[PATCH] afs: Fix uninit var in afs_alloc_anon_key()

[PATCH] afs: Fix uninit var in afs_alloc_anon_key()

[David Howells]

Fix an uninitialised variable (key) in afs_alloc_anon_key() by setting it
to cell->anonymous_key.  Without this change, the error check may return a
false failure with a bad error number.

Most of the time this is unlikely to happen because the first encounter
with afs_alloc_anon_key() will usually be from (auto)mount, for which all
subsequent operations must wait - apart from other (auto)mounts.  Once the
call->anonymous_key is allocated, all further calls to afs_request_key()
will skip the call to afs_alloc_anon_key() for that cell.

Fixes: d27c71257825 ("afs: Fix delayed allocation of a cell's anonymous key")
Reported-by: Paulo Alcantra <pc@manguebit.org>
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Paulo Alcantara <pc@manguebit.org>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: syzbot+41c68824eefb67cdf00c@syzkaller.appspotmail.com
cc: linux-afs@lists.infradead.org
cc: linux-fsdevel@vger.kernel.org
---
 fs/afs/security.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/fs/afs/security.c b/fs/afs/security.c
index ff8830e6982f..55ddce94af03 100644
--- a/fs/afs/security.c
+++ b/fs/afs/security.c
@@ -26,7 +26,8 @@ static int afs_alloc_anon_key(struct afs_cell *cell)
 	struct key *key;
 
 	mutex_lock(&afs_key_lock);
-	if (!cell->anonymous_key) {
+	key = cell->anonymous_key;
+	if (!key) {
 		key = rxrpc_get_null_key(cell->key_desc);
 		if (!IS_ERR(key))
 			cell->anonymous_key = key;

Re: [PATCH] afs: Fix uninit var in afs_alloc_anon_key()

[Linus Torvalds]

I've applied this directly, since it seems to be a recent regression
and final 6.18 is days away.

And it seems like an obvious fix. Famous last words...

            Linus

On Fri, 28 Nov 2025 at 16:40, David Howells <dhowells@redhat.com> wrote:
>
> Fix an uninitialised variable (key) in afs_alloc_anon_key() by setting it
> to cell->anonymous_key.  Without this change, the error check may return a
> false failure with a bad error number.