Merge tag 'for-4.13/dm-fixes' of...

Merge tag 'for-4.13/dm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm

Pull device mapper fixes from Mike Snitzer:

 - a few DM integrity fixes that improve performance. One that address
   inefficiencies in the on-disk journal device layout. Another that
   makes use of the block layer's on-stack plugging when writing the
   journal.

 - a dm-bufio fix for the blk_status_t conversion that went in during
   the merge window.

 - a few DM raid fixes that address correctness when suspending the
   device and a validation fix for validation that occurs during device
   activation.

 - a couple DM zoned target fixes. Important one being the fix to not
   use GFP_KERNEL in the IO path due to concerns about deadlock in
   low-memory conditions (e.g. swap over a DM zoned device, etc).

 - a DM DAX device fix to make sure dm_dax_flush() is called if the
   underlying DAX device is operating as a write cache.

* tag 'for-4.13/dm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm:
  dm, dax: Make sure dm_dax_flush() is called if device supports it
  dm verity fec: fix GFP flags used with mempool_alloc()
  dm zoned: use GFP_NOIO in I/O path
  dm zoned: remove test for impossible REQ_OP_FLUSH conditions
  dm raid: bump target version
  dm raid: avoid mddev->suspended access
  dm raid: fix activation check in validate_raid_redundancy()
  dm raid: remove WARN_ON() in raid10_md_layout_to_format()
  dm bufio: fix error code in dm_bufio_write_dirty_buffers()
  dm integrity: test for corrupted disk format during table load
  dm integrity: WARN_ON if variables representing journal usage get out of sync
  dm integrity: use plugging when writing the journal
  dm integrity: fix inefficient allocation of journal space

Documentation/device-mapper/dm-raid.txt

@@ -343,3 +343,4 @@ Version History
 1.11.0  Fix table line argument order
 	(wrong raid10_copies/raid10_format sequence)
 1.11.1  Add raid4/5/6 journal write-back support via journal_mode option
+1.12.1  fix for MD deadlock between mddev_suspend() and md_write_start() available

drivers/dax/super.c

@@ -278,6 +278,12 @@ void dax_write_cache(struct dax_device *dax_dev, bool wc)
 }
 EXPORT_SYMBOL_GPL(dax_write_cache);
 
+bool dax_write_cache_enabled(struct dax_device *dax_dev)
+{
+	return test_bit(DAXDEV_WRITE_CACHE, &dax_dev->flags);
+}
+EXPORT_SYMBOL_GPL(dax_write_cache_enabled);
+
 bool dax_alive(struct dax_device *dax_dev)
 {
 	lockdep_assert_held(&dax_srcu);

drivers/md/dm-bufio.c

@@ -1258,8 +1258,7 @@ EXPORT_SYMBOL_GPL(dm_bufio_write_dirty_buffers_async);
  */
 int dm_bufio_write_dirty_buffers(struct dm_bufio_client *c)
 {
-	blk_status_t a;
-	int f;
+	int a, f;
 	unsigned long buffers_processed = 0;
 	struct dm_buffer *b, *tmp;
 

drivers/md/dm-integrity.c

@@ -1587,16 +1587,18 @@ static void dm_integrity_map_continue(struct dm_integrity_io *dio, bool from_map
 	if (likely(ic->mode == 'J')) {
 		if (dio->write) {
 			unsigned next_entry, i, pos;
-			unsigned ws, we;
+			unsigned ws, we, range_sectors;
 
-			dio->range.n_sectors = min(dio->range.n_sectors, ic->free_sectors);
+			dio->range.n_sectors = min(dio->range.n_sectors,
+						   ic->free_sectors << ic->sb->log2_sectors_per_block);
 			if (unlikely(!dio->range.n_sectors))
 				goto sleep;
-			ic->free_sectors -= dio->range.n_sectors;
+			range_sectors = dio->range.n_sectors >> ic->sb->log2_sectors_per_block;
+			ic->free_sectors -= range_sectors;
 			journal_section = ic->free_section;
 			journal_entry = ic->free_section_entry;
 
