diff options
Diffstat (limited to 'arch/x86/kernel/cpu')
28 files changed, 1135 insertions, 674 deletions
diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c index d3153e281d7..ce8b8ff0e0e 100644 --- a/arch/x86/kernel/cpu/amd.c +++ b/arch/x86/kernel/cpu/amd.c @@ -218,7 +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) @@ -233,9 +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); } } @@ -509,14 +507,8 @@ static void early_init_amd(struct cpuinfo_x86 *c) #endif /* F16h erratum 793, CVE-2013-6885 */ - if (c->x86 == 0x16 && c->x86_model <= 0xf) { - u64 val; - - rdmsrl(MSR_AMD64_LS_CFG, val); - if (!(val & BIT(15))) - wrmsrl(MSR_AMD64_LS_CFG, val | BIT(15)); - } - + if (c->x86 == 0x16 && c->x86_model <= 0xf) + msr_set_bit(MSR_AMD64_LS_CFG, 15); } static const int amd_erratum_383[]; @@ -536,11 +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); @@ -623,14 +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"); } } } @@ -709,19 +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, @@ -733,10 +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); @@ -767,10 +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) diff --git a/arch/x86/kernel/cpu/centaur.c b/arch/x86/kernel/cpu/centaur.c index 8779edab684..d8fba5c15fb 100644 --- a/arch/x86/kernel/cpu/centaur.c +++ b/arch/x86/kernel/cpu/centaur.c @@ -8,236 +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 */ @@ -362,20 +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) { @@ -392,40 +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 = "??"; diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index 24b6fd10625..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 + } } /* @@ -948,6 +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) @@ -1020,7 +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); @@ -1073,6 +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, @@ -1089,10 +1133,6 @@ 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; @@ -1154,6 +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); @@ -1162,6 +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) { @@ -1170,6 +1212,7 @@ 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 */ diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c index 3db61c644e4..f9e4fdd3b87 100644 --- a/arch/x86/kernel/cpu/intel.c +++ b/arch/x86/kernel/cpu/intel.c @@ -31,11 +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); } @@ -129,16 +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 /* @@ -195,10 +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 @@ -225,16 +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"); } } @@ -267,10 +274,6 @@ static void intel_workarounds(struct cpuinfo_x86 *c) } #endif -#ifdef CONFIG_X86_NUMAQ - numaq_tsc_disable(); -#endif - intel_smp_check(c); } #else @@ -367,6 +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); @@ -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); @@ -640,21 +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; } diff --git a/arch/x86/kernel/cpu/intel_cacheinfo.c b/arch/x86/kernel/cpu/intel_cacheinfo.c index 0641113e296..9c8f7394c61 100644 --- a/arch/x86/kernel/cpu/intel_cacheinfo.c +++ b/arch/x86/kernel/cpu/intel_cacheinfo.c @@ -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.c b/arch/x86/kernel/cpu/mcheck/mce.c index 4d5419b249d..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)); @@ -2423,28 +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 fb6156fee6f..9a316b21df8 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_intel.c +++ b/arch/x86/kernel/cpu/mcheck/mce_intel.c @@ -9,6 +9,7 @@ #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> @@ -41,7 +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) @@ -137,6 +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); @@ -158,7 +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); @@ -194,7 +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; @@ -249,7 +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"); @@ -299,10 +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) @@ -343,9 +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/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/microcode/amd_early.c b/arch/x86/kernel/cpu/microcode/amd_early.c index 8384c0fa206..617a9e28424 100644 --- a/arch/x86/kernel/cpu/microcode/amd_early.c +++ b/arch/x86/kernel/cpu/microcode/amd_early.c @@ -285,6 +285,15 @@ static void __init collect_cpu_sig_on_bsp(void *arg) 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) { @@ -337,31 +346,37 @@ void load_ucode_amd_ap(void) int __init save_microcode_in_initrd_amd(void) { + unsigned long cont; enum ucode_state ret; u32 eax; -#ifdef CONFIG_X86_32 - 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); + if (!container) + return -EINVAL; +#ifdef CONFIG_X86_32 + get_bsp_sig(); + cont = (unsigned long)container; +#else /* - * 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. + * We need the physical address of the container for both bitness since + * boot_params.hdr.ramdisk_image is a physical address. */ - container = (u8 *)(__va((u32)relocated_ramdisk) + - ((u32)container - boot_params.hdr.ramdisk_image)); + 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); - if (!container) - return -EINVAL; - eax = cpuid_eax(0x00000001); eax = ((eax >> 8) & 0xf) + ((eax >> 20) & 0xff); diff --git a/arch/x86/kernel/cpu/microcode/core.c b/arch/x86/kernel/cpu/microcode/core.c index 15c987698b0..dd9d6190b08 100644 --- a/arch/x86/kernel/cpu/microcode/core.c +++ b/arch/x86/kernel/cpu/microcode/core.c @@ -97,6 +97,9 @@ MODULE_LICENSE("GPL"); static struct microcode_ops *microcode_ops; +bool dis_ucode_ldr; +module_param(dis_ucode_ldr, bool, 0); + /* * Synchronization. * @@ -546,6 +549,9 @@ 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) diff --git a/arch/x86/kernel/cpu/microcode/core_early.c b/arch/x86/kernel/cpu/microcode/core_early.c index be7f8514f57..5f28a64e71e 100644 --- a/arch/x86/kernel/cpu/microcode/core_early.c +++ b/arch/x86/kernel/cpu/microcode/core_early.c @@ -17,9 +17,11 @@ * 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') @@ -72,10 +74,33 @@ static int x86_family(void) 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; @@ -96,10 +121,22 @@ void __init load_ucode_bsp(void) } } +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; diff --git a/arch/x86/kernel/cpu/mshyperv.c b/arch/x86/kernel/cpu/mshyperv.c index 9f7ca266864..a450373e8e9 100644 --- a/arch/x86/kernel/cpu/mshyperv.c +++ b/arch/x86/kernel/cpu/mshyperv.c @@ -17,6 +17,7 @@ #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> @@ -26,10 +27,50 @@ #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; @@ -91,20 +132,16 @@ static void __init ms_hyperv_init_platform(void) lapic_timer_frequency = hv_lapic_frequency; printk(KERN_INFO "HyperV: LAPIC Timer Frequency: %#x\n", lapic_timer_frequency); - - /* - * On Hyper-V, when we are booting off an EFI firmware stack, - * we do not have many legacy devices including PIC, PIT etc. - */ - if (efi_enabled(EFI_BOOT)) { - printk(KERN_INFO "HyperV: Using null_legacy_pic\n"); - legacy_pic = &null_legacy_pic; - } } #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 = { @@ -113,41 +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 b88645191fe..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); } @@ -1273,7 +1287,7 @@ void perf_events_lapic_init(void) apic_write(APIC_LVTPC, APIC_DM_NMI); } -static int __kprobes +static int perf_event_nmi_handler(unsigned int cmd, struct pt_regs *regs) { u64 start_clock; @@ -1291,6 +1305,7 @@ perf_event_nmi_handler(unsigned int cmd, struct pt_regs *regs) 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; diff --git a/arch/x86/kernel/cpu/perf_event.h b/arch/x86/kernel/cpu/perf_event.h index c1a861829d8..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 */ @@ -293,14 +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) \ @@ -409,6 +413,7 @@ struct x86_pmu { /* * sysfs attrs */ + int attr_rdpmc_broken; int attr_rdpmc; struct attribute **format_attrs; struct attribute **event_attrs; diff --git a/arch/x86/kernel/cpu/perf_event_amd_ibs.c b/arch/x86/kernel/cpu/perf_event_amd_ibs.c index 4b8e4d3cd6e..cbb1be3ed9e 100644 --- a/arch/x86/kernel/cpu/perf_event_amd_ibs.c +++ b/arch/x86/kernel/cpu/perf_event_amd_ibs.c @@ -593,7 +593,7 @@ out: return 1; } -static int __kprobes +static int perf_ibs_nmi_handler(unsigned int cmd, struct pt_regs *regs) { int handled = 0; @@ -606,6 +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) { @@ -926,13 +927,13 @@ static __init int amd_ibs_init(void) goto out; perf_ibs_pm_init(); - get_online_cpus(); + 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 0fa4f242f05..