From: Yuto Ohnuki <ytohnuki@amazon.com>
To: <djwong@kernel.org>
Cc: <bfoster@redhat.com>, <cem@kernel.org>, <darrick.wong@oracle.com>,
<dchinner@redhat.com>, <linux-kernel@vger.kernel.org>,
<linux-xfs@vger.kernel.org>, <stable@vger.kernel.org>,
<syzbot+652af2b3c5569c4ab63c@syzkaller.appspotmail.com>,
<ytohnuki@amazon.com>
Subject: Re: [PATCH v3 3/4] xfs: avoid dereferencing log items after push callbacks
Date: Tue, 10 Mar 2026 17:51:00 +0000 [thread overview]
Message-ID: <20260310175059.78341-2-ytohnuki@amazon.com> (raw)
In-Reply-To: <20260309162710.GC6033@frogsfrogsfrogs>
> How difficult would it be to add a refcount to xfs_log_item so that any
> other code walking through the AIL's log item list can't accidentally
> suffer from this UAF? I keep seeing periodic log item UAF bugfixes on
> the list, which (to me anyway) suggests we ought to think about a
> struct(ural) fix to this problem.
>
> I /think/ the answer to that is "sort of nasty" because of things like
> xfs_dquot embedding its own log item. The other log item types might
> not be so bad because at least they're allocated separately. However,
> refcount_t accesses also aren't free.
Agreed that a structural fix would be the right long-term approach.
As you noted, the dquot embedding makes it non-trivial. I'd like to
keep this series focused on the immediate syzbot fix and explore a
refcount-based approach as a separate effort.
> This is true after the xfsaild_push_item call, correct? If so then I
> think the comment for the call needs updating too:
>
> /*
> * Note that iop_push may unlock and reacquire the AIL lock. We
> * rely on the AIL cursor implementation to be able to deal with
> * the dropped lock.
> *
> * The log item may have been freed by the push, so it must not
> * be accessed or dereferenced below this line.
> */
> lock_result = xfsaild_push_item(ailp, lip);
>
> Otherwise this looks ok to me.
>
> --D
Thank you. In v4, I have added the comments you suggested.
Yuto
Amazon Web Services EMEA SARL, 38 avenue John F. Kennedy, L-1855 Luxembourg, R.C.S. Luxembourg B186284
Amazon Web Services EMEA SARL, Irish Branch, One Burlington Plaza, Burlington Road, Dublin 4, Ireland, branch registration number 908705
next prev parent reply other threads:[~2026-03-10 17:51 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-03-08 18:28 [PATCH v3 0/4] xfs: fix AIL push use-after-free during shutdown Yuto Ohnuki
2026-03-08 18:28 ` [PATCH v3 1/4] xfs: stop reclaim before pushing AIL during unmount Yuto Ohnuki
2026-03-09 16:02 ` Darrick J. Wong
2026-03-10 17:33 ` Yuto Ohnuki
2026-03-08 18:28 ` [PATCH v3 2/4] xfs: refactor xfsaild_push loop into helper Yuto Ohnuki
2026-03-09 16:14 ` Darrick J. Wong
2026-03-10 17:38 ` Yuto Ohnuki
2026-03-10 5:26 ` Dave Chinner
2026-03-10 17:46 ` Yuto Ohnuki
2026-03-08 18:28 ` [PATCH v3 3/4] xfs: avoid dereferencing log items after push callbacks Yuto Ohnuki
2026-03-09 16:27 ` Darrick J. Wong
2026-03-10 5:25 ` Dave Chinner
2026-03-10 17:51 ` Yuto Ohnuki [this message]
2026-03-10 5:27 ` Dave Chinner
2026-03-10 17:56 ` Yuto Ohnuki
2026-03-08 18:28 ` [PATCH v3 4/4] xfs: save ailp before dropping the AIL lock in " Yuto Ohnuki
2026-03-09 16:28 ` Darrick J. Wong
2026-03-10 5:27 ` Dave Chinner
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20260310175059.78341-2-ytohnuki@amazon.com \
--to=ytohnuki@amazon.com \
--cc=bfoster@redhat.com \
--cc=cem@kernel.org \
--cc=darrick.wong@oracle.com \
--cc=dchinner@redhat.com \
--cc=djwong@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-xfs@vger.kernel.org \
--cc=stable@vger.kernel.org \
--cc=syzbot+652af2b3c5569c4ab63c@syzkaller.appspotmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox