* [PATCH v3 10/10] dm-zoned: Drive-managed zoned block device target
2017-05-08 23:40 [PATCH v3 00/10] dm: zoned block device support Bart Van Assche
` (8 preceding siblings ...)
2017-05-08 23:40 ` [PATCH v3 09/10] dm-kcopyd: Add sequential write feature Bart Van Assche
@ 2017-05-08 23:40 ` Bart Van Assche
2017-05-09 14:50 ` [PATCH v3 00/10] dm: zoned block device support Mike Snitzer
2017-05-16 20:03 ` Mike Snitzer
11 siblings, 0 replies; 21+ messages in thread
From: Bart Van Assche @ 2017-05-08 23:40 UTC (permalink / raw)
To: dm-devel, Mike Snitzer, Alasdair Kergon
Cc: Bart Van Assche, Damien Le Moal, Christoph Hellwig
From: Damien Le Moal <damien.lemoal@wdc.com>
The dm-zoned device mapper target provides transparent write access
to zoned block devices (ZBC and ZAC compliant block devices).
dm-zoned hides to the device user (a file system or an application
doing raw block device accesses) any constraint imposed on write
requests by the device, equivalent to a drive-managed zoned block
device model.
Write requests are processed using a combination of on-disk buffering
using the device conventional zones and direct in-place processing for
requests aligned to a zone sequential write pointer position.
A background reclaim process implemented using dm_kcopyd_copy ensures
that conventional zones are always available for executing unaligned
write requests. The reclaim process overhead is minimized by managing
buffer zones in a least-recently-written order and first targeting the
oldest buffer zones. Doing so, blocks under regular write access (such
as metadata blocks of a file system) remain stored in conventional
zones, resulting in no apparent overhead.
dm-zoned implementation focus on simplicity and on minimizing overhead
(CPU, memory and storage overhead). For a 10TB host-managed disk with
256 MB zones, dm-zoned memory usage per disk instance is at most 4.5 MB
and as little as 5 zones will be used internally for storing metadata
and performing buffer zone reclaim operations. This is achieved using
zone level indirection rather than a full block indirection system for
managing block movement between zones.
dm-zoned primary target is host-managed zoned block devices but it can
also be used with host-aware device models to mitigate potential
device-side performance degradation due to excessive random writing.
dm-zoned target devices can be formatted and checked using the dmzadm
utility available at:
https://github.com/hgst/dm-zoned-tools
Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Bart Van Assche <bart.vanassche@sandisk.com>
---
Documentation/device-mapper/dm-zoned.txt | 154 +++
drivers/md/Kconfig | 17 +
drivers/md/Makefile | 2 +
drivers/md/dm-zoned-io.c | 999 ++++++++++++++
drivers/md/dm-zoned-metadata.c | 2225 ++++++++++++++++++++++++++++++
drivers/md/dm-zoned-reclaim.c | 535 +++++++
drivers/md/dm-zoned.h | 530 +++++++
7 files changed, 4462 insertions(+)
create mode 100644 Documentation/device-mapper/dm-zoned.txt
create mode 100644 drivers/md/dm-zoned-io.c
create mode 100644 drivers/md/dm-zoned-metadata.c
create mode 100644 drivers/md/dm-zoned-reclaim.c
create mode 100644 drivers/md/dm-zoned.h
diff --git a/Documentation/device-mapper/dm-zoned.txt b/Documentation/device-mapper/dm-zoned.txt
new file mode 100644
index 0000000..d41f597
--- /dev/null
+++ b/Documentation/device-mapper/dm-zoned.txt
@@ -0,0 +1,154 @@
+dm-zoned
+========
+
+The dm-zoned device mapper exposes a zoned block device (ZBC and ZAC compliant
+devices) as a regular block device without any write pattern constraint. In
+effect, it implements a drive-managed zoned block device which hides to the
+user (a file system or an application doing raw block device accesses) the
+sequential write constraints of host-managed zoned block devices and can
+mitigate the potential device-side performance degradation due to excessive
+random writes on host-aware zoned block devices.
+
+For a more detailed description of the zoned block device models and
+their constraints see (for SCSI devices):
+
+http://www.t10.org/drafts.htm#ZBC_Family
+
+and (for ATA devices):
+
+http://www.t13.org/Documents/UploadedDocuments/docs2015/
+di537r05-Zoned_Device_ATA_Command_Set_ZAC.pdf
+
+dm-zoned implementation is simple and minimizes system overhead (CPU and
+memory usage as well as storage capacity loss). For a 10TB host-manmaged disk
+with 256 MB zones, dm-zoned memory usage per disk instance is at most 4.5 MB
+and as little as 5 zones will be used internally for storing metadata and
+performaing reclaim operations.
+
+dm-zoned targte devices can be formatted and checked using the dmzadm utility
+available at:
+
+https://github.com/hgst/dm-zoned-tools
+
+Algorithm
+=========
+
+dm-zoned implements an on-disk buffering scheme to handle non-sequential write
+accesses to the sequential zones of a zoned block device. Conventional zones
+are used for caching as well as for storing internal metadata.
+
+The zones of the device are separated into 2 types:
+
+1) Metadata zones: these are conventional zones used to store metadata.
+Metadata zones are not reported as useable capacity to the user.
+
+2) Data zones: all remaining zones, the vast majority of which will be
+sequential zones used exclusively to store user data. The conventional zones
+of the device may be used also for buffering user random writes. Data in these
+zones may be directly mapped to the conventional zone, but later moved to a
+sequential zone after so that the conventional zone can be reused for buffering
+incoming random writes.
+
+dm-zoned exposes a logical device with a sector size of 4096 bytes,
+irrespectively of the physical sector size of the backend zoned block device
+being used. This allows reducing the amount of metadata needed to manage valid
+blocks (blocks written).
+
+The on-disk metadata format is as follows:
+
+1) The first block of the first convnetional zone found contains the
+super block which describes the amount and position on disk of metadata blocks.
+
+2) Following the super block, a set of blocks is used to describe the mapping
+of the logical device blocks. The mapping is done per chunk of blocks, with
+the chunk size equal to the zoned block device size. The mapping table is
+indexed by chunk number and each mapping entry indicates the zone number of
+the device storing the chunk of data. Each mapping entry may also indicate if
+the zone number of a conventional zone used to buffer random modification to
+the data zone.
+
+3) A set of blocks used to store bitmaps indicating the validity of blocks in
+the data zones follows the mapping table. A valid block is defined as a block
+that was writen and not discarded. For a buffered data chunk, a block is
+always valid only in the data zone mapping the chunk or in the buffer zone of
+the chunk.
+
+For a logical chunk mapped to a conventional zone, all write operations are
+processed by directly writing to the zone. If the mapping zone is a
+sequential zone, the write operation is processed directly only and only if
+the write offset within the logical chunk is equal to the write pointer offset
+within of the sequential data zone (i.e. the write operation is aligned on the
+zone write pointer). Otherwise, write operations are processed indirectly
+using a buffer zone. In such case, an unused conventional zone is allocated
+and assigned to the chunk being accessed. Writing a block to the buffer zone
+of a chunk will automatically invalidate the same block in the sequential zone
+mapping the chunk. If all blocks of the sequential zone become invalid, the
+zone is freed and the chunk buffer zone becomes the primary zone mapping the
+chunk, resulting is native random write performance similar to a regular
+block device.
+
+Read operations are processed according to the block validity information
+provided by the bitmaps. Valid blocks are read either from the sequential zone
+mapping a chunk, or if the chunk is buffered, from the buffer zone assigned.
+If the accessed chunk has no mapping, or the accessed blocks are invalid, the
+read buffer is zeroed and the read operation terminated.
+
+After some time, the limited number of convnetional zones available may be
+exhausted (all used to map chunks or buffer sequential zones) and unaligned
+writes to unbuffered chunks become impossible. To avoid such situation, a
+reclaim process regularly scans used conventional zones and try to reclaim
+the least recently used ones copying the valid blocks of the buffer zone
+to a free sequential zone. Once the copy completes, the chunk mapping is
+updated to point to the sequential zone and the buffer zone freed for reuse.
+
+Metadata Protection
+===================
+
+To protect metadata against corruption in case of sudden power loss or system
+crash, 2 sets of metadata zones are used. One set, the primary set, is used as
+the main metadata region, while the secondary set is used as a staging area.
+Modified metadata are first written to the secondary set and validated by
+updating the super block in the secondary set, indicating using a generation
+counter that this set contains the newest metadata. Once this operation
+completes, updates in place of metadata blocks can be done in the primary
+metadata set, ensuring that one of the set is always consistent (all
+modifications committed or none at all). Flush operations are used as a commit
+point. Upon reception of a flush request, metadata modification activity is
+temporarily blocked (for both incoming BIO processing and reclaim process) and
+all dirty metadata blocks staged and updated. Normal operation is then resumed.
+Metadata flush thus only temporarily delays write and discard requests. Read
+requests can be concurrently processed while metadata flush is being executed.
+
+Usage
+=====
+
+A zoned block device must first be formatted using the dmzadm tool. This will
+analyze the device zone configuration, determine where to place the metadata
+sets on the device and initialize the metadata sets.
+
+Ex:
+
+dmzadm --format /dev/sdxx
+
+For a formatted device, the target can be created normally with the dmsetup
+utility. The only parameter that dm-zoned requires is the device name.
+
+Example scripts
+===============
+
+[[
+#!/bin/sh
+
+if [ $# -ne 1 ]; then
+ echo "Usage: $0 <Zoned device path>"
+ exit 1
+fi
+
+dev="${1}"
+shift
+
+modprobe dm-zoned
+
+echo "0 `blockdev --getsize ${dev}` dm-zoned ${dev}" | dmsetup create dmz-`basename ${dev}`
+]]
+
diff --git a/drivers/md/Kconfig b/drivers/md/Kconfig
index 906103c..a081b3c 100644
--- a/drivers/md/Kconfig
+++ b/drivers/md/Kconfig
@@ -521,6 +521,23 @@ config DM_INTEGRITY
To compile this code as a module, choose M here: the module will
be called dm-integrity.
+config DM_ZONED
+ tristate "Drive-managed zoned block device target support"
+ depends on BLK_DEV_DM
+ depends on BLK_DEV_ZONED
+ ---help---
+ This device-mapper target takes a host-managed or host-aware zoned
+ block device and expose most of its capacity as a regular block
+ device (drive-managed zoned block device) without any write
+ constraint. This is mainly intended for use with file systems that
+ do not natively support zoned block devices but still want to
+ benefit from the increased capacity offered by SMR disks. Other uses
+ by applications using raw block devices (for example object stores)
+ is also possible.
+
+ To compile this code as a module, choose M here: the module will
+ be called dm-zoned.
+
If unsure, say N.
endif # MD
diff --git a/drivers/md/Makefile b/drivers/md/Makefile
index 913720b..f77cc3b 100644
--- a/drivers/md/Makefile
+++ b/drivers/md/Makefile
@@ -20,6 +20,7 @@ dm-era-y += dm-era-target.o
dm-verity-y += dm-verity-target.o
md-mod-y += md.o bitmap.o
raid456-y += raid5.o raid5-cache.o raid5-ppl.o
+dm-zoned-y += dm-zoned-io.o dm-zoned-metadata.o dm-zoned-reclaim.o
# Note: link order is important. All raid personalities
# and must come before md.o, as they each initialise
@@ -60,6 +61,7 @@ obj-$(CONFIG_DM_CACHE_SMQ) += dm-cache-smq.o
obj-$(CONFIG_DM_ERA) += dm-era.o
obj-$(CONFIG_DM_LOG_WRITES) += dm-log-writes.o
obj-$(CONFIG_DM_INTEGRITY) += dm-integrity.o
+obj-$(CONFIG_DM_ZONED) += dm-zoned.o
ifeq ($(CONFIG_DM_UEVENT),y)
dm-mod-objs += dm-uevent.o
diff --git a/drivers/md/dm-zoned-io.c b/drivers/md/dm-zoned-io.c
new file mode 100644
index 0000000..0119896
--- /dev/null
+++ b/drivers/md/dm-zoned-io.c
@@ -0,0 +1,999 @@
+/*
+ * Drive-managed zoned block device target
+ * Copyright (C) 2017 Western Digital Corporation or its affiliates.
+ *
+ * Written by: Damien Le Moal <damien.lemoal@wdc.com>
+ *
+ * This software is distributed under the terms of the GNU General Public
+ * License version 2, or any later version, "as is," without technical
+ * support, and WITHOUT ANY WARRANTY, without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+ */
+
+#include <linux/module.h>
+
+#include "dm-zoned.h"
+
+/*
+ * Target BIO completion.
+ */
+static inline void dmz_bio_end(struct bio *bio, int err)
+{
+ struct dmz_bioctx *bioctx =
+ dm_per_bio_data(bio, sizeof(struct dmz_bioctx));
+
+ if (atomic_dec_and_test(&bioctx->ref)) {
+ struct dmz_target *dmz = bioctx->target;
+
+ /* User BIO Completed */
+ if (bioctx->zone)
+ dmz_deactivate_zone(dmz, bioctx->zone);
+ atomic_dec(&dmz->bio_count);
+ bio->bi_error = bioctx->error;
+ bio_endio(bio);
+ }
+}
+
+/*
+ * Partial/internal BIO completion callback.
+ * This terminates the user target BIO when there
+ * are no more references to its context.
+ */
+static void dmz_bio_end_io(struct bio *bio)
+{
+ struct dmz_bioctx *bioctx = bio->bi_private;
+ int err = bio->bi_error;
+
+ if (err) {
+ struct dm_zone *zone = bioctx->zone;
+
+ bioctx->error = err;
+ if (bio_op(bio) == REQ_OP_WRITE &&
+ dmz_is_seq(zone))
+ set_bit(DMZ_SEQ_WRITE_ERR, &zone->flags);
+ }
+
+ dmz_bio_end(bioctx->bio, err);
+
+ bio_put(bio);
+
+}
+
+/*
+ * Issue a BIO to a zone. The BIO may only partially process the
+ * original target BIO.
+ */
+static int dmz_submit_bio(struct dmz_target *dmz, struct dm_zone *zone,
+ struct bio *dmz_bio,
+ sector_t chunk_block, unsigned int nr_blocks)
+{
+ struct dmz_bioctx *bioctx
+ = dm_per_bio_data(dmz_bio, sizeof(struct dmz_bioctx));
+ unsigned int nr_sectors = dmz_blk2sect(nr_blocks);
+ unsigned int size = nr_sectors << SECTOR_SHIFT;
+ struct bio *clone;
+
+ clone = bio_clone_fast(dmz_bio, GFP_NOIO, dmz->bio_set);
+ if (!clone)
+ return -ENOMEM;
+
+ /* Setup the clone */
+ clone->bi_bdev = dmz->zbd;
+ clone->bi_opf = dmz_bio->bi_opf;
+ clone->bi_iter.bi_sector =
+ dmz_start_sect(dmz, zone) + dmz_blk2sect(chunk_block);
+ clone->bi_iter.bi_size = size;
+ clone->bi_end_io = dmz_bio_end_io;
+ clone->bi_private = bioctx;
+
+ bio_advance(dmz_bio, size);
+
+ /* Submit the clone */
+ atomic_inc(&bioctx->ref);
+ generic_make_request(clone);
+
+ return 0;
+}
+
+/*
+ * Zero out pages of discarded blocks accessed by a read BIO.
+ */
+static void dmz_handle_read_zero(struct dmz_target *dmz, struct bio *bio,
+ sector_t chunk_block, unsigned int nr_blocks)
+{
+ unsigned int size = nr_blocks << DMZ_BLOCK_SHIFT;
+
+ dmz_dev_debug(dmz,
+ "=> ZERO READ chunk %llu -> block %llu, %u blocks\n",
+ (unsigned long long)dmz_bio_chunk(dmz, bio),
+ (unsigned long long)chunk_block,
+ nr_blocks);
+
+ /* Clear nr_blocks */
+ swap(bio->bi_iter.bi_size, size);
+ zero_fill_bio(bio);
+ swap(bio->bi_iter.bi_size, size);
+
+ bio_advance(bio, size);
+}
+
+/*
+ * Process a read BIO.
+ */
+static int dmz_handle_read(struct dmz_target *dmz, struct dm_zone *zone,
+ struct bio *bio)
+{
+ sector_t block = dmz_bio_block(bio);
+ unsigned int nr_blocks = dmz_bio_blocks(bio);
+ sector_t chunk_block = dmz_chunk_block(dmz, block);
+ sector_t end_block = chunk_block + nr_blocks;
+ struct dm_zone *rzone, *bzone;
+ int ret;
+
+ /* Read into unmapped chunks need only zeroing the BIO buffer */
+ if (!zone) {
+ dmz_handle_read_zero(dmz, bio, chunk_block, nr_blocks);
+ return 0;
+ }
+
+ dmz_dev_debug(dmz, "READ %s zone %u, block %llu, %u blocks\n",
+ (dmz_is_rnd(zone) ? "RND" : "SEQ"), dmz_id(dmz, zone),
+ (unsigned long long)chunk_block, nr_blocks);
+
+ /* Check block validity to determine the read location */
+ bzone = zone->bzone;
+ while (chunk_block < end_block) {
+
+ nr_blocks = 0;
+ if (dmz_is_rnd(zone)
+ || chunk_block < zone->wp_block) {
+ /* Test block validity in the data zone */
+ ret = dmz_block_valid(dmz, zone, chunk_block);
+ if (ret < 0)
+ return ret;
+ if (ret > 0) {
+ /* Read data zone blocks */
+ nr_blocks = ret;
+ rzone = zone;
+ }
+ }
+
+ /*
+ * No valid blocks found in the data zone.
+ * Check the buffer zone, if there is one.
+ */
+ if (!nr_blocks && bzone) {
+ ret = dmz_block_valid(dmz, bzone, chunk_block);
+ if (ret < 0)
+ return ret;
+ if (ret > 0) {
+ /* Read buffer zone blocks */
+ nr_blocks = ret;
+ rzone = bzone;
+ }
+ }
+
+ if (nr_blocks) {
+
+ /* Valid blocks found: read them */
+ nr_blocks = min_t(unsigned int, nr_blocks,
+ end_block - chunk_block);
+
+ dmz_dev_debug(dmz,
+ "=> %s READ zone %u, block %llu, %u blocks\n",
+ (dmz_is_buf(rzone) ? "BUF" : "DATA"),
+ dmz_id(dmz, rzone),
+ (unsigned long long)chunk_block,
+ nr_blocks);
+
+ ret = dmz_submit_bio(dmz, rzone, bio,
+ chunk_block, nr_blocks);
+ if (ret)
+ return ret;
+ chunk_block += nr_blocks;
+
+ } else {
+
+ /* No valid block: zeroout the current BIO block */
+ dmz_handle_read_zero(dmz, bio, chunk_block, 1);
+ chunk_block++;
+
+ }
+
+ }
+
+ return 0;
+}
+
+/*
+ * Write blocks directly in a data zone, at the write pointer.
+ * If a buffer zone is assigned, invalidate the blocks written
+ * in place.
+ */
+static int dmz_handle_direct_write(struct dmz_target *dmz,
+ struct dm_zone *zone, struct bio *bio,
+ sector_t chunk_block,
+ unsigned int nr_blocks)
+{
+ struct dm_zone *bzone = zone->bzone;
+ int ret;
+
+ dmz_dev_debug(dmz, "WRITE %s zone %u, block %llu, %u blocks\n",
+ (dmz_is_rnd(zone) ? "RND" : "SEQ"), dmz_id(dmz, zone),
+ (unsigned long long)chunk_block, nr_blocks);
+
+ if (dmz_is_readonly(zone))
+ return -EROFS;
+
+ /* Submit write */
+ ret = dmz_submit_bio(dmz, zone, bio, chunk_block, nr_blocks);
+ if (ret)
+ return -EIO;
+
+ if (dmz_is_seq(zone))
+ zone->wp_block += nr_blocks;
+
+ /*
+ * Validate the blocks in the data zone and invalidate
+ * in the buffer zone, if there is one.
