mail archive of the barebox mailing list
 help / color / mirror / Atom feed
From: Ahmad Fatoum <a.fatoum@pengutronix.de>
To: barebox@lists.infradead.org
Cc: Ahmad Fatoum <a.fatoum@barebox.org>
Subject: [PATCH 4/6] fs: ubootvarfs: do not form pointers past the end of the environment
Date: Fri,  2 Oct 2026 13:40:07 +0200	[thread overview]
Message-ID: <20261002114253.2906535-5-a.fatoum@pengutronix.de> (raw)
In-Reply-To: <20261002114253.2906535-1-a.fatoum@pengutronix.de>

From: Ahmad Fatoum <a.fatoum@barebox.org>

When the environment fills its whole partition, data->end is one past
the end of the mapping. The room check in create adds the length of
the new name to it and tail relocation adds one, both yielding pointers
further past the end, which C leaves undefined even if they are only
compared.

Let's compare the room left against the length instead and only step
past data->end when it is still inside the mapping. No change in
behavior with the compilers we use.

Fixes: 8daaa21b3949 ("fs: Add a driver to access U-Boot environment variables")
Fixes: 79f048173d98 ("fs: ubootvarfs: fix tail relocation length computation")
Assisted-by: Claude:opus-5.5
Signed-off-by: Ahmad Fatoum <a.fatoum@barebox.org>
---
 fs/ubootvarfs.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/ubootvarfs.c b/fs/ubootvarfs.c
index 3b5c3039e3aa..72d0627fb0cd 100644
--- a/fs/ubootvarfs.c
+++ b/fs/ubootvarfs.c
@@ -210,7 +210,7 @@ static void ubootvarfs_relocate_tail(struct ubootvarfs_inode *node,
 	 * the last entry's NUL is the final byte and data->end points
 	 * one past the mapping, so there is no terminator to carry along.
 	 */
-	tail_end = min_t(const char *, data->end + 1, data->limit);
+	tail_end = data->end == data->limit ? data->end : data->end + 1;
 
 	memmove(src + delta, src, tail_end - src);
 
@@ -286,7 +286,7 @@ static int ubootvarfs_create(struct inode *dir, struct dentry *dentry,
 	 * we need to make sure there's enough room for it. Note that
 	 * + 3 is to accoutn for '=', and two '\0' from above
 	 */
-	if (data->end + len + 3 > data->limit)
+	if (len + 3 > data->limit - data->end)
 		return -ENOSPC;
 
 	var = xmalloc(sizeof(*var));
-- 
2.47.3




  parent reply	other threads:[~2026-10-02 11:44 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-10-02 11:40 [PATCH 0/6] fs: ubootvarfs: harden parser Ahmad Fatoum
2026-10-02 11:40 ` [PATCH 1/6] fs: skip directory entries whose name does not fit struct dirent Ahmad Fatoum
2026-10-02 11:40 ` [PATCH 2/6] fs: reject negative lengths in ftruncate() Ahmad Fatoum
2026-10-02 11:40 ` [PATCH 3/6] fs: ubootvarfs: range-check the new size in truncate Ahmad Fatoum
2026-10-02 11:40 ` Ahmad Fatoum [this message]
2026-10-02 11:40 ` [PATCH 5/6] fs: ubootvarfs: reject variable names containing '=' Ahmad Fatoum
2026-10-02 11:40 ` [PATCH 6/6] fs: ubootvarfs: handle removal of variables that are still open Ahmad Fatoum

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=20261002114253.2906535-5-a.fatoum@pengutronix.de \
    --to=a.fatoum@pengutronix.de \
    --cc=a.fatoum@barebox.org \
    --cc=barebox@lists.infradead.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