ceph: check session state after bumping session->s_seq

Some messages sent by the MDS entail a session sequence number
increment, and the MDS will drop certain types of requests on the floor
when the sequence numbers don't match.

In particular, a REQUEST_CLOSE message can cross with one of the
sequence morphing messages from the MDS which can cause the client to
stall, waiting for a response that will never come.

Originally, this meant an up to 5s delay before the recurring workqueue
job kicked in and resent the request, but a recent change made it so
that the client would never resend, causing a 60s stall unmounting and
sometimes a blockisting event.

Add a new helper for incrementing the session sequence and then testing
to see whether a REQUEST_CLOSE needs to be resent, and move the handling
of CEPH_MDS_SESSION_CLOSING into that function. Change all of the
bare sequence counter increments to use the new helper.

Reorganize check_session_state with a switch statement.  It should no
longer be called when the session is CLOSING, so throw a warning if it
ever is (but still handle that case sanely).

[ idryomov: whitespace, pr_err() call fixup ]

URL: https://tracker.ceph.com/issues/47563
Fixes: fa9967734227 ("ceph: fix potential mdsc use-after-free crash")
Reported-by: Patrick Donnelly <pdonnell@redhat.com>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: Ilya Dryomov <idryomov@gmail.com>
Reviewed-by: Xiubo Li <xiubli@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

1 files changed, 35 insertions, 15 deletions

diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c
index 08f1c0c31dc2..8f1d7500a7ec 100644
--- a/fs/ceph/mds_client.c
+++ b/fs/ceph/mds_client.c
@@ -4231,7 +4231,7 @@ static void handle_lease(struct ceph_mds_client *mdsc,
 	     dname.len, dname.name);
 
 	mutex_lock(&session->s_mutex);
-	session->s_seq++;
+	inc_session_sequence(session);
 
 	if (!inode) {
 		dout("handle_lease no inode %llx\n", vino.ino);
@@ -4385,29 +4385,49 @@ static void maybe_recover_session(struct ceph_mds_client *mdsc)
 
 bool check_session_state(struct ceph_mds_session *s)
 {
-	if (s->s_state == CEPH_MDS_SESSION_CLOSING) {
-		dout("resending session close request for mds%d\n",
-				s->s_mds);
-		request_close_session(s);
-		return false;
-	}
-	if (s->s_ttl && time_after(jiffies, s->s_ttl)) {
-		if (s->s_state == CEPH_MDS_SESSION_OPEN) {
+	switch (s->s_state) {
+	case CEPH_MDS_SESSION_OPEN:
+		if (s->s_ttl && time_after(jiffies, s->s_ttl)) {
 			s->s_state = CEPH_MDS_SESSION_HUNG;
 			pr_info("mds%d hung\n", s->s_mds);
 		}
-	}
-	if (s->s_state == CEPH_MDS_SESSION_NEW ||
-	    s->s_state == CEPH_MDS_SESSION_RESTARTING ||
-	    s->s_state == CEPH_MDS_SESSION_CLOSED ||
-	    s->s_state == CEPH_MDS_SESSION_REJECTED)
-		/* this mds is failed or recovering, just wait */
+		break;
+	case CEPH_MDS_SESSION_CLOSING:
+		/* Should never reach this when we're unmounting */
+		WARN_ON_ONCE(true);
+		fallthrough;
+	case CEPH_MDS_SESSION_NEW:
+	case CEPH_MDS_SESSION_RESTARTING:
+	case CEPH_MDS_SESSION_CLOSED:
+	case CEPH_MDS_SESSION_REJECTED:
 		return false;
+	}
 
 	return true;
 }
 
 /*
+ * If the sequence is incremented while we're waiting on a REQUEST_CLOSE reply,
+ * then we need to retransmit that request.
+ */
+void inc_session_sequence(struct ceph_mds_session *s)
+{
+	lockdep_assert_held(&s->s_mutex);
+
+	s->s_seq++;
+
+	if (s->s_state == CEPH_MDS_SESSION_CLOSING) {
+		int ret;
+
+		dout("resending session close request for mds%d\n", s->s_mds);
+		ret = request_close_session(s);
+		if (ret < 0)
+			pr_err("unable to close session to mds%d: %d\n",
+			       s->s_mds, ret);
+	}
+}
+
+/*
  * delayed work -- periodically trim expired leases, renew caps with mds
  */
 static void schedule_delayed(struct ceph_mds_client *mdsc)