X-Git-Url: http://ftp.safe.ca/?a=blobdiff_plain;f=drivers%2Fmd%2Fraid5.c;h=15348c393b5daeb59ad11a9050f66eef428b4314;hb=4434ade8c9334a3ab975d8993de456f06841899e;hp=54ef8d75541d3cc8607e37b0a128305390f3dff5;hpb=0403e3827788d878163f9ef0541b748b0f88ca5d;p=safe%2Fjmp%2Flinux-2.6 diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index 54ef8d7..15348c3 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -50,6 +50,7 @@ #include #include #include +#include #include "md.h" #include "raid5.h" #include "bitmap.h" @@ -156,13 +157,16 @@ static inline int raid6_next_disk(int disk, int raid_disks) static int raid6_idx_to_slot(int idx, struct stripe_head *sh, int *count, int syndrome_disks) { - int slot; + int slot = *count; + if (sh->ddf_layout) + (*count)++; if (idx == sh->pd_idx) return syndrome_disks; if (idx == sh->qd_idx) return syndrome_disks + 1; - slot = (*count)++; + if (!sh->ddf_layout) + (*count)++; return slot; } @@ -717,7 +721,7 @@ static int set_syndrome_sources(struct page **srcs, struct stripe_head *sh) int i; for (i = 0; i < disks; i++) - srcs[i] = (void *)raid6_empty_zero_page; + srcs[i] = NULL; count = 0; i = d0_idx; @@ -727,9 +731,8 @@ static int set_syndrome_sources(struct page **srcs, struct stripe_head *sh) srcs[slot] = sh->dev[i].page; i = raid6_next_disk(i, disks); } while (i != d0_idx); - BUG_ON(count != syndrome_disks); - return count; + return syndrome_disks; } static struct dma_async_tx_descriptor * @@ -810,11 +813,11 @@ ops_run_compute6_2(struct stripe_head *sh, struct raid5_percpu *percpu) BUG_ON(!test_bit(R5_Wantcompute, &tgt->flags)); BUG_ON(!test_bit(R5_Wantcompute, &tgt2->flags)); - /* we need to open-code set_syndrome_sources to handle to the + /* we need to open-code set_syndrome_sources to handle the * slot number conversion for 'faila' and 'failb' */ for (i = 0; i < disks ; i++) - blocks[i] = (void *)raid6_empty_zero_page; + blocks[i] = NULL; count = 0; i = d0_idx; do { @@ -828,7 +831,6 @@ ops_run_compute6_2(struct stripe_head *sh, struct raid5_percpu *percpu) failb = slot; i = raid6_next_disk(i, disks); } while (i != d0_idx); - BUG_ON(count != syndrome_disks); BUG_ON(faila == failb); if (failb < faila) @@ -845,7 +847,7 @@ ops_run_compute6_2(struct stripe_head *sh, struct raid5_percpu *percpu) init_async_submit(&submit, ASYNC_TX_FENCE, NULL, ops_complete_compute, sh, to_addr_conv(sh, percpu)); - return async_gen_syndrome(blocks, 0, count+2, + return async_gen_syndrome(blocks, 0, syndrome_disks+2, STRIPE_SIZE, &submit); } else { struct page *dest; @@ -879,18 +881,21 @@ ops_run_compute6_2(struct stripe_head *sh, struct raid5_percpu *percpu) return async_gen_syndrome(blocks, 0, count+2, STRIPE_SIZE, &submit); } - } - - init_async_submit(&submit, ASYNC_TX_FENCE, NULL, ops_complete_compute, - sh, to_addr_conv(sh, percpu)); - if (failb == syndrome_disks) { - /* We're missing D+P. */ - return async_raid6_datap_recov(syndrome_disks+2, STRIPE_SIZE, - faila, blocks, &submit); } else { - /* We're missing D+D. */ - return async_raid6_2data_recov(syndrome_disks+2, STRIPE_SIZE, - faila, failb, blocks, &submit); + init_async_submit(&submit, ASYNC_TX_FENCE, NULL, + ops_complete_compute, sh, + to_addr_conv(sh, percpu)); + if (failb == syndrome_disks) { + /* We're missing D+P. */ + return async_raid6_datap_recov(syndrome_disks+2, + STRIPE_SIZE, faila, + blocks, &submit); + } else { + /* We're missing D+D. */ + return async_raid6_2data_recov(syndrome_disks+2, + STRIPE_SIZE, faila, failb, + blocks, &submit); + } } } @@ -1136,7 +1141,7 @@ static void ops_run_check_pq(struct stripe_head *sh, struct raid5_percpu *percpu &sh->ops.zero_sum_result, percpu->spare_page, &submit); } -static void raid_run_ops(struct stripe_head *sh, unsigned long ops_request) +static void __raid_run_ops(struct stripe_head *sh, unsigned long ops_request) { int overlap_clear = 0, i, disks = sh->disks; struct dma_async_tx_descriptor *tx = NULL; @@ -1201,22 +1206,55 @@ static void raid_run_ops(struct stripe_head *sh, unsigned long ops_request) put_cpu(); } +#ifdef CONFIG_MULTICORE_RAID456 +static void async_run_ops(void *param, async_cookie_t cookie) +{ + struct stripe_head *sh = param; + unsigned long ops_request = sh->ops.request; + + clear_bit_unlock(STRIPE_OPS_REQ_PENDING, &sh->state); + wake_up(&sh->ops.wait_for_ops); + + __raid_run_ops(sh, ops_request); + release_stripe(sh); +} + +static void raid_run_ops(struct stripe_head *sh, unsigned long ops_request) +{ + /* since handle_stripe can be called outside of raid5d context + * we need to ensure sh->ops.request is de-staged before another + * request arrives + */ + wait_event(sh->ops.wait_for_ops, + !test_and_set_bit_lock(STRIPE_OPS_REQ_PENDING, &sh->state)); + sh->ops.request = ops_request; + + atomic_inc(&sh->count); + async_schedule(async_run_ops, sh); +} +#else +#define raid_run_ops __raid_run_ops +#endif + static int grow_one_stripe(raid5_conf_t *conf) { struct stripe_head *sh; + int disks = max(conf->raid_disks, conf->previous_raid_disks); sh = kmem_cache_alloc(conf->slab_cache, GFP_KERNEL); if (!sh) return 0; - memset(sh, 0, sizeof(*sh) + (conf->raid_disks-1)*sizeof(struct r5dev)); + memset(sh, 0, sizeof(*sh) + (disks-1)*sizeof(struct r5dev)); sh->raid_conf = conf; spin_lock_init(&sh->lock); + #ifdef CONFIG_MULTICORE_RAID456 + init_waitqueue_head(&sh->ops.wait_for_ops); + #endif - if (grow_buffers(sh, conf->raid_disks)) { - shrink_buffers(sh, conf->raid_disks); + if (grow_buffers(sh, disks)) { + shrink_buffers(sh, disks); kmem_cache_free(conf->slab_cache, sh); return 0; } - sh->disks = conf->raid_disks; /* we just created an active stripe so... */ atomic_set(&sh->count, 1); atomic_inc(&conf->active_stripes); @@ -1228,7 +1266,7 @@ static int grow_one_stripe(raid5_conf_t *conf) static int grow_stripes(raid5_conf_t *conf, int num) { struct kmem_cache *sc; - int devs = conf->raid_disks; + int devs = max(conf->raid_disks, conf->previous_raid_disks); sprintf(conf->cache_name[0], "raid%d-%s", conf->level, mdname(conf->mddev)); @@ -1326,6 +1364,9 @@ static int resize_stripes(raid5_conf_t *conf, int newsize) nsh->raid_conf = conf; spin_lock_init(&nsh->lock); + #ifdef CONFIG_MULTICORE_RAID456 + init_waitqueue_head(&nsh->ops.wait_for_ops); + #endif list_add(&nsh->lru, &newstripes); } @@ -1486,7 +1527,7 @@ static void raid5_end_read_request(struct bio * bi, int error) clear_bit(R5_UPTODATE, &sh->dev[i].flags); atomic_inc(&rdev->read_errors); - if (conf->mddev->degraded) + if (conf->mddev->degraded >= conf->max_degraded) printk_rl(KERN_WARNING "raid5:%s: read error not correctable " "(sector %llu on %s).