From nobody Fri Dec 19 04:58:51 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 7884AC001DB for ; Sat, 12 Aug 2023 08:52:54 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S236945AbjHLIwp (ORCPT ); Sat, 12 Aug 2023 04:52:45 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:38350 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S236086AbjHLIwD (ORCPT ); Sat, 12 Aug 2023 04:52:03 -0400 Received: from szxga01-in.huawei.com (szxga01-in.huawei.com [45.249.212.187]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 038072696; Sat, 12 Aug 2023 01:52:05 -0700 (PDT) Received: from kwepemm600003.china.huawei.com (unknown [172.30.72.56]) by szxga01-in.huawei.com (SkyGuard) with ESMTP id 4RNDpP21LFztRQh; Sat, 12 Aug 2023 16:48:29 +0800 (CST) Received: from localhost.localdomain (10.67.174.95) by kwepemm600003.china.huawei.com (7.193.23.202) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27; Sat, 12 Aug 2023 16:52:00 +0800 From: Yang Jihong To: , , , , , , , , , , , , , CC: Subject: [RFC v1 01/16] perf kwork: Fix incorrect and missing free atom in work_push_atom() Date: Sat, 12 Aug 2023 08:49:02 +0000 Message-ID: <20230812084917.169338-2-yangjihong1@huawei.com> X-Mailer: git-send-email 2.30.GIT In-Reply-To: <20230812084917.169338-1-yangjihong1@huawei.com> References: <20230812084917.169338-1-yangjihong1@huawei.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Originating-IP: [10.67.174.95] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemm600003.china.huawei.com (7.193.23.202) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" 1. Atoms are managed in page mode and should be released using atom_free() instead of free(). 2. When the event does not match, the atom needs to free. Fixes: f98919ec4fcc ("perf kwork: Implement 'report' subcommand") Signed-off-by: Yang Jihong Reviewed-by: Ian Rogers --- tools/perf/builtin-kwork.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/tools/perf/builtin-kwork.c b/tools/perf/builtin-kwork.c index 14bf7a8429e7..73b5dc099a8a 100644 --- a/tools/perf/builtin-kwork.c +++ b/tools/perf/builtin-kwork.c @@ -406,12 +406,14 @@ static int work_push_atom(struct perf_kwork *kwork, =20 work =3D work_findnew(&class->work_root, &key, &kwork->cmp_id); if (work =3D=3D NULL) { - free(atom); + atom_free(atom); return -1; } =20 - if (!profile_event_match(kwork, work, sample)) + if (!profile_event_match(kwork, work, sample)) { + atom_free(atom); return 0; + } =20 if (dst_type < KWORK_TRACE_MAX) { dst_atom =3D list_last_entry_or_null(&work->atom_list[dst_type], --=20 2.30.GIT From nobody Fri Dec 19 04:58:51 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id D02D6C04E69 for ; Sat, 12 Aug 2023 08:52:13 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S236272AbjHLIwF (ORCPT ); Sat, 12 Aug 2023 04:52:05 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:38328 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230023AbjHLIwC (ORCPT ); Sat, 12 Aug 2023 04:52:02 -0400 Received: from szxga03-in.huawei.com (szxga03-in.huawei.com [45.249.212.189]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 7C9EA268C; Sat, 12 Aug 2023 01:52:05 -0700 (PDT) Received: from kwepemm600003.china.huawei.com (unknown [172.30.72.53]) by szxga03-in.huawei.com (SkyGuard) with ESMTP id 4RNDq90mTnz2Bd18; Sat, 12 Aug 2023 16:49:09 +0800 (CST) Received: from localhost.localdomain (10.67.174.95) by kwepemm600003.china.huawei.com (7.193.23.202) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27; Sat, 12 Aug 2023 16:52:01 +0800 From: Yang Jihong To: , , , , , , , , , , , , , CC: Subject: [RFC v1 02/16] perf kwork: Add the supported subcommands to the document Date: Sat, 12 Aug 2023 08:49:03 +0000 Message-ID: <20230812084917.169338-3-yangjihong1@huawei.com> X-Mailer: git-send-email 2.30.GIT In-Reply-To: <20230812084917.169338-1-yangjihong1@huawei.com> References: <20230812084917.169338-1-yangjihong1@huawei.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Originating-IP: [10.67.174.95] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemm600003.china.huawei.com (7.193.23.202) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Add missing report, latency and timehist subcommands to the document. Fixes: f98919ec4fcc ("perf kwork: Implement 'report' subcommand") Fixes: ad3d9f7a929a ("perf kwork: Implement perf kwork latency") Fixes: bcc8b3e88d6f ("perf kwork: Implement perf kwork timehist") Signed-off-by: Yang Jihong Reviewed-by: Ian Rogers --- tools/perf/Documentation/perf-kwork.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/perf/Documentation/perf-kwork.txt b/tools/perf/Documenta= tion/perf-kwork.txt index 3c36324712b6..482d6c52e2ed 100644 --- a/tools/perf/Documentation/perf-kwork.txt +++ b/tools/perf/Documentation/perf-kwork.txt @@ -8,7 +8,7 @@ perf-kwork - Tool to trace/measure kernel work properties (= latencies) SYNOPSIS -------- [verse] -'perf kwork' {record} +'perf kwork' {record|report|latency|timehist} =20 DESCRIPTION ----------- --=20 2.30.GIT From nobody Fri Dec 19 04:58:51 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 42244C04A94 for ; Sat, 12 Aug 2023 08:52:11 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S235787AbjHLIwD (ORCPT ); Sat, 12 Aug 2023 04:52:03 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:38316 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229497AbjHLIwC (ORCPT ); Sat, 12 Aug 2023 04:52:02 -0400 Received: from szxga03-in.huawei.com (szxga03-in.huawei.com [45.249.212.189]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 7C28A2684; Sat, 12 Aug 2023 01:52:05 -0700 (PDT) Received: from kwepemm600003.china.huawei.com (unknown [172.30.72.56]) by szxga03-in.huawei.com (SkyGuard) with ESMTP id 4RNDq95Q0vz2Bd1F; Sat, 12 Aug 2023 16:49:09 +0800 (CST) Received: from localhost.localdomain (10.67.174.95) by kwepemm600003.china.huawei.com (7.193.23.202) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27; Sat, 12 Aug 2023 16:52:02 +0800 From: Yang Jihong To: , , , , , , , , , , , , , CC: Subject: [RFC v1 03/16] perf kwork: Set ordered_events for perf_tool Date: Sat, 12 Aug 2023 08:49:04 +0000 Message-ID: <20230812084917.169338-4-yangjihong1@huawei.com> X-Mailer: git-send-email 2.30.GIT In-Reply-To: <20230812084917.169338-1-yangjihong1@huawei.com> References: <20230812084917.169338-1-yangjihong1@huawei.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Originating-IP: [10.67.174.95] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemm600003.china.huawei.com (7.193.23.202) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" perf kwork processes data based on timestamps and needs to sort events. Fixes: f98919ec4fcc ("perf kwork: Implement 'report' subcommand") Signed-off-by: Yang Jihong Reviewed-by: Ian Rogers --- tools/perf/builtin-kwork.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/tools/perf/builtin-kwork.c b/tools/perf/builtin-kwork.c index 73b5dc099a8a..de2fbb7c56c3 100644 --- a/tools/perf/builtin-kwork.c +++ b/tools/perf/builtin-kwork.c @@ -1694,9 +1694,10 @@ int cmd_kwork(int argc, const char **argv) static struct perf_kwork kwork =3D { .class_list =3D LIST_HEAD_INIT(kwork.class_list), .tool =3D { - .mmap =3D perf_event__process_mmap, - .mmap2 =3D perf_event__process_mmap2, - .sample =3D perf_kwork__process_tracepoint_sample, + .mmap =3D perf_event__process_mmap, + .mmap2 =3D perf_event__process_mmap2, + .sample =3D perf_kwork__process_tracepoint_sample, + .ordered_events =3D true, }, .atom_page_list =3D LIST_HEAD_INIT(kwork.atom_page_list), .sort_list =3D LIST_HEAD_INIT(kwork.sort_list), --=20 2.30.GIT From nobody Fri Dec 19 04:58:51 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id ED59BC001E0 for ; Sat, 12 Aug 2023 08:52:19 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229609AbjHLIwM (ORCPT ); Sat, 12 Aug 2023 04:52:12 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:38336 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S235461AbjHLIwD (ORCPT ); Sat, 12 Aug 2023 04:52:03 -0400 Received: from szxga02-in.huawei.com (szxga02-in.huawei.com [45.249.212.188]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 4D80D2D44; Sat, 12 Aug 2023 01:52:06 -0700 (PDT) Received: from kwepemm600003.china.huawei.com (unknown [172.30.72.54]) by szxga02-in.huawei.com (SkyGuard) with ESMTP id 4RNDpT48HjzNmnM; Sat, 12 Aug 2023 16:48:33 +0800 (CST) Received: from localhost.localdomain (10.67.174.95) by kwepemm600003.china.huawei.com (7.193.23.202) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27; Sat, 12 Aug 2023 16:52:02 +0800 From: Yang Jihong To: , , , , , , , , , , , , , CC: Subject: [RFC v1 04/16] perf kwork: Add `kwork` and `src_type` to work_init() for struct kwork_class Date: Sat, 12 Aug 2023 08:49:05 +0000 Message-ID: <20230812084917.169338-5-yangjihong1@huawei.com> X-Mailer: git-send-email 2.30.GIT In-Reply-To: <20230812084917.169338-1-yangjihong1@huawei.com> References: <20230812084917.169338-1-yangjihong1@huawei.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Originating-IP: [10.67.174.95] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemm600003.china.huawei.com (7.193.23.202) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" To support different types of reports, two parameters `struct perf_kwork * kwork` and `enum kwork_trace_type src_type` are added to work_init() of struct kwork_class for initialization in different scenarios. No functional change. Signed-off-by: Yang Jihong Reviewed-by: Ian Rogers --- tools/perf/builtin-kwork.c | 16 +++++++++++----- tools/perf/util/kwork.h | 6 ++++-- 2 files changed, 15 insertions(+), 7 deletions(-) diff --git a/tools/perf/builtin-kwork.c b/tools/perf/builtin-kwork.c index de2fbb7c56c3..42ea59a957ae 100644 --- a/tools/perf/builtin-kwork.c +++ b/tools/perf/builtin-kwork.c @@ -398,7 +398,7 @@ static int work_push_atom(struct perf_kwork *kwork, struct kwork_work *work, key; =20 BUG_ON(class->work_init =3D=3D NULL); - class->work_init(class, &key, evsel, sample, machine); + class->work_init(kwork, class, &key, src_type, evsel, sample, machine); =20 atom =3D atom_new(kwork, sample); if (atom =3D=3D NULL) @@ -445,7 +445,7 @@ static struct kwork_atom *work_pop_atom(struct perf_kwo= rk *kwork, struct kwork_work *work, key; =20 BUG_ON(class->work_init =3D=3D NULL); - class->work_init(class, &key, evsel, sample, machine); + class->work_init(kwork, class, &key, src_type, evsel, sample, machine); =20 work =3D work_findnew(&class->work_root, &key, &kwork->cmp_id); if (ret_work !=3D NULL) @@ -821,8 +821,10 @@ static int irq_class_init(struct kwork_class *class, return 0; } =20 -static void irq_work_init(struct kwork_class *class, +static void irq_work_init(struct perf_kwork *kwork __maybe_unused, + struct kwork_class *class, struct kwork_work *work, + enum kwork_trace_type src_type __maybe_unused, struct evsel *evsel, struct perf_sample *sample, struct machine *machine __maybe_unused) @@ -940,8 +942,10 @@ static char *evsel__softirq_name(struct evsel *evsel, = u64 num) return name; } =20 -static void softirq_work_init(struct kwork_class *class, +static void softirq_work_init(struct perf_kwork *kwork __maybe_unused, + struct kwork_class *class, struct kwork_work *work, + enum kwork_trace_type src_type __maybe_unused, struct evsel *evsel, struct perf_sample *sample, struct machine *machine __maybe_unused) @@ -1031,8 +1035,10 @@ static int workqueue_class_init(struct kwork_class *= class, return 0; } =20 -static void workqueue_work_init(struct kwork_class *class, +static void workqueue_work_init(struct perf_kwork *kwork __maybe_unused, + struct kwork_class *class, struct kwork_work *work, + enum kwork_trace_type src_type __maybe_unused, struct evsel *evsel, struct perf_sample *sample, struct machine *machine) diff --git a/tools/perf/util/kwork.h b/tools/perf/util/kwork.h index 53b7327550b8..736c7a08fb19 100644 --- a/tools/perf/util/kwork.h +++ b/tools/perf/util/kwork.h @@ -91,6 +91,7 @@ struct kwork_atom_page { DECLARE_BITMAP(bitmap, NR_ATOM_PER_PAGE); }; =20 +struct perf_kwork; struct kwork_class; struct kwork_work { /* @@ -142,8 +143,10 @@ struct kwork_class { int (*class_init)(struct kwork_class *class, struct perf_session *session); =20 - void (*work_init)(struct kwork_class *class, + void (*work_init)(struct perf_kwork *kwork, + struct kwork_class *class, struct kwork_work *work, + enum kwork_trace_type src_type, struct evsel *evsel, struct perf_sample *sample, struct machine *machine); @@ -152,7 +155,6 @@ struct kwork_class { char *buf, int len); }; =20 -struct perf_kwork; struct trace_kwork_handler { int (*raise_event)(struct perf_kwork *kwork, struct kwork_class *class, struct evsel *evsel, --=20 2.30.GIT From nobody Fri Dec 19 04:58:51 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id AD724C001DB for ; Sat, 12 Aug 2023 08:52:16 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S236416AbjHLIwI (ORCPT ); Sat, 12 Aug 2023 04:52:08 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:38330 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S235319AbjHLIwD (ORCPT ); Sat, 12 Aug 2023 04:52:03 -0400 Received: from szxga02-in.huawei.com (szxga02-in.huawei.com [45.249.212.188]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 3A270271E; Sat, 12 Aug 2023 01:52:06 -0700 (PDT) Received: from kwepemm600003.china.huawei.com (unknown [172.30.72.54]) by szxga02-in.huawei.com (SkyGuard) with ESMTP id 4RNDrD54zlzTmHx; Sat, 12 Aug 2023 16:50:04 +0800 (CST) Received: from localhost.localdomain (10.67.174.95) by kwepemm600003.china.huawei.com (7.193.23.202) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27; Sat, 12 Aug 2023 16:52:03 +0800 From: Yang Jihong To: , , , , , , , , , , , , , CC: Subject: [RFC v1 05/16] perf kwork: Overwrite original atom in the list when a new atom is pushed. Date: Sat, 12 Aug 2023 08:49:06 +0000 Message-ID: <20230812084917.169338-6-yangjihong1@huawei.com> X-Mailer: git-send-email 2.30.GIT In-Reply-To: <20230812084917.169338-1-yangjihong1@huawei.com> References: <20230812084917.169338-1-yangjihong1@huawei.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Originating-IP: [10.67.174.95] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemm600003.china.huawei.com (7.193.23.202) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" work_push_atom() supports nesting. Currently, all supported kworks are not nested. A `overwrite` parameter is added to overwrite the original atom in the list. Signed-off-by: Yang Jihong Reviewed-by: Ian Rogers --- tools/perf/builtin-kwork.c | 24 ++++++++++++++++++------ 1 file changed, 18 insertions(+), 6 deletions(-) diff --git a/tools/perf/builtin-kwork.c b/tools/perf/builtin-kwork.c index 42ea59a957ae..f620911a26a2 100644 --- a/tools/perf/builtin-kwork.c +++ b/tools/perf/builtin-kwork.c @@ -392,9 +392,10 @@ static int work_push_atom(struct perf_kwork *kwork, struct evsel *evsel, struct perf_sample *sample, struct machine *machine, - struct kwork_work **ret_work) + struct kwork_work **ret_work, + bool overwrite) { - struct kwork_atom *atom, *dst_atom; + struct kwork_atom *atom, *dst_atom, *last_atom; struct kwork_work *work, key; =20 BUG_ON(class->work_init =3D=3D NULL); @@ -427,6 +428,17 @@ static int work_push_atom(struct perf_kwork *kwork, if (ret_work !