diff options
Diffstat (limited to 'arch/x86/kernel/cpu')
43 files changed, 5286 insertions, 858 deletions
diff --git a/arch/x86/kernel/cpu/Makefile b/arch/x86/kernel/cpu/Makefile index 47b56a7e99c..7fd54f09b01 100644 --- a/arch/x86/kernel/cpu/Makefile +++ b/arch/x86/kernel/cpu/Makefile @@ -36,12 +36,13 @@ obj-$(CONFIG_CPU_SUP_AMD) += perf_event_amd_iommu.o endif obj-$(CONFIG_CPU_SUP_INTEL) += perf_event_p6.o perf_event_knc.o perf_event_p4.o obj-$(CONFIG_CPU_SUP_INTEL) += perf_event_intel_lbr.o perf_event_intel_ds.o perf_event_intel.o -obj-$(CONFIG_CPU_SUP_INTEL) += perf_event_intel_uncore.o +obj-$(CONFIG_CPU_SUP_INTEL) += perf_event_intel_uncore.o perf_event_intel_rapl.o endif obj-$(CONFIG_X86_MCE) += mcheck/ obj-$(CONFIG_MTRR) += mtrr/ +obj-$(CONFIG_MICROCODE) += microcode/ obj-$(CONFIG_X86_LOCAL_APIC) += perfctr-watchdog.o perf_event_amd_ibs.o diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c index 903a264af98..ce8b8ff0e0e 100644 --- a/arch/x86/kernel/cpu/amd.c +++ b/arch/x86/kernel/cpu/amd.c @@ -1,5 +1,4 @@ #include <linux/export.h> -#include <linux/init.h> #include <linux/bitops.h> #include <linux/elf.h> #include <linux/mm.h> @@ -219,7 +218,7 @@ static void amd_k7_smp_check(struct cpuinfo_x86 *c) */ WARN_ONCE(1, "WARNING: This combination of AMD" " processors is not suitable for SMP.\n"); - add_taint(TAINT_UNSAFE_SMP, LOCKDEP_NOW_UNRELIABLE); + add_taint(TAINT_CPU_OUT_OF_SPEC, LOCKDEP_NOW_UNRELIABLE); } static void init_amd_k7(struct cpuinfo_x86 *c) @@ -234,9 +233,7 @@ static void init_amd_k7(struct cpuinfo_x86 *c) if (c->x86_model >= 6 && c->x86_model <= 10) { if (!cpu_has(c, X86_FEATURE_XMM)) { printk(KERN_INFO "Enabling disabled K7/SSE Support.\n"); - rdmsr(MSR_K7_HWCR, l, h); - l &= ~0x00008000; - wrmsr(MSR_K7_HWCR, l, h); + msr_clear_bit(MSR_K7_HWCR, 15); set_cpu_cap(c, X86_FEATURE_XMM); } } @@ -339,7 +336,7 @@ static void amd_get_topology(struct cpuinfo_x86 *c) #endif /* - * On a AMD dual core setup the lower bits of the APIC id distingush the cores. + * On a AMD dual core setup the lower bits of the APIC id distinguish the cores. * Assumes number of cores is a power of two. */ static void amd_detect_cmp(struct cpuinfo_x86 *c) @@ -487,7 +484,7 @@ static void early_init_amd(struct cpuinfo_x86 *c) set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC); if (!check_tsc_unstable()) - sched_clock_stable = 1; + set_sched_clock_stable(); } #ifdef CONFIG_X86_64 @@ -508,6 +505,10 @@ static void early_init_amd(struct cpuinfo_x86 *c) set_cpu_cap(c, X86_FEATURE_EXTD_APICID); } #endif + + /* F16h erratum 793, CVE-2013-6885 */ + if (c->x86 == 0x16 && c->x86_model <= 0xf) + msr_set_bit(MSR_AMD64_LS_CFG, 15); } static const int amd_erratum_383[]; @@ -527,11 +528,8 @@ static void init_amd(struct cpuinfo_x86 *c) * Errata 63 for SH-B3 steppings * Errata 122 for all steppings (F+ have it disabled by default) */ - if (c->x86 == 0xf) { - rdmsrl(MSR_K7_HWCR, value); - value |= 1 << 6; - wrmsrl(MSR_K7_HWCR, value); - } + if (c->x86 == 0xf) + msr_set_bit(MSR_K7_HWCR, 6); #endif early_init_amd(c); @@ -614,14 +612,11 @@ static void init_amd(struct cpuinfo_x86 *c) (c->x86_model >= 0x10) && (c->x86_model <= 0x1f) && !cpu_has(c, X86_FEATURE_TOPOEXT)) { - if (!rdmsrl_safe(0xc0011005, &value)) { - value |= 1ULL << 54; - wrmsrl_safe(0xc0011005, value); + if (msr_set_bit(0xc0011005, 54) > 0) { rdmsrl(0xc0011005, value); - if (value & (1ULL << 54)) { + if (value & BIT_64(54)) { set_cpu_cap(c, X86_FEATURE_TOPOEXT); - printk(KERN_INFO FW_INFO "CPU: Re-enabling " - "disabled Topology Extensions Support\n"); + pr_info(FW_INFO "CPU: Re-enabling disabled Topology Extensions Support.\n"); } } } @@ -700,19 +695,12 @@ static void init_amd(struct cpuinfo_x86 *c) * Disable GART TLB Walk Errors on Fam10h. We do this here * because this is always needed when GART is enabled, even in a * kernel which has no MCE support built in. - * BIOS should disable GartTlbWlk Errors themself. If - * it doesn't do it here as suggested by the BKDG. + * BIOS should disable GartTlbWlk Errors already. If + * it doesn't, do it here as suggested by the BKDG. * * Fixes: https://bugzilla.kernel.org/show_bug.cgi?id=33012 */ - u64 mask; - int err; - - err = rdmsrl_safe(MSR_AMD64_MCx_MASK(4), &mask); - if (err == 0) { - mask |= (1 << 10); - wrmsrl_safe(MSR_AMD64_MCx_MASK(4), mask); - } + msr_set_bit(MSR_AMD64_MCx_MASK(4), 10); /* * On family 10h BIOS may not have properly enabled WC+ support, @@ -724,10 +712,7 @@ static void init_amd(struct cpuinfo_x86 *c) * NOTE: we want to use the _safe accessors so as not to #GP kvm * guests on older kvm hosts. */ - - rdmsrl_safe(MSR_AMD64_BU_CFG2, &value); - value &= ~(1ULL << 24); - wrmsrl_safe(MSR_AMD64_BU_CFG2, value); + msr_clear_bit(MSR_AMD64_BU_CFG2, 24); if (cpu_has_amd_erratum(c, amd_erratum_383)) set_cpu_bug(c, X86_BUG_AMD_TLB_MMATCH); @@ -758,10 +743,7 @@ static unsigned int amd_size_cache(struct cpuinfo_x86 *c, unsigned int size) static void cpu_set_tlb_flushall_shift(struct cpuinfo_x86 *c) { - tlb_flushall_shift = 5; - - if (c->x86 <= 0x11) - tlb_flushall_shift = 4; + tlb_flushall_shift = 6; } static void cpu_detect_tlb_amd(struct cpuinfo_x86 *c) @@ -790,14 +772,10 @@ static void cpu_detect_tlb_amd(struct cpuinfo_x86 *c) } /* Handle DTLB 2M and 4M sizes, fall back to L1 if L2 is disabled */ - if (!((eax >> 16) & mask)) { - u32 a, b, c, d; - - cpuid(0x80000005, &a, &b, &c, &d); - tlb_lld_2m[ENTRIES] = (a >> 16) & 0xff; - } else { + if (!((eax >> 16) & mask)) + tlb_lld_2m[ENTRIES] = (cpuid_eax(0x80000005) >> 16) & 0xff; + else tlb_lld_2m[ENTRIES] = (eax >> 16) & mask; - } /* a 4M entry uses two 2M entries */ tlb_lld_4m[ENTRIES] = tlb_lld_2m[ENTRIES] >> 1; @@ -823,8 +801,8 @@ static const struct cpu_dev amd_cpu_dev = { .c_vendor = "AMD", .c_ident = { "AuthenticAMD" }, #ifdef CONFIG_X86_32 - .c_models = { - { .vendor = X86_VENDOR_AMD, .family = 4, .model_names = + .legacy_models = { + { .family = 4, .model_names = { [3] = "486 DX/2", [7] = "486 DX/2-WB", @@ -835,7 +813,7 @@ static const struct cpu_dev amd_cpu_dev = { } }, }, - .c_size_cache = amd_size_cache, + .legacy_cache_size = amd_size_cache, #endif .c_early_init = early_init_amd, .c_detect_tlb = cpu_detect_tlb_amd, diff --git a/arch/x86/kernel/cpu/centaur.c b/arch/x86/kernel/cpu/centaur.c index fbf6c3bc240..d8fba5c15fb 100644 --- a/arch/x86/kernel/cpu/centaur.c +++ b/arch/x86/kernel/cpu/centaur.c @@ -1,6 +1,5 @@ #include <linux/bitops.h> #include <linux/kernel.h> -#include <linux/init.h> #include <asm/processor.h> #include <asm/e820.h> @@ -9,236 +8,6 @@ #include "cpu.h" -#ifdef CONFIG_X86_OOSTORE - -static u32 power2(u32 x) -{ - u32 s = 1; - - while (s <= x) - s <<= 1; - - return s >>= 1; -} - - -/* - * Set up an actual MCR - */ -static void centaur_mcr_insert(int reg, u32 base, u32 size, int key) -{ - u32 lo, hi; - - hi = base & ~0xFFF; - lo = ~(size-1); /* Size is a power of 2 so this makes a mask */ - lo &= ~0xFFF; /* Remove the ctrl value bits */ - lo |= key; /* Attribute we wish to set */ - wrmsr(reg+MSR_IDT_MCR0, lo, hi); - mtrr_centaur_report_mcr(reg, lo, hi); /* Tell the mtrr driver */ -} - -/* - * Figure what we can cover with MCR's - * - * Shortcut: We know you can't put 4Gig of RAM on a winchip - */ -static u32 ramtop(void) -{ - u32 clip = 0xFFFFFFFFUL; - u32 top = 0; - int i; - - for (i = 0; i < e820.nr_map; i++) { - unsigned long start, end; - - if (e820.map[i].addr > 0xFFFFFFFFUL) - continue; - /* - * Don't MCR over reserved space. Ignore the ISA hole - * we frob around that catastrophe already - */ - if (e820.map[i].type == E820_RESERVED) { - if (e820.map[i].addr >= 0x100000UL && - e820.map[i].addr < clip) - clip = e820.map[i].addr; - continue; - } - start = e820.map[i].addr; - end = e820.map[i].addr + e820.map[i].size; - if (start >= end) - continue; - if (end > top) - top = end; - } - /* - * Everything below 'top' should be RAM except for the ISA hole. - * Because of the limited MCR's we want to map NV/ACPI into our - * MCR range for gunk in RAM - * - * Clip might cause us to MCR insufficient RAM but that is an - * acceptable failure mode and should only bite obscure boxes with - * a VESA hole at 15Mb - * - * The second case Clip sometimes kicks in is when the EBDA is marked - * as reserved. Again we fail safe with reasonable results - */ - if (top > clip) - top = clip; - - return top; -} - -/* - * Compute a set of MCR's to give maximum coverage - */ -static int centaur_mcr_compute(int nr, int key) -{ - u32 mem = ramtop(); - u32 root = power2(mem); - u32 base = root; - u32 top = root; - u32 floor = 0; - int ct = 0; - - while (ct < nr) { - u32 fspace = 0; - u32 high; - u32 low; - - /* - * Find the largest block we will fill going upwards - */ - high = power2(mem-top); - - /* - * Find the largest block we will fill going downwards - */ - low = base/2; - - /* - * Don't fill below 1Mb going downwards as there - * is an ISA hole in the way. - */ - if (base <= 1024*1024) - low = 0; - - /* - * See how much space we could cover by filling below - * the ISA hole - */ - - if (floor == 0) - fspace = 512*1024; - else if (floor == 512*1024) - fspace = 128*1024; - - /* And forget ROM space */ - - /* - * Now install the largest coverage we get - */ - if (fspace > high && fspace > low) { - centaur_mcr_insert(ct, floor, fspace, key); - floor += fspace; - } else if (high > low) { - centaur_mcr_insert(ct, top, high, key); - top += high; - } else if (low > 0) { - base -= low; - centaur_mcr_insert(ct, base, low, key); - } else - break; - ct++; - } - /* - * We loaded ct values. We now need to set the mask. The caller - * must do this bit. - */ - return ct; -} - -static void centaur_create_optimal_mcr(void) -{ - int used; - int i; - - /* - * Allocate up to 6 mcrs to mark as much of ram as possible - * as write combining and weak write ordered. - * - * To experiment with: Linux never uses stack operations for - * mmio spaces so we could globally enable stack operation wc - * - * Load the registers with type 31 - full write combining, all - * writes weakly ordered. - */ - used = centaur_mcr_compute(6, 31); - - /* - * Wipe unused MCRs - */ - for (i = used; i < 8; i++) - wrmsr(MSR_IDT_MCR0+i, 0, 0); -} - -static void winchip2_create_optimal_mcr(void) -{ - u32 lo, hi; - int used; - int i; - - /* - * Allocate up to 6 mcrs to mark as much of ram as possible - * as write combining, weak store ordered. - * - * Load the registers with type 25 - * 8 - weak write ordering - * 16 - weak read ordering - * 1 - write combining - */ - used = centaur_mcr_compute(6, 25); - - /* - * Mark the registers we are using. - */ - rdmsr(MSR_IDT_MCR_CTRL, lo, hi); - for (i = 0; i < used; i++) - lo |= 1<<(9+i); - wrmsr(MSR_IDT_MCR_CTRL, lo, hi); - - /* - * Wipe unused MCRs - */ - - for (i = used; i < 8; i++) - wrmsr(MSR_IDT_MCR0+i, 0, 0); -} - -/* - * Handle the MCR key on the Winchip 2. - */ -static void winchip2_unprotect_mcr(void) -{ - u32 lo, hi; - u32 key; - - rdmsr(MSR_IDT_MCR_CTRL, lo, hi); - lo &= ~0x1C0; /* blank bits 8-6 */ - key = (lo>>17) & 7; - lo |= key<<6; /* replace with unlock key */ - wrmsr(MSR_IDT_MCR_CTRL, lo, hi); -} - -static void winchip2_protect_mcr(void) -{ - u32 lo, hi; - - rdmsr(MSR_IDT_MCR_CTRL, lo, hi); - lo &= ~0x1C0; /* blank bits 8-6 */ - wrmsr(MSR_IDT_MCR_CTRL, lo, hi); -} -#endif /* CONFIG_X86_OOSTORE */ - #define ACE_PRESENT (1 << 6) #define ACE_ENABLED (1 << 7) #define ACE_FCR (1 << 28) /* MSR_VIA_FCR */ @@ -363,20 +132,6 @@ static void init_centaur(struct cpuinfo_x86 *c) fcr_clr = DPDC; printk(KERN_NOTICE "Disabling bugged TSC.\n"); clear_cpu_cap(c, X86_FEATURE_TSC); -#ifdef CONFIG_X86_OOSTORE - centaur_create_optimal_mcr(); - /* - * Enable: - * write combining on non-stack, non-string - * write combining on string, all types - * weak write ordering - * - * The C6 original lacks weak read order - * - * Note 0x120 is write only on Winchip 1 - */ - wrmsr(MSR_IDT_MCR_CTRL, 0x01F0001F, 0); -#endif break; case 8: switch (c->x86_mask) { @@ -393,40 +148,12 @@ static void init_centaur(struct cpuinfo_x86 *c) fcr_set = ECX8|DSMC|DTLOCK|EMMX|EBRPRED|ERETSTK| E2MMX|EAMD3D; fcr_clr = DPDC; -#ifdef CONFIG_X86_OOSTORE - winchip2_unprotect_mcr(); - winchip2_create_optimal_mcr(); - rdmsr(MSR_IDT_MCR_CTRL, lo, hi); - /* - * Enable: - * write combining on non-stack, non-string - * write combining on string, all types - * weak write ordering - */ - lo |= 31; - wrmsr(MSR_IDT_MCR_CTRL, lo, hi); - winchip2_protect_mcr(); -#endif break; case 9: name = "3"; fcr_set = ECX8|DSMC|DTLOCK|EMMX|EBRPRED|ERETSTK| E2MMX|EAMD3D; fcr_clr = DPDC; -#ifdef CONFIG_X86_OOSTORE - winchip2_unprotect_mcr(); - winchip2_create_optimal_mcr(); - rdmsr(MSR_IDT_MCR_CTRL, lo, hi); - /* - * Enable: - * write combining on non-stack, non-string - * write combining on string, all types - * weak write ordering - */ - lo |= 31; - wrmsr(MSR_IDT_MCR_CTRL, lo, hi); - winchip2_protect_mcr(); -#endif break; default: name = "??"; @@ -468,10 +195,10 @@ static void init_centaur(struct cpuinfo_x86 *c) #endif } +#ifdef CONFIG_X86_32 static unsigned int centaur_size_cache(struct cpuinfo_x86 *c, unsigned int size) { -#ifdef CONFIG_X86_32 /* VIA C3 CPUs (670-68F) need further shifting. */ if ((c->x86 == 6) && ((c->x86_model == 7) || (c->x86_model == 8))) size >>= 8; @@ -484,16 +211,18 @@ centaur_size_cache(struct cpuinfo_x86 *c, unsigned int size) if ((c->x86 == 6) && (c->x86_model == 9) && (c->x86_mask == 1) && (size == 65)) size -= 1; -#endif return size; } +#endif static const struct cpu_dev centaur_cpu_dev = { .c_vendor = "Centaur", .c_ident = { "CentaurHauls" }, .c_early_init = early_init_centaur, .c_init = init_centaur, - .c_size_cache = centaur_size_cache, +#ifdef CONFIG_X86_32 + .legacy_cache_size = centaur_size_cache, +#endif .c_x86_vendor = X86_VENDOR_CENTAUR, }; diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index 2793d1f095a..ef1b93f18ed 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -8,6 +8,7 @@ #include <linux/delay.h> #include <linux/sched.h> #include <linux/init.h> +#include <linux/kprobes.h> #include <linux/kgdb.h> #include <linux/smp.h> #include <linux/io.h> @@ -20,6 +21,7 @@ #include <asm/processor.h> #include <asm/debugreg.h> #include <asm/sections.h> +#include <asm/vsyscall.h> #include <linux/topology.h> #include <linux/cpumask.h> #include <asm/pgtable.h> @@ -284,8 +286,13 @@ static __always_inline void setup_smap(struct cpuinfo_x86 *c) raw_local_save_flags(eflags); BUG_ON(eflags & X86_EFLAGS_AC); - if (cpu_has(c, X86_FEATURE_SMAP)) + if (cpu_has(c, X86_FEATURE_SMAP)) { +#ifdef CONFIG_X86_SMAP set_in_cr4(X86_CR4_SMAP); +#else + clear_in_cr4(X86_CR4_SMAP); +#endif + } } /* @@ -346,7 +353,8 @@ static void filter_cpuid_features(struct cpuinfo_x86 *c, bool warn) /* Look up CPU names by table lookup. */ static const char *table_lookup_model(struct cpuinfo_x86 *c) { - const struct cpu_model_info *info; +#ifdef CONFIG_X86_32 + const struct legacy_cpu_model_info *info; if (c->x86_model >= 16) return NULL; /* Range check */ @@ -354,13 +362,14 @@ static const char *table_lookup_model(struct cpuinfo_x86 *c) if (!this_cpu) return NULL; - info = this_cpu->c_models; + info = this_cpu->legacy_models; - while (info && info->family) { + while (info->family) { if (info->family == c->x86) return info->model_names[c->x86_model]; info++; } +#endif return NULL; /* Not found */ } @@ -450,8 +459,8 @@ void cpu_detect_cache_sizes(struct cpuinfo_x86 *c) c->x86_tlbsize += ((ebx >> 16) & 0xfff) + (ebx & 0xfff); #else /* do processor-specific cache resizing */ - if (this_cpu->c_size_cache) - l2size = this_cpu->c_size_cache(c, l2size); + if (this_cpu->legacy_cache_size) + l2size = this_cpu->legacy_cache_size(c, l2size); /* Allow user to override all this if necessary. */ if (cachesize_override != -1) @@ -470,6 +479,7 @@ u16 __read_mostly tlb_lli_4m[NR_INFO]; u16 __read_mostly tlb_lld_4k[NR_INFO]; u16 __read_mostly tlb_lld_2m[NR_INFO]; u16 __read_mostly tlb_lld_4m[NR_INFO]; +u16 __read_mostly tlb_lld_1g[NR_INFO]; /* * tlb_flushall_shift shows the balance point in replacing cr3 write @@ -484,13 +494,13 @@ void cpu_detect_tlb(struct cpuinfo_x86 *c) if (this_cpu->c_detect_tlb) this_cpu->c_detect_tlb(c); - printk(KERN_INFO "Last level iTLB entries: 4KB %d, 2MB %d, 4MB %d\n" \ - "Last level dTLB entries: 4KB %d, 2MB %d, 4MB %d\n" \ + printk(KERN_INFO "Last level iTLB entries: 4KB %d, 2MB %d, 4MB %d\n" + "Last level dTLB entries: 4KB %d, 2MB %d, 4MB %d, 1GB %d\n" "tlb_flushall_shift: %d\n", tlb_lli_4k[ENTRIES], tlb_lli_2m[ENTRIES], tlb_lli_4m[ENTRIES], tlb_lld_4k[ENTRIES], tlb_lld_2m[ENTRIES], tlb_lld_4m[ENTRIES], - tlb_flushall_shift); + tlb_lld_1g[ENTRIES], tlb_flushall_shift); } void detect_ht(struct cpuinfo_x86 *c) @@ -945,6 +955,38 @@ static void vgetcpu_set_mode(void) else vgetcpu_mode = VGETCPU_LSL; } + +/* May not be __init: called during resume */ +static void syscall32_cpu_init(void) +{ + /* Load these always in case some future AMD CPU supports + SYSENTER from compat mode too. */ + wrmsrl_safe(MSR_IA32_SYSENTER_CS, (u64)__KERNEL_CS); + wrmsrl_safe(MSR_IA32_SYSENTER_ESP, 0ULL); + wrmsrl_safe(MSR_IA32_SYSENTER_EIP, (u64)ia32_sysenter_target); + + wrmsrl(MSR_CSTAR, ia32_cstar_target); +} +#endif + +#ifdef CONFIG_X86_32 +void enable_sep_cpu(void) +{ + int cpu = get_cpu(); + struct tss_struct *tss = &per_cpu(init_tss, cpu); + + if (!boot_cpu_has(X86_FEATURE_SEP)) { + put_cpu(); + return; + } + + tss->x86_tss.ss1 = __KERNEL_CS; + tss->x86_tss.sp1 = sizeof(struct tss_struct) + (unsigned long) tss; + wrmsr(MSR_IA32_SYSENTER_CS, __KERNEL_CS, 0); + wrmsr(MSR_IA32_SYSENTER_ESP, tss->x86_tss.sp1, 0); + wrmsr(MSR_IA32_SYSENTER_EIP, (unsigned long) ia32_sysenter_target, 0); + put_cpu(); +} #endif void __init identify_boot_cpu(void) @@ -1017,7 +1059,8 @@ __setup("show_msr=", setup_show_msr); static __init int setup_noclflush(char *arg) { - setup_clear_cpu_cap(X86_FEATURE_CLFLSH); + setup_clear_cpu_cap(X86_FEATURE_CLFLUSH); + setup_clear_cpu_cap(X86_FEATURE_CLFLUSHOPT); return 1; } __setup("noclflush", setup_noclflush); @@ -1070,6 +1113,10 @@ static __init int setup_disablecpuid(char *arg) } __setup("clearcpuid=", setup_disablecpuid); +DEFINE_PER_CPU(unsigned long, kernel_stack) = + (unsigned long)&init_thread_union - KERNEL_STACK_OFFSET + THREAD_SIZE; +EXPORT_PER_CPU_SYMBOL(kernel_stack); + #ifdef CONFIG_X86_64 struct desc_ptr idt_descr = { NR_VECTORS * 16 - 1, (unsigned long) idt_table }; struct desc_ptr debug_idt_descr = { NR_VECTORS * 16 - 1, @@ -1086,15 +1133,14 @@ DEFINE_PER_CPU(struct task_struct *, current_task) ____cacheline_aligned = &init_task; EXPORT_PER_CPU_SYMBOL(current_task); -DEFINE_PER_CPU(unsigned long, kernel_stack) = - (unsigned long)&init_thread_union - KERNEL_STACK_OFFSET + THREAD_SIZE; -EXPORT_PER_CPU_SYMBOL(kernel_stack); - DEFINE_PER_CPU(char *, irq_stack_ptr) = init_per_cpu_var(irq_stack_union.irq_stack) + IRQ_STACK_SIZE - 64; DEFINE_PER_CPU(unsigned int, irq_count) __visible = -1; +DEFINE_PER_CPU(int, __preempt_count) = INIT_PREEMPT_COUNT; +EXPORT_PER_CPU_SYMBOL(__preempt_count); + DEFINE_PER_CPU(struct task_struct *, fpu_owner_task); /* @@ -1148,6 +1194,7 @@ int is_debug_stack(unsigned long addr) (addr <= __get_cpu_var(debug_stack_addr) && addr > (__get_cpu_var(debug_stack_addr) - DEBUG_STKSZ)); } +NOKPROBE_SYMBOL(is_debug_stack); DEFINE_PER_CPU(u32, debug_idt_ctr); @@ -1156,6 +1203,7 @@ void debug_stack_set_zero(void) this_cpu_inc(debug_idt_ctr); load_current_idt(); } +NOKPROBE_SYMBOL(debug_stack_set_zero); void debug_stack_reset(void) { @@ -1164,11 +1212,14 @@ void debug_stack_reset(void) if (this_cpu_dec_return(debug_idt_ctr) == 0) load_current_idt(); } +NOKPROBE_SYMBOL(debug_stack_reset); #else /* CONFIG_X86_64 */ DEFINE_PER_CPU(struct task_struct *, current_task) = &init_task; EXPORT_PER_CPU_SYMBOL(current_task); +DEFINE_PER_CPU(int, __preempt_count) = INIT_PREEMPT_COUNT; +EXPORT_PER_CPU_SYMBOL(__preempt_count); DEFINE_PER_CPU(struct task_struct *, fpu_owner_task); #ifdef CONFIG_CC_STACKPROTECTOR diff --git a/arch/x86/kernel/cpu/cpu.h b/arch/x86/kernel/cpu/cpu.h index 4041c24ae7d..c37dc37e831 100644 --- a/arch/x86/kernel/cpu/cpu.h +++ b/arch/x86/kernel/cpu/cpu.h @@ -1,12 +1,6 @@ #ifndef ARCH_X86_CPU_H #define ARCH_X86_CPU_H -struct cpu_model_info { - int vendor; - int family; - const char *model_names[16]; -}; - /* attempt to consolidate cpu attributes */ struct cpu_dev { const char *c_vendor; @@ -14,15 +8,23 @@ struct cpu_dev { /* some have two possibilities for cpuid string */ const char *c_ident[2]; - struct cpu_model_info c_models[4]; - void (*c_early_init)(struct cpuinfo_x86 *); void (*c_bsp_init)(struct cpuinfo_x86 *); void (*c_init)(struct cpuinfo_x86 *); void (*c_identify)(struct cpuinfo_x86 *); void (*c_detect_tlb)(struct cpuinfo_x86 *); - unsigned int (*c_size_cache)(struct cpuinfo_x86 *, unsigned int); int c_x86_vendor; +#ifdef CONFIG_X86_32 + /* Optional vendor specific routine to obtain the cache size. */ + unsigned int (*legacy_cache_size)(struct cpuinfo_x86 *, + unsigned int); + + /* Family/stepping-based lookup table for model names. */ + struct legacy_cpu_model_info { + int family; + const char *model_names[16]; + } legacy_models[5]; +#endif }; struct _tlb_table { diff --git a/arch/x86/kernel/cpu/cyrix.c b/arch/x86/kernel/cpu/cyrix.c index d0969c75ab5..aaf152e7963 100644 --- a/arch/x86/kernel/cpu/cyrix.c +++ b/arch/x86/kernel/cpu/cyrix.c @@ -1,4 +1,3 @@ -#include <linux/init.h> #include <linux/bitops.h> #include <linux/delay.h> #include <linux/pci.h> diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c index ec7299566f7..f9e4fdd3b87 100644 --- a/arch/x86/kernel/cpu/intel.c +++ b/arch/x86/kernel/cpu/intel.c @@ -1,4 +1,3 @@ -#include <linux/init.h> #include <linux/kernel.h> #include <linux/string.h> @@ -32,11 +31,8 @@ static void early_init_intel(struct cpuinfo_x86 *c) /* Unmask CPUID levels if masked: */ if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) { - rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable); - - if (misc_enable & MSR_IA32_MISC_ENABLE_LIMIT_CPUID) { - misc_enable &= ~MSR_IA32_MISC_ENABLE_LIMIT_CPUID; - wrmsrl(MSR_IA32_MISC_ENABLE, misc_enable); + if (msr_clear_bit(MSR_IA32_MISC_ENABLE, + MSR_IA32_MISC_ENABLE_LIMIT_CPUID_BIT) > 0) { c->cpuid_level = cpuid_eax(0); get_cpu_cap(c); } @@ -93,7 +89,7 @@ static void early_init_intel(struct cpuinfo_x86 *c) set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC); set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC); if (!check_tsc_unstable()) - sched_clock_stable = 1; + set_sched_clock_stable(); } /* Penwell and Cloverview have the TSC which doesn't sleep on S3 */ @@ -130,16 +126,10 @@ static void early_init_intel(struct cpuinfo_x86 *c) * Ingo Molnar reported a Pentium D (model 6) and a Xeon * (model 2) with the same problem. */ - if (c->x86 == 15) { - rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable); - - if (misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING) { - printk(KERN_INFO "kmemcheck: Disabling fast string operations\n"); - - misc_enable &= ~MSR_IA32_MISC_ENABLE_FAST_STRING; - wrmsrl(MSR_IA32_MISC_ENABLE, misc_enable); - } - } + if (c->x86 == 15) + if (msr_clear_bit(MSR_IA32_MISC_ENABLE, + MSR_IA32_MISC_ENABLE_FAST_STRING_BIT) > 0) + pr_info("kmemcheck: Disabling fast string operations\n"); #endif /* @@ -196,10 +186,16 @@ static void intel_smp_check(struct cpuinfo_x86 *c) } } -static void intel_workarounds(struct cpuinfo_x86 *c) +static int forcepae; +static int __init forcepae_setup(char *__unused) { - unsigned long lo, hi; + forcepae = 1; + return 1; +} +__setup("forcepae", forcepae_setup); +static void intel_workarounds(struct cpuinfo_x86 *c) +{ #ifdef CONFIG_X86_F00F_BUG /* * All current models of Pentium and Pentium with MMX technology CPUs @@ -226,16 +222,26 @@ static void intel_workarounds(struct cpuinfo_x86 *c) clear_cpu_cap(c, X86_FEATURE_SEP); /* + * PAE CPUID issue: many Pentium M report no PAE but may have a + * functionally usable PAE implementation. + * Forcefully enable PAE if kernel parameter "forcepae" is present. + */ + if (forcepae) { + printk(KERN_WARNING "PAE forced!\n"); + set_cpu_cap(c, X86_FEATURE_PAE); + add_taint(TAINT_CPU_OUT_OF_SPEC, LOCKDEP_NOW_UNRELIABLE); + } + + /* * P4 Xeon errata 037 workaround. * Hardware prefetcher may cause stale data to be loaded into the cache. */ if ((c->x86 == 15) && (c->x86_model == 1) && (c->x86_mask == 1)) { - rdmsr(MSR_IA32_MISC_ENABLE, lo, hi); - if ((lo & MSR_IA32_MISC_ENABLE_PREFETCH_DISABLE) == 0) { - printk (KERN_INFO "CPU: C0 stepping P4 Xeon detected.\n"); - printk (KERN_INFO "CPU: Disabling hardware prefetching (Errata 037)\n"); - lo |= MSR_IA32_MISC_ENABLE_PREFETCH_DISABLE; - wrmsr(MSR_IA32_MISC_ENABLE, lo, hi); + if (msr_set_bit(MSR_IA32_MISC_ENABLE, + MSR_IA32_MISC_ENABLE_PREFETCH_DISABLE_BIT) + > 0) { + pr_info("CPU: C0 stepping P4 Xeon detected.