\n", @@ -1609,16 +1650,16 @@ static sector_t raid5_compute_sector(raid5_conf_t *conf, sector_t r_sector, int previous, int *dd_idx, struct stripe_head *sh) { - long stripe; - unsigned long chunk_number; + sector_t stripe, stripe2; + sector_t chunk_number; unsigned int chunk_offset; int pd_idx, qd_idx; int ddf_layout = 0; sector_t new_sector; int algorithm = previous ? conf->prev_algo : conf->algorithm; - int sectors_per_chunk = previous ? (conf->prev_chunk >> 9) - : (conf->chunk_size >> 9); + int sectors_per_chunk = previous ? conf->prev_chunk_sectors + : conf->chunk_sectors; int raid_disks = previous ? conf->previous_raid_disks : conf->raid_disks; int data_disks = raid_disks - conf->max_degraded; @@ -1630,18 +1671,13 @@ static sector_t raid5_compute_sector(raid5_conf_t *conf, sector_t r_sector, */ chunk_offset = sector_div(r_sector, sectors_per_chunk); chunk_number = r_sector; - BUG_ON(r_sector != chunk_number); /* * Compute the stripe number */ - stripe = chunk_number / data_disks; - - /* - * Compute the data disk and parity disk indexes inside the stripe - */ - *dd_idx = chunk_number % data_disks; - + stripe = chunk_number; + *dd_idx = sector_div(stripe, data_disks); + stripe2 = stripe; /* * Select the parity disk based on the user selected algorithm. */ @@ -1653,21 +1689,21 @@ static sector_t raid5_compute_sector(raid5_conf_t *conf, sector_t r_sector, case 5: switch (algorithm) { case ALGORITHM_LEFT_ASYMMETRIC: - pd_idx = data_disks - stripe % raid_disks; + pd_idx = data_disks - sector_div(stripe2, raid_disks); if (*dd_idx >= pd_idx) (*dd_idx)++; break; case ALGORITHM_RIGHT_ASYMMETRIC: - pd_idx = stripe % raid_disks; + pd_idx = sector_div(stripe2, raid_disks); if (*dd_idx >= pd_idx) (*dd_idx)++; break; case ALGORITHM_LEFT_SYMMETRIC: - pd_idx = data_disks - stripe % raid_disks; + pd_idx = data_disks - sector_div(stripe2, raid_disks); *dd_idx = (pd_idx + 1 + *dd_idx) % raid_disks; break; case ALGORITHM_RIGHT_SYMMETRIC: - pd_idx = stripe % raid_disks; + pd_idx = sector_div(stripe2, raid_disks); *dd_idx = (pd_idx + 1 + *dd_idx) % raid_disks; break; case ALGORITHM_PARITY_0: @@ -1687,7 +1723,7 @@ static sector_t raid5_compute_sector(raid5_conf_t *conf, sector_t r_sector, switch (algorithm) { case ALGORITHM_LEFT_ASYMMETRIC: - pd_idx = raid_disks - 1 - (stripe % raid_disks); + pd_idx = raid_disks - 1 - sector_div(stripe2, raid_disks); qd_idx = pd_idx + 1; if (pd_idx == raid_disks-1) { (*dd_idx)++; /* Q D D D P */ @@ -1696,7 +1732,7 @@ static sector_t raid5_compute_sector(raid5_conf_t *conf, sector_t r_sector, (*dd_idx) += 2; /* D D P Q D */ break; case ALGORITHM_RIGHT_ASYMMETRIC: - pd_idx = stripe % raid_disks; + pd_idx = sector_div(stripe2, raid_disks); qd_idx = pd_idx + 1; if (pd_idx == raid_disks-1) { (*dd_idx)++; /* Q D D D P */ @@ -1705,12 +1741,12 @@ static sector_t raid5_compute_sector(raid5_conf_t *conf, sector_t r_sector, (*dd_idx) += 2; /* D D P Q D */ break; case ALGORITHM_LEFT_SYMMETRIC: - pd_idx = raid_disks - 1 - (stripe % raid_disks); + pd_idx = raid_disks - 1 - sector_div(stripe2, raid_disks); qd_idx = (pd_idx + 1) % raid_disks; *dd_idx = (pd_idx + 2 + *dd_idx) % raid_disks; break; case ALGORITHM_RIGHT_SYMMETRIC: - pd_idx = stripe % raid_disks; + pd_idx = sector_div(stripe2, raid_disks); qd_idx = (pd_idx + 1) % raid_disks; *dd_idx = (pd_idx + 2 + *dd_idx) % raid_disks; break; @@ -1729,7 +1765,7 @@ static sector_t raid5_compute_sector(raid5_conf_t *conf, sector_t r_sector, /* Exactly the same as RIGHT_ASYMMETRIC, but or * of blocks for computing Q is different. */ - pd_idx = stripe % raid_disks; + pd_idx = sector_div(stripe2, raid_disks); qd_idx = pd_idx + 1; if (pd_idx == raid_disks-1) { (*dd_idx)++; /* Q D D D P */ @@ -1744,7 +1780,8 @@ static sector_t raid5_compute_sector(raid5_conf_t *conf, sector_t r_sector, * D D D P Q rather than * Q D D D P */ - pd_idx = raid_disks - 1 - ((stripe + 1) % raid_disks); + stripe2 += 1; + pd_idx = raid_disks - 1 - sector_div(stripe2, raid_disks); qd_idx = pd_idx + 1; if (pd_idx == raid_disks-1) { (*dd_idx)++; /* Q D D D P */ @@ -1756,7 +1793,7 @@ static sector_t raid5_compute_sector(raid5_conf_t *conf, sector_t r_sector, case ALGORITHM_ROTATING_N_CONTINUE: /* Same as left_symmetric but Q is before P */ - pd_idx = raid_disks - 1 - (stripe % raid_disks); + pd_idx = raid_disks - 1 - sector_div(stripe2, raid_disks); qd_idx = (pd_idx + raid_disks - 1) % raid_disks; *dd_idx = (pd_idx + 1 + *dd_idx) % raid_disks; ddf_layout = 1; @@ -1764,27 +1801,27 @@ static sector_t raid5_compute_sector(raid5_conf_t *conf, sector_t r_sector, case ALGORITHM_LEFT_ASYMMETRIC_6: /* RAID5 left_asymmetric, with Q on last device */ - pd_idx = data_disks - stripe % (raid_disks-1); + pd_idx = data_disks - sector_div(stripe2, raid_disks-1); if (*dd_idx >= pd_idx) (*dd_idx)++; qd_idx = raid_disks - 1; break; case ALGORITHM_RIGHT_ASYMMETRIC_6: - pd_idx = stripe % (raid_disks-1); + pd_idx = sector_div(stripe2, raid_disks-1); if (*dd_idx >= pd_idx) (*dd_idx)++; qd_idx = raid_disks - 1; break; case ALGORITHM_LEFT_SYMMETRIC_6: - pd_idx = data_disks - stripe % (raid_disks-1); + pd_idx = data_disks - sector_div(stripe2, raid_disks-1); *dd_idx = (pd_idx + 1 + *dd_idx) % (raid_disks-1); qd_idx = raid_disks - 1; break; case ALGORITHM_RIGHT_SYMMETRIC_6: - pd_idx = stripe % (raid_disks-1); + pd_idx = sector_div(stripe2, raid_disks-1); *dd_idx = (pd_idx + 1 + *dd_idx) % (raid_disks-1); qd_idx = raid_disks - 1; break; @@ -1823,20 +1860,20 @@ static sector_t compute_blocknr(struct stripe_head *sh, int i, int previous) int raid_disks = sh->disks; int data_disks = raid_disks - conf->max_degraded; sector_t new_sector = sh->sector, check; - int sectors_per_chunk = previous ? (conf->prev_chunk >> 9) - : (conf->chunk_size >> 9); + int sectors_per_chunk = previous ? conf->prev_chunk_sectors + : conf->chunk_sectors; int algorithm = previous ? conf->prev_algo : conf->algorithm; sector_t stripe; int chunk_offset; - int chunk_number, dummy1, dd_idx = i; + sector_t chunk_number; + int dummy1, dd_idx = i; sector_t r_sector; struct stripe_head sh2; chunk_offset = sector_div(new_sector, sectors_per_chunk); stripe = new_sector; - BUG_ON(new_sector != stripe); if (i == sh->pd_idx) return 0; @@ -1896,10 +1933,15 @@ static sector_t compute_blocknr(struct stripe_head *sh, int i, int previous) case ALGORITHM_PARITY_N: break; case ALGORITHM_ROTATING_N_CONTINUE: + /* Like left_symmetric, but P is before Q */ if (sh->pd_idx == 0) i--; /* P D D D Q */ - else if (i > sh->pd_idx) - i -= 2; /* D D Q P D */ + else { + /* D D Q P D */ + if (i < sh->pd_idx) + i += raid_disks; + i -= (sh->pd_idx + 1); + } break; case ALGORITHM_LEFT_ASYMMETRIC_6: case ALGORITHM_RIGHT_ASYMMETRIC_6: @@ -1924,7 +1966,7 @@ static sector_t compute_blocknr(struct stripe_head *sh, int i, int previous) } chunk_number = stripe * data_disks + i; - r_sector = (sector_t)chunk_number * sectors_per_chunk + chunk_offset; + r_sector = chunk_number * sectors_per_chunk + chunk_offset; check = raid5_compute_sector(conf, r_sector, previous, &dummy1, &sh2); @@ -2098,8 +2140,7 @@ static void stripe_set_idx(sector_t stripe, raid5_conf_t *conf, int previous, struct stripe_head *sh) { int sectors_per_chunk = - previous ? (conf->prev_chunk >> 9) - : (conf->chunk_size >> 9); + previous ? conf->prev_chunk_sectors : conf->chunk_sectors; int dd_idx; int chunk_offset = sector_div(stripe, sectors_per_chunk); int disks = previous ? conf->previous_raid_disks : conf->raid_disks; @@ -2894,7 +2935,7 @@ static void handle_stripe_expansion(raid5_conf_t *conf, struct stripe_head *sh, * */ -static bool handle_stripe5(struct stripe_head *sh) +static void handle_stripe5(struct stripe_head *sh) { raid5_conf_t *conf = sh->raid_conf; int disks = sh->disks, i; @@ -2903,6 +2944,7 @@ static bool handle_stripe5(struct stripe_head *sh) struct r5dev *dev; mdk_rdev_t *blocked_rdev = NULL; int prexor; + int dec_preread_active = 0; memset(&s, 0, sizeof(s)); pr_debug("handling stripe %llu, state=%#lx cnt=%d, pd_idx=%d check:%d " @@ -2922,7 +2964,8 @@ static bool handle_stripe5(struct stripe_head *sh) rcu_read_lock(); for (i=disks; i--; ) { mdk_rdev_t *rdev; - struct r5dev *dev = &sh->dev[i]; + + dev = &sh->dev[i]; clear_bit(R5_Insync, &dev->flags); pr_debug("check %d: state 0x%lx toread %p read %p write %p " @@ -3051,12 +3094,8 @@ static bool handle_stripe5(struct stripe_head *sh) set_bit(STRIPE_INSYNC, &sh->state); } } - if (test_and_clear_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) { - atomic_dec(&conf->preread_active_stripes); - if (atomic_read(&conf->preread_active_stripes) < - IO_THRESHOLD) - md_wakeup_thread(conf->mddev->thread); - } + if (test_and_clear_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) + dec_preread_active = 1; } /* Now to consider new write requests and what else, if anything @@ -3163,12 +3202,20 @@ static bool handle_stripe5(struct stripe_head *sh) ops_run_io(sh, &s); + if (dec_preread_active) { + /* We delay this until after ops_run_io so that if make_request + * is waiting on a barrier, it won't continue until the writes + * have actually been submitted. + */ + atomic_dec(&conf->preread_active_stripes); + if (atomic_read(&conf->preread_active_stripes) < + IO_THRESHOLD) + md_wakeup_thread(conf->mddev->thread); + } return_io(return_bi); - - return blocked_rdev == NULL; } -static bool handle_stripe6(struct stripe_head *sh) +static void handle_stripe6(struct stripe_head *sh) { raid5_conf_t *conf = sh->raid_conf; int disks = sh->disks; @@ -3178,6 +3225,7 @@ static bool handle_stripe6(struct stripe_head *sh) struct r6_state r6s; struct r5dev *dev, *pdev, *qdev; mdk_rdev_t *blocked_rdev = NULL; + int dec_preread_active = 0; pr_debug("handling stripe %llu, state=%#lx cnt=%d, " "pd_idx=%d, qd_idx=%d\n, check:%d, reconstruct:%d\n", @@ -3215,8 +3263,10 @@ static bool handle_stripe6(struct stripe_head *sh) /* now count some things */ if (test_bit(R5_LOCKED, &dev->flags)) s.locked++; if (test_bit(R5_UPTODATE, &dev->flags)) s.uptodate++; - if (test_bit(R5_Wantcompute, &dev->flags)) - BUG_ON(++s.compute > 2); + if (test_bit(R5_Wantcompute, &dev->flags)) { + s.compute++; + BUG_ON(s.compute > 2); + } if (test_bit(R5_Wantfill, &dev->flags)) { s.to_fill++; @@ -3313,7 +3363,6 @@ static bool handle_stripe6(struct stripe_head *sh) * completed */ if (sh->reconstruct_state == reconstruct_state_drain_result) { - int qd_idx = sh->qd_idx; sh->reconstruct_state = reconstruct_state_idle; /* All the 'written' buffers and the parity blocks are ready to @@ -3335,12 +3384,8 @@ static bool handle_stripe6(struct stripe_head *sh) set_bit(STRIPE_INSYNC, &sh->state); } } - if (test_and_clear_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) { - atomic_dec(&conf->preread_active_stripes); - if (atomic_read(&conf->preread_active_stripes) < - IO_THRESHOLD) - md_wakeup_thread(conf->mddev->thread); - } + if (test_and_clear_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) + dec_preread_active = 1; } /* Now to consider new write requests and what else, if anything @@ -3449,18 +3494,27 @@ static bool handle_stripe6(struct stripe_head *sh) ops_run_io(sh, &s); - return_io(return_bi); - return blocked_rdev == NULL; + if (dec_preread_active) { + /* We delay this until after ops_run_io so that if make_request + * is waiting on a barrier, it won't continue until the writes + * have actually been submitted. + */ + atomic_dec(&conf->preread_active_stripes); + if (atomic_read(&conf->preread_active_stripes) < + IO_THRESHOLD) + md_wakeup_thread(conf->mddev->thread); + } + + return_io(return_bi); } -/* returns true if the stripe was handled */ -static bool handle_stripe(struct stripe_head *sh) +static void handle_stripe(struct stripe_head *sh) { if (sh->raid_conf->level == 6) - return handle_stripe6(sh); + handle_stripe6(sh); else - return handle_stripe5(sh); + handle_stripe5(sh); } static void raid5_activate_delayed(raid5_conf_t *conf) @@ -3496,11 +3550,12 @@ static void activate_bit_delay(raid5_conf_t *conf) static void unplug_slaves(mddev_t *mddev) { - raid5_conf_t *conf = mddev_to_conf(mddev); + raid5_conf_t *conf = mddev->private; int i; + int devs = max(conf->raid_disks, conf->previous_raid_disks); rcu_read_lock(); - for (i = 0; i < conf->raid_disks; i++) { + for (i = 0; i < devs; i++) { mdk_rdev_t *rdev = rcu_dereference(conf->disks[i].rdev); if (rdev && !test_bit(Faulty, &rdev->flags) && atomic_read(&rdev->nr_pending)) { struct request_queue *r_queue = bdev_get_queue(rdev->bdev); @@ -3520,7 +3575,7 @@ static void unplug_slaves(mddev_t *mddev) static void raid5_unplug_device(struct request_queue *q) { mddev_t *mddev = q->queuedata; - raid5_conf_t *conf = mddev_to_conf(mddev); + raid5_conf_t *conf = mddev->private; unsigned long flags; spin_lock_irqsave(&conf->device_lock, flags); @@ -3539,11 +3594,14 @@ static void raid5_unplug_device(struct request_queue *q) static int raid5_congested(void *data, int bits) { mddev_t *mddev = data; - raid5_conf_t *conf = mddev_to_conf(mddev); + raid5_conf_t *conf = mddev->private; /* No difference between reads and writes. Just check * how busy the stripe_cache is */ + + if (mddev_congested(mddev, bits)) + return 1; if (conf->inactive_blocked) return 1; if (conf->quiesce) @@ -3564,14 +3622,14 @@ static int raid5_mergeable_bvec(struct request_queue *q, mddev_t *mddev = q->queuedata; sector_t sector = bvm->bi_sector + get_start_sect(bvm->bi_bdev); int max; - unsigned int chunk_sectors = mddev->chunk_size >> 9; + unsigned int chunk_sectors = mddev->chunk_sectors; unsigned int bio_sectors = bvm->bi_size >> 9; if ((bvm->bi_rw & 1) == WRITE) return biovec->bv_len; /* always allow writes to be mergeable */ - if (mddev->new_chunk < mddev->chunk_size) - chunk_sectors = mddev->new_chunk >> 9; + if (mddev->new_chunk_sectors < mddev->chunk_sectors) + chunk_sectors = mddev->new_chunk_sectors; max = (chunk_sectors - ((sector & (chunk_sectors - 1)) + bio_sectors)) << 9; if (max < 0) max = 0; if (max <= biovec->bv_len && bio_sectors == 0) @@ -3584,11 +3642,11 @@ static int raid5_mergeable_bvec(struct request_queue *q, static int in_chunk_boundary(mddev_t *mddev, struct bio *bio) { sector_t sector = bio->bi_sector + get_start_sect(bio->bi_bdev); - unsigned int chunk_sectors = mddev->chunk_size >> 9; + unsigned int chunk_sectors = mddev->chunk_sectors; unsigned int bio_sectors = bio->bi_size >> 9; - if (mddev->new_chunk < mddev->chunk_size) - chunk_sectors = mddev->new_chunk >> 9; + if (mddev->new_chunk_sectors < mddev->chunk_sectors) + chunk_sectors = mddev->new_chunk_sectors; return chunk_sectors >= ((sector & (chunk_sectors - 1)) + bio_sectors); } @@ -3652,7 +3710,7 @@ static void raid5_align_endio(struct bio *bi, int error) bio_put(bi); mddev = raid_bi->bi_bdev->bd_disk->queue->queuedata; - conf = mddev_to_conf(mddev); + conf = mddev->private; rdev = (void*)raid_bi->bi_next; raid_bi->bi_next = NULL; @@ -3675,10 +3733,10 @@ static int bio_fits_rdev(struct bio *bi) { struct request_queue *q = bdev_get_queue(bi->bi_bdev); - if ((bi->bi_size>>9) > q->max_sectors) + if ((bi->bi_size>>9) > queue_max_sectors(q)) return 0; blk_recount_segments(q, bi); - if (bi->bi_phys_segments > q->max_phys_segments) + if (bi->bi_phys_segments > queue_max_segments(q)) return 0; if (q->merge_bvec_fn) @@ -3694,8 +3752,8 @@ static int bio_fits_rdev(struct bio *bi) static int chunk_aligned_read(struct request_queue *q, struct bio * raid_bio) { mddev_t *mddev = q->queuedata; - raid5_conf_t *conf = mddev_to_conf(mddev); - unsigned int dd_idx; + raid5_conf_t *conf = mddev->private; + int dd_idx; struct bio* align_bi; mdk_rdev_t *rdev; @@ -3811,7 +3869,7 @@ static struct stripe_head *__get_priority_stripe(raid5_conf_t *conf) static int make_request(struct request_queue *q, struct bio * bi) { mddev_t *mddev = q->queuedata; - raid5_conf_t *conf = mddev_to_conf(mddev); + raid5_conf_t *conf = mddev->private; int dd_idx; sector_t new_sector; sector_t logical_sector, last_sector; @@ -3819,8 +3877,14 @@ static int make_request(struct request_queue *q, struct bio * bi) const int rw = bio_data_dir(bi); int cpu, remaining; - if (unlikely(bio_barrier(bi))) { - bio_endio(bi, -EOPNOTSUPP); + if (unlikely(bio_rw_flagged(bi, BIO_RW_BARRIER))) { + /* Drain all pending writes. We only really need + * to ensure they have been submitted, but this is + * easier. + */ + mddev->pers->quiesce(mddev, 1); + mddev->pers->quiesce(mddev, 0); + md_barrier_request(mddev, bi); return 0; } @@ -3908,16 +3972,25 @@ static int make_request(struct request_queue *q, struct bio * bi) spin_unlock_irq(&conf->device_lock); if (must_retry) { release_stripe(sh); + schedule(); goto retry; } } - /* FIXME what if we get a false positive because these - * are being updated. - */ - if (logical_sector >= mddev->suspend_lo && + + if (bio_data_dir(bi) == WRITE && + logical_sector >= mddev->suspend_lo && logical_sector < mddev->suspend_hi) { release_stripe(sh); - schedule(); + /* As the suspend_* range is controlled by + * userspace, we want an interruptible + * wait. + */ + flush_signals(current); + prepare_to_wait(&conf->wait_for_overlap, + &w, TASK_INTERRUPTIBLE); + if (logical_sector >= mddev->suspend_lo && + logical_sector < mddev->suspend_hi) + schedule(); goto retry; } @@ -3935,6 +4008,9 @@ static int make_request(struct request_queue *q, struct bio * bi) finish_wait(&conf->wait_for_overlap, &w); set_bit(STRIPE_HANDLE, &sh->state); clear_bit(STRIPE_DELAYED, &sh->state); + if (mddev->barrier && + !test_and_set_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) + atomic_inc(&conf->preread_active_stripes); release_stripe(sh); } else { /* cannot get stripe for read-ahead, just give-up */ @@ -3954,6 +4030,14 @@ static int make_request(struct request_queue *q, struct bio * bi) bio_endio(bi, 0); } + + if (mddev->barrier) { + /* We need to wait for the stripes to all be handled. + * So: wait for preread_active_stripes to drop to 0. + */ + wait_event(mddev->thread->wqueue, + atomic_read(&conf->preread_active_stripes) == 0); + } return 0; } @@ -3989,11 +4073,13 @@ static sector_t reshape_request(mddev_t *mddev, sector_t sector_nr, int *skipped conf->reshape_progress < raid5_size(mddev, 0, 0)) { sector_nr = raid5_size(mddev, 0, 0) - conf->reshape_progress; - } else if (mddev->delta_disks > 0 && + } else if (mddev->delta_disks >= 0 && conf->reshape_progress > 0) sector_nr = conf->reshape_progress; sector_div(sector_nr, new_data_disks); if (sector_nr) { + mddev->curr_resync_completed = sector_nr; + sysfs_notify(&mddev->kobj, NULL, "sync_completed"); *skipped = 1; return sector_nr; } @@ -4003,10 +4089,10 @@ static sector_t reshape_request(mddev_t *mddev, sector_t sector_nr, int *skipped * If old and new chunk sizes differ, we need to process the * largest of these */ - if (mddev->new_chunk > mddev->chunk_size) - reshape_sectors = mddev->new_chunk / 512; + if (mddev->new_chunk_sectors > mddev->chunk_sectors) + reshape_sectors = mddev->new_chunk_sectors; else - reshape_sectors = mddev->chunk_size / 512; + reshape_sectors = mddev->chunk_sectors; /* we update the metadata when there is more than 3Meg * in the block range (that is rather arbitrary, should @@ -4084,7 +4170,7 @@ static sector_t reshape_request(mddev_t *mddev, sector_t sector_nr, int *skipped INIT_LIST_HEAD(&stripes); for (i = 0; i < reshape_sectors; i += STRIPE_SECTORS) { int j; - int skipped = 0; + int skipped_disk = 0; sh = get_active_stripe(conf, stripe_addr+i, 0, 0, 1); set_bit(STRIPE_EXPANDING, &sh->state); atomic_inc(&conf->reshape_stripes); @@ -4100,14 +4186,14 @@ static sector_t reshape_request(mddev_t *mddev, sector_t sector_nr, int *skipped continue; s = compute_blocknr(sh, j, 0); if (s < raid5_size(mddev, 0, 0)) { - skipped = 1; + skipped_disk = 1; continue; } memset(page_address(sh->dev[j].page), 0, STRIPE_SIZE); set_bit(R5_Expanded, &sh->dev[j].flags); set_bit(R5_UPTODATE, &sh->dev[j].flags); } - if (!skipped) { + if (!skipped_disk) { set_bit(STRIPE_EXPAND_READY, &sh->state); set_bit(STRIPE_HANDLE, &sh->state); } @@ -4129,7 +4215,7 @@ static sector_t reshape_request(mddev_t *mddev, sector_t sector_nr, int *skipped 1, &dd_idx, NULL); last_sector = raid5_compute_sector(conf, ((stripe_addr+reshape_sectors) - *(new_data_disks) - 1), + * new_data_disks - 1), 1, &dd_idx, NULL); if (last_sector >= mddev->dev_sectors) last_sector = mddev->dev_sectors - 1; @@ -4158,7 +4244,7 @@ static sector_t reshape_request(mddev_t *mddev, sector_t sector_nr, int *skipped wait_event(conf->wait_for_overlap, atomic_read(&conf->reshape_stripes) == 0); mddev->reshape_position = conf->reshape_progress; - mddev->curr_resync_completed = mddev->curr_resync; + mddev->curr_resync_completed = mddev->curr_resync + reshape_sectors; conf->reshape_checkpoint = jiffies; set_bit(MD_CHANGE_DEVS, &mddev->flags); md_wakeup_thread(mddev->thread); @@ -4203,6 +4289,9 @@ static inline sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *ski return 0; } + /* Allow raid5_quiesce to complete */ + wait_event(conf->wait_for_overlap, conf->quiesce != 2); + if (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery)) return reshape_request(mddev, sector_nr, skipped); @@ -4257,9 +4346,7 @@ static inline sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *ski clear_bit(STRIPE_INSYNC, &sh->state); spin_unlock(&sh->lock); - /* wait for any blocked device to be handled */ - while (unlikely(!handle_stripe(sh))) - ; + handle_stripe(sh); release_stripe(sh); return STRIPE_SECTORS; @@ -4329,37 +4416,6 @@ static int retry_aligned_read(raid5_conf_t *conf, struct bio *raid_bio) return handled; } -#ifdef CONFIG_MULTICORE_RAID456 -static void __process_stripe(void *param, async_cookie_t cookie) -{ - struct stripe_head *sh = param; - - handle_stripe(sh); - release_stripe(sh); -} - -static void process_stripe(struct stripe_head *sh, struct list_head *domain) -{ - async_schedule_domain(__process_stripe, sh, domain); -} - -static void synchronize_stripe_processing(struct list_head *domain) -{ - async_synchronize_full_domain(domain); -} -#else -static void process_stripe(struct stripe_head *sh, struct list_head *domain) -{ - handle_stripe(sh); - release_stripe(sh); - cond_resched(); -} - -static void synchronize_stripe_processing(struct list_head *domain) -{ -} -#endif - /* * This is our raid5 kernel thread. @@ -4371,9 +4427,8 @@ static void synchronize_stripe_processing(struct list_head *domain) static void raid5d(mddev_t *mddev) { struct stripe_head *sh; - raid5_conf_t *conf = mddev_to_conf(mddev); + raid5_conf_t *conf = mddev->private; int handled; - LIST_HEAD(raid_domain); pr_debug("+++ raid5d active\n"); @@ -4410,7 +4465,9 @@ static void raid5d(mddev_t *mddev) spin_unlock_irq(&conf->device_lock); handled++; - process_stripe(sh, &raid_domain); + handle_stripe(sh); + release_stripe(sh); + cond_resched(); spin_lock_irq(&conf->device_lock); } @@ -4418,7 +4475,6 @@ static void raid5d(mddev_t *mddev) spin_unlock_irq(&conf->device_lock); - synchronize_stripe_processing(&raid_domain); async_tx_issue_pending_all(); unplug_slaves(mddev); @@ -4428,7 +4484,7 @@ static void raid5d(mddev_t *mddev) static ssize_t raid5_show_stripe_cache_size(mddev_t *mddev, char *page) { - raid5_conf_t *conf = mddev_to_conf(mddev); + raid5_conf_t *conf = mddev->private; if (conf) return sprintf(page, "%d\n", conf->max_nr_stripes); else @@ -4438,7 +4494,7 @@ raid5_show_stripe_cache_size(mddev_t *mddev, char *page) static ssize_t raid5_store_stripe_cache_size(mddev_t *mddev, const char *page, size_t len) { - raid5_conf_t *conf = mddev_to_conf(mddev); + raid5_conf_t *conf = mddev->private; unsigned long new; int err; @@ -4476,7 +4532,7 @@ raid5_stripecache_size = __ATTR(stripe_cache_size, S_IRUGO | S_IWUSR, static ssize_t raid5_show_preread_threshold(mddev_t *mddev, char *page) { - raid5_conf_t *conf = mddev_to_conf(mddev); + raid5_conf_t *conf = mddev->private; if (conf) return sprintf(page, "%d\n", conf->bypass_threshold); else @@ -4486,7 +4542,7 @@ raid5_show_preread_threshold(mddev_t *mddev, char *page) static ssize_t raid5_store_preread_threshold(mddev_t *mddev, const char *page, size_t len) { - raid5_conf_t *conf = mddev_to_conf(mddev); + raid5_conf_t *conf = mddev->private; unsigned long new; if (len >= PAGE_SIZE) return -EINVAL; @@ -4510,7 +4566,7 @@ raid5_preread_bypass_threshold = __ATTR(preread_bypass_threshold, static ssize_t stripe_cache_active_show(mddev_t *mddev, char *page) { - raid5_conf_t *conf = mddev_to_conf(mddev); + raid5_conf_t *conf = mddev->private; if (conf) return sprintf(page, "%d\n", atomic_read(&conf->active_stripes)); else @@ -4534,20 +4590,16 @@ static struct attribute_group raid5_attrs_group = { static sector_t raid5_size(mddev_t *mddev, sector_t sectors, int raid_disks) { - raid5_conf_t *conf = mddev_to_conf(mddev); + raid5_conf_t *conf = mddev->private; if (!sectors) sectors = mddev->dev_sectors; - if (!raid_disks) { + if (!raid_disks) /* size is defined by the smallest of previous and new size */ - if (conf->raid_disks < conf->previous_raid_disks) - raid_disks = conf->raid_disks; - else - raid_disks = conf->previous_raid_disks; - } + raid_disks = min(conf->raid_disks, conf->previous_raid_disks); - sectors &= ~((sector_t)mddev->chunk_size/512 - 1); - sectors &= ~((sector_t)mddev->new_chunk/512 - 1); + sectors &= ~((sector_t)mddev->chunk_sectors - 1); + sectors &= ~((sector_t)mddev->new_chunk_sectors - 1); return sectors * (raid_disks - conf->max_degraded); } @@ -4625,7 +4677,7 @@ static int raid5_alloc_percpu(raid5_conf_t *conf) { unsigned long cpu; struct page *spare_page; - struct raid5_percpu *allcpus; + struct raid5_percpu __percpu *allcpus; void *scribble; int err; @@ -4645,7 +4697,7 @@ static int raid5_alloc_percpu(raid5_conf_t *conf) } per_cpu_ptr(conf->percpu, cpu)->spare_page = spare_page; } - scribble = kmalloc(scribble_len(conf->raid_disks), GFP_KERNEL); + scribble = kmalloc(conf->scribble_len, GFP_KERNEL); if (!scribble) { err = -ENOMEM; break; @@ -4666,7 +4718,7 @@ static int raid5_alloc_percpu(raid5_conf_t *conf) static raid5_conf_t *setup_conf(mddev_t *mddev) { raid5_conf_t *conf; - int raid_disk, memory; + int raid_disk, memory, max_disks; mdk_rdev_t *rdev; struct disk_info *disk; @@ -4691,24 +4743,39 @@ static raid5_conf_t *setup_conf(mddev_t *mddev) return ERR_PTR(-EINVAL); } - if (!mddev->new_chunk || mddev->new_chunk % PAGE_SIZE) { + if (!mddev->new_chunk_sectors || + (mddev->new_chunk_sectors << 9) % PAGE_SIZE || + !is_power_of_2(mddev->new_chunk_sectors)) { printk(KERN_ERR "raid5: invalid chunk size %d for %s\n", - mddev->new_chunk, mdname(mddev)); + mddev->new_chunk_sectors << 9, mdname(mddev)); return ERR_PTR(-EINVAL); } conf = kzalloc(sizeof(raid5_conf_t), GFP_KERNEL); if (conf == NULL) goto abort; + spin_lock_init(&conf->device_lock); + init_waitqueue_head(&conf->wait_for_stripe); + init_waitqueue_head(&conf->wait_for_overlap); + INIT_LIST_HEAD(&conf->handle_list); + INIT_LIST_HEAD(&conf->hold_list); + INIT_LIST_HEAD(&conf->delayed_list); + INIT_LIST_HEAD(&conf->bitmap_list); + INIT_LIST_HEAD(&conf->inactive_list); + atomic_set(&conf->active_stripes, 0); + atomic_set(&conf->preread_active_stripes, 0); + atomic_set(&conf->active_aligned_reads, 0); + conf->bypass_threshold = BYPASS_THRESHOLD; conf->raid_disks = mddev->raid_disks; - conf->scribble_len = scribble_len(conf->raid_disks); if (mddev->reshape_position == MaxSector) conf->previous_raid_disks = mddev->raid_disks; else conf->previous_raid_disks = mddev->raid_disks - mddev->delta_disks; + max_disks = max(conf->raid_disks, conf->previous_raid_disks); + conf->scribble_len = scribble_len(max_disks); - conf->disks = kzalloc(conf->raid_disks * sizeof(struct disk_info), + conf->disks = kzalloc(max_disks * sizeof(struct disk_info), GFP_KERNEL); if (!conf->disks) goto abort; @@ -4722,24 +4789,11 @@ static raid5_conf_t *setup_conf(mddev_t *mddev) if (raid5_alloc_percpu(conf) != 0) goto abort; - spin_lock_init(&conf->device_lock); - init_waitqueue_head(&conf->wait_for_stripe); - init_waitqueue_head(&conf->wait_for_overlap); - INIT_LIST_HEAD(&conf->handle_list); - INIT_LIST_HEAD(&conf->hold_list); - INIT_LIST_HEAD(&conf->delayed_list); - INIT_LIST_HEAD(&conf->bitmap_list); - INIT_LIST_HEAD(&conf->inactive_list); - atomic_set(&conf->active_stripes, 0); - atomic_set(&conf->preread_active_stripes, 0); - atomic_set(&conf->active_aligned_reads, 0); - conf->bypass_threshold = BYPASS_THRESHOLD; - pr_debug("raid5: run(%s) called.