From nobody Mon Sep 15 12:54:45 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 9A34DC678D4 for ; Thu, 12 Jan 2023 21:48:59 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S233018AbjALVs5 (ORCPT ); Thu, 12 Jan 2023 16:48:57 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:40492 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S240757AbjALVrg (ORCPT ); Thu, 12 Jan 2023 16:47:36 -0500 Received: from mail-pf1-x429.google.com (mail-pf1-x429.google.com [IPv6:2607:f8b0:4864:20::429]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 17E231B9F3; Thu, 12 Jan 2023 13:40:29 -0800 (PST) Received: by mail-pf1-x429.google.com with SMTP id 20so9371786pfu.13; Thu, 12 Jan 2023 13:40:29 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:sender:from:to:cc:subject:date :message-id:reply-to; bh=cFiEc4fnXcWosSJZXgNQt+xQQqIdoE277uZe1x+iaq4=; b=bPETWCW5IbBSYQyy6jMD+fldm74bgEgo3vBdHRXKU3fnmlEGP0/eT6B0u73M/2AaA8 IyHPPe4jHkK0Mqa+tZkgNBza/2P4NC6Gs+eRQbyk/BP4gXjedz4eShGS743BR2XKU8lJ o7gjuCg2rtuHjcWL/U75YpXufj1FHXlQWL/oAQtd9HqbWs1Ta3iy7sV184SDWfXKIp/O HZvKzFb3XErNYlGNTIkXj6qNYEbh8jGJVIlb6G1Es9Ay7KvwaKzA1h46lFeFd4tuIVFj dGu1fC7IMoBpIbuqTqFrL4o6DJZJwAZCPZ4mnJ7Dw6ioHTKeQNCX4V70AMNzlsKPHMD+ 5lUQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:sender:x-gm-message-state:from :to:cc:subject:date:message-id:reply-to; bh=cFiEc4fnXcWosSJZXgNQt+xQQqIdoE277uZe1x+iaq4=; b=edBeMQHaJQEbrIDYmow1C+o8Gxv77S7kcr0Z02IHyMsZOl+rxQ8Kbj1hwpMt0XX67v 6mtHCadfLkTfYHmGltT2/52duYYGQB9AAHeryI8vPNRUdw/VzuZp3Kz1YvARsTJcFidl yNvCeAGTWGi6j210zynVmuW7CqxZLilP/Q/1v+zPyl+2UeCsWTtvXfVDhohPaXAdkKCL l0h1dJEHYq2jMusrpcLcwSQ+EY9vJ9FsvCiwAeiG0Hym03DPoLT9r0MMknRy9uLpk4ue lPJ7cSGPB3Cn2CwGiI09WdLt0gV6oSi4z4aS1vUa4+1YK602ygWwVNCrPVqnl8Z0WDOd vBRA== X-Gm-Message-State: AFqh2kqGS4FSTLWzk+Zcj5fJT49UR9K7i8+ULho25kIKhzkGNBVNtzql wKv1bbuBOE+sQcyhb/FW6Zc= X-Google-Smtp-Source: AMrXdXtl1PJTJmv7dZW6vqrL/16Dcpwl/KSf0TEBH/jIL9IX3GY2dpbk9kGVN/lQNfCJ1CQS/Sh6Vw== X-Received: by 2002:aa7:9732:0:b0:580:d722:10e7 with SMTP id k18-20020aa79732000000b00580d72210e7mr71908290pfg.24.1673559628420; Thu, 12 Jan 2023 13:40:28 -0800 (PST) Received: from youngsil.svl.corp.google.com ([2620:15c:2d4:203:b71e:23d4:aa6a:47c0]) by smtp.gmail.com with ESMTPSA id q20-20020aa79834000000b0058134d2df41sm11344605pfl.146.2023.01.12.13.40.27 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Thu, 12 Jan 2023 13:40:28 -0800 (PST) Sender: Namhyung Kim From: Namhyung Kim To: Peter Zijlstra Cc: Ingo Molnar , Mark Rutland , Alexander Shishkin , Arnaldo Carvalho de Melo , Jiri Olsa , LKML , Song Liu , linux-s390@vger.kernel.org, x86@kernel.org Subject: [PATCH 7/8] perf/core: Introduce perf_prepare_header() Date: Thu, 12 Jan 2023 13:40:14 -0800 Message-Id: <20230112214015.1014857-8-namhyung@kernel.org> X-Mailer: git-send-email 2.39.0.314.g84b9a713c41-goog In-Reply-To: <20230112214015.1014857-1-namhyung@kernel.org> References: <20230112214015.1014857-1-namhyung@kernel.org> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" Factor out perf_prepare_header() so that it can call perf_prepare_sample() without a header if not