[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20250522-exotisch-chloren-3fa7b7ce5266@brauner>
Date: Thu, 22 May 2025 14:24:10 +0200
From: Christian Brauner <brauner@...nel.org>
To: Ard Biesheuvel <ardb@...nel.org>
Cc: syzbot <syzbot+52cd651546d11d2af06b@...kaller.appspotmail.com>,
"James E.J. Bottomley" <James.Bottomley@...senpartnership.com>, Al Viro <viro@...iv.linux.org.uk>, jk@...abs.org,
linux-efi@...r.kernel.org, linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org,
syzkaller-bugs@...glegroups.com
Subject: Re: [syzbot] [fs?] [efi?] BUG: unable to handle kernel paging
request in alloc_fs_context
On Wed, May 21, 2025 at 07:28:34PM +0200, Ard Biesheuvel wrote:
> (cc James, Al, Christian)
>
> Please see the splat below.
>
> The NULL dereference is due to get_cred() in alloc_fs_context()
> attempting to increment current->cred->usage while current->cred ==
> NULL, and this is a result of the fact that PM notifier call chain is
> called while the task is exiting.
>
> IIRC, the intent was for commit
>
> 11092db5b573 efivarfs: fix NULL dereference on resume
>
> to be replaced at some point with something more robust; might that
> address this issue as well?
Yes. It's in the merge queue for v6.16 which kills off all that
vfs_kern_mount() stuff.
Powered by blists - more mailing lists