diff options
| author | Peter Zijlstra <peterz@infradead.org> | 2015-02-25 15:56:04 +0100 |
|---|---|---|
| committer | Ingo Molnar <mingo@kernel.org> | 2015-03-27 09:49:45 +0100 |
| commit | ccd41c86ad4d464d0ed4e48d80759ff85c2115b0 (patch) | |
| tree | b53b1e7507761a9144f900eff6f97cd4bb6d15c5 /kernel/events/core.c | |
| parent | 9332d250b4b4f67c633894b311e022e3cf943bd5 (diff) | |
perf: Fix racy group access
While looking at some fuzzer output I noticed that we do not hold any
locks on leader->ctx and therefore the sibling_list iteration is
unsafe.
Acquire the relevant ctx->mutex before calling into the pmu specific
code.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Sasha Levin <sasha.levin@oracle.com>
Link: http://lkml.kernel.org/r/20150225151639.GL5029@twins.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Diffstat (limited to 'kernel/events/core.c')
| -rw-r--r-- | kernel/events/core.c | 11 |
1 files changed, 11 insertions, 0 deletions
diff --git a/kernel/events/core.c b/kernel/events/core.c index b01dfb602db1..bb1a7c36e794 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -7036,12 +7036,23 @@ EXPORT_SYMBOL_GPL(perf_pmu_unregister); static int perf_try_init_event(struct pmu *pmu, struct perf_event *event) { + struct perf_event_context *ctx = NULL; int ret; if (!try_module_get(pmu->module)) return -ENODEV; + + if (event->group_leader != event) { + ctx = perf_event_ctx_lock(event->group_leader); + BUG_ON(!ctx); + } + event->pmu = pmu; ret = pmu->event_init(event); + + if (ctx) + perf_event_ctx_unlock(event->group_leader, ctx); + if (ret) module_put(pmu->module); |