+ */
+ ret = dmz_validate_blocks(dmz, zone, chunk_block, nr_blocks);
+ if (ret == 0 && bzone)
+ ret = dmz_invalidate_blocks(dmz, bzone, chunk_block, nr_blocks);
+
+ return ret;
+}
+
+/*
+ * Write blocks in the buffer zone of @zone.
+ * If no buffer zone is assigned yet, get one.
+ * Called with @zone write locked.
+ */
+static int dmz_handle_buffered_write(struct dmz_target *dmz,
+ struct dm_zone *zone, struct bio *bio,
+ sector_t chunk_block,
+ unsigned int nr_blocks)
+{
+ struct dm_zone *bzone = zone->bzone;
+ int ret;
+
+ if (!bzone) {
+ /* Get a buffer zone */
+ bzone = dmz_get_chunk_buffer(dmz, zone);
+ if (!bzone)
+ return -ENOSPC;
+ }
+
+ dmz_dev_debug(dmz, "WRITE BUF zone %u, block %llu, %u blocks\n",
+ dmz_id(dmz, bzone), (unsigned long long)chunk_block,
+ nr_blocks);
+
+ if (dmz_is_readonly(bzone))
+ return -EROFS;
+
+ /* Submit write */
+ ret = dmz_submit_bio(dmz, bzone, bio, chunk_block, nr_blocks);
+ if (ret)
+ return -EIO;
+
+ /*
+ * Validate the blocks in the buffer zone
+ * and invalidate in the data zone.
+ */
+ ret = dmz_validate_blocks(dmz, bzone, chunk_block, nr_blocks);
+ if (ret == 0 && chunk_block < zone->wp_block)
+ ret = dmz_invalidate_blocks(dmz, zone, chunk_block, nr_blocks);
+
+ return ret;
+}
+
+/*
+ * Process a write BIO.
+ */
+static int dmz_handle_write(struct dmz_target *dmz, struct dm_zone *zone,
+ struct bio *bio)
+{
+ sector_t block = dmz_bio_block(bio);
+ unsigned int nr_blocks = dmz_bio_blocks(bio);
+ sector_t chunk_block = dmz_chunk_block(dmz, block);
+
+ if (!zone)
+ return -ENOSPC;
+
+ if (dmz_is_rnd(zone) ||
+ chunk_block == zone->wp_block)
+ /*
+ * zone is a random zone, or it is a sequential zone
+ * and the BIO is aligned to the zone write pointer:
+ * direct write the zone.
+ */
+ return dmz_handle_direct_write(dmz, zone, bio,
+ chunk_block, nr_blocks);
+
+ /*
+ * This is an unaligned write in a sequential zone:
+ * use buffered write.
+ */
+ return dmz_handle_buffered_write(dmz, zone, bio,
+ chunk_block, nr_blocks);
+}
+
+/*
+ * Process a discard BIO.
+ */
+static int dmz_handle_discard(struct dmz_target *dmz, struct dm_zone *zone,
+ struct bio *bio)
+{
+ sector_t block = dmz_bio_block(bio);
+ unsigned int nr_blocks = dmz_bio_blocks(bio);
+ sector_t chunk_block = dmz_chunk_block(dmz, block);
+ int ret = 0;
+
+ /* For unmapped chunks, there is nothing to do */
+ if (!zone)
+ return 0;
+
+ if (dmz_is_readonly(zone))
+ return -EROFS;
+
+ dmz_dev_debug(dmz,
+ "DISCARD chunk %llu -> zone %u, block %llu, %u blocks\n",
+ (unsigned long long)dmz_bio_chunk(dmz, bio),
+ dmz_id(dmz, zone),
+ (unsigned long long)chunk_block, nr_blocks);
+
+ /*
+ * Invalidate blocks in the data zone and its
+ * buffer zone if one is mapped.
+ */
+ if (dmz_is_rnd(zone) ||
+ chunk_block < zone->wp_block)
+ ret = dmz_invalidate_blocks(dmz, zone,
+ chunk_block, nr_blocks);
+ if (ret == 0 && zone->bzone)
+ ret = dmz_invalidate_blocks(dmz, zone->bzone,
+ chunk_block, nr_blocks);
+
+ return ret;
+}
+
+/*
+ * Process a BIO.
+ */
+static void dmz_handle_bio(struct dmz_target *dmz, struct dm_chunk_work *cw,
+ struct bio *bio)
+{
+ struct dmz_bioctx *bioctx =
+ dm_per_bio_data(bio, sizeof(struct dmz_bioctx));
+ struct dm_zone *zone;
+ int ret;
+
+ down_read(&dmz->mblk_sem);
+
+ /*
+ * Get the data zone mapping the chunk. There may be no
+ * mapping for read and discard. If a mapping is obtained,
+ + the zone returned will be set to active state.
+ */
+ zone = dmz_get_chunk_mapping(dmz, dmz_bio_chunk(dmz, bio),
+ bio_op(bio));
+ if (IS_ERR(zone)) {
+ dmz_bio_end(bio, PTR_ERR(zone));
+ goto out;
+ }
+
+ /* Process the BIO */
+ if (zone) {
+ dmz_activate_zone(dmz, zone);
+ bioctx->zone = zone;
+ }
+
+ switch (bio_op(bio)) {
+ case REQ_OP_READ:
+ ret = dmz_handle_read(dmz, zone, bio);
+ break;
+ case REQ_OP_WRITE:
+ ret = dmz_handle_write(dmz, zone, bio);
+ break;
+ case REQ_OP_DISCARD:
+ case REQ_OP_WRITE_ZEROES:
+ ret = dmz_handle_discard(dmz, zone, bio);
+ break;
+ default:
+ dmz_dev_err(dmz,
+ "Unsupported BIO operation 0x%x\n",
+ bio_op(bio));
+ ret = -EIO;
+ }
+
+ dmz_bio_end(bio, ret);
+
+ /*
+ * Release the chunk mapping. This will check that the mapping
+ * is still valid, that is, that the zone used still has valid blocks.
+ */
+ if (zone)
+ dmz_put_chunk_mapping(dmz, zone);
+
+out:
+ up_read(&dmz->mblk_sem);
+}
+
+/*
+ * Increment a chunk reference counter.
+ */
+static inline void dmz_get_chunk_work(struct dm_chunk_work *cw)
+{
+ atomic_inc(&cw->refcount);
+}
+
+/*
+ * Decrement a chunk work reference count and
+ * free it if it becomes 0.
+ */
+static void dmz_put_chunk_work(struct dm_chunk_work *cw)
+{
+ if (atomic_dec_and_test(&cw->refcount)) {
+ atomic_dec(&cw->target->nr_active_chunks);
+ radix_tree_delete(&cw->target->chunk_rxtree, cw->chunk);
+ kfree(cw);
+ }
+}
+
+/*
+ * Chunk BIO work function.
+ */
+static void dmz_chunk_work(struct work_struct *work)
+{
+ struct dm_chunk_work *cw =
+ container_of(work, struct dm_chunk_work, work);
+ struct dmz_target *dmz = cw->target;
+ struct bio *bio;
+
+ mutex_lock(&dmz->chunk_lock);
+
+ /* Process the chunk BIOs */
+ while ((bio = bio_list_pop(&cw->bio_list))) {
+
+ mutex_unlock(&dmz->chunk_lock);
+ dmz_handle_bio(dmz, cw, bio);
+ mutex_lock(&dmz->chunk_lock);
+
+ dmz_put_chunk_work(cw);
+
+ }
+
+ /*
+ * Queueing the work added one to the work refcount.
+ * So drop this here.
+ */
+ dmz_put_chunk_work(cw);
+
+ mutex_unlock(&dmz->chunk_lock);
+}
+
+/*
+ * Flush work.
+ */
+static void dmz_flush_work(struct work_struct *work)
+{
+ struct dmz_target *dmz =
+ container_of(work, struct dmz_target, flush_work.work);
+ struct bio *bio;
+ int ret;
+
+ /* Flush metablocks */
+ ret = dmz_flush_mblocks(dmz);
+
+ /* Process queued flush requests */
+ while (1) {
+
+ spin_lock(&dmz->flush_lock);
+ bio = bio_list_pop(&dmz->flush_list);
+ spin_unlock(&dmz->flush_lock);
+
+ if (!bio)
+ break;
+
+ /* Do flush */
+ dmz_bio_end(bio, ret);
+
+ }
+
+ queue_delayed_work(dmz->flush_wq, &dmz->flush_work,
+ DMZ_FLUSH_PERIOD);
+}
+
+/*
+ * Get a chunk work and start it to process a new BIO.
+ * If the BIO chunk has no work yet, create one.
+ */
+static void dmz_queue_chunk_work(struct dmz_target *dmz,
+ struct bio *bio)
+{
+ unsigned int chunk = dmz_bio_chunk(dmz, bio);
+ struct dm_chunk_work *cw;
+
+ mutex_lock(&dmz->chunk_lock);
+
+ /* Get the BIO chunk work. If one is not active yet, create one */
+ cw = radix_tree_lookup(&dmz->chunk_rxtree, chunk);
+ if (!cw) {
+ int ret;
+
+ /* Create a new chunk work */
+ cw = kmalloc(sizeof(struct dm_chunk_work), GFP_NOFS);
+ if (!cw)
+ goto out;
+
+ INIT_WORK(&cw->work, dmz_chunk_work);
+ atomic_set(&cw->refcount, 0);
+ cw->target = dmz;
+ cw->chunk = chunk;
+ bio_list_init(&cw->bio_list);
+
+ ret = radix_tree_insert(&dmz->chunk_rxtree, chunk, cw);
+ if (unlikely(ret != 0)) {
+ kfree(cw);
+ cw = NULL;
+ goto out;
+ }
+
+ atomic_inc(&dmz->nr_active_chunks);
+ }
+
+ bio_list_add(&cw->bio_list, bio);
+ dmz_get_chunk_work(cw);
+
+ if (queue_work(dmz->chunk_wq, &cw->work))
+ dmz_get_chunk_work(cw);
+
+out:
+ mutex_unlock(&dmz->chunk_lock);
+}
+
+/*
+ * Process a new BIO.
+ */
+static int dmz_map(struct dm_target *ti, struct bio *bio)
+{
+ struct dmz_target *dmz = ti->private;
+ struct dmz_bioctx *bioctx
+ = dm_per_bio_data(bio, sizeof(struct dmz_bioctx));
+ sector_t sector = bio->bi_iter.bi_sector;
+ unsigned int nr_sectors = bio_sectors(bio);
+ sector_t chunk_sector;
+
+ dmz_dev_debug(dmz,
+ "BIO sector %llu + %u => chunk %llu, block %llu, %u blocks\n",
+ (u64)sector, nr_sectors,
+ (u64)dmz_bio_chunk(dmz, bio),
+ (u64)dmz_chunk_block(dmz, dmz_bio_block(bio)),
+ (unsigned int)dmz_bio_blocks(bio));
+
+ bio->bi_bdev = dmz->zbd;
+
+ if (!nr_sectors &&
+ (bio_op(bio) != REQ_OP_FLUSH) &&
+ (bio_op(bio) != REQ_OP_WRITE)) {
+ bio->bi_bdev = dmz->zbd;
+ return DM_MAPIO_REMAPPED;
+ }
+
+ /* The BIO should be block aligned */
+ if ((nr_sectors & DMZ_BLOCK_SECTORS_MASK) ||
+ (sector & DMZ_BLOCK_SECTORS_MASK)) {
+ dmz_dev_err(dmz,
+ "Unaligned BIO sector %llu, len %u\n",
+ (u64)sector,
+ nr_sectors);
+ return -EIO;
+ }
+
+ /* Initialize the BIO context */
+ bioctx->target = dmz;
+ bioctx->zone = NULL;
+ bioctx->bio = bio;
+ atomic_set(&bioctx->ref, 1);
+ bioctx->error = 0;
+
+ atomic_inc(&dmz->bio_count);
+ dmz->atime = jiffies;
+
+ /* Set the BIO pending in the flush list */
+ if (bio_op(bio) == REQ_OP_FLUSH ||
+ (!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);
+ dmz_trigger_flush(dmz);
+ return DM_MAPIO_SUBMITTED;
+ }
+
+ /* Split zone BIOs to fit entirely into a zone */
+ chunk_sector = dmz_chunk_sector(dmz, sector);
+ if (chunk_sector + nr_sectors > dmz->zone_nr_sectors)
+ dm_accept_partial_bio(bio,
+ dmz->zone_nr_sectors - chunk_sector);
+
+ /* Now ready to handle this BIO */
+ dmz_queue_chunk_work(dmz, bio);
+
+ return DM_MAPIO_SUBMITTED;
+}
+
+/*
+ * Setup target.
+ */
+static int dmz_ctr(struct dm_target *ti, unsigned int argc, char **argv)
+{
+ struct dmz_target *dmz;
+ int ret;
+
+ /* Check arguments */
+ if (argc != 1) {
+ ti->error = "Invalid argument count";
+ return -EINVAL;
+ }
+
+ /* Allocate and initialize the target descriptor */
+ dmz = kzalloc(sizeof(struct dmz_target), GFP_KERNEL);
+ if (!dmz) {
+ ti->error = "Allocate target descriptor failed";
+ return -ENOMEM;
+ }
+
+ /* Get the target device */
+ ret = dm_get_device(ti, argv[0], dm_table_get_mode(ti->table),
+ &dmz->ddev);
+ if (ret != 0) {
+ ti->error = "Get target device failed";
+ goto err;
+ }
+
+ dmz->zbd = dmz->ddev->bdev;
+ if (!bdev_is_zoned(dmz->zbd)) {
+ ti->error = "Not a zoned block device";
+ ret = -EINVAL;
+ goto err;
+ }
+
+ dmz->zbd_capacity = i_size_read(dmz->zbd->bd_inode) >> SECTOR_SHIFT;
+ if (ti->begin || (ti->len != dmz->zbd_capacity)) {
+ ti->error = "Partial mapping not supported";
+ ret = -EINVAL;
+ goto err;
+ }
+
+ (void)bdevname(dmz->zbd, dmz->zbd_name);
+ dmz->zbdq = bdev_get_queue(dmz->zbd);
+
+ dmz->mblk_rbtree = RB_ROOT;
+ init_rwsem(&dmz->mblk_sem);
+ spin_lock_init(&dmz->mblk_lock);
+ INIT_LIST_HEAD(&dmz->mblk_lru_list);
+ INIT_LIST_HEAD(&dmz->mblk_dirty_list);
+
+ mutex_init(&dmz->map_lock);
+ atomic_set(&dmz->dz_unmap_nr_rnd, 0);
+ INIT_LIST_HEAD(&dmz->dz_unmap_rnd_list);
+ INIT_LIST_HEAD(&dmz->dz_map_rnd_list);
+
+ atomic_set(&dmz->dz_unmap_nr_seq, 0);
+ INIT_LIST_HEAD(&dmz->dz_unmap_seq_list);
+ INIT_LIST_HEAD(&dmz->dz_map_seq_list);
+
+ init_waitqueue_head(&dmz->dz_free_wq);
+
+ atomic_set(&dmz->nr_reclaim_seq_zones, 0);
+ INIT_LIST_HEAD(&dmz->reclaim_seq_zones_list);
+
+ ret = dmz_init_meta(dmz);
+ if (ret != 0) {
+ ti->error = "Metadata initialization failed";
+ goto err;
+ }
+
+ /* Set target (no write same support) */
+ ti->private = dmz;
+ ti->max_io_len = dmz->zone_nr_sectors << 9;
+ ti->num_flush_bios = 1;
+ ti->num_discard_bios = 1;
+ ti->num_write_zeroes_bios = 1;
+ ti->per_io_data_size = sizeof(struct dmz_bioctx);
+ ti->flush_supported = true;
+ ti->discards_supported = true;
+ ti->split_discard_bios = true;
+
+ /* The exposed capacity is the number of chunks that can be mapped */
+ ti->len = dmz->nr_chunks * dmz->zone_nr_sectors;
+
+ /* Zone BIO */
+ atomic_set(&dmz->bio_count, 0);
+ dmz->atime = jiffies;
+ dmz->bio_set = bioset_create_nobvec(DMZ_MIN_BIOS, 0);
+ if (!dmz->bio_set) {
+ ti->error = "Create BIO set failed";
+ ret = -ENOMEM;
+ goto err;
+ }
+
+ /* Chunk BIO work */
+ mutex_init(&dmz->chunk_lock);
+ atomic_set(&dmz->nr_active_chunks, 0);
+ INIT_RADIX_TREE(&dmz->chunk_rxtree, GFP_NOFS);
+ dmz->chunk_wq = alloc_workqueue("dmz_cwq_%s",
+ WQ_MEM_RECLAIM | WQ_UNBOUND,
+ 0, dmz->zbd_name);
+ if (!dmz->chunk_wq) {
+ ti->error = "Create chunk workqueue failed";
+ ret = -ENOMEM;
+ goto err;
+ }
+
+ /* Flush work */
+ spin_lock_init(&dmz->flush_lock);
+ bio_list_init(&dmz->flush_list);
+ INIT_DELAYED_WORK(&dmz->flush_work, dmz_flush_work);
+ dmz->flush_wq = alloc_ordered_workqueue("dmz_fwq_%s", WQ_MEM_RECLAIM,
+ dmz->zbd_name);
+ if (!dmz->flush_wq) {
+ ti->error = "Create flush workqueue failed";
+ ret = -ENOMEM;
+ goto err;
+ }
+ mod_delayed_work(dmz->flush_wq, &dmz->flush_work, DMZ_FLUSH_PERIOD);
+
+ /* Reclaim kcopyd client */
+ dmz->reclaim_kc = dm_kcopyd_client_create(&dmz->reclaim_throttle);
+ if (IS_ERR(dmz->reclaim_kc)) {
+ ti->error = "Create kcopyd client failed";
+ ret = PTR_ERR(dmz->reclaim_kc);
+ dmz->reclaim_kc = NULL;
+ goto err;
+ }
+
+ /* Reclaim work */
+ INIT_DELAYED_WORK(&dmz->reclaim_work, dmz_reclaim_work);
+ dmz->reclaim_wq = alloc_ordered_workqueue("dmz_rwq_%s", WQ_MEM_RECLAIM,
+ dmz->zbd_name);
+ if (!dmz->reclaim_wq) {
+ ti->error = "Create reclaim workqueue failed";
+ ret = -ENOMEM;
+ goto err;
+ }
+
+ dmz_dev_info(dmz,
+ "Target device: %llu 512-byte logical sectors (%llu blocks)\n",
+ (unsigned long long)ti->len,
+ (unsigned long long)dmz_sect2blk(ti->len));
+
+ dmz_trigger_reclaim(dmz);
+
+ return 0;
+
+err:
+ if (dmz->ddev) {
+ if (dmz->reclaim_wq)
+ destroy_workqueue(dmz->reclaim_wq);
+ if (dmz->reclaim_kc)
+ dm_kcopyd_client_destroy(dmz->reclaim_kc);
+ if (dmz->flush_wq)
+ destroy_workqueue(dmz->flush_wq);
+ if (dmz->chunk_wq)
+ destroy_workqueue(dmz->chunk_wq);
+ if (dmz->bio_set)
+ bioset_free(dmz->bio_set);
+ dmz_cleanup_meta(dmz);
+ dm_put_device(ti, dmz->ddev);
+ }
+
+ kfree(dmz);
+
+ return ret;
+
+}
+
+/*
+ * Cleanup target.
+ */
+static void dmz_dtr(struct dm_target *ti)
+{
+ struct dmz_target *dmz = ti->private;
+
+ dmz_dev_info(dmz, "Removing target device\n");
+
+ flush_workqueue(dmz->chunk_wq);
+ destroy_workqueue(dmz->chunk_wq);
+
+ cancel_delayed_work_sync(&dmz->reclaim_work);
+ destroy_workqueue(dmz->reclaim_wq);
+ dm_kcopyd_client_destroy(dmz->reclaim_kc);
+
+ cancel_delayed_work_sync(&dmz->flush_work);
+ destroy_workqueue(dmz->flush_wq);
+
+ dmz_flush_mblocks(dmz);
+
+ bioset_free(dmz->bio_set);
+
+ dmz_cleanup_meta(dmz);
+
+ dm_put_device(ti, dmz->ddev);
+
+ kfree(dmz);
+}
+
+/*
+ * Setup target request queue limits.
