Skip to contentSkip to conversation

io_uring: don't touch sqd->thread off tw add

PATCH Applied Pavel Begunkov 2 participants 3 messages Jan 10, 2025 io_uring

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 io_uring: don't touch sqd->thread off tw add

With IORING_SETUP_SQPOLL all requests are created by the SQPOLL task, which means that req->task should always match sqd->thread. Since accesses to sqd->thread should be separately protected, use req->task in io_req_normal_work_add() instead.

Note, in the eyes of io_req_normal_work_add(), the SQPOLL task struct is always pinned and alive, and sqd->thread can either be the task or NULL. It's only problematic if the compiler decides to reload the value after the null check, which is not so likely.

Cc Bui Quang Minh <[email protected]>
Fixes 78f9b61bd8e54 ("io_uring: wake SQPOLL task when task_work is added to an empty queue")
---
 io_uring/io_uring.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)
io_uring/io_uring.c +1 -4
index db198bd435b5..9b83b875d2e4 100644
@@ -1225,10 +1225,7 @@ static void io_req_normal_work_add(struct io_kiocb *req)
 
 	/* SQPOLL doesn't need the task_work added, it'll run it itself */
 	if (ctx->flags & IORING_SETUP_SQPOLL) {
-		struct io_sq_data *sqd = ctx->sq_data;
-
-		if (sqd->thread)
-			__set_notify_signal(sqd->thread);
+		__set_notify_signal(tctx->task);
 		return;
 	}
 
Signature · 1 line
2.47.1
#1 lore raw
Quoted 9 lines · Pavel Begunkov

With IORING_SETUP_SQPOLL all requests are created by the SQPOLL task, which means that req->task should always match sqd->thread. Since accesses to sqd->thread should be separately protected, use req->task in io_req_normal_work_add() instead.

Note, in the eyes of io_req_normal_work_add(), the SQPOLL task struct is always pinned and alive, and sqd->thread can either be the task or NULL. It's only problematic if the compiler decides to reload the value after the null check, which is not so likely.

We don't have much time to drag it on, let's fix it up so it hopefully gets into 6.13

Quoted 25 lines
Cc Bui Quang Minh <[email protected]>
Fixes 78f9b61bd8e54 ("io_uring: wake SQPOLL task when task_work is added to an empty queue")
---
  io_uring/io_uring.c | 5 +----
  1 file changed, 1 insertion(+), 4 deletions(-)
io_uring/io_uring.c +1 -4
index db198bd435b5..9b83b875d2e4 100644
@@ -1225,10 +1225,7 @@ static void io_req_normal_work_add(struct io_kiocb *req)
 
  	/* SQPOLL doesn't need the task_work added, it'll run it itself */
  	if (ctx->flags & IORING_SETUP_SQPOLL) {
-		struct io_sq_data *sqd = ctx->sq_data;
-
-		if (sqd->thread)
-			__set_notify_signal(sqd->thread);
+		__set_notify_signal(tctx->task);
  		return;
  	}
 
Signature · 1 line
Pavel Begunkov
maintainer Applied
#2 lore raw
In reply to Pavel Begunkov
Quoted 11 lines · Pavel Begunkov

With IORING_SETUP_SQPOLL all requests are created by the SQPOLL task, which means that req->task should always match sqd->thread. Since accesses to sqd->thread should be separately protected, use req->task in io_req_normal_work_add() instead.

Note, in the eyes of io_req_normal_work_add(), the SQPOLL task struct is always pinned and alive, and sqd->thread can either be the task or NULL. It's only problematic if the compiler decides to reload the value after the null check, which is not so likely.

[...]

Applied, thanks!

[1/1] io_uring: don't touch sqd->thread off tw add

      commit: bd2703b42decebdcddf76e277ba76b4c4a142d73

Best regards,

Signature · 1 line
Jens Axboe