block: add a report_zones method

Dispatching a report zones command through the request queue is a major
pain due to the command reply payload rewriting necessary. Given that
blkdev_report_zones() is executing everything synchronously, implement
report zones as a block device file operation instead, allowing major
simplification of the code in many places.

sd, null-blk, dm-linear and dm-flakey being the only block device
drivers supporting exposing zoned block devices, these drivers are
modified to provide the device side implementation of the
report_zones() block device file operation.

For device mappers, a new report_zones() target type operation is
defined so that the upper block layer calls blkdev_report_zones() can
be propagated down to the underlying devices of the dm targets.
Implementation for this new operation is added to the dm-linear and
dm-flakey targets.

Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
[Damien]
* Changed method block_device argument to gendisk
* Various bug fixes and improvements
* Added support for null_blk, dm-linear and dm-flakey.
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

3 files changed, 126 insertions, 108 deletions

diff --git a/drivers/md/dm-flakey.c b/drivers/md/dm-flakey.c
index 32aabe27b37c..3cb97fa4c11d 100644
--- a/drivers/md/dm-flakey.c
+++ b/drivers/md/dm-flakey.c
@@ -315,10 +315,6 @@ static int flakey_map(struct dm_target *ti, struct bio *bio)
 	if (bio_op(bio) == REQ_OP_ZONE_RESET)
 		goto map_bio;
 
-	/* We need to remap reported zones, so remember the BIO iter */
-	if (bio_op(bio) == REQ_OP_ZONE_REPORT)
-		goto map_bio;
-
 	/* Are we alive ? */
 	elapsed = (jiffies - fc->start_time) / HZ;
 	if (elapsed % (fc->up_interval + fc->down_interval) >= fc->up_interval) {
@@ -380,11 +376,6 @@ static int flakey_end_io(struct dm_target *ti, struct bio *bio,
 	if (bio_op(bio) == REQ_OP_ZONE_RESET)
 		return DM_ENDIO_DONE;
 
-	if (bio_op(bio) == REQ_OP_ZONE_REPORT) {
-		dm_remap_zone_report(ti, bio, fc->start);
-		return DM_ENDIO_DONE;
-	}
-
 	if (!*error && pb->bio_submitted && (bio_data_dir(bio) == READ)) {
 		if (fc->corrupt_bio_byte && (fc->corrupt_bio_rw == READ) &&
 		    all_corrupt_bio_flags_match(bio, fc)) {
@@ -457,6 +448,26 @@ static int flakey_prepare_ioctl(struct dm_target *ti, struct block_device **bdev
 	return 0;
 }
 
+#ifdef CONFIG_BLK_DEV_ZONED
+static int flakey_report_zones(struct dm_target *ti, sector_t sector,
+			       struct blk_zone *zones, unsigned int *nr_zones,
+			       gfp_t gfp_mask)
+{
+	struct flakey_c *fc = ti->private;
+	int ret;
+
+	/* Do report and remap it */
+	ret = blkdev_report_zones(fc->dev->bdev, flakey_map_sector(ti, sector),
+				  zones, nr_zones, gfp_mask);
+	if (ret != 0)
+		return ret;
+
+	if (*nr_zones)
+		dm_remap_zone_report(ti, fc->start, zones, nr_zones);
+	return 0;
+}
+#endif
+
 static int flakey_iterate_devices(struct dm_target *ti, iterate_devices_callout_fn fn, void *data)
 {
 	struct flakey_c *fc = ti->private;
@@ -469,6 +480,7 @@ static struct target_type flakey_target = {
 	.version = {1, 5, 0},
 #ifdef CONFIG_BLK_DEV_ZONED
 	.features = DM_TARGET_ZONED_HM,
+	.report_zones = flakey_report_zones,
 #endif
 	.module = THIS_MODULE,
 	.ctr    = flakey_ctr,
diff --git a/drivers/md/dm-linear.c b/drivers/md/dm-linear.c
index 2f7c44a006c4..8d7ddee6ac4d 100644
--- a/drivers/md/dm-linear.c
+++ b/drivers/md/dm-linear.c
@@ -102,19 +102,6 @@ static int linear_map(struct dm_target *ti, struct bio *bio)
 	return DM_MAPIO_REMAPPED;
 }
 
