[PATCH 6/7] io_uring: thin down io_commit_cqring()

[Pavel Begunkov <asml.silence@xxxxxxxxx>]

io_commit_cqring() is currently always under spinlock section, so it's
always better to keep it as slim as possible. Move
__io_commit_cqring_flush() out of it into ev_posted*(). If fast checks
do fail and this post-processing is required, we'll reacquire
->completion_lock, which is fine as we don't care about performance of
draining and offset timeouts.

Signed-off-by: Pavel Begunkov <asml.silence@xxxxxxxxx>
---
 fs/io_uring.c | 23 +++++++++++++++--------
 1 file changed, 15 insertions(+), 8 deletions(-)

diff --git a/fs/io_uring.c b/fs/io_uring.c
index 5a87e0622ecb..c75a5767f58d 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -1771,20 +1771,21 @@ static __cold void io_flush_timeouts(struct io_ring_ctx *ctx)
 	spin_unlock_irq(&ctx->timeout_lock);
 }
 
+static inline void io_commit_cqring(struct io_ring_ctx *ctx)
+{
+	/* order cqe stores with ring update */
+	smp_store_release(&ctx->rings->cq.tail, ctx->cached_cq_tail);
+}
+
 static __cold void __io_commit_cqring_flush(struct io_ring_ctx *ctx)
 {
+	spin_lock(&ctx->completion_lock);
 	if (ctx->off_timeout_used)
 		io_flush_timeouts(ctx);
 	if (ctx->drain_active)
 		io_queue_deferred(ctx);
-}
-
-static inline void io_commit_cqring(struct io_ring_ctx *ctx)
-{
-	if (unlikely(ctx->off_timeout_used || ctx->drain_active))
-		__io_commit_cqring_flush(ctx);
-	/* order cqe stores with ring update */
-	smp_store_release(&ctx->rings->cq.tail, ctx->cached_cq_tail);
+	io_commit_cqring(ctx);
+	spin_unlock(&ctx->completion_lock);
 }
 
 static inline bool io_sqring_full(struct io_ring_ctx *ctx)
@@ -1852,6 +1853,9 @@ static void io_eventfd_signal(struct io_ring_ctx *ctx)
  */
 static inline void io_cqring_ev_posted(struct io_ring_ctx *ctx)
 {
+	if (unlikely(ctx->off_timeout_used || ctx->drain_active))
+		__io_commit_cqring_flush(ctx);
+
 	/*
 	 * wake_up_all() may seem excessive, but io_wake_function() and
 	 * io_should_wake() handle the termination of the loop and only
@@ -1865,6 +1869,9 @@ static inline void io_cqring_ev_posted(struct io_ring_ctx *ctx)
 
 static void io_cqring_ev_posted_iopoll(struct io_ring_ctx *ctx)
 {
+	if (unlikely(ctx->off_timeout_used || ctx->drain_active))
+		__io_commit_cqring_flush(ctx);
+
 	if (ctx->flags & IORING_SETUP_SQPOLL) {
 		if (wq_has_sleeper(&ctx->cq_wait))
 			wake_up_all(&ctx->cq_wait);
-- 
2.35.1