1 files changed, 63 insertions, 42 deletions

diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
index ca5b0e8ba707..24b32a0c95b4 100644
--- a/drivers/md/raid5.c
+++ b/drivers/md/raid5.c
@@ -3740,7 +3740,7 @@ static int want_replace(struct stripe_head *sh, int disk_idx)
 	    && !test_bit(Faulty, &rdev->flags)
 	    && !test_bit(In_sync, &rdev->flags)
 	    && (rdev->recovery_offset <= sh->sector
-		|| rdev->mddev->recovery_cp <= sh->sector))
+		|| rdev->mddev->resync_offset <= sh->sector))
 		rv = 1;
 	return rv;
 }
@@ -3832,7 +3832,7 @@ static int need_this_block(struct stripe_head *sh, struct stripe_head_state *s,
 	 * is missing/faulty, then we need to read everything we can.
 	 */
 	if (!force_rcw &&
-	    sh->sector < sh->raid_conf->mddev->recovery_cp)
+	    sh->sector < sh->raid_conf->mddev->resync_offset)
 		/* reconstruct-write isn't being forced */
 		return 0;
 	for (i = 0; i < s->failed && i < 2; i++) {
@@ -4097,7 +4097,8 @@ static int handle_stripe_dirtying(struct r5conf *conf,
 				  int disks)
 {
 	int rmw = 0, rcw = 0, i;
-	sector_t recovery_cp = conf->mddev->recovery_cp;
+	struct mddev *mddev = conf->mddev;
+	sector_t resync_offset = mddev->resync_offset;
 
 	/* Check whether resync is now happening or should start.
 	 * If yes, then the array is dirty (after unclean shutdown or
@@ -4107,15 +4108,21 @@ static int handle_stripe_dirtying(struct r5conf *conf,
 	 * generate correct data from the parity.
 	 */
 	if (conf->rmw_level == PARITY_DISABLE_RMW ||
-	    (recovery_cp < MaxSector && sh->sector >= recovery_cp &&
+	    (resync_offset < MaxSector && sh->sector >= resync_offset &&
 	     s->failed == 0)) {
 		/* Calculate the real rcw later - for now make it
 		 * look like rcw is cheaper
 		 */
 		rcw = 1; rmw = 2;
-		pr_debug("force RCW rmw_level=%u, recovery_cp=%llu sh->sector=%llu\n",
-			 conf->rmw_level, (unsigned long long)recovery_cp,
+		pr_debug("force RCW rmw_level=%u, resync_offset=%llu sh->sector=%llu\n",
+			 conf->rmw_level, (unsigned long long)resync_offset,
 			 (unsigned long long)sh->sector);
+	} else if (mddev->bitmap_ops && mddev->bitmap_ops->blocks_synced &&
+		   !mddev->bitmap_ops->blocks_synced(mddev, sh->sector)) {
+		/* The initial recover is not done, must read everything */
+		rcw = 1; rmw = 2;
+		pr_debug("force RCW by lazy recovery, sh->sector=%llu\n",
+			 sh->sector);
 	} else for (i = disks; i--; ) {
 		/* would I have to read this buffer for read_modify_write */
 		struct r5dev *dev = &sh->dev[i];
@@ -4148,7 +4155,7 @@ static int handle_stripe_dirtying(struct r5conf *conf,
 	set_bit(STRIPE_HANDLE, &sh->state);
 	if ((rmw < rcw || (rmw == rcw && conf->rmw_level == PARITY_PREFER_RMW)) && rmw > 0) {
 		/* prefer read-modify-write, but need to get some data */
-		mddev_add_trace_msg(conf->mddev, "raid5 rmw %llu %d",
+		mddev_add_trace_msg(mddev, "raid5 rmw %llu %d",
 				sh->sector, rmw);
 