-#ifdef CONFIG_BLK_DEV_ZONED
-static int linear_end_io(struct dm_target *ti, struct bio *bio,
-			 blk_status_t *error)
-{
-	struct linear_c *lc = ti->private;
-
-	if (!*error && bio_op(bio) == REQ_OP_ZONE_REPORT)
-		dm_remap_zone_report(ti, bio, lc->start);
-
-	return DM_ENDIO_DONE;
-}
-#endif
-
 static void linear_status(struct dm_target *ti, status_type_t type,
 			  unsigned status_flags, char *result, unsigned maxlen)
 {
@@ -148,6 +135,26 @@ static int linear_prepare_ioctl(struct dm_target *ti, struct block_device **bdev
 	return 0;
 }
 
+#ifdef CONFIG_BLK_DEV_ZONED
+static int linear_report_zones(struct dm_target *ti, sector_t sector,
+			       struct blk_zone *zones, unsigned int *nr_zones,
+			       gfp_t gfp_mask)
+{
+	struct linear_c *lc = (struct linear_c *) ti->private;
+	int ret;
+
+	/* Do report and remap it */
+	ret = blkdev_report_zones(lc->dev->bdev, linear_map_sector(ti, sector),
+				  zones, nr_zones, gfp_mask);
+	if (ret != 0)
+		return ret;
+
+	if (*nr_zones)
+		dm_remap_zone_report(ti, lc->start, zones, nr_zones);
+	return 0;
+}
+#endif
+
 static int linear_iterate_devices(struct dm_target *ti,
 				  iterate_devices_callout_fn fn, void *data)
 {
@@ -211,8 +218,8 @@ static struct target_type linear_target = {
 	.name   = "linear",
 	.version = {1, 4, 0},
 #ifdef CONFIG_BLK_DEV_ZONED
-	.end_io = linear_end_io,
 	.features = DM_TARGET_PASSES_INTEGRITY | DM_TARGET_ZONED_HM,
+	.report_zones = linear_report_zones,
 #else
 	.features = DM_TARGET_PASSES_INTEGRITY,
 #endif
diff --git a/drivers/md/dm.c b/drivers/md/dm.c
index 45abb54037fc..6be21dc210a1 100644
--- a/drivers/md/dm.c
+++ b/drivers/md/dm.c
@@ -458,6 +458,57 @@ static int dm_blk_getgeo(struct block_device *bdev, struct hd_geometry *geo)
 	return dm_get_geometry(md, geo);
 }
 
+static int dm_blk_report_zones(struct gendisk *disk, sector_t sector,
+			       struct blk_zone *zones, unsigned int *nr_zones,
+			       gfp_t gfp_mask)
+{
+#ifdef CONFIG_BLK_DEV_ZONED
+	struct mapped_device *md = disk->private_data;
+	struct dm_target *tgt;
+	struct dm_table *map;
+	int srcu_idx, ret;
+
+	if (dm_suspended_md(md))
+		return -EAGAIN;
+
+	map = dm_get_live_table(md, &srcu_idx);
+	if (!map)
+		return -EIO;
+
+	tgt = dm_table_find_target(map, sector);
+	if (!dm_target_is_valid(tgt)) {
+		ret = -EIO;
+		goto out;
+	}
+
+	/*
+	 * If we are executing this, we already know that the block device
+	 * is a zoned device and so each target should have support for that
+	 * type of drive. A missing report_zones method means that the target
+	 * driver has a problem.
+	 */
+	if (WARN_ON(!tgt->type->report_zones)) {
+		ret = -EIO;
+		goto out;
+	}
+
+	/*
+	 * blkdev_report_zones() will loop and call this again to cover all the
+	 * zones of the target, eventually moving on to the next target.
+	 * So there is no need to loop here trying to fill the entire array
+	 * of zones.
+	 */
+	ret = tgt->type->report_zones(tgt, sector, zones,
+				      nr_zones, gfp_mask);
+
+out:
+	dm_put_live_table(md, srcu_idx);
+	return ret;
+#else
+	return -ENOTSUPP;
+#endif
+}
+
 static int dm_prepare_ioctl(struct mapped_device *md, int *srcu_idx,
 			    struct block_device **bdev)
 	__acquires(md->io_barrier)
@@ -1155,93 +1206,49 @@ void dm_accept_partial_bio(struct bio *bio, unsigned n_sectors)
 EXPORT_SYMBOL_GPL(dm_accept_partial_bio);
 