-			next_entry = ic->free_section_entry + dio->range.n_sectors;
+			next_entry = ic->free_section_entry + range_sectors;
 			ic->free_section_entry = next_entry % ic->journal_section_entries;
 			ic->free_section += next_entry / ic->journal_section_entries;
 			ic->n_uncommitted_sections += next_entry / ic->journal_section_entries;
@@ -1727,6 +1729,8 @@ static void pad_uncommitted(struct dm_integrity_c *ic)
 		wraparound_section(ic, &ic->free_section);
 		ic->n_uncommitted_sections++;
 	}
+	WARN_ON(ic->journal_sections * ic->journal_section_entries !=
+		(ic->n_uncommitted_sections + ic->n_committed_sections) * ic->journal_section_entries + ic->free_sectors);
 }
 
 static void integrity_commit(struct work_struct *w)
@@ -1821,6 +1825,9 @@ static void do_journal_write(struct dm_integrity_c *ic, unsigned write_start,
 {
 	unsigned i, j, n;
 	struct journal_completion comp;
+	struct blk_plug plug;
+
+	blk_start_plug(&plug);
 
 	comp.ic = ic;
 	comp.in_flight = (atomic_t)ATOMIC_INIT(1);
@@ -1945,6 +1952,8 @@ static void do_journal_write(struct dm_integrity_c *ic, unsigned write_start,
 
 	dm_bufio_write_dirty_buffers_async(ic->bufio);
 
+	blk_finish_plug(&plug);
+
 	complete_journal_op(&comp);
 	wait_for_completion_io(&comp.comp);
 
@@ -3019,6 +3028,11 @@ static int dm_integrity_ctr(struct dm_target *ti, unsigned argc, char **argv)
 		ti->error = "Block size doesn't match the information in superblock";
 		goto bad;
 	}
+	if (!le32_to_cpu(ic->sb->journal_sections)) {
+		r = -EINVAL;
+		ti->error = "Corrupted superblock, journal_sections is 0";
+		goto bad;
+	}
 	/* make sure that ti->max_io_len doesn't overflow */
 	if (ic->sb->log2_interleave_sectors < MIN_LOG2_INTERLEAVE_SECTORS ||
 	    ic->sb->log2_interleave_sectors > MAX_LOG2_INTERLEAVE_SECTORS) {

drivers/md/dm-raid.c

@@ -208,6 +208,7 @@ struct raid_dev {
 #define RT_FLAG_RS_BITMAP_LOADED	2
 #define RT_FLAG_UPDATE_SBS		3
 #define RT_FLAG_RESHAPE_RS		4
+#define RT_FLAG_RS_SUSPENDED		5
 
 /* Array elements of 64 bit needed for rebuild/failed disk bits */
 #define DISKS_ARRAY_ELEMS ((MAX_RAID_DEVICES + (sizeof(uint64_t) * 8 - 1)) / sizeof(uint64_t) / 8)
@@ -564,9 +565,10 @@ static const char *raid10_md_layout_to_format(int layout)
 	if (__raid10_near_copies(layout) > 1)
 		return "near";
 
-	WARN_ON(__raid10_far_copies(layout) < 2);
+	if (__raid10_far_copies(layout) > 1)
+		return "far";
 
-	return "far";
+	return "unknown";
 }
 
 /* Return md raid10 algorithm for @name */
@@ -2540,11 +2542,6 @@ static int analyse_superblocks(struct dm_target *ti, struct raid_set *rs)
 	if (!freshest)
 		return 0;
 
-	if (validate_raid_redundancy(rs)) {
-		rs->ti->error = "Insufficient redundancy to activate array";
-		return -EINVAL;
-	}
-
 	/*
 	 * Validation of the freshest device provides the source of
 	 * validation for the remaining devices.
@@ -2553,6 +2550,11 @@ static int analyse_superblocks(struct dm_target *ti, struct raid_set *rs)
 	if (super_validate(rs, freshest))
 		return -EINVAL;
 
+	if (validate_raid_redundancy(rs)) {
+		rs->ti->error = "Insufficient redundancy to activate array";
+		return -EINVAL;
+	}
+
 	rdev_for_each(rdev, mddev)
 		if (!test_bit(Journal, &rdev->flags) &&
 		    rdev != freshest &&
@@ -3168,6 +3170,7 @@ static int raid_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 	}
 
 	mddev_suspend(&rs->md);
+	set_bit(RT_FLAG_RS_SUSPENDED, &rs->runtime_flags);
 
