2024-03-08 04:22:05

by Imran Khan

[permalink] [raw]
Subject: [RFC PATCH] eventpoll: record task that adds to monitor list.

While debugging latency issues for task waiting on epoll_wait, sometimes
it becomes necessary to see if the lower layer is invoking ep_poll_callback
timely for a given pid or not.

Recording task_struct in involved eppoll_entry's wait_queue_entry, allows us
to check this using a probe (say dtrace) at this function.
We could also achieve this by checking wait_queue_entry on eventpoll's
wait_queue_head itself, but that would involve more indirections.

Signed-off-by: Imran Khan <[email protected]>
---

Since it just helps in debugging, I have kept the patch RFC.

fs/eventpoll.c | 2 ++
1 file changed, 2 insertions(+)

diff --git a/fs/eventpoll.c b/fs/eventpoll.c
index 3534d36a1474..3d4faf0a2d25 100644
--- a/fs/eventpoll.c
+++ b/fs/eventpoll.c
@@ -1287,6 +1287,8 @@ static void ep_ptable_queue_proc(struct file *file, wait_queue_head_t *whead,
init_waitqueue_func_entry(&pwq->wait, ep_poll_callback);
pwq->whead = whead;
pwq->base = epi;
+ pwq->wait.private = current;
+
if (epi->event.events & EPOLLEXCLUSIVE)
add_wait_queue_exclusive(whead, &pwq->wait);
else
--
2.34.1