=3D NULL) *ret_work =3D work; =20 + if (overwrite) { + last_atom =3D list_last_entry_or_null(&work->atom_list[src_type], + struct kwork_atom, list); + if (last_atom) { + atom_del(last_atom); + + kwork->nr_skipped_events[src_type]++; + kwork->nr_skipped_events[KWORK_TRACE_MAX]++; + } + } + list_add_tail(&atom->list, &work->atom_list[src_type]); =20 return 0; @@ -502,7 +514,7 @@ static int report_entry_event(struct perf_kwork *kwork, { return work_push_atom(kwork, class, KWORK_TRACE_ENTRY, KWORK_TRACE_MAX, evsel, sample, - machine, NULL); + machine, NULL, true); } =20 static int report_exit_event(struct perf_kwork *kwork, @@ -557,7 +569,7 @@ static int latency_raise_event(struct perf_kwork *kwork, { return work_push_atom(kwork, class, KWORK_TRACE_RAISE, KWORK_TRACE_MAX, evsel, sample, - machine, NULL); + machine, NULL, true); } =20 static int latency_entry_event(struct perf_kwork *kwork, @@ -716,7 +728,7 @@ static int timehist_raise_event(struct perf_kwork *kwor= k, { return work_push_atom(kwork, class, KWORK_TRACE_RAISE, KWORK_TRACE_MAX, evsel, sample, - machine, NULL); + machine, NULL, true); } =20 static int timehist_entry_event(struct perf_kwork *kwork, @@ -730,7 +742,7 @@ static int timehist_entry_event(struct perf_kwork *kwor= k, =20 ret =3D work_push_atom(kwork, class, KWORK_TRACE_ENTRY, KWORK_TRACE_RAISE, evsel, sample, - machine, &work); + machine, &work, true); if (ret) return ret; =20 --=20 2.30.GIT From nobody Fri Dec 19 04:58:51 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 52A6AC04A94 for ; Sat, 12 Aug 2023 08:52:25 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S236577AbjHLIwS (ORCPT ); Sat, 12 Aug 2023 04:52:18 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:38358 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S235796AbjHLIwD (ORCPT ); Sat, 12 Aug 2023 04:52:03 -0400 Received: from szxga02-in.huawei.com (szxga02-in.huawei.com [45.249.212.188]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id BAFF42D52; Sat, 12 Aug 2023 01:52:06 -0700 (PDT) Received: from kwepemm600003.china.huawei.com (unknown [172.30.72.57]) by szxga02-in.huawei.com (SkyGuard) with ESMTP id 4RNDrF20D9zTlp3; Sat, 12 Aug 2023 16:50:05 +0800 (CST) Received: from localhost.localdomain (10.67.174.95) by kwepemm600003.china.huawei.com (7.193.23.202) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27; Sat, 12 Aug 2023 16:52:04 +0800 From: Yang Jihong To: , , , , , , , , , , , , , CC: Subject: [RFC v1 06/16] perf kwork: Set default events list if not specified in setup_event_list() Date: Sat, 12 Aug 2023 08:49:07 +0000 Message-ID: <20230812084917.169338-7-yangjihong1@huawei.com> X-Mailer: git-send-email 2.30.GIT In-Reply-To: <20230812084917.169338-1-yangjihong1@huawei.com> References: <20230812084917.169338-1-yangjihong1@huawei.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Originating-IP: [10.67.174.95] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemm600003.china.huawei.com (7.193.23.202) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Currently when no kwork event is specified, all events are configured by default. Now set to default event list string, which is more flexible and supports subsequent function extension. Also put setup_event_list() into each subcommand for different settings. Signed-off-by: Yang Jihong Reviewed-by: Ian Rogers --- tools/perf/builtin-kwork.c | 25 ++++++++++--------------- 1 file changed, 10 insertions(+), 15 deletions(-) diff --git a/tools/perf/builtin-kwork.c b/tools/perf/builtin-kwork.c index f620911a26a2..e859c34b23f3 100644 --- a/tools/perf/builtin-kwork.c +++ b/tools/perf/builtin-kwork.c @@ -1623,8 +1623,11 @@ static void setup_event_list(struct perf_kwork *kwor= k, struct kwork_class *class; char *tmp, *tok, *str; =20 + /* + * set default events list if not specified + */ if (kwork->event_list_str =3D=3D NULL) - goto null_event_list_str; + kwork->event_list_str =3D "irq, softirq, workqueue"; =20 str =3D strdup(kwork->event_list_str); for (tok =3D strtok_r(str, ", ", &tmp); @@ -1643,17 +1646,6 @@ static void setup_event_list(struct perf_kwork *kwor= k, } free(str); =20 -null_event_list_str: - /* - * config all kwork events if not specified - */ - if (list_empty(&kwork->class_list)) { - for (i =3D 0; i < KWORK_CLASS_MAX; i++) { - list_add_tail(&kwork_class_supported_list[i]->list, - &kwork->class_list); - } - } - pr_debug("Config event list:"); list_for_each_entry(class, &kwork->class_list, list) pr_debug(" %s", class->name); @@ -1835,12 +1827,12 @@ int cmd_kwork(int argc, const char **argv) if (!argc) usage_with_options(kwork_usage, kwork_options); =20 - setup_event_list(&kwork, kwork_options, kwork_usage); sort_dimension__add(&kwork, "id", &kwork.cmp_id); =20 - if (strlen(argv[0]) > 2 && strstarts("record", argv[0])) + if (strlen(argv[0]) > 2 && strstarts("record", argv[0])) { + setup_event_list(&kwork, kwork_options, kwork_usage); return perf_kwork__record(&kwork, argc, argv); - else if (strlen(argv[0]) > 2 && strstarts("report", argv[0])) { + } else if (strlen(argv[0]) > 2 && strstarts("report", argv[0])) { kwork.sort_order =3D default_report_sort_order; if (argc > 1) { argc =3D parse_options(argc, argv, report_options, report_usage, 0); @@ -1849,6 +1841,7 @@ int cmd_kwork(int argc, const char **argv) } kwork.report =3D KWORK_REPORT_RUNTIME; setup_sorting(&kwork, report_options, report_usage); + setup_event_list(&kwork, kwork_options, kwork_usage); return perf_kwork__report(&kwork); } else if (strlen(argv[0]) > 2 && strstarts("latency", argv[0])) { kwork.sort_order =3D default_latency_sort_order; @@ -1859,6 +1852,7 @@ int cmd_kwork(int argc, const char **argv) } kwork.report =3D KWORK_REPORT_LATENCY; setup_sorting(&kwork, latency_options, latency_usage); + setup_event_list(&kwork, kwork_options, kwork_usage); return perf_kwork__report(&kwork); } else if (strlen(argv[0]) > 2 && strstarts("timehist", argv[0])) { if (argc > 1) { @@ -1867,6 +1861,7 @@ int cmd_kwork(int argc, const char **argv) usage_with_options(timehist_usage, timehist_options); } kwork.report =3D KWORK_REPORT_TIMEHIST; + setup_event_list(&kwork, kwork_options, kwork_usage); return perf_kwork__timehist(&kwork); } else usage_with_options(kwork_usage, kwork_options); --=20 2.30.GIT From nobody Fri Dec 19 04:58:51 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 144C5C001DB for ; Sat, 12 Aug 2023 08:52:35 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S236837AbjHLIw3 (ORCPT ); Sat, 12 Aug 2023 04:52:29 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:38378 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S236151AbjHLIwE (ORCPT ); Sat, 12 Aug 2023 04:52:04 -0400 Received: from szxga08-in.huawei.com (szxga08-in.huawei.com [45.249.212.255]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 6167D2684; Sat, 12 Aug 2023 01:52:07 -0700 (PDT) Received: from kwepemm600003.china.huawei.com (unknown [172.30.72.56]) by szxga08-in.huawei.com (SkyGuard) with ESMTP id 4RNDs62r94z1GDXR; Sat, 12 Aug 2023 16:50:50 +0800 (CST) Received: from localhost.localdomain (10.67.174.95) by kwepemm600003.china.huawei.com (7.193.23.202) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27; Sat, 12 Aug 2023 16:52:04 +0800 From: Yang Jihong To: , , , , , , , , , , , , , CC: Subject: [RFC v1 07/16] perf kwork: Add sched record support Date: Sat, 12 Aug 2023 08:49:08 +0000 Message-ID: <20230812084917.169338-8-yangjihong1@huawei.com> X-Mailer: git-send-email 2.30.GIT In-Reply-To: <20230812084917.169338-1-yangjihong1@huawei.com> References: <20230812084917.169338-1-yangjihong1@huawei.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Originating-IP: [10.67.174.95] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemm600003.china.huawei.com (7.193.23.202) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" The kwork_class type of sched is added to support recording and parsing of sched_switch events. As follows: # perf kwork -h Usage: perf kwork [] {record|report|latency|timehist} -D, --dump-raw-trace dump raw trace in ASCII -f, --force don't complain, do it -k, --kwork list of kwork to profile (irq, softirq, workque= ue, sched, etc) -v, --verbose be more verbose (show symbol address, etc) # perf kwork -k sched record true [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.083 MB perf.data (47 samples) ] # perf evlist sched:sched_switch dummy:HG # Tip: use 'perf evlist --trace-fields' to show fields for tracepoint eve= nts Signed-off-by: Yang Jihong Reviewed-by: Ian Rogers --- tools/perf/Documentation/perf-kwork.txt | 2 +- tools/perf/builtin-kwork.c | 69 ++++++++++++++++++++++++- tools/perf/util/kwork.h | 5 ++ 3 files changed, 74 insertions(+), 2 deletions(-) diff --git a/tools/perf/Documentation/perf-kwork.txt b/tools/perf/Documenta= tion/perf-kwork.txt index 482d6c52e2ed..2092ab916ea9 100644 --- a/tools/perf/Documentation/perf-kwork.txt +++ b/tools/perf/Documentation/perf-kwork.txt @@ -66,7 +66,7 @@ OPTIONS =20 -k:: --kwork:: - List of kwork to profile (irq, softirq, workqueue, etc) + List of kwork to profile (irq, softirq, workqueue, sched, etc) =20 -v:: --verbose:: diff --git a/tools/perf/builtin-kwork.c b/tools/perf/builtin-kwork.c index e859c34b23f3..4e2b9103b9e0 100644 --- a/tools/perf/builtin-kwork.c +++ b/tools/perf/builtin-kwork.c @@ -1084,10 +1084,77 @@ static struct kwork_class kwork_workqueue =3D { .work_name =3D workqueue_work_name, }; =20 +static struct kwork_class kwork_sched; +static int process_sched_switch_event(struct perf_tool *tool, + struct evsel *evsel, + struct perf_sample *sample, + struct machine *machine) +{ + struct perf_kwork *kwork =3D container_of(tool, struct perf_kwork, tool); + + if (kwork->tp_handler->sched_switch_event) + return kwork->tp_handler->sched_switch_event(kwork, &kwork_sched, + evsel, sample, machine); + return 0; +} + +const struct evsel_str_handler sched_tp_handlers[] =3D { + { "sched:sched_switch", process_sched_switch_event, }, +}; + +static int sched_class_init(struct kwork_class *class, + struct perf_session *session) +{ + if (perf_session__set_tracepoints_handlers(session, + sched_tp_handlers)) { + pr_err("Failed to set sched tracepoints handlers\n"); + return -1; + } + + class->work_root =3D RB_ROOT_CACHED; + return 0; +} + +static void sched_work_init(struct perf_kwork *kwork __maybe_unused, + struct kwork_class *class, + struct kwork_work *work, + enum kwork_trace_type src_type, + struct evsel *evsel, + struct perf_sample *sample, + struct machine *machine __maybe_unused) +{ + work->class =3D class; + work->cpu =3D sample->cpu; + + if (src_type =3D=3D KWORK_TRACE_EXIT) { + work->id =3D evsel__intval(evsel, sample, "prev_pid"); + work->name =3D strdup(evsel__strval(evsel, sample, "prev_comm")); + } else if (src_type =3D=3D KWORK_TRACE_ENTRY) { + work->id =3D evsel__intval(evsel, sample, "next_pid"); + work->name =3D strdup(evsel__strval(evsel, sample, "next_comm")); + } +} + +static void sched_work_name(struct kwork_work *work, char *buf, int len) +{ + snprintf(buf, len, "%s", work->name); +} + +static struct kwork_class kwork_sched =3D { + .name =3D "sched", + .type =3D KWORK_CLASS_SCHED, + .nr_tracepoints =3D ARRAY_SIZE(sched_tp_handlers), + .tp_handlers =3D sched_tp_handlers, + .class_init =3D sched_class_init, + .work_init =3D sched_work_init, + .work_name =3D sched_work_name, +}; + static struct kwork_class *kwork_class_supported_list[KWORK_CLASS_MAX] =3D= { [KWORK_CLASS_IRQ] =3D &kwork_irq, [KWORK_CLASS_SOFTIRQ] =3D &kwork_softirq, [KWORK_CLASS_WORKQUEUE] =3D &kwork_workqueue, + [KWORK_CLASS_SCHED] =3D &kwork_sched, }; =20 static void print_separator(int len) @@ -1740,7 +1807,7 @@ int cmd_kwork(int argc, const char **argv) OPT_BOOLEAN('D', "dump-raw-trace", &dump_trace, "dump raw trace in ASCII"), OPT_STRING('k', "kwork", &kwork.event_list_str, "kwork", - "list of kwork to profile (irq, softirq, workqueue, etc)"), + "list of kwork to profile (irq, softirq, workqueue, sched, etc)"), OPT_BOOLEAN('f', "force", &kwork.force, "don't complain, do it"), OPT_END() }; diff --git a/tools/perf/util/kwork.h b/tools/perf/util/kwork.h index 736c7a08fb19..f8e9cdd1371a 100644 --- a/tools/perf/util/kwork.h +++ b/tools/perf/util/kwork.h @@ -16,6 +16,7 @@ enum kwork_class_type { KWORK_CLASS_IRQ, KWORK_CLASS_SOFTIRQ, KWORK_CLASS_WORKQUEUE, + KWORK_CLASS_SCHED, KWORK_CLASS_MAX, }; =20 @@ -167,6 +168,10 @@ struct trace_kwork_handler { int (*exit_event)(struct perf_kwork *kwork, struct kwork_class *class, struct evsel *evsel, struct perf_sample *sample, struct machine *machine); + + int (*sched_switch_event)(struct perf_kwork *kwork, + struct kwork_class *class, struct evsel *evsel, + struct perf_sample *sample, struct machine *machine); }; =20 struct perf_kwork { --=20 2.30.GIT From nobody Fri Dec 19 04:58:51 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 1679CC001DB for ; Sat, 12 Aug 2023 08:52:50 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S236934AbjHLIwn (ORCPT ); Sat, 12 Aug 2023 04:52:43 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:38368 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229497AbjHLIwE (ORCPT ); Sat, 12 Aug 2023 04:52:04 -0400 Received: from szxga02-in.huawei.com (szxga02-in.huawei.com [45.249.212.188]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id DB976268C; Sat, 12 Aug 2023 01:52:07 -0700 (PDT) Received: from kwepemm600003.china.huawei.com (unknown [172.30.72.53]) by szxga02-in.huawei.com (SkyGuard) with ESMTP id 4RNDpX04yWzNmSx; Sat, 12 Aug 2023 16:48:35 +0800 (CST) Received: from localhost.localdomain (10.67.174.95) by kwepemm600003.china.huawei.com (7.193.23.202) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27; Sat, 12 Aug 2023 16:52:05 +0800 From: Yang Jihong To: , , , , , , , , , , , , , CC: Subject: [RFC v1 08/16] perf kwork: Add `root` parameter to work_sort() Date: Sat, 12 Aug 2023 08:49:09 +0000 Message-ID: <20230812084917.169338-9-yangjihong1@huawei.com> X-Mailer: git-send-email 2.30.GIT In-Reply-To: <20230812084917.169338-1-yangjihong1@huawei.com> References: <20230812084917.169338-1-yangjihong1@huawei.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Originating-IP: [10.67.174.95] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemm600003.china.huawei.com (7.193.23.202) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Add a `struct rb_root_cached *root` parameter to work_sort() to sort the specified rb tree elements. No functional change. Signed-off-by: Yang Jihong Reviewed-by: Ian Rogers --- tools/perf/builtin-kwork.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/tools/perf/builtin-kwork.c b/tools/perf/builtin-kwork.c index 4e2b9103b9e0..7d93951c2ed2 100644 --- a/tools/perf/builtin-kwork.c +++ b/tools/perf/builtin-kwork.c @@ -1378,11 +1378,11 @@ static void print_bad_events(struct perf_kwork *kwo= rk) } } =20 -static void work_sort(struct perf_kwork *kwork, struct kwork_class *class) +static void work_sort(struct perf_kwork *kwork, + struct kwork_class *class, struct rb_root_cached *root) { struct rb_node *node; struct kwork_work *data; - struct rb_root_cached *root =3D &class->work_root; =20 pr_debug("Sorting %s ...\n", class->name); for (;;) { @@ -1402,7 +1402,7 @@ static void perf_kwork__sort(struct perf_kwork *kwork) struct kwork_class *class; =20 list_for_each_entry(class, &kwork->class_list, list) - work_sort(kwork, class); + work_sort(kwork, class, &class->work_root); } =20 static int perf_kwork__check_config(struct perf_kwork *kwork, --=20 2.30.GIT From nobody Fri Dec 19 04:58:51 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 2968CC001DB for ; Sat, 12 Aug 2023 08:52:44 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S236918AbjHLIwh (ORCPT ); Sat, 12 Aug 2023 04:52:37 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:52606 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S236335AbjHLIwG (ORCPT ); Sat, 12 Aug 2023 04:52:06 -0400 Received: from szxga02-in.huawei.com (szxga02-in.huawei.com [45.249.212.188]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 791BB2696; Sat, 12 Aug 2023 01:52:08 -0700 (PDT) Received: from kwepemm600003.china.huawei.com (unknown [172.30.72.54]) by szxga02-in.huawei.com (SkyGuard) with ESMTP id 4RNDpX4V2HzNmj2; Sat, 12 Aug 2023 16:48:36 +0800 (CST) Received: from localhost.localdomain (10.67.174.95) by kwepemm600003.china.huawei.com (7.193.23.202) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27; Sat, 12 Aug 2023 16:52:05 +0800 From: Yang Jihong To: , , , , , , , , , , , , , CC: Subject: [RFC v1 09/16] perf kwork: Implement perf kwork top Date: Sat, 12 Aug 2023 08:49:10 +0000 Message-ID: <20230812084917.169338-10-yangjihong1@huawei.com> X-Mailer: git-send-email 2.30.GIT In-Reply-To: <20230812084917.169338-1-yangjihong1@huawei.com> References: <20230812084917.169338-1-yangjihong1@huawei.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Originating-IP: [10.67.174.95] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemm600003.china.huawei.com (7.193.23.202) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Some common tools for collecting statistics on CPU usage, such as top, obtain statistics from timer interrupt sampling, and then periodically read statistics from /proc/stat. This method has some deviations: 1. In the tick interrupt, the time between the last tick and the current tick is counted in the current task. However, the task may be running only part of the time. 2. For each task, the top tool periodically reads the /proc/{PID}/status information. For tasks with a short life cycle, it may be missed. In conclusion, the top tool cannot accurately collect statistics on the CPU usage and running time of tasks. The statistical method based on sched_switch tracepoint can accurately calculate the CPU usage of all tasks. This method is applicable to scenarios where performance comparison data is of high precision. Example usage: # perf kwork Usage: perf kwork [] {record|report|latency|timehist|top} -D, --dump-raw-trace dump raw trace in ASCII -f, --force don't complain, do it -k, --kwork list of kwork to profile (irq, softirq, workque= ue, sched, etc) -v, --verbose be more verbose (show symbol address, etc) # perf kwork -k sched record -- perf bench sched messaging -g 1 -l 10000 # Running 'sched/messaging' benchmark: # 20 sender and receiver processes per group # 1 groups =3D=3D 40 processes run Total time: 14.074 [sec] [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 15.886 MB perf.data (129472 samples) ] # perf kwork top Total : 115708.178 ms, 8 cpus %Cpu(s): 9.78% id %Cpu0 [||||||||||||||||||||||||||| 90.55%] %Cpu1 [||||||||||||||||||||||||||| 90.51%] %Cpu2 [|||||||||||||||||||||||||| 88.57%] %Cpu3 [||||||||||||||||||||||||||| 91.18%] %Cpu4 [||||||||||||||||||||||||||| 91.09%] %Cpu5 [||||||||||||||||||||||||||| 90.88%] %Cpu6 [|||||||||||||||||||||||||| 88.64%] %Cpu7 [||||||||||||||||||||||||||| 90.28%] PID %CPU RUNTIME COMMMAND ---------------------------------------------------- 4113 22.23 3221.547 ms sched-messaging 4105 21.61 3131.495 ms sched-messaging 4119 21.53 3120.937 ms sched-messaging 4103 21.39 3101.614 ms sched-messaging 4106 21.37 3095.209 ms sched-messaging 4104 21.25 3077.269 ms sched-messaging 4115 21.21 3073.188 ms sched-messaging 4109 21.18 3069.022 ms sched-messaging 4111 20.78 3010.033 ms sched-messaging 4114 20.74 3007.073 ms sched-messaging 4108 20.73 3002.137 ms sched-messaging 4107 20.47 2967.292 ms sched-messaging 4117 20.39 2955.335 ms sched-messaging 4112 20.34 2947.080 ms sched-messaging 4118 20.32 2942.519 ms sched-messaging 4121 20.23 2929.865 ms sched-messaging 4110 20.22 2930.078 ms sched-messaging 4122 20.15 2919.542 ms sched-messaging 4120 19.77 2866.032 ms sched-messaging 4116 19.72 2857.660 ms sched-messaging 4127 16.19 2346.334 ms sched-messaging 4142 15.86 2297.600 ms sched-messaging 4141 15.62 2262.646 ms sched-messaging 4136 15.41 2231.408 ms sched-messaging 4130 15.38 2227.008 ms sched-messaging 4129 15.31 2217.692 ms sched-messaging 4126 15.21 2201.711 ms sched-messaging 4139 15.19 2200.722 ms sched-messaging 4137 15.10 2188.633 ms sched-messaging 4134 15.06 2182.082 ms sched-messaging 4132 15.02 2177.530 ms sched-messaging 4131 14.73 2131.973 ms sched-messaging 4125 14.68 2125.439 ms sched-messaging 4128 14.66 2122.255 ms sched-messaging 4123 14.65 2122.113 ms sched-messaging 4135 14.56 2107.144 ms sched-messaging 4133 14.51 2103.549 ms sched-messaging 4124 14.27 2066.671 ms sched-messaging 4140 14.17 2052.251 ms sched-messaging 4138 13.81 2000.361 ms sched-messaging 0 11.42 1652.009 ms swapper/2 0 11.35 1641.694 ms swapper/6 0 9.71 1405.108 ms swapper/7 0 9.48 1372.338 ms swapper/1 0 9.44 1366.013 ms swapper/0 0 9.11 1318.382 ms swapper/5 0 8.90 1287.582 ms swapper/4 0 8.81 1274.356 ms swapper/3 4100 2.61 379.328 ms perf 4101 1.16 169.487 ms perf-exec 151 0.65 94.741 ms systemd-resolve 249 0.36 53.030 ms sd-resolve 153 0.14 21.405 ms systemd-timesyn 1 0.10 16.200 ms systemd 16 0.09 15.785 ms rcu_preempt 4102 0.06 9.727 ms perf 4095 0.03 5.464 ms kworker/7:1 98 0.02 3.231 ms jbd2/sda-8 353 0.02 4.115 ms sshd 75 0.02 3.889 ms kworker/2:1 73 0.01 1.552 ms kworker/5:1 64 0.01 1.591 ms kworker/4:1 74 0.01 1.952 ms kworker/3:1 61 0.01 2.608 ms kcompactd0 397 0.01 1.602 ms kworker/1:1 69 0.01 1.817 ms kworker/1:1H 10 0.01 2.553 ms kworker/u16:0 2909 0.01 2.684 ms kworker/0:2 1211 0.00 0.426 ms kworker/7:0 97 0.00 0.153 ms kworker/7:1H 51 0.00 0.100 ms ksoftirqd/7 120 0.00 0.856 ms systemd-journal 76 0.00 1.414 ms kworker/6:1 46 0.00 0.246 ms ksoftirqd/6 45 0.00 0.164 ms migration/6 41 0.00 0.098 ms ksoftirqd/5 40 0.00 0.207 ms migration/5 86 0.00 1.339 ms kworker/4:1H 36 0.00 0.252 ms ksoftirqd/4 35 0.00 0.090 ms migration/4 31 0.00 0.156 ms ksoftirqd/3 30 0.00 0.073 ms migration/3 26 0.00 0.180 ms ksoftirqd/2 25 0.00 0.085 ms migration/2 21 0.00 0.106 ms ksoftirqd/1 20 0.00 0.118 ms migration/1 302 0.00 1.440 ms systemd-logind 17 0.00 0.132 ms migration/0 15 0.00 0.255 ms ksoftirqd/0 Signed-off-by: Yang Jihong Reviewed-by: Ian Rogers --- tools/perf/Documentation/perf-kwork.txt | 5 +- tools/perf/builtin-kwork.c | 387 +++++++++++++++++++++++- tools/perf/util/kwork.h | 22 ++ 3 files changed, 412 insertions(+), 2 deletions(-) diff --git a/tools/perf/Documentation/perf-kwork.txt b/tools/perf/Documenta= tion/perf-kwork.txt index 2092ab916ea9..0601fcb0feea 100644 --- a/tools/perf/Documentation/perf-kwork.txt +++ b/tools/perf/Documentation/perf-kwork.txt @@ -8,7 +8,7 @@ perf-kwork - Tool to trace/measure kernel work properties (= latencies) SYNOPSIS -------- [verse] -'perf kwork' {record|report|latency|timehist} +'perf kwork' {record|report|latency|timehist|top} =20 DESCRIPTION ----------- @@ -23,6 +23,8 @@ There are several variants of 'perf kwork': =20 'perf kwork timehist' provides an analysis of kernel work events. =20 + 'perf kwork top' to report the task cpu usage. + Example usage: perf kwork record -- sleep 1 perf kwork report @@ -30,6 +32,7 @@ There are several variants of 'perf kwork': perf kwork latency perf kwork latency -b perf kwork timehist + perf kwork top =20 By default it shows the individual work events such as irq, workqeueu, including the run time and delay (time between raise and actually entry= ): diff --git a/tools/perf/builtin-kwork.c b/tools/perf/builtin-kwork.c index 7d93951c2ed2..1b213af59135 100644 --- a/tools/perf/builtin-kwork.c +++ b/tools/perf/builtin-kwork.c @@ -45,6 +45,11 @@ #define PRINT_BRACKETPAIR_WIDTH 2 #define PRINT_TIME_UNIT_SEC_WIDTH 2 #define PRINT_TIME_UNIT_MESC_WIDTH 3 +#define PRINT_PID_WIDTH 7 +#define PRINT_TASK_NAME_WIDTH 16 +#define PRINT_CPU_USAGE_WIDTH 6 +#define PRINT_CPU_USAGE_DECIMAL_WIDTH 2 +#define PRINT_CPU_USAGE_HIST_WIDTH 30 #define PRINT_RUNTIME_HEADER_WIDTH (PRINT_RUNTIME_WIDTH + PRINT_TIME_UNIT_= MESC_WIDTH) #define PRINT_LATENCY_HEADER_WIDTH (PRINT_LATENCY_WIDTH + PRINT_TIME_UNIT_= MESC_WIDTH) #define PRINT_TIMEHIST_CPU_WIDTH (PRINT_CPU_WIDTH + PRINT_BRACKETPAIR_WIDT= H) @@ -131,6 +136,16 @@ static int max_latency_cmp(struct kwork_work *l, struc= t kwork_work *r) return 0; } =20 +static int cpu_usage_cmp(struct kwork_work *l, struct kwork_work *r) +{ + if (l->cpu_usage > r->cpu_usage) + return 1; + if (l->cpu_usage < r->cpu_usage) + return -1; + + return 0; +} + static int sort_dimension__add(struct perf_kwork *kwork __maybe_unused, const char *tok, struct list_head *list) { @@ -155,12 +170,17 @@ static int sort_dimension__add(struct perf_kwork *kwo= rk __maybe_unused, .name =3D "avg", .cmp =3D avg_latency_cmp, }; + static struct sort_dimension rate_sort_dimension =3D { + .name =3D "rate", + .cmp =3D cpu_usage_cmp, + }; struct sort_dimension *available_sorts[] =3D { &id_sort_dimension, &max_sort_dimension, &count_sort_dimension, &runtime_sort_dimension, &avg_sort_dimension, + &rate_sort_dimension, }; =20 if (kwork->report =3D=3D KWORK_REPORT_LATENCY) @@ -485,6 +505,38 @@ static struct kwork_atom *work_pop_atom(struct perf_kw= ork *kwork, return NULL; } =20 +static struct kwork_work *find_work_by_id(struct rb_root_cached *root, + u64 id, int cpu) +{ + struct rb_node *next; + struct kwork_work *work; + + next =3D rb_first_cached(root); + while (next) { + work =3D rb_entry(next, struct kwork_work, node); + if ((cpu !=3D -1 && work->id =3D=3D id && work->cpu =3D=3D cpu) || + (cpu =3D=3D -1 && work->id =3D=3D id)) + return work; + + next =3D rb_next(next); + } + + return NULL; +} + +static struct kwork_class *get_kwork_class(struct perf_kwork *kwork, + enum kwork_class_type type) +{ + struct kwork_class *class; + + list_for_each_entry(class, &kwork->class_list, list) { + if (class->type =3D=3D type) + return class; + } + + return NULL; +} + static void report_update_exit_event(struct kwork_work *work, struct kwork_atom *atom, struct perf_sample *sample) @@ -789,6 +841,54 @@ static int timehist_exit_event(struct perf_kwork *kwor= k, return ret; } =20 +static void top_update_runtime(struct kwork_work *work, + struct kwork_atom *atom, + struct perf_sample *sample) +{ + u64 delta; + u64 exit_time =3D sample->time; + u64 entry_time =3D atom->time; + + if ((entry_time !=3D 0) && (exit_time >=3D entry_time)) { + delta =3D exit_time - entry_time; + work->total_runtime +=3D delta; + } +} + +static int top_entry_event(struct perf_kwork *kwork, + struct kwork_class *class, + struct evsel *evsel, + struct perf_sample *sample, + struct machine *machine) +{ + return work_push_atom(kwork, class, KWORK_TRACE_ENTRY, + KWORK_TRACE_MAX, evsel, sample, + machine, NULL, true); +} + +static int top_sched_switch_event(struct perf_kwork *kwork, + struct kwork_class *class, + struct evsel *evsel, + struct perf_sample *sample, + struct machine *machine) +{ + struct kwork_atom *atom; + struct kwork_work *work; + + atom =3D work_pop_atom(kwork, class, KWORK_TRACE_EXIT, + KWORK_TRACE_ENTRY, evsel, sample, + machine, &work); + if (!work) + return -1; + + if (atom) { + top_update_runtime(work, atom, sample); + atom_del(atom); + } + + return top_entry_event(kwork, class, evsel, sample, machine); +} + static struct kwork_class kwork_irq; static int process_irq_handler_entry_event(struct perf_tool *tool, struct evsel *evsel, @@ -1378,6 +1478,101 @@ static void print_bad_events(struct perf_kwork *kwo= rk) } } =20 +const char *graph_load =3D "||||||||||||||||||||||||||||||||||||||||||||||= ||"; +const char *graph_idle =3D " = "; +static void top_print_per_cpu_load(struct perf_kwork *kwork) +{ + int i, load_width; + u64 total, load, load_ratio; + struct kwork_top_stat *stat =3D &kwork->top_stat; + + for (i =3D 0; i < MAX_NR_CPUS; i++) { + total =3D stat->cpus_runtime[i].total; + load =3D stat->cpus_runtime[i].load; + if (test_bit(i, stat->all_cpus_bitmap) && total) { + load_ratio =3D load * 10000 / total; + load_width =3D PRINT_CPU_USAGE_HIST_WIDTH * + load_ratio / 10000; + + printf("%%Cpu%-*d[%.*s%.*s %*.*f%%]\n", + PRINT_CPU_WIDTH, i, + load_width, graph_load, + PRINT_CPU_USAGE_HIST_WIDTH - load_width, + graph_idle, + PRINT_CPU_USAGE_WIDTH, + PRINT_CPU_USAGE_DECIMAL_WIDTH, + (double)load_ratio / 100); + } + } +} + +static void top_print_cpu_usage(struct perf_kwork *kwork) +{ + struct kwork_top_stat *stat =3D &kwork->top_stat; + u64 idle_time =3D stat->cpus_runtime[MAX_NR_CPUS].idle; + int cpus_nr =3D bitmap_weight(stat->all_cpus_bitmap, MAX_NR_CPUS); + u64 cpus_total_time =3D stat->cpus_runtime[MAX_NR_CPUS].total; + + printf("Total : %*.*f ms, %d cpus\n", + PRINT_RUNTIME_WIDTH, RPINT_DECIMAL_WIDTH, + (double)cpus_total_time / NSEC_PER_MSEC, + cpus_nr); + + printf("%%Cpu(s): %*.*f%% id\n", + PRINT_CPU_USAGE_WIDTH, PRINT_CPU_USAGE_DECIMAL_WIDTH, + cpus_total_time ? (double)idle_time * 100 / cpus_total_time : 0); + + top_print_per_cpu_load(kwork); +} + +static void top_print_header(struct perf_kwork *kwork __maybe_unused) +{ + int ret; + + printf("\n "); + ret =3D printf(" %*s %*s %*s %-*s", + PRINT_PID_WIDTH, "PID", + PRINT_CPU_USAGE_WIDTH, "%CPU", + PRINT_RUNTIME_HEADER_WIDTH + RPINT_DECIMAL_WIDTH, "RUNTIME", + PRINT_TASK_NAME_WIDTH, "COMMMAND"); + printf("\n "); + print_separator(ret); +} + +static int top_print_work(struct perf_kwork *kwork __maybe_unused, struct = kwork_work *work) +{ + int ret =3D 0; + + printf(" "); + + /* + * pid + */ + ret +=3D printf(" %*ld ", PRINT_PID_WIDTH, work->id); + + /* + * cpu usage + */ + ret +=3D printf(" %*.*f ", + PRINT_CPU_USAGE_WIDTH, PRINT_CPU_USAGE_DECIMAL_WIDTH, + (double)work->cpu_usage / 100); + + /* + * total runtime + */ + ret +=3D printf(" %*.