From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755338AbbLAA21 (ORCPT ); Mon, 30 Nov 2015 19:28:27 -0500 Received: from mga03.intel.com ([134.134.136.65]:9871 "EHLO mga03.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754458AbbLAA2X (ORCPT ); Mon, 30 Nov 2015 19:28:23 -0500 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.20,366,1444719600"; d="scan'208";a="861704897" From: Andi Kleen To: peterz@infradead.org Cc: linux-kernel@vger.kernel.org, mingo@kernel.org, Andi Kleen Subject: [PATCH 2/2] x86, perf: Use INST_RETIRED.PREC_DIST for cycles:ppp Date: Mon, 30 Nov 2015 16:28:09 -0800 Message-Id: <1448929689-13771-2-git-send-email-andi@firstfloor.org> X-Mailer: git-send-email 2.4.3 In-Reply-To: <1448929689-13771-1-git-send-email-andi@firstfloor.org> References: <1448929689-13771-1-git-send-email-andi@firstfloor.org> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org From: Andi Kleen Add a new 'three-p' precise level, that uses INST_RETIRED.PREC_DIST as base. The basic mechanism of abusing the inverse cmask to get all cycles works the same as before. PREC_DIST is available on Sandy Bridge or later. It had some problems on Sandy Bridge, so we only use it on IvyBridge and later. I tested it on Broadwell and Skylake. PREC_DIST has special support for avoiding shadow effects, which can give better results compare to UOPS_RETIRED. The drawback is that PREC_DIST can only schedule on counter 1, but that is ok for cycle sampling, as there is normally no need to do multiple cycle sampling runs in parallel. It is still possible to run perf top in parallel, as that doesn't use precise mode. Also of course the multiplexing can still allow parallel operation. :pp stays with the previous event. Example: Sample a loop with 10 sqrt with old cycles:pp 0.14 │10: sqrtps %xmm1,%xmm0 <-------------- 9.13 │ sqrtps %xmm1,%xmm0 11.58 │ sqrtps %xmm1,%xmm0 11.51 │ sqrtps %xmm1,%xmm0 6.27 │ sqrtps %xmm1,%xmm0 10.38 │ sqrtps %xmm1,%xmm0 12.20 │ sqrtps %xmm1,%xmm0 12.74 │ sqrtps %xmm1,%xmm0 5.40 │ sqrtps %xmm1,%xmm0 10.14 │ sqrtps %xmm1,%xmm0 10.51 │ ↑ jmp 10 We expect all 10 sqrt to get roughly the sample number of samples. But you can see that the instruction directly after the jmp is systematically underestimated in the result, due to sampling shadow effects. With the new PREC_DIST based sampling this problem is gone and all instructions show up roughly evenly: 9.51 │10: sqrtps %xmm1,%xmm0 11.74 │ sqrtps %xmm1,%xmm0 11.84 │ sqrtps %xmm1,%xmm0 6.05 │ sqrtps %xmm1,%xmm0 10.46 │ sqrtps %xmm1,%xmm0 12.25 │ sqrtps %xmm1,%xmm0 12.18 │ sqrtps %xmm1,%xmm0 5.26 │ sqrtps %xmm1,%xmm0 10.13 │ sqrtps %xmm1,%xmm0 10.43 │ sqrtps %xmm1,%xmm0 0.16 │ ↑ jmp 10 Even with PREC_DIST there is still sampling skid and the result is not completely even, but systematic shadow effects are significantly reduced. The improvements are mainly expected to make a difference in high IPC code. With low IPC it should be similar. v2: Change to use precise_ip == 3 for the new cycles event Split original patch into two. Use from Ivy Bridge and up. Signed-off-by: Andi Kleen --- arch/x86/kernel/cpu/perf_event.c | 3 ++ arch/x86/kernel/cpu/perf_event_intel.c | 46 ++++++++++++++++++++++++++++--- arch/x86/kernel/cpu/perf_event_intel_ds.c | 6 ++++ 3 files changed, 51 insertions(+), 4 deletions(-) diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 9dfbba5..ba41899 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -482,6 +482,9 @@ int x86_pmu_hw_config(struct perf_event *event) /* Support for IP fixup */ if (x86_pmu.lbr_nr || x86_pmu.intel_cap.pebs_format >= 2) precise++; + + if (x86_pmu.pebs_aliases) + precise++; } if (event->attr.precise_ip > precise) diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 09f4399..671c1c0 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -2473,6 +2473,44 @@ static void intel_pebs_aliases_snb(struct perf_event *event) } } +static void intel_pebs_aliases_precdist(struct perf_event *event) +{ + if ((event->hw.config & X86_RAW_EVENT_MASK) == 0x003c) { + /* + * Use an alternative encoding for CPU_CLK_UNHALTED.THREAD_P + * (0x003c) so that we can use it with PEBS. + * + * The regular CPU_CLK_UNHALTED.THREAD_P event (0x003c) isn't + * PEBS capable. However we can use INST_RETIRED.PREC_DIST + * (0x01c0), which is a PEBS capable event, to get the same + * count. + * + * The PREC_DIST event has special support to minimize sample + * shadowing effects. One drawback is that it can be + * only programmed on counter 1, but that seems like an + * acceptable trade off. + */ + u64 alt_config = X86_CONFIG(.event=0xc0, .umask=0x01, .inv=1, .cmask=16); + + alt_config |= (event->hw.config & ~X86_RAW_EVENT_MASK); + event->hw.config = alt_config; + } +} + +static void intel_pebs_aliases_ivb(struct perf_event *event) +{ + if (event->attr.precise_ip < 3) + return intel_pebs_aliases_snb(event); + return intel_pebs_aliases_precdist(event); +} + +static void intel_pebs_aliases_skl(struct perf_event *event) +{ + if (event->attr.precise_ip < 3) + return intel_pebs_aliases_core2(event); + return intel_pebs_aliases_precdist(event); +} + static unsigned long intel_pmu_free_running_flags(struct perf_event *event) { unsigned long flags = x86_pmu.free_running_flags; @@ -3434,7 +3472,7 @@ __init int intel_pmu_init(void) x86_pmu.event_constraints = intel_ivb_event_constraints; x86_pmu.pebs_constraints = intel_ivb_pebs_event_constraints; - x86_pmu.pebs_aliases = intel_pebs_aliases_snb; + x86_pmu.pebs_aliases = intel_pebs_aliases_ivb; if (boot_cpu_data.x86_model == 62) x86_pmu.extra_regs = intel_snbep_extra_regs; else @@ -3466,7 +3504,7 @@ __init int intel_pmu_init(void) x86_pmu.event_constraints = intel_hsw_event_constraints; x86_pmu.pebs_constraints = intel_hsw_pebs_event_constraints; x86_pmu.extra_regs = intel_snbep_extra_regs; - x86_pmu.pebs_aliases = intel_pebs_aliases_snb; + x86_pmu.pebs_aliases = intel_pebs_aliases_ivb; /* all extra regs are per-cpu when HT is on */ x86_pmu.flags |= PMU_FL_HAS_RSP_1; x86_pmu.flags |= PMU_FL_NO_HT_SHARING; @@ -3500,7 +3538,7 @@ __init int intel_pmu_init(void) x86_pmu.event_constraints = intel_bdw_event_constraints; x86_pmu.pebs_constraints = intel_hsw_pebs_event_constraints; x86_pmu.extra_regs = intel_snbep_extra_regs; - x86_pmu.pebs_aliases = intel_pebs_aliases_snb; + x86_pmu.pebs_aliases = intel_pebs_aliases_ivb; /* all extra regs are per-cpu when HT is on */ x86_pmu.flags |= PMU_FL_HAS_RSP_1; x86_pmu.flags |= PMU_FL_NO_HT_SHARING; @@ -3521,7 +3559,7 @@ __init int intel_pmu_init(void) x86_pmu.event_constraints = intel_skl_event_constraints; x86_pmu.pebs_constraints = intel_skl_pebs_event_constraints; x86_pmu.extra_regs = intel_skl_extra_regs; - x86_pmu.pebs_aliases = intel_pebs_aliases_core2; + x86_pmu.pebs_aliases = intel_pebs_aliases_skl; /* all extra regs are per-cpu when HT is on */ x86_pmu.flags |= PMU_FL_HAS_RSP_1; x86_pmu.flags |= PMU_FL_NO_HT_SHARING; diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index dcab005..7f11784 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -686,6 +686,8 @@ struct event_constraint intel_ivb_pebs_event_constraints[] = { INTEL_PST_CONSTRAINT(0x02cd, 0x8), /* MEM_TRANS_RETIRED.PRECISE_STORES */ /* UOPS_RETIRED.ALL, inv=1, cmask=16 (cycles:p). */ INTEL_FLAGS_EVENT_CONSTRAINT(0x108001c2, 0xf), + /* INST_RETIRED.PREC_DIST, inv=1, cmask=16 (cycles:ppp). */ + INTEL_FLAGS_EVENT_CONSTRAINT(0x108001c0, 0x2), INTEL_EXCLEVT_CONSTRAINT(0xd0, 0xf), /* MEM_UOP_RETIRED.* */ INTEL_EXCLEVT_CONSTRAINT(0xd1, 0xf), /* MEM_LOAD_UOPS_RETIRED.* */ INTEL_EXCLEVT_CONSTRAINT(0xd2, 0xf), /* MEM_LOAD_UOPS_LLC_HIT_RETIRED.* */ @@ -700,6 +702,8 @@ struct event_constraint intel_hsw_pebs_event_constraints[] = { INTEL_PLD_CONSTRAINT(0x01cd, 0xf), /* MEM_TRANS_RETIRED.* */ /* UOPS_RETIRED.ALL, inv=1, cmask=16 (cycles:p). */ INTEL_FLAGS_EVENT_CONSTRAINT(0x108001c2, 0xf), + /* INST_RETIRED.PREC_DIST, inv=1, cmask=16 (cycles:ppp). */ + INTEL_FLAGS_EVENT_CONSTRAINT(0x108001c0, 0x2), INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_NA(0x01c2, 0xf), /* UOPS_RETIRED.ALL */ INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_XLD(0x11d0, 0xf), /* MEM_UOPS_RETIRED.STLB_MISS_LOADS */ INTEL_FLAGS_UEVENT_CONSTRAINT_DATALA_XLD(0x21d0, 0xf), /* MEM_UOPS_RETIRED.LOCK_LOADS */ @@ -718,6 +722,8 @@ struct event_constraint intel_hsw_pebs_event_constraints[] = { struct event_constraint intel_skl_pebs_event_constraints[] = { INTEL_FLAGS_UEVENT_CONSTRAINT(0x1c0, 0x2), /* INST_RETIRED.PREC_DIST */ + /* INST_RETIRED.PREC_DIST, inv=1, cmask=16 (cycles:ppp). */ + INTEL_FLAGS_EVENT_CONSTRAINT(0x108001c0, 0x2), /* INST_RETIRED.TOTAL_CYCLES_PS (inv=1, cmask=16) (cycles:p). */ INTEL_FLAGS_EVENT_CONSTRAINT(0x108000c0, 0x01), INTEL_PLD_CONSTRAINT(0x1cd, 0xf), /* MEM_TRANS_RETIRED.* */ -- 2.4.3