\n"); + pr_info("CPU: Disabling hardware prefetching (Errata 037)\n"); } } @@ -268,10 +274,6 @@ static void intel_workarounds(struct cpuinfo_x86 *c) } #endif -#ifdef CONFIG_X86_NUMAQ - numaq_tsc_disable(); -#endif - intel_smp_check(c); } #else @@ -368,6 +370,17 @@ static void init_intel(struct cpuinfo_x86 *c) */ detect_extended_topology(c); + if (!cpu_has(c, X86_FEATURE_XTOPOLOGY)) { + /* + * let's use the legacy cpuid vector 0x1 and 0x4 for topology + * detection. + */ + c->x86_max_cores = intel_num_cpu_cores(c); +#ifdef CONFIG_X86_32 + detect_ht(c); +#endif + } + l2 = init_intel_cacheinfo(c); if (c->cpuid_level > 9) { unsigned eax = cpuid_eax(10); @@ -387,7 +400,8 @@ static void init_intel(struct cpuinfo_x86 *c) set_cpu_cap(c, X86_FEATURE_PEBS); } - if (c->x86 == 6 && c->x86_model == 29 && cpu_has_clflush) + if (c->x86 == 6 && cpu_has_clflush && + (c->x86_model == 29 || c->x86_model == 46 || c->x86_model == 47)) set_cpu_cap(c, X86_FEATURE_CLFLUSH_MONITOR); #ifdef CONFIG_X86_64 @@ -435,17 +449,6 @@ static void init_intel(struct cpuinfo_x86 *c) set_cpu_cap(c, X86_FEATURE_P3); #endif - if (!cpu_has(c, X86_FEATURE_XTOPOLOGY)) { - /* - * let's use the legacy cpuid vector 0x1 and 0x4 for topology - * detection. - */ - c->x86_max_cores = intel_num_cpu_cores(c); -#ifdef CONFIG_X86_32 - detect_ht(c); -#endif - } - /* Work around errata */ srat_detect_node(c); @@ -505,6 +508,7 @@ static unsigned int intel_size_cache(struct cpuinfo_x86 *c, unsigned int size) #define TLB_DATA0_2M_4M 0x23 #define STLB_4K 0x41 +#define STLB_4K_2M 0x42 static const struct _tlb_table intel_tlb_table[] = { { 0x01, TLB_INST_4K, 32, " TLB_INST 4 KByte pages, 4-way set associative" }, @@ -525,13 +529,20 @@ static const struct _tlb_table intel_tlb_table[] = { { 0x5b, TLB_DATA_4K_4M, 64, " TLB_DATA 4 KByte and 4 MByte pages" }, { 0x5c, TLB_DATA_4K_4M, 128, " TLB_DATA 4 KByte and 4 MByte pages" }, { 0x5d, TLB_DATA_4K_4M, 256, " TLB_DATA 4 KByte and 4 MByte pages" }, + { 0x61, TLB_INST_4K, 48, " TLB_INST 4 KByte pages, full associative" }, + { 0x63, TLB_DATA_1G, 4, " TLB_DATA 1 GByte pages, 4-way set associative" }, + { 0x76, TLB_INST_2M_4M, 8, " TLB_INST 2-MByte or 4-MByte pages, fully associative" }, { 0xb0, TLB_INST_4K, 128, " TLB_INST 4 KByte pages, 4-way set associative" }, { 0xb1, TLB_INST_2M_4M, 4, " TLB_INST 2M pages, 4-way, 8 entries or 4M pages, 4-way entries" }, { 0xb2, TLB_INST_4K, 64, " TLB_INST 4KByte pages, 4-way set associative" }, { 0xb3, TLB_DATA_4K, 128, " TLB_DATA 4 KByte pages, 4-way set associative" }, { 0xb4, TLB_DATA_4K, 256, " TLB_DATA 4 KByte pages, 4-way associative" }, + { 0xb5, TLB_INST_4K, 64, " TLB_INST 4 KByte pages, 8-way set ssociative" }, + { 0xb6, TLB_INST_4K, 128, " TLB_INST 4 KByte pages, 8-way set ssociative" }, { 0xba, TLB_DATA_4K, 64, " TLB_DATA 4 KByte pages, 4-way associative" }, { 0xc0, TLB_DATA_4K_4M, 8, " TLB_DATA 4 KByte and 4 MByte pages, 4-way associative" }, + { 0xc1, STLB_4K_2M, 1024, " STLB 4 KByte and 2 MByte pages, 8-way associative" }, + { 0xc2, TLB_DATA_2M_4M, 16, " DTLB 2 MByte/4MByte pages, 4-way associative" }, { 0xca, STLB_4K, 512, " STLB 4 KByte pages, 4-way associative" }, { 0x00, 0, 0 } }; @@ -557,6 +568,20 @@ static void intel_tlb_lookup(const unsigned char desc) if (tlb_lld_4k[ENTRIES] < intel_tlb_table[k].entries) tlb_lld_4k[ENTRIES] = intel_tlb_table[k].entries; break; + case STLB_4K_2M: + if (tlb_lli_4k[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_4k[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lld_4k[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_4k[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lli_2m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_2m[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lld_2m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_2m[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lli_4m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lli_4m[ENTRIES] = intel_tlb_table[k].entries; + if (tlb_lld_4m[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_4m[ENTRIES] = intel_tlb_table[k].entries; + break; case TLB_INST_ALL: if (tlb_lli_4k[ENTRIES] < intel_tlb_table[k].entries) tlb_lli_4k[ENTRIES] = intel_tlb_table[k].entries; @@ -602,6 +627,10 @@ static void intel_tlb_lookup(const unsigned char desc) if (tlb_lld_4m[ENTRIES] < intel_tlb_table[k].entries) tlb_lld_4m[ENTRIES] = intel_tlb_table[k].entries; break; + case TLB_DATA_1G: + if (tlb_lld_1g[ENTRIES] < intel_tlb_table[k].entries) + tlb_lld_1g[ENTRIES] = intel_tlb_table[k].entries; + break; } } @@ -614,21 +643,17 @@ static void intel_tlb_flushall_shift_set(struct cpuinfo_x86 *c) case 0x61d: /* six-core 45 nm xeon "Dunnington" */ tlb_flushall_shift = -1; break; + case 0x63a: /* Ivybridge */ + tlb_flushall_shift = 2; + break; case 0x61a: /* 45 nm nehalem, "Bloomfield" */ case 0x61e: /* 45 nm nehalem, "Lynnfield" */ case 0x625: /* 32 nm nehalem, "Clarkdale" */ case 0x62c: /* 32 nm nehalem, "Gulftown" */ case 0x62e: /* 45 nm nehalem-ex, "Beckton" */ case 0x62f: /* 32 nm Xeon E7 */ - tlb_flushall_shift = 6; - break; case 0x62a: /* SandyBridge */ case 0x62d: /* SandyBridge, "Romely-EP" */ - tlb_flushall_shift = 5; - break; - case 0x63a: /* Ivybridge */ - tlb_flushall_shift = 1; - break; default: tlb_flushall_shift = 6; } @@ -665,8 +690,8 @@ static const struct cpu_dev intel_cpu_dev = { .c_vendor = "Intel", .c_ident = { "GenuineIntel" }, #ifdef CONFIG_X86_32 - .c_models = { - { .vendor = X86_VENDOR_INTEL, .family = 4, .model_names = + .legacy_models = { + { .family = 4, .model_names = { [0] = "486 DX-25/33", [1] = "486 DX-50", @@ -679,7 +704,7 @@ static const struct cpu_dev intel_cpu_dev = { [9] = "486 DX/4-WB" } }, - { .vendor = X86_VENDOR_INTEL, .family = 5, .model_names = + { .family = 5, .model_names = { [0] = "Pentium 60/66 A-step", [1] = "Pentium 60/66", @@ -690,7 +715,7 @@ static const struct cpu_dev intel_cpu_dev = { [8] = "Mobile Pentium MMX" } }, - { .vendor = X86_VENDOR_INTEL, .family = 6, .model_names = + { .family = 6, .model_names = { [0] = "Pentium Pro A-step", [1] = "Pentium Pro", @@ -704,7 +729,7 @@ static const struct cpu_dev intel_cpu_dev = { [11] = "Pentium III (Tualatin)", } }, - { .vendor = X86_VENDOR_INTEL, .family = 15, .model_names = + { .family = 15, .model_names = { [0] = "Pentium 4 (Unknown)", [1] = "Pentium 4 (Willamette)", @@ -714,7 +739,7 @@ static const struct cpu_dev intel_cpu_dev = { } }, }, - .c_size_cache = intel_size_cache, + .legacy_cache_size = intel_size_cache, #endif .c_detect_tlb = intel_detect_tlb, .c_early_init = early_init_intel, diff --git a/arch/x86/kernel/cpu/intel_cacheinfo.c b/arch/x86/kernel/cpu/intel_cacheinfo.c index 1414c90feab..9c8f7394c61 100644 --- a/arch/x86/kernel/cpu/intel_cacheinfo.c +++ b/arch/x86/kernel/cpu/intel_cacheinfo.c @@ -1,5 +1,5 @@ /* - * Routines to indentify caches on Intel CPU. + * Routines to identify caches on Intel CPU. * * Changes: * Venkatesh Pallipadi : Adding cache identification through cpuid(4) @@ -730,6 +730,18 @@ unsigned int init_intel_cacheinfo(struct cpuinfo_x86 *c) #endif } +#ifdef CONFIG_X86_HT + /* + * If cpu_llc_id is not yet set, this means cpuid_level < 4 which in + * turns means that the only possibility is SMT (as indicated in + * cpuid1). Since cpuid2 doesn't specify shared caches, and we know + * that SMT shares all caches, we can unconditionally set cpu_llc_id to + * c->phys_proc_id. + */ + if (per_cpu(cpu_llc_id, cpu) == BAD_APICID) + per_cpu(cpu_llc_id, cpu) = c->phys_proc_id; +#endif + c->x86_cache_size = l3 ? l3 : (l2 ? l2 : (l1i+l1d)); return l2; @@ -1225,21 +1237,24 @@ static struct notifier_block cacheinfo_cpu_notifier = { static int __init cache_sysfs_init(void) { - int i; + int i, err = 0; if (num_cache_leaves == 0) return 0; + cpu_notifier_register_begin(); for_each_online_cpu(i) { - int err; struct device *dev = get_cpu_device(i); err = cache_add_dev(dev); if (err) - return err; + goto out; } - register_hotcpu_notifier(&cacheinfo_cpu_notifier); - return 0; + __register_hotcpu_notifier(&cacheinfo_cpu_notifier); + +out: + cpu_notifier_register_done(); + return err; } device_initcall(cache_sysfs_init); diff --git a/arch/x86/kernel/cpu/match.c b/arch/x86/kernel/cpu/match.c index 36565373af8..afa9f0d487e 100644 --- a/arch/x86/kernel/cpu/match.c +++ b/arch/x86/kernel/cpu/match.c @@ -47,45 +47,3 @@ const struct x86_cpu_id *x86_match_cpu(const struct x86_cpu_id *match) return NULL; } EXPORT_SYMBOL(x86_match_cpu); - -ssize_t arch_print_cpu_modalias(struct device *dev, - struct device_attribute *attr, - char *bufptr) -{ - int size = PAGE_SIZE; - int i, n; - char *buf = bufptr; - - n = snprintf(buf, size, "x86cpu:vendor:%04X:family:%04X:" - "model:%04X:feature:", - boot_cpu_data.x86_vendor, - boot_cpu_data.x86, - boot_cpu_data.x86_model); - size -= n; - buf += n; - size -= 1; - for (i = 0; i < NCAPINTS*32; i++) { - if (boot_cpu_has(i)) { - n = snprintf(buf, size, ",%04X", i); - if (n >= size) { - WARN(1, "x86 features overflow page\n"); - break; - } - size -= n; - buf += n; - } - } - *buf++ = '\n'; - return buf - bufptr; -} - -int arch_cpu_uevent(struct device *dev, struct kobj_uevent_env *env) -{ - char *buf = kzalloc(PAGE_SIZE, GFP_KERNEL); - if (buf) { - arch_print_cpu_modalias(NULL, NULL, buf); - add_uevent_var(env, "MODALIAS=%s", buf); - kfree(buf); - } - return 0; -} diff --git a/arch/x86/kernel/cpu/mcheck/mce-apei.c b/arch/x86/kernel/cpu/mcheck/mce-apei.c index cd8b166a173..a1aef953315 100644 --- a/arch/x86/kernel/cpu/mcheck/mce-apei.c +++ b/arch/x86/kernel/cpu/mcheck/mce-apei.c @@ -33,23 +33,28 @@ #include <linux/acpi.h> #include <linux/cper.h> #include <acpi/apei.h> +#include <acpi/ghes.h> #include <asm/mce.h> #include "mce-internal.h" -void apei_mce_report_mem_error(int corrected, struct cper_sec_mem_err *mem_err) +void apei_mce_report_mem_error(int severity, struct cper_sec_mem_err *mem_err) { struct mce m; - /* Only corrected MC is reported */ - if (!corrected || !(mem_err->validation_bits & - CPER_MEM_VALID_PHYSICAL_ADDRESS)) + if (!(mem_err->validation_bits & CPER_MEM_VALID_PA)) return; mce_setup(&m); m.bank = 1; - /* Fake a memory read corrected error with unknown channel */ + /* Fake a memory read error with unknown channel */ m.status = MCI_STATUS_VAL | MCI_STATUS_EN | MCI_STATUS_ADDRV | 0x9f; + + if (severity >= GHES_SEV_RECOVERABLE) + m.status |= MCI_STATUS_UC; + if (severity >= GHES_SEV_PANIC) + m.status |= MCI_STATUS_PCC; + m.addr = mem_err->physical_addr; mce_log(&m); mce_notify_irq(); diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c index b3218cdee95..9a79c8dbd8e 100644 --- a/arch/x86/kernel/cpu/mcheck/mce.c +++ b/arch/x86/kernel/cpu/mcheck/mce.c @@ -60,8 +60,6 @@ static DEFINE_MUTEX(mce_chrdev_read_mutex); #define SPINUNIT 100 /* 100ns */ -atomic_t mce_entry; - DEFINE_PER_CPU(unsigned, mce_exception_count); struct mce_bank *mce_banks __read_mostly; @@ -89,6 +87,9 @@ static DECLARE_WAIT_QUEUE_HEAD(mce_chrdev_wait); static DEFINE_PER_CPU(struct mce, mces_seen); static int cpu_missing; +/* CMCI storm detection filter */ +static DEFINE_PER_CPU(unsigned long, mce_polled_error); + /* * MCA banks polled by the period polling timer for corrected events. * With Intel CMCI, this only has MCA banks which do not support CMCI (if any). @@ -614,6 +615,7 @@ void machine_check_poll(enum mcp_flags flags, mce_banks_t *b) if (!(m.status & MCI_STATUS_VAL)) continue; + this_cpu_write(mce_polled_error, 1); /* * Uncorrected or signalled events are handled by the exception * handler when it is enabled, so don't process those here. @@ -700,8 +702,7 @@ static int mce_timed_out(u64 *t) if (!mca_cfg.monarch_timeout) goto out; if ((s64)*t < SPINUNIT) { - /* CHECKME: Make panic default for 1 too? */ - if (mca_cfg.tolerant < 1) + if (mca_cfg.tolerant <= 1) mce_panic("Timeout synchronizing machine check over CPUs", NULL, NULL); cpu_missing = 1; @@ -1037,8 +1038,6 @@ void do_machine_check(struct pt_regs *regs, long error_code) DECLARE_BITMAP(valid_banks, MAX_NR_BANKS); char *msg = "Unknown"; - atomic_inc(&mce_entry); - this_cpu_inc(mce_exception_count); if (!cfg->banks) @@ -1168,7 +1167,6 @@ void do_machine_check(struct pt_regs *regs, long error_code) mce_report_event(regs); mce_wrmsrl(MSR_IA32_MCG_STATUS, 0); out: - atomic_dec(&mce_entry); sync_core(); } EXPORT_SYMBOL_GPL(do_machine_check); @@ -1278,10 +1276,18 @@ static unsigned long mce_adjust_timer_default(unsigned long interval) static unsigned long (*mce_adjust_timer)(unsigned long interval) = mce_adjust_timer_default; +static int cmc_error_seen(void) +{ + unsigned long *v = &__get_cpu_var(mce_polled_error); + + return test_and_clear_bit(0, v); +} + static void mce_timer_fn(unsigned long data) { struct timer_list *t = &__get_cpu_var(mce_timer); unsigned long iv; + int notify; WARN_ON(smp_processor_id() != data); @@ -1296,7 +1302,9 @@ static void mce_timer_fn(unsigned long data) * polling interval, otherwise increase the polling interval. */ iv = __this_cpu_read(mce_next_interval); - if (mce_notify_irq()) { + notify = mce_notify_irq(); + notify |= cmc_error_seen(); + if (notify) { iv = max(iv / 2, (unsigned long) HZ/100); } else { iv = min(iv * 2, round_jiffies_relative(check_interval * HZ)); @@ -1638,15 +1646,15 @@ static void __mcheck_cpu_init_vendor(struct cpuinfo_x86 *c) static void mce_start_timer(unsigned int cpu, struct timer_list *t) { - unsigned long iv = mce_adjust_timer(check_interval * HZ); - - __this_cpu_write(mce_next_interval, iv); + unsigned long iv = check_interval * HZ; if (mca_cfg.ignore_ce || !iv) return; + per_cpu(mce_next_interval, cpu) = iv; + t->expires = round_jiffies(jiffies + iv); - add_timer_on(t, smp_processor_id()); + add_timer_on(t, cpu); } static void __mcheck_cpu_init_timer(void) @@ -2272,8 +2280,10 @@ static int mce_device_create(unsigned int cpu) dev->release = &mce_device_release; err = device_register(dev); - if (err) + if (err) { + put_device(dev); return err; + } for (i = 0; mce_device_attrs[i]; i++) { err = device_create_file(dev, mce_device_attrs[i]); @@ -2421,28 +2431,67 @@ static __init int mcheck_init_device(void) int err; int i = 0; - if (!mce_available(&boot_cpu_data)) - return -EIO; + if (!mce_available(&boot_cpu_data)) { + err = -EIO; + goto err_out; + } - zalloc_cpumask_var(&mce_device_initialized, GFP_KERNEL); + if (!zalloc_cpumask_var(&mce_device_initialized, GFP_KERNEL)) { + err = -ENOMEM; + goto err_out; + } mce_init_banks(); err = subsys_system_register(&mce_subsys, NULL); if (err) - return err; + goto err_out_mem; + cpu_notifier_register_begin(); for_each_online_cpu(i) { err = mce_device_create(i); - if (err) - return err; + if (err) { + /* + * Register notifier anyway (and do not unreg it) so + * that we don't leave undeleted timers, see notifier + * callback above. + */ + __register_hotcpu_notifier(&mce_cpu_notifier); + cpu_notifier_register_done(); + goto err_device_create; + } } + __register_hotcpu_notifier(&mce_cpu_notifier); + cpu_notifier_register_done(); + register_syscore_ops(&mce_syscore_ops); - register_hotcpu_notifier(&mce_cpu_notifier); /* register character device /dev/mcelog */ - misc_register(&mce_chrdev_device); + err = misc_register(&mce_chrdev_device); + if (err) + goto err_register; + + return 0; + +err_register: + unregister_syscore_ops(&mce_syscore_ops); + +err_device_create: + /* + * We didn't keep track of which devices were created above, but + * even if we had, the set of online cpus might have changed. + * Play safe and remove for every possible cpu, since + * mce_device_remove() will do the right thing. + */ + for_each_possible_cpu(i) + mce_device_remove(i); + +err_out_mem: + free_cpumask_var(mce_device_initialized); + +err_out: + pr_err("Unable to init device /dev/mcelog (rc: %d)\n", err); return err; } diff --git a/arch/x86/kernel/cpu/mcheck/mce_intel.c b/arch/x86/kernel/cpu/mcheck/mce_intel.c index 4cfe0458ca6..9a316b21df8 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_intel.c +++ b/arch/x86/kernel/cpu/mcheck/mce_intel.c @@ -6,10 +6,10 @@ */ #include <linux/gfp.h> -#include <linux/init.h> #include <linux/interrupt.h> #include <linux/percpu.h> #include <linux/sched.h> +#include <linux/cpumask.h> #include <asm/apic.h> #include <asm/processor.h> #include <asm/msr.h> @@ -42,7 +42,7 @@ static DEFINE_PER_CPU(mce_banks_t, mce_banks_owned); * cmci_discover_lock protects against parallel discovery attempts * which could race against each other. */ -static DEFINE_RAW_SPINLOCK(cmci_discover_lock); +static DEFINE_SPINLOCK(cmci_discover_lock); #define CMCI_THRESHOLD 1 #define CMCI_POLL_INTERVAL (30 * HZ) @@ -138,6 +138,22 @@ unsigned long mce_intel_adjust_timer(unsigned long interval) } } +static void cmci_storm_disable_banks(void) +{ + unsigned long flags, *owned; + int bank; + u64 val; + + spin_lock_irqsave(&cmci_discover_lock, flags); + owned = __get_cpu_var(mce_banks_owned); + for_each_set_bit(bank, owned, MAX_NR_BANKS) { + rdmsrl(MSR_IA32_MCx_CTL2(bank), val); + val &= ~MCI_CTL2_CMCI_EN; + wrmsrl(MSR_IA32_MCx_CTL2(bank), val); + } + spin_unlock_irqrestore(&cmci_discover_lock, flags); +} + static bool cmci_storm_detect(void) { unsigned int cnt = __this_cpu_read(cmci_storm_cnt); @@ -159,7 +175,7 @@ static bool cmci_storm_detect(void) if (cnt <= CMCI_STORM_THRESHOLD) return false; - cmci_clear(); + cmci_storm_disable_banks(); __this_cpu_write(cmci_storm_state, CMCI_STORM_ACTIVE); r = atomic_add_return(1, &cmci_storm_on_cpus); mce_timer_kick(CMCI_POLL_INTERVAL); @@ -195,7 +211,7 @@ static void cmci_discover(int banks) int i; int bios_wrong_thresh = 0; - raw_spin_lock_irqsave(&cmci_discover_lock, flags); + spin_lock_irqsave(&cmci_discover_lock, flags); for (i = 0; i < banks; i++) { u64 val; int bios_zero_thresh = 0; @@ -250,7 +266,7 @@ static void cmci_discover(int banks) WARN_ON(!test_bit(i, __get_cpu_var(mce_poll_banks))); } } - raw_spin_unlock_irqrestore(&cmci_discover_lock, flags); + spin_unlock_irqrestore(&cmci_discover_lock, flags); if (mca_cfg.bios_cmci_threshold && bios_wrong_thresh) { pr_info_once( "bios_cmci_threshold: Some banks do not have valid thresholds set\n"); @@ -300,10 +316,10 @@ void cmci_clear(void) if (!cmci_supported(&banks)) return; - raw_spin_lock_irqsave(&cmci_discover_lock, flags); + spin_lock_irqsave(&cmci_discover_lock, flags); for (i = 0; i < banks; i++) __cmci_disable_bank(i); - raw_spin_unlock_irqrestore(&cmci_discover_lock, flags); + spin_unlock_irqrestore(&cmci_discover_lock, flags); } static void cmci_rediscover_work_func(void *arg) @@ -344,9 +360,9 @@ void cmci_disable_bank(int bank) if (!cmci_supported(&banks)) return; - raw_spin_lock_irqsave(&cmci_discover_lock, flags); + spin_lock_irqsave(&cmci_discover_lock, flags); __cmci_disable_bank(bank); - raw_spin_unlock_irqrestore(&cmci_discover_lock, flags); + spin_unlock_irqrestore(&cmci_discover_lock, flags); } static void intel_init_cmci(void) diff --git a/arch/x86/kernel/cpu/mcheck/p5.c b/arch/x86/kernel/cpu/mcheck/p5.c index 1c044b1ccc5..a3042989398 100644 --- a/arch/x86/kernel/cpu/mcheck/p5.c +++ b/arch/x86/kernel/cpu/mcheck/p5.c @@ -5,7 +5,6 @@ #include <linux/interrupt.h> #include <linux/kernel.h> #include <linux/types.h> -#include <linux/init.h> #include <linux/smp.h> #include <asm/processor.h> diff --git a/arch/x86/kernel/cpu/mcheck/therm_throt.c b/arch/x86/kernel/cpu/mcheck/therm_throt.c index 3eec7de76ef..36a1bb6d1ee 100644 --- a/arch/x86/kernel/cpu/mcheck/therm_throt.c +++ b/arch/x86/kernel/cpu/mcheck/therm_throt.c @@ -271,9 +271,6 @@ static void thermal_throttle_remove_dev(struct device *dev) sysfs_remove_group(&dev->kobj, &thermal_attr_group); } -/* Mutex protecting device creation against CPU hotplug: */ -static DEFINE_MUTEX(therm_cpu_lock); - /* Get notified when a cpu comes on/off. Be hotplug friendly. */ static int thermal_throttle_cpu_callback(struct notifier_block *nfb, @@ -289,18 +286,14 @@ thermal_throttle_cpu_callback(struct notifier_block *nfb, switch (action) { case CPU_UP_PREPARE: case CPU_UP_PREPARE_FROZEN: - mutex_lock(&therm_cpu_lock); err = thermal_throttle_add_dev(dev, cpu); - mutex_unlock(&therm_cpu_lock); WARN_ON(err); break; case CPU_UP_CANCELED: case CPU_UP_CANCELED_FROZEN: case CPU_DEAD: case CPU_DEAD_FROZEN: - mutex_lock(&therm_cpu_lock); thermal_throttle_remove_dev(dev); - mutex_unlock(&therm_cpu_lock); break; } return notifier_from_errno(err); @@ -319,19 +312,16 @@ static __init int thermal_throttle_init_device(void) if (!atomic_read(&therm_throt_en)) return 0; - register_hotcpu_notifier(&thermal_throttle_cpu_notifier); + cpu_notifier_register_begin(); -#ifdef CONFIG_HOTPLUG_CPU - mutex_lock(&therm_cpu_lock); -#endif /* connect live CPUs to sysfs */ for_each_online_cpu(cpu) { err = thermal_throttle_add_dev(get_cpu_device(cpu), cpu); WARN_ON(err); } -#ifdef CONFIG_HOTPLUG_CPU - mutex_unlock(&therm_cpu_lock); -#endif + + __register_hotcpu_notifier(&thermal_throttle_cpu_notifier); + cpu_notifier_register_done(); return 0; } @@ -439,14 +429,14 @@ static inline void __smp_thermal_interrupt(void) smp_thermal_vector(); } -asmlinkage void smp_thermal_interrupt(struct pt_regs *regs) +asmlinkage __visible void smp_thermal_interrupt(struct pt_regs *regs) { entering_irq(); __smp_thermal_interrupt(); exiting_ack_irq(); } -asmlinkage void smp_trace_thermal_interrupt(struct pt_regs *regs) +asmlinkage __visible void smp_trace_thermal_interrupt(struct pt_regs *regs) { entering_irq(); trace_thermal_apic_entry(THERMAL_APIC_VECTOR); diff --git a/arch/x86/kernel/cpu/mcheck/threshold.c b/arch/x86/kernel/cpu/mcheck/threshold.c index fe6b1c86645..7245980186e 100644 --- a/arch/x86/kernel/cpu/mcheck/threshold.c +++ b/arch/x86/kernel/cpu/mcheck/threshold.c @@ -24,14 +24,14 @@ static inline void __smp_threshold_interrupt(void) mce_threshold_vector(); } -asmlinkage void smp_threshold_interrupt(void) +asmlinkage __visible void smp_threshold_interrupt(void) { entering_irq(); __smp_threshold_interrupt(); exiting_ack_irq(); } -asmlinkage void smp_trace_threshold_interrupt(void) +asmlinkage __visible void smp_trace_threshold_interrupt(void) { entering_irq(); trace_threshold_apic_entry(THRESHOLD_APIC_VECTOR); diff --git a/arch/x86/kernel/cpu/mcheck/winchip.c b/arch/x86/kernel/cpu/mcheck/winchip.c index e9a701aecaa..7dc5564d0cd 100644 --- a/arch/x86/kernel/cpu/mcheck/winchip.c +++ b/arch/x86/kernel/cpu/mcheck/winchip.c @@ -5,7 +5,6 @@ #include <linux/interrupt.h> #include <linux/kernel.h> #include <linux/types.h> -#include <linux/init.h> #include <asm/processor.h> #include <asm/mce.h> diff --git a/arch/x86/kernel/cpu/microcode/Makefile b/arch/x86/kernel/cpu/microcode/Makefile new file mode 100644 index 00000000000..285c85427c3 --- /dev/null +++ b/arch/x86/kernel/cpu/microcode/Makefile @@ -0,0 +1,7 @@ +microcode-y := core.o +obj-$(CONFIG_MICROCODE) += microcode.o +microcode-$(CONFIG_MICROCODE_INTEL) += intel.o intel_lib.o +microcode-$(CONFIG_MICROCODE_AMD) += amd.o +obj-$(CONFIG_MICROCODE_EARLY) += core_early.o +obj-$(CONFIG_MICROCODE_INTEL_EARLY) += intel_early.o +obj-$(CONFIG_MICROCODE_AMD_EARLY) += amd_early.o diff --git a/arch/x86/kernel/cpu/microcode/amd.c b/arch/x86/kernel/cpu/microcode/amd.c new file mode 100644 index 00000000000..8fffd845e22 --- /dev/null +++ b/arch/x86/kernel/cpu/microcode/amd.c @@ -0,0 +1,492 @@ +/* + * AMD CPU Microcode Update Driver for Linux + * Copyright (C) 2008-2011 Advanced Micro Devices Inc. + * + * Author: Peter Oruba <peter.oruba@amd.com> + * + * Based on work by: + * Tigran Aivazian <tigran@aivazian.fsnet.co.uk> + * + * Maintainers: + * Andreas Herrmann <herrmann.der.user@googlemail.com> + * Borislav Petkov <bp@alien8.de> + * + * This driver allows to upgrade microcode on F10h AMD + * CPUs and later. + * + * Licensed under the terms of the GNU General Public + * License version 2. See file COPYING for details. + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/firmware.h> +#include <linux/pci_ids.h> +#include <linux/uaccess.h> +#include <linux/vmalloc.h> +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/pci.h> + +#include <asm/microcode.h> +#include <asm/processor.h> +#include <asm/msr.h> +#include <asm/microcode_amd.h> + +MODULE_DESCRIPTION("AMD Microcode Update Driver"); +MODULE_AUTHOR("Peter Oruba"); +MODULE_LICENSE("GPL v2"); + +static struct equiv_cpu_entry *equiv_cpu_table; + +struct ucode_patch { + struct list_head plist; + void *data; + u32 patch_id; + u16 equiv_cpu; +}; + +static LIST_HEAD(pcache); + +static u16 __find_equiv_id(unsigned int cpu) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + return find_equiv_id(equiv_cpu_table, uci->cpu_sig.sig); +} + +static u32 find_cpu_family_by_equiv_cpu(u16 equiv_cpu) +{ + int i = 0; + + BUG_ON(!equiv_cpu_table); + + while (equiv_cpu_table[i].equiv_cpu != 0) { + if (equiv_cpu == equiv_cpu_table[i].equiv_cpu) + return equiv_cpu_table[i].installed_cpu; + i++; + } + return 0; +} + +/* + * a small, trivial cache of per-family ucode patches + */ +static struct ucode_patch *cache_find_patch(u16 equiv_cpu) +{ + struct ucode_patch *p; + + list_for_each_entry(p, &pcache, plist) + if (p->equiv_cpu == equiv_cpu) + return p; + return NULL; +} + +static void update_cache(struct ucode_patch *new_patch) +{ + struct ucode_patch *p; + + list_for_each_entry(p, &pcache, plist) { + if (p->equiv_cpu == new_patch->equiv_cpu) { + if (p->patch_id >= new_patch->patch_id) + /* we already have the latest patch */ + return; + + list_replace(&p->plist, &new_patch->plist); + kfree(p->data); + kfree(p); + return; + } + } + /* no patch found, add it */ + list_add_tail(&new_patch->plist, &pcache); +} + +static void free_cache(void) +{ + struct ucode_patch *p, *tmp; + + list_for_each_entry_safe(p, tmp, &pcache, plist) { + __list_del(p->plist.prev, p->plist.next); + kfree(p->data); + kfree(p); + } +} + +static struct ucode_patch *find_patch(unsigned int cpu) +{ + u16 equiv_id; + + equiv_id = __find_equiv_id(cpu); + if (!equiv_id) + return NULL; + + return cache_find_patch(equiv_id); +} + +static int collect_cpu_info_amd(int cpu, struct cpu_signature *csig) +{ + struct cpuinfo_x86 *c = &cpu_data(cpu); + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + struct ucode_patch *p; + + csig->sig = cpuid_eax(0x00000001); + csig->rev = c->microcode; + + /* + * a patch could have been loaded early, set uci->mc so that + * mc_bp_resume() can call apply_microcode() + */ + p = find_patch(cpu); + if (p && (p->patch_id == csig->rev)) + uci->mc = p->data; + + pr_info("CPU%d: patch_level=0x%08x\n", cpu, csig->rev); + + return 0; +} + +static unsigned int verify_patch_size(u8 family, u32 patch_size, + unsigned int size) +{ + u32 max_size; + +#define F1XH_MPB_MAX_SIZE 2048 +#define F14H_MPB_MAX_SIZE 1824 +#define F15H_MPB_MAX_SIZE 4096 +#define F16H_MPB_MAX_SIZE 3458 + + switch (family) { + case 0x14: + max_size = F14H_MPB_MAX_SIZE; + break; + case 0x15: + max_size = F15H_MPB_MAX_SIZE; + break; + case 0x16: + max_size = F16H_MPB_MAX_SIZE; + break; + default: + max_size = F1XH_MPB_MAX_SIZE; + break; + } + + if (patch_size > min_t(u32, size, max_size)) { + pr_err("patch size mismatch\n"); + return 0; + } + + return patch_size; +} + +int __apply_microcode_amd(struct microcode_amd *mc_amd) +{ + u32 rev, dummy; + + native_wrmsrl(MSR_AMD64_PATCH_LOADER, (u64)(long)&mc_amd->hdr.data_code); + + /* verify patch application was successful */ + native_rdmsr(MSR_AMD64_PATCH_LEVEL, rev, dummy); + if (rev != mc_amd->hdr.patch_id) + return -1; + + return 0; +} + +int apply_microcode_amd(int cpu) +{ + struct cpuinfo_x86 *c = &cpu_data(cpu); + struct microcode_amd *mc_amd; + struct ucode_cpu_info *uci; + struct ucode_patch *p; + u32 rev, dummy; + + BUG_ON(raw_smp_processor_id() != cpu); + + uci = ucode_cpu_info + cpu; + + p = find_patch(cpu); + if (!p) + return 0; + + mc_amd = p->data; + uci->mc = p->data; + + rdmsr(MSR_AMD64_PATCH_LEVEL, rev, dummy); + + /* need to apply patch? */ + if (rev >= mc_amd->hdr.patch_id) { + c->microcode = rev; + uci->cpu_sig.rev = rev; + return 0; + } + + if (__apply_microcode_amd(mc_amd)) { + pr_err("CPU%d: update failed for patch_level=0x%08x\n", + cpu, mc_amd->hdr.patch_id); + return -1; + } + pr_info("CPU%d: new patch_level=0x%08x\n", cpu, + mc_amd->hdr.patch_id); + + uci->cpu_sig.rev = mc_amd->hdr.patch_id; + c->microcode = mc_amd->hdr.patch_id; + + return 0; +} + +static int install_equiv_cpu_table(const u8 *buf) +{ + unsigned int *ibuf = (unsigned int *)buf; + unsigned int type = ibuf[1]; + unsigned int size = ibuf[2]; + + if (type != UCODE_EQUIV_CPU_TABLE_TYPE || !size) { + pr_err("empty section/" + "invalid type field in container file section header\n"); + return -EINVAL; + } + + equiv_cpu_table = vmalloc(size); + if (!equiv_cpu_table) { + pr_err("failed to allocate equivalent CPU table\n"); + return -ENOMEM; + } + + memcpy(equiv_cpu_table, buf + CONTAINER_HDR_SZ, size); + + /* add header length */ + return size + CONTAINER_HDR_SZ; +} + +static void free_equiv_cpu_table(void) +{ + vfree(equiv_cpu_table); + equiv_cpu_table = NULL; +} + +static void cleanup(void) +{ + free_equiv_cpu_table(); + free_cache(); +} + +/* + * We return the current size even if some of the checks failed so that + * we can skip over the next patch. If we return a negative value, we + * signal a grave error like a memory allocation has failed and the + * driver cannot continue functioning normally. In such cases, we tear + * down everything we've used up so far and exit. + */ +static int verify_and_add_patch(u8 family, u8 *fw, unsigned int leftover) +{ + struct microcode_header_amd *mc_hdr; + struct ucode_patch *patch; + unsigned int patch_size, crnt_size, ret; + u32 proc_fam; + u16 proc_id; + + patch_size = *(u32 *)(fw + 4); + crnt_size = patch_size + SECTION_HDR_SIZE; + mc_hdr = (struct microcode_header_amd *)(fw + SECTION_HDR_SIZE); + proc_id = mc_hdr->processor_rev_id; + + proc_fam = find_cpu_family_by_equiv_cpu(proc_id); + if (!proc_fam) { + pr_err("No patch family for equiv ID: 0x%04x\n", proc_id); + return crnt_size; + } + + /* check if patch is for the current family */ + proc_fam = ((proc_fam >> 8) & 0xf) + ((proc_fam >> 20) & 0xff); + if (proc_fam != family) + return crnt_size; + + if (mc_hdr->nb_dev_id || mc_hdr->sb_dev_id) { + pr_err("Patch-ID 0x%08x: chipset-specific code unsupported.