\n", mdname(mddev)); list_for_each_entry(rdev, &mddev->disks, same_set) { raid_disk = rdev->raid_disk; - if (raid_disk >= conf->raid_disks + if (raid_disk >= max_disks || raid_disk < 0) continue; disk = conf->disks + raid_disk; @@ -4756,7 +4810,8 @@ static raid5_conf_t *setup_conf(mddev_t *mddev) conf->fullsync = 1; } - conf->chunk_size = mddev->new_chunk; + conf->chunk_sectors = mddev->new_chunk_sectors; + conf->level = mddev->new_level; if (conf->level == 6) conf->max_degraded = 2; else @@ -4765,12 +4820,12 @@ static raid5_conf_t *setup_conf(mddev_t *mddev) conf->max_nr_stripes = NR_STRIPES; conf->reshape_progress = mddev->reshape_position; if (conf->reshape_progress != MaxSector) { - conf->prev_chunk = mddev->chunk_size; + conf->prev_chunk_sectors = mddev->chunk_sectors; conf->prev_algo = mddev->layout; } memory = conf->max_nr_stripes * (sizeof(struct stripe_head) + - conf->raid_disks * ((sizeof(struct bio) + PAGE_SIZE))) / 1024; + max_disks * ((sizeof(struct bio) + PAGE_SIZE))) / 1024; if (grow_stripes(conf, conf->max_nr_stripes)) { printk(KERN_ERR "raid5: couldn't allocate %dkB for buffers\n", memory); @@ -4779,7 +4834,7 @@ static raid5_conf_t *setup_conf(mddev_t *mddev) printk(KERN_INFO "raid5: allocated %dkB for %s\n", memory, mdname(mddev)); - conf->thread = md_register_thread(raid5d, mddev, "%s_raid5"); + conf->thread = md_register_thread(raid5d, mddev, NULL); if (!conf->thread) { printk(KERN_ERR "raid5: couldn't allocate thread for %s\n", @@ -4797,12 +4852,45 @@ static raid5_conf_t *setup_conf(mddev_t *mddev) return ERR_PTR(-ENOMEM); } + +static int only_parity(int raid_disk, int algo, int raid_disks, int max_degraded) +{ + switch (algo) { + case ALGORITHM_PARITY_0: + if (raid_disk < max_degraded) + return 1; + break; + case ALGORITHM_PARITY_N: + if (raid_disk >= raid_disks - max_degraded) + return 1; + break; + case ALGORITHM_PARITY_0_6: + if (raid_disk == 0 || + raid_disk == raid_disks - 1) + return 1; + break; + case ALGORITHM_LEFT_ASYMMETRIC_6: + case ALGORITHM_RIGHT_ASYMMETRIC_6: + case ALGORITHM_LEFT_SYMMETRIC_6: + case ALGORITHM_RIGHT_SYMMETRIC_6: + if (raid_disk == raid_disks - 1) + return 1; + } + return 0; +} + static int run(mddev_t *mddev) { raid5_conf_t *conf; - int working_disks = 0; + int working_disks = 0, chunk_size; + int dirty_parity_disks = 0; mdk_rdev_t *rdev; + sector_t reshape_offset = 0; + if (mddev->recovery_cp != MaxSector) + printk(KERN_NOTICE "raid5: %s is not clean" + " -- starting background reconstruction\n", + mdname(mddev)); if (mddev->reshape_position != MaxSector) { /* Check that we can continue the reshape. * Currently only disks can change, it must @@ -4825,19 +4913,39 @@ static int run(mddev_t *mddev) * geometry. */ here_new = mddev->reshape_position; - if (sector_div(here_new, (mddev->new_chunk>>9)* + if (sector_div(here_new, mddev->new_chunk_sectors * (mddev->raid_disks - max_degraded))) { printk(KERN_ERR "raid5: reshape_position not " "on a stripe boundary\n"); return -EINVAL; } + reshape_offset = here_new * mddev->new_chunk_sectors; /* here_new is the stripe we will write to */ here_old = mddev->reshape_position; - sector_div(here_old, (mddev->chunk_size>>9)* + sector_div(here_old, mddev->chunk_sectors * (old_disks-max_degraded)); /* here_old is the first stripe that we might need to read * from */ - if (here_new >= here_old) { + if (mddev->delta_disks == 0) { + /* We cannot be sure it is safe to start an in-place + * reshape. It is only safe if user-space if monitoring + * and taking constant backups. + * mdadm always starts a situation like this in + * readonly mode so it can take control before + * allowing any writes. So just check for that. + */ + if ((here_new * mddev->new_chunk_sectors != + here_old * mddev->chunk_sectors) || + mddev->ro == 0) { + printk(KERN_ERR "raid5: in-place reshape must be started" + " in read-only mode - aborting\n"); + return -EINVAL; + } + } else if (mddev->delta_disks < 0 + ? (here_new * mddev->new_chunk_sectors <= + here_old * mddev->chunk_sectors) + : (here_new * mddev->new_chunk_sectors >= + here_old * mddev->chunk_sectors)) { /* Reading from the same stripe as writing to - bad */ printk(KERN_ERR "raid5: reshape_position too early for " "auto-recovery - aborting.\n"); @@ -4848,7 +4956,7 @@ static int run(mddev_t *mddev) } else { BUG_ON(mddev->level != mddev->new_level); BUG_ON(mddev->layout != mddev->new_layout); - BUG_ON(mddev->chunk_size != mddev->new_chunk); + BUG_ON(mddev->chunk_sectors != mddev->new_chunk_sectors); BUG_ON(mddev->delta_disks != 0); } @@ -4867,12 +4975,54 @@ static int run(mddev_t *mddev) /* * 0 for a fully functional array, 1 or 2 for a degraded array. */ - list_for_each_entry(rdev, &mddev->disks, same_set) - if (rdev->raid_disk >= 0 && - test_bit(In_sync, &rdev->flags)) + list_for_each_entry(rdev, &mddev->disks, same_set) { + if (rdev->raid_disk < 0) + continue; + if (test_bit(In_sync, &rdev->flags)) working_disks++; + /* This disc is not fully in-sync. However if it + * just stored parity (beyond the recovery_offset), + * when we don't need to be concerned about the + * array being dirty. + * When reshape goes 'backwards', we never have + * partially completed devices, so we only need + * to worry about reshape going forwards. + */ + /* Hack because v0.91 doesn't store recovery_offset properly. */ + if (mddev->major_version == 0 && + mddev->minor_version > 90) + rdev->recovery_offset = reshape_offset; + + printk("%d: w=%d pa=%d pr=%d m=%d a=%d r=%d op1=%d op2=%d\n", + rdev->raid_disk, working_disks, conf->prev_algo, + conf->previous_raid_disks, conf->max_degraded, + conf->algorithm, conf->raid_disks, + only_parity(rdev->raid_disk, + conf->prev_algo, + conf->previous_raid_disks, + conf->max_degraded), + only_parity(rdev->raid_disk, + conf->algorithm, + conf->raid_disks, + conf->max_degraded)); + if (rdev->recovery_offset < reshape_offset) { + /* We need to check old and new layout */ + if (!only_parity(rdev->raid_disk, + conf->algorithm, + conf->raid_disks, + conf->max_degraded)) + continue; + } + if (!only_parity(rdev->raid_disk, + conf->prev_algo, + conf->previous_raid_disks, + conf->max_degraded)) + continue; + dirty_parity_disks++; + } - mddev->degraded = conf->raid_disks - working_disks; + mddev->degraded = (max(conf->raid_disks, conf->previous_raid_disks) + - working_disks); if (mddev->degraded > conf->max_degraded) { printk(KERN_ERR "raid5: not enough operational devices for %s" @@ -4882,10 +5032,10 @@ static int run(mddev_t *mddev) } /* device size must be a multiple of chunk size */ - mddev->dev_sectors &= ~(mddev->chunk_size / 512 - 1); + mddev->dev_sectors &= ~(mddev->chunk_sectors - 1); mddev->resync_max_sectors = mddev->dev_sectors; - if (mddev->degraded > 0 && + if (mddev->degraded > dirty_parity_disks && mddev->recovery_cp != MaxSector) { if (mddev->ok_start_degraded) printk(KERN_WARNING @@ -4922,7 +5072,7 @@ static int run(mddev_t *mddev) set_bit(MD_RECOVERY_RESHAPE, &mddev->recovery); set_bit(MD_RECOVERY_RUNNING, &mddev->recovery); mddev->sync_thread = md_register_thread(md_do_sync, mddev, - "%s_reshape"); + "reshape"); } /* read-ahead size must cover two whole stripes, which is @@ -4931,7 +5081,7 @@ static int run(mddev_t *mddev) { int data_disks = conf->previous_raid_disks - conf->max_degraded; int stripe = data_disks * - (mddev->chunk_size / PAGE_SIZE); + ((mddev->chunk_sectors << 9) / PAGE_SIZE); if (mddev->queue->backing_dev_info.ra_pages < 2 * stripe) mddev->queue->backing_dev_info.ra_pages = 2 * stripe; } @@ -4951,6 +5101,14 @@ static int run(mddev_t *mddev) md_set_array_sectors(mddev, raid5_size(mddev, 0, 0)); blk_queue_merge_bvec(mddev->queue, raid5_mergeable_bvec); + chunk_size = mddev->chunk_sectors << 9; + blk_queue_io_min(mddev->queue, chunk_size); + blk_queue_io_opt(mddev->queue, chunk_size * + (conf->raid_disks - conf->max_degraded)); + + list_for_each_entry(rdev, &mddev->disks, same_set) + disk_stack_limits(mddev->gendisk, rdev->bdev, + rdev->data_offset << 9); return 0; abort: @@ -4975,9 +5133,8 @@ static int stop(mddev_t *mddev) mddev->thread = NULL; mddev->queue->backing_dev_info.congested_fn = NULL; blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/ - sysfs_remove_group(&mddev->kobj, &raid5_attrs_group); free_conf(conf); - mddev->private = NULL; + mddev->private = &raid5_attrs_group; return 0; } @@ -5021,7 +5178,8 @@ static void status(struct seq_file *seq, mddev_t *mddev) raid5_conf_t *conf = (raid5_conf_t *) mddev->private; int i; - seq_printf (seq, " level %d, %dk chunk, algorithm %d", mddev->level, mddev->chunk_size >> 10, mddev->layout); + seq_printf(seq, " level %d, %dk chunk, algorithm %d", mddev->level, + mddev->chunk_sectors / 2, mddev->layout); seq_printf (seq, " [%d/%d] [", conf->raid_disks, conf->raid_disks - mddev->degraded); for (i = 0; i < conf->raid_disks; i++) seq_printf (seq, "%s", @@ -5169,7 +5327,7 @@ static int raid5_resize(mddev_t *mddev, sector_t sectors) * any io in the removed space completes, but it hardly seems * worth it. */ - sectors &= ~((sector_t)mddev->chunk_size/512 - 1); + sectors &= ~((sector_t)mddev->chunk_sectors - 1); md_set_array_sectors(mddev, raid5_size(mddev, sectors, mddev->raid_disks)); if (mddev->array_sectors > @@ -5177,6 +5335,7 @@ static int raid5_resize(mddev_t *mddev, sector_t sectors) return -EINVAL; set_capacity(mddev->gendisk, mddev->array_sectors); mddev->changed = 1; + revalidate_disk(mddev->gendisk); if (sectors > mddev->dev_sectors && mddev->recovery_cp == MaxSector) { mddev->recovery_cp = mddev->dev_sectors; set_bit(MD_RECOVERY_NEEDED, &mddev->recovery); @@ -5186,14 +5345,37 @@ static int raid5_resize(mddev_t *mddev, sector_t sectors) return 0; } -static int raid5_check_reshape(mddev_t *mddev) +static int check_stripe_cache(mddev_t *mddev) +{ + /* Can only proceed if there are plenty of stripe_heads. + * We need a minimum of one full stripe,, and for sensible progress + * it is best to have about 4 times that. + * If we require 4 times, then the default 256 4K stripe_heads will + * allow for chunk sizes up to 256K, which is probably OK. + * If the chunk size is greater, user-space should request more + * stripe_heads first. + */ + raid5_conf_t *conf = mddev->private; + if (((mddev->chunk_sectors << 9) / STRIPE_SIZE) * 4 + > conf->max_nr_stripes || + ((mddev->new_chunk_sectors << 9) / STRIPE_SIZE) * 4 + > conf->max_nr_stripes) { + printk(KERN_WARNING "raid5: reshape: not enough stripes. Needed %lu\n", + ((max(mddev->chunk_sectors, mddev->new_chunk_sectors) << 9) + / STRIPE_SIZE)*4); + return 0; + } + return 1; +} + +static int check_reshape(mddev_t *mddev) { - raid5_conf_t *conf = mddev_to_conf(mddev); + raid5_conf_t *conf = mddev->private; if (mddev->delta_disks == 0 && mddev->new_layout == mddev->layout && - mddev->new_chunk == mddev->chunk_size) - return -EINVAL; /* nothing to do */ + mddev->new_chunk_sectors == mddev->chunk_sectors) + return 0; /* nothing to do */ if (mddev->bitmap) /* Cannot grow a bitmap yet */ return -EBUSY; @@ -5212,28 +5394,15 @@ static int raid5_check_reshape(mddev_t *mddev) return -EINVAL; } - /* Can only proceed if there are plenty of stripe_heads. - * We need a minimum of one full stripe,, and for sensible progress - * it is best to have about 4 times that. - * If we require 4 times, then the default 256 4K stripe_heads will - * allow for chunk sizes up to 256K, which is probably OK. - * If the chunk size is greater, user-space should request more - * stripe_heads first. - */ - if ((mddev->chunk_size / STRIPE_SIZE) * 4 > conf->max_nr_stripes || - (mddev->new_chunk / STRIPE_SIZE) * 4 > conf->max_nr_stripes) { - printk(KERN_WARNING "raid5: reshape: not enough stripes. Needed %lu\n", - (max(mddev->chunk_size, mddev->new_chunk) - / STRIPE_SIZE)*4); + if (!check_stripe_cache(mddev)) return -ENOSPC; - } return resize_stripes(conf, conf->raid_disks + mddev->delta_disks); } static int raid5_start_reshape(mddev_t *mddev) { - raid5_conf_t *conf = mddev_to_conf(mddev); + raid5_conf_t *conf = mddev->private; mdk_rdev_t *rdev; int spares = 0; int added_devices = 0; @@ -5242,6 +5411,9 @@ static int raid5_start_reshape(mddev_t *mddev) if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery)) return -EBUSY; + if (!check_stripe_cache(mddev)) + return -ENOSPC; + list_for_each_entry(rdev, &mddev->disks, same_set) if (rdev->raid_disk < 0 && !test_bit(Faulty, &rdev->flags)) @@ -5268,8 +5440,8 @@ static int raid5_start_reshape(mddev_t *mddev) spin_lock_irq(&conf->device_lock); conf->previous_raid_disks = conf->raid_disks; conf->raid_disks += mddev->delta_disks; - conf->prev_chunk = conf->chunk_size; - conf->chunk_size = mddev->new_chunk; + conf->prev_chunk_sectors = conf->chunk_sectors; + conf->chunk_sectors = mddev->new_chunk_sectors; conf->prev_algo = conf->algorithm; conf->algorithm = mddev->new_layout; if (mddev->delta_disks < 0) @@ -5288,9 +5460,11 @@ static int raid5_start_reshape(mddev_t *mddev) !test_bit(Faulty, &rdev->flags)) { if (raid5_add_disk(mddev, rdev) == 0) { char nm[20]; - set_bit(In_sync, &rdev->flags); - added_devices++; - rdev->recovery_offset = 0; + if (rdev->raid_disk >= conf->previous_raid_disks) { + set_bit(In_sync, &rdev->flags); + added_devices++; + } else + rdev->recovery_offset = 0; sprintf(nm, "rd%d", rdev->raid_disk); if (sysfs_create_link(&mddev->kobj, &rdev->kobj, nm)) @@ -5302,14 +5476,17 @@ static int raid5_start_reshape(mddev_t *mddev) break; } + /* When a reshape changes the number of devices, ->degraded + * is measured against the large of the pre and post number of + * devices.