Received-SPF: pass (google.com: domain of linux-kernel+bounces-112901-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45d1:ec00::1 as permitted sender) client-ip=2604:1380:45d1:ec00::1;
From: Sasha Levin <sashal@kernel.org>
To: linux-kernel@vger.kernel.org,
	stable@vger.kernel.org
Cc: Jens Axboe <axboe@kernel.dk>,
	Sasha Levin <sashal@kernel.org>
Subject: [PATCH 6.8 014/715] io_uring: remove looping around handling traditional task_work
Date: Sun, 24 Mar 2024 18:23:13 -0400
Message-ID: <20240324223455.1342824-15-sashal@kernel.org>
In-Reply-To: <20240324223455.1342824-1-sashal@kernel.org>
References: <20240324223455.1342824-1-sashal@kernel.org>
Precedence: bulk
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit

From: Jens Axboe <axboe@kernel.dk>

[ Upstream commit 592b4805432af075468876771c0f7d41273ccb3c ]

A previous commit added looping around handling traditional task_work
as an optimization, and while that may seem like a good idea, it's also
possible to run into application starvation doing so. If the task_work
generation is bursty, we can get very deep task_work queues, and we can
end up looping in here for a very long time.

One immediately observable problem with that is handling network traffic
using provided buffers, where flooding incoming traffic and looping
task_work handling will very quickly lead to buffer starvation as we
keep running task_work rather than returning to the application so it
can handle the associated CQEs and also provide buffers back.

Fixes: 3a0c037b0e16 ("io_uring: batch task_work")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 io_uring/io_uring.c | 45 +++++++--------------------------------------
 1 file changed, 7 insertions(+), 38 deletions(-)

diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c
index cd9a137ad6cef..5233a20d01b54 100644
--- a/io_uring/io_uring.c
+++ b/io_uring/io_uring.c
@@ -1176,12 +1176,11 @@ static void ctx_flush_and_put(struct io_ring_ctx *ctx, struct io_tw_state *ts)
 
 static unsigned int handle_tw_list(struct llist_node *node,
 				   struct io_ring_ctx **ctx,
-				   struct io_tw_state *ts,
-				   struct llist_node *last)
+				   struct io_tw_state *ts)
 {
 	unsigned int count = 0;
 
-	while (node && node != last) {
+	do {
 		struct llist_node *next = node->next;
 		struct io_kiocb *req = container_of(node, struct io_kiocb,
 						    io_task_work.node);
@@ -1205,7 +1204,7 @@ static unsigned int handle_tw_list(struct llist_node *node,
 			*ctx = NULL;
 			cond_resched();
 		}
-	}
+	} while (node);
 
 	return count;
 }
@@ -1224,22 +1223,6 @@ static inline struct llist_node *io_llist_xchg(struct llist_head *head,
 	return xchg(&head->first, new);
 }
 
-/**
- * io_llist_cmpxchg - possibly swap all entries in a lock-less list
- * @head:	the head of lock-less list to delete all entries
- * @old:	expected old value of the first entry of the list
- * @new:	new entry as the head of the list
- *
- * perform a cmpxchg on the first entry of the list.
- */
-
-static inline struct llist_node *io_llist_cmpxchg(struct llist_head *head,
-						  struct llist_node *old,
-						  struct llist_node *new)
-{
-	return cmpxchg(&head->first, old, new);
-}
-
 static __cold void io_fallback_tw(struct io_uring_task *tctx, bool sync)
 {
 	struct llist_node *node = llist_del_all(&tctx->task_list);
@@ -1274,9 +1257,7 @@ void tctx_task_work(struct callback_head *cb)
 	struct io_ring_ctx *ctx = NULL;
 	struct io_uring_task *tctx = container_of(cb, struct io_uring_task,
 						  task_work);
-	struct llist_node fake = {};
 	struct llist_node *node;
-	unsigned int loops = 0;
 	unsigned int count = 0;
 
 	if (unlikely(current->flags & PF_EXITING)) {
@@ -1284,21 +1265,9 @@ void tctx_task_work(struct callback_head *cb)
 		return;
 	}
 
-	do {
-		loops++;
-		node = io_llist_xchg(&tctx->task_list, &fake);
-		count += handle_tw_list(node, &ctx, &ts, &fake);
-
-		/* skip expensive cmpxchg if there are items in the list */
-		if (READ_ONCE(tctx->task_list.first) != &fake)
-			continue;
-		if (ts.locked && !wq_list_empty(&ctx->submit_state.compl_reqs)) {
-			io_submit_flush_completions(ctx);
-			if (READ_ONCE(tctx->task_list.first) != &fake)
-				continue;
-		}
-		node = io_llist_cmpxchg(&tctx->task_list, &fake, NULL);
-	} while (node != &fake);
+	node = llist_del_all(&tctx->task_list);
+	if (node)
+		count = handle_tw_list(node, &ctx, &ts);
 
 	ctx_flush_and_put(ctx, &ts);
 
@@ -1306,7 +1275,7 @@ void tctx_task_work(struct callback_head *cb)
 	if (unlikely(atomic_read(&tctx->in_cancel)))
 		io_uring_drop_tctx_refs(current);
 
-	trace_io_uring_task_work_run(tctx, count, loops);
+	trace_io_uring_task_work_run(tctx, count, 1);
 }
 
 static inline void io_req_local_work_add(struct io_kiocb *req, unsigned flags)
-- 
2.43.0