public inbox for [email protected]
 help / color / mirror / Atom feed
* [PATCH] io_uring: only call io_cqring_ev_posted() if events were posted
@ 2021-01-25  4:50 Jens Axboe
  0 siblings, 0 replies; only message in thread
From: Jens Axboe @ 2021-01-25  4:50 UTC (permalink / raw)
  To: io-uring

This normally doesn't cause any extra harm, but it does mean that we'll
increment the eventfd notification count, if one has been registered
with the ring. This can confuse applications, when they see more
notifications on the eventfd side than are available in the ring.

Do the nice thing and only increment this count, if we actually posted
(or even overflowed) events.

Reported-and-tested-by: Dan Melnic <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>

---

diff --git a/fs/io_uring.c b/fs/io_uring.c
index 695fe00bafdc..2166c469789d 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -1779,12 +1779,13 @@ static bool __io_cqring_overflow_flush(struct io_ring_ctx *ctx, bool force,
 	struct io_kiocb *req, *tmp;
 	struct io_uring_cqe *cqe;
 	unsigned long flags;
-	bool all_flushed;
+	bool all_flushed, posted;
 	LIST_HEAD(list);
 
 	if (!force && __io_cqring_events(ctx) == rings->cq_ring_entries)
 		return false;
 
+	posted = false;
 	spin_lock_irqsave(&ctx->completion_lock, flags);
 	list_for_each_entry_safe(req, tmp, &ctx->cq_overflow_list, compl.list) {
 		if (!io_match_task(req, tsk, files))
@@ -1804,6 +1805,7 @@ static bool __io_cqring_overflow_flush(struct io_ring_ctx *ctx, bool force,
 			WRITE_ONCE(ctx->rings->cq_overflow,
 				   ctx->cached_cq_overflow);
 		}
+		posted = true;
 	}
 
 	all_flushed = list_empty(&ctx->cq_overflow_list);
@@ -1813,9 +1815,11 @@ static bool __io_cqring_overflow_flush(struct io_ring_ctx *ctx, bool force,
 		ctx->rings->sq_flags &= ~IORING_SQ_CQ_OVERFLOW;
 	}
 
-	io_commit_cqring(ctx);
+	if (posted)
+		io_commit_cqring(ctx);
 	spin_unlock_irqrestore(&ctx->completion_lock, flags);
-	io_cqring_ev_posted(ctx);
+	if (posted)
+		io_cqring_ev_posted(ctx);
 
 	while (!list_empty(&list)) {
 		req = list_first_entry(&list, struct io_kiocb, compl.list);

-- 
Jens Axboe


^ permalink raw reply related	[flat|nested] only message in thread

only message in thread, other threads:[~2021-01-25  4:50 UTC | newest]

Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2021-01-25  4:50 [PATCH] io_uring: only call io_cqring_ev_posted() if events were posted Jens Axboe

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox