diff options
Diffstat (limited to 'drivers/md/raid1.c')
| -rw-r--r-- | drivers/md/raid1.c | 2513 |
1 files changed, 1710 insertions, 803 deletions
diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c index 45f8324196e..56e24c072b6 100644 --- a/drivers/md/raid1.c +++ b/drivers/md/raid1.c @@ -34,41 +34,49 @@ #include <linux/slab.h> #include <linux/delay.h> #include <linux/blkdev.h> +#include <linux/module.h> #include <linux/seq_file.h> +#include <linux/ratelimit.h> #include "md.h" #include "raid1.h" #include "bitmap.h" -#define DEBUG 0 -#if DEBUG -#define PRINTK(x...) printk(x) -#else -#define PRINTK(x...) -#endif - /* * Number of guaranteed r1bios in case of extreme VM load: */ #define NR_RAID1_BIOS 256 +/* when we get a read error on a read-only array, we redirect to another + * device without failing the first device, or trying to over-write to + * correct the read error. To keep track of bad blocks on a per-bio + * level, we store IO_BLOCKED in the appropriate 'bios' pointer + */ +#define IO_BLOCKED ((struct bio *)1) +/* When we successfully write to a known bad-block, we need to remove the + * bad-block marking which must be done from process context. So we record + * the success by setting devs[n].bio to IO_MADE_GOOD + */ +#define IO_MADE_GOOD ((struct bio *)2) + +#define BIO_SPECIAL(bio) ((unsigned long)bio <= 2) -static void unplug_slaves(mddev_t *mddev); +/* When there are this many requests queue to be written by + * the raid1 thread, we become 'congested' to provide back-pressure + * for writeback. + */ +static int max_queued_requests = 1024; -static void allow_barrier(conf_t *conf); -static void lower_barrier(conf_t *conf); +static void allow_barrier(struct r1conf *conf, sector_t start_next_window, + sector_t bi_sector); +static void lower_barrier(struct r1conf *conf); static void * r1bio_pool_alloc(gfp_t gfp_flags, void *data) { struct pool_info *pi = data; - r1bio_t *r1_bio; - int size = offsetof(r1bio_t, bios[pi->raid_disks]); + int size = offsetof(struct r1bio, bios[pi->raid_disks]); /* allocate a r1bio with room for raid_disks entries in the bios array */ - r1_bio = kzalloc(size, gfp_flags); - if (!r1_bio && pi->mddev) - unplug_slaves(pi->mddev); - - return r1_bio; + return kzalloc(size, gfp_flags); } static void r1bio_pool_free(void *r1_bio, void *data) @@ -77,24 +85,24 @@ static void r1bio_pool_free(void *r1_bio, void *data) } #define RESYNC_BLOCK_SIZE (64*1024) -//#define RESYNC_BLOCK_SIZE PAGE_SIZE +#define RESYNC_DEPTH 32 #define RESYNC_SECTORS (RESYNC_BLOCK_SIZE >> 9) #define RESYNC_PAGES ((RESYNC_BLOCK_SIZE + PAGE_SIZE-1) / PAGE_SIZE) -#define RESYNC_WINDOW (2048*1024) +#define RESYNC_WINDOW (RESYNC_BLOCK_SIZE * RESYNC_DEPTH) +#define RESYNC_WINDOW_SECTORS (RESYNC_WINDOW >> 9) +#define NEXT_NORMALIO_DISTANCE (3 * RESYNC_WINDOW_SECTORS) static void * r1buf_pool_alloc(gfp_t gfp_flags, void *data) { struct pool_info *pi = data; - struct page *page; - r1bio_t *r1_bio; + struct r1bio *r1_bio; struct bio *bio; + int need_pages; int i, j; r1_bio = r1bio_pool_alloc(gfp_flags, pi); - if (!r1_bio) { - unplug_slaves(pi->mddev); + if (!r1_bio) return NULL; - } /* * Allocate bios : 1 for reading, n-1 for writing @@ -112,19 +120,15 @@ static void * r1buf_pool_alloc(gfp_t gfp_flags, void *data) * RESYNC_PAGES for each bio. */ if (test_bit(MD_RECOVERY_REQUESTED, &pi->mddev->recovery)) - j = pi->raid_disks; + need_pages = pi->raid_disks; else - j = 1; - while(j--) { + need_pages = 1; + for (j = 0; j < need_pages; j++) { bio = r1_bio->bios[j]; - for (i = 0; i < RESYNC_PAGES; i++) { - page = alloc_page(gfp_flags); - if (unlikely(!page)) - goto out_free_pages; + bio->bi_vcnt = RESYNC_PAGES; - bio->bi_io_vec[i].bv_page = page; - bio->bi_vcnt = i+1; - } + if (bio_alloc_pages(bio, gfp_flags)) + goto out_free_pages; } /* If not user-requests, copy the page pointers to all bios */ if (!test_bit(MD_RECOVERY_REQUESTED, &pi->mddev->recovery)) { @@ -139,12 +143,15 @@ static void * r1buf_pool_alloc(gfp_t gfp_flags, void *data) return r1_bio; out_free_pages: - for (j=0 ; j < pi->raid_disks; j++) - for (i=0; i < r1_bio->bios[j]->bi_vcnt ; i++) - put_page(r1_bio->bios[j]->bi_io_vec[i].bv_page); - j = -1; + while (--j >= 0) { + struct bio_vec *bv; + + bio_for_each_segment_all(bv, r1_bio->bios[j], i) + __free_page(bv->bv_page); + } + out_free_bio: - while ( ++j < pi->raid_disks ) + while (++j < pi->raid_disks) bio_put(r1_bio->bios[j]); r1bio_pool_free(r1_bio, data); return NULL; @@ -154,7 +161,7 @@ static void r1buf_pool_free(void *__r1_bio, void *data) { struct pool_info *pi = data; int i,j; - r1bio_t *r1bio = __r1_bio; + struct r1bio *r1bio = __r1_bio; for (i = 0; i < RESYNC_PAGES; i++) for (j = pi->raid_disks; j-- ;) { @@ -169,38 +176,32 @@ static void r1buf_pool_free(void *__r1_bio, void *data) r1bio_pool_free(r1bio, data); } -static void put_all_bios(conf_t *conf, r1bio_t *r1_bio) +static void put_all_bios(struct r1conf *conf, struct r1bio *r1_bio) { int i; - for (i = 0; i < conf->raid_disks; i++) { + for (i = 0; i < conf->raid_disks * 2; i++) { struct bio **bio = r1_bio->bios + i; - if (*bio && *bio != IO_BLOCKED) + if (!BIO_SPECIAL(*bio)) bio_put(*bio); *bio = NULL; } } -static void free_r1bio(r1bio_t *r1_bio) +static void free_r1bio(struct r1bio *r1_bio) { - conf_t *conf = r1_bio->mddev->private; - - /* - * Wake up any possible resync thread that waits for the device - * to go idle. - */ - allow_barrier(conf); + struct r1conf *conf = r1_bio->mddev->private; put_all_bios(conf, r1_bio); mempool_free(r1_bio, conf->r1bio_pool); } -static void put_buf(r1bio_t *r1_bio) +static void put_buf(struct r1bio *r1_bio) { - conf_t *conf = r1_bio->mddev->private; + struct r1conf *conf = r1_bio->mddev->private; int i; - for (i=0; i<conf->raid_disks; i++) { + for (i = 0; i < conf->raid_disks * 2; i++) { struct bio *bio = r1_bio->bios[i]; if (bio->bi_end_io) rdev_dec_pending(conf->mirrors[i].rdev, r1_bio->mddev); @@ -211,11 +212,11 @@ static void put_buf(r1bio_t *r1_bio) lower_barrier(conf); } -static void reschedule_retry(r1bio_t *r1_bio) +static void reschedule_retry(struct r1bio *r1_bio) { unsigned long flags; - mddev_t *mddev = r1_bio->mddev; - conf_t *conf = mddev->private; + struct mddev *mddev = r1_bio->mddev; + struct r1conf *conf = mddev->private; spin_lock_irqsave(&conf->device_lock, flags); list_add(&r1_bio->retry_list, &conf->retry_list); @@ -231,20 +232,52 @@ static void reschedule_retry(r1bio_t *r1_bio) * operation and are ready to return a success/failure code to the buffer * cache layer. */ -static void raid_end_bio_io(r1bio_t *r1_bio) +static void call_bio_endio(struct r1bio *r1_bio) +{ + struct bio *bio = r1_bio->master_bio; + int done; + struct r1conf *conf = r1_bio->mddev->private; + sector_t start_next_window = r1_bio->start_next_window; + sector_t bi_sector = bio->bi_iter.bi_sector; + + if (bio->bi_phys_segments) { + unsigned long flags; + spin_lock_irqsave(&conf->device_lock, flags); + bio->bi_phys_segments--; + done = (bio->bi_phys_segments == 0); + spin_unlock_irqrestore(&conf->device_lock, flags); + /* + * make_request() might be waiting for + * bi_phys_segments to decrease + */ + wake_up(&conf->wait_barrier); + } else + done = 1; + + if (!test_bit(R1BIO_Uptodate, &r1_bio->state)) + clear_bit(BIO_UPTODATE, &bio->bi_flags); + if (done) { + bio_endio(bio, 0); + /* + * Wake up any possible resync thread that waits for the device + * to go idle. + */ + allow_barrier(conf, start_next_window, bi_sector); + } +} + +static void raid_end_bio_io(struct r1bio *r1_bio) { struct bio *bio = r1_bio->master_bio; /* if nobody has done the final endio yet, do it now */ if (!test_and_set_bit(R1BIO_Returned, &r1_bio->state)) { - PRINTK(KERN_DEBUG "raid1: sync end %s on sectors %llu-%llu\n", - (bio_data_dir(bio) == WRITE) ? "write" : "read", - (unsigned long long) bio->bi_sector, - (unsigned long long) bio->bi_sector + - (bio->bi_size >> 9) - 1); + pr_debug("raid1: sync end %s on sectors %llu-%llu\n", + (bio_data_dir(bio) == WRITE) ? "write" : "read", + (unsigned long long) bio->bi_iter.bi_sector, + (unsigned long long) bio_end_sector(bio) - 1); - bio_endio(bio, - test_bit(R1BIO_Uptodate, &r1_bio->state) ? 0 : -EIO); + call_bio_endio(r1_bio); } free_r1bio(r1_bio); } @@ -252,20 +285,39 @@ static void raid_end_bio_io(r1bio_t *r1_bio) /* * Update disk head position estimator based on IRQ completion info. */ -static inline void update_head_pos(int disk, r1bio_t *r1_bio) +static inline void update_head_pos(int disk, struct r1bio *r1_bio) { - conf_t *conf = r1_bio->mddev->private; + struct r1conf *conf = r1_bio->mddev->private; conf->mirrors[disk].head_position = r1_bio->sector + (r1_bio->sectors); } +/* + * Find the disk number which triggered given bio + */ +static int find_bio_disk(struct r1bio *r1_bio, struct bio *bio) +{ + int mirror; + struct r1conf *conf = r1_bio->mddev->private; + int raid_disks = conf->raid_disks; + + for (mirror = 0; mirror < raid_disks * 2; mirror++) + if (r1_bio->bios[mirror] == bio) + break; + + BUG_ON(mirror == raid_disks * 2); + update_head_pos(mirror, r1_bio); + + return mirror; +} + static void raid1_end_read_request(struct bio *bio, int error) { int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags); - r1bio_t *r1_bio = bio->bi_private; + struct r1bio *r1_bio = bio->bi_private; int mirror; - conf_t *conf = r1_bio->mddev->private; + struct r1conf *conf = r1_bio->mddev->private; mirror = r1_bio->read_disk; /* @@ -289,68 +341,85 @@ static void raid1_end_read_request(struct bio *bio, int error) spin_unlock_irqrestore(&conf->device_lock, flags); } - if (uptodate) + if (uptodate) { raid_end_bio_io(r1_bio); - else { + rdev_dec_pending(conf->mirrors[mirror].rdev, conf->mddev); + } else { /* * oops, read error: */ char b[BDEVNAME_SIZE]; - if (printk_ratelimit()) - printk(KERN_ERR "md/raid1:%s: %s: rescheduling sector %llu\n", - mdname(conf->mddev), - bdevname(conf->mirrors[mirror].rdev->bdev,b), (unsigned long long)r1_bio->sector); + printk_ratelimited( + KERN_ERR "md/raid1:%s: %s: " + "rescheduling sector %llu\n", + mdname(conf->mddev), + bdevname(conf->mirrors[mirror].rdev->bdev, + b), + (unsigned long long)r1_bio->sector); + set_bit(R1BIO_ReadError, &r1_bio->state); reschedule_retry(r1_bio); + /* don't drop the reference on read_disk yet */ } +} - rdev_dec_pending(conf->mirrors[mirror].rdev, conf->mddev); +static void close_write(struct r1bio *r1_bio) +{ + /* it really is the end of this request */ + if (test_bit(R1BIO_BehindIO, &r1_bio->state)) { + /* free extra copy of the data pages */ + int i = r1_bio->behind_page_count; + while (i--) + safe_put_page(r1_bio->behind_bvecs[i].bv_page); + kfree(r1_bio->behind_bvecs); + r1_bio->behind_bvecs = NULL; + } + /* clear the bitmap if all writes complete successfully */ + bitmap_endwrite(r1_bio->mddev->bitmap, r1_bio->sector, + r1_bio->sectors, + !test_bit(R1BIO_Degraded, &r1_bio->state), + test_bit(R1BIO_BehindIO, &r1_bio->state)); + md_write_end(r1_bio->mddev); } -static void r1_bio_write_done(r1bio_t *r1_bio, int vcnt, struct bio_vec *bv, - int behind) +static void r1_bio_write_done(struct r1bio *r1_bio) { - if (atomic_dec_and_test(&r1_bio->remaining)) - { - /* it really is the end of this request */ - if (test_bit(R1BIO_BehindIO, &r1_bio->state)) { - /* free extra copy of the data pages */ - int i = vcnt; - while (i--) - safe_put_page(bv[i].bv_page); - } - /* clear the bitmap if all writes complete successfully */ - bitmap_endwrite(r1_bio->mddev->bitmap, r1_bio->sector, - r1_bio->sectors, - !test_bit(R1BIO_Degraded, &r1_bio->state), - behind); - md_write_end(r1_bio->mddev); - raid_end_bio_io(r1_bio); + if (!atomic_dec_and_test(&r1_bio->remaining)) + return; + + if (test_bit(R1BIO_WriteError, &r1_bio->state)) + reschedule_retry(r1_bio); + else { + close_write(r1_bio); + if (test_bit(R1BIO_MadeGood, &r1_bio->state)) + reschedule_retry(r1_bio); + else + raid_end_bio_io(r1_bio); } } static void raid1_end_write_request(struct bio *bio, int error) { int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags); - r1bio_t *r1_bio = bio->bi_private; + struct r1bio *r1_bio = bio->bi_private; int mirror, behind = test_bit(R1BIO_BehindIO, &r1_bio->state); - conf_t *conf = r1_bio->mddev->private; + struct r1conf *conf = r1_bio->mddev->private; struct bio *to_put = NULL; - - for (mirror = 0; mirror < conf->raid_disks; mirror++) - if (r1_bio->bios[mirror] == bio) - break; + mirror = find_bio_disk(r1_bio, bio); /* * 'one mirror IO has finished' event handler: */ - r1_bio->bios[mirror] = NULL; - to_put = bio; if (!uptodate) { - md_error(r1_bio->mddev, conf->mirrors[mirror].rdev); - /* an