diff options
| -rw-r--r-- | Documentation/admin-guide/md.rst | 32 | ||||
| -rw-r--r-- | Documentation/md/md-cluster.txt | 2 | ||||
| -rw-r--r-- | Documentation/md/raid5-ppl.txt | 44 | ||||
| -rw-r--r-- | block/bio.c | 61 | ||||
| -rw-r--r-- | drivers/md/Makefile | 2 | ||||
| -rw-r--r-- | drivers/md/bitmap.c | 59 | ||||
| -rw-r--r-- | drivers/md/bitmap.h | 3 | ||||
| -rw-r--r-- | drivers/md/linear.c | 75 | ||||
| -rw-r--r-- | drivers/md/md-cluster.c | 223 | ||||
| -rw-r--r-- | drivers/md/md-cluster.h | 1 | ||||
| -rw-r--r-- | drivers/md/md.c | 414 | ||||
| -rw-r--r-- | drivers/md/md.h | 71 | ||||
| -rw-r--r-- | drivers/md/raid0.c | 78 | ||||
| -rw-r--r-- | drivers/md/raid1.c | 679 | ||||
| -rw-r--r-- | drivers/md/raid1.h | 13 | ||||
| -rw-r--r-- | drivers/md/raid10.c | 736 | ||||
| -rw-r--r-- | drivers/md/raid10.h | 1 | ||||
| -rw-r--r-- | drivers/md/raid5-cache.c | 362 | ||||
| -rw-r--r-- | drivers/md/raid5-log.h | 115 | ||||
| -rw-r--r-- | drivers/md/raid5-ppl.c | 1271 | ||||
| -rw-r--r-- | drivers/md/raid5.c | 643 | ||||
| -rw-r--r-- | drivers/md/raid5.h | 106 | ||||
| -rw-r--r-- | include/linux/bio.h | 11 | ||||
| -rw-r--r-- | include/linux/percpu-refcount.h | 1 | ||||
| -rw-r--r-- | include/uapi/linux/raid/md_p.h | 45 | ||||
| -rw-r--r-- | lib/percpu-refcount.c | 17 |
26 files changed, 3582 insertions, 1483 deletions
diff --git a/Documentation/admin-guide/md.rst b/Documentation/admin-guide/md.rst index 1e61bf50595c..84de718f24a4 100644 --- a/Documentation/admin-guide/md.rst +++ b/Documentation/admin-guide/md.rst @@ -276,14 +276,14 @@ All md devices contain: array creation it will default to 0, though starting the array as ``clean`` will set it much larger. - new_dev + new_dev This file can be written but not read. The value written should be a block device number as major:minor. e.g. 8:0 This will cause that device to be attached to the array, if it is available. It will then appear at md/dev-XXX (depending on the name of the device) and further configuration is then possible. - safe_mode_delay + safe_mode_delay When an md array has seen no write requests for a certain period of time, it will be marked as ``clean``. When another write request arrives, the array is marked as ``dirty`` before the write @@ -292,7 +292,7 @@ All md devices contain: period as a number of seconds. The default is 200msec (0.200). Writing a value of 0 disables safemode. - array_state + array_state This file contains a single word which describes the current state of the array. In many cases, the state can be set by writing the word for the desired state, however some states @@ -401,7 +401,30 @@ All md devices contain: once the array becomes non-degraded, and this fact has been recorded in the metadata. + consistency_policy + This indicates how the array maintains consistency in case of unexpected + shutdown. It can be: + none + Array has no redundancy information, e.g. raid0, linear. + + resync + Full resync is performed and all redundancy is regenerated when the + array is started after unclean shutdown. + + bitmap + Resync assisted by a write-intent bitmap. + + journal + For raid4/5/6, journal device is used to log transactions and replay + after unclean shutdown. + + ppl + For raid5 only, Partial Parity Log is used to close the write hole and + eliminate resync. + + The accepted values when writing to this file are ``ppl`` and ``resync``, + used to enable and disable PPL. As component devices are added to an md array, they appear in the ``md`` @@ -563,6 +586,9 @@ Each directory contains: adds bad blocks without acknowledging them. This is largely for testing. + ppl_sector, ppl_size + Location and size (in sectors) of the space used for Partial Parity Log + on this device. An active md device will also contain an entry for each active device diff --git a/Documentation/md/md-cluster.txt b/Documentation/md/md-cluster.txt index d22103994aef..82ee51604e9a 100644 --- a/Documentation/md/md-cluster.txt +++ b/Documentation/md/md-cluster.txt @@ -321,4 +321,4 @@ The algorithm is: There are somethings which are not supported by cluster MD yet. -- update size and change array_sectors. +- change array_sectors. diff --git a/Documentation/md/raid5-ppl.txt b/Documentation/md/raid5-ppl.txt new file mode 100644 index 000000000000..127072b09363 --- /dev/null +++ b/Documentation/md/raid5-ppl.txt @@ -0,0 +1,44 @@ +Partial Parity Log + +Partial Parity Log (PPL) is a feature available for RAID5 arrays. The issue +addressed by PPL is that after a dirty shutdown, parity of a particular stripe +may become inconsistent with data on other member disks. If the array is also +in degraded state, there is no way to recalculate parity, because one of the +disks is missing. This can lead to silent data corruption when rebuilding the +array or using it is as degraded - data calculated from parity for array blocks +that have not been touched by a write request during the unclean shutdown can +be incorrect. Such condition is known as the RAID5 Write Hole. Because of +this, md by default does not allow starting a dirty degraded array. + +Partial parity for a write operation is the XOR of stripe data chunks not +modified by this write. It is just enough data needed for recovering from the +write hole. XORing partial parity with the modified chunks produces parity for +the stripe, consistent with its state before the write operation, regardless of +which chunk writes have completed. If one of the not modified data disks of +this stripe is missing, this updated parity can be used to recover its +contents. PPL recovery is also performed when starting an array after an +unclean shutdown and all disks are available, eliminating the need to resync +the array. Because of this, using write-intent bitmap and PPL together is not +supported. + +When handling a write request PPL writes partial parity before new data and +parity are dispatched to disks. PPL is a distributed log - it is stored on +array member drives in the metadata area, on the parity drive of a particular +stripe. It does not require a dedicated journaling drive. Write performance is +reduced by up to 30%-40% but it scales with the number of drives in the array +and the journaling drive does not become a bottleneck or a single point of +failure. + +Unlike raid5-cache, the other solution in md for closing the write hole, PPL is +not a true journal. It does not protect from losing in-flight data, only from +silent data corruption. If a dirty disk of a stripe is lost, no PPL recovery is +performed for this stripe (parity is not updated). So it is possible to have +arbitrary data in the written part of a stripe if that disk is lost. In such +case the behavior is the same as in plain raid5. + +PPL is available for md version-1 metadata and external (specifically IMSM) +metadata arrays. It can be enabled using mdadm option --consistency-policy=ppl. + +Currently, volatile write-back cache should be disabled on all member drives +when using PPL. Otherwise it cannot guarantee consistency in case of power +failure. diff --git a/block/bio.c b/block/bio.c index f4d207180266..888e7801c638 100644 --- a/block/bio.c +++ b/block/bio.c @@ -633,20 +633,21 @@ struct bio *bio_clone_fast(struct bio *bio, gfp_t gfp_mask, struct bio_set *bs) } EXPORT_SYMBOL(bio_clone_fast); -static struct bio *__bio_clone_bioset(struct bio *bio_src, gfp_t gfp_mask, - struct bio_set *bs, int offset, - int size) +/** + * bio_clone_bioset - clone a bio + * @bio_src: bio to clone + * @gfp_mask: allocation priority + * @bs: bio_set to allocate from + * + * Clone bio. Caller will own the returned bio, but not the actual data it + * points to. Reference count of returned bio will be one. + */ +struct bio *bio_clone_bioset(struct bio *bio_src, gfp_t gfp_mask, + struct bio_set *bs) { struct bvec_iter iter; struct bio_vec bv; struct bio *bio; - struct bvec_iter iter_src = bio_src->bi_iter; - - /* for supporting partial clone */ - if (offset || size != bio_src->bi_iter.bi_size) { - bio_advance_iter(bio_src, &iter_src, offset); - iter_src.bi_size = size; - } /* * Pre immutable biovecs, __bio_clone() used to just do a memcpy from @@ -670,8 +671,7 @@ static struct bio *__bio_clone_bioset(struct bio *bio_src, gfp_t gfp_mask, * __bio_clone_fast() anyways. */ - bio = bio_alloc_bioset(gfp_mask, __bio_segments(bio_src, - &iter_src), bs); + bio = bio_alloc_bioset(gfp_mask, bio_segments(bio_src), bs); if (!bio) return NULL; bio->bi_bdev = bio_src->bi_bdev; @@ -688,7 +688,7 @@ static struct bio *__bio_clone_bioset(struct bio *bio_src, gfp_t gfp_mask, bio->bi_io_vec[bio->bi_vcnt++] = bio_src->bi_io_vec[0]; break; default: - __bio_for_each_segment(bv, bio_src, iter, iter_src) + bio_for_each_segment(bv, bio_src, iter) bio->bi_io_vec[bio->bi_vcnt++] = bv; break; } @@ -707,44 +707,9 @@ static struct bio *__bio_clone_bioset(struct bio *bio_src, gfp_t gfp_mask, return bio; } - -/** - * bio_clone_bioset - clone a bio - * @bio_src: bio to clone - * @gfp_mask: allocation priority - * @bs: bio_set to allocate from - * - * Clone bio. Caller will own the returned bio, but not the actual data it - * points to. Reference count of returned bio will be one. - */ -struct bio *bio_clone_bioset(struct bio *bio_src, gfp_t gfp_mask, - struct bio_set *bs) -{ - return __bio_clone_bioset(bio_src, gfp_mask, bs, 0, - bio_src->bi_iter.bi_size); -} EXPORT_SYMBOL(bio_clone_bioset); /** - * bio_clone_bioset_partial - clone a partial bio - * @bio_src: bio to clone - * @gfp_mask: allocation priority - * @bs: bio_set to allocate from - * @offset: cloned starting from the offset - * @size: size for the cloned bio - * - * Clone bio. Caller will own the returned bio, but not the actual data it - * points to. Reference count of returned bio will be one. - */ -struct bio *bio_clone_bioset_partial(struct bio *bio_src, gfp_t gfp_mask, - struct bio_set *bs, int offset, - int size) -{ - return __bio_clone_bioset(bio_src, gfp_mask, bs, offset, size); -} -EXPORT_SYMBOL(bio_clone_bioset_partial); - -/** * bio_add_pc_page - attempt to add page to bio * @q: the target queue * @bio: destination bio diff --git a/drivers/md/Makefile b/drivers/md/Makefile index 3cbda1af87a0..4d48714ccc6b 100644 --- a/drivers/md/Makefile +++ b/drivers/md/Makefile @@ -18,7 +18,7 @@ dm-cache-cleaner-y += dm-cache-policy-cleaner.o 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 +raid456-y += raid5.o raid5-cache.o raid5-ppl.o # Note: link order is important. All raid personalities # and must come before md.o, as they each initialise diff --git a/drivers/md/bitmap.c b/drivers/md/bitmap.c index 9fb2ccac958a..bf7419a56454 100644 --- a/drivers/md/bitmap.c +++ b/drivers/md/bitmap.c @@ -471,6 +471,7 @@ void bitmap_update_sb(struct bitmap *bitmap) kunmap_atomic(sb); write_page(bitmap, bitmap->storage.sb_page, 1); } +EXPORT_SYMBOL(bitmap_update_sb); /* print out the bitmap file superblock */ void bitmap_print_sb(struct bitmap *bitmap) @@ -696,7 +697,7 @@ re_read: out: kunmap_atomic(sb); - /* Assiging chunksize is required for "re_read" */ + /* Assigning chunksize is required for "re_read" */ bitmap->mddev->bitmap_info.chunksize = chunksize; if (err == 0 && nodes && (bitmap->cluster_slot < 0)) { err = md_setup_cluster(bitmap->mddev, nodes); @@ -1727,7 +1728,7 @@ void bitmap_flush(struct mddev *mddev) /* * free memory that was allocated */ -static void bitmap_free(struct bitmap *bitmap) +void bitmap_free(struct bitmap *bitmap) { unsigned long k, pages; struct bitmap_page *bp; @@ -1761,6 +1762,21 @@ static void bitmap_free(struct bitmap *bitmap) kfree(bp); kfree(bitmap); } +EXPORT_SYMBOL(bitmap_free); + +void bitmap_wait_behind_writes(struct mddev *mddev) +{ + struct bitmap *bitmap = mddev->bitmap; + + /* wait for behind writes to complete */ + if (bitmap && atomic_read(&bitmap->behind_writes) > 0) { + pr_debug("md:%s: behind writes in progress - waiting to stop.\n", + mdname(mddev)); + /* need to kick something here to make sure I/O goes? */ + wait_event(bitmap->behind_wait, + atomic_read(&bitmap->behind_writes) == 0); + } +} void bitmap_destroy(struct mddev *mddev) { @@ -1769,6 +1785,8 @@ void bitmap_destroy(struct mddev *mddev) if (!bitmap) /* there was no bitmap */ return; + bitmap_wait_behind_writes(mddev); + mutex_lock(&mddev->bitmap_info.mutex); spin_lock(&mddev->lock); mddev->bitmap = NULL; /* disconnect from the md device */ @@ -1920,6 +1938,27 @@ out: } EXPORT_SYMBOL_GPL(bitmap_load); +struct bitmap *get_bitmap_from_slot(struct mddev *mddev, int slot) +{ + int rv = 0; + struct bitmap *bitmap; + + bitmap = bitmap_create(mddev, slot); + if (IS_ERR(bitmap)) { + rv = PTR_ERR(bitmap); + return ERR_PTR(rv); + } + + rv = bitmap_init_from_disk(bitmap, 0); + if (rv) { + bitmap_free(bitmap); + return ERR_PTR(rv); + } + + return bitmap; +} +EXPORT_SYMBOL(get_bitmap_from_slot); + /* Loads the bitmap associated with slot and copies the resync information * to our bitmap */ @@ -1929,14 +1968,13 @@ int bitmap_copy_from_slot(struct mddev *mddev, int slot, int rv = 0, i, j; sector_t block, lo = 0, hi = 0; struct bitmap_counts *counts; - struct bitmap *bitmap = bitmap_create(mddev, slot); - - if (IS_ERR(bitmap)) - return PTR_ERR(bitmap); + struct bitmap *bitmap; - rv = bitmap_init_from_disk(bitmap, 0); - if (rv) - goto err; + bitmap = get_bitmap_from_slot(mddev, slot); + if (IS_ERR(bitmap)) { + pr_err("%s can't get bitmap from slot %d\n", __func__, slot); + return -1; + } counts = &bitmap->counts; for (j = 0; j < counts->chunks; j++) { @@ -1963,8 +2001,7 @@ int bitmap_copy_from_slot(struct mddev *mddev, int slot, bitmap_unplug(mddev->bitmap); *low = lo; *high = hi; -err: - bitmap_free(bitmap); + return rv; } EXPORT_SYMBOL_GPL(bitmap_copy_from_slot); diff --git a/drivers/md/bitmap.h b/drivers/md/bitmap.h index 5b6dd63dda91..d15721ac07a6 100644 --- a/drivers/md/bitmap.h +++ b/drivers/md/bitmap.h @@ -267,8 +267,11 @@ void bitmap_daemon_work(struct mddev *mddev); int bitmap_resize(struct bitmap *bitmap, sector_t blocks, int chunksize, int init); +struct bitmap *get_bitmap_from_slot(struct mddev *mddev, int slot); int bitmap_copy_from_slot(struct mddev *mddev, int slot, sector_t *lo, sector_t *hi, bool clear_bits); +void bitmap_free(struct bitmap *bitmap); +void bitmap_wait_behind_writes(struct mddev *mddev); #endif #endif diff --git a/drivers/md/linear.c b/drivers/md/linear.c index 377a8a3672e3..df6f2c98eca7 100644 --- a/drivers/md/linear.c +++ b/drivers/md/linear.c @@ -249,54 +249,49 @@ static void linear_make_request(struct mddev *mddev, struct bio *bio) { char b[BDEVNAME_SIZE]; struct dev_info *tmp_dev; - struct bio *split; sector_t start_sector, end_sector, data_offset; + sector_t bio_sector = bio->bi_iter.bi_sector; if (unlikely(bio->bi_opf & REQ_PREFLUSH)) { md_flush_request(mddev, bio); return; } - do { - sector_t bio_sector = bio->bi_iter.bi_sector; - tmp_dev = which_dev(mddev, bio_sector); - start_sector = tmp_dev->end_sector - tmp_dev->rdev->sectors; - end_sector = tmp_dev->end_sector; - data_offset = tmp_dev->rdev->data_offset; - bio->bi_bdev = tmp_dev->rdev->bdev; - - if (unlikely(bio_sector >= end_sector || - bio_sector < start_sector)) - goto out_of_bounds; - - if (unlikely(bio_end_sector(bio) > end_sector)) { - /* This bio crosses a device boundary, so we have to - * split it. - */ - split = bio_split(bio, end_sector - bio_sector, - GFP_NOIO, fs_bio_set); - bio_chain(split, bio); - } else { - split = bio; - } + tmp_dev = which_dev(mddev, bio_sector); + start_sector = tmp_dev->end_sector - tmp_dev->rdev->sectors; + end_sector = tmp_dev->end_sector; + data_offset = tmp_dev->rdev->data_offset; + + if (unlikely(bio_sector >= end_sector || + bio_sector < start_sector)) + goto out_of_bounds; + + if (unlikely(bio_end_sector(bio) > end_sector)) { + /* This bio crosses a device boundary, so we have to split it */ + struct bio *split = bio_split(bio, end_sector - bio_sector, + GFP_NOIO, mddev->bio_set); + bio_chain(split, bio); + generic_make_request(bio); + bio = split; + } - split->bi_iter.bi_sector = split->bi_iter.bi_sector - - start_sector + data_offset; - - if (unlikely((bio_op(split) == REQ_OP_DISCARD) && - !blk_queue_discard(bdev_get_queue(split->bi_bdev)))) { - /* Just ignore it */ - bio_endio(split); - } else { - if (mddev->gendisk) - trace_block_bio_remap(bdev_get_queue(split->bi_bdev), - split, disk_devt(mddev->gendisk), - bio_sector); - mddev_check_writesame(mddev, split); - mddev_check_write_zeroes(mddev, split); - generic_make_request(split); - } - } while (split != bio); + bio->bi_bdev = tmp_dev->rdev->bdev; + bio->bi_iter.bi_sector = bio->bi_iter.bi_sector - + start_sector + data_offset; + + if (unlikely((bio_op(bio) == REQ_OP_DISCARD) && + !blk_queue_discard(bdev_get_queue(bio->bi_bdev)))) { + /* Just ignore it */ + bio_endio(bio); + } else { + if (mddev->gendisk) + trace_block_bio_remap(bdev_get_queue(bio->bi_bdev), + bio, disk_devt(mddev->gendisk), + bio_sector); + mddev_check_writesame(mddev, bio); + mddev_check_write_zeroes(mddev, bio); + generic_make_request(bio); + } return; out_of_bounds: diff --git a/drivers/md/md-cluster.c b/drivers/md/md-cluster.c index 321ecac23027..7299ce2f08a8 100644 --- a/drivers/md/md-cluster.c +++ b/drivers/md/md-cluster.c @@ -67,9 +67,10 @@ struct resync_info { * set up all the related infos such as bitmap and personality */ #define MD_CLUSTER_ALREADY_IN_CLUSTER 6 #define MD_CLUSTER_PENDING_RECV_EVENT 7 - +#define MD_CLUSTER_HOLDING_MUTEX_FOR_RECVD 8 struct md_cluster_info { + struct mddev *mddev; /* the md device which md_cluster_info belongs to */ /* dlm lock space and resources for clustered raid. */ dlm_lockspace_t *lockspace; int slot_number; @@ -103,6 +104,7 @@ enum msg_type { REMOVE, RE_ADD, BITMAP_NEEDS_SYNC, + CHANGE_CAPACITY, }; struct cluster_msg { @@ -523,11 +525,17 @@ static void process_add_new_disk(struct mddev *mddev, struct cluster_msg *cmsg) static void process_metadata_update(struct mddev *mddev, struct cluster_msg *msg) { + int got_lock = 0; struct md_cluster_info *cinfo = mddev->cluster_info; mddev->good_device_nr = le32_to_cpu(msg->raid_slot); - set_bit(MD_RELOAD_SB, &mddev->flags); + dlm_lock_sync(cinfo->no_new_dev_lockres, DLM_LOCK_CR); - md_wakeup_thread(mddev->thread); + wait_event(mddev->thread->wqueue, + (got_lock = mddev_trylock(mddev)) || + test_bit(MD_CLUSTER_HOLDING_MUTEX_FOR_RECVD, &cinfo->state)); + md_reload_sb(mddev, mddev->good_device_nr); + if (got_lock) + mddev_unlock(mddev); } static void process_remove_disk(struct mddev *mddev, struct cluster_msg *msg) @@ -572,6 +580,10 @@ static int process_recvd_msg(struct mddev *mddev, struct cluster_msg *msg) case METADATA_UPDATED: process_metadata_update(mddev, msg); break; + case CHANGE_CAPACITY: + set_capacity(mddev->gendisk, mddev->array_sectors); + revalidate_disk(mddev->gendisk); + break; case RESYNCING: process_suspend_info(mddev, le32_to_cpu(msg->slot), le64_to_cpu(msg->low), @@ -646,11 +658,29 @@ out: * Takes the lock on the TOKEN lock resource so no other * node can communicate while the operation is underway. */ -static int lock_token(struct md_cluster_info *cinfo) +static int lock_token(struct md_cluster_info *cinfo, bool mddev_locked) { - int error; + int error, set_bit = 0; + struct mddev *mddev = cinfo->mddev; + /* + * If resync thread run after raid1d thread, then process_metadata_update + * could not continue if raid1d held reconfig_mutex (and raid1d is blocked + * since another node already got EX on Token and waitting the EX of Ack), + * so let resync wake up thread in case flag is set. + */ + if (mddev_locked && !test_bit(MD_CLUSTER_HOLDING_MUTEX_FOR_RECVD, + &cinfo->state)) { + error = test_and_set_bit_lock(MD_CLUSTER_HOLDING_MUTEX_FOR_RECVD, + &cinfo->state); + WARN_ON_ONCE(error); + md_wakeup_thread(mddev->thread); + set_bit = 1; + } error = dlm_lock_sync(cinfo->token_lockres, DLM_LOCK_EX); + if (set_bit) + clear_bit_unlock(MD_CLUSTER_HOLDING_MUTEX_FOR_RECVD, &cinfo->state); + if (error) pr_err("md-cluster(%s:%d): failed to get EX on TOKEN (%d)\n", __func__, __LINE__, error); @@ -663,12 +693,12 @@ static int lock_token(struct md_cluster_info *cinfo) /* lock_comm() * Sets the MD_CLUSTER_SEND_LOCK bit to lock the send channel. */ -static int lock_comm(struct md_cluster_info *cinfo) +static int lock_comm(struct md_cluster_info *cinfo, bool mddev_locked) { wait_event(cinfo->wait, !test_and_set_bit(MD_CLUSTER_SEND_LOCK, &cinfo->state)); - return lock_token(cinfo); + return lock_token(cinfo, mddev_locked); } static void unlock_comm(struct md_cluster_info *cinfo) @@ -743,11 +773,12 @@ failed_message: return error; } -static int sendmsg(struct md_cluster_info *cinfo, struct cluster_msg *cmsg) +static int sendmsg(struct md_cluster_info *cinfo, struct cluster_msg *cmsg, + bool mddev_locked) { int ret; - lock_comm(cinfo); + lock_comm(cinfo, mddev_locked); ret = __sendmsg(cinfo, cmsg); unlock_comm(cinfo); return ret; @@ -834,6 +865,7 @@ static int join(struct mddev *mddev, int nodes) mutex_init(&cinfo->recv_mutex); mddev->cluster_info = cinfo; + cinfo->mddev = mddev; memset(str, 0, 64); sprintf(str, "%pU", mddev->uuid); @@ -908,6 +940,7 @@ static int join(struct mddev *mddev, int nodes) return 0; err: + set_bit(MD_CLUSTER_HOLDING_MUTEX_FOR_RECVD, &cinfo->state); md_unregister_thread(&cinfo->recovery_thread); md_unregister_thread(&cinfo->recv_thread); lockres_free(cinfo->message_lockres); @@ -943,7 +976,7 @@ static void resync_bitmap(struct mddev *mddev) int err; cmsg.type = cpu_to_le32(BITMAP_NEEDS_SYNC); - err = sendmsg(cinfo, &cmsg); + err = sendmsg(cinfo, &cmsg, 1); if (err) pr_err("%s:%d: failed to send BITMAP_NEEDS_SYNC message (%d)\n", __func__, __LINE__, err); @@ -963,6 +996,7 @@ static int leave(struct mddev *mddev) if (cinfo->slot_number > 0 && mddev->recovery_cp != MaxSector) resync_bitmap(mddev); + set_bit(MD_CLUSTER_HOLDING_MUTEX_FOR_RECVD, &cinfo->state); md_unregister_thread(&cinfo->recovery_thread); md_unregister_thread(&cinfo->recv_thread); lockres_free(cinfo->message_lockres); @@ -997,16 +1031,30 @@ static int slot_number(struct mddev *mddev) static int metadata_update_start(struct mddev *mddev) { struct md_cluster_info *cinfo = mddev->cluster_info; + int ret; + + /* + * metadata_update_start is always called with the protection of + * reconfig_mutex, so set WAITING_FOR_TOKEN here. + */ + ret = test_and_set_bit_lock(MD_CLUSTER_HOLDING_MUTEX_FOR_RECVD, + &cinfo->state); + WARN_ON_ONCE(ret); + md_wakeup_thread(mddev->thread); wait_event(cinfo->wait, !test_and_set_bit(MD_CLUSTER_SEND_LOCK, &cinfo->state) || test_and_clear_bit(MD_CLUSTER_SEND_LOCKED_ALREADY, &cinfo->state)); /* If token is already locked, return 0 */ - if (cinfo->token_lockres->mode == DLM_LOCK_EX) + if (cinfo->token_lockres->mode == DLM_LOCK_EX) { + clear_bit_unlock(MD_CLUSTER_HOLDING_MUTEX_FOR_RECVD, &cinfo->state); return 0; + } - return lock_token(cinfo); + ret = lock_token(cinfo, 1); + clear_bit_unlock(MD_CLUSTER_HOLDING_MUTEX_FOR_RECVD, &cinfo->state); + return ret; } static int metadata_update_finish(struct mddev *mddev) @@ -1043,6 +1091,141 @@ static void metadata_update_cancel(struct mddev *mddev) unlock_comm(cinfo); } +/* + * return 0 if all the bitmaps have the same sync_size + */ +int cluster_check_sync_size(struct mddev *mddev) +{ + int i, rv; + bitmap_super_t *sb; + unsigned long my_sync_size, sync_size = 0; + int node_num = mddev->bitmap_info.nodes; + int current_slot = md_cluster_ops->slot_number(mddev); + struct bitmap *bitmap = mddev->bitmap; + char str[64]; + struct dlm_lock_resource *bm_lockres; + + sb = kmap_atomic(bitmap->storage.sb_page); + my_sync_size = sb->sync_size; + kunmap_atomic(sb); + + for (i = 0; i < node_num; i++) { + if (i == current_slot) + continue; + + bitmap = get_bitmap_from_slot(mddev, i); + if (IS_ERR(bitmap)) { + pr_err("can't get bitmap from slot %d\n", i); + return -1; + } + + /* + * If we can hold the bitmap lock of one node then + * the slot is not occupied, update the sb. + */ + snprintf(str, 64, "bitmap%04d", i); + bm_lockres = lockres_init(mddev, str, NULL, 1); + if (!bm_lockres) { + pr_err("md-cluster: Cannot initialize %s\n", str); + bitmap_free(bitmap); + return -1; + } + bm_lockres->flags |= DLM_LKF_NOQUEUE; + rv = dlm_lock_sync(bm_lockres, DLM_LOCK_PW); + if (!rv) + bitmap_update_sb(bitmap); + lockres_free(bm_lockres); + + sb = kmap_atomic(bitmap->storage.sb_page); + if (sync_size == 0) + sync_size = sb->sync_size; + else if (sync_size != sb->sync_size) { + kunmap_atomic(sb); + bitmap_free(bitmap); + return -1; + } + kunmap_atomic(sb); + bitmap_free(bitmap); + } + + return (my_sync_size == sync_size) ? 0 : -1; +} + +/* + * Update the size for cluster raid is a little more complex, we perform it + * by the steps: + * 1. hold token lock and update superblock in initiator node. + * 2. send METADATA_UPDATED msg to other nodes. + * 3. The initiator node continues to check each bitmap's sync_size, if all + * bitmaps have the same value of sync_size, then we can set capacity and + * let other nodes to perform it. If one node can't update sync_size + * accordingly, we need to revert to previous value. + */ +static void update_size(struct mddev *mddev, sector_t old_dev_sectors) +{ + struct md_cluster_info *cinfo = mddev->cluster_info; + struct cluster_msg cmsg; + struct md_rdev *rdev; + int ret = 0; + int raid_slot = -1; + + md_update_sb(mddev, 1); + lock_comm(cinfo, 1); + + memset(&cmsg, 0, sizeof(cmsg)); + cmsg.type = cpu_to_le32(METADATA_UPDATED); + rdev_for_each(rdev, mddev) + if (rdev->raid_disk >= 0 && !test_bit(Faulty, &rdev->flags)) { + raid_slot = rdev->desc_nr; + break; + } + if (raid_slot >= 0) { + cmsg.raid_slot = cpu_to_le32(raid_slot); + /* + * We can only change capiticy after all the nodes can do it, + * so need to wait after other nodes already received the msg + * and handled the change + */ + ret = __sendmsg(cinfo, &cmsg); + if (ret) { + pr_err("%s:%d: failed to send METADATA_UPDATED msg\n", + __func__, __LINE__); + unlock_comm(cinfo); + return; + } + } else { + pr_err("md-cluster: No good device id found to send\n"); + unlock_comm(cinfo); + return; + } + + /* + * check the sync_size from other node's bitmap, if sync_size + * have already updated in other nodes as expected, send an + * empty metadata msg to permit the change of capacity + */ + if (cluster_check_sync_size(mddev) == 0) { + memset(&cmsg, 0, sizeof(cmsg)); + cmsg.type = cpu_to_le32(CHANGE_CAPACITY); + ret = __sendmsg(cinfo, &cmsg); + if (ret) + pr_err("%s:%d: failed to send CHANGE_CAPACITY msg\n", + __func__, __LINE__); + set_capacity(mddev->gendisk, mddev->array_sectors); + revalidate_disk(mddev->gendisk); + } else { + /* revert to previous sectors */ + ret = mddev->pers->resize(mddev, old_dev_sectors); + if (!ret) + revalidate_disk(mddev->gendisk); + ret = __sendmsg(cinfo, &cmsg); + if (ret) + pr_err("%s:%d: failed to send METADATA_UPDATED msg\n", + __func__, __LINE__); + } + unlock_comm(cinfo); +} + static int resync_start(struct mddev *mddev) { struct md_cluster_info *cinfo = mddev->cluster_info; @@ -1069,7 +1252,14 @@ static int resync_info_update(struct mddev *mddev, sector_t lo, sector_t hi) cmsg.low = cpu_to_le64(lo); cmsg.high = cpu_to_le64(hi); - return sendmsg(cinfo, &cmsg); + /* + * mddev_lock is held if resync_info_update is called from + * resync_finish (md_reap_sync_thread -> resync_finish) + */ + if (lo == 0 && hi == 0) + return sendmsg(cinfo, &cmsg, 1); + else + return sendmsg(cinfo, &cmsg, 0); } static int resync_finish(struct mddev *mddev) @@ -1119,7 +1309,7 @@ static int add_new_disk(struct mddev *mddev, struct md_rdev *rdev) cmsg.type = cpu_to_le32(NEWDISK); memcpy(cmsg.uuid, uuid, 16); cmsg.raid_slot = cpu_to_le32(rdev->desc_nr); - lock_comm(cinfo); + lock_comm(cinfo, 1); ret = __sendmsg(cinfo, &cmsg); if (ret) return ret; @@ -1179,7 +1369,7 @@ static int remove_disk(struct mddev *mddev, struct md_rdev *rdev) struct md_cluster_info *cinfo = mddev->cluster_info; cmsg.type = cpu_to_le32(REMOVE); cmsg.raid_slot = cpu_to_le32(rdev->desc_nr); - return sendmsg(cinfo, &cmsg); + return sendmsg(cinfo, &cmsg, 1); } static int lock_all_bitmaps(struct mddev *mddev) @@ -1243,7 +1433,7 @@ static int gather_bitmaps(struct md_rdev *rdev) cmsg.type = cpu_to_le32(RE_ADD); cmsg.raid_slot = cpu_to_le32(rdev->desc_nr); - err = sendmsg(cinfo, &cmsg); + err = sendmsg(cinfo, &cmsg, 1); if (err) goto out; @@ -1281,6 +1471,7 @@ static struct md_cluster_operations cluster_ops = { .gather_bitmaps = gather_bitmaps, .lock_all_bitmaps = lock_all_bitmaps, .unlock_all_bitmaps = unlock_all_bitmaps, + .update_size = update_size, }; static int __init cluster_init(void) diff --git a/drivers/md/md-cluster.h b/drivers/md/md-cluster.h index e765499ba591..274016177983 100644 --- a/drivers/md/md-cluster.h +++ b/drivers/md/md-cluster.h @@ -27,6 +27,7 @@ struct md_cluster_operations { int (*gather_bitmaps)(struct md_rdev *rdev); int (*lock_all_bitmaps)(struct mddev *mddev); void (*unlock_all_bitmaps)(struct mddev *mddev); + void (*update_size)(struct mddev *mddev, sector_t old_dev_sectors); }; #endif /* _MD_CLUSTER_H */ diff --git a/drivers/md/md.c b/drivers/md/md.c index f6ae1d67bcd0..82f798be964f 100644 --- a/ |