| Message ID | 20240827124654.3817443-1-zhengyejian@huaweicloud.com (mailing list archive) |
|---|---|
| State | Accepted |
| Commit | 49aa8a1f4d6800721c7971ed383078257f12e8f9 |
| Headers | show |
| Series | [v2] tracing: Avoid possible softlockup in tracing_iter_reset() | expand |
diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index ebe7ce2f5f4a..edf6bc817aa1 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -3958,6 +3958,8 @@ void tracing_iter_reset(struct trace_iterator *iter, int cpu) break; entries++; ring_buffer_iter_advance(buf_iter); + /* This could be a big loop */ + cond_resched(); } per_cpu_ptr(iter->array_buffer->data, cpu)->skipped_entries = entries;
In __tracing_open(), when max latency tracers took place on the cpu, the time start of its buffer would be updated, then event entries with timestamps being earlier than start of the buffer would be skipped (see tracing_iter_reset()). Softlockup will occur if the kernel is non-preemptible and too many entries were skipped in the loop that reset every cpu buffer, so add cond_resched() to avoid it. Suggested-by: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Zheng Yejian <zhengyejian@huaweicloud.com> --- kernel/trace/trace.c | 2 ++ 1 file changed, 2 insertions(+) v2: - Change to add cond_resched() in tracing_iter_reset() Link: https://lore.kernel.org/all/20240826103522.390faa85@gandalf.local.home/ - Update commit title and add suggested-by tag v1: https://lore.kernel.org/all/20240824030343.3218618-1-zhengyejian@huaweicloud.com/