This is a note to let you know that I've just added the patch titled
cifs: fix lock ordering while disabling multichannel
to the 6.6-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary
The filename of the patch is:
cifs-fix-lock-ordering-while-disabling-multichannel.patch
and it can be found in the queue-6.6 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@xxxxxxxxxxxxxxx> know about it.
>From 5eef12c4e3230f2025dc46ad8c4a3bc19978e5d7 Mon Sep 17 00:00:00 2001
From: Shyam Prasad N <sprasad@xxxxxxxxxxxxx>
Date: Tue, 14 Nov 2023 04:58:23 +0000
Subject: cifs: fix lock ordering while disabling multichannel
From: Shyam Prasad N <sprasad@xxxxxxxxxxxxx>
commit 5eef12c4e3230f2025dc46ad8c4a3bc19978e5d7 upstream.
The code to handle the case of server disabling multichannel
was picking iface_lock with chan_lock held. This goes against
the lock ordering rules, as iface_lock is a higher order lock
(even if it isn't so obvious).
This change fixes the lock ordering by doing the following in
that order for each secondary channel:
1. store iface and server pointers in local variable
2. remove references to iface and server in channels
3. unlock chan_lock
4. lock iface_lock
5. dec ref count for iface
6. unlock iface_lock
7. dec ref count for server
8. lock chan_lock again
Since this function can only be called in smb2_reconnect, and
that cannot be called by two parallel processes, we should not
have races due to dropping chan_lock between steps 3 and 8.
Fixes: ee1d21794e55 ("cifs: handle when server stops supporting multichannel")
Reported-by: Paulo Alcantara <pc@xxxxxxxxxxxxx>
Signed-off-by: Shyam Prasad N <sprasad@xxxxxxxxxxxxx>
Signed-off-by: Steve French <stfrench@xxxxxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
fs/smb/client/sess.c | 22 +++++++++++++---------
1 file changed, 13 insertions(+), 9 deletions(-)
--- a/fs/smb/client/sess.c
+++ b/fs/smb/client/sess.c
@@ -316,28 +316,32 @@ cifs_disable_secondary_channels(struct c
iface = ses->chans[i].iface;
server = ses->chans[i].server;
+ /*
+ * remove these references first, since we need to unlock
+ * the chan_lock here, since iface_lock is a higher lock
+ */
+ ses->chans[i].iface = NULL;
+ ses->chans[i].server = NULL;
+ spin_unlock(&ses->chan_lock);
+
if (iface) {
spin_lock(&ses->iface_lock);
kref_put(&iface->refcount, release_iface);
- ses->chans[i].iface = NULL;
iface->num_channels--;
if (iface->weight_fulfilled)
iface->weight_fulfilled--;
spin_unlock(&ses->iface_lock);
}
- spin_unlock(&ses->chan_lock);
- if (server && !server->terminate) {
- server->terminate = true;
- cifs_signal_cifsd_for_reconnect(server, false);
- }
- spin_lock(&ses->chan_lock);
-
if (server) {
- ses->chans[i].server = NULL;
+ if (!server->terminate) {
+ server->terminate = true;
+ cifs_signal_cifsd_for_reconnect(server, false);
+ }
cifs_put_tcp_session(server, false);
}
+ spin_lock(&ses->chan_lock);
}
done:
Patches currently in stable-queue which might be from sprasad@xxxxxxxxxxxxx are
queue-6.6/cifs-reconnect-worker-should-take-reference-on-serve.patch
queue-6.6/cifs-handle-cases-where-a-channel-is-closed.patch
queue-6.6/cifs-update-iface_last_update-on-each-query-and-upda.patch
queue-6.6/cifs-fix-lock-ordering-while-disabling-multichannel.patch
queue-6.6/cifs-reconnect-work-should-have-reference-on-server-.patch
queue-6.6/revert-cifs-reconnect-work-should-have-reference-on-.patch
queue-6.6/cifs-handle-servers-that-still-advertise-multichanne.patch
queue-6.6/cifs-handle-when-server-starts-supporting-multichann.patch
queue-6.6/cifs-handle-when-server-stops-supporting-multichanne.patch
queue-6.6/cifs-after-disabling-multichannel-mark-tcon-for-reconnect.patch
queue-6.6/cifs-fix-a-pending-undercount-of-srv_count.patch
