_irq_regs(old_regs); + return 1; +} + +/* * Interrupt statistics: */ @@ -313,16 +335,20 @@ skip: per_cpu(irq_stat,j).irq_tlb_count); seq_printf(p, " TLB shootdowns\n"); #endif +#ifdef CONFIG_X86_MCE seq_printf(p, "TRM: "); for_each_online_cpu(j) seq_printf(p, "%10u ", per_cpu(irq_stat,j).irq_thermal_count); seq_printf(p, " Thermal event interrupts\n"); +#endif +#ifdef CONFIG_X86_LOCAL_APIC seq_printf(p, "SPU: "); for_each_online_cpu(j) seq_printf(p, "%10u ", per_cpu(irq_stat,j).irq_spurious_count); seq_printf(p, " Spurious interrupts\n"); +#endif seq_printf(p, "ERR: %10u\n", atomic_read(&irq_err_count)); #if defined(CONFIG_X86_IO_APIC) seq_printf(p, "MIS: %10u\n", atomic_read(&irq_mis_count)); @@ -331,6 +357,40 @@ skip: return 0; } +/* + * /proc/stat helpers + */ +u64 arch_irq_stat_cpu(unsigned int cpu) +{ + u64 sum = nmi_count(cpu); + +#ifdef CONFIG_X86_LOCAL_APIC + sum += per_cpu(irq_stat, cpu).apic_timer_irqs; +#endif +#ifdef CONFIG_SMP + sum += per_cpu(irq_stat, cpu).irq_resched_count; + sum += per_cpu(irq_stat, cpu).irq_call_count; + sum += per_cpu(irq_stat, cpu).irq_tlb_count; +#endif +#ifdef CONFIG_X86_MCE + sum += per_cpu(irq_stat, cpu).irq_thermal_count; +#endif +#ifdef CONFIG_X86_LOCAL_APIC + sum += per_cpu(irq_stat, cpu).irq_spurious_count; +#endif + return sum; +} + +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; +} + #ifdef CONFIG_HOTPLUG_CPU #include <mach_apic.h> diff --git a/arch/x86/kernel/irq_64.c b/arch/x86/kernel/irq_64.c index 3aac15466a9..1f78b238d8d 100644 --- a/arch/x86/kernel/irq_64.c +++ b/arch/x86/kernel/irq_64.c @@ -135,6 +135,7 @@ skip: seq_printf(p, "%10u ", cpu_pda(j)->irq_tlb_count); seq_printf(p, " TLB shootdowns\n"); #endif +#ifdef CONFIG_X86_MCE seq_printf(p, "TRM: "); for_each_online_cpu(j) seq_printf(p, "%10u ", cpu_pda(j)->irq_thermal_count); @@ -143,6 +144,7 @@ skip: for_each_online_cpu(j) seq_printf(p, "%10u ", cpu_pda(j)->irq_threshold_count); seq_printf(p, " Threshold APIC interrupts\n"); +#endif seq_printf(p, "SPU: "); for_each_online_cpu(j) seq_printf(p, "%10u ", cpu_pda(j)->irq_spurious_count); @@ -153,6 +155,32 @@ skip: } /* + * /proc/stat helpers + */ +u64 arch_irq_stat_cpu(unsigned int cpu) +{ + u64 sum = cpu_pda(cpu)->__nmi_count; + + sum += cpu_pda(cpu)->apic_timer_irqs; +#ifdef CONFIG_SMP + sum += cpu_pda(cpu)->irq_resched_count; + sum += cpu_pda(cpu)->irq_call_count; + sum += cpu_pda(cpu)->irq_tlb_count; +#endif +#ifdef CONFIG_X86_MCE + sum += cpu_pda(cpu)->irq_thermal_count; + sum += cpu_pda(cpu)->irq_threshold_count; +#endif + sum += cpu_pda(cpu)->irq_spurious_count; + return sum; +} + +u64 arch_irq_stat(void) +{ + return atomic_read(&irq_err_count); +} + +/* * do_IRQ handles all normal device IRQ's (the special * SMP cross-CPU interrupts have their own specific * handlers). diff --git a/arch/x86/kernel/irqinit_32.c b/arch/x86/kernel/irqinit_32.c new file mode 100644 index 00000000000..d66914287ee --- /dev/null +++ b/arch/x86/kernel/irqinit_32.c @@ -0,0 +1,114 @@ +#include <linux/errno.h> +#include <linux/signal.h> +#include <linux/sched.h> +#include <linux/ioport.h> +#include <linux/interrupt.h> +#include <linux/slab.h> +#include <linux/random.h> +#include <linux/init.h> +#include <linux/kernel_stat.h> +#include <linux/sysdev.h> +#include <linux/bitops.h> + +#include <asm/atomic.h> +#include <asm/system.h> +#include <asm/io.h> +#include <asm/timer.h> +#include <asm/pgtable.h> +#include <asm/delay.h> +#include <asm/desc.h> +#include <asm/apic.h> +#include <asm/arch_hooks.h> +#include <asm/i8259.h> + + + +/* + * 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) +{ + extern void math_error(void __user *); + outb(0,0xF0); + if (ignore_fpu_irq || !boot_cpu_data.hard_math) + return IRQ_NONE; + math_error((void __user *)get_irq_regs()->ip); + 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, + .mask = CPU_MASK_NONE, + .name = "fpu", +}; + +void __init init_ISA_irqs (void) +{ + int i; + +#ifdef CONFIG_X86_LOCAL_APIC + init_bsp_APIC(); +#endif + init_8259A(0); + + /* + * 16 old-style INTA-cycle interrupts: + */ + for (i = 0; i < 16; i++) { + set_irq_chip_and_handler_name(i, &i8259A_chip, + handle_level_irq, "XT"); + } +} + +/* Overridden in paravirt.c */ +void init_IRQ(void) __attribute__((weak, alias("native_init_IRQ"))); + +void __init native_init_IRQ(void) +{ + int i; + + /* all the set up before the call gates are initialised */ + pre_intr_init_hook(); + + /* + * Cover the whole vector space, no vector can escape + * us. (some of these will be overridden and become + * 'special' SMP interrupts) + */ + for (i = 0; i < (NR_VECTORS - FIRST_EXTERNAL_VECTOR); i++) { + int vector = FIRST_EXTERNAL_VECTOR + i; + if (i >= NR_IRQS) + break; + /* SYSCALL_VECTOR was reserved in trap_init. */ + if (!test_bit(vector, used_vectors)) + set_intr_gate(vector, interrupt[i]); + } + + /* setup after call gates are initialised (usually add in + * the architecture specific gates) + */ + intr_init_hook(); + + /* + * 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()); +} diff --git a/arch/x86/kernel/irqinit_64.c b/arch/x86/kernel/irqinit_64.c new file mode 100644 index 00000000000..31f49e8f46a --- /dev/null +++ b/arch/x86/kernel/irqinit_64.c @@ -0,0 +1,217 @@ +#include <linux/linkage.h> +#include <linux/errno.h> +#include <linux/signal.h> +#include <linux/sched.h> +#include <linux/ioport.h> +#include <linux/interrupt.h> +#include <linux/timex.h> +#include <linux/slab.h> +#include <linux/random.h> +#include <linux/init.h> +#include <linux/kernel_stat.h> +#include <linux/sysdev.h> +#include <linux/bitops.h> + +#include <asm/acpi.h> +#include <asm/atomic.h> +#include <asm/system.h> +#include <asm/io.h> +#include <asm/hw_irq.h> +#include <asm/pgtable.h> +#include <asm/delay.h> +#include <asm/desc.h> +#include <asm/apic.h> +#include <asm/i8259.h> + +/* + * Common place to define all x86 IRQ vectors + * + * This builds up the IRQ handler stubs using some ugly macros in irq.h + * + * These macros create the low-level assembly IRQ routines that save + * register context and call do_IRQ(). do_IRQ() then does all the + * operations that are needed to keep the AT (or SMP IOAPIC) + * interrupt-controller happy. + */ + +#define IRQ_NAME2(nr) nr##_interrupt(void) +#define IRQ_NAME(nr) IRQ_NAME2(IRQ##nr) + +/* + * SMP has a few special interrupts for IPI messages + */ + +#define BUILD_IRQ(nr) \ + asmlinkage void IRQ_NAME(nr); \ + asm("\n.p2align\n" \ + "IRQ" #nr "_interrupt:\n\t" \ + "push $~(" #nr ") ; " \ + "jmp common_interrupt"); + +#define BI(x,y) \ + BUILD_IRQ(x##y) + +#define BUILD_16_IRQS(x) \ + BI(x,0) BI(x,1) BI(x,2) BI(x,3) \ + BI(x,4) BI(x,5) BI(x,6) BI(x,7) \ + BI(x,8) BI(x,9) BI(x,a) BI(x,b) \ + BI(x,c) BI(x,d) BI(x,e) BI(x,f) + +/* + * ISA PIC or low IO-APIC triggered (INTA-cycle or APIC) interrupts: + * (these are usually mapped to vectors 0x30-0x3f) + */ + +/* + * The IO-APIC gives us many more interrupt sources. Most of these + * are unused but an SMP system is supposed to have enough memory ... + * sometimes (mostly wrt. hw bugs) we get corrupted vectors all + * across the spectrum, so we really want to be prepared to get all + * of these. Plus, more powerful systems might have more than 64 + * IO-APIC registers. + * + * (these are usually mapped into the 0x30-0xff vector range) + */ + BUILD_16_IRQS(0x2) BUILD_16_IRQS(0x3) +BUILD_16_IRQS(0x4) BUILD_16_IRQS(0x5) BUILD_16_IRQS(0x6) BUILD_16_IRQS(0x7) +BUILD_16_IRQS(0x8) BUILD_16_IRQS(0x9) BUILD_16_IRQS(0xa) BUILD_16_IRQS(0xb) +BUILD_16_IRQS(0xc) BUILD_16_IRQS(0xd) BUILD_16_IRQS(0xe) BUILD_16_IRQS(0xf) + +#undef BUILD_16_IRQS +#undef BI + + +#define IRQ(x,y) \ + IRQ##x##y##_interrupt + +#define IRQLIST_16(x) \ + IRQ(x,0), IRQ(x,1), IRQ(x,2), IRQ(x,3), \ + IRQ(x,4), IRQ(x,5), IRQ(x,6), IRQ(x,7), \ + IRQ(x,8), IRQ(x,9), IRQ(x,a), IRQ(x,b), \ + IRQ(x,c), IRQ(x,d), IRQ(x,e), IRQ(x,f) + +/* for the irq vectors */ +static void (*__initdata interrupt[NR_VECTORS - FIRST_EXTERNAL_VECTOR])(void) = { + IRQLIST_16(0x2), IRQLIST_16(0x3), + IRQLIST_16(0x4), IRQLIST_16(0x5), IRQLIST_16(0x6), IRQLIST_16(0x7), + IRQLIST_16(0x8), IRQLIST_16(0x9), IRQLIST_16(0xa), IRQLIST_16(0xb), + IRQLIST_16(0xc), IRQLIST_16(0xd), IRQLIST_16(0xe), IRQLIST_16(0xf) +}; + +#undef IRQ +#undef IRQLIST_16 + + + + +/* + * IRQ2 is cascade interrupt to second interrupt controller + */ + +static struct irqaction irq2 = { + .handler = no_action, + .mask = CPU_MASK_NONE, + .name = "cascade", +}; +DEFINE_PER_CPU(vector_irq_t, vector_irq) = { + [0 ... IRQ0_VECTOR - 1] = -1, + [IRQ0_VECTOR] = 0, + [IRQ1_VECTOR] = 1, + [IRQ2_VECTOR] = 2, + [IRQ3_VECTOR] = 3, + [IRQ4_VECTOR] = 4, + [IRQ5_VECTOR] = 5, + [IRQ6_VECTOR] = 6, + [IRQ7_VECTOR] = 7, + [IRQ8_VECTOR] = 8, + [IRQ9_VECTOR] = 9, + [IRQ10_VECTOR] = 10, + [IRQ11_VECTOR] = 11, + [IRQ12_VECTOR] = 12, + [IRQ13_VECTOR] = 13, + [IRQ14_VECTOR] = 14, + [IRQ15_VECTOR] = 15, + [IRQ15_VECTOR + 1 ... NR_VECTORS - 1] = -1 +}; + +static void __init init_ISA_irqs (void) +{ + int i; + + init_bsp_APIC(); + init_8259A(0); + + for (i = 0; i < NR_IRQS; i++) { + irq_desc[i].status = IRQ_DISABLED; + irq_desc[i].action = NULL; + irq_desc[i].depth = 1; + + if (i < 16) { + /* + * 16 old-style INTA-cycle interrupts: + */ + set_irq_chip_and_handler_name(i, &i8259A_chip, + handle_level_irq, "XT"); + } else { + /* + * 'high' PCI IRQs filled in on demand + */ + irq_desc[i].chip = &no_irq_chip; + } + } +} + +void init_IRQ(void) __attribute__((weak, alias("native_init_IRQ"))); + +void __init native_init_IRQ(void) +{ + int i; + + init_ISA_irqs(); + /* + * Cover the whole vector space, no vector can escape + * us. (some of these will be overridden and become + * 'special' SMP interrupts) + */ + for (i = 0; i < (NR_VECTORS - FIRST_EXTERNAL_VECTOR); i++) { + int vector = FIRST_EXTERNAL_VECTOR + i; + if (vector != IA32_SYSCALL_VECTOR) + set_intr_gate(vector, interrupt[i]); + } + +#ifdef CONFIG_SMP + /* + * The reschedule interrupt is a CPU-to-CPU reschedule-helper + * IPI, driven by wakeup. + */ + alloc_intr_gate(RESCHEDULE_VECTOR, reschedule_interrupt); + + /* IPIs for invalidation */ + alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+0, invalidate_interrupt0); + alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+1, invalidate_interrupt1); + alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+2, invalidate_interrupt2); + alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+3, invalidate_interrupt3); + alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+4, invalidate_interrupt4); + alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+5, invalidate_interrupt5); + alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+6, invalidate_interrupt6); + alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+7, invalidate_interrupt7); + + /* IPI for generic function call */ + alloc_intr_gate(CALL_FUNCTION_VECTOR, call_function_interrupt); + + /* Low priority IPI to cleanup after moving an irq */ + set_intr_gate(IRQ_MOVE_CLEANUP_VECTOR, irq_move_cleanup_interrupt); +#endif + alloc_intr_gate(THERMAL_APIC_VECTOR, thermal_interrupt); + alloc_intr_gate(THRESHOLD_APIC_VECTOR, threshold_interrupt); + + /* self generated IPI for local APIC timer */ + alloc_intr_gate(LOCAL_TIMER_VECTOR, apic_timer_interrupt); + + /* IPI vectors for APIC spurious and error interrupts */ + alloc_intr_gate(SPURIOUS_APIC_VECTOR, spurious_interrupt); + alloc_intr_gate(ERROR_APIC_VECTOR, error_interrupt); + + if (!acpi_ioapic) + setup_irq(2, &irq2); +} diff --git a/arch/x86/kernel/ldt.c b/arch/x86/kernel/ldt.c index 0224c3637c7..21f2bae98c1 100644 --- a/arch/x86/kernel/ldt.c +++ b/arch/x86/kernel/ldt.c @@ -20,9 +20,9 @@ #include <asm/mmu_context.h> #ifdef CONFIG_SMP -static void flush_ldt(void *null) +static void flush_ldt(void *current_mm) { - if (current->active_mm) + if (current->active_mm == current_mm) load_LDT(¤t->active_mm->context); } #endif @@ -68,7 +68,7 @@ static int alloc_ldt(mm_context_t *pc, int mincount, int reload) load_LDT(pc); mask = cpumask_of_cpu(smp_processor_id()); if (!cpus_equal(current->mm->cpu_vm_mask, mask)) - smp_call_function(flush_ldt, NULL, 1, 1); + smp_call_function(flush_ldt, current->mm, 1, 1); preempt_enable(); #else load_LDT(pc); diff --git a/arch/x86/kernel/machine_kexec_32.c b/arch/x86/kernel/machine_kexec_32.c index d0b234c9fc3..f4960171bc6 100644 --- a/arch/x86/kernel/machine_kexec_32.c +++ b/arch/x86/kernel/machine_kexec_32.c @@ -39,7 +39,7 @@ static void set_idt(void *newidt, __u16 limit) curidt.address = (unsigned long)newidt; load_idt(&curidt); -}; +} static void set_gdt(void *newgdt, __u16 limit) @@ -51,7 +51,7 @@ static void set_gdt(void *newgdt, __u16 limit) curgdt.address = (unsigned long)newgdt; load_gdt(&curgdt); -}; +} static void load_segments(void) { diff --git a/arch/x86/kernel/machine_kexec_64.c b/arch/x86/kernel/machine_kexec_64.c index 576a03db451..7830dc4a838 100644 --- a/arch/x86/kernel/machine_kexec_64.c +++ b/arch/x86/kernel/machine_kexec_64.c @@ -110,7 +110,7 @@ static int init_pgtable(struct kimage *image, unsigned long start_pgtable) { pgd_t *level4p; level4p = (pgd_t *)__va(start_pgtable); - return init_level4_page(image, level4p, 0, end_pfn << PAGE_SHIFT); + return init_level4_page(image, level4p, 0, max_pfn << PAGE_SHIFT); } static void set_idt(void *newidt, u16 limit) diff --git a/arch/x86/kernel/microcode.c b/arch/x86/kernel/microcode.c index 69729e38b78..9758fea87c5 100644 --- a/arch/x86/kernel/microcode.c +++ b/arch/x86/kernel/microcode.c @@ -5,13 +5,14 @@ * 2006 Shaohua Li <shaohua.li@intel.com> * * This driver allows to upgrade microcode on Intel processors - * belonging to IA-32 family - PentiumPro, Pentium II, + * belonging to IA-32 family - PentiumPro, Pentium II, * Pentium III, Xeon, Pentium 4, etc. * - * Reference: Section 8.10 of Volume III, Intel Pentium 4 Manual, - * Order Number 245472 or free download from: - * - * http://developer.intel.com/design/pentium4/manuals/245472.htm + * 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/design/pentium4/manuals/253668.htm * * For more information, go to http://www.urbanmyth.org/microcode * @@ -58,12 +59,12 @@ * nature of implementation. * 1.11 22 Mar 2002 Tigran Aivazian <tigran@veritas.com> * Fix the panic when writing zero-length microcode chunk. - * 1.12 29 Sep 2003 Nitin Kamble <nitin.a.kamble@intel.com>, + * 1.12 29 Sep 2003 Nitin Kamble <nitin.a.kamble@intel.com>, * Jun Nakajima <jun.nakajima@intel.com> * Support for the microcode updates in the new format. * 1.13 10 Oct 2003 Tigran Aivazian <tigran@veritas.com> * Removed ->read() method and obsoleted MICROCODE_IOCFREE ioctl - * because we no longer hold a copy of applied microcode + * because we no longer hold a copy of applied microcode * in kernel memory. * 1.14 25 Jun 2004 Tigran Aivazian <tigran@veritas.com> * Fix sigmatch() macro to handle old CPUs with pf == 0. @@ -320,11 +321,11 @@ static void apply_microcode(int cpu) return; /* serialize access to the physical write to MSR 0x79 */ - spin_lock_irqsave(µcode_update_lock, flags); + spin_lock_irqsave(µcode_update_lock, flags); /* write microcode via MSR 0x79 */ wrmsr(MSR_IA32_UCODE_WRITE, - (unsigned long) uci->mc->bits, + (unsigned long) uci->mc->bits, (unsigned long) uci->mc->bits >> 16 >> 16); wrmsr(MSR_IA32_UCODE_REV, 0, 0); @@ -341,7 +342,7 @@ static void apply_microcode(int cpu) return; } printk(KERN_INFO "microcode: CPU%d updated from revision " - "0x%x to 0x%x, date = %08x \n", + "0x%x to 0x%x, date = %08x \n", cpu_num, uci->rev, val[1], uci->mc->hdr.date); uci->rev = val[1]; } @@ -534,7 +535,7 @@ static int cpu_request_microcode(int cpu) c->x86, c->x86_model, c->x86_mask); error = request_firmware(&firmware, name, µcode_pdev->dev); if (error) { - pr_debug("microcode: ucode data file %s load failed\n", name); + pr_debug("microcode: data file %s load failed\n", name); return error; } buf = firmware->data; @@ -805,6 +806,9 @@ static int __init microcode_init (void) { int error; + printk(KERN_INFO + "IA-32 Microcode Update Driver: v" MICROCODE_VERSION " <tigran@aivazian.fsnet.co.uk>\n"); + error = microcode_dev_init(); if (error) return error; @@ -825,9 +829,6 @@ static int __init microcode_init (void) } register_hotcpu_notifier(&mc_cpu_notifier); - - printk(KERN_INFO - "IA-32 Microcode Update Driver: v" MICROCODE_VERSION " <tigran@aivazian.fsnet.co.uk>\n"); return 0; } diff --git a/arch/x86/kernel/mmconf-fam10h_64.c b/arch/x86/kernel/mmconf-fam10h_64.c index edc5fbfe85c..fdfdc550b36 100644 --- a/arch/x86/kernel/mmconf-fam10h_64.c +++ b/arch/x86/kernel/mmconf-fam10h_64.c @@ -12,6 +12,7 @@ #include <asm/io.h> #include <asm/msr.h> #include <asm/acpi.h> +#include <asm/mmconfig.h> #include "../pci/pci.h" diff --git a/arch/x86/kernel/mpparse.c b/arch/x86/kernel/mpparse.c index 404683b94e7..3b25e49380c 100644 --- a/arch/x86/kernel/mpparse.c +++ b/arch/x86/kernel/mpparse.c @@ -25,6 +25,8 @@ #include <asm/proto.h> #include <asm/acpi.h> #include <asm/bios_ebda.h> +#include <asm/e820.h> +#include <asm/trampoline.h> #include <mach_apic.h> #ifdef CONFIG_X86_32 @@ -32,28 +34,6 @@ #include <mach_mpparse.h> #endif -/* Have we found an MP table */ -int smp_found_config; - -/* - * Various Linux-internal data structures created from the - * MP-table. - */ -#if defined (CONFIG_MCA) || defined (CONFIG_EISA) -int mp_bus_id_to_type[MAX_MP_BUSSES]; -#endif - -DECLARE_BITMAP(mp_bus_not_pci, MAX_MP_BUSSES); -int mp_bus_id_to_pci_bus[MAX_MP_BUSSES] = {[0 ... MAX_MP_BUSSES - 1] = -1 }; - -static int mp_current_pci_id; - -int pic_mode; - -/* - * Intel MP BIOS table parsing routines: - */ - /* * Checksum an MP configuration block. */ @@ -69,15 +49,73 @@ static int __init mpf_checksum(unsigned char *mp, int len) } #ifdef CONFIG_X86_NUMAQ +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_config_translation { + 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_config_translation *translation_table[MAX_MPC_ENTRY] __cpuinitdata; + +static inline int generate_logical_apicid(int quad, int phys_apicid) +{ + return (quad << 4) + (phys_apicid ? phys_apicid << 1 : 1); +} + + +static inline int mpc_apic_id(struct mpc_config_processor *m, + struct mpc_config_translation *translation_record) +{ + int quad = translation_record->trans_quad; + int logical_apicid = generate_logical_apicid(quad, m->mpc_apicid); + + printk(KERN_DEBUG "Processor #%d %u:%u APIC version %d (quad %d, apic %d)\n", + m->mpc_apicid, + (m->mpc_cpufeature & CPU_FAMILY_MASK) >> 8, + (m->mpc_cpufeature & CPU_MODEL_MASK) >> 4, + m->mpc_apicver, quad, logical_apicid); + return logical_apicid; +} + +int mp_bus_id_to_node[MAX_MP_BUSSES]; + +int mp_bus_id_to_local[MAX_MP_BUSSES]; + +static void mpc_oem_bus_info(struct mpc_config_bus *m, char *name, + struct mpc_config_translation *translation) +{ + int quad = translation->trans_quad; + int local = translation->trans_local; + + mp_bus_id_to_node[m->mpc_busid] = quad; + mp_bus_id_to_local[m->mpc_busid] = local; + printk(KERN_INFO "Bus #%d is %s (node %d)\n", + m->mpc_busid, name, quad); +} + +int quad_local_to_mp_bus_id [NR_CPUS/4][4]; +static void mpc_oem_pci_bus(struct mpc_config_bus *m, + struct mpc_config_translation *translation) +{ + int quad = translation->trans_quad; + int local = translation->trans_local; + + quad_local_to_mp_bus_id[quad][local] = m->mpc_busid; +} + #endif static void __cpuinit MP_processor_info(struct mpc_config_processor *m) @@ -90,7 +128,10 @@ static void __cpuinit MP_processor_info(struct mpc_config_processor *m) return; } #ifdef CONFIG_X86_NUMAQ - apicid = mpc_apic_id(m, translation_table[mpc_record]); + if (found_numaq) + apicid = mpc_apic_id(m, translation_table[mpc_record]); + else + apicid = m->mpc_apicid; #else apicid = m->mpc_apicid; #endif @@ -103,17 +144,18 @@ static void __cpuinit MP_processor_info(struct mpc_config_processor *m) generic_processor_info(apicid, m->mpc_apicver); } +#ifdef CONFIG_X86_IO_APIC static void __init MP_bus_info(struct mpc_config_bus *m) { char str[7]; - memcpy(str, m->mpc_bustype, 6); str[6] = 0; #ifdef CONFIG_X86_NUMAQ - mpc_oem_bus_info(m, str, translation_table[mpc_record]); + if (found_numaq) + mpc_oem_bus_info(m, str, translation_table[mpc_record]); #else - Dprintk("Bus #%d is %s\n", m->mpc_busid, str); + printk(KERN_INFO "Bus #%d is %s\n", m->mpc_busid, str); #endif #if MAX_MP_BUSSES < 256 @@ -132,11 +174,10 @@ static void __init MP_bus_info(struct mpc_config_bus *m) #endif } else if (strncmp(str, BUSTYPE_PCI, sizeof(BUSTYPE_PCI) - 1) == 0) { #ifdef CONFIG_X86_NUMAQ - mpc_oem_pci_bus(m, translation_table[mpc_record]); + if (found_numaq) + mpc_oem_pci_bus(m, translation_table[mpc_record]); #endif clear_bit(m->mpc_busid, mp_bus_not_pci); - mp_bus_id_to_pci_bus[m->mpc_busid] = mp_current_pci_id; - mp_current_pci_id++; #if defined(CONFIG_EISA) || defined (CONFIG_MCA) mp_bus_id_to_type[m->mpc_busid] = MP_BUS_PCI; } else if (strncmp(str, BUSTYPE_EISA, sizeof(BUSTYPE_EISA) - 1) == 0) { @@ -147,6 +188,7 @@ static void __init MP_bus_info(struct mpc_config_bus *m) } else printk(KERN_WARNING "Unknown bustype %s - ignoring\n", str); } +#endif #ifdef CONFIG_X86_IO_APIC @@ -176,18 +218,89 @@ static void __init MP_ioapic_info(struct mpc_config_ioapic *m) if (bad_ioapic(m->mpc_apicaddr)) return; - mp_ioapics[nr_ioapics] = *m; + mp_ioapics[nr_ioapics].mp_apicaddr = m->mpc_apicaddr; + mp_ioapics[nr_ioapics].mp_apicid = m->mpc_apicid; + mp_ioapics[nr_ioapics].mp_type = m->mpc_type; + mp_ioapics[nr_ioapics].mp_apicver = m->mpc_apicver; + mp_ioapics[nr_ioapics].mp_flags = m->mpc_flags; nr_ioapics++; } -static void __init MP_intsrc_info(struct mpc_config_intsrc *m) +static void print_MP_intsrc_info(struct mpc_config_intsrc *m) { - mp_irqs[mp_irq_entries] = *m; - Dprintk("Int: type %d, pol %d, trig %d, bus %d," + printk(KERN_CONT "Int: type %d, pol %d, trig %d, bus %02x," " IRQ %02x, APIC ID %x, APIC INT %02x\n", m->mpc_irqtype, m->mpc_irqflag & 3, (m->mpc_irqflag >> 2) & 3, m->mpc_srcbus, m->mpc_srcbusirq, m->mpc_dstapic, m->mpc_dstirq); +} + +static void __init print_mp_irq_info(struct mp_config_intsrc *mp_irq) +{ + printk(KERN_CONT "Int: type %d, pol %d, trig %d, bus %02x," + " IRQ %02x, APIC ID %x, APIC INT %02x\n", + mp_irq->mp_irqtype, mp_irq->mp_irqflag & 3, + (mp_irq->mp_irqflag >> 2) & 3, mp_irq->mp_srcbus, + mp_irq->mp_srcbusirq, mp_irq->mp_dstapic, mp_irq->mp_dstirq); +} + +static void __init assign_to_mp_irq(struct mpc_config_intsrc *m, + struct mp_config_intsrc *mp_irq) +{ + mp_irq->mp_dstapic = m->mpc_dstapic; + mp_irq->mp_type = m->mpc_type; + mp_irq->mp_irqtype = m->mpc_irqtype; + mp_irq->mp_irqflag = m->mpc_irqflag; + mp_irq->mp_srcbus = m->mpc_srcbus; + mp_irq->mp_srcbusirq = m->mpc_srcbusirq; + mp_irq->mp_dstirq = m->mpc_dstirq; +} + +static void __init assign_to_mpc_intsrc(struct mp_config_intsrc *mp_irq, + struct mpc_config_intsrc *m) +{ + m->mpc_dstapic = mp_irq->mp_dstapic; + m->mpc_type = mp_irq->mp_type; + m->mpc_irqtype = mp_irq->mp_irqtype; + m->mpc_irqflag = mp_irq->mp_irqflag; + m->mpc_srcbus = mp_irq->mp_srcbus; + m->mpc_srcbusirq = mp_irq->mp_srcbusirq; + m->mpc_dstirq = mp_irq->mp_dstirq; +} + +static int __init mp_irq_mpc_intsrc_cmp(struct mp_config_intsrc *mp_irq, + struct mpc_config_intsrc *m) +{ + if (mp_irq->mp_dstapic != m->mpc_dstapic) + return 1; + if (mp_irq->mp_type != m->mpc_type) + return 2; + if (mp_irq->mp_irqtype != m->mpc_irqtype) + return 3; + if (mp_irq->mp_irqflag != m->mpc_irqflag) + return 4; + if (mp_irq->mp_srcbus != m->mpc_srcbus) + return 5; + if (mp_irq->mp_srcbusirq != m->mpc_srcbusirq) + return 6; + if (mp_irq->mp_dstirq != m->mpc_dstirq) + return 7; + + return 0; +} + +static void __init MP_intsrc_info(struct mpc_config_intsrc *m) +{ + int i; + + print_MP_intsrc_info(m); + + for (i = 0; i < mp_irq_entries; i++) { + if (!mp_irq_mpc_intsrc_cmp(&mp_irqs[i], m)) + return; + } + + assign_to_mp_irq(m, &mp_irqs[mp_irq_entries]); if (++mp_irq_entries == MAX_IRQ_SOURCES) panic("Max # of irq sources exceeded!!\n"); } @@ -196,7 +309,7 @@ static void __init MP_intsrc_info(struct mpc_config_intsrc *m) static void __init MP_lintsrc_info(struct mpc_config_lintsrc *m) { - Dprintk("Lint: type %d, pol %d, trig %d, bus %d," + printk(KERN_INFO "Lint: type %d, pol %d, trig %d, bus %02x," " IRQ %02x, APIC ID %x, APIC LINT %02x\n", m->mpc_irqtype, m->mpc_irqflag & 3, (m->mpc_irqflag >> 2) & 3, m->mpc_srcbusid, @@ -266,11 +379,14 @@ static void __init smp_read_mpc_oem(struct mp_config_oemtable *oemtable, } } -static inline void mps_oem_check(struct mp_config_table *mpc, char *oem, +void numaq_mps_oem_check(struct mp_config_table *mpc, char *oem, char *productid) { if (strncmp(oem, "IBM NUMA", 8)) - printk("Warning! May not be a NUMA-Q system!\n"); + printk("Warning! Not a NUMA-Q system!\n"); + else + found_numaq = 1; + if (mpc->mpc_oemptr) smp_read_mpc_oem((struct mp_config_oemtable *)mpc->mpc_oemptr, mpc->mpc_oemsize); @@ -281,12 +397,9 @@ static inline void mps_oem_check(struct mp_config_table *mpc, char *oem, * Read/parse the MPC */ -static int __init smp_read_mpc(struct mp_config_table *mpc, unsigned early) +static int __init smp_check_mpc(struct mp_config_table *mpc, char *oem, + char *str) { - char str[16]; - char oem[10]; - int count = sizeof(*mpc); - unsigned char *mpt = ((unsigned char *)mpc) + count; if (memcmp(mpc->mpc_signature, MPC_SIGNATURE, 4)) { printk(KERN_ERR "MPTABLE: bad signature [%c%c%c%c]!\n", @@ -309,19 +422,42 @@ static int __init smp_read_mpc(struct mp_config_table *mpc, unsigned early) } memcpy(oem, mpc->mpc_oem, 8); oem[8] = 0; - printk(KERN_INFO "MPTABLE: OEM ID: %s ", oem); + printk(KERN_INFO "MPTABLE: OEM ID: %s\n", oem); memcpy(str, mpc->mpc_productid, 12); str[12] = 0; - printk("Product ID: %s ", str); -#ifdef CONFIG_X86_32 - mps_oem_check(mpc, oem, str); -#endif - printk(KERN_INFO "MPTABLE: Product ID: %s ", str); + printk(KERN_INFO "MPTABLE: Product ID: %s\n", str); printk(KERN_INFO "MPTABLE: APIC at: 0x%X\n", mpc->mpc_lapic); + return 1; +} + +static int __init smp_read_mpc(struct mp_config_table *mpc, unsigned early) +{ + char str[16]; + char oem[10]; + + int count = sizeof(*mpc); + unsigned char *mpt = ((unsigned char *)mpc) + count; + + if (!smp_check_mpc(mpc, oem, str)) + return 0; + +#ifdef CONFIG_X86_32 + /* + * need to make sure summit and es7000's mps_oem_check is safe to be + * called early via genericarch 's mps_oem_check + */ + if (early) { +#ifdef CONFIG_X86_NUMAQ + numaq_mps_oem_check(mpc, oem, str); +#endif + } else + mps_oem_check(mpc, oem, str); +#endif + /* save the local APIC address, it might be non-default */ if (!acpi_lapic) mp_lapic_addr = mpc->mpc_lapic; @@ -352,7 +488,9 @@ static int __init smp_read_mpc(struct mp_config_table *mpc, unsigned early) { struct mpc_config_bus *m = (struct mpc_config_bus *)mpt; +#ifdef CONFIG_X86_IO_APIC MP_bus_info(m); +#endif mpt += sizeof(*m); count += sizeof(*m); break; @@ -402,6 +540,11 @@ static int __init smp_read_mpc(struct mp_config_table *mpc, unsigned early) ++mpc_record; #endif } + +#ifdef CONFIG_X86_GENERICARCH + generic_bigsmp_probe(); +#endif + setup_apic_routing(); if (!num_processors) printk(KERN_ERR "MPTABLE: no processors registered!\n"); @@ -427,7 +570,7 @@ static void __init construct_default_ioirq_mptable(int mpc_default_type) intsrc.mpc_type = MP_INTSRC; intsrc.mpc_irqflag = 0; /* conforming */ intsrc.mpc_srcbus = 0; - intsrc.mpc_dstapic = mp_ioapics[0].mpc_apicid; + intsrc.mpc_dstapic = mp_ioapics[0].mp_apicid; intsrc.mpc_irqtype = mp_INT; @@ -488,40 +631,11 @@ static void __init construct_default_ioirq_mptable(int mpc_default_type) MP_intsrc_info(&intsrc); } -#endif -static inline void __init construct_default_ISA_mptable(int mpc_default_type) +static void construct_ioapic_table(int mpc_default_type) { - struct mpc_config_processor processor; - struct mpc_config_bus bus; -#ifdef CONFIG_X86_IO_APIC struct mpc_config_ioapic ioapic; -#endif - struct mpc_config_lintsrc lintsrc; - int linttypes[2] = { mp_ExtINT, mp_NMI }; - int i; - - /* - * local APIC has default address - */ - mp_lapic_addr = APIC_DEFAULT_PHYS_BASE; - - /* - * 2 CPUs, numbered 0 & 1. - */ - processor.mpc_type = MP_PROCESSOR; - /* Either an integrated APIC or a discrete 82489DX. */ - processor.mpc_apicver = mpc_default_type > 4 ? 0x10 : 0x01; - processor.mpc_cpuflag = CPU_ENABLED; - processor.mpc_cpufeature = (boot_cpu_data.x86 << 8) | - (boot_cpu_data.x86_model << 4) | boot_cpu_data.x86_mask; - processor.mpc_featureflag = boot_cpu_data.x86_capability[0]; - processor.mpc_reserved[0] = 0; - processor.mpc_reserved[1] = 0; - for (i = 0; i < 2; i++) { - processor.mpc_apicid = i; - MP_processor_info(&processor); - } + struct mpc_config_bus bus; bus.mpc_type = MP_BUS; bus.mpc_busid = 0; @@ -550,7 +664,6 @@ static inline void __init construct_default_ISA_mptable(int mpc_default_type) MP_bus_info(&bus); } -#ifdef CONFIG_X86_IO_APIC ioapic.mpc_type = MP_IOAPIC; ioapic.mpc_apicid = 2; ioapic.mpc_apicver = mpc_default_type > 4 ? 0x10 : 0x01; @@ -562,7 +675,42 @@ static inline void __init construct_default_ISA_mptable(int mpc_default_type) * We set up most of the low 16 IO-APIC pins according to MPS rules. */ construct_default_ioirq_mptable(mpc_default_type); +} +#else +static inline void construct_ioapic_table(int mpc_default_type) { } #endif + +static inline void __init construct_default_ISA_mptable(int mpc_default_type) +{ + struct mpc_config_processor processor; + struct mpc_config_lintsrc lintsrc; + int linttypes[2] = { mp_ExtINT, mp_NMI }; + int i; + + /* + * local APIC has default address + */ + mp_lapic_addr = APIC_DEFAULT_PHYS_BASE; + + /* + * 2 CPUs, numbered 0 & 1. + */ + processor.mpc_type = MP_PROCESSOR; + /* Either an integrated APIC or a discrete 82489DX. */ + processor.mpc_apicver = mpc_default_type > 4 ? 0x10 : 0x01; + processor.mpc_cpuflag = CPU_ENABLED; + processor.mpc_cpufeature = (boot_cpu_data.x86 << 8) | + (boot_cpu_data.x86_model << 4) | boot_cpu_data.x86_mask; + processor.mpc_featureflag = boot_cpu_data.x86_capability[0]; + processor.mpc_reserved[0] = 0; + processor.mpc_reserved[1] = 0; + for (i = 0; i < 2; i++) { + processor.mpc_apicid = i; + MP_processor_info(&processor); + } + + construct_ioapic_table(mpc_default_type); + lintsrc.mpc_type = MP_LINTSRC; lintsrc.mpc_irqflag = 0; /* conforming */ lintsrc.mpc_srcbusid = 0; @@ -578,12 +726,22 @@ static inline void __init construct_default_ISA_mptable(int mpc_default_type) static struct intel_mp_floating *mpf_found; /* + * Machine specific quirk for finding the SMP config before other setup + * activities destroy the table: + */ +int (*mach_get_smp_config_quirk)(unsigned int early); + +/* * Scan the memory blocks for an SMP configuration block. */ -static void __init __get_smp_config(unsigned early) +static void __init __get_smp_config(unsigned int early) { struct intel_mp_floating *mpf = mpf_found; + if (mach_get_smp_config_quirk) { + if (mach_get_smp_config_quirk(early)) + return; + } if (acpi_lapic && early) return; /* @@ -600,7 +758,7 @@ static void __init __get_smp_config(unsigned early) printk(KERN_INFO "Intel MultiProcessor Specification v1.%d\n", mpf->mpf_specification); -#ifdef CONFIG_X86_32 +#if defined(CONFIG_X86_LOCAL_APIC) && defined(CONFIG_X86_32) if (mpf->mpf_feature2 & (1 << 7)) { printk(KERN_INFO " IMCR and PIC compatibility mode.\n"); pic_mode = 1; @@ -632,7 +790,9 @@ static void __init __get_smp_config(unsigned early) * override the defaults. */ if (!smp_read_mpc(phys_to_virt(mpf->mpf_physptr), early)) { +#ifdef CONFIG_X86_LOCAL_APIC smp_found_config = 0; +#endif printk(KERN_ERR "BIOS bug, MP table errors detected!...\n"); printk(KERN_ERR "... disabling SMP support. " @@ -689,7 +849,7 @@ static int __init smp_scan_config(unsigned long base, unsigned long length, unsigned int *bp = phys_to_virt(base); struct intel_mp_floating *mpf; - Dprintk("Scan SMP from %p for %ld bytes.\n", bp, length); + printk(KERN_DEBUG "Scan SMP from %p for %ld bytes.\n", bp, length); BUILD_BUG_ON(sizeof(*mpf) != 16); while (length > 0) { @@ -699,15 +859,21 @@ static int __init smp_scan_config(unsigned long base, unsigned long length, !mpf_checksum((unsigned char *)bp, 16) && ((mpf->mpf_specification == 1) || (mpf->mpf_specification == 4))) { - +#ifdef CONFIG_X86_LOCAL_APIC smp_found_config = 1; +#endif mpf_found = mpf; -#ifdef CONFIG_X86_32 + printk(KERN_INFO "found SMP MP-table at [%p] %08lx\n", mpf, virt_to_phys(mpf)); - reserve_bootmem(virt_to_phys(mpf), PAGE_SIZE, + + if (!reserve) + return 1; + reserve_bootmem_generic(virt_to_phys(mpf), PAGE_SIZE, BOOTMEM_DEFAULT); if (mpf->mpf_physptr) { + unsigned long size = PAGE_SIZE; +#ifdef CONFIG_X86_32 /* * We cannot access to MPC table to compute * table size yet, as only few megabytes from @@ -717,24 +883,15 @@ static int __init smp_scan_config(unsigned long base, unsigned long length, * PAGE_SIZE from mpg->mpf_physptr yields BUG() * in reserve_bootmem. */ - unsigned long size = PAGE_SIZE; unsigned long end = max_low_pfn * PAGE_SIZE; if (mpf->mpf_physptr + size > end) size = end - mpf->mpf_physptr; - reserve_bootmem(mpf->mpf_physptr, size, +#endif + reserve_bootmem_generic(mpf->mpf_physptr, size, BOOTMEM_DEFAULT); } -#else - if (!reserve) - return 1; - - reserve_bootmem_generic(virt_to_phys(mpf), PAGE_SIZE); - if (mpf->mpf_physptr) - reserve_bootmem_generic(mpf->mpf_physptr, - PAGE_SIZE); -#endif - return 1; + return 1; } bp += 4; length -= 16; @@ -742,10 +899,16 @@ static int __init smp_scan_config(unsigned long base, unsigned long length, return 0; } -static void __init __find_smp_config(unsigned reserve) +int (*mach_find_smp_config_quirk)(unsigned int reserve); + +static void __init __find_smp_config(unsigned int reserve) { unsigned int address; + if (mach_find_smp_config_quirk) { + if (mach_find_smp_config_quirk(reserve)) + return; + } /* * FIXME: Linux assumes you have 640K of base ram.. * this continues the error... @@ -790,298 +953,294 @@ void __init find_smp_config(void) __find_smp_config(1); } -/* -------------------------------------------------------------------------- - ACPI-based MP Configuration - -------------------------------------------------------------------------- */ +#ifdef CONFIG_X86_IO_APIC +static u8 __initdata irq_used[MAX_IRQ_SOURCES]; -/* - * Keep this outside and initialized to 0, for !CONFIG_ACPI builds: - */ -int es7000_plat; +static int __init get_MP_intsrc_index(struct mpc_config_intsrc *m) +{ + int i; -#ifdef CONFIG_ACPI + if (m->mpc_irqtype != mp_INT) + return 0; -#ifdef CONFIG_X86_IO_APIC + if (m->mpc_irqflag != 0x0f) + return 0; -#define MP_ISA_BUS 0 + /* not legacy */ -extern struct mp_ioapic_routing mp_ioapic_routing[MAX_IO_APICS]; + for (i = 0; i < mp_irq_entries; i++) { + if (mp_irqs[i].mp_irqtype != mp_INT) + continue; -static int mp_find_ioapic(int gsi) -{ - int i = 0; + if (mp_irqs[i].mp_irqflag != 0x0f) + continue; - /* Find the IOAPIC that manages this GSI. */ - for (i = 0; i < nr_ioapics; i++) { - if ((gsi >= mp_ioapic_routing[i].gsi_base) - && (gsi <= mp_ioapic_routing[i].gsi_end)) - return i; + if (mp_irqs[i].mp_srcbus != m->mpc_srcbus) + continue; + if (mp_irqs[i].mp_srcbusirq != m->mpc_srcbusirq) + continue; + if (irq_used[i]) { + /* already claimed */ + return -2; + } + irq_used[i] = 1; + return i; } - printk(KERN_ERR "ERROR: Unable to locate IOAPIC for GSI %d\n", gsi); + /* not found */ return -1; } -static u8 __init uniq_ioapic_id(u8 id) -{ -#ifdef CONFIG_X86_32 - if ((boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) && - !APIC_XAPIC(apic_version[boot_cpu_physical_apicid])) - return io_apic_get_unique_id(nr_ioapics, id); - else - return id; -#else - int i; - DECLARE_BITMAP(used, 256); - bitmap_zero(used, 256); - for (i = 0; i < nr_ioapics; i++) { - struct mpc_config_ioapic *ia = &mp_ioapics[i]; - __set_bit(ia->mpc_apicid, used); - } - if (!test_bit(id, used)) - return id; - return find_first_zero_bit(used, 256); +#define SPARE_SLOT_NUM 20 + +static struct mpc_config_intsrc __initdata *m_spare[SPARE_SLOT_NUM]; #endif -} -void __init mp_register_ioapic(int id, u32 address, u32 gsi_base) +static int __init replace_intsrc_all(struct mp_config_table *mpc, + unsigned long mpc_new_phys, + unsigned long mpc_new_length) { - int idx = 0; - - if (bad_ioapic(address)) - return; +#ifdef CONFIG_X86_IO_APIC + int i; + int nr_m_spare = 0; +#endif - idx = nr_ioapics; + int count = sizeof(*mpc); + unsigned char *mpt = ((unsigned char *)mpc) + count; - mp_ioapics[idx].mpc_type = MP_IOAPIC; - mp_ioapics[idx].mpc_flags = MPC_APIC_USABLE; - mp_ioapics[idx].mpc_apicaddr = address; + printk(KERN_INFO "mpc_length %x\n", mpc->mpc_length); + while (count < mpc->mpc_length) { + switch (*mpt) { + case MP_PROCESSOR: + { + struct mpc_config_processor *m = + (struct mpc_config_processor *)mpt; + mpt += sizeof(*m); + count += sizeof(*m); + break; + } + case MP_BUS: + { + struct mpc_config_bus *m = + (struct mpc_config_bus *)mpt; + mpt += sizeof(*m); + count += sizeof(*m); + break; + } + case MP_IOAPIC: + { + mpt += sizeof(struct mpc_config_ioapic); + count += sizeof(struct mpc_config_ioapic); + break; + } + case MP_INTSRC: + { +#ifdef CONFIG_X86_IO_APIC + struct mpc_config_intsrc *m = + (struct mpc_config_intsrc *)mpt; - set_fixmap_nocache(FIX_IO_APIC_BASE_0 + idx, address); - mp_ioapics[idx].mpc_apicid = uniq_ioapic_id(id); -#ifdef CONFIG_X86_32 - mp_ioapics[idx].mpc_apicver = io_apic_get_version(idx); -#else - mp_ioapics[idx].mpc_apicver = 0; + printk(KERN_INFO "OLD "); + print_MP_intsrc_info(m); + i = get_MP_intsrc_index(m); + if (i > 0) { + assign_to_mpc_intsrc(&mp_irqs[i], m); + printk(KERN_INFO "NEW "); + print_mp_irq_info(&mp_irqs[i]); + } else if (!i) { + /* legacy, do nothing */ + } else if (nr_m_spare < SPARE_SLOT_NUM) { + /* + * not found (-1), or duplicated (-2) + * are invalid entries, + * we need to use the slot later + */ + m_spare[nr_m_spare] = m; + nr_m_spare++; + } #endif - /* - * Build basic GSI lookup table to facilitate gsi->io_apic lookups - * and to prevent reprogramming of IOAPIC pins (PCI GSIs). - */ - mp_ioapic_routing[idx].apic_id = mp_ioapics[idx].mpc_apicid; - mp_ioapic_routing[idx].gsi_base = gsi_base; - mp_ioapic_routing[idx].gsi_end = gsi_base + - io_apic_get_redir_entries(idx); - - printk(KERN_INFO "IOAPIC[%d]: apic_id %d, version %d, address 0x%x, " - "GSI %d-%d\n", idx, mp_ioapics[idx].mpc_apicid, - mp_ioapics[idx].mpc_apicver, mp_ioapics[idx].mpc_apicaddr, - mp_ioapic_routing[idx].gsi_base, mp_ioapic_routing[idx].gsi_end); - - nr_ioapics++; -} + mpt += sizeof(struct mpc_config_intsrc); + count += sizeof(struct mpc_config_intsrc); + break; + } + case MP_LINTSRC: + { + struct mpc_config_lintsrc *m = + (struct mpc_config_lintsrc *)mpt; + mpt += sizeof(*m); + count += sizeof(*m); + break; + } + default: + /* wrong mptable */ + printk(KERN_ERR "Your mptable is wrong, contact your HW vendor!\n"); + printk(KERN_ERR "type %x\n", *mpt); + print_hex_dump(KERN_ERR, " ", DUMP_PREFIX_ADDRESS, 16, + 1, mpc, mpc->mpc_length, 1); + goto out; + } + } -void __init mp_override_legacy_irq(u8 bus_irq, u8 polarity, u8 trigger, u32 gsi) -{ - struct mpc_config_intsrc intsrc; - int ioapic = -1; - int pin = -1; +#ifdef CONFIG_X86_IO_APIC + for (i = 0; i < mp_irq_entries; i++) { + if (irq_used[i]) + continue; - /* - * Convert 'gsi' to 'ioapic.pin'. - */ - ioapic = mp_find_ioapic(gsi); - if (ioapic < 0) - return; - pin = gsi - mp_ioapic_routing[ioapic].gsi_base; + if (mp_irqs[i].mp_irqtype != mp_INT) + continue; - /* - * TBD: This check is for faulty timer entries, where the override - * erroneously sets the trigger to level, resulting in a HUGE - * increase of timer interrupts! - */ - if ((bus_irq == 0) && (trigger == 3)) - trigger = 1; + if (mp_irqs[i].mp_irqflag != 0x0f) + continue; - intsrc.mpc_type = MP_INTSRC; - intsrc.mpc_irqtype = mp_INT; - intsrc.mpc_irqflag = (trigger << 2) | polarity; - intsrc.mpc_srcbus = MP_ISA_BUS; - intsrc.mpc_srcbusirq = bus_irq; /* IRQ */ - intsrc.mpc_dstapic = mp_ioapics[ioapic].mpc_apicid; /* APIC ID */ - intsrc.mpc_dstirq = pin; /* INTIN# */ + if (nr_m_spare > 0) { + printk(KERN_INFO "*NEW* found "); + nr_m_spare--; + assign_to_mpc_intsrc(&mp_irqs[i], m_spare[nr_m_spare]); + m_spare[nr_m_spare] = NULL; + } else { + struct mpc_config_intsrc *m = + (struct mpc_config_intsrc *)mpt; + count += sizeof(struct mpc_config_intsrc); + if (!mpc_new_phys) { + printk(KERN_INFO "No spare slots, try to append...take your risk, new mpc_length %x\n", count); + } else { + if (count <= mpc_new_length) + printk(KERN_INFO "No spare slots, try to append..., new mpc_length %x\n", count); + else { + printk(KERN_ERR "mpc_new_length %lx is too small\n", mpc_new_length); + goto out; + } + } + assign_to_mpc_intsrc(&mp_irqs[i], m); + mpc->mpc_length = count; + mpt += sizeof(struct mpc_config_intsrc); + } + print_mp_irq_info(&mp_irqs[i]); + } +#endif +out: + /* update checksum */ + mpc->mpc_checksum = 0; + mpc->mpc_checksum -= mpf_checksum((unsigned char *)mpc, + mpc->mpc_length); - MP_intsrc_info(&intsrc); + return 0; } -void __init mp_config_acpi_legacy_irqs(void) -{ - struct mpc_config_intsrc intsrc; - int i = 0; - int ioapic = -1; +static int __initdata enable_update_mptable; -#if defined (CONFIG_MCA) || defined (CONFIG_EISA) - /* - * Fabricate the legacy ISA bus (bus #31). - */ - mp_bus_id_to_type[MP_ISA_BUS] = MP_BUS_ISA; -#endif - set_bit(MP_ISA_BUS, mp_bus_not_pci); - Dprintk("Bus #%d is ISA\n", MP_ISA_BUS); +static int __init update_mptable_setup(char *str) +{ + enable_update_mptable = 1; + return 0; +} +early_param("update_mptable", update_mptable_setup); - /* - * Older generations of ES7000 have no legacy identity mappings - */ - if (es7000_plat == 1) - return; +static unsigned long __initdata mpc_new_phys; +static unsigned long mpc_new_length __initdata = 4096; - /* - * Locate the IOAPIC that manages the ISA IRQs (0-15). - */ - ioapic = mp_find_ioapic(0); - if (ioapic < 0) - return; +/* alloc_mptable or alloc_mptable=4k */ +static int __initdata alloc_mptable; +static int __init parse_alloc_mptable_opt(char *p) +{ + enable_update_mptable = 1; + alloc_mptable = 1; + if (!p) + return 0; + mpc_new_length = memparse(p, &p); + return 0; +} +early_param("alloc_mptable", parse_alloc_mptable_opt); - intsrc.mpc_type = MP_INTSRC; - intsrc.mpc_irqflag = 0; /* Conforming */ - intsrc.mpc_srcbus = MP_ISA_BUS; -#ifdef CONFIG_X86_IO_APIC - intsrc.mpc_dstapic = mp_ioapics[ioapic].mpc_apicid; +void __init early_reserve_e820_mpc_new(void) +{ + if (enable_update_mptable && alloc_mptable) { + u64 startt = 0; +#ifdef CONFIG_X86_TRAMPOLINE + startt = TRAMPOLINE_BASE; #endif - /* - * Use the default configuration for the IRQs 0-15. Unless - * overridden by (MADT) interrupt source override entries. - */ - for (i = 0; i < 16; i++) { - int idx; - - for (idx = 0; idx < mp_irq_entries; idx++) { - struct mpc_config_intsrc *irq = mp_irqs + idx; - - /* Do we already have a mapping for this ISA IRQ? */ - if (irq->mpc_srcbus == MP_ISA_BUS - && irq->mpc_srcbusirq == i) - break; - - /* Do we already have a mapping for this IOAPIC pin */ - if ((irq->mpc_dstapic == intsrc.mpc_dstapic) && - (irq->mpc_dstirq == i)) - break; - } - - if (idx != mp_irq_entries) { - printk(KERN_DEBUG "ACPI: IRQ%d used by override.