 	/* Try to adjust the raid4/5/6 stripe cache size to the stripe size */
 	if (rs_is_raid456(rs)) {
@@ -3625,7 +3628,7 @@ static void raid_postsuspend(struct dm_target *ti)
 {
 	struct raid_set *rs = ti->private;
 
-	if (!rs->md.suspended)
+	if (!test_and_set_bit(RT_FLAG_RS_SUSPENDED, &rs->runtime_flags))
 		mddev_suspend(&rs->md);
 
 	rs->md.ro = 1;
@@ -3759,7 +3762,7 @@ static int rs_start_reshape(struct raid_set *rs)
 		return r;
 
 	/* Need to be resumed to be able to start reshape, recovery is frozen until raid_resume() though */
-	if (mddev->suspended)
+	if (test_and_clear_bit(RT_FLAG_RS_SUSPENDED, &rs->runtime_flags))
 		mddev_resume(mddev);
 
 	/*
@@ -3786,8 +3789,8 @@ static int rs_start_reshape(struct raid_set *rs)
 	}
 
 	/* Suspend because a resume will happen in raid_resume() */
-	if (!mddev->suspended)
-		mddev_suspend(mddev);
+	set_bit(RT_FLAG_RS_SUSPENDED, &rs->runtime_flags);
+	mddev_suspend(mddev);
 
 	/*
 	 * Now reshape got set up, update superblocks to
@@ -3883,13 +3886,13 @@ static void raid_resume(struct dm_target *ti)
 	if (!(rs->ctr_flags & RESUME_STAY_FROZEN_FLAGS))
 		clear_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
 
-	if (mddev->suspended)
+	if (test_and_clear_bit(RT_FLAG_RS_SUSPENDED, &rs->runtime_flags))
 		mddev_resume(mddev);
 }
 
 static struct target_type raid_target = {
 	.name = "raid",
-	.version = {1, 11, 1},
+	.version = {1, 12, 1},
 	.module = THIS_MODULE,
 	.ctr = raid_ctr,
 	.dtr = raid_dtr,

drivers/md/dm-table.c

@@ -20,6 +20,7 @@
 #include <linux/atomic.h>
 #include <linux/blk-mq.h>
 #include <linux/mount.h>
+#include <linux/dax.h>
 
 #define DM_MSG_PREFIX "table"
 
@@ -1630,6 +1631,37 @@ static bool dm_table_supports_flush(struct dm_table *t, unsigned long flush)
 	return false;
 }
 
+static int device_dax_write_cache_enabled(struct dm_target *ti,
+					  struct dm_dev *dev, sector_t start,
+					  sector_t len, void *data)
+{
+	struct dax_device *dax_dev = dev->dax_dev;
+
+	if (!dax_dev)
+		return false;
+
+	if (dax_write_cache_enabled(dax_dev))
+		return true;
+	return false;
+}
+
+static int dm_table_supports_dax_write_cache(struct dm_table *t)
+{
+	struct dm_target *ti;
+	unsigned i;
+
+	for (i = 0; i < dm_table_get_num_targets(t); i++) {
+		ti = dm_table_get_target(t, i);
+
+		if (ti->type->iterate_devices &&
+		    ti->type->iterate_devices(ti,
+				device_dax_write_cache_enabled, NULL))
+			return true;
+	}
+
+	return false;
+}
+
 static int device_is_nonrot(struct dm_target *ti, struct dm_dev *dev,
 			    sector_t start, sector_t len, void *data)
 {
@@ -1785,6 +1817,9 @@ void dm_table_set_restrictions(struct dm_table *t, struct request_queue *q,
 	}
 	blk_queue_write_cache(q, wc, fua);
 
+	if (dm_table_supports_dax_write_cache(t))
+		dax_write_cache(t->md->dax_dev, true);
+
 	/* Ensure that all underlying devices are non-rotational. */
 	if (dm_table_all_devices_attribute(t, device_is_nonrot))
 		queue_flag_set_unlocked(QUEUE_FLAG_NONROT, q);

drivers/md/dm-verity-fec.c

@@ -308,19 +308,14 @@ static int fec_alloc_bufs(struct dm_verity *v, struct dm_verity_fec_io *fio)
 {
 	unsigned n;
 
