Re: [PATCH v7 2/5] perf cpumap: Add reference count checking

From: Ian Rogers
Date: Wed Apr 12 2023 - 12:08:13 EST


On Wed, Apr 12, 2023 at 8:50 AM Arnaldo Carvalho de Melo
<acme@xxxxxxxxxx> wrote:
>
> Em Tue, Apr 11, 2023 at 03:19:06PM -0300, Arnaldo Carvalho de Melo escreveu:
> > Em Fri, Apr 07, 2023 at 04:04:02PM -0700, Ian Rogers escreveu:
> > > This change is intended to catch:
> > > - use after put: using a cpumap after you have put it will cause a
> > > segv.
> > > - unbalanced puts: two puts for a get will result in a double free
> > > that can be captured and reported by tools like address sanitizer,
> > > including with the associated stack traces of allocation and frees.
> > > - missing puts: if a put is missing then the get turns into a memory
> > > leak that can be reported by leak sanitizer, including the stack
> > > trace at the point the get occurs.
>
> > I think this should be further split into self contained patches as it
> > does:
>
> > These are missing convertions that should be in a separate patch, no?
> >
> > > @@ -239,7 +236,7 @@ struct cpu_aggr_map *cpu_aggr_map__new(const struct perf_cpu_map *cpus,
> > > {
> > > int idx;
> > > struct perf_cpu cpu;
> > > - struct cpu_aggr_map *c = cpu_aggr_map__empty_new(cpus->nr);
> > > + struct cpu_aggr_map *c = cpu_aggr_map__empty_new(perf_cpu_map__nr(cpus));
> > >
> > > if (!c)
> > > return NULL;
>
> Extracted this from your larger patch:
>
> From f8a23fce48400168be0cc078a0b0bd0e7d4a889f Mon Sep 17 00:00:00 2001
> From: Ian Rogers <irogers@xxxxxxxxxx>
> Date: Wed, 12 Apr 2023 12:45:45 -0300
> Subject: [PATCH 1/1] perf cpumap: Use perf_cpu_map__nr(cpus) to access
> cpus->nr
>
> So that we can have a single point where to refcount check 'struct perf_cpu_map'
> instances for use after free, etc.

Thanks! In one of your previous messages you mentioned adding some
setters for cpu_map to bury the macro usage. As cpu maps are
immutable, can we make these internal APIs? Other than that, thanks
for helping to get this landed!

Ian

> Signed-off-by: Ian Rogers <irogers@xxxxxxxxxx>
> Cc: Adrian Hunter <adrian.hunter@xxxxxxxxx>
> Cc: Alexey Bayduraev <alexey.v.bayduraev@xxxxxxxxxxxxxxx>
> Cc: Dmitriy Vyukov <dvyukov@xxxxxxxxxx>
> Cc: Jiri Olsa <jolsa@xxxxxxxxxx>
> Cc: Namhyung Kim <namhyung@xxxxxxxxxx>
> Cc: Riccardo Mancini <rickyman7@xxxxxxxxx>
> Cc: Stephane Eranian <eranian@xxxxxxxxxx>
> Cc: Stephen Brennan <stephen.s.brennan@xxxxxxxxxx>
> Link: https://lore.kernel.org/lkml/
> Signed-off-by: Arnaldo Carvalho de Melo <acme@xxxxxxxxxx>
> ---
> tools/perf/util/cpumap.c | 12 ++++++------
> 1 file changed, 6 insertions(+), 6 deletions(-)
>
> diff --git a/tools/perf/util/cpumap.c b/tools/perf/util/cpumap.c
> index 5e564974fba4ffab..c8484b75413ef709 100644
> --- a/tools/perf/util/cpumap.c
> +++ b/tools/perf/util/cpumap.c
> @@ -239,7 +239,7 @@ struct cpu_aggr_map *cpu_aggr_map__new(const struct perf_cpu_map *cpus,
> {
> int idx;
> struct perf_cpu cpu;
> - struct cpu_aggr_map *c = cpu_aggr_map__empty_new(cpus->nr);
> + struct cpu_aggr_map *c = cpu_aggr_map__empty_new(perf_cpu_map__nr(cpus));
>
> if (!c)
> return NULL;
> @@ -263,7 +263,7 @@ struct cpu_aggr_map *cpu_aggr_map__new(const struct perf_cpu_map *cpus,
> }
> }
> /* Trim. */
> - if (c->nr != cpus->nr) {
> + if (c->nr != perf_cpu_map__nr(cpus)) {
> struct cpu_aggr_map *trimmed_c =
> realloc(c,
> sizeof(struct cpu_aggr_map) + sizeof(struct aggr_cpu_id) * c->nr);
> @@ -582,9 +582,9 @@ size_t cpu_map__snprint(struct perf_cpu_map *map, char *buf, size_t size)
>
> #define COMMA first ? "" : ","
>
> - for (i = 0; i < map->nr + 1; i++) {
> + for (i = 0; i < perf_cpu_map__nr(map) + 1; i++) {
> struct perf_cpu cpu = { .cpu = INT_MAX };
> - bool last = i == map->nr;
> + bool last = i == perf_cpu_map__nr(map);
>
> if (!last)
> cpu = map->map[i];
> @@ -633,7 +633,7 @@ size_t cpu_map__snprint_mask(struct perf_cpu_map *map, char *buf, size_t size)
> int i, cpu;
> char *ptr = buf;
> unsigned char *bitmap;
> - struct perf_cpu last_cpu = perf_cpu_map__cpu(map, map->nr - 1);
> + struct perf_cpu last_cpu = perf_cpu_map__cpu(map, perf_cpu_map__nr(map) - 1);
>
> if (buf == NULL)
> return 0;
> @@ -644,7 +644,7 @@ size_t cpu_map__snprint_mask(struct perf_cpu_map *map, char *buf, size_t size)
> return 0;
> }
>
> - for (i = 0; i < map->nr; i++) {
> + for (i = 0; i < perf_cpu_map__nr(map); i++) {
> cpu = perf_cpu_map__cpu(map, i).cpu;
> bitmap[cpu / 8] |= 1 << (cpu % 8);
> }
> --
> 2.39.2
>