+ */
+static void dmz_io_hints(struct dm_target *ti,
+ struct queue_limits *limits)
+{
+ struct dmz_target *dmz = ti->private;
+ unsigned int chunk_sectors = dmz->zone_nr_sectors;
+
+ limits->logical_block_size = DMZ_BLOCK_SIZE;
+ limits->physical_block_size = DMZ_BLOCK_SIZE;
+
+ blk_limits_io_min(limits, DMZ_BLOCK_SIZE);
+ blk_limits_io_opt(limits, DMZ_BLOCK_SIZE);
+
+ limits->discard_alignment = DMZ_BLOCK_SIZE;
+ limits->discard_granularity = DMZ_BLOCK_SIZE;
+ limits->max_discard_sectors = chunk_sectors;
+ limits->max_hw_discard_sectors = chunk_sectors;
+ limits->max_write_zeroes_sectors = chunk_sectors;
+
+ /* FS hint to try to align to the device zone size */
+ limits->chunk_sectors = chunk_sectors;
+ limits->max_sectors = chunk_sectors;
+
+ /* We are exposing a drive-managed zone model */
+ limits->zoned = BLK_ZONED_NONE;
+}
+
+/*
+ * Pass on ioctl to the backend device.
+ */
+static int dmz_prepare_ioctl(struct dm_target *ti,
+ struct block_device **bdev, fmode_t *mode)
+{
+ struct dmz_target *dmz = ti->private;
+
+ *bdev = dmz->zbd;
+
+ return 0;
+}
+
+/*
+ * Stop reclaim before suspend.
+ */
+static void dmz_presuspend(struct dm_target *ti)
+{
+ struct dmz_target *dmz = ti->private;
+
+ dmz_dev_debug(dmz, "Pre-suspend\n");
+
+ /* Enter suspend state */
+ set_bit(DMZ_SUSPENDED, &dmz->flags);
+ smp_mb__after_atomic();
+
+ /* Stop reclaim */
+ cancel_delayed_work_sync(&dmz->reclaim_work);
+}
+
+/*
+ * Restart reclaim if suspend failed.
+ */
+static void dmz_presuspend_undo(struct dm_target *ti)
+{
+ struct dmz_target *dmz = ti->private;
+
+ dmz_dev_debug(dmz, "Pre-suspend undo\n");
+
+ /* Clear suspend state */
+ clear_bit_unlock(DMZ_SUSPENDED, &dmz->flags);
+ smp_mb__after_atomic();
+
+ /* Restart reclaim */
+ mod_delayed_work(dmz->reclaim_wq, &dmz->reclaim_work, 0);
+}
+
+/*
+ * Stop works and flush on suspend.
+ */
+static void dmz_postsuspend(struct dm_target *ti)
+{
+ struct dmz_target *dmz = ti->private;
+
+ dmz_dev_debug(dmz, "Post-suspend\n");
+
+ /* Stop works */
+ flush_workqueue(dmz->chunk_wq);
+ flush_workqueue(dmz->flush_wq);
+}
+
+/*
+ * Refresh zone information before resuming.
+ */
+static int dmz_preresume(struct dm_target *ti)
+{
+ struct dmz_target *dmz = ti->private;
+
+ if (!test_bit(DMZ_SUSPENDED, &dmz->flags))
+ return 0;
+
+ dmz_dev_debug(dmz, "Pre-resume\n");
+
+ /* Refresh zone information */
+ return dmz_resume_meta(dmz);
+}
+
+/*
+ * Resume.
+ */
+static void dmz_resume(struct dm_target *ti)
+{
+ struct dmz_target *dmz = ti->private;
+
+ if (!test_bit(DMZ_SUSPENDED, &dmz->flags))
+ return;
+
+ dmz_dev_debug(dmz, "Resume\n");
+
+ /* Clear suspend state */
+ clear_bit_unlock(DMZ_SUSPENDED, &dmz->flags);
+ smp_mb__after_atomic();
+
+ /* Restart reclaim */
+ mod_delayed_work(dmz->reclaim_wq, &dmz->reclaim_work, 0);
+
+}
+
+static int dmz_iterate_devices(struct dm_target *ti,
+ iterate_devices_callout_fn fn, void *data)
+{
+ struct dmz_target *dmz = ti->private;
+ sector_t offset = dmz->zbd_capacity -
+ ((sector_t)dmz->nr_chunks * dmz->zone_nr_sectors);
+
+ return fn(ti, dmz->ddev, offset, ti->len, data);
+}
+
+static struct target_type dmz_type = {
+ .name = "dm-zoned",
+ .version = {1, 0, 0},
+ .features = DM_TARGET_SINGLETON | DM_TARGET_ZONED_HM,
+ .module = THIS_MODULE,
+ .ctr = dmz_ctr,
+ .dtr = dmz_dtr,
+ .map = dmz_map,
+ .io_hints = dmz_io_hints,
+ .prepare_ioctl = dmz_prepare_ioctl,
+ .presuspend = dmz_presuspend,
+ .presuspend_undo = dmz_presuspend_undo,
+ .postsuspend = dmz_postsuspend,
+ .preresume = dmz_preresume,
+ .resume = dmz_resume,
+ .iterate_devices = dmz_iterate_devices,
+};
+
+static int __init dmz_init(void)
+{
+ dmz_info("Zoned block device target (C) Western Digital\n");
+
+ return dm_register_target(&dmz_type);
+}
+
+static void __exit dmz_exit(void)
+{
+ dm_unregister_target(&dmz_type);
+}
+
+module_init(dmz_init);
+module_exit(dmz_exit);
+
+MODULE_DESCRIPTION(DM_NAME " target for zoned block devices");
+MODULE_AUTHOR("Damien Le Moal <damien.lemoal@wdc.com>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/md/dm-zoned-metadata.c b/drivers/md/dm-zoned-metadata.c
new file mode 100644
index 0000000..7440b47
--- /dev/null
+++ b/drivers/md/dm-zoned-metadata.c
@@ -0,0 +1,2225 @@
+/*
+ * Drive-managed zoned block device target
+ * Copyright (C) 2017 Western Digital Corporation or its affiliates.
+ *
+ * Written by: Damien Le Moal <damien.lemoal@wdc.com>
+ *
+ * This software is distributed under the terms of the GNU General Public
+ * License version 2, or any later version, "as is," without technical
+ * support, and WITHOUT ANY WARRANTY, without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+ */
+
+#include <linux/module.h>
+#include <linux/crc32.h>
+
+#include "dm-zoned.h"
+
+/*
+ * Allocate a metadata block.
+ */
+static struct dmz_mblock *dmz_alloc_mblock(struct dmz_target *dmz,
+ sector_t mblk_no)
+{
+ struct dmz_mblock *mblk = NULL;
+
+ /* See if we can reuse cached blocks */
+ if (dmz->max_nr_mblks &&
+ atomic_read(&dmz->nr_mblks) > dmz->max_nr_mblks) {
+
+ spin_lock(&dmz->mblk_lock);
+
+ if (list_empty(&dmz->mblk_lru_list) &&
+ !list_empty(&dmz->mblk_dirty_list))
+ /* Cleanup dirty blocks */
+ dmz_trigger_flush(dmz);
+
+ mblk = list_first_entry_or_null(&dmz->mblk_lru_list,
+ struct dmz_mblock, link);
+ if (mblk) {
+ list_del_init(&mblk->link);
+ rb_erase(&mblk->node, &dmz->mblk_rbtree);
+ mblk->no = mblk_no;
+ }
+
+ spin_unlock(&dmz->mblk_lock);
+
+ if (mblk)
+ return mblk;
+ }
+
+ /* Allocate a new block */
+ mblk = kmalloc(sizeof(struct dmz_mblock), GFP_NOIO);
+ if (!mblk)
+ return NULL;
+
+ mblk->page = alloc_page(GFP_NOIO);
+ if (!mblk->page) {
+ kfree(mblk);
+ return NULL;
+ }
+
+ RB_CLEAR_NODE(&mblk->node);
+ INIT_LIST_HEAD(&mblk->link);
+ atomic_set(&mblk->ref, 0);
+ mblk->state = 0;
+ mblk->no = mblk_no;
+ mblk->data = page_address(mblk->page);
+
+ atomic_inc(&dmz->nr_mblks);
+
+ return mblk;
+}
+
+/*
+ * Free a metadata block.
+ */
+static void dmz_free_mblock(struct dmz_target *dmz, struct dmz_mblock *mblk)
+{
+ __free_pages(mblk->page, 0);
+ kfree(mblk);
+
+ atomic_dec(&dmz->nr_mblks);
+}
+
+/*
+ * Insert a metadata block in the rbtree.
+ */
+static void dmz_insert_mblock(struct dmz_target *dmz,
+ struct dmz_mblock *mblk)
+{
+ struct rb_root *root = &dmz->mblk_rbtree;
+ struct rb_node **new = &(root->rb_node), *parent = NULL;
+ struct dmz_mblock *b;
+
+ /* Figure out where to put the new node */
+ while (*new) {
+ b = container_of(*new, struct dmz_mblock, node);
+ parent = *new;
+ new = (b->no < mblk->no) ?
+ &((*new)->rb_left) : &((*new)->rb_right);
+ }
+
+ /* Add new node and rebalance tree */
+ rb_link_node(&mblk->node, parent, new);
+ rb_insert_color(&mblk->node, root);
+}
+
+/*
+ * Lookup a metadata block in the rbtree.
+ */
+static struct dmz_mblock *dmz_lookup_mblock(struct dmz_target *dmz,
+ sector_t mblk_no)
+{
+ struct rb_root *root = &dmz->mblk_rbtree;
+ struct rb_node *node = root->rb_node;
+ struct dmz_mblock *mblk;
+
+ while (node) {
+ mblk = container_of(node, struct dmz_mblock, node);
+ if (mblk->no == mblk_no)
+ return mblk;
+ node = (mblk->no < mblk_no) ? node->rb_left : node->rb_right;
+ }
+
+ return NULL;
+}
+
+/*
+ * Metadata block BIO end callback.
+ */
+static void dmz_mblock_bio_end_io(struct bio *bio)
+{
+ struct dmz_mblock *mblk = bio->bi_private;
+ int flag;
+
+ if (bio->bi_error)
+ set_bit(DMZ_META_ERROR, &mblk->state);
+
+ if (bio_op(bio) == REQ_OP_WRITE)
+ flag = DMZ_META_WRITING;
+ else
+ flag = DMZ_META_READING;
+
+ clear_bit_unlock(flag, &mblk->state);
+ smp_mb__after_atomic();
+ wake_up_bit(&mblk->state, flag);
+
+ bio_put(bio);
+}
+
+/*
+ * Read a metadata block from disk.
+ */
+static struct dmz_mblock *dmz_fetch_mblock(struct dmz_target *dmz,
+ sector_t mblk_no)
+{
+ struct dmz_mblock *mblk;
+ sector_t block = dmz->sb[dmz->mblk_primary].block + mblk_no;
+ struct bio *bio;
+
+ /* Get block and insert it */
+ mblk = dmz_alloc_mblock(dmz, mblk_no);
+ if (!mblk)
+ return NULL;
+
+ spin_lock(&dmz->mblk_lock);
+ atomic_inc(&mblk->ref);
+ set_bit(DMZ_META_READING, &mblk->state);
+ dmz_insert_mblock(dmz, mblk);
+ spin_unlock(&dmz->mblk_lock);
+
+ bio = bio_alloc(GFP_NOIO, 1);
+ if (!bio) {
+ dmz_free_mblock(dmz, mblk);
+ return NULL;
+ }
+
+ bio->bi_iter.bi_sector = dmz_blk2sect(block);
+ bio->bi_bdev = dmz->zbd;
+ bio->bi_private = mblk;
+ bio->bi_end_io = dmz_mblock_bio_end_io;
+ bio_set_op_attrs(bio, REQ_OP_READ, REQ_META | REQ_PRIO);
+ bio_add_page(bio, mblk->page, DMZ_BLOCK_SIZE, 0);
+ submit_bio(bio);
+
+ return mblk;
+}
+
+/*
+ * Free metadata blocks.
+ */
+static unsigned long dmz_shrink_mblock_cache(struct dmz_target *dmz,
+ unsigned long limit)
+{
+ struct dmz_mblock *mblk;
+ unsigned long count = 0;
+
+ if (!dmz->max_nr_mblks)
+ return 0;
+
+ while (!list_empty(&dmz->mblk_lru_list) &&
+ atomic_read(&dmz->nr_mblks) > dmz->min_nr_mblks &&
+ count < limit) {
+ mblk = list_first_entry(&dmz->mblk_lru_list,
+ struct dmz_mblock, link);
+ list_del_init(&mblk->link);
+ rb_erase(&mblk->node, &dmz->mblk_rbtree);
+ dmz_free_mblock(dmz, mblk);
+ count++;
+ }
+
+ return count;
+}
+
+/*
+ * For mblock shrinker: get the number of unused metadata blocks in the cache.
+ */
+static unsigned long dmz_mblock_shrinker_count(struct shrinker *shrink,
+ struct shrink_control *sc)
+{
+ struct dmz_target *dmz =
+ container_of(shrink, struct dmz_target, mblk_shrinker);
+
+ return atomic_read(&dmz->nr_mblks);
+}
+
+/*
+ * For mblock shrinker: scan unused metadata blocks and shrink the cache.
+ */
+static unsigned long dmz_mblock_shrinker_scan(struct shrinker *shrink,
+ struct shrink_control *sc)
+{
+ struct dmz_target *dmz =
+ container_of(shrink, struct dmz_target, mblk_shrinker);
+ unsigned long count;
+
+ spin_lock(&dmz->mblk_lock);
+ count = dmz_shrink_mblock_cache(dmz, sc->nr_to_scan);
+ spin_unlock(&dmz->mblk_lock);
+
+ return count ? count : SHRINK_STOP;
+}
+
+/*
+ * Release a metadata block.
+ */
+static void dmz_release_mblock(struct dmz_target *dmz, struct dmz_mblock *mblk)
+{
+
+ if (!mblk)
+ return;
+
+ spin_lock(&dmz->mblk_lock);
+
+ if (atomic_dec_and_test(&mblk->ref)) {
+ if (test_bit(DMZ_META_ERROR, &mblk->state)) {
+ rb_erase(&mblk->node, &dmz->mblk_rbtree);
+ dmz_free_mblock(dmz, mblk);
+ } else if (!test_bit(DMZ_META_DIRTY, &mblk->state)) {
+ list_add_tail(&mblk->link, &dmz->mblk_lru_list);
+ dmz_shrink_mblock_cache(dmz, 1);
+ }
+ }
+
+ spin_unlock(&dmz->mblk_lock);
+}
+
+/*
+ * Get a metadata block from the rbtree. If the block
+ * is not present, read it from disk.
+ */
+static struct dmz_mblock *dmz_get_mblock(struct dmz_target *dmz,
+ sector_t mblk_no)
+{
+ struct dmz_mblock *mblk;
+
+ /* Check rbtree */
+ spin_lock(&dmz->mblk_lock);
+ mblk = dmz_lookup_mblock(dmz, mblk_no);
+ if (mblk) {
+ /* Cache hit: remove block from LRU list */
+ if (atomic_inc_return(&mblk->ref) == 1 &&
+ !test_bit(DMZ_META_DIRTY, &mblk->state))
+ list_del_init(&mblk->link);
+ }
+ spin_unlock(&dmz->mblk_lock);
+
+ if (!mblk) {
+ /* Cache miss: read the block from disk */
+ mblk = dmz_fetch_mblock(dmz, mblk_no);
+ if (!mblk)
+ return ERR_PTR(-ENOMEM);
+ }
+
+ /* Wait for on-going read I/O and check for error */
+ wait_on_bit_io(&mblk->state, DMZ_META_READING,
+ TASK_UNINTERRUPTIBLE);
+ if (test_bit(DMZ_META_ERROR, &mblk->state)) {
+ dmz_release_mblock(dmz, mblk);
+ return ERR_PTR(-EIO);
+ }
+
+ return mblk;
+}
+
+/*
+ * Mark a metadata block dirty.
+ */
+static void dmz_dirty_mblock(struct dmz_target *dmz, struct dmz_mblock *mblk)
+{
+ spin_lock(&dmz->mblk_lock);
+ if (!test_and_set_bit(DMZ_META_DIRTY, &mblk->state))
+ list_add_tail(&mblk->link, &dmz->mblk_dirty_list);
+ spin_unlock(&dmz->mblk_lock);
+}
+
+/*
+ * Issue a metadata block write BIO.
+ */
+static void dmz_write_mblock(struct dmz_target *dmz, struct dmz_mblock *mblk,
+ unsigned int set)
+{
+ sector_t block = dmz->sb[set].block + mblk->no;
+ struct bio *bio;
+
+ bio = bio_alloc(GFP_NOIO, 1);
+ if (!bio) {
+ set_bit(DMZ_META_ERROR, &mblk->state);
+ return;
+ }
+
+ set_bit(DMZ_META_WRITING, &mblk->state);
+
+ bio->bi_iter.bi_sector = dmz_blk2sect(block);
+ bio->bi_bdev = dmz->zbd;
+ bio->bi_private = mblk;
+ bio->bi_end_io = dmz_mblock_bio_end_io;
+ bio_set_op_attrs(bio, REQ_OP_WRITE, REQ_META | REQ_PRIO);
+ bio_add_page(bio, mblk->page, DMZ_BLOCK_SIZE, 0);
+ submit_bio(bio);
+}
+
+/*
+ * Sync read/write a block.
+ */
+static int dmz_rdwr_block_sync(struct dmz_target *dmz, int op, sector_t block,
+ struct page *page)
+{
+ struct bio *bio;
+ int ret;
+
+ bio = bio_alloc(GFP_NOIO, 1);
+ if (!bio)
+ return -ENOMEM;
+
+ bio->bi_iter.bi_sector = dmz_blk2sect(block);
+ bio->bi_bdev = dmz->zbd;
+ bio_set_op_attrs(bio, op, REQ_SYNC | REQ_META | REQ_PRIO);
+ bio_add_page(bio, page, DMZ_BLOCK_SIZE, 0);
+ ret = submit_bio_wait(bio);
+ bio_put(bio);
+
+ return ret;
+}
+
+/*
+ * Write super block of the specified metadata set.
+ */
+static int dmz_write_sb(struct dmz_target *dmz, unsigned int set)
+{
+ sector_t block = dmz->sb[set].block;
+ struct dmz_mblock *mblk = dmz->sb[set].mblk;
+ struct dmz_super *sb = dmz->sb[set].sb;
+ u64 sb_gen = dmz->sb_gen + 1;
+ int ret;
+
+ sb->magic = cpu_to_le32(DMZ_MAGIC);
+ sb->version = cpu_to_le32(DMZ_META_VER);
+
+ sb->gen = cpu_to_le64(sb_gen);
+
+ sb->sb_block = cpu_to_le64(block);
+ sb->nr_meta_blocks = cpu_to_le32(dmz->nr_meta_blocks);
+ sb->nr_reserved_seq = cpu_to_le32(dmz->nr_reserved_seq);
+ sb->nr_chunks = cpu_to_le32(dmz->nr_chunks);
+
+ sb->nr_map_blocks = cpu_to_le32(dmz->nr_map_blocks);
+ sb->nr_bitmap_blocks = cpu_to_le32(dmz->nr_bitmap_blocks);
+
+ sb->crc = 0;
+ sb->crc = cpu_to_le32(crc32_le(sb_gen,
+ (unsigned char *)sb, DMZ_BLOCK_SIZE));
+
+ ret = dmz_rdwr_block_sync(dmz, REQ_OP_WRITE, block, mblk->page);
+ if (ret == 0)
+ ret = blkdev_issue_flush(dmz->zbd, GFP_KERNEL, NULL);
+
+ return ret;
+}
+
+/*
+ * Write dirty metadata blocks to the specified set.
