From ded80bda8bc9bb65a344b79b36d5acf45a907b25 Mon Sep 17 00:00:00 2001 From: Ian Rogers Date: Fri, 15 May 2020 15:17:32 -0700 Subject: perf expr: Migrate expr ids table to a hashmap Use a hashmap between a char* string and a double* value. While bpf's hashmap entries are size_t in size, we can't guarantee sizeof(size_t) >= sizeof(double). Avoid a memory allocation when gathering ids by making 0.0 a special value encoded as NULL. Original map suggestion by Andi Kleen: https://lore.kernel.org/lkml/20200224210308.GQ160988@tassilo.jf.intel.com/ and seconded by Jiri Olsa: https://lore.kernel.org/lkml/20200423112915.GH1136647@krava/ Committer notes: There are fixes that need to land upstream before we can use libbpf's headers, for now use our copy unconditionally, since the data structures at this point are exactly the same, no problem. When the fixes for libbpf's hashmap land upstream, we can fix this up. Testing it: Building with LIBBPF=1, i.e. the default: $ perf -vv | grep -i bpf bpf: [ on ] # HAVE_LIBBPF_SUPPORT $ nm ~/bin/perf | grep -i libbpf_ | wc -l 39 $ nm ~/bin/perf | grep -i hashmap_ | wc -l 17 $ Explicitely building without LIBBPF: $ perf -vv | grep -i bpf bpf: [ OFF ] # HAVE_LIBBPF_SUPPORT $ $ nm ~/bin/perf | grep -i libbpf_ | wc -l 0 $ nm ~/bin/perf | grep -i hashmap_ | wc -l 9 $ Signed-off-by: Ian Rogers Tested-by: Arnaldo Carvalho de Melo Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Alexei Starovoitov Cc: Andi Kleen Cc: Andrii Nakryiko Cc: Cong Wang Cc: Daniel Borkmann Cc: Jin Yao Cc: Jiri Olsa Cc: John Fastabend Cc: John Garry Cc: Kajol Jain Cc: Kan Liang Cc: Kim Phillips Cc: Leo Yan Cc: Mark Rutland Cc: Martin KaFai Lau Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Song Liu Cc: Stephane Eranian Cc: Yonghong Song Cc: bpf@vger.kernel.org Cc: kp singh Cc: netdev@vger.kernel.org Link: http://lore.kernel.org/lkml/20200515221732.44078-8-irogers@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/stat-shadow.c | 49 ++++++++++++++++++++++++++++--------------- 1 file changed, 32 insertions(+), 17 deletions(-) (limited to 'tools/perf/util/stat-shadow.c') diff --git a/tools/perf/util/stat-shadow.c b/tools/perf/util/stat-shadow.c index 9bd7a8d2a858..c44dc814b377 100644 --- a/tools/perf/util/stat-shadow.c +++ b/tools/perf/util/stat-shadow.c @@ -323,35 +323,46 @@ void perf_stat__collect_metric_expr(struct evlist *evsel_list) { struct evsel *counter, *leader, **metric_events, *oc; bool found; - const char **metric_names; + struct expr_parse_ctx ctx; + struct hashmap_entry *cur; + size_t bkt; int i; - int num_metric_names; + expr__ctx_init(&ctx); evlist__for_each_entry(evsel_list, counter) { bool invalid = false; leader = counter->leader; if (!counter->metric_expr) continue; + + expr__ctx_clear(&ctx); metric_events = counter->metric_events; if (!metric_events) { - if (expr__find_other(counter->metric_expr, counter->name, - &metric_names, &num_metric_names, 1) < 0) + if (expr__find_other(counter->metric_expr, + counter->name, + &ctx, 1) < 0) continue; metric_events = calloc(sizeof(struct evsel *), - num_metric_names + 1); - if (!metric_events) + hashmap__size(&ctx.ids) + 1); + if (!metric_events) { + expr__ctx_clear(&ctx); return; + } counter->metric_events = metric_events; } - for (i = 0; i < num_metric_names; i++) { + i = 0; + hashmap__for_each_entry((&ctx.ids), cur, bkt) { + const char *metric_name = (const char *)cur->key; + found = false; if (leader) { /* Search in group */ for_each_group_member (oc, leader) { - if (!strcasecmp(oc->name, metric_names[i]) && + if (!strcasecmp(oc->name, + metric_name) && !oc->collect_stat) { found = true; break; @@ -360,7 +371,8 @@ void perf_stat__collect_metric_expr(struct evlist *evsel_list) } if (!found) { /* Search ignoring groups */ - oc = perf_stat__find_event(evsel_list, metric_names[i]); + oc = perf_stat__find_event(evsel_list, + metric_name); } if (!oc) { /* Deduping one is good enough to handle duplicated PMUs. */ @@ -373,27 +385,28 @@ void perf_stat__collect_metric_expr(struct evlist *evsel_list) * of events. So we ask the user instead to add the missing * events. */ - if (!printed || strcasecmp(printed, metric_names[i])) { + if (!printed || + strcasecmp(printed, metric_name)) { fprintf(stderr, "Add %s event to groups to get metric expression for %s\n", - metric_names[i], + metric_name, counter->name); - printed = strdup(metric_names[i]); + printed = strdup(metric_name); } invalid = true; continue; } - metric_events[i] = oc; + metric_events[i++] = oc; oc->collect_stat = true; } metric_events[i] = NULL; - free(metric_names); if (invalid) { free(metric_events); counter->metric_events = NULL; counter->metric_expr = NULL; } } + expr__ctx_clear(&ctx); } static double runtime_stat_avg(struct runtime_stat *st, @@ -738,7 +751,10 @@ static void generic_metric(struct perf_stat_config *config, expr__ctx_init(&pctx); /* Must be first id entry */ - expr__add_id(&pctx, name, avg); + n = strdup(name); + if (!n) + return; + expr__add_id(&pctx, n, avg); for (i = 0; metric_events[i]; i++) { struct saved_value *v; struct stats *stats; @@ -814,8 +830,7 @@ static void generic_metric(struct perf_stat_config *config, (metric_name ? metric_name : name) : "", 0); } - for (i = 1; i < pctx.num_ids; i++) - zfree(&pctx.ids[i].name); + expr__ctx_clear(&pctx); } void perf_stat__print_shadow_stats(struct perf_stat_config *config, -- cgit v1.2.3