public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Arnaldo Carvalho de Melo <acme@kernel.org>
To: kan.liang@linux.intel.com
Cc: mingo@redhat.com, peterz@infradead.org,
	linux-kernel@vger.kernel.org, jolsa@redhat.com,
	namhyung@kernel.org, ganapatrao.kulkarni@cavium.com,
	zhangshaokun@hisilicon.com, yao.jin@linux.intel.com,
	will.deacon@arm.com, ak@linux.intel.com, agustinv@codeaurora.org
Subject: Re: [PATCH 4/5] perf parse-events: Specially handle uncore event alias in small groups
Date: Tue, 24 Apr 2018 16:17:21 -0300	[thread overview]
Message-ID: <20180424191721.GB4427@kernel.org> (raw)
In-Reply-To: <1524594014-79243-4-git-send-email-kan.liang@linux.intel.com>

Em Tue, Apr 24, 2018 at 11:20:13AM -0700, kan.liang@linux.intel.com escreveu:
> From: Kan Liang <kan.liang@linux.intel.com>
> 
> Perf stat doesn't count the uncore event aliases from the same uncore
> block in a group, for example:

This one is not applying to acme/perf/urgent, all the rest I got merged
there, the last one with that change from using strstr() to a new bool
in perf_evsel for the uniquifying operation having being performed.