I/O failed, we can't clear the bitmap */ - set_bit(R1BIO_Degraded, &r1_bio->state); - } else + set_bit(WriteErrorSeen, + &conf->mirrors[mirror].rdev->flags); + if (!test_and_set_bit(WantReplacement, + &conf->mirrors[mirror].rdev->flags)) + set_bit(MD_RECOVERY_NEEDED, & + conf->mddev->recovery); + + set_bit(R1BIO_WriteError, &r1_bio->state); + } else { /* * Set R1BIO_Uptodate in our master bio, so that we * will return a good error code for to the higher @@ -361,9 +430,31 @@ static void raid1_end_write_request(struct bio *bio, int error) * to user-side. So if something waits for IO, then it * will wait for the 'master' bio. */ - set_bit(R1BIO_Uptodate, &r1_bio->state); + sector_t first_bad; + int bad_sectors; - update_head_pos(mirror, r1_bio); + r1_bio->bios[mirror] = NULL; + to_put = bio; + /* + * Do not set R1BIO_Uptodate if the current device is + * rebuilding or Faulty. This is because we cannot use + * such device for properly reading the data back (we could + * potentially use it, if the current write would have felt + * before rdev->recovery_offset, but for simplicity we don't + * check this here. + */ + if (test_bit(In_sync, &conf->mirrors[mirror].rdev->flags) && + !test_bit(Faulty, &conf->mirrors[mirror].rdev->flags)) + set_bit(R1BIO_Uptodate, &r1_bio->state); + + /* Maybe we can clear some bad blocks. */ + if (is_badblock(conf->mirrors[mirror].rdev, + r1_bio->sector, r1_bio->sectors, + &first_bad, &bad_sectors)) { + r1_bio->bios[mirror] = IO_MADE_GOOD; + set_bit(R1BIO_MadeGood, &r1_bio->state); + } + } if (behind) { if (test_bit(WriteMostly, &conf->mirrors[mirror].rdev->flags)) @@ -381,21 +472,23 @@ static void raid1_end_write_request(struct bio *bio, int error) /* Maybe we can return now */ if (!test_and_set_bit(R1BIO_Returned, &r1_bio->state)) { struct bio *mbio = r1_bio->master_bio; - PRINTK(KERN_DEBUG "raid1: behind end write sectors %llu-%llu\n", - (unsigned long long) mbio->bi_sector, - (unsigned long long) mbio->bi_sector + - (mbio->bi_size >> 9) - 1); - bio_endio(mbio, 0); + pr_debug("raid1: behind end write sectors" + " %llu-%llu\n", + (unsigned long long) mbio->bi_iter.bi_sector, + (unsigned long long) bio_end_sector(mbio) - 1); + call_bio_endio(r1_bio); } } } - rdev_dec_pending(conf->mirrors[mirror].rdev, conf->mddev); + if (r1_bio->bios[mirror] == NULL) + rdev_dec_pending(conf->mirrors[mirror].rdev, + conf->mddev); /* * Let's see if all mirrored write operations have finished * already. */ - r1_bio_write_done(r1_bio, bio->bi_vcnt, bio->bi_io_vec, behind); + r1_bio_write_done(r1_bio); if (to_put) bio_put(to_put); @@ -416,16 +509,19 @@ static void raid1_end_write_request(struct bio *bio, int error) * * The rdev for the device selected will have nr_pending incremented. */ -static int read_balance(conf_t *conf, r1bio_t *r1_bio) +static int read_balance(struct r1conf *conf, struct r1bio *r1_bio, int *max_sectors) { const sector_t this_sector = r1_bio->sector; - const int sectors = r1_bio->sectors; - int new_disk = -1; - int start_disk; - int i; - sector_t new_distance, current_distance; - mdk_rdev_t *rdev; + int sectors; + int best_good_sectors; + int best_disk, best_dist_disk, best_pending_disk; + int has_nonrot_disk; + int disk; + sector_t best_dist; + unsigned int min_pending; + struct md_rdev *rdev; int choose_first; + int choose_next_idle; rcu_read_lock(); /* @@ -434,138 +530,232 @@ static int read_balance(conf_t *conf, r1bio_t *r1_bio) * We take the first readable disk when above the resync window. */ retry: + sectors = r1_bio->sectors; + best_disk = -1; + best_dist_disk = -1; + best_dist = MaxSector; + best_pending_disk = -1; + min_pending = UINT_MAX; + best_good_sectors = 0; + has_nonrot_disk = 0; + choose_next_idle = 0; + if (conf->mddev->recovery_cp < MaxSector && - (this_sector + sectors >= conf->next_resync)) { + (this_sector + sectors >= conf->next_resync)) choose_first = 1; - start_disk = 0; - } else { + else choose_first = 0; - start_disk = conf->last_used; - } - /* make sure the disk is operational */ - for (i = 0 ; i < conf->raid_disks ; i++) { - int disk = start_disk + i; - if (disk >= conf->raid_disks) - disk -= conf->raid_disks; + for (disk = 0 ; disk < conf->raid_disks * 2 ; disk++) { + sector_t dist; + sector_t first_bad; + int bad_sectors; + unsigned int pending; + bool nonrot; rdev = rcu_dereference(conf->mirrors[disk].rdev); if (r1_bio->bios[disk] == IO_BLOCKED || rdev == NULL - || !test_bit(In_sync, &rdev->flags)) + || test_bit(Unmerged, &rdev->flags) + || test_bit(Faulty, &rdev->flags)) continue; + if (!test_bit(In_sync, &rdev->flags) && + rdev->recovery_offset < this_sector + sectors) + continue; + if (test_bit(WriteMostly, &rdev->flags)) { + /* Don't balance among write-mostly, just + * use the first as a last resort */ + if (best_disk < 0) { + if (is_badblock(rdev, this_sector, sectors, + &first_bad, &bad_sectors)) { + if (first_bad < this_sector) + /* Cannot use this */ + continue; + best_good_sectors = first_bad - this_sector; + } else + best_good_sectors = sectors; + best_disk = disk; + } + continue; + } + /* This is a reasonable device to use. It might + * even be best. + */ + if (is_badblock(rdev, this_sector, sectors, + &first_bad, &bad_sectors)) { + if (best_dist < MaxSector) + /* already have a better device */ + continue; + if (first_bad <= this_sector) { + /* cannot read here. If this is the 'primary' + * device, then we must not read beyond + * bad_sectors from another device.. + */ + bad_sectors -= (this_sector - first_bad); + if (choose_first && sectors > bad_sectors) + sectors = bad_sectors; + if (best_good_sectors > sectors) + best_good_sectors = sectors; - new_disk = disk; - if (!test_bit(WriteMostly, &rdev->flags)) + } else { + sector_t good_sectors = first_bad - this_sector; + if (good_sectors > best_good_sectors) { + best_good_sectors = good_sectors; + best_disk = disk; + } + if (choose_first) + break; + } + continue; + } else + best_good_sectors = sectors; + + nonrot = blk_queue_nonrot(bdev_get_queue(rdev->bdev)); + has_nonrot_disk |= nonrot; + pending = atomic_read(&rdev->nr_pending); + dist = abs(this_sector - conf->mirrors[disk].head_position); + if (choose_first) { + best_disk = disk; break; - } - - if (new_disk < 0 || choose_first) - goto rb_out; - - /* - * Don't change to another disk for sequential reads: - */ - if (conf->next_seq_sect == this_sector) - goto rb_out; - if (this_sector == conf->mirrors[new_disk].head_position) - goto rb_out; - - current_distance = abs(this_sector - - conf->mirrors[new_disk].head_position); + } + /* Don't change to another disk for sequential reads */ + if (conf->mirrors[disk].next_seq_sect == this_sector + || dist == 0) { + int opt_iosize = bdev_io_opt(rdev->bdev) >> 9; + struct raid1_info *mirror = &conf->mirrors[disk]; - /* look for a better disk - i.e. head is closer */ - start_disk = new_disk; - for (i = 1; i < conf->raid_disks; i++) { - int disk = start_disk + 1; - if (disk >= conf->raid_disks) - disk -= conf->raid_disks; + best_disk = disk; + /* + * If buffered sequential IO size exceeds optimal + * iosize, check if there is idle disk. If yes, choose + * the idle disk. read_balance could already choose an + * idle disk before noticing it's a sequential IO in + * this disk. This doesn't matter because this disk + * will idle, next time it will be utilized after the + * first disk has IO size exceeds optimal iosize. In + * this way, iosize of the first disk will be optimal + * iosize at least. iosize of the second disk might be + * small, but not a big deal since when the second disk + * starts IO, the first disk is likely still busy. + */ + if (nonrot && opt_iosize > 0 && + mirror->seq_start != MaxSector && + mirror->next_seq_sect > opt_iosize && + mirror->next_seq_sect - opt_iosize >= + mirror->seq_start) { + choose_next_idle = 1; + continue; + } + break; + } + /* If device is idle, use it */ + if (pending == 0) { + best_disk = disk; + break; + } - rdev = rcu_dereference(conf->mirrors[disk].rdev); - if (r1_bio->bios[disk] == IO_BLOCKED - || rdev == NULL - || !test_bit(In_sync, &rdev->flags) - || test_bit(WriteMostly, &rdev->flags)) + if (choose_next_idle) continue; - if (!atomic_read(&rdev->nr_pending)) { - new_disk = disk; - break; + if (min_pending > pending) { + min_pending = pending; + best_pending_disk = disk; } - new_distance = abs(this_sector - conf->mirrors[disk].head_position); - if (new_distance < current_distance) { - current_distance = new_distance; - new_disk = disk; + + if (dist < best_dist) { + best_dist = dist; + best_dist_disk = disk; } } - rb_out: - if (new_disk >= 0) { - rdev = rcu_dereference(conf->mirrors[new_disk].rdev); + /* + * If all disks are rotational, choose the closest disk. If any disk is + * non-rotational, choose the disk with less pending request even the + * disk is rotational, which might/might not be optimal for raids with + * mixed ratation/non-rotational disks depending on workload. + */ + if (best_disk == -1) { + if (has_nonrot_disk) + best_disk = best_pending_disk; + else + best_disk = best_dist_disk; + } + + if (best_disk >= 0) { + rdev = rcu_dereference(conf->mirrors[best_disk].rdev); if (!rdev) goto retry; atomic_inc(&rdev->nr_pending); - if (!test_bit(In_sync, &rdev->flags)) { + if (test_bit(Faulty, &rdev->flags)) { /* cannot risk returning a device that failed * before we inc'ed nr_pending */ rdev_dec_pending(rdev, conf->mddev); goto retry; } - conf->next_seq_sect = this_sector + sectors; - conf->last_used = new_disk; + sectors = best_good_sectors; + + if (conf->mirrors[best_disk].next_seq_sect != this_sector) + conf->mirrors[best_disk].seq_start = this_sector; + + conf->mirrors[best_disk].next_seq_sect = this_sector + sectors; } rcu_read_unlock(); + *max_sectors = sectors; - return new_disk; + return best_disk; } -static void unplug_slaves(mddev_t *mddev) +static int raid1_mergeable_bvec(struct request_queue *q, + struct bvec_merge_data *bvm, + struct bio_vec *biovec) { - conf_t *conf = mddev->private; - int i; - - rcu_read_lock(); - for (i=0; i<mddev->raid_disks; i++) { - mdk_rdev_t *rdev = rcu_dereference(conf->mirrors[i].rdev); - if (rdev && !test_bit(Faulty, &rdev->flags) && atomic_read(&rdev->nr_pending)) { - struct request_queue *r_queue = bdev_get_queue(rdev->bdev); - - atomic_inc(&rdev->nr_pending); - rcu_read_unlock(); - - blk_unplug(r_queue); - - rdev_dec_pending(rdev, mddev); - rcu_read_lock(); + struct mddev *mddev = q->queuedata; + struct r1conf *conf = mddev->private; + sector_t sector = bvm->bi_sector + get_start_sect(bvm->bi_bdev); + int max = biovec->bv_len; + + if (mddev->merge_check_needed) { + int disk; + rcu_read_lock(); + for (disk = 0; disk < conf->raid_disks * 2; disk++) { + struct md_rdev *rdev = rcu_dereference( + conf->mirrors[disk].rdev); + if (rdev && !test_bit(Faulty, &rdev->flags)) { + struct request_queue *q = + bdev_get_queue(rdev->bdev); + if (q->merge_bvec_fn) { + bvm->bi_sector = sector + + rdev->data_offset; + bvm->bi_bdev = rdev->bdev; + max = min(max, q->merge_bvec_fn( + q, bvm, biovec)); + } + } } + rcu_read_unlock(); } - rcu_read_unlock(); -} + return max; -static void raid1_unplug(struct request_queue *q) -{ - mddev_t *mddev = q->queuedata; - - unplug_slaves(mddev); - md_wakeup_thread(mddev->thread); } -static int raid1_congested(void *data, int bits) +int md_raid1_congested(struct mddev *mddev, int bits) { - mddev_t *mddev = data; - conf_t *conf = mddev->private; + struct r1conf *conf = mddev->private; int i, ret = 0; - if (mddev_congested(mddev, bits)) + if ((bits & (1 << BDI_async_congested)) && + conf->pending_count >= max_queued_requests) return 1; rcu_read_lock(); - for (i = 0; i < mddev->raid_disks; i++) { - mdk_rdev_t *rdev = rcu_dereference(conf->mirrors[i].rdev); + for (i = 0; i < conf->raid_disks * 2; i++) { + struct md_rdev *rdev = rcu_dereference(conf->mirrors[i].rdev); if (rdev && !test_bit(Faulty, &rdev->flags)) { struct request_queue *q = bdev_get_queue(rdev->bdev); + BUG_ON(!q); + /* Note the '|| 1' - when read_balance prefers * non-congested targets, it can be removed */ @@ -578,37 +768,46 @@ static int raid1_congested(void *data, int bits) rcu_read_unlock(); return ret; } +EXPORT_SYMBOL_GPL(md_raid1_congested); + +static int raid1_congested(void *data, int bits) +{ + struct mddev *mddev = data; + return mddev_congested(mddev, bits) || + md_raid1_congested(mddev, bits); +} -static int flush_pending_writes(conf_t *conf) +static void flush_pending_writes(struct r1conf *conf) { /* Any writes that have been queued but are awaiting * bitmap updates get flushed here. - * We return 1 if any