Re: [PATCH v4 1/6] perf cpumap: Switch to using perf_cpu_map API

From: Namhyung Kim <namhyung@kernel.org>
To: Ian Rogers <irogers@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>,
	Ingo Molnar <mingo@redhat.com>,
	Arnaldo Carvalho de Melo <acme@kernel.org>,
	Mark Rutland <mark.rutland@arm.com>,
	Alexander Shishkin <alexander.shishkin@linux.intel.com>,
	Jiri Olsa <jolsa@kernel.org>,
	Mathieu Poirier <mathieu.poirier@linaro.org>,
	Suzuki K Poulose <suzuki.poulose@arm.com>,
	Mike Leach <mike.leach@linaro.org>, Leo Yan <leo.yan@linaro.org>,
	John Garry <john.garry@huawei.com>, Will Deacon <will@kernel.org>,
	Alexei Starovoitov <ast@kernel.org>,
	Daniel Borkmann <daniel@iogearbox.net>,
	Andrii Nakryiko <andrii@kernel.org>,
	Martin KaFai Lau <kafai@fb.com>, Song Liu <songliubraving@fb.com>,
	Yonghong Song <yhs@fb.com>,
	John Fastabend <john.fastabend@gmail.com>,
	KP Singh <kpsingh@kernel.org>, Kajol Jain <kjain@linux.ibm.com>,
	James Clark <james.clark@arm.com>,
	German Gomez <german.gomez@arm.com>,
	Adrian Hunter <adrian.hunter@intel.com>,
	Riccardo Mancini <rickyman7@gmail.com>,
	Andi Kleen <ak@linux.intel.com>,
	Alexey Bayduraev <alexey.v.bayduraev@linux.intel.com>,
	Alexander Antonov <alexander.antonov@linux.intel.com>,
	linux-perf-users <linux-perf-users@vger.kernel.org>,
	linux-kernel <linux-kernel@vger.kernel.org>,
	Stephane Eranian <eranian@google.com>
Subject: Re: [PATCH v4 1/6] perf cpumap: Switch to using perf_cpu_map API
Date: Mon, 2 May 2022 16:51:30 -0700	[thread overview]
Message-ID: <CAM9d7cir7=okKG0gNDkCKQQYNWRNA2aUz9+o3vtJFK+xKNjYew@mail.gmail.com> (raw)
In-Reply-To: <20220430062324.1565215-2-irogers@google.com>

Hi Ian,

On Fri, Apr 29, 2022 at 11:23 PM Ian Rogers <irogers@google.com> wrote:
>
> Switch some raw accesses to the cpu map to using the library API. This
> can help with reference count checking.
>
> Signed-off-by: Ian Rogers <irogers@google.com>
> ---
[...]
> diff --git a/tools/perf/util/bpf_counter_cgroup.c b/tools/perf/util/bpf_counter_cgroup.c
> index ac60c08e8e2a..a4b676920da0 100644
> --- a/tools/perf/util/bpf_counter_cgroup.c
> +++ b/tools/perf/util/bpf_counter_cgroup.c
> @@ -47,7 +47,7 @@ static int bperf_load_program(struct evlist *evlist)
>         struct evsel *evsel;
>         struct cgroup *cgrp, *leader_cgrp;
>         __u32 i, cpu;
> -       __u32 nr_cpus = evlist->core.all_cpus->nr;
> +       __u32 nr_cpus = perf_cpu_map__nr(evlist->core.all_cpus);
>         int total_cpus = cpu__max_cpu().cpu;
>         int map_size, map_fd;
>         int prog_fd, err;
> @@ -125,7 +125,7 @@ static int bperf_load_program(struct evlist *evlist)
>                         for (cpu = 0; cpu < nr_cpus; cpu++) {
>                                 int fd = FD(evsel, cpu);
>                                 __u32 idx = evsel->core.idx * total_cpus +
> -                                       evlist->core.all_cpus->map[cpu].cpu;
> +                                       perf_cpu_map__cpu(evlist->core.all_cpus, cpu).cpu;
>
>                                 err = bpf_map_update_elem(map_fd, &idx, &fd,
>                                                           BPF_ANY);
> @@ -207,13 +207,13 @@ static int bperf_cgrp__install_pe(struct evsel *evsel __maybe_unused,
>   */
>  static int bperf_cgrp__sync_counters(struct evlist *evlist)
>  {
> -       int i, cpu;
> -       int nr_cpus = evlist->core.all_cpus->nr;
> +       struct perf_cpu cpu;
> +       int idx;
>         int prog_fd = bpf_program__fd(skel->progs.trigger_read);
>
> -       for (i = 0; i < nr_cpus; i++) {
> -               cpu = evlist->core.all_cpus->map[i].cpu;
> -               bperf_trigger_reading(prog_fd, cpu);
> +       perf_cpu_map__for_each_cpu(cpu, idx, cpus) {

s/cpus/evlist->core.all_cpus/?

> +               cpu = perf_cpu_map__cpu(evlist->core.all_cpus, i);

I don't think we need this line anymore.

> +               bperf_trigger_reading(prog_fd, cpu.cpu);
>         }
>
>         return 0;
> @@ -244,12 +244,10 @@ static int bperf_cgrp__disable(struct evsel *evsel)
>  static int bperf_cgrp__read(struct evsel *evsel)
>  {
>         struct evlist *evlist = evsel->evlist;
> -       int i, cpu, nr_cpus = evlist->core.all_cpus->nr;
>         int total_cpus = cpu__max_cpu().cpu;
>         struct perf_counts_values *counts;
>         struct bpf_perf_event_value *values;
>         int reading_map_fd, err = 0;
> -       __u32 idx;
>
>         if (evsel->core.idx)
>                 return 0;
> @@ -263,7 +261,10 @@ static int bperf_cgrp__read(struct evsel *evsel)
>         reading_map_fd = bpf_map__fd(skel->maps.cgrp_readings);
>
>         evlist__for_each_entry(evlist, evsel) {
> -               idx = evsel->core.idx;
> +               __u32 idx = evsel->core.idx;
> +               int i;
> +               struct perf_cpu_map cpu;
> +
>                 err = bpf_map_lookup_elem(reading_map_fd, &idx, values);
>                 if (err) {
>                         pr_err("bpf map lookup failed: idx=%u, event=%s, cgrp=%s\n",
> @@ -271,13 +272,11 @@ static int bperf_cgrp__read(struct evsel *evsel)
>                         goto out;
>                 }
>
> -               for (i = 0; i < nr_cpus; i++) {
> -                       cpu = evlist->core.all_cpus->map[i].cpu;
> -
> +               perf_cpu_map__for_each_cpu(cpu, i, evlist->core.all_cpu) {

evlist->core.all_cpus ?

Thanks,
Namhyung

>                         counts = perf_counts(evsel->counts, i, 0);
> -                       counts->val = values[cpu].counter;
> -                       counts->ena = values[cpu].enabled;
> -                       counts->run = values[cpu].running;
> +                       counts->val = values[cpu.cpu].counter;
> +                       counts->ena = values[cpu.cpu].enabled;
> +                       counts->run = values[cpu.cpu].running;
>                 }
>         }
>
> --
> 2.36.0.464.gb9c8b46e94-goog
>