\n", i); - continue; /* IRQ already used */ - } - - intsrc.mpc_irqtype = mp_INT; - intsrc.mpc_srcbusirq = i; /* Identity mapped */ - intsrc.mpc_dstirq = i; - - MP_intsrc_info(&intsrc); + mpc_new_phys = early_reserve_e820(startt, mpc_new_length, 4); } } -int mp_register_gsi(u32 gsi, int triggering, int polarity) +static int __init update_mp_table(void) { - int ioapic; - int ioapic_pin; -#ifdef CONFIG_X86_32 -#define MAX_GSI_NUM 4096 -#define IRQ_COMPRESSION_START 64 + char str[16]; + char oem[10]; + struct intel_mp_floating *mpf; + struct mp_config_table *mpc; + struct mp_config_table *mpc_new; + + if (!enable_update_mptable) + return 0; + + mpf = mpf_found; + if (!mpf) + return 0; - static int pci_irq = IRQ_COMPRESSION_START; /* - * Mapping between Global System Interrupts, which - * represent all possible interrupts, and IRQs - * assigned to actual devices. + * Now see if we need to go further. */ - static int gsi_to_irq[MAX_GSI_NUM]; -#else - - if (acpi_irq_model != ACPI_IRQ_MODEL_IOAPIC) - return gsi; -#endif + if (mpf->mpf_feature1 != 0) + return 0; - /* Don't set up the ACPI SCI because it's already set up */ - if (acpi_gbl_FADT.sci_interrupt == gsi) - return gsi; + if (!mpf->mpf_physptr) + return 0; - ioapic = mp_find_ioapic(gsi); - if (ioapic < 0) { - printk(KERN_WARNING "No IOAPIC for GSI %u\n", gsi); - return gsi; - } + mpc = phys_to_virt(mpf->mpf_physptr); - ioapic_pin = gsi - mp_ioapic_routing[ioapic].gsi_base; + if (!smp_check_mpc(mpc, oem, str)) + return 0; -#ifdef CONFIG_X86_32 - if (ioapic_renumber_irq) - gsi = ioapic_renumber_irq(ioapic, gsi); -#endif + printk(KERN_INFO "mpf: %lx\n", virt_to_phys(mpf)); + printk(KERN_INFO "mpf_physptr: %x\n", mpf->mpf_physptr); - /* - * Avoid pin reprogramming. PRTs typically include entries - * with redundant pin->gsi mappings (but unique PCI devices); - * we only program the IOAPIC on the first. - */ - if (ioapic_pin > MP_MAX_IOAPIC_PIN) { - printk(KERN_ERR "Invalid reference to IOAPIC pin " - "%d-%d\n", mp_ioapic_routing[ioapic].apic_id, - ioapic_pin); - return gsi; + if (mpc_new_phys && mpc->mpc_length > mpc_new_length) { + mpc_new_phys = 0; + printk(KERN_INFO "mpc_new_length is %ld, please use alloc_mptable=8k\n", + mpc_new_length); } - if (test_bit(ioapic_pin, mp_ioapic_routing[ioapic].pin_programmed)) { - Dprintk(KERN_DEBUG "Pin %d-%d already programmed\n", - mp_ioapic_routing[ioapic].apic_id, ioapic_pin); -#ifdef CONFIG_X86_32 - return (gsi < IRQ_COMPRESSION_START ? gsi : gsi_to_irq[gsi]); -#else - return gsi; -#endif + + if (!mpc_new_phys) { + unsigned char old, new; + /* check if we can change the postion */ + mpc->mpc_checksum = 0; + old = mpf_checksum((unsigned char *)mpc, mpc->mpc_length); + mpc->mpc_checksum = 0xff; + new = mpf_checksum((unsigned char *)mpc, mpc->mpc_length); + if (old == new) { + printk(KERN_INFO "mpc is readonly, please try alloc_mptable instead\n"); + return 0; + } + printk(KERN_INFO "use in-positon replacing\n"); + } else { + mpf->mpf_physptr = mpc_new_phys; + mpc_new = phys_to_virt(mpc_new_phys); + memcpy(mpc_new, mpc, mpc->mpc_length); + mpc = mpc_new; + /* check if we can modify that */ + if (mpc_new_phys - mpf->mpf_physptr) { + struct intel_mp_floating *mpf_new; + /* steal 16 bytes from [0, 1k) */ + printk(KERN_INFO "mpf new: %x\n", 0x400 - 16); + mpf_new = phys_to_virt(0x400 - 16); + memcpy(mpf_new, mpf, 16); + mpf = mpf_new; + mpf->mpf_physptr = mpc_new_phys; + } + mpf->mpf_checksum = 0; + mpf->mpf_checksum -= mpf_checksum((unsigned char *)mpf, 16); + printk(KERN_INFO "mpf_physptr new: %x\n", mpf->mpf_physptr); } - set_bit(ioapic_pin, mp_ioapic_routing[ioapic].pin_programmed); -#ifdef CONFIG_X86_32 /* - * For GSI >= 64, use IRQ compression + * only replace the one with mp_INT and + * MP_IRQ_TRIGGER_LEVEL|MP_IRQ_POLARITY_LOW, + * already in mp_irqs , stored by ... and mp_config_acpi_gsi, + * may need pci=routeirq for all coverage */ - if ((gsi >= IRQ_COMPRESSION_START) - && (triggering == ACPI_LEVEL_SENSITIVE)) { - /* - * For PCI devices assign IRQs in order, avoiding gaps - * due to unused I/O APIC pins. - */ - int irq = gsi; - if (gsi < MAX_GSI_NUM) { - /* - * Retain the VIA chipset work-around (gsi > 15), but - * avoid a problem where the 8254 timer (IRQ0) is setup - * via an override (so it's not on pin 0 of the ioapic), - * and at the same time, the pin 0 interrupt is a PCI - * type. The gsi > 15 test could cause these two pins - * to be shared as IRQ0, and they are not shareable. - * So test for this condition, and if necessary, avoid - * the pin collision. - */ - gsi = pci_irq++; - /* - * Don't assign IRQ used by ACPI SCI - */ - if (gsi == acpi_gbl_FADT.sci_interrupt) - gsi = pci_irq++; - gsi_to_irq[irq] = gsi; - } else { - printk(KERN_ERR "GSI %u is too high\n", gsi); - return gsi; - } - } -#endif - io_apic_set_pci_routing(ioapic, ioapic_pin, gsi, - triggering == ACPI_EDGE_SENSITIVE ? 0 : 1, - polarity == ACPI_ACTIVE_HIGH ? 0 : 1); - return gsi; + replace_intsrc_all(mpc, mpc_new_phys, mpc_new_length); + + return 0; } -#endif /* CONFIG_X86_IO_APIC */ -#endif /* CONFIG_ACPI */ +late_initcall(update_mp_table); diff --git a/arch/x86/kernel/nmi_32.c b/arch/x86/kernel/nmi.c index 84160f74eeb..716b89284be 100644 --- a/arch/x86/kernel/nmi_32.c +++ b/arch/x86/kernel/nmi.c @@ -11,10 +11,13 @@ * Mikael Pettersson : PM converted to driver model. Disable/enable API. */ +#include <asm/apic.h> + +#include <linux/nmi.h> +#include <linux/mm.h> #include <linux/delay.h> #include <linux/interrupt.h> #include <linux/module.h> -#include <linux/nmi.h> #include <linux/sysdev.h> #include <linux/sysctl.h> #include <linux/percpu.h> @@ -22,12 +25,18 @@ #include <linux/cpumask.h> #include <linux/kernel_stat.h> #include <linux/kdebug.h> -#include <linux/slab.h> +#include <linux/smp.h> +#include <asm/i8259.h> +#include <asm/io_apic.h> #include <asm/smp.h> #include <asm/nmi.h> +#include <asm/proto.h> +#include <asm/timer.h> -#include "mach_traps.h" +#include <asm/mce.h> + +#include <mach_traps.h> int unknown_nmi_panic; int nmi_watchdog_enabled; @@ -41,28 +50,65 @@ static cpumask_t backtrace_mask = CPU_MASK_NONE; * 0: the lapic NMI watchdog is disabled, but can be enabled */ atomic_t nmi_active = ATOMIC_INIT(0); /* oprofile uses this */ +EXPORT_SYMBOL(nmi_active); -unsigned int nmi_watchdog = NMI_DEFAULT; -static unsigned int nmi_hz = HZ; +unsigned int nmi_watchdog = NMI_NONE; +EXPORT_SYMBOL(nmi_watchdog); +static int panic_on_timeout; + +static unsigned int nmi_hz = HZ; static DEFINE_PER_CPU(short, wd_enabled); +static int endflag __initdata; -static int endflag __initdata = 0; +static inline unsigned int get_nmi_count(int cpu) +{ +#ifdef CONFIG_X86_64 + return cpu_pda(cpu)->__nmi_count; +#else + return nmi_count(cpu); +#endif +} + +static inline int mce_in_progress(void) +{ +#if defined(CONFIG_X86_64) && defined(CONFIG_X86_MCE) + return atomic_read(&mce_entry) > 0; +#endif + return 0; +} + +/* + * Take the local apic timer and PIT/HPET into account. We don't + * know which one is active, when we have highres/dyntick on + */ +static inline unsigned int get_timer_irqs(int cpu) +{ +#ifdef CONFIG_X86_64 + return read_pda(apic_timer_irqs) + read_pda(irq0_irqs); +#else + return per_cpu(irq_stat, cpu).apic_timer_irqs + + per_cpu(irq_stat, cpu).irq0_irqs; +#endif +} #ifdef CONFIG_SMP -/* The performance counters used by NMI_LOCAL_APIC don't trigger when +/* + * The performance counters used by NMI_LOCAL_APIC don't trigger when * the CPU is idle. To make sure the NMI watchdog really ticks on all * CPUs during the test make them busy. */ static __init void nmi_cpu_busy(void *data) { local_irq_enable_in_hardirq(); - /* Intentionally don't use cpu_relax here. This is - to make sure that the performance counter really ticks, - even if there is a simulator or similar that catches the - pause instruction. On a real HT machine this is fine because - all other CPUs are busy with "useless" delay loops and don't - care if they get somewhat less cycles. */ + /* + * Intentionally don't use cpu_relax here. This is + * to make sure that the performance counter really ticks, + * even if there is a simulator or similar that catches the + * pause instruction. On a real HT machine this is fine because + * all other CPUs are busy with "useless" delay loops and don't + * care if they get somewhat less cycles. + */ while (endflag == 0) mb(); } @@ -73,15 +119,12 @@ int __init check_nmi_watchdog(void) unsigned int *prev_nmi_count; int cpu; - if ((nmi_watchdog == NMI_NONE) || (nmi_watchdog == NMI_DISABLED)) - return 0; - - if (!atomic_read(&nmi_active)) + if (!nmi_watchdog_active() || !atomic_read(&nmi_active)) return 0; - prev_nmi_count = kmalloc(NR_CPUS * sizeof(int), GFP_KERNEL); + prev_nmi_count = kmalloc(nr_cpu_ids * sizeof(int), GFP_KERNEL); if (!prev_nmi_count) - return -1; + goto error; printk(KERN_INFO "Testing NMI watchdog ... "); @@ -91,25 +134,19 @@ int __init check_nmi_watchdog(void) #endif for_each_possible_cpu(cpu) - prev_nmi_count[cpu] = nmi_count(cpu); + prev_nmi_count[cpu] = get_nmi_count(cpu); local_irq_enable(); - mdelay((20*1000)/nmi_hz); // wait 20 ticks + mdelay((20 * 1000) / nmi_hz); /* wait 20 ticks */ - for_each_possible_cpu(cpu) { -#ifdef CONFIG_SMP - /* Check cpu_callin_map here because that is set - after the timer is started. */ - if (!cpu_isset(cpu, cpu_callin_map)) - continue; -#endif + for_each_online_cpu(cpu) { if (!per_cpu(wd_enabled, cpu)) continue; - if (nmi_count(cpu) - prev_nmi_count[cpu] <= 5) { + if (get_nmi_count(cpu) - prev_nmi_count[cpu] <= 5) { printk(KERN_WARNING "WARNING: CPU#%d: NMI " "appears to be stuck (%d->%d)!\n", cpu, prev_nmi_count[cpu], - nmi_count(cpu)); + get_nmi_count(cpu)); per_cpu(wd_enabled, cpu) = 0; atomic_dec(&nmi_active); } @@ -118,37 +155,53 @@ int __init check_nmi_watchdog(void) if (!atomic_read(&nmi_active)) { kfree(prev_nmi_count); atomic_set(&nmi_active, -1); - return -1; + goto error; } printk("OK.\n"); - /* now that we know it works we can reduce NMI frequency to - something more reasonable; makes a difference in some configs */ + /* + * now that we know it works we can reduce NMI frequency to + * something more reasonable; makes a difference in some configs + */ if (nmi_watchdog == NMI_LOCAL_APIC) nmi_hz = lapic_adjust_nmi_hz(1); kfree(prev_nmi_count); return 0; +error: + if (nmi_watchdog == NMI_IO_APIC && !timer_through_8259) + disable_8259A_irq(0); +#ifdef CONFIG_X86_32 + timer_ack = 0; +#endif + return -1; } static int __init setup_nmi_watchdog(char *str) { - int nmi; + unsigned int nmi; + + if (!strncmp(str, "panic", 5)) { + panic_on_timeout = 1; + str = strchr(str, ','); + if (!str) + return 1; + ++str; + } get_option(&str, &nmi); - if ((nmi >= NMI_INVALID) || (nmi < NMI_NONE)) + if (nmi >= NMI_INVALID) return 0; nmi_watchdog = nmi; return 1; } - __setup("nmi_watchdog=", setup_nmi_watchdog); - -/* Suspend/resume support */ - +/* + * Suspend/resume support + */ #ifdef CONFIG_PM static int nmi_pm_active; /* nmi_active before suspend */ @@ -172,7 +225,6 @@ static int lapic_nmi_resume(struct sys_device *dev) return 0; } - static struct sysdev_class nmi_sysclass = { .name = "lapic_nmi", .resume = lapic_nmi_resume, @@ -188,7 +240,8 @@ static int __init init_lapic_nmi_sysfs(void) { int error; - /* should really be a BUG_ON but b/c this is an + /* + * should really be a BUG_ON but b/c this is an * init call, it just doesn't work. -dcz */ if (nmi_watchdog != NMI_LOCAL_APIC) @@ -202,6 +255,7 @@ static int __init init_lapic_nmi_sysfs(void) error = sysdev_register(&device_lapic_nmi); return error; } + /* must come after the local APIC's device_initcall() */ late_initcall(init_lapic_nmi_sysfs); @@ -223,7 +277,7 @@ void acpi_nmi_enable(void) static void __acpi_nmi_disable(void *__unused) { - apic_write(APIC_LVT0, APIC_DM_NMI | APIC_LVT_MASKED); + apic_write_around(APIC_LVT0, APIC_DM_NMI | APIC_LVT_MASKED); } /* @@ -242,12 +296,13 @@ void setup_apic_nmi_watchdog(void *unused) /* cheap hack to support suspend/resume */ /* if cpu0 is not active neither should the other cpus */ - if ((smp_processor_id() != 0) && (atomic_read(&nmi_active) <= 0)) + if (smp_processor_id() != 0 && atomic_read(&nmi_active) <= 0) return; switch (nmi_watchdog) { case NMI_LOCAL_APIC: - __get_cpu_var(wd_enabled) = 1; /* enable it before to avoid race with handler */ + /* enable it before to avoid race with handler */ + __get_cpu_var(wd_enabled) = 1; if (lapic_watchdog_init(nmi_hz) < 0) { __get_cpu_var(wd_enabled) = 0; return; @@ -262,9 +317,8 @@ void setup_apic_nmi_watchdog(void *unused) void stop_apic_nmi_watchdog(void *unused) { /* only support LOCAL and IO APICs for now */ - if ((nmi_watchdog != NMI_LOCAL_APIC) && - (nmi_watchdog != NMI_IO_APIC)) - return; + if (!nmi_watchdog_active()) + return; if (__get_cpu_var(wd_enabled) == 0) return; if (nmi_watchdog == NMI_LOCAL_APIC) @@ -284,26 +338,26 @@ void stop_apic_nmi_watchdog(void *unused) * since NMIs don't listen to _any_ locks, we have to be extremely * careful not to rely on unsafe variables. The printk might lock * up though, so we have to break up any console locks first ... - * [when there will be more tty-related locks, break them up - * here too!] + * [when there will be more tty-related locks, break them up here too!] */ -static unsigned int - last_irq_sums [NR_CPUS], - alert_counter [NR_CPUS]; +static DEFINE_PER_CPU(unsigned, last_irq_sum); +static DEFINE_PER_CPU(local_t, alert_counter); +static DEFINE_PER_CPU(int, nmi_touch); void touch_nmi_watchdog(void) { - if (nmi_watchdog > 0) { + if (nmi_watchdog_active()) { unsigned cpu; /* - * Just reset the alert counters, (other CPUs might be - * spinning on locks we hold): + * Tell other CPUs to reset their alert counters. We cannot + * do it ourselves because the alert count increase is not + * atomic. */ for_each_present_cpu(cpu) { - if (alert_counter[cpu]) - alert_counter[cpu] = 0; + if (per_cpu(nmi_touch, cpu) != 1) + per_cpu(nmi_touch, cpu) = 1; } } @@ -314,12 +368,9 @@ void touch_nmi_watchdog(void) } EXPORT_SYMBOL(touch_nmi_watchdog); -extern void die_nmi(struct pt_regs *, const char *msg); - notrace __kprobes int nmi_watchdog_tick(struct pt_regs *regs, unsigned reason) { - /* * Since current_thread_info()-> is always on the stack, and we * always switch the stack NMI-atomically, it's safe to use @@ -337,39 +388,45 @@ nmi_watchdog_tick(struct pt_regs *regs, unsigned reason) touched = 1; } + sum = get_timer_irqs(cpu); + + if (__get_cpu_var(nmi_touch)) { + __get_cpu_var(nmi_touch) = 0; + touched = 1; + } + if (cpu_isset(cpu, backtrace_mask)) { static DEFINE_SPINLOCK(lock); /* Serialise the printks */ spin_lock(&lock); - printk("NMI backtrace for cpu %d\n", cpu); + printk(KERN_WARNING "NMI backtrace for cpu %d\n", cpu); dump_stack(); spin_unlock(&lock); cpu_clear(cpu, backtrace_mask); } - /* - * Take the local apic timer and PIT/HPET into account. We don't - * know which one is active, when we have highres/dyntick on - */ - sum = per_cpu(irq_stat, cpu).apic_timer_irqs + - per_cpu(irq_stat, cpu).irq0_irqs; + /* Could check oops_in_progress here too, but it's safer not to */ + if (mce_in_progress()) + touched = 1; /* if the none of the timers isn't firing, this cpu isn't doing much */ - if (!touched && last_irq_sums[cpu] == sum) { + if (!touched && __get_cpu_var(last_irq_sum) == sum) { /* * Ayiee, looks like this CPU is stuck ... * wait a few IRQs (5 seconds) before doing the oops ... */ - alert_counter[cpu]++; - if (alert_counter[cpu] == 5*nmi_hz) + local_inc(&__get_cpu_var(alert_counter)); + if (local_read(&__get_cpu_var(alert_counter)) == 5 * nmi_hz) /* * die_nmi will return ONLY if NOTIFY_STOP happens.. */ - die_nmi(regs, "BUG: NMI Watchdog detected LOCKUP"); + die_nmi("BUG: NMI Watchdog detected LOCKUP", + regs, panic_on_timeout); } else { - last_irq_sums[cpu] = sum; - alert_counter[cpu] = 0; + __get_cpu_var(last_irq_sum) = sum; + local_set(&__get_cpu_var(alert_counter), 0); } + /* see if the nmi watchdog went off */ if (!__get_cpu_var(wd_enabled)) return rc; @@ -378,7 +435,8 @@ nmi_watchdog_tick(struct pt_regs *regs, unsigned reason) rc |= lapic_wd_event(nmi_hz); break; case NMI_IO_APIC: - /* don't know how to accurately check for this. + /* + * don't know how to accurately check for this. * just assume it was a watchdog timer interrupt * This matches the old behaviour. */ @@ -396,7 +454,7 @@ static int unknown_nmi_panic_callback(struct pt_regs *regs, int cpu) char buf[64]; sprintf(buf, "NMI received for unknown reason %02x\n", reason); - die_nmi(regs, buf); + die_nmi(buf, regs, 1); /* Always panic here */ return 0; } @@ -414,32 +472,26 @@ int proc_nmi_enabled(struct ctl_table *table, int write, struct file *file, if (!!old_state == !!nmi_watchdog_enabled) return 0; - if (atomic_read(&nmi_active) < 0 || nmi_watchdog == NMI_DISABLED) { - printk( KERN_WARNING "NMI watchdog is permanently disabled\n"); + if (atomic_read(&nmi_active) < 0 || !nmi_watchdog_active()) { + printk(KERN_WARNING + "NMI watchdog is permanently disabled\n"); return -EIO; } - if (nmi_watchdog == NMI_DEFAULT) { - if (lapic_watchdog_ok()) - nmi_watchdog = NMI_LOCAL_APIC; - else - nmi_watchdog = NMI_IO_APIC; - } - if (nmi_watchdog == NMI_LOCAL_APIC) { if (nmi_watchdog_enabled) enable_lapic_nmi_watchdog(); else disable_lapic_nmi_watchdog(); } else { - printk( KERN_WARNING + printk(KERN_WARNING "NMI watchdog doesn't know what hardware to touch\n"); return -EIO; } return 0; } -#endif +#endif /* CONFIG_SYSCTL */ int do_nmi_callback(struct pt_regs *regs, int cpu) { @@ -462,6 +514,3 @@ void __trigger_all_cpu_backtrace(void) mdelay(1); } } - -EXPORT_SYMBOL(nmi_active); -EXPORT_SYMBOL(nmi_watchdog); diff --git a/arch/x86/kernel/nmi_64.c b/arch/x86/kernel/nmi_64.c deleted file mode 100644 index 5a29ded994f..00000000000 --- a/arch/x86/kernel/nmi_64.c +++ /dev/null @@ -1,482 +0,0 @@ -/* - * NMI watchdog support on APIC systems - * - * Started by Ingo Molnar <mingo@redhat.com> - * - * Fixes: - * Mikael Pettersson : AMD K7 support for local APIC NMI watchdog. - * Mikael Pettersson : Power Management for local APIC NMI watchdog. - * Pavel Machek and - * Mikael Pettersson : PM converted to driver model. Disable/enable API. - */ - -#include <linux/nmi.h> -#include <linux/mm.h> -#include <linux/delay.h> -#include <linux/interrupt.h> -#include <linux/module.h> -#include <linux/sysdev.h> -#include <linux/sysctl.h> -#include <linux/kprobes.h> -#include <linux/cpumask.h> -#include <linux/kdebug.h> - -#include <asm/smp.h> -#include <asm/nmi.h> -#include <asm/proto.h> -#include <asm/mce.h> - -#include <mach_traps.h> - -int unknown_nmi_panic; -int nmi_watchdog_enabled; -int panic_on_unrecovered_nmi; - -static cpumask_t backtrace_mask = CPU_MASK_NONE; - -/* nmi_active: - * >0: the lapic NMI watchdog is active, but can be disabled - * <0: the lapic NMI watchdog has not been set up, and cannot - * be enabled - * 0: the lapic NMI watchdog is disabled, but can be enabled - */ -atomic_t nmi_active = ATOMIC_INIT(0); /* oprofile uses this */ -static int panic_on_timeout; - -unsigned int nmi_watchdog = NMI_DEFAULT; -static unsigned int nmi_hz = HZ; - -static DEFINE_PER_CPU(short, wd_enabled); - -/* Run after command line and cpu_init init, but before all other checks */ -void nmi_watchdog_default(void) -{ - if (nmi_watchdog != NMI_DEFAULT) - return; - nmi_watchdog = NMI_NONE; -} - -static int endflag __initdata = 0; - -#ifdef CONFIG_SMP -/* The performance counters used by NMI_LOCAL_APIC don't trigger when - * the CPU is idle. To make sure the NMI watchdog really ticks on all - * CPUs during the test make them busy. - */ -static __init void nmi_cpu_busy(void *data) -{ - local_irq_enable_in_hardirq(); - /* Intentionally don't use cpu_relax here. This is - to make sure that the performance counter really ticks, - even if there is a simulator or similar that catches the - pause instruction. On a real HT machine this is fine because - all other CPUs are busy with "useless" delay loops and don't - care if they get somewhat less cycles. */ - while (endflag == 0) - mb(); -} -#endif - -int __init check_nmi_watchdog(void) -{ - int *prev_nmi_count; - int cpu; - - if ((nmi_watchdog == NMI_NONE) || (nmi_watchdog == NMI_DISABLED)) - return 0; - - if (!atomic_read(&nmi_active)) - return 0; - - prev_nmi_count = kmalloc(NR_CPUS * sizeof(int), GFP_KERNEL); - if (!prev_nmi_count) - return -1; - - printk(KERN_INFO "Testing NMI watchdog ... "); - -#ifdef CONFIG_SMP - if (nmi_watchdog == NMI_LOCAL_APIC) - smp_call_function(nmi_cpu_busy, (void *)&endflag, 0, 0); -#endif - - for (cpu = 0; cpu < NR_CPUS; cpu++) - prev_nmi_count[cpu] = cpu_pda(cpu)->__nmi_count; - local_irq_enable(); - mdelay((20*1000)/nmi_hz); // wait 20 ticks - - for_each_online_cpu(cpu) { - if (!per_cpu(wd_enabled, cpu)) - continue; - if (cpu_pda(cpu)->__nmi_count - prev_nmi_count[cpu] <= 5) { - printk(KERN_WARNING "WARNING: CPU#%d: NMI " - "appears to be stuck (%d->%d)!\n", - cpu, - prev_nmi_count[cpu], - cpu_pda(cpu)->__nmi_count); - per_cpu(wd_enabled, cpu) = 0; - atomic_dec(&nmi_active); - } - } - endflag = 1; - if (!atomic_read(&nmi_active)) { - kfree(prev_nmi_count); - atomic_set(&nmi_active, -1); - return -1; - } - printk("OK.\n"); - - /* now that we know it works we can reduce NMI frequency to - something more reasonable; makes a difference in some configs */ - if (nmi_watchdog == NMI_LOCAL_APIC) - nmi_hz = lapic_adjust_nmi_hz(1); - - kfree(prev_nmi_count); - return 0; -} - -static int __init setup_nmi_watchdog(char *str) -{ - int nmi; - - if (!strncmp(str,"panic",5)) { - panic_on_timeout = 1; - str = strchr(str, ','); - if (!str) - return 1; - ++str; - } - - get_option(&str, &nmi); - - if ((nmi >= NMI_INVALID) || (nmi < NMI_NONE)) - return 0; - - nmi_watchdog = nmi; - return 1; -} - -__setup("nmi_watchdog=", setup_nmi_watchdog); - -#ifdef CONFIG_PM - -static int nmi_pm_active; /* nmi_active before suspend */ - -static int lapic_nmi_suspend(struct sys_device *dev, pm_message_t state) -{ - /* only CPU0 goes here, other CPUs should be offline */ - nmi_pm_active = atomic_read(&nmi_active); - stop_apic_nmi_watchdog(NULL); - BUG_ON(atomic_read(&nmi_active) != 0); - return 0; -} - -static int lapic_nmi_resume(struct sys_device *dev) -{ - /* only CPU0 goes here, other CPUs should be offline */ - if (nmi_pm_active > 0) { - setup_apic_nmi_watchdog(NULL); - touch_nmi_watchdog(); - } - return 0; -} - -static struct sysdev_class nmi_sysclass = { - .name = "lapic_nmi", - .resume = lapic_nmi_resume, - .suspend = lapic_nmi_suspend, -}; - -static struct sys_device device_lapic_nmi = { - .id = 0, - .cls = &nmi_sysclass, -}; - -static int __init init_lapic_nmi_sysfs(void) -{ - int error; - - /* should really be a BUG_ON but b/c this is an - * init call, it just doesn't work. -dcz - */ - if (nmi_watchdog != NMI_LOCAL_APIC) - return 0; - - if (atomic_read(&nmi_active) < 0) - return 0; - - error = sysdev_class_register(&nmi_sysclass); - if (!error) - error = sysdev_register(&device_lapic_nmi); - return error; -} -/* must come after the local APIC's device_initcall() */ -late_initcall(init_lapic_nmi_sysfs); - -#endif /* CONFIG_PM */ - -static void __acpi_nmi_enable(void *__unused) -{ - apic_write(APIC_LVT0, APIC_DM_NMI); -} - -/* - * Enable timer based NMIs on all CPUs: - */ -void acpi_nmi_enable(void) -{ - if (atomic_read(&nmi_active) && nmi_watchdog == NMI_IO_APIC) - on_each_cpu(__acpi_nmi_enable, NULL, 0, 1); -} - -static void __acpi_nmi_disable(void *__unused) -{ - apic_write(APIC_LVT0, APIC_DM_NMI | APIC_LVT_MASKED); -} - -/* - * Disable timer based NMIs on all CPUs: - */ -void acpi_nmi_disable(void) -{ - if (atomic_read(&nmi_active) && nmi_watchdog == NMI_IO_APIC) - on_each_cpu(__acpi_nmi_disable, NULL, 0, 1); -} - -void setup_apic_nmi_watchdog(void *unused) -{ - if (__get_cpu_var(wd_enabled)) - return; - - /* cheap hack to support suspend/resume */ - /* if cpu0 is not active neither should the other cpus */ - if ((smp_processor_id() != 0) && (atomic_read(&nmi_active) <= 0)) - return; - - switch (nmi_watchdog) { - case NMI_LOCAL_APIC: - __get_cpu_var(wd_enabled) = 1; - if (lapic_watchdog_init(nmi_hz) < 0) { - __get_cpu_var(wd_enabled) = 0; - return; - } - /* FALL THROUGH */ - case NMI_IO_APIC: - __get_cpu_var(wd_enabled) = 1; - atomic_inc(&nmi_active); - } -} - -void stop_apic_nmi_watchdog(void *unused) -{ - /* only support LOCAL and IO APICs for now */ - if ((nmi_watchdog != NMI_LOCAL_APIC) && - (nmi_watchdog != NMI_IO_APIC)) - return; - if (__get_cpu_var(wd_enabled) == 0) - return; - if (nmi_watchdog == NMI_LOCAL_APIC) - lapic_watchdog_stop(); - __get_cpu_var(wd_enabled) = 0; - atomic_dec(&nmi_active); -} - -/* - * the best way to detect whether a CPU has a 'hard lockup' problem - * is to check it's local APIC timer IRQ counts. If they are not - * changing then that CPU has some problem. - * - * as these watchdog NMI IRQs are generated on every CPU, we only - * have to check the current processor. - */ - -static DEFINE_PER_CPU(unsigned, last_irq_sum); -static DEFINE_PER_CPU(local_t, alert_counter); -static DEFINE_PER_CPU(int, nmi_touch); - -void touch_nmi_watchdog(void) -{ - if (nmi_watchdog > 0) { - unsigned cpu; - - /* - * Tell other CPUs to reset their alert counters. We cannot - * do it ourselves because the alert count increase is not - * atomic. - */ - for_each_present_cpu(cpu) { - if (per_cpu(nmi_touch, cpu) != 1) - per_cpu(nmi_touch, cpu) = 1; - } - } - - touch_softlockup_watchdog(); -} -EXPORT_SYMBOL(touch_nmi_watchdog); - -notrace __kprobes int -nmi_watchdog_tick(struct pt_regs *regs, unsigned reason) -{ - int sum; - int touched = 0; - int cpu = smp_processor_id(); - int rc = 0; - - /* check for other users first */ - if (notify_die(DIE_NMI, "nmi", regs, reason, 2, SIGINT) - == NOTIFY_STOP) { - rc = 1; - touched = 1; - } - - sum = read_pda(apic_timer_irqs) + read_pda(irq0_irqs); - if (__get_cpu_var(nmi_touch)) { - __get_cpu_var(nmi_touch) = 0; - touched = 1; - } - - if (cpu_isset(cpu, backtrace_mask)) { - static DEFINE_SPINLOCK(lock); /* Serialise the printks */ - - spin_lock(&lock); - printk("NMI backtrace for cpu %d\n", cpu); - dump_stack(); - spin_unlock(&lock); - cpu_clear(cpu, backtrace_mask); - } - -#ifdef CONFIG_X86_MCE - /* Could check oops_in_progress here too, but it's safer - not too */ - if (atomic_read(&mce_entry) > 0) - touched = 1; -#endif - /* if the apic timer isn't firing, this cpu isn't doing much */ - if (!touched && __get_cpu_var(last_irq_sum) == sum) { - /* - * Ayiee, looks like this CPU is stuck ... - * wait a few IRQs (5 seconds) before doing the oops ... - */ - local_inc(&__get_cpu_var(alert_counter)); - if (local_read(&__get_cpu_var(alert_counter)) == 5*nmi_hz) - die_nmi("NMI Watchdog detected LOCKUP on CPU %d\n", regs, - panic_on_timeout); - } else { - __get_cpu_var(last_irq_sum) = sum; - local_set(&__get_cpu_var(alert_counter), 0); - } - - /* see if the nmi watchdog went off */ - if (!__get_cpu_var(wd_enabled)) - return rc; - switch (nmi_watchdog) { - case NMI_LOCAL_APIC: - rc |= lapic_wd_event(nmi_hz); - break; - case NMI_IO_APIC: - /* don't know how to accurately check for this. - * just assume it was a watchdog timer interrupt - * This matches the old behaviour. - */ - rc = 1; - break; - } - return rc; -} - -static unsigned ignore_nmis; - -asmlinkage notrace __kprobes void -do_nmi(struct pt_regs *regs, long error_code) -{ - nmi_enter(); - add_pda(__nmi_count,1); - if (!ignore_nmis) - default_do_nmi(regs); - nmi_exit(); -} - -void stop_nmi(void) -{ - acpi_nmi_disable(); - ignore_nmis++; -} - -void restart_nmi(void) -{ - ignore_nmis--; - acpi_nmi_enable(); -} - -#ifdef CONFIG_SYSCTL - -static int unknown_nmi_panic_callback(struct pt_regs *regs, int cpu) -{ - unsigned char reason = get_nmi_reason(); - char buf[64]; - - sprintf(buf, "NMI received for unknown reason %02x\n", reason); - die_nmi(buf, regs, 1); /* Always panic here */ - return 0; -} - -/* - * proc handler for /proc/sys/kernel/nmi - */ -int proc_nmi_enabled(struct ctl_table *table, int write, struct file *file, - void __user *buffer, size_t *length, loff_t *ppos) -{ - int old_state; - - nmi_watchdog_enabled = (atomic_read(&nmi_active) > 0) ? 1 : 0; - old_state = nmi_watchdog_enabled; - proc_dointvec(table, write, file, buffer, length, ppos); - if (!!old_state == !!nmi_watchdog_enabled) - return 0; - - if (atomic_read(&nmi_active) < 0 || nmi_watchdog == NMI_DISABLED) { - printk( KERN_WARNING "NMI watchdog is permanently disabled\n"); - return -EIO; - } - - /* if nmi_watchdog is not set yet, then set it */ - nmi_watchdog_default(); - - if (nmi_watchdog == NMI_LOCAL_APIC) { - if (nmi_watchdog_enabled) - enable_lapic_nmi_watchdog(); - else - disable_lapic_nmi_watchdog(); - } else { - printk( KERN_WARNING - "NMI watchdog doesn't know what hardware to touch\n"); - return -EIO; - } - return 0; -} - -#endif - -int do_nmi_callback(struct pt_regs *regs, int cpu) -{ -#ifdef CONFIG_SYSCTL - if (unknown_nmi_panic) - return unknown_nmi_panic_callback(regs, cpu); -#endif - return 0; -} - -void __trigger_all_cpu_backtrace(void) -{ - int i; - - backtrace_mask = cpu_online_map; - /* Wait for up to 10 seconds for all CPUs to do the backtrace */ - for (i = 0; i < 10 * 1000; i++) { - if (cpus_empty(backtrace_mask)) - break; - mdelay(1); - } -} - -EXPORT_SYMBOL(nmi_active); -EXPORT_SYMBOL(nmi_watchdog); diff --git a/arch/x86/kernel/numaq_32.c b/arch/x86/kernel/numaq_32.c index e65281b1634..5b20a5e7ac2 100644 --- a/arch/x86/kernel/numaq_32.c +++ b/arch/x86/kernel/numaq_32.c @@ -31,6 +31,8 @@ #include <asm/numaq.h> #include <asm/topology.h> #include <asm/processor.h> +#include <asm/mpspec.h> +#include <asm/e820.h> #define MB_TO_PAGES(addr) ((addr) << (20 - PAGE_SHIFT)) @@ -58,6 +60,8 @@ static void __init smp_dump_qct(void) node_end_pfn[node] = MB_TO_PAGES( eq->hi_shrd_mem_start + eq->hi_shrd_mem_size); + e820_register_active_regions(node, node_start_pfn[node], + node_end_pfn[node]); memory_present(node, node_start_pfn[node], node_end_pfn[node]); node_remap_size[node] = node_memmap_size_bytes(node, @@ -67,23 +71,35 @@ static void __init smp_dump_qct(void) } } -/* - * Unlike Summit, we don't really care to let the NUMA-Q - * fall back to flat mode. Don't compile for NUMA-Q - * unless you really need it! - */ +static __init void early_check_numaq(void) +{ + /* + * Find possible boot-time SMP configuration: + */ + early_find_smp_config(); + /* + * get boot-time SMP configuration: + */ + if (smp_found_config) + early_get_smp_config(); +} + int __init get_memcfg_numaq(void) { + early_check_numaq(); + if (!found_numaq) + return 0; smp_dump_qct(); return 1; } -static int __init numaq_tsc_disable(void) +void __init numaq_tsc_disable(void) { + if (!found_numaq) + return -1; + if (num_online_nodes() > 1) { printk(KERN_DEBUG "NUMAQ: disabling TSC\n"); setup_clear_cpu_cap(X86_FEATURE_TSC); } - return 0; } -arch_initcall(numaq_tsc_disable); diff --git a/arch/x86/kernel/paravirt.c b/arch/x86/kernel/paravirt.c index 74f0c5ea2a0..e0f571d58c1 100644 --- a/arch/x86/kernel/paravirt.c +++ b/arch/x86/kernel/paravirt.c @@ -30,6 +30,7 @@ #include <asm/setup.h> #include <asm/arch_hooks.h> #include <asm/time.h> +#include <asm/pgalloc.h> #include <asm/irq.h> #include <asm/delay.h> #include <asm/fixmap.h> @@ -139,7 +140,9 @@ unsigned paravirt_patch_default(u8 type, u16 clobbers, void *insnbuf, /* If the operation is a nop, then nop the callsite */ ret = paravirt_patch_nop(); else if (type == PARAVIRT_PATCH(pv_cpu_ops.iret) || - type == PARAVIRT_PATCH(pv_cpu_ops.irq_enable_syscall_ret)) + type == PARAVIRT_PATCH(pv_cpu_ops.irq_enable_sysexit) || + type == PARAVIRT_PATCH(pv_cpu_ops.usergs_sysret32) || + type == PARAVIRT_PATCH(pv_cpu_ops.usergs_sysret64)) /* If operation requires a jmp, then jmp */ ret = paravirt_patch_jmp(insnbuf, opfunc, addr, len); else @@ -190,7 +193,9 @@ static void native_flush_tlb_single(unsigned long addr) /* These are in entry.S */ extern void native_iret(void); -extern void native_irq_enable_syscall_ret(void); +extern void native_irq_enable_sysexit(void); +extern void native_usergs_sysret32(void); +extern void native_usergs_sysret64(void); static int __init print_banner(void) { @@ -280,7 +285,7 @@ struct pv_time_ops pv_time_ops = { .get_wallclock = native_get_wallclock, .set_wallclock = native_set_wallclock, .sched_clock = native_sched_clock, - .get_cpu_khz = native_calculate_cpu_khz, + .get_tsc_khz = native_calibrate_tsc, }; struct pv_irq_ops pv_irq_ops = { @@ -291,6 +296,9 @@ struct pv_irq_ops pv_irq_ops = { .irq_enable = native_irq_enable, .safe_halt = native_safe_halt, .halt = native_halt, +#ifdef CONFIG_X86_64 + .adjust_exception_frame = paravirt_nop, +#endif }; struct pv_cpu_ops pv_cpu_ops = { @@ -321,12 +329,23 @@ struct pv_cpu_ops pv_cpu_ops = { .store_idt = native_store_idt, .store_tr = native_store_tr, .load_tls = native_load_tls, +#ifdef CONFIG_X86_64 + .load_gs_index = native_load_gs_index, +#endif .write_ldt_entry = native_write_ldt_entry, .write_gdt_entry = native_write_gdt_entry, .write_idt_entry = native_write_idt_entry, .load_sp0 = native_load_sp0, - .irq_enable_syscall_ret = native_irq_enable_syscall_ret, +#if defined(CONFIG_X86_32) || defined(CONFIG_IA32_EMULATION) + .irq_enable_sysexit = native_irq_enable_sysexit, +#endif +#ifdef CONFIG_X86_64 +#ifdef CONFIG_IA32_EMULATION + .usergs_sysret32 = native_usergs_sysret32, +#endif + .usergs_sysret64 = native_usergs_sysret64, +#endif .iret = native_iret, .swapgs = native_swapgs, @@ -366,6 +385,9 @@ struct pv_mmu_ops pv_mmu_ops = { .flush_tlb_single = native_flush_tlb_single, .flush_tlb_others = native_flush_tlb_others, + .pgd_alloc = __paravirt_pgd_alloc, + .pgd_free = paravirt_nop, + .alloc_pte = paravirt_nop, .alloc_pmd = paravirt_nop, .alloc_pmd_clone = paravirt_nop, @@ -380,6 +402,9 @@ struct pv_mmu_ops pv_mmu_ops = { .pte_update = paravirt_nop, .pte_update_defer = paravirt_nop, + .ptep_modify_prot_start = __ptep_modify_prot_start, + .ptep_modify_prot_commit = __ptep_modify_prot_commit, + #ifdef CONFIG_HIGHPTE .kmap_atomic_pte = kmap_atomic, #endif @@ -403,6 +428,7 @@ struct pv_mmu_ops pv_mmu_ops = { #endif /* PAGETABLE_LEVELS >= 3 */ .pte_val = native_pte_val, + .pte_flags = native_pte_val, .pgd_val = native_pgd_val, .make_pte = native_make_pte, @@ -416,6 +442,8 @@ struct pv_mmu_ops pv_mmu_ops = { .enter = paravirt_nop, .leave = paravirt_nop, }, + + .set_fixmap = native_set_fixmap, }; EXPORT_SYMBOL_GPL(pv_time_ops); diff --git a/arch/x86/kernel/paravirt_patch_32.c b/arch/x86/kernel/paravirt_patch_32.c index 82fc5fcab4f..58262218781 100644 --- a/arch/x86/kernel/paravirt_patch_32.c +++ b/arch/x86/kernel/paravirt_patch_32.c @@ -5,7 +5,7 @@ DEF_NATIVE(pv_irq_ops, irq_enable, "sti"); DEF_NATIVE(pv_irq_ops, restore_fl, "push %eax; popf"); DEF_NATIVE(pv_irq_ops, save_fl, "pushf; pop %eax"); DEF_NATIVE(pv_cpu_ops, iret, "iret"); -DEF_NATIVE(pv_cpu_ops, irq_enable_syscall_ret, "sti; sysexit"); +DEF_NATIVE(pv_cpu_ops, irq_enable_sysexit, "sti; sysexit"); DEF_NATIVE(pv_mmu_ops, read_cr2, "mov %cr2, %eax"); DEF_NATIVE(pv_mmu_ops, write_cr3, "mov %eax, %cr3"); DEF_NATIVE(pv_mmu_ops, read_cr3, "mov %cr3, %eax"); @@ -29,7 +29,7 @@ unsigned native_patch(u8 type, u16 clobbers, void *ibuf, PATCH_SITE(pv_irq_ops, restore_fl); PATCH_SITE(pv_irq_ops, save_fl); PATCH_SITE(pv_cpu_ops, iret); - PATCH_SITE(pv_cpu_ops, irq_enable_syscall_ret); + PATCH_SITE(pv_cpu_ops, irq_enable_sysexit); PATCH_SITE(pv_mmu_ops, read_cr2); PATCH_SITE(pv_mmu_ops, read_cr3); PATCH_SITE(pv_mmu_ops, write_cr3); diff --git a/arch/x86/kernel/paravirt_patch_64.c b/arch/x86/kernel/paravirt_patch_64.c index 7d904e138d7..061d01df9ae 100644 --- a/arch/x86/kernel/paravirt_patch_64.c +++ b/arch/x86/kernel/paravirt_patch_64.c @@ -14,8 +14,9 @@ DEF_NATIVE(pv_mmu_ops, flush_tlb_single, "invlpg (%rdi)"); DEF_NATIVE(pv_cpu_ops, clts, "clts"); DEF_NATIVE(pv_cpu_ops, wbinvd, "wbinvd"); -/* the three commands give us more control to how to return from a syscall */ -DEF_NATIVE(pv_cpu_ops, irq_enable_syscall_ret, "movq %gs:" __stringify(pda_oldrsp) ", %rsp; swapgs; sysretq;"); +DEF_NATIVE(pv_cpu_ops, irq_enable_sysexit, "swapgs; sti; sysexit"); +DEF_NATIVE(pv_cpu_ops, usergs_sysret64, "swapgs; sysretq"); +DEF_NATIVE(pv_cpu_ops, usergs_sysret32, "swapgs; sysretl"); DEF_NATIVE(pv_cpu_ops, swapgs, "swapgs"); unsigned native_patch(u8 type, u16 clobbers, void *ibuf, @@ -35,7 +36,9 @@ unsigned native_patch(u8 type, u16 clobbers, void *ibuf, 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_syscall_ret); + PATCH_SITE(pv_cpu_ops, irq_enable_sysexit); + PATCH_SITE(pv_cpu_ops, usergs_sysret32); + PATCH_SITE(pv_cpu_ops, usergs_sysret64); PATCH_SITE(pv_cpu_ops, swapgs); PATCH_SITE(pv_mmu_ops, read_cr2); PATCH_SITE(pv_mmu_ops, read_cr3); diff --git a/arch/x86/kernel/pci-calgary_64.c b/arch/x86/kernel/pci-calgary_64.c index e28ec497e14..6959b5c45df 100644 --- a/arch/x86/kernel/pci-calgary_64.c +++ b/arch/x86/kernel/pci-calgary_64.c @@ -1394,7 +1394,7 @@ void __init detect_calgary(void) return; } - specified_table_size = determine_tce_table_size(end_pfn * PAGE_SIZE); + specified_table_size = determine_tce_table_size(max_pfn * PAGE_SIZE); for (bus = 0; bus < MAX_PHB_BUS_NUM; bus++) { struct calgary_bus_info *info = &bus_info[bus]; @@ -1459,7 +1459,7 @@ int __init calgary_iommu_init(void) if (ret) { printk(KERN_ERR "PCI-DMA: Calgary init failed %d, " "falling back to no_iommu\n", ret); - if (end_pfn > MAX_DMA32_PFN) + if (max_pfn > MAX_DMA32_PFN) printk(KERN_ERR "WARNING more than 4GB of memory, " "32bit PCI may malfunction.\n"); return ret; diff --git a/arch/x86/kernel/pci-dma.c b/arch/x86/kernel/pci-dma.c index dc00a1331ac..8467ec2320f 100644 --- a/arch/x86/kernel/pci-dma.c +++ b/arch/x86/kernel/pci-dma.c @@ -7,6 +7,7 @@ #include <asm/dma.h> #include <asm/gart.h> #include <asm/calgary.h> +#include <asm/amd_iommu.h> int forbid_dac __read_mostly; EXPORT_SYMBOL(forbid_dac); @@ -74,13 +75,17 @@ early_param("dma32_size", parse_dma32_size_opt); void __init dma32_reserve_bootmem(void) { unsigned long size, align; - if (end_pfn <= MAX_DMA32_PFN) + if (max_pfn <= MAX_DMA32_PFN) return; + /* + * check aperture_64.c allocate_aperture() for reason about + * using 512M as goal + */ align = 64ULL<<20; size = round_up(dma32_bootmem_size, align); dma32_bootmem_ptr = __alloc_bootmem_nopanic(size, align, - __pa(MAX_DMA_ADDRESS)); + 512ULL<<20); if (dma32_bootmem_ptr) dma32_bootmem_size = size; else @@ -88,17 +93,14 @@ void __init dma32_reserve_bootmem(void) } static void __init dma32_free_bootmem(void) { - int node; - if (end_pfn <= MAX_DMA32_PFN) + if (max_pfn <= MAX_DMA32_PFN) return; if (!dma32_bootmem_ptr) return; - for_each_online_node(node) - free_bootmem_node(NODE_DATA(node), __pa(dma32_bootmem_ptr), - dma32_bootmem_size); + free_bootmem(__pa(dma32_bootmem_ptr), dma32_bootmem_size); dma32_bootmem_ptr = NULL; dma32_bootmem_size = 0; @@ -122,6 +124,8 @@ void __init pci_iommu_alloc(void) detect_intel_iommu(); + amd_iommu_detect(); + #ifdef CONFIG_SWIOTLB pci_swiotlb_init(); #endif @@ -357,7 +361,7 @@ int dma_supported(struct device *dev, u64 mask) EXPORT_SYMBOL(dma_supported); /* Allocate DMA memory on node near device */ -noinline struct page * +static noinline struct page * dma_alloc_pages(struct device *dev, gfp_t gfp, unsigned order) { int node; @@ -502,6 +506,8 @@ static int __init pci_iommu_init(void) intel_iommu_init(); + amd_iommu_init(); + #ifdef CONFIG_GART_IOMMU gart_iommu_init(); #endif diff --git a/arch/x86/kernel/pci-gart_64.c b/arch/x86/kernel/pci-gart_64.c index aa8ec928caa..c3fe78406d1 100644 --- a/arch/x86/kernel/pci-gart_64.c +++ b/arch/x86/kernel/pci-gart_64.c @@ -104,7 +104,6 @@ static unsigned long alloc_iommu(struct device *dev, int size) size, base_index, boundary_size, 0); } if (offset != -1) { - set_bit_string(iommu_gart_bitmap, offset, size); next_bit = offset+size; if (next_bit >= iommu_pages) { next_bit = 0; @@ -534,8 +533,8 @@ static __init unsigned read_aperture(struct pci_dev *dev, u32 *size) unsigned aper_size = 0, aper_base_32, aper_order; u64 aper_base; - pci_read_config_dword(dev, 0x94, &aper_base_32); - pci_read_config_dword(dev, 0x90, &aper_order); + pci_read_config_dword(dev, AMD64_GARTAPERTUREBASE, &aper_base_32); + pci_read_config_dword(dev, AMD64_GARTAPERTURECTL, &aper_order); aper_order = (aper_order >> 1) & 7; aper_base = aper_base_32 & 0x7fff; @@ -549,14 +548,63 @@ static __init unsigned read_aperture(struct pci_dev *dev, u32 *size) return aper_base; } +static void enable_gart_translations(void) +{ + int i; + + for (i = 0; i < num_k8_northbridges; i++) { + struct pci_dev *dev = k8_northbridges[i]; + + enable_gart_translation(dev, __pa(agp_gatt_table)); + } +} + +/* + * If fix_up_north_bridges is set, the north bridges have to be fixed up on + * resume in the same way as they are handled in gart_iommu_hole_init(). + */ +static bool fix_up_north_bridges; +static u32 aperture_order; +static u32 aperture_alloc; + +void set_up_gart_resume(u32 aper_order, u32 aper_alloc) +{ + fix_up_north_bridges = true; + aperture_order = aper_order; + aperture_alloc = aper_alloc; +} + static int gart_resume(struct sys_device *dev) { + printk(KERN_INFO "PCI-DMA: Resuming GART IOMMU\n"); + + if (fix_up_north_bridges) { + int i; + + printk(KERN_INFO "PCI-DMA: Restoring GART aperture settings\n"); + + for (i = 0; i < num_k8_northbridges; i++) { + struct pci_dev *dev = k8_northbridges[i]; + + /* + * Don't enable translations just yet. That is the next + * step. Restore the pre-suspend aperture settings. + */ + pci_write_config_dword(dev, AMD64_GARTAPERTURECTL, + aperture_order << 1); + pci_write_config_dword(dev, AMD64_GARTAPERTUREBASE, + aperture_alloc >> 25); + } + } + + enable_gart_translations(); + return 0; } static int gart_suspend(struct sys_device *dev, pm_message_t state) { - return -EINVAL; + return 0; } static struct sysdev_class gart_sysdev_class = { @@ -582,6 +630,7 @@ static __init int init_k8_gatt(struct agp_kern_info *info) struct pci_dev *dev; void *gatt; int i, error; + unsigned long start_pfn, end_pfn; printk(KERN_INFO "PCI-DMA: Disabling AGP.\n"); aper_size = aper_base = info->aper_size = 0; @@ -614,31 +663,25 @@ static __init int init_k8_gatt(struct agp_kern_info *info) memset(gatt, 0, gatt_size); agp_gatt_table = gatt; - for (i = 0; i < num_k8_northbridges; i++) { - u32 gatt_reg; - u32 ctl; - - dev = k8_northbridges[i]; - gatt_reg = __pa(gatt) >> 12; - gatt_reg <<= 4; - pci_write_config_dword(dev, 0x98, gatt_reg); - pci_read_config_dword(dev, 0x90, &ctl); - - ctl |= 1; - ctl &= ~((1<<4) | (1<<5)); - - pci_write_config_dword(dev, 0x90, ctl); - } + enable_gart_translations(); error = sysdev_class_register(&gart_sysdev_class); if (!error) error = sysdev_register(&device_gart); if (error) panic("Could not register gart_sysdev -- would corrupt data on next suspend"); + flush_gart(); printk(KERN_INFO "PCI-DMA: aperture base @ %x size %u KB\n", aper_base, aper_size>>10); + + /* need to map that range */ + end_pfn = (aper_base>>PAGE_SHIFT) + (aper_size>>PAGE_SHIFT); + if (end_pfn > max_low_pfn_mapped) { + start_pfn = (aper_base>>PAGE_SHIFT); + init_memory_mapping(start_pfn<<PAGE_SHIFT, end_pfn<<PAGE_SHIFT); + } return 0; nommu: @@ -677,11 +720,11 @@ void gart_iommu_shutdown(void) u32 ctl; dev = k8_northbridges[i]; - pci_read_config_dword(dev, 0x90, &ctl); + pci_read_config_dword(dev, AMD64_GARTAPERTURECTL, &ctl); - ctl &= ~1; + ctl &= ~GARTEN; - pci_write_config_dword(dev, 0x90, ctl); + pci_write_config_dword(dev, AMD64_GARTAPERTURECTL, ctl); } } @@ -716,10 +759,10 @@ void __init gart_iommu_init(void) return; if (no_iommu || - (!force_iommu && end_pfn <= MAX_DMA32_PFN) || + (!force_iommu && max_pfn <= MAX_DMA32_PFN) || !gart_iommu_aperture || (no_agp && init_k8_gatt(&info) < 0)) { - if (end_pfn > MAX_DMA32_PFN) { + if (max_pfn > MAX_DMA32_PFN) { printk(KERN_WARNING "More than 4GB of memory " "but GART IOMMU not available.\n" KERN_WARNING "falling back to iommu=soft.