smb: client: let smbd_destroy() wait for SMBDIRECT_SOCKET_DISCONNECTED

We should wait for the rdma_cm to become SMBDIRECT_SOCKET_DISCONNECTED,
it turns out that (at least running some xfstests e.g. cifs/001)
often triggers the case where wait_event_interruptible() returns
with -ERESTARTSYS instead of waiting for SMBDIRECT_SOCKET_DISCONNECTED
to be reached.

Or we are already in SMBDIRECT_SOCKET_DISCONNECTING and never wait
for SMBDIRECT_SOCKET_DISCONNECTED.

Fixes: 050b8c3 ("smbd: Make upper layer decide when to destroy the transport")
Fixes: e8b3bfe ("cifs: smbd: Don't destroy transport on RDMA disconnect")
Fixes: b0aa92a ("smb: client: make sure smbd_disconnect_rdma_work() doesn't run after smbd_destroy() took over")
Cc: Steve French <[email protected]>
Cc: Tom Talpey <[email protected]>
Cc: Long Li <[email protected]>
Cc: Namjae Jeon <[email protected]>
Cc: [email protected]
Cc: [email protected]
Signed-off-by: Stefan Metzmacher <[email protected]>
Signed-off-by: Steve French <[email protected]>

Author: metze-samba

fs/smb/client/smbdirect.c

@@ -1575,12 +1575,12 @@ void smbd_destroy(struct TCP_Server_Info *server)
 	disable_work_sync(&sc->disconnect_work);
 
 	log_rdma_event(INFO, "destroying rdma session\n");
-	if (sc->status < SMBDIRECT_SOCKET_DISCONNECTING) {
+	if (sc->status < SMBDIRECT_SOCKET_DISCONNECTING)
 		smbd_disconnect_rdma_work(&sc->disconnect_work);
+	if (sc->status < SMBDIRECT_SOCKET_DISCONNECTED) {
 		log_rdma_event(INFO, "wait for transport being disconnected\n");
-		wait_event_interruptible(
-			sc->status_wait,
-			sc->status == SMBDIRECT_SOCKET_DISCONNECTED);
+		wait_event(sc->status_wait, sc->status == SMBDIRECT_SOCKET_DISCONNECTED);
+		log_rdma_event(INFO, "waited for transport being disconnected\n");
 	}
 
 	/*