[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAP-5=fWz9Put0QznfXSf1ATXjcfQj1T2toZmXrXp502qU4fRSw@mail.gmail.com>
Date: Mon, 15 Sep 2025 08:00:53 -0700
From: Ian Rogers <irogers@...gle.com>
To: Peter Zijlstra <peterz@...radead.org>, Ingo Molnar <mingo@...hat.com>,
Arnaldo Carvalho de Melo <acme@...nel.org>, Namhyung Kim <namhyung@...nel.org>,
Mark Rutland <mark.rutland@....com>,
Alexander Shishkin <alexander.shishkin@...ux.intel.com>, Jiri Olsa <jolsa@...nel.org>,
Ian Rogers <irogers@...gle.com>, Adrian Hunter <adrian.hunter@...el.com>,
"Liang, Kan" <kan.liang@...ux.intel.com>, Chun-Tse Shao <ctshao@...gle.com>,
Christophe Leroy <christophe.leroy@...roup.eu>, linux-perf-users@...r.kernel.org,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH v1] perf maps: Ensure kmap is set up for all inserts
On Sun, Sep 14, 2025 at 11:18 AM Ian Rogers <irogers@...gle.com> wrote:
>
> __maps__fixup_overlap_and_insert may split or directly insert a map,
> when doing this the map may need to have a kmap set up for the sake of
> the kmaps. The missing kmap set up fails the check_invariants test in
> maps, later "Internal error" reports from map__kmap and ultimately
> causes segfaults.
>
> Similar fixes were added in commit e0e4e0b8b7fa ("perf maps: Add
> missing map__set_kmap_maps() when replacing a kernel map") and commit
> 25d9c0301d36 ("perf maps: Set the kmaps for newly created/added kernel
> maps") but they missed cases. To try to reduce the risk of this,
> update the kmap directly following any manual insert. This identified
> another problem in maps__copy_from.
>
> Fixes: e0e4e0b8b7fa ("perf maps: Add missing map__set_kmap_maps() when replacing a kernel map")
> Fixes: 25d9c0301d36 ("perf maps: Set the kmaps for newly created/added kernel maps")
> Signed-off-by: Ian Rogers <irogers@...gle.com>
This may be worth picking up for v6.17. It was a fairly regular source
of crashes on basic commands like `perf record` for me on certain
kernels. `perf record` by default is processing the whole perf.data
file and if there were overlapping maps, from BPF symbol events
typically, the broken kmap would trigger assertion failures and seg
faults.
Thanks,
Ian
> ---
> tools/perf/util/maps.c | 9 +++++----
> 1 file changed, 5 insertions(+), 4 deletions(-)
>
> diff --git a/tools/perf/util/maps.c b/tools/perf/util/maps.c
> index 85b2a93a59ac..779f6230130a 100644
> --- a/tools/perf/util/maps.c
> +++ b/tools/perf/util/maps.c
> @@ -477,6 +477,7 @@ static int __maps__insert(struct maps *maps, struct map *new)
> }
> /* Insert the value at the end. */
> maps_by_address[nr_maps] = map__get(new);
> + map__set_kmap_maps(new, maps);
> if (maps_by_name)
> maps_by_name[nr_maps] = map__get(new);
>
> @@ -502,8 +503,6 @@ static int __maps__insert(struct maps *maps, struct map *new)
> if (map__end(new) < map__start(new))
> RC_CHK_ACCESS(maps)->ends_broken = true;
>
> - map__set_kmap_maps(new, maps);
> -
> return 0;
> }
>
> @@ -891,6 +890,7 @@ static int __maps__fixup_overlap_and_insert(struct maps *maps, struct map *new)
> if (before) {
> map__put(maps_by_address[i]);
> maps_by_address[i] = before;
> + map__set_kmap_maps(before, maps);
>
> if (maps_by_name) {
> map__put(maps_by_name[ni]);
> @@ -918,6 +918,7 @@ static int __maps__fixup_overlap_and_insert(struct maps *maps, struct map *new)
> */
> map__put(maps_by_address[i]);
> maps_by_address[i] = map__get(new);
> + map__set_kmap_maps(new, maps);
>
> if (maps_by_name) {
> map__put(maps_by_name[ni]);
> @@ -942,14 +943,13 @@ static int __maps__fixup_overlap_and_insert(struct maps *maps, struct map *new)
> */
> map__put(maps_by_address[i]);
> maps_by_address[i] = map__get(new);
> + map__set_kmap_maps(new, maps);
>
> if (maps_by_name) {
> map__put(maps_by_name[ni]);
> maps_by_name[ni] = map__get(new);
> }
>
> - map__set_kmap_maps(new, maps);
> -
> check_invariants(maps);
> return err;
> }
> @@ -1019,6 +1019,7 @@ int maps__copy_from(struct maps *dest, struct maps *parent)
> err = unwind__prepare_access(dest, new, NULL);
> if (!err) {
> dest_maps_by_address[i] = new;
> + map__set_kmap_maps(new, dest);
> if (dest_maps_by_name)
> dest_maps_by_name[i] = map__get(new);
> RC_CHK_ACCESS(dest)->nr_maps = i + 1;
> --
> 2.51.0.384.g4c02a37b29-goog
>
Powered by blists - more mailing lists