\n"); @@ -788,10 +831,10 @@ void __init gart_iommu_init(void) wbinvd(); /* - * Try to workaround a bug (thanks to BenH) + * Try to workaround a bug (thanks to BenH): * Set unmapped entries to a scratch page instead of 0. * Any prefetches that hit unmapped entries won't get an bus abort - * then. + * then. (P2P bridge may be prefetching on DMA reads). */ scratch = get_zeroed_page(GFP_KERNEL); if (!scratch) diff --git a/arch/x86/kernel/pci-swiotlb_64.c b/arch/x86/kernel/pci-swiotlb_64.c index 490da7f4b8d..82299cd1d04 100644 --- a/arch/x86/kernel/pci-swiotlb_64.c +++ b/arch/x86/kernel/pci-swiotlb_64.c @@ -38,7 +38,7 @@ const struct dma_mapping_ops swiotlb_dma_ops = { void __init pci_swiotlb_init(void) { /* don't initialize swiotlb if iommu=off (no_iommu=1) */ - if (!iommu_detected && !no_iommu && end_pfn > MAX_DMA32_PFN) + if (!iommu_detected && !no_iommu && max_pfn > MAX_DMA32_PFN) swiotlb = 1; if (swiotlb_force) swiotlb = 1; diff --git a/arch/x86/kernel/probe_roms_32.c b/arch/x86/kernel/probe_roms_32.c new file mode 100644 index 00000000000..675a48c404a --- /dev/null +++ b/arch/x86/kernel/probe_roms_32.c @@ -0,0 +1,166 @@ +#include <linux/sched.h> +#include <linux/mm.h> +#include <linux/uaccess.h> +#include <linux/mmzone.h> +#include <linux/ioport.h> +#include <linux/seq_file.h> +#include <linux/console.h> +#include <linux/init.h> +#include <linux/edd.h> +#include <linux/dmi.h> +#include <linux/pfn.h> +#include <linux/pci.h> +#include <asm/pci-direct.h> + + +#include <asm/e820.h> +#include <asm/mmzone.h> +#include <asm/setup.h> +#include <asm/sections.h> +#include <asm/io.h> +#include <setup_arch.h> + +static struct resource system_rom_resource = { + .name = "System ROM", + .start = 0xf0000, + .end = 0xfffff, + .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM +}; + +static struct resource extension_rom_resource = { + .name = "Extension ROM", + .start = 0xe0000, + .end = 0xeffff, + .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM +}; + +static struct resource adapter_rom_resources[] = { { + .name = "Adapter ROM", + .start = 0xc8000, + .end = 0, + .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM +}, { + .name = "Adapter ROM", + .start = 0, + .end = 0, + .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM +}, { + .name = "Adapter ROM", + .start = 0, + .end = 0, + .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM +}, { + .name = "Adapter ROM", + .start = 0, + .end = 0, + .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM +}, { + .name = "Adapter ROM", + .start = 0, + .end = 0, + .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM +}, { + .name = "Adapter ROM", + .start = 0, + .end = 0, + .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM +} }; + +static struct resource video_rom_resource = { + .name = "Video ROM", + .start = 0xc0000, + .end = 0xc7fff, + .flags = IORESOURCE_BUSY | IORESOURCE_READONLY | IORESOURCE_MEM +}; + +#define ROMSIGNATURE 0xaa55 + +static int __init romsignature(const unsigned char *rom) +{ + const unsigned short * const ptr = (const unsigned short *)rom; + unsigned short sig; + + return probe_kernel_address(ptr, sig) == 0 && sig == ROMSIGNATURE; +} + +static int __init romchecksum(const unsigned char *rom, unsigned long length) +{ + unsigned char sum, c; + + for (sum = 0; length && probe_kernel_address(rom++, c) == 0; length--) + sum += c; + return !length && !sum; +} + +void __init probe_roms(void) +{ + const unsigned char *rom; + unsigned long start, length, upper; + unsigned char c; + int i; + + /* video rom */ + upper = adapter_rom_resources[0].start; + for (start = video_rom_resource.start; start < upper; start += 2048) { + rom = isa_bus_to_virt(start); + if (!romsignature(rom)) + continue; + + video_rom_resource.start = start; + + if (probe_kernel_address(rom + 2, c) != 0) + continue; + + /* 0 < length <= 0x7f * 512, historically */ + length = c * 512; + + /* if checksum okay, trust length byte */ + if (length && romchecksum(rom, length)) + video_rom_resource.end = start + length - 1; + + request_resource(&iomem_resource, &video_rom_resource); + break; + } + + start = (video_rom_resource.end + 1 + 2047) & ~2047UL; + if (start < upper) + start = upper; + + /* system rom */ + request_resource(&iomem_resource, &system_rom_resource); + upper = system_rom_resource.start; + + /* check for extension rom (ignore length byte!) */ + rom = isa_bus_to_virt(extension_rom_resource.start); + if (romsignature(rom)) { + length = extension_rom_resource.end - extension_rom_resource.start + 1; + if (romchecksum(rom, length)) { + request_resource(&iomem_resource, &extension_rom_resource); + upper = extension_rom_resource.start; + } + } + + /* check for adapter roms on 2k boundaries */ + for (i = 0; i < ARRAY_SIZE(adapter_rom_resources) && start < upper; start += 2048) { + rom = isa_bus_to_virt(start); + if (!romsignature(rom)) + continue; + + if (probe_kernel_address(rom + 2, c) != 0) + continue; + + /* 0 < length <= 0x7f * 512, historically */ + length = c * 512; + + /* but accept any length that fits if checksum okay */ + if (!length || start + length > upper || !romchecksum(rom, length)) + continue; + + adapter_rom_resources[i].start = start; + adapter_rom_resources[i].end = start + length - 1; + request_resource(&iomem_resource, &adapter_rom_resources[i]); + + start = adapter_rom_resources[i++].end & ~2047UL; + } +} + diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c index ba370dc8685..4061d63aabe 100644 --- a/arch/x86/kernel/process.c +++ b/arch/x86/kernel/process.c @@ -6,6 +6,7 @@ #include <linux/sched.h> #include <linux/module.h> #include <linux/pm.h> +#include <linux/clockchips.h> struct kmem_cache *task_xstate_cachep; @@ -45,6 +46,76 @@ void arch_task_cache_init(void) SLAB_PANIC, NULL); } +/* + * Idle related variables and functions + */ +unsigned long boot_option_idle_override = 0; +EXPORT_SYMBOL(boot_option_idle_override); + +/* + * Powermanagement idle function, if any.. + */ +void (*pm_idle)(void); +EXPORT_SYMBOL(pm_idle); + +#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) +{ + hlt_counter++; +} +EXPORT_SYMBOL(disable_hlt); + +void enable_hlt(void) +{ + hlt_counter--; +} +EXPORT_SYMBOL(enable_hlt); + +static inline int hlt_use_halt(void) +{ + return (!hlt_counter && boot_cpu_data.hlt_works_ok); +} +#else +static inline int hlt_use_halt(void) +{ + return 1; +} +#endif + +/* + * We use this if we don't have any better + * idle routine.. + */ +void default_idle(void) +{ + if (hlt_use_halt()) { + 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; + } else { + local_irq_enable(); + /* loop is done by the caller */ + cpu_relax(); + } +} +#ifdef CONFIG_APM_MODULE +EXPORT_SYMBOL(default_idle); +#endif + static void do_nothing(void *unused) { } @@ -122,44 +193,129 @@ static void poll_idle(void) * * 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 + static int __cpuinit mwait_usable(const struct cpuinfo_x86 *c) { + u32 eax, ebx, ecx, edx; + if (force_mwait) return 1; - if (c->x86_vendor == X86_VENDOR_AMD) { - switch(c->x86) { - case 0x10: - case 0x11: - return 0; - } - } + 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); +} + +/* + * Check for AMD CPUs, which have potentially C1E support + */ +static int __cpuinit check_c1e_idle(const struct cpuinfo_x86 *c) +{ + if (c->x86_vendor != X86_VENDOR_AMD) + return 0; + + if (c->x86 < 0x0F) + return 0; + + /* Family 0x0f models < rev F do not have C1E */ + if (c->x86 == 0x0f && c->x86_model < 0x40) + return 0; + return 1; } -void __cpuinit select_idle_routine(const struct cpuinfo_x86 *c) +/* + * C1E aware idle routine. We check for C1E active in the interrupt + * pending message MSR. If we detect C1E, then we handle it the same + * way as C3 power states (local apic timer and TSC stop) + */ +static void c1e_idle(void) { - static int selected; + static cpumask_t c1e_mask = CPU_MASK_NONE; + static int c1e_detected; - if (selected) + if (need_resched()) return; + + if (!c1e_detected) { + u32 lo, hi; + + rdmsr(MSR_K8_INT_PENDING_MSG, lo, hi); + if (lo & K8_INTP_C1E_ACTIVE_MASK) { + c1e_detected = 1; + mark_tsc_unstable("TSC halt in C1E"); + printk(KERN_INFO "System has C1E enabled\n"); + } + } + + if (c1e_detected) { + int cpu = smp_processor_id(); + + if (!cpu_isset(cpu, c1e_mask)) { + cpu_set(cpu, c1e_mask); + /* + * Force broadcast so ACPI can not interfere. Needs + * to run with interrupts enabled as it uses + * smp_function_call. + */ + local_irq_enable(); + clockevents_notify(CLOCK_EVT_NOTIFY_BROADCAST_FORCE, + &cpu); + printk(KERN_INFO "Switch to broadcast mode on CPU%d\n", + cpu); + local_irq_disable(); + } + clockevents_notify(CLOCK_EVT_NOTIFY_BROADCAST_ENTER, &cpu); + + default_idle(); + + /* + * 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(); + } else + default_idle(); +} + +void __cpuinit select_idle_routine(const struct cpuinfo_x86 *c) +{ #ifdef CONFIG_X86_SMP if (pm_idle == poll_idle && smp_num_siblings > 1) { printk(KERN_WARNING "WARNING: polling idle and HT enabled," " performance may degrade.\n"); } #endif + if (pm_idle) + return; + if (cpu_has(c, X86_FEATURE_MWAIT) && mwait_usable(c)) { /* - * Skip, if setup has overridden idle. * One CPU supports mwait => All CPUs supports mwait */ - if (!pm_idle) { - printk(KERN_INFO "using mwait in idle threads.\n"); - pm_idle = mwait_idle; - } - } - selected = 1; + printk(KERN_INFO "using mwait in idle threads.\n"); + pm_idle = mwait_idle; + } else if (check_c1e_idle(c)) { + printk(KERN_INFO "using C1E aware idle routine\n"); + pm_idle = c1e_idle; + } else + pm_idle = default_idle; } static int __init idle_setup(char *str) diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c index e2db9ac5c61..9a139f6c9df 100644 --- a/arch/x86/kernel/process_32.c +++ b/arch/x86/kernel/process_32.c @@ -58,11 +58,6 @@ asmlinkage void ret_from_fork(void) __asm__("ret_from_fork"); -static int hlt_counter; - -unsigned long boot_option_idle_override = 0; -EXPORT_SYMBOL(boot_option_idle_override); - DEFINE_PER_CPU(struct task_struct *, current_task) = &init_task; EXPORT_PER_CPU_SYMBOL(current_task); @@ -77,57 +72,24 @@ unsigned long thread_saved_pc(struct task_struct *tsk) return ((unsigned long *)tsk->thread.sp)[3]; } -/* - * Powermanagement idle function, if any.. - */ -void (*pm_idle)(void); -EXPORT_SYMBOL(pm_idle); +#ifdef CONFIG_HOTPLUG_CPU +#include <asm/nmi.h> -void disable_hlt(void) +static void cpu_exit_clear(void) { - hlt_counter++; -} + int cpu = raw_smp_processor_id(); -EXPORT_SYMBOL(disable_hlt); + idle_task_exit(); -void enable_hlt(void) -{ - hlt_counter--; -} + cpu_uninit(); + irq_ctx_exit(cpu); -EXPORT_SYMBOL(enable_hlt); - -/* - * We use this if we don't have any better - * idle routine.. - */ -void default_idle(void) -{ - if (!hlt_counter && boot_cpu_data.hlt_works_ok) { - current_thread_info()->status &= ~TS_POLLING; - /* - * TS_POLLING-cleared state must be visible before we - * test NEED_RESCHED: - */ - smp_mb(); + cpu_clear(cpu, cpu_callout_map); + cpu_clear(cpu, cpu_callin_map); - if (!need_resched()) - safe_halt(); /* enables interrupts racelessly */ - else - local_irq_enable(); - current_thread_info()->status |= TS_POLLING; - } else { - local_irq_enable(); - /* loop is done by the caller */ - cpu_relax(); - } + numa_remove_cpu(cpu); } -#ifdef CONFIG_APM_MODULE -EXPORT_SYMBOL(default_idle); -#endif -#ifdef CONFIG_HOTPLUG_CPU -#include <asm/nmi.h> /* We don't actually take CPU down, just spin without interrupts. */ static inline void play_dead(void) { @@ -168,24 +130,19 @@ void cpu_idle(void) while (1) { tick_nohz_stop_sched_tick(); while (!need_resched()) { - void (*idle)(void); check_pgt_cache(); rmb(); - idle = pm_idle; if (rcu_pending(cpu)) rcu_check_callbacks(cpu, 0); - if (!idle) - idle = default_idle; - if (cpu_is_offline(cpu)) play_dead(); local_irq_disable(); __get_cpu_var(irq_stat).idle_timestamp = jiffies; - idle(); + pm_idle(); } tick_nohz_restart_sched_tick(); preempt_enable_no_resched(); diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c index c6eb5c91e5f..db5eb963e4d 100644 --- a/arch/x86/kernel/process_64.c +++ b/arch/x86/kernel/process_64.c @@ -56,15 +56,6 @@ asmlinkage extern void ret_from_fork(void); unsigned long kernel_thread_flags = CLONE_VM | CLONE_UNTRACED; -unsigned long boot_option_idle_override = 0; -EXPORT_SYMBOL(boot_option_idle_override); - -/* - * Powermanagement idle function, if any.. - */ -void (*pm_idle)(void); -EXPORT_SYMBOL(pm_idle); - static ATOMIC_NOTIFIER_HEAD(idle_notifier); void idle_notifier_register(struct notifier_block *n) @@ -94,25 +85,6 @@ void exit_idle(void) __exit_idle(); } -/* - * We use this if we don't have any better - * idle routine.. - */ -void default_idle(void) -{ - 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; -} - #ifdef CONFIG_HOTPLUG_CPU DECLARE_PER_CPU(int, cpu_state); @@ -150,12 +122,9 @@ void cpu_idle(void) while (1) { tick_nohz_stop_sched_tick(); while (!need_resched()) { - void (*idle)(void); rmb(); - idle = pm_idle; - if (!idle) - idle = default_idle; + if (cpu_is_offline(smp_processor_id())) play_dead(); /* @@ -165,7 +134,7 @@ void cpu_idle(void) */ local_irq_disable(); enter_idle(); - idle(); + pm_idle(); /* In many cases the interrupt that ended idle has already called exit_idle. But some idle loops can be woken up without interrupt. */ @@ -366,10 +335,10 @@ int copy_thread(int nr, unsigned long clone_flags, unsigned long sp, p->thread.fs = me->thread.fs; p->thread.gs = me->thread.gs; - asm("mov %%gs,%0" : "=m" (p->thread.gsindex)); - asm("mov %%fs,%0" : "=m" (p->thread.fsindex)); - asm("mov %%es,%0" : "=m" (p->thread.es)); - asm("mov %%ds,%0" : "=m" (p->thread.ds)); + savesegment(gs, p->thread.gsindex); + savesegment(fs, p->thread.fsindex); + savesegment(es, p->thread.es); + savesegment(ds, p->thread.ds); if (unlikely(test_tsk_thread_flag(me, TIF_IO_BITMAP))) { p->thread.io_bitmap_ptr = kmalloc(IO_BITMAP_BYTES, GFP_KERNEL); @@ -408,7 +377,9 @@ out: void start_thread(struct pt_regs *regs, unsigned long new_ip, unsigned long new_sp) { - asm volatile("movl %0, %%fs; movl %0, %%es; movl %0, %%ds" :: "r"(0)); + loadsegment(fs, 0); + loadsegment(es, 0); + loadsegment(ds, 0); load_gs_index(0); regs->ip = new_ip; regs->sp = new_sp; @@ -567,6 +538,7 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) *next = &next_p->thread; int cpu = smp_processor_id(); struct tss_struct *tss = &per_cpu(init_tss, cpu); + unsigned fsindex, gsindex; /* we're going to use this soon, after a few expensive things */ if (next_p->fpu_counter>5) @@ -581,22 +553,38 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) * Switch DS and ES. * This won't pick up thread selector changes, but I guess that is ok. */ - asm volatile("mov %%es,%0" : "=m" (prev->es)); + savesegment(es, prev->es); if (unlikely(next->es | prev->es)) loadsegment(es, next->es); - - asm volatile ("mov %%ds,%0" : "=m" (prev->ds)); + + savesegment(ds, prev->ds); if (unlikely(next->ds | prev->ds)) loadsegment(ds, next->ds); + + /* We must save %fs and %gs before load_TLS() because + * %fs and %gs may be cleared by load_TLS(). + * + * (e.g. xen_load_tls()) + */ + savesegment(fs, fsindex); + savesegment(gs, gsindex); + load_TLS(next, cpu); + /* + * Leave lazy mode, flushing any hypercalls made here. + * This must be done before restoring TLS segments so + * the GDT and LDT are properly updated, and must be + * done before math_state_restore, so the TS bit is up + * to date. + */ + arch_leave_lazy_cpu_mode(); + /* * Switch FS and GS. */ { - unsigned fsindex; - asm volatile("movl %%fs,%0" : "=r" (fsindex)); /* segment register != 0 always requires a reload. also reload when it has changed. when prev process used 64bit base always reload @@ -614,10 +602,7 @@ __switch_to(struct task_struct *prev_p, struct task_struct *next_p) if (next->fs) wrmsrl(MSR_FS_BASE, next->fs); prev->fsindex = fsindex; - } - { - unsigned gsindex; - asm volatile("movl %%gs,%0" : "=r" (gsindex)); + if (unlikely(gsindex | next->gsindex | prev->gs)) { load_gs_index(next->gsindex); if (gsindex) @@ -798,7 +783,7 @@ long do_arch_prctl(struct task_struct *task, int code, unsigned long addr) set_32bit_tls(task, FS_TLS, addr); if (doit) { load_TLS(&task->thread, cpu); - asm volatile("movl %0,%%fs" :: "r"(FS_TLS_SEL)); + loadsegment(fs, FS_TLS_SEL); } task->thread.fsindex = FS_TLS_SEL; task->thread.fs = 0; @@ -808,7 +793,7 @@ long do_arch_prctl(struct task_struct *task, int code, unsigned long addr) if (doit) { /* set the selector to 0 to not confuse __switch_to */ - asm volatile("movl %0,%%fs" :: "r" (0)); + loadsegment(fs, 0); ret = checking_wrmsrl(MSR_FS_BASE, addr); } } @@ -831,7 +816,7 @@ long do_arch_prctl(struct task_struct *task, int code, unsigned long addr) if (task->thread.gsindex == GS_TLS_SEL) base = read_32bit_tls(task, GS_TLS); else if (doit) { - asm("movl %%gs,%0" : "=r" (gsindex)); + savesegment(gs, gsindex); if (gsindex) rdmsrl(MSR_KERNEL_GS_BASE, base); else diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c index a7835f28293..77040b6070e 100644 --- a/arch/x86/kernel/ptrace.c +++ b/arch/x86/kernel/ptrace.c @@ -943,13 +943,13 @@ long arch_ptrace(struct task_struct *child, long request, long addr, long data) return copy_regset_to_user(child, &user_x86_32_view, REGSET_XFP, 0, sizeof(struct user_fxsr_struct), - datap); + datap) ? -EIO : 0; case PTRACE_SETFPXREGS: /* Set the child extended FPU state. */ return copy_regset_from_user(child, &user_x86_32_view, REGSET_XFP, 0, sizeof(struct user_fxsr_struct), - datap); + datap) ? -EIO : 0; #endif #if defined CONFIG_X86_32 || defined CONFIG_IA32_EMULATION diff --git a/arch/x86/kernel/quirks.c b/arch/x86/kernel/quirks.c index d89a648fe71..79bdcd11c66 100644 --- a/arch/x86/kernel/quirks.c +++ b/arch/x86/kernel/quirks.c @@ -65,6 +65,7 @@ static enum { ICH_FORCE_HPET_RESUME, VT8237_FORCE_HPET_RESUME, NVIDIA_FORCE_HPET_RESUME, + ATI_FORCE_HPET_RESUME, } force_hpet_resume_type; static void __iomem *rcba_base; @@ -158,6 +159,8 @@ static void ich_force_enable_hpet(struct pci_dev *dev) DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_ESB2_0, ich_force_enable_hpet); +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_ICH6_0, + ich_force_enable_hpet); DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_ICH6_1, ich_force_enable_hpet); DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_ICH7_0, @@ -174,6 +177,12 @@ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_ICH9_7, static struct pci_dev *cached_dev; +static void hpet_print_force_info(void) +{ + printk(KERN_INFO "HPET not enabled in BIOS. " + "You might try hpet=force boot option\n"); +} + static void old_ich_force_hpet_resume(void) { u32 val; @@ -253,6 +262,8 @@ static void old_ich_force_enable_hpet_user(struct pci_dev *dev) { if (hpet_force_user) old_ich_force_enable_hpet(dev); + else + hpet_print_force_info(); } DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82801CA_0, @@ -290,8 +301,13 @@ static void vt8237_force_enable_hpet(struct pci_dev *dev) { u32 uninitialized_var(val); - if (!hpet_force_user || hpet_address || force_hpet_address) + if (hpet_address || force_hpet_address) + return; + + if (!hpet_force_user) { + hpet_print_force_info(); return; + } pci_read_config_dword(dev, 0x68, &val); /* @@ -330,6 +346,36 @@ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_8235, DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_8237, vt8237_force_enable_hpet); +static void ati_force_hpet_resume(void) +{ + pci_write_config_dword(cached_dev, 0x14, 0xfed00000); + printk(KERN_DEBUG "Force enabled HPET at resume\n"); +} + +static void ati_force_enable_hpet(struct pci_dev *dev) +{ + u32 uninitialized_var(val); + + if (hpet_address || force_hpet_address) + return; + + if (!hpet_force_user) { + hpet_print_force_info(); + return; + } + + pci_write_config_dword(dev, 0x14, 0xfed00000); + pci_read_config_dword(dev, 0x14, &val); + force_hpet_address = val; + force_hpet_resume_type = ATI_FORCE_HPET_RESUME; + dev_printk(KERN_DEBUG, &dev->dev, "Force enabled HPET at 0x%lx\n", + force_hpet_address); + cached_dev = dev; + return; +} +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_ATI, PCI_DEVICE_ID_ATI_IXP400_SMBUS, + ati_force_enable_hpet); + /* * Undocumented chipset feature taken from LinuxBIOS. */ @@ -343,8 +389,13 @@ static void nvidia_force_enable_hpet(struct pci_dev *dev) { u32 uninitialized_var(val); - if (!hpet_force_user || hpet_address || force_hpet_address) + if (hpet_address || force_hpet_address) + return; + + if (!hpet_force_user) { + hpet_print_force_info(); return; + } pci_write_config_dword(dev, 0x44, 0xfed00001); pci_read_config_dword(dev, 0x44, &val); @@ -397,6 +448,9 @@ void force_hpet_resume(void) case NVIDIA_FORCE_HPET_RESUME: nvidia_force_hpet_resume(); return; + case ATI_FORCE_HPET_RESUME: + ati_force_hpet_resume(); + return; default: break; } diff --git a/arch/x86/kernel/reboot.c b/arch/x86/kernel/reboot.c index f6be7d5f82f..f8a62160e15 100644 --- a/arch/x86/kernel/reboot.c +++ b/arch/x86/kernel/reboot.c @@ -27,7 +27,7 @@ void (*pm_power_off)(void); EXPORT_SYMBOL(pm_power_off); -static long no_idt[3]; +static const struct desc_ptr no_idt = {}; static int reboot_mode; enum reboot_type reboot_type = BOOT_KBD; int reboot_force; @@ -201,15 +201,15 @@ core_initcall(reboot_init); 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. */ -static unsigned long long +static const unsigned long long real_mode_gdt_entries [3] = { 0x0000000000000000ULL, /* Null descriptor */ - 0x00009a000000ffffULL, /* 16-bit real-mode 64k code at 0x00000000 */ - 0x000092000100ffffULL /* 16-bit real-mode 64k data at 0x00000100 */ + 0x00009b000000ffffULL, /* 16-bit real-mode 64k code at 0x00000000 */ + 0x000093000100ffffULL /* 16-bit real-mode 64k data at 0x00000100 */ }; -static struct desc_ptr +static const struct desc_ptr real_mode_gdt = { sizeof (real_mode_gdt_entries) - 1, (long)real_mode_gdt_entries }, real_mode_idt = { 0x3ff, 0 }; @@ -231,7 +231,7 @@ real_mode_idt = { 0x3ff, 0 }; More could be done here to set up the registers as if a CPU reset had occurred; hopefully real BIOSs don't assume much. */ -static unsigned char real_mode_switch [] = +static const unsigned char real_mode_switch [] = { 0x66, 0x0f, 0x20, 0xc0, /* movl %cr0,%eax */ 0x66, 0x83, 0xe0, 0x11, /* andl $0x00000011,%eax */ @@ -245,7 +245,7 @@ static unsigned char real_mode_switch [] = 0x24, 0x10, /* f: andb $0x10,al */ 0x66, 0x0f, 0x22, 0xc0 /* movl %eax,%cr0 */ }; -static unsigned char jump_to_bios [] = +static const unsigned char jump_to_bios [] = { 0xea, 0x00, 0x00, 0xff, 0xff /* ljmp $0xffff,$0x0000 */ }; @@ -255,7 +255,7 @@ static unsigned char jump_to_bios [] = * specified by the code and length parameters. * We assume that length will aways be less that 100! */ -void machine_real_restart(unsigned char *code, int length) +void machine_real_restart(const unsigned char *code, int length) { local_irq_disable(); @@ -368,7 +368,7 @@ static void native_machine_emergency_restart(void) } case BOOT_TRIPLE: - load_idt((const struct desc_ptr *)&no_idt); + load_idt(&no_idt); __asm__ __volatile__("int3"); reboot_type = BOOT_KBD; diff --git a/arch/x86/kernel/reboot_fixups_32.c b/arch/x86/kernel/reboot_fixups_32.c index dec0b5ec25c..61a837743fe 100644 --- a/arch/x86/kernel/reboot_fixups_32.c +++ b/arch/x86/kernel/reboot_fixups_32.c @@ -49,7 +49,7 @@ struct device_fixup { void (*reboot_fixup)(struct pci_dev *); }; -static struct device_fixup fixups_table[] = { +static const struct device_fixup fixups_table[] = { { PCI_VENDOR_ID_CYRIX, PCI_DEVICE_ID_CYRIX_5530_LEGACY, cs5530a_warm_reset }, { PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_CS5536_ISA, cs5536_warm_reset }, { PCI_VENDOR_ID_NS, PCI_DEVICE_ID_NS_SC1100_BRIDGE, cs5530a_warm_reset }, @@ -64,7 +64,7 @@ static struct device_fixup fixups_table[] = { */ void mach_reboot_fixups(void) { - struct device_fixup *cur; + const struct device_fixup *cur; struct pci_dev *dev; int i; diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c index 6f80b852a19..36c540d4ac4 100644 --- a/arch/x86/kernel/setup.c +++ b/arch/x86/kernel/setup.c @@ -1,139 +1,889 @@ -#include <linux/kernel.h> +/* + * Copyright (C) 1995 Linus Torvalds + * + * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999 + * + * Memory region support + * David Parsons <orc@pell.chi.il.us>, July-August 1999 + * + * Added E820 sanitization routine (removes overlapping memory regions); + * Brian Moyle <bmoyle@mvista.com>, February 2001 + * + * Moved CPU detection code to cpu/${cpu}.c + * Patrick Mochel <mochel@osdl.org>, March 2002 + * + * Provisions for empty E820 memory regions (reported by certain BIOSes). + * Alex Achenbach <xela@slit.de>, December 2002. + * + */ + +/* + * This file handles the architecture-dependent parts of initialization + */ + +#include <linux/sched.h> +#include <linux/mm.h> +#include <linux/mmzone.h> +#include <linux/screen_info.h> +#include <linux/ioport.h> +#include <linux/acpi.h> +#include <linux/apm_bios.h> +#include <linux/initrd.h> +#include <linux/bootmem.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> +#include <linux/efi.h> #include <linux/init.h> -#include <linux/bootmem.h> +#include <linux/edd.h> +#include <linux/iscsi_ibft.h> +#include <linux/nodemask.h> +#include <linux/kexec.h> +#include <linux/dmi.h> +#include <linux/pfn.h> +#include <linux/pci.h> +#include <asm/pci-direct.h> +#include <linux/init_ohci1394_dma.h> +#include <linux/kvm_para.h> + +#include <linux/errno.h> +#include <linux/kernel.h> +#include <linux/stddef.h> +#include <linux/unistd.h> +#include <linux/ptrace.h> +#include <linux/slab.h> +#include <linux/user.h> +#include <linux/delay.h> +#include <linux/highmem.h> + +#include <linux/kallsyms.h> +#include <linux/edd.h> +#include <linux/iscsi_ibft.h> +#include <linux/kexec.h> +#include <linux/cpufreq.h> +#include <linux/dma-mapping.h> +#include <linux/ctype.h> +#include <linux/uaccess.h> + #include <linux/percpu.h> +#include <linux/crash_dump.h> + +#include <video/edid.h> + +#include <asm/mtrr.h> +#include <asm/apic.h> +#include <asm/e820.h> +#include <asm/mpspec.h> +#include <asm/setup.h> +#include <asm/arch_hooks.h> +#include <asm/efi.h> +#include <asm/sections.h> +#include <asm/dmi.h> +#include <asm/io_apic.h> +#include <asm/ist.h> +#include <asm/vmi.h> +#include <setup_arch.h> +#include <asm/bios_ebda.h> +#include <asm/cacheflush.h> +#include <asm/processor.h> +#include <asm/bugs.h> + +#include <asm/system.h> +#include <asm/vsyscall.h> #include <asm/smp.h> +#include <asm/desc.h> +#include <asm/dma.h> +#include <asm/gart.h> +#include <asm/mmu_context.h> +#include <asm/proto.h> + +#include <mach_apic.h> +#include <asm/paravirt.h> + #include <asm/percpu.h> #include <asm/sections.h> -#include <asm/processor.h> -#include <asm/setup.h> #include <asm/topology.h> -#include <asm/mpspec.h> #include <asm/apicdef.h> +#ifdef CONFIG_X86_64 +#include <asm/numa_64.h> +#endif -#ifdef CONFIG_X86_LOCAL_APIC -unsigned int num_processors; -unsigned disabled_cpus __cpuinitdata; -/* Processor that is doing the boot up */ -unsigned int boot_cpu_physical_apicid = -1U; -EXPORT_SYMBOL(boot_cpu_physical_apicid); +#ifndef ARCH_SETUP +#define ARCH_SETUP +#endif + +#ifndef CONFIG_DEBUG_BOOT_PARAMS +struct boot_params __initdata boot_params; +#else +struct boot_params boot_params; +#endif + +/* + * Machine setup.. + */ +static struct resource data_resource = { + .name = "Kernel data", + .start = 0, + .end = 0, + .flags = IORESOURCE_BUSY | IORESOURCE_MEM +}; + +static struct resource code_resource = { + .name = "Kernel code", + .start = 0, + .end = 0, + .flags = IORESOURCE_BUSY | IORESOURCE_MEM +}; + +static struct resource bss_resource = { + .name = "Kernel bss", + .start = 0, + .end = 0, + .flags = IORESOURCE_BUSY | IORESOURCE_MEM +}; + + +#ifdef CONFIG_X86_32 +/* This value is set up by the early boot code to point to the value + immediately after the boot time page tables. It contains a *physical* + address, and must not be in the .bss segment! */ +unsigned long init_pg_tables_start __initdata = ~0UL; +unsigned long init_pg_tables_end __initdata = ~0UL; + +static struct resource video_ram_resource = { + .name = "Video RAM area", + .start = 0xa0000, + .end = 0xbffff, + .flags = IORESOURCE_BUSY | IORESOURCE_MEM +}; + +/* 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}; +/* common cpu data for all cpus */ +struct cpuinfo_x86 boot_cpu_data __read_mostly = {0, 0, 0, 0, -1, 1, 0, 0, -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; + +/* for MCA, but anyone else can use it if they want */ +unsigned int machine_id; +unsigned int machine_submodel_id; +unsigned int BIOS_revision; + +struct apm_info apm_info; +EXPORT_SYMBOL(apm_info); + +#if defined(CONFIG_X86_SPEEDSTEP_SMI) || \ + defined(CONFIG_X86_SPEEDSTEP_SMI_MODULE) +struct ist_info ist_info; +EXPORT_SYMBOL(ist_info); +#else +struct ist_info ist_info; +#endif + +#else +struct cpuinfo_x86 boot_cpu_data __read_mostly; +EXPORT_SYMBOL(boot_cpu_data); +#endif -DEFINE_PER_CPU(u16, x86_cpu_to_apicid) = BAD_APICID; -EXPORT_PER_CPU_SYMBOL(x86_cpu_to_apicid); -/* Bitmask of physically existing CPUs */ -physid_mask_t phys_cpu_present_map; +#if !defined(CONFIG_X86_PAE) || defined(CONFIG_X86_64) +unsigned long mmu_cr4_features; +#else +unsigned long mmu_cr4_features = X86_CR4_PAE; #endif -#if defined(CONFIG_HAVE_SETUP_PER_CPU_AREA) && defined(CONFIG_X86_SMP) +/* Boot loader ID as an integer, for the benefit of proc_dointvec */ +int bootloader_type; + /* - * Copy data used in early init routines from the initial arrays to the - * per cpu data areas. These arrays then become expendable and the - * *_early_ptr's are zeroed indicating that the static arrays are gone. + * Early DMI memory */ -static void __init setup_per_cpu_maps(void) +int dmi_alloc_index; +char dmi_alloc_data[DMI_MAX_DATA]; + +/* + * Setup options + */ +struct screen_info screen_info; +EXPORT_SYMBOL(screen_info); +struct edid_info edid_info; +EXPORT_SYMBOL_GPL(edid_info); + +extern int root_mountflags; + +unsigned long saved_video_mode; + +#define RAMDISK_IMAGE_START_MASK 0x07FF +#define RAMDISK_PROMPT_FLAG 0x8000 +#define RAMDISK_LOAD_FLAG 0x4000 + +static char __initdata command_line[COMMAND_LINE_SIZE]; + +#if defined(CONFIG_EDD) || defined(CONFIG_EDD_MODULE) +struct edd edd; +#ifdef CONFIG_EDD_MODULE +EXPORT_SYMBOL(edd); +#endif +/** + * copy_edd() - Copy the BIOS EDD information + * from boot_params into a safe place. + * + */ +static inline void copy_edd(void) +{ + memcpy(edd.mbr_signature, boot_params.edd_mbr_sig_buffer, + sizeof(edd.mbr_signature)); + memcpy(edd.edd_info, boot_params.eddbuf, sizeof(edd.edd_info)); + edd.mbr_signature_nr = boot_params.edd_mbr_sig_buf_entries; + edd.edd_info_nr = boot_params.eddbuf_entries; +} +#else +static inline void copy_edd(void) +{ +} +#endif + +#ifdef CONFIG_BLK_DEV_INITRD + +#ifdef CONFIG_X86_32 + +#define MAX_MAP_CHUNK (NR_FIX_BTMAPS << PAGE_SHIFT) +static void __init relocate_initrd(void) { - int cpu; - for_each_possible_cpu(cpu) { - per_cpu(x86_cpu_to_apicid, cpu) = x86_cpu_to_apicid_init[cpu]; - per_cpu(x86_bios_cpu_apicid, cpu) = - x86_bios_cpu_apicid_init[cpu]; -#ifdef CONFIG_NUMA - per_cpu(x86_cpu_to_node_map, cpu) = - x86_cpu_to_node_map_init[cpu]; + u64 ramdisk_image = boot_params.hdr.ramdisk_image; + u64 ramdisk_size = boot_params.hdr.ramdisk_size; + u64 end_of_lowmem = max_low_pfn << PAGE_SHIFT; + u64 ramdisk_here; + unsigned long slop, clen, mapaddr; + char *p, *q; + + /* We need to move the initrd down into lowmem */ + ramdisk_here = find_e820_area(0, end_of_lowmem, ramdisk_size, + PAGE_SIZE); + + if (ramdisk_here == -1ULL) + panic("Cannot find place for new RAMDISK of size %lld\n", + ramdisk_size); + + /* Note: this includes all the lowmem currently occupied by + the initrd, we rely on that fact to keep the data intact. */ + reserve_early(ramdisk_here, ramdisk_here + ramdisk_size, + "NEW RAMDISK"); + initrd_start = ramdisk_here + PAGE_OFFSET; + initrd_end = initrd_start + ramdisk_size; + printk(KERN_INFO "Allocated new RAMDISK: %08llx - %08llx\n", + ramdisk_here, ramdisk_here + ramdisk_size); + + 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 */ + while (ramdisk_size) { + slop = ramdisk_image & ~PAGE_MASK; + clen = ramdisk_size; + if (clen > MAX_MAP_CHUNK-slop) + clen = MAX_MAP_CHUNK-slop; + mapaddr = ramdisk_image & PAGE_MASK; + p = early_ioremap(mapaddr, clen+slop); + memcpy(q, p+slop, clen); + early_iounmap(p, clen+slop); + q += clen; + 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, ramdisk_image + ramdisk_size - 1, + ramdisk_here, ramdisk_here + ramdisk_size - 1); +} #endif + +static void __init reserve_initrd(void) +{ + u64 ramdisk_image = boot_params.hdr.ramdisk_image; + u64 ramdisk_size = boot_params.hdr.ramdisk_size; + u64 ramdisk_end = ramdisk_image + ramdisk_size; + u64 end_of_lowmem = max_low_pfn << PAGE_SHIFT; + + if (!boot_params.hdr.type_of_loader || + !ramdisk_image || !ramdisk_size) + return; /* No initrd provided by bootloader */ + + initrd_start = 0; + + if (ramdisk_size >= (end_of_lowmem>>1)) { + free_early(ramdisk_image, ramdisk_end); + printk(KERN_ERR "initrd too large to handle, " + "disabling initrd\n"); + return; + } + + printk(KERN_INFO "RAMDISK: %08llx - %08llx\n", ramdisk_image, + ramdisk_end); + + + if (ramdisk_end <= end_of_lowmem) { + /* All in lowmem, easy case */ + /* + * don't need to reserve again, already reserved early + * in i386_start_kernel + */ + initrd_start = ramdisk_image + PAGE_OFFSET; + initrd_end = initrd_start + ramdisk_size; + return; } - /* indicate the early static arrays will soon be gone */ - x86_cpu_to_apicid_early_ptr = NULL; - x86_bios_cpu_apicid_early_ptr = NULL; -#ifdef CONFIG_NUMA - x86_cpu_to_node_map_early_ptr = NULL; +#ifdef CONFIG_X86_32 + relocate_initrd(); +#else + printk(KERN_ERR "initrd extends beyond end of memory " + "(0x%08llx > 0x%08llx)\ndisabling initrd\n", + ramdisk_end, end_of_lowmem); + initrd_start = 0; #endif + free_early(ramdisk_image, ramdisk_end); } +#else +static void __init reserve_initrd(void) +{ +} +#endif /* CONFIG_BLK_DEV_INITRD */ + +static void __init parse_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_ioremap(pa_data, PAGE_SIZE); + switch (data->type) { + case SETUP_E820_EXT: + parse_e820_ext(data, pa_data); + break; + default: + break; + } + pa_data = data->next; + early_iounmap(data, PAGE_SIZE); + } +} + +static void __init e820_reserve_setup_data(void) +{ + struct setup_data *data; + 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_ioremap(pa_data, sizeof(*data)); + e820_update_range(pa_data, sizeof(*data)+data->len, + E820_RAM, E820_RESERVED_KERN); + found = 1; + pa_data = data->next; + early_iounmap(data, sizeof(*data)); + } + if (!found) + return; -#ifdef CONFIG_HAVE_CPUMASK_OF_CPU_MAP -cpumask_t *cpumask_of_cpu_map __read_mostly; -EXPORT_SYMBOL(cpumask_of_cpu_map); + sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &e820.nr_map); + memcpy(&e820_saved, &e820, sizeof(struct e820map)); + printk(KERN_INFO "extended physical RAM map:\n"); + e820_print_map("reserve setup_data"); +} -/* requires nr_cpu_ids to be initialized */ -static void __init setup_cpumask_of_cpu(void) +static void __init reserve_early_setup_data(void) { - int i; + struct setup_data *data; + u64 pa_data; + char buf[32]; + + if (boot_params.hdr.version < 0x0209) + return; + pa_data = boot_params.hdr.setup_data; + while (pa_data) { + data = early_ioremap(pa_data, sizeof(*data)); + sprintf(buf, "setup data %x", data->type); + reserve_early(pa_data, pa_data+sizeof(*data)+data->len, buf); + pa_data = data->next; + early_iounmap(data, sizeof(*data)); + } +} + +/* + * --------- Crashkernel reservation ------------------------------ + */ + +#ifdef CONFIG_KEXEC + +/** + * Reserve @size bytes of crashkernel memory at any suitable offset. + * + * @size: Size of the crashkernel memory to reserve. + * Returns the base address on success, and -1ULL on failure. + */ +unsigned long long find_and_reserve_crashkernel(unsigned long long size) +{ + const unsigned long long alignment = 16<<20; /* 16M */ + unsigned long long start = 0LL; + + while (1) { + int ret; + + start = find_e820_area(start, ULONG_MAX, size, alignment); + if (start == -1ULL) + return start; + + /* try to reserve it */ + ret = reserve_bootmem_generic(start, size, BOOTMEM_EXCLUSIVE); + if (ret >= 0) + return start; + + start += alignment; + } +} - /* alloc_bootmem zeroes memory */ - cpumask_of_cpu_map = alloc_bootmem_low(sizeof(cpumask_t) * nr_cpu_ids); - for (i = 0; i < nr_cpu_ids; i++) - cpu_set(i, cpumask_of_cpu_map[i]); +static inline unsigned long long get_total_mem(void) +{ + unsigned long long total; + + total = max_low_pfn - min_low_pfn; +#ifdef CONFIG_HIGHMEM + total += highend_pfn - highstart_pfn; +#endif + + return total << PAGE_SHIFT; +} + +static void __init reserve_crashkernel(void) +{ + unsigned long long total_mem; + unsigned long long crash_size, crash_base; + int ret; + + total_mem = get_total_mem(); + + ret = parse_crashkernel(boot_command_line, total_mem, + &crash_size, &crash_base); + if (ret != 0 || crash_size <= 0) + return; + + /* 0 means: find the address automatically */ + if (crash_base <= 0) { + crash_base = find_and_reserve_crashkernel(crash_size); + if (crash_base == -1ULL) { + pr_info("crashkernel reservation failed. " + "No suitable area found.\n"); + return; + } + } else { + ret = reserve_bootmem_generic(crash_base, crash_size, + BOOTMEM_EXCLUSIVE); + if (ret < 0) { + pr_info("crashkernel reservation failed - " + "memory is in use\n"); + return; + } + } + + printk(KERN_INFO "Reserving %ldMB of memory at %ldMB " + "for crashkernel (System RAM: %ldMB)\n", + (unsigned long)(crash_size >> 20), + (unsigned long)(crash_base >> 20), + (unsigned long)(total_mem >> 20)); + + crashk_res.start = crash_base; + crashk_res.end = crash_base + crash_size - 1; + insert_resource(&iomem_resource, &crashk_res); } #else -static inline void setup_cpumask_of_cpu(void) { } +static void __init reserve_crashkernel(void) +{ +} #endif -#ifdef CONFIG_X86_32 -/* - * Great future not-so-futuristic plan: make i386 and x86_64 do it - * the same way +static struct resource standard_io_resources[] = { + { .name = "dma1", .start = 0x00, .end = 0x1f, + .flags = IORESOURCE_BUSY | IORESOURCE_IO }, + { .name = "pic1", .start = 0x20, .end = 0x21, + .flags = IORESOURCE_BUSY | IORESOURCE_IO }, + { .name = "timer0", .start = 0x40, .end = 0x43, + .flags = IORESOURCE_BUSY | IORESOURCE_IO }, + { .name = "timer1", .start = 0x50, .end = 0x53, + .flags = IORESOURCE_BUSY | IORESOURCE_IO }, + { .name = "keyboard", .start = 0x60, .end = 0x60, + .flags = IORESOURCE_BUSY | IORESOURCE_IO }, + { .name = "keyboard", .start = 0x64, .end = 0x64, + .flags = IORESOURCE_BUSY | IORESOURCE_IO }, + { .name = "dma page reg", .start = 0x80, .end = 0x8f, + .flags = IORESOURCE_BUSY | IORESOURCE_IO }, + { .name = "pic2", .start = 0xa0, .end = 0xa1, + .flags = IORESOURCE_BUSY | IORESOURCE_IO }, + { .name = "dma2", .start = 0xc0, .end = 0xdf, + .flags = IORESOURCE_BUSY | IORESOURCE_IO }, + { .name = "fpu", .start = 0xf0, .end = 0xff, + .flags = IORESOURCE_BUSY | IORESOURCE_IO } +}; + +static void __init reserve_standard_io_resources(void) +{ + int i; + + /* request I/O space for devices used on all i[345]86 PCs */ + for (i = 0; i < ARRAY_SIZE(standard_io_resources); i++) + request_resource(&ioport_resource, &standard_io_resources[i]); + +} + +#ifdef CONFIG_PROC_VMCORE +/* elfcorehdr= specifies the location of elf core header + * stored by the crashed kernel. This option will be passed + * by kexec loader to the capture kernel. */ -unsigned long __per_cpu_offset[NR_CPUS] __read_mostly; -EXPORT_SYMBOL(__per_cpu_offset); +static int __init setup_elfcorehdr(char *arg) +{ + char *end; + if (!arg) + return -EINVAL; + elfcorehdr_addr = memparse(arg, &end); + return end > arg ? 0 : -EINVAL; +} +early_param("elfcorehdr", setup_elfcorehdr); #endif /* - * Great future plan: - * Declare PDA itself and support (irqstack,tss,pgd) as per cpu data. - * Always point %gs to its beginning + * 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 + * for initialization. Note, the efi init code path is determined by the + * global efi_enabled. This allows the same kernel image to be used on existing + * systems (with a traditional BIOS) as well as on EFI systems. */ -void __init setup_per_cpu_areas(void) +/* + * setup_arch - architecture-specific boot-time initializations + * + * Note: On x86_64, fixmaps are ready for use even before this is called. + */ + +void __init setup_arch(char **cmdline_p) { - int i, highest_cpu = 0; - unsigned long size; +#ifdef CONFIG_X86_32 + memcpy(&boot_cpu_data, &new_cpu_data, sizeof(new_cpu_data)); + visws_early_detect(); + pre_setup_arch_hook(); + early_cpu_init(); +#else + printk(KERN_INFO "Command line: %s\n", boot_command_line); +#endif -#ifdef CONFIG_HOTPLUG_CPU - prefill_possible_map(); + early_ioremap_init(); + + ROOT_DEV = old_decode_dev(boot_params.hdr.root_dev); + screen_info = boot_params.screen_info; + edid_info = boot_params.edid_info; +#ifdef CONFIG_X86_32 + 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]; + } +#endif + saved_video_mode = boot_params.hdr.vid_mode; + bootloader_type = boot_params.hdr.type_of_loader; + +#ifdef CONFIG_BLK_DEV_RAM + rd_image_start = boot_params.hdr.ram_size & RAMDISK_IMAGE_START_MASK; + rd_prompt = ((boot_params.hdr.ram_size & RAMDISK_PROMPT_FLAG) != 0); + rd_doload = ((boot_params.hdr.ram_size & RAMDISK_LOAD_FLAG) != 0); #endif +#ifdef CONFIG_EFI + if (!strncmp((char *)&boot_params.efi_info.efi_loader_signature, +#ifdef CONFIG_X86_32 + "EL32", +#else + "EL64", +#endif + 4)) { + efi_enabled = 1; + efi_reserve_early(); + } +#endif + + ARCH_SETUP - /* Copy section for each CPU (we discard the original) */ - size = PERCPU_ENOUGH_ROOM; - printk(KERN_INFO "PERCPU: Allocating %lu bytes of per cpu data\n", - size); + setup_memory_map(); + parse_setup_data(); + /* update the e820_saved too */ + e820_reserve_setup_data(); - for_each_possible_cpu(i) { - char *ptr; -#ifndef CONFIG_NEED_MULTIPLE_NODES - ptr = alloc_bootmem_pages(size); + copy_edd(); + + if (!boot_params.hdr.root_flags) + root_mountflags &= ~MS_RDONLY; + init_mm.start_code = (unsigned long) _text; + init_mm.end_code = (unsigned long) _etext; + init_mm.end_data = (unsigned long) _edata; +#ifdef CONFIG_X86_32 + init_mm.brk = init_pg_tables_end + PAGE_OFFSET; #else - int node = early_cpu_to_node(i); - if (!node_online(node) || !NODE_DATA(node)) { - ptr = alloc_bootmem_pages(size); - printk(KERN_INFO - "cpu %d has no node or node-local memory\n", i); - } - else - ptr = alloc_bootmem_pages_node(NODE_DATA(node), size); + init_mm.brk = (unsigned long) &_end; #endif - if (!ptr) - panic("Cannot allocate cpu data for CPU %d\n", i); + + 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; + #ifdef CONFIG_X86_64 - cpu_pda(i)->data_offset = ptr - __per_cpu_start; + early_cpu_init(); +#endif + strlcpy(command_line, boot_command_line, COMMAND_LINE_SIZE); + *cmdline_p = command_line; + + parse_early_param(); + + /* after early param, so could get panic from serial */ + reserve_early_setup_data(); + + if (acpi_mps_check()) { +#ifdef CONFIG_X86_LOCAL_APIC + disable_apic = 1; +#endif + clear_cpu_cap(&boot_cpu_data, X86_FEATURE_APIC); + } + + finish_e820_parsing(); + +#ifdef CONFIG_X86_32 + probe_roms(); +#endif + + /* after parse_early_param, so could debug it */ + insert_resource(&iomem_resource, &code_resource); + insert_resource(&iomem_resource, &data_resource); + insert_resource(&iomem_resource, &bss_resource); + + if (efi_enabled) + efi_init(); + +#ifdef CONFIG_X86_32 + if (ppro_with_ram_bug()) { + e820_update_range(0x70000000ULL, 0x40000ULL, E820_RAM, + E820_RESERVED); + sanitize_e820_map(e820.map, ARRAY_SIZE(e820.map), &e820.nr_map); + printk(KERN_INFO "fixed physical RAM map:\n"); + e820_print_map("bad_ppro"); + } +#else + early_gart_iommu_check(); +#endif + + /* + * partially used pages are not usable - thus + * we are rounding upwards: + */ + max_pfn = e820_end_of_ram_pfn(); + + /* preallocate 4k for mptable mpc */ + early_reserve_e820_mpc_new(); + /* update e820 for memory not covered by WB MTRRs */ + mtrr_bp_init(); + if (mtrr_trim_uncached_memory(max_pfn)) + max_pfn = e820_end_of_ram_pfn(); + +#ifdef CONFIG_X86_32 + /* max_low_pfn get updated here */ + find_low_pfn_range(); #else - __per_cpu_offset[i] = ptr - __per_cpu_start; + num_physpages = max_pfn; + + check_efer(); + + /* How many end-of-memory variables you have, grandma! */ + /* need this before calling reserve_initrd */ + if (max_pfn > (1UL<<(32 - PAGE_SHIFT))) + max_low_pfn = e820_end_of_low_ram_pfn(); + else + max_low_pfn = max_pfn; + + high_memory = (void *)__va(max_pfn * PAGE_SIZE - 1) + 1; #endif - memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start); - highest_cpu = i; + /* 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; + +#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 + + /* + * NOTE: On x86-32, only from this point on, fixmaps are ready for use. + */ - nr_cpu_ids = highest_cpu + 1; - printk(KERN_DEBUG "NR_CPUS: %d, nr_cpu_ids: %d\n", NR_CPUS, nr_cpu_ids); +#ifdef CONFIG_PROVIDE_OHCI1394_DMA_INIT + if (init_ohci1394_dma_early) + init_ohci1394_dma_on_all_controllers(); +#endif - /* Setup percpu data maps */ - setup_per_cpu_maps(); + reserve_initrd(); - /* Setup cpumask_of_cpu map */ - setup_cpumask_of_cpu(); -} +#ifdef CONFIG_X86_64 + vsmp_init(); +#endif + + dmi_scan_machine(); + + io_delay_init(); + + /* + * Parse the ACPI tables for possible boot-time SMP configuration. + */ + acpi_boot_table_init(); + +#ifdef CONFIG_ACPI_NUMA + /* + * Parse SRAT to discover nodes. + */ + acpi_numa_init(); +#endif + + initmem_init(0, max_pfn); + +#ifdef CONFIG_X86_64 + dma32_reserve_bootmem(); +#endif +#ifdef CONFIG_ACPI_SLEEP + /* + * Reserve low memory region for sleep support. + */ + acpi_reserve_bootmem(); +#endif +#ifdef CONFIG_X86_FIND_SMP_CONFIG + /* + * Find and reserve possible boot-time SMP configuration: + */ + find_smp_config(); #endif + reserve_crashkernel(); + + reserve_ibft_region(); + +#ifdef CONFIG_KVM_CLOCK + kvmclock_init(); +#endif + +#if defined(CONFIG_VMI) && defined(CONFIG_X86_32) + /* + * Must be after max_low_pfn is determined, and before kernel + * pagetables are setup. + */ + vmi_init(); +#endif + + paging_init(); + +#ifdef CONFIG_X86_64 + map_vsyscall(); +#endif + +#ifdef CONFIG_X86_GENERICARCH + generic_apic_probe(); +#endif + + early_quirks(); + + /* + * Read APIC and some other early information from ACPI tables. + */ + acpi_boot_init(); + +#if defined(CONFIG_X86_MPPARSE) || defined(CONFIG_X86_VISWS) + /* + * get boot-time SMP configuration: + */ + if (smp_found_config) + get_smp_config(); +#endif + + prefill_possible_map(); +#ifdef CONFIG_X86_64 + init_cpu_to_node(); +#endif + +#ifdef CONFIG_X86_NUMAQ + /* + * need to check online nodes num, call it + * here before time_init/tsc_init + */ + numaq_tsc_disable(); +#endif + + init_apic_mappings(); + ioapic_init_mappings(); + +#if defined(CONFIG_SMP) && defined(CONFIG_X86_PC) && defined(CONFIG_X86_32) + if (def_to_bigsmp) + printk(KERN_WARNING "More than 8 CPUs detected and " + "CONFIG_X86_PC cannot handle it.\nUse " + "CONFIG_X86_GENERICARCH or CONFIG_X86_BIGSMP.\n"); +#endif + kvm_guest_init(); + + e820_reserve_resources(); + e820_mark_nosave_regions(max_low_pfn); + +#ifdef CONFIG_X86_32 + request_resource(&iomem_resource, &video_ram_resource); +#endif + reserve_standard_io_resources(); + + e820_setup_gap(); + +#ifdef CONFIG_VT +#if defined(CONFIG_VGA_CONSOLE) + if (!efi_enabled || (efi_mem_type(0xa0000) != EFI_CONVENTIONAL_MEMORY)) + conswitchp = &vga_con; +#elif defined(CONFIG_DUMMY_CONSOLE) + conswitchp = &dummy_con; +#endif +#endif +} diff --git a/arch/x86/kernel/setup64.c b/arch/x86/kernel/setup64.c deleted file mode 100644 index aee0e820077..00000000000 --- a/arch/x86/kernel/setup64.c +++ /dev/null @@ -1,287 +0,0 @@ -/* - * X86-64 specific CPU setup. - * Copyright (C) 1995 Linus Torvalds - * Copyright 2001, 2002, 2003 SuSE Labs / Andi Kleen. - * See setup.c for older changelog. - */ -#include <linux/init.h> -#include <linux/kernel.h> -#include <linux/sched.h> -#include <linux/string.h> -#include <linux/bootmem.h> -#include <linux/bitops.h> -#include <linux/module.h> -#include <linux/kgdb.h> -#include <asm/pda.h> -#include <asm/pgtable.h> -#include <asm/processor.h> -#include <asm/desc.h> -#include <asm/atomic.h> -#include <asm/mmu_context.h> -#include <asm/smp.h> -#include <asm/i387.h> -#include <asm/percpu.h> -#include <asm/proto.h> -#include <asm/sections.h> -#include <asm/setup.h> -#include <asm/genapic.h> - -#ifndef CONFIG_DEBUG_BOOT_PARAMS -struct boot_params __initdata boot_params; -#else -struct boot_params boot_params; -#endif - -cpumask_t cpu_initialized __cpuinitdata = CPU_MASK_NONE; - -struct x8664_pda *_cpu_pda[NR_CPUS] __read_mostly; -EXPORT_SYMBOL(_cpu_pda); -struct x8664_pda boot_cpu_pda[NR_CPUS] __cacheline_aligned; - -struct desc_ptr idt_descr = { 256 * 16 - 1, (unsigned long) idt_table }; - -char boot_cpu_stack[IRQSTACKSIZE] __attribute__((section(".bss.page_aligned"))); - -unsigned long __supported_pte_mask __read_mostly = ~0UL; -EXPORT_SYMBOL_GPL(__supported_pte_mask); - -static int do_not_nx __cpuinitdata = 0; - -/* noexec=on|off -Control non executable mappings for 64bit processes. - -on Enable(default) -off Disable -*/ -static int __init nonx_setup(char *str) -{ - if (!str) - return -EINVAL; - if (!strncmp(str, "on", 2)) { - __supported_pte_mask |= _PAGE_NX; - do_not_nx = 0; - } else if (!strncmp(str, "off", 3)) { - do_not_nx = 1; - __supported_pte_mask &= ~_PAGE_NX; - } - return 0; -} -early_param("noexec", nonx_setup); - -int force_personality32 = 0; - -/* noexec32=on|off -Control non executable heap for 32bit processes. -To control the stack too use noexec=off - -on PROT_READ does not imply PROT_EXEC for 32bit processes (default) -off PROT_READ implies PROT_EXEC -*/ -static int __init nonx32_setup(char *str) -{ - if (!strcmp(str, "on")) - force_personality32 &= ~READ_IMPLIES_EXEC; - else if (!strcmp(str, "off")) - force_personality32 |= READ_IMPLIES_EXEC; - return 1; -} -__setup("noexec32=", nonx32_setup); - -void pda_init(int cpu) -{ - struct x8664_pda *pda = cpu_pda(cpu); - - /* Setup up data that may be needed in __get_free_pages early */ - asm volatile("movl %0,%%fs ; movl %0,%%gs" :: "r" (0)); - /* Memory clobbers used to order PDA accessed */ - mb(); - wrmsrl(MSR_GS_BASE, pda); - mb(); - - pda->cpunumber = cpu; - pda->irqcount = -1; - pda->kernelstack = - (unsigned long)stack_thread_info() - PDA_STACKOFFSET + THREAD_SIZE; - pda->active_mm = &init_mm; - pda->mmu_state = 0; - - if (cpu == 0) { - /* others are initialized in smpboot.c */ - pda->pcurrent = &init_task; - pda->irqstackptr = boot_cpu_stack; - } else { - pda->irqstackptr = (char *) - __get_free_pages(GFP_ATOMIC, IRQSTACK_ORDER); - if (!pda->irqstackptr) - panic("cannot allocate irqstack for cpu %d", cpu); - } - - - pda->irqstackptr += IRQSTACKSIZE-64; -} - -char boot_exception_stacks[(N_EXCEPTION_STACKS - 1) * EXCEPTION_STKSZ + DEBUG_STKSZ] -__attribute__((section(".bss.page_aligned"))); - -extern asmlinkage void ignore_sysret(void); - -/* May not be marked __init: used by software suspend */ -void syscall_init(void) -{ - /* - * LSTAR and STAR live in a bit strange symbiosis. - * They both write to the same internal register. STAR allows to set CS/DS - * but only a 32bit target. LSTAR sets the 64bit rip. - */ - wrmsrl(MSR_STAR, ((u64)__USER32_CS)<<48 | ((u64)__KERNEL_CS)<<32); - wrmsrl(MSR_LSTAR, system_call); - wrmsrl(MSR_CSTAR, ignore_sysret); - -#ifdef CONFIG_IA32_EMULATION - syscall32_cpu_init (); -#endif - - /* Flags to clear on syscall */ - wrmsrl(MSR_SYSCALL_MASK, - X86_EFLAGS_TF|X86_EFLAGS_DF|X86_EFLAGS_IF|X86_EFLAGS_IOPL); -} - -void __cpuinit check_efer(void) -{ - unsigned long efer; - - rdmsrl(MSR_EFER, efer); - if (!(efer & EFER_NX) || do_not_nx) { - __supported_pte_mask &= ~_PAGE_NX; - } -} - -unsigned long kernel_eflags; - -/* - * Copies of the original ist values from the tss are only accessed during - * debugging, no special alignment required. - */ -DEFINE_PER_CPU(struct orig_ist, orig_ist); - -/* - * 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 - * 'CPU state barrier', nothing should get across. - * A lot of state is already set up in PDA init. - */ -void __cpuinit cpu_init (void) -{ - int cpu = stack_smp_processor_id(); - struct tss_struct *t = &per_cpu(init_tss, cpu); - struct orig_ist *orig_ist = &per_cpu(orig_ist, cpu); - unsigned long v; - char *estacks = NULL; - struct task_struct *me; - int i; - - /* CPU 0 is initialised in head64.c */ - if (cpu != 0) { - pda_init(cpu); - } else - estacks = boot_exception_stacks; - - me = current; - - if (cpu_test_and_set(cpu, cpu_initialized)) - panic("CPU#%d already initialized!\n", cpu); - - printk("Initializing CPU#%d\n", cpu); - - clear_in_cr4(X86_CR4_VME|X86_CR4_PVI|X86_CR4_TSD|X86_CR4_DE); - - /* - * Initialize the per-CPU GDT with the boot GDT, - * and set up the GDT descriptor: - */ - if (cpu) - memcpy(get_cpu_gdt_table(cpu), cpu_gdt_table, GDT_SIZE); - - cpu_gdt_descr[cpu].size = GDT_SIZE; - load_gdt((const struct desc_ptr *)&cpu_gdt_descr[cpu]); - load_idt((const struct desc_ptr *)&idt_descr); - - memset(me->thread.tls_array, 0, GDT_ENTRY_TLS_ENTRIES * 8); - syscall_init(); - - wrmsrl(MSR_FS_BASE, 0); - wrmsrl(MSR_KERNEL_GS_BASE, 0); - barrier(); - - check_efer(); - - /* - * set up and load the per-CPU TSS - */ - for (v = 0; v < N_EXCEPTION_STACKS; v++) { - static const unsigned int order[N_EXCEPTION_STACKS] = { - [0 ... N_EXCEPTION_STACKS - 1] = EXCEPTION_STACK_ORDER, - [DEBUG_STACK - 1] = DEBUG_STACK_ORDER - }; - if (cpu) { - estacks = (char *)__get_free_pages(GFP_ATOMIC, order[v]); - if (!estacks) - panic("Cannot allocate exception stack %ld %d\n", - v, cpu); - } - estacks += PAGE_SIZE << order[v]; - orig_ist->ist[v] = t->x86_tss.ist[v] = (unsigned long)estacks; - } - - t->x86_tss.io_bitmap_base = offsetof(struct tss_struct, io_bitmap); - /* - * <= is required because the CPU will access up to - * 8 bits beyond the end of the IO permission bitmap. - */ - for (i = 0; i <= IO_BITMAP_LONGS; i++) - t->io_bitmap[i] = ~0UL; - - atomic_inc(&init_mm.mm_count); - me->active_mm = &init_mm; - if (me->mm) - BUG(); - enter_lazy_tlb(&init_mm, me); - - set_tss_desc(cpu, t); - load_TR_desc(); - load_LDT(&init_mm.context); - -#ifdef CONFIG_KGDB - /* - * If the kgdb is connected no debug regs should be altered. This - * is only applicable when KGDB and a KGDB I/O module are built - * into the kernel and you are using early debugging with - * kgdbwait. KGDB will control the kernel HW breakpoint registers. - */ - if (kgdb_connected && arch_kgdb_ops.correct_hw_break) - arch_kgdb_ops.correct_hw_break(); - else { -#endif - /* - * Clear all 6 debug registers: - */ - - set_debugreg(0UL, 0); - set_debugreg(0UL, 1); - set_debugreg(0UL, 2); - set_debugreg(0UL, 3); - set_debugreg(0UL, 6); - set_debugreg(0UL, 7); -#ifdef CONFIG_KGDB - /* If the kgdb is connected no debug regs should be altered. */ - } -#endif - - fpu_init(); - - raw_local_save_flags(kernel_eflags); - - if (is_uv_system()) - uv_cpu_init(); -} diff --git a/arch/x86/kernel/setup_32.c b/arch/x86/kernel/setup_32.c deleted file mode 100644 index 5a2f8e06388..00000000000 --- a/arch/x86/kernel/setup_32.c +++ /dev/null @@ -1,964 +0,0 @@ -/* - * Copyright (C) 1995 Linus Torvalds - * - * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999 - * - * Memory region support - * David Parsons <orc@pell.chi.il.us>, July-August 1999 - * - * Added E820 sanitization routine (removes overlapping memory regions); - * Brian Moyle <bmoyle@mvista.com>, February 2001 - * - * Moved CPU detection code to cpu/${cpu}.c - * Patrick Mochel <mochel@osdl.org>, March 2002 - * - * Provisions for empty E820 memory regions (reported by certain BIOSes). - * Alex Achenbach <xela@slit.de>, December 2002. - * - */ - -/* - * This file handles the architecture-dependent parts of initialization - */ - -#include <linux/sched.h> -#include <linux/mm.h> -#include <linux/mmzone.h> -#include <linux/screen_info.h> -#include <linux/ioport.h> -#include <linux/acpi.h> -#include <linux/apm_bios.h> -#include <linux/initrd.h> -#include <linux/bootmem.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> -#include <linux/efi.h> -#include <linux/init.h> -#include <linux/edd.h> -#include <linux/iscsi_ibft.h> -#include <linux/nodemask.h> -#include <linux/kexec.h> -#include <linux/crash_dump.h> -#include <linux/dmi.h> -#include <linux/pfn.h> -#include <linux/pci.h> -#include <linux/init_ohci1394_dma.h> -#include <linux/kvm_para.h> - -#include <video/edid.h> - -#include <asm/mtrr.h> -#include <asm/apic.h> -#include <asm/e820.h> -#include <asm/mpspec.h> -#include <asm/mmzone.h> -#include <asm/setup.h> -#include <asm/arch_hooks.h> -#include <asm/sections.h> -#include <asm/io_apic.h> -#include <asm/ist.h> -#include <asm/io.h> -#include <asm/vmi.h> -#include <setup_arch.h> -#include <asm/bios_ebda.h> -#include <asm/cacheflush.h> -#include <asm/processor.h> - -/* This value is set up by the early boot code to point to the value - immediately after the boot time page tables. It contains a *physical* - address, and must not be in the .bss segment! */ -unsigned long init_pg_tables_end __initdata = ~0UL; - -/* - * Machine setup.. - */ -static struct resource data_resource = { - .name = "Kernel data", - .start = 0, - .end = 0, - .flags = IORESOURCE_BUSY | IORESOURCE_MEM -}; - -static struct resource code_resource = { - .name = "Kernel code", - .start = 0, - .end = 0, - .flags = IORESOURCE_BUSY | IORESOURCE_MEM -}; - -static struct resource bss_resource = { - .name = "Kernel bss", - .start = 0, - .end = 0, - .flags = IORESOURCE_BUSY | IORESOURCE_MEM -}; - -static struct resource video_ram_resource = { - .name = "Video RAM area", - .start = 0xa0000, - .end = 0xbffff, - .flags = IORESOURCE_BUSY | IORESOURCE_MEM -}; - -static struct resource standard_io_resources[] = { { - .name = "dma1", - .start = 0x0000, - .end = 0x001f, - .flags = IORESOURCE_BUSY | IORESOURCE_IO -}, { - .name = "pic1", - .start = 0x0020, - .end = 0x0021, - .flags = IORESOURCE_BUSY | IORESOURCE_IO -}, { - .name = "timer0", - .start = 0x0040, - .end = 0x0043, - .flags = IORESOURCE_BUSY | IORESOURCE_IO -}, { - .name = "timer1", - .start = 0x0050, - .end = 0x0053, - .flags = IORESOURCE_BUSY | IORESOURCE_IO -}, { - .name = "keyboard", - .start = 0x0060, - .end = 0x0060, - .flags = IORESOURCE_BUSY | IORESOURCE_IO -}, { - .name = "keyboard", - .start = 0x0064, - .end = 0x0064, - .flags = IORESOURCE_BUSY | IORESOURCE_IO -}, { - .name = "dma page reg", - .start = 0x0080, - .end = 0x008f, - .flags = IORESOURCE_BUSY | IORESOURCE_IO -}, { - .name = "pic2", - .start = 0x00a0, - .end = 0x00a1, - .flags = IORESOURCE_BUSY | IORESOURCE_IO -}, { - .name = "dma2", - .start = 0x00c0, - .end = 0x00df, - .flags = IORESOURCE_BUSY | IORESOURCE_IO -}, { - .name = "fpu", - .start = 0x00f0, - .end = 0x00ff, - .flags = IORESOURCE_BUSY | IORESOURCE_IO -} }; - -/* 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 }; -/* common cpu data for all cpus */ -struct cpuinfo_x86 boot_cpu_data __read_mostly = { 0, 0, 0, 0, -1, 1, 0, 0, -1 }; -EXPORT_SYMBOL(boot_cpu_data); - -unsigned int def_to_bigsmp; - -#ifndef CONFIG_X86_PAE -unsigned long mmu_cr4_features; -#else -unsigned long mmu_cr4_features = X86_CR4_PAE; -#endif - -/* for MCA, but anyone else can use it if they want */ -unsigned int machine_id; -unsigned int machine_submodel_id; -unsigned int BIOS_revision; - -/* Boot loader ID as an integer, for the benefit of proc_dointvec */ -int bootloader_type; - -/* user-defined highmem size */ -static unsigned int highmem_pages = -1; - -/* - * Setup options - */ -struct screen_info screen_info; -EXPORT_SYMBOL(screen_info); -struct apm_info apm_info; -EXPORT_SYMBOL(apm_info); -struct edid_info edid_info; -EXPORT_SYMBOL_GPL(edid_info); -struct ist_info ist_info; -#if defined(CONFIG_X86_SPEEDSTEP_SMI) || \ - defined(CONFIG_X86_SPEEDSTEP_SMI_MODULE) -EXPORT_SYMBOL(ist_info); -#endif - -extern void early_cpu_init(void); -extern int root_mountflags; - -unsigned long saved_video_mode; - -#define RAMDISK_IMAGE_START_MASK 0x07FF -#define RAMDISK_PROMPT_FLAG 0x8000 -#define RAMDISK_LOAD_FLAG 0x4000 - -static char __initdata command_line[COMMAND_LINE_SIZE]; - -#ifndef CONFIG_DEBUG_BOOT_PARAMS -struct boot_params __initdata boot_params; -#else -struct boot_params boot_params; -#endif - -#if defined(CONFIG_EDD) || defined(CONFIG_EDD_MODULE) -struct edd edd; -#ifdef CONFIG_EDD_MODULE -EXPORT_SYMBOL(edd); -#endif -/** - * copy_edd() - Copy the BIOS EDD information - * from boot_params into a safe place. - * - */ -static inline void copy_edd(void) -{ - memcpy(edd.mbr_signature, boot_params.edd_mbr_sig_buffer, - sizeof(edd.mbr_signature)); - memcpy(edd.edd_info, boot_params.eddbuf, sizeof(edd.edd_info)); - edd.mbr_signature_nr = boot_params.edd_mbr_sig_buf_entries; - edd.edd_info_nr = boot_params.eddbuf_entries; -} -#else -static inline void copy_edd(void) -{ -} -#endif - -int __initdata user_defined_memmap; - -/* - * "mem=nopentium" disables the 4MB page tables. - * "mem=XXX[kKmM]" defines a memory region from HIGH_MEM - * to <mem>, overriding the bios size. - * "memmap=XXX[KkmM]@XXX[KkmM]" defines a memory region from - * <start> to <start>+<mem>, overriding the bios size. - * - * HPA tells me bootloaders need to parse mem=, so no new - * option should be mem= [also see Documentation/i386/boot.txt] - */ -static int __init parse_mem(char *arg) -{ - if (!arg) - return -EINVAL; - - if (strcmp(arg, "nopentium") == 0) { - setup_clear_cpu_cap(X86_FEATURE_PSE); - } else { - /* If the user specifies memory size, we - * limit the BIOS-provided memory map to - * that size. exactmap can be used to specify - * the exact map. mem=number can be used to - * trim the existing memory map. - */ - unsigned long long mem_size; - - mem_size = memparse(arg, &arg); - limit_regions(mem_size); - user_defined_memmap = 1; - } - return 0; -} -early_param("mem", parse_mem); - -#ifdef CONFIG_PROC_VMCORE -/* elfcorehdr= specifies the location of elf core header - * stored by the crashed kernel. - */ -static int __init parse_elfcorehdr(char *arg) -{ - if (!arg) - return -EINVAL; - - elfcorehdr_addr = memparse(arg, &arg); - return 0; -} -early_param("elfcorehdr", parse_elfcorehdr); -#endif /* CONFIG_PROC_VMCORE */ - -/* - * highmem=size forces highmem to be exactly 'size' bytes. - * This works even on boxes that have no highmem otherwise. - * This also works to reduce highmem size on bigger boxes. - */ -static int __init parse_highmem(char *arg) -{ - if (!arg) - return -EINVAL; - - highmem_pages = memparse(arg, &arg) >> PAGE_SHIFT; - return 0; -} -early_param("highmem", parse_highmem); - -/* - * vmalloc=size forces the vmalloc area to be exactly 'size' - * bytes. This can be used to increase (or decrease) the - * vmalloc area - the default is 128m. - */ -static int __init parse_vmalloc(char *arg) -{ - if (!arg) - return -EINVAL; - - __VMALLOC_RESERVE = memparse(arg, &arg); - return 0; -} -early_param("vmalloc", parse_vmalloc); - -/* - * reservetop=size reserves a hole at the top of the kernel address space which - * a hypervisor can load into later. Needed for dynamically loaded hypervisors, - * so relocating the fixmap can be done before paging initialization. - */ -static int __init parse_reservetop(char *arg) -{ - unsigned long address; - - if (!arg) - return -EINVAL; - - address = memparse(arg, &arg); - reserve_top_address(address); - return 0; -} -early_param("reservetop", parse_reservetop); - -/* - * Determine low and high memory ranges: - */ -unsigned long __init find_max_low_pfn(void) -{ - unsigned long max_low_pfn; - - max_low_pfn = max_pfn; - if (max_low_pfn > MAXMEM_PFN) { - if (highmem_pages == -1) - highmem_pages = max_pfn - MAXMEM_PFN; - if (highmem_pages + MAXMEM_PFN < max_pfn) - max_pfn = MAXMEM_PFN + highmem_pages; - if (highmem_pages + MAXMEM_PFN > max_pfn) { - printk("only %luMB highmem pages available, ignoring highmem size of %uMB.\n", pages_to_mb(max_pfn - MAXMEM_PFN), pages_to_mb(highmem_pages)); - highmem_pages = 0; - } - max_low_pfn = MAXMEM_PFN; -#ifndef CONFIG_HIGHMEM - /* Maximum memory usable is what is directly addressable */ - printk(KERN_WARNING "Warning only %ldMB will be used.\n", - MAXMEM>>20); - if (max_pfn > MAX_NONPAE_PFN) - printk(KERN_WARNING "Use a HIGHMEM64G enabled kernel.\n"); - else - printk(KERN_WARNING "Use a HIGHMEM enabled kernel.\n"); - max_pfn = MAXMEM_PFN; -#else /* !CONFIG_HIGHMEM */ -#ifndef CONFIG_HIGHMEM64G - if (max_pfn > MAX_NONPAE_PFN) { - max_pfn = MAX_NONPAE_PFN; - printk(KERN_WARNING "Warning only 4GB will be used.\n"); - printk(KERN_WARNING "Use a HIGHMEM64G enabled kernel.\n"); - } -#endif /* !CONFIG_HIGHMEM64G */ -#endif /* !CONFIG_HIGHMEM */ - } else { - if (highmem_pages == -1) - highmem_pages = 0; -#ifdef CONFIG_HIGHMEM - if (highmem_pages >= max_pfn) { - printk(KERN_ERR "highmem size specified (%uMB) is bigger than pages available (%luMB)!.\n", pages_to_mb(highmem_pages), pages_to_mb(max_pfn)); - highmem_pages = 0; - } - if (highmem_pages) { - if (max_low_pfn-highmem_pages < 64*1024*1024/PAGE_SIZE){ - printk(KERN_ERR "highmem size %uMB results in smaller than 64MB lowmem, ignoring it.\n", pages_to_mb(highmem_pages)); - highmem_pages = 0; - } - max_low_pfn -= highmem_pages; - } -#else - if (highmem_pages) - printk(KERN_ERR "ignoring highmem size on non-highmem kernel!\n"); -#endif - } - return max_low_pfn; -} - -#define BIOS_LOWMEM_KILOBYTES 0x413 - -/* - * The BIOS places the EBDA/XBDA at the top of conventional - * memory, and usually decreases the reported amount of - * conventional memory (int 0x12) too. This also contains a - * workaround for Dell systems that neglect to reserve EBDA. - * The same workaround also avoids a problem with the AMD768MPX - * 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. - */ -static 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. */ - if (paravirt_enabled()) - return; - - /* end of low (conventional) memory */ - lowmem = *(unsigned short *)__va(BIOS_LOWMEM_KILOBYTES); - lowmem <<= 10; - - /* 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; - - /* 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; - - /* Paranoia: should never happen, but... */ - if ((lowmem == 0) || (lowmem >= 0x100000)) - lowmem = 0x9f000; - - /* reserve all memory between lowmem and the 1MB mark */ - reserve_bootmem(lowmem, 0x100000 - lowmem, BOOTMEM_DEFAULT); -} - -#ifndef CONFIG_NEED_MULTIPLE_NODES -static void __init setup_bootmem_allocator(void); -static unsigned long __init setup_memory(void) -{ - /* - * partially used pages are not usable - thus - * we are rounding upwards: - */ - min_low_pfn = PFN_UP(init_pg_tables_end); - - max_low_pfn = find_max_low_pfn(); - -#ifdef CONFIG_HIGHMEM - highstart_pfn = highend_pfn = max_pfn; - if (max_pfn > max_low_pfn) { - highstart_pfn = max_low_pfn; - } - printk(KERN_NOTICE "%ldMB HIGHMEM available.\n", - pages_to_mb(highend_pfn - highstart_pfn)); - num_physpages = highend_pfn; - high_memory = (void *) __va(highstart_pfn * PAGE_SIZE - 1) + 1; -#else - num_physpages = max_low_pfn; - high_memory = (void *) __va(max_low_pfn * PAGE_SIZE - 1) + 1; -#endif -#ifdef CONFIG_FLATMEM - max_mapnr = num_physpages; -#endif - printk(KERN_NOTICE "%ldMB LOWMEM available.\n", - pages_to_mb(max_low_pfn)); - - setup_bootmem_allocator(); - - return max_low_pfn; -} - -static void __init zone_sizes_init(void) -{ - unsigned long max_zone_pfns[MAX_NR_ZONES]; - memset(max_zone_pfns, 0, sizeof(max_zone_pfns)); - max_zone_pfns[ZONE_DMA] = - virt_to_phys((char *)MAX_DMA_ADDRESS) >> PAGE_SHIFT; - max_zone_pfns[ZONE_NORMAL] = max_low_pfn; -#ifdef CONFIG_HIGHMEM - max_zone_pfns[ZONE_HIGHMEM] = highend_pfn; - add_active_range(0, 0, highend_pfn); -#else - add_active_range(0, 0, max_low_pfn); -#endif - - free_area_init_nodes(max_zone_pfns); -} -#else -extern unsigned long __init setup_memory(void); -extern void zone_sizes_init(void); -#endif /* !CONFIG_NEED_MULTIPLE_NODES */ - -static inline unsigned long long get_total_mem(void) -{ - unsigned long long total; - - total = max_low_pfn - min_low_pfn; -#ifdef CONFIG_HIGHMEM - total += highend_pfn - highstart_pfn; -#endif - - return total << PAGE_SHIFT; -} - -#ifdef CONFIG_KEXEC -static void __init reserve_crashkernel(void) -{ - unsigned long long total_mem; - unsigned long long crash_size, crash_base; - int ret; - - total_mem = get_total_mem(); - - ret = parse_crashkernel(boot_command_line, total_mem, - &crash_size, &crash_base); - if (ret == 0 && crash_size > 0) { - if (crash_base > 0) { - printk(KERN_INFO "Reserving %ldMB of memory at %ldMB " - "for crashkernel (System RAM: %ldMB)\n", - (unsigned long)(crash_size >> 20), - (unsigned long)(crash_base >> 20), - (unsigned long)(total_mem >> 20)); - - if (reserve_bootmem(crash_base, crash_size, - BOOTMEM_EXCLUSIVE) < 0) { - printk(KERN_INFO "crashkernel reservation " - "failed - memory is in use\n"); - return; - } - - crashk_res.start = crash_base; - crashk_res.end = crash_base + crash_size - 1; - } else - printk(KERN_INFO "crashkernel reservation failed - " - "you have to specify a base address\n"); - } -} -#else -static inline void __init reserve_crashkernel(void) -{} -#endif - -#ifdef CONFIG_BLK_DEV_INITRD - -static bool do_relocate_initrd = false; - -static void __init reserve_initrd(void) -{ - unsigned long ramdisk_image = boot_params.hdr.ramdisk_image; - unsigned long ramdisk_size = boot_params.hdr.ramdisk_size; - unsigned long ramdisk_end = ramdisk_image + ramdisk_size; - unsigned long end_of_lowmem = max_low_pfn << PAGE_SHIFT; - unsigned long ramdisk_here; - - initrd_start = 0; - - if (!boot_params.hdr.type_of_loader || - !ramdisk_image || !ramdisk_size) - return; /* No initrd provided by bootloader */ - - if (ramdisk_end < ramdisk_image) { - printk(KERN_ERR "initrd wraps around end of memory, " - "disabling initrd\n"); - return; - } - if (ramdisk_size >= end_of_lowmem/2) { - printk(KERN_ERR "initrd too large to handle, " - "disabling initrd\n"); - return; - } - if (ramdisk_end <= end_of_lowmem) { - /* All in lowmem, easy case */ - reserve_bootmem(ramdisk_image, ramdisk_size, BOOTMEM_DEFAULT); - initrd_start = ramdisk_image + PAGE_OFFSET; - initrd_end = initrd_start+ramdisk_size; - return; - } - - /* We need to move the initrd down into lowmem */ - ramdisk_here = (end_of_lowmem - ramdisk_size) & PAGE_MASK; - - /* Note: this includes all the lowmem currently occupied by - the initrd, we rely on that fact to keep the data intact. */ - reserve_bootmem(ramdisk_here, ramdisk_size, BOOTMEM_DEFAULT); - initrd_start = ramdisk_here + PAGE_OFFSET; - initrd_end = initrd_start + ramdisk_size; - - do_relocate_initrd = true; -} - -#define MAX_MAP_CHUNK (NR_FIX_BTMAPS << PAGE_SHIFT) - -static void __init relocate_initrd(void) -{ - unsigned long ramdisk_image = boot_params.hdr.ramdisk_image; - unsigned long ramdisk_size = boot_params.hdr.ramdisk_size; - unsigned long end_of_lowmem = max_low_pfn << PAGE_SHIFT; - unsigned long ramdisk_here; - unsigned long slop, clen, mapaddr; - char *p, *q; - - if (!do_relocate_initrd) - return; - - ramdisk_here = initrd_start - PAGE_OFFSET; - - 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 */ - while (ramdisk_size) { - slop = ramdisk_image & ~PAGE_MASK; - clen = ramdisk_size; - if (clen > MAX_MAP_CHUNK-slop) - clen = MAX_MAP_CHUNK-slop; - mapaddr = ramdisk_image & PAGE_MASK; - p = early_ioremap(mapaddr, clen+slop); - memcpy(q, p+slop, clen); - early_iounmap(p, clen+slop); - q += clen; - ramdisk_image += clen; - ramdisk_size -= clen; - } -} - -#endif /* CONFIG_BLK_DEV_INITRD */ - -void __init setup_bootmem_allocator(void) -{ - unsigned long bootmap_size; - /* - * Initialize the boot-time allocator (with low memory only): - */ - bootmap_size = init_bootmem(min_low_pfn, max_low_pfn); - - register_bootmem_low_pages(max_low_pfn); - - /* - * Reserve the bootmem bitmap itself as well. We do this in two - * steps (first step was init_bootmem()) because this catches - * the (very unlikely) case of us accidentally initializing the - * bootmem allocator with an invalid RAM area. - */ - reserve_bootmem(__pa_symbol(_text), (PFN_PHYS(min_low_pfn) + - bootmap_size + PAGE_SIZE-1) - __pa_symbol(_text), - BOOTMEM_DEFAULT); - - /* - * reserve physical page 0 - it's a special BIOS page on many boxes, - * enabling clean reboots, SMP operation, laptop functions. - */ - reserve_bootmem(0, PAGE_SIZE, BOOTMEM_DEFAULT); - - /* reserve EBDA region */ - reserve_ebda_region(); - -#ifdef CONFIG_SMP - /* - * But first pinch a few for the stack/trampoline stuff - * FIXME: Don't need the extra page at 4K, but need to fix - * trampoline before removing it. (see the GDT stuff) - */ - reserve_bootmem(PAGE_SIZE, PAGE_SIZE, BOOTMEM_DEFAULT); -#endif -#ifdef CONFIG_ACPI_SLEEP - /* - * Reserve low memory region for sleep support. - */ - acpi_reserve_bootmem(); -#endif -#ifdef CONFIG_X86_FIND_SMP_CONFIG - /* - * Find and reserve possible boot-time SMP configuration: - */ - find_smp_config(); -#endif -#ifdef CONFIG_BLK_DEV_INITRD - reserve_initrd(); -#endif - numa_kva_reserve(); - reserve_crashkernel(); - - reserve_ibft_region(); -} - -/* - * The node 0 pgdat is initialized before all of these because - * it's needed for bootmem. node>0 pgdats have their virtual - * space allocated before the pagetables are in place to access - * them, so they can't be cleared then. - * - * This should all compile down to nothing when NUMA is off. - */ -static void __init remapped_pgdat_init(void) -{ - int nid; - - for_each_online_node(nid) { - if (nid != 0) - memset(NODE_DATA(nid), 0, sizeof(struct pglist_data)); - } -} - -#ifdef CONFIG_MCA -static void set_mca_bus(int x) -{ - MCA_bus = x; -} -#else -static void set_mca_bus(int x) { } -#endif - -/* Overridden in paravirt.c if CONFIG_PARAVIRT */ -char * __init __attribute__((weak)) memory_setup(void) -{ - return machine_specific_memory_setup(); -} - -#ifdef CONFIG_NUMA -/* - * In the golden day, when everything among i386 and x86_64 will be - * integrated, this will not live here - */ -void *x86_cpu_to_node_map_early_ptr; -int x86_cpu_to_node_map_init[NR_CPUS] = { - [0 ... NR_CPUS-1] = NUMA_NO_NODE -}; -DEFINE_PER_CPU(int, x86_cpu_to_node_map) = NUMA_NO_NODE; -#endif - -/* - * 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 - * for initialization. Note, the efi init code path is determined by the - * global efi_enabled. This allows the same kernel image to be used on existing - * systems (with a traditional BIOS) as well as on EFI systems. - */ -void __init setup_arch(char **cmdline_p) -{ - unsigned long max_low_pfn; - - memcpy(&boot_cpu_data, &new_cpu_data, sizeof(new_cpu_data)); - pre_setup_arch_hook(); - early_cpu_init(); - early_ioremap_init(); - -#ifdef CONFIG_EFI - if (!strncmp((char *)&boot_params.efi_info.efi_loader_signature, - "EL32", 4)) - efi_enabled = 1; -#endif - - ROOT_DEV = old_decode_dev(boot_params.hdr.root_dev); - screen_info = boot_params.screen_info; - edid_info = boot_params.edid_info; - apm_info.bios = boot_params.apm_bios_info; - ist_info = boot_params.ist_info; - saved_video_mode = boot_params.hdr.vid_mode; - 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]; - } - bootloader_type = boot_params.hdr.type_of_loader; - -#ifdef CONFIG_BLK_DEV_RAM - rd_image_start = boot_params.hdr.ram_size & RAMDISK_IMAGE_START_MASK; - rd_prompt = ((boot_params.hdr.ram_size & RAMDISK_PROMPT_FLAG) != 0); - rd_doload = ((boot_params.hdr.ram_size & RAMDISK_LOAD_FLAG) != 0); -#endif - ARCH_SETUP - - printk(KERN_INFO "BIOS-provided physical RAM map:\n"); - print_memory_map(memory_setup()); - - copy_edd(); - - if (!boot_params.hdr.root_flags) - root_mountflags &= ~MS_RDONLY; - init_mm.start_code = (unsigned long) _text; - init_mm.end_code = (unsigned long) _etext; - init_mm.end_data = (unsigned long) _edata; - init_mm.brk = init_pg_tables_end + PAGE_OFFSET; - - 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; - - parse_early_param(); - - if (user_defined_memmap) { - printk(KERN_INFO "user-defined physical RAM map:\n"); - print_memory_map("user"); - } - - strlcpy(command_line, boot_command_line, COMMAND_LINE_SIZE); - *cmdline_p = command_line; - - if (efi_enabled) - efi_init(); - - /* update e820 for memory not covered by WB MTRRs */ - propagate_e820_map(); - mtrr_bp_init(); - if (mtrr_trim_uncached_memory(max_pfn)) - propagate_e820_map(); - - max_low_pfn = setup_memory(); - -#ifdef CONFIG_KVM_CLOCK - kvmclock_init(); -#endif - -#ifdef CONFIG_VMI - /* - * Must be after max_low_pfn is determined, and before kernel - * pagetables are setup. - */ - vmi_init(); -#endif - kvm_guest_init(); - - /* - * NOTE: before this point _nobody_ is allowed to allocate - * any memory using the bootmem allocator. Although the - * allocator is now initialised only the first 8Mb of the kernel - * virtual address space has been mapped. All allocations before - * paging_init() has completed must use the alloc_bootmem_low_pages() - * variant (which allocates DMA'able memory) and care must be taken - * not to exceed the 8Mb limit. - */ - -#ifdef CONFIG_SMP - smp_alloc_memory(); /* AP processor realmode stacks in low memory*/ -#endif - paging_init(); - - /* - * NOTE: On x86-32, only from this point on, fixmaps are ready for use. - */ - -#ifdef CONFIG_PROVIDE_OHCI1394_DMA_INIT - if (init_ohci1394_dma_early) - init_ohci1394_dma_on_all_controllers(); -#endif - - remapped_pgdat_init(); - sparse_init(); - zone_sizes_init(); - - /* - * NOTE: at this point the bootmem allocator is fully available. - */ - -#ifdef CONFIG_BLK_DEV_INITRD - relocate_initrd(); -#endif - - paravirt_post_allocator_init(); - - dmi_scan_machine(); - - io_delay_init(); - -#ifdef CONFIG_X86_SMP - /* - * setup to use the early static init tables during kernel startup - * X86_SMP will exclude sub-arches that don't deal well with it. - */ - x86_cpu_to_apicid_early_ptr = (void *)x86_cpu_to_apicid_init; - x86_bios_cpu_apicid_early_ptr = (void *)x86_bios_cpu_apicid_init; -#ifdef CONFIG_NUMA - x86_cpu_to_node_map_early_ptr = (void *)x86_cpu_to_node_map_init; -#endif -#endif - -#ifdef CONFIG_X86_GENERICARCH - generic_apic_probe(); -#endif - -#ifdef CONFIG_ACPI - /* - * Parse the ACPI tables for possible boot-time SMP configuration. - */ - acpi_boot_table_init(); -#endif - - early_quirks(); - -#ifdef CONFIG_ACPI - acpi_boot_init(); - -#if defined(CONFIG_SMP) && defined(CONFIG_X86_PC) - if (def_to_bigsmp) - printk(KERN_WARNING "More than 8 CPUs detected and " - "CONFIG_X86_PC cannot handle it.\nUse " - "CONFIG_X86_GENERICARCH or CONFIG_X86_BIGSMP.\n"); -#endif -#endif -#ifdef CONFIG_X86_LOCAL_APIC - if (smp_found_config) - get_smp_config(); -#endif - - e820_register_memory(); - e820_mark_nosave_regions(); - -#ifdef CONFIG_VT -#if defined(CONFIG_VGA_CONSOLE) - if (!efi_enabled || (efi_mem_type(0xa0000) != EFI_CONVENTIONAL_MEMORY)) - conswitchp = &vga_con; -#elif defined(CONFIG_DUMMY_CONSOLE) - conswitchp = &dummy_con; -#endif -#endif -} - -/* - * Request address space for all standard resources - * - * This is called just before pcibios_init(), which is also a - * subsys_initcall, but is linked in later (in arch/i386/pci/common.c). - */ -static int __init request_standard_resources(void) -{ - int i; - - printk(KERN_INFO "Setting up standard PCI resources\n"); - init_iomem_resources(&code_resource, &data_resource, &bss_resource); - - request_resource(&iomem_resource, &video_ram_resource); - - /* request I/O space for devices used on all i[345]86 PCs */ - for (i = 0; i < ARRAY_SIZE(standard_io_resources); i++) - request_resource(&ioport_resource, &standard_io_resources[i]); - return 0; -} - -subsys_initcall(request_standard_resources); diff --git a/arch/x86/kernel/setup_64.c b/arch/x86/kernel/setup_64.c deleted file mode 100644 index 6dff1286ad8..00000000000 --- a/arch/x86/kernel/setup_64.c +++ /dev/null @@ -1,1194 +0,0 @@ -/* - * Copyright (C) 1995 Linus Torvalds - */ - -/* - * This file handles the architecture-dependent parts of initialization - */ - -#include <linux/errno.h> -#include <linux/sched.h> -#include <linux/kernel.h> -#include <linux/mm.h> -#include <linux/stddef.h> -#include <linux/unistd.h> -#include <linux/ptrace.h> -#include <linux/slab.h> -#include <linux/user.h> -#include <linux/screen_info.h> -#include <linux/ioport.h> -#include <linux/delay.h> -#include <linux/init.h> -#include <linux/initrd.h> -#include <linux/highmem.h> -#include <linux/bootmem.h> -#include <linux/module.h> -#include <asm/processor.h> -#include <linux/console.h> -#include <linux/seq_file.h> -#include <linux/crash_dump.h> -#include <linux/root_dev.h> -#include <linux/pci.h> -#include <asm/pci-direct.h> -#include <linux/efi.h> -#include <linux/acpi.h> -#include <linux/kallsyms.h> -#include <linux/edd.h> -#include <linux/iscsi_ibft.h> -#include <linux/mmzone.h> -#include <linux/kexec.h> -#include <linux/cpufreq.h> -#include <linux/dmi.h> -#include <linux/dma-mapping.h> -#include <linux/ctype.h> -#include <linux/sort.h> -#include <linux/uaccess.h> -#include <linux/init_ohci1394_dma.h> -#include <linux/kvm_para.h> - -#include <asm/mtrr.h> -#include <asm/uaccess.h> -#include <asm/system.h> -#include <asm/vsyscall.h> -#include <asm/io.h> -#include <asm/smp.h> -#include <asm/msr.h> -#include <asm/desc.h> -#include <video/edid.h> -#include <asm/e820.h> -#include <asm/dma.h> -#include <asm/gart.h> -#include <asm/mpspec.h> -#include <asm/mmu_context.h> -#include <asm/proto.h> -#include <asm/setup.h> -#include <asm/numa.h> -#include <asm/sections.h> -#include <asm/dmi.h> -#include <asm/cacheflush.h> -#include <asm/mce.h> -#include <asm/ds.h> -#include <asm/topology.h> -#include <asm/trampoline.h> -#include <asm/pat.h> - -#include <mach_apic.h> -#ifdef CONFIG_PARAVIRT -#include <asm/paravirt.h> -#else -#define ARCH_SETUP -#endif - -/* - * Machine setup.. - */ - -struct cpuinfo_x86 boot_cpu_data __read_mostly; -EXPORT_SYMBOL(boot_cpu_data); - -__u32 cleared_cpu_caps[NCAPINTS] __cpuinitdata; - -unsigned long mmu_cr4_features; - -/* Boot loader ID as an integer, for the benefit of proc_dointvec */ -int bootloader_type; - -unsigned long saved_video_mode; - -int force_mwait __cpuinitdata; - -/* - * Early DMI memory - */ -int dmi_alloc_index; -char dmi_alloc_data[DMI_MAX_DATA]; - -/* - * Setup options - */ -struct screen_info screen_info; -EXPORT_SYMBOL(screen_info); -struct sys_desc_table_struct { - unsigned short length; - unsigned char table[0]; -}; - -struct edid_info edid_info; -EXPORT_SYMBOL_GPL(edid_info); - -extern int root_mountflags; - -char __initdata command_line[COMMAND_LINE_SIZE]; - -static struct resource standard_io_resources[] = { - { .name = "dma1", .start = 0x00, .end = 0x1f, - .flags = IORESOURCE_BUSY | IORESOURCE_IO }, - { .name = "pic1", .start = 0x20, .end = 0x21, - .flags = IORESOURCE_BUSY | IORESOURCE_IO }, - { .name = "timer0", .start = 0x40, .end = 0x43, - .flags = IORESOURCE_BUSY | IORESOURCE_IO }, - { .name = "timer1", .start = 0x50, .end = 0x53, - .flags = IORESOURCE_BUSY | IORESOURCE_IO }, - { .name = "keyboard", .start = 0x60, .end = 0x60, - .flags = IORESOURCE_BUSY | IORESOURCE_IO }, - { .name = "keyboard", .start = 0x64, .end = 0x64, - .flags = IORESOURCE_BUSY | IORESOURCE_IO }, - { .name = "dma page reg", .start = 0x80, .end = 0x8f, - .flags = IORESOURCE_BUSY | IORESOURCE_IO }, - { .name = "pic2", .start = 0xa0, .end = 0xa1, - .flags = IORESOURCE_BUSY | IORESOURCE_IO }, - { .name = "dma2", .start = 0xc0, .end = 0xdf, - .flags = IORESOURCE_BUSY | IORESOURCE_IO }, - { .name = "fpu", .start = 0xf0, .end = 0xff, - .flags = IORESOURCE_BUSY | IORESOURCE_IO } -}; - -#define IORESOURCE_RAM (IORESOURCE_BUSY | IORESOURCE_MEM) - -static struct resource data_resource = { - .name = "Kernel data", - .start = 0, - .end = 0, - .flags = IORESOURCE_RAM, -}; -static struct resource code_resource = { - .name = "Kernel code", - .start = 0, - .end = 0, - .flags = IORESOURCE_RAM, -}; -static struct resource bss_resource = { - .name = "Kernel bss", - .start = 0, - .end = 0, - .flags = IORESOURCE_RAM, -}; - -static void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c); - -#ifdef CONFIG_PROC_VMCORE -/* elfcorehdr= specifies the location of elf core header - * stored by the crashed kernel. This option will be passed - * by kexec loader to the capture kernel. - */ -static int __init setup_elfcorehdr(char *arg) -{ - char *end; - if (!arg) - return -EINVAL; - elfcorehdr_addr = memparse(arg, &end); - return end > arg ? 0 : -EINVAL; -} -early_param("elfcorehdr", setup_elfcorehdr); -#endif - -#ifndef CONFIG_NUMA -static void __init -contig_initmem_init(unsigned long start_pfn, unsigned long end_pfn) -{ - unsigned long bootmap_size, bootmap; - - bootmap_size = bootmem_bootmap_pages(end_pfn)<<PAGE_SHIFT; - bootmap = find_e820_area(0, end_pfn<<PAGE_SHIFT, bootmap_size, - PAGE_SIZE); - if (bootmap == -1L) - panic("Cannot find bootmem map of size %ld\n", bootmap_size); - bootmap_size = init_bootmem(bootmap >> PAGE_SHIFT, end_pfn); - e820_register_active_regions(0, start_pfn, end_pfn); - free_bootmem_with_active_regions(0, end_pfn); - early_res_to_bootmem(0, end_pfn<<PAGE_SHIFT); - reserve_bootmem(bootmap, bootmap_size, BOOTMEM_DEFAULT); -} -#endif - -#if defined(CONFIG_EDD) || defined(CONFIG_EDD_MODULE) -struct edd edd; -#ifdef CONFIG_EDD_MODULE -EXPORT_SYMBOL(edd); -#endif -/** - * copy_edd() - Copy the BIOS EDD information - * from boot_params into a safe place. - * - */ -static inline void copy_edd(void) -{ - memcpy(edd.mbr_signature, boot_params.edd_mbr_sig_buffer, - sizeof(edd.mbr_signature)); - memcpy(edd.edd_info, boot_params.eddbuf, sizeof(edd.edd_info)); - edd.mbr_signature_nr = boot_params.edd_mbr_sig_buf_entries; - edd.edd_info_nr = boot_params.eddbuf_entries; -} -#else -static inline void copy_edd(void) -{ -} -#endif - -#ifdef CONFIG_KEXEC -static void __init reserve_crashkernel(void) -{ - unsigned long long total_mem; - unsigned long long crash_size, crash_base; - int ret; - - total_mem = ((unsigned long long)max_low_pfn - min_low_pfn) << PAGE_SHIFT; - - ret = parse_crashkernel(boot_command_line, total_mem, - &crash_size, &crash_base); - if (ret == 0 && crash_size) { - if (crash_base <= 0) { - printk(KERN_INFO "crashkernel reservation failed - " - "you have to specify a base address\n"); - return; - } - - if (reserve_bootmem(crash_base, crash_size, - BOOTMEM_EXCLUSIVE) < 0) { - printk(KERN_INFO "crashkernel reservation failed - " - "memory is in use\n"); - return; - } - - printk(KERN_INFO "Reserving %ldMB of memory at %ldMB " - "for crashkernel (System RAM: %ldMB)\n", - (unsigned long)(crash_size >> 20), - (unsigned long)(crash_base >> 20), - (unsigned long)(total_mem >> 20)); - crashk_res.start = crash_base; - crashk_res.end = crash_base + crash_size - 1; - insert_resource(&iomem_resource, &crashk_res); - } -} -#else -static inline void __init reserve_crashkernel(void) -{} -#endif - -/* Overridden in paravirt.c if CONFIG_PARAVIRT */ -void __attribute__((weak)) __init memory_setup(void) -{ - machine_specific_memory_setup(); -} - -static void __init parse_setup_data(void) -{ - struct setup_data *data; - unsigned long pa_data; - - if (boot_params.hdr.version < 0x0209) - return; - pa_data = boot_params.hdr.setup_data; - while (pa_data) { - data = early_ioremap(pa_data, PAGE_SIZE); - switch (data->type) { - default: - break; - } -#ifndef CONFIG_DEBUG_BOOT_PARAMS - free_early(pa_data, pa_data+sizeof(*data)+data->len); -#endif - pa_data = data->next; - early_iounmap(data, PAGE_SIZE); - } -} - -#ifdef CONFIG_PCI_MMCONFIG -extern void __cpuinit fam10h_check_enable_mmcfg(void); -extern void __init check_enable_amd_mmconf_dmi(void); -#else -void __cpuinit fam10h_check_enable_mmcfg(void) -{ -} -void __init check_enable_amd_mmconf_dmi(void) -{ -} -#endif - -/* - * setup_arch - architecture-specific boot-time initializations - * - * Note: On x86_64, fixmaps are ready for use even before this is called. - */ -void __init setup_arch(char **cmdline_p) -{ - unsigned i; - - printk(KERN_INFO "Command line: %s\n", boot_command_line); - - ROOT_DEV = old_decode_dev(boot_params.hdr.root_dev); - screen_info = boot_params.screen_info; - edid_info = boot_params.edid_info; - saved_video_mode = boot_params.hdr.vid_mode; - bootloader_type = boot_params.hdr.type_of_loader; - -#ifdef CONFIG_BLK_DEV_RAM - rd_image_start = boot_params.hdr.ram_size & RAMDISK_IMAGE_START_MASK; - rd_prompt = ((boot_params.hdr.ram_size & RAMDISK_PROMPT_FLAG) != 0); - rd_doload = ((boot_params.hdr.ram_size & RAMDISK_LOAD_FLAG) != 0); -#endif -#ifdef CONFIG_EFI - if (!strncmp((char *)&boot_params.efi_info.efi_loader_signature, - "EL64", 4)) - efi_enabled = 1; -#endif - - ARCH_SETUP - - memory_setup(); - copy_edd(); - - if (!boot_params.hdr.root_flags) - root_mountflags &= ~MS_RDONLY; - init_mm.start_code = (unsigned long) &_text; - init_mm.end_code = (unsigned long) &_etext; - init_mm.end_data = (unsigned long) &_edata; - init_mm.brk = (unsigned long) &_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; - - early_identify_cpu(&boot_cpu_data); - - strlcpy(command_line, boot_command_line, COMMAND_LINE_SIZE); - *cmdline_p = command_line; - - parse_setup_data(); - - parse_early_param(); - -#ifdef CONFIG_PROVIDE_OHCI1394_DMA_INIT - if (init_ohci1394_dma_early) - init_ohci1394_dma_on_all_controllers(); -#endif - - finish_e820_parsing(); - - /* after parse_early_param, so could debug it */ - insert_resource(&iomem_resource, &code_resource); - insert_resource(&iomem_resource, &data_resource); - insert_resource(&iomem_resource, &bss_resource); - - early_gart_iommu_check(); - - e820_register_active_regions(0, 0, -1UL); - /* - * partially used pages are not usable - thus - * we are rounding upwards: - */ - end_pfn = e820_end_of_ram(); - /* update e820 for memory not covered by WB MTRRs */ - mtrr_bp_init(); - if (mtrr_trim_uncached_memory(end_pfn)) { - e820_register_active_regions(0, 0, -1UL); - end_pfn = e820_end_of_ram(); - } - - num_physpages = end_pfn; - - check_efer(); - - max_pfn_mapped = init_memory_mapping(0, (max_pfn_mapped << PAGE_SHIFT)); - if (efi_enabled) - efi_init(); - - vsmp_init(); - - dmi_scan_machine(); - - io_delay_init(); - -#ifdef CONFIG_KVM_CLOCK - kvmclock_init(); -#endif - -#ifdef CONFIG_SMP - /* setup to use the early static init tables during kernel startup */ - x86_cpu_to_apicid_early_ptr = (void *)x86_cpu_to_apicid_init; - x86_bios_cpu_apicid_early_ptr = (void *)x86_bios_cpu_apicid_init; -#ifdef CONFIG_NUMA - x86_cpu_to_node_map_early_ptr = (void *)x86_cpu_to_node_map_init; -#endif -#endif - -#ifdef CONFIG_ACPI - /* - * Initialize the ACPI boot-time table parser (gets the RSDP and SDT). - * Call this early for SRAT node setup. - */ - acpi_boot_table_init(); -#endif - - /* How many end-of-memory variables you have, grandma! */ - max_low_pfn = end_pfn; - max_pfn = end_pfn; - high_memory = (void *)__va(end_pfn * PAGE_SIZE - 1) + 1; - - /* Remove active ranges so rediscovery with NUMA-awareness happens */ - remove_all_active_ranges(); - -#ifdef CONFIG_ACPI_NUMA - /* - * Parse SRAT to discover nodes. - */ - acpi_numa_init(); -#endif - -#ifdef CONFIG_NUMA - numa_initmem_init(0, end_pfn); -#else - contig_initmem_init(0, end_pfn); -#endif - - dma32_reserve_bootmem(); - -#ifdef CONFIG_ACPI_SLEEP - /* - * Reserve low memory region for sleep support. - */ - acpi_reserve_bootmem(); -#endif - - if (efi_enabled) - efi_reserve_bootmem(); - - /* - * Find and reserve possible boot-time SMP configuration: - */ - find_smp_config(); -#ifdef CONFIG_BLK_DEV_INITRD - if (boot_params.hdr.type_of_loader && boot_params.hdr.ramdisk_image) { - unsigned long ramdisk_image = boot_params.hdr.ramdisk_image; - unsigned long ramdisk_size = boot_params.hdr.ramdisk_size; - unsigned long ramdisk_end = ramdisk_image + ramdisk_size; - unsigned long end_of_mem = end_pfn << PAGE_SHIFT; - - if (ramdisk_end <= end_of_mem) { - /* - * don't need to reserve again, already reserved early - * in x86_64_start_kernel, and early_res_to_bootmem - * convert that to reserved in bootmem - */ - initrd_start = ramdisk_image + PAGE_OFFSET; - initrd_end = initrd_start+ramdisk_size; - } else { - free_bootmem(ramdisk_image, ramdisk_size); - printk(KERN_ERR "initrd extends beyond end of memory " - "(0x%08lx > 0x%08lx)\ndisabling initrd\n", - ramdisk_end, end_of_mem); - initrd_start = 0; - } - } -#endif - reserve_crashkernel(); - - reserve_ibft_region(); - - paging_init(); - map_vsyscall(); - - early_quirks(); - -#ifdef CONFIG_ACPI - /* - * Read APIC and some other early information from ACPI tables. - */ - acpi_boot_init(); -#endif - - init_cpu_to_node(); - - /* - * get boot-time SMP configuration: - */ - if (smp_found_config) - get_smp_config(); - init_apic_mappings(); - ioapic_init_mappings(); - - kvm_guest_init(); - - /* - * We trust e820 completely. No explicit ROM probing in memory. - */ - e820_reserve_resources(); - e820_mark_nosave_regions(); - - /* request I/O space for devices used on all i[345]86 PCs */ - for (i = 0; i < ARRAY_SIZE(standard_io_resources); i++) - request_resource(&ioport_resource, &standard_io_resources[i]); - - e820_setup_gap(); - -#ifdef CONFIG_VT -#if defined(CONFIG_VGA_CONSOLE) - if (!efi_enabled || (efi_mem_type(0xa0000) != EFI_CONVENTIONAL_MEMORY)) - conswitchp = &vga_con; -#elif defined(CONFIG_DUMMY_CONSOLE) - conswitchp = &dummy_con; -#endif -#endif - - /* do this before identify_cpu for boot cpu */ - check_enable_amd_mmconf_dmi(); -} - -static int __cpuinit get_model_name(struct cpuinfo_x86 *c) -{ - unsigned int *v; - - if (c->extended_cpuid_level < 0x80000004) - return 0; - - v = (unsigned int *) c->x86_model_id; - cpuid(0x80000002, &v[0], &v[1], &v[2], &v[3]); - cpuid(0x80000003, &v[4], &v[5], &v[6], &v[7]); - cpuid(0x80000004, &v[8], &v[9], &v[10], &v[11]); - c->x86_model_id[48] = 0; - return 1; -} - - -static void __cpuinit display_cacheinfo(struct cpuinfo_x86 *c) -{ - unsigned int n, dummy, eax, ebx, ecx, edx; - - n = c->extended_cpuid_level; - - if (n >= 0x80000005) { - cpuid(0x80000005, &dummy, &ebx, &ecx, &edx); - printk(KERN_INFO "CPU: L1 I Cache: %dK (%d bytes/line), " - "D cache %dK (%d bytes/line)\n", - edx>>24, edx&0xFF, ecx>>24, ecx&0xFF); - c->x86_cache_size = (ecx>>24) + (edx>>24); - /* On K8 L1 TLB is inclusive, so don't count it */ - c->x86_tlbsize = 0; - } - - if (n >= 0x80000006) { - cpuid(0x80000006, &dummy, &ebx, &ecx, &edx); - ecx = cpuid_ecx(0x80000006); - c->x86_cache_size = ecx >> 16; - c->x86_tlbsize += ((ebx >> 16) & 0xfff) + (ebx & 0xfff); - - printk(KERN_INFO "CPU: L2 Cache: %dK (%d bytes/line)\n", - c->x86_cache_size, ecx & 0xFF); - } - if (n >= 0x80000008) { - cpuid(0x80000008, &eax, &dummy, &dummy, &dummy); - c->x86_virt_bits = (eax >> 8) & 0xff; - c->x86_phys_bits = eax & 0xff; - } -} - -#ifdef CONFIG_NUMA -static int __cpuinit nearby_node(int apicid) -{ - int i, node; - - for (i = apicid - 1; i >= 0; i--) { - node = apicid_to_node[i]; - if (node != NUMA_NO_NODE && node_online(node)) - return node; - } - for (i = apicid + 1; i < MAX_LOCAL_APIC; i++) { - node = apicid_to_node[i]; - if (node != NUMA_NO_NODE && node_online(node)) - return node; - } - return first_node(node_online_map); /* Shouldn't happen */ -} -#endif - -/* - * On a AMD dual core setup the lower bits of the APIC id distingush the cores. - * Assumes number of cores is a power of two. - */ -static void __cpuinit amd_detect_cmp(struct cpuinfo_x86 *c) -{ -#ifdef CONFIG_SMP - unsigned bits; -#ifdef CONFIG_NUMA - int cpu = smp_processor_id(); - int node = 0; - unsigned apicid = hard_smp_processor_id(); -#endif - bits = c->x86_coreid_bits; - - /* Low order bits define the core id (index of core in socket) */ - c->cpu_core_id = c->initial_apicid & ((1 << bits)-1); - /* Convert the initial APIC ID into the socket ID */ - c->phys_proc_id = c->initial_apicid >> bits; - -#ifdef CONFIG_NUMA - node = c->phys_proc_id; - if (apicid_to_node[apicid] != NUMA_NO_NODE) - node = apicid_to_node[apicid]; - if (!node_online(node)) { - /* Two possibilities here: - - The CPU is missing memory and no node was created. - In that case try picking one from a nearby CPU - - The APIC IDs differ from the HyperTransport node IDs - which the K8 northbridge parsing fills in. - Assume they are all increased by a constant offset, - but in the same order as the HT nodeids. - If that doesn't result in a usable node fall back to the - path for the previous case. */ - - int ht_nodeid = c->initial_apicid; - - if (ht_nodeid >= 0 && - apicid_to_node[ht_nodeid] != NUMA_NO_NODE) - node = apicid_to_node[ht_nodeid]; - /* Pick a nearby node */ - if (!node_online(node)) - node = nearby_node(apicid); - } - numa_set_node(cpu, node); - - printk(KERN_INFO "CPU %d/%x -> Node %d\n", cpu, apicid, node); -#endif -#endif -} - -static void __cpuinit early_init_amd_mc(struct cpuinfo_x86 *c) -{ -#ifdef CONFIG_SMP - unsigned bits, ecx; - - /* Multi core CPU? */ - if (c->extended_cpuid_level < 0x80000008) - return; - - ecx = cpuid_ecx(0x80000008); - - c->x86_max_cores = (ecx & 0xff) + 1; - - /* CPU telling us the core id bits shift? */ - bits = (ecx >> 12) & 0xF; - - /* Otherwise recompute */ - if (bits == 0) { - while ((1 << bits) < c->x86_max_cores) - bits++; - } - - c->x86_coreid_bits = bits; - -#endif -} - -#define ENABLE_C1E_MASK 0x18000000 -#define CPUID_PROCESSOR_SIGNATURE 1 -#define CPUID_XFAM 0x0ff00000 -#define CPUID_XFAM_K8 0x00000000 -#define CPUID_XFAM_10H 0x00100000 -#define CPUID_XFAM_11H 0x00200000 -#define CPUID_XMOD 0x000f0000 -#define CPUID_XMOD_REV_F 0x00040000 - -/* AMD systems with C1E don't have a working lAPIC timer. Check for that. */ -static __cpuinit int amd_apic_timer_broken(void) -{ - u32 lo, hi, eax = cpuid_eax(CPUID_PROCESSOR_SIGNATURE); - - switch (eax & CPUID_XFAM) { - case CPUID_XFAM_K8: - if ((eax & CPUID_XMOD) < CPUID_XMOD_REV_F) - break; - case CPUID_XFAM_10H: - case CPUID_XFAM_11H: - rdmsr(MSR_K8_ENABLE_C1E, lo, hi); - if (lo & ENABLE_C1E_MASK) - return 1; - break; - default: - /* err on the side of caution */ - return 1; - } - return 0; -} - -static void __cpuinit early_init_amd(struct cpuinfo_x86 *c) -{ - early_init_amd_mc(c); - - /* c->x86_power is 8000_0007 edx. Bit 8 is constant TSC */ - if (c->x86_power & (1<<8)) - set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); -} - -static void __cpuinit init_amd(struct cpuinfo_x86 *c) -{ - unsigned level; - -#ifdef CONFIG_SMP - unsigned long value; - - /* - * Disable TLB flush filter by setting HWCR.FFDIS on K8 - * bit 6 of msr C001_0015 - * - * Errata 63 for SH-B3 steppings - * Errata 122 for all steppings (F+ have it disabled by default) - */ - if (c->x86 == 15) { - rdmsrl(MSR_K8_HWCR, value); - value |= 1 << 6; - wrmsrl(MSR_K8_HWCR, value); - } -#endif - - /* Bit 31 in normal CPUID used for nonstandard 3DNow ID; - 3DNow is IDd by bit 31 in extended CPUID (1*32+31) anyway */ - clear_cpu_cap(c, 0*32+31); - - /* On C+ stepping K8 rep microcode works well for copy/memset */ - level = cpuid_eax(1); - if (c->x86 == 15 && ((level >= 0x0f48 && level < 0x0f50) || - level >= 0x0f58)) - set_cpu_cap(c, X86_FEATURE_REP_GOOD); - if (c->x86 == 0x10 || c->x86 == 0x11) - set_cpu_cap(c, X86_FEATURE_REP_GOOD); - - /* Enable workaround for FXSAVE leak */ - if (c->x86 >= 6) - set_cpu_cap(c, X86_FEATURE_FXSAVE_LEAK); - - level = get_model_name(c); - if (!level) { - switch (c->x86) { - case 15: - /* Should distinguish Models here, but this is only - a fallback anyways. */ - strcpy(c->x86_model_id, "Hammer"); - break; - } - } - display_cacheinfo(c); - - /* Multi core CPU? */ - if (c->extended_cpuid_level >= 0x80000008) - amd_detect_cmp(c); - - if (c->extended_cpuid_level >= 0x80000006 && - (cpuid_edx(0x80000006) & 0xf000)) - num_cache_leaves = 4; - else - num_cache_leaves = 3; - - if (c->x86 == 0xf || c->x86 == 0x10 || c->x86 == 0x11) - set_cpu_cap(c, X86_FEATURE_K8); - - /* MFENCE stops RDTSC speculation */ - set_cpu_cap(c, X86_FEATURE_MFENCE_RDTSC); - - if (c->x86 == 0x10) - fam10h_check_enable_mmcfg(); - - if (amd_apic_timer_broken()) - disable_apic_timer = 1; - - if (c == &boot_cpu_data && c->x86 >= 0xf && c->x86 <= 0x11) { - unsigned long long tseg; - - /* - * Split up direct mapping around the TSEG SMM area. - * Don't do it for gbpages because there seems very little - * benefit in doing so. - */ - if (!rdmsrl_safe(MSR_K8_TSEG_ADDR, &tseg) && - (tseg >> PMD_SHIFT) < (max_pfn_mapped >> (PMD_SHIFT-PAGE_SHIFT))) - set_memory_4k((unsigned long)__va(tseg), 1); - } -} - -void __cpuinit detect_ht(struct cpuinfo_x86 *c) -{ -#ifdef CONFIG_SMP - u32 eax, ebx, ecx, edx; - int index_msb, core_bits; - - cpuid(1, &eax, &ebx, &ecx, &edx); - - - if (!cpu_has(c, X86_FEATURE_HT)) - return; - if (cpu_has(c, X86_FEATURE_CMP_LEGACY)) - goto out; - - smp_num_siblings = (ebx & 0xff0000) >> 16; - - if (smp_num_siblings == 1) { - printk(KERN_INFO "CPU: Hyper-Threading is disabled\n"); - } else if (smp_num_siblings > 1) { - - if (smp_num_siblings > NR_CPUS) { - printk(KERN_WARNING "CPU: Unsupported number of " - "siblings %d", smp_num_siblings); - smp_num_siblings = 1; - return; - } - - index_msb = get_count_order(smp_num_siblings); - c->phys_proc_id = phys_pkg_id(index_msb); - - smp_num_siblings = smp_num_siblings / c->x86_max_cores; - - index_msb = get_count_order(smp_num_siblings); - - core_bits = get_count_order(c->x86_max_cores); - - c->cpu_core_id = phys_pkg_id(index_msb) & - ((1 << core_bits) - 1); - } -out: - if ((c->x86_max_cores * smp_num_siblings) > 1) { - printk(KERN_INFO "CPU: Physical Processor ID: %d\n", - c->phys_proc_id); - printk(KERN_INFO "CPU: Processor Core ID: %d\n", - c->cpu_core_id); - } - -#endif -} - -/* - * find out the number of processor cores on the die - */ -static int __cpuinit intel_num_cpu_cores(struct cpuinfo_x86 *c) -{ - unsigned int eax, t; - - if (c->cpuid_level < 4) - return 1; - - cpuid_count(4, 0, &eax, &t, &t, &t); - - if (eax & 0x1f) - return ((eax >> 26) + 1); - else - return 1; -} - -static void __cpuinit srat_detect_node(void) -{ -#ifdef CONFIG_NUMA - unsigned node; - int cpu = smp_processor_id(); - int apicid = hard_smp_processor_id(); - - /* Don't do the funky fallback heuristics the AMD version employs - for now. */ - node = apicid_to_node[apicid]; - if (node == NUMA_NO_NODE || !node_online(node)) - node = first_node(node_online_map); - numa_set_node(cpu, node); - - printk(KERN_INFO "CPU %d/%x -> Node %d\n", cpu, apicid, node); -#endif -} - -static void __cpuinit early_init_intel(struct cpuinfo_x86 *c) -{ - if ((c->x86 == 0xf && c->x86_model >= 0x03) || - (c->x86 == 0x6 && c->x86_model >= 0x0e)) - set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); -} - -static void __cpuinit init_intel(struct cpuinfo_x86 *c) -{ - /* Cache sizes */ - unsigned n; - - init_intel_cacheinfo(c); - if (c->cpuid_level > 9) { - unsigned eax = cpuid_eax(10); - /* Check for version and the number of counters */ - if ((eax & 0xff) && (((eax>>8) & 0xff) > 1)) - set_cpu_cap(c, X86_FEATURE_ARCH_PERFMON); - } - - if (cpu_has_ds) { - unsigned int l1, l2; - rdmsr(MSR_IA32_MISC_ENABLE, l1, l2); - if (!(l1 & (1<<11))) - set_cpu_cap(c, X86_FEATURE_BTS); - if (!(l1 & (1<<12))) - set_cpu_cap(c, X86_FEATURE_PEBS); - } - - - if (cpu_has_bts) - ds_init_intel(c); - - n = c->extended_cpuid_level; - if (n >= 0x80000008) { - unsigned eax = cpuid_eax(0x80000008); - c->x86_virt_bits = (eax >> 8) & 0xff; - c->x86_phys_bits = eax & 0xff; - /* CPUID workaround for Intel 0F34 CPU */ - if (c->x86_vendor == X86_VENDOR_INTEL && - c->x86 == 0xF && c->x86_model == 0x3 && - c->x86_mask == 0x4) - c->x86_phys_bits = 36; - } - - if (c->x86 == 15) - c->x86_cache_alignment = c->x86_clflush_size * 2; - if (c->x86 == 6) - set_cpu_cap(c, X86_FEATURE_REP_GOOD); - set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC); - c->x86_max_cores = intel_num_cpu_cores(c); - - srat_detect_node(); -} - -static void __cpuinit early_init_centaur(struct cpuinfo_x86 *c) -{ - if (c->x86 == 0x6 && c->x86_model >= 0xf) - set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); -} - -static void __cpuinit init_centaur(struct cpuinfo_x86 *c) -{ - /* Cache sizes */ - unsigned n; - - n = c->extended_cpuid_level; - if (n >= 0x80000008) { - unsigned eax = cpuid_eax(0x80000008); - c->x86_virt_bits = (eax >> 8) & 0xff; - c->x86_phys_bits = eax & 0xff; - } - - if (c->x86 == 0x6 && c->x86_model >= 0xf) { - c->x86_cache_alignment = c->x86_clflush_size * 2; - set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); - set_cpu_cap(c, X86_FEATURE_REP_GOOD); - } - set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC); -} - -static void __cpuinit get_cpu_vendor(struct cpuinfo_x86 *c) -{ - char *v = c->x86_vendor_id; - - if (!strcmp(v, "AuthenticAMD")) - c->x86_vendor = X86_VENDOR_AMD; - else if (!strcmp(v, "GenuineIntel")) - c->x86_vendor = X86_VENDOR_INTEL; - else if (!strcmp(v, "CentaurHauls")) - c->x86_vendor = X86_VENDOR_CENTAUR; - else - c->x86_vendor = X86_VENDOR_UNKNOWN; -} - -/* Do some early cpuid on the boot CPU to get some parameter that are - needed before check_bugs. Everything advanced is in identify_cpu - below. */ -static void __cpuinit early_identify_cpu(struct cpuinfo_x86 *c) -{ - u32 tfms, xlvl; - - c->loops_per_jiffy = loops_per_jiffy; - c->x86_cache_size = -1; - c->x86_vendor = X86_VENDOR_UNKNOWN; - c->x86_model = c->x86_mask = 0; /* So far unknown... */ - c->x86_vendor_id[0] = '\0'; /* Unset */ - c->x86_model_id[0] = '\0'; /* Unset */ - c->x86_clflush_size = 64; - c->x86_cache_alignment = c->x86_clflush_size; - c->x86_max_cores = 1; - c->x86_coreid_bits = 0; - c->extended_cpuid_level = 0; - memset(&c->x86_capability, 0, sizeof c->x86_capability); - - /* Get vendor name */ - cpuid(0x00000000, (unsigned int *)&c->cpuid_level, - (unsigned int *)&c->x86_vendor_id[0], - (unsigned int *)&c->x86_vendor_id[8], - (unsigned int *)&c->x86_vendor_id[4]); - - get_cpu_vendor(c); - - /* Initialize the standard set of capabilities */ - /* Note that the vendor-specific code below might override */ - - /* Intel-defined flags: level 0x00000001 */ - if (c->cpuid_level >= 0x00000001) { - __u32 misc; - cpuid(0x00000001, &tfms, &misc, &c->x86_capability[4], - &c->x86_capability[0]); - c->x86 = (tfms >> 8) & 0xf; - c->x86_model = (tfms >> 4) & 0xf; - c->x86_mask = tfms & 0xf; - if (c->x86 == 0xf) - c->x86 += (tfms >> 20) & 0xff; - if (c->x86 >= 0x6) - c->x86_model += ((tfms >> 16) & 0xF) << 4; - if (test_cpu_cap(c, X86_FEATURE_CLFLSH)) - c->x86_clflush_size = ((misc >> 8) & 0xff) * 8; - } else { - /* Have CPUID level 0 only - unheard of */ - c->x86 = 4; - } - - c->initial_apicid = (cpuid_ebx(1) >> 24) & 0xff; -#ifdef CONFIG_SMP - c->phys_proc_id = c->initial_apicid; -#endif - /* AMD-defined flags: level 0x80000001 */ - xlvl = cpuid_eax(0x80000000); - c->extended_cpuid_level = xlvl; - if ((xlvl & 0xffff0000) == 0x80000000) { - if (xlvl >= 0x80000001) { - c->x86_capability[1] = cpuid_edx(0x80000001); - c->x86_capability[6] = cpuid_ecx(0x80000001); - } - if (xlvl >= 0x80000004) - get_model_name(c); /* Default name */ - } - - /* Transmeta-defined flags: level 0x80860001 */ - xlvl = cpuid_eax(0x80860000); - if ((xlvl & 0xffff0000) == 0x80860000) { - /* Don't set x86_cpuid_level here for now to not confuse. */ - if (xlvl >= 0x80860001) - c->x86_capability[2] = cpuid_edx(0x80860001); - } - - c->extended_cpuid_level = cpuid_eax(0x80000000); - if (c->extended_cpuid_level >= 0x80000007) - c->x86_power = cpuid_edx(0x80000007); - - switch (c->x86_vendor) { - case X86_VENDOR_AMD: - early_init_amd(c); - break; - case X86_VENDOR_INTEL: - early_init_intel(c); - break; - case X86_VENDOR_CENTAUR: - early_init_centaur(c); - break; - } - - validate_pat_support(c); -} - -/* - * This does the hard work of actually picking apart the CPU stuff... - */ -void __cpuinit identify_cpu(struct cpuinfo_x86 *c) -{ - int i; - - early_identify_cpu(c); - - init_scattered_cpuid_features(c); - - c->apicid = phys_pkg_id(0); - - /* - * Vendor-specific initialization. In this section we - * canonicalize the feature flags, meaning if there are - * features a certain CPU supports which CPUID doesn't - * tell us, CPUID claiming incorrect flags, or other bugs, - * we handle them here. - * - * At the end of this section, c->x86_capability better - * indicate the features this CPU genuinely supports! - */ - switch (c->x86_vendor) { - case X86_VENDOR_AMD: - init_amd(c); - break; - - case X86_VENDOR_INTEL: - init_intel(c); - break; - - case X86_VENDOR_CENTAUR: - init_centaur(c); - break; - - case X86_VENDOR_UNKNOWN: - default: - display_cacheinfo(c); - break; - } - - detect_ht(c); - - /* - * On SMP, boot_cpu_data holds the common feature set between - * all CPUs; so make sure that we indicate which features are - * common between the CPUs. The first time this routine gets - * executed, c == &boot_cpu_data. - */ - if (c != &boot_cpu_data) { - /* AND the already accumulated flags with these */ - for (i = 0; i < NCAPINTS; i++) - boot_cpu_data.x86_capability[i] &= c->x86_capability[i]; - } - - /* Clear all flags overriden by options */ - for (i = 0; i < NCAPINTS; i++) - c->x86_capability[i] &= ~cleared_cpu_caps[i]; - -#ifdef CONFIG_X86_MCE - mcheck_init(c); -#endif - select_idle_routine(c); - -#ifdef CONFIG_NUMA - numa_add_cpu(smp_processor_id()); -#endif - -} - -void __cpuinit identify_boot_cpu(void) -{ - identify_cpu(&boot_cpu_data); -} - -void __cpuinit identify_secondary_cpu(struct cpuinfo_x86 *c) -{ - BUG_ON(c == &boot_cpu_data); - identify_cpu(c); - mtrr_ap_init(); -} - -static __init int setup_noclflush(char *arg) -{ - setup_clear_cpu_cap(X86_FEATURE_CLFLSH); - return 1; -} -__setup("noclflush", setup_noclflush); - -void __cpuinit print_cpu_info(struct cpuinfo_x86 *c) -{ - if (c->x86_model_id[0]) - printk(KERN_CONT "%s", c->x86_model_id); - - if (c->x86_mask || c->cpuid_level >= 0) - printk(KERN_CONT " stepping %02x\n", c->x86_mask); - else - printk(KERN_CONT "\n"); -} - -static __init int setup_disablecpuid(char *arg) -{ - int bit; - if (get_option(&arg, &bit) && bit < NCAPINTS*32) - setup_clear_cpu_cap(bit); - else - return 0; - return 1; -} -__setup("clearcpuid=", setup_disablecpuid); diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c new file mode 100644 index 00000000000..cac68430d31 --- /dev/null +++ b/arch/x86/kernel/setup_percpu.c @@ -0,0 +1,399 @@ +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/init.h> +#include <linux/bootmem.h> +#include <linux/percpu.h> +#include <linux/kexec.h> +#include <linux/crash_dump.h> +#include <asm/smp.h> +#include <asm/percpu.h> +#include <asm/sections.h> +#include <asm/processor.h> +#include <asm/setup.h> +#include <asm/topology.h> +#include <asm/mpspec.h> +#include <asm/apicdef.h> +#include <asm/highmem.h> + +#ifdef CONFIG_X86_LOCAL_APIC +unsigned int num_processors; +unsigned disabled_cpus __cpuinitdata; +/* Processor that is doing the boot up */ +unsigned int boot_cpu_physical_apicid = -1U; +unsigned int max_physical_apicid; +EXPORT_SYMBOL(boot_cpu_physical_apicid); + +/* Bitmask of physically existing CPUs */ +physid_mask_t phys_cpu_present_map; +#endif + +/* 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); +EXPORT_EARLY_PER_CPU_SYMBOL(x86_cpu_to_apicid); +EXPORT_EARLY_PER_CPU_SYMBOL(x86_bios_cpu_apicid); + +#if defined(CONFIG_NUMA) && defined(CONFIG_X86_64) +#define X86_64_NUMA 1 + +/* map cpu index to node index */ +DEFINE_EARLY_PER_CPU(int, x86_cpu_to_node_map, NUMA_NO_NODE); +EXPORT_EARLY_PER_CPU_SYMBOL(x86_cpu_to_node_map); + +/* which logical CPUs are on which nodes */ +cpumask_t *node_to_cpumask_map; +EXPORT_SYMBOL(node_to_cpumask_map); + +/* setup node_to_cpumask_map */ +static void __init setup_node_to_cpumask_map(void); + +#else +static inline void setup_node_to_cpumask_map(void) { } +#endif + +#if defined(CONFIG_HAVE_SETUP_PER_CPU_AREA) && defined(CONFIG_X86_SMP) +/* + * Copy data used in early init routines from the initial arrays to the + * per cpu data areas. These arrays then become expendable and the + * *_early_ptr's are zeroed indicating that the static arrays are gone. + */ +static void __init setup_per_cpu_maps(void) +{ + int cpu; + + for_each_possible_cpu(cpu) { + per_cpu(x86_cpu_to_apicid, cpu) = + early_per_cpu_map(x86_cpu_to_apicid, cpu); + per_cpu(x86_bios_cpu_apicid, cpu) = + early_per_cpu_map(x86_bios_cpu_apicid, cpu); +#ifdef X86_64_NUMA + per_cpu(x86_cpu_to_node_map, cpu) = + early_per_cpu_map(x86_cpu_to_node_map, cpu); +#endif + } + + /* indicate the early static arrays will soon be gone */ + early_per_cpu_ptr(x86_cpu_to_apicid) = NULL; + early_per_cpu_ptr(x86_bios_cpu_apicid) = NULL; +#ifdef X86_64_NUMA + early_per_cpu_ptr(x86_cpu_to_node_map) = NULL; +#endif +} + +#ifdef CONFIG_HAVE_CPUMASK_OF_CPU_MAP +cpumask_t *cpumask_of_cpu_map __read_mostly; +EXPORT_SYMBOL(cpumask_of_cpu_map); + +/* requires nr_cpu_ids to be initialized */ +static void __init setup_cpumask_of_cpu(void) +{ + int i; + + /* alloc_bootmem zeroes memory */ + cpumask_of_cpu_map = alloc_bootmem_low(sizeof(cpumask_t) * nr_cpu_ids); + for (i = 0; i < nr_cpu_ids; i++) + cpu_set(i, cpumask_of_cpu_map[i]); +} +#else +static inline void setup_cpumask_of_cpu(void) { } +#endif + +#ifdef CONFIG_X86_32 +/* + * Great future not-so-futuristic plan: make i386 and x86_64 do it + * the same way + */ +unsigned long __per_cpu_offset[NR_CPUS] __read_mostly; +EXPORT_SYMBOL(__per_cpu_offset); +static inline void setup_cpu_pda_map(void) { } + +#elif !defined(CONFIG_SMP) +static inline void setup_cpu_pda_map(void) { } + +#else /* CONFIG_SMP && CONFIG_X86_64 */ + +/* + * Allocate cpu_pda pointer table and array via alloc_bootmem. + */ +static void __init setup_cpu_pda_map(void) +{ + char *pda; + struct x8664_pda **new_cpu_pda; + unsigned long size; + int cpu; + + size = roundup(sizeof(struct x8664_pda), cache_line_size()); + + /* allocate cpu_pda array and pointer table */ + { + unsigned long tsize = nr_cpu_ids * sizeof(void *); + unsigned long asize = size * (nr_cpu_ids - 1); + + tsize = roundup(tsize, cache_line_size()); + new_cpu_pda = alloc_bootmem(tsize + asize); + pda = (char *)new_cpu_pda + tsize; + } + + /* initialize pointer table to static pda's */ + for_each_possible_cpu(cpu) { + if (cpu == 0) { + /* leave boot cpu pda in place */ + new_cpu_pda[0] = cpu_pda(0); + continue; + } + new_cpu_pda[cpu] = (struct x8664_pda *)pda; + new_cpu_pda[cpu]->in_bootmem = 1; + pda += size; + } + + /* point to new pointer table */ + _cpu_pda = new_cpu_pda; +} +#endif + +/* + * Great future plan: + * Declare PDA itself and support (irqstack,tss,pgd) as per cpu data. + * Always point %gs to its beginning + */ +void __init setup_per_cpu_areas(void) +{ + ssize_t size = PERCPU_ENOUGH_ROOM; + char *ptr; + int cpu; + + /* Setup cpu_pda map */ + setup_cpu_pda_map(); + + /* Copy section for each CPU (we discard the original) */ + size = PERCPU_ENOUGH_ROOM; + printk(KERN_INFO "PERCPU: Allocating %zd bytes of per cpu data\n", + size); + + for_each_possible_cpu(cpu) { +#ifndef CONFIG_NEED_MULTIPLE_NODES + ptr = alloc_bootmem_pages(size); +#else + int node = early_cpu_to_node(cpu); + if (!node_online(node) || !NODE_DATA(node)) { + ptr = alloc_bootmem_pages(size); + printk(KERN_INFO + "cpu %d has no node %d or node-local memory\n", + cpu, node); + } + else + ptr = alloc_bootmem_pages_node(NODE_DATA(node), size); +#endif + per_cpu_offset(cpu) = ptr - __per_cpu_start; + memcpy(ptr, __per_cpu_start, __per_cpu_end - __per_cpu_start); + + } + + printk(KERN_DEBUG "NR_CPUS: %d, nr_cpu_ids: %d, nr_node_ids %d\n", + NR_CPUS, nr_cpu_ids, nr_node_ids); + + /* Setup percpu data maps */ + setup_per_cpu_maps(); + + /* Setup node to cpumask map */ + setup_node_to_cpumask_map(); + + /* Setup cpumask_of_cpu map */ + setup_cpumask_of_cpu(); +} + +#endif + +#ifdef X86_64_NUMA + +/* + * Allocate node_to_cpumask_map based on number of available nodes + * Requires node_possible_map to be valid. + * + * Note: node_to_cpumask() is not valid until after this is done. + */ +static void __init setup_node_to_cpumask_map(void) +{ + unsigned int node, num = 0; + cpumask_t *map; + + /* setup nr_node_ids if not done yet */ + if (nr_node_ids == MAX_NUMNODES) { + for_each_node_mask(node, node_possible_map) + num = node; + nr_node_ids = num + 1; + } + + /* allocate the map */ + map = alloc_bootmem_low(nr_node_ids * sizeof(cpumask_t)); + + Dprintk(KERN_DEBUG "Node to cpumask map at %p for %d nodes\n", + map, nr_node_ids); + + /* node_to_cpumask() will now work */ + node_to_cpumask_map = map; +} + +void __cpuinit numa_set_node(int cpu, int node) +{ + int *cpu_to_node_map = early_per_cpu_ptr(x86_cpu_to_node_map); + + if (cpu_pda(cpu) && node != NUMA_NO_NODE) + cpu_pda(cpu)->nodenumber = node; + + if (cpu_to_node_map) + cpu_to_node_map[cpu] = node; + + else if (per_cpu_offset(cpu)) + per_cpu(x86_cpu_to_node_map, cpu) = node; + + else + Dprintk(KERN_INFO "Setting node for non-present cpu %d\n", cpu); +} + +void __cpuinit numa_clear_node(int cpu) +{ + numa_set_node(cpu, NUMA_NO_NODE); +} + +#ifndef CONFIG_DEBUG_PER_CPU_MAPS + +void __cpuinit numa_add_cpu(int cpu) +{ + cpu_set(cpu, node_to_cpumask_map[early_cpu_to_node(cpu)]); +} + +void __cpuinit numa_remove_cpu(int cpu) +{ + cpu_clear(cpu, node_to_cpumask_map[cpu_to_node(cpu)]); +} + +#else /* CONFIG_DEBUG_PER_CPU_MAPS */ + +/* + * --------- debug versions of the numa functions --------- + */ +static void __cpuinit numa_set_cpumask(int cpu, int enable) +{ + int node = cpu_to_node(cpu); + cpumask_t *mask; + char buf[64]; + + if (node_to_cpumask_map == NULL) { + printk(KERN_ERR "node_to_cpumask_map NULL\n"); + dump_stack(); + return; + } + + mask = &node_to_cpumask_map[node]; + if (enable) + cpu_set(cpu, *mask); + else + cpu_clear(cpu, *mask); + + cpulist_scnprintf(buf, sizeof(buf), *mask); + printk(KERN_DEBUG "%s cpu %d node %d: mask now %s\n", + enable? "numa_add_cpu":"numa_remove_cpu", cpu, node, buf); + } + +void __cpuinit numa_add_cpu(int cpu) +{ + numa_set_cpumask(cpu, 1); +} + +void __cpuinit numa_remove_cpu(int cpu) +{ + numa_set_cpumask(cpu, 0); +} + +int cpu_to_node(int cpu) +{ + if (early_per_cpu_ptr(x86_cpu_to_node_map)) { + printk(KERN_WARNING + "cpu_to_node(%d): usage too early!\n", cpu); + dump_stack(); + return early_per_cpu_ptr(x86_cpu_to_node_map)[cpu]; + } + return per_cpu(x86_cpu_to_node_map, cpu); +} +EXPORT_SYMBOL(cpu_to_node); + +/* + * Same function as cpu_to_node() but used if called before the + * per_cpu areas are setup. + */ +int early_cpu_to_node(int cpu) +{ + if (early_per_cpu_ptr(x86_cpu_to_node_map)) + return early_per_cpu_ptr(x86_cpu_to_node_map)[cpu]; + + if (!per_cpu_offset(cpu)) { + printk(KERN_WARNING + "early_cpu_to_node(%d): no per_cpu area!\n", cpu); + dump_stack(); + return NUMA_NO_NODE; + } + return per_cpu(x86_cpu_to_node_map, cpu); +} + + +/* empty cpumask */ +static const cpumask_t cpu_mask_none; + +/* + * Returns a pointer to the bitmask of CPUs on Node 'node'. + */ +const cpumask_t *_node_to_cpumask_ptr(int node) +{ + if (node_to_cpumask_map == NULL) { + printk(KERN_WARNING + "_node_to_cpumask_ptr(%d): no node_to_cpumask_map!\n", + node); + dump_stack(); + return (const cpumask_t *)&cpu_online_map; + } + if (node >= nr_node_ids) { + printk(KERN_WARNING + "_node_to_cpumask_ptr(%d): node > nr_node_ids(%d)\n", + node, nr_node_ids); + dump_stack(); + return &cpu_mask_none; + } + return &node_to_cpumask_map[node]; +} +EXPORT_SYMBOL(_node_to_cpumask_ptr); + +/* + * Returns a bitmask of CPUs on Node 'node'. + * + * Side note: this function creates the returned cpumask on the stack + * so with a high NR_CPUS count, excessive stack space is used. The + * node_to_cpumask_ptr function should be used whenever possible. + */ +cpumask_t node_to_cpumask(int node) +{ + if (node_to_cpumask_map == NULL) { + printk(KERN_WARNING + "node_to_cpumask(%d): no node_to_cpumask_map!\n", node); + dump_stack(); + return cpu_online_map; + } + if (node >= nr_node_ids) { + printk(KERN_WARNING + "node_to_cpumask(%d): node > nr_node_ids(%d)\n", + node, nr_node_ids); + dump_stack(); + return cpu_mask_none; + } + return node_to_cpumask_map[node]; +} +EXPORT_SYMBOL(node_to_cpumask); + +/* + * --------- end of debug versions of the numa functions --------- + */ + +#endif /* CONFIG_DEBUG_PER_CPU_MAPS */ + +#endif /* X86_64_NUMA */ + diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 3e1cecedde4..f35c2d8016a 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -59,7 +59,6 @@ #include <asm/pgtable.h> #include <asm/tlbflush.h> #include <asm/mtrr.h> -#include <asm/nmi.h> #include <asm/vmi.h> #include <asm/genapic.h> #include <linux/mc146818rtc.h> @@ -68,22 +67,6 @@ #include <mach_wakecpu.h> #include <smpboot_hooks.h> -/* - * FIXME: For x86_64, those are defined in other files. But moving them here, - * would make the setup areas dependent on smp, which is a loss. When we - * integrate apic between arches, we can probably do a better job, but - * right now, they'll stay here -- glommer - */ - -/* which logical CPU number maps to which CPU (physical APIC ID) */ -u16 x86_cpu_to_apicid_init[NR_CPUS] __initdata = - { [0 ... NR_CPUS-1] = BAD_APICID }; -void *x86_cpu_to_apicid_early_ptr; - -u16 x86_bios_cpu_apicid_init[NR_CPUS] __initdata - = { [0 ... NR_CPUS-1] = BAD_APICID }; -void *x86_bios_cpu_apicid_early_ptr; - #ifdef CONFIG_X86_32 u8 apicid_2_node[MAX_APICID]; static int low_mappings; @@ -198,13 +181,12 @@ static void map_cpu_to_logical_apicid(void) map_cpu_to_node(cpu, node); } -static void unmap_cpu_to_logical_apicid(int cpu) +void numa_remove_cpu(int cpu) { cpu_2_logical_apicid[cpu] = BAD_APICID; unmap_cpu_to_node(cpu); } #else -#define unmap_cpu_to_logical_apicid(cpu) do {} while (0) #define map_cpu_to_logical_apicid() do {} while (0) #endif @@ -346,15 +328,8 @@ static void __cpuinit start_secondary(void *unused) * smp_call_function(). */ lock_ipi_call_lock(); -#ifdef CONFIG_X86_64 - spin_lock(&vector_lock); - - /* Setup the per cpu irq handling data structures */ - __setup_vector_irq(smp_processor_id()); - /* - * Allow the master to continue. - */ - spin_unlock(&vector_lock); +#ifdef CONFIG_X86_IO_APIC + setup_vector_irq(smp_processor_id()); #endif cpu_set(smp_processor_id(), cpu_online_map); unlock_ipi_call_lock(); @@ -366,31 +341,8 @@ static void __cpuinit start_secondary(void *unused) cpu_idle(); } -#ifdef CONFIG_X86_32 -/* - * Everything has been set up for the secondary - * CPUs - they just need to reload everything - * from the task structure - * This function must not return. - */ -void __devinit initialize_secondary(void) -{ - /* - * We don't actually need to load the full TSS, - * basically just the stack pointer and the ip. - */ - - asm volatile( - "movl %0,%%esp\n\t" - "jmp *%1" - : - :"m" (current->thread.sp), "m" (current->thread.ip)); -} -#endif - static void __cpuinit smp_apply_quirks(struct cpuinfo_x86 *c) { -#ifdef CONFIG_X86_32 /* * Mask B, Pentium, but not Pentium MMX */ @@ -440,7 +392,6 @@ static void __cpuinit smp_apply_quirks(struct cpuinfo_x86 *c) valid_k7: ; -#endif } static void __cpuinit smp_checks(void) @@ -555,23 +506,6 @@ cpumask_t cpu_coregroup_map(int cpu) return c->llc_shared_map; } -#ifdef CONFIG_X86_32 -/* - * We are called very early to get the low memory for the - * SMP bootup trampoline page. - */ -void __init smp_alloc_memory(void) -{ - trampoline_base = alloc_bootmem_low_pages(PAGE_SIZE); - /* - * Has to be in very low memory so we can execute - * real-mode AP code. - */ - if (__pa(trampoline_base) >= 0x9F000) - BUG(); -} -#endif - static void impress_friends(void) { int cpu; @@ -748,11 +682,7 @@ wakeup_secondary_cpu(int phys_apicid, unsigned long start_eip) * target processor state. */ startup_ipi_hook(phys_apicid, (unsigned long) start_secondary, -#ifdef CONFIG_X86_64 - (unsigned long)init_rsp); -#else (unsigned long)stack_start.sp); -#endif /* * Run STARTUP IPI loop. @@ -832,6 +762,45 @@ static void __cpuinit do_fork_idle(struct work_struct *work) complete(&c_idle->done); } +#ifdef CONFIG_X86_64 +/* + * Allocate node local memory for the AP pda. + * + * Must be called after the _cpu_pda pointer table is initialized. + */ +static int __cpuinit get_local_pda(int cpu) +{ + struct x8664_pda *oldpda, *newpda; + unsigned long size = sizeof(struct x8664_pda); + int node = cpu_to_node(cpu); + + if (cpu_pda(cpu) && !cpu_pda(cpu)->in_bootmem) + return 0; + + oldpda = cpu_pda(cpu); + newpda = kmalloc_node(size, GFP_ATOMIC, node); + if (!newpda) { + printk(KERN_ERR "Could not allocate node local PDA " + "for CPU %d on node %d\n", cpu, node); + + if (oldpda) + return 0; /* have a usable pda */ + else + return -1; + } + + if (oldpda) { + memcpy(newpda, oldpda, size); + if (!after_bootmem) + free_bootmem((unsigned long)oldpda, size); + } + + newpda->in_bootmem = 0; + cpu_pda(cpu) = newpda; + return 0; +} +#endif /* CONFIG_X86_64 */ + static int __cpuinit do_boot_cpu(int apicid, int cpu) /* * NOTE - on most systems this is a PHYSICAL apic ID, but on multiquad @@ -848,28 +817,14 @@ static int __cpuinit do_boot_cpu(int apicid, int cpu) .done = COMPLETION_INITIALIZER_ONSTACK(c_idle.done), }; INIT_WORK(&c_idle.work, do_fork_idle); -#ifdef CONFIG_X86_64 - /* allocate memory for gdts of secondary cpus. Hotplug is considered */ - if (!cpu_gdt_descr[cpu].address && - !(cpu_gdt_descr[cpu].address = get_zeroed_page(GFP_KERNEL))) { - printk(KERN_ERR "Failed to allocate GDT for CPU %d\n", cpu); - return -1; - } +#ifdef CONFIG_X86_64 /* Allocate node local memory for AP pdas */ - if (cpu_pda(cpu) == &boot_cpu_pda[cpu]) { - struct x8664_pda *newpda, *pda; - int node = cpu_to_node(cpu); - pda = cpu_pda(cpu); - newpda = kmalloc_node(sizeof(struct x8664_pda), GFP_ATOMIC, - node); - if (newpda) { - memcpy(newpda, pda, sizeof(struct x8664_pda)); - cpu_pda(cpu) = newpda; - } else - printk(KERN_ERR - "Could not allocate node local PDA for CPU %d on node %d\n", - cpu, node); + if (cpu > 0) { + boot_error = get_local_pda(cpu); + if (boot_error) + goto restore_state; + /* if can't get pda memory, can't start cpu */ } #endif @@ -905,18 +860,15 @@ do_rest: #ifdef CONFIG_X86_32 per_cpu(current_task, cpu) = c_idle.idle; init_gdt(cpu); - early_gdt_descr.address = (unsigned long)get_cpu_gdt_table(cpu); - c_idle.idle->thread.ip = (unsigned long) start_secondary; /* Stack for startup_32 can be just as for start_secondary onwards */ - stack_start.sp = (void *) c_idle.idle->thread.sp; irq_ctx_init(cpu); #else cpu_pda(cpu)->pcurrent = c_idle.idle; - init_rsp = c_idle.idle->thread.sp; - load_sp0(&per_cpu(init_tss, cpu), &c_idle.idle->thread); - initial_code = (unsigned long)start_secondary; clear_tsk_thread_flag(c_idle.idle, TIF_FORK); #endif + early_gdt_descr.address = (unsigned long)get_cpu_gdt_table(cpu); + initial_code = (unsigned long)start_secondary; + stack_start.sp = (void *) c_idle.idle->thread.sp; /* start_ip had better be page-aligned! */ start_ip = setup_trampoline(); @@ -987,13 +939,12 @@ do_rest: inquire_remote_apic(apicid); } } - - if (boot_error) { - /* Try to put things back the way they were before ... */ - unmap_cpu_to_logical_apicid(cpu); #ifdef CONFIG_X86_64 - clear_node_cpumask(cpu); /* was set by numa_add_cpu */ +restore_state: #endif + if (boot_error) { + /* Try to put things back the way they were before ... */ + numa_remove_cpu(cpu); /* was set by numa_add_cpu */ cpu_clear(cpu, cpu_callout_map); /* was set by do_boot_cpu() */ cpu_clear(cpu, cpu_initialized); /* was set by cpu_init() */ cpu_clear(cpu, cpu_present_map); @@ -1087,14 +1038,12 @@ static __init void disable_smp(void) { cpu_present_map = cpumask_of_cpu(0); cpu_possible_map = cpumask_of_cpu(0); -#ifdef CONFIG_X86_32 smpboot_clear_io_apic_irqs(); -#endif + if (smp_found_config) - phys_cpu_present_map = - physid_mask_of_physid(boot_cpu_physical_apicid); + physid_set_mask_of_physid(boot_cpu_physical_apicid, &phys_cpu_present_map); else - phys_cpu_present_map = physid_mask_of_physid(0); + physid_set_mask_of_physid(0, &phys_cpu_present_map); map_cpu_to_logical_apicid(); cpu_set(0, per_cpu(cpu_sibling_map, 0)); cpu_set(0, per_cpu(cpu_core_map, 0)); @@ -1157,12 +1106,12 @@ 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," - "forcing use of dummy APIC emulation.\n"); + printk(KERN_INFO "SMP mode deactivated.\n"); smpboot_clear_io_apic(); -#ifdef CONFIG_X86_32 + + localise_nmi_watchdog(); + connect_bsp_APIC(); -#endif setup_local_APIC(); end_local_APIC_setup(); return -1; @@ -1190,7 +1139,6 @@ static void __init smp_cpu_index_default(void) void __init native_smp_prepare_cpus(unsigned int max_cpus) { preempt_disable(); - nmi_watchdog_default(); smp_cpu_index_default(); current_cpu_data = boot_cpu_data; cpu_callin_map = cpumask_of_cpu(0); @@ -1217,9 +1165,8 @@ void __init native_smp_prepare_cpus(unsigned int max_cpus) } preempt_enable(); -#ifdef CONFIG_X86_32 connect_bsp_APIC(); -#endif + /* * Switch from PIC to APIC mode. */ @@ -1257,8 +1204,8 @@ void __init native_smp_prepare_boot_cpu(void) int me = smp_processor_id(); #ifdef CONFIG_X86_32 init_gdt(me); - switch_to_new_gdt(); #endif + switch_to_new_gdt(); /* already set me in cpu_online_map in boot_cpu_init() */ cpu_set(me, cpu_callout_map); per_cpu(cpu_state, me) = CPU_ONLINE; @@ -1278,23 +1225,6 @@ void __init native_smp_cpus_done(unsigned int max_cpus) #ifdef CONFIG_HOTPLUG_CPU -# ifdef CONFIG_X86_32 -void cpu_exit_clear(void) -{ - int cpu = raw_smp_processor_id(); - - idle_task_exit(); - - cpu_uninit(); - irq_ctx_exit(cpu); - - cpu_clear(cpu, cpu_callout_map); - cpu_clear(cpu, cpu_callin_map); - - unmap_cpu_to_logical_apicid(cpu); -} -# endif /* CONFIG_X86_32 */ - static void remove_siblinginfo(int cpu) { int sibling; @@ -1348,12 +1278,20 @@ __init void prefill_possible_map(void) int i; int possible; + /* no processor from mptable or madt */ + if (!num_processors) + num_processors = 1; + +#ifdef CONFIG_HOTPLUG_CPU if (additional_cpus == -1) { if (disabled_cpus > 0) additional_cpus = disabled_cpus; else additional_cpus = 0; } +#else + additional_cpus = 0; +#endif possible = num_processors + additional_cpus; if (possible > NR_CPUS) possible = NR_CPUS; @@ -1363,18 +1301,18 @@ __init void prefill_possible_map(void) for (i = 0; i < possible; i++) cpu_set(i, cpu_possible_map); + + nr_cpu_ids = possible; } static void __ref remove_cpu_from_maps(int cpu) { cpu_clear(cpu, cpu_online_map); -#ifdef CONFIG_X86_64 cpu_clear(cpu, cpu_callout_map); cpu_clear(cpu, cpu_callin_map); /* was set by cpu_init() */ clear_bit(cpu, (unsigned long *)&cpu_initialized); - clear_node_cpumask(cpu); -#endif + numa_remove_cpu(cpu); } int __cpu_disable(void) diff --git a/arch/x86/kernel/srat_32.c b/arch/x86/kernel/srat_32.c deleted file mode 100644 index 70e4a374b4e..00000000000 --- a/arch/x86/kernel/srat_32.c +++ /dev/null @@ -1,358 +0,0 @@ -/* - * Some of the code in this file has been gleaned from the 64 bit - * discontigmem support code base. - * - * Copyright (C) 2002, 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 Pat Gaughen <gone@us.ibm.com> - */ -#include <linux/mm.h> -#include <linux/bootmem.h> -#include <linux/mmzone.h> -#include <linux/acpi.h> -#include <linux/nodemask.h> -#include <asm/srat.h> -#include <asm/topology.h> -#include <asm/smp.h> - -/* - * proximity macros and definitions - */ -#define NODE_ARRAY_INDEX(x) ((x) / 8) /* 8 bits/char */ -#define NODE_ARRAY_OFFSET(x) ((x) % 8) /* 8 bits/char */ -#define BMAP_SET(bmap, bit) ((bmap)[NODE_ARRAY_INDEX(bit)] |= 1 << NODE_ARRAY_OFFSET(bit)) -#define BMAP_TEST(bmap, bit) ((bmap)[NODE_ARRAY_INDEX(bit)] & (1 << NODE_ARRAY_OFFSET(bit))) -/* bitmap length; _PXM is at most 255 */ -#define PXM_BITMAP_LEN (MAX_PXM_DOMAINS / 8) -static u8 pxm_bitmap[PXM_BITMAP_LEN]; /* bitmap of proximity domains */ - -#define MAX_CHUNKS_PER_NODE 3 -#define MAXCHUNKS (MAX_CHUNKS_PER_NODE * MAX_NUMNODES) -struct node_memory_chunk_s { - unsigned long start_pfn; - unsigned long end_pfn; - u8 pxm; // proximity domain of node - u8 nid; // which cnode contains this chunk? - u8 bank; // which mem bank on this node -}; -static struct node_memory_chunk_s node_memory_chunk[MAXCHUNKS]; - -static int num_memory_chunks; /* total number of memory chunks */ -static u8 __initdata apicid_to_pxm[MAX_APICID]; - -/* Identify CPU proximity domains */ -static void __init parse_cpu_affinity_structure(char *p) -{ - struct acpi_srat_cpu_affinity *cpu_affinity = - (struct acpi_srat_cpu_affinity *) p; - - if ((cpu_affinity->flags & ACPI_SRAT_CPU_ENABLED) == 0) - return; /* empty entry */ - - /* mark this node as "seen" in node bitmap */ - BMAP_SET(pxm_bitmap, cpu_affinity->proximity_domain_lo); - - apicid_to_pxm[cpu_affinity->apic_id] = cpu_affinity->proximity_domain_lo; - - printk("CPU 0x%02X in proximity domain 0x%02X\n", - cpu_affinity->apic_id, cpu_affinity->proximity_domain_lo); -} - -/* - * Identify memory proximity domains and hot-remove capabilities. - * Fill node memory chunk list structure. - */ -static void __init parse_memory_affinity_structure (char *sratp) -{ - unsigned long long paddr, size; - unsigned long start_pfn, end_pfn; - u8 pxm; - struct node_memory_chunk_s *p, *q, *pend; - struct acpi_srat_mem_affinity *memory_affinity = - (struct acpi_srat_mem_affinity *) sratp; - - if ((memory_affinity->flags & ACPI_SRAT_MEM_ENABLED) == 0) - return; /* empty entry */ - - pxm = memory_affinity->proximity_domain & 0xff; - - /* mark this node as "seen" in node bitmap */ - BMAP_SET(pxm_bitmap, pxm); - - /* calculate info for memory chunk structure */ - paddr = memory_affinity->base_address; - size = memory_affinity->length; - - start_pfn = paddr >> PAGE_SHIFT; - end_pfn = (paddr + size) >> PAGE_SHIFT; - - - if (num_memory_chunks >= MAXCHUNKS) { - printk("Too many mem chunks in SRAT. Ignoring %lld MBytes at %llx\n", - size/(1024*1024), paddr); - return; - } - - /* Insertion sort based on base address */ - pend = &node_memory_chunk[num_memory_chunks]; - for (p = &node_memory_chunk[0]; p < pend; p++) { - if (start_pfn < p->start_pfn) - break; - } - if (p < pend) { - for (q = pend; q >= p; q--) - *(q + 1) = *q; - } - p->start_pfn = start_pfn; - p->end_pfn = end_pfn; - p->pxm = pxm; - - num_memory_chunks++; - - printk("Memory range 0x%lX to 0x%lX (type 0x%X) in proximity domain 0x%02X %s\n", - start_pfn, end_pfn, - memory_affinity->memory_type, - pxm, - ((memory_affinity->flags & ACPI_SRAT_MEM_HOT_PLUGGABLE) ? - "enabled and removable" : "enabled" ) ); -} - -/* - * The SRAT table always lists ascending addresses, so can always - * assume that the first "start" address that you see is the real - * start of the node, and that the current "end" address is after - * the previous one. - */ -static __init void node_read_chunk(int nid, struct node_memory_chunk_s *memory_chunk) -{ - /* - * Only add present memory as told by the e820. - * There is no guarantee from the SRAT that the memory it - * enumerates is present at boot time because it represents - * *possible* memory hotplug areas the same as normal RAM. - */ - if (memory_chunk->start_pfn >= max_pfn) { - printk (KERN_INFO "Ignoring SRAT pfns: 0x%08lx -> %08lx\n", - memory_chunk->start_pfn, memory_chunk->end_pfn); - return; - } - if (memory_chunk->nid != nid) - return; - - if (!node_has_online_mem(nid)) - node_start_pfn[nid] = memory_chunk->start_pfn; - - if (node_start_pfn[nid] > memory_chunk->start_pfn) - node_start_pfn[nid] = memory_chunk->start_pfn; - - if (node_end_pfn[nid] < memory_chunk->end_pfn) - node_end_pfn[nid] = memory_chunk->end_pfn; -} - -/* Parse the ACPI Static Resource Affinity Table */ -static int __init acpi20_parse_srat(struct acpi_table_srat *sratp) -{ - u8 *start, *end, *p; - int i, j, nid; - - start = (u8 *)(&(sratp->reserved) + 1); /* skip header */ - p = start; - end = (u8 *)sratp + sratp->header.length; - - memset(pxm_bitmap, 0, sizeof(pxm_bitmap)); /* init proximity domain bitmap */ - memset(node_memory_chunk, 0, sizeof(node_memory_chunk)); - - num_memory_chunks = 0; - while (p < end) { - switch (*p) { - case ACPI_SRAT_TYPE_CPU_AFFINITY: - parse_cpu_affinity_structure(p); - break; - case ACPI_SRAT_TYPE_MEMORY_AFFINITY: - parse_memory_affinity_structure(p); - break; - default: - printk("ACPI 2.0 SRAT: unknown entry skipped: type=0x%02X, len=%d\n", p[0], p[1]); - break; - } - p += p[1]; - if (p[1] == 0) { - printk("acpi20_parse_srat: Entry length value is zero;" - " can't parse any further!\n"); - break; - } - } - - if (num_memory_chunks == 0) { - printk("could not finy any ACPI SRAT memory areas.\n"); - goto out_fail; - } - - /* Calculate total number of nodes in system from PXM bitmap and create - * a set of sequential node IDs starting at zero. (ACPI doesn't seem - * to specify the range of _PXM values.) - */ - /* - * MCD - we no longer HAVE to number nodes sequentially. PXM domain - * numbers could go as high as 256, and MAX_NUMNODES for i386 is typically - * 32, so we will continue numbering them in this manner until MAX_NUMNODES - * approaches MAX_PXM_DOMAINS for i386. - */ - nodes_clear(node_online_map); - for (i = 0; i < MAX_PXM_DOMAINS; i++) { - if (BMAP_TEST(pxm_bitmap, i)) { - int nid = acpi_map_pxm_to_node(i); - node_set_online(nid); - } - } - BUG_ON(num_online_nodes() == 0); - - /* set cnode id in memory chunk structure */ - for (i = 0; i < num_memory_chunks; i++) - node_memory_chunk[i].nid = pxm_to_node(node_memory_chunk[i].pxm); - - printk("pxm bitmap: "); - for (i = 0; i < sizeof(pxm_bitmap); i++) { - printk("%02X ", pxm_bitmap[i]); - } - printk("\n"); - printk("Number of logical nodes in system = %d\n", num_online_nodes()); - printk("Number of memory chunks in system = %d\n", num_memory_chunks); - - for (i = 0; i < MAX_APICID; i++) - apicid_2_node[i] = pxm_to_node(apicid_to_pxm[i]); - - for (j = 0; j < num_memory_chunks; j++){ - struct node_memory_chunk_s * chunk = &node_memory_chunk[j]; - printk("chunk %d nid %d start_pfn %08lx end_pfn %08lx\n", - j, chunk->nid, chunk->start_pfn, chunk->end_pfn); - node_read_chunk(chunk->nid, chunk); - add_active_range(chunk->nid, chunk->start_pfn, chunk->end_pfn); - } - - for_each_online_node(nid) { - unsigned long start = node_start_pfn[nid]; - unsigned long end = node_end_pfn[nid]; - - memory_present(nid, start, end); - node_remap_size[nid] = node_memmap_size_bytes(nid, start, end); - } - return 1; -out_fail: - return 0; -} - -struct acpi_static_rsdt { - struct acpi_table_rsdt table; - u32 padding[7]; /* Allow for 7 more table entries */ -}; - -int __init get_memcfg_from_srat(void) -{ - struct acpi_table_header *header = NULL; - struct acpi_table_rsdp *rsdp = NULL; - struct acpi_table_rsdt *rsdt = NULL; - acpi_native_uint rsdp_address = 0; - struct acpi_static_rsdt saved_rsdt; - int tables = 0; - int i = 0; - - rsdp_address = acpi_os_get_root_pointer(); - if (!rsdp_address) { - printk("%s: System description tables not found\n", - __func__); - goto out_err; - } - - printk("%s: assigning address to rsdp\n", __func__); - rsdp = (struct acpi_table_rsdp *)(u32)rsdp_address; - if (!rsdp) { - printk("%s: Didn't find ACPI root!\n", __func__); - goto out_err; - } - - printk(KERN_INFO "%.8s v%d [%.6s]\n", rsdp->signature, rsdp->revision, - rsdp->oem_id); - - if (strncmp(rsdp->signature, ACPI_SIG_RSDP,strlen(ACPI_SIG_RSDP))) { - printk(KERN_WARNING "%s: RSDP table signature incorrect\n", __func__); - goto out_err; - } - - rsdt = (struct acpi_table_rsdt *) - early_ioremap(rsdp->rsdt_physical_address, sizeof(struct acpi_table_rsdt)); - - if (!rsdt) { - printk(KERN_WARNING - "%s: ACPI: Invalid root system description tables (RSDT)\n", - __func__); - goto out_err; - } - - header = &rsdt->header; - - if (strncmp(header->signature, ACPI_SIG_RSDT, strlen(ACPI_SIG_RSDT))) { - printk(KERN_WARNING "ACPI: RSDT signature incorrect\n"); - goto out_err; - } - - /* - * The number of tables is computed by taking the - * size of all entries (header size minus total - * size of RSDT) divided by the size of each entry - * (4-byte table pointers). - */ - tables = (header->length - sizeof(struct acpi_table_header)) / 4; - - if (!tables) - goto out_err; - - memcpy(&saved_rsdt, rsdt, sizeof(saved_rsdt)); - - if (saved_rsdt.table.header.length > sizeof(saved_rsdt)) { - printk(KERN_WARNING "ACPI: Too big length in RSDT: %d\n", - saved_rsdt.table.header.length); - goto out_err; - } - - printk("Begin SRAT table scan....\n"); - - for (i = 0; i < tables; i++) { - /* Map in header, then map in full table length. */ - header = (struct acpi_table_header *) - early_ioremap(saved_rsdt.table.table_offset_entry[i], sizeof(struct acpi_table_header)); - if (!header) - break; - header = (struct acpi_table_header *) - early_ioremap(saved_rsdt.table.table_offset_entry[i], header->length); - if (!header) - break; - - if (strncmp((char *) &header->signature, ACPI_SIG_SRAT, 4)) - continue; - - /* we've found the srat table. don't need to look at any more tables */ - return acpi20_parse_srat((struct acpi_table_srat *)header); - } -out_err: - remove_all_active_ranges(); - printk("failed to get NUMA memory information from SRAT table\n"); - return 0; -} diff --git a/arch/x86/kernel/summit_32.c b/arch/x86/kernel/summit_32.c index ae751094eba..d67ce5f044b 100644 --- a/arch/x86/kernel/summit_32.c +++ b/arch/x86/kernel/summit_32.c @@ -36,7 +36,9 @@ static struct rio_table_hdr *rio_table_hdr __initdata; static struct scal_detail *scal_devs[MAX_NUMNODES] __initdata; static struct rio_detail *rio_devs[MAX_NUMNODES*4] __initdata; +#ifndef CONFIG_X86_NUMAQ static int mp_bus_id_to_node[MAX_MP_BUSSES] __initdata; +#endif static int __init setup_pci_node_map_for_wpeg(int wpeg_num, int last_bus) { diff --git a/arch/x86/kernel/sys_i386_32.c b/arch/x86/kernel/sys_i386_32.c index d2ab52cc1d6..7066cb855a6 100644 --- a/arch/x86/kernel/sys_i386_32.c +++ b/arch/x86/kernel/sys_i386_32.c @@ -19,8 +19,8 @@ #include <linux/utsname.h> #include <linux/ipc.h> -#include <asm/uaccess.h> -#include <asm/unistd.h> +#include <linux/uaccess.h> +#include <linux/unistd.h> asmlinkage long sys_mmap2(unsigned long addr, unsigned long len, unsigned long prot, unsigned long flags, @@ -103,7 +103,7 @@ asmlinkage int old_select(struct sel_arg_struct __user *arg) * * This is really horribly ugly. */ -asmlinkage int sys_ipc (uint call, int first, int second, +asmlinkage int sys_ipc(uint call, int first, int second, int third, void __user *ptr, long fifth) { int version, ret; @@ -113,24 +113,24 @@ asmlinkage int sys_ipc (uint call, int first, int second, switch (call) { case SEMOP: - return sys_semtimedop (first, (struct sembuf __user *)ptr, second, NULL); + return sys_semtimedop(first, (struct sembuf __user *)ptr, second, NULL); case SEMTIMEDOP: return sys_semtimedop(first, (struct sembuf __user *)ptr, second, (const struct timespec __user *)fifth); case SEMGET: - return sys_semget (first, second, third); + return sys_semget(first, second, third); case SEMCTL: { union semun fourth; if (!ptr) return -EINVAL; if (get_user(fourth.__pad, (void __user * __user *) ptr)) return -EFAULT; - return sys_semctl (first, second, third, fourth); + return sys_semctl(first, second, third, fourth); } case MSGSND: - return sys_msgsnd (first, (struct msgbuf __user *) ptr, + return sys_msgsnd(first, (struct msgbuf __user *) ptr, second, third); case MSGRCV: switch (version) { @@ -138,45 +138,45 @@ asmlinkage int sys_ipc (uint call, int first, int second, struct ipc_kludge tmp; if (!ptr) return -EINVAL; - + if (copy_from_user(&tmp, - (struct ipc_kludge __user *) ptr, - sizeof (tmp))) + (struct ipc_kludge __user *) ptr, + sizeof(tmp))) return -EFAULT; - return sys_msgrcv (first, tmp.msgp, second, + return sys_msgrcv(first, tmp.msgp, second, tmp.msgtyp, third); } default: - return sys_msgrcv (first, + return sys_msgrcv(first, (struct msgbuf __user *) ptr, second, fifth, third); } case MSGGET: - return sys_msgget ((key_t) first, second); + return sys_msgget((key_t) first, second); case MSGCTL: - return sys_msgctl (first, second, (struct msqid_ds __user *) ptr); + return sys_msgctl(first, second, (struct msqid_ds __user *) ptr); case SHMAT: switch (version) { default: { ulong raddr; - ret = do_shmat (first, (char __user *) ptr, second, &raddr); + ret = do_shmat(first, (char __user *) ptr, second, &raddr); if (ret) return ret; - return put_user (raddr, (ulong __user *) third); + return put_user(raddr, (ulong __user *) third); } case 1: /* iBCS2 emulator entry point */ if (!segment_eq(get_fs(), get_ds())) return -EINVAL; /* The "(ulong *) third" is valid _only_ because of the kernel segment thing */ - return do_shmat (first, (char __user *) ptr, second, (ulong *) third); + return do_shmat(first, (char __user *) ptr, second, (ulong *) third); } - case SHMDT: - return sys_shmdt ((char __user *)ptr); + case SHMDT: + return sys_shmdt((char __user *)ptr); case SHMGET: - return sys_shmget (first, second, third); + return sys_shmget(first, second, third); case SHMCTL: - return sys_shmctl (first, second, + return sys_shmctl(first, second, (struct shmid_ds __user *) ptr); default: return -ENOSYS; @@ -186,28 +186,28 @@ asmlinkage int sys_ipc (uint call, int first, int second, /* * Old cruft */ -asmlinkage int sys_uname(struct old_utsname __user * name) +asmlinkage int sys_uname(struct old_utsname __user *name) { int err; if (!name) return -EFAULT; down_read(&uts_sem); - err = copy_to_user(name, utsname(), sizeof (*name)); + err = copy_to_user(name, utsname(), sizeof(*name)); up_read(&uts_sem); - return err?-EFAULT:0; + return err? -EFAULT:0; } -asmlinkage int sys_olduname(struct oldold_utsname __user * name) +asmlinkage int sys_olduname(struct oldold_utsname __user *name) { int error; if (!name) return -EFAULT; - if (!access_ok(VERIFY_WRITE,name,sizeof(struct oldold_utsname))) + if (!access_ok(VERIFY_WRITE, name, sizeof(struct oldold_utsname))) return -EFAULT; - - down_read(&uts_sem); - + + down_read(&uts_sem); + error = __copy_to_user(&name->sysname, &utsname()->sysname, __OLD_UTS_LEN); error |= __put_user(0, name->sysname + __OLD_UTS_LEN); @@ -223,9 +223,9 @@ asmlinkage int sys_olduname(struct oldold_utsname __user * name) error |= __copy_to_user(&name->machine, &utsname()->machine, __OLD_UTS_LEN); error |= __put_user(0, name->machine + __OLD_UTS_LEN); - + up_read(&uts_sem); - + error = error ? -EFAULT : 0; return error; @@ -241,6 +241,6 @@ int kernel_execve(const char *filename, char *const argv[], char *const envp[]) long __res; asm volatile ("push %%ebx ; movl %2,%%ebx ; int $0x80 ; pop %%ebx" : "=a" (__res) - : "0" (__NR_execve),"ri" (filename),"c" (argv), "d" (envp) : "memory"); + : "0" (__NR_execve), "ri" (filename), "c" (argv), "d" (envp) : "memory"); return __res; } diff --git a/arch/x86/kernel/time_32.c b/arch/x86/kernel/time_32.c index 2ff21f39893..059ca6ee59b 100644 --- a/arch/x86/kernel/time_32.c +++ b/arch/x86/kernel/time_32.c @@ -39,9 +39,6 @@ #include "do_timer.h" -unsigned int cpu_khz; /* Detected as we calibrate the TSC */ -EXPORT_SYMBOL(cpu_khz); - int timer_ack; unsigned long profile_pc(struct pt_regs *regs) @@ -84,8 +81,7 @@ irqreturn_t timer_interrupt(int irq, void *dev_id) if (timer_ack) { /* * Subtle, when I/O APICs are used we have to ack timer IRQ - * manually to reset the IRR bit for do_slow_gettimeoffset(). - * This will also deassert NMI lines for the watchdog if run + * manually to deassert NMI lines for the watchdog if run * on an 82489DX-based system. */ spin_lock(&i8259A_lock); diff --git a/arch/x86/kernel/time_64.c b/arch/x86/kernel/time_64.c index c737849e2ef..e3d49c553af 100644 --- a/arch/x86/kernel/time_64.c +++ b/arch/x86/kernel/time_64.c @@ -56,7 +56,7 @@ static irqreturn_t timer_event_interrupt(int irq, void *dev_id) /* calibrate_cpu is used on systems with fixed rate TSCs to determine * processor frequency */ #define TICK_COUNT 100000000 -unsigned long __init native_calculate_cpu_khz(void) +unsigned long __init calibrate_cpu(void) { int tsc_start, tsc_now; int i, no_ctr_free; @@ -116,23 +116,11 @@ void __init hpet_time_init(void) void __init time_init(void) { - tsc_calibrate(); - - cpu_khz = tsc_khz; - if (cpu_has(&boot_cpu_data, X86_FEATURE_CONSTANT_TSC) && - (boot_cpu_data.x86_vendor == X86_VENDOR_AMD)) - cpu_khz = calculate_cpu_khz(); - - if (unsynchronized_tsc()) - mark_tsc_unstable("TSCs unsynchronized"); - + tsc_init(); if (cpu_has(&boot_cpu_data, X86_FEATURE_RDTSCP)) vgetcpu_mode = VGETCPU_RDTSCP; else vgetcpu_mode = VGETCPU_LSL; - printk(KERN_INFO "time.c: Detected %d.%03d MHz processor.\n", - cpu_khz / 1000, cpu_khz % 1000); - init_tsc_clocksource(); late_time_init = choose_time_init(); } diff --git a/arch/x86/kernel/tlb_64.c b/arch/x86/kernel/tlb_64.c index a1f07d79320..5039d0f097a 100644 --- a/arch/x86/kernel/tlb_64.c +++ b/arch/x86/kernel/tlb_64.c @@ -15,6 +15,8 @@ #include <asm/proto.h> #include <asm/apicdef.h> #include <asm/idle.h> +#include <asm/uv/uv_hub.h> +#include <asm/uv/uv_bau.h> #include <mach_ipi.h> /* @@ -162,6 +164,9 @@ void native_flush_tlb_others(const cpumask_t *cpumaskp, struct mm_struct *mm, union smp_flush_state *f; cpumask_t cpumask = *cpumaskp; + if (is_uv_system() && uv_flush_tlb_others(&cpumask, mm, va)) + return; + /* Caller has disabled preemption */ sender = smp_processor_id() % NUM_INVALIDATE_TLB_VECTORS; f = &per_cpu(flush_state, sender); diff --git a/arch/x86/kernel/tlb_uv.c b/arch/x86/kernel/tlb_uv.c new file mode 100644 index 00000000000..d0fbb7712ab --- /dev/null +++ b/arch/x86/kernel/tlb_uv.c @@ -0,0 +1,792 @@ +/* + * SGI UltraViolet TLB flush routines. + * + * (c) 2008 Cliff Wickman <cpw@sgi.com>, SGI. + * + * This code is released under the GNU General Public License version 2 or + * later. + */ +#include <linux/mc146818rtc.h> +#include <linux/proc_fs.h> +#include <linux/kernel.h> + +#include <asm/mmu_context.h> +#include <asm/uv/uv_mmrs.h> +#include <asm/uv/uv_hub.h> +#include <asm/uv/uv_bau.h> +#include <asm/genapic.h> +#include <asm/idle.h> +#include <asm/tsc.h> + +#include <mach_apic.h> + +static struct bau_control **uv_bau_table_bases __read_mostly; +static int uv_bau_retry_limit __read_mostly; + +/* position of pnode (which is nasid>>1): */ +static int uv_nshift __read_mostly; + +static unsigned long uv_mmask __read_mostly; + +static DEFINE_PER_CPU(struct ptc_stats, ptcstats); +static DEFINE_PER_CPU(struct bau_control, bau_control); + +/* + * Free a software acknowledge hardware resource by clearing its Pending + * bit. This will return a reply to the sender. + * If the message has timed out, a reply has already been sent by the + * hardware but the resource has not been released. In that case our + * clear of the Timeout bit (as well) will free the resource. No reply will + * be sent (the hardware will only do one reply per message). + */ +static void uv_reply_to_message(int resource, + struct bau_payload_queue_entry *msg, + struct bau_msg_status *msp) +{ + unsigned long dw; + + dw = (1 << (resource + UV_SW_ACK_NPENDING)) | (1 << resource); + msg->replied_to = 1; + msg->sw_ack_vector = 0; + if (msp) + msp->seen_by.bits = 0; + uv_write_local_mmr(UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_ALIAS, dw); +} + +/* + * Do all the things a cpu should do for a TLB shootdown message. + * Other cpu's may come here at the same time for this message. + */ +static void uv_bau_process_message(struct bau_payload_queue_entry *msg, + int msg_slot, int sw_ack_slot) +{ + unsigned long this_cpu_mask; + struct bau_msg_status *msp; + int cpu; + + msp = __get_cpu_var(bau_control).msg_statuses + msg_slot; + cpu = uv_blade_processor_id(); + msg->number_of_cpus = + uv_blade_nr_online_cpus(uv_node_to_blade_id(numa_node_id())); + this_cpu_mask = 1UL << cpu; + if (msp->seen_by.bits & this_cpu_mask) + return; + atomic_or_long(&msp->seen_by.bits, this_cpu_mask); + + if (msg->replied_to == 1) + return; + + if (msg->address == TLB_FLUSH_ALL) { + local_flush_tlb(); + __get_cpu_var(ptcstats).alltlb++; + } else { + __flush_tlb_one(msg->address); + __get_cpu_var(ptcstats).onetlb++; + } + + __get_cpu_var(ptcstats).requestee++; + + atomic_inc_short(&msg->acknowledge_count); + if (msg->number_of_cpus == msg->acknowledge_count) + uv_reply_to_message(sw_ack_slot, msg, msp); +} + +/* + * Examine the payload queue on one distribution node to see + * which messages have not been seen, and which cpu(s) have not seen them. + * + * Returns the number of cpu's that have not responded. + */ +static int uv_examine_destination(struct bau_control *bau_tablesp, int sender) +{ + struct bau_payload_queue_entry *msg; + struct bau_msg_status *msp; + int count = 0; + int i; + int j; + + for (msg = bau_tablesp->va_queue_first, i = 0; i < DEST_Q_SIZE; + msg++, i++) { + if ((msg->sending_cpu == sender) && (!msg->replied_to)) { + msp = bau_tablesp->msg_statuses + i; + printk(KERN_DEBUG + "blade %d: address:%#lx %d of %d, not cpu(s): ", + i, msg->address, msg->acknowledge_count, + msg->number_of_cpus); + for (j = 0; j < msg->number_of_cpus; j++) { + if (!((1L << j) & msp->seen_by.bits)) { + count++; + printk("%d ", j); + } + } + printk("\n"); + } + } + return count; +} + +/* + * Examine the payload queue on all the distribution nodes to see + * which messages have not been seen, and which cpu(s) have not seen them. + * + * Returns the number of cpu's that have not responded. + */ +static int uv_examine_destinations(struct bau_target_nodemask *distribution) +{ + int sender; + int i; + int count = 0; + + sender = smp_processor_id(); + for (i = 0; i < sizeof(struct bau_target_nodemask) * BITSPERBYTE; i++) { + if (!bau_node_isset(i, distribution)) + continue; + count += uv_examine_destination(uv_bau_table_bases[i], sender); + } + return count; +} + +/* + * wait for completion of a broadcast message + * + * return COMPLETE, RETRY or GIVEUP + */ +static int uv_wait_completion(struct bau_desc *bau_desc, + unsigned long mmr_offset, int right_shift) +{ + int exams = 0; + long destination_timeouts = 0; + long source_timeouts = 0; + unsigned long descriptor_status; + + while ((descriptor_status = (((unsigned long) + uv_read_local_mmr(mmr_offset) >> + right_shift) & UV_ACT_STATUS_MASK)) != + DESC_STATUS_IDLE) { + if (descriptor_status == DESC_STATUS_SOURCE_TIMEOUT) { + source_timeouts++; + if (source_timeouts > SOURCE_TIMEOUT_LIMIT) + source_timeouts = 0; + __get_cpu_var(ptcstats).s_retry++; + return FLUSH_RETRY; + } + /* + * spin here looking for progress at the destinations + */ + if (descriptor_status == DESC_STATUS_DESTINATION_TIMEOUT) { + destination_timeouts++; + if (destination_timeouts > DESTINATION_TIMEOUT_LIMIT) { + /* + * returns number of cpus not responding + */ + if (uv_examine_destinations + (&bau_desc->distribution) == 0) { + __get_cpu_var(ptcstats).d_retry++; + return FLUSH_RETRY; + } + exams++; + if (exams >= uv_bau_retry_limit) { + printk(KERN_DEBUG + "uv_flush_tlb_others"); + printk("giving up on cpu %d\n", + smp_processor_id()); + return FLUSH_GIVEUP; + } + /* + * delays can hang the simulator + udelay(1000); + */ + destination_timeouts = 0; + } + } + } + return FLUSH_COMPLETE; +} + +/** + * uv_flush_send_and_wait + * + * Send a broadcast and wait for a broadcast message to complete. + * + * The cpumaskp mask contains the cpus the broadcast was sent to. + * + * Returns 1 if all remote flushing was done. The mask is zeroed. + * Returns 0 if some remote flushing remains to be done. The mask is left + * unchanged. + */ +int uv_flush_send_and_wait(int cpu, int this_blade, struct bau_desc *bau_desc, + cpumask_t *cpumaskp) +{ + int completion_status = 0; + int right_shift; + int tries = 0; + int blade; + int bit; + unsigned long mmr_offset; + unsigned long index; + cycles_t time1; + cycles_t time2; + + if (cpu < UV_CPUS_PER_ACT_STATUS) { + mmr_offset = UVH_LB_BAU_SB_ACTIVATION_STATUS_0; + right_shift = cpu * UV_ACT_STATUS_SIZE; + } else { + mmr_offset = UVH_LB_BAU_SB_ACTIVATION_STATUS_1; + right_shift = + ((cpu - UV_CPUS_PER_ACT_STATUS) * UV_ACT_STATUS_SIZE); + } + time1 = get_cycles(); + do { + tries++; + index = (1UL << UVH_LB_BAU_SB_ACTIVATION_CONTROL_PUSH_SHFT) | + cpu; + uv_write_local_mmr(UVH_LB_BAU_SB_ACTIVATION_CONTROL, index); + completion_status = uv_wait_completion(bau_desc, mmr_offset, + right_shift); + } while (completion_status == FLUSH_RETRY); + time2 = get_cycles(); + __get_cpu_var(ptcstats).sflush += (time2 - time1); + if (tries > 1) + __get_cpu_var(ptcstats).retriesok++; + + if (completion_status == FLUSH_GIVEUP) { + /* + * Cause the caller to do an IPI-style TLB shootdown on + * the cpu's, all of which are still in the mask. + */ + __get_cpu_var(ptcstats).ptc_i++; + return 0; + } + + /* + * Success, so clear the remote cpu's from the mask so we don't + * use the IPI method of shootdown on them. + */ + for_each_cpu_mask(bit, *cpumaskp) { + blade = uv_cpu_to_blade_id(bit); + if (blade == this_blade) + continue; + cpu_clear(bit, *cpumaskp); + } + if (!cpus_empty(*cpumaskp)) + return 0; + return 1; +} + +/** + * uv_flush_tlb_others - globally purge translation cache of a virtual + * address or all TLB's + * @cpumaskp: mask of all cpu's in which the address is to be removed + * @mm: mm_struct containing virtual address range + * @va: virtual address to be removed (or TLB_FLUSH_ALL for all TLB's on cpu) + * + * This is the entry point for initiating any UV global TLB shootdown. + * + * Purges the translation caches of all specified processors of the given + * virtual address, or purges all TLB's on specified processors. + * + * The caller has derived the cpumaskp from the mm_struct and has subtracted + * the local cpu from the mask. This function is called only if there + * are bits set in the mask. (e.g. flush_tlb_page()) + * + * The cpumaskp is converted into a nodemask of the nodes containing + * the cpus. + * + * Returns 1 if all remote flushing was done. + * Returns 0 if some remote flushing remains to be done. + */ +int uv_flush_tlb_others(cpumask_t *cpumaskp, struct mm_struct *mm, + unsigned long va) +{ + int i; + int bit; + int blade; + int cpu; + int this_blade; + int locals = 0; + struct bau_desc *bau_desc; + + cpu = uv_blade_processor_id(); + this_blade = uv_numa_blade_id(); + bau_desc = __get_cpu_var(bau_control).descriptor_base; + bau_desc += UV_ITEMS_PER_DESCRIPTOR * cpu; + + bau_nodes_clear(&bau_desc->distribution, UV_DISTRIBUTION_SIZE); + + i = 0; + for_each_cpu_mask(bit, *cpumaskp) { + blade = uv_cpu_to_blade_id(bit); + BUG_ON(blade > (UV_DISTRIBUTION_SIZE - 1)); + if (blade == this_blade) { + locals++; + continue; + } + bau_node_set(blade, &bau_desc->distribution); + i++; + } + if (i == 0) { + /* + * no off_node flushing; return status for local node + */ + if (locals) + return 0; + else + return 1; + } + __get_cpu_var(ptcstats).requestor++; + __get_cpu_var(ptcstats).ntargeted += i; + + bau_desc->payload.address = va; + bau_desc->payload.sending_cpu = smp_processor_id(); + + return uv_flush_send_and_wait(cpu, this_blade, bau_desc, cpumaskp); +} + +/* + * The BAU message interrupt comes here. (registered by set_intr_gate) + * See entry_64.S + * + * We received a broadcast assist message. + * + * Interrupts may have been disabled; this interrupt could represent + * the receipt of several messages. + * + * All cores/threads on this node get this interrupt. + * The last one to see it does the s/w ack. + * (the resource will not be freed until noninterruptable cpus see this + * interrupt; hardware will timeout the s/w ack and reply ERROR) + */ +void uv_bau_message_interrupt(struct pt_regs *regs) +{ + struct bau_payload_queue_entry *va_queue_first; + struct bau_payload_queue_entry *va_queue_last; + struct bau_payload_queue_entry *msg; + struct pt_regs *old_regs = set_irq_regs(regs); + cycles_t time1; + cycles_t time2; + int msg_slot; + int sw_ack_slot; + int fw; + int count = 0; + unsigned long local_pnode; + + ack_APIC_irq(); + exit_idle(); + irq_enter(); + + time1 = get_cycles(); + + local_pnode = uv_blade_to_pnode(uv_numa_blade_id()); + + va_queue_first = __get_cpu_var(bau_control).va_queue_first; + va_queue_last = __get_cpu_var(bau_control).va_queue_last; + + msg = __get_cpu_var(bau_control).bau_msg_head; + while (msg->sw_ack_vector) { + count++; + fw = msg->sw_ack_vector; + msg_slot = msg - va_queue_first; + sw_ack_slot = ffs(fw) - 1; + + uv_bau_process_message(msg, msg_slot, sw_ack_slot); + + msg++; + if (msg > va_queue_last) + msg = va_queue_first; + __get_cpu_var(bau_control).bau_msg_head = msg; + } + if (!count) + __get_cpu_var(ptcstats).nomsg++; + else if (count > 1) + __get_cpu_var(ptcstats).multmsg++; + + time2 = get_cycles(); + __get_cpu_var(ptcstats).dflush += (time2 - time1); + + irq_exit(); + set_irq_regs(old_regs); +} + +static void uv_enable_timeouts(void) +{ + int i; + int blade; + int last_blade; + int pnode; + int cur_cpu = 0; + unsigned long apicid; + + last_blade = -1; + for_each_online_node(i) { + blade = uv_node_to_blade_id(i); + if (blade == last_blade) + continue; + last_blade = blade; + apicid = per_cpu(x86_cpu_to_apicid, cur_cpu); + pnode = uv_blade_to_pnode(blade); + cur_cpu += uv_blade_nr_possible_cpus(i); + } +} + +static void *uv_ptc_seq_start(struct seq_file *file, loff_t *offset) +{ + if (*offset < num_possible_cpus()) + return offset; + return NULL; +} + +static void *uv_ptc_seq_next(struct seq_file *file, void *data, loff_t *offset) +{ + (*offset)++; + if (*offset < num_possible_cpus()) + return offset; + return NULL; +} + +static void uv_ptc_seq_stop(struct seq_file *file, void *data) +{ +} + +/* + * Display the statistics thru /proc + * data points to the cpu number + */ +static int uv_ptc_seq_show(struct seq_file *file, void *data) +{ + struct ptc_stats *stat; + int cpu; + + cpu = *(loff_t *)data; + + if (!cpu) { + seq_printf(file, + "# cpu requestor requestee one all sretry dretry ptc_i "); + seq_printf(file, + "sw_ack sflush dflush sok dnomsg dmult starget\n"); + } + if (cpu < num_possible_cpus() && cpu_online(cpu)) { + stat = &per_cpu(ptcstats, cpu); + seq_printf(file, "cpu %d %ld %ld %ld %ld %ld %ld %ld ", + cpu, stat->requestor, + stat->requestee, stat->onetlb, stat->alltlb, + stat->s_retry, stat->d_retry, stat->ptc_i); + seq_printf(file, "%lx %ld %ld %ld %ld %ld %ld\n", + uv_read_global_mmr64(uv_blade_to_pnode + (uv_cpu_to_blade_id(cpu)), + UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE), + stat->sflush, stat->dflush, + stat->retriesok, stat->nomsg, + stat->multmsg, stat->ntargeted); + } + + return 0; +} + +/* + * 0: display meaning of the statistics + * >0: retry limit + */ +static ssize_t uv_ptc_proc_write(struct file *file, const char __user *user, + size_t count, loff_t *data) +{ + long newmode; + char optstr[64]; + + if (count == 0 || count > sizeof(optstr)) + return -EINVAL; + if (copy_from_user(optstr, user, count)) + return -EFAULT; + optstr[count - 1] = '\0'; + if (strict_strtoul(optstr, 10, &newmode) < 0) { + printk(KERN_DEBUG "%s is invalid\n", optstr); + return -EINVAL; + } + + if (newmode == 0) { + printk(KERN_DEBUG "# cpu: cpu number\n"); + printk(KERN_DEBUG + "requestor: times this cpu was the flush requestor\n"); + printk(KERN_DEBUG + "requestee: times this cpu was requested to flush its TLBs\n"); + printk(KERN_DEBUG + "one: times requested to flush a single address\n"); + printk(KERN_DEBUG + "all: times requested to flush all TLB's\n"); + printk(KERN_DEBUG + "sretry: number of retries of source-side timeouts\n"); + printk(KERN_DEBUG + "dretry: number of retries of destination-side timeouts\n"); + printk(KERN_DEBUG + "ptc_i: times UV fell through to IPI-style flushes\n"); + printk(KERN_DEBUG + "sw_ack: image of UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE\n"); + printk(KERN_DEBUG + "sflush_us: cycles spent in uv_flush_tlb_others()\n"); + printk(KERN_DEBUG + "dflush_us: cycles spent in handling flush requests\n"); + printk(KERN_DEBUG "sok: successes on retry\n"); + printk(KERN_DEBUG "dnomsg: interrupts with no message\n"); + printk(KERN_DEBUG + "dmult: interrupts with multiple messages\n"); + printk(KERN_DEBUG "starget: nodes targeted\n"); + } else { + uv_bau_retry_limit = newmode; + printk(KERN_DEBUG "timeout retry limit:%d\n", + uv_bau_retry_limit); + } + + return count; +} + +static const struct seq_operations uv_ptc_seq_ops = { + .start = uv_ptc_seq_start, + .next = uv_ptc_seq_next, + .stop = uv_ptc_seq_stop, + .show = uv_ptc_seq_show +}; + +static int uv_ptc_proc_open(struct inode *inode, struct file *file) +{ + return seq_open(file, &uv_ptc_seq_ops); +} + +static const struct file_operations proc_uv_ptc_operations = { + .open = uv_ptc_proc_open, + .read = seq_read, + .write = uv_ptc_proc_write, + .llseek = seq_lseek, + .release = seq_release, +}; + +static int __init uv_ptc_init(void) +{ + struct proc_dir_entry *proc_uv_ptc; + + if (!is_uv_system()) + return 0; + + if (!proc_mkdir("sgi_uv", NULL)) + return -EINVAL; + + proc_uv_ptc = create_proc_entry(UV_PTC_BASENAME, 0444, NULL); + if (!proc_uv_ptc) { + printk(KERN_ERR "unable to create %s proc entry\n", + UV_PTC_BASENAME); + remove_proc_entry("sgi_uv", NULL); + return -EINVAL; + } + proc_uv_ptc->proc_fops = &proc_uv_ptc_operations; + return 0; +} + +/* + * begin the initialization of the per-blade control structures + */ +static struct bau_control * __init uv_table_bases_init(int blade, int node) +{ + int i; + int *ip; + struct bau_msg_status *msp; + struct bau_control *bau_tabp; + + bau_tabp = + kmalloc_node(sizeof(struct bau_control), GFP_KERNEL, node); + BUG_ON(!bau_tabp); + + bau_tabp->msg_statuses = + kmalloc_node(sizeof(struct bau_msg_status) * + DEST_Q_SIZE, GFP_KERNEL, node); + BUG_ON(!bau_tabp->msg_statuses); + + for (i = 0, msp = bau_tabp->msg_statuses; i < DEST_Q_SIZE; i++, msp++) + bau_cpubits_clear(&msp->seen_by, (int) + uv_blade_nr_possible_cpus(blade)); + + bau_tabp->watching = + kmalloc_node(sizeof(int) * DEST_NUM_RESOURCES, GFP_KERNEL, node); + BUG_ON(!bau_tabp->watching); + + for (i = 0, ip = bau_tabp->watching; i < DEST_Q_SIZE; i++, ip++) + *ip = 0; + + uv_bau_table_bases[blade] = bau_tabp; + + return bau_tabp; +} + +/* + * finish the initialization of the per-blade control structures + */ +static void __init +uv_table_bases_finish(int blade, int node, int cur_cpu, + struct bau_control *bau_tablesp, + struct bau_desc *adp) +{ + struct bau_control *bcp; + int i; + + for (i = cur_cpu; i < cur_cpu + uv_blade_nr_possible_cpus(blade); i++) { + bcp = (struct bau_control *)&per_cpu(bau_control, i); + + bcp->bau_msg_head = bau_tablesp->va_queue_first; + bcp->va_queue_first = bau_tablesp->va_queue_first; + bcp->va_queue_last = bau_tablesp->va_queue_last; + bcp->watching = bau_tablesp->watching; + bcp->msg_statuses = bau_tablesp->msg_statuses; + bcp->descriptor_base = adp; + } +} + +/* + * initialize the sending side's sending buffers + */ +static struct bau_desc * __init +uv_activation_descriptor_init(int node, int pnode) +{ + int i; + unsigned long pa; + unsigned long m; + unsigned long n; + unsigned long mmr_image; + struct bau_desc *adp; + struct bau_desc *ad2; + + adp = (struct bau_desc *) + kmalloc_node(16384, GFP_KERNEL, node); + BUG_ON(!adp); + + pa = __pa((unsigned long)adp); + n = pa >> uv_nshift; + m = pa & uv_mmask; + + mmr_image = uv_read_global_mmr64(pnode, UVH_LB_BAU_SB_DESCRIPTOR_BASE); + if (mmr_image) { + uv_write_global_mmr64(pnode, (unsigned long) + UVH_LB_BAU_SB_DESCRIPTOR_BASE, + (n << UV_DESC_BASE_PNODE_SHIFT | m)); + } + + for (i = 0, ad2 = adp; i < UV_ACTIVATION_DESCRIPTOR_SIZE; i++, ad2++) { + memset(ad2, 0, sizeof(struct bau_desc)); + ad2->header.sw_ack_flag = 1; + ad2->header.base_dest_nodeid = + uv_blade_to_pnode(uv_cpu_to_blade_id(0)); + ad2->header.command = UV_NET_ENDPOINT_INTD; + ad2->header.int_both = 1; + /* + * all others need to be set to zero: + * fairness chaining multilevel count replied_to + */ + } + return adp; +} + +/* + * initialize the destination side's receiving buffers + */ +static struct bau_payload_queue_entry * __init +uv_payload_queue_init(int node, int pnode, struct bau_control *bau_tablesp) +{ + struct bau_payload_queue_entry *pqp; + char *cp; + + pqp = (struct bau_payload_queue_entry *) kmalloc_node( + (DEST_Q_SIZE + 1) * sizeof(struct bau_payload_queue_entry), + GFP_KERNEL, node); + BUG_ON(!pqp); + + cp = (char *)pqp + 31; + pqp = (struct bau_payload_queue_entry *)(((unsigned long)cp >> 5) << 5); + bau_tablesp->va_queue_first = pqp; + uv_write_global_mmr64(pnode, + UVH_LB_BAU_INTD_PAYLOAD_QUEUE_FIRST, + ((unsigned long)pnode << + UV_PAYLOADQ_PNODE_SHIFT) | + uv_physnodeaddr(pqp)); + uv_write_global_mmr64(pnode, UVH_LB_BAU_INTD_PAYLOAD_QUEUE_TAIL, + uv_physnodeaddr(pqp)); + bau_tablesp->va_queue_last = pqp + (DEST_Q_SIZE - 1); + uv_write_global_mmr64(pnode, UVH_LB_BAU_INTD_PAYLOAD_QUEUE_LAST, + (unsigned long) + uv_physnodeaddr(bau_tablesp->va_queue_last)); + memset(pqp, 0, sizeof(struct bau_payload_queue_entry) * DEST_Q_SIZE); + + return pqp; +} + +/* + * Initialization of each UV blade's structures + */ +static int __init uv_init_blade(int blade, int node, int cur_cpu) +{ + int pnode; + unsigned long pa; + unsigned long apicid; + struct bau_desc *adp; + struct bau_payload_queue_entry *pqp; + struct bau_control *bau_tablesp; + + bau_tablesp = uv_table_bases_init(blade, node); + pnode = uv_blade_to_pnode(blade); + adp = uv_activation_descriptor_init(node, pnode); + pqp = uv_payload_queue_init(node, pnode, bau_tablesp); + uv_table_bases_finish(blade, node, cur_cpu, bau_tablesp, adp); + /* + * the below initialization can't be in firmware because the + * messaging IRQ will be determined by the OS + */ + apicid = per_cpu(x86_cpu_to_apicid, cur_cpu); + pa = uv_read_global_mmr64(pnode, UVH_BAU_DATA_CONFIG); + if ((pa & 0xff) != UV_BAU_MESSAGE) { + uv_write_global_mmr64(pnode, UVH_BAU_DATA_CONFIG, + ((apicid << 32) | UV_BAU_MESSAGE)); + } + return 0; +} + +/* + * Initialization of BAU-related structures + */ +static int __init uv_bau_init(void) +{ + int blade; + int node; + int nblades; + int last_blade; + int cur_cpu = 0; + + if (!is_uv_system()) + return 0; + + uv_bau_retry_limit = 1; + uv_nshift = uv_hub_info->n_val; + uv_mmask = (1UL << uv_hub_info->n_val) - 1; + nblades = 0; + last_blade = -1; + for_each_online_node(node) { + blade = uv_node_to_blade_id(node); + if (blade == last_blade) + continue; + last_blade = blade; + nblades++; + } + uv_bau_table_bases = (struct bau_control **) + kmalloc(nblades * sizeof(struct bau_control *), GFP_KERNEL); + BUG_ON(!uv_bau_table_bases); + + last_blade = -1; + for_each_online_node(node) { + blade = uv_node_to_blade_id(node); + if (blade == last_blade) + continue; + last_blade = blade; + uv_init_blade(blade, node, cur_cpu); + cur_cpu += uv_blade_nr_possible_cpus(blade); + } + set_intr_gate(UV_BAU_MESSAGE, uv_bau_message_intr1); + uv_enable_timeouts(); + + return 0; +} +__initcall(uv_bau_init); +__initcall(uv_ptc_init); diff --git a/arch/x86/kernel/trampoline.c b/arch/x86/kernel/trampoline.c index abbf199adeb..1106fac6024 100644 --- a/arch/x86/kernel/trampoline.c +++ b/arch/x86/kernel/trampoline.c @@ -2,7 +2,7 @@ #include <asm/trampoline.h> -/* ready for x86_64, no harm for x86, since it will overwrite after alloc */ +/* ready for x86_64 and x86 */ unsigned char *trampoline_base = __va(TRAMPOLINE_BASE); /* diff --git a/arch/x86/kernel/traps_32.c b/arch/x86/kernel/traps_32.c index 08d752de4ee..8a768973c4f 100644 --- a/arch/x86/kernel/traps_32.c +++ b/arch/x86/kernel/traps_32.c @@ -1,5 +1,6 @@ /* * Copyright (C) 1991, 1992 Linus Torvalds + * Copyright (C) 2000, 2001, 2002 Andi Kleen, SuSE Labs * * Pentium III FXSR, SSE support * Gareth Hughes <gareth@valinux.com>, May 2000 @@ -60,8 +61,6 @@ #include "mach_traps.h" -int panic_on_unrecovered_nmi; - DECLARE_BITMAP(used_vectors, NR_VECTORS); EXPORT_SYMBOL_GPL(used_vectors); @@ -98,19 +97,22 @@ asmlinkage void alignment_check(void); asmlinkage void spurious_interrupt_bug(void); asmlinkage void machine_check(void); +int panic_on_unrecovered_nmi; int kstack_depth_to_print = 24; static unsigned int code_bytes = 64; +static int ignore_nmis; +static int die_counter; void printk_address(unsigned long address, int reliable) { #ifdef CONFIG_KALLSYMS - char namebuf[KSYM_NAME_LEN]; unsigned long offset = 0; unsigned long symsize; const char *symname; - char reliab[4] = ""; - char *delim = ":"; char *modname; + char *delim = ":"; + char namebuf[KSYM_NAME_LEN]; + char reliab[4] = ""; symname = kallsyms_lookup(address, &symsize, &offset, &modname, namebuf); @@ -130,22 +132,23 @@ void printk_address(unsigned long address, int reliable) #endif } -static inline int valid_stack_ptr(struct thread_info *tinfo, void *p, unsigned size) +static inline int valid_stack_ptr(struct thread_info *tinfo, + void *p, unsigned int size) { - return p > (void *)tinfo && - p <= (void *)tinfo + THREAD_SIZE - size; + void *t = tinfo; + return p > t && p <= t + THREAD_SIZE - size; } /* The form of the top of the frame on the stack */ struct stack_frame { - struct stack_frame *next_frame; - unsigned long return_address; + struct stack_frame *next_frame; + unsigned long return_address; }; static inline unsigned long print_context_stack(struct thread_info *tinfo, - unsigned long *stack, unsigned long bp, - const struct stacktrace_ops *ops, void *data) + unsigned long *stack, unsigned long bp, + const struct stacktrace_ops *ops, void *data) { struct stack_frame *frame = (struct stack_frame *)bp; @@ -167,8 +170,6 @@ print_context_stack(struct thread_info *tinfo, return bp; } -#define MSG(msg) ops->warning(data, msg) - void dump_trace(struct task_struct *task, struct pt_regs *regs, unsigned long *stack, unsigned long bp, const struct stacktrace_ops *ops, void *data) @@ -178,7 +179,6 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, if (!stack) { unsigned long dummy; - stack = &dummy; if (task != current) stack = (unsigned long *)task->thread.sp; @@ -196,7 +196,7 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs, } #endif - while (1) { + for (;;) { struct thread_info *context; context = (struct thread_info *) @@ -248,10 +248,10 @@ static void print_trace_address(void *data, unsigned long addr, int reliable) } static const struct stacktrace_ops print_trace_ops = { - .warning = print_trace_warning, - .warning_symbol = print_trace_warning_symbol, - .stack = print_trace_stack, - .address = print_trace_address, + .warning = print_trace_warning, + .warning_symbol = print_trace_warning_symbol, + .stack = print_trace_stack, + .address = print_trace_address, }; static void @@ -351,15 +351,14 @@ void show_registers(struct pt_regs *regs) printk(KERN_EMERG "Code: "); ip = (u8 *)regs->ip - code_prologue; - if (ip < (u8 *)PAGE_OFFSET || - probe_kernel_address(ip, c)) { + if (ip < (u8 *)PAGE_OFFSET || probe_kernel_address(ip, c)) { /* try starting at EIP */ ip = (u8 *)regs->ip; code_len = code_len - code_prologue + 1; } for (i = 0; i < code_len; i++, ip++) { if (ip < (u8 *)PAGE_OFFSET || - probe_kernel_address(ip, c)) { + probe_kernel_address(ip, c)) { printk(" Bad EIP value."); break; } @@ -384,8 +383,6 @@ int is_valid_bugaddr(unsigned long ip) return ud2 == 0x0b0f; } -static int die_counter; - int __kprobes __die(const char *str, struct pt_regs *regs, long err) { unsigned short ss; @@ -402,26 +399,22 @@ int __kprobes __die(const char *str, struct pt_regs *regs, long err) printk("DEBUG_PAGEALLOC"); #endif printk("\n"); - if (notify_die(DIE_OOPS, str, regs, err, - current->thread.trap_no, SIGSEGV) != NOTIFY_STOP) { - - show_registers(regs); - /* Executive summary in case the oops scrolled away */ - sp = (unsigned long) (®s->sp); - savesegment(ss, ss); - if (user_mode(regs)) { - sp = regs->sp; - ss = regs->ss & 0xffff; - } - printk(KERN_EMERG "EIP: [<%08lx>] ", regs->ip); - print_symbol("%s", regs->ip); - printk(" SS:ESP %04x:%08lx\n", ss, sp); + current->thread.trap_no, SIGSEGV) == NOTIFY_STOP) + return 1; - return 0; + show_registers(regs); + /* Executive summary in case the oops scrolled away */ + sp = (unsigned long) (®s->sp); + savesegment(ss, ss); + if (user_mode(regs)) { + sp = regs->sp; + ss = regs->ss & 0xffff; } - - return 1; + printk(KERN_EMERG "EIP: [<%08lx>] ", regs->ip); + print_symbol("%s", regs->ip); + printk(" SS:ESP %04x:%08lx\n", ss, sp); + return 0; } /* @@ -546,7 +539,7 @@ void do_##name(struct pt_regs *regs, long error_code) \ { \ trace_hardirqs_fixup(); \ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \ - == NOTIFY_STOP) \ + == NOTIFY_STOP) \ return; \ do_trap(trapnr, signr, str, 0, regs, error_code, NULL); \ } @@ -562,7 +555,7 @@ void do_##name(struct pt_regs *regs, long error_code) \ info.si_code = sicode; \ info.si_addr = (void __user *)siaddr; \ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \ - == NOTIFY_STOP) \ + == NOTIFY_STOP) \ return; \ do_trap(trapnr, signr, str, 0, regs, error_code, &info); \ } @@ -571,7 +564,7 @@ void do_##name(struct pt_regs *regs, long error_code) \ void do_##name(struct pt_regs *regs, long error_code) \ { \ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \ - == NOTIFY_STOP) \ + == NOTIFY_STOP) \ return; \ do_trap(trapnr, signr, str, 1, regs, error_code, NULL); \ } @@ -586,27 +579,29 @@ void do_##name(struct pt_regs *regs, long error_code) \ info.si_addr = (void __user *)siaddr; \ trace_hardirqs_fixup(); \ if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \ - == NOTIFY_STOP) \ + == NOTIFY_STOP) \ return; \ do_trap(trapnr, signr, str, 1, regs, error_code, &info); \ } -DO_VM86_ERROR_INFO(0, SIGFPE, "divide error", divide_error, FPE_INTDIV, regs->ip) +DO_VM86_ERROR_INFO(0, SIGFPE, "divide error", divide_error, FPE_INTDIV, regs->ip) #ifndef CONFIG_KPROBES DO_VM86_ERROR(3, SIGTRAP, "int3", int3) #endif DO_VM86_ERROR(4, SIGSEGV, "overflow", overflow) DO_VM86_ERROR(5, SIGSEGV, "bounds", bounds) -DO_ERROR_INFO(6, SIGILL, "invalid opcode", invalid_op, ILL_ILLOPN, regs->ip, 0) -DO_ERROR(9, SIGFPE, "coprocessor segment overrun", coprocessor_segment_overrun) +DO_ERROR_INFO(6, SIGILL, "invalid opcode", invalid_op, ILL_ILLOPN, regs->ip, 0) +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(12, SIGBUS, "stack segment", stack_segment) +DO_ERROR(11, SIGBUS, "segment not present", segment_not_present) +DO_ERROR(12, SIGBUS, "stack segment", stack_segment) DO_ERROR_INFO(17, SIGBUS, "alignment check", alignment_check, BUS_ADRALN, 0, 0) DO_ERROR_INFO(32, SIGILL, "iret exception", iret_error, ILL_BADSTK, 0, 1) -void __kprobes do_general_protection(struct pt_regs *regs, long error_code) +void __kprobes +do_general_protection(struct pt_regs *regs, long error_code) { + struct task_struct *tsk; struct thread_struct *thread; struct tss_struct *tss; int cpu; @@ -647,23 +642,24 @@ void __kprobes do_general_protection(struct pt_regs *regs, long error_code) if (regs->flags & X86_VM_MASK) goto gp_in_vm86; + tsk = current; if (!user_mode(regs)) goto gp_in_kernel; - current->thread.error_code = error_code; - current->thread.trap_no = 13; + tsk->thread.error_code = error_code; + tsk->thread.trap_no = 13; - if (show_unhandled_signals && unhandled_signal(current, SIGSEGV) && - printk_ratelimit()) { + if (show_unhandled_signals && unhandled_signal(tsk, SIGSEGV) && + printk_ratelimit()) { printk(KERN_INFO - "%s[%d] general protection ip:%lx sp:%lx error:%lx", - current->comm, task_pid_nr(current), - regs->ip, regs->sp, error_code); + "%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"); } - force_sig(SIGSEGV, current); + force_sig(SIGSEGV, tsk); return; gp_in_vm86: @@ -672,14 +668,15 @@ gp_in_vm86: return; gp_in_kernel: - if (!fixup_exception(regs)) { - current->thread.error_code = error_code; - current->thread.trap_no = 13; - if (notify_die(DIE_GPF, "general protection fault", regs, + if (fixup_exception(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) - return; - die("general protection fault", regs, error_code); - } + return; + die("general protection fault", regs, error_code); } static notrace __kprobes void @@ -756,9 +753,9 @@ unknown_nmi_error(unsigned char reason, struct pt_regs *regs) static DEFINE_SPINLOCK(nmi_print_lock); -void notrace __kprobes die_nmi(struct pt_regs *regs, const char *msg) +void notrace __kprobes die_nmi(char *str, struct pt_regs *regs, int do_panic) { - if (notify_die(DIE_NMIWATCHDOG, msg, regs, 0, 2, SIGINT) == NOTIFY_STOP) + if (notify_die(DIE_NMIWATCHDOG, str, regs, 0, 2, SIGINT) == NOTIFY_STOP) return; spin_lock(&nmi_print_lock); @@ -767,10 +764,12 @@ void notrace __kprobes die_nmi(struct pt_regs *regs, const char *msg) * to get a message out: */ bust_spinlocks(1); - printk(KERN_EMERG "%s", msg); + printk(KERN_EMERG "%s", str); printk(" on CPU%d, ip %08lx, registers:\n", smp_processor_id(), regs->ip); show_registers(regs); + if (do_panic) + panic("Non maskable interrupt"); console_silent(); spin_unlock(&nmi_print_lock); bust_spinlocks(0); @@ -790,14 +789,17 @@ void notrace __kprobes die_nmi(struct pt_regs *regs, const char *msg) static notrace __kprobes void default_do_nmi(struct pt_regs *regs) { unsigned char reason = 0; + int cpu; + + cpu = smp_processor_id(); - /* Only the BSP gets external NMIs from the system: */ - if (!smp_processor_id()) + /* Only the BSP gets external NMIs from the system. */ + if (!cpu) reason = get_nmi_reason(); if (!