- Arnaldo
 
>   perf stat -e '{unc_m_cas_count.all,unc_m_clockticks}' -a -I 1000
>   #           time             counts unit events
>        1.000447342      <not counted>      unc_m_cas_count.all
>        1.000447342      <not counted>      unc_m_clockticks
>        2.000740654      <not counted>      unc_m_cas_count.all
>        2.000740654      <not counted>      unc_m_clockticks
> 
> The output is very misleading. It gives a wrong impression that the
> uncore event doesn't work.
> 
> An uncore block could be composed by several PMUs. An uncore event alias
> is a joint name which means the same event runs on all PMUs of a block.
> Perf doesn't support mixed events from different PMUs in the same group.
> It is wrong to put uncore event aliases in a big group.
> 
> The right way is to split the big group into multiple small groups which
> only include the events from the same PMU.
> Only uncore event aliases from the same uncore block should be specially
> handled here. It doesn't make sense to mix the uncore events with other
> uncore events from different blocks or even core events in a group.
> 
> With the patch:
>   #           time             counts unit events
>      1.001557653            140,833      unc_m_cas_count.all
>      1.001557653      1,330,231,332      unc_m_clockticks
>      2.002709483             85,007      unc_m_cas_count.all
>      2.002709483      1,429,494,563      unc_m_clockticks
> 
> Reported-by: Andi Kleen <ak@linux.intel.com>
> Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
> ---
>  tools/perf/util/evsel.h        |  1 +
>  tools/perf/util/parse-events.c | 94 ++++++++++++++++++++++++++++++++++++++++--
>  tools/perf/util/parse-events.h |  7 +++-
>  tools/perf/util/parse-events.y |  8 ++--
>  4 files changed, 101 insertions(+), 9 deletions(-)
> 
> diff --git a/tools/perf/util/evsel.h b/tools/perf/util/evsel.h
> index d3ee3af..b8a1055 100644
> --- a/tools/perf/util/evsel.h
> +++ b/tools/perf/util/evsel.h
> @@ -126,6 +126,7 @@ struct perf_evsel {
>  	bool			precise_max;
>  	bool			ignore_missing_thread;
>  	bool			forced_leader;
> +	bool			use_uncore_alias;
>  	/* parse modifier helper */
>  	int			exclude_GH;
>  	int			nr_members;
> diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
> index 2fb0272..7a836fe 100644
> --- a/tools/perf/util/parse-events.c
> +++ b/tools/perf/util/parse-events.c
> @@ -1219,13 +1219,16 @@ int parse_events_add_numeric(struct parse_events_state *parse_state,
>  
>  int parse_events_add_pmu(struct parse_events_state *parse_state,
>  			 struct list_head *list, char *name,
> -			 struct list_head *head_config, bool auto_merge_stats)
> +			 struct list_head *head_config,
> +			 bool auto_merge_stats,
> +			 bool use_alias)
>  {
>  	struct perf_event_attr attr;
>  	struct perf_pmu_info info;
>  	struct perf_pmu *pmu;
>  	struct perf_evsel *evsel;
>  	struct parse_events_error *err = parse_state->error;
> +	bool use_uncore_alias;
>  	LIST_HEAD(config_terms);
>  
>  	pmu = perf_pmu__find(name);
> @@ -1244,11 +1247,14 @@ int parse_events_add_pmu(struct parse_events_state *parse_state,
>  		memset(&attr, 0, sizeof(attr));
>  	}
>  
> +	use_uncore_alias = (pmu->is_uncore && use_alias);
> +
>  	if (!head_config) {
>  		attr.type = pmu->type;
>  		evsel = __add_event(list, &parse_state->idx, &attr, NULL, pmu, NULL, auto_merge_stats);
>  		if (evsel) {
>  			evsel->pmu_name = name;
> +			evsel->use_uncore_alias = use_uncore_alias;
>  			return 0;
>  		} else {
>  			return -ENOMEM;
> @@ -1282,6 +1288,7 @@ int parse_events_add_pmu(struct parse_events_state *parse_state,
>  		evsel->metric_expr = info.metric_expr;
>  		evsel->metric_name = info.metric_name;
>  		evsel->pmu_name = name;
> +		evsel->use_uncore_alias = use_uncore_alias;
>  	}
>  
>  	return evsel ? 0 : -ENOMEM;
> @@ -1317,7 +1324,7 @@ int parse_events_multi_pmu_add(struct parse_events_state *parse_state,
>  				list_add_tail(&term->list, head);
>  
>  				if (!parse_events_add_pmu(parse_state, list,
> -							  pmu->name, head, true)) {
> +							  pmu->name, head, true, true)) {
>  					pr_debug("%s -> %s/%s/\n", str,
>  						 pmu->name, alias->str);
>  					ok++;
> @@ -1339,7 +1346,85 @@ int parse_events__modifier_group(struct list_head *list,
>  	return parse_events__modifier_event(list, event_mod, true);
>  }
>  
> -void parse_events__set_leader(char *name, struct list_head *list)
> +/*
> + * Check if the two uncore PMUs are from the same uncore block
> + * The format of the uncore PMU name is uncore_#blockname_#pmuidx
> + */
> +static bool is_same_uncore_block(const char *pmu_name_a, const char *pmu_name_b)
> +{
> +	char *end_a, *end_b;
> +
> +	end_a = strrchr(pmu_name_a, '_');
> +	end_b = strrchr(pmu_name_b, '_');
> +
> +	if (!end_a || !end_b)
> +		return false;
> +
> +	if ((end_a - pmu_name_a) != (end_b - pmu_name_b))
> +		return false;
> +
> +	return (0 == strncmp(pmu_name_a, pmu_name_b, end_a - pmu_name_a));
> +}
> +
> +static int
> +parse_events__set_leader_for_uncore_aliase(char *name, struct list_head *list,
> +					   struct parse_events_state *parse_state)
> +{
> +	struct perf_evsel *evsel, *leader;
> +	uintptr_t *leaders;
> +	int i = 0, nr_pmu = 0, total_members, ret = 0;
> +
> +	leader = list_entry(list->next, struct perf_evsel, node);
> +	evsel = list_entry(list->prev, struct perf_evsel, node);
> +	total_members = evsel->idx - leader->idx + 1;
> +
> +	leaders = calloc(total_members, sizeof(uintptr_t));
> +	if (!leaders)
> +		return ret;
> +
> +	 __evlist__for_each_entry(list, evsel) {
> +
> +		/* Only split the uncore group which members use alias */
> +		if (!evsel->use_uncore_alias)
> +			goto out;
> +
> +		/* The events must be from the same uncore block */
> +		if (!is_same_uncore_block(leader->pmu_name, evsel->pmu_name))
> +			goto out;
> +
> +		if (!strcmp(leader->name, evsel->name))
> +			leaders[nr_pmu++] = (uintptr_t) evsel;
> +	}
> +
> +	/* only one event alias */
> +	if (nr_pmu == total_members) {
> +		parse_state->nr_groups--;
> +		goto handled;
> +	}
> +
> +	__evlist__for_each_entry(list, evsel) {
> +		if (i >= nr_pmu)
> +			i = 0;
> +		evsel->leader = (struct perf_evsel *) leaders[i++];
> +	}
> +
> +	for (i = 0; i < nr_pmu; i++) {
> +		evsel = (struct perf_evsel *) leaders[i];
> +		evsel->nr_members = total_members / nr_pmu;
> +		evsel->group_name = name ? strdup(name) : NULL;
> +	}
> +
> +	parse_state->nr_groups += nr_pmu - 1;
> +
> +handled:
> +	ret = 1;
> +out:
> +	free(leaders);
> +	return ret;
> +}
> +
> +void parse_events__set_leader(char *name, struct list_head *list,
> +			      struct parse_events_state *parse_state)
>  {
>  	struct perf_evsel *leader;
>  
> @@ -1348,6 +1433,9 @@ void parse_events__set_leader(char *name, struct list_head *list)
>  		return;
>  	}
>  
> +	if (parse_events__set_leader_for_uncore_aliase(name, list, parse_state))
> +		return;
> +
>  	__perf_evlist__set_leader(list);
>  	leader = list_entry(list->next, struct perf_evsel, node);
>  	leader->group_name = name ? strdup(name) : NULL;
> diff --git a/tools/perf/util/parse-events.h b/tools/perf/util/parse-events.h
> index 5015cfd..4473dac 100644
> --- a/tools/perf/util/parse-events.h
> +++ b/tools/perf/util/parse-events.h
> @@ -167,7 +167,9 @@ int parse_events_add_breakpoint(struct list_head *list, int *idx,
>  				void *ptr, char *type, u64 len);
>  int parse_events_add_pmu(struct parse_events_state *parse_state,
>  			 struct list_head *list, char *name,
> -			 struct list_head *head_config, bool auto_merge_stats);
> +			 struct list_head *head_config,
> +			 bool auto_merge_stats,
> +			 bool use_alias);
>  
>  int parse_events_multi_pmu_add(struct parse_events_state *parse_state,
>  			       char *str,
> @@ -178,7 +180,8 @@ int parse_events_copy_term_list(struct list_head *old,
>  
>  enum perf_pmu_event_symbol_type
>  perf_pmu__parse_check(const char *name);
> -void parse_events__set_leader(char *name, struct list_head *list);
> +void parse_events__set_leader(char *name, struct list_head *list,
> +			      struct parse_events_state *parse_state);
>  void parse_events_update_lists(struct list_head *list_event,
>  			       struct list_head *list_all);
>  void parse_events_evlist_error(struct parse_events_state *parse_state,
> diff --git a/tools/perf/util/parse-events.y b/tools/perf/util/parse-events.y
> index 7afeb80..e37608a 100644
> --- a/tools/perf/util/parse-events.y
> +++ b/tools/perf/util/parse-events.y
> @@ -161,7 +161,7 @@ PE_NAME '{' events '}'
>  	struct list_head *list = $3;
>  
>  	inc_group_count(list, _parse_state);
> -	parse_events__set_leader($1, list);
> +	parse_events__set_leader($1, list, _parse_state);
>  	$$ = list;
>  }
>  |
> @@ -170,7 +170,7 @@ PE_NAME '{' events '}'
>  	struct list_head *list = $2;
>  
>  	inc_group_count(list, _parse_state);
> -	parse_events__set_leader(NULL, list);
> +	parse_events__set_leader(NULL, list, _parse_state);
>  	$$ = list;
>  }
>  
> @@ -232,7 +232,7 @@ PE_NAME opt_event_config
>  		YYABORT;
>  
>  	ALLOC_LIST(list);
> -	if (parse_events_add_pmu(_parse_state, list, $1, $2, false)) {
> +	if (parse_events_add_pmu(_parse_state, list, $1, $2, false, false)) {
>  		struct perf_pmu *pmu = NULL;
>  		int ok = 0;
>  		char *pattern;
> @@ -251,7 +251,7 @@ PE_NAME opt_event_config
>  					free(pattern);
>  					YYABORT;
>  				}
> -				if (!parse_events_add_pmu(_parse_state, list, pmu->name, terms, true))
> +				if (!parse_events_add_pmu(_parse_state, list, pmu->name, terms, true, false))
>  					ok++;
>  				parse_events_terms__delete(terms);
>  			}
> -- 
> 2.7.4

