From 0fbb4d93b38bce1f8235aacfa37e90ad8f011473 Mon Sep 17 00:00:00 2001 From: Mike Snitzer Date: Thu, 17 Feb 2022 23:40:32 -0500 Subject: dm: add dm_submit_bio_remap interface Where possible, switch from early bio-based IO accounting (at the time DM clones each incoming bio) to late IO accounting just before each remapped bio is issued to underlying device via submit_bio_noacct(). Allows more precise bio-based IO accounting for DM targets that use their own workqueues to perform additional processing of each bio in conjunction with their DM_MAPIO_SUBMITTED return from their map function. When a target is updated to use dm_submit_bio_remap() they must also set ti->accounts_remapped_io to true. Use xchg() in start_io_acct(), as suggested by Mikulas, to ensure each IO is only started once. The xchg race only happens if __send_duplicate_bios() sends multiple bios -- that case is reflected via tio->is_duplicate_bio. Given the niche nature of this race, it is best to avoid any xchg performance penalty for normal IO. For IO that was never submitted with dm_bio_submit_remap(), but the target completes the clone with bio_endio, accounting is started then ended and pending_io counter decremented. Reviewed-by: Mikulas Patocka Signed-off-by: Mike Snitzer --- include/linux/device-mapper.h | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'include/linux') diff --git a/include/linux/device-mapper.h b/include/linux/device-mapper.h index b26fecf6c8e8..7752d14d13f8 100644 --- a/include/linux/device-mapper.h +++ b/include/linux/device-mapper.h @@ -362,6 +362,12 @@ struct dm_target { * zone append operations using regular writes. */ bool emulate_zone_append:1; + + /* + * Set if the target will submit IO using dm_submit_bio_remap() + * after returning DM_MAPIO_SUBMITTED from its map function. + */ + bool accounts_remapped_io:1; }; void *dm_per_bio_data(struct bio *bio, size_t data_size); @@ -465,6 +471,7 @@ int dm_suspended(struct dm_target *ti); int dm_post_suspending(struct dm_target *ti); int dm_noflush_suspending(struct dm_target *ti); void dm_accept_partial_bio(struct bio *bio, unsigned n_sectors); +void dm_submit_bio_remap(struct bio *clone, struct bio *tgt_clone, bool from_wq); union map_info *dm_get_rq_mapinfo(struct request *rq); #ifdef CONFIG_BLK_DEV_ZONED -- cgit From a8b9d116cda047f38ba29ce26df49c57479ffaa2 Mon Sep 17 00:00:00 2001 From: Tom Rix Date: Wed, 26 Jan 2022 05:18:26 -0800 Subject: dm: cleanup double word in comment Remove the second 'a'. Signed-off-by: Tom Rix Signed-off-by: Mike Snitzer --- include/linux/device-mapper.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/device-mapper.h b/include/linux/device-mapper.h index 7752d14d13f8..70cd9449275a 100644 --- a/include/linux/device-mapper.h +++ b/include/linux/device-mapper.h @@ -358,7 +358,7 @@ struct dm_target { bool limit_swap_bios:1; /* - * Set if this target implements a a zoned device and needs emulation of + * Set if this target implements a zoned device and needs emulation of * zone append operations using regular writes. */ bool emulate_zone_append:1; -- cgit From 69fe0f29892077f14b56e2a479b6bcf533209d53 Mon Sep 17 00:00:00 2001 From: Ming Lei Date: Fri, 4 Mar 2022 21:08:03 -0500 Subject: block: add ->poll_bio to block_device_operations Prepare for supporting IO polling for bio-based driver. Add ->poll_bio callback so that bio-based driver can provide their own logic for polling bio. Also fix ->submit_bio_bio typo in comment block above __submit_bio_noacct. Reviewed-by: Christoph Hellwig Reviewed-by: Jens Axboe Signed-off-by: Ming Lei Signed-off-by: Mike Snitzer --- block/blk-core.c | 14 +++++++++----- block/genhd.c | 4 ++++ include/linux/blkdev.h | 2 ++ 3 files changed, 15 insertions(+), 5 deletions(-) (limited to 'include/linux') diff --git a/block/blk-core.c b/block/blk-core.c index 94bf37f8e61d..ce08f0aa9dfc 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -708,7 +708,7 @@ static void __submit_bio(struct