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 }; @@ -1361,10 +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: @@ -1385,6 +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)) { @@ -2176,6 +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; @@ -2265,7 +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) { @@ -2310,10 +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); @@ -2471,6 +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)); @@ -2571,6 +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 ae96cfa5edd..696ade311de 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -108,15 +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; } @@ -295,9 +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; @@ -887,7 +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); } diff --git a/arch/x86/kernel/cpu/perf_event_intel_lbr.c b/arch/x86/kernel/cpu/perf_event_intel_lbr.c index d82d155aca8..9dd2459a4c7 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_lbr.c +++ b/arch/x86/kernel/cpu/perf_event_intel_lbr.c @@ -384,6 +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 @@ -678,6 +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] = { @@ -689,6 +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 index 5ad35ad94d0..619f7699487 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_rapl.c +++ b/arch/x86/kernel/cpu/perf_event_intel_rapl.c @@ -59,7 +59,7 @@ #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 /* DRAM */ +#define RAPL_IDX_PP1_NRG_STAT 3 /* gpu */ #define INTEL_RAPL_PP1 0x4 /* pseudo-encoding */ /* Clients have PP0, PKG */ @@ -72,6 +72,12 @@ 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 @@ -425,6 +431,24 @@ static struct attribute *rapl_events_cln_attr[] = { 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 */ @@ -511,6 +535,7 @@ 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; @@ -518,6 +543,10 @@ static int rapl_cpu_prepare(int cpu) 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; @@ -531,8 +560,7 @@ static int rapl_cpu_prepare(int cpu) * * we cache in local PMU instance */ - rdmsrl(MSR_RAPL_POWER_UNIT, pmu->hw_unit); - pmu->hw_unit = (pmu->hw_unit >> 8) & 0x1FULL; + pmu->hw_unit = (msr_rapl_power_unit_bits >> 8) & 0x1FULL; pmu->pmu = &rapl_pmu_class; /* @@ -631,11 +659,14 @@ static int __init rapl_pmu_init(void) switch (boot_cpu_data.x86_model) { case 42: /* Sandy Bridge */ case 58: /* Ivy Bridge */ - case 60: /* Haswell */ - case 69: /* Haswell-Celeron */ 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; @@ -646,19 +677,22 @@ static int __init rapl_pmu_init(void) /* unsupported */ return 0; } - get_online_cpus(); + + cpu_notifier_register_begin(); for_each_online_cpu(cpu) { - rapl_cpu_prepare(cpu); + ret = rapl_cpu_prepare(cpu); + if (ret) + goto out; rapl_cpu_init(cpu); } - perf_cpu_notifier(rapl_cpu_notifier); + __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); - put_online_cpus(); + cpu_notifier_register_done(); return -1; } @@ -672,7 +706,8 @@ static int __init rapl_pmu_init(void) hweight32(rapl_cntr_mask), ktime_to_ms(pmu->timer_interval)); - put_online_cpus(); +out: + cpu_notifier_register_done(); return 0; } diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore.c b/arch/x86/kernel/cpu/perf_event_intel_uncore.c index 29c248799ce..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), @@ -1178,10 +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), @@ -1196,7 +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), @@ -1631,6 +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 */ @@ -2781,6 +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; @@ -2793,19 +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); } @@ -2839,43 +3241,12 @@ static struct intel_uncore_box *uncore_alloc_box(struct intel_uncore_type *type, box->cpu = -1; box->phys_id = -1; - return box; -} + /* set default hrtimer timeout */ + box->hrtimer_duration = UNCORE_PMU_HRTIMER_INTERVAL; -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; + INIT_LIST_HEAD(&box->active_list); - 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_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_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 @@ -3271,16 +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) @@ -3326,6 +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); @@ -3361,7 +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); @@ -3493,6 +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; } @@ -3764,7 +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) { @@ -3808,29 +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; } @@ -3859,6 +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; @@ -3877,6 +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/rdrand.c b/arch/x86/kernel/cpu/rdrand.c index 384df5105fb..136ac74dee8 100644 --- a/arch/x86/kernel/cpu/rdrand.c +++ b/arch/x86/kernel/cpu/rdrand.c @@ -27,6 +27,7 @@ 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); |