Received: by 2002:a05:6a10:1287:0:0:0:0 with SMTP id d7csp6567726pxv; Thu, 29 Jul 2021 18:46:20 -0700 (PDT) X-Google-Smtp-Source: ABdhPJwT4q5e/GTK1iwLQos3TtlWCyOXpEaLx5jhmMmOBrx6fZkhowXlHfzbSQrN5QWTrLEx6xvJ X-Received: by 2002:a05:6402:10cc:: with SMTP id p12mr53371edu.328.1627609579952; Thu, 29 Jul 2021 18:46:19 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1627609579; cv=none; d=google.com; s=arc-20160816; b=gInxp2PAPrezth5dxi4rSdFbqLeHrC0xHzTni0zdKwgx8wR9ybzP9SohUXNeBB0s7z d5gFAPzkeyZVhlqo3FqiakshwsN907luQ5i+rgIa7YLyro2ImRz0bu8Zq6d5DFc6cg5L gDL+1ntmjpqT9r1vj1nIe4Om1W6BnibepBXw63h4eIggPE4dV+4LLy4UJYDBJkNLAcZv jOAkKyYDIHfEWKm2v7Wf8PoBf5/RBVuyvEpy/WqzzuRyHCoMWAo8RHqjHsWwtdjWWBsB EeeJ8w2exmcq69Qn27pA4YhALBOHTm9mPXy85BVL87WhKeTmX2va6aZTwtZLlvlCYvC4 EDxQ== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=list-id:precedence:content-transfer-encoding:content-language :in-reply-to:mime-version:user-agent:date:message-id:from:references :cc:to:subject; bh=/NvwIOWQinjGn49eOwgnz8pYG+aAIzd+25QcAi4I2To=; b=Aeoy0fWc1XncJ2T1kw53KQE2h+G98cC0y9gaEPx9+KTRGHgBoX33wDKUxe7D+Tn84o oBbFwskmMDmKBwNEU2Ohg0rpHo+t1UpHU3/CuAM09d4NIJso2BLx2RswNW/elYwR9O8p Yhi0DlM6Hl1UpPOYcclI44z3OCtlUnryOnyRBlA+3UifYqRBJoH6yw6zXPYvtZ7yCEIS EOJ5Jr6ahCgD70nPcixCthG5gX+zaYJpy6Q05pvbLIXN8slLo+ncNa+RLsdyahQPEHqH T+Oxd4WNIAKkc6b1rfGgTZdNXkngxjtVYprOyolC21Ins0YVwPpoeGZ6iKUA29H62i0I 2Pog== ARC-Authentication-Results: i=1; mx.google.com; spf=pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=fail (p=NONE sp=NONE dis=NONE) header.from=intel.com Return-Path: Received: from vger.kernel.org (vger.kernel.org. [23.128.96.18]) by mx.google.com with ESMTP id i8si210259ejj.636.2021.07.29.18.45.57; Thu, 29 Jul 2021 18:46:19 -0700 (PDT) Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18; Authentication-Results: mx.google.com; spf=pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=fail (p=NONE sp=NONE dis=NONE) header.from=intel.com Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S234659AbhG3Bnb (ORCPT + 99 others); Thu, 29 Jul 2021 21:43:31 -0400 Received: from mga03.intel.com ([134.134.136.65]:2632 "EHLO mga03.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S234495AbhG3BnT (ORCPT ); Thu, 29 Jul 2021 21:43:19 -0400 X-IronPort-AV: E=McAfee;i="6200,9189,10060"; a="213042952" X-IronPort-AV: E=Sophos;i="5.84,280,1620716400"; d="scan'208";a="213042952" Received: from fmsmga003.fm.intel.com ([10.253.24.29]) by orsmga103.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 29 Jul 2021 18:43:14 -0700 X-IronPort-AV: E=Sophos;i="5.84,280,1620716400"; d="scan'208";a="507550294" Received: from yjin15-mobl1.ccr.corp.intel.com (HELO [10.238.4.147]) ([10.238.4.147]) by fmsmga003-auth.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 29 Jul 2021 18:43:12 -0700 Subject: Re: [PATCH v2 1/2] perf pmu: Add PMU alias support To: Riccardo Mancini Cc: acme@kernel.org, jolsa@kernel.org, peterz@infradead.org, mingo@redhat.com, alexander.shishkin@linux.intel.com, Linux-kernel@vger.kernel.org, linux-perf-users@vger.kernel.org, ak@linux.intel.com, Kan Liang References: <20210729070619.20726-1-yao.jin@linux.intel.com> <20210729070619.20726-2-yao.jin@linux.intel.com> <862ca312fe2b500e2d26bbfdfdadec3d2a705ac1.camel@gmail.com> From: "Jin, Yao" Message-ID: Date: Fri, 30 Jul 2021 09:43:09 +0800 