Merge tag 'md-3.11' of git://neil.brown.name/md

Pull md updates from NeilBrown:
 "Mostly fixes, with a few minor features (eg 'last_sync_action' sysfs
  file)

  A couple marked for -stable including one recent bug which causes a
  RAID10 reshape to complete without moving any data :-(

  A couple more bugfixes (at least) to come, but haven't confirmed the
  right solution yet."

* tag 'md-3.11' of git://neil.brown.name/md:
  md/raid10: fix bug which causes all RAID10 reshapes to move no data.
  md/raid5: allow 5-device RAID6 to be reshaped to 4-device.
  md/raid10: fix two bugs affecting RAID10 reshape.
  md: remove doubled description for sync_max, merging it within sync_min/sync_max
  MD: Remember the last sync operation that was performed
  md: fix buglet in RAID5 -> RAID0 conversion.
  md/raid10: check In_sync flag in 'enough()'.
  md/raid10: locking changes for 'enough()'.
  md: replace strict_strto*() with kstrto*()
  md: Wait for md_check_recovery before attempting device removal.
  dm-raid: silence compiler warning on rebuilds_per_group.
  DM RAID: Fix raid_resume not reviving failed devices in all cases
  DM RAID: Break-up untidy function
  DM RAID: Add ability to restore transiently failed devices on resume

Author: torvalds

Documentation/device-mapper/dm-raid.txt

@@ -222,3 +222,5 @@ Version History
 1.4.2   Add RAID10 "far" and "offset" algorithm support.
 1.5.0   Add message interface to allow manipulation of the sync_action.
 	New status (STATUSTYPE_INFO) fields: sync_action and mismatch_cnt.
+1.5.1   Add ability to restore transiently failed devices on resume.
+1.5.2   'mismatch_cnt' is zero unless [last_]sync_action is "check".

Documentation/md.txt

@@ -566,13 +566,6 @@ also have
      when it reaches the current sync_max (below) and possibly at
      other times.
 
-   sync_max
-     This is a number of sectors at which point a resync/recovery
-     process will pause.  When a resync is active, the value can
-     only ever be increased, never decreased.  The value of 'max'
-     effectively disables the limit.
-
-
    sync_speed
      This shows the current actual speed, in K/sec, of the current
      sync_action.  It is averaged over the last 30 seconds.
@@ -593,6 +586,12 @@ also have
      that number to reach sync_max.  Then you can either increase
      "sync_max", or can write 'idle' to "sync_action".
 
+     The value of 'max' for "sync_max" effectively disables the limit.
+     When a resync is active, the value can only ever be increased,
+     never decreased.
+     The value of '0' is the minimum for "sync_min".
+
+
 
 Each active md device may also have attributes specific to the
 personality module that manages it.

drivers/md/bitmap.c

@@ -2002,9 +2002,9 @@ location_store(struct mddev *mddev, const char *buf, size_t len)
 		} else {
 			int rv;
 			if (buf[0] == '+')
-				rv = strict_strtoll(buf+1, 10, &offset);
+				rv = kstrtoll(buf+1, 10, &offset);
 			else
-				rv = strict_strtoll(buf, 10, &offset);
+				rv = kstrtoll(buf, 10, &offset);
 			if (rv)
 				return rv;
 			if (offset == 0)
@@ -2139,7 +2139,7 @@ static ssize_t
 backlog_store(struct mddev *mddev, const char *buf, size_t len)
 {
 	unsigned long backlog;
-	int rv = strict_strtoul(buf, 10, &backlog);
+	int rv = kstrtoul(buf, 10, &backlog);
 	if (rv)
 		return rv;
 	if (backlog > COUNTER_MAX)
@@ -2165,7 +2165,7 @@ chunksize_store(struct mddev *mddev, const char *buf, size_t len)
 	unsigned long csize;
 	if (mddev->bitmap)
 		return -EBUSY;
-	rv = strict_strtoul(buf, 10, &csize);
+	rv = kstrtoul(buf, 10, &csize);
 	if (rv)
 		return rv;
 	if (csize < 512 ||

drivers/md/dm-raid.c

@@ -380,7 +380,7 @@ static int validate_region_size(struct raid_set *rs, unsigned long region_size)
 static int validate_raid_redundancy(struct raid_set *rs)
 {
 	unsigned i, rebuild_cnt = 0;
-	unsigned rebuilds_per_group, copies, d;
+	unsigned rebuilds_per_group = 0, copies, d;
 	unsigned group_size, last_group_start;
 
 	for (i = 0; i < rs->md.raid_disks; i++)
@@ -504,7 +504,7 @@ static int parse_raid_params(struct raid_set *rs, char **argv,
 	 * First, parse the in-order required arguments
 	 * "chunk_size" is the only argument of this type.
 	 */
-	if ((strict_strtoul(argv[0], 10, &value) < 0)) {
+	if ((kstrtoul(argv[0], 10, &value) < 0)) {
 		rs->ti->error = "Bad chunk size";
 		return -EINVAL;
 	} else if (rs->raid_type->level == 1) {
@@ -585,7 +585,7 @@ static int parse_raid_params(struct raid_set *rs, char **argv,
 			continue;
 		}
 