bio *bio) * * bio_list_on_stack[0] contains bios submitted by the current ->submit_bio. * bio_list_on_stack[1] contains bios that were submitted before the current - * ->submit_bio_bio, but that haven't been processed yet. + * ->submit_bio, but that haven't been processed yet. */ static void __submit_bio_noacct(struct bio *bio) { @@ -975,7 +975,7 @@ int bio_poll(struct bio *bio, struct io_comp_batch *iob, unsigned int flags) { struct request_queue *q = bdev_get_queue(bio->bi_bdev); blk_qc_t cookie = READ_ONCE(bio->bi_cookie); - int ret; + int ret = 0; if (cookie == BLK_QC_T_NONE || !test_bit(QUEUE_FLAG_POLL, &q->queue_flags)) @@ -985,10 +985,14 @@ int bio_poll(struct bio *bio, struct io_comp_batch *iob, unsigned int flags) if (blk_queue_enter(q, BLK_MQ_REQ_NOWAIT)) return 0; - if (WARN_ON_ONCE(!queue_is_mq(q))) - ret = 0; /* not yet implemented, should not happen */ - else + if (queue_is_mq(q)) { ret = blk_mq_poll(q, cookie, iob, flags); + } else { + struct gendisk *disk = q->disk; + + if (disk && disk->fops->poll_bio) + ret = disk->fops->poll_bio(bio, iob, flags); + } blk_queue_exit(q); return ret; } diff --git a/block/genhd.c b/block/genhd.c index e351fac41bf2..1ed46a6f94f5 100644 --- a/block/genhd.c +++ b/block/genhd.c @@ -410,6 +410,10 @@ int __must_check device_add_disk(struct device *parent, struct gendisk *disk, struct device *ddev = disk_to_dev(disk); int ret; + /* Only makes sense for bio-based to set ->poll_bio */ + if (queue_is_mq(disk->queue) && disk->fops->poll_bio) + return -EINVAL; + /* * The disk queue should now be all set with enough information about * the device for the elevator code to pick an adequate default diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index f757f9c2871f..51f1b1ddbed2 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -1455,6 +1455,8 @@ enum blk_unique_id { struct block_device_operations { void (*submit_bio)(struct bio *bio); + int (*poll_bio)(struct bio *bio, struct io_comp_batch *iob, + unsigned int flags); int (*open) (struct block_device *, fmode_t); void (*release) (struct gendisk *, fmode_t); int (*rw_page)(struct block_device *, sector_t, struct page *, unsigned int); -- cgit From b7f8dff09827c96032c34a945ee7757e394b5952 Mon Sep 17 00:00:00 2001 From: Mike Snitzer Date: Thu, 10 Mar 2022 11:45:58 -0500 Subject: dm: simplify dm_sumbit_bio_remap interface Remove the from_wq argument from dm_sumbit_bio_remap(). Eliminates the need for dm_sumbit_bio_remap() callers to know whether they are calling for a workqueue or from the original dm_submit_bio(). Add map_task to dm_io struct, record the map_task in alloc_io and clear it after all target ->map() calls have completed. Update dm_sumbit_bio_remap to check if 'current' matches io->map_task rather than rely on passed 'from_rq' argument. This change really simplifies the chore of porting each DM target to using dm_sumbit_bio_remap() because there is no longer the risk of programming error by not completely knowing all the different contexts a particular method that calls dm_sumbit_bio_remap() might be used in. Signed-off-by: Mike Snitzer --- drivers/md/dm-core.h | 1 + drivers/md/dm-crypt.c | 6 +++--- drivers/md/dm-delay.c | 2 +- drivers/md/dm-thin.c | 4 ++-- drivers/md/dm.c | 8 ++++---- include/linux/device-mapper.h | 2 +- 6 files changed, 12 insertions(+), 11 deletions(-) (limited to 'include/linux') diff --git a/drivers/md/dm-core.h b/drivers/md/dm-core.h index 8cc03c0c262e..8d3d11887343 100644 --- a/drivers/md/dm-core.h +++ b/drivers/md/dm-core.h @@ -237,6 +237,7 @@ struct dm_io { unsigned long start_time; void *data; struct hlist_node node; + struct task_struct *map_task; spinlock_t endio_lock; struct dm_stats_aux stats_aux; /* last member of dm_target_io is 'struct bio' */ diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c index ec746cc4b1f8..a1730cc8ae27 100644 --- a/drivers/md/dm-crypt.c +++ b/drivers/md/dm-crypt.c @@ -1857,7 +1857,7 @@ static int