[SCSI] libsas: fix sas_unregister_ports vs sas_drain_work

We need to hold drain_mutex across the unregistration as port down events
queue device removal as chained events, so we need to make sure no other
drainers are active.

[ 1118.673968] WARNING: at kernel/workqueue.c:996 __queue_work+0x11a/0x326()
[ 1118.681982] Hardware name: S2600CP
[ 1118.686193] Modules linked in: isci(-) libsas scsi_transport_sas nls_utf8
ipv6 uinput sg iTCO_wdt iTCO_vendor_support i2c_i801 i2c_core ioatdma dca
sd_mod sr_mod cdrom ahci libahci libata [last unloaded: scsi_transport_sas]
[ 1118.709893] Pid: 6831, comm: rmmod Not tainted 3.2.0-isci+ #1
[ 1118.716727] Call Trace:
[ 1118.719867]  [<ffffffff8103e9f5>] warn_slowpath_common+0x85/0x9d
[ 1118.727000]  [<ffffffff8103ea27>] warn_slowpath_null+0x1a/0x1c
[ 1118.733942]  [<ffffffff81056d44>] __queue_work+0x11a/0x326
[ 1118.740481]  [<ffffffff81056f99>] queue_work_on+0x1b/0x22
[ 1118.746925]  [<ffffffff81057106>] queue_work+0x37/0x3e
[ 1118.753105]  [<ffffffffa0120e05>] ? sas_discover_event+0x55/0x82 [libsas]
[ 1118.761094]  [<ffffffff813217c3>] scsi_queue_work+0x42/0x44
[ 1118.767717]  [<ffffffffa0120e19>] sas_discover_event+0x69/0x82 [libsas]
[ 1118.775509]  [<ffffffffa0120f5b>] sas_unregister_dev+0xc3/0xcc [libsas]
[ 1118.783319]  [<ffffffffa0120fae>] sas_unregister_domain_devices+0x4a/0xc8 [libsas]
[ 1118.792731]  [<ffffffffa0120071>] sas_deform_port+0x60/0x1a6 [libsas]
[ 1118.800339]  [<ffffffffa01201ea>] sas_unregister_ports+0x33/0x44 [libsas]
[ 1118.808342]  [<ffffffffa011f7e5>] sas_unregister_ha+0x41/0x6b [libsas]
[ 1118.816055]  [<ffffffffa0134055>] isci_unregister+0x22/0x4d [isci]
[ 1118.823384]  [<ffffffffa0143040>] isci_pci_remove+0x2e/0x60 [isci]

Reported-by: Jacek Danecki <jacek.danecki@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>

drivers/scsi/libsas/sas_event.c

@@ -51,15 +51,11 @@ static void sas_queue_event(int event, unsigned long *pending,
 	}
 }
 
-int sas_drain_work(struct sas_ha_struct *ha)
+
+void __sas_drain_work(struct sas_ha_struct *ha)
 {
 	struct workqueue_struct *wq = ha->core.shost->work_q;
 	struct work_struct *w, *_w;
-	int err;
-
-	err = mutex_lock_interruptible(&ha->drain_mutex);
-	if (err)
-		return err;
 
 	set_bit(SAS_HA_DRAINING, &ha->state);
 	/* flush submitters */
@@ -75,6 +71,17 @@ int sas_drain_work(struct sas_ha_struct *ha)
 		sas_queue_work(ha, w);
 	}
 	spin_unlock_irq(&ha->state_lock);
+}
+
+int sas_drain_work(struct sas_ha_struct *ha)
+{
+	int err;
+
+	err = mutex_lock_interruptible(&ha->drain_mutex);
+	if (err)
+		return err;
+	if (test_bit(SAS_HA_REGISTERED, &ha->state))
+		__sas_drain_work(ha);
 	mutex_unlock(&ha->drain_mutex);
 
 	return 0;

drivers/scsi/libsas/sas_init.c

@@ -160,18 +160,22 @@ Undo_phys:
 
 int sas_unregister_ha(struct sas_ha_struct *sas_ha)
 {
-	unsigned long flags;
-
 	/* Set the state to unregistered to avoid further unchained
-	 * events to be queued
+	 * events to be queued, and flush any in-progress drainers
 	 */
-	spin_lock_irqsave(&sas_ha->state_lock, flags);
+	mutex_lock(&sas_ha->drain_mutex);
+	spin_lock_irq(&sas_ha->state_lock);
 	clear_bit(SAS_HA_REGISTERED, &sas_ha->state);
-	spin_unlock_irqrestore(&sas_ha->state_lock, flags);
-	sas_drain_work(sas_ha);
+	spin_unlock_irq(&sas_ha->state_lock);
+	__sas_drain_work(sas_ha);
+	mutex_unlock(&sas_ha->drain_mutex);
 
 	sas_unregister_ports(sas_ha);
-	sas_drain_work(sas_ha);
+
+	/* flush unregistration work */
+	mutex_lock(&sas_ha->drain_mutex);
+	__sas_drain_work(sas_ha);
+	mutex_unlock(&sas_ha->drain_mutex);
 
 	if (sas_ha->lldd_max_execute_num > 1) {
 		sas_shutdown_queue(sas_ha);

drivers/scsi/libsas/sas_internal.h

@@ -70,6 +70,7 @@ int  sas_init_events(struct sas_ha_struct *sas_ha);
 void sas_shutdown_queue(struct sas_ha_struct *sas_ha);
 void sas_disable_revalidation(struct sas_ha_struct *ha);
 void sas_enable_revalidation(struct sas_ha_struct *ha);
+void __sas_drain_work(struct sas_ha_struct *ha);
 
 void sas_deform_port(struct asd_sas_phy *phy, int gone);