All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Darrick J. Wong" <djwong@kernel.org>
To: Raphael Pinsonneault-Thibeault <rpthibeault@gmail.com>
Cc: cem@kernel.org, chandanbabu@kernel.org, bfoster@redhat.com,
	david@fromorbit.com, hch@infradead.org,
	linux-xfs@vger.kernel.org, linux-kernel@vger.kernel.org,
	linux-kernel-mentees@lists.linux.dev,
	syzbot+9f6d080dece587cfdd4c@syzkaller.appspotmail.com
Subject: Re: [PATCH v5] xfs: validate log record version against superblock log version
Date: Mon, 24 Nov 2025 10:52:03 -0800	[thread overview]
Message-ID: <20251124185203.GA6076@frogsfrogsfrogs> (raw)
In-Reply-To: <20251124174658.59275-3-rpthibeault@gmail.com>

On Mon, Nov 24, 2025 at 12:47:00PM -0500, Raphael Pinsonneault-Thibeault wrote:
> Syzbot creates a fuzzed record where xfs_has_logv2() but the
> xlog_rec_header h_version != XLOG_VERSION_2. This causes a
> KASAN: slab-out-of-bounds read in xlog_do_recovery_pass() ->
> xlog_recover_process() -> xlog_cksum().
> 
> Fix by adding a check to xlog_valid_rec_header() to abort journal
> recovery if the xlog_rec_header h_version does not match the super
> block log version.
> 
> A file system with a version 2 log will only ever set
> XLOG_VERSION_2 in its headers (and v1 will only ever set V_1), so if
> there is any mismatch, either the journal or the superblock has been
> corrupted and therefore we abort processing with a -EFSCORRUPTED error
> immediately.
> 
> Also, refactor the structure of the validity checks for better
> readability. At the default error level (LOW), XFS_IS_CORRUPT() emits
> the condition that failed, the file and line number it is
> located at, then dumps the stack. This gives us everything we need
> to know about the failure if we do a single validity check per
> XFS_IS_CORRUPT().
> 
> Reported-by: syzbot+9f6d080dece587cfdd4c@syzkaller.appspotmail.com
> Closes: https://syzkaller.appspot.com/bug?extid=9f6d080dece587cfdd4c
> Tested-by: syzbot+9f6d080dece587cfdd4c@syzkaller.appspotmail.com
> Fixes: 45cf976008dd ("xfs: fix log recovery buffer allocation for the legacy h_size fixup")
> Signed-off-by: Raphael Pinsonneault-Thibeault <rpthibeault@gmail.com>
> ---
> Changelog
> v1 -> v2: 
> - reject the mount for h_size > XLOG_HEADER_CYCLE_SIZE && !XLOG_VERSION_2
> v2 -> v3: 
> - abort journal recovery if the xlog_rec_header h_version does not 
> match the super block log version
> v3 -> v4: 
> - refactor for readability
> v4 -> v5:
> - stop pretending h_version is a bitmap, remove check using
> XLOG_VERSION_OKBITS

Hrmm maybe we ought to reserve XLOG_VERSION==0x3 so that whenever we do
log v3 we don't accidentally write logs with bits that won't be
validated quite right on old kernels?

Though I suppose logv3 would have a separate superblock bit an old
kernel just plain wouldn't mount a logv3 filesystem, let alone look at
its log.

>  fs/xfs/xfs_log_recover.c | 27 ++++++++++++++++-----------
>  1 file changed, 16 insertions(+), 11 deletions(-)
> 
> diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
> index e6ed9e09c027..2ed94be010d0 100644
> --- a/fs/xfs/xfs_log_recover.c
> +++ b/fs/xfs/xfs_log_recover.c
> @@ -2950,18 +2950,23 @@ xlog_valid_rec_header(
>  	xfs_daddr_t		blkno,
>  	int			bufsize)
>  {
> +	struct xfs_mount	*mp = log->l_mp;
> +	u32			h_version = be32_to_cpu(rhead->h_version);
>  	int			hlen;
>  
> -	if (XFS_IS_CORRUPT(log->l_mp,
> +	if (XFS_IS_CORRUPT(mp,
>  			   rhead->h_magicno != cpu_to_be32(XLOG_HEADER_MAGIC_NUM)))
>  		return -EFSCORRUPTED;
> -	if (XFS_IS_CORRUPT(log->l_mp,
> -			   (!rhead->h_version ||
> -			   (be32_to_cpu(rhead->h_version) &
> -			    (~XLOG_VERSION_OKBITS))))) {
> -		xfs_warn(log->l_mp, "%s: unrecognised log version (%d).",
> -			__func__, be32_to_cpu(rhead->h_version));
> -		return -EFSCORRUPTED;
> +
> +	/*
> +	 * The log version must match the superblock
> +	 */
> +	if (xfs_has_logv2(mp)) {
> +		if (XFS_IS_CORRUPT(mp, h_version != XLOG_VERSION_2))

