diff options
Diffstat (limited to 'drivers/md')
| -rw-r--r-- | drivers/md/bcache/writeback.h | 3 | ||||
| -rw-r--r-- | drivers/md/dm-kcopyd.c | 19 | ||||
| -rw-r--r-- | drivers/md/dm-snap.c | 22 | ||||
| -rw-r--r-- | drivers/md/dm-thin-metadata.c | 4 | ||||
| -rw-r--r-- | drivers/md/dm-thin-metadata.h | 2 | ||||
| -rw-r--r-- | drivers/md/dm-thin.c | 65 | ||||
| -rw-r--r-- | drivers/md/raid10.c | 3 | ||||
| -rw-r--r-- | drivers/md/raid5.c | 2 |
8 files changed, 101 insertions, 19 deletions
diff --git a/drivers/md/bcache/writeback.h b/drivers/md/bcache/writeback.h index cdf8d253209e..6fb6dd600970 100644 --- a/drivers/md/bcache/writeback.h +++ b/drivers/md/bcache/writeback.h @@ -68,6 +68,9 @@ static inline bool should_writeback(struct cached_dev *dc, struct bio *bio, in_use > CUTOFF_WRITEBACK_SYNC) return false; + if (bio_op(bio) == REQ_OP_DISCARD) + return false; + if (dc->partial_stripes_expensive && bcache_dev_stripe_dirty(dc, bio->bi_iter.bi_sector, bio_sectors(bio))) diff --git a/drivers/md/dm-kcopyd.c b/drivers/md/dm-kcopyd.c index 56fcccc30554..e0cfde3501e0 100644 --- a/drivers/md/dm-kcopyd.c +++ b/drivers/md/dm-kcopyd.c @@ -55,15 +55,17 @@ struct dm_kcopyd_client { struct dm_kcopyd_throttle *throttle; /* - * We maintain three lists of jobs: + * We maintain four lists of jobs: * * i) jobs waiting for pages * ii) jobs that have pages, and are waiting for the io to be issued. - * iii) jobs that have completed. + * iii) jobs that don't need to do any IO and just run a callback + * iv) jobs that have completed. * - * All three of these are protected by job_lock. + * All four of these are protected by job_lock. */ spinlock_t job_lock; + struct list_head callback_jobs; struct list_head complete_jobs; struct list_head io_jobs; struct list_head pages_jobs; @@ -584,6 +586,7 @@ static void do_work(struct work_struct *work) struct dm_kcopyd_client *kc = container_of(work, struct dm_kcopyd_client, kcopyd_work); struct blk_plug plug; + unsigned long flags; /* * The order that these are called is *very* important. @@ -592,6 +595,10 @@ static void do_work(struct work_struct *work) * list. io jobs call wake when they complete and it all * starts again. */ + spin_lock_irqsave(&kc->job_lock, flags); + list_splice_tail_init(&kc->callback_jobs, &kc->complete_jobs); + spin_unlock_irqrestore(&kc->job_lock, flags); + blk_start_plug(&plug); process_jobs(&kc->complete_jobs, kc, run_complete_job); process_jobs(&kc->pages_jobs, kc, run_pages_job); @@ -609,7 +616,7 @@ static void dispatch_job(struct kcopyd_job *job) struct dm_kcopyd_client *kc = job->kc; atomic_inc(&kc->nr_jobs); if (unlikely(!job->source.count)) - push(&kc->complete_jobs, job); + push(&kc->callback_jobs, job); else if (job->pages == &zero_page_list) push(&kc->io_jobs, job); else @@ -796,7 +803,7 @@ void dm_kcopyd_do_callback(void *j, int read_err, unsigned long write_err) job->read_err = read_err; job->write_err = write_err; - push(&kc->complete_jobs, job); + push(&kc->callback_jobs, job); wake(kc); } EXPORT_SYMBOL(dm_kcopyd_do_callback); @@ -826,6 +833,7 @@ struct dm_kcopyd_client *dm_kcopyd_client_create(struct dm_kcopyd_throttle *thro return ERR_PTR(-ENOMEM); spin_lock_init(&kc->job_lock); + INIT_LIST_HEAD(&kc->callback_jobs); INIT_LIST_HEAD(&kc->complete_jobs); INIT_LIST_HEAD(&kc->io_jobs); INIT_LIST_HEAD(&kc->pages_jobs); @@ -875,6 +883,7 @@ void dm_kcopyd_client_destroy(struct dm_kcopyd_client *kc) /* Wait for completion of all jobs submitted by this client. */ wait_event(kc->destroyq, !atomic_read(&kc->nr_jobs)); + BUG_ON(!list_empty(&kc->callback_jobs)); BUG_ON(!list_empty(&kc->complete_jobs)); BUG_ON(!list_empty(&kc->io_jobs)); BUG_ON(!list_empty(&kc->pages_jobs)); diff --git a/drivers/md/dm-snap.c b/drivers/md/dm-snap.c index c65feeada864..2da0b9b213c7 100644 --- a/drivers/md/dm-snap.c +++ b/drivers/md/dm-snap.c @@ -19,6 +19,7 @@ #include <linux/vmalloc.h> #include <linux/log2.h> #include <linux/dm-kcopyd.h> +#include <linux/semaphore.h> #include "dm.h" @@ -105,6 +106,9 @@ struct dm_snapshot { /* The on disk metadata handler */ struct dm_exception_store *store; + /* Maximum number of in-flight COW jobs. */ + struct semaphore cow_count; + struct dm_kcopyd_client *kcopyd_client; /* Wait for events based on state_bits */ @@ -145,6 +149,19 @@ struct dm_snapshot { #define RUNNING_MERGE 0 #define SHUTDOWN_MERGE 1 +/* + * Maximum number of chunks being copied on write. + * + * The value was decided experimentally as a trade-off between memory + * consumption, stalling the kernel's workqueues and maintaining a high enough + * throughput. + */ +#define DEFAULT_COW_THRESHOLD 2048 + +static int cow_threshold = DEFAULT_COW_THRESHOLD; +module_param_named(snapshot_cow_threshold, cow_threshold, int, 0644); +MODULE_PARM_DESC(snapshot_cow_threshold, "Maximum number of chunks being copied on write"); + DECLARE_DM_KCOPYD_THROTTLE_WITH_MODULE_PARM(snapshot_copy_throttle, "A percentage of time allocated for copy on write"); @@ -1189,6 +1206,8 @@ static int snapshot_ctr(struct dm_target *ti, unsigned int argc, char **argv) goto bad_hash_tables; } + sema_init(&s->cow_count, (cow_threshold > 0) ? cow_threshold : INT_MAX); + s->kcopyd_client = dm_kcopyd_client_create(&dm_kcopyd_throttle); if (IS_ERR(s->kcopyd_client)) { r = PTR_ERR(s->kcopyd_client); @@ -1560,6 +1579,7 @@ static void copy_callback(int read_err, unsigned long write_err, void *context) } list_add(&pe->out_of_order_entry, lh); } + up(&s->cow_count); } /* @@ -1583,6 +1603,7 @@ static void start_copy(struct dm_snap_pending_exception *pe) dest.count = src.count; /* Hand over to kcopyd */ + down(&s->cow_count); dm_kcopyd_copy(s->kcopyd_client, &src, 1, &dest, 0, copy_callback, pe); } @@ -1602,6 +1623,7 @@ static void start_full_bio(struct dm_snap_pending_exception *pe, pe->full_bio = bio; pe->full_bio_end_io = bio->bi_end_io; + down(&s->cow_count); callback_data = dm_kcopyd_prepare_callback(s->kcopyd_client, copy_callback, pe); diff --git a/drivers/md/dm-thin-metadata.c b/drivers/md/dm-thin-metadata.c index 149fbac97cb6..d20f4023f6c1 100644 --- a/drivers/md/dm-thin-metadata.c +++ b/drivers/md/dm-thin-metadata.c @@ -1689,7 +1689,7 @@ int dm_thin_remove_range(struct dm_thin_device *td, return r; } -int dm_pool_block_is_used(struct dm_pool_metadata *pmd, dm_block_t b, bool *result) +int dm_pool_block_is_shared(struct dm_pool_metadata *pmd, dm_block_t b, bool *result) { int r; uint32_t ref_count; @@ -1697,7 +1697,7 @@ int dm_pool_block_is_used(struct dm_pool_metadata *pmd, dm_block_t b, bool *resu down_read(&pmd->root_lock); r = dm_sm_get_count(pmd->data_sm, b, &ref_count); if (!r) - *result = (ref_count != 0); + *result = (ref_count > 1); up_read(&pmd->root_lock); return r; diff --git a/drivers/md/dm-thin-metadata.h b/drivers/md/dm-thin-metadata.h index 35e954ea20a9..f6be0d733c20 100644 --- a/drivers/md/dm-thin-metadata.h +++ b/drivers/md/dm-thin-metadata.h @@ -195,7 +195,7 @@ int dm_pool_get_metadata_dev_size(struct dm_pool_metadata *pmd, int dm_pool_get_data_dev_size(struct dm_pool_metadata *pmd, dm_block_t *result); -int dm_pool_block_is_used(struct dm_pool_metadata *pmd, dm_block_t b, bool *result); +int dm_pool_block_is_shared(struct dm_pool_metadata *pmd, dm_block_t b, bool *result); int dm_pool_inc_data_range(struct dm_pool_metadata *pmd, dm_block_t b, dm_block_t e); int dm_pool_dec_data_range(struct dm_pool_metadata *pmd, dm_block_t b, dm_block_t e); diff --git a/drivers/md/dm-thin.c b/drivers/md/dm-thin.c index 81309d7836c5..345f4d81ba07 100644 --- a/drivers/md/dm-thin.c +++ b/drivers/md/dm-thin.c @@ -257,6 +257,7 @@ struct pool { spinlock_t lock; struct bio_list deferred_flush_bios; + struct bio_list deferred_flush_completions; struct list_head prepared_mappings; struct list_head prepared_discards; struct list_head prepared_discards_pt2; @@ -925,6 +926,39 @@ static void process_prepared_mapping_fail(struct dm_thin_new_mapping *m) mempool_free(m, m->tc->pool->mapping_pool); } +static void complete_overwrite_bio(struct thin_c *tc, struct bio *bio) +{ + struct pool *pool = tc->pool; + unsigned long flags; + + /* + * If the bio has the REQ_FUA flag set we must commit the metadata + * before signaling its completion. + */ + if (!bio_triggers_commit(tc, bio)) { + bio_endio(bio); + return; + } + + /* + * Complete bio with an error if earlier I/O caused changes to the + * metadata that can't be committed, e.g, due to I/O errors on the + * metadata device. + */ + if (dm_thin_aborted_changes(tc->td)) { + bio_io_error(bio); + return; + } + + /* + * Batch together any bios that trigger commits and then issue a + * single commit for them in process_deferred_bios(). + */ + spin_lock_irqsave(&pool->lock, flags); + bio_list_add(&pool->deferred_flush_completions, bio); + spin_unlock_irqrestore(&pool->lock, flags); +} + static void process_prepared_mapping(struct dm_thin_new_mapping *m) { struct thin_c *tc = m->tc; @@ -957,7 +991,7 @@ static void process_prepared_mapping(struct dm_thin_new_mapping *m) */ if (bio) { inc_remap_and_issue_cell(tc, m->cell, m->data_block); - bio_endio(bio); + complete_overwrite_bio(tc, bio); } else { inc_all_io_entry(tc->pool, m->cell->holder); remap_and_issue(tc, m->cell->holder, m->data_block); @@ -1017,7 +1051,7 @@ static void passdown_double_checking_shared_status(struct dm_thin_new_mapping *m * passdown we have to check that these blocks are now unused. */ int r = 0; - bool used = true; + bool shared = true; struct thin_c *tc = m->tc; struct pool *pool = tc->pool; dm_block_t b = m->data_block, e, end = m->data_block + m->virt_end - m->virt_begin; @@ -1027,11 +1061,11 @@ static void passdown_double_checking_shared_status(struct dm_thin_new_mapping *m while (b != end) { /* find start of unmapped run */ for (; b < end; b++) { - r = dm_pool_block_is_used(pool->pmd, b, &used); + r = dm_pool_block_is_shared(pool->pmd, b, &shared); if (r) goto out; - if (!used) + if (!shared) break; } @@ -1040,11 +1074,11 @@ static void passdown_double_checking_shared_status(struct dm_thin_new_mapping *m /* find end of run */ for (e = b + 1; e != end; e++) { - r = dm_pool_block_is_used(pool->pmd, e, &used); + r = dm_pool_block_is_shared(pool->pmd, e, &shared); if (r) goto out; - if (used) + if (shared) break; } @@ -2303,7 +2337,7 @@ static void process_deferred_bios(struct pool *pool) { unsigned long flags; struct bio *bio; - struct bio_list bios; + struct bio_list bios, bio_completions; struct thin_c *tc; tc = get_first_thin(pool); @@ -2314,26 +2348,36 @@ static void process_deferred_bios(struct pool *pool) } /* - * If there are any deferred flush bios, we must commit - * the metadata before issuing them. + * If there are any deferred flush bios, we must commit the metadata + * before issuing them or signaling their completion. */ bio_list_init(&bios); + bio_list_init(&bio_completions); + spin_lock_irqsave(&pool->lock, flags); bio_list_merge(&bios, &pool->deferred_flush_bios); bio_list_init(&pool->deferred_flush_bios); + + bio_list_merge(&bio_completions, &pool->deferred_flush_completions); + bio_list_init(&pool->deferred_flush_completions); spin_unlock_irqrestore(&pool->lock, flags); - if (bio_list_empty(&bios) && + if (bio_list_empty(&bios) && bio_list_empty(&bio_completions) && !(dm_pool_changed_this_transaction(pool->pmd) && need_commit_due_to_time(pool))) return; if (commit(pool)) { + bio_list_merge(&bios, &bio_completions); + while ((bio = bio_list_pop(&bios))) bio_io_error(bio); return; } pool->last_commit_jiffies = jiffies; + while ((bio = bio_list_pop(&bio_completions))) + bio_endio(bio); + while ((bio = bio_list_pop(&bios))) generic_make_request(bio); } @@ -2968,6 +3012,7 @@ static struct pool *pool_create(struct mapped_device *pool_md, INIT_DELAYED_WORK(&pool->no_space_timeout, do_no_space_timeout); spin_lock_init(&pool->lock); bio_list_init(&pool->deferred_flush_bios); + bio_list_init(&pool->deferred_flush_completions); INIT_LIST_HEAD(&pool->prepared_mappings); INIT_LIST_HEAD(&pool->prepared_discards); INIT_LIST_HEAD(&pool->prepared_discards_pt2); diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c index b62e6ab66b31..717787d09e0f 100644 --- a/drivers/md/raid10.c +++ b/drivers/md/raid10.c @@ -3798,6 +3798,8 @@ static int raid10_run(struct mddev *mddev) set_bit(MD_RECOVERY_RUNNING, &mddev->recovery); mddev->sync_thread = md_register_thread(md_do_sync, mddev, "reshape"); + if (!mddev->sync_thread) + goto out_free_conf; } return 0; @@ -4489,7 +4491,6 @@ bio_full: atomic_inc(&r10_bio->remaining); read_bio->bi_next = NULL; generic_make_request(read_bio); - sector_nr += nr_sectors; sectors_done += nr_sectors; if (sector_nr <= last) goto read_more; diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index 4afc419da60f..9ec74dfe94f4 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -6977,6 +6977,8 @@ static int raid5_run(struct mddev *mddev) set_bit(MD_RECOVERY_RUNNING, &mddev->recovery); mddev->sync_thread = md_register_thread(md_do_sync, mddev, "reshape"); + if (!mddev->sync_thread) + goto abort; } /* Ok, everything is just fine now */ |