 /*
- * The zone descriptors obtained with a zone report indicate zone positions
- * within the target backing device, regardless of that device is a partition
- * and regardless of the target mapping start sector on the device or partition.
- * The zone descriptors start sector and write pointer position must be adjusted
- * to match their relative position within the dm device.
- * A target may call dm_remap_zone_report() after completion of a
- * REQ_OP_ZONE_REPORT bio to remap the zone descriptors obtained from the
- * backing device.
+ * The zone descriptors obtained with a zone report indicate
+ * zone positions within the underlying device of the target. The zone
+ * descriptors must be remapped to match their position within the dm device.
+ * The caller target should obtain the zones information using
+ * blkdev_report_zones() to ensure that remapping for partition offset is
+ * already handled.
  */
-void dm_remap_zone_report(struct dm_target *ti, struct bio *bio, sector_t start)
+void dm_remap_zone_report(struct dm_target *ti, sector_t start,
+			  struct blk_zone *zones, unsigned int *nr_zones)
 {
 #ifdef CONFIG_BLK_DEV_ZONED
-	struct dm_target_io *tio = container_of(bio, struct dm_target_io, clone);
-	struct bio *report_bio = tio->io->orig_bio;
-	struct blk_zone_report_hdr *hdr = NULL;
 	struct blk_zone *zone;
-	unsigned int nr_rep = 0;
-	unsigned int ofst;
-	sector_t part_offset;
-	struct bio_vec bvec;
-	struct bvec_iter iter;
-	void *addr;
-
-	if (bio->bi_status)
-		return;
-
-	/*
-	 * bio sector was incremented by the request size on completion. Taking
-	 * into account the original request sector, the target start offset on
-	 * the backing device and the target mapping offset (ti->begin), the
-	 * start sector of the backing device. The partition offset is always 0
-	 * if the target uses a whole device.
-	 */
-	part_offset = bio->bi_iter.bi_sector + ti->begin - (start + bio_end_sector(report_bio));
+	unsigned int nrz = *nr_zones;
+	int i;
 
 	/*
-	 * Remap the start sector of the reported zones. For sequential zones,
-	 * also remap the write pointer position.
+	 * Remap the start sector and write pointer position of the zones in
+	 * the array. Since we may have obtained from the target underlying
+	 * device more zones that the target size, also adjust the number
+	 * of zones.
 	 */
-	bio_for_each_segment(bvec, report_bio, iter) {
-		addr = kmap_atomic(bvec.bv_page);
-
-		/* Remember the report header in the first page */
-		if (!hdr) {
-			hdr = addr;
-			ofst = sizeof(struct blk_zone_report_hdr);
-		} else
-			ofst = 0;
-
-		/* Set zones start sector */
-		while (hdr->nr_zones && ofst < bvec.bv_len) {
-			zone = addr + ofst;
-			zone->start -= part_offset;
-			if (zone->start >= start + ti->len) {
-				hdr->nr_zones = 0;
-				break;
-			}
-			zone->start = zone->start + ti->begin - start;
-			if (zone->type != BLK_ZONE_TYPE_CONVENTIONAL) {
-				if (zone->cond == BLK_ZONE_COND_FULL)
-					zone->wp = zone->start + zone->len;
-				else if (zone->cond == BLK_ZONE_COND_EMPTY)
-					zone->wp = zone->start;
-				else
-					zone->wp = zone->wp + ti->begin - start - part_offset;
-			}
-			ofst += sizeof(struct blk_zone);
-			hdr->nr_zones--;
-			nr_rep++;
+	for (i = 0; i < nrz; i++) {
+		zone = zones + i;
+		if (zone->start >= start + ti->len) {
+			memset(zone, 0, sizeof(struct blk_zone) * (nrz - i));
+			break;
 		}
 
