From: Joseph Qi <joseph.qi@linux.alibaba.com>
To: Sasha Levin <sashal@kernel.org>, Chuck Lever <cel@kernel.org>,
Jeff Layton <jlayton@kernel.org>
Cc: stable@vger.kernel.org, linux-nfs@vger.kernel.org,
linux-kernel@vger.kernel.org,
Joseph Qi <joseph.qi@linux.alibaba.com>
Subject: [PATCH 6.6.y] nfsd: fix NULL dereference in nfsd4_free_layout_stateid()
Date: Mon, 9 Nov 2026 16:49:32 +0800 [thread overview]
Message-ID: <20261109084932.3861080-1-joseph.qi@linux.alibaba.com> (raw)
Commit d12af460df27 ("nfsd: fix BUG_ON in nfsd4_alloc_layout_stateid on
racing delegation revoke") replaced BUG_ON(!ls->ls_file) with a bail-out
that calls nfs4_put_stid() on the partially initialized layout stateid.
nfs4_alloc_stid() leaves sc_count at 1 and nothing takes another
reference before that bail-out, so the put runs sc_free, which is
nfsd4_free_layout_stateid(). That function uses ls->ls_file
unconditionally, both for vfs_setlease() and for nfsd_file_put(), and
nfsd_file_put() does not accept NULL either. The new error path thus
turns the race it was meant to handle gracefully into a NULL pointer
dereference. The nfsd4_layout_setlease() failure path that already
called nfs4_put_stid() was safe only because the BUG_ON guaranteed
ls_file was non-NULL.
Upstream does not have this problem because its destructor goes through
nfsd4_close_layout(), which tests the pointer before using it. That
helper came in with the admin-revoke series, which is not in 6.6, so
just add the NULL test here.
Fixes: d12af460df27 ("nfsd: fix BUG_ON in nfsd4_alloc_layout_stateid on racing delegation revoke")
Signed-off-by: Joseph Qi <joseph.qi@linux.alibaba.com>
---
fs/nfsd/nfs4layouts.c | 10 +++++++---
1 file changed, 7 insertions(+), 3 deletions(-)
diff --git a/fs/nfsd/nfs4layouts.c b/fs/nfsd/nfs4layouts.c
index bea7513a0de90..7c3f0948a610c 100644
--- a/fs/nfsd/nfs4layouts.c
+++ b/fs/nfsd/nfs4layouts.c
@@ -168,9 +168,13 @@ nfsd4_free_layout_stateid(struct nfs4_stid *stid)
list_del_init(&ls->ls_perfile);
spin_unlock(&fp->fi_lock);
- if (!nfsd4_layout_ops[ls->ls_layout_type]->disable_recalls)
- vfs_setlease(ls->ls_file->nf_file, F_UNLCK, NULL, (void **)&ls);
- nfsd_file_put(ls->ls_file);
+ /* ls_file is NULL if nfsd4_alloc_layout_stateid() bailed out early */
+ if (ls->ls_file) {
+ if (!nfsd4_layout_ops[ls->ls_layout_type]->disable_recalls)
+ vfs_setlease(ls->ls_file->nf_file, F_UNLCK, NULL,
+ (void **)&ls);
+ nfsd_file_put(ls->ls_file);
+ }
if (ls->ls_recalled)
atomic_dec(&ls->ls_stid.sc_file->fi_lo_recalls);
--
2.39.3
reply other threads:[~2026-11-09 8:49 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=20261109084932.3861080-1-joseph.qi@linux.alibaba.com \
--to=joseph.qi@linux.alibaba.com \
--cc=cel@kernel.org \
--cc=jlayton@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-nfs@vger.kernel.org \
--cc=sashal@kernel.org \
--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
all inboxes | Powered by JetHome®