User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:78.0) Gecko/20100101 Thunderbird/78.12.0 MIME-Version: 1.0 In-Reply-To: <862ca312fe2b500e2d26bbfdfdadec3d2a705ac1.camel@gmail.com> Content-Type: text/plain; charset=utf-8; format=flowed Content-Language: en-US Content-Transfer-Encoding: 8bit Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi Riccardo, On 7/29/2021 9:13 PM, Riccardo Mancini wrote: > Hi, > > On Thu, 2021-07-29 at 15:06 +0800, Jin Yao wrote: >> From: Kan Liang >> >> A perf uncore PMU may have two PMU names, a real name and an alias. The >> alias is exported at /sys/bus/event_source/devices/uncore_*/alias. >> The perf tool should support the alias as well. >> >> Add alias_name in the struct perf_pmu to store the alias. For the PMU >> which doesn't have an alias. It's NULL. >> >> Introduce two X86 specific functions to retrieve the real name and the >> alias separately. >> >> Only go through the sysfs to retrieve the mapping between the real name >> and the alias once. The result is cached in a list, uncore_pmu_list. >> >> Nothing changed for the other ARCHs. >> >> With the patch, the perf tool can monitor the PMU with either the real >> name or the alias. >> >> Use the real name, >>  $ perf stat -e uncore_cha_2/event=1/ -x, >>    4044879584,,uncore_cha_2/event=1/,2528059205,100.00,, >> >> Use the alias, >>  $ perf stat -e uncore_type_0_2/event=1/ -x, >>    3659675336,,uncore_type_0_2/event=1/,2287306455,100.00,, >> >> Co-developed-by: Jin Yao >> Signed-off-by: Jin Yao >> Signed-off-by: Kan Liang >> --- >> v2: >>  - No change. >> >>  tools/perf/arch/x86/util/pmu.c | 109 ++++++++++++++++++++++++++++++++- >>  tools/perf/util/parse-events.y |   3 +- >>  tools/perf/util/pmu.c          |  26 +++++++- >>  tools/perf/util/pmu.h          |   5 ++ >>  4 files changed, 138 insertions(+), 5 deletions(-) >> >> diff --git a/tools/perf/arch/x86/util/pmu.c b/tools/perf/arch/x86/util/pmu.c >> index d48d608517fd..bb79b1d19b96 100644 >> --- a/tools/perf/arch/x86/util/pmu.c >> +++ b/tools/perf/arch/x86/util/pmu.c >> @@ -1,12 +1,28 @@ >>  // SPDX-License-Identifier: GPL-2.0 >>  #include >> - >> +#include >> +#include >> +#include >> +#include >>  #include >>  #include >> +#include >> +#include >> >>  #include "../../../util/intel-pt.h" >>  #include "../../../util/intel-bts.h" >>  #include "../../../util/pmu.h" >> +#include "../../../util/fncache.h" >> + >> +#define TEMPLATE_ALIAS "%s/bus/event_source/devices/%s/alias" >> + >> +struct perf_pmu_alias_name { >> +       char *name; >> +       char *alias; >> +       struct list_head list; >> +}; >> + >> +static LIST_HEAD(pmu_alias_name_list); >> >>  struct perf_event_attr *perf_pmu__get_default_config(struct perf_pmu *pmu >> __maybe_unused) >>  { >> @@ -18,3 +34,94 @@ struct perf_event_attr *perf_pmu__get_default_config(struct >> perf_pmu *pmu __mayb >>  #endif >>         return NULL; >>  } >> + >> +static void setup_pmu_alias_list(void) >> +{ >> +       char path[PATH_MAX]; >> +       DIR *dir; >> +       struct dirent *dent; >> +       const char *sysfs = sysfs__mountpoint(); >> +       struct perf_pmu_alias_name *pmu; >> +       char buf[MAX_PMU_NAME_LEN]; >> +       FILE *file; >> + >> +       if (!sysfs) >> +               return; >> + >> +       snprintf(path, PATH_MAX, >> +                "%s" EVENT_SOURCE_DEVICE_PATH, sysfs); >> + >> +       