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

[Jens Axboe]

On 7/24/24 3:03 PM, syzbot wrote:
> Hello,
> 
> syzbot has tested the proposed patch and the reproducer did not trigger any issue:
> 
> Reported-by: syzbot+82609b8937a4458106ca@xxxxxxxxxxxxxxxxxxxxxxxxx
> Tested-by: syzbot+82609b8937a4458106ca@xxxxxxxxxxxxxxxxxxxxxxxxx
> 
> Tested on:
> 
> commit:         c33ffdb7 Merge tag 'phy-for-6.11' of git://git.kernel...
> git tree:       upstream
> console output: https://syzkaller.appspot.com/x/log.txt?x=17d29bb5980000
> kernel config:  https://syzkaller.appspot.com/x/.config?x=f26f43c6f7db5ad2
> dashboard link: https://syzkaller.appspot.com/bug?extid=82609b8937a4458106ca
> compiler:       Debian clang version 15.0.6, GNU ld (GNU Binutils for Debian) 2.40
> patch:          https://syzkaller.appspot.com/x/patch.diff?x=120b7dad980000
> 
> Note: testing is done by a robot and is best-effort only.

As I figured... No idea what's going on here, nr_tw is what is being
complained about and it's _clearly_ initialized. Randomly disabling a
branch that won't touch it off this path (hence it should be a no-op in
terms of code executed messing with nr_tw) makes it go away.

-- 
Jens Axboe