(reason & 0xc0)) { if (notify_die(DIE_NMI_IPI, "nmi_ipi", regs, reason, 2, SIGINT) - == NOTIFY_STOP) + == NOTIFY_STOP) return; #ifdef CONFIG_X86_LOCAL_APIC /* @@ -806,7 +808,7 @@ static notrace __kprobes void default_do_nmi(struct pt_regs *regs) */ if (nmi_watchdog_tick(regs, reason)) return; - if (!do_nmi_callback(regs, smp_processor_id())) + if (!do_nmi_callback(regs, cpu)) unknown_nmi_error(reason, regs); #else unknown_nmi_error(reason, regs); @@ -816,6 +818,8 @@ static notrace __kprobes void default_do_nmi(struct pt_regs *regs) } if (notify_die(DIE_NMI, "nmi", regs, reason, 2, SIGINT) == NOTIFY_STOP) return; + + /* AK: following checks seem to be broken on modern chipsets. FIXME */ if (reason & 0x80) mem_parity_error(reason, regs); if (reason & 0x40) @@ -827,8 +831,6 @@ static notrace __kprobes void default_do_nmi(struct pt_regs *regs) reassert_nmi(); } -static int ignore_nmis; - notrace __kprobes void do_nmi(struct pt_regs *regs, long error_code) { int cpu; @@ -913,7 +915,7 @@ void __kprobes do_debug(struct pt_regs *regs, long error_code) tsk->thread.debugctlmsr = 0; if (notify_die(DIE_DEBUG, "debug", regs, condition, error_code, - SIGTRAP) == NOTIFY_STOP) + SIGTRAP) == NOTIFY_STOP) return; /* It's safe to allow irq's after DR6 has been saved */ if (regs->flags & X86_EFLAGS_IF) @@ -974,9 +976,8 @@ clear_TF_reenable: void math_error(void __user *ip) { struct task_struct *task; - unsigned short cwd; - unsigned short swd; siginfo_t info; + unsigned short cwd, swd; /* * Save the info for the exception handler and clear the error. @@ -995,7 +996,7 @@ void math_error(void __user *ip) * C1 reg you need in case of a stack fault, 0x040 is the stack * fault bit. We should only be taking one exception at a time, * so if this combination doesn't produce any single exception, - * then we have a bad program that isn't syncronizing its FPU usage + * then we have a bad program that isn't synchronizing its FPU usage * and it will suffer the consequences since we won't be able to * fully reproduce the context of the exception */ @@ -1004,7 +1005,7 @@ void math_error(void __user *ip) switch (swd & ~cwd & 0x3f) { case 0x000: /* No unmasked exception */ return; - default: /* Multiple exceptions */ + default: /* Multiple exceptions */ break; case 0x001: /* Invalid Op */ /* @@ -1040,8 +1041,8 @@ void do_coprocessor_error(struct pt_regs *regs, long error_code) static void simd_math_error(void __user *ip) { struct task_struct *task; - unsigned short mxcsr; siginfo_t info; + unsigned short mxcsr; /* * Save the info for the exception handler and clear the error. @@ -1117,7 +1118,7 @@ void do_spurious_interrupt_bug(struct pt_regs *regs, long error_code) unsigned long patch_espfix_desc(unsigned long uesp, unsigned long kesp) { - struct desc_struct *gdt = __get_cpu_var(gdt_page).gdt; + struct desc_struct *gdt = get_cpu_gdt_table(smp_processor_id()); unsigned long base = (kesp - uesp) & -THREAD_SIZE; unsigned long new_kesp = kesp - base; unsigned long lim_pages = (new_kesp | (THREAD_SIZE - 1)) >> PAGE_SHIFT; @@ -1196,19 +1197,16 @@ void __init trap_init(void) early_iounmap(p, 4); #endif -#ifdef CONFIG_X86_LOCAL_APIC - init_apic_mappings(); -#endif - set_trap_gate(0, ÷_error); - set_intr_gate(1, &debug); - set_intr_gate(2, &nmi); - set_system_intr_gate(3, &int3); /* int3/4 can be called from all */ - set_system_gate(4, &overflow); - set_trap_gate(5, &bounds); - set_trap_gate(6, &invalid_op); - set_trap_gate(7, &device_not_available); - set_task_gate(8, GDT_ENTRY_DOUBLEFAULT_TSS); - set_trap_gate(9, &coprocessor_segment_overrun); + set_trap_gate(0, ÷_error); + set_intr_gate(1, &debug); + set_intr_gate(2, &nmi); + set_system_intr_gate(3, &int3); /* int3 can be called from all */ + set_system_gate(4, &overflow); /* int4 can be called from all */ + set_trap_gate(5, &bounds); + set_trap_gate(6, &invalid_op); + set_trap_gate(7, &device_not_available); + set_task_gate(8, GDT_ENTRY_DOUBLEFAULT_TSS); + set_trap_gate(9, &coprocessor_segment_overrun); set_trap_gate(10, &invalid_TSS); set_trap_gate(11, &segment_not_present); set_trap_gate(12, &stack_segment); diff --git a/arch/x86/kernel/traps_64.c b/arch/x86/kernel/traps_64.c index f1a95d10595..2696a683778 100644 --- a/arch/x86/kernel/traps_64.c +++ b/arch/x86/kernel/traps_64.c @@ -10,49 +10,49 @@ * 'Traps.c' handles hardware traps and faults after we have saved some * state in 'entry.S'. */ -#include <linux/sched.h> +#include <linux/moduleparam.h> +#include <linux/interrupt.h> +#include <linux/kallsyms.h> +#include <linux/spinlock.h> +#include <linux/kprobes.h> +#include <linux/uaccess.h> +#include <linux/utsname.h> +#include <linux/kdebug.h> #include <linux/kernel.h> +#include <linux/module.h> +#include <linux/ptrace.h> #include <linux/string.h> +#include <linux/unwind.h> +#include <linux/delay.h> #include <linux/errno.h> -#include <linux/ptrace.h> +#include <linux/kexec.h> +#include <linux/sched.h> #include <linux/timer.h> -#include <linux/mm.h> #include <linux/init.h> -#include <linux/delay.h> -#include <linux/spinlock.h> -#include <linux/interrupt.h> -#include <linux/kallsyms.h> -#include <linux/module.h> -#include <linux/moduleparam.h> -#include <linux/nmi.h> -#include <linux/kprobes.h> -#include <linux/kexec.h> -#include <linux/unwind.h> -#include <linux/uaccess.h> #include <linux/bug.h> -#include <linux/kdebug.h> -#include <linux/utsname.h> - -#include <mach_traps.h> +#include <linux/nmi.h> +#include <linux/mm.h> #if defined(CONFIG_EDAC) #include <linux/edac.h> #endif -#include <asm/system.h> -#include <asm/io.h> -#include <asm/atomic.h> +#include <asm/stacktrace.h> +#include <asm/processor.h> #include <asm/debugreg.h> +#include <asm/atomic.h> +#include <asm/system.h> +#include <asm/unwind.h> #include <asm/desc.h> #include <asm/i387.h> -#include <asm/processor.h> -#include <asm/unwind.h> +#include <asm/nmi.h> #include <asm/smp.h> +#include <asm/io.h> #include <asm/pgalloc.h> -#include <asm/pda.h> #include <asm/proto.h> -#include <asm/nmi.h> -#include <asm/stacktrace.h> +#include <asm/pda.h> + +#include <mach_traps.h> asmlinkage void divide_error(void); asmlinkage void debug(void); @@ -71,12 +71,15 @@ asmlinkage void general_protection(void); asmlinkage void page_fault(void); asmlinkage void coprocessor_error(void); asmlinkage void simd_coprocessor_error(void); -asmlinkage void reserved(void); asmlinkage void alignment_check(void); -asmlinkage void machine_check(void); asmlinkage void spurious_interrupt_bug(void); +asmlinkage void machine_check(void); +int panic_on_unrecovered_nmi; +int kstack_depth_to_print = 12; static unsigned int code_bytes = 64; +static int ignore_nmis; +static int die_counter; static inline void conditional_sti(struct pt_regs *regs) { @@ -100,8 +103,6 @@ static inline void preempt_conditional_cli(struct pt_regs *regs) dec_preempt_count(); } -int kstack_depth_to_print = 12; - void printk_address(unsigned long address, int reliable) { printk(" [<%016lx>] %s%pS\n", address, reliable ? "": "? ", (void *) address); @@ -181,8 +182,6 @@ static unsigned long *in_exception_stack(unsigned cpu, unsigned long stack, return NULL; } -#define MSG(txt) ops->warning(data, txt) - /* * x86-64 can have up to three kernel stacks: * process stack @@ -209,11 +208,11 @@ struct stack_frame { unsigned long return_address; }; - -static inline unsigned long print_context_stack(struct thread_info *tinfo, - unsigned long *stack, unsigned long bp, - const struct stacktrace_ops *ops, void *data, - unsigned long *end) +static inline unsigned long +print_context_stack(struct thread_info *tinfo, + unsigned long *stack, unsigned long bp, + const struct stacktrace_ops *ops, void *data, + unsigned long *end) { struct stack_frame *frame = (struct stack_frame *)bp; @@ -235,7 +234,7 @@ static inline unsigned long print_context_stack(struct thread_info *tinfo, return bp; } -void dump_trace(struct task_struct *tsk, struct pt_regs *regs, +void dump_trace(struct task_struct *task, struct pt_regs *regs, unsigned long *stack, unsigned long bp, const struct stacktrace_ops *ops, void *data) { @@ -244,36 +243,34 @@ void dump_trace(struct task_struct *tsk, struct pt_regs *regs, unsigned used = 0; struct thread_info *tinfo; - if (!tsk) - tsk = current; - tinfo = task_thread_info(tsk); + if (!task) + task = current; if (!stack) { unsigned long dummy; stack = &dummy; - if (tsk && tsk != current) - stack = (unsigned long *)tsk->thread.sp; + if (task && task != current) + stack = (unsigned long *)task->thread.sp; } #ifdef CONFIG_FRAME_POINTER if (!bp) { - if (tsk == current) { + if (task == current) { /* Grab bp right from our regs */ - asm("movq %%rbp, %0" : "=r" (bp):); + asm("movq %%rbp, %0" : "=r" (bp) :); } else { /* bp is the last reg pushed by switch_to */ - bp = *(unsigned long *) tsk->thread.sp; + bp = *(unsigned long *) task->thread.sp; } } #endif - - /* * Print function call entries in all stacks, starting at the * current stack address. If the stacks consist of nested * exceptions */ + tinfo = task_thread_info(task); for (;;) { char *id; unsigned long *estack_end; @@ -358,18 +355,17 @@ static const struct stacktrace_ops print_trace_ops = { .address = print_trace_address, }; -void -show_trace(struct task_struct *tsk, struct pt_regs *regs, unsigned long *stack, - unsigned long bp) +void show_trace(struct task_struct *task, struct pt_regs *regs, + unsigned long *stack, unsigned long bp) { printk("\nCall Trace:\n"); - dump_trace(tsk, regs, stack, bp, &print_trace_ops, NULL); + dump_trace(task, regs, stack, bp, &print_trace_ops, NULL); printk("\n"); } static void -_show_stack(struct task_struct *tsk, struct pt_regs *regs, unsigned long *sp, - unsigned long bp) +_show_stack(struct task_struct *task, struct pt_regs *regs, + unsigned long *sp, unsigned long bp) { unsigned long *stack; int i; @@ -381,14 +377,14 @@ _show_stack(struct task_struct *tsk, struct pt_regs *regs, unsigned long *sp, // back trace for this cpu. if (sp == NULL) { - if (tsk) - sp = (unsigned long *)tsk->thread.sp; + if (task) + sp = (unsigned long *)task->thread.sp; else sp = (unsigned long *)&sp; } stack = sp; - for(i=0; i < kstack_depth_to_print; i++) { + for (i = 0; i < kstack_depth_to_print; i++) { if (stack >= irqstack && stack <= irqstack_end) { if (stack == irqstack_end) { stack = (unsigned long *) (irqstack_end[-1]); @@ -403,12 +399,12 @@ _show_stack(struct task_struct *tsk, struct pt_regs *regs, unsigned long *sp, printk(" %016lx", *stack++); touch_nmi_watchdog(); } - show_trace(tsk, regs, sp, bp); + show_trace(task, regs, sp, bp); } -void show_stack(struct task_struct *tsk, unsigned long * sp) +void show_stack(struct task_struct *task, unsigned long *sp) { - _show_stack(tsk, NULL, sp, 0); + _show_stack(task, NULL, sp, 0); } /* @@ -416,8 +412,8 @@ void show_stack(struct task_struct *tsk, unsigned long * sp) */ void dump_stack(void) { - unsigned long dummy; unsigned long bp = 0; + unsigned long stack; #ifdef CONFIG_FRAME_POINTER if (!bp) @@ -429,7 +425,7 @@ void dump_stack(void) init_utsname()->release, (int)strcspn(init_utsname()->version, " "), init_utsname()->version); - show_trace(NULL, NULL, &dummy, bp); + show_trace(NULL, NULL, &stack, bp); } EXPORT_SYMBOL(dump_stack); @@ -440,12 +436,8 @@ void show_registers(struct pt_regs *regs) unsigned long sp; const int cpu = smp_processor_id(); struct task_struct *cur = cpu_pda(cpu)->pcurrent; - u8 *ip; - unsigned int code_prologue = code_bytes * 43 / 64; - unsigned int code_len = code_bytes; sp = regs->sp; - ip = (u8 *) regs->ip - code_prologue; printk("CPU %d ", cpu); __show_regs(regs); printk("Process %s (pid: %d, threadinfo %p, task %p)\n", @@ -456,15 +448,21 @@ void show_registers(struct pt_regs *regs) * time of the fault.. */ if (!user_mode(regs)) { + unsigned int code_prologue = code_bytes * 43 / 64; + unsigned int code_len = code_bytes; unsigned char c; + u8 *ip; + printk("Stack: "); _show_stack(NULL, regs, (unsigned long *)sp, regs->bp); printk("\n"); printk(KERN_EMERG "Code: "); + + ip = (u8 *)regs->ip - code_prologue; if (ip < (u8 *)PAGE_OFFSET || probe_kernel_address(ip, c)) { /* try starting at RIP */ - ip = (u8 *) regs->ip; + ip = (u8 *)regs->ip; code_len = code_len - code_prologue + 1; } for (i = 0; i < code_len; i++, ip++) { @@ -480,7 +478,7 @@ void show_registers(struct pt_regs *regs) } } printk("\n"); -} +} int is_valid_bugaddr(unsigned long ip) { @@ -538,10 +536,9 @@ void __kprobes oops_end(unsigned long flags, struct pt_regs *regs, int signr) do_exit(signr); } -int __kprobes __die(const char * str, struct pt_regs * regs, long err) +int __kprobes __die(const char *str, struct pt_regs *regs, long err) { - static int die_counter; - printk(KERN_EMERG "%s: %04lx [%u] ", str, err & 0xffff,++die_counter); + printk(KERN_EMERG "%s: %04lx [%u] ", str, err & 0xffff, ++die_counter); #ifdef CONFIG_PREEMPT printk("PREEMPT "); #endif @@ -552,8 +549,10 @@ int __kprobes __die(const char * str, struct pt_regs * regs, long err) printk("DEBUG_PAGEALLOC"); #endif printk("\n"); - if (notify_die(DIE_OOPS, str, regs, err, current->thread.trap_no, SIGSEGV) == NOTIFY_STOP) + if (notify_die(DIE_OOPS, str, regs, err, + current->thread.trap_no, SIGSEGV) == NOTIFY_STOP) return 1; + show_registers(regs); add_taint(TAINT_DIE); /* Executive summary in case the oops scrolled away */ @@ -565,7 +564,7 @@ int __kprobes __die(const char * str, struct pt_regs * regs, long err) return 0; } -void die(const char * str, struct pt_regs * regs, long err) +void die(const char *str, struct pt_regs *regs, long err) { unsigned long flags = oops_begin(); @@ -582,8 +581,7 @@ die_nmi(char *str, struct pt_regs *regs, int do_panic) { unsigned long flags; - if (notify_die(DIE_NMIWATCHDOG, str, regs, 0, 2, SIGINT) == - NOTIFY_STOP) + if (notify_die(DIE_NMIWATCHDOG, str, regs, 0, 2, SIGINT) == NOTIFY_STOP) return; flags = oops_begin(); @@ -591,7 +589,9 @@ die_nmi(char *str, struct pt_regs *regs, int do_panic) * We are in trouble anyway, lets at least try * to get a message out. */ - printk(str, smp_processor_id()); + printk(KERN_EMERG "%s", str); + printk(" on CPU%d, ip %08lx, registers:\n", + smp_processor_id(), regs->ip); show_registers(regs); if (kexec_should_crash(current)) crash_kexec(regs); @@ -603,44 +603,44 @@ die_nmi(char *str, struct pt_regs *regs, int do_panic) do_exit(SIGBUS); } -static void __kprobes do_trap(int trapnr, int signr, char *str, - struct pt_regs * regs, long error_code, - siginfo_t *info) +static void __kprobes +do_trap(int trapnr, int signr, char *str, struct pt_regs *regs, + long error_code, siginfo_t *info) { struct task_struct *tsk = current; - if (user_mode(regs)) { - /* - * We want error_code and trap_no 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 kernel fault information, - * so that won't result in polluting the information - * about previously queued, but not yet delivered, - * faults. See also do_general_protection below. - */ - tsk->thread.error_code = error_code; - tsk->thread.trap_no = trapnr; - - 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); - print_vma_addr(" in ", regs->ip); - printk("\n"); - } + if (!user_mode(regs)) + goto kernel_trap; - if (info) - force_sig_info(signr, info, tsk); - else - force_sig(signr, tsk); - return; + /* + * We want error_code and trap_no 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 + * kernel fault information, so that won't result in polluting + * the information about previously queued, but not yet + * delivered, faults. See also do_general_protection below. + */ + tsk->thread.error_code = error_code; + tsk->thread.trap_no = trapnr; + + 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); + print_vma_addr(" in ", regs->ip); + printk("\n"); } + if (info) + force_sig_info(signr, info, tsk); + else + force_sig(signr, tsk); + return; +kernel_trap: if (!fixup_exception(regs)) { tsk->thread.error_code = error_code; tsk->thread.trap_no = trapnr; @@ -650,41 +650,39 @@ static void __kprobes do_trap(int trapnr, int signr, char *str, } #define DO_ERROR(trapnr, signr, str, name) \ -asmlinkage void do_##name(struct pt_regs * regs, long error_code) \ -{ \ - if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \ - == NOTIFY_STOP) \ - return; \ +asmlinkage 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); \ + do_trap(trapnr, signr, str, regs, error_code, NULL); \ } -#define DO_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr) \ -asmlinkage 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; \ - trace_hardirqs_fixup(); \ - if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \ - == NOTIFY_STOP) \ - return; \ +#define DO_ERROR_INFO(trapnr, signr, str, name, sicode, siaddr) \ +asmlinkage 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; \ + trace_hardirqs_fixup(); \ + 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_trap(trapnr, signr, str, regs, error_code, &info); \ } -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( 7, SIGSEGV, "device not available", device_not_available) -DO_ERROR( 9, SIGFPE, "coprocessor segment overrun", coprocessor_segment_overrun) +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(11, SIGBUS, "segment not present", segment_not_present) DO_ERROR_INFO(17, SIGBUS, "alignment check", alignment_check, BUS_ADRALN, 0) -DO_ERROR(18, SIGSEGV, "reserved", reserved) /* Runs on IST stack */ asmlinkage void do_stack_segment(struct pt_regs *regs, long error_code) @@ -714,31 +712,34 @@ asmlinkage void do_double_fault(struct pt_regs * regs, long error_code) die(str, regs, error_code); } -asmlinkage void __kprobes do_general_protection(struct pt_regs * regs, - long error_code) +asmlinkage void __kprobes +do_general_protection(struct pt_regs *regs, long error_code) { - struct task_struct *tsk = current; + struct task_struct *tsk; conditional_sti(regs); - if (user_mode(regs)) { - tsk->thread.error_code = error_code; - tsk->thread.trap_no = 13; - - if (show_unhandled_signals && unhandled_signal(tsk, SIGSEGV) && - printk_ratelimit()) { - printk(KERN_INFO - "%s[%d] general protection ip:%lx sp:%lx error:%lx", - tsk->comm, tsk->pid, - regs->ip, regs->sp, error_code); - print_vma_addr(" in ", regs->ip); - printk("\n"); - } + tsk = current; + if (!user_mode(regs)) + goto gp_in_kernel; - force_sig(SIGSEGV, tsk); - return; - } + tsk->thread.error_code = error_code; + tsk->thread.trap_no = 13; + + if (show_unhandled_signals && unhandled_signal(tsk, SIGSEGV) && + printk_ratelimit()) { + printk(KERN_INFO + "%s[%d] general protection ip:%lx sp:%lx error:%lx", + tsk->comm, tsk->pid, + regs->ip, regs->sp, error_code); + print_vma_addr(" in ", regs->ip); + printk("\n"); + } + + force_sig(SIGSEGV, tsk); + return; +gp_in_kernel: if (fixup_exception(regs)) return; @@ -751,14 +752,14 @@ asmlinkage void __kprobes do_general_protection(struct pt_regs * regs, } static notrace __kprobes void -mem_parity_error(unsigned char reason, struct pt_regs * regs) +mem_parity_error(unsigned char reason, struct pt_regs *regs) { printk(KERN_EMERG "Uhhuh. NMI received for unknown reason %02x.\n", reason); printk(KERN_EMERG "You have some hardware problem, likely on the PCI bus.\n"); #if defined(CONFIG_EDAC) - if(edac_handler_set()) { + if (edac_handler_set()) { edac_atomic_assert_error(); return; } @@ -775,7 +776,7 @@ mem_parity_error(unsigned char reason, struct pt_regs * regs) } static notrace __kprobes void -io_check_error(unsigned char reason, struct pt_regs * regs) +io_check_error(unsigned char reason, struct pt_regs *regs) { printk("NMI: IOCK error (debug interrupt?)\n"); show_registers(regs); @@ -805,14 +806,14 @@ unknown_nmi_error(unsigned char reason, struct pt_regs * regs) /* Runs on IST stack. This code must keep interrupts off all the time. Nested NMIs are prevented by the CPU. */ -asmlinkage notrace __kprobes void default_do_nmi(struct pt_regs *regs) +asmlinkage notrace __kprobes void default_do_nmi(struct pt_regs *regs) { unsigned char reason = 0; int cpu; cpu = smp_processor_id(); - /* Only the BSP gets external NMIs from the system. */ + /* Only the BSP gets external NMIs from the system. */ if (!cpu) reason = get_nmi_reason(); @@ -824,32 +825,57 @@ asmlinkage notrace __kprobes void default_do_nmi(struct pt_regs *regs) * Ok, so this is none of the documented NMI sources, * so it must be the NMI watchdog. */ - if (nmi_watchdog_tick(regs,reason)) + if (nmi_watchdog_tick(regs, reason)) return; - if (!do_nmi_callback(regs,cpu)) + if (!do_nmi_callback(regs, cpu)) unknown_nmi_error(reason, regs); return; } if (notify_die(DIE_NMI, "nmi", regs, reason, 2, SIGINT) == NOTIFY_STOP) - return; + return; /* AK: following checks seem to be broken on modern chipsets. FIXME */ - if (reason & 0x80) mem_parity_error(reason, regs); if (reason & 0x40) io_check_error(reason, regs); } +asmlinkage notrace __kprobes void +do_nmi(struct pt_regs *regs, long error_code) +{ + nmi_enter(); + + add_pda(__nmi_count, 1); + + if (!ignore_nmis) + default_do_nmi(regs); + + nmi_exit(); +} + +void stop_nmi(void) +{ + acpi_nmi_disable(); + ignore_nmis++; +} + +void restart_nmi(void) +{ + ignore_nmis--; + acpi_nmi_enable(); +} + /* runs on IST stack. */ -asmlinkage void __kprobes do_int3(struct pt_regs * regs, long error_code) +asmlinkage void __kprobes do_int3(struct pt_regs *regs, long error_code) { trace_hardirqs_fixup(); - if (notify_die(DIE_INT3, "int3", regs, error_code, 3, SIGTRAP) == NOTIFY_STOP) { + if (notify_die(DIE_INT3, "int3", regs, error_code, 3, SIGTRAP) + == NOTIFY_STOP) return; - } + preempt_conditional_sti(regs); do_trap(3, SIGTRAP, "int3", regs, error_code, NULL); preempt_conditional_cli(regs); @@ -880,8 +906,8 @@ asmlinkage __kprobes struct pt_regs *sync_regs(struct pt_regs *eregs) asmlinkage void __kprobes do_debug(struct pt_regs * regs, unsigned long error_code) { - unsigned long condition; struct task_struct *tsk = current; + unsigned long condition; siginfo_t info; trace_hardirqs_fixup(); @@ -902,21 +928,19 @@ asmlinkage void __kprobes do_debug(struct pt_regs * regs, /* Mask out spurious debug traps due to lazy DR7 setting */ if (condition & (DR_TRAP0|DR_TRAP1|DR_TRAP2|DR_TRAP3)) { - if (!tsk->thread.debugreg7) { + if (!tsk->thread.debugreg7) goto clear_dr7; - } } tsk->thread.debugreg6 = condition; - /* * Single-stepping through TF: make sure we ignore any events in * kernel space (but re-enable TF when returning to user mode). */ if (condition & DR_STEP) { - if (!user_mode(regs)) - goto clear_TF_reenable; + if (!user_mode(regs)) + goto clear_TF_reenable; } /* Ok, finally something we can handle */ @@ -929,7 +953,7 @@ asmlinkage void __kprobes do_debug(struct pt_regs * regs, force_sig_info(SIGTRAP, &info, tsk); clear_dr7: - set_debugreg(0UL, 7); + set_debugreg(0, 7); preempt_conditional_cli(regs); return; @@ -937,6 +961,7 @@ clear_TF_reenable: set_tsk_thread_flag(tsk, TIF_SINGLESTEP); regs->flags &= ~X86_EFLAGS_TF; preempt_conditional_cli(regs); + return; } static int kernel_math_error(struct pt_regs *regs, const char *str, int trapnr) @@ -959,7 +984,7 @@ static int kernel_math_error(struct pt_regs *regs, const char *str, int trapnr) asmlinkage void do_coprocessor_error(struct pt_regs *regs) { void __user *ip = (void __user *)(regs->ip); - struct task_struct * task; + struct task_struct *task; siginfo_t info; unsigned short cwd, swd; @@ -992,30 +1017,30 @@ asmlinkage void do_coprocessor_error(struct pt_regs *regs) cwd = get_fpu_cwd(task); swd = get_fpu_swd(task); switch (swd & ~cwd & 0x3f) { - case 0x000: - default: - break; - case 0x001: /* Invalid Op */ - /* - * swd & 0x240 == 0x040: Stack Underflow - * swd & 0x240 == 0x240: Stack Overflow - * User must clear the SF bit (0x40) if set - */ - info.si_code = FPE_FLTINV; - break; - case 0x002: /* Denormalize */ - case 0x010: /* Underflow */ - info.si_code = FPE_FLTUND; - break; - case 0x004: /* Zero Divide */ - info.si_code = FPE_FLTDIV; - break; - case 0x008: /* Overflow */ - info.si_code = FPE_FLTOVF; - break; - case 0x020: /* Precision */ - info.si_code = FPE_FLTRES; - break; + case 0x000: /* No unmasked exception */ + default: /* Multiple exceptions */ + break; + case 0x001: /* Invalid Op */ + /* + * swd & 0x240 == 0x040: Stack Underflow + * swd & 0x240 == 0x240: Stack Overflow + * User must clear the SF bit (0x40) if set + */ + info.si_code = FPE_FLTINV; + break; + case 0x002: /* Denormalize */ + case 0x010: /* Underflow */ + info.si_code = FPE_FLTUND; + break; + case 0x004: /* Zero Divide */ + info.si_code = FPE_FLTDIV; + break; + case 0x008: /* Overflow */ + info.si_code = FPE_FLTOVF; + break; + case 0x020: /* Precision */ + info.si_code = FPE_FLTRES; + break; } force_sig_info(SIGFPE, &info, task); } @@ -1028,7 +1053,7 @@ asmlinkage void bad_intr(void) asmlinkage void do_simd_coprocessor_error(struct pt_regs *regs) { void __user *ip = (void __user *)(regs->ip); - struct task_struct * task; + struct task_struct *task; siginfo_t info; unsigned short mxcsr; @@ -1056,25 +1081,25 @@ asmlinkage void do_simd_coprocessor_error(struct pt_regs *regs) */ mxcsr = get_fpu_mxcsr(task); switch (~((mxcsr & 0x1f80) >> 7) & (mxcsr & 0x3f)) { - case 0x000: - default: - break; - case 0x001: /* Invalid Op */ - info.si_code = FPE_FLTINV; - break; - case 0x002: /* Denormalize */ - case 0x010: /* Underflow */ - info.si_code = FPE_FLTUND; - break; - case 0x004: /* Zero Divide */ - info.si_code = FPE_FLTDIV; - break; - case 0x008: /* Overflow */ - info.si_code = FPE_FLTOVF; - break; - case 0x020: /* Precision */ - info.si_code = FPE_FLTRES; - break; + case 0x000: + default: + break; + case 0x001: /* Invalid Op */ + info.si_code = FPE_FLTINV; + break; + case 0x002: /* Denormalize */ + case 0x010: /* Underflow */ + info.si_code = FPE_FLTUND; + break; + case 0x004: /* Zero Divide */ + info.si_code = FPE_FLTDIV; + break; + case 0x008: /* Overflow */ + info.si_code = FPE_FLTOVF; + break; + case 0x020: /* Precision */ + info.si_code = FPE_FLTRES; + break; } force_sig_info(SIGFPE, &info, task); } @@ -1092,7 +1117,7 @@ asmlinkage void __attribute__((weak)) mce_threshold_interrupt(void) } /* - * 'math_state_restore()' saves the current math information in the + * 'math_state_restore()' saves the current math information in the * old math state array, and gets the new ones from the current task * * Careful.. There are problems with IBM-designed IRQ13 behaviour. @@ -1117,7 +1142,7 @@ asmlinkage void math_state_restore(void) local_irq_disable(); } - clts(); /* Allow maths ops (or we recurse) */ + clts(); /* Allow maths ops (or we recurse) */ restore_fpu_checking(&me->thread.xstate->fxsave); task_thread_info(me)->status |= TS_USEDFPU; me->fpu_counter++; @@ -1126,64 +1151,61 @@ EXPORT_SYMBOL_GPL(math_state_restore); void __init trap_init(void) { - set_intr_gate(0,÷_error); - set_intr_gate_ist(1,&debug,DEBUG_STACK); - set_intr_gate_ist(2,&nmi,NMI_STACK); - set_system_gate_ist(3,&int3,DEBUG_STACK); /* int3 can be called from all */ - set_system_gate(4,&overflow); /* int4 can be called from all */ - set_intr_gate(5,&bounds); - set_intr_gate(6,&invalid_op); - set_intr_gate(7,&device_not_available); - set_intr_gate_ist(8,&double_fault, DOUBLEFAULT_STACK); - 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(14,&page_fault); - set_intr_gate(15,&spurious_interrupt_bug); - set_intr_gate(16,&coprocessor_error); - set_intr_gate(17,&alignment_check); + set_intr_gate(0, ÷_error); + set_intr_gate_ist(1, &debug, DEBUG_STACK); + set_intr_gate_ist(2, &nmi, NMI_STACK); + set_system_gate_ist(3, &int3, DEBUG_STACK); /* int3 can be called from all */ + set_system_gate(4, &overflow); /* int4 can be called from all */ + set_intr_gate(5, &bounds); + set_intr_gate(6, &invalid_op); + set_intr_gate(7, &device_not_available); + set_intr_gate_ist(8, &double_fault, DOUBLEFAULT_STACK); + 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(14, &page_fault); + set_intr_gate(15, &spurious_interrupt_bug); + set_intr_gate(16, &coprocessor_error); + set_intr_gate(17, &alignment_check); #ifdef CONFIG_X86_MCE - set_intr_gate_ist(18,&machine_check, MCE_STACK); + set_intr_gate_ist(18, &machine_check, MCE_STACK); #endif - set_intr_gate(19,&simd_coprocessor_error); + set_intr_gate(19, &simd_coprocessor_error); #ifdef CONFIG_IA32_EMULATION set_system_gate(IA32_SYSCALL_VECTOR, ia32_syscall); #endif - /* * initialize the per thread extended state: */ - init_thread_xstate(); + init_thread_xstate(); /* - * Should be a barrier for any external CPU state. + * Should be a barrier for any external CPU state: */ cpu_init(); } - static int __init oops_setup(char *s) -{ +{ if (!s) return -EINVAL; if (!strcmp(s, "panic")) panic_on_oops = 1; return 0; -} +} early_param("oops", oops_setup); static int __init kstack_setup(char *s) { if (!s) return -EINVAL; - kstack_depth_to_print = simple_strtoul(s,NULL,0); + kstack_depth_to_print = simple_strtoul(s, NULL, 0); return 0; } early_param("kstack", kstack_setup); - static int __init code_bytes_setup(char *s) { code_bytes = simple_strtoul(s, NULL, 0); diff --git a/arch/x86/kernel/tsc.c b/arch/x86/kernel/tsc.c new file mode 100644 index 00000000000..3c36f92160c --- /dev/null +++ b/arch/x86/kernel/tsc.c @@ -0,0 +1,533 @@ +#include <linux/kernel.h> +#include <linux/sched.h> +#include <linux/init.h> +#include <linux/module.h> +#include <linux/timer.h> +#include <linux/acpi_pmtmr.h> +#include <linux/cpufreq.h> +#include <linux/dmi.h> +#include <linux/delay.h> +#include <linux/clocksource.h> +#include <linux/percpu.h> + +#include <asm/hpet.h> +#include <asm/timer.h> +#include <asm/vgtod.h> +#include <asm/time.h> +#include <asm/delay.h> + +unsigned int cpu_khz; /* TSC clocks / usec, not used here */ +EXPORT_SYMBOL(cpu_khz); +unsigned int tsc_khz; +EXPORT_SYMBOL(tsc_khz); + +/* + * TSC can be unstable due to cpufreq or due to unsynced TSCs + */ +static int tsc_unstable; + +/* native_sched_clock() is called before tsc_init(), so + we must start with the TSC soft disabled to prevent + erroneous rdtsc usage on !cpu_has_tsc processors */ +static int tsc_disabled = -1; + +/* + * Scheduler clock - returns current time in nanosec units. + */ +u64 native_sched_clock(void) +{ + u64 this_offset; + + /* + * Fall back to jiffies if there's no TSC available: + * ( But note that we still use it if the TSC is marked + * unstable. We do this because unlike Time Of Day, + * the scheduler clock tolerates small errors and it's + * very important for it to be as fast as the platform + * can achive it. ) + */ + if (unlikely(tsc_disabled)) { + /* 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); + + /* return the value in ns */ + return cycles_2_ns(this_offset); +} + +/* We need to define a real function for sched_clock, to override the + weak default version */ +#ifdef CONFIG_PARAVIRT +unsigned long long sched_clock(void) +{ + return paravirt_sched_clock(); +} +#else +unsigned long long +sched_clock(void) __attribute__((alias("native_sched_clock"))); +#endif + +int check_tsc_unstable(void) +{ + return tsc_unstable; +} +EXPORT_SYMBOL_GPL(check_tsc_unstable); + +#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"); + tsc_disabled = 1; + return 1; +} +#else +/* + * disable flag for tsc. Takes effect by clearing the TSC cpu flag + * in cpu/common.c + */ +int __init notsc_setup(char *str) +{ + setup_clear_cpu_cap(X86_FEATURE_TSC); + return 1; +} +#endif + +__setup("notsc", notsc_setup); + +#define MAX_RETRIES 5 +#define SMI_TRESHOLD 50000 + +/* + * Read TSC and the reference counters. Take care of SMI disturbance + */ +static u64 __init tsc_read_refs(u64 *pm, u64 *hpet) +{ + u64 t1, t2; + int i; + + for (i = 0; i < MAX_RETRIES; i++) { + t1 = get_cycles(); + if (hpet) + *hpet = hpet_readl(HPET_COUNTER) & 0xFFFFFFFF; + else + *pm = acpi_pm_read_early(); + t2 = get_cycles(); + if ((t2 - t1) < SMI_TRESHOLD) + return t2; + } + return ULLONG_MAX; +} + +/** + * native_calibrate_tsc - calibrate the tsc on boot + */ +unsigned long native_calibrate_tsc(void) +{ + unsigned long flags; + u64 tsc1, tsc2, tr1, tr2, delta, pm1, pm2, hpet1, hpet2; + int hpet = is_hpet_enabled(); + unsigned int tsc_khz_val = 0; + + local_irq_save(flags); + + tsc1 = tsc_read_refs(&pm1, hpet ? &hpet1 : NULL); + + outb((inb(0x61) & ~0x02) | 0x01, 0x61); + + outb(0xb0, 0x43); + outb((CLOCK_TICK_RATE / (1000 / 50)) & 0xff, 0x42); + outb((CLOCK_TICK_RATE / (1000 / 50)) >> 8, 0x42); + tr1 = get_cycles(); + while ((inb(0x61) & 0x20) == 0); + tr2 = get_cycles(); + + tsc2 = tsc_read_refs(&pm2, hpet ? &hpet2 : NULL); + + local_irq_restore(flags); + + /* + * Preset the result with the raw and inaccurate PIT + * calibration value + */ + delta = (tr2 - tr1); + do_div(delta, 50); + tsc_khz_val = delta; + + /* hpet or pmtimer available ? */ + if (!hpet && !pm1 && !pm2) { + printk(KERN_INFO "TSC calibrated against PIT\n"); + goto out; + } + + /* Check, whether the sampling was disturbed by an SMI */ + if (tsc1 == ULLONG_MAX || tsc2 == ULLONG_MAX) { + printk(KERN_WARNING "TSC calibration disturbed by SMI, " + "using PIT calibration result\n"); + goto out; + } + + tsc2 = (tsc2 - tsc1) * 1000000LL; + + if (hpet) { + printk(KERN_INFO "TSC calibrated against HPET\n"); + if (hpet2 < hpet1) + hpet2 += 0x100000000ULL; + hpet2 -= hpet1; + tsc1 = ((u64)hpet2 * hpet_readl(HPET_PERIOD)); + do_div(tsc1, 1000000); + } else { + printk(KERN_INFO "TSC calibrated against PM_TIMER\n"); + if (pm2 < pm1) + pm2 += (u64)ACPI_PM_OVRRUN; + pm2 -= pm1; + tsc1 = pm2 * 1000000000LL; + do_div(tsc1, PMTMR_TICKS_PER_SEC); + } + + do_div(tsc2, tsc1); + tsc_khz_val = tsc2; + +out: + return tsc_khz_val; +} + + +#ifdef CONFIG_X86_32 +/* Only called from the Powernow K7 cpu freq driver */ +int recalibrate_cpu_khz(void) +{ +#ifndef CONFIG_SMP + unsigned long cpu_khz_old = cpu_khz; + + if (cpu_has_tsc) { + tsc_khz = calibrate_tsc(); + cpu_khz = tsc_khz; + cpu_data(0).loops_per_jiffy = + cpufreq_scale(cpu_data(0).loops_per_jiffy, + cpu_khz_old, cpu_khz); + return 0; + } else + return -ENODEV; +#else + return -ENODEV; +#endif +} + +EXPORT_SYMBOL(recalibrate_cpu_khz); + +#endif /* CONFIG_X86_32 */ + +/* 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); + +static void set_cyc2ns_scale(unsigned long cpu_khz, int cpu) +{ + unsigned long long tsc_now, ns_now; + unsigned long flags, *scale; + + local_irq_save(flags); + sched_clock_idle_sleep_event(); + + scale = &per_cpu(cyc2ns, cpu); + + rdtscll(tsc_now); + ns_now = __cycles_2_ns(tsc_now); + + if (cpu_khz) + *scale = (NSEC_PER_MSEC << CYC2NS_SCALE_FACTOR)/cpu_khz; + + sched_clock_idle_wakeup_event(0); + local_irq_restore(flags); +} + +#ifdef CONFIG_CPU_FREQ + +/* Frequency scaling support. Adjust the TSC based timer when the cpu frequency + * changes. + * + * RED-PEN: On SMP we assume all CPUs run with the same frequency. It's + * not that important because current Opteron setups do not support + * scaling on SMP anyroads. + * + * Should fix up last_tsc too. Currently gettimeofday in the + * first tick after the change will be slightly wrong. + */ + +static unsigned int ref_freq; +static unsigned long loops_per_jiffy_ref; +static unsigned long tsc_khz_ref; + +static int time_cpufreq_notifier(struct notifier_block *nb, unsigned long val, + void *data) +{ + struct cpufreq_freqs *freq = data; + unsigned long *lpj, dummy; + + if (cpu_has(&cpu_data(freq->cpu), X86_FEATURE_CONSTANT_TSC)) + return 0; + + lpj = &dummy; + if (!(freq->flags & CPUFREQ_CONST_LOOPS)) +#ifdef CONFIG_SMP + lpj = &cpu_data(freq->cpu).loops_per_jiffy; +#else + lpj = &boot_cpu_data.loops_per_jiffy; +#endif + + if (!ref_freq) { + ref_freq = freq->old; + loops_per_jiffy_ref = *lpj; + tsc_khz_ref = tsc_khz; + } + if ((val == CPUFREQ_PRECHANGE && freq->old < freq->new) || + (val == CPUFREQ_POSTCHANGE && freq->old > freq->new) || + (val == CPUFREQ_RESUMECHANGE)) { + *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_ref, freq->cpu); + + return 0; +} + +static struct notifier_block time_cpufreq_notifier_block = { + .notifier_call = time_cpufreq_notifier +}; + +static int __init cpufreq_tsc(void) +{ + cpufreq_register_notifier(&time_cpufreq_notifier_block, + CPUFREQ_TRANSITION_NOTIFIER); + return 0; +} + +core_initcall(cpufreq_tsc); + +#endif /* CONFIG_CPU_FREQ */ + +/* clocksource code */ + +static struct clocksource clocksource_tsc; + +/* + * We compare the TSC to the cycle_last value in the clocksource + * structure to avoid a nasty time-warp. This can be observed in a + * very small window right after one CPU updated cycle_last under + * xtime/vsyscall_gtod lock and the other CPU reads a TSC value which + * is smaller than the cycle_last reference value due to a TSC which + * is slighty behind. This delta is nowhere else observable, but in + * that case it results in a forward time jump in the range of hours + * due to the unsigned delta calculation of the time keeping core + * code, which is necessary to support wrapping clocksources like pm + * timer. + */ +static cycle_t read_tsc(void) +{ + cycle_t ret = (cycle_t)get_cycles(); + + return ret >= clocksource_tsc.cycle_last ? + ret : clocksource_tsc.cycle_last; +} + +static cycle_t __vsyscall_fn vread_tsc(void) +{ + cycle_t ret = (cycle_t)vget_cycles(); + + return ret >= __vsyscall_gtod_data.clock.cycle_last ? + ret : __vsyscall_gtod_data.clock.cycle_last; +} + +static struct clocksource clocksource_tsc = { + .name = "tsc", + .rating = 300, + .read = read_tsc, + .mask = CLOCKSOURCE_MASK(64), + .shift = 22, + .flags = CLOCK_SOURCE_IS_CONTINUOUS | + CLOCK_SOURCE_MUST_VERIFY, +#ifdef CONFIG_X86_64 + .vread = vread_tsc, +#endif +}; + +void mark_tsc_unstable(char *reason) +{ + if (!tsc_unstable) { + tsc_unstable = 1; + printk("Marking TSC unstable due to %s\n", reason); + /* Change only the rating, when not registered */ + if (clocksource_tsc.mult) + clocksource_change_rating(&clocksource_tsc, 0); + else + clocksource_tsc.rating = 0; + } +} + +EXPORT_SYMBOL_GPL(mark_tsc_unstable); + +static int __init dmi_mark_tsc_unstable(const struct dmi_system_id *d) +{ + printk(KERN_NOTICE "%s detected: marking TSC unstable.