dir = opendir(path); >> +       if (!dir) >> +               return; >> + >> +       while ((dent = readdir(dir))) { >> +               if (!strcmp(dent->d_name, ".") || >> +                   !strcmp(dent->d_name, "..")) >> +                       continue; >> + >> +               snprintf(path, PATH_MAX, >> +                        TEMPLATE_ALIAS, sysfs, dent->d_name); >> + >> +               if (!file_available(path)) >> +                       continue; >> + >> +               file = fopen(path, "r"); >> +               if (!file) >> +                       continue; >> + >> +               if (fscanf(file, "%s", buf) != 1) >> +                       continue; > > Using fscanf with "%s" can lead to a buffer overflow. > It's better to use fgets, if possible, or to specify a maximum length (better if > through a macro, e.g. "%" STR(MAX_PMU_NAME_LEN) "s"). > Yes, fgets should be better. I will use 'fgets(buf, sizeof(buf), file)' instead in v3. >> + >> +               pmu = zalloc(sizeof(*pmu)); >> +               if (!pmu) >> +                       continue; >> + >> +               pmu->alias = strdup(buf); >> +               if (!pmu->alias) { >> +                       free(pmu); >> +                       continue; >> +               } >> +               pmu->name = strdup(dent->d_name); > > The result of strdup is not checked. > It's also probably better to return an error in case of ENOMEM, instead of > continuing iteration, as also previous checks do. > Yes, strictly we should check the return value, I will fix that. Thanks Jin Yao > Thanks, > Riccardo > >> +               list_add_tail(&pmu->list, &pmu_alias_name_list); >> +               fclose(file); >> +       } >> + >> +       closedir(dir); >> +} >> + >> +static char *__pmu_find_real_name(const char *name) >> +{ >> +       struct perf_pmu_alias_name *pmu; >> + >> +       list_for_each_entry(pmu, &pmu_alias_name_list, list) { >> +               if (!strcmp(name, pmu->alias)) >> +                       return strdup(pmu->name); >> +       } >> + >> +       return strdup(name); >> +} >> + >> +char *pmu_find_real_name(const char *name) >> +{ >> +       static bool cached_list; >> + >> +       if (cached_list) >> +               return __pmu_find_real_name(name); >> + >> +       setup_pmu_alias_list(); >> +       cached_list = true; >> + >> +       return __pmu_find_real_name(name); >> +} >> + >> +char *pmu_find_alias_name(const char *name) >> +{ >> +       struct perf_pmu_alias_name *pmu; >> + >> +       list_for_each_entry(pmu, &pmu_alias_name_list, list) { >> +               if (!strcmp(name, pmu->name)) >> +                       return strdup(pmu->alias); >> +       } >> +       return NULL; >> +} >> diff --git a/tools/perf/util/parse-events.y b/tools/perf/util/parse-events.y >> index 9321bd0e2f76..d94e48e1ff9b 100644 >> --- a/tools/perf/util/parse-events.y >> +++ b/tools/perf/util/parse-events.y >> @@ -316,7 +316,8 @@ event_pmu_name opt_pmu_config >>                         if (!strncmp(name, "uncore_", 7) && >>                             strncmp($1, "uncore_", 7)) >>                                 name += 7; >> -                       if (!perf_pmu__match(pattern, name, $1)) { >> +                       if (!perf_pmu__match(pattern, name, $1) || >> +                           !perf_pmu__match(pattern, pmu->alias_name, $1)) { >>                                 if (parse_events_copy_term_list(orig_terms, >> &terms)) >>                                         CLEANUP_YYABORT; >>                                 if (!parse_events_add_pmu(_parse_state, list, >> pmu->name, terms, true, false)) >> diff --git a/tools/perf/util/pmu.c b/tools/perf/util/pmu.c >> index 44b90d638ad5..cc9af7942e7b 100644 >> --- a/tools/perf/util/pmu.c >> +++ b/tools/perf/util/pmu.c >> @@ -944,13 +944,28 @@ static int pmu_max_precise(const char *name) >>         return max_precise; >>  } >> >> -static struct perf_pmu *pmu_lookup(const char *name) >> +char * __weak >> +pmu_find_real_name(const char *name) >> +{ >> +       return strdup(name); >> +} >> + >> +char * __weak >> +pmu_find_alias_name(const char *name __maybe_unused) >> +{ >> +       return NULL; >> +} >> + >> +static struct perf_pmu *pmu_lookup(const char *lookup_name) >>  { >>         struct perf_pmu *pmu; >> +       char *name; >>         LIST_HEAD(format); >>         LIST_HEAD(aliases); >>         __u32 type; >> >> +       name = pmu_find_real_name(lookup_name); >> + >>         /* >>          * The pmu data we store & need consists of the pmu >>          * type value and format definitions. Load both right >> @@ -973,7 +988,8 @@ static struct perf_pmu *pmu_lookup(const char *name) >>                 return NULL; >> >>         pmu->cpus = pmu_cpumask(name); >> -       pmu->name = strdup(name); >> +       pmu->name = name; >> +       pmu->alias_name = pmu_find_alias_name(name); >>         pmu->type = type; >>         pmu->is_uncore = pmu_is_uncore(name); >>         if (pmu->is_uncore) >> @@ -1003,7 +1019,8 @@ static struct perf_pmu *pmu_find(const char *name) >>         struct perf_pmu *pmu; >> >>         list_for_each_entry(pmu, &pmus, list) >> -               if (!strcmp(pmu->name, name)) >> +               if (!strcmp(pmu->name, name) || >> +                   (pmu->alias_name && !strcmp(pmu->alias_name, name))) >>                         return pmu; >> >>         return NULL; >> @@ -1898,6 +1915,9 @@ bool perf_pmu__has_hybrid(void) >> >>  int perf_pmu__match(char *pattern, char *name, char *tok) >>  { >> +       if (!name) >> +               return -1; >> + >>         if (fnmatch(pattern, name, 0)) >>                 return -1; >> >> diff --git a/tools/perf/util/pmu.h b/tools/perf/util/pmu.h >> index 926da483a141..f6ca9f6a06ef 100644 >> --- a/tools/perf/util/pmu.h >> +++ b/tools/perf/util/pmu.h >> @@ -21,6 +21,7 @@ enum { >>  #define PERF_PMU_FORMAT_BITS 64 >>  #define EVENT_SOURCE_DEVICE_PATH "/bus/event_source/devices/" >>  #define CPUS_TEMPLATE_CPU      "%s/bus/event_source/devices/%s/cpus" >> +#define MAX_PMU_NAME_LEN 128 >> >>  struct perf_event_attr; >> >> @@ -32,6 +33,7 @@ struct perf_pmu_caps { >> >>  struct perf_pmu { >>         char *name; >> +       char *alias_name;       /* PMU alias name */ >>         char *id; >>         __u32 type; >>         bool selectable; >> @@ -135,4 +137,7 @@ void perf_pmu__warn_invalid_config(struct perf_pmu *pmu, >> __u64 config, >>  bool perf_pmu__has_hybrid(void); >>  int perf_pmu__match(char *pattern, char *name, char *tok); >> >> +char *pmu_find_real_name(const char *name); >> +char *pmu_find_alias_name(const char *name); >> + >>  #endif /* __PMU_H */ > >