\n", + mc_hdr->patch_id); + return crnt_size; + } + + ret = verify_patch_size(family, patch_size, leftover); + if (!ret) { + pr_err("Patch-ID 0x%08x: size mismatch.\n", mc_hdr->patch_id); + return crnt_size; + } + + patch = kzalloc(sizeof(*patch), GFP_KERNEL); + if (!patch) { + pr_err("Patch allocation failure.\n"); + return -EINVAL; + } + + patch->data = kzalloc(patch_size, GFP_KERNEL); + if (!patch->data) { + pr_err("Patch data allocation failure.\n"); + kfree(patch); + return -EINVAL; + } + + /* All looks ok, copy patch... */ + memcpy(patch->data, fw + SECTION_HDR_SIZE, patch_size); + INIT_LIST_HEAD(&patch->plist); + patch->patch_id = mc_hdr->patch_id; + patch->equiv_cpu = proc_id; + + pr_debug("%s: Added patch_id: 0x%08x, proc_id: 0x%04x\n", + __func__, patch->patch_id, proc_id); + + /* ... and add to cache. */ + update_cache(patch); + + return crnt_size; +} + +static enum ucode_state __load_microcode_amd(u8 family, const u8 *data, + size_t size) +{ + enum ucode_state ret = UCODE_ERROR; + unsigned int leftover; + u8 *fw = (u8 *)data; + int crnt_size = 0; + int offset; + + offset = install_equiv_cpu_table(data); + if (offset < 0) { + pr_err("failed to create equivalent cpu table\n"); + return ret; + } + fw += offset; + leftover = size - offset; + + if (*(u32 *)fw != UCODE_UCODE_TYPE) { + pr_err("invalid type field in container file section header\n"); + free_equiv_cpu_table(); + return ret; + } + + while (leftover) { + crnt_size = verify_and_add_patch(family, fw, leftover); + if (crnt_size < 0) + return ret; + + fw += crnt_size; + leftover -= crnt_size; + } + + return UCODE_OK; +} + +enum ucode_state load_microcode_amd(u8 family, const u8 *data, size_t size) +{ + enum ucode_state ret; + + /* free old equiv table */ + free_equiv_cpu_table(); + + ret = __load_microcode_amd(family, data, size); + + if (ret != UCODE_OK) + cleanup(); + +#if defined(CONFIG_MICROCODE_AMD_EARLY) && defined(CONFIG_X86_32) + /* save BSP's matching patch for early load */ + if (cpu_data(smp_processor_id()).cpu_index == boot_cpu_data.cpu_index) { + struct ucode_patch *p = find_patch(smp_processor_id()); + if (p) { + memset(amd_ucode_patch, 0, PATCH_MAX_SIZE); + memcpy(amd_ucode_patch, p->data, min_t(u32, ksize(p->data), + PATCH_MAX_SIZE)); + } + } +#endif + return ret; +} + +/* + * AMD microcode firmware naming convention, up to family 15h they are in + * the legacy file: + * + * amd-ucode/microcode_amd.bin + * + * This legacy file is always smaller than 2K in size. + * + * Beginning with family 15h, they are in family-specific firmware files: + * + * amd-ucode/microcode_amd_fam15h.bin + * amd-ucode/microcode_amd_fam16h.bin + * ... + * + * These might be larger than 2K. + */ +static enum ucode_state request_microcode_amd(int cpu, struct device *device, + bool refresh_fw) +{ + char fw_name[36] = "amd-ucode/microcode_amd.bin"; + struct cpuinfo_x86 *c = &cpu_data(cpu); + enum ucode_state ret = UCODE_NFOUND; + const struct firmware *fw; + + /* reload ucode container only on the boot cpu */ + if (!refresh_fw || c->cpu_index != boot_cpu_data.cpu_index) + return UCODE_OK; + + if (c->x86 >= 0x15) + snprintf(fw_name, sizeof(fw_name), "amd-ucode/microcode_amd_fam%.2xh.bin", c->x86); + + if (request_firmware_direct(&fw, (const char *)fw_name, device)) { + pr_debug("failed to load file %s\n", fw_name); + goto out; + } + + ret = UCODE_ERROR; + if (*(u32 *)fw->data != UCODE_MAGIC) { + pr_err("invalid magic value (0x%08x)\n", *(u32 *)fw->data); + goto fw_release; + } + + ret = load_microcode_amd(c->x86, fw->data, fw->size); + + fw_release: + release_firmware(fw); + + out: + return ret; +} + +static enum ucode_state +request_microcode_user(int cpu, const void __user *buf, size_t size) +{ + return UCODE_ERROR; +} + +static void microcode_fini_cpu_amd(int cpu) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + + uci->mc = NULL; +} + +static struct microcode_ops microcode_amd_ops = { + .request_microcode_user = request_microcode_user, + .request_microcode_fw = request_microcode_amd, + .collect_cpu_info = collect_cpu_info_amd, + .apply_microcode = apply_microcode_amd, + .microcode_fini_cpu = microcode_fini_cpu_amd, +}; + +struct microcode_ops * __init init_amd_microcode(void) +{ + struct cpuinfo_x86 *c = &cpu_data(0); + + if (c->x86_vendor != X86_VENDOR_AMD || c->x86 < 0x10) { + pr_warning("AMD CPU family 0x%x not supported\n", c->x86); + return NULL; + } + + return µcode_amd_ops; +} + +void __exit exit_amd_microcode(void) +{ + cleanup(); +} diff --git a/arch/x86/kernel/cpu/microcode/amd_early.c b/arch/x86/kernel/cpu/microcode/amd_early.c new file mode 100644 index 00000000000..617a9e28424 --- /dev/null +++ b/arch/x86/kernel/cpu/microcode/amd_early.c @@ -0,0 +1,395 @@ +/* + * Copyright (C) 2013 Advanced Micro Devices, Inc. + * + * Author: Jacob Shin <jacob.shin@amd.com> + * Fixes: Borislav Petkov <bp@suse.de> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/earlycpio.h> +#include <linux/initrd.h> + +#include <asm/cpu.h> +#include <asm/setup.h> +#include <asm/microcode_amd.h> + +/* + * This points to the current valid container of microcode patches which we will + * save from the initrd before jettisoning its contents. + */ +static u8 *container; +static size_t container_size; + +static u32 ucode_new_rev; +u8 amd_ucode_patch[PATCH_MAX_SIZE]; +static u16 this_equiv_id; + +struct cpio_data ucode_cpio; + +/* + * Microcode patch container file is prepended to the initrd in cpio format. + * See Documentation/x86/early-microcode.txt + */ +static __initdata char ucode_path[] = "kernel/x86/microcode/AuthenticAMD.bin"; + +static struct cpio_data __init find_ucode_in_initrd(void) +{ + long offset = 0; + char *path; + void *start; + size_t size; + +#ifdef CONFIG_X86_32 + struct boot_params *p; + + /* + * On 32-bit, early load occurs before paging is turned on so we need + * to use physical addresses. + */ + p = (struct boot_params *)__pa_nodebug(&boot_params); + path = (char *)__pa_nodebug(ucode_path); + start = (void *)p->hdr.ramdisk_image; + size = p->hdr.ramdisk_size; +#else + path = ucode_path; + start = (void *)(boot_params.hdr.ramdisk_image + PAGE_OFFSET); + size = boot_params.hdr.ramdisk_size; +#endif + + return find_cpio_data(path, start, size, &offset); +} + +static size_t compute_container_size(u8 *data, u32 total_size) +{ + size_t size = 0; + u32 *header = (u32 *)data; + + if (header[0] != UCODE_MAGIC || + header[1] != UCODE_EQUIV_CPU_TABLE_TYPE || /* type */ + header[2] == 0) /* size */ + return size; + + size = header[2] + CONTAINER_HDR_SZ; + total_size -= size; + data += size; + + while (total_size) { + u16 patch_size; + + header = (u32 *)data; + + if (header[0] != UCODE_UCODE_TYPE) + break; + + /* + * Sanity-check patch size. + */ + patch_size = header[1]; + if (patch_size > PATCH_MAX_SIZE) + break; + + size += patch_size + SECTION_HDR_SIZE; + data += patch_size + SECTION_HDR_SIZE; + total_size -= patch_size + SECTION_HDR_SIZE; + } + + return size; +} + +/* + * Early load occurs before we can vmalloc(). So we look for the microcode + * patch container file in initrd, traverse equivalent cpu table, look for a + * matching microcode patch, and update, all in initrd memory in place. + * When vmalloc() is available for use later -- on 64-bit during first AP load, + * and on 32-bit during save_microcode_in_initrd_amd() -- we can call + * load_microcode_amd() to save equivalent cpu table and microcode patches in + * kernel heap memory. + */ +static void apply_ucode_in_initrd(void *ucode, size_t size) +{ + struct equiv_cpu_entry *eq; + size_t *cont_sz; + u32 *header; + u8 *data, **cont; + u16 eq_id = 0; + int offset, left; + u32 rev, eax, ebx, ecx, edx; + u32 *new_rev; + +#ifdef CONFIG_X86_32 + new_rev = (u32 *)__pa_nodebug(&ucode_new_rev); + cont_sz = (size_t *)__pa_nodebug(&container_size); + cont = (u8 **)__pa_nodebug(&container); +#else + new_rev = &ucode_new_rev; + cont_sz = &container_size; + cont = &container; +#endif + + data = ucode; + left = size; + header = (u32 *)data; + + /* find equiv cpu table */ + if (header[0] != UCODE_MAGIC || + header[1] != UCODE_EQUIV_CPU_TABLE_TYPE || /* type */ + header[2] == 0) /* size */ + return; + + eax = 0x00000001; + ecx = 0; + native_cpuid(&eax, &ebx, &ecx, &edx); + + while (left > 0) { + eq = (struct equiv_cpu_entry *)(data + CONTAINER_HDR_SZ); + + *cont = data; + + /* Advance past the container header */ + offset = header[2] + CONTAINER_HDR_SZ; + data += offset; + left -= offset; + + eq_id = find_equiv_id(eq, eax); + if (eq_id) { + this_equiv_id = eq_id; + *cont_sz = compute_container_size(*cont, left + offset); + + /* + * truncate how much we need to iterate over in the + * ucode update loop below + */ + left = *cont_sz - offset; + break; + } + + /* + * support multiple container files appended together. if this + * one does not have a matching equivalent cpu entry, we fast + * forward to the next container file. + */ + while (left > 0) { + header = (u32 *)data; + if (header[0] == UCODE_MAGIC && + header[1] == UCODE_EQUIV_CPU_TABLE_TYPE) + break; + + offset = header[1] + SECTION_HDR_SIZE; + data += offset; + left -= offset; + } + + /* mark where the next microcode container file starts */ + offset = data - (u8 *)ucode; + ucode = data; + } + + if (!eq_id) { + *cont = NULL; + *cont_sz = 0; + return; + } + + /* find ucode and update if needed */ + + native_rdmsr(MSR_AMD64_PATCH_LEVEL, rev, eax); + + while (left > 0) { + struct microcode_amd *mc; + + header = (u32 *)data; + if (header[0] != UCODE_UCODE_TYPE || /* type */ + header[1] == 0) /* size */ + break; + + mc = (struct microcode_amd *)(data + SECTION_HDR_SIZE); + + if (eq_id == mc->hdr.processor_rev_id && rev < mc->hdr.patch_id) { + + if (!__apply_microcode_amd(mc)) { + rev = mc->hdr.patch_id; + *new_rev = rev; + + /* save ucode patch */ + memcpy(amd_ucode_patch, mc, + min_t(u32, header[1], PATCH_MAX_SIZE)); + } + } + + offset = header[1] + SECTION_HDR_SIZE; + data += offset; + left -= offset; + } +} + +void __init load_ucode_amd_bsp(void) +{ + struct cpio_data cp; + void **data; + size_t *size; + +#ifdef CONFIG_X86_32 + data = (void **)__pa_nodebug(&ucode_cpio.data); + size = (size_t *)__pa_nodebug(&ucode_cpio.size); +#else + data = &ucode_cpio.data; + size = &ucode_cpio.size; +#endif + + cp = find_ucode_in_initrd(); + if (!cp.data) + return; + + *data = cp.data; + *size = cp.size; + + apply_ucode_in_initrd(cp.data, cp.size); +} + +#ifdef CONFIG_X86_32 +/* + * On 32-bit, since AP's early load occurs before paging is turned on, we + * cannot traverse cpu_equiv_table and pcache in kernel heap memory. So during + * cold boot, AP will apply_ucode_in_initrd() just like the BSP. During + * save_microcode_in_initrd_amd() BSP's patch is copied to amd_ucode_patch, + * which is used upon resume from suspend. + */ +void load_ucode_amd_ap(void) +{ + struct microcode_amd *mc; + size_t *usize; + void **ucode; + + mc = (struct microcode_amd *)__pa(amd_ucode_patch); + if (mc->hdr.patch_id && mc->hdr.processor_rev_id) { + __apply_microcode_amd(mc); + return; + } + + ucode = (void *)__pa_nodebug(&container); + usize = (size_t *)__pa_nodebug(&container_size); + + if (!*ucode || !*usize) + return; + + apply_ucode_in_initrd(*ucode, *usize); +} + +static void __init collect_cpu_sig_on_bsp(void *arg) +{ + unsigned int cpu = smp_processor_id(); + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + + uci->cpu_sig.sig = cpuid_eax(0x00000001); +} + +static void __init get_bsp_sig(void) +{ + unsigned int bsp = boot_cpu_data.cpu_index; + struct ucode_cpu_info *uci = ucode_cpu_info + bsp; + + if (!uci->cpu_sig.sig) + smp_call_function_single(bsp, collect_cpu_sig_on_bsp, NULL, 1); +} +#else +void load_ucode_amd_ap(void) +{ + unsigned int cpu = smp_processor_id(); + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + struct equiv_cpu_entry *eq; + struct microcode_amd *mc; + u32 rev, eax; + u16 eq_id; + + /* Exit if called on the BSP. */ + if (!cpu) + return; + + if (!container) + return; + + rdmsr(MSR_AMD64_PATCH_LEVEL, rev, eax); + + uci->cpu_sig.rev = rev; + uci->cpu_sig.sig = eax; + + eax = cpuid_eax(0x00000001); + eq = (struct equiv_cpu_entry *)(container + CONTAINER_HDR_SZ); + + eq_id = find_equiv_id(eq, eax); + if (!eq_id) + return; + + if (eq_id == this_equiv_id) { + mc = (struct microcode_amd *)amd_ucode_patch; + + if (mc && rev < mc->hdr.patch_id) { + if (!__apply_microcode_amd(mc)) + ucode_new_rev = mc->hdr.patch_id; + } + + } else { + if (!ucode_cpio.data) + return; + + /* + * AP has a different equivalence ID than BSP, looks like + * mixed-steppings silicon so go through the ucode blob anew. + */ + apply_ucode_in_initrd(ucode_cpio.data, ucode_cpio.size); + } +} +#endif + +int __init save_microcode_in_initrd_amd(void) +{ + unsigned long cont; + enum ucode_state ret; + u32 eax; + + if (!container) + return -EINVAL; + +#ifdef CONFIG_X86_32 + get_bsp_sig(); + cont = (unsigned long)container; +#else + /* + * We need the physical address of the container for both bitness since + * boot_params.hdr.ramdisk_image is a physical address. + */ + cont = __pa(container); +#endif + + /* + * Take into account the fact that the ramdisk might get relocated and + * therefore we need to recompute the container's position in virtual + * memory space. + */ + if (relocated_ramdisk) + container = (u8 *)(__va(relocated_ramdisk) + + (cont - boot_params.hdr.ramdisk_image)); + + if (ucode_new_rev) + pr_info("microcode: updated early to new patch_level=0x%08x\n", + ucode_new_rev); + + eax = cpuid_eax(0x00000001); + eax = ((eax >> 8) & 0xf) + ((eax >> 20) & 0xff); + + ret = load_microcode_amd(eax, container, container_size); + if (ret != UCODE_OK) + return -EINVAL; + + /* + * This will be freed any msec now, stash patches for the current + * family and switch to patch cache for cpu hotplug, etc later. + */ + container = NULL; + container_size = 0; + + return 0; +} diff --git a/arch/x86/kernel/cpu/microcode/core.c b/arch/x86/kernel/cpu/microcode/core.c new file mode 100644 index 00000000000..dd9d6190b08 --- /dev/null +++ b/arch/x86/kernel/cpu/microcode/core.c @@ -0,0 +1,651 @@ +/* + * Intel CPU Microcode Update Driver for Linux + * + * Copyright (C) 2000-2006 Tigran Aivazian <tigran@aivazian.fsnet.co.uk> + * 2006 Shaohua Li <shaohua.li@intel.com> + * + * This driver allows to upgrade microcode on Intel processors + * belonging to IA-32 family - PentiumPro, Pentium II, + * Pentium III, Xeon, Pentium 4, etc. + * + * Reference: Section 8.11 of Volume 3a, IA-32 Intel? Architecture + * Software Developer's Manual + * Order Number 253668 or free download from: + * + * http://developer.intel.com/Assets/PDF/manual/253668.pdf + * + * For more information, go to http://www.urbanmyth.org/microcode + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + * + * 1.0 16 Feb 2000, Tigran Aivazian <tigran@sco.com> + * Initial release. + * 1.01 18 Feb 2000, Tigran Aivazian <tigran@sco.com> + * Added read() support + cleanups. + * 1.02 21 Feb 2000, Tigran Aivazian <tigran@sco.com> + * Added 'device trimming' support. open(O_WRONLY) zeroes + * and frees the saved copy of applied microcode. + * 1.03 29 Feb 2000, Tigran Aivazian <tigran@sco.com> + * Made to use devfs (/dev/cpu/microcode) + cleanups. + * 1.04 06 Jun 2000, Simon Trimmer <simon@veritas.com> + * Added misc device support (now uses both devfs and misc). + * Added MICROCODE_IOCFREE ioctl to clear memory. + * 1.05 09 Jun 2000, Simon Trimmer <simon@veritas.com> + * Messages for error cases (non Intel & no suitable microcode). + * 1.06 03 Aug 2000, Tigran Aivazian <tigran@veritas.com> + * Removed ->release(). Removed exclusive open and status bitmap. + * Added microcode_rwsem to serialize read()/write()/ioctl(). + * Removed global kernel lock usage. + * 1.07 07 Sep 2000, Tigran Aivazian <tigran@veritas.com> + * Write 0 to 0x8B msr and then cpuid before reading revision, + * so that it works even if there were no update done by the + * BIOS. Otherwise, reading from 0x8B gives junk (which happened + * to be 0 on my machine which is why it worked even when I + * disabled update by the BIOS) + * Thanks to Eric W. Biederman <ebiederman@lnxi.com> for the fix. + * 1.08 11 Dec 2000, Richard Schaal <richard.schaal@intel.com> and + * Tigran Aivazian <tigran@veritas.com> + * Intel Pentium 4 processor support and bugfixes. + * 1.09 30 Oct 2001, Tigran Aivazian <tigran@veritas.com> + * Bugfix for HT (Hyper-Threading) enabled processors + * whereby processor resources are shared by all logical processors + * in a single CPU package. + * 1.10 28 Feb 2002 Asit K Mallick <asit.k.mallick@intel.com> and + * Tigran Aivazian <tigran@veritas.com>, + * Serialize updates as required on HT processors due to + * speculative 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>, + * 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 + * in kernel memory. + * 1.14 25 Jun 2004 Tigran Aivazian <tigran@veritas.com> + * Fix sigmatch() macro to handle old CPUs with pf == 0. + * Thanks to Stuart Swales for pointing out this bug. + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/platform_device.h> +#include <linux/miscdevice.h> +#include <linux/capability.h> +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/mutex.h> +#include <linux/cpu.h> +#include <linux/fs.h> +#include <linux/mm.h> +#include <linux/syscore_ops.h> + +#include <asm/microcode.h> +#include <asm/processor.h> +#include <asm/cpu_device_id.h> +#include <asm/perf_event.h> + +MODULE_DESCRIPTION("Microcode Update Driver"); +MODULE_AUTHOR("Tigran Aivazian <tigran@aivazian.fsnet.co.uk>"); +MODULE_LICENSE("GPL"); + +#define MICROCODE_VERSION "2.00" + +static struct microcode_ops *microcode_ops; + +bool dis_ucode_ldr; +module_param(dis_ucode_ldr, bool, 0); + +/* + * Synchronization. + * + * All non cpu-hotplug-callback call sites use: + * + * - microcode_mutex to synchronize with each other; + * - get/put_online_cpus() to synchronize with + * the cpu-hotplug-callback call sites. + * + * We guarantee that only a single cpu is being + * updated at any particular moment of time. + */ +static DEFINE_MUTEX(microcode_mutex); + +struct ucode_cpu_info ucode_cpu_info[NR_CPUS]; +EXPORT_SYMBOL_GPL(ucode_cpu_info); + +/* + * Operations that are run on a target cpu: + */ + +struct cpu_info_ctx { + struct cpu_signature *cpu_sig; + int err; +}; + +static void collect_cpu_info_local(void *arg) +{ + struct cpu_info_ctx *ctx = arg; + + ctx->err = microcode_ops->collect_cpu_info(smp_processor_id(), + ctx->cpu_sig); +} + +static int collect_cpu_info_on_target(int cpu, struct cpu_signature *cpu_sig) +{ + struct cpu_info_ctx ctx = { .cpu_sig = cpu_sig, .err = 0 }; + int ret; + + ret = smp_call_function_single(cpu, collect_cpu_info_local, &ctx, 1); + if (!ret) + ret = ctx.err; + + return ret; +} + +static int collect_cpu_info(int cpu) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + int ret; + + memset(uci, 0, sizeof(*uci)); + + ret = collect_cpu_info_on_target(cpu, &uci->cpu_sig); + if (!ret) + uci->valid = 1; + + return ret; +} + +struct apply_microcode_ctx { + int err; +}; + +static void apply_microcode_local(void *arg) +{ + struct apply_microcode_ctx *ctx = arg; + + ctx->err = microcode_ops->apply_microcode(smp_processor_id()); +} + +static int apply_microcode_on_target(int cpu) +{ + struct apply_microcode_ctx ctx = { .err = 0 }; + int ret; + + ret = smp_call_function_single(cpu, apply_microcode_local, &ctx, 1); + if (!ret) + ret = ctx.err; + + return ret; +} + +#ifdef CONFIG_MICROCODE_OLD_INTERFACE +static int do_microcode_update(const void __user *buf, size_t size) +{ + int error = 0; + int cpu; + + for_each_online_cpu(cpu) { + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + enum ucode_state ustate; + + if (!uci->valid) + continue; + + ustate = microcode_ops->request_microcode_user(cpu, buf, size); + if (ustate == UCODE_ERROR) { + error = -1; + break; + } else if (ustate == UCODE_OK) + apply_microcode_on_target(cpu); + } + + return error; +} + +static int microcode_open(struct inode *inode, struct file *file) +{ + return capable(CAP_SYS_RAWIO) ? nonseekable_open(inode, file) : -EPERM; +} + +static ssize_t microcode_write(struct file *file, const char __user *buf, + size_t len, loff_t *ppos) +{ + ssize_t ret = -EINVAL; + + if ((len >> PAGE_SHIFT) > totalram_pages) { + pr_err("too much data (max %ld pages)\n", totalram_pages); + return ret; + } + + get_online_cpus(); + mutex_lock(µcode_mutex); + + if (do_microcode_update(buf, len) == 0) + ret = (ssize_t)len; + + if (ret > 0) + perf_check_microcode(); + + mutex_unlock(µcode_mutex); + put_online_cpus(); + + return ret; +} + +static const struct file_operations microcode_fops = { + .owner = THIS_MODULE, + .write = microcode_write, + .open = microcode_open, + .llseek = no_llseek, +}; + +static struct miscdevice microcode_dev = { + .minor = MICROCODE_MINOR, + .name = "microcode", + .nodename = "cpu/microcode", + .fops = µcode_fops, +}; + +static int __init microcode_dev_init(void) +{ + int error; + + error = misc_register(µcode_dev); + if (error) { + pr_err("can't misc_register on minor=%d\n", MICROCODE_MINOR); + return error; + } + + return 0; +} + +static void __exit microcode_dev_exit(void) +{ + misc_deregister(µcode_dev); +} + +MODULE_ALIAS_MISCDEV(MICROCODE_MINOR); +MODULE_ALIAS("devname:cpu/microcode"); +#else +#define microcode_dev_init() 0 +#define microcode_dev_exit() do { } while (0) +#endif + +/* fake device for request_firmware */ +static struct platform_device *microcode_pdev; + +static int reload_for_cpu(int cpu) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + enum ucode_state ustate; + int err = 0; + + if (!uci->valid) + return err; + + ustate = microcode_ops->request_microcode_fw(cpu, µcode_pdev->dev, true); + if (ustate == UCODE_OK) + apply_microcode_on_target(cpu); + else + if (ustate == UCODE_ERROR) + err = -EINVAL; + return err; +} + +static ssize_t reload_store(struct device *dev, + struct device_attribute *attr, + const char *buf, size_t size) +{ + unsigned long val; + int cpu; + ssize_t ret = 0, tmp_ret; + + ret = kstrtoul(buf, 0, &val); + if (ret) + return ret; + + if (val != 1) + return size; + + get_online_cpus(); + mutex_lock(µcode_mutex); + for_each_online_cpu(cpu) { + tmp_ret = reload_for_cpu(cpu); + if (tmp_ret != 0) + pr_warn("Error reloading microcode on CPU %d\n", cpu); + + /* save retval of the first encountered reload error */ + if (!ret) + ret = tmp_ret; + } + if (!ret) + perf_check_microcode(); + mutex_unlock(µcode_mutex); + put_online_cpus(); + + if (!ret) + ret = size; + + return ret; +} + +static ssize_t version_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + dev->id; + + return sprintf(buf, "0x%x\n", uci->cpu_sig.rev); +} + +static ssize_t pf_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + dev->id; + + return sprintf(buf, "0x%x\n", uci->cpu_sig.pf); +} + +static DEVICE_ATTR(reload, 0200, NULL, reload_store); +static DEVICE_ATTR(version, 0400, version_show, NULL); +static DEVICE_ATTR(processor_flags, 0400, pf_show, NULL); + +static struct attribute *mc_default_attrs[] = { + &dev_attr_version.attr, + &dev_attr_processor_flags.attr, + NULL +}; + +static struct attribute_group mc_attr_group = { + .attrs = mc_default_attrs, + .name = "microcode", +}; + +static void microcode_fini_cpu(int cpu) +{ + microcode_ops->microcode_fini_cpu(cpu); +} + +static enum ucode_state microcode_resume_cpu(int cpu) +{ + pr_debug("CPU%d updated upon resume\n", cpu); + + if (apply_microcode_on_target(cpu)) + return UCODE_ERROR; + + return UCODE_OK; +} + +static enum ucode_state microcode_init_cpu(int cpu, bool refresh_fw) +{ + enum ucode_state ustate; + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + + if (uci && uci->valid) + return UCODE_OK; + + if (collect_cpu_info(cpu)) + return UCODE_ERROR; + + /* --dimm. Trigger a delayed update? */ + if (system_state != SYSTEM_RUNNING) + return UCODE_NFOUND; + + ustate = microcode_ops->request_microcode_fw(cpu, µcode_pdev->dev, + refresh_fw); + + if (ustate == UCODE_OK) { + pr_debug("CPU%d updated upon init\n", cpu); + apply_microcode_on_target(cpu); + } + + return ustate; +} + +static enum ucode_state microcode_update_cpu(int cpu) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + + if (uci->valid) + return microcode_resume_cpu(cpu); + + return microcode_init_cpu(cpu, false); +} + +static int mc_device_add(struct device *dev, struct subsys_interface *sif) +{ + int err, cpu = dev->id; + + if (!cpu_online(cpu)) + return 0; + + pr_debug("CPU%d added\n", cpu); + + err = sysfs_create_group(&dev->kobj, &mc_attr_group); + if (err) + return err; + + if (microcode_init_cpu(cpu, true) == UCODE_ERROR) + return -EINVAL; + + return err; +} + +static int mc_device_remove(struct device *dev, struct subsys_interface *sif) +{ + int cpu = dev->id; + + if (!cpu_online(cpu)) + return 0; + + pr_debug("CPU%d removed\n", cpu); + microcode_fini_cpu(cpu); + sysfs_remove_group(&dev->kobj, &mc_attr_group); + return 0; +} + +static struct subsys_interface mc_cpu_interface = { + .name = "microcode", + .subsys = &cpu_subsys, + .add_dev = mc_device_add, + .remove_dev = mc_device_remove, +}; + +/** + * mc_bp_resume - Update boot CPU microcode during resume. + */ +static void mc_bp_resume(void) +{ + int cpu = smp_processor_id(); + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + + if (uci->valid && uci->mc) + microcode_ops->apply_microcode(cpu); +} + +static struct syscore_ops mc_syscore_ops = { + .resume = mc_bp_resume, +}; + +static int +mc_cpu_callback(struct notifier_block *nb, unsigned long action, void *hcpu) +{ + unsigned int cpu = (unsigned long)hcpu; + struct device *dev; + + dev = get_cpu_device(cpu); + + switch (action & ~CPU_TASKS_FROZEN) { + case CPU_ONLINE: + microcode_update_cpu(cpu); + pr_debug("CPU%d added\n", cpu); + /* + * "break" is missing on purpose here because we want to fall + * through in order to create the sysfs group. + */ + + case CPU_DOWN_FAILED: + if (sysfs_create_group(&dev->kobj, &mc_attr_group)) + pr_err("Failed to create group for CPU%d\n", cpu); + break; + + case CPU_DOWN_PREPARE: + /* Suspend is in progress, only remove the interface */ + sysfs_remove_group(&dev->kobj, &mc_attr_group); + pr_debug("CPU%d removed\n", cpu); + break; + + /* + * case CPU_DEAD: + * + * When a CPU goes offline, don't free up or invalidate the copy of + * the microcode in kernel memory, so that we can reuse it when the + * CPU comes back online without unnecessarily requesting the userspace + * for it again. + */ + } + + /* The CPU refused to come up during a system resume */ + if (action == CPU_UP_CANCELED_FROZEN) + microcode_fini_cpu(cpu); + + return NOTIFY_OK; +} + +static struct notifier_block __refdata mc_cpu_notifier = { + .notifier_call = mc_cpu_callback, +}; + +#ifdef MODULE +/* Autoload on Intel and AMD systems */ +static const struct x86_cpu_id __initconst microcode_id[] = { +#ifdef CONFIG_MICROCODE_INTEL + { X86_VENDOR_INTEL, X86_FAMILY_ANY, X86_MODEL_ANY, }, +#endif +#ifdef CONFIG_MICROCODE_AMD + { X86_VENDOR_AMD, X86_FAMILY_ANY, X86_MODEL_ANY, }, +#endif + {} +}; +MODULE_DEVICE_TABLE(x86cpu, microcode_id); +#endif + +static struct attribute *cpu_root_microcode_attrs[] = { + &dev_attr_reload.attr, + NULL +}; + +static struct attribute_group cpu_root_microcode_group = { + .name = "microcode", + .attrs = cpu_root_microcode_attrs, +}; + +static int __init microcode_init(void) +{ + struct cpuinfo_x86 *c = &cpu_data(0); + int error; + + if (dis_ucode_ldr) + return 0; + + if (c->x86_vendor == X86_VENDOR_INTEL) + microcode_ops = init_intel_microcode(); + else if (c->x86_vendor == X86_VENDOR_AMD) + microcode_ops = init_amd_microcode(); + else + pr_err("no support for this CPU vendor\n"); + + if (!microcode_ops) + return -ENODEV; + + microcode_pdev = platform_device_register_simple("microcode", -1, + NULL, 0); + if (IS_ERR(microcode_pdev)) + return PTR_ERR(microcode_pdev); + + get_online_cpus(); + mutex_lock(µcode_mutex); + + error = subsys_interface_register(&mc_cpu_interface); + if (!error) + perf_check_microcode(); + mutex_unlock(µcode_mutex); + put_online_cpus(); + + if (error) + goto out_pdev; + + error = sysfs_create_group(&cpu_subsys.dev_root->kobj, + &cpu_root_microcode_group); + + if (error) { + pr_err("Error creating microcode group!