-		if (strict_strtoul(argv[i], 10, &value) < 0) {
+		if (kstrtoul(argv[i], 10, &value) < 0) {
 			rs->ti->error = "Bad numerical argument given in raid params";
 			return -EINVAL;
 		}
@@ -1181,7 +1181,7 @@ static int raid_ctr(struct dm_target *ti, unsigned argc, char **argv)
 	argv++;
 
 	/* number of RAID parameters */
-	if (strict_strtoul(argv[0], 10, &num_raid_params) < 0) {
+	if (kstrtoul(argv[0], 10, &num_raid_params) < 0) {
 		ti->error = "Cannot understand number of RAID parameters";
 		return -EINVAL;
 	}
@@ -1194,7 +1194,7 @@ static int raid_ctr(struct dm_target *ti, unsigned argc, char **argv)
 		return -EINVAL;
 	}
 
-	if ((strict_strtoul(argv[num_raid_params], 10, &num_raid_devs) < 0) ||
+	if ((kstrtoul(argv[num_raid_params], 10, &num_raid_devs) < 0) ||
 	    (num_raid_devs >= INT_MAX)) {
 		ti->error = "Cannot understand number of raid devices";
 		return -EINVAL;
@@ -1388,6 +1388,7 @@ static void raid_status(struct dm_target *ti, status_type_t type,
 		 *   performing a "check" of the array.
 		 */
 		DMEMIT(" %llu",
+		       (strcmp(rs->md.last_sync_action, "check")) ? 0 :
 		       (unsigned long long)
 		       atomic64_read(&rs->md.resync_mismatches));
 		break;
@@ -1572,6 +1573,62 @@ static void raid_postsuspend(struct dm_target *ti)
 	mddev_suspend(&rs->md);
 }
 
+static void attempt_restore_of_faulty_devices(struct raid_set *rs)
+{
+	int i;
+	uint64_t failed_devices, cleared_failed_devices = 0;
+	unsigned long flags;
+	struct dm_raid_superblock *sb;
+	struct md_rdev *r;
+
+	for (i = 0; i < rs->md.raid_disks; i++) {
+		r = &rs->dev[i].rdev;
+		if (test_bit(Faulty, &r->flags) && r->sb_page &&
+		    sync_page_io(r, 0, r->sb_size, r->sb_page, READ, 1)) {
+			DMINFO("Faulty %s device #%d has readable super block."
+			       "  Attempting to revive it.",
+			       rs->raid_type->name, i);
+
+			/*
+			 * Faulty bit may be set, but sometimes the array can
+			 * be suspended before the personalities can respond
+			 * by removing the device from the array (i.e. calling
+			 * 'hot_remove_disk').  If they haven't yet removed
+			 * the failed device, its 'raid_disk' number will be
+			 * '>= 0' - meaning we must call this function
+			 * ourselves.
+			 */
+			if ((r->raid_disk >= 0) &&
+			    (r->mddev->pers->hot_remove_disk(r->mddev, r) != 0))
+				/* Failed to revive this device, try next */
+				continue;
+
+			r->raid_disk = i;
+			r->saved_raid_disk = i;
+			flags = r->flags;
+			clear_bit(Faulty, &r->flags);
+			clear_bit(WriteErrorSeen, &r->flags);
+			clear_bit(In_sync, &r->flags);
+			if (r->mddev->pers->hot_add_disk(r->mddev, r)) {
+				r->raid_disk = -1;
+				r->saved_raid_disk = -1;
+				r->flags = flags;
+			} else {
+				r->recovery_offset = 0;
+				cleared_failed_devices |= 1 << i;
+			}
+		}
+	}
+	if (cleared_failed_devices) {
+		rdev_for_each(r, &rs->md) {
+			sb = page_address(r->sb_page);
+			failed_devices = le64_to_cpu(sb->failed_devices);
+			failed_devices &= ~cleared_failed_devices;
+			sb->failed_devices = cpu_to_le64(failed_devices);
+		}
+	}
+}
+
 static void raid_resume(struct dm_target *ti)
 {
 	struct raid_set *rs = ti->private;
@@ -1580,6 +1637,13 @@ static void raid_resume(struct dm_target *ti)
 	if (!rs->bitmap_loaded) {
 		bitmap_load(&rs->md);
 		rs->bitmap_loaded = 1;
+	} else {
+		/*
+		 * A secondary resume while the device is active.
+		 * Take this opportunity to check whether any failed
+		 * devices are reachable again.
+		 */
+		attempt_restore_of_faulty_devices(rs);
 	}
 