*f ms ", + PRINT_RUNTIME_WIDTH + RPINT_DECIMAL_WIDTH, RPINT_DECIMAL_WIDTH, + (double)work->total_runtime / NSEC_PER_MSEC); + + /* + * command + */ + ret +=3D printf(" %-*s", PRINT_TASK_NAME_WIDTH, work->name); + + printf("\n"); + return ret; +} + static void work_sort(struct perf_kwork *kwork, struct kwork_class *class, struct rb_root_cached *root) { @@ -1425,6 +1620,9 @@ static int perf_kwork__check_config(struct perf_kwork= *kwork, .entry_event =3D timehist_entry_event, .exit_event =3D timehist_exit_event, }; + static struct trace_kwork_handler top_ops =3D { + .sched_switch_event =3D top_sched_switch_event, + }; =20 switch (kwork->report) { case KWORK_REPORT_RUNTIME: @@ -1436,6 +1634,9 @@ static int perf_kwork__check_config(struct perf_kwork= *kwork, case KWORK_REPORT_TIMEHIST: kwork->tp_handler =3D &timehist_ops; break; + case KWORK_REPORT_TOP: + kwork->tp_handler =3D &top_ops; + break; default: pr_debug("Invalid report type %d\n", kwork->report); return -1; @@ -1682,6 +1883,169 @@ static int perf_kwork__timehist(struct perf_kwork *= kwork) return perf_kwork__read_events(kwork); } =20 +static void top_calc_total_runtime(struct perf_kwork *kwork) +{ + struct kwork_class *class; + struct kwork_work *work; + struct rb_node *next; + struct kwork_top_stat *stat =3D &kwork->top_stat; + + class =3D get_kwork_class(kwork, KWORK_CLASS_SCHED); + if (!class) + return; + + next =3D rb_first_cached(&class->work_root); + while (next) { + work =3D rb_entry(next, struct kwork_work, node); + BUG_ON(work->cpu >=3D MAX_NR_CPUS); + stat->cpus_runtime[work->cpu].total +=3D work->total_runtime; + stat->cpus_runtime[MAX_NR_CPUS].total +=3D work->total_runtime; + next =3D rb_next(next); + } +} + +static void top_calc_idle_time(struct perf_kwork *kwork, + struct kwork_work *work) +{ + struct kwork_top_stat *stat =3D &kwork->top_stat; + + if (work->id =3D=3D 0) { + stat->cpus_runtime[work->cpu].idle +=3D work->total_runtime; + stat->cpus_runtime[MAX_NR_CPUS].idle +=3D work->total_runtime; + } +} + +static void top_calc_cpu_usage(struct perf_kwork *kwork) +{ + struct kwork_class *class; + struct kwork_work *work; + struct rb_node *next; + struct kwork_top_stat *stat =3D &kwork->top_stat; + + class =3D get_kwork_class(kwork, KWORK_CLASS_SCHED); + if (!class) + return; + + next =3D rb_first_cached(&class->work_root); + while (next) { + work =3D rb_entry(next, struct kwork_work, node); + + if (work->total_runtime =3D=3D 0) + goto next; + + __set_bit(work->cpu, stat->all_cpus_bitmap); + + work->cpu_usage =3D work->total_runtime * 10000 / + stat->cpus_runtime[work->cpu].total; + + top_calc_idle_time(kwork, work); +next: + next =3D rb_next(next); + } +} + +static void top_calc_load_runtime(struct perf_kwork *kwork, + struct kwork_work *work) +{ + struct kwork_top_stat *stat =3D &kwork->top_stat; + + if (work->id !=3D 0) { + stat->cpus_runtime[work->cpu].load +=3D work->total_runtime; + stat->cpus_runtime[MAX_NR_CPUS].load +=3D work->total_runtime; + } +} + +static void top_merge_tasks(struct perf_kwork *kwork) +{ + struct kwork_work *merged_work, *data; + struct kwork_class *class; + struct rb_node *node; + int cpu; + struct rb_root_cached merged_root =3D RB_ROOT_CACHED; + + class =3D get_kwork_class(kwork, KWORK_CLASS_SCHED); + if (!class) + return; + + for (;;) { + node =3D rb_first_cached(&class->work_root); + if (!node) + break; + + rb_erase_cached(node, &class->work_root); + data =3D rb_entry(node, struct kwork_work, node); + + cpu =3D data->cpu; + merged_work =3D find_work_by_id(&merged_root, data->id, + data->id =3D=3D 0 ? cpu : -1); + if (!merged_work) { + work_insert(&merged_root, data, &kwork->cmp_id); + } else { + merged_work->total_runtime +=3D data->total_runtime; + merged_work->cpu_usage +=3D data->cpu_usage; + } + + top_calc_load_runtime(kwork, data); + } + + work_sort(kwork, class, &merged_root); +} + +static void perf_kwork__top_report(struct perf_kwork *kwork) +{ + struct kwork_work *work; + struct rb_node *next; + + printf("\n"); + + top_print_cpu_usage(kwork); + top_print_header(kwork); + next =3D rb_first_cached(&kwork->sorted_work_root); + while (next) { + work =3D rb_entry(next, struct kwork_work, node); + process_skipped_events(kwork, work); + + if (work->total_runtime =3D=3D 0) + goto next; + + top_print_work(kwork, work); + +next: + next =3D rb_next(next); + } + + printf("\n"); +} + +static int perf_kwork__top(struct perf_kwork *kwork) +{ + struct __top_cpus_runtime *cpus_runtime; + int ret =3D 0; + + cpus_runtime =3D zalloc(sizeof(struct __top_cpus_runtime) * (MAX_NR_CPUS = + 1)); + if (!cpus_runtime) + return -1; + + kwork->top_stat.cpus_runtime =3D cpus_runtime; + bitmap_zero(kwork->top_stat.all_cpus_bitmap, MAX_NR_CPUS); + + ret =3D perf_kwork__read_events(kwork); + if (ret) + goto out; + + top_calc_total_runtime(kwork); + top_calc_cpu_usage(kwork); + top_merge_tasks(kwork); + + setup_pager(); + + perf_kwork__top_report(kwork); + +out: + free(kwork->top_stat.cpus_runtime); + return ret; +} + static void setup_event_list(struct perf_kwork *kwork, const struct option *options, const char * const usage_msg[]) @@ -1801,6 +2165,7 @@ int cmd_kwork(int argc, const char **argv) }; static const char default_report_sort_order[] =3D "runtime, max, count"; static const char default_latency_sort_order[] =3D "avg, max, count"; + static const char default_top_sort_order[] =3D "rate, runtime"; const struct option kwork_options[] =3D { OPT_INCR('v', "verbose", &verbose, "be more verbose (show symbol address, etc)"), @@ -1868,6 +2233,9 @@ int cmd_kwork(int argc, const char **argv) "input file name"), OPT_PARENT(kwork_options) }; + const struct option top_options[] =3D { + OPT_PARENT(kwork_options) + }; const char *kwork_usage[] =3D { NULL, NULL @@ -1884,8 +2252,12 @@ int cmd_kwork(int argc, const char **argv) "perf kwork timehist []", NULL }; + const char * const top_usage[] =3D { + "perf kwork top []", + NULL + }; const char *const kwork_subcommands[] =3D { - "record", "report", "latency", "timehist", NULL + "record", "report", "latency", "timehist", "top", NULL }; =20 argc =3D parse_options_subcommand(argc, argv, kwork_options, @@ -1930,6 +2302,19 @@ int cmd_kwork(int argc, const char **argv) kwork.report =3D KWORK_REPORT_TIMEHIST; setup_event_list(&kwork, kwork_options, kwork_usage); return perf_kwork__timehist(&kwork); + } else if (strlen(argv[0]) > 2 && strstarts("top", argv[0])) { + kwork.sort_order =3D default_top_sort_order; + if (argc > 1) { + argc =3D parse_options(argc, argv, top_options, top_usage, 0); + if (argc) + usage_with_options(top_usage, top_options); + } + kwork.report =3D KWORK_REPORT_TOP; + if (!kwork.event_list_str) + kwork.event_list_str =3D "sched"; + setup_event_list(&kwork, kwork_options, kwork_usage); + setup_sorting(&kwork, top_options, top_usage); + return perf_kwork__top(&kwork); } else usage_with_options(kwork_usage, kwork_options); =20 diff --git a/tools/perf/util/kwork.h b/tools/perf/util/kwork.h index f8e9cdd1371a..41ed193d5d8b 100644 --- a/tools/perf/util/kwork.h +++ b/tools/perf/util/kwork.h @@ -24,6 +24,7 @@ enum kwork_report_type { KWORK_REPORT_RUNTIME, KWORK_REPORT_LATENCY, KWORK_REPORT_TIMEHIST, + KWORK_REPORT_TOP, }; =20 enum kwork_trace_type { @@ -129,6 +130,11 @@ struct kwork_work { u64 max_latency_start; u64 max_latency_end; u64 total_latency; + + /* + * top report + */ + u32 cpu_usage; }; =20 struct kwork_class { @@ -174,6 +180,17 @@ struct trace_kwork_handler { struct perf_sample *sample, struct machine *machine); }; =20 +struct __top_cpus_runtime { + u64 load; + u64 idle; + u64 total; +}; + +struct kwork_top_stat { + DECLARE_BITMAP(all_cpus_bitmap, MAX_NR_CPUS); + struct __top_cpus_runtime *cpus_runtime; +}; + struct perf_kwork { /* * metadata @@ -225,6 +242,11 @@ struct perf_kwork { u64 all_runtime; u64 all_count; u64 nr_skipped_events[KWORK_TRACE_MAX + 1]; + + /* + * perf kwork top data + */ + struct kwork_top_stat top_stat; }; =20 struct kwork_work *perf_kwork_add_work(struct perf_kwork *kwork, --=20 2.30.GIT From nobody Fri Dec 19 04:58:51 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id B406FC001E0 for ; Sat, 12 Aug 2023 08:52:31 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S236811AbjHLIwZ (ORCPT ); Sat, 12 Aug 2023 04:52:25 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:52600 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S236285AbjHLIwG (ORCPT ); Sat, 12 Aug 2023 04:52:06 -0400 Received: from szxga01-in.huawei.com (szxga01-in.huawei.com [45.249.212.187]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 3C6C12684; Sat, 12 Aug 2023 01:52:09 -0700 (PDT) Received: from kwepemm600003.china.huawei.com (unknown [172.30.72.54]) by szxga01-in.huawei.com (SkyGuard) with ESMTP id 4RNDs73NHFzrS89; Sat, 12 Aug 2023 16:50:51 +0800 (CST) Received: from localhost.localdomain (10.67.174.95) by kwepemm600003.china.huawei.com (7.193.23.202) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27; Sat, 12 Aug 2023 16:52:06 +0800 From: Yang Jihong To: , , , , , , , , , , , , , CC: Subject: [RFC v1 10/16] perf evsel: Add evsel__intval_common() helper Date: Sat, 12 Aug 2023 08:49:11 +0000 Message-ID: <20230812084917.169338-11-yangjihong1@huawei.com> X-Mailer: git-send-email 2.30.GIT In-Reply-To: <20230812084917.169338-1-yangjihong1@huawei.com> References: <20230812084917.169338-1-yangjihong1@huawei.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Originating-IP: [10.67.174.95] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemm600003.china.huawei.com (7.193.23.202) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Add evsel__intval_common() helper to search for common_field in tracepoint format. Signed-off-by: Yang Jihong Reviewed-by: Ian Rogers --- tools/perf/util/evsel.c | 13 +++++++++++++ tools/perf/util/evsel.h | 2 ++ 2 files changed, 15 insertions(+) diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c index e41bc4d9925f..f3ad2fa862bf 100644 --- a/tools/perf/util/evsel.c +++ b/tools/perf/util/evsel.c @@ -2764,6 +2764,11 @@ struct tep_format_field *evsel__field(struct evsel *= evsel, const char *name) return tep_find_field(evsel->tp_format, name); } =20 +struct tep_format_field *evsel__common_field(struct evsel *evsel, const ch= ar *name) +{ + return tep_find_common_field(evsel->tp_format, name); +} + void *evsel__rawptr(struct evsel *evsel, struct perf_sample *sample, const= char *name) { struct tep_format_field *field =3D evsel__field(evsel, name); @@ -2832,6 +2837,14 @@ u64 evsel__intval(struct evsel *evsel, struct perf_s= ample *sample, const char *n =20 return field ? format_field__intval(field, sample, evsel->needs_swap) : 0; } + +u64 evsel__intval_common(struct evsel *evsel, struct perf_sample *sample, = const char *name) +{ + struct tep_format_field *field =3D evsel__common_field(evsel, name); + + return field ? format_field__intval(field, sample, evsel->needs_swap) : 0; +} + #endif =20 bool evsel__fallback(struct evsel *evsel, int err, char *msg, size_t msgsi= ze) diff --git a/tools/perf/util/evsel.h b/tools/perf/util/evsel.h index 848534ec74fa..815be2491938 100644 --- a/tools/perf/util/evsel.h +++ b/tools/perf/util/evsel.h @@ -340,6 +340,7 @@ struct perf_sample; #ifdef HAVE_LIBTRACEEVENT void *evsel__rawptr(struct evsel *evsel, struct perf_sample *sample, const= char *name); u64 evsel__intval(struct evsel *evsel, struct perf_sample *sample, const c= har *name); +u64 evsel__intval_common(struct evsel *evsel, struct perf_sample *sample, = const char *name); =20 static inline char *evsel__strval(struct evsel *evsel, struct perf_sample = *sample, const char *name) { @@ -352,6 +353,7 @@ struct tep_format_field; u64 format_field__intval(struct tep_format_field *field, struct perf_sampl= e *sample, bool needs_swap); =20 struct tep_format_field *evsel__field(struct evsel *evsel, const char *nam= e); +struct tep_format_field *evsel__common_field(struct evsel *evsel, const ch= ar *name); =20 static inline bool __evsel__match(const struct evsel *evsel, u32 type, u64= config) { --=20 2.30.GIT From nobody Fri Dec 19 04:58:51 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 58377C001E0 for ; Sat, 12 Aug 2023 08:52:47 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S236928AbjHLIwk (ORCPT ); Sat, 12 Aug 2023 04:52:40 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:52622 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230023AbjHLIwH (ORCPT ); Sat, 12 Aug 2023 04:52:07 -0400 Received: from szxga03-in.huawei.com (szxga03-in.huawei.com [45.249.212.189]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 9A870268C; Sat, 12 Aug 2023 01:52:09 -0700 (PDT) Received: from kwepemm600003.china.huawei.com (unknown [172.30.72.57]) by szxga03-in.huawei.com (SkyGuard) with ESMTP id 4RNDqG4PBmz2Bd1n; Sat, 12 Aug 2023 16:49:14 +0800 (CST) Received: from localhost.localdomain (10.67.174.95) by kwepemm600003.china.huawei.com (7.193.23.202) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27; Sat, 12 Aug 2023 16:52:07 +0800 From: Yang Jihong To: , , , , , , , , , , , , , CC: Subject: [RFC v1 11/16] perf kwork top: Add statistics on hardirq event support Date: Sat, 12 Aug 2023 08:49:12 +0000 Message-ID: <20230812084917.169338-12-yangjihong1@huawei.com> X-Mailer: git-send-email 2.30.GIT In-Reply-To: <20230812084917.169338-1-yangjihong1@huawei.com> References: <20230812084917.169338-1-yangjihong1@huawei.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Originating-IP: [10.67.174.95] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemm600003.china.huawei.com (7.193.23.202) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Calculate the runtime of the hardirq events and subtract it from the corresponding task runtime to improve the precision. Example usage: # perf kwork -k sched,irq record -- perf record -o perf_record.data -a sl= eep 10 [ perf record: Woken up 2 times to write data ] [ perf record: Captured and wrote 1.054 MB perf_record.data (18019 sample= s) ] [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 1.798 MB perf.data (16334 samples) ] # # perf kwork top Total : 139240.869 ms, 8 cpus %Cpu(s): 94.91% id, 0.05% hi %Cpu0 [ 0.05%] %Cpu1 [| 5.00%] %Cpu2 [ 0.43%] %Cpu3 [ 0.57%] %Cpu4 [ 1.19%] %Cpu5 [|||||| 20.46%] %Cpu6 [ 0.48%] %Cpu7 [||| 12.10%] PID %CPU RUNTIME COMMMAND ---------------------------------------------------- 0 99.54 17325.622 ms swapper/2 0 99.54 17327.527 ms swapper/0 0 99.51 17319.909 ms swapper/6 0 99.42 17304.934 ms swapper/3 0 98.80 17197.385 ms swapper/4 0 94.99 16534.991 ms swapper/1 0 87.89 15295.264 ms swapper/7 0 79.53 13843.182 ms swapper/5 4252 36.50 6361.768 ms perf 4256 1.17 205.215 ms bash 151 0.53 93.298 ms systemd-resolve 4254 0.39 69.468 ms perf 423 0.34 59.368 ms bash 412 0.29 51.204 ms sshd 249 0.20 35.288 ms sd-resolve 16 0.17 30.287 ms rcu_preempt 153 0.09 17.266 ms systemd-timesyn 1 0.09 17.078 ms systemd 4253 0.07 12.457 ms perf 4255 0.06 11.559 ms perf 4234 0.03 6.105 ms kworker/u16:1 69 0.03 6.259 ms kworker/1:1H 4251 0.02 4.615 ms perf 4095 0.02 4.890 ms kworker/7:1 61 0.02 4.005 ms kcompactd0 75 0.02 3.546 ms kworker/2:1 97 0.01 3.106 ms kworker/7:1H 98 0.01 1.995 ms jbd2/sda-8 4088 0.01 1.779 ms kworker/u16:3 2909 0.01 1.795 ms kworker/0:2 4246 0.00 1.117 ms kworker/7:2 51 0.00 0.327 ms ksoftirqd/7 50 0.00 0.369 ms migration/7 102 0.00 0.160 ms kworker/6:1H 76 0.00 0.609 ms kworker/6:1 45 0.00 0.779 ms migration/6 87 0.00 0.504 ms kworker/5:1H 73 0.00 1.130 ms kworker/5:1 41 0.00 0.152 ms ksoftirqd/5 40 0.00 0.702 ms migration/5 64 0.00 0.316 ms kworker/4:1 35 0.00 0.791 ms migration/4 353 0.00 2.211 ms sshd 74 0.00 0.272 ms kworker/3:1 30 0.00 0.819 ms migration/3 25 0.00 0.784 ms migration/2 397 0.00 0.539 ms kworker/1:1 21 0.00 1.600 ms ksoftirqd/1 20 0.00 0.773 ms migration/1 17 0.00 1.682 ms migration/0 15 0.00 0.076 ms ksoftirqd/0 Signed-off-by: Yang Jihong Reviewed-by: Ian Rogers --- tools/perf/builtin-kwork.c | 93 +++++++++++++++++++++++++++++++++++--- tools/perf/util/kwork.h | 1 + 2 files changed, 88 insertions(+), 6 deletions(-) diff --git a/tools/perf/builtin-kwork.c b/tools/perf/builtin-kwork.c index 1b213af59135..66ad8c2e8020 100644 --- a/tools/perf/builtin-kwork.c +++ b/tools/perf/builtin-kwork.c @@ -866,6 +866,36 @@ static int top_entry_event(struct perf_kwork *kwork, machine, NULL, true); } =20 +static int top_exit_event(struct perf_kwork *kwork, + struct kwork_class *class, + struct evsel *evsel, + struct perf_sample *sample, + struct machine *machine) +{ + struct kwork_work *work, *sched_work; + struct kwork_class *sched_class; + struct kwork_atom *atom; + + atom =3D work_pop_atom(kwork, class, KWORK_TRACE_EXIT, + KWORK_TRACE_ENTRY, evsel, sample, + machine, &work); + if (!work) + return -1; + + if (atom) { + sched_class =3D get_kwork_class(kwork, KWORK_CLASS_SCHED); + if (sched_class) { + sched_work =3D find_work_by_id(&sched_class->work_root, + work->id, work->cpu); + if (sched_work) + top_update_runtime(work, atom, sample); + } + atom_del(atom); + } + + return 0; +} + static int top_sched_switch_event(struct perf_kwork *kwork, struct kwork_class *class, struct evsel *evsel, @@ -933,7 +963,7 @@ static int irq_class_init(struct kwork_class *class, return 0; } =20 -static void irq_work_init(struct perf_kwork *kwork __maybe_unused, +static void irq_work_init(struct perf_kwork *kwork, struct kwork_class *class, struct kwork_work *work, enum kwork_trace_type src_type __maybe_unused, @@ -943,8 +973,14 @@ static void irq_work_init(struct perf_kwork *kwork __m= aybe_unused, { work->class =3D class; work->cpu =3D sample->cpu; - work->id =3D evsel__intval(evsel, sample, "irq"); - work->name =3D evsel__strval(evsel, sample, "name"); + + if (kwork->report =3D=3D KWORK_REPORT_TOP) { + work->id =3D evsel__intval_common(evsel, sample, "common_pid"); + work->name =3D NULL; + } else { + work->id =3D evsel__intval(evsel, sample, "irq"); + work->name =3D evsel__strval(evsel, sample, "name"); + } } =20 static void irq_work_name(struct kwork_work *work, char *buf, int len) @@ -1510,6 +1546,7 @@ static void top_print_cpu_usage(struct perf_kwork *kw= ork) { struct kwork_top_stat *stat =3D &kwork->top_stat; u64 idle_time =3D stat->cpus_runtime[MAX_NR_CPUS].idle; + u64 hardirq_time =3D stat->cpus_runtime[MAX_NR_CPUS].irq; int cpus_nr =3D bitmap_weight(stat->all_cpus_bitmap, MAX_NR_CPUS); u64 cpus_total_time =3D stat->cpus_runtime[MAX_NR_CPUS].total; =20 @@ -1518,9 +1555,12 @@ static void top_print_cpu_usage(struct perf_kwork *k= work) (double)cpus_total_time / NSEC_PER_MSEC, cpus_nr); =20 - printf("%%Cpu(s): %*.*f%% id\n", + printf("%%Cpu(s): %*.*f%% id, %*.*f%% hi\n", + PRINT_CPU_USAGE_WIDTH, PRINT_CPU_USAGE_DECIMAL_WIDTH, + cpus_total_time ? (double)idle_time * 100 / cpus_total_time : 0, + PRINT_CPU_USAGE_WIDTH, PRINT_CPU_USAGE_DECIMAL_WIDTH, - cpus_total_time ? (double)idle_time * 100 / cpus_total_time : 0); + cpus_total_time ? (double)hardirq_time * 100 / cpus_total_time : 0= ); =20 top_print_per_cpu_load(kwork); } @@ -1621,6 +1661,8 @@ static int perf_kwork__check_config(struct perf_kwork= *kwork, .exit_event =3D timehist_exit_event, }; static struct trace_kwork_handler top_ops =3D { + .entry_event =3D timehist_entry_event, + .exit_event =3D top_exit_event, .sched_switch_event =3D top_sched_switch_event, }; =20 @@ -1915,6 +1957,43 @@ static void top_calc_idle_time(struct perf_kwork *kw= ork, } } =20 +static void top_calc_irq_runtime(struct perf_kwork *kwork, + enum kwork_class_type type, + struct kwork_work *work) +{ + struct kwork_top_stat *stat =3D &kwork->top_stat; + + if (type =3D=3D KWORK_CLASS_IRQ) { + stat->cpus_runtime[work->cpu].irq +=3D work->total_runtime; + stat->cpus_runtime[MAX_NR_CPUS].irq +=3D work->total_runtime; + } +} + +static void top_subtract_irq_runtime(struct perf_kwork *kwork, + struct kwork_work *work) +{ + struct kwork_class *class; + struct kwork_work *data; + unsigned int i; + int irq_class_list[] =3D {KWORK_CLASS_IRQ}; + + for (i =3D 0; i < ARRAY_SIZE(irq_class_list); i++) { + class =3D get_kwork_class(kwork, irq_class_list[i]); + if (!class) + continue; + + data =3D find_work_by_id(&class->work_root, + work->id, work->cpu); + if (!data) + continue; + + if (work->total_runtime > data->total_runtime) { + work->total_runtime -=3D data->total_runtime; + top_calc_irq_runtime(kwork, irq_class_list[i], data); + } + } +} + static void top_calc_cpu_usage(struct perf_kwork *kwork) { struct kwork_class *class; @@ -1935,6 +2014,8 @@ static void top_calc_cpu_usage(struct perf_kwork *kwo= rk) =20 __set_bit(work->cpu, stat->all_cpus_bitmap); =20 + top_subtract_irq_runtime(kwork, work); + work->cpu_usage =3D work->total_runtime * 10000 / stat->cpus_runtime[work->cpu].total; =20 @@ -2311,7 +2392,7 @@ int cmd_kwork(int argc, const char **argv) } kwork.report =3D KWORK_REPORT_TOP; if (!kwork.event_list_str) - kwork.event_list_str =3D "sched"; + kwork.event_list_str =3D "sched, irq"; setup_event_list(&kwork, kwork_options, kwork_usage); setup_sorting(&kwork, top_options, top_usage); return perf_kwork__top(&kwork); diff --git a/tools/perf/util/kwork.h b/tools/perf/util/kwork.h index 41ed193d5d8b..b3e410d02043 100644 --- a/tools/perf/util/kwork.h +++ b/tools/perf/util/kwork.h @@ -183,6 +183,7 @@ struct trace_kwork_handler { struct __top_cpus_runtime { u64 load; u64 idle; + u64 irq; u64 total; }; =20 --=20 2.30.GIT From nobody Fri Dec 19 04:58:51 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id A8288C04A94 for ; Sat, 12 Aug 2023 08:52:39 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S236905AbjHLIwd (ORCPT ); Sat, 12 Aug 2023 04:52:33 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:52634 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S236370AbjHLIwH (ORCPT ); Sat, 12 Aug 2023 04:52:07 -0400 Received: from szxga01-in.huawei.com (szxga01-in.huawei.com [45.249.212.187]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 6CF4B2694; Sat, 12 Aug 2023 01:52:10 -0700 (PDT) Received: from kwepemm600003.china.huawei.com (unknown [172.30.72.54]) by szxga01-in.huawei.com (SkyGuard) with ESMTP id 4RNDpW6wdxztRhT; Sat, 12 Aug 2023 16:48:35 +0800 (CST) Received: from localhost.localdomain (10.67.174.95) by kwepemm600003.china.huawei.com (7.193.23.202) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27; Sat, 12 Aug 2023 16:52:07 +0800 From: Yang Jihong To: , , , , , , , , , , , , , CC: Subject: [RFC v1 12/16] perf kwork top: Add statistics on softirq event support Date: Sat, 12 Aug 2023 08:49:13 +0000 Message-ID: <20230812084917.169338-13-yangjihong1@huawei.com> X-Mailer: git-send-email 2.30.GIT In-Reply-To: <20230812084917.169338-1-yangjihong1@huawei.com> References: <20230812084917.169338-1-yangjihong1@huawei.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Originating-IP: [10.67.174.95] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemm600003.china.huawei.com (7.193.23.202) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Calculate the runtime of the softirq events and subtract it from the corresponding task runtime to improve the precision. Example usage: # perf kwork -k sched,irq,softirq record -- perf record -e cpu-clock -o p= erf_record.data -a sleep 10 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.467 MB perf_record.data (7154 samples= ) ] [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 2.152 MB perf.data (22846 samples) ] # perf kwork top Total : 136601.588 ms, 8 cpus %Cpu(s): 95.66% id, 0.04% hi, 0.05% si %Cpu0 [ 0.02%] %Cpu1 [ 0.01%] %Cpu2 [| 4.61%] %Cpu3 [ 0.04%] %Cpu4 [ 0.01%] %Cpu5 [||||| 17.31%] %Cpu6 [ 0.51%] %Cpu7 [||| 11.42%] PID %CPU RUNTIME COMMMAND ---------------------------------------------------- 0 99.98 17073.515 ms swapper/4 0 99.98 17072.173 ms swapper/1 0 99.93 17064.229 ms swapper/3 0 99.62 17011.013 ms swapper/0 0 99.47 16985.180 ms swapper/6 0 95.17 16250.874 ms swapper/2 0 88.51 15111.684 ms swapper/7 0 82.62 14108.577 ms swapper/5 4342 33.00 5644.045 ms perf 4344 0.43 74.351 ms perf 16 0.13 22.296 ms rcu_preempt 4345 0.05 10.093 ms perf 4343 0.05 8.769 ms perf 4341 0.02 4.882 ms perf 4095 0.02 4.605 ms kworker/7:1 75 0.02 4.261 ms kworker/2:1 120 0.01 1.909 ms systemd-journal 98 0.01 2.540 ms jbd2/sda-8 61 0.01 3.404 ms kcompactd0 667 0.01 2.542 ms kworker/u16:2 4340 0.00 1.052 ms kworker/7:2 97 0.00 0.489 ms kworker/7:1H 51 0.00 0.209 ms ksoftirqd/7 50 0.00 0.646 ms migration/7 76 0.00 0.753 ms kworker/6:1 45 0.00 0.572 ms migration/6 87 0.00 0.145 ms kworker/5:1H 73 0.00 0.596 ms kworker/5:1 41 0.00 0.041 ms ksoftirqd/5 40 0.00 0.718 ms migration/5 64 0.00 0.115 ms kworker/4:1 35 0.00 0.556 ms migration/4 353 0.00 2.600 ms sshd 74 0.00 0.205 ms kworker/3:1 33 0.00 1.576 ms kworker/3:0H 30 0.00 0.996 ms migration/3 26 0.00 1.665 ms ksoftirqd/2 25 0.00 0.662 ms migration/2 397 0.00 0.057 ms kworker/1:1 20 0.00 1.005 ms migration/1 2909 0.00 1.053 ms kworker/0:2 17 0.00 0.720 ms migration/0 15 0.00 0.039 ms ksoftirqd/0 Signed-off-by: Yang Jihong Reviewed-by: Ian Rogers --- tools/perf/builtin-kwork.c | 30 ++++++++++++++++++++++-------- tools/perf/util/kwork.h | 1 + 2 files changed, 23 insertions(+), 8 deletions(-) diff --git a/tools/perf/builtin-kwork.c b/tools/perf/builtin-kwork.c index 66ad8c2e8020..c741cc1a543f 100644 --- a/tools/perf/builtin-kwork.c +++ b/tools/perf/builtin-kwork.c @@ -1090,7 +1090,7 @@ static char *evsel__softirq_name(struct evsel *evsel,= u64 num) return name; } =20 -static void softirq_work_init(struct perf_kwork *kwork __maybe_unused, +static void softirq_work_init(struct perf_kwork *kwork, struct kwork_class *class, struct kwork_work *work, enum kwork_trace_type src_type __maybe_unused, @@ -1098,12 +1098,19 @@ static void softirq_work_init(struct perf_kwork *kw= ork __maybe_unused, struct perf_sample *sample, struct machine *machine __maybe_unused) { - u64 num =3D evsel__intval(evsel, sample, "vec"); + u64 num; =20 - work->id =3D num; work->class =3D class; work->cpu =3D sample->cpu; - work->name =3D evsel__softirq_name(evsel, num); + + if (kwork->report =3D=3D KWORK_REPORT_TOP) { + work->id =3D evsel__intval_common(evsel, sample, "common_pid"); + work->name =3D NULL; + } else { + num =3D evsel__intval(evsel, sample, "vec"); + work->id =3D num; + work->name =3D evsel__softirq_name(evsel, num); + } } =20 static void softirq_work_name(struct kwork_work *work, char *buf, int len) @@ -1547,6 +1554,7 @@ static void top_print_cpu_usage(struct perf_kwork *kw= ork) struct kwork_top_stat *stat =3D &kwork->top_stat; u64 idle_time =3D stat->cpus_runtime[MAX_NR_CPUS].idle; u64 hardirq_time =3D stat->cpus_runtime[MAX_NR_CPUS].irq; + u64 softirq_time =3D stat->cpus_runtime[MAX_NR_CPUS].softirq; int cpus_nr =3D bitmap_weight(stat->all_cpus_bitmap, MAX_NR_CPUS); u64 cpus_total_time =3D stat->cpus_runtime[MAX_NR_CPUS].total; =20 @@ -1555,12 +1563,15 @@ static void top_print_cpu_usage(struct perf_kwork *= kwork) (double)cpus_total_time / NSEC_PER_MSEC, cpus_nr); =20 - printf("%%Cpu(s): %*.*f%% id, %*.*f%% hi\n", + printf("%%Cpu(s): %*.*f%% id, %*.*f%% hi, %*.*f%% si\n", PRINT_CPU_USAGE_WIDTH, PRINT_CPU_USAGE_DECIMAL_WIDTH, cpus_total_time ? (double)idle_time * 100 / cpus_total_time : 0, =20 PRINT_CPU_USAGE_WIDTH, PRINT_CPU_USAGE_DECIMAL_WIDTH, - cpus_total_time ? (double)hardirq_time * 100 / cpus_total_time : 0= ); + cpus_total_time ? (double)hardirq_time * 100 / cpus_total_time : 0, + + PRINT_CPU_USAGE_WIDTH, PRINT_CPU_USAGE_DECIMAL_WIDTH, + cpus_total_time ? (double)softirq_time * 100 / cpus_total_time : 0= ); =20 top_print_per_cpu_load(kwork); } @@ -1966,6 +1977,9 @@ static void top_calc_irq_runtime(struct perf_kwork *k= work, if (type =3D=3D KWORK_CLASS_IRQ) { stat->cpus_runtime[work->cpu].irq +=3D work->total_runtime; stat->cpus_runtime[MAX_NR_CPUS].irq +=3D work->total_runtime; + } else if (type =3D=3D KWORK_CLASS_SOFTIRQ) { + stat->cpus_runtime[work->cpu].softirq +=3D work->total_runtime; + stat->cpus_runtime[MAX_NR_CPUS].softirq +=3D work->total_runtime; } } =20 @@ -1975,7 +1989,7 @@ static void top_subtract_irq_runtime(struct perf_kwor= k *kwork, struct kwork_class *class; struct kwork_work *data; unsigned int i; - int irq_class_list[] =3D {KWORK_CLASS_IRQ}; + int irq_class_list[] =3D {KWORK_CLASS_IRQ, KWORK_CLASS_SOFTIRQ}; =20 for (i =3D 0; i < ARRAY_SIZE(irq_class_list); i++) { class =3D get_kwork_class(kwork, irq_class_list[i]); @@ -2392,7 +2406,7 @@ int cmd_kwork(int argc, const char **argv) } kwork.report =3D KWORK_REPORT_TOP; if (!kwork.event_list_str) - kwork.event_list_str =3D "sched, irq"; + kwork.event_list_str =3D "sched, irq, softirq"; setup_event_list(&kwork, kwork_options, kwork_usage); setup_sorting(&kwork, top_options, top_usage); return perf_kwork__top(&kwork); diff --git a/tools/perf/util/kwork.h b/tools/perf/util/kwork.h index b3e410d02043..723b34385df6 100644 --- a/tools/perf/util/kwork.h +++ b/tools/perf/util/kwork.h @@ -184,6 +184,7 @@ struct __top_cpus_runtime { u64 load; u64 idle; u64 irq; + u64 softirq; u64 total; }; =20 --=20 2.30.GIT From nobody Fri Dec 19 04:58:51 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 49021C001DB for ; Sat, 12 Aug 2023 08:52:38 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S234509AbjHLIwb (ORCPT ); Sat, 12 Aug 2023 04:52:31 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:52648 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S236443AbjHLIwI (ORCPT ); Sat, 12 Aug 2023 04:52:08 -0400 Received: from szxga01-in.huawei.com (szxga01-in.huawei.com [45.249.212.187]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 344732684; Sat, 12 Aug 2023 01:52:11 -0700 (PDT) Received: from kwepemm600003.china.huawei.com (unknown [172.30.72.57]) by szxga01-in.huawei.com (SkyGuard) with ESMTP id 4RNDpX4Jw2ztRm8; Sat, 12 Aug 2023 16:48:36 +0800 (CST) Received: from localhost.localdomain (10.67.174.95) by kwepemm600003.china.huawei.com (7.193.23.202) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27; Sat, 12 Aug 2023 16:52:08 +0800 From: Yang Jihong To: , , , , , , , , , , , , , CC: Subject: [RFC v1 13/16] perf kwork top: Add -C/--cpu -i/--input -n/--name -s/--sort --time options Date: Sat, 12 Aug 2023 08:49:14 +0000 Message-ID: <20230812084917.169338-14-yangjihong1@huawei.com> X-Mailer: git-send-email 2.30.GIT In-Reply-To: <20230812084917.169338-1-yangjihong1@huawei.com> References: <20230812084917.169338-1-yangjihong1@huawei.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Originating-IP: [10.67.174.95] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemm600003.china.huawei.com (7.193.23.202) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Provide the following options for perf kwork top: 1. -C, --cpu list of cpus to profile 2. -i, --input input file name 3. -n, --name event name to profile 4. -s, --sort sort by key(s): rate, runtime, tid 5. --time Time span for analysis (start,stop) Example usage: # perf kwork top -h Usage: perf kwork top [] -C, --cpu list of cpus to profile -i, --input input file name -n, --name event name to profile -s, --sort sort by key(s): rate, runtime, tid --time Time span for analysis (start,stop) # perf kwork top -C 2,4,5 Total : 51226.940 ms, 3 cpus %Cpu(s): 92.59% id, 0.00% hi, 0.09% si %Cpu2 [| 4.61%] %Cpu4 [ 0.01%] %Cpu5 [||||| 17.31%] PID %CPU RUNTIME COMMMAND ---------------------------------------------------- 0 99.98 17073.515 ms swapper/4 0 95.17 16250.874 ms swapper/2 0 82.62 14108.577 ms swapper/5 4342 21.70 3708.358 ms perf 16 0.13 22.296 ms rcu_preempt 75 0.02 4.261 ms kworker/2:1 98 0.01 2.540 ms jbd2/sda-8 61 0.01 3.404 ms kcompactd0 87 0.00 0.145 ms kworker/5:1H 73 0.00 0.596 ms kworker/5:1 41 0.00 0.041 ms ksoftirqd/5 40 0.00 0.718 ms migration/5 64 0.00 0.115 ms kworker/4:1 35 0.00 0.556 ms migration/4 353 0.00 1.143 ms sshd 26 0.00 1.665 ms ksoftirqd/2 25 0.00 0.662 ms migration/2 # perf kwork top -i perf.data Total : 136601.588 ms, 8 cpus %Cpu(s): 95.66% id, 0.04% hi, 0.05% si %Cpu0 [ 0.02%] %Cpu1 [ 0.01%] %Cpu2 [| 4.61%] %Cpu3 [ 0.04%] %Cpu4 [ 0.01%] %Cpu5 [||||| 17.31%] %Cpu6 [ 0.51%] %Cpu7 [||| 11.42%] PID %CPU RUNTIME COMMMAND ---------------------------------------------------- 0 99.98 17073.515 ms swapper/4 0 99.98 17072.173 ms swapper/1 0 99.93 17064.229 ms swapper/3 0 99.62 17011.013 ms swapper/0 0 99.47 16985.180 ms swapper/6 0 95.17 16250.874 ms swapper/2 0 88.51 15111.684 ms swapper/7 0 82.62 14108.577 ms swapper/5 4342 33.00 5644.045 ms perf 4344 0.43 74.351 ms perf 16 0.13 22.296 ms rcu_preempt 4345 0.05 10.093 ms perf 4343 0.05 8.769 ms perf 4341 0.02 4.882 ms perf 4095 0.02 4.605 ms kworker/7:1 75 0.02 4.261 ms kworker/2:1 120 0.01 1.909 ms systemd-journal 98 0.01 2.540 ms jbd2/sda-8 61 0.01 3.404 ms kcompactd0 667 0.01 2.542 ms kworker/u16:2 4340 0.00 1.052 ms kworker/7:2 97 0.00 0.489 ms kworker/7:1H 51 0.00 0.209 ms ksoftirqd/7 50 0.00 0.646 ms migration/7 76 0.00 0.753 ms kworker/6:1 45 0.00 0.572 ms migration/6 87 0.00 0.145 ms kworker/5:1H 73 0.00 0.596 ms kworker/5:1 41 0.00 0.041 ms ksoftirqd/5 40 0.00 0.718 ms migration/5 64 0.00 0.115 ms kworker/4:1 35 0.00 0.556 ms migration/4 353 0.00 2.600 ms sshd 74 0.00 0.205 ms kworker/3:1 33 0.00 1.576 ms kworker/3:0H 30 0.00 0.996 ms migration/3 26 0.00 1.665 ms ksoftirqd/2 25 0.00 0.662 ms migration/2 397 0.00 0.057 ms kworker/1:1 20 0.00 1.005 ms migration/1 2909 0.00 1.053 ms kworker/0:2 17 0.00 0.720 ms migration/0 15 0.00 0.039 ms ksoftirqd/0 # perf kwork top -n perf Total : 136601.588 ms, 8 cpus %Cpu(s): 95.66% id, 0.04% hi, 0.05% si %Cpu0 [ 0.01%] %Cpu1 [ 0.00%] %Cpu2 [| 4.44%] %Cpu3 [ 0.00%] %Cpu4 [ 0.00%] %Cpu5 [ 0.00%] %Cpu6 [ 0.49%] %Cpu7 [||| 11.38%] PID %CPU RUNTIME COMMMAND ---------------------------------------------------- 4342 15.74 2695.516 ms perf 4344 0.43 74.351 ms perf 4345 0.05 10.093 ms perf 4343 0.05 8.769 ms perf 4341 0.02 4.882 ms perf # perf kwork top -s tid Total : 136601.588 ms, 8 cpus %Cpu(s): 95.66% id, 0.04% hi, 0.05% si %Cpu0 [ 0.02%] %Cpu1 [ 0.01%] %Cpu2 [| 4.61%] %Cpu3 [ 0.04%] %Cpu4 [ 0.01%] %Cpu5 [||||| 17.31%] %Cpu6 [ 0.51%] %Cpu7 [||| 11.42%] PID %CPU RUNTIME COMMMAND ---------------------------------------------------- 0 99.62 17011.013 ms swapper/0 0 99.98 17072.173 ms swapper/1 0 95.17 16250.874 ms swapper/2 0 99.93 17064.229 ms swapper/3 0 99.98 17073.515 ms swapper/4 0 82.62 14108.577 ms swapper/5 0 99.47 16985.180 ms swapper/6 0 88.51 15111.684 ms swapper/7 15 0.00 0.039 ms ksoftirqd/0 16 0.13 22.296 ms rcu_preempt 17 0.00 0.720 ms migration/0 20 0.00 1.005 ms migration/1 25 0.00 0.662 ms migration/2 26 0.00 1.665 ms ksoftirqd/2 30 0.00 0.996 ms migration/3 33 0.00 1.576 ms kworker/3:0H 35 0.00 0.556 ms migration/4 40 0.00 0.718 ms migration/5 41 0.00 0.041 ms ksoftirqd/5 45 0.00 0.572 ms migration/6 50 0.00 0.646 ms migration/7 51 0.00 0.209 ms ksoftirqd/7 61 0.01 3.404 ms kcompactd0 64 0.00 0.115 ms kworker/4:1 73 0.00 0.596 ms kworker/5:1 74 0.00 0.205 ms kworker/3:1 75 0.02 4.261 ms kworker/2:1 76 0.00 0.753 ms kworker/6:1 87 0.00 0.145 ms kworker/5:1H 97 0.00 0.489 ms kworker/7:1H 98 0.01 2.540 ms jbd2/sda-8 120 0.01 1.909 ms systemd-journal 353 0.00 2.600 ms sshd 397 0.00 0.057 ms kworker/1:1 667 0.01 2.542 ms kworker/u16:2 2909 0.00 1.053 ms kworker/0:2 4095 0.02 4.605 ms kworker/7:1 4340 0.00 1.052 ms kworker/7:2 4341 0.02 4.882 ms perf 4342 33.00 5644.045 ms perf 4343 0.05 8.769 ms perf 4344 0.43 74.351 ms perf 4345 0.05 10.093 ms perf # perf kwork top --time 128800, Total : 53495.122 ms, 8 cpus %Cpu(s): 94.71% id, 0.09% hi, 0.09% si %Cpu0 [ 0.07%] %Cpu1 [ 0.04%] %Cpu2 [|| 8.49%] %Cpu3 [ 0.09%] %Cpu4 [ 0.02%] %Cpu5 [ 0.06%] %Cpu6 [ 0.12%] %Cpu7 [|||||| 21.24%] PID %CPU RUNTIME COMMMAND ---------------------------------------------------- 0 99.96 3981.363 ms swapper/4 0 99.94 3978.955 ms swapper/1 0 99.91 9329.375 ms swapper/5 0 99.87 4906.829 ms swapper/3 0 99.86 9028.064 ms swapper/6 0 98.67 3928.161 ms swapper/0 0 91.17 8388.432 ms swapper/2 0 78.65 7125.602 ms swapper/7 4342 29.42 2675.198 ms perf 16 0.18 16.817 ms rcu_preempt 4345 0.09 8.183 ms perf 4344 0.04 4.290 ms perf 4343 0.03 2.844 ms perf 353 0.03 2.600 ms sshd 4095 0.02 2.702 ms kworker/7:1 120 0.02 1.909 ms systemd-journal 98 0.02 2.540 ms jbd2/sda-8 61 0.02 1.886 ms kcompactd0 667 0.02 1.011 ms kworker/u16:2 75 0.02 2.693 ms kworker/2:1 4341 0.01 1.838 ms perf 30 0.01 0.788 ms migration/3 26 0.01 1.665 ms ksoftirqd/2 20 0.01 0.752 ms migration/1 2909 0.01 0.604 ms kworker/0:2 4340 0.00 0.635 ms kworker/7:2 97 0.00 0.214 ms kworker/7:1H 51 0.00 0.209 ms ksoftirqd/7 50 0.00 0.646 ms migration/7 76 0.00 0.602 ms kworker/6:1 45 0.00 0.366 ms migration/6 87 0.00 0.145 ms kworker/5:1H 40 0.00 0.446 ms migration/5 35 0.00 0.318 ms migration/4 74 0.00 0.205 ms kworker/3:1 33 0.00 0.080 ms kworker/3:0H 25 0.00 0.448 ms migration/2 397 0.00 0.057 ms kworker/1:1 17 0.00 0.365 ms migration/0 Signed-off-by: Yang Jihong Reviewed-by: Ian Rogers --- tools/perf/Documentation/perf-kwork.txt | 26 +++++++++++ tools/perf/builtin-kwork.c | 57 +++++++++++++++++++++++-- 2 files changed, 80 insertions(+), 3 deletions(-) diff --git a/tools/perf/Documentation/perf-kwork.txt b/tools/perf/Documenta= tion/perf-kwork.txt index 0601fcb0feea..34d6c285e527 100644 --- a/tools/perf/Documentation/perf-kwork.txt +++ b/tools/perf/Documentation/perf-kwork.txt @@ -178,6 +178,32 @@ OPTIONS for 'perf kwork timehist' stop time is not given (i.e, time string is 'x.y,') then analysis goes to end of file. =20 +OPTIONS for 'perf kwork top' +--------------------------------- + +-C:: +--cpu:: + Only show events for the given CPU(s) (comma separated list). + +-i:: +--input:: + Input file name. (default: perf.data unless stdin is a fifo) + +-n:: +--name:: + Only show events for the given name. + +-s:: +--sort:: + Sort by key(s): rate, runtime, tid + +--time:: + Only analyze samples within given time window: ,. Times + have the format seconds.microseconds. If start is not given (i.e., time + string is ',x.y') then analysis starts at the beginning of the file. If + stop time is not given (i.e, time string is 'x.y,') then analysis goes + to end of file. + SEE ALSO -------- linkperf:perf-record[1] diff --git a/tools/perf/builtin-kwork.c b/tools/perf/builtin-kwork.c index c741cc1a543f..d5949ff4bd15 100644 --- a/tools/perf/builtin-kwork.c +++ b/tools/perf/builtin-kwork.c @@ -146,6 +146,24 @@ static int cpu_usage_cmp(struct kwork_work *l, struct = kwork_work *r) return 0; } =20 +static int id_or_cpu_r_cmp(struct kwork_work *l, struct kwork_work *r) +{ + if (l->id < r->id) + return 1; + if (l->id > r->id) + return -1; + + if (l->id !=3D 0) + return 0; + + if (l->cpu < r->cpu) + return 1; + if (l->cpu > r->cpu) + return -1; + + return 0; +} + static int sort_dimension__add(struct perf_kwork *kwork __maybe_unused, const char *tok, struct list_head *list) { @@ -174,6 +192,10 @@ static int sort_dimension__add(struct perf_kwork *kwor= k __maybe_unused, .name =3D "rate", .cmp =3D cpu_usage_cmp, }; + static struct sort_dimension tid_sort_dimension =3D { + .name =3D "tid", + .cmp =3D id_or_cpu_r_cmp, + }; struct sort_dimension *available_sorts[] =3D { &id_sort_dimension, &max_sort_dimension, @@ -181,6 +203,7 @@ static int sort_dimension__add(struct perf_kwork *kwork= __maybe_unused, &runtime_sort_dimension, &avg_sort_dimension, &rate_sort_dimension, + &tid_sort_dimension, }; =20 if (kwork->report =3D=3D KWORK_REPORT_LATENCY) @@ -381,6 +404,17 @@ static void profile_update_timespan(struct perf_kwork = *kwork, kwork->timeend =3D sample->time; } =20 +static bool profile_name_match(struct perf_kwork *kwork, + struct kwork_work *work) +{ + if (kwork->profile_name && work->name && + (strcmp(work->name, kwork->profile_name) !=3D 0)) { + return false; + } + + return true; +} + static bool profile_event_match(struct perf_kwork *kwork, struct kwork_work *work, struct perf_sample *sample) @@ -396,10 +430,14 @@ static bool profile_event_match(struct perf_kwork *kw= ork, ((ptime->end !=3D 0) && (ptime->end < time))) return false; =20 - if ((kwork->profile_name !=3D NULL) && - (work->name !=3D NULL) && - (strcmp(work->name, kwork->profile_name) !=3D 0)) + /* + * report top needs to collect the runtime of all tasks to + * calculate the load of each core. + */ + if ((kwork->report !=3D KWORK_REPORT_TOP) && + !profile_name_match(kwork, work)) { return false; + } =20 profile_update_timespan(kwork, sample); return true; @@ -2070,6 +2108,9 @@ static void top_merge_tasks(struct perf_kwork *kwork) rb_erase_cached(node, &class->work_root); data =3D rb_entry(node, struct kwork_work, node); =20 + if (!profile_name_match(kwork, data)) + continue; + cpu =3D data->cpu; merged_work =3D find_work_by_id(&merged_root, data->id, data->id =3D=3D 0 ? cpu : -1); @@ -2329,6 +2370,16 @@ int cmd_kwork(int argc, const char **argv) OPT_PARENT(kwork_options) }; const struct option top_options[] =3D { + OPT_STRING('s', "sort", &kwork.sort_order, "key[,key2...]", + "sort by key(s): rate, runtime, tid"), + OPT_STRING('C', "cpu", &kwork.cpu_list, "cpu", + "list of cpus to profile"), + OPT_STRING('n', "name", &kwork.profile_name, "name", + "event name to profile"), + OPT_STRING(0, "time", &kwork.time_str, "str", + "Time span for analysis (start,stop)"), + OPT_STRING('i', "input", &input_name, "file", + "input file name"), OPT_PARENT(kwork_options) }; const char *kwork_usage[] =3D { --=20 2.30.GIT From nobody Fri Dec 19 04:58:51 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id DE5E8C001DB for ; Sat, 12 Aug 2023 08:52:56 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S236990AbjHLIwt (ORCPT ); Sat, 12 Aug 2023 04:52:49 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:52656 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S236458AbjHLIwK (ORCPT ); Sat, 12 Aug 2023 04:52:10 -0400 Received: from szxga08-in.huawei.com (szxga08-in.huawei.com [45.249.212.255]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 4C4E8271E; Sat, 12 Aug 2023 01:52:11 -0700 (PDT) Received: from kwepemm600003.china.huawei.com (unknown [172.30.72.56]) by szxga08-in.huawei.com (SkyGuard) with ESMTP id 4RNDsB4Xtbz1GDYQ; Sat, 12 Aug 2023 16:50:54 +0800 (CST) Received: from localhost.localdomain (10.67.174.95) by kwepemm600003.china.huawei.com (7.193.23.202) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27; Sat, 12 Aug 2023 16:52:08 +0800 From: Yang Jihong To: , , , , , , , , , , , , , CC: Subject: [RFC v1 14/16] perf kwork top: Implements BPF-based cpu usage statistics Date: Sat, 12 Aug 2023 08:49:15 +0000 Message-ID: <20230812084917.169338-15-yangjihong1@huawei.com> X-Mailer: git-send-email 2.30.GIT In-Reply-To: <20230812084917.169338-1-yangjihong1@huawei.com> References: <20230812084917.169338-1-yangjihong1@huawei.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Originating-IP: [10.67.174.95] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemm600003.china.huawei.com (7.193.23.202) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Use bpf to collect statistics on the CPU usage based on perf BPF skeletons. Example usage: # perf kwork top -h Usage: perf kwork top [] -b, --use-bpf Use BPF to measure task cpu usage -C, --cpu list of cpus to profile -i, --input input file name -n, --name event name to profile -s, --sort sort by key(s): rate, runtime, tid --time Time span for analysis (start,stop) # # perf kwork -k sched top -b Starting trace, Hit to stop and report ^C Total : 160702.425 ms, 8 cpus %Cpu(s): 36.00% id, 0.00% hi, 0.00% si %Cpu0 [|||||||||||||||||| 61.66%] %Cpu1 [|||||||||||||||||| 61.27%] %Cpu2 [||||||||||||||||||| 66.40%] %Cpu3 [|||||||||||||||||| 61.28%] %Cpu4 [|||||||||||||||||| 61.82%] %Cpu5 [||||||||||||||||||||||| 77.41%] %Cpu6 [|||||||||||||||||| 61.73%] %Cpu7 [|||||||||||||||||| 63.25%] PID SPID %CPU RUNTIME COMMMAND ------------------------------------------------------------- 0 0 38.72 8089.463 ms [swapper/1] 0 0 38.71 8084.547 ms [swapper/3] 0 0 38.33 8007.532 ms [swapper/0] 0 0 38.26 7992.985 ms [swapper/6] 0 0 38.17 7971.865 ms [swapper/4] 0 0 36.74 7447.765 ms [swapper/7] 0 0 33.59 6486.942 ms [swapper/2] 0 0 22.58 3771.268 ms [swapper/5] 9545 9351 2.48 447.136 ms sched-messaging 9574 9351 2.09 418.583 ms sched-messaging 9724 9351 2.05 372.407 ms sched-messaging 9531 9351 2.01 368.804 ms sched-messaging 9512 9351 2.00 362.250 ms sched-messaging 9514 9351 1.95 357.767 ms sched-messaging 9538 9351 1.86 384.476 ms sched-messaging 9712 9351 1.84 386.490 ms sched-messaging 9723 9351 1.83 380.021 ms sched-messaging 9722 9351 1.82 382.738 ms sched-messaging 9517 9351 1.81 354.794 ms sched-messaging 9559 9351 1.79 344.305 ms sched-messaging 9725 9351 1.77 365.315 ms sched-messaging # perf kwork -k sched top -b -n perf Starting trace, Hit to stop and report ^C Total : 151563.332 ms, 8 cpus %Cpu(s): 26.49% id, 0.00% hi, 0.00% si %Cpu0 [ 0.01%] %Cpu1 [ 0.00%] %Cpu2 [ 0.00%] %Cpu3 [ 0.00%] %Cpu4 [ 0.00%] %Cpu5 [ 0.00%] %Cpu6 [ 0.00%] %Cpu7 [ 0.00%] PID SPID %CPU RUNTIME COMMMAND ------------------------------------------------------------- 9754 9754 0.01 2.303 ms perf # # perf kwork -k sched top -b -C 2,3,4 Starting trace, Hit to stop and report ^C Total : 48016.721 ms, 3 cpus %Cpu(s): 27.82% id, 0.00% hi, 0.00% si %Cpu2 [|||||||||||||||||||||| 74.68%] %Cpu3 [||||||||||||||||||||| 71.06%] %Cpu4 [||||||||||||||||||||| 70.91%] PID SPID %CPU RUNTIME COMMMAND ------------------------------------------------------------- 0 0 29.08 4734.998 ms [swapper/4] 0 0 28.93 4710.029 ms [swapper/3] 0 0 25.31 3912.363 ms [swapper/2] 10248 10158 1.62 264.931 ms sched-messaging 10253 10158 1.62 265.136 ms sched-messaging 10158 10158 1.60 263.013 ms bash 10360 10158 1.49 243.639 ms sched-messaging 10413 10158 1.48 238.604 ms sched-messaging 10531 10158 1.47 234.067 ms sched-messaging 10400 10158 1.47 240.631 ms sched-messaging 10355 10158 1.47 230.586 ms sched-messaging 10377 10158 1.43 234.835 ms sched-messaging 10526 10158 1.42 232.045 ms sched-messaging 10298 10158 1.41 222.396 ms sched-messaging 10410 10158 1.38 221.853 ms sched-messaging 10364 10158 1.38 226.042 ms sched-messaging 10480 10158 1.36 213.633 ms sched-messaging 10370 10158 1.36 223.620 ms sched-messaging 10553 10158 1.34 217.169 ms sched-messaging 10291 10158 1.34 211.516 ms sched-messaging 10251 10158 1.34 218.813 ms sched-messaging 10522 10158 1.33 218.498 ms sched-messaging 10288 10158 1.33 