\n"); + goto out_driver; + } + + error = microcode_dev_init(); + if (error) + goto out_ucode_group; + + register_syscore_ops(&mc_syscore_ops); + register_hotcpu_notifier(&mc_cpu_notifier); + + pr_info("Microcode Update Driver: v" MICROCODE_VERSION + " <tigran@aivazian.fsnet.co.uk>, Peter Oruba\n"); + + return 0; + + out_ucode_group: + sysfs_remove_group(&cpu_subsys.dev_root->kobj, + &cpu_root_microcode_group); + + out_driver: + get_online_cpus(); + mutex_lock(µcode_mutex); + + subsys_interface_unregister(&mc_cpu_interface); + + mutex_unlock(µcode_mutex); + put_online_cpus(); + + out_pdev: + platform_device_unregister(microcode_pdev); + return error; + +} +module_init(microcode_init); + +static void __exit microcode_exit(void) +{ + struct cpuinfo_x86 *c = &cpu_data(0); + + microcode_dev_exit(); + + unregister_hotcpu_notifier(&mc_cpu_notifier); + unregister_syscore_ops(&mc_syscore_ops); + + sysfs_remove_group(&cpu_subsys.dev_root->kobj, + &cpu_root_microcode_group); + + get_online_cpus(); + mutex_lock(µcode_mutex); + + subsys_interface_unregister(&mc_cpu_interface); + + mutex_unlock(µcode_mutex); + put_online_cpus(); + + platform_device_unregister(microcode_pdev); + + microcode_ops = NULL; + + if (c->x86_vendor == X86_VENDOR_AMD) + exit_amd_microcode(); + + pr_info("Microcode Update Driver: v" MICROCODE_VERSION " removed.\n"); +} +module_exit(microcode_exit); diff --git a/arch/x86/kernel/cpu/microcode/core_early.c b/arch/x86/kernel/cpu/microcode/core_early.c new file mode 100644 index 00000000000..5f28a64e71e --- /dev/null +++ b/arch/x86/kernel/cpu/microcode/core_early.c @@ -0,0 +1,178 @@ +/* + * X86 CPU microcode early update for Linux + * + * Copyright (C) 2012 Fenghua Yu <fenghua.yu@intel.com> + * H Peter Anvin" <hpa@zytor.com> + * + * This driver allows to early upgrade microcode on Intel processors + * belonging to IA-32 family - PentiumPro, Pentium II, + * Pentium III, Xeon, Pentium 4, etc. + * + * Reference: Section 9.11 of Volume 3, IA-32 Intel Architecture + * Software Developer's Manual. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ +#include <linux/module.h> +#include <asm/microcode.h> +#include <asm/microcode_intel.h> +#include <asm/microcode_amd.h> +#include <asm/processor.h> +#include <asm/cmdline.h> + +#define QCHAR(a, b, c, d) ((a) + ((b) << 8) + ((c) << 16) + ((d) << 24)) +#define CPUID_INTEL1 QCHAR('G', 'e', 'n', 'u') +#define CPUID_INTEL2 QCHAR('i', 'n', 'e', 'I') +#define CPUID_INTEL3 QCHAR('n', 't', 'e', 'l') +#define CPUID_AMD1 QCHAR('A', 'u', 't', 'h') +#define CPUID_AMD2 QCHAR('e', 'n', 't', 'i') +#define CPUID_AMD3 QCHAR('c', 'A', 'M', 'D') + +#define CPUID_IS(a, b, c, ebx, ecx, edx) \ + (!((ebx ^ (a))|(edx ^ (b))|(ecx ^ (c)))) + +/* + * In early loading microcode phase on BSP, boot_cpu_data is not set up yet. + * x86_vendor() gets vendor id for BSP. + * + * In 32 bit AP case, accessing boot_cpu_data needs linear address. To simplify + * coding, we still use x86_vendor() to get vendor id for AP. + * + * x86_vendor() gets vendor information directly through cpuid. + */ +static int x86_vendor(void) +{ + u32 eax = 0x00000000; + u32 ebx, ecx = 0, edx; + + native_cpuid(&eax, &ebx, &ecx, &edx); + + if (CPUID_IS(CPUID_INTEL1, CPUID_INTEL2, CPUID_INTEL3, ebx, ecx, edx)) + return X86_VENDOR_INTEL; + + if (CPUID_IS(CPUID_AMD1, CPUID_AMD2, CPUID_AMD3, ebx, ecx, edx)) + return X86_VENDOR_AMD; + + return X86_VENDOR_UNKNOWN; +} + +static int x86_family(void) +{ + u32 eax = 0x00000001; + u32 ebx, ecx = 0, edx; + int x86; + + native_cpuid(&eax, &ebx, &ecx, &edx); + + x86 = (eax >> 8) & 0xf; + if (x86 == 15) + x86 += (eax >> 20) & 0xff; + + return x86; +} + +static bool __init check_loader_disabled_bsp(void) +{ +#ifdef CONFIG_X86_32 + const char *cmdline = (const char *)__pa_nodebug(boot_command_line); + const char *opt = "dis_ucode_ldr"; + const char *option = (const char *)__pa_nodebug(opt); + bool *res = (bool *)__pa_nodebug(&dis_ucode_ldr); + +#else /* CONFIG_X86_64 */ + const char *cmdline = boot_command_line; + const char *option = "dis_ucode_ldr"; + bool *res = &dis_ucode_ldr; +#endif + + if (cmdline_find_option_bool(cmdline, option)) + *res = true; + + return *res; +} + +void __init load_ucode_bsp(void) +{ + int vendor, x86; + + if (check_loader_disabled_bsp()) + return; + + if (!have_cpuid_p()) + return; + + vendor = x86_vendor(); + x86 = x86_family(); + + switch (vendor) { + case X86_VENDOR_INTEL: + if (x86 >= 6) + load_ucode_intel_bsp(); + break; + case X86_VENDOR_AMD: + if (x86 >= 0x10) + load_ucode_amd_bsp(); + break; + default: + break; + } +} + +static bool check_loader_disabled_ap(void) +{ +#ifdef CONFIG_X86_32 + return __pa_nodebug(dis_ucode_ldr); +#else + return dis_ucode_ldr; +#endif +} + +void load_ucode_ap(void) +{ + int vendor, x86; + + if (check_loader_disabled_ap()) + return; + + if (!have_cpuid_p()) + return; + + vendor = x86_vendor(); + x86 = x86_family(); + + switch (vendor) { + case X86_VENDOR_INTEL: + if (x86 >= 6) + load_ucode_intel_ap(); + break; + case X86_VENDOR_AMD: + if (x86 >= 0x10) + load_ucode_amd_ap(); + break; + default: + break; + } +} + +int __init save_microcode_in_initrd(void) +{ + struct cpuinfo_x86 *c = &boot_cpu_data; + + switch (c->x86_vendor) { + case X86_VENDOR_INTEL: + if (c->x86 >= 6) + save_microcode_in_initrd_intel(); + break; + case X86_VENDOR_AMD: + if (c->x86 >= 0x10) + save_microcode_in_initrd_amd(); + break; + default: + break; + } + + return 0; +} diff --git a/arch/x86/kernel/cpu/microcode/intel.c b/arch/x86/kernel/cpu/microcode/intel.c new file mode 100644 index 00000000000..a276fa75d9b --- /dev/null +++ b/arch/x86/kernel/cpu/microcode/intel.c @@ -0,0 +1,333 @@ +/* + * Intel CPU Microcode Update Driver for Linux + * + * Copyright (C) 2000-2006 Tigran Aivazian <tigran@aivazian.fsnet.co.uk> + * 2006 Shaohua Li <shaohua.li@intel.com> + * + * This driver allows to upgrade microcode on Intel processors + * belonging to IA-32 family - PentiumPro, Pentium II, + * Pentium III, Xeon, Pentium 4, etc. + * + * Reference: Section 8.11 of Volume 3a, IA-32 Intel? Architecture + * Software Developer's Manual + * Order Number 253668 or free download from: + * + * http://developer.intel.com/Assets/PDF/manual/253668.pdf + * + * For more information, go to http://www.urbanmyth.org/microcode + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + * + * 1.0 16 Feb 2000, Tigran Aivazian <tigran@sco.com> + * Initial release. + * 1.01 18 Feb 2000, Tigran Aivazian <tigran@sco.com> + * Added read() support + cleanups. + * 1.02 21 Feb 2000, Tigran Aivazian <tigran@sco.com> + * Added 'device trimming' support. open(O_WRONLY) zeroes + * and frees the saved copy of applied microcode. + * 1.03 29 Feb 2000, Tigran Aivazian <tigran@sco.com> + * Made to use devfs (/dev/cpu/microcode) + cleanups. + * 1.04 06 Jun 2000, Simon Trimmer <simon@veritas.com> + * Added misc device support (now uses both devfs and misc). + * Added MICROCODE_IOCFREE ioctl to clear memory. + * 1.05 09 Jun 2000, Simon Trimmer <simon@veritas.com> + * Messages for error cases (non Intel & no suitable microcode). + * 1.06 03 Aug 2000, Tigran Aivazian <tigran@veritas.com> + * Removed ->release(). Removed exclusive open and status bitmap. + * Added microcode_rwsem to serialize read()/write()/ioctl(). + * Removed global kernel lock usage. + * 1.07 07 Sep 2000, Tigran Aivazian <tigran@veritas.com> + * Write 0 to 0x8B msr and then cpuid before reading revision, + * so that it works even if there were no update done by the + * BIOS. Otherwise, reading from 0x8B gives junk (which happened + * to be 0 on my machine which is why it worked even when I + * disabled update by the BIOS) + * Thanks to Eric W. Biederman <ebiederman@lnxi.com> for the fix. + * 1.08 11 Dec 2000, Richard Schaal <richard.schaal@intel.com> and + * Tigran Aivazian <tigran@veritas.com> + * Intel Pentium 4 processor support and bugfixes. + * 1.09 30 Oct 2001, Tigran Aivazian <tigran@veritas.com> + * Bugfix for HT (Hyper-Threading) enabled processors + * whereby processor resources are shared by all logical processors + * in a single CPU package. + * 1.10 28 Feb 2002 Asit K Mallick <asit.k.mallick@intel.com> and + * Tigran Aivazian <tigran@veritas.com>, + * Serialize updates as required on HT processors due to + * speculative 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>, + * 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 + * in kernel memory. + * 1.14 25 Jun 2004 Tigran Aivazian <tigran@veritas.com> + * Fix sigmatch() macro to handle old CPUs with pf == 0. + * Thanks to Stuart Swales for pointing out this bug. + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/firmware.h> +#include <linux/uaccess.h> +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/vmalloc.h> + +#include <asm/microcode_intel.h> +#include <asm/processor.h> +#include <asm/msr.h> + +MODULE_DESCRIPTION("Microcode Update Driver"); +MODULE_AUTHOR("Tigran Aivazian <tigran@aivazian.fsnet.co.uk>"); +MODULE_LICENSE("GPL"); + +static int collect_cpu_info(int cpu_num, struct cpu_signature *csig) +{ + struct cpuinfo_x86 *c = &cpu_data(cpu_num); + unsigned int val[2]; + + memset(csig, 0, sizeof(*csig)); + + csig->sig = cpuid_eax(0x00000001); + + if ((c->x86_model >= 5) || (c->x86 > 6)) { + /* get processor flags from MSR 0x17 */ + rdmsr(MSR_IA32_PLATFORM_ID, val[0], val[1]); + csig->pf = 1 << ((val[1] >> 18) & 7); + } + + csig->rev = c->microcode; + pr_info("CPU%d sig=0x%x, pf=0x%x, revision=0x%x\n", + cpu_num, csig->sig, csig->pf, csig->rev); + + return 0; +} + +/* + * return 0 - no update found + * return 1 - found update + */ +static int get_matching_mc(struct microcode_intel *mc_intel, int cpu) +{ + struct cpu_signature cpu_sig; + unsigned int csig, cpf, crev; + + collect_cpu_info(cpu, &cpu_sig); + + csig = cpu_sig.sig; + cpf = cpu_sig.pf; + crev = cpu_sig.rev; + + return get_matching_microcode(csig, cpf, mc_intel, crev); +} + +int apply_microcode(int cpu) +{ + struct microcode_intel *mc_intel; + struct ucode_cpu_info *uci; + unsigned int val[2]; + int cpu_num = raw_smp_processor_id(); + struct cpuinfo_x86 *c = &cpu_data(cpu_num); + + uci = ucode_cpu_info + cpu; + mc_intel = uci->mc; + + /* We should bind the task to the CPU */ + BUG_ON(cpu_num != cpu); + + if (mc_intel == NULL) + return 0; + + /* + * Microcode on this CPU could be updated earlier. Only apply the + * microcode patch in mc_intel when it is newer than the one on this + * CPU. + */ + if (get_matching_mc(mc_intel, cpu) == 0) + return 0; + + /* write microcode via MSR 0x79 */ + wrmsr(MSR_IA32_UCODE_WRITE, + (unsigned long) mc_intel->bits, + (unsigned long) mc_intel->bits >> 16 >> 16); + wrmsr(MSR_IA32_UCODE_REV, 0, 0); + + /* As documented in the SDM: Do a CPUID 1 here */ + sync_core(); + + /* get the current revision from MSR 0x8B */ + rdmsr(MSR_IA32_UCODE_REV, val[0], val[1]); + + if (val[1] != mc_intel->hdr.rev) { + pr_err("CPU%d update to revision 0x%x failed\n", + cpu_num, mc_intel->hdr.rev); + return -1; + } + pr_info("CPU%d updated to revision 0x%x, date = %04x-%02x-%02x\n", + cpu_num, val[1], + mc_intel->hdr.date & 0xffff, + mc_intel->hdr.date >> 24, + (mc_intel->hdr.date >> 16) & 0xff); + + uci->cpu_sig.rev = val[1]; + c->microcode = val[1]; + + return 0; +} + +static enum ucode_state generic_load_microcode(int cpu, void *data, size_t size, + int (*get_ucode_data)(void *, const void *, size_t)) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + u8 *ucode_ptr = data, *new_mc = NULL, *mc = NULL; + int new_rev = uci->cpu_sig.rev; + unsigned int leftover = size; + enum ucode_state state = UCODE_OK; + unsigned int curr_mc_size = 0; + unsigned int csig, cpf; + + while (leftover) { + struct microcode_header_intel mc_header; + unsigned int mc_size; + + if (get_ucode_data(&mc_header, ucode_ptr, sizeof(mc_header))) + break; + + mc_size = get_totalsize(&mc_header); + if (!mc_size || mc_size > leftover) { + pr_err("error! Bad data in microcode data file\n"); + break; + } + + /* For performance reasons, reuse mc area when possible */ + if (!mc || mc_size > curr_mc_size) { + vfree(mc); + mc = vmalloc(mc_size); + if (!mc) + break; + curr_mc_size = mc_size; + } + + if (get_ucode_data(mc, ucode_ptr, mc_size) || + microcode_sanity_check(mc, 1) < 0) { + break; + } + + csig = uci->cpu_sig.sig; + cpf = uci->cpu_sig.pf; + if (get_matching_microcode(csig, cpf, mc, new_rev)) { + vfree(new_mc); + new_rev = mc_header.rev; + new_mc = mc; + mc = NULL; /* trigger new vmalloc */ + } + + ucode_ptr += mc_size; + leftover -= mc_size; + } + + vfree(mc); + + if (leftover) { + vfree(new_mc); + state = UCODE_ERROR; + goto out; + } + + if (!new_mc) { + state = UCODE_NFOUND; + goto out; + } + + vfree(uci->mc); + uci->mc = (struct microcode_intel *)new_mc; + + /* + * If early loading microcode is supported, save this mc into + * permanent memory. So it will be loaded early when a CPU is hot added + * or resumes. + */ + save_mc_for_early(new_mc); + + pr_debug("CPU%d found a matching microcode update with version 0x%x (current=0x%x)\n", + cpu, new_rev, uci->cpu_sig.rev); +out: + return state; +} + +static int get_ucode_fw(void *to, const void *from, size_t n) +{ + memcpy(to, from, n); + return 0; +} + +static enum ucode_state request_microcode_fw(int cpu, struct device *device, + bool refresh_fw) +{ + char name[30]; + struct cpuinfo_x86 *c = &cpu_data(cpu); + const struct firmware *firmware; + enum ucode_state ret; + + sprintf(name, "intel-ucode/%02x-%02x-%02x", + c->x86, c->x86_model, c->x86_mask); + + if (request_firmware_direct(&firmware, name, device)) { + pr_debug("data file %s load failed\n", name); + return UCODE_NFOUND; + } + + ret = generic_load_microcode(cpu, (void *)firmware->data, + firmware->size, &get_ucode_fw); + + release_firmware(firmware); + + return ret; +} + +static int get_ucode_user(void *to, const void *from, size_t n) +{ + return copy_from_user(to, from, n); +} + +static enum ucode_state +request_microcode_user(int cpu, const void __user *buf, size_t size) +{ + return generic_load_microcode(cpu, (void *)buf, size, &get_ucode_user); +} + +static void microcode_fini_cpu(int cpu) +{ + struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + + vfree(uci->mc); + uci->mc = NULL; +} + +static struct microcode_ops microcode_intel_ops = { + .request_microcode_user = request_microcode_user, + .request_microcode_fw = request_microcode_fw, + .collect_cpu_info = collect_cpu_info, + .apply_microcode = apply_microcode, + .microcode_fini_cpu = microcode_fini_cpu, +}; + +struct microcode_ops * __init init_intel_microcode(void) +{ + struct cpuinfo_x86 *c = &cpu_data(0); + + if (c->x86_vendor != X86_VENDOR_INTEL || c->x86 < 6 || + cpu_has(c, X86_FEATURE_IA64)) { + pr_err("Intel CPU family 0x%x not supported\n", c->x86); + return NULL; + } + + return µcode_intel_ops; +} + diff --git a/arch/x86/kernel/cpu/microcode/intel_early.c b/arch/x86/kernel/cpu/microcode/intel_early.c new file mode 100644 index 00000000000..18f739129e7 --- /dev/null +++ b/arch/x86/kernel/cpu/microcode/intel_early.c @@ -0,0 +1,787 @@ +/* + * Intel CPU microcode early update for Linux + * + * Copyright (C) 2012 Fenghua Yu <fenghua.yu@intel.com> + * H Peter Anvin" <hpa@zytor.com> + * + * This allows to early upgrade microcode on Intel processors + * belonging to IA-32 family - PentiumPro, Pentium II, + * Pentium III, Xeon, Pentium 4, etc. + * + * Reference: Section 9.11 of Volume 3, IA-32 Intel Architecture + * Software Developer's Manual. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ +#include <linux/module.h> +#include <linux/mm.h> +#include <linux/slab.h> +#include <linux/earlycpio.h> +#include <linux/initrd.h> +#include <linux/cpu.h> +#include <asm/msr.h> +#include <asm/microcode_intel.h> +#include <asm/processor.h> +#include <asm/tlbflush.h> +#include <asm/setup.h> + +unsigned long mc_saved_in_initrd[MAX_UCODE_COUNT]; +struct mc_saved_data { + unsigned int mc_saved_count; + struct microcode_intel **mc_saved; +} mc_saved_data; + +static enum ucode_state +generic_load_microcode_early(struct microcode_intel **mc_saved_p, + unsigned int mc_saved_count, + struct ucode_cpu_info *uci) +{ + struct microcode_intel *ucode_ptr, *new_mc = NULL; + int new_rev = uci->cpu_sig.rev; + enum ucode_state state = UCODE_OK; + unsigned int mc_size; + struct microcode_header_intel *mc_header; + unsigned int csig = uci->cpu_sig.sig; + unsigned int cpf = uci->cpu_sig.pf; + int i; + + for (i = 0; i < mc_saved_count; i++) { + ucode_ptr = mc_saved_p[i]; + + mc_header = (struct microcode_header_intel *)ucode_ptr; + mc_size = get_totalsize(mc_header); + if (get_matching_microcode(csig, cpf, ucode_ptr, new_rev)) { + new_rev = mc_header->rev; + new_mc = ucode_ptr; + } + } + + if (!new_mc) { + state = UCODE_NFOUND; + goto out; + } + + uci->mc = (struct microcode_intel *)new_mc; +out: + return state; +} + +static void +microcode_pointer(struct microcode_intel **mc_saved, + unsigned long *mc_saved_in_initrd, + unsigned long initrd_start, int mc_saved_count) +{ + int i; + + for (i = 0; i < mc_saved_count; i++) + mc_saved[i] = (struct microcode_intel *) + (mc_saved_in_initrd[i] + initrd_start); +} + +#ifdef CONFIG_X86_32 +static void +microcode_phys(struct microcode_intel **mc_saved_tmp, + struct mc_saved_data *mc_saved_data) +{ + int i; + struct microcode_intel ***mc_saved; + + mc_saved = (struct microcode_intel ***) + __pa_nodebug(&mc_saved_data->mc_saved); + for (i = 0; i < mc_saved_data->mc_saved_count; i++) { + struct microcode_intel *p; + + p = *(struct microcode_intel **) + __pa_nodebug(mc_saved_data->mc_saved + i); + mc_saved_tmp[i] = (struct microcode_intel *)__pa_nodebug(p); + } +} +#endif + +static enum ucode_state +load_microcode(struct mc_saved_data *mc_saved_data, + unsigned long *mc_saved_in_initrd, + unsigned long initrd_start, + struct ucode_cpu_info *uci) +{ + struct microcode_intel *mc_saved_tmp[MAX_UCODE_COUNT]; + unsigned int count = mc_saved_data->mc_saved_count; + + if (!mc_saved_data->mc_saved) { + microcode_pointer(mc_saved_tmp, mc_saved_in_initrd, + initrd_start, count); + + return generic_load_microcode_early(mc_saved_tmp, count, uci); + } else { +#ifdef CONFIG_X86_32 + microcode_phys(mc_saved_tmp, mc_saved_data); + return generic_load_microcode_early(mc_saved_tmp, count, uci); +#else + return generic_load_microcode_early(mc_saved_data->mc_saved, + count, uci); +#endif + } +} + +static u8 get_x86_family(unsigned long sig) +{ + u8 x86; + + x86 = (sig >> 8) & 0xf; + + if (x86 == 0xf) + x86 += (sig >> 20) & 0xff; + + return x86; +} + +static u8 get_x86_model(unsigned long sig) +{ + u8 x86, x86_model; + + x86 = get_x86_family(sig); + x86_model = (sig >> 4) & 0xf; + + if (x86 == 0x6 || x86 == 0xf) + x86_model += ((sig >> 16) & 0xf) << 4; + + return x86_model; +} + +/* + * Given CPU signature and a microcode patch, this function finds if the + * microcode patch has matching family and model with the CPU. + */ +static enum ucode_state +matching_model_microcode(struct microcode_header_intel *mc_header, + unsigned long sig) +{ + u8 x86, x86_model; + u8 x86_ucode, x86_model_ucode; + struct extended_sigtable *ext_header; + unsigned long total_size = get_totalsize(mc_header); + unsigned long data_size = get_datasize(mc_header); + int ext_sigcount, i; + struct extended_signature *ext_sig; + + x86 = get_x86_family(sig); + x86_model = get_x86_model(sig); + + x86_ucode = get_x86_family(mc_header->sig); + x86_model_ucode = get_x86_model(mc_header->sig); + + if (x86 == x86_ucode && x86_model == x86_model_ucode) + return UCODE_OK; + + /* Look for ext. headers: */ + if (total_size <= data_size + MC_HEADER_SIZE) + return UCODE_NFOUND; + + ext_header = (struct extended_sigtable *) + mc_header + data_size + MC_HEADER_SIZE; + ext_sigcount = ext_header->count; + ext_sig = (void *)ext_header + EXT_HEADER_SIZE; + + for (i = 0; i < ext_sigcount; i++) { + x86_ucode = get_x86_family(ext_sig->sig); + x86_model_ucode = get_x86_model(ext_sig->sig); + + if (x86 == x86_ucode && x86_model == x86_model_ucode) + return UCODE_OK; + + ext_sig++; + } + + return UCODE_NFOUND; +} + +static int +save_microcode(struct mc_saved_data *mc_saved_data, + struct microcode_intel **mc_saved_src, + unsigned int mc_saved_count) +{ + int i, j; + struct microcode_intel **mc_saved_p; + int ret; + + if (!mc_saved_count) + return -EINVAL; + + /* + * Copy new microcode data. + */ + mc_saved_p = kmalloc(mc_saved_count*sizeof(struct microcode_intel *), + GFP_KERNEL); + if (!mc_saved_p) + return -ENOMEM; + + for (i = 0; i < mc_saved_count; i++) { + struct microcode_intel *mc = mc_saved_src[i]; + struct microcode_header_intel *mc_header = &mc->hdr; + unsigned long mc_size = get_totalsize(mc_header); + mc_saved_p[i] = kmalloc(mc_size, GFP_KERNEL); + if (!mc_saved_p[i]) { + ret = -ENOMEM; + goto err; + } + if (!mc_saved_src[i]) { + ret = -EINVAL; + goto err; + } + memcpy(mc_saved_p[i], mc, mc_size); + } + + /* + * Point to newly saved microcode. + */ + mc_saved_data->mc_saved = mc_saved_p; + mc_saved_data->mc_saved_count = mc_saved_count; + + return 0; + +err: + for (j = 0; j <= i; j++) + kfree(mc_saved_p[j]); + kfree(mc_saved_p); + + return ret; +} + +/* + * A microcode patch in ucode_ptr is saved into mc_saved + * - if it has matching signature and newer revision compared to an existing + * patch mc_saved. + * - or if it is a newly discovered microcode patch. + * + * The microcode patch should have matching model with CPU. + */ +static void _save_mc(struct microcode_intel **mc_saved, u8 *ucode_ptr, + unsigned int *mc_saved_count_p) +{ + int i; + int found = 0; + unsigned int mc_saved_count = *mc_saved_count_p; + struct microcode_header_intel *mc_header; + + mc_header = (struct microcode_header_intel *)ucode_ptr; + for (i = 0; i < mc_saved_count; i++) { + unsigned int sig, pf; + unsigned int new_rev; + struct microcode_header_intel *mc_saved_header = + (struct microcode_header_intel *)mc_saved[i]; + sig = mc_saved_header->sig; + pf = mc_saved_header->pf; + new_rev = mc_header->rev; + + if (get_matching_sig(sig, pf, ucode_ptr, new_rev)) { + found = 1; + if (update_match_revision(mc_header, new_rev)) { + /* + * Found an older ucode saved before. + * Replace the older one with this newer + * one. + */ + mc_saved[i] = + (struct microcode_intel *)ucode_ptr; + break; + } + } + } + if (i >= mc_saved_count && !found) + /* + * This ucode is first time discovered in ucode file. + * Save it to memory. + */ + mc_saved[mc_saved_count++] = + (struct microcode_intel *)ucode_ptr; + + *mc_saved_count_p = mc_saved_count; +} + +/* + * Get microcode matching with BSP's model. Only CPUs with the same model as + * BSP can stay in the platform. + */ +static enum ucode_state __init +get_matching_model_microcode(int cpu, unsigned long start, + void *data, size_t size, + struct mc_saved_data *mc_saved_data, + unsigned long *mc_saved_in_initrd, + struct ucode_cpu_info *uci) +{ + u8 *ucode_ptr = data; + unsigned int leftover = size; + enum ucode_state state = UCODE_OK; + unsigned int mc_size; + struct microcode_header_intel *mc_header; + struct microcode_intel *mc_saved_tmp[MAX_UCODE_COUNT]; + unsigned int mc_saved_count = mc_saved_data->mc_saved_count; + int i; + + while (leftover) { + mc_header = (struct microcode_header_intel *)ucode_ptr; + + mc_size = get_totalsize(mc_header); + if (!mc_size || mc_size > leftover || + microcode_sanity_check(ucode_ptr, 0) < 0) + break; + + leftover -= mc_size; + + /* + * Since APs with same family and model as the BSP may boot in + * the platform, we need to find and save microcode patches + * with the same family and model as the BSP. + */ + if (matching_model_microcode(mc_header, uci->cpu_sig.sig) != + UCODE_OK) { + ucode_ptr += mc_size; + continue; + } + + _save_mc(mc_saved_tmp, ucode_ptr, &mc_saved_count); + + ucode_ptr += mc_size; + } + + if (leftover) { + state = UCODE_ERROR; + goto out; + } + + if (mc_saved_count == 0) { + state = UCODE_NFOUND; + goto out; + } + + for (i = 0; i < mc_saved_count; i++) + mc_saved_in_initrd[i] = (unsigned long)mc_saved_tmp[i] - start; + + mc_saved_data->mc_saved_count = mc_saved_count; +out: + return state; +} + +static int collect_cpu_info_early(struct ucode_cpu_info *uci) +{ + unsigned int val[2]; + u8 x86, x86_model; + struct cpu_signature csig; + unsigned int eax, ebx, ecx, edx; + + csig.sig = 0; + csig.pf = 0; + csig.rev = 0; + + memset(uci, 0, sizeof(*uci)); + + eax = 0x00000001; + ecx = 0; + native_cpuid(&eax, &ebx, &ecx, &edx); + csig.sig = eax; + + x86 = get_x86_family(csig.sig); + x86_model = get_x86_model(csig.sig); + + if ((x86_model >= 5) || (x86 > 6)) { + /* get processor flags from MSR 0x17 */ + native_rdmsr(MSR_IA32_PLATFORM_ID, val[0], val[1]); + csig.pf = 1 << ((val[1] >> 18) & 7); + } + native_wrmsr(MSR_IA32_UCODE_REV, 0, 0); + + /* As documented in the SDM: Do a CPUID 1 here */ + sync_core(); + + /* get the current revision from MSR 0x8B */ + native_rdmsr(MSR_IA32_UCODE_REV, val[0], val[1]); + + csig.rev = val[1]; + + uci->cpu_sig = csig; + uci->valid = 1; + + return 0; +} + +#ifdef DEBUG +static void __ref show_saved_mc(void) +{ + int i, j; + unsigned int sig, pf, rev, total_size, data_size, date; + struct ucode_cpu_info uci; + + if (mc_saved_data.mc_saved_count == 0) { + pr_debug("no micorcode data saved.