-	if (!fio->rs) {
-		fio->rs = mempool_alloc(v->fec->rs_pool, 0);
-		if (unlikely(!fio->rs)) {
-			DMERR("failed to allocate RS");
-			return -ENOMEM;
-		}
-	}
+	if (!fio->rs)
+		fio->rs = mempool_alloc(v->fec->rs_pool, GFP_NOIO);
 
 	fec_for_each_prealloc_buffer(n) {
 		if (fio->bufs[n])
 			continue;
 
-		fio->bufs[n] = mempool_alloc(v->fec->prealloc_pool, GFP_NOIO);
+		fio->bufs[n] = mempool_alloc(v->fec->prealloc_pool, GFP_NOWAIT);
 		if (unlikely(!fio->bufs[n])) {
 			DMERR("failed to allocate FEC buffer");
 			return -ENOMEM;
@@ -332,22 +327,16 @@ static int fec_alloc_bufs(struct dm_verity *v, struct dm_verity_fec_io *fio)
 		if (fio->bufs[n])
 			continue;
 
-		fio->bufs[n] = mempool_alloc(v->fec->extra_pool, GFP_NOIO);
+		fio->bufs[n] = mempool_alloc(v->fec->extra_pool, GFP_NOWAIT);
 		/* we can manage with even one buffer if necessary */
 		if (unlikely(!fio->bufs[n]))
 			break;
 	}
 	fio->nbufs = n;
 
-	if (!fio->output) {
+	if (!fio->output)
 		fio->output = mempool_alloc(v->fec->output_pool, GFP_NOIO);
 
-		if (!fio->output) {
-			DMERR("failed to allocate FEC page");
-			return -ENOMEM;
-		}
-	}
-
 	return 0;
 }
 

drivers/md/dm-zoned-metadata.c

@@ -624,7 +624,7 @@ static int dmz_write_sb(struct dmz_metadata *zmd, unsigned int set)
 
 	ret = dmz_rdwr_block(zmd, REQ_OP_WRITE, block, mblk->page);
 	if (ret == 0)
-		ret = blkdev_issue_flush(zmd->dev->bdev, GFP_KERNEL, NULL);
+		ret = blkdev_issue_flush(zmd->dev->bdev, GFP_NOIO, NULL);
 
 	return ret;
 }
@@ -658,7 +658,7 @@ static int dmz_write_dirty_mblocks(struct dmz_metadata *zmd,
 
 	/* Flush drive cache (this will also sync data) */
 	if (ret == 0)
-		ret = blkdev_issue_flush(zmd->dev->bdev, GFP_KERNEL, NULL);
+		ret = blkdev_issue_flush(zmd->dev->bdev, GFP_NOIO, NULL);
 
 	return ret;
 }
@@ -722,7 +722,7 @@ int dmz_flush_metadata(struct dmz_metadata *zmd)
 
 	/* If there are no dirty metadata blocks, just flush the device cache */
 	if (list_empty(&write_list)) {
-		ret = blkdev_issue_flush(zmd->dev->bdev, GFP_KERNEL, NULL);
+		ret = blkdev_issue_flush(zmd->dev->bdev, GFP_NOIO, NULL);
 		goto out;
 	}
 
@@ -927,7 +927,7 @@ static int dmz_recover_mblocks(struct dmz_metadata *zmd, unsigned int dst_set)
 			(zmd->nr_meta_zones << zmd->dev->zone_nr_blocks_shift);
 	}
 
-	page = alloc_page(GFP_KERNEL);
+	page = alloc_page(GFP_NOIO);
 	if (!page)
 		return -ENOMEM;
 
@@ -1183,7 +1183,7 @@ static int dmz_update_zone(struct dmz_metadata *zmd, struct dm_zone *zone)
 