 	clear_bit(MD_RECOVERY_FROZEN, &rs->md.recovery);
@@ -1588,7 +1652,7 @@ static void raid_resume(struct dm_target *ti)
 
 static struct target_type raid_target = {
 	.name = "raid",
-	.version = {1, 5, 0},
+	.version = {1, 5, 2},
 	.module = THIS_MODULE,
 	.ctr = raid_ctr,
 	.dtr = raid_dtr,

drivers/md/md.c

@@ -521,6 +521,7 @@ void mddev_init(struct mddev *mddev)
 	init_waitqueue_head(&mddev->recovery_wait);
 	mddev->reshape_position = MaxSector;
 	mddev->reshape_backwards = 0;
+	mddev->last_sync_action = "none";
 	mddev->resync_min = 0;
 	mddev->resync_max = MaxSector;
 	mddev->level = LEVEL_NONE;
@@ -2867,7 +2868,7 @@ static ssize_t
 offset_store(struct md_rdev *rdev, const char *buf, size_t len)
 {
 	unsigned long long offset;
-	if (strict_strtoull(buf, 10, &offset) < 0)
+	if (kstrtoull(buf, 10, &offset) < 0)
 		return -EINVAL;
 	if (rdev->mddev->pers && rdev->raid_disk >= 0)
 		return -EBUSY;
@@ -2895,7 +2896,7 @@ static ssize_t new_offset_store(struct md_rdev *rdev,
 	unsigned long long new_offset;
 	struct mddev *mddev = rdev->mddev;
 
-	if (strict_strtoull(buf, 10, &new_offset) < 0)
+	if (kstrtoull(buf, 10, &new_offset) < 0)
 		return -EINVAL;
 
 	if (mddev->sync_thread)
@@ -2961,7 +2962,7 @@ static int strict_blocks_to_sectors(const char *buf, sector_t *sectors)
 	unsigned long long blocks;
 	sector_t new;
 
-	if (strict_strtoull(buf, 10, &blocks) < 0)
+	if (kstrtoull(buf, 10, &blocks) < 0)
 		return -EINVAL;
 
 	if (blocks & 1ULL << (8 * sizeof(blocks) - 1))
@@ -3069,7 +3070,7 @@ static ssize_t recovery_start_store(struct md_rdev *rdev, const char *buf, size_
 
 	if (cmd_match(buf, "none"))
 		recovery_start = MaxSector;
-	else if (strict_strtoull(buf, 10, &recovery_start))
+	else if (kstrtoull(buf, 10, &recovery_start))
 		return -EINVAL;
 
 	if (rdev->mddev->pers &&
@@ -3497,7 +3498,7 @@ level_store(struct mddev *mddev, const char *buf, size_t len)
 	if (clevel[len-1] == '\n')
 		len--;
 	clevel[len] = 0;
-	if (strict_strtol(clevel, 10, &level))
+	if (kstrtol(clevel, 10, &level))
 		level = LEVEL_NONE;
 
 	if (request_module("md-%s", clevel) != 0)
@@ -4272,6 +4273,17 @@ action_store(struct mddev *mddev, const char *page, size_t len)
 	return len;
 }
 
+static struct md_sysfs_entry md_scan_mode =
+__ATTR(sync_action, S_IRUGO|S_IWUSR, action_show, action_store);
+
+static ssize_t
+last_sync_action_show(struct mddev *mddev, char *page)
+{
+	return sprintf(page, "%s\n", mddev->last_sync_action);
+}
+
+static struct md_sysfs_entry md_last_scan_mode = __ATTR_RO(last_sync_action);
+
 static ssize_t
 mismatch_cnt_show(struct mddev *mddev, char *page)
 {
@@ -4280,10 +4292,6 @@ mismatch_cnt_show(struct mddev *mddev, char *page)
 		       atomic64_read(&mddev->resync_mismatches));
 }
 
-static struct md_sysfs_entry md_scan_mode =
-__ATTR(sync_action, S_IRUGO|S_IWUSR, action_show, action_store);
-
-
 static struct md_sysfs_entry md_mismatches = __ATTR_RO(mismatch_cnt);
 
 static ssize_t
@@ -4356,7 +4364,7 @@ sync_force_parallel_store(struct mddev *mddev, const char *buf, size_t len)
 {
 	long n;
 
