Commit 6f8d0c77 authored by NeilBrown's avatar NeilBrown
Browse files

md: make error_handler functions more uniform and correct. 



- there is no need to test_bit Faulty, as that was already done in
  md_error which is the only caller of these functions.
- MD_CHANGE_DEVS should be set *after* faulty is set to ensure
  metadata is updated correctly.
- spinlock should be held while updating ->degraded.

Signed-off-by: default avatarNeilBrown <neilb@suse.de>
parent 92f861a7

drivers/md/multipath.c

+22 −18
Original line number Original line Diff line number Diff line
@@ -186,6 +186,7 @@ static int multipath_congested(void *data, int bits) 
static void multipath_error (mddev_t *mddev, mdk_rdev_t *rdev)

static void multipath_error (mddev_t *mddev, mdk_rdev_t *rdev)

{

{

	multipath_conf_t *conf = mddev->private;

	multipath_conf_t *conf = mddev->private;

	char b[BDEVNAME_SIZE];





	if (conf->raid_disks - mddev->degraded <= 1) {

	if (conf->raid_disks - mddev->degraded <= 1) {

		/*

		/*
@@ -196,16 +197,19 @@ static void multipath_error (mddev_t *mddev, mdk_rdev_t *rdev) 
		printk(KERN_ALERT 

		printk(KERN_ALERT 

		       "multipath: only one IO path left and IO error.\n");

		       "multipath: only one IO path left and IO error.\n");

		/* leave it active... it's all we have */

		/* leave it active... it's all we have */

	} else {

		return;

	}

	/*

	/*

	 * Mark disk as unusable

	 * Mark disk as unusable

	 */

	 */

		if (!test_bit(Faulty, &rdev->flags)) {

	if (test_and_clear_bit(In_sync, &rdev->flags)) {

			char b[BDEVNAME_SIZE];

		unsigned long flags;

			clear_bit(In_sync, &rdev->flags);

		spin_lock_irqsave(&conf->device_lock, flags);

		mddev->degraded++;

		spin_unlock_irqrestore(&conf->device_lock, flags);

	}

	set_bit(Faulty, &rdev->flags);

	set_bit(Faulty, &rdev->flags);

	set_bit(MD_CHANGE_DEVS, &mddev->flags);

	set_bit(MD_CHANGE_DEVS, &mddev->flags);

			mddev->degraded++;

	printk(KERN_ALERT "multipath: IO failure on %s,"

	printk(KERN_ALERT "multipath: IO failure on %s,"

	       " disabling IO path.\n"

	       " disabling IO path.\n"

	       "multipath: Operation continuing"

	       "multipath: Operation continuing"
@@ -213,8 +217,6 @@ static void multipath_error (mddev_t *mddev, mdk_rdev_t *rdev) 
	       bdevname(rdev->bdev, b),

	       bdevname(rdev->bdev, b),

	       conf->raid_disks - mddev->degraded);

	       conf->raid_disks - mddev->degraded);

}

}

	}

}





static void print_multipath_conf (multipath_conf_t *conf)

static void print_multipath_conf (multipath_conf_t *conf)

{

{
@@ -273,9 +275,11 @@ static int multipath_add_disk(mddev_t *mddev, mdk_rdev_t *rdev) 
							   PAGE_CACHE_SIZE - 1);

							   PAGE_CACHE_SIZE - 1);

			}

			}





			spin_lock_irq(&conf->device_lock);

			mddev->degraded--;

			mddev->degraded--;

			rdev->raid_disk = path;

			rdev->raid_disk = path;

			set_bit(In_sync, &rdev->flags);

			set_bit(In_sync, &rdev->flags);

			spin_unlock_irq(&conf->device_lock);

			rcu_assign_pointer(p->rdev, rdev);

			rcu_assign_pointer(p->rdev, rdev);

			err = 0;

			err = 0;

			md_integrity_add_rdev(rdev, mddev);

			md_integrity_add_rdev(rdev, mddev);

drivers/md/raid5.c

+18 −20
Original line number Original line Diff line number Diff line
@@ -1700,8 +1700,6 @@ static void error(mddev_t *mddev, mdk_rdev_t *rdev) 
	raid5_conf_t *conf = mddev->private;

	raid5_conf_t *conf = mddev->private;

	pr_debug("raid456: error called\n");

	pr_debug("raid456: error called\n");





	if (!test_bit(Faulty, &rdev->flags)) {

		set_bit(MD_CHANGE_DEVS, &mddev->flags);

	if (test_and_clear_bit(In_sync, &rdev->flags)) {

	if (test_and_clear_bit(In_sync, &rdev->flags)) {

		unsigned long flags;

		unsigned long flags;

		spin_lock_irqsave(&conf->device_lock, flags);

		spin_lock_irqsave(&conf->device_lock, flags);
@@ -1713,6 +1711,7 @@ static void error(mddev_t *mddev, mdk_rdev_t *rdev) 
		set_bit(MD_RECOVERY_INTR, &mddev->recovery);

		set_bit(MD_RECOVERY_INTR, &mddev->recovery);

	}

	}

	set_bit(Faulty, &rdev->flags);

	set_bit(Faulty, &rdev->flags);

	set_bit(MD_CHANGE_DEVS, &mddev->flags);

	printk(KERN_ALERT

	printk(KERN_ALERT

	       "md/raid:%s: Disk failure on %s, disabling device.\n"

	       "md/raid:%s: Disk failure on %s, disabling device.\n"

	       "md/raid:%s: Operation continuing on %d devices.\n",

	       "md/raid:%s: Operation continuing on %d devices.\n",
@@ -1721,7 +1720,6 @@ static void error(mddev_t *mddev, mdk_rdev_t *rdev) 
	       mdname(mddev),

	       mdname(mddev),

	       conf->raid_disks - mddev->degraded);

	       conf->raid_disks - mddev->degraded);

}

}

}





/*

/*

 * Input: a 'big' sector number,

 * Input: a 'big' sector number,

CRA Git | Maintained and supported by SUSTech CRA and CCSE