[PATCH] md: Improve locking around error handling

The error handling routines don't use proper locking, and so two concurrent
errors could trigger a problem.

So:
- use test-and-set and test-and-clear to synchonise
the In_sync bits with the ->degraded count
- use the spinlock to protect updates to the
degraded count (could use an atomic_t but that
would be a bigger change in code, and isn't
really justified)
- remove un-necessary locking in raid5

Signed-off-by: Neil Brown <[e-mail]>
Signed-off-by: Andrew Morton <[e-mail]>
Signed-off-by: Linus Torvalds <[e-mail]>

drivers/md/raid1.c

···
 		 * normal single drive
 		 */
 		return;
-	if (test_bit(In_sync, &rdev->flags)) {
+	if (test_and_clear_bit(In_sync, &rdev->flags)) {
+		unsigned long flags;
+		spin_lock_irqsave(&conf->device_lock, flags);
 		mddev->degraded++;
+		spin_unlock_irqrestore(&conf->device_lock, flags);
 		/*
 		 * if recovery is running, make sure it aborts.
 		 */
 		set_bit(MD_RECOVERY_ERR, &mddev->recovery);
 	}
-	clear_bit(In_sync, &rdev->flags);
 	set_bit(Faulty, &rdev->flags);
 	set_bit(MD_CHANGE_DEVS, &mddev->flags);
 	printk(KERN_ALERT "raid1: Disk failure on %s, disabling device. \n"
···
 		mdk_rdev_t *rdev = conf->mirrors[i].rdev;
 		if (rdev
 		    && !test_bit(Faulty, &rdev->flags)
-		    && !test_bit(In_sync, &rdev->flags)) {
+		    && !test_and_set_bit(In_sync, &rdev->flags)) {
+			unsigned long flags;
+			spin_lock_irqsave(&conf->device_lock, flags);
 			mddev->degraded--;
-			set_bit(In_sync, &rdev->flags);
+			spin_unlock_irqrestore(&conf->device_lock, flags);
 		}
 	}
 
···
 	mirror_info_t *newmirrors;
 	conf_t *conf = mddev_to_conf(mddev);
 	int cnt, raid_disks;
-
+	unsigned long flags;
 	int d, d2;
 
 	/* Cannot change chunk_size, layout, or level */
···
 	kfree(conf->poolinfo);
 	conf->poolinfo = newpoolinfo;
 
+	spin_lock_irqsave(&conf->device_lock, flags);
 	mddev->degraded += (raid_disks - conf->raid_disks);
+	spin_unlock_irqrestore(&conf->device_lock, flags);
 	conf->raid_disks = mddev->raid_disks = raid_disks;
 	mddev->delta_disks = 0;
 

drivers/md/raid10.c

···
 		 * really dead" tests...
 		 */
 		return;
-	if (test_bit(In_sync, &rdev->flags)) {
+	if (test_and_clear_bit(In_sync, &rdev->flags)) {
+		unsigned long flags;
+		spin_lock_irqsave(&conf->device_lock, flags);
 		mddev->degraded++;
+		spin_unlock_irqrestore(&conf->device_lock, flags);
 		/*
 		 * if recovery is running, make sure it aborts.
 		 */
 		set_bit(MD_RECOVERY_ERR, &mddev->recovery);
 	}
-	clear_bit(In_sync, &rdev->flags);
 	set_bit(Faulty, &rdev->flags);
 	set_bit(MD_CHANGE_DEVS, &mddev->flags);
 	printk(KERN_ALERT "raid10: Disk failure on %s, disabling device. \n"
···
 		tmp = conf->mirrors + i;
 		if (tmp->rdev
 		    && !test_bit(Faulty, &tmp->rdev->flags)
-		    && !test_bit(In_sync, &tmp->rdev->flags)) {
+		    && !test_and_set_bit(In_sync, &tmp->rdev->flags)) {
+			unsigned long flags;
+			spin_lock_irqsave(&conf->device_lock, flags);
 			mddev->degraded--;
-			set_bit(In_sync, &tmp->rdev->flags);
+			spin_unlock_irqrestore(&conf->device_lock, flags);
 		}
 	}
 

drivers/md/raid5.c

···
  	struct stripe_head *sh = bi->bi_private;
 	raid5_conf_t *conf = sh->raid_conf;
 	int disks = sh->disks, i;
-	unsigned long flags;
 	int uptodate = test_bit(BIO_UPTODATE, &bi->bi_flags);
 
 	if (bi->bi_size)
···
 		return 0;
 	}
 
-	spin_lock_irqsave(&conf->device_lock, flags);
 	if (!uptodate)
 		md_error(conf->mddev, conf->disks[i].rdev);
 
···
 	
 	clear_bit(R5_LOCKED, &sh->dev[i].flags);
 	set_bit(STRIPE_HANDLE, &sh->state);
-	__release_stripe(conf, sh);
-	spin_unlock_irqrestore(&conf->device_lock, flags);
+	release_stripe(sh);
 	return 0;
 }
 
···
 
 	if (!test_bit(Faulty, &rdev->flags)) {
 		set_bit(MD_CHANGE_DEVS, &mddev->flags);
-		if (test_bit(In_sync, &rdev->flags)) {
+		if (test_and_clear_bit(In_sync, &rdev->flags)) {
+			unsigned long flags;
+			spin_lock_irqsave(&conf->device_lock, flags);
 			mddev->degraded++;
-			clear_bit(In_sync, &rdev->flags);
+			spin_unlock_irqrestore(&conf->device_lock, flags);
 			/*
 			 * if recovery was running, make sure it aborts.
 			 */
···
 		tmp = conf->disks + i;
 		if (tmp->rdev
 		    && !test_bit(Faulty, &tmp->rdev->flags)
-		    && !test_bit(In_sync, &tmp->rdev->flags)) {
+		    && !test_and_set_bit(In_sync, &tmp->rdev->flags)) {
+			unsigned long flags;
+			spin_lock_irqsave(&conf->device_lock, flags);
 			mddev->degraded--;
-			set_bit(In_sync, &tmp->rdev->flags);
+			spin_unlock_irqrestore(&conf->device_lock, flags);
 		}
 	}
 	print_raid5_conf(conf);
···
 	struct list_head *rtmp;
 	int spares = 0;
 	int added_devices = 0;
+	unsigned long flags;
 
 	if (mddev->degraded ||
 	    test_bit(MD_RECOVERY_RUNNING, &mddev->recovery))
···
 				break;
 		}
 
+	spin_lock_irqsave(&conf->device_lock, flags);
 	mddev->degraded = (conf->raid_disks - conf->previous_raid_disks) - added_devices;
+	spin_unlock_irqrestore(&conf->device_lock, flags);
 	mddev->raid_disks = conf->raid_disks;
 	mddev->reshape_position = 0;
 	set_bit(MD_CHANGE_DEVS, &mddev->flags);