5 files changed, 43 insertions, 16 deletions

diff --git a/drivers/md/md.c b/drivers/md/md.c
index 295be1a68806..51c19f86ff99 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -5434,7 +5434,7 @@ void md_done_sync(mddev_t *mddev, int blocks, int ok)
         atomic_sub(blocks, &mddev->recovery_active);
         wake_up(&mddev->recovery_wait);
         if (!ok) {
-                set_bit(MD_RECOVERY_ERR, &mddev->recovery);
+                set_bit(MD_RECOVERY_INTR, &mddev->recovery);
                 md_wakeup_thread(mddev->thread);
                 // stop recovery, signal do_sync ....
         }
@@ -5690,7 +5690,7 @@ void md_do_sync(mddev_t *mddev)
                 sectors = mddev->pers->sync_request(mddev, j, &skipped,
                                                   currspeed < speed_min(mddev));
                 if (sectors == 0) {
-                        set_bit(MD_RECOVERY_ERR, &mddev->recovery);
+                        set_bit(MD_RECOVERY_INTR, &mddev->recovery);
                         goto out;
                 }
 
@@ -5713,8 +5713,7 @@ void md_do_sync(mddev_t *mddev)
 
                 last_check = io_sectors;
 
-                if (test_bit(MD_RECOVERY_INTR, &mddev->recovery) ||
-                    test_bit(MD_RECOVERY_ERR, &mddev->recovery))
+                if (test_bit(MD_RECOVERY_INTR, &mddev->recovery))
                         break;
 
         repeat:
@@ -5768,8 +5767,7 @@ void md_do_sync(mddev_t *mddev)
         /* tell personality that we are finished */
         mddev->pers->sync_request(mddev, max_sectors, &skipped, 1);
 
-        if (!test_bit(MD_RECOVERY_ERR, &mddev->recovery) &&
-            !test_bit(MD_RECOVERY_CHECK, &mddev->recovery) &&
+        if (!test_bit(MD_RECOVERY_CHECK, &mddev->recovery) &&
             mddev->curr_resync > 2) {
                 if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery)) {
                         if (test_bit(MD_RECOVERY_INTR, &mddev->recovery)) {
@@ -5838,7 +5836,10 @@ static int remove_and_add_spares(mddev_t *mddev)
                 }
 
         if (mddev->degraded) {
-                rdev_for_each(rdev, rtmp, mddev)
+                rdev_for_each(rdev, rtmp, mddev) {
+                        if (rdev->raid_disk >= 0 &&
+                            !test_bit(In_sync, &rdev->flags))
+                                spares++;
                         if (rdev->raid_disk < 0
                             && !test_bit(Faulty, &rdev->flags)) {
                                 rdev->recovery_offset = 0;
@@ -5856,6 +5857,7 @@ static int remove_and_add_spares(mddev_t *mddev)
                                 } else
                                         break;
                         }
+                }
         }
         return spares;
 }
@@ -5869,7 +5871,7 @@ static int remove_and_add_spares(mddev_t *mddev)
  * to do that as needed.
  * When it is determined that resync is needed, we set MD_RECOVERY_RUNNING in
  * "->recovery" and create a thread at ->sync_thread.
- * When the thread finishes it sets MD_RECOVERY_DONE (and might set MD_RECOVERY_ERR)
+ * When the thread finishes it sets MD_RECOVERY_DONE
  * and wakeups up this thread which will reap the thread and finish up.
  * This thread also removes any faulty devices (with nr_pending == 0).
  *
@@ -5944,8 +5946,7 @@ void md_check_recovery(mddev_t *mddev)
                         /* resync has finished, collect result */
                         md_unregister_thread(mddev->sync_thread);
                         mddev->sync_thread = NULL;
-                        if (!test_bit(MD_RECOVERY_ERR, &mddev->recovery) &&
-                            !test_bit(MD_RECOVERY_INTR, &mddev->recovery)) {
+                        if (!test_bit(MD_RECOVERY_INTR, &mddev->recovery)) {
                                 /* success...*/
                                 /* activate any spares */
                                 mddev->pers->spare_active(mddev);
@@ -5969,7 +5970,6 @@ void md_check_recovery(mddev_t *mddev)
                  * might be left set
                  */
                 clear_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
-                clear_bit(MD_RECOVERY_ERR, &mddev->recovery);
                 clear_bit(MD_RECOVERY_INTR, &mddev->recovery);
                 clear_bit(MD_RECOVERY_DONE, &mddev->recovery);
 
