Merge tag 'md/3.12-fixes' of git://neil.brown.name/md

Pull md bugfixes from Neil Brown:
 "Assorted md bug-fixes for 3.12.

  All tagged for -stable releases too"

* tag 'md/3.12-fixes' of git://neil.brown.name/md:
  raid5: avoid finding "discard" stripe
  raid5: set bio bi_vcnt 0 for discard request
  md: avoid deadlock when md_set_badblocks.
  md: Fix skipping recovery for read-only arrays.

drivers/md/md.c

@@ -8111,6 +8111,7 @@ static int md_set_badblocks(struct badblocks *bb, sector_t s, int sectors,
 	u64 *p;
 	int lo, hi;
 	int rv = 1;
+	unsigned long flags;
 
 	if (bb->shift < 0)
 		/* badblocks are disabled */
@@ -8125,7 +8126,7 @@ static int md_set_badblocks(struct badblocks *bb, sector_t s, int sectors,
 		sectors = next - s;
 	}
 
-	write_seqlock_irq(&bb->lock);
+	write_seqlock_irqsave(&bb->lock, flags);
 
 	p = bb->page;
 	lo = 0;
@@ -8241,7 +8242,7 @@ static int md_set_badblocks(struct badblocks *bb, sector_t s, int sectors,
 	bb->changed = 1;
 	if (!acknowledged)
 		bb->unacked_exist = 1;
-	write_sequnlock_irq(&bb->lock);
+	write_sequnlock_irqrestore(&bb->lock, flags);
 
 	return rv;
 }

drivers/md/raid1.c

@@ -1479,6 +1479,7 @@ static int raid1_spare_active(struct mddev *mddev)
 			}
 		}
 		if (rdev
+		    && rdev->recovery_offset == MaxSector
 		    && !test_bit(Faulty, &rdev->flags)
 		    && !test_and_set_bit(In_sync, &rdev->flags)) {
 			count++;

drivers/md/raid10.c

@@ -1782,6 +1782,7 @@ static int raid10_spare_active(struct mddev *mddev)
 			}
 			sysfs_notify_dirent_safe(tmp->replacement->sysfs_state);
 		} else if (tmp->rdev
+			   && tmp->rdev->recovery_offset == MaxSector
 			   && !test_bit(Faulty, &tmp->rdev->flags)
 			   && !test_and_set_bit(In_sync, &tmp->rdev->flags)) {
 			count++;

drivers/md/raid5.c

@@ -778,6 +778,12 @@ static void ops_run_io(struct stripe_head *sh, struct stripe_head_state *s)
 			bi->bi_io_vec[0].bv_len = STRIPE_SIZE;
 			bi->bi_io_vec[0].bv_offset = 0;
 			bi->bi_size = STRIPE_SIZE;
+			/*
+			 * If this is discard request, set bi_vcnt 0. We don't
+			 * want to confuse SCSI because SCSI will replace payload
+			 */
+			if (rw & REQ_DISCARD)
+				bi->bi_vcnt = 0;
 			if (rrdev)
 				set_bit(R5_DOUBLE_LOCKED, &sh->dev[i].flags);
 
@@ -816,6 +822,12 @@ static void ops_run_io(struct stripe_head *sh, struct stripe_head_state *s)
 			rbi->bi_io_vec[0].bv_len = STRIPE_SIZE;
 			rbi->bi_io_vec[0].bv_offset = 0;
 			rbi->bi_size = STRIPE_SIZE;
+			/*
+			 * If this is discard request, set bi_vcnt 0. We don't
+			 * want to confuse SCSI because SCSI will replace payload
+			 */
+			if (rw & REQ_DISCARD)
+				rbi->bi_vcnt = 0;
 			if (conf->mddev->gendisk)
 				trace_block_bio_remap(bdev_get_queue(rbi->bi_bdev),
 						      rbi, disk_devt(conf->mddev->gendisk),
@@ -2910,6 +2922,14 @@ static void handle_stripe_clean_event(struct r5conf *conf,
 		}
 		/* now that discard is done we can proceed with any sync */
 		clear_bit(STRIPE_DISCARD, &sh->state);
+		/*
+		 * SCSI discard will change some bio fields and the stripe has
+		 * no updated data, so remove it from hash list and the stripe
+		 * will be reinitialized
+		 */
+		spin_lock_irq(&conf->device_lock);
+		remove_hash(sh);
+		spin_unlock_irq(&conf->device_lock);
 		if (test_bit(STRIPE_SYNC_REQUESTED, &sh->state))
 			set_bit(STRIPE_HANDLE, &sh->state);