216.787 ms sched-messaging Signed-off-by: Yang Jihong Reviewed-by: Ian Rogers --- tools/perf/Documentation/perf-kwork.txt | 5 + tools/perf/Makefile.perf | 2 +- tools/perf/builtin-kwork.c | 62 ++++- tools/perf/util/Build | 1 + tools/perf/util/bpf_kwork_top.c | 286 +++++++++++++++++++++++ tools/perf/util/bpf_skel/kwork_top.bpf.c | 187 +++++++++++++++ tools/perf/util/kwork.h | 26 +++ 7 files changed, 565 insertions(+), 4 deletions(-) create mode 100644 tools/perf/util/bpf_kwork_top.c create mode 100644 tools/perf/util/bpf_skel/kwork_top.bpf.c diff --git a/tools/perf/Documentation/perf-kwork.txt b/tools/perf/Documenta= tion/perf-kwork.txt index 34d6c285e527..109ace1d5e90 100644 --- a/tools/perf/Documentation/perf-kwork.txt +++ b/tools/perf/Documentation/perf-kwork.txt @@ -33,6 +33,7 @@ There are several variants of 'perf kwork': perf kwork latency -b perf kwork timehist perf kwork top + perf kwork top -b =20 By default it shows the individual work events such as irq, workqeueu, including the run time and delay (time between raise and actually entry= ): @@ -181,6 +182,10 @@ OPTIONS for 'perf kwork timehist' OPTIONS for 'perf kwork top' --------------------------------- =20 +-b:: +--use-bpf:: + Use BPF to measure task cpu usage. + -C:: --cpu:: Only show events for the given CPU(s) (comma separated list). diff --git a/tools/perf/Makefile.perf b/tools/perf/Makefile.perf index 0ed7ee0c1665..bc61c0d0aae4 100644 --- a/tools/perf/Makefile.perf +++ b/tools/perf/Makefile.perf @@ -1054,7 +1054,7 @@ SKELETONS +=3D $(SKEL_OUT)/bperf_leader.skel.h $(SKEL= _OUT)/bperf_follower.skel.h SKELETONS +=3D $(SKEL_OUT)/bperf_cgroup.skel.h $(SKEL_OUT)/func_latency.sk= el.h SKELETONS +=3D $(SKEL_OUT)/off_cpu.skel.h $(SKEL_OUT)/lock_contention.skel= .h SKELETONS +=3D $(SKEL_OUT)/kwork_trace.skel.h $(SKEL_OUT)/sample_filter.sk= el.h -SKELETONS +=3D $(SKEL_OUT)/bench_uprobe.skel.h +SKELETONS +=3D $(SKEL_OUT)/bench_uprobe.skel.h $(SKEL_OUT)/kwork_top.skel.h =20 $(SKEL_TMP_OUT) $(LIBAPI_OUTPUT) $(LIBBPF_OUTPUT) $(LIBPERF_OUTPUT) $(LIBS= UBCMD_OUTPUT) $(LIBSYMBOL_OUTPUT): $(Q)$(MKDIR) -p $@ diff --git a/tools/perf/builtin-kwork.c b/tools/perf/builtin-kwork.c index d5949ff4bd15..04b966801643 100644 --- a/tools/perf/builtin-kwork.c +++ b/tools/perf/builtin-kwork.c @@ -1619,8 +1619,14 @@ static void top_print_header(struct perf_kwork *kwor= k __maybe_unused) int ret; =20 printf("\n "); - ret =3D printf(" %*s %*s %*s %-*s", + ret =3D printf(" %*s %s%*s%s %*s %*s %-*s", PRINT_PID_WIDTH, "PID", + + kwork->use_bpf ? " " : "", + kwork->use_bpf ? PRINT_PID_WIDTH : 0, + kwork->use_bpf ? "SPID" : "", + kwork->use_bpf ? " " : "", + PRINT_CPU_USAGE_WIDTH, "%CPU", PRINT_RUNTIME_HEADER_WIDTH + RPINT_DECIMAL_WIDTH, "RUNTIME", PRINT_TASK_NAME_WIDTH, "COMMMAND"); @@ -1639,6 +1645,12 @@ static int top_print_work(struct perf_kwork *kwork _= _maybe_unused, struct kwork_ */ ret +=3D printf(" %*ld ", PRINT_PID_WIDTH, work->id); =20 + /* + * tgid + */ + if (kwork->use_bpf) + ret +=3D printf(" %*d ", PRINT_PID_WIDTH, work->tgid); + /* * cpu usage */ @@ -1656,7 +1668,13 @@ static int top_print_work(struct perf_kwork *kwork _= _maybe_unused, struct kwork_ /* * command */ - ret +=3D printf(" %-*s", PRINT_TASK_NAME_WIDTH, work->name); + if (kwork->use_bpf) + ret +=3D printf(" %s%s%s", + work->is_kthread ? "[" : "", + work->name, + work->is_kthread ? "]" : ""); + else + ret +=3D printf(" %-*s", PRINT_TASK_NAME_WIDTH, work->name); =20 printf("\n"); return ret; @@ -2153,6 +2171,36 @@ static void perf_kwork__top_report(struct perf_kwork= *kwork) printf("\n"); } =20 +static int perf_kwork__top_bpf(struct perf_kwork *kwork) +{ + int ret; + + signal(SIGINT, sig_handler); + signal(SIGTERM, sig_handler); + + ret =3D perf_kwork__top_prepare_bpf(kwork); + if (ret) + return -1; + + printf("Starting trace, Hit to stop and report\n"); + + perf_kwork__top_start(); + + /* + * a simple pause, wait here for stop signal + */ + pause(); + + perf_kwork__top_finish(); + + perf_kwork__top_read_bpf(kwork); + + perf_kwork__top_cleanup_bpf(); + + return 0; + +} + static int perf_kwork__top(struct perf_kwork *kwork) { struct __top_cpus_runtime *cpus_runtime; @@ -2165,7 +2213,11 @@ static int perf_kwork__top(struct perf_kwork *kwork) kwork->top_stat.cpus_runtime =3D cpus_runtime; bitmap_zero(kwork->top_stat.all_cpus_bitmap, MAX_NR_CPUS); =20 - ret =3D perf_kwork__read_events(kwork); + if (kwork->use_bpf) + ret =3D perf_kwork__top_bpf(kwork); + else + ret =3D perf_kwork__read_events(kwork); + if (ret) goto out; =20 @@ -2380,6 +2432,10 @@ int cmd_kwork(int argc, const char **argv) "Time span for analysis (start,stop)"), OPT_STRING('i', "input", &input_name, "file", "input file name"), +#ifdef HAVE_BPF_SKEL + OPT_BOOLEAN('b', "use-bpf", &kwork.use_bpf, + "Use BPF to measure task cpu usage"), +#endif OPT_PARENT(kwork_options) }; const char *kwork_usage[] =3D { diff --git a/tools/perf/util/Build b/tools/perf/util/Build index d487aec0b458..a46ac80de366 100644 --- a/tools/perf/util/Build +++ b/tools/perf/util/Build @@ -165,6 +165,7 @@ endif =20 ifeq ($(CONFIG_LIBTRACEEVENT),y) perf-$(CONFIG_PERF_BPF_SKEL) +=3D bpf_kwork.o + perf-$(CONFIG_PERF_BPF_SKEL) +=3D bpf_kwork_top.o endif =20 perf-$(CONFIG_BPF_PROLOGUE) +=3D bpf-prologue.o diff --git a/tools/perf/util/bpf_kwork_top.c b/tools/perf/util/bpf_kwork_to= p.c new file mode 100644 index 000000000000..42897ea22c61 --- /dev/null +++ b/tools/perf/util/bpf_kwork_top.c @@ -0,0 +1,286 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * bpf_kwork_top.c + * + * Copyright (c) 2022 Huawei Inc, Yang Jihong + */ + +#include +#include +#include +#include +#include + +#include + +#include "util/debug.h" +#include "util/evsel.h" +#include "util/kwork.h" + +#include +#include + +#include "util/bpf_skel/kwork_top.skel.h" + +/* + * This should be in sync with "util/kwork_top.bpf.c" + */ +#define MAX_COMMAND_LEN 16 + +struct time_data { + __u64 timestamp; +}; + +struct work_data { + __u64 runtime; +}; + +struct task_data { + __u32 tgid; + __u32 is_kthread; + char comm[MAX_COMMAND_LEN]; +}; + +struct work_key { + __u32 type; + __u32 pid; + __u64 task_p; +}; + +struct task_key { + __u32 pid; + __u32 cpu; +}; + +struct kwork_class_bpf { + struct kwork_class *class; + void (*load_prepare)(void); +}; + +static struct kwork_top_bpf *skel; + +void perf_kwork__top_start(void) +{ + struct timespec ts; + + clock_gettime(CLOCK_MONOTONIC, &ts); + skel->bss->from_timestamp =3D (u64)ts.tv_sec * NSEC_PER_SEC + ts.tv_nsec; + skel->bss->enabled =3D 1; + pr_debug("perf kwork top start at: %lld\n", skel->bss->from_timestamp); +} + +void perf_kwork__top_finish(void) +{ + struct timespec ts; + + skel->bss->enabled =3D 0; + clock_gettime(CLOCK_MONOTONIC, &ts); + skel->bss->to_timestamp =3D (u64)ts.tv_sec * NSEC_PER_SEC + ts.tv_nsec; + pr_debug("perf kwork top finish at: %lld\n", skel->bss->to_timestamp); +} + +static void sched_load_prepare(void) +{ + bpf_program__set_autoload(skel->progs.on_switch, true); +} + +static struct kwork_class_bpf kwork_sched_bpf =3D { + .load_prepare =3D sched_load_prepare, +}; + +static struct kwork_class_bpf * +kwork_class_bpf_supported_list[KWORK_CLASS_MAX] =3D { + [KWORK_CLASS_SCHED] =3D &kwork_sched_bpf, +}; + +static bool valid_kwork_class_type(enum kwork_class_type type) +{ + return type >=3D 0 && type < KWORK_CLASS_MAX ? true : false; +} + +static int setup_filters(struct perf_kwork *kwork) +{ + u8 val =3D 1; + int i, nr_cpus, fd; + struct perf_cpu_map *map; + + if (kwork->cpu_list) { + fd =3D bpf_map__fd(skel->maps.kwork_top_cpu_filter); + if (fd < 0) { + pr_debug("Invalid cpu filter fd\n"); + return -1; + } + + map =3D perf_cpu_map__new(kwork->cpu_list); + if (!map) { + pr_debug("Invalid cpu_list\n"); + return -1; + } + + nr_cpus =3D libbpf_num_possible_cpus(); + for (i =3D 0; i < perf_cpu_map__nr(map); i++) { + struct perf_cpu cpu =3D perf_cpu_map__cpu(map, i); + + if (cpu.cpu >=3D nr_cpus) { + perf_cpu_map__put(map); + pr_err("Requested cpu %d too large\n", cpu.cpu); + return -1; + } + bpf_map_update_elem(fd, &cpu.cpu, &val, BPF_ANY); + } + perf_cpu_map__put(map); + + skel->bss->has_cpu_filter =3D 1; + } + + return 0; +} + +int perf_kwork__top_prepare_bpf(struct perf_kwork *kwork __maybe_unused) +{ + struct bpf_program *prog; + struct kwork_class *class; + struct kwork_class_bpf *class_bpf; + enum kwork_class_type type; + + skel =3D kwork_top_bpf__open(); + if (!skel) { + pr_debug("Failed to open kwork top skeleton\n"); + return -1; + } + + /* + * set all progs to non-autoload, + * then set corresponding progs according to config + */ + bpf_object__for_each_program(prog, skel->obj) + bpf_program__set_autoload(prog, false); + + list_for_each_entry(class, &kwork->class_list, list) { + type =3D class->type; + if (!valid_kwork_class_type(type) || + !kwork_class_bpf_supported_list[type]) { + pr_err("Unsupported bpf trace class %s\n", class->name); + goto out; + } + + class_bpf =3D kwork_class_bpf_supported_list[type]; + class_bpf->class =3D class; + + if (class_bpf->load_prepare) + class_bpf->load_prepare(); + } + + if (kwork_top_bpf__load(skel)) { + pr_debug("Failed to load kwork top skeleton\n"); + goto out; + } + + if (setup_filters(kwork)) + goto out; + + if (kwork_top_bpf__attach(skel)) { + pr_debug("Failed to attach kwork top skeleton\n"); + goto out; + } + + return 0; + +out: + kwork_top_bpf__destroy(skel); + return -1; +} + +static void read_task_info(struct kwork_work *work) +{ + int fd; + struct task_data data; + struct task_key key =3D { + .pid =3D work->id, + .cpu =3D work->cpu, + }; + + fd =3D bpf_map__fd(skel->maps.kwork_top_tasks); + if (fd < 0) { + pr_debug("Invalid top tasks map fd\n"); + return; + } + + if (!bpf_map_lookup_elem(fd, &key, &data)) { + work->tgid =3D data.tgid; + work->is_kthread =3D data.is_kthread; + work->name =3D strdup(data.comm); + } +} +static int add_work(struct perf_kwork *kwork, struct work_key *key, + struct work_data *data, int cpu) +{ + struct kwork_class_bpf *bpf_trace; + struct kwork_work *work; + struct kwork_work tmp =3D { + .id =3D key->pid, + .cpu =3D cpu, + .name =3D NULL, + }; + enum kwork_class_type type =3D key->type; + + if (!valid_kwork_class_type(type)) { + pr_debug("Invalid class type %d to add work\n", type); + return -1; + } + + bpf_trace =3D kwork_class_bpf_supported_list[type]; + tmp.class =3D bpf_trace->class; + + work =3D perf_kwork_add_work(kwork, tmp.class, &tmp); + if (!work) + return -1; + + work->total_runtime =3D data->runtime; + read_task_info(work); + + return 0; +} + +int perf_kwork__top_read_bpf(struct perf_kwork *kwork) +{ + int i, fd, nr_cpus; + struct work_data *data; + struct work_key key, prev; + + fd =3D bpf_map__fd(skel->maps.kwork_top_works); + if (fd < 0) { + pr_debug("Invalid top runtime fd\n"); + return -1; + } + + nr_cpus =3D libbpf_num_possible_cpus(); + data =3D calloc(nr_cpus, sizeof(struct work_data)); + if (!data) + return -1; + + memset(&prev, 0, sizeof(prev)); + while (!bpf_map_get_next_key(fd, &prev, &key)) { + if ((bpf_map_lookup_elem(fd, &key, data)) !=3D 0) { + pr_debug("Failed to lookup top elem\n"); + return -1; + } + + for (i =3D 0; i < nr_cpus; i++) { + if (data[i].runtime =3D=3D 0) + continue; + + if (add_work(kwork, &key, &data[i], i)) + return -1; + } + prev =3D key; + } + free(data); + + return 0; +} + +void perf_kwork__top_cleanup_bpf(void) +{ + kwork_top_bpf__destroy(skel); +} diff --git a/tools/perf/util/bpf_skel/kwork_top.bpf.c b/tools/perf/util/bpf= _skel/kwork_top.bpf.c new file mode 100644 index 000000000000..47ad61608ec7 --- /dev/null +++ b/tools/perf/util/bpf_skel/kwork_top.bpf.c @@ -0,0 +1,187 @@ +// SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) +// Copyright (c) 2022, Huawei + +#include "vmlinux.h" +#include +#include +#include + +/* + * This should be in sync with "util/kwork.h" + */ +enum kwork_class_type { + KWORK_CLASS_IRQ, + KWORK_CLASS_SOFTIRQ, + KWORK_CLASS_WORKQUEUE, + KWORK_CLASS_SCHED, + KWORK_CLASS_MAX, +}; + +#define MAX_ENTRIES 102400 +#define MAX_NR_CPUS 2048 +#define PF_KTHREAD 0x00200000 +#define MAX_COMMAND_LEN 16 + +struct time_data { + __u64 timestamp; +}; + +struct work_data { + __u64 runtime; +}; + +struct task_data { + __u32 tgid; + __u32 is_kthread; + char comm[MAX_COMMAND_LEN]; +}; + +struct work_key { + __u32 type; + __u32 pid; + __u64 task_p; +}; + +struct task_key { + __u32 pid; + __u32 cpu; +}; + +struct { + __uint(type, BPF_MAP_TYPE_TASK_STORAGE); + __uint(map_flags, BPF_F_NO_PREALLOC); + __type(key, int); + __type(value, struct time_data); +} kwork_top_task_time SEC(".maps"); + +struct { + __uint(type, BPF_MAP_TYPE_HASH); + __uint(key_size, sizeof(struct task_key)); + __uint(value_size, sizeof(struct task_data)); + __uint(max_entries, MAX_ENTRIES); +} kwork_top_tasks SEC(".maps"); + +struct { + __uint(type, BPF_MAP_TYPE_PERCPU_HASH); + __uint(key_size, sizeof(struct work_key)); + __uint(value_size, sizeof(struct work_data)); + __uint(max_entries, MAX_ENTRIES); +} kwork_top_works SEC(".maps"); + +struct { + __uint(type, BPF_MAP_TYPE_HASH); + __uint(key_size, sizeof(u32)); + __uint(value_size, sizeof(u8)); + __uint(max_entries, MAX_NR_CPUS); +} kwork_top_cpu_filter SEC(".maps"); + +int enabled =3D 0; + +int has_cpu_filter =3D 0; + +__u64 from_timestamp =3D 0; +__u64 to_timestamp =3D 0; + +static __always_inline int cpu_is_filtered(__u32 cpu) +{ + __u8 *cpu_val; + + if (has_cpu_filter) { + cpu_val =3D bpf_map_lookup_elem(&kwork_top_cpu_filter, &cpu); + if (!cpu_val) + return 1; + } + + return 0; +} + +static __always_inline void update_task_info(struct task_struct *task, __u= 32 cpu) +{ + struct task_key key =3D { + .pid =3D task->pid, + .cpu =3D cpu, + }; + + if (!bpf_map_lookup_elem(&kwork_top_tasks, &key)) { + struct task_data data =3D { + .tgid =3D task->tgid, + .is_kthread =3D task->flags & PF_KTHREAD ? 