-	if (strict_strtol(buf, 10, &n))
+	if (kstrtol(buf, 10, &n))
 		return -EINVAL;
 
 	if (n != 0 && n != 1)
@@ -4424,7 +4432,7 @@ static ssize_t
 min_sync_store(struct mddev *mddev, const char *buf, size_t len)
 {
 	unsigned long long min;
-	if (strict_strtoull(buf, 10, &min))
+	if (kstrtoull(buf, 10, &min))
 		return -EINVAL;
 	if (min > mddev->resync_max)
 		return -EINVAL;
@@ -4461,7 +4469,7 @@ max_sync_store(struct mddev *mddev, const char *buf, size_t len)
 		mddev->resync_max = MaxSector;
 	else {
 		unsigned long long max;
-		if (strict_strtoull(buf, 10, &max))
+		if (kstrtoull(buf, 10, &max))
 			return -EINVAL;
 		if (max < mddev->resync_min)
 			return -EINVAL;
@@ -4686,6 +4694,7 @@ static struct attribute *md_default_attrs[] = {
 
 static struct attribute *md_redundancy_attrs[] = {
 	&md_scan_mode.attr,
+	&md_last_scan_mode.attr,
 	&md_mismatches.attr,
 	&md_sync_min.attr,
 	&md_sync_max.attr,
@@ -6405,6 +6414,12 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode,
 		/* need to ensure md_delayed_delete() has completed */
 		flush_workqueue(md_misc_wq);
 
+	if (cmd == HOT_REMOVE_DISK)
+		/* need to ensure recovery thread has run */
+		wait_event_interruptible_timeout(mddev->sb_wait,
+						 !test_bit(MD_RECOVERY_NEEDED,
+							   &mddev->flags),
+						 msecs_to_jiffies(5000));
 	err = mddev_lock(mddev);
 	if (err) {
 		printk(KERN_INFO 
@@ -7323,7 +7338,7 @@ void md_do_sync(struct md_thread *thread)
 	sector_t last_check;
 	int skipped = 0;
 	struct md_rdev *rdev;
-	char *desc;
+	char *desc, *action = NULL;
 	struct blk_plug plug;
 
 	/* just incase thread restarts... */
@@ -7333,17 +7348,21 @@ void md_do_sync(struct md_thread *thread)
 		return;
 
 	if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery)) {
-		if (test_bit(MD_RECOVERY_CHECK, &mddev->recovery))
+		if (test_bit(MD_RECOVERY_CHECK, &mddev->recovery)) {
 			desc = "data-check";
-		else if (test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery))
+			action = "check";
+		} else if (test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) {
 			desc = "requested-resync";
-		else
+			action = "repair";
+		} else
 			desc = "resync";
 	} else if (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery))
 		desc = "reshape";
 	else
 		desc = "recovery";
 
+	mddev->last_sync_action = action ?: desc;
+
 	/* we overload curr_resync somewhat here.
 	 * 0 == not engaged in resync at all
 	 * 2 == checking that there is no conflict with another sync
@@ -7892,6 +7911,8 @@ void md_check_recovery(struct mddev *mddev)
 			md_new_event(mddev);
 		}
 	unlock:
+		wake_up(&mddev->sb_wait);
+
 		if (!mddev->sync_thread) {
 			clear_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
 			if (test_and_clear_bit(MD_RECOVERY_RECOVER,

drivers/md/md.h

@@ -268,6 +268,14 @@ struct mddev {
 
 	struct md_thread		*thread;	/* management thread */
 	struct md_thread		*sync_thread;	/* doing resync or reconstruct */
+
+	/* 'last_sync_action' is initialized to "none".  It is set when a
+	 * sync operation (i.e "data-check", "requested-resync", "resync",
+	 * "recovery", or "reshape") is started.  It holds this value even
+	 * when the sync thread is "frozen" (interrupted) or "idle" (stopped
+	 * or finished).  It is overwritten when a new sync operation is begun.
+	 */
+	char				*last_sync_action;
 	sector_t			curr_resync;	/* last block scheduled */
 	/* As resync requests can complete out of order, we cannot easily track
 	 * how much resync has been completed.  So we occasionally pause until

drivers/md/raid0.c

@@ -597,6 +597,7 @@ static void *raid0_takeover_raid45(struct mddev *mddev)
 			       mdname(mddev));
 			return ERR_PTR(-EINVAL);
 		}
+		rdev->sectors = mddev->dev_sectors;
 	}
 
 	/* Set new parameters */