Being pedantic here, but the kernel cpu_to_be32 wrappers are magic in
that they compile to byteswapped constants so you can avoid the runtime
overhead of byteswapping rhead->h_version by doing:

	if (XFS_IS_CORRUPT(mp,
	    rhead->h_version != cpu_to_be32(XLOG_VERSION_2)))
		return -EFSCORRUPTED;

But seeing as this is log validation for recovery, I think the
performance implications are vanishingly small.

I suppose if we /really/ want to be pedantic then the change of bitmask
test to direct comparison ought to be a separate patch in case someone
some day ends up bisecting a log recovery problem to this patch.

Either way,
Reviewed-by: "Darrick J. Wong" <djwong@kernel.org>

--D


> +			return -EFSCORRUPTED;
> +	} else {
> +		if (XFS_IS_CORRUPT(mp, h_version != XLOG_VERSION_1))
> +			return -EFSCORRUPTED;
>  	}
>  
>  	/*
> @@ -2969,12 +2974,12 @@ xlog_valid_rec_header(
>  	 * and h_len must not be greater than LR buffer size.
>  	 */
>  	hlen = be32_to_cpu(rhead->h_len);
> -	if (XFS_IS_CORRUPT(log->l_mp, hlen <= 0 || hlen > bufsize))
> +	if (XFS_IS_CORRUPT(mp, hlen <= 0 || hlen > bufsize))
>  		return -EFSCORRUPTED;
>  
> -	if (XFS_IS_CORRUPT(log->l_mp,
> -			   blkno > log->l_logBBsize || blkno > INT_MAX))
> +	if (XFS_IS_CORRUPT(mp, blkno > log->l_logBBsize || blkno > INT_MAX))
>  		return -EFSCORRUPTED;
> +
>  	return 0;
>  }
>  
> -- 
> 2.43.0
> 

  reply	other threads:[~2025-11-24 18:52 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-11-12 14:10 [PATCH] xfs: ensure log recovery buffer is resized to avoid OOB Raphael Pinsonneault-Thibeault
2025-11-12 15:28 ` Christoph Hellwig
2025-11-12 18:18   ` [PATCH] xfs: reject log records with v2 size but v1 header version " Raphael Pinsonneault-Thibeault
2025-11-12 18:45     ` Darrick J. Wong
2025-11-13  6:55       ` Christoph Hellwig
2025-11-12 22:19 ` [PATCH] xfs: ensure log recovery buffer is resized " Dave Chinner
2025-11-13 19:01   ` [PATCH v3] xfs: validate log record version against superblock log version Raphael Pinsonneault-Thibeault
2025-11-18 20:19     ` Dave Chinner
2025-11-19 15:37       ` [PATCH v4] " Raphael Pinsonneault-Thibeault
2025-11-19 20:16         ` Dave Chinner
2025-11-20  6:57         ` Christoph Hellwig
2025-11-24 17:47           ` [PATCH v5] " Raphael Pinsonneault-Thibeault
2025-11-24 18:52             ` Darrick J. Wong [this message]
2025-11-25  6:31               ` Christoph Hellwig
2025-11-25 17:06                 ` Darrick J. Wong
2025-11-25  6:31             ` Christoph Hellwig

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=20251124185203.GA6076@frogsfrogsfrogs \
    --to=djwong@kernel.org \
    --cc=bfoster@redhat.com \
    --cc=cem@kernel.org \
    --cc=chandanbabu@kernel.org \
    --cc=david@fromorbit.com \
    --cc=hch@infradead.org \
    --cc=linux-kernel-mentees@lists.linux.dev \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-xfs@vger.kernel.org \
    --cc=rpthibeault@gmail.com \
    --cc=syzbot+9f6d080dece587cfdd4c@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.