1 : 0, + }; + BPF_CORE_READ_STR_INTO(&data.comm, task, comm); + + bpf_map_update_elem(&kwork_top_tasks, &key, &data, BPF_ANY); + } +} + +static __always_inline void update_work(struct work_key *key, __u64 delta) +{ + struct work_data *data; + + data =3D bpf_map_lookup_elem(&kwork_top_works, key); + if (data) { + data->runtime +=3D delta; + } else { + struct work_data new_data =3D { + .runtime =3D delta, + }; + + bpf_map_update_elem(&kwork_top_works, key, &new_data, BPF_ANY); + } +} + +static void on_sched_out(struct task_struct *task, __u64 ts, __u32 cpu) +{ + __u64 delta; + struct time_data *pelem; + + pelem =3D bpf_task_storage_get(&kwork_top_task_time, task, NULL, 0); + if (pelem) + delta =3D ts - pelem->timestamp; + else + delta =3D ts - from_timestamp; + + struct work_key key =3D { + .type =3D KWORK_CLASS_SCHED, + .pid =3D task->pid, + .task_p =3D (__u64)task, + }; + + update_work(&key, delta); + update_task_info(task, cpu); +} + +static void on_sched_in(struct task_struct *task, __u64 ts) +{ + struct time_data *pelem; + + pelem =3D bpf_task_storage_get(&kwork_top_task_time, task, NULL, + BPF_LOCAL_STORAGE_GET_F_CREATE); + if (pelem) + pelem->timestamp =3D ts; +} + +SEC("tp_btf/sched_switch") +int on_switch(u64 *ctx) +{ + struct task_struct *prev, *next; + + prev =3D (struct task_struct *)ctx[1]; + next =3D (struct task_struct *)ctx[2]; + + if (!enabled) + return 0; + + __u32 cpu =3D bpf_get_smp_processor_id(); + + if (cpu_is_filtered(cpu)) + return 0; + + __u64 ts =3D bpf_ktime_get_ns(); + + on_sched_out(prev, ts, cpu); + on_sched_in(next, ts); + + return 0; +} + +char LICENSE[] SEC("license") =3D "Dual BSD/GPL"; diff --git a/tools/perf/util/kwork.h b/tools/perf/util/kwork.h index 723b34385df6..76fe2a821bcf 100644 --- a/tools/perf/util/kwork.h +++ b/tools/perf/util/kwork.h @@ -135,6 +135,8 @@ struct kwork_work { * top report */ u32 cpu_usage; + u32 tgid; + bool is_kthread; }; =20 struct kwork_class { @@ -264,6 +266,13 @@ void perf_kwork__report_cleanup_bpf(void); void perf_kwork__trace_start(void); void perf_kwork__trace_finish(void); =20 +int perf_kwork__top_prepare_bpf(struct perf_kwork *kwork); +int perf_kwork__top_read_bpf(struct perf_kwork *kwork); +void perf_kwork__top_cleanup_bpf(void); + +void perf_kwork__top_start(void); +void perf_kwork__top_finish(void); + #else /* !HAVE_BPF_SKEL */ =20 static inline int @@ -283,6 +292,23 @@ static inline void perf_kwork__report_cleanup_bpf(void= ) {} static inline void perf_kwork__trace_start(void) {} static inline void perf_kwork__trace_finish(void) {} =20 +static inline int +perf_kwork__top_prepare_bpf(struct perf_kwork *kwork __maybe_unused) +{ + return -1; +} + +static inline int +perf_kwork__top_read_bpf(struct perf_kwork *kwork __maybe_unused) +{ + return -1; +} + +static inline void perf_kwork__top_cleanup_bpf(void) {} + +static inline void perf_kwork__top_start(void) {} +static inline void perf_kwork__top_finish(void) {} + #endif /* HAVE_BPF_SKEL */ =20 #endif /* PERF_UTIL_KWORK_H */ --=20 2.30.GIT From nobody Fri Dec 19 04:58:51 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id B5838C001DB for ; Sat, 12 Aug 2023 08:53:02 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S237053AbjHLIwx (ORCPT ); Sat, 12 Aug 2023 04:52:53 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:52668 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S236490AbjHLIwK (ORCPT ); Sat, 12 Aug 2023 04:52:10 -0400 Received: from szxga02-in.huawei.com (szxga02-in.huawei.com [45.249.212.188]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 2B5FD2D52; Sat, 12 Aug 2023 01:52:12 -0700 (PDT) Received: from kwepemm600003.china.huawei.com (unknown [172.30.72.57]) by szxga02-in.huawei.com (SkyGuard) with ESMTP id 4RNDrL5DLrzTmPP; Sat, 12 Aug 2023 16:50:10 +0800 (CST) Received: from localhost.localdomain (10.67.174.95) by kwepemm600003.china.huawei.com (7.193.23.202) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27; Sat, 12 Aug 2023 16:52:09 +0800 From: Yang Jihong To: , , , , , , , , , , , , , CC: Subject: [RFC v1 15/16] perf kwork top: Add BPF-based statistics on hardirq event support Date: Sat, 12 Aug 2023 08:49:16 +0000 Message-ID: <20230812084917.169338-16-yangjihong1@huawei.com> X-Mailer: git-send-email 2.30.GIT In-Reply-To: <20230812084917.169338-1-yangjihong1@huawei.com> References: <20230812084917.169338-1-yangjihong1@huawei.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Originating-IP: [10.67.174.95] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemm600003.china.huawei.com (7.193.23.202) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Use bpf to collect statistics on hardirq events based on perf BPF skeletons. Example usage: # perf kwork top -k sched,irq -b Starting trace, Hit to stop and report ^C Total : 136717.945 ms, 8 cpus %Cpu(s): 17.10% id, 0.01% hi, 0.00% si %Cpu0 [||||||||||||||||||||||||| 84.26%] %Cpu1 [||||||||||||||||||||||||| 84.77%] %Cpu2 [|||||||||||||||||||||||| 83.22%] %Cpu3 [|||||||||||||||||||||||| 80.37%] %Cpu4 [|||||||||||||||||||||||| 81.49%] %Cpu5 [||||||||||||||||||||||||| 84.68%] %Cpu6 [||||||||||||||||||||||||| 84.48%] %Cpu7 [|||||||||||||||||||||||| 80.21%] PID SPID %CPU RUNTIME COMMMAND ------------------------------------------------------------- 0 0 19.78 3482.833 ms [swapper/7] 0 0 19.62 3454.219 ms [swapper/3] 0 0 18.50 3258.339 ms [swapper/4] 0 0 16.76 2842.749 ms [swapper/2] 0 0 15.71 2627.905 ms [swapper/0] 0 0 15.51 2598.206 ms [swapper/6] 0 0 15.31 2561.820 ms [swapper/5] 0 0 15.22 2548.708 ms [swapper/1] 13253 13018 2.95 513.108 ms sched-messaging 13092 13018 2.67 454.167 ms sched-messaging 13401 13018 2.66 454.790 ms sched-messaging 13240 13018 2.64 454.587 ms sched-messaging 13251 13018 2.61 442.273 ms sched-messaging 13075 13018 2.61 438.932 ms sched-messaging 13220 13018 2.60 443.245 ms sched-messaging 13235 13018 2.59 443.268 ms sched-messaging 13222 13018 2.50 426.344 ms sched-messaging 13410 13018 2.49 426.191 ms sched-messaging 13228 13018 2.46 425.121 ms sched-messaging 13379 13018 2.38 409.950 ms sched-messaging 13236 13018 2.37 413.159 ms sched-messaging 13095 13018 2.36 396.572 ms sched-messaging 13325 13018 2.35 408.089 ms sched-messaging 13242 13018 2.32 394.750 ms sched-messaging 13386 13018 2.31 396.997 ms sched-messaging 13046 13018 2.29 383.833 ms sched-messaging 13109 13018 2.28 388.482 ms sched-messaging 13388 13018 2.28 393.576 ms sched-messaging 13238 13018 2.26 388.487 ms sched-messaging Signed-off-by: Yang Jihong Reviewed-by: Ian Rogers --- tools/perf/util/bpf_kwork_top.c | 11 ++++ tools/perf/util/bpf_skel/kwork_top.bpf.c | 79 ++++++++++++++++++++++++ 2 files changed, 90 insertions(+) diff --git a/tools/perf/util/bpf_kwork_top.c b/tools/perf/util/bpf_kwork_to= p.c index 42897ea22c61..3998bd2a938f 100644 --- a/tools/perf/util/bpf_kwork_top.c +++ b/tools/perf/util/bpf_kwork_top.c @@ -79,6 +79,16 @@ void perf_kwork__top_finish(void) pr_debug("perf kwork top finish at: %lld\n", skel->bss->to_timestamp); } =20 +static void irq_load_prepare(void) +{ + bpf_program__set_autoload(skel->progs.on_irq_handler_entry, true); + bpf_program__set_autoload(skel->progs.on_irq_handler_exit, true); +} + +static struct kwork_class_bpf kwork_irq_bpf =3D { + .load_prepare =3D irq_load_prepare, +}; + static void sched_load_prepare(void) { bpf_program__set_autoload(skel->progs.on_switch, true); @@ -90,6 +100,7 @@ static struct kwork_class_bpf kwork_sched_bpf =3D { =20 static struct kwork_class_bpf * kwork_class_bpf_supported_list[KWORK_CLASS_MAX] =3D { + [KWORK_CLASS_IRQ] =3D &kwork_irq_bpf, [KWORK_CLASS_SCHED] =3D &kwork_sched_bpf, }; =20 diff --git a/tools/perf/util/bpf_skel/kwork_top.bpf.c b/tools/perf/util/bpf= _skel/kwork_top.bpf.c index 47ad61608ec7..9c7dc62386c7 100644 --- a/tools/perf/util/bpf_skel/kwork_top.bpf.c +++ b/tools/perf/util/bpf_skel/kwork_top.bpf.c @@ -54,6 +54,13 @@ struct { __type(value, struct time_data); } kwork_top_task_time SEC(".maps"); =20 +struct { + __uint(type, BPF_MAP_TYPE_PERCPU_HASH); + __uint(key_size, sizeof(struct work_key)); + __uint(value_size, sizeof(struct time_data)); + __uint(max_entries, MAX_ENTRIES); +} kwork_top_irq_time SEC(".maps"); + struct { __uint(type, BPF_MAP_TYPE_HASH); __uint(key_size, sizeof(struct task_key)); @@ -184,4 +191,76 @@ int on_switch(u64 *ctx) return 0; } =20 +SEC("tp_btf/irq_handler_entry") +int on_irq_handler_entry(u64 *cxt) +{ + struct task_struct *task; + + if (!enabled) + return 0; + + __u32 cpu =3D bpf_get_smp_processor_id(); + + if (cpu_is_filtered(cpu)) + return 0; + + __u64 ts =3D bpf_ktime_get_ns(); + + task =3D (struct task_struct *)bpf_get_current_task(); + if (!task) + return 0; + + struct work_key key =3D { + .type =3D KWORK_CLASS_IRQ, + .pid =3D BPF_CORE_READ(task, pid), + .task_p =3D (__u64)task, + }; + + struct time_data data =3D { + .timestamp =3D ts, + }; + + bpf_map_update_elem(&kwork_top_irq_time, &key, &data, BPF_ANY); + + return 0; +} + +SEC("tp_btf/irq_handler_exit") +int on_irq_handler_exit(u64 *cxt) +{ + __u64 delta; + struct task_struct *task; + struct time_data *pelem; + + if (!enabled) + return 0; + + __u32 cpu =3D bpf_get_smp_processor_id(); + + if (cpu_is_filtered(cpu)) + return 0; + + __u64 ts =3D bpf_ktime_get_ns(); + + task =3D (struct task_struct *)bpf_get_current_task(); + if (!task) + return 0; + + struct work_key key =3D { + .type =3D KWORK_CLASS_IRQ, + .pid =3D BPF_CORE_READ(task, pid), + .task_p =3D (__u64)task, + }; + + pelem =3D bpf_map_lookup_elem(&kwork_top_irq_time, &key); + if (pelem && pelem->timestamp !=3D 0) + delta =3D ts - pelem->timestamp; + else + delta =3D ts - from_timestamp; + + update_work(&key, delta); + + return 0; +} + char LICENSE[] SEC("license") =3D "Dual BSD/GPL"; --=20 2.30.GIT From nobody Fri Dec 19 04:58:51 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 5829BC001DB for ; Sat, 12 Aug 2023 08:53:08 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S236842AbjHLIw6 (ORCPT ); Sat, 12 Aug 2023 04:52:58 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:52674 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S236495AbjHLIwK (ORCPT ); Sat, 12 Aug 2023 04:52:10 -0400 Received: from szxga02-in.huawei.com (szxga02-in.huawei.com [45.249.212.188]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id D30BD2D5B; Sat, 12 Aug 2023 01:52:12 -0700 (PDT) Received: from kwepemm600003.china.huawei.com (unknown [172.30.72.57]) by szxga02-in.huawei.com (SkyGuard) with ESMTP id 4RNDpc5sZLzNmjN; Sat, 12 Aug 2023 16:48:40 +0800 (CST) Received: from localhost.localdomain (10.67.174.95) by kwepemm600003.china.huawei.com (7.193.23.202) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27; Sat, 12 Aug 2023 16:52:10 +0800 From: Yang Jihong To: , , , , , , , , , , , , , CC: Subject: [RFC v1 16/16] perf kwork top: Add BPF-based statistics on softirq event support Date: Sat, 12 Aug 2023 08:49:17 +0000 Message-ID: <20230812084917.169338-17-yangjihong1@huawei.com> X-Mailer: git-send-email 2.30.GIT In-Reply-To: <20230812084917.169338-1-yangjihong1@huawei.com> References: <20230812084917.169338-1-yangjihong1@huawei.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Originating-IP: [10.67.174.95] X-ClientProxiedBy: dggems706-chm.china.huawei.com (10.3.19.183) To kwepemm600003.china.huawei.com (7.193.23.202) X-CFilter-Loop: Reflected Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Use bpf to collect statistics on softirq events based on perf BPF skeletons. Example usage: # perf kwork top -b Starting trace, Hit to stop and report ^C Total : 135445.704 ms, 8 cpus %Cpu(s): 28.35% id, 0.00% hi, 0.25% si %Cpu0 [|||||||||||||||||||| 69.85%] %Cpu1 [|||||||||||||||||||||| 74.10%] %Cpu2 [||||||||||||||||||||| 71.18%] %Cpu3 [|||||||||||||||||||| 69.61%] %Cpu4 [|||||||||||||||||||||| 74.05%] %Cpu5 [|||||||||||||||||||| 69.33%] %Cpu6 [|||||||||||||||||||| 69.71%] %Cpu7 [|||||||||||||||||||||| 73.77%] PID SPID %CPU RUNTIME COMMMAND ------------------------------------------------------------- 0 0 30.43 5271.005 ms [swapper/5] 0 0 30.17 5226.644 ms [swapper/3] 0 0 30.08 5210.257 ms [swapper/6] 0 0 29.89 5177.177 ms [swapper/0] 0 0 28.51 4938.672 ms [swapper/2] 0 0 25.93 4223.464 ms [swapper/7] 0 0 25.69 4181.411 ms [swapper/4] 0 0 25.63 4173.804 ms [swapper/1] 16665 16265 2.16 360.600 ms sched-messaging 16537 16265 2.05 356.275 ms sched-messaging 16503 16265 2.01 343.063 ms sched-messaging 16424 16265 1.97 336.876 ms sched-messaging 16580 16265 1.94 323.658 ms sched-messaging 16515 16265 1.92 321.616 ms sched-messaging 16659 16265 1.91 325.538 ms sched-messaging 16634 16265 1.88 327.766 ms sched-messaging 16454 16265 1.87 326.843 ms sched-messaging 16382 16265 1.87 322.591 ms sched-messaging 16642 16265 1.86 320.506 ms sched-messaging 16582 16265 1.86 320.164 ms sched-messaging 16315 16265 1.86 326.872 ms sched-messaging 16637 16265 1.85 323.766 ms sched-messaging 16506 16265 1.82 311.688 ms sched-messaging 16512 16265 1.81 304.643 ms sched-messaging 16560 16265 1.80 314.751 ms sched-messaging 16320 16265 1.80 313.405 ms sched-messaging 16442 16265 1.80 314.403 ms sched-messaging 16626 16265 1.78 295.380 ms sched-messaging 16600 16265 1.77 309.444 ms sched-messaging 16550 16265 1.76 301.161 ms sched-messaging 16525 16265 1.75 296.560 ms sched-messaging 16314 16265 1.75 298.338 ms sched-messaging 16595 16265 1.74 304.390 ms sched-messaging 16555 16265 1.74 287.564 ms sched-messaging 16520 16265 1.74 295.734 ms sched-messaging 16507 16265 1.73 293.956 ms sched-messaging 16593 16265 1.72 296.443 ms sched-messaging 16531 16265 1.72 299.950 ms sched-messaging 16281 16265 1.72 301.339 ms sched-messaging Signed-off-by: Yang Jihong Reviewed-by: Ian Rogers --- tools/perf/util/bpf_kwork_top.c | 11 ++++ tools/perf/util/bpf_skel/kwork_top.bpf.c | 72 ++++++++++++++++++++++++ 2 files changed, 83 insertions(+) diff --git a/tools/perf/util/bpf_kwork_top.c b/tools/perf/util/bpf_kwork_to= p.c index 3998bd2a938f..1a607b94f44d 100644 --- a/tools/perf/util/bpf_kwork_top.c +++ b/tools/perf/util/bpf_kwork_top.c @@ -89,6 +89,16 @@ static struct kwork_class_bpf kwork_irq_bpf =3D { .load_prepare =3D irq_load_prepare, }; =20 +static void softirq_load_prepare(void) +{ + bpf_program__set_autoload(skel->progs.on_softirq_entry, true); + bpf_program__set_autoload(skel->progs.on_softirq_exit, true); +} + +static struct kwork_class_bpf kwork_softirq_bpf =3D { + .load_prepare =3D softirq_load_prepare, +}; + static void sched_load_prepare(void) { bpf_program__set_autoload(skel->progs.on_switch, true); @@ -101,6 +111,7 @@ static struct kwork_class_bpf kwork_sched_bpf =3D { static struct kwork_class_bpf * kwork_class_bpf_supported_list[KWORK_CLASS_MAX] =3D { [KWORK_CLASS_IRQ] =3D &kwork_irq_bpf, + [KWORK_CLASS_SOFTIRQ] =3D &kwork_softirq_bpf, [KWORK_CLASS_SCHED] =3D &kwork_sched_bpf, }; =20 diff --git a/tools/perf/util/bpf_skel/kwork_top.bpf.c b/tools/perf/util/bpf= _skel/kwork_top.bpf.c index 9c7dc62386c7..84c15ccbab44 100644 --- a/tools/perf/util/bpf_skel/kwork_top.bpf.c +++ b/tools/perf/util/bpf_skel/kwork_top.bpf.c @@ -263,4 +263,76 @@ int on_irq_handler_exit(u64 *cxt) return 0; } =20 +SEC("tp_btf/softirq_entry") +int on_softirq_entry(u64 *cxt) +{ + struct task_struct *task; + + if (!enabled) + return 0; + + __u32 cpu =3D bpf_get_smp_processor_id(); + + if (cpu_is_filtered(cpu)) + return 0; + + __u64 ts =3D bpf_ktime_get_ns(); + + task =3D (struct task_struct *)bpf_get_current_task(); + if (!task) + return 0; + + struct work_key key =3D { + .type =3D KWORK_CLASS_SOFTIRQ, + .pid =3D BPF_CORE_READ(task, pid), + .task_p =3D (__u64)task, + }; + + struct time_data data =3D { + .timestamp =3D ts, + }; + + bpf_map_update_elem(&kwork_top_irq_time, &key, &data, BPF_ANY); + + return 0; +} + +SEC("tp_btf/softirq_exit") +int on_softirq_exit(u64 *cxt) +{ + __u64 delta; + struct task_struct *task; + struct time_data *pelem; + + if (!enabled) + return 0; + + __u32 cpu =3D bpf_get_smp_processor_id(); + + if (cpu_is_filtered(cpu)) + return 0; + + __u64 ts =3D bpf_ktime_get_ns(); + + task =3D (struct task_struct *)bpf_get_current_task(); + if (!task) + return 0; + + struct work_key key =3D { + .type =3D KWORK_CLASS_SOFTIRQ, + .pid =3D BPF_CORE_READ(task, pid), + .task_p =3D (__u64)task, + }; + + pelem =3D bpf_map_lookup_elem(&kwork_top_irq_time, &key); + if (pelem) + delta =3D ts - pelem->timestamp; + else + delta =3D ts - from_timestamp; + + update_work(&key, delta); + + return 0; +} + char LICENSE[] SEC("license") =3D "Dual BSD/GPL"; --=20 2.30.GIT