  reply	other threads:[~2018-04-24 19:17 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-04-24 18:20 [PATCH 1/5] perf pmu: Fix core PMU alias list for X86 platform kan.liang
2018-04-24 18:20 ` [PATCH 2/5] perf stat: Print out hint for mixed PMU group error kan.liang
2018-04-26  5:56   ` [tip:perf/urgent] " tip-bot for Kan Liang
2018-04-24 18:20 ` [PATCH 3/5] perf evsel: Only fall back group read for leader kan.liang
2018-04-26  5:57   ` [tip:perf/urgent] " tip-bot for Kan Liang
2018-04-24 18:20 ` [PATCH 4/5] perf parse-events: Specially handle uncore event alias in small groups kan.liang
2018-04-24 19:17   ` Arnaldo Carvalho de Melo [this message]
2018-04-24 19:23     ` Liang, Kan
2018-04-24 19:29       ` Arnaldo Carvalho de Melo
2018-04-25 12:27         ` Liang, Kan
2018-04-25 12:59           ` Arnaldo Carvalho de Melo
2018-04-25 13:39             ` Liang, Kan
2018-04-24 18:20 ` [PATCH 5/5] perf stat: Fix duplicate PMU name for interval print kan.liang
2018-04-24 18:53   ` Arnaldo Carvalho de Melo
2018-04-24 19:18     ` Liang, Kan
2018-04-24 19:29       ` Arnaldo Carvalho de Melo
2018-04-26  5:57   ` [tip:perf/urgent] " tip-bot for Kan Liang
2018-04-26  5:56 ` [tip:perf/urgent] perf pmu: Fix core PMU alias list for X86 platform tip-bot for Kan Liang

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20180424191721.GB4427@kernel.org \
    --to=acme@kernel.org \
    --cc=agustinv@codeaurora.org \
    --cc=ak@linux.intel.com \
    --cc=ganapatrao.kulkarni@cavium.com \
    --cc=jolsa@redhat.com \
    --cc=kan.liang@linux.intel.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=namhyung@kernel.org \
    --cc=peterz@infradead.org \
    --cc=will.deacon@arm.com \
    --cc=yao.jin@linux.intel.com \
    --cc=zhangshaokun@hisilicon.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox