linux-kernel - Re: [PATCH v2] perf s390-sample-raw: Cache counter names

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAP-5=fXvRK6PpYctLa4UXFbr5xzGDCAUkFuc4JWCEtd0s5QRaQ@mail.gmail.com>
Date: Fri, 31 Oct 2025 09:10:26 -0700
From: Ian Rogers <irogers@...gle.com>
To: Thomas Richter <tmricht@...ux.ibm.com>, Peter Zijlstra <peterz@...radead.org>, 
	Ingo Molnar <mingo@...hat.com>, Arnaldo Carvalho de Melo <acme@...nel.org>, Namhyung Kim <namhyung@...nel.org>, 
	Alexander Shishkin <alexander.shishkin@...ux.intel.com>, Jiri Olsa <jolsa@...nel.org>, 
	Ian Rogers <irogers@...gle.com>, Adrian Hunter <adrian.hunter@...el.com>, 
	linux-perf-users@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] perf s390-sample-raw: Cache counter names

On Thu, Oct 30, 2025 at 10:21 AM Ian Rogers <irogers@...gle.com> wrote:
>
> Searching all event names is slower now that legacy names are
> included. Add a cache to avoid long iterative searches.
>
> Reported-by: Thomas Richter <tmricht@...ux.ibm.com>
> Closes: https://lore.kernel.org/linux-perf-users/09943f4f-516c-4b93-877c-e4a64ed61d38@linux.ibm.com/
> Signed-off-by: Ian Rogers <irogers@...gle.com>
> --
> v2: Small tweak to the cache_key, just make it match the wanted event value.

In this thread Thomas added his Tested-by tag:
https://lore.kernel.org/linux-perf-users/0d4b605b-64ab-48a0-bd95-8465dabbbae3@linux.ibm.com/
Tested-by: Thomas Richter <tmricht@...ux.ibm.com>

Thanks!
Ian

> ---
>  tools/perf/util/s390-sample-raw.c | 56 ++++++++++++++++++++++++++++---
>  1 file changed, 51 insertions(+), 5 deletions(-)
>
> diff --git a/tools/perf/util/s390-sample-raw.c b/tools/perf/util/s390-sample-raw.c
> index 335217bb532b..f5acf6dfa8d0 100644
> --- a/tools/perf/util/s390-sample-raw.c
> +++ b/tools/perf/util/s390-sample-raw.c
> @@ -19,12 +19,14 @@
>
>  #include <sys/stat.h>
>  #include <linux/compiler.h>
> +#include <linux/err.h>
>  #include <asm/byteorder.h>
>
>  #include "debug.h"
>  #include "session.h"
>  #include "evlist.h"
>  #include "color.h"
> +#include "hashmap.h"
>  #include "sample-raw.h"
>  #include "s390-cpumcf-kernel.h"
>  #include "util/pmu.h"
> @@ -132,8 +134,8 @@ static int get_counterset_start(int setnr)
>  }
>
>  struct get_counter_name_data {
> -       int wanted;
> -       char *result;
> +       long wanted;
> +       const char *result;
>  };
>
>  static int get_counter_name_callback(void *vdata, struct pmu_event_info *info)
> @@ -151,12 +153,22 @@ static int get_counter_name_callback(void *vdata, struct pmu_event_info *info)
>
>         rc = sscanf(event_str, "event=%x", &event_nr);
>         if (rc == 1 && event_nr == data->wanted) {
> -               data->result = strdup(info->name);
> +               data->result = info->name;
>                 return 1; /* Terminate the search. */
>         }
>         return 0;
>  }
>
> +static size_t get_counter_name_hash_fn(long key, void *ctx __maybe_unused)
> +{
> +       return key;
> +}
> +
> +static bool get_counter_name_hashmap_equal_fn(long key1, long key2, void *ctx __maybe_unused)
> +{
> +       return key1 == key2;
> +}
> +
>  /* Scan the PMU and extract the logical name of a counter from the event. Input
>   * is the counter set and counter number with in the set. Construct the event
>   * number and use this as key. If they match return the name of this counter.
> @@ -164,17 +176,51 @@ static int get_counter_name_callback(void *vdata, struct pmu_event_info *info)
>   */
>  static char *get_counter_name(int set, int nr, struct perf_pmu *pmu)
>  {
> +       static struct hashmap *cache;
> +       static struct perf_pmu *cache_pmu;
> +       long cache_key = get_counterset_start(set) + nr;
>         struct get_counter_name_data data = {
> -               .wanted = get_counterset_start(set) + nr,
> +               .wanted = cache_key,
>                 .result = NULL,
>         };
> +       char *result = NULL;
>
>         if (!pmu)
>                 return NULL;
>
> +       if (cache_pmu == pmu && hashmap__find(cache, cache_key, &result))
> +               return strdup(result);
> +
>         perf_pmu__for_each_event(pmu, /*skip_duplicate_pmus=*/ true,
>                                  &data, get_counter_name_callback);
> -       return data.result;
> +
> +       if (data.result)
> +               result = strdup(data.result);
> +
> +       if (cache_pmu == NULL) {
> +               struct hashmap *tmp = hashmap__new(get_counter_name_hash_fn,
> +                                                  get_counter_name_hashmap_equal_fn,
> +                                                  /*ctx=*/NULL);
> +
> +               if (!IS_ERR(cache)) {
> +                       cache = tmp;
> +                       cache_pmu = pmu;
> +               }
> +       }
> +
> +       if (cache_pmu == pmu) {
> +               char *old_value = NULL, *new_value = strdup(result);
> +
> +               if (new_value) {
> +                       hashmap__set(cache, cache_key, new_value, /*old_key=*/NULL, &old_value);
> +                        /*
> +                         * Free in case of a race, but resizing would be broken
> +                       f  * in that case.

Oh, there's a typo here. I'll send v3 with just this comment fix.

Thanks,
Ian

> +                         */
> +                       free(old_value);
> +               }
> +       }
> +       return result;
>  }
>
>  static void s390_cpumcfdg_dump(struct perf_pmu *pmu, struct perf_sample *sample)
> --
> 2.51.1.930.gacf6e81ea2-goog
>