Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756778AbcLBVW1 (ORCPT ); Fri, 2 Dec 2016 16:22:27 -0500 Received: from mga05.intel.com ([192.55.52.43]:26610 "EHLO mga05.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751577AbcLBVUm (ORCPT ); Fri, 2 Dec 2016 16:20:42 -0500 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.33,288,1477983600"; d="scan'208";a="1093847502" From: kan.liang@intel.com To: peterz@infradead.org, mingo@redhat.com, acme@kernel.org, linux-kernel@vger.kernel.org Cc: alexander.shishkin@linux.intel.com, tglx@linutronix.de, namhyung@kernel.org, jolsa@kernel.org, adrian.hunter@intel.com, wangnan0@huawei.com, mark.rutland@arm.com, andi@firstfloor.org, Kan Liang Subject: [PATCH V2 04/13] perf/core: output multiplexing overhead Date: Fri, 2 Dec 2016 16:19:12 -0500 Message-Id: <1480713561-6617-5-git-send-email-kan.liang@intel.com> X-Mailer: git-send-email 2.5.5 In-Reply-To: <1480713561-6617-1-git-send-email-kan.liang@intel.com> References: <1480713561-6617-1-git-send-email-kan.liang@intel.com> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2641 Lines: 91 From: Kan Liang Multiplexing overhead is one of the key overhead when the number of events is more than available counters. The multiplexing overhead PERF_CORE_MUX_OVERHEAD is a common overhead type. Signed-off-by: Kan Liang --- include/linux/perf_event.h | 2 ++ include/uapi/linux/perf_event.h | 1 + kernel/events/core.c | 14 ++++++++++++++ 3 files changed, 17 insertions(+) diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h index ebd356e..351d321 100644 --- a/include/linux/perf_event.h +++ b/include/linux/perf_event.h @@ -794,6 +794,8 @@ struct perf_cpu_context { struct list_head sched_cb_entry; int sched_cb_usage; + + struct perf_overhead_entry mux_overhead; }; struct perf_output_handle { diff --git a/include/uapi/linux/perf_event.h b/include/uapi/linux/perf_event.h index fe7b1fb..355086f 100644 --- a/include/uapi/linux/perf_event.h +++ b/include/uapi/linux/perf_event.h @@ -999,6 +999,7 @@ struct perf_branch_entry { */ enum perf_record_overhead_type { PERF_CORE_OVERHEAD = 0, + PERF_CORE_MUX_OVERHEAD = 0, PERF_PMU_OVERHEAD = 20, PERF_PMU_SAMPLE_OVERHEAD = 20, diff --git a/kernel/events/core.c b/kernel/events/core.c index 306bc92..025a19d 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -1830,6 +1830,9 @@ event_sched_out(struct perf_event *event, if (event->attr.exclusive || !cpuctx->active_oncpu) cpuctx->exclusive = 0; + if (log_overhead && cpuctx->mux_overhead.nr) + perf_log_overhead(event, PERF_CORE_MUX_OVERHEAD, &cpuctx->mux_overhead); + perf_pmu_enable(event->pmu); } @@ -3330,9 +3333,17 @@ static void rotate_ctx(struct perf_event_context *ctx) list_rotate_left(&ctx->flexible_groups); } +static void +perf_calculate_mux_overhead(struct perf_cpu_context *cpuctx, u64 time) +{ + cpuctx->mux_overhead.nr++; + cpuctx->mux_overhead.time += time; +} + static int perf_rotate_context(struct perf_cpu_context *cpuctx) { struct perf_event_context *ctx = NULL; + u64 start_clock, end_clock; int rotate = 0; if (cpuctx->ctx.nr_events) { @@ -3349,6 +3360,7 @@ static int perf_rotate_context(struct perf_cpu_context *cpuctx) if (!rotate) goto done; + start_clock = perf_clock(); perf_ctx_lock(cpuctx, cpuctx->task_ctx); perf_pmu_disable(cpuctx->ctx.pmu); @@ -3364,6 +3376,8 @@ static int perf_rotate_context(struct perf_cpu_context *cpuctx) perf_pmu_enable(cpuctx->ctx.pmu); perf_ctx_unlock(cpuctx, cpuctx->task_ctx); + end_clock = perf_clock(); + perf_calculate_mux_overhead(cpuctx, end_clock - start_clock); done: return rotate; -- 2.5.5