scsi: libfc: hold disc_mutex in fc_disc_stop_rports()

fc_disc_stop_rports() is calling fc_rport_logoff(), which in turn is
acquiring the rport mutex. So we cannot use RCU list traversal here, but
rather need to hold the disc mutex to avoid list corruption while
traversing.

Fixes: a407c593398c ("scsi: libfc: Fixup disc_mutex handling")
Signed-off-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

diff --git a/drivers/scsi/libfc/fc_disc.c b/drivers/scsi/libfc/fc_disc.c
index c1756b9..f969a71 100644 (file)
--- a/drivers/scsi/libfc/fc_disc.c
+++ b/drivers/scsi/libfc/fc_disc.c
@@ -62,20 +62,16 @@ static void fc_disc_restart(struct fc_disc *);
  */
 static void fc_disc_stop_rports(struct fc_disc *disc)
 {
-       struct fc_lport *lport;
        struct fc_rport_priv *rdata;
 
-       lport = fc_disc_lport(disc);
-       lockdep_assert_held(&lport->lp_mutex);
+       lockdep_assert_held(&disc->disc_mutex);
 
-       rcu_read_lock();
-       list_for_each_entry_rcu(rdata, &disc->rports, peers) {
+       list_for_each_entry(rdata, &disc->rports, peers) {
                if (kref_get_unless_zero(&rdata->kref)) {
                        fc_rport_logoff(rdata);
                        kref_put(&rdata->kref, fc_rport_destroy);
                }
        }
-       rcu_read_unlock();
 }
 
 /**
@@ -699,7 +695,9 @@ static void fc_disc_stop(struct fc_lport *lport)
 
        if (disc->pending)
                cancel_delayed_work_sync(&disc->disc_work);
+       mutex_lock(&disc->disc_mutex);
        fc_disc_stop_rports(disc);
+       mutex_unlock(&disc->disc_mutex);
 }
 
 /**