 		for (i = disks; i--; ) {
@@ -4227,8 +4234,8 @@ static int handle_stripe_dirtying(struct r5conf *conf,
 					set_bit(STRIPE_DELAYED, &sh->state);
 			}
 		}
-		if (rcw && !mddev_is_dm(conf->mddev))
-			blk_add_trace_msg(conf->mddev->gendisk->queue,
+		if (rcw && !mddev_is_dm(mddev))
+			blk_add_trace_msg(mddev->gendisk->queue,
 				"raid5 rcw %llu %d %d %d",
 				(unsigned long long)sh->sector, rcw, qread,
 				test_bit(STRIPE_DELAYED, &sh->state));
@@ -4698,10 +4705,21 @@ static void analyse_stripe(struct stripe_head *sh, struct stripe_head_state *s)
 			}
 		} else if (test_bit(In_sync, &rdev->flags))
 			set_bit(R5_Insync, &dev->flags);
-		else if (sh->sector + RAID5_STRIPE_SECTORS(conf) <= rdev->recovery_offset)
-			/* in sync if before recovery_offset */
-			set_bit(R5_Insync, &dev->flags);
-		else if (test_bit(R5_UPTODATE, &dev->flags) &&
+		else if (sh->sector + RAID5_STRIPE_SECTORS(conf) <=
+			 rdev->recovery_offset) {
+			/*
+			 * in sync if:
+			 *  - normal IO, or
+			 *  - resync IO that is not lazy recovery
+			 *
+			 * For lazy recovery, we have to mark the rdev without
+			 * In_sync as failed, to build initial xor data.
+			 */
+			if (!test_bit(STRIPE_SYNCING, &sh->state) ||
+			    !test_bit(MD_RECOVERY_LAZY_RECOVER,
+				      &conf->mddev->recovery))
+				set_bit(R5_Insync, &dev->flags);
+		} else if (test_bit(R5_UPTODATE, &dev->flags) &&
 			 test_bit(R5_Expanded, &dev->flags))
 			/* If we've reshaped into here, we assume it is Insync.
 			 * We will shortly update recovery_offset to make
@@ -4770,14 +4788,14 @@ static void analyse_stripe(struct stripe_head *sh, struct stripe_head_state *s)
 	if (test_bit(STRIPE_SYNCING, &sh->state)) {
 		/* If there is a failed device being replaced,
 		 *     we must be recovering.
-		 * else if we are after recovery_cp, we must be syncing
+		 * else if we are after resync_offset, we must be syncing
 		 * else if MD_RECOVERY_REQUESTED is set, we also are syncing.
 		 * else we can only be replacing
 		 * sync and recovery both need to read all devices, and so
 		 * use the same flag.
 		 */
 		if (do_recovery ||
-		    sh->sector >= conf->mddev->recovery_cp ||
+		    sh->sector >= conf->mddev->resync_offset ||
 		    test_bit(MD_RECOVERY_REQUESTED, &(conf->mddev->recovery)))
 			s->syncing = 1;
 		else
@@ -5468,17 +5486,17 @@ static int raid5_read_one_chunk(struct mddev *mddev, struct bio *raid_bio)
 
 static struct bio *chunk_aligned_read(struct mddev *mddev, struct bio *raid_bio)
 {
-	struct bio *split;
 	sector_t sector = raid_bio->bi_iter.bi_sector;
 	unsigned chunk_sects = mddev->chunk_sectors;
 	unsigned sectors = chunk_sects - (sector & (chunk_sects-1));
 
 	if (sectors < bio_sectors(raid_bio)) {
 		struct r5conf *conf = mddev->private;
-		split = bio_split(raid_bio, sectors, GFP_NOIO, &conf->bio_split);
-		bio_chain(split, raid_bio);
-		submit_bio_noacct(raid_bio);
-		raid_bio = split;
+
+		raid_bio = bio_submit_split_bioset(raid_bio, sectors,
+						   &conf->bio_split);
+		if (!raid_bio)
+			return NULL;
 	}
 
 	if (!raid5_read_one_chunk(mddev, raid_bio))
@@ -6492,11 +6510,12 @@ static inline sector_t raid5_sync_request(struct mddev *mddev, sector_t sector_n
 		}
 
 		if (mddev->curr_resync < max_sector) /* aborted */
-			mddev->bitmap_ops->end_sync(mddev, mddev->curr_resync,
-						    &sync_blocks);
+			md_bitmap_end_sync(mddev, mddev->curr_resync,
+					   &sync_blocks);
 		else /* completed sync */
 			conf->fullsync = 0;
-		mddev->bitmap_ops->close_sync(mddev);
+		if (md_bitmap_enabled(mddev, false))
+			mddev->bitmap_ops->close_sync(mddev);
 
 		return 0;
 	}
@@ -6525,8 +6544,7 @@ static inline sector_t raid5_sync_request(struct mddev *mddev, sector_t sector_n
 	}
 	if (!test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery) &&
 	    !conf->fullsync &&
-	    !mddev->bitmap_ops->start_sync(mddev, sector_nr, &sync_blocks,
-					   true) &&
+	    !md_bitmap_start_sync(mddev, sector_nr, &sync_blocks, true) &&
 	    sync_blocks >= RAID5_STRIPE_SECTORS(conf)) {
 		/* we can skip this block, and probably more */
 		do_div(sync_blocks, RAID5_STRIPE_SECTORS(conf));
@@ -6535,7 +6553,8 @@ static inline sector_t raid5_sync_request(struct mddev *mddev, sector_t sector_n
 		return sync_blocks * RAID5_STRIPE_SECTORS(conf);
 	}
 
-	mddev->bitmap_ops->cond_end_sync(mddev, sector_nr, false);
+	if (md_bitmap_enabled(mddev, false))
+		mddev->bitmap_ops->cond_end_sync(mddev, sector_nr, false);
 
 	sh = raid5_get_active_stripe(conf, NULL, sector_nr,
 				     R5_GAS_NOBLOCK);
@@ -6557,9 +6576,7 @@ static inline sector_t raid5_sync_request(struct mddev *mddev, sector_t sector_n
 			still_degraded = true;
 	}
 
-	mddev->bitmap_ops->start_sync(mddev, sector_nr, &sync_blocks,
-				      still_degraded);
-
+	md_bitmap_start_sync(mddev, sector_nr, &sync_blocks, still_degraded);
 	set_bit(STRIPE_SYNC_REQUESTED, &sh->state);
 	set_bit(STRIPE_HANDLE, &sh->state);
 