+ */
+static int dmz_write_dirty_mblocks(struct dmz_target *dmz,
+ struct list_head *write_list,
+ unsigned int set)
+{
+ struct dmz_mblock *mblk;
+ struct blk_plug plug;
+ int ret = 0;
+
+ /* Issue writes */
+ blk_start_plug(&plug);
+ list_for_each_entry(mblk, write_list, link)
+ dmz_write_mblock(dmz, mblk, set);
+ blk_finish_plug(&plug);
+
+ /* Wait for completion */
+ list_for_each_entry(mblk, write_list, link) {
+ wait_on_bit_io(&mblk->state, DMZ_META_WRITING,
+ TASK_UNINTERRUPTIBLE);
+ if (test_bit(DMZ_META_ERROR, &mblk->state)) {
+ dmz_dev_err(dmz, "Write metablock %u/%llu failed\n",
+ set, (u64)mblk->no);
+ clear_bit(DMZ_META_ERROR, &mblk->state);
+ ret = -EIO;
+ }
+ }
+
+ /* Flush drive cache (this will also sync data) */
+ if (ret == 0)
+ ret = blkdev_issue_flush(dmz->zbd, GFP_KERNEL, NULL);
+
+ return ret;
+}
+
+/*
+ * Log dirty metadata blocks.
+ */
+static int dmz_log_dirty_mblocks(struct dmz_target *dmz,
+ struct list_head *write_list)
+{
+ unsigned int log_set = dmz->mblk_primary ^ 0x1;
+ int ret;
+
+ dmz_dev_debug(dmz, "Log metadata to set %u, gen %llu\n",
+ log_set, dmz->sb_gen + 1);
+
+ /* Write dirty blocks to the log */
+ ret = dmz_write_dirty_mblocks(dmz, write_list, log_set);
+ if (ret)
+ return ret;
+
+ /*
+ * No error so far: now validate the log by updating the
+ * log index super block generation.
+ */
+ ret = dmz_write_sb(dmz, log_set);
+ if (ret)
+ return ret;
+
+ return 0;
+}
+
+/*
+ * Flush dirty metadata blocks.
+ */
+int dmz_flush_mblocks(struct dmz_target *dmz)
+{
+ struct dmz_mblock *mblk;
+ struct list_head write_list;
+ int ret;
+
+ INIT_LIST_HEAD(&write_list);
+
+ /*
+ * Prevent BIOs to zones and reclaim. This ensure exclusive
+ * access to metadata.
+ */
+ down_write(&dmz->mblk_sem);
+
+ /* If there are no dirty metadata blocks, just flush the device cache */
+ if (list_empty(&dmz->mblk_dirty_list)) {
+ ret = blkdev_issue_flush(dmz->zbd, GFP_KERNEL, NULL);
+ goto out;
+ }
+
+ /*
+ * The primary metadata set is still clean. Keep it this way until
+ * all updates are successful in the secondary set. That is, use
+ * the secondary set as a log.
+ */
+ list_splice_init(&dmz->mblk_dirty_list, &write_list);
+ ret = dmz_log_dirty_mblocks(dmz, &write_list);
+ if (ret)
+ goto out;
+
+ /*
+ * The log is on disk. It is now safe to update in place
+ * in the primary metadata set.
+ */
+ dmz_dev_debug(dmz, "Commit metadata to set %u, gen %llu\n",
+ dmz->mblk_primary, dmz->sb_gen + 1);
+ ret = dmz_write_dirty_mblocks(dmz, &write_list, dmz->mblk_primary);
+ if (ret)
+ goto out;
+
+ ret = dmz_write_sb(dmz, dmz->mblk_primary);
+ if (ret)
+ goto out;
+
+ while (!list_empty(&write_list)) {
+ mblk = list_first_entry(&write_list,
+ struct dmz_mblock, link);
+ list_del_init(&mblk->link);
+
+ clear_bit(DMZ_META_DIRTY, &mblk->state);
+ if (atomic_read(&mblk->ref) == 0)
+ list_add_tail(&mblk->link, &dmz->mblk_lru_list);
+
+ }
+
+ dmz->sb_gen++;
+
+out:
+ if (ret && !list_empty(&write_list))
+ list_splice(&write_list, &dmz->mblk_dirty_list);
+
+ up_write(&dmz->mblk_sem);
+
+ return ret;
+}
+
+/*
+ * Check super block.
+ */
+static int dmz_check_sb(struct dmz_target *dmz, struct dmz_super *sb)
+{
+ unsigned int nr_meta_zones, nr_data_zones;
+ u32 crc, stored_crc;
+ u64 gen;
+
+ gen = le64_to_cpu(sb->gen);
+ stored_crc = le32_to_cpu(sb->crc);
+ sb->crc = 0;
+ crc = crc32_le(gen, (unsigned char *)sb, DMZ_BLOCK_SIZE);
+ if (crc != stored_crc) {
+ dmz_dev_err(dmz,
+ "Invalid checksum (needed 0x%08x, got 0x%08x)\n",
+ crc, stored_crc);
+ return -ENXIO;
+ }
+
+ if (le32_to_cpu(sb->magic) != DMZ_MAGIC) {
+ dmz_dev_err(dmz,
+ "Invalid meta magic (need 0x%08x, got 0x%08x)\n",
+ DMZ_MAGIC, le32_to_cpu(sb->magic));
+ return -ENXIO;
+ }
+
+ if (le32_to_cpu(sb->version) != DMZ_META_VER) {
+ dmz_dev_err(dmz, "Invalid meta version (need %d, got %d)\n",
+ DMZ_META_VER, le32_to_cpu(sb->version));
+ return -ENXIO;
+ }
+
+ nr_meta_zones =
+ (le32_to_cpu(sb->nr_meta_blocks) + dmz->zone_nr_blocks - 1)
+ >> dmz->zone_nr_blocks_shift;
+ if (!nr_meta_zones ||
+ nr_meta_zones >= dmz->nr_rnd_zones) {
+ dmz_dev_err(dmz, "Invalid number of metadata blocks\n");
+ return -ENXIO;
+ }
+
+ if (!le32_to_cpu(sb->nr_reserved_seq) ||
+ le32_to_cpu(sb->nr_reserved_seq) >=
+ (dmz->nr_useable_zones - nr_meta_zones)) {
+ dmz_dev_err(dmz,
+ "Invalid number of reserved sequential zones\n");
+ return -ENXIO;
+ }
+
+ nr_data_zones = dmz->nr_useable_zones -
+ (nr_meta_zones * 2 + le32_to_cpu(sb->nr_reserved_seq));
+ if (le32_to_cpu(sb->nr_chunks) > nr_data_zones) {
+ dmz_dev_err(dmz, "Invalid number of chunks %u / %u\n",
+ le32_to_cpu(sb->nr_chunks), nr_data_zones);
+ return -ENXIO;
+ }
+
+ /* OK */
+ dmz->nr_meta_blocks = le32_to_cpu(sb->nr_meta_blocks);
+ dmz->nr_reserved_seq = le32_to_cpu(sb->nr_reserved_seq);
+ dmz->nr_chunks = le32_to_cpu(sb->nr_chunks);
+ dmz->nr_map_blocks = le32_to_cpu(sb->nr_map_blocks);
+ dmz->nr_bitmap_blocks = le32_to_cpu(sb->nr_bitmap_blocks);
+ dmz->nr_meta_zones = nr_meta_zones;
+ dmz->nr_data_zones = nr_data_zones;
+
+ return 0;
+}
+
+/*
+ * Read the first or second super block from disk.
+ */
+static int dmz_read_sb(struct dmz_target *dmz, unsigned int set)
+{
+ return dmz_rdwr_block_sync(dmz, REQ_OP_READ,
+ dmz->sb[set].block,
+ dmz->sb[set].mblk->page);
+}
+
+/*
+ * Determine the position of the secondary super blocks on disk.
+ * This is used only if a corruption of the primary super block
+ * is detected.
+ */
+static int dmz_lookup_secondary_sb(struct dmz_target *dmz)
+{
+ struct dmz_mblock *mblk;
+ int i;
+
+ /* Allocate a block */
+ mblk = dmz_alloc_mblock(dmz, 0);
+ if (!mblk)
+ return -ENOMEM;
+
+ dmz->sb[1].mblk = mblk;
+ dmz->sb[1].sb = mblk->data;
+
+ /* Bad first super block: search for the second one */
+ dmz->sb[1].block = dmz->sb[0].block + dmz->zone_nr_blocks;
+ for (i = 0; i < dmz->nr_rnd_zones - 1; i++) {
+ if (dmz_read_sb(dmz, 1) != 0)
+ break;
+ if (le32_to_cpu(dmz->sb[1].sb->magic) == DMZ_MAGIC)
+ return 0;
+ dmz->sb[1].block += dmz->zone_nr_blocks;
+ }
+
+ dmz_free_mblock(dmz, mblk);
+ dmz->sb[1].mblk = NULL;
+
+ return -EIO;
+}
+
+/*
+ * Read the first or second super block from disk.
+ */
+static int dmz_get_sb(struct dmz_target *dmz, unsigned int set)
+{
+ struct dmz_mblock *mblk;
+ int ret;
+
+ /* Allocate a block */
+ mblk = dmz_alloc_mblock(dmz, 0);
+ if (!mblk)
+ return -ENOMEM;
+
+ dmz->sb[set].mblk = mblk;
+ dmz->sb[set].sb = mblk->data;
+
+ /* Read super block */
+ ret = dmz_read_sb(dmz, set);
+ if (ret) {
+ dmz_free_mblock(dmz, mblk);
+ dmz->sb[set].mblk = NULL;
+ return ret;
+ }
+
+ return 0;
+}
+
+/*
+ * Recover a metadata set.
+ */
+static int dmz_recover_mblocks(struct dmz_target *dmz, unsigned int dst_set)
+{
+ unsigned int src_set = dst_set ^ 0x1;
+ struct page *page;
+ int i, ret;
+
+ dmz_dev_warn(dmz, "Metadata set %u invalid: recovering\n",
+ dst_set);
+
+ if (dst_set == 0)
+ dmz->sb[0].block = dmz_start_block(dmz, dmz->sb_zone);
+ else
+ dmz->sb[1].block = dmz->sb[0].block +
+ (dmz->nr_meta_zones * dmz->zone_nr_blocks);
+
+ page = alloc_page(GFP_KERNEL);
+ if (!page)
+ return -ENOMEM;
+
+ /* Copy metadata blocks */
+ for (i = 1; i < dmz->nr_meta_blocks; i++) {
+ ret = dmz_rdwr_block_sync(dmz, REQ_OP_READ,
+ dmz->sb[src_set].block + i,
+ page);
+ if (ret)
+ goto out;
+ ret = dmz_rdwr_block_sync(dmz, REQ_OP_WRITE,
+ dmz->sb[dst_set].block + i,
+ page);
+ if (ret)
+ goto out;
+ }
+
+ /* Finalize with the super block */
+ if (!dmz->sb[dst_set].mblk) {
+ dmz->sb[dst_set].mblk = dmz_alloc_mblock(dmz, 0);
+ if (!dmz->sb[dst_set].mblk) {
+ ret = -ENOMEM;
+ goto out;
+ }
+ dmz->sb[dst_set].sb = dmz->sb[dst_set].mblk->data;
+ }
+
+ ret = dmz_write_sb(dmz, dst_set);
+
+out:
+ __free_pages(page, 0);
+
+ return ret;
+}
+
+/*
+ * Get super block from disk.
+ */
+static int dmz_load_sb(struct dmz_target *dmz)
+{
+ bool sb_good[2] = {false, false};
+ u64 sb_gen[2] = {0, 0};
+ int ret;
+
+ /* Read and check the primary super block */
+ dmz->sb[0].block = dmz_start_block(dmz, dmz->sb_zone);
+ ret = dmz_get_sb(dmz, 0);
+ if (ret) {
+ dmz_dev_err(dmz, "Read primary super block failed\n");
+ return ret;
+ }
+
+ ret = dmz_check_sb(dmz, dmz->sb[0].sb);
+
+ /* Read and check secondary super block */
+ if (ret == 0) {
+ sb_good[0] = true;
+ dmz->sb[1].block = dmz->sb[0].block +
+ (dmz->nr_meta_zones * dmz->zone_nr_blocks);
+ ret = dmz_get_sb(dmz, 1);
+ } else {
+ ret = dmz_lookup_secondary_sb(dmz);
+ }
+ if (ret) {
+ dmz_dev_err(dmz, "Read secondary super block failed\n");
+ return ret;
+ }
+
+ ret = dmz_check_sb(dmz, dmz->sb[1].sb);
+ if (ret == 0)
+ sb_good[1] = true;
+
+ /* Use highest generation sb first */
+ if (!sb_good[0] && !sb_good[1]) {
+ dmz_dev_err(dmz, "No valid super block found\n");
+ return -EIO;
+ }
+
+ if (sb_good[0])
+ sb_gen[0] = le64_to_cpu(dmz->sb[0].sb->gen);
+ else
+ ret = dmz_recover_mblocks(dmz, 0);
+
+ if (sb_good[1])
+ sb_gen[1] = le64_to_cpu(dmz->sb[1].sb->gen);
+ else
+ ret = dmz_recover_mblocks(dmz, 1);
+
+ if (ret) {
+ dmz_dev_err(dmz, "Recovery failed\n");
+ return -EIO;
+ }
+
+ if (sb_gen[0] >= sb_gen[1]) {
+ dmz->sb_gen = sb_gen[0];
+ dmz->mblk_primary = 0;
+ } else {
+ dmz->sb_gen = sb_gen[1];
+ dmz->mblk_primary = 1;
+ }
+
+ dmz_dev_debug(dmz, "Using super block %u (gen %llu)\n",
+ dmz->mblk_primary, dmz->sb_gen);
+
+ return 0;
+}
+
+/*
+ * Initialize a zone descriptor.
+ */
+static int dmz_init_zone(struct dmz_target *dmz, struct dm_zone *zone,
+ struct blk_zone *blkz)
+{
+
+ /* Ignore the eventual last runt (smaller) zone */
+ if (blkz->len != dmz->zone_nr_sectors) {
+ if (blkz->start + blkz->len == dmz->zbd_capacity)
+ return 0;
+ return -ENXIO;
+ }
+
+ INIT_LIST_HEAD(&zone->link);
+ atomic_set(&zone->refcount, 0);
+ zone->chunk = DMZ_MAP_UNMAPPED;
+
+ if (blkz->type == BLK_ZONE_TYPE_CONVENTIONAL) {
+ set_bit(DMZ_RND, &zone->flags);
+ dmz->nr_rnd_zones++;
+ } else if (blkz->type == BLK_ZONE_TYPE_SEQWRITE_REQ ||
+ blkz->type == BLK_ZONE_TYPE_SEQWRITE_PREF) {
+ set_bit(DMZ_SEQ, &zone->flags);
+ } else {
+ return -ENXIO;
+ }
+
+ if (blkz->cond == BLK_ZONE_COND_OFFLINE)
+ set_bit(DMZ_OFFLINE, &zone->flags);
+ else if (blkz->cond == BLK_ZONE_COND_READONLY)
+ set_bit(DMZ_READ_ONLY, &zone->flags);
+
+ if (dmz_is_rnd(zone))
+ zone->wp_block = 0;
+ else
+ zone->wp_block = dmz_sect2blk(blkz->wp - blkz->start);
+
+ if (!dmz_is_offline(zone) && !dmz_is_readonly(zone)) {
+ dmz->nr_useable_zones++;
+ if (dmz_is_rnd(zone)) {
+ dmz->nr_rnd_zones++;
+ if (!dmz->sb_zone) {
+ /* Super block zone */
+ dmz->sb_zone = zone;
+ }
+ }
+ }
+
+ return 0;
+}
+
+/*
+ * Free zones descriptors.
+ */
+static void dmz_drop_zones(struct dmz_target *dmz)
+{
+ kfree(dmz->zones);
+ dmz->zones = NULL;
+}
+
+/*
+ * Allocate and initialize zone descriptors using the zone
+ * information from disk.
+ */
+static int dmz_init_zones(struct dmz_target *dmz)
+{
+ struct dm_zone *zone;
+ struct blk_zone *blkz;
+ unsigned int nr_blkz;
+ sector_t sector = 0;
+ int i, ret = 0;
+
+ /* Init */
+ dmz->zone_nr_sectors = dmz->zbdq->limits.chunk_sectors;
+ dmz->zone_nr_sectors_shift = ilog2(dmz->zone_nr_sectors);
+
+ dmz->zone_nr_blocks = dmz_sect2blk(dmz->zone_nr_sectors);
+ dmz->zone_nr_blocks_shift = ilog2(dmz->zone_nr_blocks);
+
+ dmz->zone_bitmap_size = dmz->zone_nr_blocks >> 3;
+ dmz->zone_nr_bitmap_blocks =
+ dmz->zone_bitmap_size >> DMZ_BLOCK_SHIFT;
+
+ dmz->nr_zones = (dmz->zbd_capacity + dmz->zone_nr_sectors - 1)
+ >> dmz->zone_nr_sectors_shift;
+
+ /* Allocate zone array */
+ dmz->zones = kcalloc(dmz->nr_zones, sizeof(struct dm_zone), GFP_KERNEL);
+ if (!dmz->zones)
+ return -ENOMEM;
+
+ dmz_dev_info(dmz, "Using %zu B for zone information\n",
+ sizeof(struct dm_zone) * dmz->nr_zones);
+
+ /* Get zone information */
+ nr_blkz = DMZ_REPORT_NR_ZONES;
+ blkz = kcalloc(nr_blkz, sizeof(struct blk_zone), GFP_KERNEL);
+ if (!blkz) {
+ ret = -ENOMEM;
+ goto out;
+ }
+
+ /*
+ * Get zone information and initialize zone descriptors.
+ * At the same time, determine where the super block
+ * should be: first block of the first randomly writable
+ * zone.
+ */
+ zone = dmz->zones;
+ while (sector < dmz->zbd_capacity) {
+
+ /* Get zone information */
+ nr_blkz = DMZ_REPORT_NR_ZONES;
+ ret = blkdev_report_zones(dmz->zbd, sector,
+ blkz, &nr_blkz,
+ GFP_KERNEL);
+ if (ret) {
+ dmz_dev_err(dmz, "Report zones failed %d\n", ret);
+ goto out;
+ }
+
+ /* Process report */
+ for (i = 0; i < nr_blkz; i++) {
+ ret = dmz_init_zone(dmz, zone, &blkz[i]);
+ if (ret)
+ goto out;
+ sector += dmz->zone_nr_sectors;
+ zone++;
+ }
+
+ }
+
+ /* The entire zone configuration of the disk should now be known */
+ if (sector < dmz->zbd_capacity) {
+ dmz_dev_err(dmz, "Failed to get zone information\n");
+ ret = -ENXIO;
+ goto out;
+ }
+
+out:
+ kfree(blkz);
+
+ if (ret)
+ dmz_drop_zones(dmz);
+
+ return ret;
+}
+
+/*
+ * Update a zone information.
+ */
+static int dmz_update_zone(struct dmz_target *dmz, struct dm_zone *zone)
+{
+ unsigned int nr_blkz = 1;
+ struct blk_zone blkz;
+ int ret;
+
+ /* Get zone information from disk */
+ ret = blkdev_report_zones(dmz->zbd, dmz_start_sect(dmz, zone),
+ &blkz, &nr_blkz,
+ GFP_KERNEL);
+ if (ret) {
+ dmz_dev_err(dmz, "Get zone %u report failed\n",
+ dmz_id(dmz, zone));
+ return ret;
+ }
+
+ clear_bit(DMZ_OFFLINE, &zone->flags);
+ clear_bit(DMZ_READ_ONLY, &zone->flags);
+ if (blkz.cond == BLK_ZONE_COND_OFFLINE)
+ set_bit(DMZ_OFFLINE, &zone->flags);
+ else if (blkz.cond == BLK_ZONE_COND_READONLY)
+ set_bit(DMZ_READ_ONLY, &zone->flags);
+
+ if (dmz_is_seq(zone))
+ zone->wp_block = dmz_sect2blk(blkz.wp - blkz.start);
+ else
+ zone->wp_block = 0;
+
+ return 0;
+}
+
+/*
+ * Check a zone write pointer position when the zone is marked
+ * with the sequential write error flag.
+ */
+static int dmz_handle_seq_write_err(struct dmz_target *dmz,
+ struct dm_zone *zone)
+{
+ unsigned int wp = 0;
+ int ret = 0;
+
+ wp = zone->wp_block;
+ ret = dmz_update_zone(dmz, zone);
+ if (ret != 0)
+ return ret;
+
+ dmz_dev_warn(dmz, "Processing zone %u write error (zone wp %u/%u)\n",
+ dmz_id(dmz, zone), zone->wp_block, wp);
+
+ if (zone->wp_block < wp)
+ dmz_invalidate_blocks(dmz, zone,
+ zone->wp_block,
+ wp - zone->wp_block);
+
+ return 0;
+}
+
+/*
+ * Check zone information after a resume.
+ */
+static int dmz_check_zones(struct dmz_target *dmz)
+{
+ struct dm_zone *zone;
+ sector_t wp_block;
+ unsigned int i;
+ int ret;
+
+ /* Check zones */
+ for (i = 0; i < dmz->nr_zones; i++) {
+
+ zone = dmz_get(dmz, i);
+ if (!zone) {
+ dmz_dev_err(dmz, "Unable to get zone %u\n", i);
+ return -EIO;
+ }
+
+ wp_block = zone->wp_block;
+
+ ret = dmz_update_zone(dmz, zone);
+ if (ret) {
+ dmz_dev_err(dmz, "Broken zone %u\n", i);
+ return ret;
+ }
+
+ if (dmz_is_offline(zone)) {
+ dmz_dev_warn(dmz, "Zone %u is offline\n", i);
+ continue;
+ }
+
+ /* Check write pointer */
+ if (!dmz_is_seq(zone))
+ zone->wp_block = 0;
+ else if (zone->wp_block != wp_block) {
+ dmz_dev_err(dmz, "Zone %u: Invalid wp (%llu / %llu)\n",
+ i, (u64)zone->wp_block, (u64)wp_block);
+ zone->wp_block = wp_block;
+ dmz_invalidate_blocks(dmz, zone, zone->wp_block,
+ dmz->zone_nr_blocks - zone->wp_block);
+ }
+
+ }
+
+ return 0;
+}
+
+/*
+ * Reset a zone write pointer.
+ */
+static int dmz_reset_zone(struct dmz_target *dmz, struct dm_zone *zone)
+{
+ int ret;
+
+ /*
+ * Ignore offline zones, read only zones,
+ * and conventional zones.
+ */
+ if (dmz_is_offline(zone) ||
+ dmz_is_readonly(zone) ||
+ dmz_is_rnd(zone))
+ return 0;
+
+ if (!dmz_is_empty(zone) || dmz_seq_write_err(zone)) {
+ ret = blkdev_reset_zones(dmz->zbd,
+ dmz_start_sect(dmz, zone),
+ dmz->zone_nr_sectors,
+ GFP_KERNEL);
+ if (ret) {
+ dmz_dev_err(dmz, "Reset zone %u failed %d\n",
+ dmz_id(dmz, zone), ret);
+ return ret;
+ }
+ }
+
+ /* Clear write error bit and rewind write pointer position */
+ clear_bit(DMZ_SEQ_WRITE_ERR, &zone->flags);
+ zone->wp_block = 0;
+
+ return 0;
+}
+
+static void dmz_get_zone_weight(struct dmz_target *dmz, struct dm_zone *zone);
+
+/*
+ * Initialize chunk mapping.
+ */
+static int dmz_load_mapping(struct dmz_target *dmz)
+{
+ struct dm_zone *dzone, *bzone;
+ struct dmz_mblock *dmap_mblk = NULL;
+ struct dmz_map *dmap;
+ unsigned int i = 0, e = 0, chunk = 0;
+ unsigned int dzone_id;
+ unsigned int bzone_id;
+
+ /* Metadata block array for the chunk mapping table */
+ dmz->dz_map_mblk = kcalloc(dmz->nr_map_blocks,
+ sizeof(struct dmz_mblk *),
+ GFP_KERNEL);
+ if (!dmz->dz_map_mblk)
+ return -ENOMEM;
+
+ /* Get chunk mapping table blocks and initialize zone mapping */
+ while (chunk < dmz->nr_chunks) {
+
+ if (!dmap_mblk) {
+ /* Get mapping block */
+ dmap_mblk = dmz_get_mblock(dmz, i + 1);
+ if (IS_ERR(dmap_mblk))
+ return PTR_ERR(dmap_mblk);
+ dmz->dz_map_mblk[i] = dmap_mblk;
+ dmap = (struct dmz_map *) dmap_mblk->data;
+ i++;
+ e = 0;
+ }
+
+ /* Check data zone */
+ dzone_id = le32_to_cpu(dmap[e].dzone_id);
+ if (dzone_id == DMZ_MAP_UNMAPPED)
+ goto next;
+
+ if (dzone_id >= dmz->nr_zones) {
+ dmz_dev_err(dmz,
+ "Chunk %u mapping: invalid data zone ID %u\n",
+ chunk, dzone_id);
+ return -EIO;
+ }
+
+ dzone = dmz_get(dmz, dzone_id);
+ set_bit(DMZ_DATA, &dzone->flags);
+ dzone->chunk = chunk;
+ dmz_get_zone_weight(dmz, dzone);
+
+ if (dmz_is_rnd(dzone))
+ list_add_tail(&dzone->link, &dmz->dz_map_rnd_list);
+ else
+ list_add_tail(&dzone->link, &dmz->dz_map_seq_list);
+
+ /* Check buffer zone */
+ bzone_id = le32_to_cpu(dmap[e].bzone_id);
+ if (bzone_id == DMZ_MAP_UNMAPPED)
+ goto next;
+
+ if (bzone_id >= dmz->nr_zones) {
+ dmz_dev_err(dmz,
+ "Chunk %u mapping: invalid buffer zone ID %u\n",
+ chunk, bzone_id);
+ return -EIO;
+ }
+
+ bzone = dmz_get(dmz, bzone_id);
+ if (!dmz_is_rnd(bzone)) {
+ dmz_dev_err(dmz,
+ "Chunk %u mapping: invalid buffer zone %u\n",
+ chunk, bzone_id);
+ return -EIO;
+ }
+
+ set_bit(DMZ_DATA, &bzone->flags);
+ set_bit(DMZ_BUF, &bzone->flags);
+ bzone->chunk = chunk;
+ bzone->bzone = dzone;
+ dzone->bzone = bzone;
+ dmz_get_zone_weight(dmz, bzone);
+ list_add_tail(&bzone->link, &dmz->dz_map_rnd_list);
+
+next:
+ chunk++;
+ e++;
+ if (e >= DMZ_MAP_ENTRIES)
+ dmap_mblk = NULL;
+
+ }
+
+ /*
+ * At this point, only meta zones and mapped data zones were
+ * fully initialized. All remaining zones are unmapped data
+ * zones. Finish initializing those here.
+ */
+ for (i = 0; i < dmz->nr_zones; i++) {
+
+ dzone = dmz_get(dmz, i);
+ if (dmz_is_meta(dzone))
+ continue;
+
+ if (dmz_is_rnd(dzone))
+ dmz->dz_nr_rnd++;
+ else
+ dmz->dz_nr_seq++;
+
+ if (dmz_is_data(dzone))
+ /* Already initialized */
+ continue;
+
+ /* Unmapped data zone */
+ set_bit(DMZ_DATA, &dzone->flags);
+ dzone->chunk = DMZ_MAP_UNMAPPED;
+ if (dmz_is_rnd(dzone)) {
+ list_add_tail(&dzone->link,
+ &dmz->dz_unmap_rnd_list);
+ atomic_inc(&dmz->dz_unmap_nr_rnd);
+ } else if (atomic_read(&dmz->nr_reclaim_seq_zones) <
+ dmz->nr_reserved_seq) {
+ list_add_tail(&dzone->link,
+ &dmz->reclaim_seq_zones_list);
+ atomic_inc(&dmz->nr_reclaim_seq_zones);
+ dmz->dz_nr_seq--;
+ } else {
+ list_add_tail(&dzone->link,
+ &dmz->dz_unmap_seq_list);
+ atomic_inc(&dmz->dz_unmap_nr_seq);
+ }
+ }
+
+ return 0;
+}
+
+/*
+ * Set a data chunk mapping.
+ */
+static void dmz_set_chunk_mapping(struct dmz_target *dmz,
+ unsigned int chunk,
+ unsigned int dzone_id,
+ unsigned int bzone_id)
+{
+ struct dmz_mblock *dmap_mblk =
+ dmz->dz_map_mblk[chunk >> DMZ_MAP_ENTRIES_SHIFT];
+ struct dmz_map *dmap = (struct dmz_map *) dmap_mblk->data;
+ int map_idx = chunk & DMZ_MAP_ENTRIES_MASK;
+
+ dmap[map_idx].dzone_id = cpu_to_le32(dzone_id);
+ dmap[map_idx].bzone_id = cpu_to_le32(bzone_id);
+ dmz_dirty_mblock(dmz, dmap_mblk);
+}
+
+/*
+ * The list of mapped zones is maintained in LRU order.
+ * This rotates a zone at the end of its map list.
+ */
+static void __dmz_lru_zone(struct dmz_target *dmz,
+ struct dm_zone *zone)
+{
+ if (list_empty(&zone->link))
+ return;
+
+ list_del_init(&zone->link);
+ if (dmz_is_seq(zone))
+ /* LRU rotate sequential zone */
+ list_add_tail(&zone->link, &dmz->dz_map_seq_list);
+ else
+ /* LRU rotate random zone */
+ list_add_tail(&zone->link, &dmz->dz_map_rnd_list);
+}
+
+/*
+ * The list of mapped random zones is maintained
+ * in LRU order. This rotates a zone at the end of the list.
+ */
+static void dmz_lru_zone(struct dmz_target *dmz,
+ struct dm_zone *zone)
+{
+ __dmz_lru_zone(dmz, zone);
+ if (zone->bzone)
+ __dmz_lru_zone(dmz, zone->bzone);
+}
+
+/*
+ * Wait for any zone to be freed.
+ */
+static void dmz_wait_for_free_zones(struct dmz_target *dmz)
+{
+ DEFINE_WAIT(wait);
+
+ dmz_trigger_reclaim(dmz);
+
+ prepare_to_wait(&dmz->dz_free_wq, &wait, TASK_UNINTERRUPTIBLE);
+ dmz_unlock_map(dmz);
+ up_read(&dmz->mblk_sem);
+
+ io_schedule_timeout(HZ);
+
+ down_read(&dmz->mblk_sem);
+ dmz_lock_map(dmz);
+ finish_wait(&dmz->dz_free_wq, &wait);
+}
+
+/*
+ * Wait for a zone reclaim to complete.
+ */
+static void dmz_wait_for_reclaim(struct dmz_target *dmz,
+ struct dm_zone *zone)
+{
+ dmz_unlock_map(dmz);
+ wait_on_bit_timeout(&zone->flags, DMZ_RECLAIM,
+ TASK_UNINTERRUPTIBLE,
+ HZ);
+ dmz_lock_map(dmz);
+}
+
+/*
+ * Activate a zone (increment its reference count).
+ */
+void dmz_activate_zone(struct dmz_target *dmz, struct dm_zone *zone)
+{
+ set_bit(DMZ_ACTIVE, &zone->flags);
+ atomic_inc(&zone->refcount);
+}
+
+/*
+ * Deactivate a zone. This decrement the zone reference counter
+ * and clears the active state of the zone once the count reaches 0,
+ * indicating that all BIOs to the zone have completed. Returns
+ * true if the zone was deactivated.
+ */
+void dmz_deactivate_zone(struct dmz_target *dmz, struct dm_zone *zone)
+{
+ if (atomic_dec_and_test(&zone->refcount)) {
+ WARN_ON(!test_bit(DMZ_ACTIVE, &zone->flags));
+ clear_bit_unlock(DMZ_ACTIVE, &zone->flags);
+ smp_mb__after_atomic();
+ }
+}
+
+/*
+ * Get the zone mapping a chunk, if the chunk is mapped already.
+ * If no mapping exist and the operation is WRITE, a zone is
+ * allocated and used to map the chunk.
+ * The zone returned will be set to the active state.
+ */
+struct dm_zone *dmz_get_chunk_mapping(struct dmz_target *dmz,
+ unsigned int chunk, int op)
+{
+ struct dmz_mblock *dmap_mblk =
+ dmz->dz_map_mblk[chunk >> DMZ_MAP_ENTRIES_SHIFT];
+ struct dmz_map *dmap = (struct dmz_map *) dmap_mblk->data;
+ int dmap_idx = chunk & DMZ_MAP_ENTRIES_MASK;
+ unsigned int dzone_id;
+ struct dm_zone *dzone = NULL;
+ int ret = 0;
+
+ dmz_lock_map(dmz);
+
+again:
+
+ /* Get the chunk mapping */
+ dzone_id = le32_to_cpu(dmap[dmap_idx].dzone_id);
+ if (dzone_id == DMZ_MAP_UNMAPPED) {
+
+ /*
+ * Read or discard in unmapped chunks are fine. But for
+ * writes, we need a mapping, so get one.
+ */
+ if (op != REQ_OP_WRITE)
+ goto out;
+
+ /* Alloate a random zone */
+ dzone = dmz_alloc_zone(dmz, DMZ_ALLOC_RND);
+ if (!dzone) {
+ dmz_wait_for_free_zones(dmz);
+ goto again;
+ }
+
+ dmz_map_zone(dmz, dzone, chunk);
+
+ } else {
+
+ /* The chunk is already mapped: get the mapping zone */
+ dzone = dmz_get(dmz, dzone_id);
+ if (dzone->chunk != chunk) {
+ dzone = ERR_PTR(-EIO);
+ goto out;
+ }
+
+ /* Repair write pointer if the sequential dzone has error */
+ if (dmz_seq_write_err(dzone)) {
+ ret = dmz_handle_seq_write_err(dmz, dzone);
+ if (ret) {
+ dzone = ERR_PTR(-EIO);
+ goto out;
+ }
+ clear_bit(DMZ_SEQ_WRITE_ERR, &dzone->flags);
+ }
+ }
+
+ /*
+ * If the zone is being reclaimed, the chunk mapping may change
+ * to a different zone. So wait for reclaim and retry. Otherwise,
+ * activate the zone (this will prevent reclaim from touching it).
+ */
+ if (dmz_in_reclaim(dzone)) {
+ dmz_wait_for_reclaim(dmz, dzone);
+ goto again;
+ }
+ dmz_activate_zone(dmz, dzone);
+ dmz_lru_zone(dmz, dzone);
+
+out:
+ dmz_unlock_map(dmz);
+
+ return dzone;
+}
+
+/*
+ * Write and discard change the block validity of data zones and their buffer
+ * zones. Check here that valid blocks are still present. If all blocks are
+ * invalid, the zones can be unmapped on the fly without waiting for reclaim
+ * to do it.
+ */
+void dmz_put_chunk_mapping(struct dmz_target *dmz, struct dm_zone *dzone)
+{
+ struct dm_zone *bzone;
+
+ dmz_lock_map(dmz);
+
+ bzone = dzone->bzone;
+ if (bzone) {
+ if (dmz_weight(bzone)) {
+ dmz_lru_zone(dmz, bzone);
+ } else {
+ /* Empty buffer zone: reclaim it */
+ dmz_unmap_zone(dmz, bzone);
+ dmz_free_zone(dmz, bzone);
+ bzone = NULL;
+ }
+ }
+
+ /* Deactivate the data zone */
+ dmz_deactivate_zone(dmz, dzone);
+ if (dmz_is_active(dzone) || bzone || dmz_weight(dzone)) {
+ dmz_lru_zone(dmz, dzone);
+ } else {
+ /* Unbuffered inactive empty data zone: reclaim it */
+ dmz_unmap_zone(dmz, dzone);
+ dmz_free_zone(dmz, dzone);
+ }
+
+ dmz_unlock_map(dmz);
+}
+
+/*
+ * Allocate and map a random zone to buffer a chunk
+ * already mapped to a sequential zone.
+ */
+struct dm_zone *dmz_get_chunk_buffer(struct dmz_target *dmz,
+ struct dm_zone *dzone)
+{
+ struct dm_zone *bzone;
+ unsigned int chunk;
+
+ dmz_lock_map(dmz);
+
+ chunk = dzone->chunk;
+
+ /* Alloate a random zone */
+ do {
+ bzone = dmz_alloc_zone(dmz, DMZ_ALLOC_RND);
+ if (!bzone)
+ dmz_wait_for_free_zones(dmz);
+ } while (!bzone);
+
+ /* Update the chunk mapping */
+ dmz_set_chunk_mapping(dmz, chunk,
+ dmz_id(dmz, dzone),
+ dmz_id(dmz, bzone));
+
+ set_bit(DMZ_BUF, &bzone->flags);
+ bzone->chunk = chunk;
+ bzone->bzone = dzone;
+ dzone->bzone = bzone;
+ list_add_tail(&bzone->link, &dmz->dz_map_rnd_list);
+
+ dmz_unlock_map(dmz);
+
+ return bzone;
+}
+
+/*
+ * Get an unmapped (free) zone.
+ * This must be called with the mapping lock held.
+ */
+struct dm_zone *dmz_alloc_zone(struct dmz_target *dmz, unsigned long flags)
+{
+ struct list_head *list;
+ struct dm_zone *zone;
+
+ if (flags & DMZ_ALLOC_RND)
+ list = &dmz->dz_unmap_rnd_list;
+ else
+ list = &dmz->dz_unmap_seq_list;
+
+again:
+ if (list_empty(list)) {
+
+ /*
+ * No free zone: if this is for reclaim, allow using the
+ * reserved sequential zones.
+ */
+ if (!(flags & DMZ_ALLOC_RECLAIM) ||
+ list_empty(&dmz->reclaim_seq_zones_list))
+ return NULL;
+
+ zone = list_first_entry(&dmz->reclaim_seq_zones_list,
+ struct dm_zone, link);
+ list_del_init(&zone->link);
+ atomic_dec(&dmz->nr_reclaim_seq_zones);
+ return zone;
+
+ }
+
+ zone = list_first_entry(list, struct dm_zone, link);
+ list_del_init(&zone->link);
+
+ if (dmz_is_rnd(zone))
+ atomic_dec(&dmz->dz_unmap_nr_rnd);
+ else
+ atomic_dec(&dmz->dz_unmap_nr_seq);
+
+ if (dmz_is_offline(zone)) {
+ dmz_dev_warn(dmz, "Zone %u is offline\n",
+ dmz_id(dmz, zone));
+ zone = NULL;
+ goto again;
+ }
+
+ if (dmz_should_reclaim(dmz))
+ dmz_trigger_reclaim(dmz);
+
+ return zone;
+}
+
+/*
+ * Free a zone.
+ * This must be called with the mapping lock held.
+ */
+void dmz_free_zone(struct dmz_target *dmz, struct dm_zone *zone)
+{
+
+
+
+ /* If this is a sequential zone, reset it */
+ if (dmz_is_seq(zone))
+ dmz_reset_zone(dmz, zone);
+
+ /* Return the zone to its type unmap list */
+ if (dmz_is_rnd(zone)) {
+ list_add_tail(&zone->link, &dmz->dz_unmap_rnd_list);
+ atomic_inc(&dmz->dz_unmap_nr_rnd);
+ } else if (atomic_read(&dmz->nr_reclaim_seq_zones) <
+ dmz->nr_reserved_seq) {
+ list_add_tail(&zone->link, &dmz->reclaim_seq_zones_list);
+ atomic_inc(&dmz->nr_reclaim_seq_zones);
+ } else {
+ list_add_tail(&zone->link, &dmz->dz_unmap_seq_list);
+ atomic_inc(&dmz->dz_unmap_nr_seq);
+ }
+
+ wake_up_all(&dmz->dz_free_wq);
+}
+
+/*
+ * Map a chunk to a zone.
+ * This must be called with the mapping lock held.
+ */
+void dmz_map_zone(struct dmz_target *dmz, struct dm_zone *dzone,
+ unsigned int chunk)
+{
+
+ /* Set the chunk mapping */
+ dmz_set_chunk_mapping(dmz, chunk,
+ dmz_id(dmz, dzone),
+ DMZ_MAP_UNMAPPED);
+ dzone->chunk = chunk;
+ if (dmz_is_rnd(dzone))
+ list_add_tail(&dzone->link, &dmz->dz_map_rnd_list);
+ else
+ list_add_tail(&dzone->link, &dmz->dz_map_seq_list);
+}
+
+/*
+ * Unmap a zone.
+ * This must be called with the mapping lock held.
+ */
+void dmz_unmap_zone(struct dmz_target *dmz, struct dm_zone *zone)
+{
+ unsigned int chunk = zone->chunk;
+ unsigned int dzone_id;
+
+ if (chunk == DMZ_MAP_UNMAPPED)
+ /* Already unmapped */
+ return;
+
+ if (test_and_clear_bit(DMZ_BUF, &zone->flags)) {
+
+ /*
+ * Unmapping the chunk buffer zone: clear only
+ * the chunk buffer mapping
+ */
+ dzone_id = dmz_id(dmz, zone->bzone);
+ zone->bzone->bzone = NULL;
+ zone->bzone = NULL;
+
+ } else {
+
+ /*
+ * Unmapping the chunk data zone: the zone must
+ * not be buffered.
+ */
+ if (WARN_ON(zone->bzone)) {
+ zone->bzone->bzone = NULL;
+ zone->bzone = NULL;
+ }
+ dzone_id = DMZ_MAP_UNMAPPED;
+
+ }
+
+ dmz_set_chunk_mapping(dmz, chunk, dzone_id,
+ DMZ_MAP_UNMAPPED);
+
+ zone->chunk = DMZ_MAP_UNMAPPED;
+ list_del_init(&zone->link);
+}
+
+/*
+ * Set @nr_bits bits in @bitmap starting from @bit.
+ * Return the number of bits changed from 0 to 1.
+ */
+static unsigned int dmz_set_bits(unsigned long *bitmap,
+ unsigned int bit, unsigned int nr_bits)
+{
+ unsigned long *addr;
+ unsigned int end = bit + nr_bits;
+ unsigned int n = 0;
+
+ while (bit < end) {
+
+ if (((bit & (BITS_PER_LONG - 1)) == 0) &&
+ ((end - bit) >= BITS_PER_LONG)) {
+ /* Try to set the whole word at once */
+ addr = bitmap + BIT_WORD(bit);
+ if (*addr == 0) {
+ *addr = ULONG_MAX;
+ n += BITS_PER_LONG;
+ bit += BITS_PER_LONG;
+ continue;
+ }
+ }
+
+ if (!test_and_set_bit(bit, bitmap))
+ n++;
+ bit++;
+
+ }
+
+ return n;
+
+}
+
+/*
+ * Get the bitmap block storing the bit for chunk_block in zone.
+ */
+static struct dmz_mblock *dmz_get_bitmap(struct dmz_target *dmz,
+ struct dm_zone *zone,
+ sector_t chunk_block)
+{
+ sector_t bitmap_block = 1 + dmz->nr_map_blocks
+ + (sector_t)(dmz_id(dmz, zone) * dmz->zone_nr_bitmap_blocks)
+ + (chunk_block >> DMZ_BLOCK_SHIFT_BITS);
+
+ return dmz_get_mblock(dmz, bitmap_block);
+}
+
+/*
+ * Copy the bitmap of from_zone to the bitmap of to_zone.
+ */
+int dmz_valid_copy(struct dmz_target *dmz, struct dm_zone *from_zone,
+ struct dm_zone *to_zone)
+{
+ struct dmz_mblock *from_mblk, *to_mblk;
+ sector_t chunk_block = 0;
+
+ /* Get the zones bitmap blocks */
+ while (chunk_block < dmz->zone_nr_blocks) {
+
+ from_mblk = dmz_get_bitmap(dmz, from_zone, chunk_block);
+ if (IS_ERR(from_mblk))
+ return PTR_ERR(from_mblk);
+ to_mblk = dmz_get_bitmap(dmz, to_zone, chunk_block);
+ if (IS_ERR(to_mblk)) {
+ dmz_release_mblock(dmz, from_mblk);
+ return PTR_ERR(to_mblk);
+ }
+
+ memcpy(to_mblk->data, from_mblk->data, DMZ_BLOCK_SIZE);
+ dmz_dirty_mblock(dmz, to_mblk);
+
+ dmz_release_mblock(dmz, to_mblk);
+ dmz_release_mblock(dmz, from_mblk);
+
+ chunk_block += DMZ_BLOCK_SIZE_BITS;
+
+ }
+
+ to_zone->weight = from_zone->weight;
+
+ return 0;
+}
+
+/*
+ * Merge the valid blocks of from_zone into the bitmap of to_zone.
+ */
+int dmz_valid_merge(struct dmz_target *dmz, struct dm_zone *from_zone,
+ struct dm_zone *to_zone, sector_t chunk_block)
+{
+ unsigned int nr_blocks;
+ int ret;
+
+ /* Get the zones bitmap blocks */
+ while (chunk_block < dmz->zone_nr_blocks) {
+
+ /* Get a valid region from the source zone */
+ ret = dmz_first_valid_block(dmz, from_zone, &chunk_block);
+ if (ret < 0)
+ return ret;
+
+ /* Are we done ? */
+ nr_blocks = ret;
+ if (!nr_blocks)
+ return 0;
+
+ ret = dmz_validate_blocks(dmz, to_zone, chunk_block, nr_blocks);
+ if (ret != 0)
+ return ret;
+
+ chunk_block += nr_blocks;
+
+ }
+
+ return 0;
+}
+
+/*
+ * Validate all the blocks in the range [block..block+nr_blocks-1].
+ */
+int dmz_validate_blocks(struct dmz_target *dmz, struct dm_zone *zone,
+ sector_t chunk_block, unsigned int nr_blocks)
+{
+ unsigned int count, bit, nr_bits;
+ struct dmz_mblock *mblk;
+ unsigned int n = 0;
+
+ dmz_dev_debug(dmz, "=> VALIDATE zone %u, block %llu, %u blocks\n",
+ dmz_id(dmz, zone), (u64)chunk_block, nr_blocks);
+
+ WARN_ON(chunk_block + nr_blocks > dmz->zone_nr_blocks);
+
+ while (nr_blocks) {
+
+ /* Get bitmap block */
+ mblk = dmz_get_bitmap(dmz, zone, chunk_block);
+ if (IS_ERR(mblk))
+ return PTR_ERR(mblk);
+
+ /* Set bits */
+ bit = chunk_block & DMZ_BLOCK_MASK_BITS;
+ nr_bits = min(nr_blocks, DMZ_BLOCK_SIZE_BITS - bit);
+
+ count = dmz_set_bits((unsigned long *) mblk->data,
+ bit, nr_bits);
+ if (count) {
+ dmz_dirty_mblock(dmz, mblk);
+ n += count;
+ }
+ dmz_release_mblock(dmz, mblk);
+
+ nr_blocks -= nr_bits;
+ chunk_block += nr_bits;
+
+ }
+
+ if (likely(zone->weight + n <= dmz->zone_nr_blocks)) {
+ zone->weight += n;
+ } else {
+ dmz_dev_warn(dmz, "Zone %u: weight %u should be <= %llu\n",
+ dmz_id(dmz, zone), zone->weight,
+ (u64)dmz->zone_nr_blocks - n);
+ zone->weight = dmz->zone_nr_blocks;
+ }
+
+ return 0;
+}
+
+/*
+ * Clear nr_bits bits in bitmap starting from bit.
+ * Return the number of bits cleared.
+ */
+static int dmz_clear_bits(unsigned long *bitmap, int bit, int nr_bits)
+{
+ unsigned long *addr;
+ int end = bit + nr_bits;
+ int n = 0;
+
+ while (bit < end) {
+
+ if (((bit & (BITS_PER_LONG - 1)) == 0) &&
+ ((end - bit) >= BITS_PER_LONG)) {
+ /* Try to clear whole word at once */
+ addr = bitmap + BIT_WORD(bit);
+ if (*addr == ULONG_MAX) {
+ *addr = 0;
+ n += BITS_PER_LONG;
+ bit += BITS_PER_LONG;
+ continue;
+ }
+ }
+
+ if (test_and_clear_bit(bit, bitmap))
+ n++;
+ bit++;
+
+ }
+
+ return n;
+
+}
+
+/*
+ * Invalidate all the blocks in the range [block..block+nr_blocks-1].
+ */
+int dmz_invalidate_blocks(struct dmz_target *dmz, struct dm_zone *zone,
+ sector_t chunk_block, unsigned int nr_blocks)
+{
+ unsigned int count, bit, nr_bits;
+ struct dmz_mblock *mblk;
+ unsigned int n = 0;
+
+ dmz_dev_debug(dmz, "=> INVALIDATE zone %u, block %llu, %u blocks\n",
+ dmz_id(dmz, zone), (u64)chunk_block, nr_blocks);
+
+ WARN_ON(chunk_block + nr_blocks > dmz->zone_nr_blocks);
+
+ while (nr_blocks) {
+
+ /* Get bitmap block */
+ mblk = dmz_get_bitmap(dmz, zone, chunk_block);
+ if (IS_ERR(mblk))
+ return PTR_ERR(mblk);
+
+ /* Clear bits */
+ bit = chunk_block & DMZ_BLOCK_MASK_BITS;
+ nr_bits = min(nr_blocks, DMZ_BLOCK_SIZE_BITS - bit);
+
+ count = dmz_clear_bits((unsigned long *) mblk->data,
+ bit, nr_bits);
+ if (count) {
+ dmz_dirty_mblock(dmz, mblk);
+ n += count;
+ }
+ dmz_release_mblock(dmz, mblk);
+
+ nr_blocks -= nr_bits;
+ chunk_block += nr_bits;
+
+ }
+
+ if (zone->weight >= n) {
+ zone->weight -= n;
+ } else {
+ dmz_dev_warn(dmz, "Zone %u: weight %u should be >= %u\n",
+ dmz_id(dmz, zone), zone->weight, n);
+ zone->weight = 0;
+ }
+
+ return 0;
+}
+
+/*
+ * Get a block bit value.
+ */
+static int dmz_test_block(struct dmz_target *dmz, struct dm_zone *zone,
+ sector_t chunk_block)
+{
+ struct dmz_mblock *mblk;
+ int ret;
+
+ WARN_ON(chunk_block >= dmz->zone_nr_blocks);
+
+ /* Get bitmap block */
+ mblk = dmz_get_bitmap(dmz, zone, chunk_block);
+ if (IS_ERR(mblk))
+ return PTR_ERR(mblk);
+
+ /* Get offset */
+ ret = test_bit(chunk_block & DMZ_BLOCK_MASK_BITS,
+ (unsigned long *) mblk->data) != 0;
+
+ dmz_release_mblock(dmz, mblk);
+
+ return ret;
+}
+
+/*
+ * Return the number of blocks from chunk_block to the first block with a bit
+ * value specified by set. Search at most nr_blocks blocks from chunk_block.
+ */
+static int dmz_to_next_set_block(struct dmz_target *dmz, struct dm_zone *zone,
+ sector_t chunk_block, unsigned int nr_blocks,
+ int set)
+{
+ struct dmz_mblock *mblk;
+ unsigned int bit, set_bit, nr_bits;
+ unsigned long *bitmap;
+ int n = 0;
+
+ WARN_ON(chunk_block + nr_blocks > dmz->zone_nr_blocks);
+
+ while (nr_blocks) {
+
+ /* Get bitmap block */
+ mblk = dmz_get_bitmap(dmz, zone, chunk_block);
+ if (IS_ERR(mblk))
+ return PTR_ERR(mblk);
+
+ /* Get offset */
+ bitmap = (unsigned long *) mblk->data;
+ bit = chunk_block & DMZ_BLOCK_MASK_BITS;
+ nr_bits = min(nr_blocks, DMZ_BLOCK_SIZE_BITS - bit);
+ if (set)
+ set_bit = find_next_bit(bitmap,
+ DMZ_BLOCK_SIZE_BITS,
+ bit);
+ else
+ set_bit = find_next_zero_bit(bitmap,
+ DMZ_BLOCK_SIZE_BITS,
+ bit);
+ dmz_release_mblock(dmz, mblk);
+
+ n += set_bit - bit;
+ if (set_bit < DMZ_BLOCK_SIZE_BITS)
+ break;
+
+ nr_blocks -= nr_bits;
+ chunk_block += nr_bits;
+
+ }
+
+ return n;
+}
+
+/*
+ * Test if chunk_block is valid. If it is, the number of consecutive
+ * valid blocks from chunk_block will be returned.
+ */
+int dmz_block_valid(struct dmz_target *dmz, struct dm_zone *zone,
+ sector_t chunk_block)
+{
+ int valid;
+
+ /* Test block */
+ valid = dmz_test_block(dmz, zone, chunk_block);
+ if (valid <= 0)
+ return valid;
+
+ /* The block is valid: get the number of valid blocks from block */
+ return dmz_to_next_set_block(dmz, zone, chunk_block,
+ dmz->zone_nr_blocks - chunk_block,
+ 0);
+}
+
+/*
+ * Find the first valid block from @chunk_block in @zone.
+ * If such a block is found, its number is returned using
+ * @chunk_block and the total number of valid blocks from @chunk_block
+ * is returned.
+ */
+int dmz_first_valid_block(struct dmz_target *dmz, struct dm_zone *zone,
+ sector_t *chunk_block)
+{
+ sector_t start_block = *chunk_block;
+ int ret;
+
+ ret = dmz_to_next_set_block(dmz, zone, start_block,
+ dmz->zone_nr_blocks - start_block, 1);
+ if (ret < 0)
+ return ret;
+
+ start_block += ret;
+ *chunk_block = start_block;
+
+ return dmz_to_next_set_block(dmz, zone, start_block,
+ dmz->zone_nr_blocks - start_block, 0);
+}
+
+/*
+ * Count the number of bits set starting from bit up to bit + nr_bits - 1.
+ */
+static int dmz_count_bits(void *bitmap, int bit, int nr_bits)
+{
+ unsigned long *addr;
+ int end = bit + nr_bits;
+ int n = 0;
+
+ while (bit < end) {
+
+ if (((bit & (BITS_PER_LONG - 1)) == 0) &&
+ ((end - bit) >= BITS_PER_LONG)) {
+ addr = (unsigned long *)bitmap + BIT_WORD(bit);
+ if (*addr == ULONG_MAX) {
+ n += BITS_PER_LONG;
+ bit += BITS_PER_LONG;
+ continue;
+ }
+ }
+
+ if (test_bit(bit, bitmap))
+ n++;
+ bit++;
+
+ }
+
+ return n;
+
+}
+
+/*
+ * Get a zone weight.
+ */
+static void dmz_get_zone_weight(struct dmz_target *dmz, struct dm_zone *zone)
+{
+ struct dmz_mblock *mblk;
+ sector_t chunk_block = 0;
+ unsigned int bit, nr_bits;
+ unsigned int nr_blocks = dmz->zone_nr_blocks;
+ void *bitmap;
+ int n = 0;
+
+ while (nr_blocks) {
+
+ /* Get bitmap block */
+ mblk = dmz_get_bitmap(dmz, zone, chunk_block);
+ if (IS_ERR(mblk)) {
+ n = 0;
+ break;
+ }
+
+ /* Count bits in this block */
+ bitmap = mblk->data;
+ bit = chunk_block & DMZ_BLOCK_MASK_BITS;
+ nr_bits = min(nr_blocks, DMZ_BLOCK_SIZE_BITS - bit);
+ n += dmz_count_bits(bitmap, bit, nr_bits);
+
+ dmz_release_mblock(dmz, mblk);
+
+ nr_blocks -= nr_bits;
+ chunk_block += nr_bits;
+
+ }
+
+ zone->weight = n;
+}
+
+/*
+ * Initialize the target metadata.
+ */
+int dmz_init_meta(struct dmz_target *dmz)
+{
+ unsigned int i, zid;
+ struct dm_zone *zone;
+ int ret;
+
+ /* Initialize zone descriptors */
+ ret = dmz_init_zones(dmz);
+ if (ret)
+ return ret;
+
+ /* Get super block */
+ ret = dmz_load_sb(dmz);
+ if (ret)
+ goto out;
+
+ /* Set metadata zones starting from sb_zone */
+ zid = dmz_id(dmz, dmz->sb_zone);
+ for (i = 0; i < dmz->nr_meta_zones << 1; i++) {
+ zone = dmz_get(dmz, zid + i);
+ if (!dmz_is_rnd(zone))
+ return -ENXIO;
+ set_bit(DMZ_META, &zone->flags);
+ }
+
+ /* Load mapping table */
+ ret = dmz_load_mapping(dmz);
+ if (ret)
+ goto out;
+
+ /*
+ * Cache size boundaries: allow at least 2 super blocks, the chunk map
+ * blocks and enough blocks to be able to cache the bitmap blocks of
+ * up to 16 zones when idle (min_nr_mblks). Otherwise, if busy, allow
+ * the cache to add 512 more metadata blocks.
+ */
+ dmz->min_nr_mblks = 2 + dmz->nr_map_blocks +
+ dmz->zone_nr_bitmap_blocks * 16;
+ dmz->max_nr_mblks = dmz->min_nr_mblks + 512;
+ dmz->mblk_shrinker.count_objects = dmz_mblock_shrinker_count;
+ dmz->mblk_shrinker.scan_objects = dmz_mblock_shrinker_scan;
+ dmz->mblk_shrinker.seeks = DEFAULT_SEEKS;
+ ret = register_shrinker(&dmz->mblk_shrinker);
+ if (ret)
+ goto out;
+
+ dmz_dev_info(dmz, "Host-%s zoned block device\n",
+ bdev_zoned_model(dmz->zbd) == BLK_ZONED_HA ?
+ "aware" : "managed");
+ dmz_dev_info(dmz, " %llu 512-byte logical sectors\n",
+ (u64)dmz->nr_zones
+ << dmz->zone_nr_sectors_shift);
+ dmz_dev_info(dmz, " %u zones of %llu 512-byte logical sectors\n",
+ dmz->nr_zones,
+ (u64)dmz->zone_nr_sectors);
+ dmz_dev_info(dmz, " %u metadata zones\n",
+ dmz->nr_meta_zones * 2);
+ dmz_dev_info(dmz, " %u data zones for %u chunks\n",
+ dmz->nr_data_zones,
+ dmz->nr_chunks);
+ dmz_dev_info(dmz, " %u random zones (%u unmapped)\n",
+ dmz->dz_nr_rnd,
+ atomic_read(&dmz->dz_unmap_nr_rnd));
+ dmz_dev_info(dmz, " %u sequential zones (%u unmapped)\n",
+ dmz->dz_nr_seq,
+ atomic_read(&dmz->dz_unmap_nr_seq));
+ dmz_dev_info(dmz, " %u reserved sequential data zones\n",
+ dmz->nr_reserved_seq);
+
+ dmz_dev_debug(dmz, "Format:\n");
+ dmz_dev_debug(dmz, "%u metadata blocks per set (%u max cache)\n",
+ dmz->nr_meta_blocks,
+ dmz->max_nr_mblks);
+ dmz_dev_debug(dmz, " %u data zone mapping blocks\n",
+ dmz->nr_map_blocks);
+ dmz_dev_debug(dmz, " %u bitmap blocks\n",
+ dmz->nr_bitmap_blocks);
+
+out:
+ if (ret)
+ dmz_cleanup_meta(dmz);
+
+ return ret;
+}
+
+/*
+ * Cleanup the target metadata resources.