needed. Also it checks the filtered_sample_type to avoid duplicate work when perf_prepare_sample() is called twice (or more). Cc: linux-s390@vger.kernel.org Cc: x86@kernel.org Suggested-by: Peter Zijlstr Signed-off-by: Namhyung Kim Acked-by: Song Liu --- arch/s390/kernel/perf_cpum_sf.c | 3 ++- arch/x86/events/intel/ds.c | 3 ++- include/linux/perf_event.h | 16 +++++++++++++- kernel/events/core.c | 38 +++++++++++++++++++++------------ 4 files changed, 43 insertions(+), 17 deletions(-) diff --git a/arch/s390/kernel/perf_cpum_sf.c b/arch/s390/kernel/perf_cpum_s= f.c index 332a49965130..fd02f8423243 100644 --- a/arch/s390/kernel/perf_cpum_sf.c +++ b/arch/s390/kernel/perf_cpum_sf.c @@ -671,7 +671,8 @@ static void cpumsf_output_event_pid(struct perf_event *= event, /* Protect callchain buffers, tasks */ rcu_read_lock(); =20 - perf_prepare_sample(&header, data, event, regs); + perf_prepare_sample(data, event, regs); + perf_prepare_header(&header, data, event, regs); if (perf_output_begin(&handle, data, event, header.size)) goto out; =20 diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c index 07c8a2cdc3ee..183efa914b99 100644 --- a/arch/x86/events/intel/ds.c +++ b/arch/x86/events/intel/ds.c @@ -807,7 +807,8 @@ int intel_pmu_drain_bts_buffer(void) * the sample. */ rcu_read_lock(); - perf_prepare_sample(&header, &data, event, ®s); + perf_prepare_sample(&data, event, ®s); + perf_prepare_header(&header, &data, event, ®s); =20 if (perf_output_begin(&handle, &data, event, header.size * (top - base - skip))) diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h index 7db0e9cc2682..d5628a7b5eaa 100644 --- a/include/linux/perf_event.h +++ b/include/linux/perf_event.h @@ -1250,6 +1250,17 @@ static inline void perf_sample_save_brstack(struct p= erf_sample_data *data, data->sample_flags |=3D PERF_SAMPLE_BRANCH_STACK; } =20 +static inline u32 perf_sample_data_size(struct perf_sample_data *data, + struct perf_event *event) +{ + u32 size =3D sizeof(struct perf_event_header); + + size +=3D event->header_size + event->id_header_size; + size +=3D data->dyn_size; + + return size; +} + /* * Clear all bitfields in the perf_branch_entry. * The to and from fields are not cleared because they are @@ -1271,7 +1282,10 @@ extern void perf_output_sample(struct perf_output_ha= ndle *handle, struct perf_event_header *header, struct perf_sample_data *data, struct perf_event *event); -extern void perf_prepare_sample(struct perf_event_header *header, +extern void perf_prepare_sample(struct perf_sample_data *data, + struct perf_event *event, + struct pt_regs *regs); +extern void perf_prepare_header(struct perf_event_header *header, struct perf_sample_data *data, struct perf_event *event, struct pt_regs *regs); diff --git a/kernel/events/core.c b/kernel/events/core.c index 127dddd20f93..5c4f3fa3d2b7 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -7574,20 +7574,13 @@ static __always_inline u64 __cond_set(u64 flags, u6= 4 s, u64 d) return d * !!