Re: [syzbot] [io-uring?] KMSAN: uninit-value in io_req_task_work_add_remote

[Jens Axboe <axboe@kernel.dk>]

On 7/24/24 4:51 AM, syzbot wrote:
> Hello,
> 
> syzbot found the following issue on:
> 
> HEAD commit:    933069701c1b Merge tag '6.11-rc-smb3-server-fixes' of git:..
> git tree:       upstream
> console+strace: https://syzkaller.appspot.com/x/log.txt?x=16e38d5e980000
> kernel config:  https://syzkaller.appspot.com/x/.config?x=c062b3d00b275b52
> dashboard link: https://syzkaller.appspot.com/bug?extid=82609b8937a4458106ca
> compiler:       Debian clang version 15.0.6, GNU ld (GNU Binutils for Debian) 2.40
> syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=149e5245980000
> C reproducer:   https://syzkaller.appspot.com/x/repro.c?x=1388c55e980000

#syz test

diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c
index 2626424f5d73..1aaab21e1574 100644
--- a/io_uring/io_uring.c
+++ b/io_uring/io_uring.c
@@ -1137,9 +1137,10 @@ static inline void io_req_local_work_add(struct io_kiocb *req,
 	BUILD_BUG_ON(IO_CQ_WAKE_FORCE <= IORING_MAX_CQ_ENTRIES);
 
 	/*
-	 * We don't know how many reuqests is there in the link and whether
-	 * they can even be queued lazily, fall back to non-lazy.
+	 * We don't know how many requests are in the link and whether they can
+	 * even be queued lazily, fall back to non-lazy.
 	 */
+	req->nr_tw = 0;
 	if (req->flags & (REQ_F_LINK | REQ_F_HARDLINK))
 		flags &= ~IOU_F_TWQ_LAZY_WAKE;
 

-- 
Jens Axboe