From c10a8de0d32e95b0b8c7c17b6dc09baea5a5a899 Mon Sep 17 00:00:00 2001 From: Kan Liang Date: Fri, 19 Oct 2018 10:04:18 -0700 Subject: perf/x86/intel/uncore: Add more IMC PCI IDs for KabyLake and CoffeeLake CPUs KabyLake and CoffeeLake CPUs have the same client uncore events as SkyLake. Add the PCI IDs for the KabyLake Y, U, S processor lines and CoffeeLake U, H, S processor lines. Signed-off-by: Kan Liang Signed-off-by: Peter Zijlstra (Intel) Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Stephane Eranian Cc: Thomas Gleixner Cc: Vince Weaver Link: http://lkml.kernel.org/r/20181019170419.378-1-kan.liang@linux.intel.com Signed-off-by: Ingo Molnar --- arch/x86/events/intel/uncore_snb.c | 115 ++++++++++++++++++++++++++++++++++++- 1 file changed, 114 insertions(+), 1 deletion(-) (limited to 'arch/x86/events/intel') diff --git a/arch/x86/events/intel/uncore_snb.c b/arch/x86/events/intel/uncore_snb.c index 8527c3e1038b..bfa25814fe5f 100644 --- a/arch/x86/events/intel/uncore_snb.c +++ b/arch/x86/events/intel/uncore_snb.c @@ -15,6 +15,25 @@ #define PCI_DEVICE_ID_INTEL_SKL_HQ_IMC 0x1910 #define PCI_DEVICE_ID_INTEL_SKL_SD_IMC 0x190f #define PCI_DEVICE_ID_INTEL_SKL_SQ_IMC 0x191f +#define PCI_DEVICE_ID_INTEL_KBL_Y_IMC 0x590c +#define PCI_DEVICE_ID_INTEL_KBL_U_IMC 0x5904 +#define PCI_DEVICE_ID_INTEL_KBL_UQ_IMC 0x5914 +#define PCI_DEVICE_ID_INTEL_KBL_SD_IMC 0x590f +#define PCI_DEVICE_ID_INTEL_KBL_SQ_IMC 0x591f +#define PCI_DEVICE_ID_INTEL_CFL_2U_IMC 0x3ecc +#define PCI_DEVICE_ID_INTEL_CFL_4U_IMC 0x3ed0 +#define PCI_DEVICE_ID_INTEL_CFL_4H_IMC 0x3e10 +#define PCI_DEVICE_ID_INTEL_CFL_6H_IMC 0x3ec4 +#define PCI_DEVICE_ID_INTEL_CFL_2S_D_IMC 0x3e0f +#define PCI_DEVICE_ID_INTEL_CFL_4S_D_IMC 0x3e1f +#define PCI_DEVICE_ID_INTEL_CFL_6S_D_IMC 0x3ec2 +#define PCI_DEVICE_ID_INTEL_CFL_8S_D_IMC 0x3e30 +#define PCI_DEVICE_ID_INTEL_CFL_4S_W_IMC 0x3e18 +#define PCI_DEVICE_ID_INTEL_CFL_6S_W_IMC 0x3ec6 +#define PCI_DEVICE_ID_INTEL_CFL_8S_W_IMC 0x3e31 +#define PCI_DEVICE_ID_INTEL_CFL_4S_S_IMC 0x3e33 +#define PCI_DEVICE_ID_INTEL_CFL_6S_S_IMC 0x3eca +#define PCI_DEVICE_ID_INTEL_CFL_8S_S_IMC 0x3e32 /* SNB event control */ #define SNB_UNC_CTL_EV_SEL_MASK 0x000000ff @@ -569,7 +588,82 @@ static const struct pci_device_id skl_uncore_pci_ids[] = { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_SKL_SQ_IMC), .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), }, - + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_KBL_Y_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_KBL_U_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_KBL_UQ_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_KBL_SD_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_KBL_SQ_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_CFL_2U_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_CFL_4U_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_CFL_4H_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_CFL_6H_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_CFL_2S_D_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_CFL_4S_D_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_CFL_6S_D_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_CFL_8S_D_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_CFL_4S_W_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_CFL_6S_W_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_CFL_8S_W_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_CFL_4S_S_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_CFL_6S_S_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, + { /* IMC */ + PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_CFL_8S_S_IMC), + .driver_data = UNCORE_PCI_DEV_DATA(SNB_PCI_UNCORE_IMC, 0), + }, { /* end: all zeroes */ }, }; @@ -618,6 +712,25 @@ static const struct imc_uncore_pci_dev desktop_imc_pci_ids[] = { IMC_DEV(SKL_HQ_IMC, &skl_uncore_pci_driver), /* 6th Gen Core H Quad Core */ IMC_DEV(SKL_SD_IMC, &skl_uncore_pci_driver), /* 6th Gen Core S Dual Core */ IMC_DEV(SKL_SQ_IMC, &skl_uncore_pci_driver), /* 6th Gen Core S Quad Core */ + IMC_DEV(KBL_Y_IMC, &skl_uncore_pci_driver), /* 7th Gen Core Y */ + IMC_DEV(KBL_U_IMC, &skl_uncore_pci_driver), /* 7th Gen Core U */ + IMC_DEV(KBL_UQ_IMC, &skl_uncore_pci_driver), /* 7th Gen Core U Quad Core */ + IMC_DEV(KBL_SD_IMC, &skl_uncore_pci_driver), /* 7th Gen Core S Dual Core */ + IMC_DEV(KBL_SQ_IMC, &skl_uncore_pci_driver), /* 7th Gen Core S Quad Core */ + IMC_DEV(CFL_2U_IMC, &skl_uncore_pci_driver), /* 8th Gen Core U 2 Cores */ + IMC_DEV(CFL_4U_IMC, &skl_uncore_pci_driver), /* 8th Gen Core U 4 Cores */ + IMC_DEV(CFL_4H_IMC, &skl_uncore_pci_driver), /* 8th Gen Core H 4 Cores */ + IMC_DEV(CFL_6H_IMC, &skl_uncore_pci_driver), /* 8th Gen Core H 6 Cores */ + IMC_DEV(CFL_2S_D_IMC, &skl_uncore_pci_driver), /* 8th Gen Core S 2 Cores Desktop */ + IMC_DEV(CFL_4S_D_IMC, &skl_uncore_pci_driver), /* 8th Gen Core S 4 Cores Desktop */ + IMC_DEV(CFL_6S_D_IMC, &skl_uncore_pci_driver), /* 8th Gen Core S 6 Cores Desktop */ + IMC_DEV(CFL_8S_D_IMC, &skl_uncore_pci_driver), /* 8th Gen Core S 8 Cores Desktop */ + IMC_DEV(CFL_4S_W_IMC, &skl_uncore_pci_driver), /* 8th Gen Core S 4 Cores Work Station */ + IMC_DEV(CFL_6S_W_IMC, &skl_uncore_pci_driver), /* 8th Gen Core S 6 Cores Work Station */ + IMC_DEV(CFL_8S_W_IMC, &skl_uncore_pci_driver), /* 8th Gen Core S 8 Cores Work Station */ + IMC_DEV(CFL_4S_S_IMC, &skl_uncore_pci_driver), /* 8th Gen Core S 4 Cores Server */ + IMC_DEV(CFL_6S_S_IMC, &skl_uncore_pci_driver), /* 8th Gen Core S 6 Cores Server */ + IMC_DEV(CFL_8S_S_IMC, &skl_uncore_pci_driver), /* 8th Gen Core S 8 Cores Server */ { /* end marker */ } }; -- cgit v1.2.3-55-g7522 From 4d47d6407ac7b4b442a4e717488a3bb137398b6c Mon Sep 17 00:00:00 2001 From: Kan Liang Date: Fri, 19 Oct 2018 10:04:19 -0700 Subject: perf/x86/intel/uncore: Support CoffeeLake 8th CBOX Coffee Lake has 8 core products which has 8 Cboxes. The 8th CBOX is mapped into different MSR space. Increase the num_boxes to 8 to handle the new products. It will not impact the previous platforms, SkyLake, KabyLake and earlier CoffeeLake. Because the num_boxes will be recalculated in uncore_cpu_init and doesn't exceed the x86_max_cores. Introduce a new box flag bit to indicate the 8th CBOX. Signed-off-by: Kan Liang Signed-off-by: Peter Zijlstra (Intel) Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Stephane Eranian Cc: Thomas Gleixner Cc: Vince Weaver Link: http://lkml.kernel.org/r/20181019170419.378-2-kan.liang@linux.intel.com Signed-off-by: Ingo Molnar --- arch/x86/events/intel/uncore.h | 33 +++++++++++++++++++++++++-------- arch/x86/events/intel/uncore_snb.c | 6 +++++- 2 files changed, 30 insertions(+), 9 deletions(-) (limited to 'arch/x86/events/intel') diff --git a/arch/x86/events/intel/uncore.h b/arch/x86/events/intel/uncore.h index e17ab885b1e9..cb46d602a6b8 100644 --- a/arch/x86/events/intel/uncore.h +++ b/arch/x86/events/intel/uncore.h @@ -129,8 +129,15 @@ struct intel_uncore_box { struct intel_uncore_extra_reg shared_regs[0]; }; -#define UNCORE_BOX_FLAG_INITIATED 0 -#define UNCORE_BOX_FLAG_CTL_OFFS8 1 /* event config registers are 8-byte apart */ +/* CFL uncore 8th cbox MSRs */ +#define CFL_UNC_CBO_7_PERFEVTSEL0 0xf70 +#define CFL_UNC_CBO_7_PER_CTR0 0xf76 + +#define UNCORE_BOX_FLAG_INITIATED 0 +/* event config registers are 8-byte apart */ +#define UNCORE_BOX_FLAG_CTL_OFFS8 1 +/* CFL 8th CBOX has different MSR space */ +#define UNCORE_BOX_FLAG_CFL8_CBOX_MSR_OFFS 2 struct uncore_event_desc { struct kobj_attribute attr; @@ -297,17 +304,27 @@ unsigned int uncore_freerunning_counter(struct intel_uncore_box *box, static inline unsigned uncore_msr_event_ctl(struct intel_uncore_box *box, int idx) { - return box->pmu->type->event_ctl + - (box->pmu->type->pair_ctr_ctl ? 2 * idx : idx) + - uncore_msr_box_offset(box); + if (test_bit(UNCORE_BOX_FLAG_CFL8_CBOX_MSR_OFFS, &box->flags)) { + return CFL_UNC_CBO_7_PERFEVTSEL0 + + (box->pmu->type->pair_ctr_ctl ? 2 * idx : idx); + } else { + return box->pmu->type->event_ctl + + (box->pmu->type->pair_ctr_ctl ? 2 * idx : idx) + + uncore_msr_box_offset(box); + } } static inline unsigned uncore_msr_perf_ctr(struct intel_uncore_box *box, int idx) { - return box->pmu->type->perf_ctr + - (box->pmu->type->pair_ctr_ctl ? 2 * idx : idx) + - uncore_msr_box_offset(box); + if (test_bit(UNCORE_BOX_FLAG_CFL8_CBOX_MSR_OFFS, &box->flags)) { + return CFL_UNC_CBO_7_PER_CTR0 + + (box->pmu->type->pair_ctr_ctl ? 2 * idx : idx); + } else { + return box->pmu->type->perf_ctr + + (box->pmu->type->pair_ctr_ctl ? 2 * idx : idx) + + uncore_msr_box_offset(box); + } } static inline diff --git a/arch/x86/events/intel/uncore_snb.c b/arch/x86/events/intel/uncore_snb.c index bfa25814fe5f..2593b0d7aeee 100644 --- a/arch/x86/events/intel/uncore_snb.c +++ b/arch/x86/events/intel/uncore_snb.c @@ -221,6 +221,10 @@ static void skl_uncore_msr_init_box(struct intel_uncore_box *box) wrmsrl(SKL_UNC_PERF_GLOBAL_CTL, SNB_UNC_GLOBAL_CTL_EN | SKL_UNC_GLOBAL_CTL_CORE_ALL); } + + /* The 8th CBOX has different MSR space */ + if (box->pmu->pmu_idx == 7) + __set_bit(UNCORE_BOX_FLAG_CFL8_CBOX_MSR_OFFS, &box->flags); } static void skl_uncore_msr_enable_box(struct intel_uncore_box *box) @@ -247,7 +251,7 @@ static struct intel_uncore_ops skl_uncore_msr_ops = { static struct intel_uncore_type skl_uncore_cbox = { .name = "cbox", .num_counters = 4, - .num_boxes = 5, + .num_boxes = 8, .perf_ctr_bits = 44, .fixed_ctr_bits = 48, .perf_ctr = SNB_UNC_CBO_0_PER_CTR0, -- cgit v1.2.3-55-g7522 From 2a5bf23d5b795d5df33dc284e8f5cf8b6a5b4042 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 20 Nov 2018 18:08:42 +0100 Subject: perf/x86/intel: Fix regression by default disabling perfmon v4 interrupt handling Kyle Huey reported that 'rr', a replay debugger, broke due to the following commit: af3bdb991a5c ("perf/x86/intel: Add a separate Arch Perfmon v4 PMI handler") Rework the 'disable_counter_freezing' __setup() parameter such that we can explicitly enable/disable it and switch to default disabled. To this purpose, rename the parameter to "perf_v4_pmi=" which is a much better description and allows requiring a bool argument. [ mingo: Improved the changelog some more. ] Reported-by: Kyle Huey Signed-off-by: Peter Zijlstra (Intel) Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Kan Liang Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Robert O'Callahan Cc: Stephane Eranian Cc: Thomas Gleixner Cc: Vince Weaver Cc: acme@kernel.org Link: http://lkml.kernel.org/r/20181120170842.GZ2131@hirez.programming.kicks-ass.net Signed-off-by: Ingo Molnar --- Documentation/admin-guide/kernel-parameters.txt | 3 ++- arch/x86/events/intel/core.c | 12 ++++++++---- 2 files changed, 10 insertions(+), 5 deletions(-) (limited to 'arch/x86/events/intel') diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt index 81d1d5a74728..5463d5a4d85c 100644 --- a/Documentation/admin-guide/kernel-parameters.txt +++ b/Documentation/admin-guide/kernel-parameters.txt @@ -856,7 +856,8 @@ causing system reset or hang due to sending INIT from AP to BSP. - disable_counter_freezing [HW] + perf_v4_pmi= [X86,INTEL] + Format: Disable Intel PMU counter freezing feature. The feature only exists starting from Arch Perfmon v4 (Skylake and newer). diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index 273c62e81546..af8bea9d4006 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -2306,14 +2306,18 @@ static int handle_pmi_common(struct pt_regs *regs, u64 status) return handled; } -static bool disable_counter_freezing; +static bool disable_counter_freezing = true; static int __init intel_perf_counter_freezing_setup(char *s) { - disable_counter_freezing = true; - pr_info("Intel PMU Counter freezing feature disabled\n"); + bool res; + + if (kstrtobool(s, &res)) + return -EINVAL; + + disable_counter_freezing = !res; return 1; } -__setup("disable_counter_freezing", intel_perf_counter_freezing_setup); +__setup("perf_v4_pmi=", intel_perf_counter_freezing_setup); /* * Simplified handler for Arch Perfmon v4: -- cgit v1.2.3-55-g7522 From ed6101bbf6266ee83e620b19faa7c6ad56bb41ab Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Wed, 21 Nov 2018 11:16:10 +0100 Subject: perf/x86/intel: Move branch tracing setup to the Intel-specific source file Moving branch tracing setup to Intel core object into separate intel_pmu_bts_config function, because it's Intel specific. Suggested-by: Peter Zijlstra Signed-off-by: Jiri Olsa Acked-by: Peter Zijlstra Cc: Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Linus Torvalds Cc: Stephane Eranian Cc: Thomas Gleixner Cc: Vince Weaver Link: http://lkml.kernel.org/r/20181121101612.16272-1-jolsa@kernel.org Signed-off-by: Ingo Molnar --- arch/x86/events/core.c | 20 -------------------- arch/x86/events/intel/core.c | 41 ++++++++++++++++++++++++++++++++++++++++- 2 files changed, 40 insertions(+), 21 deletions(-) (limited to 'arch/x86/events/intel') diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c index 106911b603bd..374a19712e20 100644 --- a/arch/x86/events/core.c +++ b/arch/x86/events/core.c @@ -438,26 +438,6 @@ int x86_setup_perfctr(struct perf_event *event) if (config == -1LL) return -EINVAL; - /* - * Branch tracing: - */ - if (attr->config == PERF_COUNT_HW_BRANCH_INSTRUCTIONS && - !attr->freq && hwc->sample_period == 1) { - /* BTS is not supported by this architecture. */ - if (!x86_pmu.bts_active) - return -EOPNOTSUPP; - - /* BTS is currently only allowed for user-mode. */ - if (!attr->exclude_kernel) - return -EOPNOTSUPP; - - /* disallow bts if conflicting events are present */ - if (x86_add_exclusive(x86_lbr_exclusive_lbr)) - return -EBUSY; - - event->destroy = hw_perf_lbr_event_destroy; - } - hwc->config |= config; return 0; diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index af8bea9d4006..a95079abeb03 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -3102,10 +3102,49 @@ static unsigned long intel_pmu_large_pebs_flags(struct perf_event *event) return flags; } +static int intel_pmu_bts_config(struct perf_event *event) +{ + struct perf_event_attr *attr = &event->attr; + struct hw_perf_event *hwc = &event->hw; + + if (attr->config == PERF_COUNT_HW_BRANCH_INSTRUCTIONS && + !attr->freq && hwc->sample_period == 1) { + /* BTS is not supported by this architecture. */ + if (!x86_pmu.bts_active) + return -EOPNOTSUPP; + + /* BTS is currently only allowed for user-mode. */ + if (!attr->exclude_kernel) + return -EOPNOTSUPP; + + /* disallow bts if conflicting events are present */ + if (x86_add_exclusive(x86_lbr_exclusive_lbr)) + return -EBUSY; + + event->destroy = hw_perf_lbr_event_destroy; + } + + return 0; +} + +static int core_pmu_hw_config(struct perf_event *event) +{ + int ret = x86_pmu_hw_config(event); + + if (ret) + return ret; + + return intel_pmu_bts_config(event); +} + static int intel_pmu_hw_config(struct perf_event *event) { int ret = x86_pmu_hw_config(event); + if (ret) + return ret; + + ret = intel_pmu_bts_config(event); if (ret) return ret; @@ -3600,7 +3639,7 @@ static __initconst const struct x86_pmu core_pmu = { .enable_all = core_pmu_enable_all, .enable = core_pmu_enable_event, .disable = x86_pmu_disable_event, - .hw_config = x86_pmu_hw_config, + .hw_config = core_pmu_hw_config, .schedule_events = x86_schedule_events, .eventsel = MSR_ARCH_PERFMON_EVENTSEL0, .perfctr = MSR_ARCH_PERFMON_PERFCTR0, -- cgit v1.2.3-55-g7522 From 67266c1080ad56c31af72b9c18355fde8ccc124a Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Wed, 21 Nov 2018 11:16:11 +0100 Subject: perf/x86/intel: Add generic branch tracing check to intel_pmu_has_bts() Currently we check the branch tracing only by checking for the PERF_COUNT_HW_BRANCH_INSTRUCTIONS event of PERF_TYPE_HARDWARE type. But we can define the same event with the PERF_TYPE_RAW type. Changing the intel_pmu_has_bts() code to check on event's final hw config value, so both HW types are covered. Adding unlikely to intel_pmu_has_bts() condition calls, because it was used in the original code in intel_bts_constraints. Signed-off-by: Jiri Olsa Acked-by: Peter Zijlstra Cc: Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Stephane Eranian Cc: Thomas Gleixner Cc: Vince Weaver Link: http://lkml.kernel.org/r/20181121101612.16272-2-jolsa@kernel.org Signed-off-by: Ingo Molnar --- arch/x86/events/intel/core.c | 17 +++-------------- arch/x86/events/perf_event.h | 13 +++++++++---- 2 files changed, 12 insertions(+), 18 deletions(-) (limited to 'arch/x86/events/intel') diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index a95079abeb03..a15a73788693 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -2474,16 +2474,7 @@ done: static struct event_constraint * intel_bts_constraints(struct perf_event *event) { - struct hw_perf_event *hwc = &event->hw; - unsigned int hw_event, bts_event; - - if (event->attr.freq) - return NULL; - - hw_event = hwc->config & INTEL_ARCH_EVENT_MASK; - bts_event = x86_pmu.event_map(PERF_COUNT_HW_BRANCH_INSTRUCTIONS); - - if (unlikely(hw_event == bts_event && hwc->sample_period == 1)) + if (unlikely(intel_pmu_has_bts(event))) return &bts_constraint; return NULL; @@ -3105,10 +3096,8 @@ static unsigned long intel_pmu_large_pebs_flags(struct perf_event *event) static int intel_pmu_bts_config(struct perf_event *event) { struct perf_event_attr *attr = &event->attr; - struct hw_perf_event *hwc = &event->hw; - if (attr->config == PERF_COUNT_HW_BRANCH_INSTRUCTIONS && - !attr->freq && hwc->sample_period == 1) { + if (unlikely(intel_pmu_has_bts(event))) { /* BTS is not supported by this architecture. */ if (!x86_pmu.bts_active) return -EOPNOTSUPP; @@ -3170,7 +3159,7 @@ static int intel_pmu_hw_config(struct perf_event *event) /* * BTS is set up earlier in this path, so don't account twice */ - if (!intel_pmu_has_bts(event)) { + if (!unlikely(intel_pmu_has_bts(event))) { /* disallow lbr if conflicting events are present */ if (x86_add_exclusive(x86_lbr_exclusive_lbr)) return -EBUSY; diff --git a/arch/x86/events/perf_event.h b/arch/x86/events/perf_event.h index adae087cecdd..78d7b7031bfc 100644 --- a/arch/x86/events/perf_event.h +++ b/arch/x86/events/perf_event.h @@ -859,11 +859,16 @@ static inline int amd_pmu_init(void) static inline bool intel_pmu_has_bts(struct perf_event *event) { - if (event->attr.config == PERF_COUNT_HW_BRANCH_INSTRUCTIONS && - !event->attr.freq && event->hw.sample_period == 1) - return true; + struct hw_perf_event *hwc = &event->hw; + unsigned int hw_event, bts_event; + + if (event->attr.freq) + return false; + + hw_event = hwc->config & INTEL_ARCH_EVENT_MASK; + bts_event = x86_pmu.event_map(PERF_COUNT_HW_BRANCH_INSTRUCTIONS); - return false; + return hw_event == bts_event && hwc->sample_period == 1; } int intel_pmu_save_and_restart(struct perf_event *event); -- cgit v1.2.3-55-g7522 From 472de49fdc53365c880ab81ae2b5cfdd83db0b06 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Wed, 21 Nov 2018 11:16:12 +0100 Subject: perf/x86/intel: Disallow precise_ip on BTS events Vince reported a crash in the BTS flush code when touching the callchain data, which was supposed to be initialized as an 'early' callchain, but intel_pmu_drain_bts_buffer() does not do that: BUG: unable to handle kernel NULL pointer dereference at 0000000000000000 ... Call Trace: intel_pmu_drain_bts_buffer+0x151/0x220 ? intel_get_event_constraints+0x219/0x360 ? perf_assign_events+0xe2/0x2a0 ? select_idle_sibling+0x22/0x3a0 ? __update_load_avg_se+0x1ec/0x270 ? enqueue_task_fair+0x377/0xdd0 ? cpumask_next_and+0x19/0x20 ? load_balance+0x134/0x950 ? check_preempt_curr+0x7a/0x90 ? ttwu_do_wakeup+0x19/0x140 x86_pmu_stop+0x3b/0x90 x86_pmu_del+0x57/0x160 event_sched_out.isra.106+0x81/0x170 group_sched_out.part.108+0x51/0xc0 __perf_event_disable+0x7f/0x160 event_function+0x8c/0xd0 remote_function+0x3c/0x50 flush_smp_call_function_queue+0x35/0xe0 smp_call_function_single_interrupt+0x3a/0xd0 call_function_single_interrupt+0xf/0x20 It was triggered by fuzzer but can be easily reproduced by: # perf record -e cpu/branch-instructions/pu -g -c 1 Peter suggested not to allow branch tracing for precise events: > Now arguably, this is really stupid behaviour. Who in his right mind > wants callchain output on BTS entries. And even if they do, BTS + > precise_ip is nonsensical. > > So in my mind disallowing precise_ip on BTS would be the simplest fix. Suggested-by: Peter Zijlstra Reported-by: Vince Weaver Signed-off-by: Jiri Olsa Acked-by: Peter Zijlstra Cc: Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Linus Torvalds Cc: Stephane Eranian Cc: Thomas Gleixner Fixes: 6cbc304f2f36 ("perf/x86/intel: Fix unwind errors from PEBS entries (mk-II)") Link: http://lkml.kernel.org/r/20181121101612.16272-3-jolsa@kernel.org Signed-off-by: Ingo Molnar --- arch/x86/events/intel/core.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'arch/x86/events/intel') diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index a15a73788693..ecc3e34ca955 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -3106,6 +3106,10 @@ static int intel_pmu_bts_config(struct perf_event *event) if (!attr->exclude_kernel) return -EOPNOTSUPP; + /* BTS is not allowed for precise events. */ + if (attr->precise_ip) + return -EOPNOTSUPP; + /* disallow bts if conflicting events are present */ if (x86_add_exclusive(x86_lbr_exclusive_lbr)) return -EBUSY; -- cgit v1.2.3-55-g7522