(flags & s); } =20 -void perf_prepare_sample(struct perf_event_header *header, - struct perf_sample_data *data, +void perf_prepare_sample(struct perf_sample_data *data, struct perf_event *event, struct pt_regs *regs) { u64 sample_type =3D event->attr.sample_type; u64 filtered_sample_type; =20 - header->type =3D PERF_RECORD_SAMPLE; - header->size =3D sizeof(*header) + event->header_size + event->id_header_= size; - - header->misc =3D 0; - header->misc |=3D perf_misc_flags(regs); - /* * Add the sample flags that are dependent to others. And clear the * sample flags that have already been done by the PMU driver. @@ -7601,6 +7594,12 @@ void perf_prepare_sample(struct perf_event_header *h= eader, PERF_SAMPLE_REGS_USER); filtered_sample_type &=3D ~data->sample_flags; =20 + if (filtered_sample_type =3D=3D 0) { + /* Make sure it has the correct data->type for output */ + data->type =3D event->attr.sample_type; + return; + } + __perf_event_header__init_id(data, event, filtered_sample_type); data->sample_flags |=3D sample_type & PERF_SAMPLE_ID_ALL; =20 @@ -7653,9 +7652,10 @@ void perf_prepare_sample(struct perf_event_header *h= eader, * up the rest of the sample size. */ u16 stack_size =3D event->attr.sample_stack_user; + u16 header_size =3D perf_sample_data_size(data, event); u16 size =3D sizeof(u64); =20 - stack_size =3D perf_sample_ustack_size(stack_size, header->size, + stack_size =3D perf_sample_ustack_size(stack_size, header_size, data->regs_user.regs); =20 /* @@ -7740,8 +7740,9 @@ void perf_prepare_sample(struct perf_event_header *he= ader, =20 if (filtered_sample_type & PERF_SAMPLE_AUX) { u64 size; + u16 header_size =3D perf_sample_data_size(data, event); =20 - header->size +=3D sizeof(u64); /* size */ + header_size +=3D sizeof(u64); /* size */ =20 /* * Given the 16bit nature of header::size, an AUX sample can @@ -7749,17 +7750,25 @@ void perf_prepare_sample(struct perf_event_header *= header, * Make sure this doesn't happen by using up to U16_MAX bytes * per sample in total (rounded down to 8 byte boundary). */ - size =3D min_t(size_t, U16_MAX - header->size, + size =3D min_t(size_t, U16_MAX - header_size, event->attr.aux_sample_size); size =3D rounddown(size, 8); size =3D perf_prepare_sample_aux(event, data, size); =20 - WARN_ON_ONCE(size + header->size > U16_MAX); + WARN_ON_ONCE(size + header_size > U16_MAX); data->dyn_size +=3D size + sizeof(u64); /* size above */ data->sample_flags |=3D PERF_SAMPLE_AUX; } +} =20 - header->size +=3D data->dyn_size; +void perf_prepare_header(struct perf_event_header *header, + struct perf_sample_data *data, + struct perf_event *event, + struct pt_regs *regs) +{ + header->type =3D PERF_RECORD_SAMPLE; + header->size =3D perf_sample_data_size(data, event); + header->misc =3D perf_misc_flags(regs); =20 /* * If you're adding more sample types here, you likely need to do @@ -7788,7 +7797,8 @@ __perf_event_output(struct perf_event *event, /* protect the callchain buffers */ rcu_read_lock(); =20 - perf_prepare_sample(&header, data, event, regs); + perf_prepare_sample(data, event, regs); + perf_prepare_header(&header, data, event, regs); =20 err =3D output_begin(&handle, data, event, header.size); if (err) --=20 2.39.0.314.g84b9a713c41-goog