public inbox for linux-erofs@ozlabs.org
 help / color / mirror / Atom feed
From: Utkal Singh <singhutkal015@gmail.com>
To: hsiangkao@linux.alibaba.com
Cc: linux-erofs@lists.ozlabs.org, Utkal Singh <singhutkal015@gmail.com>
Subject: [PATCH] fsck.erofs: xattr: validate h_shared_count before shared entry loop
Date: Sun, 15 Mar 2026 07:38:11 +0000	[thread overview]
Message-ID: <20260315073811.26665-1-singhutkal015@gmail.com> (raw)

The h_shared_count field in erofs_xattr_ibody_header is a raw u8 read
directly from the on-disk image.  erofs_verify_xattr() subtracts
h_shared_count * sizeof(erofs_xattr_entry) from 'remaining' without
first checking whether that product fits within the available space:

  remaining -= xattr_hdr_size;           /* remaining = xattr_isize - 12 */
  for (i = 0; i < xattr_shared_count; ++i) {
          ...
          remaining -= xattr_entry_size; /* can go deeply negative */
  }
  while (remaining > 0) {               /* silently skipped! */
          /* inline xattr validation never runs */
  }

A crafted image with h_shared_count=100 and xattr_isize=12 causes
remaining to reach -400.  The subsequent while(remaining > 0) loop
is never entered, so all inline xattr entries are silently skipped
and fsck.erofs reports the image as clean despite corruption.

Fix by checking that h_shared_count * xattr_entry_size does not
exceed the remaining space before the loop, returning -EFSCORRUPTED
on violation.

Signed-off-by: Utkal Singh <singhutkal015@gmail.com>
---
 fsck/main.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

diff --git a/fsck/main.c b/fsck/main.c
index 16cc627..a3cd50d 100644
--- a/fsck/main.c
+++ b/fsck/main.c
@@ -378,6 +378,20 @@ static int erofs_verify_xattr(struct erofs_inode *inode)
 	ofs = erofs_blkoff(sbi, addr) + xattr_hdr_size;
 	addr += xattr_hdr_size;
 	remaining -= xattr_hdr_size;
+	/* Validate shared count fits within remaining xattr space.
+	 * h_shared_count is a raw u8 from disk; if it exceeds the
+	 * available space, remaining goes negative and the inline
+	 * xattr validation loop at line 396 is silently skipped,
+	 * causing fsck to miss structurally corrupt xattr data.
+	 */
+	if ((unsigned int)xattr_shared_count * xattr_entry_size >
+	    (unsigned int)remaining) {
+		erofs_err("nid %llu: h_shared_count %u exceeds xattr_isize %u",
+			  (unsigned long long)inode->nid,
+			  xattr_shared_count, inode->xattr_isize);
+		ret = -EFSCORRUPTED;
+		goto out;
+	}
 	for (i = 0; i < xattr_shared_count; ++i) {
 		if (ofs >= erofs_blksiz(sbi)) {
 			if (ofs != erofs_blksiz(sbi)) {
-- 
2.43.0



                 reply	other threads:[~2026-03-15  7:38 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=20260315073811.26665-1-singhutkal015@gmail.com \
    --to=singhutkal015@gmail.com \
    --cc=hsiangkao@linux.alibaba.com \
    --cc=linux-erofs@lists.ozlabs.org \
    /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