From: Dave Chinner <dgc@kernel.org>
To: Eduard Zateev <hackerowskiy@gmail.com>
Cc: cem@kernel.org, linux-xfs@vger.kernel.org,
linux-kernel@vger.kernel.org,
syzbot+9d8e687768d0f38c1f09@syzkaller.appspotmail.com
Subject: Re: [PATCH] xfs: preserve buffer item with outstanding references
Date: Wed, 21 Oct 2026 07:55:54 +1100 [thread overview]
Message-ID: <atfVWiDCE2LSOBMU@dread> (raw)
In-Reply-To: <20261020200433.3298861-1-hackerowskiy@gmail.com>
On Tue, Oct 20, 2026 at 10:04:31PM +0200, Eduard Zateev wrote:
> The buffer log item can still be referenced by the CIL after the buffer's
> synchronous write completes. In this case xfs_buf_item_relse() is called
> from xfs_buf_ioend() with bli_refcount non-zero. Freeing the item leaves
> a stale pointer in the CIL, which xlog_cil_committed() later dereferences.
>
> Restore the reference-count check removed by commit d2fe5c4c8d25
> ("xfs: rearrange code in xfs_buf_item.c"). If references remain, keep
> the log item attached to the buffer so that its final reference release
> can perform cleanup.
>
> Build-tested: fs/xfs/xfs_buf_item.o (ARCH=arm64, W=1).
> The syzbot report does not include a reproducer, so runtime testing was
> not possible.
>
> Fixes: d2fe5c4c8d25 ("xfs: rearrange code in xfs_buf_item.c")
> Reported-by: syzbot+9d8e687768d0f38c1f09@syzkaller.appspotmail.com
> Closes: https://syzbot.org/bug?extid=9d8e687768d0f38c1f09
> Assisted-by: LLM OpenAI Codex
>
> ---
> fs/xfs/xfs_buf_item.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c
> index 1a4ef34af8d..80f0cf12825 100644
> --- a/fs/xfs/xfs_buf_item.c
> +++ b/fs/xfs/xfs_buf_item.c
> @@ -80,7 +80,8 @@ xfs_buf_item_relse(
> trace_xfs_buf_item_relse(bp, _RET_IP_);
>
> ASSERT(!test_bit(XFS_LI_IN_AIL, &bip->bli_item.li_flags));
> - ASSERT(atomic_read(&bip->bli_refcount) == 0);
> + if (atomic_read(&bip->bli_refcount))
> + return;
>
> bp->b_log_item = NULL;
> xfs_buf_rele(bp);
>
> Signed-off-by: Eduard Zateev <hackerowskiy@gmail.com>
You didn't bother to check the mailing list or syzbot for
duplicates or prior discussion, did you?
What I said about this less than two days ago:
https://lore.kernel.org/linux-xfs/atUsQVuCKnAIZc3X@dread/
-Dave.
--
Dave Chinner
dgc@kernel.org
prev parent reply other threads:[~2026-10-20 20:56 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-10-20 20:04 Eduard Zateev
2026-10-20 20:55 ` Dave Chinner [this message]
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=atfVWiDCE2LSOBMU@dread \
--to=dgc@kernel.org \
--cc=cem@kernel.org \
--cc=hackerowskiy@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-xfs@vger.kernel.org \
--cc=syzbot+9d8e687768d0f38c1f09@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
all inboxes | Powered by JetHome®