Merge tag 'ceph-for-6.10-rc8' of https://github.com/ceph/ceph-client

Pull ceph fixes from Ilya Dryomov:
 "A fix for a possible use-after-free following "rbd unmap" or "umount"
  marked for stable and two kernel-doc fixups"

* tag 'ceph-for-6.10-rc8' of https://github.com/ceph/ceph-client:
  libceph: fix crush_choose_firstn() kernel-doc warnings
  libceph: suppress crush_choose_indep() kernel-doc warnings
  libceph: fix race between delayed_work() and ceph_monc_stop()

2 files changed, 17 insertions, 4 deletions

diff --git a/net/ceph/crush/mapper.c b/net/ceph/crush/mapper.c
index 1daf95e17d67..3a5bd1cd1e99 100644
--- a/net/ceph/crush/mapper.c
+++ b/net/ceph/crush/mapper.c
@@ -429,7 +429,10 @@ static int is_out(const struct crush_map *map,
 /**
  * crush_choose_firstn - choose numrep distinct items of given type
  * @map: the crush_map
+ * @work: working space initialized by crush_init_workspace()
  * @bucket: the bucket we are choose an item from
+ * @weight: weight vector (for map leaves)
+ * @weight_max: size of weight vector
  * @x: crush input value
  * @numrep: the number of items to choose
  * @type: the type of item to choose
@@ -445,6 +448,7 @@ static int is_out(const struct crush_map *map,
  * @vary_r: pass r to recursive calls
  * @out2: second output vector for leaf items (if @recurse_to_leaf)
  * @parent_r: r value passed from the parent
+ * @choose_args: weights and ids for each known bucket
  */
 static int crush_choose_firstn(const struct crush_map *map,
 			       struct crush_work *work,
@@ -636,9 +640,8 @@ reject:
 }
 
 
-/**
+/*
  * crush_choose_indep: alternative breadth-first positionally stable mapping
- *
  */
 static void crush_choose_indep(const struct crush_map *map,
 			       struct crush_work *work,
diff --git a/net/ceph/mon_client.c b/net/ceph/mon_client.c
index f263f7e91a21..ab66b599ac47 100644
--- a/net/ceph/mon_client.c
+++ b/net/ceph/mon_client.c
@@ -1085,13 +1085,19 @@ static void delayed_work(struct work_struct *work)
 	struct ceph_mon_client *monc =
 		container_of(work, struct ceph_mon_client, delayed_work.work);
 
-	dout("monc delayed_work\n");
 	mutex_lock(&monc->mutex);
+	dout("%s mon%d\n", __func__, monc->cur_mon);
+	if (monc->cur_mon < 0) {
+		goto out;
+	}
+
 	if (monc->hunting) {
 		dout("%s continuing hunt\n", __func__);
 		reopen_session(monc);
 	} else {
 		int is_auth = ceph_auth_is_authenticated(monc->auth);
+
+		dout("%s is_authed %d\n", __func__, is_auth);
 		if (ceph_con_keepalive_expired(&monc->con,
 					       CEPH_MONC_PING_TIMEOUT)) {
 			dout("monc keepalive timeout\n");
@@ -1116,6 +1122,8 @@ static void delayed_work(struct work_struct *work)
 		}
 	}
 	__schedule_delayed(monc);
+
+out:
 	mutex_unlock(&monc->mutex);
 }
 
@@ -1232,13 +1240,15 @@ EXPORT_SYMBOL(ceph_monc_init);
 void ceph_monc_stop(struct ceph_mon_client *monc)
 {
 	dout("stop\n");
-	cancel_delayed_work_sync(&monc->delayed_work);
 
 	mutex_lock(&monc->mutex);
 	__close_session(monc);
+	monc->hunting = false;
 	monc->cur_mon = -1;
 	mutex_unlock(&monc->mutex);
 
+	cancel_delayed_work_sync(&monc->delayed_work);
+
 	/*
 	 * flush msgr queue before we destroy ourselves to ensure that:
 	 *  - any work that references our embedded con is finished.