\n"); + return; + } + pr_debug("Total microcode saved: %d\n", mc_saved_data.mc_saved_count); + + collect_cpu_info_early(&uci); + + sig = uci.cpu_sig.sig; + pf = uci.cpu_sig.pf; + rev = uci.cpu_sig.rev; + pr_debug("CPU%d: sig=0x%x, pf=0x%x, rev=0x%x\n", + smp_processor_id(), sig, pf, rev); + + for (i = 0; i < mc_saved_data.mc_saved_count; i++) { + struct microcode_header_intel *mc_saved_header; + struct extended_sigtable *ext_header; + int ext_sigcount; + struct extended_signature *ext_sig; + + mc_saved_header = (struct microcode_header_intel *) + mc_saved_data.mc_saved[i]; + sig = mc_saved_header->sig; + pf = mc_saved_header->pf; + rev = mc_saved_header->rev; + total_size = get_totalsize(mc_saved_header); + data_size = get_datasize(mc_saved_header); + date = mc_saved_header->date; + + pr_debug("mc_saved[%d]: sig=0x%x, pf=0x%x, rev=0x%x, toal size=0x%x, date = %04x-%02x-%02x\n", + i, sig, pf, rev, total_size, + date & 0xffff, + date >> 24, + (date >> 16) & 0xff); + + /* Look for ext. headers: */ + if (total_size <= data_size + MC_HEADER_SIZE) + continue; + + ext_header = (struct extended_sigtable *) + mc_saved_header + data_size + MC_HEADER_SIZE; + ext_sigcount = ext_header->count; + ext_sig = (void *)ext_header + EXT_HEADER_SIZE; + + for (j = 0; j < ext_sigcount; j++) { + sig = ext_sig->sig; + pf = ext_sig->pf; + + pr_debug("\tExtended[%d]: sig=0x%x, pf=0x%x\n", + j, sig, pf); + + ext_sig++; + } + + } +} +#else +static inline void show_saved_mc(void) +{ +} +#endif + +#if defined(CONFIG_MICROCODE_INTEL_EARLY) && defined(CONFIG_HOTPLUG_CPU) +static DEFINE_MUTEX(x86_cpu_microcode_mutex); +/* + * Save this mc into mc_saved_data. So it will be loaded early when a CPU is + * hot added or resumes. + * + * Please make sure this mc should be a valid microcode patch before calling + * this function. + */ +int save_mc_for_early(u8 *mc) +{ + struct microcode_intel *mc_saved_tmp[MAX_UCODE_COUNT]; + unsigned int mc_saved_count_init; + unsigned int mc_saved_count; + struct microcode_intel **mc_saved; + int ret = 0; + int i; + + /* + * Hold hotplug lock so mc_saved_data is not accessed by a CPU in + * hotplug. + */ + mutex_lock(&x86_cpu_microcode_mutex); + + mc_saved_count_init = mc_saved_data.mc_saved_count; + mc_saved_count = mc_saved_data.mc_saved_count; + mc_saved = mc_saved_data.mc_saved; + + if (mc_saved && mc_saved_count) + memcpy(mc_saved_tmp, mc_saved, + mc_saved_count * sizeof(struct mirocode_intel *)); + /* + * Save the microcode patch mc in mc_save_tmp structure if it's a newer + * version. + */ + + _save_mc(mc_saved_tmp, mc, &mc_saved_count); + + /* + * Save the mc_save_tmp in global mc_saved_data. + */ + ret = save_microcode(&mc_saved_data, mc_saved_tmp, mc_saved_count); + if (ret) { + pr_err("Cannot save microcode patch.\n"); + goto out; + } + + show_saved_mc(); + + /* + * Free old saved microcod data. + */ + if (mc_saved) { + for (i = 0; i < mc_saved_count_init; i++) + kfree(mc_saved[i]); + kfree(mc_saved); + } + +out: + mutex_unlock(&x86_cpu_microcode_mutex); + + return ret; +} +EXPORT_SYMBOL_GPL(save_mc_for_early); +#endif + +static __initdata char ucode_name[] = "kernel/x86/microcode/GenuineIntel.bin"; +static __init enum ucode_state +scan_microcode(unsigned long start, unsigned long end, + struct mc_saved_data *mc_saved_data, + unsigned long *mc_saved_in_initrd, + struct ucode_cpu_info *uci) +{ + unsigned int size = end - start + 1; + struct cpio_data cd; + long offset = 0; +#ifdef CONFIG_X86_32 + char *p = (char *)__pa_nodebug(ucode_name); +#else + char *p = ucode_name; +#endif + + cd.data = NULL; + cd.size = 0; + + cd = find_cpio_data(p, (void *)start, size, &offset); + if (!cd.data) + return UCODE_ERROR; + + + return get_matching_model_microcode(0, start, cd.data, cd.size, + mc_saved_data, mc_saved_in_initrd, + uci); +} + +/* + * Print ucode update info. + */ +static void +print_ucode_info(struct ucode_cpu_info *uci, unsigned int date) +{ + int cpu = smp_processor_id(); + + pr_info("CPU%d microcode updated early to revision 0x%x, date = %04x-%02x-%02x\n", + cpu, + uci->cpu_sig.rev, + date & 0xffff, + date >> 24, + (date >> 16) & 0xff); +} + +#ifdef CONFIG_X86_32 + +static int delay_ucode_info; +static int current_mc_date; + +/* + * Print early updated ucode info after printk works. This is delayed info dump. + */ +void show_ucode_info_early(void) +{ + struct ucode_cpu_info uci; + + if (delay_ucode_info) { + collect_cpu_info_early(&uci); + print_ucode_info(&uci, current_mc_date); + delay_ucode_info = 0; + } +} + +/* + * At this point, we can not call printk() yet. Keep microcode patch number in + * mc_saved_data.mc_saved and delay printing microcode info in + * show_ucode_info_early() until printk() works. + */ +static void print_ucode(struct ucode_cpu_info *uci) +{ + struct microcode_intel *mc_intel; + int *delay_ucode_info_p; + int *current_mc_date_p; + + mc_intel = uci->mc; + if (mc_intel == NULL) + return; + + delay_ucode_info_p = (int *)__pa_nodebug(&delay_ucode_info); + current_mc_date_p = (int *)__pa_nodebug(¤t_mc_date); + + *delay_ucode_info_p = 1; + *current_mc_date_p = mc_intel->hdr.date; +} +#else + +/* + * Flush global tlb. We only do this in x86_64 where paging has been enabled + * already and PGE should be enabled as well. + */ +static inline void flush_tlb_early(void) +{ + __native_flush_tlb_global_irq_disabled(); +} + +static inline void print_ucode(struct ucode_cpu_info *uci) +{ + struct microcode_intel *mc_intel; + + mc_intel = uci->mc; + if (mc_intel == NULL) + return; + + print_ucode_info(uci, mc_intel->hdr.date); +} +#endif + +static int apply_microcode_early(struct mc_saved_data *mc_saved_data, + struct ucode_cpu_info *uci) +{ + struct microcode_intel *mc_intel; + unsigned int val[2]; + + mc_intel = uci->mc; + if (mc_intel == NULL) + return 0; + + /* write microcode via MSR 0x79 */ + native_wrmsr(MSR_IA32_UCODE_WRITE, + (unsigned long) mc_intel->bits, + (unsigned long) mc_intel->bits >> 16 >> 16); + native_wrmsr(MSR_IA32_UCODE_REV, 0, 0); + + /* As documented in the SDM: Do a CPUID 1 here */ + sync_core(); + + /* get the current revision from MSR 0x8B */ + native_rdmsr(MSR_IA32_UCODE_REV, val[0], val[1]); + if (val[1] != mc_intel->hdr.rev) + return -1; + +#ifdef CONFIG_X86_64 + /* Flush global tlb. This is precaution. */ + flush_tlb_early(); +#endif + uci->cpu_sig.rev = val[1]; + + print_ucode(uci); + + return 0; +} + +/* + * This function converts microcode patch offsets previously stored in + * mc_saved_in_initrd to pointers and stores the pointers in mc_saved_data. + */ +int __init save_microcode_in_initrd_intel(void) +{ + unsigned int count = mc_saved_data.mc_saved_count; + struct microcode_intel *mc_saved[MAX_UCODE_COUNT]; + int ret = 0; + + if (count == 0) + return ret; + + microcode_pointer(mc_saved, mc_saved_in_initrd, initrd_start, count); + ret = save_microcode(&mc_saved_data, mc_saved, count); + if (ret) + pr_err("Cannot save microcode patches from initrd.\n"); + + show_saved_mc(); + + return ret; +} + +static void __init +_load_ucode_intel_bsp(struct mc_saved_data *mc_saved_data, + unsigned long *mc_saved_in_initrd, + unsigned long initrd_start_early, + unsigned long initrd_end_early, + struct ucode_cpu_info *uci) +{ + collect_cpu_info_early(uci); + scan_microcode(initrd_start_early, initrd_end_early, mc_saved_data, + mc_saved_in_initrd, uci); + load_microcode(mc_saved_data, mc_saved_in_initrd, + initrd_start_early, uci); + apply_microcode_early(mc_saved_data, uci); +} + +void __init +load_ucode_intel_bsp(void) +{ + u64 ramdisk_image, ramdisk_size; + unsigned long initrd_start_early, initrd_end_early; + struct ucode_cpu_info uci; +#ifdef CONFIG_X86_32 + struct boot_params *boot_params_p; + + boot_params_p = (struct boot_params *)__pa_nodebug(&boot_params); + ramdisk_image = boot_params_p->hdr.ramdisk_image; + ramdisk_size = boot_params_p->hdr.ramdisk_size; + initrd_start_early = ramdisk_image; + initrd_end_early = initrd_start_early + ramdisk_size; + + _load_ucode_intel_bsp( + (struct mc_saved_data *)__pa_nodebug(&mc_saved_data), + (unsigned long *)__pa_nodebug(&mc_saved_in_initrd), + initrd_start_early, initrd_end_early, &uci); +#else + ramdisk_image = boot_params.hdr.ramdisk_image; + ramdisk_size = boot_params.hdr.ramdisk_size; + initrd_start_early = ramdisk_image + PAGE_OFFSET; + initrd_end_early = initrd_start_early + ramdisk_size; + + _load_ucode_intel_bsp(&mc_saved_data, mc_saved_in_initrd, + initrd_start_early, initrd_end_early, &uci); +#endif +} + +void load_ucode_intel_ap(void) +{ + struct mc_saved_data *mc_saved_data_p; + struct ucode_cpu_info uci; + unsigned long *mc_saved_in_initrd_p; + unsigned long initrd_start_addr; +#ifdef CONFIG_X86_32 + unsigned long *initrd_start_p; + + mc_saved_in_initrd_p = + (unsigned long *)__pa_nodebug(mc_saved_in_initrd); + mc_saved_data_p = (struct mc_saved_data *)__pa_nodebug(&mc_saved_data); + initrd_start_p = (unsigned long *)__pa_nodebug(&initrd_start); + initrd_start_addr = (unsigned long)__pa_nodebug(*initrd_start_p); +#else + mc_saved_data_p = &mc_saved_data; + mc_saved_in_initrd_p = mc_saved_in_initrd; + initrd_start_addr = initrd_start; +#endif + + /* + * If there is no valid ucode previously saved in memory, no need to + * update ucode on this AP. + */ + if (mc_saved_data_p->mc_saved_count == 0) + return; + + collect_cpu_info_early(&uci); + load_microcode(mc_saved_data_p, mc_saved_in_initrd_p, + initrd_start_addr, &uci); + apply_microcode_early(mc_saved_data_p, &uci); +} diff --git a/arch/x86/kernel/cpu/microcode/intel_lib.c b/arch/x86/kernel/cpu/microcode/intel_lib.c new file mode 100644 index 00000000000..ce69320d017 --- /dev/null +++ b/arch/x86/kernel/cpu/microcode/intel_lib.c @@ -0,0 +1,174 @@ +/* + * Intel CPU Microcode Update Driver for Linux + * + * Copyright (C) 2012 Fenghua Yu <fenghua.yu@intel.com> + * H Peter Anvin" <hpa@zytor.com> + * + * This driver allows to upgrade microcode on Intel processors + * belonging to IA-32 family - PentiumPro, Pentium II, + * Pentium III, Xeon, Pentium 4, etc. + * + * Reference: Section 8.11 of Volume 3a, IA-32 Intel? Architecture + * Software Developer's Manual + * Order Number 253668 or free download from: + * + * http://developer.intel.com/Assets/PDF/manual/253668.pdf + * + * For more information, go to http://www.urbanmyth.org/microcode + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + * + */ +#include <linux/firmware.h> +#include <linux/uaccess.h> +#include <linux/kernel.h> +#include <linux/module.h> + +#include <asm/microcode_intel.h> +#include <asm/processor.h> +#include <asm/msr.h> + +static inline int +update_match_cpu(unsigned int csig, unsigned int cpf, + unsigned int sig, unsigned int pf) +{ + return (!sigmatch(sig, csig, pf, cpf)) ? 0 : 1; +} + +int +update_match_revision(struct microcode_header_intel *mc_header, int rev) +{ + return (mc_header->rev <= rev) ? 0 : 1; +} + +int microcode_sanity_check(void *mc, int print_err) +{ + unsigned long total_size, data_size, ext_table_size; + struct microcode_header_intel *mc_header = mc; + struct extended_sigtable *ext_header = NULL; + int sum, orig_sum, ext_sigcount = 0, i; + struct extended_signature *ext_sig; + + total_size = get_totalsize(mc_header); + data_size = get_datasize(mc_header); + + if (data_size + MC_HEADER_SIZE > total_size) { + if (print_err) + pr_err("error! Bad data size in microcode data file\n"); + return -EINVAL; + } + + if (mc_header->ldrver != 1 || mc_header->hdrver != 1) { + if (print_err) + pr_err("error! Unknown microcode update format\n"); + return -EINVAL; + } + ext_table_size = total_size - (MC_HEADER_SIZE + data_size); + if (ext_table_size) { + if ((ext_table_size < EXT_HEADER_SIZE) + || ((ext_table_size - EXT_HEADER_SIZE) % EXT_SIGNATURE_SIZE)) { + if (print_err) + pr_err("error! Small exttable size in microcode data file\n"); + return -EINVAL; + } + ext_header = mc + MC_HEADER_SIZE + data_size; + if (ext_table_size != exttable_size(ext_header)) { + if (print_err) + pr_err("error! Bad exttable size in microcode data file\n"); + return -EFAULT; + } + ext_sigcount = ext_header->count; + } + + /* check extended table checksum */ + if (ext_table_size) { + int ext_table_sum = 0; + int *ext_tablep = (int *)ext_header; + + i = ext_table_size / DWSIZE; + while (i--) + ext_table_sum += ext_tablep[i]; + if (ext_table_sum) { + if (print_err) + pr_warn("aborting, bad extended signature table checksum\n"); + return -EINVAL; + } + } + + /* calculate the checksum */ + orig_sum = 0; + i = (MC_HEADER_SIZE + data_size) / DWSIZE; + while (i--) + orig_sum += ((int *)mc)[i]; + if (orig_sum) { + if (print_err) + pr_err("aborting, bad checksum\n"); + return -EINVAL; + } + if (!ext_table_size) + return 0; + /* check extended signature checksum */ + for (i = 0; i < ext_sigcount; i++) { + ext_sig = (void *)ext_header + EXT_HEADER_SIZE + + EXT_SIGNATURE_SIZE * i; + sum = orig_sum + - (mc_header->sig + mc_header->pf + mc_header->cksum) + + (ext_sig->sig + ext_sig->pf + ext_sig->cksum); + if (sum) { + if (print_err) + pr_err("aborting, bad checksum\n"); + return -EINVAL; + } + } + return 0; +} +EXPORT_SYMBOL_GPL(microcode_sanity_check); + +/* + * return 0 - no update found + * return 1 - found update + */ +int get_matching_sig(unsigned int csig, int cpf, void *mc, int rev) +{ + struct microcode_header_intel *mc_header = mc; + struct extended_sigtable *ext_header; + unsigned long total_size = get_totalsize(mc_header); + int ext_sigcount, i; + struct extended_signature *ext_sig; + + if (update_match_cpu(csig, cpf, mc_header->sig, mc_header->pf)) + return 1; + + /* Look for ext. headers: */ + if (total_size <= get_datasize(mc_header) + MC_HEADER_SIZE) + return 0; + + ext_header = mc + get_datasize(mc_header) + MC_HEADER_SIZE; + ext_sigcount = ext_header->count; + ext_sig = (void *)ext_header + EXT_HEADER_SIZE; + + for (i = 0; i < ext_sigcount; i++) { + if (update_match_cpu(csig, cpf, ext_sig->sig, ext_sig->pf)) + return 1; + ext_sig++; + } + return 0; +} + +/* + * return 0 - no update found + * return 1 - found update + */ +int get_matching_microcode(unsigned int csig, int cpf, void *mc, int rev) +{ + struct microcode_header_intel *mc_header = mc; + + if (!update_match_revision(mc_header, rev)) + return 0; + + return get_matching_sig(csig, cpf, mc, rev); +} +EXPORT_SYMBOL_GPL(get_matching_microcode); diff --git a/arch/x86/kernel/cpu/mshyperv.c b/arch/x86/kernel/cpu/mshyperv.c index 71a39f3621b..a450373e8e9 100644 --- a/arch/x86/kernel/cpu/mshyperv.c +++ b/arch/x86/kernel/cpu/mshyperv.c @@ -15,7 +15,9 @@ #include <linux/clocksource.h> #include <linux/module.h> #include <linux/hardirq.h> +#include <linux/efi.h> #include <linux/interrupt.h> +#include <linux/irq.h> #include <asm/processor.h> #include <asm/hypervisor.h> #include <asm/hyperv.h> @@ -23,10 +25,52 @@ #include <asm/desc.h> #include <asm/idle.h> #include <asm/irq_regs.h> +#include <asm/i8259.h> +#include <asm/apic.h> +#include <asm/timer.h> struct ms_hyperv_info ms_hyperv; EXPORT_SYMBOL_GPL(ms_hyperv); +#if IS_ENABLED(CONFIG_HYPERV) +static void (*vmbus_handler)(void); + +void hyperv_vector_handler(struct pt_regs *regs) +{ + struct pt_regs *old_regs = set_irq_regs(regs); + + irq_enter(); + exit_idle(); + + inc_irq_stat(irq_hv_callback_count); + if (vmbus_handler) + vmbus_handler(); + + irq_exit(); + set_irq_regs(old_regs); +} + +void hv_setup_vmbus_irq(void (*handler)(void)) +{ + vmbus_handler = handler; + /* + * Setup the IDT for hypervisor callback. Prevent reallocation + * at module reload. + */ + if (!test_bit(HYPERVISOR_CALLBACK_VECTOR, used_vectors)) + alloc_intr_gate(HYPERVISOR_CALLBACK_VECTOR, + hyperv_callback_vector); +} + +void hv_remove_vmbus_irq(void) +{ + /* We have no way to deallocate the interrupt gate */ + vmbus_handler = NULL; +} +EXPORT_SYMBOL_GPL(hv_setup_vmbus_irq); +EXPORT_SYMBOL_GPL(hv_remove_vmbus_irq); +#endif + static uint32_t __init ms_hyperv_platform(void) { u32 eax; @@ -76,8 +120,28 @@ static void __init ms_hyperv_init_platform(void) printk(KERN_INFO "HyperV: features 0x%x, hints 0x%x\n", ms_hyperv.features, ms_hyperv.hints); +#ifdef CONFIG_X86_LOCAL_APIC + if (ms_hyperv.features & HV_X64_MSR_APIC_FREQUENCY_AVAILABLE) { + /* + * Get the APIC frequency. + */ + u64 hv_lapic_frequency; + + rdmsrl(HV_X64_MSR_APIC_FREQUENCY, hv_lapic_frequency); + hv_lapic_frequency = div_u64(hv_lapic_frequency, HZ); + lapic_timer_frequency = hv_lapic_frequency; + printk(KERN_INFO "HyperV: LAPIC Timer Frequency: %#x\n", + lapic_timer_frequency); + } +#endif + if (ms_hyperv.features & HV_X64_MSR_TIME_REF_COUNT_AVAILABLE) clocksource_register_hz(&hyperv_cs, NSEC_PER_SEC/100); + +#ifdef CONFIG_X86_IO_APIC + no_timer_check = 1; +#endif + } const __refconst struct hypervisor_x86 x86_hyper_ms_hyperv = { @@ -86,41 +150,3 @@ const __refconst struct hypervisor_x86 x86_hyper_ms_hyperv = { .init_platform = ms_hyperv_init_platform, }; EXPORT_SYMBOL(x86_hyper_ms_hyperv); - -#if IS_ENABLED(CONFIG_HYPERV) -static int vmbus_irq = -1; -static irq_handler_t vmbus_isr; - -void hv_register_vmbus_handler(int irq, irq_handler_t handler) -{ - /* - * Setup the IDT for hypervisor callback. - */ - alloc_intr_gate(HYPERVISOR_CALLBACK_VECTOR, hyperv_callback_vector); - - vmbus_irq = irq; - vmbus_isr = handler; -} - -void hyperv_vector_handler(struct pt_regs *regs) -{ - struct pt_regs *old_regs = set_irq_regs(regs); - struct irq_desc *desc; - - irq_enter(); - exit_idle(); - - desc = irq_to_desc(vmbus_irq); - - if (desc) - generic_handle_irq_desc(vmbus_irq, desc); - - irq_exit(); - set_irq_regs(old_regs); -} -#else -void hv_register_vmbus_handler(int irq, irq_handler_t handler) -{ -} -#endif -EXPORT_SYMBOL_GPL(hv_register_vmbus_handler); diff --git a/arch/x86/kernel/cpu/mtrr/generic.c b/arch/x86/kernel/cpu/mtrr/generic.c index ce2d0a2c3e4..0e25a1bc5ab 100644 --- a/arch/x86/kernel/cpu/mtrr/generic.c +++ b/arch/x86/kernel/cpu/mtrr/generic.c @@ -683,7 +683,7 @@ static void prepare_set(void) __acquires(set_atomicity_lock) } /* Flush all TLBs via a mov %cr3, %reg; mov %reg, %cr3 */ - count_vm_event(NR_TLB_LOCAL_FLUSH_ALL); + count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ALL); __flush_tlb(); /* Save MTRR state */ @@ -697,7 +697,7 @@ static void prepare_set(void) __acquires(set_atomicity_lock) static void post_set(void) __releases(set_atomicity_lock) { /* Flush TLBs (no need to flush caches - they are disabled) */ - count_vm_event(NR_TLB_LOCAL_FLUSH_ALL); + count_vm_tlb_event(NR_TLB_LOCAL_FLUSH_ALL); __flush_tlb(); /* Intel (P6) standard MTRRs */ diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 897783b3302..2879ecdaac4 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -118,6 +118,9 @@ static int x86_pmu_extra_regs(u64 config, struct perf_event *event) continue; if (event->attr.config1 & ~er->valid_mask) return -EINVAL; + /* Check if the extra msrs can be safely accessed*/ + if (!er->extra_msr_access) + return -ENXIO; reg->idx = er->idx; reg->config = event->attr.config1; @@ -303,15 +306,6 @@ int x86_setup_perfctr(struct perf_event *event) hwc->sample_period = x86_pmu.max_period; hwc->last_period = hwc->sample_period; local64_set(&hwc->period_left, hwc->sample_period); - } else { - /* - * If we have a PMU initialized but no APIC - * interrupts, we cannot sample hardware - * events (user-space has to fall back and - * sample via a hrtimer based software event): - */ - if (!x86_pmu.apic) - return -EOPNOTSUPP; } if (attr->type == PERF_TYPE_RAW) @@ -721,6 +715,7 @@ int perf_assign_events(struct perf_event **events, int n, return sched.state.unassigned; } +EXPORT_SYMBOL_GPL(perf_assign_events); int x86_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign) { @@ -892,7 +887,6 @@ static void x86_pmu_enable(struct pmu *pmu) * hw_perf_group_sched_in() or x86_pmu_enable() * * step1: save events moving to new counters - * step2: reprogram moved events into new counters */ for (i = 0; i < n_running; i++) { event = cpuc->event_list[i]; @@ -918,6 +912,9 @@ static void x86_pmu_enable(struct pmu *pmu) x86_pmu_stop(event, PERF_EF_UPDATE); } + /* + * step2: reprogram moved events into new counters + */ for (i = 0; i < cpuc->n_events; i++) { event = cpuc->event_list[i]; hwc = &event->hw; @@ -1043,7 +1040,7 @@ static int x86_pmu_add(struct perf_event *event, int flags) /* * If group events scheduling transaction was started, * skip the schedulability test here, it will be performed - * at commit time (->commit_txn) as a whole + * at commit time (->commit_txn) as a whole. */ if (cpuc->group_flag & PERF_EVENT_TXN) goto done_collect; @@ -1058,6 +1055,10 @@ static int x86_pmu_add(struct perf_event *event, int flags) memcpy(cpuc->assign, assign, n*sizeof(int)); done_collect: + /* + * Commit the collect_events() state. See x86_pmu_del() and + * x86_pmu_*_txn(). + */ cpuc->n_events = n; cpuc->n_added += n - n0; cpuc->n_txn += n - n0; @@ -1183,25 +1184,38 @@ static void x86_pmu_del(struct perf_event *event, int flags) * If we're called during a txn, we don't need to do anything. * The events never got scheduled and ->cancel_txn will truncate * the event_list. + * + * XXX assumes any ->del() called during a TXN will only be on + * an event added during that same TXN. */ if (cpuc->group_flag & PERF_EVENT_TXN) return; + /* + * Not a TXN, therefore cleanup properly. + */ x86_pmu_stop(event, PERF_EF_UPDATE); for (i = 0; i < cpuc->n_events; i++) { - if (event == cpuc->event_list[i]) { + if (event == cpuc->event_list[i]) + break; + } - if (x86_pmu.put_event_constraints) - x86_pmu.put_event_constraints(cpuc, event); + if (WARN_ON_ONCE(i == cpuc->n_events)) /* called ->del() without ->add() ? */ + return; - while (++i < cpuc->n_events) - cpuc->event_list[i-1] = cpuc->event_list[i]; + /* If we have a newly added event; make sure to decrease n_added. */ + if (i >= cpuc->n_events - cpuc->n_added) + --cpuc->n_added; + + if (x86_pmu.put_event_constraints) + x86_pmu.put_event_constraints(cpuc, event); + + /* Delete the array entry. */ + while (++i < cpuc->n_events) + cpuc->event_list[i-1] = cpuc->event_list[i]; + --cpuc->n_events; - --cpuc->n_events; - break; - } - } perf_event_update_userpage(event); } @@ -1273,24 +1287,25 @@ void perf_events_lapic_init(void) apic_write(APIC_LVTPC, APIC_DM_NMI); } -static int __kprobes +static int perf_event_nmi_handler(unsigned int cmd, struct pt_regs *regs) { - int ret; u64 start_clock; u64 finish_clock; + int ret; if (!atomic_read(&active_events)) return NMI_DONE; - start_clock = local_clock(); + start_clock = sched_clock(); ret = x86_pmu.handle_irq(regs); - finish_clock = local_clock(); + finish_clock = sched_clock(); perf_sample_event_took(finish_clock - start_clock); return ret; } +NOKPROBE_SYMBOL(perf_event_nmi_handler); struct event_constraint emptyconstraint; struct event_constraint unconstrained; @@ -1346,6 +1361,15 @@ static void __init pmu_check_apic(void) x86_pmu.apic = 0; pr_info("no APIC, boot with the \"lapic\" boot parameter to force-enable it.\n"); pr_info("no hardware sampling interrupt available.\n"); + + /* + * If we have a PMU initialized but no APIC + * interrupts, we cannot sample hardware + * events (user-space has to fall back and + * sample via a hrtimer based software event): + */ + pmu.capabilities |= PERF_PMU_CAP_NO_INTERRUPT; + } static struct attribute_group x86_pmu_format_group = { @@ -1521,6 +1545,8 @@ static int __init init_hw_perf_events(void) pr_cont("%s PMU driver.