-		if (addr != hdr)
-			kunmap_atomic(addr);
+		zone->start = zone->start + ti->begin - start;
+		if (zone->type == BLK_ZONE_TYPE_CONVENTIONAL)
+			continue;
 
-		if (!hdr->nr_zones)
-			break;
-	}
-
-	if (hdr) {
-		hdr->nr_zones = nr_rep;
-		kunmap_atomic(hdr);
+		if (zone->cond == BLK_ZONE_COND_FULL)
+			zone->wp = zone->start + zone->len;
+		else if (zone->cond == BLK_ZONE_COND_EMPTY)
+			zone->wp = zone->start;
+		else
+			zone->wp = zone->wp + ti->begin - start;
 	}
 
-	bio_advance(report_bio, report_bio->bi_iter.bi_size);
-
+	*nr_zones = i;
 #else /* !CONFIG_BLK_DEV_ZONED */
-	bio->bi_status = BLK_STS_NOTSUPP;
+	*nr_zones = 0;
 #endif
 }
 EXPORT_SYMBOL_GPL(dm_remap_zone_report);
@@ -1327,8 +1334,7 @@ static int clone_bio(struct dm_target_io *tio, struct bio *bio,
 			return r;
 	}
 
-	if (bio_op(bio) != REQ_OP_ZONE_REPORT)
-		bio_advance(clone, to_bytes(sector - clone->bi_iter.bi_sector));
+	bio_advance(clone, to_bytes(sector - clone->bi_iter.bi_sector));
 	clone->bi_iter.bi_size = to_bytes(len);
 
 	if (unlikely(bio_integrity(bio) != NULL))
@@ -1541,7 +1547,6 @@ static bool __process_abnormal_io(struct clone_info *ci, struct dm_target *ti,
  */
 static int __split_and_process_non_flush(struct clone_info *ci)
 {
-	struct bio *bio = ci->bio;
 	struct dm_target *ti;
 	unsigned len;
 	int r;
@@ -1553,11 +1558,7 @@ static int __split_and_process_non_flush(struct clone_info *ci)
 	if (unlikely(__process_abnormal_io(ci, ti, &r)))
 		return r;
 
-	if (bio_op(bio) == REQ_OP_ZONE_REPORT)
-		len = ci->sector_count;
-	else
-		len = min_t(sector_t, max_io_len(ci->sector, ti),
-			    ci->sector_count);
+	len = min_t(sector_t, max_io_len(ci->sector, ti), ci->sector_count);
 
 	r = __clone_and_map_data_bio(ci, ti, ci->sector, &len);
 	if (r < 0)
@@ -1616,9 +1617,6 @@ static blk_qc_t __split_and_process_bio(struct mapped_device *md,
 				 * We take a clone of the original to store in
 				 * ci.io->orig_bio to be used by end_io_acct() and
 				 * for dec_pending to use for completion handling.
-				 * As this path is not used for REQ_OP_ZONE_REPORT,
-				 * the usage of io->orig_bio in dm_remap_zone_report()
-				 * won't be affected by this reassignment.
 				 */
 				struct bio *b = bio_split(bio, bio_sectors(bio) - ci.sector_count,
 							  GFP_NOIO, &md->queue->bio_split);
@@ -3167,6 +3165,7 @@ static const struct block_device_operations dm_blk_dops = {
 	.release = dm_blk_close,
 	.ioctl = dm_blk_ioctl,
 	.getgeo = dm_blk_getgeo,
+	.report_zones = dm_blk_report_zones,
 	.pr_ops = &dm_pr_ops,
 	.owner = THIS_MODULE
 };