On Mon, 19 Jul 2021 06:42:25 -0700
syzbot <syzbot+cc699626e48a6ebaf295@syzkaller.appspotmail.com> wrote:

> Hello,
> 
> syzbot found the following issue on:
> 
> HEAD commit:    8096acd7442e Merge tag 'net-5.14-rc2' of
> git://git.kernel... git tree:       upstream
> console output:
> https://syzkaller.appspot.com/x/log.txt?x=167ca94a300000 kernel
> config:  https://syzkaller.appspot.com/x/.config?x=5294764a378649cb
> dashboard link:
> https://syzkaller.appspot.com/bug?extid=cc699626e48a6ebaf295
> compiler:       Debian clang version 11.0.1-2 syz repro:
> https://syzkaller.appspot.com/x/repro.syz?x=14d68024300000 C
> reproducer:   https://syzkaller.appspot.com/x/repro.c?x=16e4f180300000
> 
> IMPORTANT: if you fix the issue, please add the following tag to the
> commit: Reported-by:
> syzbot+cc699626e48a6ebaf295@syzkaller.appspotmail.com
> 
> ==================================================================
> BUG: KASAN: slab-out-of-bounds in debug_spin_lock_before
> kernel/locking/spinlock_debug.c:83 [inline] BUG: KASAN:
> slab-out-of-bounds in do_raw_spin_lock+0x4f5/0x8e0
> kernel/locking/spinlock_debug.c:112 Read of size 4 at addr
> ffff88802b46ce14 by task kworker/0:6/8471
> 

I don't see any reason behind doing clean up stuff in firmware
callback. In my prevoius patch to this driver I fixed memory leak and, I
believed, that this approach won't trigger anything else... 

Let's just call device_release_driver() under parent lock in case of
firmware load failure and see if it works.

#syz test
git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master


With regards,
Pavel Skripkin