Linux kernel -stable discussions
 help / color / mirror / Atom feed
From: Kyle Zeng <kylebot@openai.com>
To: stable@vger.kernel.org
Cc: outbounddisclosures@openai.com, Kyle Zeng <kylebot@openai.com>
Subject: [PATCH 6.12.y] reiserfs: validate tail item bounds before conversion
Date: Thu, 11 Jun 2026 15:40:27 -0700	[thread overview]
Message-ID: <20260611224027.74281-1-kylebot@openai.com> (raw)

direct2indirect() moves direct items into an unformatted-node page while
deleting them from the tree. It assumes the item body copied by
reiserfs_delete_item() and the aggregate number of copied tail bytes both
fit in the target tail page.

A corrupted filesystem can violate those assumptions. Multiple direct
items can each fit in the page while their aggregate length exceeds the
filesystem block size, which underflows the final zero-fill length. A
malformed item can also describe an offset and length that cross the page
boundary before the aggregate length is checked.

Validate the per-item page copy and aggregate tail length before calling
reiserfs_delete_item(), and reject corrupt metadata with -EIO after
reporting a filesystem error.

Assisted-by: Codex:gpt-5.5
Signed-off-by: Kyle Zeng <kylebot@openai.com>
---
 fs/reiserfs/tail_conversion.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)

diff --git a/fs/reiserfs/tail_conversion.c b/fs/reiserfs/tail_conversion.c
index 2cec61af2a9e..ba3dd6770d46 100644
--- a/fs/reiserfs/tail_conversion.c
+++ b/fs/reiserfs/tail_conversion.c
@@ -106,6 +106,8 @@ int direct2indirect(struct reiserfs_transaction_handle *th, struct inode *inode,
 	 * and delete them.
 	 */
 	while (1) {
+		unsigned long item_len;
+		unsigned long pgoff;
 		int tail_size;
 
 		/*
@@ -120,6 +122,17 @@ int direct2indirect(struct reiserfs_transaction_handle *th, struct inode *inode,
 		RFALSE(!is_direct_le_ih(p_le_ih),
 		       "vs-14055: direct item expected(%K), found %h",
 		       &end_key, p_le_ih);
+		item_len = ih_item_len(p_le_ih);
+		pgoff = (le_ih_k_offset(p_le_ih) - 1) & (PAGE_SIZE - 1);
+		if (item_len > PAGE_SIZE - pgoff ||
+		    item_len > blk_size ||
+		    total_tail > blk_size - item_len) {
+			reiserfs_error(sb, "PAP-14060",
+				       "direct item %h does not fit in tail page",
+				       p_le_ih);
+			pathrelse(path);
+			return -EIO;
+		}
 		tail_size = (le_ih_k_offset(p_le_ih) & (blk_size - 1))
 		    + ih_item_len(p_le_ih) - 1;
 
-- 
2.54.0


             reply	other threads:[~2026-06-11 22:40 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-06-11 22:40 Kyle Zeng [this message]
2026-06-12 14:41 ` [PATCH 6.12.y] reiserfs: validate tail item bounds before conversion Sasha Levin

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=20260611224027.74281-1-kylebot@openai.com \
    --to=kylebot@openai.com \
    --cc=outbounddisclosures@openai.com \
    --cc=stable@vger.kernel.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