+ */
+void dmz_cleanup_meta(struct dmz_target *dmz)
+{
+ struct rb_root *root = &dmz->mblk_rbtree;
+ struct dmz_mblock *mblk, *next;
+ int i;
+
+ /* Release zone mapping resources */
+ if (dmz->dz_map_mblk) {
+ for (i = 0; i < dmz->nr_map_blocks; i++)
+ dmz_release_mblock(dmz, dmz->dz_map_mblk[i]);
+ kfree(dmz->dz_map_mblk);
+ dmz->dz_map_mblk = NULL;
+ }
+
+ /* Release super blocks */
+ for (i = 0; i < 2; i++) {
+ if (dmz->sb[i].mblk) {
+ dmz_free_mblock(dmz, dmz->sb[i].mblk);
+ dmz->sb[i].mblk = NULL;
+ }
+ }
+
+ /* Free cached blocks */
+ while (!list_empty(&dmz->mblk_dirty_list)) {
+ mblk = list_first_entry(&dmz->mblk_dirty_list,
+ struct dmz_mblock, link);
+ dmz_dev_warn(dmz, "mblock %llu still in dirty list (ref %u)\n",
+ (u64)mblk->no,
+ atomic_read(&mblk->ref));
+ list_del_init(&mblk->link);
+ rb_erase(&mblk->node, &dmz->mblk_rbtree);
+ dmz_free_mblock(dmz, mblk);
+ }
+
+ while (!list_empty(&dmz->mblk_lru_list)) {
+ mblk = list_first_entry(&dmz->mblk_lru_list,
+ struct dmz_mblock, link);
+ list_del_init(&mblk->link);
+ rb_erase(&mblk->node, &dmz->mblk_rbtree);
+ dmz_free_mblock(dmz, mblk);
+ }
+
+ /* Sanity checks: the mblock rbtree should now be empty */
+ rbtree_postorder_for_each_entry_safe(mblk, next, root, node) {
+ dmz_dev_warn(dmz, "mblock %llu ref %u still in rbtree\n",
+ (u64)mblk->no,
+ atomic_read(&mblk->ref));
+ atomic_set(&mblk->ref, 0);
+ dmz_free_mblock(dmz, mblk);
+ }
+
+ /* Free the zone descriptors */
+ dmz_drop_zones(dmz);
+}
+
+/*
+ * Check metadata on resume.
+ */
+int dmz_resume_meta(struct dmz_target *dmz)
+{
+ return dmz_check_zones(dmz);
+}
+
diff --git a/drivers/md/dm-zoned-reclaim.c b/drivers/md/dm-zoned-reclaim.c
new file mode 100644
index 0000000..aa76692
--- /dev/null
+++ b/drivers/md/dm-zoned-reclaim.c
@@ -0,0 +1,535 @@
+/*
+ * Drive-managed zoned block device target
+ * Copyright (C) 2017 Western Digital Corporation or its affiliates.
+ *
+ * Written by: Damien Le Moal <damien.lemoal@wdc.com>
+ *
+ * This software is distributed under the terms of the GNU General PUBLIC
+ * License version 2, or any later version, "as is," without technical
+ * support, and WITHOUT ANY WARRANTY, without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+ */
+
+#include <linux/module.h>
+
+#include "dm-zoned.h"
+
+/*
+ * Align a sequential zone write pointer to chunk_block.
+ */
+static int dmz_reclaim_align_wp(struct dmz_target *dmz, struct dm_zone *zone,
+ sector_t chunk_block)
+{
+ sector_t wp_block = zone->wp_block;
+ unsigned int nr_blocks;
+ int ret;
+
+ if (wp_block > chunk_block)
+ return -EIO;
+
+ /*
+ * Zeroout the space between the write
+ * pointer and the requested position.
+ */
+ nr_blocks = chunk_block - zone->wp_block;
+ if (!nr_blocks)
+ return 0;
+
+ ret = blkdev_issue_zeroout(dmz->zbd,
+ dmz_start_sect(dmz, zone) + dmz_blk2sect(wp_block),
+ dmz_blk2sect(nr_blocks),
+ GFP_NOFS, false);
+ if (ret) {
+ dmz_dev_err(dmz,
+ "Align zone %u wp %llu to +%u blocks failed %d\n",
+ dmz_id(dmz, zone),
+ (unsigned long long)wp_block,
+ nr_blocks,
+ ret);
+ return ret;
+ }
+
+ zone->wp_block += nr_blocks;
+
+ return 0;
+}
+
+/*
+ * dm_kcopyd_copy notification.
+ */
+static void dmz_reclaim_copy_end(int read_err, unsigned long write_err,
+ void *context)
+{
+ struct dmz_target *dmz = context;
+
+ if (read_err || write_err)
+ dmz->reclaim_err = -EIO;
+ else
+ dmz->reclaim_err = 0;
+
+ clear_bit_unlock(DMZ_RECLAIM_COPY, &dmz->flags);
+ smp_mb__after_atomic();
+ wake_up_bit(&dmz->flags, DMZ_RECLAIM_COPY);
+}
+
+/*
+ * Copy valid blocks of src_zone into dst_zone.
+ */
+static int dmz_reclaim_copy(struct dmz_target *dmz,
+ struct dm_zone *src_zone, struct dm_zone *dst_zone)
+{
+ struct dm_io_region src, dst;
+ sector_t block = 0, end_block;
+ sector_t nr_blocks;
+ sector_t src_zone_block;
+ sector_t dst_zone_block;
+ unsigned long flags = 0;
+ int ret;
+
+ if (dmz_is_seq(src_zone))
+ end_block = src_zone->wp_block;
+ else
+ end_block = dmz->zone_nr_blocks;
+ src_zone_block = dmz_start_block(dmz, src_zone);
+ dst_zone_block = dmz_start_block(dmz, dst_zone);
+
+ if (dmz_is_seq(dst_zone))
+ set_bit(DM_KCOPYD_WRITE_SEQ, &flags);
+
+ while (block < end_block) {
+
+ /* Get a valid region from the source zone */
+ ret = dmz_first_valid_block(dmz, src_zone, &block);
+ if (ret < 0)
+ return ret;
+
+ /* Are we done ? */
+ nr_blocks = ret;
+ if (!nr_blocks)
+ return 0;
+
+ /*
+ * If we are writing in a sequential zone, we must make sure
+ * that writes are sequential. So Zero out any eventual hole
+ * between writes.
+ */
+ if (dmz_is_seq(dst_zone)) {
+ ret = dmz_reclaim_align_wp(dmz, dst_zone, block);
+ if (ret)
+ return ret;
+ }
+
+ src.bdev = dmz->zbd;
+ src.sector = dmz_blk2sect(src_zone_block + block);
+ src.count = dmz_blk2sect(nr_blocks);
+
+ dst.bdev = dmz->zbd;
+ dst.sector = dmz_blk2sect(dst_zone_block + block);
+ dst.count = src.count;
+
+ dmz_dev_debug(dmz,
+ "Reclaim: Copy %s zone %u, block %llu+%llu to %s zone %u\n",
+ dmz_is_rnd(src_zone) ? "RND" : "SEQ",
+ dmz_id(dmz, src_zone),
+ (unsigned long long)block,
+ (unsigned long long)nr_blocks,
+ dmz_is_rnd(dst_zone) ? "RND" : "SEQ",
+ dmz_id(dmz, dst_zone));
+
+ /* Copy the valid region */
+ set_bit(DMZ_RECLAIM_COPY, &dmz->flags);
+ ret = dm_kcopyd_copy(dmz->reclaim_kc, &src, 1, &dst, flags,
+ dmz_reclaim_copy_end, dmz);
+ if (ret != 0)
+ return ret;
+
+ /* Wait for copy to complete */
+ wait_on_bit_io(&dmz->flags, DMZ_RECLAIM_COPY,
+ TASK_UNINTERRUPTIBLE);
+ if (dmz->reclaim_err)
+ return dmz->reclaim_err;
+
+ if (dmz_is_seq(dst_zone))
+ dst_zone->wp_block += nr_blocks;
+
+ block += nr_blocks;
+
+ }
+
+ return 0;
+}
+
+/*
+ * Clear a zone reclaim flag.
+ */
+static inline void dmz_reclaim_put_zone(struct dmz_target *dmz,
+ struct dm_zone *zone)
+{
+ WARN_ON(dmz_is_active(zone));
+ WARN_ON(!dmz_in_reclaim(zone));
+
+ clear_bit_unlock(DMZ_RECLAIM, &zone->flags);
+ smp_mb__after_atomic();
+ wake_up_bit(&zone->flags, DMZ_RECLAIM);
+}
+
+/*
+ * Move valid blocks of dzone buffer zone into dzone (after its write pointer)
+ * and free the buffer zone.
+ */
+static int dmz_reclaim_buf(struct dmz_target *dmz, struct dm_zone *dzone)
+{
+ struct dm_zone *bzone = dzone->bzone;
+ sector_t chunk_block = dzone->wp_block;
+ int ret;
+
+ dmz_dev_debug(dmz,
+ "Chunk %u, move buf zone %u (weight %u) to data zone %u (weight %u)\n",
+ dzone->chunk, dmz_id(dmz, bzone), dmz_weight(bzone),
+ dmz_id(dmz, dzone), dmz_weight(dzone));
+
+ /* Flush data zone into the buffer zone */
+ ret = dmz_reclaim_copy(dmz, bzone, dzone);
+ if (ret < 0)
+ return ret;
+
+ down_read(&dmz->mblk_sem);
+
+ /* Validate copied blocks */
+ ret = dmz_valid_merge(dmz, bzone, dzone, chunk_block);
+ if (ret == 0) {
+ /* Free the buffer zone */
+ dmz_invalidate_zone(dmz, bzone);
+ dmz_lock_map(dmz);
+ dmz_unmap_zone(dmz, bzone);
+ dmz_reclaim_put_zone(dmz, dzone);
+ dmz_free_zone(dmz, bzone);
+ dmz_unlock_map(dmz);
+ }
+
+ up_read(&dmz->mblk_sem);
+
+ return 0;
+}
+
+/*
+ * Merge valid blocks of dzone into its buffer zone and free dzone.
+ */
+static int dmz_reclaim_seq_data(struct dmz_target *dmz, struct dm_zone *dzone)
+{
+ unsigned int chunk = dzone->chunk;
+ struct dm_zone *bzone = dzone->bzone;
+ int ret = 0;
+
+ dmz_dev_debug(dmz,
+ "Chunk %u, move data zone %u (weight %u) to buf zone %u (weight %u)\n",
+ chunk, dmz_id(dmz, dzone), dmz_weight(dzone),
+ dmz_id(dmz, bzone), dmz_weight(bzone));
+
+ /* Flush data zone into the buffer zone */
+ ret = dmz_reclaim_copy(dmz, dzone, bzone);
+ if (ret < 0)
+ return ret;
+
+ down_read(&dmz->mblk_sem);
+
+ /* Validate copied blocks */
+ ret = dmz_valid_merge(dmz, dzone, bzone, 0);
+ if (ret == 0) {
+ /*
+ * Free the data zone and remap the chunk to
+ * the buffer zone.
+ */
+ dmz_invalidate_zone(dmz, dzone);
+ dmz_lock_map(dmz);
+ dmz_unmap_zone(dmz, bzone);
+ dmz_unmap_zone(dmz, dzone);
+ dmz_reclaim_put_zone(dmz, dzone);
+ dmz_free_zone(dmz, dzone);
+ dmz_map_zone(dmz, bzone, chunk);
+ dmz_unlock_map(dmz);
+ }
+
+ up_read(&dmz->mblk_sem);
+
+ return 0;
+}
+
+/*
+ * Move valid blocks of the random data zone dzone into a free sequential zone.
+ * Once blocks are moved, remap the zone chunk to the sequential zone.
+ */
+static int dmz_reclaim_rnd_data(struct dmz_target *dmz, struct dm_zone *dzone)
+{
+ unsigned int chunk = dzone->chunk;
+ struct dm_zone *szone = NULL;
+ int ret;
+
+ /* Get a free sequential zone */
+ dmz_lock_map(dmz);
+ szone = dmz_alloc_zone(dmz, DMZ_ALLOC_RECLAIM);
+ dmz_unlock_map(dmz);
+ if (!szone)
+ return -ENOSPC;
+
+ dmz_dev_debug(dmz,
+ "Chunk %u, move rnd zone %u (weight %u) to seq zone %u\n",
+ chunk, dmz_id(dmz, dzone), dmz_weight(dzone),
+ dmz_id(dmz, szone));
+
+ /* Flush the random data zone into the sequential zone */
+ ret = dmz_reclaim_copy(dmz, dzone, szone);
+
+ down_read(&dmz->mblk_sem);
+
+ if (ret == 0)
+ /* Validate copied blocks */
+ ret = dmz_valid_copy(dmz, dzone, szone);
+
+ if (ret) {
+ /* Free the sequential zone */
+ dmz_lock_map(dmz);
+ dmz_free_zone(dmz, szone);
+ dmz_unlock_map(dmz);
+ } else {
+ /* Free the data zone and remap the chunk */
+ dmz_invalidate_zone(dmz, dzone);
+ dmz_lock_map(dmz);
+ dmz_unmap_zone(dmz, dzone);
+ dmz_reclaim_put_zone(dmz, dzone);
+ dmz_free_zone(dmz, dzone);
+ dmz_map_zone(dmz, szone, chunk);
+ dmz_unlock_map(dmz);
+ }
+
+ up_read(&dmz->mblk_sem);
+
+ return 0;
+}
+
+/*
+ * Reclaim an empty zone.
+ */
+static void dmz_reclaim_empty(struct dmz_target *dmz, struct dm_zone *dzone)
+{
+ down_read(&dmz->mblk_sem);
+ dmz_lock_map(dmz);
+ dmz_unmap_zone(dmz, dzone);
+ dmz_reclaim_put_zone(dmz, dzone);
+ dmz_free_zone(dmz, dzone);
+ dmz_unlock_map(dmz);
+ up_read(&dmz->mblk_sem);
+}
+
+/*
+ * Lock a zone for reclaim. Returns 0 if the zone cannot be locked or if it is
+ * already locked and 1 otherwise.
+ */
+static inline int dmz_reclaim_lock_zone(struct dmz_target *dmz,
+ struct dm_zone *zone)
+{
+ /* Active zones cannot be reclaimed */
+ if (dmz_is_active(zone))
+ return 0;
+
+ return !test_and_set_bit(DMZ_RECLAIM, &zone->flags);
+}
+
+/*
+ * Select a random zone for reclaim.
+ */
+static struct dm_zone *dmz_reclaim_get_rnd_zone(struct dmz_target *dmz)
+{
+ struct dm_zone *dzone = NULL;
+ struct dm_zone *zone;
+
+ if (list_empty(&dmz->dz_map_rnd_list))
+ return NULL;
+
+ list_for_each_entry(zone, &dmz->dz_map_rnd_list, link) {
+ if (dmz_is_buf(zone))
+ dzone = zone->bzone;
+ else
+ dzone = zone;
+ if (dmz_reclaim_lock_zone(dmz, dzone))
+ return dzone;
+ }
+
+ return NULL;
+}
+
+/*
+ * Select a buffered sequential zone for reclaim.
+ */
+static struct dm_zone *dmz_reclaim_get_seq_zone(struct dmz_target *dmz)
+{
+ struct dm_zone *zone;
+
+ if (list_empty(&dmz->dz_map_seq_list))
+ return NULL;
+
+ list_for_each_entry(zone, &dmz->dz_map_seq_list, link) {
+ if (!zone->bzone)
+ continue;
+ if (dmz_reclaim_lock_zone(dmz, zone))
+ return zone;
+ }
+
+ return NULL;
+}
+
+/*
+ * Select a zone for reclaim.
+ */
+static struct dm_zone *dmz_reclaim_get_zone(struct dmz_target *dmz)
+{
+ struct dm_zone *zone = NULL;
+
+ /*
+ * Search for a zone candidate to reclaim: 2 cases are possible.
+ * (1) There is no free sequential zones. Then a random data zone
+ * cannot be reclaimed. So choose a sequential zone to reclaim so
+ * that afterward a random zone can be reclaimed.
+ * (2) At least one free sequential zone is available, then choose
+ * the oldest random zone (data or buffer) that can be locked.
+ */
+ dmz_lock_map(dmz);
+ if (list_empty(&dmz->reclaim_seq_zones_list))
+ zone = dmz_reclaim_get_seq_zone(dmz);
+ else
+ zone = dmz_reclaim_get_rnd_zone(dmz);
+ dmz_unlock_map(dmz);
+
+ return zone;
+}
+
+/*
+ * Find a reclaim candidate zone and reclaim it.
+ */
+static void dmz_reclaim(struct dmz_target *dmz)
+{
+ struct dm_zone *dzone;
+ struct dm_zone *rzone;
+ unsigned long start;
+ int ret;
+
+ /* Get a data zone */
+ dzone = dmz_reclaim_get_zone(dmz);
+ if (!dzone)
+ return;
+
+ start = jiffies;
+
+ if (dmz_is_rnd(dzone)) {
+
+ rzone = dzone;
+ if (!dmz_weight(dzone)) {
+ /* Empty zone */
+ dmz_reclaim_empty(dmz, dzone);
+ ret = 0;
+ } else {
+ /*
+ * Reclaim the random data zone by moving its
+ * valid data blocks to a free sequential zone.
+ */
+ ret = dmz_reclaim_rnd_data(dmz, dzone);
+ }
+
+ } else {
+
+ struct dm_zone *bzone = dzone->bzone;
+ sector_t chunk_block = 0;
+
+ ret = dmz_first_valid_block(dmz, bzone, &chunk_block);
+ if (ret < 0)
+ goto out;
+
+ if (chunk_block >= dzone->wp_block) {
+ /*
+ * Valid blocks in the buffer zone are after
+ * the data zone write pointer: copy them there.
+ */
+ ret = dmz_reclaim_buf(dmz, dzone);
+ rzone = bzone;
+ } else {
+ /*
+ * Reclaim the data zone by merging it into the
+ * buffer zone so that the buffer zone itself can
+ * be later reclaimed.
+ */
+ ret = dmz_reclaim_seq_data(dmz, dzone);
+ rzone = dzone;
+ }
+
+ }
+
+out:
+ if (ret) {
+ dmz_reclaim_put_zone(dmz, dzone);
+ return;
+ }
+
+ dmz_dev_debug(dmz, "Reclaimed zone %u in %u ms\n",
+ dmz_id(dmz, rzone), jiffies_to_msecs(jiffies - start));
+
+ dmz_trigger_flush(dmz);
+}
+
+/*
+ * Zone reclaim work.
+ */
+void dmz_reclaim_work(struct work_struct *work)
+{
+ struct dmz_target *dmz =
+ container_of(work, struct dmz_target, reclaim_work.work);
+ unsigned long next_reclaim = DMZ_RECLAIM_PERIOD;
+ unsigned int unmap_nr_rnd = atomic_read(&dmz->dz_unmap_nr_rnd);
+ unsigned int throttle, unmap_perc;
+
+ /* If there are still plenty of random zones, do not reclaim */
+ unmap_perc = unmap_nr_rnd * 100 / dmz->dz_nr_rnd;
+ if (unmap_perc >= DMZ_RECLAIM_HIGH_FREE_RND)
+ goto out;
+
+ /*
+ * If we are not idle and still have unmapped random zones,
+ * do not reclaim.
+ */
+ if (!dmz_idle(dmz) && unmap_perc > DMZ_RECLAIM_LOW_FREE_RND)
+ goto out;
+
+ /*
+ * We need to start reclaiming random zones: set up zone copy
+ * throttling to either go fast if we are very low on random zones
+ * and slower if there are still some free random zones to avoid
+ * as much as possible to negatively impact the user workload.