requests were actually submitted. */ - int rv = 0; - spin_lock_irq(&conf->device_lock); if (conf->pending_bio_list.head) { struct bio *bio; bio = bio_list_get(&conf->pending_bio_list); - blk_remove_plug(conf->mddev->queue); + conf->pending_count = 0; spin_unlock_irq(&conf->device_lock); /* flush any pending bitmap writes to * disk before proceeding w/ I/O */ bitmap_unplug(conf->mddev->bitmap); + wake_up(&conf->wait_barrier); while (bio) { /* submit pending writes */ struct bio *next = bio->bi_next; bio->bi_next = NULL; - generic_make_request(bio); + if (unlikely((bio->bi_rw & REQ_DISCARD) && + !blk_queue_discard(bdev_get_queue(bio->bi_bdev)))) + /* Just ignore it */ + bio_endio(bio, 0); + else + generic_make_request(bio); bio = next; } - rv = 1; } else spin_unlock_irq(&conf->device_lock); - return rv; } /* Barriers.... @@ -632,30 +831,36 @@ static int flush_pending_writes(conf_t *conf) * there is no normal IO happeing. It must arrange to call * lower_barrier when the particular background IO completes. */ -#define RESYNC_DEPTH 32 - -static void raise_barrier(conf_t *conf) +static void raise_barrier(struct r1conf *conf) { spin_lock_irq(&conf->resync_lock); /* Wait until no block IO is waiting */ wait_event_lock_irq(conf->wait_barrier, !conf->nr_waiting, - conf->resync_lock, - raid1_unplug(conf->mddev->queue)); + conf->resync_lock); /* block any new IO from starting */ conf->barrier++; - /* Now wait for all pending IO to complete */ + /* For these conditions we must wait: + * A: while the array is in frozen state + * B: while barrier >= RESYNC_DEPTH, meaning resync reach + * the max count which allowed. + * C: next_resync + RESYNC_SECTORS > start_next_window, meaning + * next resync will reach to the window which normal bios are + * handling. + */ wait_event_lock_irq(conf->wait_barrier, - !conf->nr_pending && conf->barrier < RESYNC_DEPTH, - conf->resync_lock, - raid1_unplug(conf->mddev->queue)); + !conf->array_frozen && + conf->barrier < RESYNC_DEPTH && + (conf->start_next_window >= + conf->next_resync + RESYNC_SECTORS), + conf->resync_lock); spin_unlock_irq(&conf->resync_lock); } -static void lower_barrier(conf_t *conf) +static void lower_barrier(struct r1conf *conf) { unsigned long flags; BUG_ON(conf->barrier <= 0); @@ -665,112 +870,240 @@ static void lower_barrier(conf_t *conf) wake_up(&conf->wait_barrier); } -static void wait_barrier(conf_t *conf) +static bool need_to_wait_for_sync(struct r1conf *conf, struct bio *bio) +{ + bool wait = false; + + if (conf->array_frozen || !bio) + wait = true; + else if (conf->barrier && bio_data_dir(bio) == WRITE) { + if (conf->next_resync < RESYNC_WINDOW_SECTORS) + wait = true; + else if ((conf->next_resync - RESYNC_WINDOW_SECTORS + >= bio_end_sector(bio)) || + (conf->next_resync + NEXT_NORMALIO_DISTANCE + <= bio->bi_iter.bi_sector)) + wait = false; + else + wait = true; + } + + return wait; +} + +static sector_t wait_barrier(struct r1conf *conf, struct bio *bio) { + sector_t sector = 0; + spin_lock_irq(&conf->resync_lock); - if (conf->barrier) { + if (need_to_wait_for_sync(conf, bio)) { conf->nr_waiting++; - wait_event_lock_irq(conf->wait_barrier, !conf->barrier, - conf->resync_lock, - raid1_unplug(conf->mddev->queue)); + /* Wait for the barrier to drop. + * However if there are already pending + * requests (preventing the barrier from + * rising completely), and the + * pre-process bio queue isn't empty, + * then don't wait, as we need to empty + * that queue to get the nr_pending + * count down. + */ + wait_event_lock_irq(conf->wait_barrier, + !conf->array_frozen && + (!conf->barrier || + ((conf->start_next_window < + conf->next_resync + RESYNC_SECTORS) && + current->bio_list && + !bio_list_empty(current->bio_list))), + conf->resync_lock); conf->nr_waiting--; } + + if (bio && bio_data_dir(bio) == WRITE) { + if (conf->next_resync + NEXT_NORMALIO_DISTANCE + <= bio->bi_iter.bi_sector) { + if (conf->start_next_window == MaxSector) + conf->start_next_window = + conf->next_resync + + NEXT_NORMALIO_DISTANCE; + + if ((conf->start_next_window + NEXT_NORMALIO_DISTANCE) + <= bio->bi_iter.bi_sector) + conf->next_window_requests++; + else + conf->current_window_requests++; + sector = conf->start_next_window; + } + } + conf->nr_pending++; spin_unlock_irq(&conf->resync_lock); + return sector; } -static void allow_barrier(conf_t *conf) +static void allow_barrier(struct r1conf *conf, sector_t start_next_window, + sector_t bi_sector) { unsigned long flags; + spin_lock_irqsave(&conf->resync_lock, flags); conf->nr_pending--; + if (start_next_window) { + if (start_next_window == conf->start_next_window) { + if (conf->start_next_window + NEXT_NORMALIO_DISTANCE + <= bi_sector) + conf->next_window_requests--; + else + conf->current_window_requests--; + } else + conf->current_window_requests--; + + if (!conf->current_window_requests) { + if (conf->next_window_requests) { + conf->current_window_requests = + conf->next_window_requests; + conf->next_window_requests = 0; + conf->start_next_window += + NEXT_NORMALIO_DISTANCE; + } else + conf->start_next_window = MaxSector; + } + } spin_unlock_irqrestore(&conf->resync_lock, flags); wake_up(&conf->wait_barrier); } -static void freeze_array(conf_t *conf) +static void freeze_array(struct r1conf *conf, int extra) { /* stop syncio and normal IO and wait for everything to * go quite. - * We increment barrier and nr_waiting, and then - * wait until nr_pending match nr_queued+1 + * We wait until nr_pending match nr_queued+extra * This is called in the context of one normal IO request * that has failed. Thus any sync request that might be pending * will be blocked by nr_pending, and we need to wait for * pending IO requests to complete or be queued for re-try. - * Thus the number queued (nr_queued) plus this request (1) + * Thus the number queued (nr_queued) plus this request (extra) * must match the number of pending IOs (nr_pending) before * we continue. */ spin_lock_irq(&conf->resync_lock); - conf->barrier++; - conf->nr_waiting++; - wait_event_lock_irq(conf->wait_barrier, - conf->nr_pending == conf->nr_queued+1, - conf->resync_lock, - ({ flush_pending_writes(conf); - raid1_unplug(conf->mddev->queue); })); + conf->array_frozen = 1; + wait_event_lock_irq_cmd(conf->wait_barrier, + conf->nr_pending == conf->nr_queued+extra, + conf->resync_lock, + flush_pending_writes(conf)); spin_unlock_irq(&conf->resync_lock); } -static void unfreeze_array(conf_t *conf) +static void unfreeze_array(struct r1conf *conf) { /* reverse the effect of the freeze */ spin_lock_irq(&conf->resync_lock); - conf->barrier--; - conf->nr_waiting--; + conf->array_frozen = 0; wake_up(&conf->wait_barrier); spin_unlock_irq(&conf->resync_lock); } /* duplicate the data pages for behind I/O - * We return a list of bio_vec rather than just page pointers - * as it makes freeing easier */ -static struct bio_vec *alloc_behind_pages(struct bio *bio) +static void alloc_behind_pages(struct bio *bio, struct r1bio *r1_bio) { int i; struct bio_vec *bvec; - struct bio_vec *pages = kzalloc(bio->bi_vcnt * sizeof(struct bio_vec), + struct bio_vec *bvecs = kzalloc(bio->bi_vcnt * sizeof(struct bio_vec), GFP_NOIO); - if (unlikely(!pages)) - goto do_sync_io; + if (unlikely(!bvecs)) + return; - bio_for_each_segment(bvec, bio, i) { - pages[i].bv_page = alloc_page(GFP_NOIO); - if (unlikely(!pages[i].bv_page)) + bio_for_each_segment_all(bvec, bio, i) { + bvecs[i] = *bvec; + bvecs[i].bv_page = alloc_page(GFP_NOIO); + if (unlikely(!bvecs[i].bv_page)) goto do_sync_io; - memcpy(kmap(pages[i].bv_page) + bvec->bv_offset, - kmap(bvec->bv_page) + bvec->bv_offset, bvec->bv_len); - kunmap(pages[i].bv_page); + memcpy(kmap(bvecs[i].bv_page) + bvec->bv_offset, + kmap(bvec->bv_page) + bvec->bv_offset, bvec->bv_len); + kunmap(bvecs[i].bv_page); kunmap(bvec->bv_page); } - - return pages; + r1_bio->behind_bvecs = bvecs; + r1_bio->behind_page_count = bio->bi_vcnt; + set_bit(R1BIO_BehindIO, &r1_bio->state); + return; do_sync_io: - if (pages) - for (i = 0; i < bio->bi_vcnt && pages[i].bv_page; i++) - put_page(pages[i].bv_page); - kfree(pages); - PRINTK("%dB behind alloc failed, doing sync I/O\n", bio->bi_size); - return NULL; + for (i = 0; i < bio->bi_vcnt; i++) + if (bvecs[i].bv_page) + put_page(bvecs[i].bv_page); + kfree(bvecs); + pr_debug("%dB behind alloc failed, doing sync I/O\n", + bio->bi_iter.bi_size); } -static int make_request(mddev_t *mddev, struct bio * bio) +struct raid1_plug_cb { + struct blk_plug_cb cb; + struct bio_list pending; + int pending_cnt; +}; + +static void raid1_unplug(struct blk_plug_cb *cb, bool from_schedule) { - conf_t *conf = mddev->private; - mirror_info_t *mirror; - r1bio_t *r1_bio; + struct raid1_plug_cb *plug = container_of(cb, struct raid1_plug_cb, + cb); + struct mddev *mddev = plug->cb.data; + struct r1conf *conf = mddev->private; + struct bio *bio; + + if (from_schedule || current->bio_list) { + spin_lock_irq(&conf->device_lock); + bio_list_merge(&conf->pending_bio_list, &plug->pending); + conf->pending_count += plug->pending_cnt; + spin_unlock_irq(&conf->device_lock); + wake_up(&conf->wait_barrier); + md_wakeup_thread(mddev->thread); + kfree(plug); + return; + } + + /* we aren't scheduling, so we can do the write-out directly. */ + bio = bio_list_get(&plug->pending); + bitmap_unplug(mddev->bitmap); + wake_up(&conf->wait_barrier); + + while (bio) { /* submit pending writes */ + struct bio *next = bio->bi_next; + bio->bi_next = NULL; + if (unlikely((bio->bi_rw & REQ_DISCARD) && + !blk_queue_discard(bdev_get_queue(bio->bi_bdev)))) + /* Just ignore it */ + bio_endio(bio, 0); + else + generic_make_request(bio); + bio = next; + } + kfree(plug); +} + +static void make_request(struct mddev *mddev, struct bio * bio) +{ + struct r1conf *conf = mddev->private; + struct raid1_info *mirror; + struct r1bio *r1_bio; struct bio *read_bio; - int i, targets = 0, disks; + int i, disks; struct bitmap *bitmap; unsigned long flags; - struct bio_vec *behind_pages = NULL; const int rw = bio_data_dir(bio); const unsigned long do_sync = (bio->bi_rw & REQ_SYNC); const unsigned long do_flush_fua = (bio->bi_rw & (REQ_FLUSH | REQ_FUA)); - mdk_rdev_t *blocked_rdev; + const unsigned long do_discard = (bio->bi_rw + & (REQ_DISCARD | REQ_SECURE)); + const unsigned long do_same = (bio->bi_rw & REQ_WRITE_SAME); + struct md_rdev *blocked_rdev; + struct blk_plug_cb *cb; + struct raid1_plug_cb *plug = NULL; + int first_clone; + int sectors_handled; + int max_sectors; + sector_t start_next_window; /* * Register the new request and wait if the reconstruction @@ -781,8 +1114,8 @@ static int make_request(mddev_t *mddev, struct bio * bio) md_write_start(mddev, bio); /* wait on superblock update early */ if (bio_data_dir(bio) == WRITE && - bio->bi_sector + bio->bi_size/512 > mddev->suspend_lo && - bio->bi_sector < mddev->suspend_hi) { + bio_end_sector(bio) > mddev->suspend_lo && + bio->bi_iter.bi_sector < mddev->suspend_hi) { /* As the suspend_* range is controlled by * userspace, we want an interruptible * wait. @@ -792,15 +1125,15 @@ static int make_request(mddev_t *mddev, struct bio * bio) flush_signals(current); prepare_to_wait(&conf->wait_barrier, &w, TASK_INTERRUPTIBLE); - if (bio->bi_sector + bio->bi_size/512 <= mddev->suspend_lo || - bio->bi_sector >= mddev->suspend_hi) + if (bio_end_sector(bio) <= mddev->suspend_lo || + bio->bi_iter.bi_sector >= mddev->suspend_hi) break; schedule(); } finish_wait(&conf->wait_barrier, &w); } - wait_barrier(conf); + start_next_window = wait_barrier(conf, bio); bitmap = mddev->bitmap; @@ -812,21 +1145,34 @@ static int make_request(mddev_t *mddev, struct bio * bio) r1_bio = mempool_alloc(conf->r1bio_pool, GFP_NOIO); r1_bio->master_bio = bio; - r1_bio->sectors = bio->bi_size >> 9; + r1_bio->sectors = bio_sectors(bio); r1_bio->state = 0; r1_bio->mddev = mddev; - r1_bio->sector = bio->bi_sector; + r1_bio->sector = bio->bi_iter.bi_sector; + + /* We might need to issue multiple reads to different + * devices if there are bad blocks around, so we keep + * track of the number of reads in bio->bi_phys_segments. + * If this is 0, there is only one r1_bio and no locking + * will be needed when requests complete. If it is + * non-zero, then it is the number of not-completed requests. + */ + bio->bi_phys_segments = 0; + clear_bit(BIO_SEG_VALID, &bio->bi_flags); if (rw == READ) { /* * read balancing logic: */ - int rdisk = read_balance(conf, r1_bio); + int rdisk; + +read_again: + rdisk = read_balance(conf, r1_bio, &max_sectors); if (rdisk < 0) { /* couldn't find anywhere to read from */ raid_end_bio_io(r1_bio); - return 0; + return; } mirror = conf->mirrors + rdisk; @@ -842,148 +1188,282 @@ static int make_request(mddev_t *mddev, struct bio * bio) r1_bio->read_disk = rdisk; read_bio = bio_clone_mddev(bio, GFP_NOIO, mddev); + bio_trim(read_bio, r1_bio->sector - bio->bi_iter.bi_sector, + max_sectors); r1_bio->bios[rdisk] = read_bio; - read_bio->bi_sector = r1_bio->sector + mirror->rdev->data_offset; + read_bio->bi_iter.bi_sector = r1_bio->sector + + mirror->rdev->data_offset; read_bio->bi_bdev = mirror->rdev->bdev; read_bio->bi_end_io = raid1_end_read_request; read_bio->bi_rw = READ | do_sync; read_bio->bi_private = r1_bio; - generic_make_request(read_bio); - return 0; + if (max_sectors < r1_bio->sectors) { + /* could not read all from this device, so we will + * need another r1_bio. + */ + + sectors_handled = (r1_bio->sector + max_sectors + - bio->bi_iter.bi_sector); + r1_bio->sectors = max_sectors; + spin_lock_irq(&conf->device_lock); + if (bio->bi_phys_segments == 0) + bio->bi_phys_segments = 2; + else + bio->bi_phys_segments++; + spin_unlock_irq(&conf->device_lock); + /* Cannot call generic_make_request directly + * as that will be queued in __make_request + * and subsequent mempool_alloc might block waiting + * for it. So hand bio over to raid1d. + */ + reschedule_retry(r1_bio); + + r1_bio = mempool_alloc(conf->r1bio_pool, GFP_NOIO); + + r1_bio->master_bio = bio; + r1_bio->sectors = bio_sectors(bio) - sectors_handled; + r1_bio->state = 0; + r1_bio->mddev = mddev; + r1_bio->sector = bio->bi_iter.bi_sector + + sectors_handled; + goto read_again; + } else + generic_make_request(read_bio); + return; } /* * WRITE: */ - /* first select target devices under spinlock and + if (conf->pending_count >= max_queued_requests) { + md_wakeup_thread(mddev->thread); + wait_event(conf->wait_barrier, + conf->pending_count < max_queued_requests); + } + /* first select target devices under rcu_lock and * inc refcount on their rdev. Record them by setting * bios[x] to bio + * If there are known/acknowledged bad blocks on any device on + * which we have seen a write error, we want to avoid writing those + * blocks. + * This potentially requires several writes to write around + * the bad blocks. Each set of writes gets it's own r1bio + * with a set of bios attached. */ - disks = conf->raid_disks; + + disks = conf->raid_disks * 2; retry_write: + r1_bio->start_next_window = start_next_window; blocked_rdev = NULL; rcu_read_lock(); + max_sectors = r1_bio->sectors; for (i = 0; i < disks; i++) { - mdk_rdev_t *rdev = rcu_dereference(conf->mirrors[i].rdev); + struct md_rdev *rdev = rcu_dereference(conf->mirrors[i].rdev); if (rdev && unlikely(test_bit(Blocked, &rdev->flags))) { atomic_inc(&rdev->nr_pending); blocked_rdev = rdev; break; } - if (rdev && !test_bit(Faulty, &rdev->flags)) { - atomic_inc(&rdev->nr_pending); - if (test_bit(Faulty, &rdev->flags)) { + r1_bio->bios[i] = NULL; + if (!rdev || test_bit(Faulty, &rdev->flags) + || test_bit(Unmerged, &rdev->flags)) { + if (i < conf->raid_disks) + set_bit(R1BIO_Degraded, &r1_bio->state); + continue; + } + + atomic_inc(&rdev->nr_pending); + if (test_bit(WriteErrorSeen, &rdev->flags)) { + sector_t first_bad; + int bad_sectors; + int is_bad; + + is_bad = is_badblock(rdev, r1_bio->sector, + max_sectors, + &first_bad, &bad_sectors); + if (is_bad < 0) { + /* mustn't write here until the bad block is + * acknowledged*/ + set_bit(BlockedBadBlocks, &rdev->flags); + blocked_rdev = rdev; + break; + } + if (is_bad && first_bad <= r1_bio->sector) { + /* Cannot write here at all */ + bad_sectors -= (r1_bio->sector - first_bad); + if (bad_sectors < max_sectors) + /* mustn't write more than bad_sectors + * to other devices yet + */ + max_sectors = bad_sectors; rdev_dec_pending(rdev, mddev); - r1_bio->bios[i] = NULL; - } else { - r1_bio->bios[i] = bio; - targets++; + /* We don't set R1BIO_Degraded as that + * only applies if the disk is + * missing, so it might be re-added, + * and we want to know to recover this + * chunk. + * In this case the device is here, + * and the fact that this chunk is not + * in-sync is recorded in the bad + * block log + */ + continue; } - } else - r1_bio->bios[i] = NULL; + if (is_bad) { + int good_sectors = first_bad - r1_bio->sector; + if (good_sectors < max_sectors) + max_sectors = good_sectors; + } + } + r1_bio->bios[i] = bio; } rcu_read_unlock(); if (unlikely(blocked_rdev)) { /* Wait for this device to become unblocked */ int j; + sector_t old = start_next_window; for (j = 0; j < i; j++) if (r1_bio->bios[j]) rdev_dec_pending(conf->mirrors[j].rdev, mddev); - - allow_barrier(conf); + r1_bio->state = 0; + allow_barrier(conf, start_next_window, bio->bi_iter.bi_sector); md_wait_for_blocked_rdev(blocked_rdev, mddev); - wait_barrier(conf); + start_next_window = wait_barrier(conf, bio); + /* + * We must make sure the multi r1bios of bio have + * the same value of bi_phys_segments + */ + if (bio->bi_phys_segments && old && + old != start_next_window) + /* Wait for the former r1bio(s) to complete */ + wait_event(conf->wait_barrier, + bio->bi_phys_segments == 1); goto retry_write; } - BUG_ON(targets == 0); /* we never fail the last device */ - - if (targets < conf->raid_disks) { - /* array is degraded, we will not clear the bitmap - * on I/O completion (see raid1_end_write_request) */ - set_bit(R1BIO_Degraded, &r1_bio->state); + if (max_sectors < r1_bio->sectors) { + /* We are splitting this write into multiple parts, so + * we need to prepare for allocating another r1_bio. + */ + r1_bio->sectors = max_sectors; + spin_lock_irq(&conf->device_lock); + if (bio->bi_phys_segments == 0) + bio->bi_phys_segments = 2; + else + bio->bi_phys_segments++; + spin_unlock_irq(&conf->device_lock); } - - /* do behind I/O ? - * Not if there are too many, or cannot allocate memory, - * or a reader on WriteMostly is waiting for behind writes - * to flush */ - if (bitmap && - (atomic_read(&bitmap->behind_writes) - < mddev->bitmap_info.max_write_behind) && - !waitqueue_active(&bitmap->behind_wait) && - (behind_pages = alloc_behind_pages(bio)) != NULL) - set_bit(R1BIO_BehindIO, &r1_bio->state); + sectors_handled = r1_bio->sector + max_sectors - bio->bi_iter.bi_sector; atomic_set(&r1_bio->remaining, 1); atomic_set(&r1_bio->behind_remaining, 0); - bitmap_startwrite(bitmap, bio->bi_sector, r1_bio->sectors, - test_bit(R1BIO_BehindIO, &r1_bio->state)); + first_clone = 1; for (i = 0; i < disks; i++) { struct bio *mbio; if (!r1_bio->bios[i]) continue; mbio = bio_clone_mddev(bio, GFP_NOIO, mddev); - r1_bio->bios[i] = mbio; - - mbio->bi_sector = r1_bio->sector + conf->mirrors[i].rdev->data_offset; - mbio->bi_bdev = conf->mirrors[i].rdev->bdev; - mbio->bi_end_io = raid1_end_write_request; - mbio->bi_rw = WRITE | do_flush_fua | do_sync; - mbio->bi_private = r1_bio; - - if (behind_pages) { + bio_trim(mbio, r1_bio->sector - bio->bi_iter.bi_sector, max_sectors); + + if (first_clone) { + /* do behind I/O ? + * Not if there are too many, or cannot + * allocate memory, or a reader on WriteMostly + * is waiting for behind writes to flush */ + if (bitmap && + (atomic_read(&bitmap->behind_writes) + < mddev->bitmap_info.max_write_behind) && + !waitqueue_active(&bitmap->behind_wait)) + alloc_behind_pages(mbio, r1_bio); + + bitmap_startwrite(bitmap, r1_bio->sector, + r1_bio->sectors, + test_bit(R1BIO_BehindIO, + &r1_bio->state)); + first_clone = 0; + } + if (r1_bio->behind_bvecs) { struct bio_vec *bvec; int j; - /* Yes, I really want the '__' version so that - * we clear any unused pointer in the io_vec, rather - * than leave them unchanged. This is important - * because when we come to free the pages, we won't - * know the original bi_idx, so we just free - * them all + /* + * We trimmed the bio, so _all is legit */ - __bio_for_each_segment(bvec, mbio, j, 0) - bvec->bv_page = behind_pages[j].bv_page; + bio_for_each_segment_all(bvec, mbio, j) + bvec->bv_page = r1_bio->behind_bvecs[j].bv_page; if (test_bit(WriteMostly, &conf->mirrors[i].rdev->flags)) atomic_inc(&r1_bio->behind_remaining); } + r1_bio->bios[i] = mbio; + + mbio->bi_iter.bi_sector = (r1_bio->sector + + conf->mirrors[i].rdev->data_offset); + mbio->bi_bdev = conf->mirrors[i].rdev->bdev; + mbio->bi_end_io = raid1_end_write_request; + mbio->bi_rw = + WRITE | do_flush_fua | do_sync | do_discard | do_same; + mbio->bi_private = r1_bio; + atomic_inc(&r1_bio->remaining); + + cb = blk_check_plugged(raid1_unplug, mddev, sizeof(*plug)); + if (cb) + plug = container_of(cb, struct raid1_plug_cb, cb); + else + plug = NULL; spin_lock_irqsave(&conf->device_lock, flags); - bio_list_add(&conf->pending_bio_list, mbio); - blk_plug_device(mddev->queue); + if (plug) { + bio_list_add(&plug->pending, mbio); + plug->pending_cnt++; + } else { + bio_list_add(&conf->pending_bio_list, mbio); + conf->pending_count++; + } spin_unlock_irqrestore(&conf->device_lock, flags); + if (!plug) + md_wakeup_thread(mddev->thread); + } + /* Mustn't call r1_bio_write_done before this next test, + * as it could result in the bio being freed. + */ + if (sectors_handled < bio_sectors(bio)) { + r1_bio_write_done(r1_bio); + /* We need another r1_bio. It has already been counted + * in bio->bi_phys_segments + */ + r1_bio = mempool_alloc(conf->r1bio_pool, GFP_NOIO); + r1_bio->master_bio = bio; + r1_bio->sectors = bio_sectors(bio) - sectors_handled; + r1_bio->state = 0; + r1_bio->mddev = mddev; + r1_bio->sector = bio->bi_iter.bi_sector + sectors_handled; + goto retry_write; } - r1_bio_write_done(r1_bio, bio->bi_vcnt, behind_pages, behind_pages != NULL); - kfree(behind_pages); /* the behind pages are attached to the bios now */ + + r1_bio_write_done(r1_bio); /* In case raid1d snuck in to freeze_array */ wake_up(&conf->wait_barrier); - - if (do_sync) - md_wakeup_thread(mddev->thread); - - return 0; } -static void status(struct seq_file *seq, mddev_t *mddev) +static void status(struct seq_file *seq, struct mddev *mddev) { - conf_t *conf = mddev->private; + struct r1conf *conf = mddev->private; int i; seq_printf(seq, " [%d/%d] [", conf->raid_disks, conf->raid_disks - mddev->degraded); rcu_read_lock(); for (i = 0; i < conf->raid_disks; i++) { - mdk_rdev_t *rdev = rcu_dereference(conf->mirrors[i].rdev); + struct md_rdev *rdev = rcu_dereference(conf->mirrors[i].rdev); seq_printf(seq, "%s", rdev && test_bit(In_sync, &rdev->flags) ? "U" : "_"); } @@ -992,10 +1472,10 @@ static void status(struct seq_file *seq, mddev_t *mddev) } -static void error(mddev_t *mddev, mdk_rdev_t *rdev) +static void error(struct mddev *mddev, struct md_rdev *rdev) { char b[BDEVNAME_SIZE]; - conf_t *conf = mddev->private; + struct r1conf *conf = mddev->private; /* * If it is not operational, then we have already marked it as dead @@ -1011,9 +1491,10 @@ static void error(mddev_t *mddev, mdk_rdev_t *rdev) * However don't try a recovery from this drive as * it is very likely to fail. */ - mddev->recovery_disabled = 1; + conf->recovery_disabled = mddev->recovery_disabled; return; } + set_bit(Blocked, &rdev->flags); if (test_and_clear_bit(In_sync, &rdev->flags)) { unsigned long flags; spin_lock_irqsave(&conf->device_lock, flags); @@ -1027,13 +1508,14 @@ static void error(mddev_t *mddev, mdk_rdev_t *rdev) } else set_bit(Faulty, &rdev->flags); set_bit(MD_CHANGE_DEVS, &mddev->flags); - printk(KERN_ALERT "md/raid1:%s: Disk failure on %s, disabling device.\n" - KERN_ALERT "md/raid1:%s: Operation continuing on %d devices.\n", + printk(KERN_ALERT + "md/raid1:%s: Disk failure on %s, disabling device.\n" + "md/raid1:%s: Operation continuing on %d devices.