\n", x86_pmu.name); + x86_pmu.attr_rdpmc = 1; /* enable userspace RDPMC usage by default */ + for (quirk = x86_pmu.quirks; quirk; quirk = quirk->next) quirk->func(); @@ -1534,7 +1560,6 @@ static int __init init_hw_perf_events(void) __EVENT_CONSTRAINT(0, (1ULL << x86_pmu.num_counters) - 1, 0, x86_pmu.num_counters, 0, 0); - x86_pmu.attr_rdpmc = 1; /* enable userspace RDPMC usage by default */ x86_pmu_format_group.attrs = x86_pmu.format_attrs; if (x86_pmu.event_attrs) @@ -1594,7 +1619,8 @@ static void x86_pmu_cancel_txn(struct pmu *pmu) { __this_cpu_and(cpu_hw_events.group_flag, ~PERF_EVENT_TXN); /* - * Truncate the collected events. + * Truncate collected array by the number of events added in this + * transaction. See x86_pmu_add() and x86_pmu_*_txn(). */ __this_cpu_sub(cpu_hw_events.n_added, __this_cpu_read(cpu_hw_events.n_txn)); __this_cpu_sub(cpu_hw_events.n_events, __this_cpu_read(cpu_hw_events.n_txn)); @@ -1605,6 +1631,8 @@ static void x86_pmu_cancel_txn(struct pmu *pmu) * Commit group events scheduling transaction * Perform the group schedulability test as a whole * Return 0 if success + * + * Does not cancel the transaction on failure; expects the caller to do this. */ static int x86_pmu_commit_txn(struct pmu *pmu) { @@ -1820,9 +1848,12 @@ static ssize_t set_attr_rdpmc(struct device *cdev, if (ret) return ret; + if (x86_pmu.attr_rdpmc_broken) + return -ENOTSUPP; + if (!!val != !!x86_pmu.attr_rdpmc) { x86_pmu.attr_rdpmc = !!val; - smp_call_function(change_rdpmc, (void *)val, 1); + on_each_cpu(change_rdpmc, (void *)val, 1); } return count; @@ -1883,26 +1914,27 @@ static struct pmu pmu = { void arch_perf_update_userpage(struct perf_event_mmap_page *userpg, u64 now) { + struct cyc2ns_data *data; + userpg->cap_user_time = 0; userpg->cap_user_time_zero = 0; userpg->cap_user_rdpmc = x86_pmu.attr_rdpmc; userpg->pmc_width = x86_pmu.cntval_bits; - if (!boot_cpu_has(X86_FEATURE_CONSTANT_TSC)) + if (!sched_clock_stable()) return; - if (!boot_cpu_has(X86_FEATURE_NONSTOP_TSC)) - return; + data = cyc2ns_read_begin(); userpg->cap_user_time = 1; - userpg->time_mult = this_cpu_read(cyc2ns); - userpg->time_shift = CYC2NS_SCALE_FACTOR; - userpg->time_offset = this_cpu_read(cyc2ns_offset) - now; + userpg->time_mult = data->cyc2ns_mul; + userpg->time_shift = data->cyc2ns_shift; + userpg->time_offset = data->cyc2ns_offset - now; - if (sched_clock_stable && !check_tsc_disabled()) { - userpg->cap_user_time_zero = 1; - userpg->time_zero = this_cpu_read(cyc2ns_offset); - } + userpg->cap_user_time_zero = 1; + userpg->time_zero = data->cyc2ns_offset; + + cyc2ns_read_end(data); } /* @@ -1994,7 +2026,7 @@ perf_callchain_user32(struct pt_regs *regs, struct perf_callchain_entry *entry) frame.return_address = 0; bytes = copy_from_user_nmi(&frame, fp, sizeof(frame)); - if (bytes != sizeof(frame)) + if (bytes != 0) break; if (!valid_user_frame(fp, sizeof(frame))) @@ -2046,7 +2078,7 @@ perf_callchain_user(struct perf_callchain_entry *entry, struct pt_regs *regs) frame.return_address = 0; bytes = copy_from_user_nmi(&frame, fp, sizeof(frame)); - if (bytes != sizeof(frame)) + if (bytes != 0) break; if (!valid_user_frame(fp, sizeof(frame))) diff --git a/arch/x86/kernel/cpu/perf_event.h b/arch/x86/kernel/cpu/perf_event.h index cc16faae053..8ade93111e0 100644 --- a/arch/x86/kernel/cpu/perf_event.h +++ b/arch/x86/kernel/cpu/perf_event.h @@ -130,9 +130,11 @@ struct cpu_hw_events { unsigned long running[BITS_TO_LONGS(X86_PMC_IDX_MAX)]; int enabled; - int n_events; - int n_added; - int n_txn; + int n_events; /* the # of events in the below arrays */ + int n_added; /* the # last events in the below arrays; + they've never been enabled yet */ + int n_txn; /* the # last events in the below arrays; + added in the current transaction */ int assign[X86_PMC_IDX_MAX]; /* event to counter assignment */ u64 tags[X86_PMC_IDX_MAX]; struct perf_event *event_list[X86_PMC_IDX_MAX]; /* in enabled order */ @@ -164,6 +166,11 @@ struct cpu_hw_events { struct perf_guest_switch_msr guest_switch_msrs[X86_PMC_IDX_MAX]; /* + * Intel checkpoint mask + */ + u64 intel_cp_status; + + /* * manage shared (per-core, per-cpu) registers * used on Intel NHM/WSM/SNB */ @@ -257,11 +264,20 @@ struct cpu_hw_events { __EVENT_CONSTRAINT(c, n, INTEL_ARCH_EVENT_MASK, \ HWEIGHT(n), 0, PERF_X86_EVENT_PEBS_ST_HSW) -#define EVENT_CONSTRAINT_END \ - EVENT_CONSTRAINT(0, 0, 0) +/* + * We define the end marker as having a weight of -1 + * to enable blacklisting of events using a counter bitmask + * of zero and thus a weight of zero. + * The end marker has a weight that cannot possibly be + * obtained from counting the bits in the bitmask. + */ +#define EVENT_CONSTRAINT_END { .weight = -1 } +/* + * Check for end marker with weight == -1 + */ #define for_each_event_constraint(e, c) \ - for ((e) = (c); (e)->weight; (e)++) + for ((e) = (c); (e)->weight != -1; (e)++) /* * Extra registers for specific events. @@ -279,14 +295,16 @@ struct extra_reg { u64 config_mask; u64 valid_mask; int idx; /* per_xxx->regs[] reg index */ + bool extra_msr_access; }; #define EVENT_EXTRA_REG(e, ms, m, vm, i) { \ - .event = (e), \ - .msr = (ms), \ - .config_mask = (m), \ - .valid_mask = (vm), \ - .idx = EXTRA_REG_##i, \ + .event = (e), \ + .msr = (ms), \ + .config_mask = (m), \ + .valid_mask = (vm), \ + .idx = EXTRA_REG_##i, \ + .extra_msr_access = true, \ } #define INTEL_EVENT_EXTRA_REG(event, msr, vm, idx) \ @@ -395,6 +413,7 @@ struct x86_pmu { /* * sysfs attrs */ + int attr_rdpmc_broken; int attr_rdpmc; struct attribute **format_attrs; struct attribute **event_attrs; @@ -440,6 +459,7 @@ struct x86_pmu { int lbr_nr; /* hardware stack size */ u64 lbr_sel_mask; /* LBR_SELECT valid bits */ const int *lbr_sel_map; /* lbr_select mappings */ + bool lbr_double_abort; /* duplicated lbr aborts */ /* * Extra registers for events diff --git a/arch/x86/kernel/cpu/perf_event_amd_ibs.c b/arch/x86/kernel/cpu/perf_event_amd_ibs.c index e09f0bfb7b8..cbb1be3ed9e 100644 --- a/arch/x86/kernel/cpu/perf_event_amd_ibs.c +++ b/arch/x86/kernel/cpu/perf_event_amd_ibs.c @@ -10,6 +10,7 @@ #include <linux/module.h> #include <linux/pci.h> #include <linux/ptrace.h> +#include <linux/syscore_ops.h> #include <asm/apic.h> @@ -592,7 +593,7 @@ out: return 1; } -static int __kprobes +static int perf_ibs_nmi_handler(unsigned int cmd, struct pt_regs *regs) { int handled = 0; @@ -605,6 +606,7 @@ perf_ibs_nmi_handler(unsigned int cmd, struct pt_regs *regs) return handled; } +NOKPROBE_SYMBOL(perf_ibs_nmi_handler); static __init int perf_ibs_pmu_init(struct perf_ibs *perf_ibs, char *name) { @@ -816,6 +818,18 @@ out: return ret; } +static void ibs_eilvt_setup(void) +{ + /* + * Force LVT offset assignment for family 10h: The offsets are + * not assigned by the BIOS for this family, so the OS is + * responsible for doing it. If the OS assignment fails, fall + * back to BIOS settings and try to setup this. + */ + if (boot_cpu_data.x86 == 0x10) + force_ibs_eilvt_setup(); +} + static inline int get_ibs_lvt_offset(void) { u64 val; @@ -851,6 +865,36 @@ static void clear_APIC_ibs(void *dummy) setup_APIC_eilvt(offset, 0, APIC_EILVT_MSG_FIX, 1); } +#ifdef CONFIG_PM + +static int perf_ibs_suspend(void) +{ + clear_APIC_ibs(NULL); + return 0; +} + +static void perf_ibs_resume(void) +{ + ibs_eilvt_setup(); + setup_APIC_ibs(NULL); +} + +static struct syscore_ops perf_ibs_syscore_ops = { + .resume = perf_ibs_resume, + .suspend = perf_ibs_suspend, +}; + +static void perf_ibs_pm_init(void) +{ + register_syscore_ops(&perf_ibs_syscore_ops); +} + +#else + +static inline void perf_ibs_pm_init(void) { } + +#endif + static int perf_ibs_cpu_notifier(struct notifier_block *self, unsigned long action, void *hcpu) { @@ -877,25 +921,19 @@ static __init int amd_ibs_init(void) if (!caps) return -ENODEV; /* ibs not supported by the cpu */ - /* - * Force LVT offset assignment for family 10h: The offsets are - * not assigned by the BIOS for this family, so the OS is - * responsible for doing it. If the OS assignment fails, fall - * back to BIOS settings and try to setup this. - */ - if (boot_cpu_data.x86 == 0x10) - force_ibs_eilvt_setup(); + ibs_eilvt_setup(); if (!ibs_eilvt_valid()) goto out; - get_online_cpus(); + perf_ibs_pm_init(); + cpu_notifier_register_begin(); ibs_caps = caps; /* make ibs_caps visible to other cpus: */ smp_mb(); - perf_cpu_notifier(perf_ibs_cpu_notifier); smp_call_function(setup_APIC_ibs, NULL, 1); - put_online_cpus(); + __perf_cpu_notifier(perf_ibs_cpu_notifier); + cpu_notifier_register_done(); ret = perf_event_ibs_init(); out: diff --git a/arch/x86/kernel/cpu/perf_event_amd_uncore.c b/arch/x86/kernel/cpu/perf_event_amd_uncore.c index 754291adec3..3bbdf4cd38b 100644 --- a/arch/x86/kernel/cpu/perf_event_amd_uncore.c +++ b/arch/x86/kernel/cpu/perf_event_amd_uncore.c @@ -531,15 +531,16 @@ static int __init amd_uncore_init(void) if (ret) return -ENODEV; - get_online_cpus(); + cpu_notifier_register_begin(); + /* init cpus already online before registering for hotplug notifier */ for_each_online_cpu(cpu) { amd_uncore_cpu_up_prepare(cpu); smp_call_function_single(cpu, init_cpu_already_online, NULL, 1); } - register_cpu_notifier(&amd_uncore_cpu_notifier_block); - put_online_cpus(); + __register_cpu_notifier(&amd_uncore_cpu_notifier_block); + cpu_notifier_register_done(); return 0; } diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index f31a1655d1f..2502d0d9d24 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -169,7 +169,6 @@ static struct event_constraint intel_slm_event_constraints[] __read_mostly = { FIXED_EVENT_CONSTRAINT(0x00c0, 0), /* INST_RETIRED.ANY */ FIXED_EVENT_CONSTRAINT(0x003c, 1), /* CPU_CLK_UNHALTED.CORE */ - FIXED_EVENT_CONSTRAINT(0x013c, 2), /* CPU_CLK_UNHALTED.REF */ FIXED_EVENT_CONSTRAINT(0x0300, 2), /* pseudo CPU_CLK_UNHALTED.REF */ EVENT_CONSTRAINT_END }; @@ -190,9 +189,9 @@ static struct extra_reg intel_snbep_extra_regs[] __read_mostly = { EVENT_EXTRA_END }; -EVENT_ATTR_STR(mem-loads, mem_ld_nhm, "event=0x0b,umask=0x10,ldlat=3"); -EVENT_ATTR_STR(mem-loads, mem_ld_snb, "event=0xcd,umask=0x1,ldlat=3"); -EVENT_ATTR_STR(mem-stores, mem_st_snb, "event=0xcd,umask=0x2"); +EVENT_ATTR_STR(mem-loads, mem_ld_nhm, "event=0x0b,umask=0x10,ldlat=3"); +EVENT_ATTR_STR(mem-loads, mem_ld_snb, "event=0xcd,umask=0x1,ldlat=3"); +EVENT_ATTR_STR(mem-stores, mem_st_snb, "event=0xcd,umask=0x2"); struct attribute *nhm_events_attrs[] = { EVENT_PTR(mem_ld_nhm), @@ -1184,6 +1183,11 @@ static void intel_pmu_disable_fixed(struct hw_perf_event *hwc) wrmsrl(hwc->config_base, ctrl_val); } +static inline bool event_is_checkpointed(struct perf_event *event) +{ + return (event->hw.config & HSW_IN_TX_CHECKPOINTED) != 0; +} + static void intel_pmu_disable_event(struct perf_event *event) { struct hw_perf_event *hwc = &event->hw; @@ -1197,6 +1201,7 @@ static void intel_pmu_disable_event(struct perf_event *event) cpuc->intel_ctrl_guest_mask &= ~(1ull << hwc->idx); cpuc->intel_ctrl_host_mask &= ~(1ull << hwc->idx); + cpuc->intel_cp_status &= ~(1ull << hwc->idx); /* * must disable before any actual event @@ -1271,6 +1276,9 @@ static void intel_pmu_enable_event(struct perf_event *event) if (event->attr.exclude_guest) cpuc->intel_ctrl_host_mask |= (1ull << hwc->idx); + if (unlikely(event_is_checkpointed(event))) + cpuc->intel_cp_status |= (1ull << hwc->idx); + if (unlikely(hwc->config_base == MSR_ARCH_PERFMON_FIXED_CTR_CTRL)) { intel_pmu_enable_fixed(hwc); return; @@ -1289,6 +1297,17 @@ static void intel_pmu_enable_event(struct perf_event *event) int intel_pmu_save_and_restart(struct perf_event *event) { x86_perf_event_update(event); + /* + * For a checkpointed counter always reset back to 0. This + * avoids a situation where the counter overflows, aborts the + * transaction and is then set back to shortly before the + * overflow, and overflows and aborts again. + */ + if (unlikely(event_is_checkpointed(event))) { + /* No race with NMIs because the counter should not be armed */ + wrmsrl(event->hw.event_base, 0); + local64_set(&event->hw.prev_count, 0); + } return x86_perf_event_set_period(event); } @@ -1341,10 +1360,8 @@ static int intel_pmu_handle_irq(struct pt_regs *regs) intel_pmu_disable_all(); handled = intel_pmu_drain_bts_buffer(); status = intel_pmu_get_status(); - if (!status) { - intel_pmu_enable_all(0); - return handled; - } + if (!status) + goto done; loops = 0; again: @@ -1365,6 +1382,15 @@ again: intel_pmu_lbr_read(); /* + * CondChgd bit 63 doesn't mean any overflow status. Ignore + * and clear the bit. + */ + if (__test_and_clear_bit(63, (unsigned long *)&status)) { + if (!status) + goto done; + } + + /* * PEBS overflow sets bit 62 in the global status register */ if (__test_and_clear_bit(62, (unsigned long *)&status)) { @@ -1372,6 +1398,13 @@ again: x86_pmu.drain_pebs(regs); } + /* + * Checkpointed counters can lead to 'spurious' PMIs because the + * rollback caused by the PMI will have cleared the overflow status + * bit. Therefore always force probe these counters. + */ + status |= cpuc->intel_cp_status; + for_each_set_bit(bit, (unsigned long *)&status, X86_PMC_IDX_MAX) { struct perf_event *event = cpuc->events[bit]; @@ -1837,6 +1870,20 @@ static int hsw_hw_config(struct perf_event *event) event->attr.precise_ip > 0)) return -EOPNOTSUPP; + if (event_is_checkpointed(event)) { + /* + * Sampling of checkpointed events can cause situations where + * the CPU constantly aborts because of a overflow, which is + * then checkpointed back and ignored. Forbid checkpointing + * for sampling. + * + * But still allow a long sampling period, so that perf stat + * from KVM works. + */ + if (event->attr.sample_period > 0 && + event->attr.sample_period < 0x7fffffff) + return -EOPNOTSUPP; + } return 0; } @@ -2135,6 +2182,41 @@ static void intel_snb_check_microcode(void) } } +/* + * Under certain circumstances, access certain MSR may cause #GP. + * The function tests if the input MSR can be safely accessed. + */ +static bool check_msr(unsigned long msr, u64 mask) +{ + u64 val_old, val_new, val_tmp; + + /* + * Read the current value, change it and read it back to see if it + * matches, this is needed to detect certain hardware emulators + * (qemu/kvm) that don't trap on the MSR access and always return 0s. + */ + if (rdmsrl_safe(msr, &val_old)) + return false; + + /* + * Only change the bits which can be updated by wrmsrl. + */ + val_tmp = val_old ^ mask; + if (wrmsrl_safe(msr, val_tmp) || + rdmsrl_safe(msr, &val_new)) + return false; + + if (val_new != val_tmp) + return false; + + /* Here it's sure that the MSR can be safely accessed. + * Restore the old value and return. + */ + wrmsrl(msr, val_old); + + return true; +} + static __init void intel_sandybridge_quirk(void) { x86_pmu.check_microcode = intel_snb_check_microcode; @@ -2182,10 +2264,36 @@ static __init void intel_nehalem_quirk(void) } } -EVENT_ATTR_STR(mem-loads, mem_ld_hsw, "event=0xcd,umask=0x1,ldlat=3"); -EVENT_ATTR_STR(mem-stores, mem_st_hsw, "event=0xd0,umask=0x82") +EVENT_ATTR_STR(mem-loads, mem_ld_hsw, "event=0xcd,umask=0x1,ldlat=3"); +EVENT_ATTR_STR(mem-stores, mem_st_hsw, "event=0xd0,umask=0x82") + +/* Haswell special events */ +EVENT_ATTR_STR(tx-start, tx_start, "event=0xc9,umask=0x1"); +EVENT_ATTR_STR(tx-commit, tx_commit, "event=0xc9,umask=0x2"); +EVENT_ATTR_STR(tx-abort, tx_abort, "event=0xc9,umask=0x4"); +EVENT_ATTR_STR(tx-capacity, tx_capacity, "event=0x54,umask=0x2"); +EVENT_ATTR_STR(tx-conflict, tx_conflict, "event=0x54,umask=0x1"); +EVENT_ATTR_STR(el-start, el_start, "event=0xc8,umask=0x1"); +EVENT_ATTR_STR(el-commit, el_commit, "event=0xc8,umask=0x2"); +EVENT_ATTR_STR(el-abort, el_abort, "event=0xc8,umask=0x4"); +EVENT_ATTR_STR(el-capacity, el_capacity, "event=0x54,umask=0x2"); +EVENT_ATTR_STR(el-conflict, el_conflict, "event=0x54,umask=0x1"); +EVENT_ATTR_STR(cycles-t, cycles_t, "event=0x3c,in_tx=1"); +EVENT_ATTR_STR(cycles-ct, cycles_ct, "event=0x3c,in_tx=1,in_tx_cp=1"); static struct attribute *hsw_events_attrs[] = { + EVENT_PTR(tx_start), + EVENT_PTR(tx_commit), + EVENT_PTR(tx_abort), + EVENT_PTR(tx_capacity), + EVENT_PTR(tx_conflict), + EVENT_PTR(el_start), + EVENT_PTR(el_commit), + EVENT_PTR(el_abort), + EVENT_PTR(el_capacity), + EVENT_PTR(el_conflict), + EVENT_PTR(cycles_t), + EVENT_PTR(cycles_ct), EVENT_PTR(mem_ld_hsw), EVENT_PTR(mem_st_hsw), NULL @@ -2198,7 +2306,8 @@ __init int intel_pmu_init(void) union cpuid10_ebx ebx; struct event_constraint *c; unsigned int unused; - int version; + struct extra_reg *er; + int version, i; if (!cpu_has(&boot_cpu_data, X86_FEATURE_ARCH_PERFMON)) { switch (boot_cpu_data.x86) { @@ -2243,10 +2352,7 @@ __init int intel_pmu_init(void) if (version > 1) x86_pmu.num_counters_fixed = max((int)edx.split.num_counters_fixed, 3); - /* - * v2 and above have a perf capabilities MSR - */ - if (version > 1) { + if (boot_cpu_has(X86_FEATURE_PDCM)) { u64 capabilities; rdmsrl(MSR_IA32_PERF_CAPABILITIES, capabilities); @@ -2404,6 +2510,9 @@ __init int intel_pmu_init(void) case 62: /* IvyBridge EP */ memcpy(hw_cache_event_ids, snb_hw_cache_event_ids, sizeof(hw_cache_event_ids)); + /* dTLB-load-misses on IVB is different than SNB */ + hw_cache_event_ids[C(DTLB)][C(OP_READ)][C(RESULT_MISS)] = 0x8108; /* DTLB_LOAD_MISSES.DEMAND_LD_MISS_CAUSES_A_WALK */ + memcpy(hw_cache_extra_regs, snb_hw_cache_extra_regs, sizeof(hw_cache_extra_regs)); @@ -2452,6 +2561,7 @@ __init int intel_pmu_init(void) x86_pmu.hw_config = hsw_hw_config; x86_pmu.get_event_constraints = hsw_get_event_constraints; x86_pmu.cpu_events = hsw_events_attrs; + x86_pmu.lbr_double_abort = true; pr_cont("Haswell events, "); break; @@ -2503,6 +2613,34 @@ __init int intel_pmu_init(void) } } + /* + * Access LBR MSR may cause #GP under certain circumstances. + * E.g. KVM doesn't support LBR MSR + * Check all LBT MSR here. + * Disable LBR access if any LBR MSRs can not be accessed. + */ + if (x86_pmu.lbr_nr && !check_msr(x86_pmu.lbr_tos, 0x3UL)) + x86_pmu.lbr_nr = 0; + for (i = 0; i < x86_pmu.lbr_nr; i++) { + if (!(check_msr(x86_pmu.lbr_from + i, 0xffffUL) && + check_msr(x86_pmu.lbr_to + i, 0xffffUL))) + x86_pmu.lbr_nr = 0; + } + + /* + * Access extra MSR may cause #GP under certain circumstances. + * E.g. KVM doesn't support offcore event + * Check all extra_regs here. + */ + if (x86_pmu.extra_regs) { + for (er = x86_pmu.extra_regs; er->msr; er++) { + er->extra_msr_access = check_msr(er->msr, 0x1ffUL); + /* Disable LBR select mapping */ + if ((er->idx == EXTRA_REG_LBR) && !er->extra_msr_access) + x86_pmu.lbr_sel_map = NULL; + } + } + /* Support full width counters using alternative MSR range */ if (x86_pmu.intel_cap.full_width_write) { x86_pmu.max_period = x86_pmu.cntval_mask; diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index ab3ba1c1b7d..696ade311de 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -12,6 +12,7 @@ #define BTS_BUFFER_SIZE (PAGE_SIZE << 4) #define PEBS_BUFFER_SIZE PAGE_SIZE +#define PEBS_FIXUP_SIZE PAGE_SIZE /* * pebs_record_32 for p4 and core not supported @@ -107,15 +108,31 @@ static u64 precise_store_data(u64 status) return val; } -static u64 precise_store_data_hsw(u64 status) +static u64 precise_store_data_hsw(struct perf_event *event, u64 status) { union perf_mem_data_src dse; + u64 cfg = event->hw.config & INTEL_ARCH_EVENT_MASK; dse.val = 0; dse.mem_op = PERF_MEM_OP_STORE; dse.mem_lvl = PERF_MEM_LVL_NA; + + /* + * L1 info only valid for following events: + * + * MEM_UOPS_RETIRED.STLB_MISS_STORES + * MEM_UOPS_RETIRED.LOCK_STORES + * MEM_UOPS_RETIRED.SPLIT_STORES + * MEM_UOPS_RETIRED.ALL_STORES + */ + if (cfg != 0x12d0 && cfg != 0x22d0 && cfg != 0x42d0 && cfg != 0x82d0) + return dse.mem_lvl; + if (status & 1) - dse.mem_lvl = PERF_MEM_LVL_L1; + dse.mem_lvl = PERF_MEM_LVL_L1 | PERF_MEM_LVL_HIT; + else + dse.mem_lvl = PERF_MEM_LVL_L1 | PERF_MEM_LVL_MISS; + /* Nothing else supported. Sorry. */ return dse.val; } @@ -182,18 +199,32 @@ struct pebs_record_nhm { * Same as pebs_record_nhm, with two additional fields. */ struct pebs_record_hsw { - struct pebs_record_nhm nhm; - /* - * Real IP of the event. In the Intel documentation this - * is called eventingrip. - */ - u64 real_ip; - /* - * TSX tuning information field: abort cycles and abort flags. - */ - u64 tsx_tuning; + u64 flags, ip; + u64 ax, bx, cx, dx; + u64 si, di, bp, sp; + u64 r8, r9, r10, r11; + u64 r12, r13, r14, r15; + u64 status, dla, dse, lat; + u64 real_ip, tsx_tuning; +}; + +union hsw_tsx_tuning { + struct { + u32 cycles_last_block : 32, + hle_abort : 1, + rtm_abort : 1, + instruction_abort : 1, + non_instruction_abort : 1, + retry : 1, + data_conflict : 1, + capacity_writes : 1, + capacity_reads : 1; + }; + u64 value; }; +#define PEBS_HSW_TSX_FLAGS 0xff00000000ULL + void init_debug_store_on_cpu(int cpu) { struct debug_store *ds = per_cpu(cpu_hw_events, cpu).ds; @@ -214,12 +245,14 @@ void fini_debug_store_on_cpu(int cpu) wrmsr_on_cpu(cpu, MSR_IA32_DS_AREA, 0, 0); } +static DEFINE_PER_CPU(void *, insn_buffer); + static int alloc_pebs_buffer(int cpu) { struct debug_store *ds = per_cpu(cpu_hw_events, cpu).ds; int node = cpu_to_node(cpu); int max, thresh = 1; /* always use a single PEBS record */ - void *buffer; + void *buffer, *ibuffer; if (!x86_pmu.pebs) return 0; @@ -228,6 +261,19 @@ static int alloc_pebs_buffer(int cpu) if (unlikely(!buffer)) return -ENOMEM; + /* + * HSW+ already provides us the eventing ip; no need to allocate this + * buffer then. + */ + if (x86_pmu.intel_cap.pebs_format < 2) { + ibuffer = kzalloc_node(PEBS_FIXUP_SIZE, GFP_KERNEL, node); + if (!ibuffer) { + kfree(buffer); + return -ENOMEM; + } + per_cpu(insn_buffer, cpu) = ibuffer; + } + max = PEBS_BUFFER_SIZE / x86_pmu.pebs_record_size; ds->pebs_buffer_base = (u64)(unsigned long)buffer; @@ -248,6 +294,9 @@ static void release_pebs_buffer(int cpu) if (!ds || !x86_pmu.pebs) return; + kfree(per_cpu(insn_buffer, cpu)); + per_cpu(insn_buffer, cpu) = NULL; + kfree((void *)(unsigned long)ds->pebs_buffer_base); ds->pebs_buffer_base = 0; } @@ -262,9 +311,11 @@ static int alloc_bts_buffer(int cpu) if (!x86_pmu.bts) return 0; - buffer = kzalloc_node(BTS_BUFFER_SIZE, GFP_KERNEL, node); - if (unlikely(!buffer)) + buffer = kzalloc_node(BTS_BUFFER_SIZE, GFP_KERNEL | __GFP_NOWARN, node); + if (unlikely(!buffer)) { + WARN_ONCE(1, "%s: BTS buffer allocation failure\n", __func__); return -ENOMEM; + } max = BTS_BUFFER_SIZE / BTS_RECORD_SIZE; thresh = max / 16; @@ -715,6 +766,7 @@ static int intel_pmu_pebs_fixup_ip(struct pt_regs *regs) unsigned long old_to, to = cpuc->lbr_entries[0].to; unsigned long ip = regs->ip; int is_64bit = 0; + void *kaddr; /* * We don't need to fixup if the PEBS assist is fault like @@ -738,7 +790,7 @@ static int intel_pmu_pebs_fixup_ip(struct pt_regs *regs) * unsigned math, either ip is before the start (impossible) or * the basic block is larger than 1 page (sanity) */ - if ((ip - to) > PAGE_SIZE) + if ((ip - to) > PEBS_FIXUP_SIZE) return 0; /* @@ -749,29 +801,33 @@ static int intel_pmu_pebs_fixup_ip(struct pt_regs *regs) return 1; } + if (!kernel_ip(ip)) { + int size, bytes; + u8 *buf = this_cpu_read(insn_buffer); + + size = ip - to; /* Must fit our buffer, see above */ + bytes = copy_from_user_nmi(buf, (void __user *)to, size); + if (bytes != 0) + return 0; + + kaddr = buf; + } else { + kaddr = (void *)to; + } + do { struct insn insn; - u8 buf[MAX_INSN_SIZE]; - void *kaddr; old_to = to; - if (!kernel_ip(ip)) { - int bytes, size = MAX_INSN_SIZE; - - bytes = copy_from_user_nmi(buf, (void __user *)to, size); - if (bytes != size) - return 0; - - kaddr = buf; - } else - kaddr = (void *)to; #ifdef CONFIG_X86_64 is_64bit = kernel_ip(to) || !test_thread_flag(TIF_IA32); #endif insn_init(&insn, kaddr, is_64bit); insn_get_length(&insn); + to += insn.length; + kaddr += insn.length; } while (to < ip); if (to == ip) { @@ -786,16 +842,34 @@ static int intel_pmu_pebs_fixup_ip(struct pt_regs *regs) return 0; } +static inline u64 intel_hsw_weight(struct pebs_record_hsw *pebs) +{ + if (pebs->tsx_tuning) { + union hsw_tsx_tuning tsx = { .value = pebs->tsx_tuning }; + return tsx.cycles_last_block; + } + return 0; +} + +static inline u64 intel_hsw_transaction(struct pebs_record_hsw *pebs) +{ + u64 txn = (pebs->tsx_tuning & PEBS_HSW_TSX_FLAGS) >> 32; + + /* For RTM XABORTs also log the abort code from AX */ + if ((txn & PERF_TXN_TRANSACTION) && (pebs->ax & 1)) + txn |= ((pebs->ax >> 24) & 0xff) << PERF_TXN_ABORT_SHIFT; + return txn; +} + static void __intel_pmu_pebs_event(struct perf_event *event, struct pt_regs *iregs, void *__pebs) { /* - * We cast to pebs_record_nhm to get the load latency data - * if extra_reg MSR_PEBS_LD_LAT_THRESHOLD used + * We cast to the biggest pebs_record but are careful not to + * unconditionally access the 'extra' entries. */ struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); - struct pebs_record_nhm *pebs = __pebs; - struct pebs_record_hsw *pebs_hsw = __pebs; + struct pebs_record_hsw *pebs = __pebs; struct perf_sample_data data; struct pt_regs regs; u64 sample_type; @@ -831,7 +905,7 @@ static void __intel_pmu_pebs_event(struct perf_event *event, data.data_src.val = load_latency_data(pebs->dse); else if (event->hw.flags & PERF_X86_EVENT_PEBS_ST_HSW) data.data_src.val = - precise_store_data_hsw(pebs->dse); + precise_store_data_hsw(event, pebs->dse); else data.data_src.val = precise_store_data(pebs->dse); } @@ -854,7 +928,7 @@ static void __intel_pmu_pebs_event(struct perf_event *event, regs.sp = pebs->sp; if (event->attr.precise_ip > 1 && x86_pmu.intel_cap.pebs_format >= 2) { - regs.ip = pebs_hsw->real_ip; + regs.ip = pebs->real_ip; regs.flags |= PERF_EFLAGS_EXACT; } else if (event->attr.precise_ip > 1 && intel_pmu_pebs_fixup_ip(®s)) regs.flags |= PERF_EFLAGS_EXACT; @@ -862,9 +936,18 @@ static void __intel_pmu_pebs_event(struct perf_event *event, regs.flags &= ~PERF_EFLAGS_EXACT; if ((event->attr.sample_type & PERF_SAMPLE_ADDR) && - x86_pmu.intel_cap.pebs_format >= 1) + x86_pmu.intel_cap.pebs_format >= 1) data.addr = pebs->dla; + if (x86_pmu.intel_cap.pebs_format >= 2) { + /* Only set the TSX weight when no memory weight. */ + if ((event->attr.sample_type & PERF_SAMPLE_WEIGHT) && !fll) + data.weight = intel_hsw_weight(pebs); + + if (event->attr.sample_type & PERF_SAMPLE_TRANSACTION) + data.txn = intel_hsw_transaction(pebs); + } + if (has_branch_stack(event)) data.br_stack = &cpuc->lbr_stack; @@ -913,17 +996,34 @@ static void intel_pmu_drain_pebs_core(struct pt_regs *iregs) __intel_pmu_pebs_event(event, iregs, at); } -static void __intel_pmu_drain_pebs_nhm(struct pt_regs *iregs, void *at, - void *top) +static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) { struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); struct debug_store *ds = cpuc->ds; struct perf_event *event = NULL; + void *at, *top; u64 status = 0; int bit; + if (!x86_pmu.pebs_active) + return; + + at = (struct pebs_record_nhm *)(unsigned long)ds->pebs_buffer_base; + top = (struct pebs_record_nhm *)(unsigned long)ds->pebs_index; + ds->pebs_index = ds->pebs_buffer_base; + if (unlikely(at > top)) + return; + + /* + * Should not happen, we program the threshold at 1 and do not + * set a reset value. + */ + WARN_ONCE(top - at > x86_pmu.max_pebs_events * x86_pmu.pebs_record_size, + "Unexpected number of pebs records %ld\n", + (long)(top - at) / x86_pmu.pebs_record_size); + for (; at < top; at += x86_pmu.pebs_record_size) { struct pebs_record_nhm *p = at; @@ -951,61 +1051,6 @@ static void __intel_pmu_drain_pebs_nhm(struct pt_regs *iregs, void *at, } } -static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) -{ - struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); - struct debug_store *ds = cpuc->ds; - struct pebs_record_nhm *at, *top; - int n; - - if (!x86_pmu.pebs_active) - return; - - at = (struct pebs_record_nhm *)(unsigned long)ds->pebs_buffer_base; - top = (struct pebs_record_nhm *)(unsigned long)ds->pebs_index; - - ds->pebs_index = ds->pebs_buffer_base; - - n = top - at; - if (n <= 0) - return; - - /* - * Should not happen, we program the threshold at 1 and do not - * set a reset value. - */ - WARN_ONCE(n > x86_pmu.max_pebs_events, - "Unexpected number of pebs records %d\n", n); - - return __intel_pmu_drain_pebs_nhm(iregs, at, top); -} - -static void intel_pmu_drain_pebs_hsw(struct pt_regs *iregs) -{ - struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); - struct debug_store *ds = cpuc->ds; - struct pebs_record_hsw *at, *top; - int n; - - if (!x86_pmu.pebs_active) - return; - - at = (struct pebs_record_hsw *)(unsigned long)ds->pebs_buffer_base; - top = (struct pebs_record_hsw *)(unsigned long)ds->pebs_index; - - n = top - at; - if (n <= 0) - return; - /* - * Should not happen, we program the threshold at 1 and do not - * set a reset value. - */ - WARN_ONCE(n > x86_pmu.max_pebs_events, - "Unexpected number of pebs records %d\n", n); - - return __intel_pmu_drain_pebs_nhm(iregs, at, top); -} - /* * BTS, PEBS probe and setup */ @@ -1040,7 +1085,7 @@ void intel_ds_init(void) case 2: pr_cont("PEBS fmt2%c, ", pebs_type); x86_pmu.pebs_record_size = sizeof(struct pebs_record_hsw); - x86_pmu.drain_pebs = intel_pmu_drain_pebs_hsw; + x86_pmu.drain_pebs = intel_pmu_drain_pebs_nhm; break; default: diff --git a/arch/x86/kernel/cpu/perf_event_intel_lbr.c b/arch/x86/kernel/cpu/perf_event_intel_lbr.c index d5be06a5005..9dd2459a4c7 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_lbr.c +++ b/arch/x86/kernel/cpu/perf_event_intel_lbr.c @@ -284,6 +284,7 @@ static void intel_pmu_lbr_read_64(struct cpu_hw_events *cpuc) int lbr_format = x86_pmu.intel_cap.lbr_format; u64 tos = intel_pmu_lbr_tos(); int i; + int out = 0; for (i = 0; i < x86_pmu.lbr_nr; i++) { unsigned long lbr_idx = (tos - i) & mask; @@ -306,15 +307,27 @@ static void intel_pmu_lbr_read_64(struct cpu_hw_events *cpuc) } from = (u64)((((s64)from) << skip) >> skip); - cpuc->lbr_entries[i].from = from; - cpuc->lbr_entries[i].to = to; - cpuc->lbr_entries[i].mispred = mis; - cpuc->lbr_entries[i].predicted = pred; - cpuc->lbr_entries[i].in_tx = in_tx; - cpuc->lbr_entries[i].abort = abort; - cpuc->lbr_entries[i].reserved = 0; + /* + * Some CPUs report duplicated abort records, + * with the second entry not having an abort bit set. + * Skip them here. This loop runs backwards, + * so we need to undo the previous record. + * If the abort just happened outside the window + * the extra entry cannot be removed. + */ + if (abort && x86_pmu.lbr_double_abort && out > 0) + out--; + + cpuc->lbr_entries[out].from = from; + cpuc->lbr_entries[out].to = to; + cpuc->lbr_entries[out].mispred = mis; + cpuc->lbr_entries[out].predicted = pred; + cpuc->lbr_entries[out].in_tx = in_tx; + cpuc->lbr_entries[out].abort = abort; + cpuc->lbr_entries[out].reserved = 0; + out++; } - cpuc->lbr_stack.nr = i; + cpuc->lbr_stack.nr = out; } void intel_pmu_lbr_read(void) @@ -371,6 +384,9 @@ static void intel_pmu_setup_sw_lbr_filter(struct perf_event *event) if (br_type & PERF_SAMPLE_BRANCH_NO_TX) mask |= X86_BR_NO_TX; + if (br_type & PERF_SAMPLE_BRANCH_COND) + mask |= X86_BR_JCC; + /* * stash actual user request into reg, it may * be used by fixup code for some CPU @@ -478,7 +494,7 @@ static int branch_type(unsigned long from, unsigned long to, int abort) /* may fail if text not present */ bytes = copy_from_user_nmi(buf, (void __user *)from, size); - if (bytes != size) + if (bytes != 0) return X86_BR_NONE; addr = buf; @@ -665,6 +681,7 @@ static const int nhm_lbr_sel_map[PERF_SAMPLE_BRANCH_MAX] = { * NHM/WSM erratum: must include IND_JMP to capture IND_CALL */ [PERF_SAMPLE_BRANCH_IND_CALL] = LBR_IND_CALL | LBR_IND_JMP, + [PERF_SAMPLE_BRANCH_COND] = LBR_JCC, }; static const int snb_lbr_sel_map[PERF_SAMPLE_BRANCH_MAX] = { @@ -676,6 +693,7 @@ static const int snb_lbr_sel_map[PERF_SAMPLE_BRANCH_MAX] = { [PERF_SAMPLE_BRANCH_ANY_CALL] = LBR_REL_CALL | LBR_IND_CALL | LBR_FAR, [PERF_SAMPLE_BRANCH_IND_CALL] = LBR_IND_CALL, + [PERF_SAMPLE_BRANCH_COND] = LBR_JCC, }; /* core */ diff --git a/arch/x86/kernel/cpu/perf_event_intel_rapl.c b/arch/x86/kernel/cpu/perf_event_intel_rapl.c new file mode 100644 index 00000000000..619f7699487 --- /dev/null +++ b/arch/x86/kernel/cpu/perf_event_intel_rapl.c @@ -0,0 +1,714 @@ +/* + * perf_event_intel_rapl.c: support Intel RAPL energy consumption counters + * Copyright (C) 2013 Google, Inc., Stephane Eranian + * + * Intel RAPL interface is specified in the IA-32 Manual Vol3b + * section 14.7.1 (September 2013) + * + * RAPL provides more controls than just reporting energy consumption + * however here we only expose the 3 energy consumption free running + * counters (pp0, pkg, dram). + * + * Each of those counters increments in a power unit defined by the + * RAPL_POWER_UNIT MSR. On SandyBridge, this unit is 1/(2^16) Joules + * but it can vary. + * + * Counter to rapl events mappings: + * + * pp0 counter: consumption of all physical cores (power plane 0) + * event: rapl_energy_cores + * perf code: 0x1 + * + * pkg counter: consumption of the whole processor package + * event: rapl_energy_pkg + * perf code: 0x2 + * + * dram counter: consumption of the dram domain (servers only) + * event: rapl_energy_dram + * perf code: 0x3 + * + * dram counter: consumption of the builtin-gpu domain (client only) + * event: rapl_energy_gpu + * perf code: 0x4 + * + * We manage those counters as free running (read-only). They may be + * use simultaneously by other tools, such as turbostat. + * + * The events only support system-wide mode counting. There is no + * sampling support because it does not make sense and is not + * supported by the RAPL hardware. + * + * Because we want to avoid floating-point operations in the kernel, + * the events are all reported in fixed point arithmetic (32.32). + * Tools must adjust the counts to convert them to Watts using + * the duration of the measurement. Tools may use a function such as + * ldexp(raw_count, -32); + */ +#include <linux/module.h> +#include <linux/slab.h> +#include <linux/perf_event.h> +#include <asm/cpu_device_id.h> +#include "perf_event.h" + +/* + * RAPL energy status counters + */ +#define RAPL_IDX_PP0_NRG_STAT 0 /* all cores */ +#define INTEL_RAPL_PP0 0x1 /* pseudo-encoding */ +#define RAPL_IDX_PKG_NRG_STAT 1 /* entire package */ +#define INTEL_RAPL_PKG 0x2 /* pseudo-encoding */ +#define RAPL_IDX_RAM_NRG_STAT 2 /* DRAM */ +#define INTEL_RAPL_RAM 0x3 /* pseudo-encoding */ +#define RAPL_IDX_PP1_NRG_STAT 3 /* gpu */ +#define INTEL_RAPL_PP1 0x4 /* pseudo-encoding */ + +/* Clients have PP0, PKG */ +#define RAPL_IDX_CLN (1<<RAPL_IDX_PP0_NRG_STAT|\ + 1<<RAPL_IDX_PKG_NRG_STAT|\ + 1<<RAPL_IDX_PP1_NRG_STAT) + +/* Servers have PP0, PKG, RAM */ +#define RAPL_IDX_SRV (1<<RAPL_IDX_PP0_NRG_STAT|\ + 1<<RAPL_IDX_PKG_NRG_STAT|\ + 1<<RAPL_IDX_RAM_NRG_STAT) + +/* Servers have PP0, PKG, RAM, PP1 */ +#define RAPL_IDX_HSW (1<<RAPL_IDX_PP0_NRG_STAT|\ + 1<<RAPL_IDX_PKG_NRG_STAT|\ + 1<<RAPL_IDX_RAM_NRG_STAT|\ + 1<<RAPL_IDX_PP1_NRG_STAT) + +/* + * event code: LSB 8 bits, passed in attr->config + * any other bit is reserved + */ +#define RAPL_EVENT_MASK 0xFFULL + +#define DEFINE_RAPL_FORMAT_ATTR(_var, _name, _format) \ +static ssize_t __rapl_##_var##_show(struct kobject *kobj, \ + struct kobj_attribute *attr, \ + char *page) \ +{ \ + BUILD_BUG_ON(sizeof(_format) >= PAGE_SIZE); \ + return sprintf(page, _format "\n"); \ +} \ +static struct kobj_attribute format_attr_##_var = \ + __ATTR(_name, 0444, __rapl_##_var##_show, NULL) + +#define RAPL_EVENT_DESC(_name, _config) \ +{ \ + .attr = __ATTR(_name, 0444, rapl_event_show, NULL), \ + .config = _config, \ +} + +#define RAPL_CNTR_WIDTH 32 /* 32-bit rapl counters */ + +struct rapl_pmu { + spinlock_t lock; + int hw_unit; /* 1/2^hw_unit Joule */ + int n_active; /* number of active events */ + struct list_head active_list; + struct pmu *pmu; /* pointer to rapl_pmu_class */ + ktime_t timer_interval; /* in ktime_t unit */ + struct hrtimer hrtimer; +}; + +static struct pmu rapl_pmu_class; +static cpumask_t rapl_cpu_mask; +static int rapl_cntr_mask; + +static DEFINE_PER_CPU(struct rapl_pmu *, rapl_pmu); +static DEFINE_PER_CPU(struct rapl_pmu *, rapl_pmu_to_free); + +static inline u64 rapl_read_counter(struct perf_event *event) +{ + u64 raw; + rdmsrl(event->hw.event_base, raw); + return raw; +} + +static inline u64 rapl_scale(u64 v) +{ + /* + * scale delta to smallest unit (1/2^32) + * users must then scale back: count * 1/(1e9*2^32) to get Joules + * or use ldexp(count, -32). + * Watts = Joules/Time delta + */ + return v << (32 - __get_cpu_var(rapl_pmu)->hw_unit); +} + +static u64 rapl_event_update(struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + u64 prev_raw_count, new_raw_count; + s64 delta, sdelta; + int shift = RAPL_CNTR_WIDTH; + +again: + prev_raw_count = local64_read(&hwc->prev_count); + rdmsrl(event->hw.event_base, new_raw_count); + + if (local64_cmpxchg(&hwc->prev_count, prev_raw_count, + new_raw_count) != prev_raw_count) { + cpu_relax(); + goto again; + } + + /* + * Now we have the new raw value and have updated the prev + * timestamp already. We can now calculate the elapsed delta + * (event-)time and add that to the generic event. + * + * Careful, not all hw sign-extends above the physical width + * of the count. + */ + delta = (new_raw_count << shift) - (prev_raw_count << shift); + delta >>= shift; + + sdelta = rapl_scale(delta); + + local64_add(sdelta, &event->count); + + return new_raw_count; +} + +static void rapl_start_hrtimer(struct rapl_pmu *pmu) +{ + __hrtimer_start_range_ns(&pmu->hrtimer, + pmu->timer_interval, 0, + HRTIMER_MODE_REL_PINNED, 0); +} + +static void rapl_stop_hrtimer(struct rapl_pmu *pmu) +{ + hrtimer_cancel(&pmu->hrtimer); +} + +static enum hrtimer_restart rapl_hrtimer_handle(struct hrtimer *hrtimer) +{ + struct rapl_pmu *pmu = __get_cpu_var(rapl_pmu); + struct perf_event *event; + unsigned long flags; + + if (!pmu->n_active) + return HRTIMER_NORESTART; + + spin_lock_irqsave(&pmu->lock, flags); + + list_for_each_entry(event, &pmu->active_list, active_entry) { + rapl_event_update(event); + } + + spin_unlock_irqrestore(&pmu->lock, flags); + + hrtimer_forward_now(hrtimer, pmu->timer_interval); + + return HRTIMER_RESTART; +} + +static void rapl_hrtimer_init(struct rapl_pmu *pmu) +{ + struct hrtimer *hr = &pmu->hrtimer; + + hrtimer_init(hr, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + hr->function = rapl_hrtimer_handle; +} + +static void __rapl_pmu_event_start(struct rapl_pmu *pmu, + struct perf_event *event) +{ + if (WARN_ON_ONCE(!(event->hw.state & PERF_HES_STOPPED))) + return; + + event->hw.state = 0; + + list_add_tail(&event->active_entry, &pmu->active_list); + + local64_set(&event->hw.prev_count, rapl_read_counter(event)); + + pmu->n_active++; + if (pmu->n_active == 1) + rapl_start_hrtimer(pmu); +} + +static void rapl_pmu_event_start(struct perf_event *event, int mode) +{ + struct rapl_pmu *pmu = __get_cpu_var(rapl_pmu); + unsigned long flags; + + spin_lock_irqsave(&pmu->lock, flags); + __rapl_pmu_event_start(pmu, event); + spin_unlock_irqrestore(&pmu->lock, flags); +} + +static void rapl_pmu_event_stop(struct perf_event *event, int mode) +{ + struct rapl_pmu *pmu = __get_cpu_var(rapl_pmu); + struct hw_perf_event *hwc = &event->hw; + unsigned long flags; + + spin_lock_irqsave(&pmu->lock, flags); + + /* mark event as deactivated and stopped */ + if (!(hwc->state & PERF_HES_STOPPED)) { + WARN_ON_ONCE(pmu->n_active <= 0); + pmu->n_active--; + if (pmu->n_active == 0) + rapl_stop_hrtimer(pmu); + + list_del(&event->active_entry); + + WARN_ON_ONCE(hwc->state & PERF_HES_STOPPED); + hwc->state |= PERF_HES_STOPPED; + } + + /* check if update of sw counter is necessary */ + if ((mode & PERF_EF_UPDATE) && !(hwc->state & PERF_HES_UPTODATE)) { + /* + * Drain the remaining delta count out of a event + * that we are disabling: + */ + rapl_event_update(event); + hwc->state |= PERF_HES_UPTODATE; + } + + spin_unlock_irqrestore(&pmu->lock, flags); +} + +static int rapl_pmu_event_add(struct perf_event *event, int mode) +{ + struct rapl_pmu *pmu = __get_cpu_var(rapl_pmu); + struct hw_perf_event *hwc = &event->hw; + unsigned long flags; + + spin_lock_irqsave(&pmu->lock, flags); + + hwc->state = PERF_HES_UPTODATE | PERF_HES_STOPPED; + + if (mode & PERF_EF_START) + __rapl_pmu_event_start(pmu, event); + + spin_unlock_irqrestore(&pmu->lock, flags); + + return 0; +} + +static void rapl_pmu_event_del(struct perf_event *event, int flags) +{ + rapl_pmu_event_stop(event, PERF_EF_UPDATE); +} + +static int rapl_pmu_event_init(struct perf_event *event) +{ + u64 cfg = event->attr.config & RAPL_EVENT_MASK; + int bit, msr, ret = 0; + + /* only look at RAPL events */ + if (event->attr.type != rapl_pmu_class.type) + return -ENOENT; + + /* check only supported bits are set */ + if (event->attr.config & ~RAPL_EVENT_MASK) + return -EINVAL; + + /* + * check event is known (determines counter) + */ + switch (cfg) { + case INTEL_RAPL_PP0: + bit = RAPL_IDX_PP0_NRG_STAT; + msr = MSR_PP0_ENERGY_STATUS; + break; + case INTEL_RAPL_PKG: + bit = RAPL_IDX_PKG_NRG_STAT; + msr = MSR_PKG_ENERGY_STATUS; + break; + case INTEL_RAPL_RAM: + bit = RAPL_IDX_RAM_NRG_STAT; + msr = MSR_DRAM_ENERGY_STATUS; + break; + case INTEL_RAPL_PP1: + bit = RAPL_IDX_PP1_NRG_STAT; + msr = MSR_PP1_ENERGY_STATUS; + break; + default: + return -EINVAL; + } + /* check event supported */ + if (!(rapl_cntr_mask & (1 << bit))) + return -EINVAL; + + /* unsupported modes and filters */ + if (event->attr.exclude_user || + event->attr.exclude_kernel || + event->attr.exclude_hv || + event->attr.exclude_idle || + event->attr.exclude_host || + event->attr.exclude_guest || + event->attr.sample_period) /* no sampling */ + return -EINVAL; + + /* must be done before validate_group */ + event->hw.event_base = msr; + event->hw.config = cfg; + event->hw.idx = bit; + + return ret; +} + +static void rapl_pmu_event_read(struct perf_event *event) +{ + rapl_event_update(event); +} + +static ssize_t rapl_get_attr_cpumask(struct device *dev, + struct device_attribute *attr, char *buf) +{ + int n = cpulist_scnprintf(buf, PAGE_SIZE - 2, &rapl_cpu_mask); + + buf[n++] = '\n'; + buf[n] = '\0'; + return n; +} + +static DEVICE_ATTR(cpumask, S_IRUGO, rapl_get_attr_cpumask, NULL); + +static struct attribute *rapl_pmu_attrs[] = { + &dev_attr_cpumask.attr, + NULL, +}; + +static struct attribute_group rapl_pmu_attr_group = { + .attrs = rapl_pmu_attrs, +}; + +EVENT_ATTR_STR(energy-cores, rapl_cores, "event=0x01"); +EVENT_ATTR_STR(energy-pkg , rapl_pkg, "event=0x02"); +EVENT_ATTR_STR(energy-ram , rapl_ram, "event=0x03"); +EVENT_ATTR_STR(energy-gpu , rapl_gpu, "event=0x04"); + +EVENT_ATTR_STR(energy-cores.unit, rapl_cores_unit, "Joules"); +EVENT_ATTR_STR(energy-pkg.unit , rapl_pkg_unit, "Joules"); +EVENT_ATTR_STR(energy-ram.unit , rapl_ram_unit, "Joules"); +EVENT_ATTR_STR(energy-gpu.unit , rapl_gpu_unit, "Joules"); + +/* + * we compute in 0.23 nJ increments regardless of MSR + */ +EVENT_ATTR_STR(energy-cores.scale, rapl_cores_scale, "2.3283064365386962890625e-10"); +EVENT_ATTR_STR(energy-pkg.scale, rapl_pkg_scale, "2.3283064365386962890625e-10"); +EVENT_ATTR_STR(energy-ram.scale, rapl_ram_scale, "2.3283064365386962890625e-10"); +EVENT_ATTR_STR(energy-gpu.scale, rapl_gpu_scale, "2.3283064365386962890625e-10"); + +static struct attribute *rapl_events_srv_attr[] = { + EVENT_PTR(rapl_cores), + EVENT_PTR(rapl_pkg), + EVENT_PTR(rapl_ram), + + EVENT_PTR(rapl_cores_unit), + EVENT_PTR(rapl_pkg_unit), + EVENT_PTR(rapl_ram_unit), + + EVENT_PTR(rapl_cores_scale), + EVENT_PTR(rapl_pkg_scale), + EVENT_PTR(rapl_ram_scale), + NULL, +}; + +static struct attribute *rapl_events_cln_attr[] = { + EVENT_PTR(rapl_cores), + EVENT_PTR(rapl_pkg), + EVENT_PTR(rapl_gpu), + + EVENT_PTR(rapl_cores_unit), + EVENT_PTR(rapl_pkg_unit), + EVENT_PTR(rapl_gpu_unit), + + EVENT_PTR(rapl_cores_scale), + EVENT_PTR(rapl_pkg_scale), + EVENT_PTR(rapl_gpu_scale), + NULL, +}; + +static struct attribute *rapl_events_hsw_attr[] = { + EVENT_PTR(rapl_cores), + EVENT_PTR(rapl_pkg), + EVENT_PTR(rapl_gpu), + EVENT_PTR(rapl_ram), + + EVENT_PTR(rapl_cores_unit), + EVENT_PTR(rapl_pkg_unit), + EVENT_PTR(rapl_gpu_unit), + EVENT_PTR(rapl_ram_unit), + + EVENT_PTR(rapl_cores_scale), + EVENT_PTR(rapl_pkg_scale), + EVENT_PTR(rapl_gpu_scale), + EVENT_PTR(rapl_ram_scale), + NULL, +}; + +static struct attribute_group rapl_pmu_events_group = { + .name = "events", + .attrs = NULL, /* patched at runtime */ +}; + +DEFINE_RAPL_FORMAT_ATTR(event, event, "config:0-7"); +static struct attribute *rapl_formats_attr[] = { + &format_attr_event.attr, + NULL, +}; + +static struct attribute_group rapl_pmu_format_group = { + .name = "format", + .attrs = rapl_formats_attr, +}; + +const struct attribute_group *rapl_attr_groups[] = { + &rapl_pmu_attr_group, + &rapl_pmu_format_group, + &rapl_pmu_events_group, + NULL, +}; + +static struct pmu rapl_pmu_class = { + .attr_groups = rapl_attr_groups, + .task_ctx_nr = perf_invalid_context, /* system-wide only */ + .event_init = rapl_pmu_event_init, + .add = rapl_pmu_event_add, /* must have */ + .del = rapl_pmu_event_del, /* must have */ + .start = rapl_pmu_event_start, + .stop = rapl_pmu_event_stop, + .read = rapl_pmu_event_read, +}; + +static void rapl_cpu_exit(int cpu) +{ + struct rapl_pmu *pmu = per_cpu(rapl_pmu, cpu); + int i, phys_id = topology_physical_package_id(cpu); + int target = -1; + + /* find a new cpu on same package */ + for_each_online_cpu(i) { + if (i == cpu) + continue; + if (phys_id == topology_physical_package_id(i)) { + target = i; + break; + } + } + /* + * clear cpu from cpumask + * if was set in cpumask and still some cpu on package, + * then move to new cpu + */ + if (cpumask_test_and_clear_cpu(cpu, &rapl_cpu_mask) && target >= 0) + cpumask_set_cpu(target, &rapl_cpu_mask); + + WARN_ON(cpumask_empty(&rapl_cpu_mask)); + /* + * migrate events and context to new cpu + */ + if (target >= 0) + perf_pmu_migrate_context(pmu->pmu, cpu, target); + + /* cancel overflow polling timer for CPU */ + rapl_stop_hrtimer(pmu); +} + +static void rapl_cpu_init(int cpu) +{ + int i, phys_id = topology_physical_package_id(cpu); + + /* check if phys_is is already covered */ + for_each_cpu(i, &rapl_cpu_mask) { + if (phys_id == topology_physical_package_id(i)) + return; + } + /* was not found, so add it */ + cpumask_set_cpu(cpu, &rapl_cpu_mask); +} + +static int rapl_cpu_prepare(int cpu) +{ + struct rapl_pmu *pmu = per_cpu(rapl_pmu, cpu); + int phys_id = topology_physical_package_id(cpu); + u64 ms; + u64 msr_rapl_power_unit_bits; + + if (pmu) + return 0; + + if (phys_id < 0) + return -1; + + /* protect rdmsrl() to handle virtualization */ + if (rdmsrl_safe(MSR_RAPL_POWER_UNIT, &msr_rapl_power_unit_bits)) + return -1; + + pmu = kzalloc_node(sizeof(*pmu), GFP_KERNEL, cpu_to_node(cpu)); + if (!pmu) + return -1; + + spin_lock_init(&pmu->lock); + + INIT_LIST_HEAD(&pmu->active_list); + + /* + * grab power unit as: 1/2^unit Joules + * + * we cache in local PMU instance + */ + pmu->hw_unit = (msr_rapl_power_unit_bits >> 8) & 0x1FULL; + pmu->pmu = &rapl_pmu_class; + + /* + * use reference of 200W for scaling the timeout + * to avoid missing counter overflows. + * 200W = 200 Joules/sec + * divide interval by 2 to avoid lockstep (2 * 100) + * if hw unit is 32, then we use 2 ms 1/200/2 + */ + if (pmu->hw_unit < 32) + ms = (1000 / (2 * 100)) * (1ULL << (32 - pmu->hw_unit - 1)); + else + ms = 2; + + pmu->timer_interval = ms_to_ktime(ms); + + rapl_hrtimer_init(pmu); + + /* set RAPL pmu for this cpu for now */ + per_cpu(rapl_pmu, cpu) = pmu; + per_cpu(rapl_pmu_to_free, cpu) = NULL; + + return 0; +} + +static void rapl_cpu_kfree(int cpu) +{ + struct rapl_pmu *pmu = per_cpu(rapl_pmu_to_free, cpu); + + kfree(pmu); + + per_cpu(rapl_pmu_to_free, cpu) = NULL; +} + +static int rapl_cpu_dying(int cpu) +{ + struct rapl_pmu *pmu = per_cpu(rapl_pmu, cpu); + + if (!pmu) + return 0; + + per_cpu(rapl_pmu, cpu) = NULL; + + per_cpu(rapl_pmu_to_free, cpu) = pmu; + + return 0; +} + +static int rapl_cpu_notifier(struct notifier_block *self, + unsigned long action, void *hcpu) +{ + unsigned int cpu = (long)hcpu; + + switch (action & ~CPU_TASKS_FROZEN) { + case CPU_UP_PREPARE: + rapl_cpu_prepare(cpu); + break; + case CPU_STARTING: + rapl_cpu_init(cpu); + break; + case CPU_UP_CANCELED: + case CPU_DYING: + rapl_cpu_dying(cpu); + break; + case CPU_ONLINE: + case CPU_DEAD: + rapl_cpu_kfree(cpu); + break; + case CPU_DOWN_PREPARE: + rapl_cpu_exit(cpu); + break; + default: + break; + } + + return NOTIFY_OK; +} + +static const struct x86_cpu_id rapl_cpu_match[] = { + [0] = { .vendor = X86_VENDOR_INTEL, .family = 6 }, + [1] = {}, +}; + +static int __init rapl_pmu_init(void) +{ + struct rapl_pmu *pmu; + int cpu, ret; + + /* + * check for Intel processor family 6 + */ + if (!x86_match_cpu(rapl_cpu_match)) + return 0; + + /* check supported CPU */ + switch (boot_cpu_data.x86_model) { + case 42: /* Sandy Bridge */ + case 58: /* Ivy Bridge */ + rapl_cntr_mask = RAPL_IDX_CLN; + rapl_pmu_events_group.attrs = rapl_events_cln_attr; + break; + case 60: /* Haswell */ + case 69: /* Haswell-Celeron */ + rapl_cntr_mask = RAPL_IDX_HSW; + rapl_pmu_events_group.attrs = rapl_events_hsw_attr; + break; + case 45: /* Sandy Bridge-EP */ + case 62: /* IvyTown */ + rapl_cntr_mask = RAPL_IDX_SRV; + rapl_pmu_events_group.attrs = rapl_events_srv_attr; + break; + + default: + /* unsupported */ + return 0; + } + + cpu_notifier_register_begin(); + + for_each_online_cpu(cpu) { + ret = rapl_cpu_prepare(cpu); + if (ret) + goto out; + rapl_cpu_init(cpu); + } + + __perf_cpu_notifier(rapl_cpu_notifier); + + ret = perf_pmu_register(&rapl_pmu_class, "power", -1); + if (WARN_ON(ret)) { + pr_info("RAPL PMU detected, registration failed (%d), RAPL PMU disabled\n", ret); + cpu_notifier_register_done(); + return -1; + } + + pmu = __get_cpu_var(rapl_pmu); + + pr_info("RAPL PMU detected, hw unit 2^-%d Joules," + " API unit is 2^-32 Joules," + " %d fixed counters" + " %llu ms ovfl timer\n", + pmu->hw_unit, + hweight32(rapl_cntr_mask), + ktime_to_ms(pmu->timer_interval)); + +out: + cpu_notifier_register_done(); + + return 0; +} +device_initcall(rapl_pmu_init); diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore.c b/arch/x86/kernel/cpu/perf_event_intel_uncore.c index 4118f9f6831..ae6552a0701 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_uncore.c +++ b/arch/x86/kernel/cpu/perf_event_intel_uncore.c @@ -66,6 +66,47 @@ DEFINE_UNCORE_FORMAT_ATTR(mask_vnw, mask_vnw, "config2:3-4"); DEFINE_UNCORE_FORMAT_ATTR(mask0, mask0, "config2:0-31"); DEFINE_UNCORE_FORMAT_ATTR(mask1, mask1, "config2:32-63"); +static void uncore_pmu_start_hrtimer(struct intel_uncore_box *box); +static void uncore_pmu_cancel_hrtimer(struct intel_uncore_box *box); +static void uncore_perf_event_update(struct intel_uncore_box *box, struct perf_event *event); +static void uncore_pmu_event_read(struct perf_event *event); + +static struct intel_uncore_pmu *uncore_event_to_pmu(struct perf_event *event) +{ + return container_of(event->pmu, struct intel_uncore_pmu, pmu); +} + +static struct intel_uncore_box * +uncore_pmu_to_box(struct intel_uncore_pmu *pmu, int cpu) +{ + struct intel_uncore_box *box; + + box = *per_cpu_ptr(pmu->box, cpu); + if (box) + return box; + + raw_spin_lock(&uncore_box_lock); + list_for_each_entry(box, &pmu->box_list, list) { + if (box->phys_id == topology_physical_package_id(cpu)) { + atomic_inc(&box->refcnt); + *per_cpu_ptr(pmu->box, cpu) = box; + break; + } + } + raw_spin_unlock(&uncore_box_lock); + + return *per_cpu_ptr(pmu->box, cpu); +} + +static struct intel_uncore_box *uncore_event_to_box(struct perf_event *event) +{ + /* + * perf core schedules event on the basis of cpu, uncore events are + * collected by one of the cpus inside a physical package. + */ + return uncore_pmu_to_box(uncore_event_to_pmu(event), smp_processor_id()); +} + static u64 uncore_msr_read_counter(struct intel_uncore_box *box, struct perf_event *event) { u64 count; @@ -501,21 +542,24 @@ static struct extra_reg snbep_uncore_cbox_extra_regs[] = { SNBEP_CBO_EVENT_EXTRA_REG(SNBEP_CBO_PMON_CTL_TID_EN, SNBEP_CBO_PMON_CTL_TID_EN, 0x1), SNBEP_CBO_EVENT_EXTRA_REG(0x0334, 0xffff, 0x4), + SNBEP_CBO_EVENT_EXTRA_REG(0x4334, 0xffff, 0x6), SNBEP_CBO_EVENT_EXTRA_REG(0x0534, 0xffff, 0x4), + SNBEP_CBO_EVENT_EXTRA_REG(0x4534, 0xffff, 0x6), SNBEP_CBO_EVENT_EXTRA_REG(0x0934, 0xffff, 0x4), + SNBEP_CBO_EVENT_EXTRA_REG(0x4934, 0xffff, 0x6), SNBEP_CBO_EVENT_EXTRA_REG(0x4134, 0xffff, 0x6), SNBEP_CBO_EVENT_EXTRA_REG(0x0135, 0xffff, 0x8), SNBEP_CBO_EVENT_EXTRA_REG(0x0335, 0xffff, 0x8), - SNBEP_CBO_EVENT_EXTRA_REG(0x4135, 0xffff, 0xc), - SNBEP_CBO_EVENT_EXTRA_REG(0x4335, 0xffff, 0xc), + SNBEP_CBO_EVENT_EXTRA_REG(0x4135, 0xffff, 0xa), + SNBEP_CBO_EVENT_EXTRA_REG(0x4335, 0xffff, 0xa), SNBEP_CBO_EVENT_EXTRA_REG(0x4435, 0xffff, 0x2), SNBEP_CBO_EVENT_EXTRA_REG(0x4835, 0xffff, 0x2), SNBEP_CBO_EVENT_EXTRA_REG(0x4a35, 0xffff, 0x2), SNBEP_CBO_EVENT_EXTRA_REG(0x5035, 0xffff, 0x2), SNBEP_CBO_EVENT_EXTRA_REG(0x0136, 0xffff, 0x8), SNBEP_CBO_EVENT_EXTRA_REG(0x0336, 0xffff, 0x8), - SNBEP_CBO_EVENT_EXTRA_REG(0x4136, 0xffff, 0xc), - SNBEP_CBO_EVENT_EXTRA_REG(0x4336, 0xffff, 0xc), + SNBEP_CBO_EVENT_EXTRA_REG(0x4136, 0xffff, 0xa), + SNBEP_CBO_EVENT_EXTRA_REG(0x4336, 0xffff, 0xa), SNBEP_CBO_EVENT_EXTRA_REG(0x4436, 0xffff, 0x2), SNBEP_CBO_EVENT_EXTRA_REG(0x4836, 0xffff, 0x2), SNBEP_CBO_EVENT_EXTRA_REG(0x4a36, 0xffff, 0x2), @@ -997,6 +1041,20 @@ static int snbep_pci2phy_map_init(int devid) } } + if (!err) { + /* + * For PCI bus with no UBOX device, find the next bus + * that has UBOX device and use its mapping. + */ + i = -1; + for (bus = 255; bus >= 0; bus--) { + if (pcibus_to_physid[bus] >= 0) + i = pcibus_to_physid[bus]; + else + pcibus_to_physid[bus] = i; + } + } + if (ubox_dev) pci_dev_put(ubox_dev); @@ -1099,6 +1157,24 @@ static struct attribute *ivt_uncore_qpi_formats_attr[] = { &format_attr_umask.attr, &format_attr_edge.attr, &format_attr_thresh8.attr, + &format_attr_match_rds.attr, + &format_attr_match_rnid30.attr, + &format_attr_match_rnid4.attr, + &format_attr_match_dnid.attr, + &format_attr_match_mc.attr, + &format_attr_match_opc.attr, + &format_attr_match_vnw.attr, + &format_attr_match0.attr, + &format_attr_match1.attr, + &format_attr_mask_rds.attr, + &format_attr_mask_rnid30.attr, + &format_attr_mask_rnid4.attr, + &format_attr_mask_dnid.attr, + &format_attr_mask_mc.attr, + &format_attr_mask_opc.attr, + &format_attr_mask_vnw.attr, + &format_attr_mask0.attr, + &format_attr_mask1.attr, NULL, }; @@ -1146,10 +1222,16 @@ static struct extra_reg ivt_uncore_cbox_extra_regs[] = { SNBEP_CBO_EVENT_EXTRA_REG(SNBEP_CBO_PMON_CTL_TID_EN, SNBEP_CBO_PMON_CTL_TID_EN, 0x1), SNBEP_CBO_EVENT_EXTRA_REG(0x1031, 0x10ff, 0x2), + + SNBEP_CBO_EVENT_EXTRA_REG(0x1134, 0xffff, 0x4), + SNBEP_CBO_EVENT_EXTRA_REG(0x4134, 0xffff, 0xc), + SNBEP_CBO_EVENT_EXTRA_REG(0x5134, 0xffff, 0xc), SNBEP_CBO_EVENT_EXTRA_REG(0x0334, 0xffff, 0x4), + SNBEP_CBO_EVENT_EXTRA_REG(0x4334, 0xffff, 0xc), SNBEP_CBO_EVENT_EXTRA_REG(0x0534, 0xffff, 0x4), + SNBEP_CBO_EVENT_EXTRA_REG(0x4534, 0xffff, 0xc), SNBEP_CBO_EVENT_EXTRA_REG(0x0934, 0xffff, 0x4), - SNBEP_CBO_EVENT_EXTRA_REG(0x4134, 0xffff, 0xc), + SNBEP_CBO_EVENT_EXTRA_REG(0x4934, 0xffff, 0xc), SNBEP_CBO_EVENT_EXTRA_REG(0x0135, 0xffff, 0x10), SNBEP_CBO_EVENT_EXTRA_REG(0x0335, 0xffff, 0x10), SNBEP_CBO_EVENT_EXTRA_REG(0x2135, 0xffff, 0x10), @@ -1164,7 +1246,7 @@ static struct extra_reg ivt_uncore_cbox_extra_regs[] = { SNBEP_CBO_EVENT_EXTRA_REG(0x8335, 0xffff, 0x10), SNBEP_CBO_EVENT_EXTRA_REG(0x0136, 0xffff, 0x10), SNBEP_CBO_EVENT_EXTRA_REG(0x0336, 0xffff, 0x10), - SNBEP_CBO_EVENT_EXTRA_REG(0x2336, 0xffff, 0x10), + SNBEP_CBO_EVENT_EXTRA_REG(0x2136, 0xffff, 0x10), SNBEP_CBO_EVENT_EXTRA_REG(0x2336, 0xffff, 0x10), SNBEP_CBO_EVENT_EXTRA_REG(0x4136, 0xffff, 0x18), SNBEP_CBO_EVENT_EXTRA_REG(0x4336, 0xffff, 0x18), @@ -1312,17 +1394,83 @@ static struct intel_uncore_type ivt_uncore_imc = { IVT_UNCORE_PCI_COMMON_INIT(), }; +/* registers in IRP boxes are not properly aligned */ +static unsigned ivt_uncore_irp_ctls[] = {0xd8, 0xdc, 0xe0, 0xe4}; +static unsigned ivt_uncore_irp_ctrs[] = {0xa0, 0xb0, 0xb8, 0xc0}; + +static void ivt_uncore_irp_enable_event(struct intel_uncore_box *box, struct perf_event *event) +{ + struct pci_dev *pdev = box->pci_dev; + struct hw_perf_event *hwc = &event->hw; + + pci_write_config_dword(pdev, ivt_uncore_irp_ctls[hwc->idx], + hwc->config | SNBEP_PMON_CTL_EN); +} + +static void ivt_uncore_irp_disable_event(struct intel_uncore_box *box, struct perf_event *event) +{ + struct pci_dev *pdev = box->pci_dev; + struct hw_perf_event *hwc = &event->hw; + + pci_write_config_dword(pdev, ivt_uncore_irp_ctls[hwc->idx], hwc->config); +} + +static u64 ivt_uncore_irp_read_counter(struct intel_uncore_box *box, struct perf_event *event) +{ + struct pci_dev *pdev = box->pci_dev; + struct hw_perf_event *hwc = &event->hw; + u64 count = 0; + + pci_read_config_dword(pdev, ivt_uncore_irp_ctrs[hwc->idx], (u32 *)&count); + pci_read_config_dword(pdev, ivt_uncore_irp_ctrs[hwc->idx] + 4, (u32 *)&count + 1); + + return count; +} + +static struct intel_uncore_ops ivt_uncore_irp_ops = { + .init_box = ivt_uncore_pci_init_box, + .disable_box = snbep_uncore_pci_disable_box, + .enable_box = snbep_uncore_pci_enable_box, + .disable_event = ivt_uncore_irp_disable_event, + .enable_event = ivt_uncore_irp_enable_event, + .read_counter = ivt_uncore_irp_read_counter, +}; + +static struct intel_uncore_type ivt_uncore_irp = { + .name = "irp", + .num_counters = 4, + .num_boxes = 1, + .perf_ctr_bits = 48, + .event_mask = IVT_PMON_RAW_EVENT_MASK, + .box_ctl = SNBEP_PCI_PMON_BOX_CTL, + .ops = &ivt_uncore_irp_ops, + .format_group = &ivt_uncore_format_group, +}; + +static struct intel_uncore_ops ivt_uncore_qpi_ops = { + .init_box = ivt_uncore_pci_init_box, + .disable_box = snbep_uncore_pci_disable_box, + .enable_box = snbep_uncore_pci_enable_box, + .disable_event = snbep_uncore_pci_disable_event, + .enable_event = snbep_qpi_enable_event, + .read_counter = snbep_uncore_pci_read_counter, + .hw_config = snbep_qpi_hw_config, + .get_constraint = uncore_get_constraint, + .put_constraint = uncore_put_constraint, +}; + static struct intel_uncore_type ivt_uncore_qpi = { - .name = "qpi", - .num_counters = 4, - .num_boxes = 3, - .perf_ctr_bits = 48, - .perf_ctr = SNBEP_PCI_PMON_CTR0, - .event_ctl = SNBEP_PCI_PMON_CTL0, - .event_mask = IVT_QPI_PCI_PMON_RAW_EVENT_MASK, - .box_ctl = SNBEP_PCI_PMON_BOX_CTL, - .ops = &ivt_uncore_pci_ops, - .format_group = &ivt_uncore_qpi_format_group, + .name = "qpi", + .num_counters = 4, + .num_boxes = 3, + .perf_ctr_bits = 48, + .perf_ctr = SNBEP_PCI_PMON_CTR0, + .event_ctl = SNBEP_PCI_PMON_CTL0, + .event_mask = IVT_QPI_PCI_PMON_RAW_EVENT_MASK, + .box_ctl = SNBEP_PCI_PMON_BOX_CTL, + .num_shared_regs = 1, + .ops = &ivt_uncore_qpi_ops, + .format_group = &ivt_uncore_qpi_format_group, }; static struct intel_uncore_type ivt_uncore_r2pcie = { @@ -1346,6 +1494,7 @@ static struct intel_uncore_type ivt_uncore_r3qpi = { enum { IVT_PCI_UNCORE_HA, IVT_PCI_UNCORE_IMC, + IVT_PCI_UNCORE_IRP, IVT_PCI_UNCORE_QPI, IVT_PCI_UNCORE_R2PCIE, IVT_PCI_UNCORE_R3QPI, @@ -1354,6 +1503,7 @@ enum { static struct intel_uncore_type *ivt_pci_uncores[] = { [IVT_PCI_UNCORE_HA] = &ivt_uncore_ha, [IVT_PCI_UNCORE_IMC] = &ivt_uncore_imc, + [IVT_PCI_UNCORE_IRP] = &ivt_uncore_irp, [IVT_PCI_UNCORE_QPI] = &ivt_uncore_qpi, [IVT_PCI_UNCORE_R2PCIE] = &ivt_uncore_r2pcie, [IVT_PCI_UNCORE_R3QPI] = &ivt_uncore_r3qpi, @@ -1401,6 +1551,10 @@ static DEFINE_PCI_DEVICE_TABLE(ivt_uncore_pci_ids) = { PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xef1), .