@@ -6763,7 +6780,8 @@ static void raid5d(struct md_thread *thread)
 			/* Now is a good time to flush some bitmap updates */
 			conf->seq_flush++;
 			spin_unlock_irq(&conf->device_lock);
-			mddev->bitmap_ops->unplug(mddev, true);
+			if (md_bitmap_enabled(mddev, true))
+				mddev->bitmap_ops->unplug(mddev, true);
 			spin_lock_irq(&conf->device_lock);
 			conf->seq_write = conf->seq_flush;
 			activate_bit_delay(conf, conf->temp_inactive_list);
@@ -7732,6 +7750,7 @@ static int raid5_set_limits(struct mddev *mddev)
 	lim.features |= BLK_FEAT_RAID_PARTIAL_STRIPES_EXPENSIVE;
 	lim.discard_granularity = stripe;
 	lim.max_write_zeroes_sectors = 0;
+	lim.max_hw_wzeroes_unmap_sectors = 0;
 	mddev_stack_rdev_limits(mddev, &lim, 0);
 	rdev_for_each(rdev, mddev)
 		queue_limits_stack_bdev(&lim, rdev->bdev, rdev->new_data_offset,
@@ -7780,7 +7799,7 @@ static int raid5_run(struct mddev *mddev)
 	int first = 1;
 	int ret = -EIO;
 
-	if (mddev->recovery_cp != MaxSector)
+	if (mddev->resync_offset != MaxSector)
 		pr_notice("md/raid:%s: not clean -- starting background reconstruction\n",
 			  mdname(mddev));
 
@@ -7921,7 +7940,7 @@ static int raid5_run(struct mddev *mddev)
 				mdname(mddev));
 			mddev->ro = 1;
 			set_disk_ro(mddev->gendisk, 1);
-		} else if (mddev->recovery_cp == MaxSector)
+		} else if (mddev->resync_offset == MaxSector)
 			set_bit(MD_JOURNAL_CLEAN, &mddev->flags);
 	}
 
@@ -7988,7 +8007,7 @@ static int raid5_run(struct mddev *mddev)
 	mddev->resync_max_sectors = mddev->dev_sectors;
 
 	if (mddev->degraded > dirty_parity_disks &&
-	    mddev->recovery_cp != MaxSector) {
+	    mddev->resync_offset != MaxSector) {
 		if (test_bit(MD_HAS_PPL, &mddev->flags))
 			pr_crit("md/raid:%s: starting dirty degraded array with PPL.\n",
 				mdname(mddev));
@@ -8312,7 +8331,6 @@ static int raid5_resize(struct mddev *mddev, sector_t sectors)
 	 */
 	sector_t newsize;
 	struct r5conf *conf = mddev->private;
-	int ret;
 
 	if (raid5_has_log(conf) || raid5_has_ppl(conf))
 		return -EINVAL;
@@ -8322,14 +8340,17 @@ static int raid5_resize(struct mddev *mddev, sector_t sectors)
 	    mddev->array_sectors > newsize)
 		return -EINVAL;
 
-	ret = mddev->bitmap_ops->resize(mddev, sectors, 0, false);
-	if (ret)
-		return ret;
+	if (md_bitmap_enabled(mddev, false)) {
+		int ret = mddev->bitmap_ops->resize(mddev, sectors, 0);
+
+		if (ret)
+			return ret;
+	}
 
 	md_set_array_sectors(mddev, newsize);
 	if (sectors > mddev->dev_sectors &&
-	    mddev->recovery_cp > mddev->dev_sectors) {
-		mddev->recovery_cp = mddev->dev_sectors;
+	    mddev->resync_offset > mddev->dev_sectors) {
+		mddev->resync_offset = mddev->dev_sectors;
 		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
 	}
 	mddev->dev_sectors = sectors;
@@ -8423,7 +8444,7 @@ static int raid5_start_reshape(struct mddev *mddev)
 		return -EINVAL;
 
 	/* raid5 can't handle concurrent reshape and recovery */
-	if (mddev->recovery_cp < MaxSector)
+	if (mddev->resync_offset < MaxSector)
 		return -EBUSY;
 	for (i = 0; i < conf->raid_disks; i++)
 		if (conf->disks[i].replacement)
@@ -8648,7 +8669,7 @@ static void *raid45_takeover_raid0(struct mddev *mddev, int level)
 	mddev->raid_disks += 1;
 	mddev->delta_disks = 1;
 	/* make sure it will be not marked as dirty */
-	mddev->recovery_cp = MaxSector;
+	mddev->resync_offset = MaxSector;
 
 	return setup_conf(mddev);
 }
@@ -9040,7 +9061,7 @@ static int __init raid5_init(void)
 	int ret;
 
 	raid5_wq = alloc_workqueue("raid5wq",
-		WQ_UNBOUND|WQ_MEM_RECLAIM|WQ_CPU_INTENSIVE|WQ_SYSFS, 0);
+		WQ_UNBOUND|WQ_MEM_RECLAIM|WQ_SYSFS, 0);
 	if (!raid5_wq)
 		return -ENOMEM;