diff --git a/drivers/md/multipath.c b/drivers/md/multipath.c
index 4f4d1f383842..e968116e0de9 100644
--- a/drivers/md/multipath.c
+++ b/drivers/md/multipath.c
@@ -327,7 +327,8 @@ static int multipath_remove_disk(mddev_t *mddev, int number)
         if (rdev) {
                 if (test_bit(In_sync, &rdev->flags) ||
                     atomic_read(&rdev->nr_pending)) {
-                        printk(KERN_ERR "hot-remove-disk, slot %d is identified"                                " but is still operational!\n", number);
+                        printk(KERN_ERR "hot-remove-disk, slot %d is identified"
+                               " but is still operational!\n", number);
                         err = -EBUSY;
                         goto abort;
                 }
diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index d0f4021bbc2e..c610b947218a 100644
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -1027,7 +1027,7 @@ static void error(mddev_t *mddev, mdk_rdev_t *rdev)
                 /*
                  * if recovery is running, make sure it aborts.
                  */
-                set_bit(MD_RECOVERY_ERR, &mddev->recovery);
+                set_bit(MD_RECOVERY_INTR, &mddev->recovery);
         } else
                 set_bit(Faulty, &rdev->flags);
         set_bit(MD_CHANGE_DEVS, &mddev->flags);
@@ -1148,6 +1148,14 @@ static int raid1_remove_disk(mddev_t *mddev, int number)
                         err = -EBUSY;
                         goto abort;
                 }
+                /* Only remove non-faulty devices is recovery
+                 * is not possible.
+                 */
+                if (!test_bit(Faulty, &rdev->flags) &&
+                    mddev->degraded < conf->raid_disks) {
+                        err = -EBUSY;
+                        goto abort;
+                }
                 p->rdev = NULL;
                 synchronize_rcu();
                 if (atomic_read(&rdev->nr_pending)) {
diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index 8536ede1e712..1de17da34a95 100644
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -1020,7 +1020,7 @@ static void error(mddev_t *mddev, mdk_rdev_t *rdev)
                 /*
                  * if recovery is running, make sure it aborts.
                  */
-                set_bit(MD_RECOVERY_ERR, &mddev->recovery);
+                set_bit(MD_RECOVERY_INTR, &mddev->recovery);
         }
         set_bit(Faulty, &rdev->flags);
         set_bit(MD_CHANGE_DEVS, &mddev->flags);
@@ -1171,6 +1171,14 @@ static int raid10_remove_disk(mddev_t *mddev, int number)
                         err = -EBUSY;
                         goto abort;
                 }
+                /* Only remove faulty devices in recovery
+                 * is not possible.
+                 */
+                if (!test_bit(Faulty, &rdev->flags) &&
+                    enough(conf)) {
+                        err = -EBUSY;
+                        goto abort;
+                }
                 p->rdev = NULL;
                 synchronize_rcu();
                 if (atomic_read(&rdev->nr_pending)) {
@@ -1237,6 +1245,7 @@ static void end_sync_write(struct bio *bio, int error)
 
         if (!uptodate)
                 md_error(mddev, conf->mirrors[d].rdev);
+
         update_head_pos(i, r10_bio);
 
         while (atomic_dec_and_test(&r10_bio->remaining)) {
@@ -1844,7 +1853,8 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i
                                         if (rb2)
                                                 atomic_dec(&rb2->remaining);
                                         r10_bio = rb2;
-                                        if (!test_and_set_bit(MD_RECOVERY_ERR, &mddev->recovery))
+                                        if (!test_and_set_bit(MD_RECOVERY_INTR,
+                                                              &mddev->recovery))
                                                 printk(KERN_INFO "raid10: %s: insufficient working devices for recovery.\n",
                                                        mdname(mddev));
                                         break;
diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index 2f28745dacf9..425958a76b84 100644
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -1268,7 +1268,7 @@ static void error(mddev_t *mddev, mdk_rdev_t *rdev)
                         /*
                          * if recovery was running, make sure it aborts.
                          */
-                        set_bit(MD_RECOVERY_ERR, &mddev->recovery);
+                        set_bit(MD_RECOVERY_INTR, &mddev->recovery);
                 }
                 set_bit(Faulty, &rdev->flags);
                 printk (KERN_ALERT
@@ -4574,6 +4574,14 @@ static int raid5_remove_disk(mddev_t *mddev, int number)
                         err = -EBUSY;
                         goto abort;
                 }
+                /* Only remove non-faulty devices if recovery
+                 * isn't possible.
+                 */
+                if (!test_bit(Faulty, &rdev->flags) &&
+                    mddev->degraded <= conf->max_degraded) {
+                        err = -EBUSY;
+                        goto abort;
+                }
                 p->rdev = NULL;
                 synchronize_rcu();
                 if (atomic_read(&rdev->nr_pending)) {