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_IMC, 7), }, + { /* IRP */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xe39), + .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_IRP, 0), + }, { /* QPI0 Port 0 */ PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xe32), .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_QPI, 0), @@ -1429,6 +1583,16 @@ static DEFINE_PCI_DEVICE_TABLE(ivt_uncore_pci_ids) = { PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xe3e), .driver_data = UNCORE_PCI_DEV_DATA(IVT_PCI_UNCORE_R3QPI, 2), }, + { /* QPI Port 0 filter */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xe86), + .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, + SNBEP_PCI_QPI_PORT0_FILTER), + }, + { /* QPI Port 0 filter */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, 0xe96), + .driver_data = UNCORE_PCI_DEV_DATA(UNCORE_EXTRA_PCI_DEV, + SNBEP_PCI_QPI_PORT1_FILTER), + }, { /* end: all zeroes */ } }; @@ -1517,6 +1681,349 @@ static struct intel_uncore_type *snb_msr_uncores[] = { &snb_uncore_cbox, NULL, }; + +enum { + SNB_PCI_UNCORE_IMC, +}; + +static struct uncore_event_desc snb_uncore_imc_events[] = { + INTEL_UNCORE_EVENT_DESC(data_reads, "event=0x01"), + INTEL_UNCORE_EVENT_DESC(data_reads.scale, "6.103515625e-5"), + INTEL_UNCORE_EVENT_DESC(data_reads.unit, "MiB"), + + INTEL_UNCORE_EVENT_DESC(data_writes, "event=0x02"), + INTEL_UNCORE_EVENT_DESC(data_writes.scale, "6.103515625e-5"), + INTEL_UNCORE_EVENT_DESC(data_writes.unit, "MiB"), + + { /* end: all zeroes */ }, +}; + +#define SNB_UNCORE_PCI_IMC_EVENT_MASK 0xff +#define SNB_UNCORE_PCI_IMC_BAR_OFFSET 0x48 + +/* page size multiple covering all config regs */ +#define SNB_UNCORE_PCI_IMC_MAP_SIZE 0x6000 + +#define SNB_UNCORE_PCI_IMC_DATA_READS 0x1 +#define SNB_UNCORE_PCI_IMC_DATA_READS_BASE 0x5050 +#define SNB_UNCORE_PCI_IMC_DATA_WRITES 0x2 +#define SNB_UNCORE_PCI_IMC_DATA_WRITES_BASE 0x5054 +#define SNB_UNCORE_PCI_IMC_CTR_BASE SNB_UNCORE_PCI_IMC_DATA_READS_BASE + +static struct attribute *snb_uncore_imc_formats_attr[] = { + &format_attr_event.attr, + NULL, +}; + +static struct attribute_group snb_uncore_imc_format_group = { + .name = "format", + .attrs = snb_uncore_imc_formats_attr, +}; + +static void snb_uncore_imc_init_box(struct intel_uncore_box *box) +{ + struct pci_dev *pdev = box->pci_dev; + int where = SNB_UNCORE_PCI_IMC_BAR_OFFSET; + resource_size_t addr; + u32 pci_dword; + + pci_read_config_dword(pdev, where, &pci_dword); + addr = pci_dword; + +#ifdef CONFIG_PHYS_ADDR_T_64BIT + pci_read_config_dword(pdev, where + 4, &pci_dword); + addr |= ((resource_size_t)pci_dword << 32); +#endif + + addr &= ~(PAGE_SIZE - 1); + + box->io_addr = ioremap(addr, SNB_UNCORE_PCI_IMC_MAP_SIZE); + box->hrtimer_duration = UNCORE_SNB_IMC_HRTIMER_INTERVAL; +} + +static void snb_uncore_imc_enable_box(struct intel_uncore_box *box) +{} + +static void snb_uncore_imc_disable_box(struct intel_uncore_box *box) +{} + +static void snb_uncore_imc_enable_event(struct intel_uncore_box *box, struct perf_event *event) +{} + +static void snb_uncore_imc_disable_event(struct intel_uncore_box *box, struct perf_event *event) +{} + +static u64 snb_uncore_imc_read_counter(struct intel_uncore_box *box, struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + + return (u64)*(unsigned int *)(box->io_addr + hwc->event_base); +} + +/* + * custom event_init() function because we define our own fixed, free + * running counters, so we do not want to conflict with generic uncore + * logic. Also simplifies processing + */ +static int snb_uncore_imc_event_init(struct perf_event *event) +{ + struct intel_uncore_pmu *pmu; + struct intel_uncore_box *box; + struct hw_perf_event *hwc = &event->hw; + u64 cfg = event->attr.config & SNB_UNCORE_PCI_IMC_EVENT_MASK; + int idx, base; + + if (event->attr.type != event->pmu->type) + return -ENOENT; + + pmu = uncore_event_to_pmu(event); + /* no device found for this pmu */ + if (pmu->func_id < 0) + return -ENOENT; + + /* Sampling not supported yet */ + if (hwc->sample_period) + return -EINVAL; + + /* unsupported modes and filters */ + if (event->attr.exclude_user || + event->attr.exclude_kernel || + event->attr.exclude_hv || + event->attr.exclude_idle || + event->attr.exclude_host || + event->attr.exclude_guest || + event->attr.sample_period) /* no sampling */ + return -EINVAL; + + /* + * Place all uncore events for a particular physical package + * onto a single cpu + */ + if (event->cpu < 0) + return -EINVAL; + + /* check only supported bits are set */ + if (event->attr.config & ~SNB_UNCORE_PCI_IMC_EVENT_MASK) + return -EINVAL; + + box = uncore_pmu_to_box(pmu, event->cpu); + if (!box || box->cpu < 0) + return -EINVAL; + + event->cpu = box->cpu; + + event->hw.idx = -1; + event->hw.last_tag = ~0ULL; + event->hw.extra_reg.idx = EXTRA_REG_NONE; + event->hw.branch_reg.idx = EXTRA_REG_NONE; + /* + * check event is known (whitelist, determines counter) + */ + switch (cfg) { + case SNB_UNCORE_PCI_IMC_DATA_READS: + base = SNB_UNCORE_PCI_IMC_DATA_READS_BASE; + idx = UNCORE_PMC_IDX_FIXED; + break; + case SNB_UNCORE_PCI_IMC_DATA_WRITES: + base = SNB_UNCORE_PCI_IMC_DATA_WRITES_BASE; + idx = UNCORE_PMC_IDX_FIXED + 1; + break; + default: + return -EINVAL; + } + + /* must be done before validate_group */ + event->hw.event_base = base; + event->hw.config = cfg; + event->hw.idx = idx; + + /* no group validation needed, we have free running counters */ + + return 0; +} + +static int snb_uncore_imc_hw_config(struct intel_uncore_box *box, struct perf_event *event) +{ + return 0; +} + +static void snb_uncore_imc_event_start(struct perf_event *event, int flags) +{ + struct intel_uncore_box *box = uncore_event_to_box(event); + u64 count; + + if (WARN_ON_ONCE(!(event->hw.state & PERF_HES_STOPPED))) + return; + + event->hw.state = 0; + box->n_active++; + + list_add_tail(&event->active_entry, &box->active_list); + + count = snb_uncore_imc_read_counter(box, event); + local64_set(&event->hw.prev_count, count); + + if (box->n_active == 1) + uncore_pmu_start_hrtimer(box); +} + +static void snb_uncore_imc_event_stop(struct perf_event *event, int flags) +{ + struct intel_uncore_box *box = uncore_event_to_box(event); + struct hw_perf_event *hwc = &event->hw; + + if (!(hwc->state & PERF_HES_STOPPED)) { + box->n_active--; + + WARN_ON_ONCE(hwc->state & PERF_HES_STOPPED); + hwc->state |= PERF_HES_STOPPED; + + list_del(&event->active_entry); + + if (box->n_active == 0) + uncore_pmu_cancel_hrtimer(box); + } + + if ((flags & PERF_EF_UPDATE) && !(hwc->state & PERF_HES_UPTODATE)) { + /* + * Drain the remaining delta count out of a event + * that we are disabling: + */ + uncore_perf_event_update(box, event); + hwc->state |= PERF_HES_UPTODATE; + } +} + +static int snb_uncore_imc_event_add(struct perf_event *event, int flags) +{ + struct intel_uncore_box *box = uncore_event_to_box(event); + struct hw_perf_event *hwc = &event->hw; + + if (!box) + return -ENODEV; + + hwc->state = PERF_HES_UPTODATE | PERF_HES_STOPPED; + if (!(flags & PERF_EF_START)) + hwc->state |= PERF_HES_ARCH; + + snb_uncore_imc_event_start(event, 0); + + box->n_events++; + + return 0; +} + +static void snb_uncore_imc_event_del(struct perf_event *event, int flags) +{ + struct intel_uncore_box *box = uncore_event_to_box(event); + int i; + + snb_uncore_imc_event_stop(event, PERF_EF_UPDATE); + + for (i = 0; i < box->n_events; i++) { + if (event == box->event_list[i]) { + --box->n_events; + break; + } + } +} + +static int snb_pci2phy_map_init(int devid) +{ + struct pci_dev *dev = NULL; + int bus; + + dev = pci_get_device(PCI_VENDOR_ID_INTEL, devid, dev); + if (!dev) + return -ENOTTY; + + bus = dev->bus->number; + + pcibus_to_physid[bus] = 0; + + pci_dev_put(dev); + + return 0; +} + +static struct pmu snb_uncore_imc_pmu = { + .task_ctx_nr = perf_invalid_context, + .event_init = snb_uncore_imc_event_init, + .add = snb_uncore_imc_event_add, + .del = snb_uncore_imc_event_del, + .start = snb_uncore_imc_event_start, + .stop = snb_uncore_imc_event_stop, + .read = uncore_pmu_event_read, +}; + +static struct intel_uncore_ops snb_uncore_imc_ops = { + .init_box = snb_uncore_imc_init_box, + .enable_box = snb_uncore_imc_enable_box, + .disable_box = snb_uncore_imc_disable_box, + .disable_event = snb_uncore_imc_disable_event, + .enable_event = snb_uncore_imc_enable_event, + .hw_config = snb_uncore_imc_hw_config, + .read_counter = snb_uncore_imc_read_counter, +}; + +static struct intel_uncore_type snb_uncore_imc = { + .name = "imc", + .num_counters = 2, + .num_boxes = 1, + .fixed_ctr_bits = 32, + .fixed_ctr = SNB_UNCORE_PCI_IMC_CTR_BASE, + .event_descs = snb_uncore_imc_events, + .format_group = &snb_uncore_imc_format_group, + .perf_ctr = SNB_UNCORE_PCI_IMC_DATA_READS_BASE, + .event_mask = SNB_UNCORE_PCI_IMC_EVENT_MASK, + .ops = &snb_uncore_imc_ops, + .pmu = &snb_uncore_imc_pmu, +}; + +static struct intel_uncore_type *snb_pci_uncores[] = { + [SNB_PCI_UNCORE_IMC] = &snb_uncore_imc, + NULL, +}; + +static DEFINE_PCI_DEVICE_TABLE(snb_uncore_pci_ids) = { + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_SNB_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* end: all zeroes */ }, +}; + +static DEFINE_PCI_DEVICE_TABLE(ivb_uncore_pci_ids) = { + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_IVB_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* end: all zeroes */ }, +}; + +static DEFINE_PCI_DEVICE_TABLE(hsw_uncore_pci_ids) = { + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_HSW_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* end: all zeroes */ }, +}; + +static struct pci_driver snb_uncore_pci_driver = { + .name = "snb_uncore", + .id_table = snb_uncore_pci_ids, +}; + +static struct pci_driver ivb_uncore_pci_driver = { + .name = "ivb_uncore", + .id_table = ivb_uncore_pci_ids, +}; + +static struct pci_driver hsw_uncore_pci_driver = { + .name = "hsw_uncore", + .id_table = hsw_uncore_pci_ids, +}; + /* end of Sandy Bridge uncore support */ /* Nehalem uncore support */ @@ -2667,6 +3174,7 @@ again: static enum hrtimer_restart uncore_pmu_hrtimer(struct hrtimer *hrtimer) { struct intel_uncore_box *box; + struct perf_event *event; unsigned long flags; int bit; @@ -2679,19 +3187,27 @@ static enum hrtimer_restart uncore_pmu_hrtimer(struct hrtimer *hrtimer) */ local_irq_save(flags); + /* + * handle boxes with an active event list as opposed to active + * counters + */ + list_for_each_entry(event, &box->active_list, active_entry) { + uncore_perf_event_update(box, event); + } + for_each_set_bit(bit, box->active_mask, UNCORE_PMC_IDX_MAX) uncore_perf_event_update(box, box->events[bit]); local_irq_restore(flags); - hrtimer_forward_now(hrtimer, ns_to_ktime(UNCORE_PMU_HRTIMER_INTERVAL)); + hrtimer_forward_now(hrtimer, ns_to_ktime(box->hrtimer_duration)); return HRTIMER_RESTART; } static void uncore_pmu_start_hrtimer(struct intel_uncore_box *box) { __hrtimer_start_range_ns(&box->hrtimer, - ns_to_ktime(UNCORE_PMU_HRTIMER_INTERVAL), 0, + ns_to_ktime(box->hrtimer_duration), 0, HRTIMER_MODE_REL_PINNED, 0); } @@ -2725,43 +3241,12 @@ static struct intel_uncore_box *uncore_alloc_box(struct intel_uncore_type *type, box->cpu = -1; box->phys_id = -1; - return box; -} - -static struct intel_uncore_box * -uncore_pmu_to_box(struct intel_uncore_pmu *pmu, int cpu) -{ - struct intel_uncore_box *box; - - box = *per_cpu_ptr(pmu->box, cpu); - if (box) - return box; - - raw_spin_lock(&uncore_box_lock); - list_for_each_entry(box, &pmu->box_list, list) { - if (box->phys_id == topology_physical_package_id(cpu)) { - atomic_inc(&box->refcnt); - *per_cpu_ptr(pmu->box, cpu) = box; - break; - } - } - raw_spin_unlock(&uncore_box_lock); - - return *per_cpu_ptr(pmu->box, cpu); -} + /* set default hrtimer timeout */ + box->hrtimer_duration = UNCORE_PMU_HRTIMER_INTERVAL; -static struct intel_uncore_pmu *uncore_event_to_pmu(struct perf_event *event) -{ - return container_of(event->pmu, struct intel_uncore_pmu, pmu); -} + INIT_LIST_HEAD(&box->active_list); -static struct intel_uncore_box *uncore_event_to_box(struct perf_event *event) -{ - /* - * perf core schedules event on the basis of cpu, uncore events are - * collected by one of the cpus inside a physical package. - */ - return uncore_pmu_to_box(uncore_event_to_pmu(event), smp_processor_id()); + return box; } static int @@ -3157,16 +3642,21 @@ static int __init uncore_pmu_register(struct intel_uncore_pmu *pmu) { int ret; - pmu->pmu = (struct pmu) { - .attr_groups = pmu->type->attr_groups, - .task_ctx_nr = perf_invalid_context, - .event_init = uncore_pmu_event_init, - .add = uncore_pmu_event_add, - .del = uncore_pmu_event_del, - .start = uncore_pmu_event_start, - .stop = uncore_pmu_event_stop, - .read = uncore_pmu_event_read, - }; + if (!pmu->type->pmu) { + pmu->pmu = (struct pmu) { + .attr_groups = pmu->type->attr_groups, + .task_ctx_nr = perf_invalid_context, + .event_init = uncore_pmu_event_init, + .add = uncore_pmu_event_add, + .del = uncore_pmu_event_del, + .start = uncore_pmu_event_start, + .stop = uncore_pmu_event_stop, + .read = uncore_pmu_event_read, + }; + } else { + pmu->pmu = *pmu->type->pmu; + pmu->pmu.attr_groups = pmu->type->attr_groups; + } if (pmu->type->num_boxes == 1) { if (strlen(pmu->type->name) > 0) @@ -3212,6 +3702,8 @@ static int __init uncore_type_init(struct intel_uncore_type *type) if (!pmus) return -ENOMEM; + type->pmus = pmus; + type->unconstrainted = (struct event_constraint) __EVENT_CONSTRAINT(0, (1ULL << type->num_counters) - 1, 0, type->num_counters, 0, 0); @@ -3247,7 +3739,6 @@ static int __init uncore_type_init(struct intel_uncore_type *type) } type->pmu_group = &uncore_pmu_attr_group; - type->pmus = pmus; return 0; fail: uncore_type_exit(type); @@ -3379,6 +3870,28 @@ static int __init uncore_pci_init(void) pci_uncores = ivt_pci_uncores; uncore_pci_driver = &ivt_uncore_pci_driver; break; + case 42: /* Sandy Bridge */ + ret = snb_pci2phy_map_init(PCI_DEVICE_ID_INTEL_SNB_IMC); + if (ret) + return ret; + pci_uncores = snb_pci_uncores; + uncore_pci_driver = &snb_uncore_pci_driver; + break; + case 58: /* Ivy Bridge */ + ret = snb_pci2phy_map_init(PCI_DEVICE_ID_INTEL_IVB_IMC); + if (ret) + return ret; + pci_uncores = snb_pci_uncores; + uncore_pci_driver = &ivb_uncore_pci_driver; + break; + case 60: /* Haswell */ + case 69: /* Haswell Celeron */ + ret = snb_pci2phy_map_init(PCI_DEVICE_ID_INTEL_HSW_IMC); + if (ret) + return ret; + pci_uncores = snb_pci_uncores; + uncore_pci_driver = &hsw_uncore_pci_driver; + break; default: return 0; } @@ -3650,7 +4163,7 @@ static void __init uncore_cpu_setup(void *dummy) static int __init uncore_cpu_init(void) { - int ret, cpu, max_cores; + int ret, max_cores; max_cores = boot_cpu_data.x86_max_cores; switch (boot_cpu_data.x86_model) { @@ -3694,29 +4207,6 @@ static int __init uncore_cpu_init(void) if (ret) return ret; - get_online_cpus(); - - for_each_online_cpu(cpu) { - int i, phys_id = topology_physical_package_id(cpu); - - for_each_cpu(i, &uncore_cpu_mask) { - if (phys_id == topology_physical_package_id(i)) { - phys_id = -1; - break; - } - } - if (phys_id < 0) - continue; - - uncore_cpu_prepare(cpu, phys_id); - uncore_event_init_cpu(cpu); - } - on_each_cpu(uncore_cpu_setup, NULL, 1); - - register_cpu_notifier(&uncore_cpu_nb); - - put_online_cpus(); - return 0; } @@ -3745,6 +4235,41 @@ static int __init uncore_pmus_register(void) return 0; } +static void __init uncore_cpumask_init(void) +{ + int cpu; + + /* + * ony invoke once from msr or pci init code + */ + if (!cpumask_empty(&uncore_cpu_mask)) + return; + + cpu_notifier_register_begin(); + + for_each_online_cpu(cpu) { + int i, phys_id = topology_physical_package_id(cpu); + + for_each_cpu(i, &uncore_cpu_mask) { + if (phys_id == topology_physical_package_id(i)) { + phys_id = -1; + break; + } + } + if (phys_id < 0) + continue; + + uncore_cpu_prepare(cpu, phys_id); + uncore_event_init_cpu(cpu); + } + on_each_cpu(uncore_cpu_setup, NULL, 1); + + __register_cpu_notifier(&uncore_cpu_nb); + + cpu_notifier_register_done(); +} + + static int __init intel_uncore_init(void) { int ret; @@ -3763,6 +4288,7 @@ static int __init intel_uncore_init(void) uncore_pci_exit(); goto fail; } + uncore_cpumask_init(); uncore_pmus_register(); return 0; diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore.h b/arch/x86/kernel/cpu/perf_event_intel_uncore.h index a80ab71a883..90236f0c94a 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_uncore.h +++ b/arch/x86/kernel/cpu/perf_event_intel_uncore.h @@ -6,6 +6,7 @@ #define UNCORE_PMU_NAME_LEN 32 #define UNCORE_PMU_HRTIMER_INTERVAL (60LL * NSEC_PER_SEC) +#define UNCORE_SNB_IMC_HRTIMER_INTERVAL (5ULL * NSEC_PER_SEC) #define UNCORE_FIXED_EVENT 0xff #define UNCORE_PMC_IDX_MAX_GENERIC 8 @@ -440,6 +441,7 @@ struct intel_uncore_type { struct intel_uncore_ops *ops; struct uncore_event_desc *event_descs; const struct attribute_group *attr_groups[4]; + struct pmu *pmu; /* for custom pmu ops */ }; #define pmu_group attr_groups[0] @@ -488,8 +490,11 @@ struct intel_uncore_box { u64 tags[UNCORE_PMC_IDX_MAX]; struct pci_dev *pci_dev; struct intel_uncore_pmu *pmu; + u64 hrtimer_duration; /* hrtimer timeout for this box */ struct hrtimer hrtimer; struct list_head list; + struct list_head active_list; + void *io_addr; struct intel_uncore_extra_reg shared_regs[0]; }; diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c index 3486e666035..5d466b7d860 100644 --- a/arch/x86/kernel/cpu/perf_event_p4.c +++ b/arch/x86/kernel/cpu/perf_event_p4.c @@ -1257,7 +1257,24 @@ again: pass++; goto again; } - + /* + * Perf does test runs to see if a whole group can be assigned + * together succesfully. There can be multiple rounds of this. + * Unfortunately, p4_pmu_swap_config_ts touches the hwc->config + * bits, such that the next round of group assignments will + * cause the above p4_should_swap_ts to pass instead of fail. + * This leads to counters exclusive to thread0 being used by + * thread1. + * + * Solve this with a cheap hack, reset the idx back to -1 to + * force a new lookup (p4_next_cntr) to get the right counter + * for the right thread. + * + * This probably doesn't comply with the general spirit of how + * perf wants to work, but P4 is special. :-( + */ + if (p4_should_swap_ts(hwc->config, cpu)) + hwc->idx = -1; p4_pmu_swap_config_ts(hwc, cpu); if (assign) assign[i] = cntr_idx; @@ -1322,6 +1339,7 @@ static __initconst const struct x86_pmu p4_pmu = { __init int p4_pmu_init(void) { unsigned int low, high; + int i, reg; /* If we get stripped -- indexing fails */ BUILD_BUG_ON(ARCH_P4_MAX_CCCR > INTEL_PMC_MAX_GENERIC); @@ -1340,5 +1358,19 @@ __init int p4_pmu_init(void) x86_pmu = p4_pmu; + /* + * Even though the counters are configured to interrupt a particular + * logical processor when an overflow happens, testing has shown that + * on kdump kernels (which uses a single cpu), thread1's counter + * continues to run and will report an NMI on thread0. Due to the + * overflow bug, this leads to a stream of unknown NMIs. + * + * Solve this by zero'ing out the registers to mimic a reset. + */ + for (i = 0; i < x86_pmu.num_counters; i++) { + reg = x86_pmu_config_addr(i); + wrmsrl_safe(reg, 0ULL); + } + return 0; } diff --git a/arch/x86/kernel/cpu/perf_event_p6.c b/arch/x86/kernel/cpu/perf_event_p6.c index b1e2fe11532..7c1a0c07b60 100644 --- a/arch/x86/kernel/cpu/perf_event_p6.c +++ b/arch/x86/kernel/cpu/perf_event_p6.c @@ -231,31 +231,49 @@ static __initconst const struct x86_pmu p6_pmu = { }; +static __init void p6_pmu_rdpmc_quirk(void) +{ + if (boot_cpu_data.x86_mask < 9) { + /* + * PPro erratum 26; fixed in stepping 9 and above. + */ + pr_warn("Userspace RDPMC support disabled due to a CPU erratum\n"); + x86_pmu.attr_rdpmc_broken = 1; + x86_pmu.attr_rdpmc = 0; + } +} + __init int p6_pmu_init(void) { + x86_pmu = p6_pmu; + switch (boot_cpu_data.x86_model) { - case 1: - case 3: /* Pentium Pro */ - case 5: - case 6: /* Pentium II */ - case 7: - case 8: - case 11: /* Pentium III */ - case 9: - case 13: - /* Pentium M */ + case 1: /* Pentium Pro */ + x86_add_quirk(p6_pmu_rdpmc_quirk); + break; + + case 3: /* Pentium II - Klamath */ + case 5: /* Pentium II - Deschutes */ + case 6: /* Pentium II - Mendocino */ break; + + case 7: /* Pentium III - Katmai */ + case 8: /* Pentium III - Coppermine */ + case 10: /* Pentium III Xeon */ + case 11: /* Pentium III - Tualatin */ + break; + + case 9: /* Pentium M - Banias */ + case 13: /* Pentium M - Dothan */ + break; + default: - pr_cont("unsupported p6 CPU model %d ", - boot_cpu_data.x86_model); + pr_cont("unsupported p6 CPU model %d ", boot_cpu_data.x86_model); return -ENODEV; } - x86_pmu = p6_pmu; - memcpy(hw_cache_event_ids, p6_hw_cache_event_ids, sizeof(hw_cache_event_ids)); - return 0; } diff --git a/arch/x86/kernel/cpu/proc.c b/arch/x86/kernel/cpu/proc.c index aee6317b902..06fe3ed8b85 100644 --- a/arch/x86/kernel/cpu/proc.c +++ b/arch/x86/kernel/cpu/proc.c @@ -11,15 +11,12 @@ static void show_cpuinfo_core(struct seq_file *m, struct cpuinfo_x86 *c, unsigned int cpu) { #ifdef CONFIG_SMP - if (c->x86_max_cores * smp_num_siblings > 1) { - seq_printf(m, "physical id\t: %d\n", c->phys_proc_id); - seq_printf(m, "siblings\t: %d\n", - cpumask_weight(cpu_core_mask(cpu))); - seq_printf(m, "core id\t\t: %d\n", c->cpu_core_id); - seq_printf(m, "cpu cores\t: %d\n", c->booted_cores); - seq_printf(m, "apicid\t\t: %d\n", c->apicid); - seq_printf(m, "initial apicid\t: %d\n", c->initial_apicid); - } + seq_printf(m, "physical id\t: %d\n", c->phys_proc_id); + seq_printf(m, "siblings\t: %d\n", cpumask_weight(cpu_core_mask(cpu))); + seq_printf(m, "core id\t\t: %d\n", c->cpu_core_id); + seq_printf(m, "cpu cores\t: %d\n", c->booted_cores); + seq_printf(m, "apicid\t\t: %d\n", c->apicid); + seq_printf(m, "initial apicid\t: %d\n", c->initial_apicid); #endif } diff --git a/arch/x86/kernel/cpu/rdrand.c b/arch/x86/kernel/cpu/rdrand.c index 88db010845c..136ac74dee8 100644 --- a/arch/x86/kernel/cpu/rdrand.c +++ b/arch/x86/kernel/cpu/rdrand.c @@ -27,24 +27,11 @@ static int __init x86_rdrand_setup(char *s) { setup_clear_cpu_cap(X86_FEATURE_RDRAND); + setup_clear_cpu_cap(X86_FEATURE_RDSEED); return 1; } __setup("nordrand", x86_rdrand_setup); -/* We can't use arch_get_random_long() here since alternatives haven't run */ -static inline int rdrand_long(unsigned long *v) -{ - int ok; - asm volatile("1: " RDRAND_LONG "\n\t" - "jc 2f\n\t" - "decl %0\n\t" - "jnz 1b\n\t" - "2:" - : "=r" (ok), "=a" (*v) - : "0" (RDRAND_RETRY_LOOPS)); - return ok; -} - /* * Force a reseed cycle; we are architecturally guaranteed a reseed * after no more than 512 128-bit chunks of random data. This also diff --git a/arch/x86/kernel/cpu/scattered.c b/arch/x86/kernel/cpu/scattered.c index f2cc63e9cf0..b6f794aa169 100644 --- a/arch/x86/kernel/cpu/scattered.c +++ b/arch/x86/kernel/cpu/scattered.c @@ -1,5 +1,5 @@ /* - * Routines to indentify additional cpu features that are scattered in + * Routines to identify additional cpu features that are scattered in * cpuid space. */ #include <linux/cpu.h> diff --git a/arch/x86/kernel/cpu/transmeta.c b/arch/x86/kernel/cpu/transmeta.c index aa0430d69b9..3fa0e5ad86b 100644 --- a/arch/x86/kernel/cpu/transmeta.c +++ b/arch/x86/kernel/cpu/transmeta.c @@ -1,6 +1,5 @@ #include <linux/kernel.h> #include <linux/mm.h> -#include <linux/init.h> #include <asm/processor.h> #include <asm/msr.h> #include "cpu.h" diff --git a/arch/x86/kernel/cpu/umc.c b/arch/x86/kernel/cpu/umc.c index 202759a1412..ef9c2a0078b 100644 --- a/arch/x86/kernel/cpu/umc.c +++ b/arch/x86/kernel/cpu/umc.c @@ -1,5 +1,4 @@ #include <linux/kernel.h> -#include <linux/init.h> #include <asm/processor.h> #include "cpu.h" @@ -11,8 +10,8 @@ static const struct cpu_dev umc_cpu_dev = { .c_vendor = "UMC", .c_ident = { "UMC UMC UMC" }, - .c_models = { - { .vendor = X86_VENDOR_UMC, .family = 4, .model_names = + .legacy_models = { + { .family = 4, .model_names = { [1] = "U5D", [2] = "U5S", |