+ */
+ if (dmz_idle(dmz) ||
+ unmap_nr_rnd < atomic_read(&dmz->nr_active_chunks))
+ /* Idle or very low: go fast */
+ throttle = 100;
+ else
+ /* Busy but we still have some random zone: go slower */
+ throttle = min(75U, 100U - unmap_perc / 2);
+ dmz->reclaim_throttle.throttle = throttle;
+
+ dmz_dev_debug(dmz,
+ "Reclaim (%u): %s (%u BIOs, %u active chunks), %u%% free rnd zones (%u/%u)\n",
+ dmz->reclaim_throttle.throttle,
+ (dmz_idle(dmz) ? "Idle" : "Busy"),
+ atomic_read(&dmz->bio_count),
+ atomic_read(&dmz->nr_active_chunks),
+ unmap_perc,
+ unmap_nr_rnd, dmz->dz_nr_rnd);
+
+ dmz_reclaim(dmz);
+
+ if ((dmz_should_reclaim(dmz)
+ && atomic_read(&dmz->nr_reclaim_seq_zones)))
+ /* Run again immmediately */
+ next_reclaim = 0;
+
+out:
+ dmz_schedule_reclaim(dmz, next_reclaim);
+}
+
diff --git a/drivers/md/dm-zoned.h b/drivers/md/dm-zoned.h
new file mode 100644
index 0000000..bdcd538
--- /dev/null
+++ b/drivers/md/dm-zoned.h
@@ -0,0 +1,530 @@
+/*
+ * Drive-managed zoned block device target
+ * Copyright (C) 2017 Western Digital Corporation or its affiliates.
+ *
+ * This software is distributed under the terms of the GNU General Public
+ * License version 2, or any later version, "as is," without technical
+ * support, and WITHOUT ANY WARRANTY, without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+ */
+#include <linux/types.h>
+#include <linux/blkdev.h>
+#include <linux/device-mapper.h>
+#include <linux/dm-kcopyd.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+#include <linux/mutex.h>
+#include <linux/workqueue.h>
+#include <linux/rwsem.h>
+#include <linux/rbtree.h>
+#include <linux/radix-tree.h>
+#include <linux/shrinker.h>
+
+#ifndef __DM_ZONED_H__
+#define __DM_ZONED_H__
+
+/*
+ * Metadata version.
+ */
+#define DMZ_META_VER 1
+
+/*
+ * On-disk super block magic.
+ */
+#define DMZ_MAGIC ((((unsigned int)('D')) << 24) | \
+ (((unsigned int)('Z')) << 16) | \
+ (((unsigned int)('B')) << 8) | \
+ ((unsigned int)('D')))
+
+/*
+ * On disk super block.
+ * This uses only 512 B but uses on disk a full 4KB block. This block is
+ * followed on disk by the mapping table of chunks to zones and the bitmap
+ * blocks indicating zone block validity.
+ * The overall resulting metadata format is:
+ * (1) Super block (1 block)
+ * (2) Chunk mapping table (nr_map_blocks)
+ * (3) Bitmap blocks (nr_bitmap_blocks)
+ * All metadata blocks are stored in conventional zones, starting from the
+ * the first conventional zone found on disk.
+ */
+struct dmz_super {
+
+ /* Magic number */
+ __le32 magic; /* 4 */
+
+ /* Metadata version number */
+ __le32 version; /* 8 */
+
+ /* Generation number */
+ __le64 gen; /* 16 */
+
+ /* This block number */
+ __le64 sb_block; /* 24 */
+
+ /* The number of metadata blocks, including this super block */
+ __le32 nr_meta_blocks; /* 28 */
+
+ /* The number of sequential zones reserved for reclaim */
+ __le32 nr_reserved_seq; /* 32 */
+
+ /* The number of entries in the mapping table */
+ __le32 nr_chunks; /* 36 */
+
+ /* The number of blocks used for the chunk mapping table */
+ __le32 nr_map_blocks; /* 40 */
+
+ /* The number of blocks used for the block bitmaps */
+ __le32 nr_bitmap_blocks; /* 44 */
+
+ /* Checksum */
+ __le32 crc; /* 48 */
+
+ /* Padding to full 512B sector */
+ u8 reserved[464]; /* 512 */
+
+};
+
+/*
+ * Chunk mapping entry: entries are indexed by chunk number
+ * and give the zone ID (dzone_id) mapping the chunk on disk.
+ * This zone may be sequential or random. If it is a sequential
+ * zone, a second zone (bzone_id) used as a write buffer may
+ * also be specified. This second zone will always be a randomly
+ * writeable zone.
+ */
+struct dmz_map {
+ __le32 dzone_id;
+ __le32 bzone_id;
+};
+
+/*
+ * dm-zoned creates block devices with 4KB blocks, always.
+ */
+#define DMZ_BLOCK_SHIFT 12
+#define DMZ_BLOCK_SIZE (1 << DMZ_BLOCK_SHIFT)
+#define DMZ_BLOCK_MASK (DMZ_BLOCK_SIZE - 1)
+
+#define DMZ_BLOCK_SHIFT_BITS (DMZ_BLOCK_SHIFT + 3)
+#define DMZ_BLOCK_SIZE_BITS (1 << DMZ_BLOCK_SHIFT_BITS)
+#define DMZ_BLOCK_MASK_BITS (DMZ_BLOCK_SIZE_BITS - 1)
+
+#define DMZ_BLOCK_SECTORS_SHIFT (DMZ_BLOCK_SHIFT - SECTOR_SHIFT)
+#define DMZ_BLOCK_SECTORS (DMZ_BLOCK_SIZE >> SECTOR_SHIFT)
+#define DMZ_BLOCK_SECTORS_MASK (DMZ_BLOCK_SECTORS - 1)
+
+/*
+ * Chunk mapping table metadata: 512 8-bytes entries per 4KB block.
+ */
+#define DMZ_MAP_ENTRIES (DMZ_BLOCK_SIZE / sizeof(struct dmz_map))
+#define DMZ_MAP_ENTRIES_SHIFT (ilog2(DMZ_MAP_ENTRIES))
+#define DMZ_MAP_ENTRIES_MASK (DMZ_MAP_ENTRIES - 1)
+#define DMZ_MAP_UNMAPPED UINT_MAX
+
+/*
+ * Block <-> 512B sector conversion.
+ */
+#define dmz_blk2sect(b) ((b) << DMZ_BLOCK_SECTORS_SHIFT)
+#define dmz_sect2blk(s) ((s) >> DMZ_BLOCK_SECTORS_SHIFT)
+
+#define DMZ_MIN_BIOS 8192
+
+/*
+ * The size of a zone report in number of zones.
+ * This results in 4096*64B=256KB report zones commands.
+ */
+#define DMZ_REPORT_NR_ZONES 4096
+
+/*
+ * Zone flags.
+ */
+enum {
+
+ /* Zone write type */
+ DMZ_RND,
+ DMZ_SEQ,
+
+ /* Zone critical condition */
+ DMZ_OFFLINE,
+ DMZ_READ_ONLY,
+
+ /* How the zone is being used */
+ DMZ_META,
+ DMZ_DATA,
+ DMZ_BUF,
+
+ /* Zone internal state */
+ DMZ_ACTIVE,
+ DMZ_RECLAIM,
+ DMZ_SEQ_WRITE_ERR,
+
+};
+
+/*
+ * Zone descriptor.
+ */
+struct dm_zone {
+
+ /* For listing the zone depending on its state */
+ struct list_head link;
+
+ /* Zone type and state */
+ unsigned long flags;
+
+ /* Zone activation reference count */
+ atomic_t refcount;
+
+ /* Zone write pointer block (relative to the zone start block) */
+ unsigned int wp_block;
+
+ /* Zone weight (number of valid blocks in the zone) */
+ unsigned int weight;
+
+ /* The chunk that the zone maps */
+ unsigned int chunk;
+
+ /*
+ * For a sequential data zone, pointer to the random zone
+ * used as a buffer for processing unaligned writes.
+ * For a buffer zone, this points back to the data zone.
+ */
+ struct dm_zone *bzone;
+
+};
+
+/*
+ * Meta data block descriptor (for cached metadata blocks).
+ */
+struct dmz_mblock {
+
+ struct rb_node node;
+ struct list_head link;
+ sector_t no;
+ atomic_t ref;
+ unsigned long state;
+ struct page *page;
+ void *data;
+
+};
+
+/*
+ * Super block information (one per metadata set).
+ */
+struct dmz_sb {
+ sector_t block;
+ struct dmz_mblock *mblk;
+ struct dmz_super *sb;
+};
+
+/*
+ * Metadata block state flags.
+ */
+enum {
+ DMZ_META_DIRTY,
+ DMZ_META_READING,
+ DMZ_META_WRITING,
+ DMZ_META_ERROR,
+};
+
+/*
+ * Target flags.
+ */
+enum {
+ DMZ_RECLAIM_COPY,
+ DMZ_SUSPENDED,
+};
+
+/*
+ * Target descriptor.
+ */
+struct dmz_target {
+
+ struct dm_dev *ddev;
+
+ /* Zoned block device information */
+ char zbd_name[BDEVNAME_SIZE];
+ struct block_device *zbd;
+ sector_t zbd_capacity;
+ struct request_queue *zbdq;
+ unsigned long flags;
+
+ unsigned int nr_zones;
+ unsigned int nr_useable_zones;
+ unsigned int nr_meta_blocks;
+ unsigned int nr_meta_zones;
+ unsigned int nr_data_zones;
+ unsigned int nr_rnd_zones;
+ unsigned int nr_reserved_seq;
+ unsigned int nr_chunks;
+
+ sector_t zone_nr_sectors;
+ unsigned int zone_nr_sectors_shift;
+
+ sector_t zone_nr_blocks;
+ sector_t zone_nr_blocks_shift;
+
+ sector_t zone_bitmap_size;
+ unsigned int zone_nr_bitmap_blocks;
+
+ unsigned int nr_bitmap_blocks;
+ unsigned int nr_map_blocks;
+
+ /* Zone information array */
+ struct dm_zone *zones;
+
+ /* For metadata handling */
+ struct dm_zone *sb_zone;
+ struct dmz_sb sb[2];
+ unsigned int mblk_primary;
+ u64 sb_gen;
+ unsigned int min_nr_mblks;
+ unsigned int max_nr_mblks;
+ atomic_t nr_mblks;
+ struct rw_semaphore mblk_sem;
+ spinlock_t mblk_lock;
+ struct rb_root mblk_rbtree;
+ struct list_head mblk_lru_list;
+ struct list_head mblk_dirty_list;
+ struct shrinker mblk_shrinker;
+
+ /* Zone allocation management */
+ struct mutex map_lock;
+ struct dmz_mblock **dz_map_mblk;
+ unsigned int dz_nr_rnd;
+ atomic_t dz_unmap_nr_rnd;
+ struct list_head dz_unmap_rnd_list;
+ struct list_head dz_map_rnd_list;
+
+ unsigned int dz_nr_seq;
+ atomic_t dz_unmap_nr_seq;
+ struct list_head dz_unmap_seq_list;
+ struct list_head dz_map_seq_list;
+
+ wait_queue_head_t dz_free_wq;
+
+ /* For chunk work */
+ struct mutex chunk_lock;
+ struct radix_tree_root chunk_rxtree;
+ struct workqueue_struct *chunk_wq;
+ atomic_t nr_active_chunks;
+
+ /* For chunk BIOs to zones */
+ struct bio_set *bio_set;
+ atomic_t bio_count;
+ unsigned long atime;
+
+ /* For flush */
+ spinlock_t flush_lock;
+ struct bio_list flush_list;
+ struct delayed_work flush_work;
+ struct workqueue_struct *flush_wq;
+
+ /* For reclaim */
+ struct delayed_work reclaim_work;
+ struct workqueue_struct *reclaim_wq;
+ atomic_t nr_reclaim_seq_zones;
+ struct list_head reclaim_seq_zones_list;
+ struct dm_kcopyd_client *reclaim_kc;
+ struct dm_kcopyd_throttle reclaim_throttle;
+ int reclaim_err;
+
+};
+
+/*
+ * Chunk work descriptor.
+ */
+struct dm_chunk_work {
+ struct work_struct work;
+ atomic_t refcount;
+ struct dmz_target *target;
+ unsigned int chunk;
+ struct bio_list bio_list;
+};
+
+#define dmz_id(dmz, z) ((unsigned int)((z) - (dmz)->zones))
+#define dmz_get(dmz, z) (&(dmz)->zones[z])
+#define dmz_start_sect(dmz, z) (dmz_id(dmz, z) << (dmz)->zone_nr_sectors_shift)
+#define dmz_start_block(dmz, z) (dmz_id(dmz, z) << (dmz)->zone_nr_blocks_shift)
+#define dmz_is_rnd(z) test_bit(DMZ_RND, &(z)->flags)
+#define dmz_is_seq(z) test_bit(DMZ_SEQ, &(z)->flags)
+#define dmz_is_empty(z) ((z)->wp_block == 0)
+#define dmz_is_offline(z) test_bit(DMZ_OFFLINE, &(z)->flags)
+#define dmz_is_readonly(z) test_bit(DMZ_READ_ONLY, &(z)->flags)
+#define dmz_is_active(z) test_bit(DMZ_ACTIVE, &(z)->flags)
+#define dmz_in_reclaim(z) test_bit(DMZ_RECLAIM, &(z)->flags)
+#define dmz_seq_write_err(z) test_bit(DMZ_SEQ_WRITE_ERR, &(z)->flags)
+
+#define dmz_is_meta(z) test_bit(DMZ_META, &(z)->flags)
+#define dmz_is_buf(z) test_bit(DMZ_BUF, &(z)->flags)
+#define dmz_is_data(z) test_bit(DMZ_DATA, &(z)->flags)
+
+#define dmz_weight(z) ((z)->weight)
+
+#define dmz_chunk_sector(dmz, s) ((s) & ((dmz)->zone_nr_sectors - 1))
+#define dmz_chunk_block(dmz, b) ((b) & ((dmz)->zone_nr_blocks - 1))
+
+#define dmz_bio_block(bio) dmz_sect2blk((bio)->bi_iter.bi_sector)
+#define dmz_bio_blocks(bio) dmz_sect2blk(bio_sectors(bio))
+#define dmz_bio_chunk(dmz, bio) ((bio)->bi_iter.bi_sector >> \
+ (dmz)->zone_nr_sectors_shift)
+
+#define dmz_lock_map(dmz) mutex_lock(&(dmz)->map_lock)
+#define dmz_unlock_map(dmz) mutex_unlock(&(dmz)->map_lock)
+
+/*
+ * Flush intervals (seconds).
+ */
+#define DMZ_FLUSH_PERIOD (10 * HZ)
+
+/*
+ * Trigger flush.
+ */
+static inline void dmz_trigger_flush(struct dmz_target *dmz)
+{
+ mod_delayed_work(dmz->flush_wq, &dmz->flush_work, 0);
+}
+
+/*
+ * Number of seconds without BIO to consider the target device idle.
+ */
+#define DMZ_IDLE_PERIOD (10UL * HZ)
+
+/*
+ * Zone reclaim check period.
+ */
+#define DMZ_RECLAIM_PERIOD (HZ)
+
+/*
+ * Percentage of unmapped (free) random zones below which reclaim starts
+ * even if the device is not idle.
+ */
+#define DMZ_RECLAIM_LOW_FREE_RND 50
+
+/*
+ * Percentage of unmapped (free) random zones above which reclaim stops
+ * * even if the device is idle.
+ */
+#define DMZ_RECLAIM_HIGH_FREE_RND 75
+
+/*
+ * Test if the target device is idle.
+ */
+static inline int dmz_idle(struct dmz_target *dmz)
+{
+ return atomic_read(&(dmz)->bio_count) == 0 &&
+ time_is_before_jiffies(dmz->atime + DMZ_IDLE_PERIOD);
+}
+
+/*
+ * Test if triggerring reclaim is necessary.
+ */
+static inline bool dmz_should_reclaim(struct dmz_target *dmz)
+{
+ unsigned int unmap_rnd = atomic_read(&dmz->dz_unmap_nr_rnd);
+
+ if (dmz_idle(dmz) && unmap_rnd < dmz->dz_nr_rnd)
+ return true;
+
+ /* Percentage of unmappped random zones low ? */
+ return ((unmap_rnd * 100) / dmz->dz_nr_rnd) <= DMZ_RECLAIM_LOW_FREE_RND;
+}
+
+/*
+ * Schedule reclaim (delay in jiffies).
+ */
+static inline void dmz_schedule_reclaim(struct dmz_target *dmz,
+ unsigned long delay)
+{
+ mod_delayed_work(dmz->reclaim_wq, &dmz->reclaim_work, delay);
+}
+
+/*
+ * Trigger reclaim.
+ */
+static inline void dmz_trigger_reclaim(struct dmz_target *dmz)
+{
+ dmz_schedule_reclaim(dmz, 0);
+}
+
+extern void dmz_reclaim_work(struct work_struct *work);
+
+/*
+ * Zone BIO context.
+ */
+struct dmz_bioctx {
+ struct dmz_target *target;
+ struct dm_zone *zone;
+ struct bio *bio;
+ atomic_t ref;
+ int error;
+};
+
+#define dmz_info(format, args...) \
+ pr_info("dm-zoned: " format, \
+ ## args)
+
+#define dmz_dev_info(target, format, args...) \
+ pr_info("dm-zoned (%s): " format, \
+ (dmz)->zbd_name, ## args)
+
+#define dmz_dev_err(dmz, format, args...) \
+ pr_err("dm-zoned (%s): " format, \
+ (dmz)->zbd_name, ## args)
+
+#define dmz_dev_warn(dmz, format, args...) \
+ pr_warn("dm-zoned (%s): " format, \
+ (dmz)->zbd_name, ## args)
+
+#define dmz_dev_debug(dmz, format, args...) \
+ pr_debug("dm-zoned (%s): " format, \
+ (dmz)->zbd_name, ## args)
+
+extern int dmz_init_meta(struct dmz_target *dmz);
+extern int dmz_resume_meta(struct dmz_target *dmz);
+extern void dmz_cleanup_meta(struct dmz_target *dmz);
+
+extern int dmz_flush_mblocks(struct dmz_target *dmz);
+
+#define DMZ_ALLOC_RND 0x01
+#define DMZ_ALLOC_RECLAIM 0x02
+
+struct dm_zone *dmz_alloc_zone(struct dmz_target *dmz, unsigned long flags);
+extern void dmz_free_zone(struct dmz_target *dmz, struct dm_zone *zone);
+
+extern void dmz_map_zone(struct dmz_target *dmz, struct dm_zone *zone,
+ unsigned int chunk);
+extern void dmz_unmap_zone(struct dmz_target *dmz, struct dm_zone *zone);
+
+extern void dmz_activate_zone(struct dmz_target *dmz, struct dm_zone *zone);
+extern void dmz_deactivate_zone(struct dmz_target *dmz, struct dm_zone *zone);
+
+extern struct dm_zone *dmz_get_chunk_mapping(struct dmz_target *dmz,
+ unsigned int chunk, int op);
+extern void dmz_put_chunk_mapping(struct dmz_target *dmz,
+ struct dm_zone *zone);
+
+extern struct dm_zone *dmz_get_chunk_buffer(struct dmz_target *dmz,
+ struct dm_zone *dzone);
+
+extern int dmz_valid_copy(struct dmz_target *dmz, struct dm_zone *from_zone,
+ struct dm_zone *to_zone);
+extern int dmz_valid_merge(struct dmz_target *dmz, struct dm_zone *from_zone,
+ struct dm_zone *to_zone, sector_t chunk_block);
+
+extern int dmz_validate_blocks(struct dmz_target *dmz, struct dm_zone *zone,
+ sector_t chunk_block, unsigned int nr_blocks);
+extern int dmz_invalidate_blocks(struct dmz_target *dmz, struct dm_zone *zone,
+ sector_t chunk_block, unsigned int nr_blocks);
+static inline int dmz_invalidate_zone(struct dmz_target *dmz,
+ struct dm_zone *zone)
+{
+ return dmz_invalidate_blocks(dmz, zone, 0, dmz->zone_nr_blocks);
+}
+
+extern int dmz_block_valid(struct dmz_target *dmz, struct dm_zone *zone,
+ sector_t chunk_block);
+
+extern int dmz_first_valid_block(struct dmz_target *dmz, struct dm_zone *zone,
+ sector_t *chunk_block);
+
+#endif /* __DM_ZONED_H__ */
--
2.9.3
^ permalink raw reply related [flat|nested] 21+ messages in thread