\n", mdname(mddev), bdevname(rdev->bdev, b), mdname(mddev), conf->raid_disks - mddev->degraded); } -static void print_conf(conf_t *conf) +static void print_conf(struct r1conf *conf) { int i; @@ -1048,7 +1530,7 @@ static void print_conf(conf_t *conf) rcu_read_lock(); for (i = 0; i < conf->raid_disks; i++) { char b[BDEVNAME_SIZE]; - mdk_rdev_t *rdev = rcu_dereference(conf->mirrors[i].rdev); + struct md_rdev *rdev = rcu_dereference(conf->mirrors[i].rdev); if (rdev) printk(KERN_DEBUG " disk %d, wo:%d, o:%d, dev:%s\n", i, !test_bit(In_sync, &rdev->flags), @@ -1058,19 +1540,22 @@ static void print_conf(conf_t *conf) rcu_read_unlock(); } -static void close_sync(conf_t *conf) +static void close_sync(struct r1conf *conf) { - wait_barrier(conf); - allow_barrier(conf); + wait_barrier(conf, NULL); + allow_barrier(conf, 0, 0); mempool_destroy(conf->r1buf_pool); conf->r1buf_pool = NULL; + + conf->next_resync = 0; + conf->start_next_window = MaxSector; } -static int raid1_spare_active(mddev_t *mddev) +static int raid1_spare_active(struct mddev *mddev) { int i; - conf_t *conf = mddev->private; + struct r1conf *conf = mddev->private; int count = 0; unsigned long flags; @@ -1080,12 +1565,32 @@ static int raid1_spare_active(mddev_t *mddev) * Called under mddev lock, so rcu protection not needed. */ for (i = 0; i < conf->raid_disks; i++) { - mdk_rdev_t *rdev = conf->mirrors[i].rdev; + struct md_rdev *rdev = conf->mirrors[i].rdev; + struct md_rdev *repl = conf->mirrors[conf->raid_disks + i].rdev; + if (repl + && repl->recovery_offset == MaxSector + && !test_bit(Faulty, &repl->flags) + && !test_and_set_bit(In_sync, &repl->flags)) { + /* replacement has just become active */ + if (!rdev || + !test_and_clear_bit(In_sync, &rdev->flags)) + count++; + if (rdev) { + /* Replaced device not technically + * faulty, but we need to be sure + * it gets removed and never re-added + */ + set_bit(Faulty, &rdev->flags); + sysfs_notify_dirent_safe( + rdev->sysfs_state); + } + } if (rdev + && rdev->recovery_offset == MaxSector && !test_bit(Faulty, &rdev->flags) && !test_and_set_bit(In_sync, &rdev->flags)) { count++; - sysfs_notify_dirent(rdev->sysfs_state); + sysfs_notify_dirent_safe(rdev->sysfs_state); } } spin_lock_irqsave(&conf->device_lock, flags); @@ -1097,34 +1602,34 @@ static int raid1_spare_active(mddev_t *mddev) } -static int raid1_add_disk(mddev_t *mddev, mdk_rdev_t *rdev) +static int raid1_add_disk(struct mddev *mddev, struct md_rdev *rdev) { - conf_t *conf = mddev->private; + struct r1conf *conf = mddev->private; int err = -EEXIST; int mirror = 0; - mirror_info_t *p; + struct raid1_info *p; int first = 0; - int last = mddev->raid_disks - 1; + int last = conf->raid_disks - 1; + struct request_queue *q = bdev_get_queue(rdev->bdev); + + if (mddev->recovery_disabled == conf->recovery_disabled) + return -EBUSY; if (rdev->raid_disk >= 0) first = last = rdev->raid_disk; - for (mirror = first; mirror <= last; mirror++) - if ( !(p=conf->mirrors+mirror)->rdev) { + if (q->merge_bvec_fn) { + set_bit(Unmerged, &rdev->flags); + mddev->merge_check_needed = 1; + } - disk_stack_limits(mddev->gendisk, rdev->bdev, - rdev->data_offset << 9); - /* as we don't honour merge_bvec_fn, we must - * never risk violating it, so limit - * ->max_segments to one lying with a single - * page, as a one page request is never in - * violation. - */ - if (rdev->bdev->bd_disk->queue->merge_bvec_fn) { - blk_queue_max_segments(mddev->queue, 1); - blk_queue_segment_boundary(mddev->queue, - PAGE_CACHE_SIZE - 1); - } + for (mirror = first; mirror <= last; mirror++) { + p = conf->mirrors+mirror; + if (!p->rdev) { + + if (mddev->gendisk) + disk_stack_limits(mddev->gendisk, rdev->bdev, + rdev->data_offset << 9); p->head_position = 0; rdev->raid_disk = mirror; @@ -1137,21 +1642,50 @@ static int raid1_add_disk(mddev_t *mddev, mdk_rdev_t *rdev) rcu_assign_pointer(p->rdev, rdev); break; } + if (test_bit(WantReplacement, &p->rdev->flags) && + p[conf->raid_disks].rdev == NULL) { + /* Add this device as a replacement */ + clear_bit(In_sync, &rdev->flags); + set_bit(Replacement, &rdev->flags); + rdev->raid_disk = mirror; + err = 0; + conf->fullsync = 1; + rcu_assign_pointer(p[conf->raid_disks].rdev, rdev); + break; + } + } + if (err == 0 && test_bit(Unmerged, &rdev->flags)) { + /* Some requests might not have seen this new + * merge_bvec_fn. We must wait for them to complete + * before merging the device fully. + * First we make sure any code which has tested + * our function has submitted the request, then + * we wait for all outstanding requests to complete. + */ + synchronize_sched(); + freeze_array(conf, 0); + unfreeze_array(conf); + clear_bit(Unmerged, &rdev->flags); + } md_integrity_add_rdev(rdev, mddev); + if (mddev->queue && blk_queue_discard(bdev_get_queue(rdev->bdev))) + queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, mddev->queue); print_conf(conf); return err; } -static int raid1_remove_disk(mddev_t *mddev, int number) +static int raid1_remove_disk(struct mddev *mddev, struct md_rdev *rdev) { - conf_t *conf = mddev->private; + struct r1conf *conf = mddev->private; int err = 0; - mdk_rdev_t *rdev; - mirror_info_t *p = conf->mirrors+ number; + int number = rdev->raid_disk; + struct raid1_info *p = conf->mirrors + number; + + if (rdev != p->rdev) + p = conf->mirrors + conf->raid_disks + number; print_conf(conf); - rdev = p->rdev; - if (rdev) { + if (rdev == p->rdev) { if (test_bit(In_sync, &rdev->flags) || atomic_read(&rdev->nr_pending)) { err = -EBUSY; @@ -1161,6 +1695,7 @@ static int raid1_remove_disk(mddev_t *mddev, int number) * is not possible. */ if (!test_bit(Faulty, &rdev->flags) && + mddev->recovery_disabled != conf->recovery_disabled && mddev->degraded < conf->raid_disks) { err = -EBUSY; goto abort; @@ -1172,8 +1707,22 @@ static int raid1_remove_disk(mddev_t *mddev, int number) err = -EBUSY; p->rdev = rdev; goto abort; - } - md_integrity_register(mddev); + } else if (conf->mirrors[conf->raid_disks + number].rdev) { + /* We just removed a device that is being replaced. + * Move down the replacement. We drain all IO before + * doing this to avoid confusion. + */ + struct md_rdev *repl = + conf->mirrors[conf->raid_disks + number].rdev; + freeze_array(conf, 0); + clear_bit(Replacement, &repl->flags); + p->rdev = repl; + conf->mirrors[conf->raid_disks + number].rdev = NULL; + unfreeze_array(conf); + clear_bit(WantReplacement, &rdev->flags); + } else + clear_bit(WantReplacement, &rdev->flags); + err = md_integrity_register(mddev); } abort: @@ -1184,14 +1733,10 @@ abort: static void end_sync_read(struct bio *bio, int error) { - r1bio_t *r1_bio = bio->bi_private; - int i; + struct r1bio *r1_bio = bio->bi_private; + + update_head_pos(r1_bio->read_disk, r1_bio); - for (i=r1_bio->mddev->raid_disks; i--; ) - if (r1_bio->bios[i] == bio) - break; - BUG_ON(i < 0); - update_head_pos(i, r1_bio); /* * we have read a block, now it needs to be re-written, * or re-read if the read failed. @@ -1207,17 +1752,15 @@ static void end_sync_read(struct bio *bio, int error) static void end_sync_write(struct bio *bio, int error) { int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags); - r1bio_t *r1_bio = bio->bi_private; - mddev_t *mddev = r1_bio->mddev; - conf_t *conf = mddev->private; - int i; + struct r1bio *r1_bio = bio->bi_private; + struct mddev *mddev = r1_bio->mddev; + struct r1conf *conf = mddev->private; int mirror=0; + sector_t first_bad; + int bad_sectors; + + mirror = find_bio_disk(r1_bio, bio); - for (i = 0; i < conf->raid_disks; i++) - if (r1_bio->bios[i] == bio) { - mirror = i; - break; - } if (!uptodate) { sector_t sync_blocks = 0; sector_t s = r1_bio->sector; @@ -1229,206 +1772,288 @@ static void end_sync_write(struct bio *bio, int error) s += sync_blocks; sectors_to_go -= sync_blocks; } while (sectors_to_go > 0); - md_error(mddev, conf->mirrors[mirror].rdev); - } - - update_head_pos(mirror, r1_bio); + set_bit(WriteErrorSeen, + &conf->mirrors[mirror].rdev->flags); + if (!test_and_set_bit(WantReplacement, + &conf->mirrors[mirror].rdev->flags)) + set_bit(MD_RECOVERY_NEEDED, & + mddev->recovery); + set_bit(R1BIO_WriteError, &r1_bio->state); + } else if (is_badblock(conf->mirrors[mirror].rdev, + r1_bio->sector, + r1_bio->sectors, + &first_bad, &bad_sectors) && + !is_badblock(conf->mirrors[r1_bio->read_disk].rdev, + r1_bio->sector, + r1_bio->sectors, + &first_bad, &bad_sectors) + ) + set_bit(R1BIO_MadeGood, &r1_bio->state); if (atomic_dec_and_test(&r1_bio->remaining)) { - sector_t s = r1_bio->sectors; - put_buf(r1_bio); - md_done_sync(mddev, s, uptodate); + int s = r1_bio->sectors; + if (test_bit(R1BIO_MadeGood, &r1_bio->state) || + test_bit(R1BIO_WriteError, &r1_bio->state)) + reschedule_retry(r1_bio); + else { + put_buf(r1_bio); + md_done_sync(mddev, s, uptodate); + } } } -static void sync_request_write(mddev_t *mddev, r1bio_t *r1_bio) +static int r1_sync_page_io(struct md_rdev *rdev, sector_t sector, + int sectors, struct page *page, int rw) { - conf_t *conf = mddev->private; - int i; - int disks = conf->raid_disks; - struct bio *bio, *wbio; - - bio = r1_bio->bios[r1_bio->read_disk]; + if (sync_page_io(rdev, sector, sectors << 9, page, rw, false)) + /* success */ + return 1; + if (rw == WRITE) { + set_bit(WriteErrorSeen, &rdev->flags); + if (!test_and_set_bit(WantReplacement, + &rdev->flags)) + set_bit(MD_RECOVERY_NEEDED, & + rdev->mddev->recovery); + } + /* need to record an error - either for the block or the device */ + if (!rdev_set_badblocks(rdev, sector, sectors, 0)) + md_error(rdev->mddev, rdev); + return 0; +} +static int fix_sync_read_error(struct r1bio *r1_bio) +{ + /* Try some synchronous reads of other devices to get + * good data, much like with normal read errors. Only + * read into the pages we already have so we don't + * need to re-issue the read request. + * We don't need to freeze the array, because being in an + * active sync request, there is no normal IO, and + * no overlapping syncs. + * We don't need to check is_badblock() again as we + * made sure that anything with a bad block in range + * will have bi_end_io clear. + */ + struct mddev *mddev = r1_bio->mddev; + struct r1conf *conf = mddev->private; + struct bio *bio = r1_bio->bios[r1_bio->read_disk]; + sector_t sect = r1_bio->sector; + int sectors = r1_bio->sectors; + int idx = 0; - if (test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) { - /* We have read all readable devices. If we haven't - * got the block, then there is no hope left. - * If we have, then we want to do a comparison - * and skip the write if everything is the same. - * If any blocks failed to read, then we need to - * attempt an over-write - */ - int primary; - if (!test_bit(R1BIO_Uptodate, &r1_bio->state)) { - for (i=0; i<mddev->raid_disks; i++) - if (r1_bio->bios[i]->bi_end_io == end_sync_read) - md_error(mddev, conf->mirrors[i].rdev); + while(sectors) { + int s = sectors; + int d = r1_bio->read_disk; + int success = 0; + struct md_rdev *rdev; + int start; - md_done_sync(mddev, r1_bio->sectors, 1); - put_buf(r1_bio); - return; - } - for (primary=0; primary<mddev->raid_disks; primary++) - if (r1_bio->bios[primary]->bi_end_io == end_sync_read && - test_bit(BIO_UPTODATE, &r1_bio->bios[primary]->bi_flags)) { - r1_bio->bios[primary]->bi_end_io = NULL; - rdev_dec_pending(conf->mirrors[primary].rdev, mddev); - break; + if (s > (PAGE_SIZE>>9)) + s = PAGE_SIZE >> 9; + do { + if (r1_bio->bios[d]->bi_end_io == end_sync_read) { + /* No rcu protection needed here devices + * can only be removed when no resync is + * active, and resync is currently active + */ + rdev = conf->mirrors[d].rdev; + if (sync_page_io(rdev, sect, s<<9, + bio->bi_io_vec[idx].bv_page, + READ, false)) { + success = 1; + break; + } } - r1_bio->read_disk = primary; - for (i=0; i<mddev->raid_disks; i++) - if (r1_bio->bios[i]->bi_end_io == end_sync_read) { - int j; - int vcnt = r1_bio->sectors >> (PAGE_SHIFT- 9); - struct bio *pbio = r1_bio->bios[primary]; - struct bio *sbio = r1_bio->bios[i]; - - if (test_bit(BIO_UPTODATE, &sbio->bi_flags)) { - for (j = vcnt; j-- ; ) { - struct page *p, *s; - p = pbio->bi_io_vec[j].bv_page; - s = sbio->bi_io_vec[j].bv_page; - if (memcmp(page_address(p), - page_address(s), - PAGE_SIZE)) - break; - } - } else - j = 0; - if (j >= 0) - mddev->resync_mismatches += r1_bio->sectors; - if (j < 0 || (test_bit(MD_RECOVERY_CHECK, &mddev->recovery) - && test_bit(BIO_UPTODATE, &sbio->bi_flags))) { - sbio->bi_end_io = NULL; - rdev_dec_pending(conf->mirrors[i].rdev, mddev); - } else { - /* fixup the bio for reuse */ - int size; - sbio->bi_vcnt = vcnt; - sbio->bi_size = r1_bio->sectors << 9; - sbio->bi_idx = 0; - sbio->bi_phys_segments = 0; - sbio->bi_flags &= ~(BIO_POOL_MASK - 1); - sbio->bi_flags |= 1 << BIO_UPTODATE; - sbio->bi_next = NULL; - sbio->bi_sector = r1_bio->sector + - conf->mirrors[i].rdev->data_offset; - sbio->bi_bdev = conf->mirrors[i].rdev->bdev; - size = sbio->bi_size; - for (j = 0; j < vcnt ; j++) { - struct bio_vec *bi; - bi = &sbio->bi_io_vec[j]; - bi->bv_offset = 0; - if (size > PAGE_SIZE) - bi->bv_len = PAGE_SIZE; - else - bi->bv_len = size; - size -= PAGE_SIZE; - memcpy(page_address(bi->bv_page), - page_address(pbio->bi_io_vec[j].bv_page), - PAGE_SIZE); - } + d++; + if (d == conf->raid_disks * 2) + d = 0; + } while (!success && d != r1_bio->read_disk); - } + if (!success) { + char b[BDEVNAME_SIZE]; + int abort = 0; + /* Cannot read from anywhere, this block is lost. + * Record a bad block on each device. If that doesn't + * work just disable and interrupt the recovery. + * Don't fail devices as that won't really help. + */ + printk(KERN_ALERT "md/raid1:%s: %s: unrecoverable