kcryptd_io_read(struct dm_crypt_io *io, gfp_t gfp) return 1; } - dm_submit_bio_remap(io->base_bio, clone, (gfp != CRYPT_MAP_READ_GFP)); + dm_submit_bio_remap(io->base_bio, clone); return 0; } @@ -1883,7 +1883,7 @@ static void kcryptd_io_write(struct dm_crypt_io *io) { struct bio *clone = io->ctx.bio_out; - dm_submit_bio_remap(io->base_bio, clone, true); + dm_submit_bio_remap(io->base_bio, clone); } #define crypt_io_from_node(node) rb_entry((node), struct dm_crypt_io, rb_node) @@ -1962,7 +1962,7 @@ static void kcryptd_crypt_write_io_submit(struct dm_crypt_io *io, int async) if ((likely(!async) && test_bit(DM_CRYPT_NO_OFFLOAD, &cc->flags)) || test_bit(DM_CRYPT_NO_WRITE_WORKQUEUE, &cc->flags)) { - dm_submit_bio_remap(io->base_bio, clone, true); + dm_submit_bio_remap(io->base_bio, clone); return; } diff --git a/drivers/md/dm-delay.c b/drivers/md/dm-delay.c index b25b45011b11..9a51bf51a859 100644 --- a/drivers/md/dm-delay.c +++ b/drivers/md/dm-delay.c @@ -72,7 +72,7 @@ static void flush_bios(struct bio *bio) while (bio) { n = bio->bi_next; bio->bi_next = NULL; - dm_submit_bio_remap(bio, NULL, true); + dm_submit_bio_remap(bio, NULL); bio = n; } } diff --git a/drivers/md/dm-thin.c b/drivers/md/dm-thin.c index ba74bc22ba42..4d25d0e27031 100644 --- a/drivers/md/dm-thin.c +++ b/drivers/md/dm-thin.c @@ -755,7 +755,7 @@ static void issue(struct thin_c *tc, struct bio *bio) struct pool *pool = tc->pool; if (!bio_triggers_commit(tc, bio)) { - dm_submit_bio_remap(bio, NULL, true); + dm_submit_bio_remap(bio, NULL); return; } @@ -2383,7 +2383,7 @@ static void process_deferred_bios(struct pool *pool) if (bio->bi_opf & REQ_PREFLUSH) bio_endio(bio); else - dm_submit_bio_remap(bio, NULL, true); + dm_submit_bio_remap(bio, NULL); } } diff --git a/drivers/md/dm.c b/drivers/md/dm.c index 1c5c9036a20e..c470f54f9193 100644 --- a/drivers/md/dm.c +++ b/drivers/md/dm.c @@ -574,6 +574,7 @@ static struct dm_io *alloc_io(struct mapped_device *md, struct bio *bio) this_cpu_inc(*md->pending_io); io->orig_bio = NULL; io->md = md; + io->map_task = current; spin_lock_init(&io->endio_lock); io->start_time = jiffies; @@ -1189,15 +1190,13 @@ static inline void __dm_submit_bio_remap(struct bio *clone, /* * @clone: clone bio that DM core passed to target's .map function * @tgt_clone: clone of @clone bio that target needs submitted - * @from_wq: caller is a workqueue thread managed by DM target * * Targets should use this interface to submit bios they take * ownership of when returning DM_MAPIO_SUBMITTED. * * Target should also enable ti->accounts_remapped_io */ -void dm_submit_bio_remap(struct bio *clone, struct bio *tgt_clone, - bool from_wq) +void dm_submit_bio_remap(struct bio *clone, struct bio *tgt_clone) { struct dm_target_io *tio = clone_to_tio(clone); struct dm_io *io = tio->io; @@ -1212,7 +1211,7 @@ void dm_submit_bio_remap(struct bio *clone, struct bio *tgt_clone, * Account io->origin_bio to DM dev on behalf of target * that took ownership of IO with DM_MAPIO_SUBMITTED. */ - if (!from_wq) { + if (io->map_task == current) { /* Still in target's map function */ io->start_io_acct = true; } else { @@ -1568,6 +1567,7 @@ static void dm_split_and_process_bio(struct mapped_device *md, } error = __split_and_process_bio(&ci); + ci.io->map_task = NULL; if (error || !ci.sector_count) goto out; diff --git a/include/linux/device-mapper.h b/include/linux/device-mapper.h index 70cd9449275a..901ec191250c 100644 --- a/include/linux/device-mapper.h +++ b/include/linux/device-mapper.h @@ -471,7 +471,7 @@ int dm_suspended(struct dm_target *ti); int dm_post_suspending(struct dm_target *ti); int dm_noflush_suspending(struct dm_target *ti); void dm_accept_partial_bio(struct bio *bio, unsigned n_sectors); -void dm_submit_bio_remap(struct bio *clone, struct bio *tgt_clone, bool from_wq); +void dm_submit_bio_remap(struct bio *clone, struct bio *tgt_clone); union map_info *dm_get_rq_mapinfo(struct request *rq); #ifdef CONFIG_BLK_DEV_ZONED -- cgit