\n", + d->ident); + tsc_unstable = 1; + return 0; +} + +/* List of systems that have known TSC problems */ +static struct dmi_system_id __initdata bad_tsc_dmi_table[] = { + { + .callback = dmi_mark_tsc_unstable, + .ident = "IBM Thinkpad 380XD", + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "IBM"), + DMI_MATCH(DMI_BOARD_NAME, "2635FA0"), + }, + }, + {} +}; + +/* + * Geode_LX - the OLPC CPU has a possibly a very reliable TSC + */ +#ifdef CONFIG_MGEODE_LX +/* RTSC counts during suspend */ +#define RTSC_SUSP 0x100 + +static void __init check_geode_tsc_reliable(void) +{ + unsigned long res_low, res_high; + + rdmsr_safe(MSR_GEODE_BUSCONT_CONF0, &res_low, &res_high); + if (res_low & RTSC_SUSP) + clocksource_tsc.flags &= ~CLOCK_SOURCE_MUST_VERIFY; +} +#else +static inline void check_geode_tsc_reliable(void) { } +#endif + +/* + * Make an educated guess if the TSC is trustworthy and synchronized + * over all CPUs. + */ +__cpuinit int unsynchronized_tsc(void) +{ + if (!cpu_has_tsc || tsc_unstable) + return 1; + +#ifdef CONFIG_SMP + if (apic_is_clustered_box()) + return 1; +#endif + + if (boot_cpu_has(X86_FEATURE_CONSTANT_TSC)) + return 0; + /* + * Intel systems are normally all synchronized. + * Exceptions must mark TSC as unstable: + */ + if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL) { + /* assume multi socket systems are not synchronized: */ + if (num_possible_cpus() > 1) + tsc_unstable = 1; + } + + return tsc_unstable; +} + +static void __init init_tsc_clocksource(void) +{ + clocksource_tsc.mult = clocksource_khz2mult(tsc_khz, + clocksource_tsc.shift); + /* lower the rating if we already know its unstable: */ + if (check_tsc_unstable()) { + clocksource_tsc.rating = 0; + clocksource_tsc.flags &= ~CLOCK_SOURCE_IS_CONTINUOUS; + } + clocksource_register(&clocksource_tsc); +} + +void __init tsc_init(void) +{ + u64 lpj; + int cpu; + + if (!cpu_has_tsc) + return; + + tsc_khz = calibrate_tsc(); + cpu_khz = tsc_khz; + + if (!tsc_khz) { + mark_tsc_unstable("could not calculate TSC khz"); + return; + } + +#ifdef CONFIG_X86_64 + if (cpu_has(&boot_cpu_data, X86_FEATURE_CONSTANT_TSC) && + (boot_cpu_data.x86_vendor == X86_VENDOR_AMD)) + cpu_khz = calibrate_cpu(); +#endif + + lpj = ((u64)tsc_khz * 1000); + do_div(lpj, HZ); + lpj_fine = lpj; + + printk("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 + * all the scale factors for all CPUs, assuming the same + * speed as the bootup CPU. (cpufreq notifiers will fix this + * up if their speed diverges) + */ + for_each_possible_cpu(cpu) + set_cyc2ns_scale(cpu_khz, cpu); + + if (tsc_disabled > 0) + return; + + /* now allow native_sched_clock() to use rdtsc */ + tsc_disabled = 0; + + use_tsc_delay(); + /* Check and install the TSC clocksource */ + dmi_check_system(bad_tsc_dmi_table); + + if (unsynchronized_tsc()) + mark_tsc_unstable("TSCs unsynchronized"); + + check_geode_tsc_reliable(); + init_tsc_clocksource(); +} + diff --git a/arch/x86/kernel/tsc_32.c b/arch/x86/kernel/tsc_32.c deleted file mode 100644 index 65b70637ad9..00000000000 --- a/arch/x86/kernel/tsc_32.c +++ /dev/null @@ -1,451 +0,0 @@ -#include <linux/sched.h> -#include <linux/clocksource.h> -#include <linux/workqueue.h> -#include <linux/cpufreq.h> -#include <linux/jiffies.h> -#include <linux/init.h> -#include <linux/dmi.h> -#include <linux/percpu.h> - -#include <asm/delay.h> -#include <asm/tsc.h> -#include <asm/io.h> -#include <asm/timer.h> - -#include "mach_timer.h" - -/* native_sched_clock() is called before tsc_init(), so - we must start with the TSC soft disabled to prevent - erroneous rdtsc usage on !cpu_has_tsc processors */ -static int tsc_disabled = -1; - -/* - * On some systems the TSC frequency does not - * change with the cpu frequency. So we need - * an extra value to store the TSC freq - */ -unsigned int tsc_khz; -EXPORT_SYMBOL_GPL(tsc_khz); - -#ifdef CONFIG_X86_TSC -static int __init tsc_setup(char *str) -{ - printk(KERN_WARNING "notsc: Kernel compiled with CONFIG_X86_TSC, " - "cannot disable TSC completely.\n"); - tsc_disabled = 1; - return 1; -} -#else -/* - * disable flag for tsc. Takes effect by clearing the TSC cpu flag - * in cpu/common.c - */ -static int __init tsc_setup(char *str) -{ - setup_clear_cpu_cap(X86_FEATURE_TSC); - return 1; -} -#endif - -__setup("notsc", tsc_setup); - -/* - * code to mark and check if the TSC is unstable - * due to cpufreq or due to unsynced TSCs - */ -static int tsc_unstable; - -int check_tsc_unstable(void) -{ - return tsc_unstable; -} -EXPORT_SYMBOL_GPL(check_tsc_unstable); - -/* 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); - -static void set_cyc2ns_scale(unsigned long cpu_khz, int cpu) -{ - unsigned long long tsc_now, ns_now; - unsigned long flags, *scale; - - local_irq_save(flags); - sched_clock_idle_sleep_event(); - - scale = &per_cpu(cyc2ns, cpu); - - rdtscll(tsc_now); - ns_now = __cycles_2_ns(tsc_now); - - if (cpu_khz) - *scale = (NSEC_PER_MSEC << CYC2NS_SCALE_FACTOR)/cpu_khz; - - /* - * Start smoothly with the new frequency: - */ - sched_clock_idle_wakeup_event(0); - local_irq_restore(flags); -} - -/* - * Scheduler clock - returns current time in nanosec units. - */ -unsigned long long native_sched_clock(void) -{ - unsigned long long this_offset; - - /* - * Fall back to jiffies if there's no TSC available: - * ( But note that we still use it if the TSC is marked - * unstable. We do this because unlike Time Of Day, - * the scheduler clock tolerates small errors and it's - * very important for it to be as fast as the platform - * can achive it. ) - */ - if (unlikely(tsc_disabled)) - /* 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); - - /* return the value in ns */ - return cycles_2_ns(this_offset); -} - -/* We need to define a real function for sched_clock, to override the - weak default version */ -#ifdef CONFIG_PARAVIRT -unsigned long long sched_clock(void) -{ - return paravirt_sched_clock(); -} -#else -unsigned long long sched_clock(void) - __attribute__((alias("native_sched_clock"))); -#endif - -unsigned long native_calculate_cpu_khz(void) -{ - unsigned long long start, end; - unsigned long count; - u64 delta64 = (u64)ULLONG_MAX; - int i; - unsigned long flags; - - local_irq_save(flags); - - /* run 3 times to ensure the cache is warm and to get an accurate reading */ - for (i = 0; i < 3; i++) { - mach_prepare_counter(); - rdtscll(start); - mach_countup(&count); - rdtscll(end); - - /* - * Error: ECTCNEVERSET - * The CTC wasn't reliable: we got a hit on the very first read, - * or the CPU was so fast/slow that the quotient wouldn't fit in - * 32 bits.. - */ - if (count <= 1) - continue; - - /* cpu freq too slow: */ - if ((end - start) <= CALIBRATE_TIME_MSEC) - continue; - - /* - * We want the minimum time of all runs in case one of them - * is inaccurate due to SMI or other delay - */ - delta64 = min(delta64, (end - start)); - } - - /* cpu freq too fast (or every run was bad): */ - if (delta64 > (1ULL<<32)) - goto err; - - delta64 += CALIBRATE_TIME_MSEC/2; /* round for do_div */ - do_div(delta64,CALIBRATE_TIME_MSEC); - - local_irq_restore(flags); - return (unsigned long)delta64; -err: - local_irq_restore(flags); - return 0; -} - -int recalibrate_cpu_khz(void) -{ -#ifndef CONFIG_SMP - unsigned long cpu_khz_old = cpu_khz; - - if (cpu_has_tsc) { - cpu_khz = calculate_cpu_khz(); - tsc_khz = cpu_khz; - cpu_data(0).loops_per_jiffy = - cpufreq_scale(cpu_data(0).loops_per_jiffy, - cpu_khz_old, cpu_khz); - return 0; - } else - return -ENODEV; -#else - return -ENODEV; -#endif -} - -EXPORT_SYMBOL(recalibrate_cpu_khz); - -#ifdef CONFIG_CPU_FREQ - -/* - * if the CPU frequency is scaled, TSC-based delays will need a different - * loops_per_jiffy value to function properly. - */ -static unsigned int ref_freq; -static unsigned long loops_per_jiffy_ref; -static unsigned long cpu_khz_ref; - -static int -time_cpufreq_notifier(struct notifier_block *nb, unsigned long val, void *data) -{ - struct cpufreq_freqs *freq = data; - - if (!ref_freq) { - if (!freq->old){ - ref_freq = freq->new; - return 0; - } - ref_freq = freq->old; - loops_per_jiffy_ref = cpu_data(freq->cpu).loops_per_jiffy; - cpu_khz_ref = cpu_khz; - } - - if ((val == CPUFREQ_PRECHANGE && freq->old < freq->new) || - (val == CPUFREQ_POSTCHANGE && freq->old > freq->new) || - (val == CPUFREQ_RESUMECHANGE)) { - if (!(freq->flags & CPUFREQ_CONST_LOOPS)) - cpu_data(freq->cpu).loops_per_jiffy = - cpufreq_scale(loops_per_jiffy_ref, - ref_freq, freq->new); - - if (cpu_khz) { - - if (num_online_cpus() == 1) - cpu_khz = cpufreq_scale(cpu_khz_ref, - ref_freq, freq->new); - if (!(freq->flags & CPUFREQ_CONST_LOOPS)) { - tsc_khz = cpu_khz; - set_cyc2ns_scale(cpu_khz, freq->cpu); - /* - * TSC based sched_clock turns - * to junk w/ cpufreq - */ - mark_tsc_unstable("cpufreq changes"); - } - } - } - - return 0; -} - -static struct notifier_block time_cpufreq_notifier_block = { - .notifier_call = time_cpufreq_notifier -}; - -static int __init cpufreq_tsc(void) -{ - return cpufreq_register_notifier(&time_cpufreq_notifier_block, - CPUFREQ_TRANSITION_NOTIFIER); -} -core_initcall(cpufreq_tsc); - -#endif - -/* clock source code */ - -static unsigned long current_tsc_khz; -static struct clocksource clocksource_tsc; - -/* - * We compare the TSC to the cycle_last value in the clocksource - * structure to avoid a nasty time-warp issue. This can be observed in - * a very small window right after one CPU updated cycle_last under - * xtime lock and the other CPU reads a TSC value which is smaller - * than the cycle_last reference value due to a TSC which is slighty - * behind. This delta is nowhere else observable, but in that case it - * results in a forward time jump in the range of hours due to the - * unsigned delta calculation of the time keeping core code, which is - * necessary to support wrapping clocksources like pm timer. - */ -static cycle_t read_tsc(void) -{ - cycle_t ret; - - rdtscll(ret); - - return ret >= clocksource_tsc.cycle_last ? - ret : clocksource_tsc.cycle_last; -} - -static struct clocksource clocksource_tsc = { - .name = "tsc", - .rating = 300, - .read = read_tsc, - .mask = CLOCKSOURCE_MASK(64), - .mult = 0, /* to be set */ - .shift = 22, - .flags = CLOCK_SOURCE_IS_CONTINUOUS | - CLOCK_SOURCE_MUST_VERIFY, -}; - -void mark_tsc_unstable(char *reason) -{ - if (!tsc_unstable) { - tsc_unstable = 1; - printk("Marking TSC unstable due to: %s.\n", reason); - /* Can be called before registration */ - if (clocksource_tsc.mult) - clocksource_change_rating(&clocksource_tsc, 0); - else - clocksource_tsc.rating = 0; - } -} -EXPORT_SYMBOL_GPL(mark_tsc_unstable); - -static int __init dmi_mark_tsc_unstable(const struct dmi_system_id *d) -{ - printk(KERN_NOTICE "%s detected: marking TSC unstable.\n", - d->ident); - tsc_unstable = 1; - return 0; -} - -/* List of systems that have known TSC problems */ -static struct dmi_system_id __initdata bad_tsc_dmi_table[] = { - { - .callback = dmi_mark_tsc_unstable, - .ident = "IBM Thinkpad 380XD", - .matches = { - DMI_MATCH(DMI_BOARD_VENDOR, "IBM"), - DMI_MATCH(DMI_BOARD_NAME, "2635FA0"), - }, - }, - {} -}; - -/* - * Make an educated guess if the TSC is trustworthy and synchronized - * over all CPUs. - */ -__cpuinit int unsynchronized_tsc(void) -{ - if (!cpu_has_tsc || tsc_unstable) - return 1; - - /* Anything with constant TSC should be synchronized */ - if (boot_cpu_has(X86_FEATURE_CONSTANT_TSC)) - return 0; - - /* - * Intel systems are normally all synchronized. - * Exceptions must mark TSC as unstable: - */ - if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL) { - /* assume multi socket systems are not synchronized: */ - if (num_possible_cpus() > 1) - tsc_unstable = 1; - } - return tsc_unstable; -} - -/* - * Geode_LX - the OLPC CPU has a possibly a very reliable TSC - */ -#ifdef CONFIG_MGEODE_LX -/* RTSC counts during suspend */ -#define RTSC_SUSP 0x100 - -static void __init check_geode_tsc_reliable(void) -{ - unsigned long res_low, res_high; - - rdmsr_safe(MSR_GEODE_BUSCONT_CONF0, &res_low, &res_high); - if (res_low & RTSC_SUSP) - clocksource_tsc.flags &= ~CLOCK_SOURCE_MUST_VERIFY; -} -#else -static inline void check_geode_tsc_reliable(void) { } -#endif - - -void __init tsc_init(void) -{ - int cpu; - - if (!cpu_has_tsc || tsc_disabled > 0) - return; - - cpu_khz = calculate_cpu_khz(); - tsc_khz = cpu_khz; - - if (!cpu_khz) { - mark_tsc_unstable("could not calculate TSC khz"); - return; - } - - /* now allow native_sched_clock() to use rdtsc */ - tsc_disabled = 0; - - printk("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 - * all the scale factors for all CPUs, assuming the same - * speed as the bootup CPU. (cpufreq notifiers will fix this - * up if their speed diverges) - */ - for_each_possible_cpu(cpu) - set_cyc2ns_scale(cpu_khz, cpu); - - use_tsc_delay(); - - /* Check and install the TSC clocksource */ - dmi_check_system(bad_tsc_dmi_table); - - unsynchronized_tsc(); - check_geode_tsc_reliable(); - current_tsc_khz = tsc_khz; - clocksource_tsc.mult = clocksource_khz2mult(current_tsc_khz, - clocksource_tsc.shift); - /* lower the rating if we already know its unstable: */ - if (check_tsc_unstable()) { - clocksource_tsc.rating = 0; - clocksource_tsc.flags &= ~CLOCK_SOURCE_IS_CONTINUOUS; - } - clocksource_register(&clocksource_tsc); -} diff --git a/arch/x86/kernel/tsc_64.c b/arch/x86/kernel/tsc_64.c deleted file mode 100644 index 1784b8077a1..00000000000 --- a/arch/x86/kernel/tsc_64.c +++ /dev/null @@ -1,357 +0,0 @@ -#include <linux/kernel.h> -#include <linux/sched.h> -#include <linux/interrupt.h> -#include <linux/init.h> -#include <linux/clocksource.h> -#include <linux/time.h> -#include <linux/acpi.h> -#include <linux/cpufreq.h> -#include <linux/acpi_pmtmr.h> - -#include <asm/hpet.h> -#include <asm/timex.h> -#include <asm/timer.h> -#include <asm/vgtod.h> - -static int notsc __initdata = 0; - -unsigned int cpu_khz; /* TSC clocks / usec, not used here */ -EXPORT_SYMBOL(cpu_khz); -unsigned int tsc_khz; -EXPORT_SYMBOL(tsc_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); - -static void set_cyc2ns_scale(unsigned long cpu_khz, int cpu) -{ - unsigned long long tsc_now, ns_now; - unsigned long flags, *scale; - - local_irq_save(flags); - sched_clock_idle_sleep_event(); - - scale = &per_cpu(cyc2ns, cpu); - - rdtscll(tsc_now); - ns_now = __cycles_2_ns(tsc_now); - - if (cpu_khz) - *scale = (NSEC_PER_MSEC << CYC2NS_SCALE_FACTOR)/cpu_khz; - - sched_clock_idle_wakeup_event(0); - local_irq_restore(flags); -} - -unsigned long long native_sched_clock(void) -{ - unsigned long a = 0; - - /* Could do CPU core sync here. Opteron can execute rdtsc speculatively, - * which means it is not completely exact and may not be monotonous - * between CPUs. But the errors should be too small to matter for - * scheduling purposes. - */ - - rdtscll(a); - return cycles_2_ns(a); -} - -/* We need to define a real function for sched_clock, to override the - weak default version */ -#ifdef CONFIG_PARAVIRT -unsigned long long sched_clock(void) -{ - return paravirt_sched_clock(); -} -#else -unsigned long long -sched_clock(void) __attribute__((alias("native_sched_clock"))); -#endif - - -static int tsc_unstable; - -int check_tsc_unstable(void) -{ - return tsc_unstable; -} -EXPORT_SYMBOL_GPL(check_tsc_unstable); - -#ifdef CONFIG_CPU_FREQ - -/* Frequency scaling support. Adjust the TSC based timer when the cpu frequency - * changes. - * - * RED-PEN: On SMP we assume all CPUs run with the same frequency. It's - * not that important because current Opteron setups do not support - * scaling on SMP anyroads. - * - * Should fix up last_tsc too. Currently gettimeofday in the - * first tick after the change will be slightly wrong. - */ - -static unsigned int ref_freq; -static unsigned long loops_per_jiffy_ref; -static unsigned long tsc_khz_ref; - -static int time_cpufreq_notifier(struct notifier_block *nb, unsigned long val, - void *data) -{ - struct cpufreq_freqs *freq = data; - unsigned long *lpj, dummy; - - if (cpu_has(&cpu_data(freq->cpu), X86_FEATURE_CONSTANT_TSC)) - return 0; - - lpj = &dummy; - if (!(freq->flags & CPUFREQ_CONST_LOOPS)) -#ifdef CONFIG_SMP - lpj = &cpu_data(freq->cpu).loops_per_jiffy; -#else - lpj = &boot_cpu_data.loops_per_jiffy; -#endif - - if (!ref_freq) { - ref_freq = freq->old; - loops_per_jiffy_ref = *lpj; - tsc_khz_ref = tsc_khz; - } - if ((val == CPUFREQ_PRECHANGE && freq->old < freq->new) || - (val == CPUFREQ_POSTCHANGE && freq->old > freq->new) || - (val == CPUFREQ_RESUMECHANGE)) { - *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_ref, freq->cpu); - - return 0; -} - -static struct notifier_block time_cpufreq_notifier_block = { - .notifier_call = time_cpufreq_notifier -}; - -static int __init cpufreq_tsc(void) -{ - cpufreq_register_notifier(&time_cpufreq_notifier_block, - CPUFREQ_TRANSITION_NOTIFIER); - return 0; -} - -core_initcall(cpufreq_tsc); - -#endif - -#define MAX_RETRIES 5 -#define SMI_TRESHOLD 50000 - -/* - * Read TSC and the reference counters. Take care of SMI disturbance - */ -static unsigned long __init tsc_read_refs(unsigned long *pm, - unsigned long *hpet) -{ - unsigned long t1, t2; - int i; - - for (i = 0; i < MAX_RETRIES; i++) { - t1 = get_cycles(); - if (hpet) - *hpet = hpet_readl(HPET_COUNTER) & 0xFFFFFFFF; - else - *pm = acpi_pm_read_early(); - t2 = get_cycles(); - if ((t2 - t1) < SMI_TRESHOLD) - return t2; - } - return ULONG_MAX; -} - -/** - * tsc_calibrate - calibrate the tsc on boot - */ -void __init tsc_calibrate(void) -{ - unsigned long flags, tsc1, tsc2, tr1, tr2, pm1, pm2, hpet1, hpet2; - int hpet = is_hpet_enabled(), cpu; - - local_irq_save(flags); - - tsc1 = tsc_read_refs(&pm1, hpet ? &hpet1 : NULL); - - outb((inb(0x61) & ~0x02) | 0x01, 0x61); - - outb(0xb0, 0x43); - outb((CLOCK_TICK_RATE / (1000 / 50)) & 0xff, 0x42); - outb((CLOCK_TICK_RATE / (1000 / 50)) >> 8, 0x42); - tr1 = get_cycles(); - while ((inb(0x61) & 0x20) == 0); - tr2 = get_cycles(); - - tsc2 = tsc_read_refs(&pm2, hpet ? &hpet2 : NULL); - - local_irq_restore(flags); - - /* - * Preset the result with the raw and inaccurate PIT - * calibration value - */ - tsc_khz = (tr2 - tr1) / 50; - - /* hpet or pmtimer available ? */ - if (!hpet && !pm1 && !pm2) { - printk(KERN_INFO "TSC calibrated against PIT\n"); - goto out; - } - - /* Check, whether the sampling was disturbed by an SMI */ - if (tsc1 == ULONG_MAX || tsc2 == ULONG_MAX) { - printk(KERN_WARNING "TSC calibration disturbed by SMI, " - "using PIT calibration result\n"); - goto out; - } - - tsc2 = (tsc2 - tsc1) * 1000000L; - - if (hpet) { - printk(KERN_INFO "TSC calibrated against HPET\n"); - if (hpet2 < hpet1) - hpet2 += 0x100000000; - hpet2 -= hpet1; - tsc1 = (hpet2 * hpet_readl(HPET_PERIOD)) / 1000000; - } else { - printk(KERN_INFO "TSC calibrated against PM_TIMER\n"); - if (pm2 < pm1) - pm2 += ACPI_PM_OVRRUN; - pm2 -= pm1; - tsc1 = (pm2 * 1000000000) / PMTMR_TICKS_PER_SEC; - } - - tsc_khz = tsc2 / tsc1; - -out: - for_each_possible_cpu(cpu) - set_cyc2ns_scale(tsc_khz, cpu); -} - -/* - * Make an educated guess if the TSC is trustworthy and synchronized - * over all CPUs. - */ -__cpuinit int unsynchronized_tsc(void) -{ - if (tsc_unstable) - return 1; - -#ifdef CONFIG_SMP - if (apic_is_clustered_box()) - return 1; -#endif - - if (boot_cpu_has(X86_FEATURE_CONSTANT_TSC)) - return 0; - - /* Assume multi socket systems are not synchronized */ - return num_present_cpus() > 1; -} - -int __init notsc_setup(char *s) -{ - notsc = 1; - return 1; -} - -__setup("notsc", notsc_setup); - -static struct clocksource clocksource_tsc; - -/* - * We compare the TSC to the cycle_last value in the clocksource - * structure to avoid a nasty time-warp. This can be observed in a - * very small window right after one CPU updated cycle_last under - * xtime/vsyscall_gtod lock and the other CPU reads a TSC value which - * is smaller than the cycle_last reference value due to a TSC which - * is slighty behind. This delta is nowhere else observable, but in - * that case it results in a forward time jump in the range of hours - * due to the unsigned delta calculation of the time keeping core - * code, which is necessary to support wrapping clocksources like pm - * timer. - */ -static cycle_t read_tsc(void) -{ - cycle_t ret = (cycle_t)get_cycles(); - - return ret >= clocksource_tsc.cycle_last ? - ret : clocksource_tsc.cycle_last; -} - -static cycle_t __vsyscall_fn vread_tsc(void) -{ - cycle_t ret = (cycle_t)vget_cycles(); - - return ret >= __vsyscall_gtod_data.clock.cycle_last ? - ret : __vsyscall_gtod_data.clock.cycle_last; -} - -static struct clocksource clocksource_tsc = { - .name = "tsc", - .rating = 300, - .read = read_tsc, - .mask = CLOCKSOURCE_MASK(64), - .shift = 22, - .flags = CLOCK_SOURCE_IS_CONTINUOUS | - CLOCK_SOURCE_MUST_VERIFY, - .vread = vread_tsc, -}; - -void mark_tsc_unstable(char *reason) -{ - if (!tsc_unstable) { - tsc_unstable = 1; - printk("Marking TSC unstable due to %s\n", reason); - /* Change only the rating, when not registered */ - if (clocksource_tsc.mult) - clocksource_change_rating(&clocksource_tsc, 0); - else - clocksource_tsc.rating = 0; - } -} -EXPORT_SYMBOL_GPL(mark_tsc_unstable); - -void __init init_tsc_clocksource(void) -{ - if (!notsc) { - clocksource_tsc.mult = clocksource_khz2mult(tsc_khz, - clocksource_tsc.shift); - if (check_tsc_unstable()) - clocksource_tsc.rating = 0; - - clocksource_register(&clocksource_tsc); - } -} diff --git a/arch/x86/kernel/visws_quirks.c b/arch/x86/kernel/visws_quirks.c new file mode 100644 index 00000000000..e94bdb6add1 --- /dev/null +++ b/arch/x86/kernel/visws_quirks.c @@ -0,0 +1,709 @@ +/* + * SGI Visual Workstation support and quirks, unmaintained. + * + * Split out from setup.c by davej@suse.de + * + * Copyright (C) 1999 Bent Hagemark, Ingo Molnar + * + * SGI Visual Workstation interrupt controller + * + * The Cobalt system ASIC in the Visual Workstation contains a "Cobalt" APIC + * which serves as the main interrupt controller in the system. Non-legacy + * hardware in the system uses this controller directly. Legacy devices + * are connected to the PIIX4 which in turn has its 8259(s) connected to + * a of the Cobalt APIC entry. + * + * 09/02/2000 - Updated for 2.4 by jbarnes@sgi.com + * + * 25/11/2002 - Updated for 2.5 by Andrey Panin <pazke@orbita1.ru> + */ +#include <linux/interrupt.h> +#include <linux/module.h> +#include <linux/init.h> +#include <linux/smp.h> + +#include <asm/visws/cobalt.h> +#include <asm/visws/piix4.h> +#include <asm/arch_hooks.h> +#include <asm/fixmap.h> +#include <asm/reboot.h> +#include <asm/setup.h> +#include <asm/e820.h> +#include <asm/smp.h> +#include <asm/io.h> + +#include <mach_ipi.h> + +#include "mach_apic.h" + +#include <linux/init.h> +#include <linux/smp.h> + +#include <linux/kernel_stat.h> +#include <linux/interrupt.h> +#include <linux/init.h> + +#include <asm/io.h> +#include <asm/apic.h> +#include <asm/i8259.h> +#include <asm/irq_vectors.h> +#include <asm/visws/cobalt.h> +#include <asm/visws/lithium.h> +#include <asm/visws/piix4.h> + +#include <linux/sched.h> +#include <linux/kernel.h> +#include <linux/init.h> +#include <linux/pci.h> +#include <linux/pci_ids.h> + +extern int no_broadcast; + +#include <asm/io.h> +#include <asm/apic.h> +#include <asm/arch_hooks.h> +#include <asm/visws/cobalt.h> +#include <asm/visws/lithium.h> + +char visws_board_type = -1; +char visws_board_rev = -1; + +int is_visws_box(void) +{ + return visws_board_type >= 0; +} + +static int __init visws_time_init_quirk(void) +{ + printk(KERN_INFO "Starting Cobalt Timer system clock\n"); + + /* Set the countdown value */ + co_cpu_write(CO_CPU_TIMEVAL, CO_TIME_HZ/HZ); + + /* Start the timer */ + co_cpu_write(CO_CPU_CTRL, co_cpu_read(CO_CPU_CTRL) | CO_CTRL_TIMERUN); + + /* Enable (unmask) the timer interrupt */ + co_cpu_write(CO_CPU_CTRL, co_cpu_read(CO_CPU_CTRL) & ~CO_CTRL_TIMEMASK); + + /* + * Zero return means the generic timer setup code will set up + * the standard vector: + */ + return 0; +} + +static int __init visws_pre_intr_init_quirk(void) +{ + init_VISWS_APIC_irqs(); + + /* + * We dont want ISA irqs to be set up by the generic code: + */ + return 1; +} + +/* Quirk for machine specific memory setup. */ + +#define MB (1024 * 1024) + +unsigned long sgivwfb_mem_phys; +unsigned long sgivwfb_mem_size; +EXPORT_SYMBOL(sgivwfb_mem_phys); +EXPORT_SYMBOL(sgivwfb_mem_size); + +long long mem_size __initdata = 0; + +static char * __init visws_memory_setup_quirk(void) +{ + long long gfx_mem_size = 8 * MB; + + mem_size = boot_params.alt_mem_k; + + if (!mem_size) { + printk(KERN_WARNING "Bootloader didn't set memory size, upgrade it !\n"); + mem_size = 128 * MB; + } + + /* + * this hardcodes the graphics memory to 8 MB + * it really should be sized dynamically (or at least + * set as a boot param) + */ + if (!sgivwfb_mem_size) { + printk(KERN_WARNING "Defaulting to 8 MB framebuffer size\n"); + sgivwfb_mem_size = 8 * MB; + } + + /* + * Trim to nearest MB + */ + sgivwfb_mem_size &= ~((1 << 20) - 1); + sgivwfb_mem_phys = mem_size - gfx_mem_size; + + e820_add_region(0, LOWMEMSIZE(), E820_RAM); + e820_add_region(HIGH_MEMORY, mem_size - sgivwfb_mem_size - HIGH_MEMORY, E820_RAM); + e820_add_region(sgivwfb_mem_phys, sgivwfb_mem_size, E820_RESERVED); + + return "PROM"; +} + +static void visws_machine_emergency_restart(void) +{ + /* + * Visual Workstations restart after this + * register is poked on the PIIX4 + */ + outb(PIIX4_RESET_VAL, PIIX4_RESET_PORT); +} + +static void visws_machine_power_off(void) +{ + unsigned short pm_status; +/* extern unsigned int pci_bus0; */ + + while ((pm_status = inw(PMSTS_PORT)) & 0x100) + outw(pm_status, PMSTS_PORT); + + outw(PM_SUSPEND_ENABLE, PMCNTRL_PORT); + + mdelay(10); + +#define PCI_CONF1_ADDRESS(bus, devfn, reg) \ + (0x80000000 | (bus << 16) | (devfn << 8) | (reg & ~3)) + +/* outl(PCI_CONF1_ADDRESS(pci_bus0, SPECIAL_DEV, SPECIAL_REG), 0xCF8); */ + outl(PIIX_SPECIAL_STOP, 0xCFC); +} + +static int __init visws_get_smp_config_quirk(unsigned int early) +{ + /* + * Prevent MP-table parsing by the generic code: + */ + return 1; +} + +extern unsigned int __cpuinitdata maxcpus; + +/* + * The Visual Workstation is Intel MP compliant in the hardware + * sense, but it doesn't have a BIOS(-configuration table). + * No problem for Linux. + */ + +static void __init MP_processor_info (struct mpc_config_processor *m) +{ + int ver, logical_apicid; + physid_mask_t apic_cpus; + + if (!(m->mpc_cpuflag & CPU_ENABLED)) + return; + + logical_apicid = m->mpc_apicid; + printk(KERN_INFO "%sCPU #%d %u:%u APIC version %d\n", + m->mpc_cpuflag & CPU_BOOTPROCESSOR ? "Bootup " : "", + m->mpc_apicid, + (m->mpc_cpufeature & CPU_FAMILY_MASK) >> 8, + (m->mpc_cpufeature & CPU_MODEL_MASK) >> 4, + m->mpc_apicver); + + if (m->mpc_cpuflag & CPU_BOOTPROCESSOR) + boot_cpu_physical_apicid = m->mpc_apicid; + + ver = m->mpc_apicver; + if ((ver >= 0x14 && m->mpc_apicid >= 0xff) || m->mpc_apicid >= 0xf) { + printk(KERN_ERR "Processor #%d INVALID. (Max ID: %d).\n", + m->mpc_apicid, MAX_APICS); + return; + } + + apic_cpus = apicid_to_cpu_present(m->mpc_apicid); + physids_or(phys_cpu_present_map, phys_cpu_present_map, apic_cpus); + /* + * Validate version + */ + if (ver == 0x0) { + printk(KERN_ERR "BIOS bug, APIC version is 0 for CPU#%d! " + "fixing up to 0x10. (tell your hw vendor)\n", + m->mpc_apicid); + ver = 0x10; + } + apic_version[m->mpc_apicid] = ver; +} + +int __init visws_find_smp_config_quirk(unsigned int reserve) +{ + struct mpc_config_processor *mp = phys_to_virt(CO_CPU_TAB_PHYS); + unsigned short ncpus = readw(phys_to_virt(CO_CPU_NUM_PHYS)); + + if (ncpus > CO_CPU_MAX) { + printk(KERN_WARNING "find_visws_smp: got cpu count of %d at %p\n", + ncpus, mp); + + ncpus = CO_CPU_MAX; + } + + if (ncpus > maxcpus) + ncpus = maxcpus; + +#ifdef CONFIG_X86_LOCAL_APIC + smp_found_config = 1; +#endif + while (ncpus--) + MP_processor_info(mp++); + + mp_lapic_addr = APIC_DEFAULT_PHYS_BASE; + + return 1; +} + +extern int visws_trap_init_quirk(void); + +void __init visws_early_detect(void) +{ + int raw; + + visws_board_type = (char)(inb_p(PIIX_GPI_BD_REG) & PIIX_GPI_BD_REG) + >> PIIX_GPI_BD_SHIFT; + + if (visws_board_type < 0) + return; + + /* + * Install special quirks for timer, interrupt and memory setup: + */ + arch_time_init_quirk = visws_time_init_quirk; + arch_pre_intr_init_quirk = visws_pre_intr_init_quirk; + arch_memory_setup_quirk = visws_memory_setup_quirk; + + /* + * Fall back to generic behavior for traps: + */ + arch_intr_init_quirk = NULL; + arch_trap_init_quirk = visws_trap_init_quirk; + + /* + * Install reboot quirks: + */ + pm_power_off = visws_machine_power_off; + machine_ops.emergency_restart = visws_machine_emergency_restart; + + /* + * Do not use broadcast IPIs: + */ + no_broadcast = 0; + + /* + * Override generic MP-table parsing: + */ + mach_get_smp_config_quirk = visws_get_smp_config_quirk; + mach_find_smp_config_quirk = visws_find_smp_config_quirk; + +#ifdef CONFIG_X86_IO_APIC + /* + * Turn off IO-APIC detection and initialization: + */ + skip_ioapic_setup = 1; +#endif + + /* + * Get Board rev. + * First, we have to initialize the 307 part to allow us access + * to the GPIO registers. Let's map them at 0x0fc0 which is right + * after the PIIX4 PM section. + */ + outb_p(SIO_DEV_SEL, SIO_INDEX); + outb_p(SIO_GP_DEV, SIO_DATA); /* Talk to GPIO regs. */ + + outb_p(SIO_DEV_MSB, SIO_INDEX); + outb_p(SIO_GP_MSB, SIO_DATA); /* MSB of GPIO base address */ + + outb_p(SIO_DEV_LSB, SIO_INDEX); + outb_p(SIO_GP_LSB, SIO_DATA); /* LSB of GPIO base address */ + + outb_p(SIO_DEV_ENB, SIO_INDEX); + outb_p(1, SIO_DATA); /* Enable GPIO registers. */ + + /* + * Now, we have to map the power management section to write + * a bit which enables access to the GPIO registers. + * What lunatic came up with this shit? + */ + outb_p(SIO_DEV_SEL, SIO_INDEX); + outb_p(SIO_PM_DEV, SIO_DATA); /* Talk to GPIO regs. */ + + outb_p(SIO_DEV_MSB, SIO_INDEX); + outb_p(SIO_PM_MSB, SIO_DATA); /* MSB of PM base address */ + + outb_p(SIO_DEV_LSB, SIO_INDEX); + outb_p(SIO_PM_LSB, SIO_DATA); /* LSB of PM base address */ + + outb_p(SIO_DEV_ENB, SIO_INDEX); + outb_p(1, SIO_DATA); /* Enable PM registers. */ + + /* + * Now, write the PM register which enables the GPIO registers. + */ + outb_p(SIO_PM_FER2, SIO_PM_INDEX); + outb_p(SIO_PM_GP_EN, SIO_PM_DATA); + + /* + * Now, initialize the GPIO registers. + * We want them all to be inputs which is the + * power on default, so let's leave them alone. + * So, let's just read the board rev! + */ + raw = inb_p(SIO_GP_DATA1); + raw &= 0x7f; /* 7 bits of valid board revision ID. */ + + if (visws_board_type == VISWS_320) { + if (raw < 0x6) { + visws_board_rev = 4; + } else if (raw < 0xc) { + visws_board_rev = 5; + } else { + visws_board_rev = 6; + } + } else if (visws_board_type == VISWS_540) { + visws_board_rev = 2; + } else { + visws_board_rev = raw; + } + + printk(KERN_INFO "Silicon Graphics Visual Workstation %s (rev %d) detected\n", + (visws_board_type == VISWS_320 ? "320" : + (visws_board_type == VISWS_540 ? "540" : + "unknown")), visws_board_rev); +} + +#define A01234 (LI_INTA_0 | LI_INTA_1 | LI_INTA_2 | LI_INTA_3 | LI_INTA_4) +#define BCD (LI_INTB | LI_INTC | LI_INTD) +#define ALLDEVS (A01234 | BCD) + +static __init void lithium_init(void) +{ + set_fixmap(FIX_LI_PCIA, LI_PCI_A_PHYS); + set_fixmap(FIX_LI_PCIB, LI_PCI_B_PHYS); + + if ((li_pcia_read16(PCI_VENDOR_ID) != PCI_VENDOR_ID_SGI) || + (li_pcia_read16(PCI_DEVICE_ID) != PCI_DEVICE_ID_SGI_LITHIUM)) { + printk(KERN_EMERG "Lithium hostbridge %c not found\n", 'A'); +/* panic("This machine is not SGI Visual Workstation 320/540"); */ + } + + if ((li_pcib_read16(PCI_VENDOR_ID) != PCI_VENDOR_ID_SGI) || + (li_pcib_read16(PCI_DEVICE_ID) != PCI_DEVICE_ID_SGI_LITHIUM)) { + printk(KERN_EMERG "Lithium hostbridge %c not found\n", 'B'); +/* panic("This machine is not SGI Visual Workstation 320/540"); */ + } + + li_pcia_write16(LI_PCI_INTEN, ALLDEVS); + li_pcib_write16(LI_PCI_INTEN, ALLDEVS); +} + +static __init void cobalt_init(void) +{ + /* + * On normal SMP PC this is used only with SMP, but we have to + * use it and set it up here to start the Cobalt clock + */ + set_fixmap(FIX_APIC_BASE, APIC_DEFAULT_PHYS_BASE); + setup_local_APIC(); + printk(KERN_INFO "Local APIC Version %#x, ID %#x\n", + (unsigned int)apic_read(APIC_LVR), + (unsigned int)apic_read(APIC_ID)); + + set_fixmap(FIX_CO_CPU, CO_CPU_PHYS); + set_fixmap(FIX_CO_APIC, CO_APIC_PHYS); + printk(KERN_INFO "Cobalt Revision %#lx, APIC ID %#lx\n", + co_cpu_read(CO_CPU_REV), co_apic_read(CO_APIC_ID)); + + /* Enable Cobalt APIC being careful to NOT change the ID! */ + co_apic_write(CO_APIC_ID, co_apic_read(CO_APIC_ID) | CO_APIC_ENABLE); + + printk(KERN_INFO "Cobalt APIC enabled: ID reg %#lx\n", + co_apic_read(CO_APIC_ID)); +} + +int __init visws_trap_init_quirk(void) +{ + lithium_init(); + cobalt_init(); + + return 1; +} + +/* + * IRQ controller / APIC support: + */ + +static DEFINE_SPINLOCK(cobalt_lock); + +/* + * Set the given Cobalt APIC Redirection Table entry to point + * to the given IDT vector/index. + */ +static inline void co_apic_set(int entry, int irq) +{ + co_apic_write(CO_APIC_LO(entry), CO_APIC_LEVEL | (irq + FIRST_EXTERNAL_VECTOR)); + co_apic_write(CO_APIC_HI(entry), 0); +} + +/* + * Cobalt (IO)-APIC functions to handle PCI devices. + */ +static inline int co_apic_ide0_hack(void) +{ + extern char visws_board_type; + extern char visws_board_rev; + + if (visws_board_type == VISWS_320 && visws_board_rev == 5) + return 5; + return CO_APIC_IDE0; +} + +static int is_co_apic(unsigned int irq) +{ + if (IS_CO_APIC(irq)) + return CO_APIC(irq); + + switch (irq) { + case 0: return CO_APIC_CPU; + case CO_IRQ_IDE0: return co_apic_ide0_hack(); + case CO_IRQ_IDE1: return CO_APIC_IDE1; + default: return -1; + } +} + + +/* + * This is the SGI Cobalt (IO-)APIC: + */ + +static void enable_cobalt_irq(unsigned int irq) +{ + co_apic_set(is_co_apic(irq), irq); +} + +static void disable_cobalt_irq(unsigned int irq) +{ + int entry = is_co_apic(irq); + + co_apic_write(CO_APIC_LO(entry), CO_APIC_MASK); + co_apic_read(CO_APIC_LO(entry)); +} + +/* + * "irq" really just serves to identify the device. Here is where we + * map this to the Cobalt APIC entry where it's physically wired. + * This is called via request_irq -> setup_irq -> irq_desc->startup() + */ +static unsigned int startup_cobalt_irq(unsigned int irq) +{ + unsigned long flags; + + spin_lock_irqsave(&cobalt_lock, flags); + if ((irq_desc[irq].status & (IRQ_DISABLED | IRQ_INPROGRESS | IRQ_WAITING))) + irq_desc[irq].status &= ~(IRQ_DISABLED | IRQ_INPROGRESS | IRQ_WAITING); + enable_cobalt_irq(irq); + spin_unlock_irqrestore(&cobalt_lock, flags); + return 0; +} + +static void ack_cobalt_irq(unsigned int irq) +{ + unsigned long flags; + + spin_lock_irqsave(&cobalt_lock, flags); + disable_cobalt_irq(irq); + apic_write(APIC_EOI, APIC_EIO_ACK); + spin_unlock_irqrestore(&cobalt_lock, flags); +} + +static void end_cobalt_irq(unsigned int irq) +{ + unsigned long flags; + + spin_lock_irqsave(&cobalt_lock, flags); + if (!(irq_desc[irq].status & (IRQ_DISABLED | IRQ_INPROGRESS))) + enable_cobalt_irq(irq); + spin_unlock_irqrestore(&cobalt_lock, flags); +} + +static struct irq_chip cobalt_irq_type = { + .typename = "Cobalt-APIC", + .startup = startup_cobalt_irq, + .shutdown = disable_cobalt_irq, + .enable = enable_cobalt_irq, + .disable = disable_cobalt_irq, + .ack = ack_cobalt_irq, + .end = end_cobalt_irq, +}; + + +/* + * This is the PIIX4-based 8259 that is wired up indirectly to Cobalt + * -- not the manner expected by the code in i8259.c. + * + * there is a 'master' physical interrupt source that gets sent to + * the CPU. But in the chipset there are various 'virtual' interrupts + * waiting to be handled. We represent this to Linux through a 'master' + * interrupt controller type, and through a special virtual interrupt- + * controller. Device drivers only see the virtual interrupt sources. + */ +static unsigned int startup_piix4_master_irq(unsigned int irq) +{ + init_8259A(0); + + return startup_cobalt_irq(irq); +} + +static void end_piix4_master_irq(unsigned int irq) +{ + unsigned long flags; + + spin_lock_irqsave(&cobalt_lock, flags); + enable_cobalt_irq(irq); + spin_unlock_irqrestore(&cobalt_lock, flags); +} + +static struct irq_chip piix4_master_irq_type = { + .typename = "PIIX4-master", + .startup = startup_piix4_master_irq, + .ack = ack_cobalt_irq, + .end = end_piix4_master_irq, +}; + + +static struct irq_chip piix4_virtual_irq_type = { + .typename = "PIIX4-virtual", + .shutdown = disable_8259A_irq, + .enable = enable_8259A_irq, + .disable = disable_8259A_irq, +}; + + +/* + * PIIX4-8259 master/virtual functions to handle interrupt requests + * from legacy devices: floppy, parallel, serial, rtc. + * + * None of these get Cobalt APIC entries, neither do they have IDT + * entries. These interrupts are purely virtual and distributed from + * the 'master' interrupt source: CO_IRQ_8259. + * + * When the 8259 interrupts its handler figures out which of these + * devices is interrupting and dispatches to its handler. + * + * CAREFUL: devices see the 'virtual' interrupt only. Thus disable/ + * enable_irq gets the right irq. This 'master' irq is never directly + * manipulated by any driver. + */ +static irqreturn_t piix4_master_intr(int irq, void *dev_id) +{ + int realirq; + irq_desc_t *desc; + unsigned long flags; + + spin_lock_irqsave(&i8259A_lock, flags); + + /* Find out what's interrupting in the PIIX4 master 8259 */ + outb(0x0c, 0x20); /* OCW3 Poll command */ + realirq = inb(0x20); + + /* + * Bit 7 == 0 means invalid/spurious + */ + if (unlikely(!(realirq & 0x80))) + goto out_unlock; + + realirq &= 7; + + if (unlikely(realirq == 2)) { + outb(0x0c, 0xa0); + realirq = inb(0xa0); + + if (unlikely(!(realirq & 0x80))) + goto out_unlock; + + realirq = (realirq & 7) + 8; + } + + /* mask and ack interrupt */ + cached_irq_mask |= 1 << realirq; + if (unlikely(realirq > 7)) { + inb(0xa1); + outb(cached_slave_mask, 0xa1); + outb(0x60 + (realirq & 7), 0xa0); + outb(0x60 + 2, 0x20); + } else { + inb(0x21); + outb(cached_master_mask, 0x21); + outb(0x60 + realirq, 0x20); + } + + spin_unlock_irqrestore(&i8259A_lock, flags); + + desc = irq_desc + realirq; + + /* + * handle this 'virtual interrupt' as a Cobalt one now. + */ + kstat_cpu(smp_processor_id()).irqs[realirq]++; + + if (likely(desc->action != NULL)) + handle_IRQ_event(realirq, desc->action); + + if (!(desc->status & IRQ_DISABLED)) + enable_8259A_irq(realirq); + + return IRQ_HANDLED; + +out_unlock: + spin_unlock_irqrestore(&i8259A_lock, flags); + return IRQ_NONE; +} + +static struct irqaction master_action = { + .handler = piix4_master_intr, + .name = "PIIX4-8259", +}; + +static struct irqaction cascade_action = { + .handler = no_action, + .name = "cascade", +}; + + +void init_VISWS_APIC_irqs(void) +{ + int i; + + for (i = 0; i < CO_IRQ_APIC0 + CO_APIC_LAST + 1; i++) { + irq_desc[i].status = IRQ_DISABLED; + irq_desc[i].action = 0; + irq_desc[i].depth = 1; + + if (i == 0) { + irq_desc[i].chip = &cobalt_irq_type; + } + else if (i == CO_IRQ_IDE0) { + irq_desc[i].chip = &cobalt_irq_type; + } + else if (i == CO_IRQ_IDE1) { + irq_desc[i].chip = &cobalt_irq_type; + } + else if (i == CO_IRQ_8259) { + irq_desc[i].chip = &piix4_master_irq_type; + } + else if (i < CO_IRQ_APIC0) { + irq_desc[i].chip = &piix4_virtual_irq_type; + } + else if (IS_CO_APIC(i)) { + irq_desc[i].chip = &cobalt_irq_type; + } + } + + setup_irq(CO_IRQ_8259, &master_action); + setup_irq(2, &cascade_action); +} diff --git a/arch/x86/kernel/vmi_32.c b/arch/x86/kernel/vmi_32.c index 956f38927aa..b15346092b7 100644 --- a/arch/x86/kernel/vmi_32.c +++ b/arch/x86/kernel/vmi_32.c @@ -151,7 +151,7 @@ static unsigned vmi_patch(u8 type, u16 clobbers, void *insns, insns, ip); case PARAVIRT_PATCH(pv_cpu_ops.iret): return patch_internal(VMI_CALL_IRET, len, insns, ip); - case PARAVIRT_PATCH(pv_cpu_ops.irq_enable_syscall_ret): + case PARAVIRT_PATCH(pv_cpu_ops.irq_enable_sysexit): return patch_internal(VMI_CALL_SYSEXIT, len, insns, ip); default: break; @@ -896,7 +896,7 @@ static inline int __init activate_vmi(void) * the backend. They are performance critical anyway, so requiring * a patch is not a big problem. */ - pv_cpu_ops.irq_enable_syscall_ret = (void *)0xfeedbab0; + pv_cpu_ops.irq_enable_sysexit = (void *)0xfeedbab0; pv_cpu_ops.iret = (void *)0xbadbab0; #ifdef CONFIG_SMP @@ -932,7 +932,7 @@ static inline int __init activate_vmi(void) pv_apic_ops.setup_secondary_clock = vmi_time_ap_init; #endif pv_time_ops.sched_clock = vmi_sched_clock; - pv_time_ops.get_cpu_khz = vmi_cpu_khz; + pv_time_ops.get_tsc_khz = vmi_tsc_khz; /* We have true wallclock functions; disable CMOS clock sync */ no_sync_cmos_clock = 1; diff --git a/arch/x86/kernel/vmiclock_32.c b/arch/x86/kernel/vmiclock_32.c index a2b030780aa..6953859fe28 100644 --- a/arch/x86/kernel/vmiclock_32.c +++ b/arch/x86/kernel/vmiclock_32.c @@ -33,8 +33,7 @@ #include <asm/apic.h> #include <asm/timer.h> #include <asm/i8253.h> - -#include <irq_vectors.h> +#include <asm/irq_vectors.h> #define VMI_ONESHOT (VMI_ALARM_IS_ONESHOT | VMI_CYCLES_REAL | vmi_get_alarm_wiring()) #define VMI_PERIODIC (VMI_ALARM_IS_PERIODIC | VMI_CYCLES_REAL | vmi_get_alarm_wiring()) @@ -70,8 +69,8 @@ unsigned long long vmi_sched_clock(void) return cycles_2_ns(vmi_timer_ops.get_cycle_counter(VMI_CYCLES_AVAILABLE)); } -/* paravirt_ops.get_cpu_khz = vmi_cpu_khz */ -unsigned long vmi_cpu_khz(void) +/* paravirt_ops.get_tsc_khz = vmi_tsc_khz */ +unsigned long vmi_tsc_khz(void) { unsigned long long khz; khz = vmi_timer_ops.get_cycle_frequency(); diff --git a/arch/x86/kernel/vmlinux_32.lds.S b/arch/x86/kernel/vmlinux_32.lds.S index ce5ed083a1e..2674f579627 100644 --- a/arch/x86/kernel/vmlinux_32.lds.S +++ b/arch/x86/kernel/vmlinux_32.lds.S @@ -60,13 +60,6 @@ SECTIONS BUG_TABLE :text - . = ALIGN(4); - .tracedata : AT(ADDR(.tracedata) - LOAD_OFFSET) { - __tracedata_start = .; - *(.tracedata) - __tracedata_end = .; - } - RODATA /* writeable */ diff --git a/arch/x86/kernel/vmlinux_64.lds.S b/arch/x86/kernel/vmlinux_64.lds.S index fad3674b06a..fd246e22fe6 100644 --- a/arch/x86/kernel/vmlinux_64.lds.S +++ b/arch/x86/kernel/vmlinux_64.lds.S @@ -53,13 +53,6 @@ SECTIONS RODATA - . = ALIGN(4); - .tracedata : AT(ADDR(.tracedata) - LOAD_OFFSET) { - __tracedata_start = .; - *(.tracedata) - __tracedata_end = .; - } - . = ALIGN(PAGE_SIZE); /* Align data segment to page size boundary */ /* Data */ .data : AT(ADDR(.data) - LOAD_OFFSET) { @@ -177,6 +170,7 @@ SECTIONS *(.con_initcall.init) } __con_initcall_end = .; + . = ALIGN(16); __x86cpuvendor_start = .; .x86cpuvendor.init : AT(ADDR(.x86cpuvendor.init) - LOAD_OFFSET) { *(.x86cpuvendor.init) diff --git a/arch/x86/kernel/vsmp_64.c b/arch/x86/kernel/vsmp_64.c index ba8c0b75ab0..0c029e8959c 100644 --- a/arch/x86/kernel/vsmp_64.c +++ b/arch/x86/kernel/vsmp_64.c @@ -15,9 +15,12 @@ #include <linux/init.h> #include <linux/pci_ids.h> #include <linux/pci_regs.h> + +#include <asm/apic.h> #include <asm/pci-direct.h> #include <asm/io.h> #include <asm/paravirt.h> +#include <asm/setup.h> #if defined CONFIG_PCI && defined CONFIG_PARAVIRT /* diff --git a/arch/x86/kernel/vsyscall_64.c b/arch/x86/kernel/vsyscall_64.c index 61efa2f7d56..c87cbd84c3e 100644 --- a/arch/x86/kernel/vsyscall_64.c +++ b/arch/x86/kernel/vsyscall_64.c @@ -249,7 +249,7 @@ static ctl_table kernel_root_table2[] = { doesn't violate that. We'll find out if it does. */ static void __cpuinit vsyscall_set_cpu(int cpu) { - unsigned long *d; + unsigned long d; unsigned long node = 0; #ifdef CONFIG_NUMA node = cpu_to_node(cpu); @@ -260,11 +260,11 @@ static void __cpuinit vsyscall_set_cpu(int cpu) /* Store cpu number in limit so that it can be loaded quickly in user space in vgetcpu. 12 bits for the CPU and 8 bits for the node. */ - d = (unsigned long *)(get_cpu_gdt_table(cpu) + GDT_ENTRY_PER_CPU); - *d = 0x0f40000000000ULL; - *d |= cpu; - *d |= (node & 0xf) << 12; - *d |= (node >> 4) << 48; + d = 0x0f40000000000ULL; + d |= cpu; + d |= (node & 0xf) << 12; + d |= (node >> 4) << 48; + write_gdt_entry(get_cpu_gdt_table(cpu), GDT_ENTRY_PER_CPU, &d, DESCTYPE_S); } static void __cpuinit cpu_vsyscall_init(void *arg) diff --git a/arch/x86/kernel/x8664_ksyms_64.c b/arch/x86/kernel/x8664_ksyms_64.c index f6c05d0410f..2f306a82689 100644 --- a/arch/x86/kernel/x8664_ksyms_64.c +++ b/arch/x86/kernel/x8664_ksyms_64.c @@ -53,8 +53,3 @@ EXPORT_SYMBOL(init_level4_pgt); EXPORT_SYMBOL(load_gs_index); EXPORT_SYMBOL(_proxy_pda); - -#ifdef CONFIG_PARAVIRT -/* Virtualized guests may want to use it */ -EXPORT_SYMBOL_GPL(cpu_gdt_descr); -#endif |