scsi: libsas: Use _safe() loop in sas_resume_port()
authorDan Carpenter <dan.carpenter@oracle.com>
Wed, 19 May 2021 14:20:27 +0000 (17:20 +0300)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Thu, 3 Jun 2021 06:36:25 +0000 (08:36 +0200)
[ Upstream commit 8c7e7b8486cda21269d393245883c5e4737d5ee7 ]

If sas_notify_lldd_dev_found() fails then this code calls:

sas_unregister_dev(port, dev);

which removes "dev", our list iterator, from the list.  This could lead to
an endless loop.  We need to use list_for_each_entry_safe().

Link: https://lore.kernel.org/r/YKUeq6gwfGcvvhty@mwanda
Fixes: 303694eeee5e ("[SCSI] libsas: suspend / resume support")
Reviewed-by: John Garry <john.garry@huawei.com>
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
drivers/scsi/libsas/sas_port.c

index d3c5297c6c89e242d845d16142f7a309ed4ef773..30e0730f613e83c6e7e485351a7b954a2715fa95 100644 (file)
@@ -41,7 +41,7 @@ static bool phy_is_wideport_member(struct asd_sas_port *port, struct asd_sas_phy
 
 static void sas_resume_port(struct asd_sas_phy *phy)
 {
-       struct domain_device *dev;
+       struct domain_device *dev, *n;
        struct asd_sas_port *port = phy->port;
        struct sas_ha_struct *sas_ha = phy->ha;
        struct sas_internal *si = to_sas_internal(sas_ha->core.shost->transportt);
@@ -60,7 +60,7 @@ static void sas_resume_port(struct asd_sas_phy *phy)
         * 1/ presume every device came back
         * 2/ force the next revalidation to check all expander phys
         */
-       list_for_each_entry(dev, &port->dev_list, dev_list_node) {
+       list_for_each_entry_safe(dev, n, &port->dev_list, dev_list_node) {
                int i, rc;
 
                rc = sas_notify_lldd_dev_found(dev);