I/O read error" + " for block %llu\n", + mdname(mddev), + bdevname(bio->bi_bdev, b), + (unsigned long long)r1_bio->sector); + for (d = 0; d < conf->raid_disks * 2; d++) { + rdev = conf->mirrors[d].rdev; + if (!rdev || test_bit(Faulty, &rdev->flags)) + continue; + if (!rdev_set_badblocks(rdev, sect, s, 0)) + abort = 1; } - } - if (!test_bit(R1BIO_Uptodate, &r1_bio->state)) { - /* ouch - failed to read all of that. - * Try some synchronous reads of other devices to get - * good data, much like with normal read errors. Only - * read into the pages we already have so we don't - * need to re-issue the read request. - * We don't need to freeze the array, because being in an - * active sync request, there is no normal IO, and - * no overlapping syncs. - */ - sector_t sect = r1_bio->sector; - int sectors = r1_bio->sectors; - int idx = 0; - - while(sectors) { - int s = sectors; - int d = r1_bio->read_disk; - int success = 0; - mdk_rdev_t *rdev; - - if (s > (PAGE_SIZE>>9)) - s = PAGE_SIZE >> 9; - do { - if (r1_bio->bios[d]->bi_end_io == end_sync_read) { - /* No rcu protection needed here devices - * can only be removed when no resync is - * active, and resync is currently active - */ - rdev = conf->mirrors[d].rdev; - if (sync_page_io(rdev, - sect + rdev->data_offset, - s<<9, - bio->bi_io_vec[idx].bv_page, - READ)) { - success = 1; - break; - } - } - d++; - if (d == conf->raid_disks) - d = 0; - } while (!success && d != r1_bio->read_disk); - - if (success) { - int start = d; - /* write it back and re-read */ - set_bit(R1BIO_Uptodate, &r1_bio->state); - while (d != r1_bio->read_disk) { - if (d == 0) - d = conf->raid_disks; - d--; - if (r1_bio->bios[d]->bi_end_io != end_sync_read) - continue; - rdev = conf->mirrors[d].rdev; - atomic_add(s, &rdev->corrected_errors); - if (sync_page_io(rdev, - sect + rdev->data_offset, - s<<9, - bio->bi_io_vec[idx].bv_page, - WRITE) == 0) - md_error(mddev, rdev); - } - d = start; - while (d != r1_bio->read_disk) { - if (d == 0) - d = conf->raid_disks; - d--; - if (r1_bio->bios[d]->bi_end_io != end_sync_read) - continue; - rdev = conf->mirrors[d].rdev; - if (sync_page_io(rdev, - sect + rdev->data_offset, - s<<9, - bio->bi_io_vec[idx].bv_page, - READ) == 0) - md_error(mddev, rdev); - } - } else { - char b[BDEVNAME_SIZE]; - /* Cannot read from anywhere, array is toast */ - md_error(mddev, conf->mirrors[r1_bio->read_disk].rdev); - printk(KERN_ALERT "md/raid1:%s: %s: unrecoverable I/O read error" - " for block %llu\n", - mdname(mddev), - bdevname(bio->bi_bdev, b), - (unsigned long long)r1_bio->sector); + if (abort) { + conf->recovery_disabled = + mddev->recovery_disabled; + set_bit(MD_RECOVERY_INTR, &mddev->recovery); md_done_sync(mddev, r1_bio->sectors, 0); put_buf(r1_bio); - return; + return 0; } + /* Try next page */ sectors -= s; sect += s; - idx ++; + idx++; + continue; + } + + start = d; + /* write it back and re-read */ + while (d != r1_bio->read_disk) { + if (d == 0) + d = conf->raid_disks * 2; + d--; + if (r1_bio->bios[d]->bi_end_io != end_sync_read) + continue; + rdev = conf->mirrors[d].rdev; + if (r1_sync_page_io(rdev, sect, s, + bio->bi_io_vec[idx].bv_page, + WRITE) == 0) { + r1_bio->bios[d]->bi_end_io = NULL; + rdev_dec_pending(rdev, mddev); + } + } + d = start; + while (d != r1_bio->read_disk) { + if (d == 0) + d = conf->raid_disks * 2; + d--; + if (r1_bio->bios[d]->bi_end_io != end_sync_read) + continue; + rdev = conf->mirrors[d].rdev; + if (r1_sync_page_io(rdev, sect, s, + bio->bi_io_vec[idx].bv_page, + READ) != 0) + atomic_add(s, &rdev->corrected_errors); + } + sectors -= s; + sect += s; + idx ++; + } + set_bit(R1BIO_Uptodate, &r1_bio->state); + set_bit(BIO_UPTODATE, &bio->bi_flags); + return 1; +} + +static int process_checks(struct r1bio *r1_bio) +{ + /* We have read all readable devices. If we haven't + * got the block, then there is no hope left. + * If we have, then we want to do a comparison + * and skip the write if everything is the same. + * If any blocks failed to read, then we need to + * attempt an over-write + */ + struct mddev *mddev = r1_bio->mddev; + struct r1conf *conf = mddev->private; + int primary; + int i; + int vcnt; + + /* Fix variable parts of all bios */ + vcnt = (r1_bio->sectors + PAGE_SIZE / 512 - 1) >> (PAGE_SHIFT - 9); + for (i = 0; i < conf->raid_disks * 2; i++) { + int j; + int size; + int uptodate; + struct bio *b = r1_bio->bios[i]; + if (b->bi_end_io != end_sync_read) + continue; + /* fixup the bio for reuse, but preserve BIO_UPTODATE */ + uptodate = test_bit(BIO_UPTODATE, &b->bi_flags); + bio_reset(b); + if (!uptodate) + clear_bit(BIO_UPTODATE, &b->bi_flags); + b->bi_vcnt = vcnt; + b->bi_iter.bi_size = r1_bio->sectors << 9; + b->bi_iter.bi_sector = r1_bio->sector + + conf->mirrors[i].rdev->data_offset; + b->bi_bdev = conf->mirrors[i].rdev->bdev; + b->bi_end_io = end_sync_read; + b->bi_private = r1_bio; + + size = b->bi_iter.bi_size; + for (j = 0; j < vcnt ; j++) { + struct bio_vec *bi; + bi = &b->bi_io_vec[j]; + bi->bv_offset = 0; + if (size > PAGE_SIZE) + bi->bv_len = PAGE_SIZE; + else + bi->bv_len = size; + size -= PAGE_SIZE; } } + for (primary = 0; primary < conf->raid_disks * 2; primary++) + if (r1_bio->bios[primary]->bi_end_io == end_sync_read && + test_bit(BIO_UPTODATE, &r1_bio->bios[primary]->bi_flags)) { + r1_bio->bios[primary]->bi_end_io = NULL; + rdev_dec_pending(conf->mirrors[primary].rdev, mddev); + break; + } + r1_bio->read_disk = primary; + for (i = 0; i < conf->raid_disks * 2; i++) { + int j; + struct bio *pbio = r1_bio->bios[primary]; + struct bio *sbio = r1_bio->bios[i]; + int uptodate = test_bit(BIO_UPTODATE, &sbio->bi_flags); + + if (sbio->bi_end_io != end_sync_read) + continue; + /* Now we can 'fixup' the BIO_UPTODATE flag */ + set_bit(BIO_UPTODATE, &sbio->bi_flags); + + if (uptodate) { + for (j = vcnt; j-- ; ) { + struct page *p, *s; + p = pbio->bi_io_vec[j].bv_page; + s = sbio->bi_io_vec[j].bv_page; + if (memcmp(page_address(p), + page_address(s), + sbio->bi_io_vec[j].bv_len)) + break; + } + } else + j = 0; + if (j >= 0) + atomic64_add(r1_bio->sectors, &mddev->resync_mismatches); + if (j < 0 || (test_bit(MD_RECOVERY_CHECK, &mddev->recovery) + && uptodate)) { + /* No need to write to this device. */ + sbio->bi_end_io = NULL; + rdev_dec_pending(conf->mirrors[i].rdev, mddev); + continue; + } + bio_copy_data(sbio, pbio); + } + return 0; +} + +static void sync_request_write(struct mddev *mddev, struct r1bio *r1_bio) +{ + struct r1conf *conf = mddev->private; + int i; + int disks = conf->raid_disks * 2; + struct bio *bio, *wbio; + + bio = r1_bio->bios[r1_bio->read_disk]; + + if (!test_bit(R1BIO_Uptodate, &r1_bio->state)) + /* ouch - failed to read all of that. */ + if (!fix_sync_read_error(r1_bio)) + return; + + if (test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) + if (process_checks(r1_bio) < 0) + return; /* * schedule writes */ @@ -1444,15 +2069,21 @@ static void sync_request_write(mddev_t *mddev, r1bio_t *r1_bio) wbio->bi_rw = WRITE; wbio->bi_end_io = end_sync_write; atomic_inc(&r1_bio->remaining); - md_sync_acct(conf->mirrors[i].rdev->bdev, wbio->bi_size >> 9); + md_sync_acct(conf->mirrors[i].rdev->bdev, bio_sectors(wbio)); generic_make_request(wbio); } if (atomic_dec_and_test(&r1_bio->remaining)) { /* if we're here, all write(s) have completed, so clean up */ - md_done_sync(mddev, r1_bio->sectors, 1); - put_buf(r1_bio); + int s = r1_bio->sectors; + if (test_bit(R1BIO_MadeGood, &r1_bio->state) || + test_bit(R1BIO_WriteError, &r1_bio->state)) + reschedule_retry(r1_bio); + else { + put_buf(r1_bio); + md_done_sync(mddev, s, 1); + } } } @@ -1461,19 +2092,19 @@ static void sync_request_write(mddev_t *mddev, r1bio_t *r1_bio) * * 1. Retries failed read operations on working mirrors. * 2. Updates the raid superblock when problems encounter. - * 3. Performs writes following reads for array syncronising. + * 3. Performs writes following reads for array synchronising. */ -static void fix_read_error(conf_t *conf, int read_disk, +static void fix_read_error(struct r1conf *conf, int read_disk, sector_t sect, int sectors) { - mddev_t *mddev = conf->mddev; + struct mddev *mddev = conf->mddev; while(sectors) { int s = sectors; int d = read_disk; int success = 0; int start; - mdk_rdev_t *rdev; + struct md_rdev *rdev; if (s > (PAGE_SIZE>>9)) s = PAGE_SIZE >> 9; @@ -1484,59 +2115,56 @@ static void fix_read_error(conf_t *conf, int read_disk, * which is the thread that might remove * a device. If raid1d ever becomes multi-threaded.... */ + sector_t first_bad; + int bad_sectors; + rdev = conf->mirrors[d].rdev; if (rdev && - test_bit(In_sync, &rdev->flags) && - sync_page_io(rdev, - sect + rdev->data_offset, - s<<9, - conf->tmppage, READ)) + (test_bit(In_sync, &rdev->flags) || + (!test_bit(Faulty, &rdev->flags) && + rdev->recovery_offset >= sect + s)) && + is_badblock(rdev, sect, s, + &first_bad, &bad_sectors) == 0 && + sync_page_io(rdev, sect, s<<9, + conf->tmppage, READ, false)) success = 1; else { d++; - if (d == conf->raid_disks) + if (d == conf->raid_disks * 2) d = 0; } } while (!success && d != read_disk); if (!success) { - /* Cannot read from anywhere -- bye bye array */ - md_error(mddev, conf->mirrors[read_disk].rdev); + /* Cannot read from anywhere - mark it bad */ + struct md_rdev *rdev = conf->mirrors[read_disk].rdev; + if (!rdev_set_badblocks(rdev, sect, s, 0)) + md_error(mddev, rdev); break; } /* write it back and re-read */ start = d; while (d != read_disk) { if (d==0) - d = conf->raid_disks; + d = conf->raid_disks * 2; d--; rdev = conf->mirrors[d].rdev; if (rdev && - test_bit(In_sync, &rdev->flags)) { - if (sync_page_io(rdev, - sect + rdev->data_offset, - s<<9, conf->tmppage, WRITE) - == 0) - /* Well, this device is dead */ - md_error(mddev, rdev); - } + test_bit(In_sync, &rdev->flags)) + r1_sync_page_io(rdev, sect, s, + conf->tmppage, WRITE); } d = start; while (d != read_disk) { char b[BDEVNAME_SIZE]; if (d==0) - d = conf->raid_disks; + d = conf->raid_disks * 2; d--; rdev = conf->mirrors[d].rdev; if (rdev && test_bit(In_sync, &rdev->flags)) { - if (sync_page_io(rdev, - sect + rdev->data_offset, - s<<9, conf->tmppage, READ) - == 0) - /* Well, this device is dead */ - md_error(mddev, rdev); - else { + if (r1_sync_page_io(rdev, sect, s, + conf->tmppage, READ)) { atomic_add(s, &rdev->corrected_errors); printk(KERN_INFO "md/raid1:%s: read error corrected " @@ -1553,29 +2181,248 @@ static void fix_read_error(conf_t *conf, int read_disk, } } -static void raid1d(mddev_t *mddev) +static int narrow_write_error(struct r1bio *r1_bio, int i) +{ + struct mddev *mddev = r1_bio->mddev; + struct r1conf *conf = mddev->private; + struct md_rdev *rdev = conf->mirrors[i].rdev; + + /* bio has the data to be written to device 'i' where + * we just recently had a write error. + * We repeatedly clone the bio and trim down to one block, + * then try the write. Where the write fails we record + * a bad block. + * It is conceivable that the bio doesn't exactly align with + * blocks. We must handle this somehow. + * + * We currently own a reference on the rdev. + */ + + int block_sectors; + sector_t sector; + int sectors; + int sect_to_write = r1_bio->sectors; + int ok = 1; + + if (rdev->badblocks.shift < 0) + return 0; + + block_sectors = 1 << rdev->badblocks.shift; + sector = r1_bio->sector; + sectors = ((sector + block_sectors) + & ~(sector_t)(block_sectors - 1)) + - sector; + + while (sect_to_write) { + struct bio *wbio; + if (sectors > sect_to_write) + sectors = sect_to_write; + /* Write at 'sector' for 'sectors'*/ + + if (test_bit(R1BIO_BehindIO, &r1_bio->state)) { + unsigned vcnt = r1_bio->behind_page_count; + struct bio_vec *vec = r1_bio->behind_bvecs; + + while (!vec->bv_page) { + vec++; + vcnt--; + } + + wbio = bio_alloc_mddev(GFP_NOIO, vcnt, mddev); + memcpy(wbio->bi_io_vec, vec, vcnt * sizeof(struct bio_vec)); + + wbio->bi_vcnt = vcnt; + } else { + wbio = bio_clone_mddev(r1_bio->master_bio, GFP_NOIO, mddev); + } + + wbio->bi_rw = WRITE; + wbio->bi_iter.bi_sector = r1_bio->sector; + wbio->bi_iter.bi_size = r1_bio->sectors << 9; + + bio_trim(wbio, sector - r1_bio->sector, sectors); + wbio->bi_iter.bi_sector += rdev->data_offset; + wbio->bi_bdev = rdev->bdev; + if (submit_bio_wait(WRITE, wbio) == 0) + /* failure! */ + ok = rdev_set_badblocks(rdev, sector, + sectors, 0) + && ok; + + bio_put(wbio); + sect_to_write -= sectors; + sector += sectors; + sectors = block_sectors; + } + return ok; +} + +static void handle_sync_write_finished(struct r1conf *conf, struct r1bio *r1_bio) { - r1bio_t *r1_bio; + int m; + int s = r1_bio->sectors; + for (m = 0; m < conf->raid_disks * 2 ; m++) { + struct md_rdev *rdev = conf->mirrors[m].rdev; + struct bio *bio = r1_bio->bios[m]; + if (bio->bi_end_io == NULL) + continue; + if (test_bit(BIO_UPTODATE, &bio->bi_flags) && + test_bit(R1BIO_MadeGood, &r1_bio->state)) { + rdev_clear_badblocks(rdev, r1_bio->sector, s, 0); + } + if (!test_bit(BIO_UPTODATE, &bio->bi_flags) && + test_bit(R1BIO_WriteError, &r1_bio->state)) { + if (!rdev_set_badblocks(rdev, r1_bio->sector, s, 0)) + md_error(conf->mddev, rdev); + } + } + put_buf(r1_bio); + md_done_sync(conf->mddev, s, 1); +} + +static void handle_write_finished(struct r1conf *conf, struct r1bio *r1_bio) +{ + int m; + for (m = 0; m < conf->raid_disks * 2 ; m++) + if (r1_bio->bios[m] == IO_MADE_GOOD) { + struct md_rdev *rdev = conf->mirrors[m].rdev; + rdev_clear_badblocks(rdev, + r1_bio->sector, + r1_bio->sectors, 0); + rdev_dec_pending(rdev, conf->mddev); + } else if (r1_bio->bios[m] != NULL) { + /* This drive got a write error. We need to + * narrow down and record precise write + * errors. + */ + if (!narrow_write_error(r1_bio, m)) { + md_error(conf->mddev, + conf->mirrors[m].rdev); + /* an I/O failed, we can't clear the bitmap */ + set_bit(R1BIO_Degraded, &r1_bio->state); + } + rdev_dec_pending(conf->mirrors[m].rdev, + conf->mddev); + } + if (test_bit(R1BIO_WriteError, &r1_bio->state)) + close_write(r1_bio); + raid_end_bio_io(r1_bio); +} + +static void handle_read_error(struct r1conf *conf, struct r1bio *r1_bio) +{ + int disk; + int max_sectors; + struct mddev *mddev = conf->mddev; struct bio *bio; + char b[BDEVNAME_SIZE]; + struct md_rdev *rdev; + + clear_bit(R1BIO_ReadError, &r1_bio->state); + /* we got a read error. Maybe the drive is bad. Maybe just + * the block and we can fix it. + * We freeze all other IO, and try reading the block from + * other devices. When we find one, we re-write + * and check it that fixes the read error. + * This is all done synchronously while the array is + * frozen + */ + if (mddev->ro == 0) { + freeze_array(conf, 1); + fix_read_error(conf, r1_bio->read_disk, + r1_bio->sector, r1_bio->sectors); + unfreeze_array(conf); + } else + md_error(mddev, conf->mirrors[r1_bio->read_disk].rdev); + rdev_dec_pending(conf->mirrors[r1_bio->read_disk].rdev, conf->mddev); + + bio = r1_bio->bios[r1_bio->read_disk]; + bdevname(bio->bi_bdev, b); +read_more: + disk = read_balance(conf, r1_bio, &max_sectors); + if (disk == -1) { + printk(KERN_ALERT "md/raid1:%s: %s: unrecoverable I/O" + " read error for block %llu\n", + mdname(mddev), b, (unsigned long long)r1_bio->sector); + raid_end_bio_io(r1_bio); + } else { + const unsigned long do_sync + = r1_bio->master_bio->bi_rw & REQ_SYNC; + if (bio) { + r1_bio->bios[r1_bio->read_disk] = + mddev->ro ? IO_BLOCKED : NULL; + bio_put(bio); + } + r1_bio->read_disk = disk; + bio = bio_clone_mddev(r1_bio->master_bio, GFP_NOIO, mddev); + bio_trim(bio, r1_bio->sector - bio->bi_iter.bi_sector, + max_sectors); + r1_bio->bios[r1_bio->read_disk] = bio; + rdev = conf->mirrors[disk].rdev; + printk_ratelimited(KERN_ERR + "md/raid1:%s: redirecting sector %llu" + " to other mirror: %s\n", + mdname(mddev), + (unsigned long long)r1_bio->sector, + bdevname(rdev->bdev, b)); + bio->bi_iter.bi_sector = r1_bio->sector + rdev->data_offset; + bio->bi_bdev = rdev->bdev; + bio->bi_end_io = raid1_end_read_request; + bio->bi_rw = READ | do_sync; + bio->bi_private = r1_bio; + if (max_sectors < r1_bio->sectors) { + /* Drat - have to split this up more */ + struct bio *mbio = r1_bio->master_bio; + int sectors_handled = (r1_bio->sector + max_sectors + - mbio->bi_iter.bi_sector); + r1_bio->sectors = max_sectors; + spin_lock_irq(&conf->device_lock); + if (mbio->bi_phys_segments == 0) + mbio->bi_phys_segments = 2; + else + mbio->bi_phys_segments++; + spin_unlock_irq(&conf->device_lock); + generic_make_request(bio); + bio = NULL; + + r1_bio = mempool_alloc(conf->r1bio_pool, GFP_NOIO); + + r1_bio->master_bio = mbio; + r1_bio->sectors = bio_sectors(mbio) - sectors_handled; + r1_bio->state = 0; + set_bit(R1BIO_ReadError, &r1_bio->state); + r1_bio->mddev = mddev; + r1_bio->sector = mbio->bi_iter.bi_sector + + sectors_handled; + + goto read_more; + } else + generic_make_request(bio); + } +} + +static void raid1d(struct md_thread *thread) +{ + struct mddev *mddev = thread->mddev; + struct r1bio *r1_bio; unsigned long flags; - conf_t *conf = mddev->private; + struct r1conf *conf = mddev->private; struct list_head *head = &conf->retry_list; - int unplug=0; - mdk_rdev_t *rdev; + struct blk_plug plug; md_check_recovery(mddev); - + + blk_start_plug(&plug); for (;;) { - char b[BDEVNAME_SIZE]; - unplug += flush_pending_writes(conf); + flush_pending_writes(conf); spin_lock_irqsave(&conf->device_lock, flags); if (list_empty(head)) { spin_unlock_irqrestore(&conf->device_lock, flags); break; } - r1_bio = list_entry(head->prev, r1bio_t, retry_list); + r1_bio = list_entry(head->prev, struct r1bio, retry_list); list_del(head->prev); conf->nr_queued--; spin_unlock_irqrestore(&conf->device_lock, flags); @@ -1583,70 +2430,31 @@ static void raid1d(mddev_t *mddev) mddev = r1_bio->mddev; conf = mddev->private; if (test_bit(R1BIO_IsSync, &r1_bio->state)) { - sync_request_write(mddev, r1_bio); - unplug = 1; - } else { - int disk; - - /* we got a read error. Maybe the drive is bad. Maybe just - * the block and we can fix it. - * We freeze all other IO, and try reading the block from - * other devices. When we find one, we re-write - * and check it that fixes the read error. - * This is all done synchronously while the array is - * frozen + if (test_bit(R1BIO_MadeGood, &r1_bio->state) || + test_bit(R1BIO_WriteError, &r1_bio->state)) + handle_sync_write_finished(conf, r1_bio); + else + sync_request_write(mddev, r1_bio); + } else if (test_bit(R1BIO_MadeGood, &r1_bio->state) || + test_bit(R1BIO_WriteError, &r1_bio->state)) + handle_write_finished(conf, r1_bio); + else if (test_bit(R1BIO_ReadError, &r1_bio->state)) + handle_read_error(conf, r1_bio); + else + /* just a partial read to be scheduled from separate + * context */ - if (mddev->ro == 0) { - freeze_array(conf); - fix_read_error(conf, r1_bio->read_disk, - r1_bio->sector, - r1_bio->sectors); - unfreeze_array(conf); - } else - md_error(mddev, - conf->mirrors[r1_bio->read_disk].rdev); - - bio = r1_bio->bios[r1_bio->read_disk]; - if ((disk=read_balance(conf, r1_bio)) == -1) { - printk(KERN_ALERT "md/raid1:%s: %s: unrecoverable I/O" - " read error for block %llu\n", - mdname(mddev), - bdevname(bio->bi_bdev,b), - (unsigned long long)r1_bio->sector); - raid_end_bio_io(r1_bio); - } else { - const unsigned long do_sync = r1_bio->master_bio->bi_rw & REQ_SYNC; - r1_bio->bios[r1_bio->read_disk] = - mddev->ro ? IO_BLOCKED : NULL; - r1_bio->read_disk = disk; - bio_put(bio); - bio = bio_clone_mddev(r1_bio->master_bio, - GFP_NOIO, mddev); - r1_bio->bios[r1_bio->read_disk] = bio; - rdev = conf->mirrors[disk].rdev; - if (printk_ratelimit()) - printk(KERN_ERR "md/raid1:%s: redirecting sector %llu to" - " other mirror: %s\n", - mdname(mddev), - (unsigned long long)r1_bio->sector, - bdevname(rdev->bdev,b)); - bio->bi_sector = r1_bio->sector + rdev->data_offset; - bio->bi_bdev = rdev->bdev; - bio->bi_end_io = raid1_end_read_request; - bio->bi_rw = READ | do_sync; - bio->bi_private = r1_bio; - unplug = 1; - generic_make_request(bio); - } - } + generic_make_request(r1_bio->bios[r1_bio->read_disk]); + cond_resched(); + if (mddev->flags & ~(1<<MD_CHANGE_PENDING)) + md_check_recovery(mddev); } - if (unplug) - unplug_slaves(mddev); + blk_finish_plug(&plug); } -static int init_resync(conf_t *conf) +static int init_resync(struct r1conf *conf) { int buffs; @@ -1670,10 +2478,10 @@ static int init_resync(conf_t *conf) * that can be installed to exclude normal IO requests. */ -static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, int go_faster) +static sector_t sync_request(struct mddev *mddev, sector_t sector_nr, int *skipped, int go_faster) { - conf_t *conf = mddev->private; - r1bio_t *r1_bio; + struct r1conf *conf = mddev->private; + struct r1bio *r1_bio; struct bio *bio; sector_t max_sector, nr_sectors; int disk = -1; @@ -1682,6 +2490,8 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i int write_targets = 0, read_targets = 0; sector_t sync_blocks; int still_degraded = 0; + int good_sectors = RESYNC_SECTORS; + int min_bad = 0; /* number of sectors that are bad in all devices */ if (!conf->r1buf_pool) if (init_resync(conf)) @@ -1750,55 +2560,108 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i r1_bio->state = 0; set_bit(R1BIO_IsSync, &r1_bio->state); - for (i=0; i < conf->raid_disks; i++) { - mdk_rdev_t *rdev; + for (i = 0; i < conf->raid_disks * 2; i++) { + struct md_rdev *rdev; bio = r1_bio->bios[i]; - - /* take from bio_init */ - bio->bi_next = NULL; - bio->bi_flags &= ~(BIO_POOL_MASK-1); - bio->bi_flags |= 1 << BIO_UPTODATE; - bio->bi_comp_cpu = -1; - bio->bi_rw = READ; - bio->bi_vcnt = 0; - bio->bi_idx = 0; - bio->bi_phys_segments = 0; - bio->bi_size = 0; - bio->bi_end_io = NULL; - bio->bi_private = NULL; + bio_reset(bio); rdev = rcu_dereference(conf->mirrors[i].rdev); if (rdev == NULL || - test_bit(Faulty, &rdev->flags)) { - still_degraded = 1; - continue; + test_bit(Faulty, &rdev->flags)) { + if (i < conf->raid_disks) + still_degraded = 1; } else if (!test_bit(In_sync, &rdev->flags)) { bio->bi_rw = WRITE; bio->bi_end_io = end_sync_write; write_targets ++; } else { /* may need to read from here */ - bio->bi_rw = READ; - bio->bi_end_io = end_sync_read; - if (test_bit(WriteMostly, &rdev->flags)) { - if (wonly < 0) - wonly = i; - } else { - if (disk < 0) - disk = i; + sector_t first_bad = MaxSector; + int bad_sectors; + + if (is_badblock(rdev, sector_nr, good_sectors, + &first_bad, &bad_sectors)) { + if (first_bad > sector_nr) + good_sectors = first_bad - sector_nr; + else { + bad_sectors -= (sector_nr - first_bad); + if (min_bad == 0 || + min_bad > bad_sectors) + min_bad = bad_sectors; + } + } + if (sector_nr < first_bad) { + if (test_bit(WriteMostly, &rdev->flags)) { + if (wonly < 0) + wonly = i; + } else { + if (disk < 0) + disk = i; + } + bio->bi_rw = READ; + bio->bi_end_io = end_sync_read; + read_targets++; + } else if (!test_bit(WriteErrorSeen, &rdev->flags) && + test_bit(MD_RECOVERY_SYNC, &mddev->recovery) && + !test_bit(MD_RECOVERY_CHECK, &mddev->recovery)) { + /* + * The device is suitable for reading (InSync), + * but has bad block(s) here. Let's try to correct them, + * if we are doing resync or repair. Otherwise, leave + * this device alone for this sync request. + */ + bio->bi_rw = WRITE; + bio->bi_end_io = end_sync_write; + write_targets++; } - read_targets++; } - atomic_inc(&rdev->nr_pending); - bio->bi_sector = sector_nr + rdev->data_offset; - bio->bi_bdev = rdev->bdev; - bio->bi_private = r1_bio; + if (bio->bi_end_io) { + atomic_inc(&rdev->nr_pending); + bio->bi_iter.bi_sector = sector_nr + rdev->data_offset; + bio->bi_bdev = rdev->bdev; + bio->bi_private = r1_bio; + } } rcu_read_unlock(); if (disk < 0) disk = wonly; r1_bio->read_disk = disk; + if (read_targets == 0 && min_bad > 0) { + /* These sectors are bad on all InSync devices, so we + * need to mark them bad on all write targets + */ + int ok = 1; + for (i = 0 ; i < conf->raid_disks * 2 ; i++) + if (r1_bio->bios[i]->bi_end_io == end_sync_write) { + struct md_rdev *rdev = conf->mirrors[i].rdev; + ok = rdev_set_badblocks(rdev, sector_nr, + min_bad, 0 + ) && ok; + } + set_bit(MD_CHANGE_DEVS, &mddev->flags); + *skipped = 1; + put_buf(r1_bio); + + if (!ok) { + /* Cannot record the badblocks, so need to + * abort the resync. + * If there are multiple read targets, could just + * fail the really bad ones ??? + */ + conf->recovery_disabled = mddev->recovery_disabled; + set_bit(MD_RECOVERY_INTR, &mddev->recovery); + return 0; + } else + return min_bad; + + } + if (min_bad > 0 && min_bad < good_sectors) { + /* only resync enough to reach the next bad->good + * transition */ + good_sectors = min_bad; + } + if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery) && read_targets > 0) /* extra read targets are also write targets */ write_targets += read_targets-1; @@ -1807,7 +2670,10 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i /* There is nowhere to write, so all non-sync * drives must be failed - so we are finished */ - sector_t rv = max_sector - sector_nr; + sector_t rv; + if (min_bad > 0) + max_sector = sector_nr + min_bad; + rv = max_sector - sector_nr; *skipped = 1; put_buf(r1_bio); return rv; @@ -1815,6 +2681,8 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i if (max_sector > mddev->resync_max) max_sector = mddev->resync_max; /* Don't do IO beyond here */ + if (max_sector > sector_nr + good_sectors) + max_sector = sector_nr + good_sectors; nr_sectors = 0; sync_blocks = 0; do { @@ -1835,7 +2703,7 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i len = sync_blocks<<9; } - for (i=0 ; i < conf->raid_disks; i++) { + for (i = 0 ; i < conf->raid_disks * 2; i++) { bio = r1_bio->bios[i]; if (bio->bi_end_io) { page = bio->bi_io_vec[bio->bi_vcnt].bv_page; @@ -1849,7 +2717,7 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i continue; /* remove last page from this bio */ bio->bi_vcnt--; - bio->bi_size -= len; + bio->bi_iter.bi_size -= len; bio->bi_flags &= ~(1<< BIO_SEG_VALID); } goto bio_full; @@ -1868,9 +2736,10 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i */ if (test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) { atomic_set(&r1_bio->remaining, read_targets); - for (i=0; i<conf->raid_disks; i++) { + for (i = 0; i < conf->raid_disks * 2 && read_targets; i++) { bio = r1_bio->bios[i]; if (bio->bi_end_io == end_sync_read) { + read_targets--; md_sync_acct(bio->bi_bdev, nr_sectors); generic_make_request(bio); } @@ -1885,7 +2754,7 @@ static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, i return nr_sectors; } -static sector_t raid1_size(mddev_t *mddev, sector_t sectors, int raid_disks) +static sector_t raid1_size(struct mddev *mddev, sector_t sectors, int raid_disks) { if (sectors) return sectors; @@ -1893,19 +2762,20 @@ static sector_t raid1_size(mddev_t *mddev, sector_t sectors, int raid_disks) return mddev->dev_sectors; } -static conf_t *setup_conf(mddev_t *mddev) +static struct r1conf *setup_conf(struct mddev *mddev) { - conf_t *conf; + struct r1conf *conf; int i; - mirror_info_t *disk; - mdk_rdev_t *rdev; + struct raid1_info *disk; + struct md_rdev *rdev; int err = -ENOMEM; - conf = kzalloc(sizeof(conf_t), GFP_KERNEL); + conf = kzalloc(sizeof(struct r1conf), GFP_KERNEL); if (!conf) goto abort; - conf->mirrors = kzalloc(sizeof(struct mirror_info)*mddev->raid_disks, + conf->mirrors = kzalloc(sizeof(struct raid1_info) + * mddev->raid_disks * 2, GFP_KERNEL); if (!conf->mirrors) goto abort; @@ -1917,7 +2787,7 @@ static conf_t *setup_conf(mddev_t *mddev) conf->poolinfo = kzalloc(sizeof(*conf->poolinfo), GFP_KERNEL); if (!conf->poolinfo) goto abort; - conf->poolinfo->raid_disks = mddev->raid_disks; + conf->poolinfo->raid_disks = mddev->raid_disks * 2; conf->r1bio_pool = mempool_create(NR_RAID1_BIOS, r1bio_pool_alloc, r1bio_pool_free, conf->poolinfo); @@ -1926,17 +2796,28 @@ static conf_t *setup_conf(mddev_t *mddev) conf->poolinfo->mddev = mddev; + err = -EINVAL; spin_lock_init(&conf->device_lock); - list_for_each_entry(rdev, &mddev->disks, same_set) { + rdev_for_each(rdev, mddev) { + struct request_queue *q; int disk_idx = rdev->raid_disk; if (disk_idx >= mddev->raid_disks || disk_idx < 0) continue; - disk = conf->mirrors + disk_idx; + if (test_bit(Replacement, &rdev->flags)) + disk = conf->mirrors + mddev->raid_disks + disk_idx; + else + disk = conf->mirrors + disk_idx; + if (disk->rdev) + goto abort; disk->rdev = rdev; + q = bdev_get_queue(rdev->bdev); + if (q->merge_bvec_fn) + mddev->merge_check_needed = 1; disk->head_position = 0; + disk->seq_start = MaxSector; } conf->raid_disks = mddev->raid_disks; conf->mddev = mddev; @@ -1946,33 +2827,43 @@ static conf_t *setup_conf(mddev_t *mddev) init_waitqueue_head(&conf->wait_barrier); bio_list_init(&conf->pending_bio_list); + conf->pending_count = 0; + conf->recovery_disabled = mddev->recovery_disabled - 1; - conf->last_used = -1; - for (i = 0; i < conf->raid_disks; i++) { + conf->start_next_window = MaxSector; + conf->current_window_requests = conf->next_window_requests = 0; + + err = -EIO; + for (i = 0; i < conf->raid_disks * 2; i++) { disk = conf->mirrors + i; + if (i < conf->raid_disks && + disk[conf->raid_disks].rdev) { + /* This slot has a replacement. */ + if (!disk->rdev) { + /* No original, just make the replacement + * a recovering spare + */ + disk->rdev = + disk[conf->raid_disks].rdev; + disk[conf->raid_disks].rdev = NULL; + } else if (!test_bit(In_sync, &disk->rdev->flags)) + /* Original is not in_sync - bad */ + goto abort; + } + if (!disk->rdev || !test_bit(In_sync, &disk->rdev->flags)) { disk->head_position = 0; - if (disk->rdev) + if (disk->rdev && + (disk->rdev->saved_raid_disk < 0)) conf->fullsync = 1; - } else if (conf->last_used < 0) - /* - * The first working device is used as a - * starting point to read balancing. - */ - conf->last_used = i; + } } - err = -EIO; - if (conf->last_used < 0) { - printk(KERN_ERR "md/raid1:%s: no operational mirrors\n", - mdname(mddev)); - goto abort; - } err = -ENOMEM; - conf->thread = md_register_thread(raid1d, mddev, NULL); + conf->thread = md_register_thread(raid1d, mddev, "raid1"); if (!conf->thread) { printk(KERN_ERR "md/raid1:%s: couldn't allocate thread\n", @@ -1994,11 +2885,14 @@ static conf_t *setup_conf(mddev_t *mddev) return ERR_PTR(err); } -static int run(mddev_t *mddev) +static int stop(struct mddev *mddev); +static int run(struct mddev *mddev) { - conf_t *conf; + struct r1conf *conf; int i; - mdk_rdev_t *rdev; + struct md_rdev *rdev; + int ret; + bool discard_supported = false; if (mddev->level != 1) { printk(KERN_ERR "md/raid1:%s: raid level not set to mirroring (%d)\n", @@ -2023,19 +2917,16 @@ static int run(mddev_t *mddev) if (IS_ERR(conf)) return PTR_ERR(conf); - mddev->queue->queue_lock = &conf->device_lock; - list_for_each_entry(rdev, &mddev->disks, same_set) { + if (mddev->queue) + blk_queue_max_write_same_sectors(mddev->queue, 0); + + rdev_for_each(rdev, mddev) { + if (!mddev->gendisk) + continue; disk_stack_limits(mddev->gendisk, rdev->bdev, rdev->data_offset << 9); - /* as we don't honour merge_bvec_fn, we must never risk - * violating it, so limit ->max_segments to 1 lying within - * a single page, as a one page request is never in violation. - */ - if (rdev->bdev->bd_disk->queue->merge_bvec_fn) { - blk_queue_max_segments(mddev->queue, 1); - blk_queue_segment_boundary(mddev->queue, - PAGE_CACHE_SIZE - 1); - } + if (blk_queue_discard(bdev_get_queue(rdev->bdev))) + discard_supported = true; } mddev->degraded = 0; @@ -2066,16 +2957,28 @@ static int run(mddev_t *mddev) md_set_array_sectors(mddev, raid1_size(mddev, 0, 0)); - mddev->queue->unplug_fn = raid1_unplug; - mddev->queue->backing_dev_info.congested_fn = raid1_congested; - mddev->queue->backing_dev_info.congested_data = mddev; - md_integrity_register(mddev); - return 0; + if (mddev->queue) { + mddev->queue->backing_dev_info.congested_fn = raid1_congested; + mddev->queue->backing_dev_info.congested_data = mddev; + blk_queue_merge_bvec(mddev->queue, raid1_mergeable_bvec); + + if (discard_supported) + queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, + mddev->queue); + else + queue_flag_clear_unlocked(QUEUE_FLAG_DISCARD, + mddev->queue); + } + + ret = md_integrity_register(mddev); + if (ret) + stop(mddev); + return ret; } -static int stop(mddev_t *mddev) +static int stop(struct mddev *mddev) { - conf_t *conf = mddev->private; + struct r1conf *conf = mddev->private; struct bitmap *bitmap = mddev->bitmap; /* wait for behind writes to complete */ @@ -2087,22 +2990,21 @@ static int stop(mddev_t *mddev) atomic_read(&bitmap->behind_writes) == 0); } - raise_barrier(conf); - lower_barrier(conf); + freeze_array(conf, 0); + unfreeze_array(conf); - md_unregister_thread(mddev->thread); - mddev->thread = NULL; - blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/ + md_unregister_thread(&mddev->thread); if (conf->r1bio_pool) mempool_destroy(conf->r1bio_pool); kfree(conf->mirrors); + safe_put_page(conf->tmppage); kfree(conf->poolinfo); kfree(conf); mddev->private = NULL; return 0; } -static int raid1_resize(mddev_t *mddev, sector_t sectors) +static int raid1_resize(struct mddev *mddev, sector_t sectors) { /* no resync is happening, and there is enough space * on all devices, so we can resize. @@ -2111,13 +3013,20 @@ static int raid1_resize(mddev_t *mddev, sector_t sectors) * any io in the removed space completes, but it hardly seems * worth it. */ - md_set_array_sectors(mddev, raid1_size(mddev, sectors, 0)); - if (mddev->array_sectors > raid1_size(mddev, sectors, 0)) + sector_t newsize = raid1_size(mddev, sectors, 0); + if (mddev->external_size && + mddev->array_sectors > newsize) return -EINVAL; + if (mddev->bitmap) { + int ret = bitmap_resize(mddev->bitmap, newsize, 0, 0); + if (ret) + return ret; + } + md_set_array_sectors(mddev, newsize); set_capacity(mddev->gendisk, mddev->array_sectors); revalidate_disk(mddev->gendisk); if (sectors > mddev->dev_sectors && - mddev->recovery_cp == MaxSector) { + mddev->recovery_cp > mddev->dev_sectors) { mddev->recovery_cp = mddev->dev_sectors; set_bit(MD_RECOVERY_NEEDED, &mddev->recovery); } @@ -2126,7 +3035,7 @@ static int raid1_resize(mddev_t *mddev, sector_t sectors) return 0; } -static int raid1_reshape(mddev_t *mddev) +static int raid1_reshape(struct mddev *mddev) { /* We need to: * 1/ resize the r1bio_pool @@ -2141,8 +3050,8 @@ static int raid1_reshape(mddev_t *mddev) */ mempool_t *newpool, *oldpool; struct pool_info *newpoolinfo; - mirror_info_t *newmirrors; - conf_t *conf = mddev->private; + struct raid1_info *newmirrors; + struct r1conf *conf = mddev->private; int cnt, raid_disks; unsigned long flags; int d, d2, err; @@ -2176,7 +3085,7 @@ static int raid1_reshape(mddev_t *mddev) if (!newpoolinfo) return -ENOMEM; newpoolinfo->mddev = mddev; - newpoolinfo->raid_disks = raid_disks; + newpoolinfo->raid_disks = raid_disks * 2; newpool = mempool_create(NR_RAID1_BIOS, r1bio_pool_alloc, r1bio_pool_free, newpoolinfo); @@ -2184,34 +3093,30 @@ static int raid1_reshape(mddev_t *mddev) kfree(newpoolinfo); return -ENOMEM; } - newmirrors = kzalloc(sizeof(struct mirror_info) * raid_disks, GFP_KERNEL); + newmirrors = kzalloc(sizeof(struct raid1_info) * raid_disks * 2, + GFP_KERNEL); if (!newmirrors) { kfree(newpoolinfo); mempool_destroy(newpool); return -ENOMEM; } - raise_barrier(conf); + freeze_array(conf, 0); /* ok, everything is stopped */ oldpool = conf->r1bio_pool; conf->r1bio_pool = newpool; for (d = d2 = 0; d < conf->raid_disks; d++) { - mdk_rdev_t *rdev = conf->mirrors[d].rdev; + struct md_rdev *rdev = conf->mirrors[d].rdev; if (rdev && rdev->raid_disk != d2) { - char nm[20]; - sprintf(nm, "rd%d", rdev->raid_disk); - sysfs_remove_link(&mddev->kobj, nm); + sysfs_unlink_rdev(mddev, rdev); rdev->raid_disk = d2; - sprintf(nm, "rd%d", rdev->raid_disk); - sysfs_remove_link(&mddev->kobj, nm); - if (sysfs_create_link(&mddev->kobj, - &rdev->kobj, nm)) + sysfs_unlink_rdev(mddev, rdev); + if (sysfs_link_rdev(mddev, rdev)) printk(KERN_WARNING - "md/raid1:%s: cannot register " - "%s\n", - mdname(mddev), nm); + "md/raid1:%s: cannot register rd%d\n", + mdname(mddev), rdev->raid_disk); } if (rdev) newmirrors[d2++].rdev = rdev; @@ -2227,8 +3132,7 @@ static int raid1_reshape(mddev_t *mddev) conf->raid_disks = mddev->raid_disks = raid_disks; mddev->delta_disks = 0; - conf->last_used = 0; /* just make sure it is in-range */ - lower_barrier(conf); + unfreeze_array(conf); set_bit(MD_RECOVERY_NEEDED, &mddev->recovery); md_wakeup_thread(mddev->thread); @@ -2237,42 +3141,43 @@ static int raid1_reshape(mddev_t *mddev) return 0; } -static void raid1_quiesce(mddev_t *mddev, int state) +static void raid1_quiesce(struct mddev *mddev, int state) { - conf_t *conf = mddev->private; + struct r1conf *conf = mddev->private; switch(state) { case 2: /* wake for suspend */ wake_up(&conf->wait_barrier); break; case 1: - raise_barrier(conf); + freeze_array(conf, 0); break; case 0: - lower_barrier(conf); + unfreeze_array(conf); break; } } -static void *raid1_takeover(mddev_t *mddev) +static void *raid1_takeover(struct mddev *mddev) { /* raid1 can take over: * raid5 with 2 devices, any layout or chunk size */ if (mddev->level == 5 && mddev->raid_disks == 2) { - conf_t *conf; + struct r1conf *conf; mddev->new_level = 1; mddev->new_layout = 0; mddev->new_chunk_sectors = 0; conf = setup_conf(mddev); if (!IS_ERR(conf)) - conf->barrier = 1; + /* Array must appear to be quiesced */ + conf->array_frozen = 1; return conf; } return ERR_PTR(-EINVAL); } -static struct mdk_personality raid1_personality = +static struct md_personality raid1_personality = { .name = "raid1", .level = 1, @@ -2310,3 +3215,5 @@ MODULE_DESCRIPTION("RAID1 (mirroring) personality for MD"); MODULE_ALIAS("md-personality-3"); /* RAID1 */ MODULE_ALIAS("md-raid1"); MODULE_ALIAS("md-level-1"); + +module_param(max_queued_requests, int, S_IRUGO|S_IWUSR); |