 	/* Get zone information from disk */
 	ret = blkdev_report_zones(zmd->dev->bdev, dmz_start_sect(zmd, zone),
-				  &blkz, &nr_blkz, GFP_KERNEL);
+				  &blkz, &nr_blkz, GFP_NOIO);
 	if (ret) {
 		dmz_dev_err(zmd->dev, "Get zone %u report failed",
 			    dmz_id(zmd, zone));
@@ -1257,7 +1257,7 @@ static int dmz_reset_zone(struct dmz_metadata *zmd, struct dm_zone *zone)
 
 		ret = blkdev_reset_zones(dev->bdev,
 					 dmz_start_sect(zmd, zone),
-					 dev->zone_nr_sectors, GFP_KERNEL);
+					 dev->zone_nr_sectors, GFP_NOIO);
 		if (ret) {
 			dmz_dev_err(dev, "Reset zone %u failed %d",
 				    dmz_id(zmd, zone), ret);

drivers/md/dm-zoned-reclaim.c

@@ -75,7 +75,7 @@ static int dmz_reclaim_align_wp(struct dmz_reclaim *zrc, struct dm_zone *zone,
 	nr_blocks = block - wp_block;
 	ret = blkdev_issue_zeroout(zrc->dev->bdev,
 				   dmz_start_sect(zmd, zone) + dmz_blk2sect(wp_block),
-				   dmz_blk2sect(nr_blocks), GFP_NOFS, false);
+				   dmz_blk2sect(nr_blocks), GFP_NOIO, 0);
 	if (ret) {
 		dmz_dev_err(zrc->dev,
 			    "Align zone %u wp %llu to %llu (wp+%u) blocks failed %d",

drivers/md/dm-zoned-target.c

@@ -541,7 +541,7 @@ static void dmz_queue_chunk_work(struct dmz_target *dmz, struct bio *bio)
 		int ret;
 
 		/* Create a new chunk work */
-		cw = kmalloc(sizeof(struct dm_chunk_work), GFP_NOFS);
+		cw = kmalloc(sizeof(struct dm_chunk_work), GFP_NOIO);
 		if (!cw)
 			goto out;
 
@@ -588,7 +588,7 @@ static int dmz_map(struct dm_target *ti, struct bio *bio)
 
 	bio->bi_bdev = dev->bdev;
 
-	if (!nr_sectors && (bio_op(bio) != REQ_OP_FLUSH) && (bio_op(bio) != REQ_OP_WRITE))
+	if (!nr_sectors && bio_op(bio) != REQ_OP_WRITE)
 		return DM_MAPIO_REMAPPED;
 
 	/* The BIO should be block aligned */
@@ -603,7 +603,7 @@ static int dmz_map(struct dm_target *ti, struct bio *bio)
 	bioctx->status = BLK_STS_OK;
 
 	/* Set the BIO pending in the flush list */
-	if (bio_op(bio) == REQ_OP_FLUSH || (!nr_sectors && bio_op(bio) == REQ_OP_WRITE)) {
+	if (!nr_sectors && bio_op(bio) == REQ_OP_WRITE) {
 		spin_lock(&dmz->flush_lock);
 		bio_list_add(&dmz->flush_list, bio);
 		spin_unlock(&dmz->flush_lock);
@@ -785,7 +785,7 @@ static int dmz_ctr(struct dm_target *ti, unsigned int argc, char **argv)
 
 	/* Chunk BIO work */
 	mutex_init(&dmz->chunk_lock);
-	INIT_RADIX_TREE(&dmz->chunk_rxtree, GFP_NOFS);
+	INIT_RADIX_TREE(&dmz->chunk_rxtree, GFP_KERNEL);
 	dmz->chunk_wq = alloc_workqueue("dmz_cwq_%s", WQ_MEM_RECLAIM | WQ_UNBOUND,
 					0, dev->name);
 	if (!dmz->chunk_wq) {

include/linux/dax.h

@@ -87,6 +87,7 @@ size_t dax_copy_from_iter(struct dax_device *dax_dev, pgoff_t pgoff, void *addr,
 void dax_flush(struct dax_device *dax_dev, pgoff_t pgoff, void *addr,
 		size_t size);
 void dax_write_cache(struct dax_device *dax_dev, bool wc);
+bool dax_write_cache_enabled(struct dax_device *dax_dev);
 
 /*
  * We use lowest available bit in exceptional entry for locking, one bit for