*/ if (mddev->delta_disks > 0) { spin_lock_irqsave(&conf->device_lock, flags); - mddev->degraded = (conf->raid_disks - conf->previous_raid_disks) + mddev->degraded += (conf->raid_disks - conf->previous_raid_disks) - added_devices; spin_unlock_irqrestore(&conf->device_lock, flags); } mddev->raid_disks = conf->raid_disks; - mddev->reshape_position = 0; + mddev->reshape_position = conf->reshape_progress; set_bit(MD_CHANGE_DEVS, &mddev->flags); clear_bit(MD_RECOVERY_SYNC, &mddev->recovery); @@ -5317,7 +5494,7 @@ static int raid5_start_reshape(mddev_t *mddev) set_bit(MD_RECOVERY_RESHAPE, &mddev->recovery); set_bit(MD_RECOVERY_RUNNING, &mddev->recovery); mddev->sync_thread = md_register_thread(md_do_sync, mddev, - "%s_reshape"); + "reshape"); if (!mddev->sync_thread) { mddev->recovery = 0; spin_lock_irq(&conf->device_lock); @@ -5351,7 +5528,7 @@ static void end_reshape(raid5_conf_t *conf) */ { int data_disks = conf->raid_disks - conf->max_degraded; - int stripe = data_disks * (conf->chunk_size + int stripe = data_disks * ((conf->chunk_sectors << 9) / PAGE_SIZE); if (conf->mddev->queue->backing_dev_info.ra_pages < 2 * stripe) conf->mddev->queue->backing_dev_info.ra_pages = 2 * stripe; @@ -5364,8 +5541,7 @@ static void end_reshape(raid5_conf_t *conf) */ static void raid5_finish_reshape(mddev_t *mddev) { - struct block_device *bdev; - raid5_conf_t *conf = mddev_to_conf(mddev); + raid5_conf_t *conf = mddev->private; if (!test_bit(MD_RECOVERY_INTR, &mddev->recovery)) { @@ -5373,15 +5549,7 @@ static void raid5_finish_reshape(mddev_t *mddev) md_set_array_sectors(mddev, raid5_size(mddev, 0, 0)); set_capacity(mddev->gendisk, mddev->array_sectors); mddev->changed = 1; - - bdev = bdget_disk(mddev->gendisk, 0); - if (bdev) { - mutex_lock(&bdev->bd_inode->i_mutex); - i_size_write(bdev->bd_inode, - (loff_t)mddev->array_sectors << 9); - mutex_unlock(&bdev->bd_inode->i_mutex); - bdput(bdev); - } + revalidate_disk(mddev->gendisk); } else { int d; mddev->degraded = conf->raid_disks; @@ -5392,11 +5560,18 @@ static void raid5_finish_reshape(mddev_t *mddev) mddev->degraded--; for (d = conf->raid_disks ; d < conf->raid_disks - mddev->delta_disks; - d++) - raid5_remove_disk(mddev, d); + d++) { + mdk_rdev_t *rdev = conf->disks[d].rdev; + if (rdev && raid5_remove_disk(mddev, d) == 0) { + char nm[20]; + sprintf(nm, "rd%d", rdev->raid_disk); + sysfs_remove_link(&mddev->kobj, nm); + rdev->raid_disk = -1; + } + } } mddev->layout = conf->algorithm; - mddev->chunk_size = conf->chunk_size; + mddev->chunk_sectors = conf->chunk_sectors; mddev->reshape_position = MaxSector; mddev->delta_disks = 0; } @@ -5404,7 +5579,7 @@ static void raid5_finish_reshape(mddev_t *mddev) static void raid5_quiesce(mddev_t *mddev, int state) { - raid5_conf_t *conf = mddev_to_conf(mddev); + raid5_conf_t *conf = mddev->private; switch(state) { case 2: /* resume for a suspend */ @@ -5413,12 +5588,18 @@ static void raid5_quiesce(mddev_t *mddev, int state) case 1: /* stop all writes */ spin_lock_irq(&conf->device_lock); - conf->quiesce = 1; + /* '2' tells resync/reshape to pause so that all + * active stripes can drain + */ + conf->quiesce = 2; wait_event_lock_irq(conf->wait_for_stripe, atomic_read(&conf->active_stripes) == 0 && atomic_read(&conf->active_aligned_reads) == 0, conf->device_lock, /* nothing */); + conf->quiesce = 1; spin_unlock_irq(&conf->device_lock); + /* allow reshape to continue */ + wake_up(&conf->wait_for_overlap); break; case 0: /* re-enable writes */ @@ -5454,7 +5635,7 @@ static void *raid5_takeover_raid1(mddev_t *mddev) mddev->new_level = 5; mddev->new_layout = ALGORITHM_LEFT_SYMMETRIC; - mddev->new_chunk = chunksect << 9; + mddev->new_chunk_sectors = chunksect; return setup_conf(mddev); } @@ -5493,24 +5674,24 @@ static void *raid5_takeover_raid6(mddev_t *mddev) } -static int raid5_reconfig(mddev_t *mddev, int new_layout, int new_chunk) +static int raid5_check_reshape(mddev_t *mddev) { /* For a 2-drive array, the layout and chunk size can be changed * immediately as not restriping is needed. * For larger arrays we record the new value - after validation * to be used by a reshape pass. */ - raid5_conf_t *conf = mddev_to_conf(mddev); + raid5_conf_t *conf = mddev->private; + int new_chunk = mddev->new_chunk_sectors; - if (new_layout >= 0 && !algorithm_valid_raid5(new_layout)) + if (mddev->new_layout >= 0 && !algorithm_valid_raid5(mddev->new_layout)) return -EINVAL; if (new_chunk > 0) { - if (new_chunk & (new_chunk-1)) - /* not a power of 2 */ + if (!is_power_of_2(new_chunk)) return -EINVAL; - if (new_chunk < PAGE_SIZE) + if (new_chunk < (PAGE_SIZE>>9)) return -EINVAL; - if (mddev->array_sectors & ((new_chunk>>9)-1)) + if (mddev->array_sectors & (new_chunk-1)) /* not factor of array size */ return -EINVAL; } @@ -5518,49 +5699,39 @@ static int raid5_reconfig(mddev_t *mddev, int new_layout, int new_chunk) /* They look valid */ if (mddev->raid_disks == 2) { - - if (new_layout >= 0) { - conf->algorithm = new_layout; - mddev->layout = mddev->new_layout = new_layout; + /* can make the change immediately */ + if (mddev->new_layout >= 0) { + conf->algorithm = mddev->new_layout; + mddev->layout = mddev->new_layout; } if (new_chunk > 0) { - conf->chunk_size = new_chunk; - mddev->chunk_size = mddev->new_chunk = new_chunk; + conf->chunk_sectors = new_chunk ; + mddev->chunk_sectors = new_chunk; } set_bit(MD_CHANGE_DEVS, &mddev->flags); md_wakeup_thread(mddev->thread); - } else { - if (new_layout >= 0) - mddev->new_layout = new_layout; - if (new_chunk > 0) - mddev->new_chunk = new_chunk; } - return 0; + return check_reshape(mddev); } -static int raid6_reconfig(mddev_t *mddev, int new_layout, int new_chunk) +static int raid6_check_reshape(mddev_t *mddev) { - if (new_layout >= 0 && !algorithm_valid_raid6(new_layout)) + int new_chunk = mddev->new_chunk_sectors; + + if (mddev->new_layout >= 0 && !algorithm_valid_raid6(mddev->new_layout)) return -EINVAL; if (new_chunk > 0) { - if (new_chunk & (new_chunk-1)) - /* not a power of 2 */ + if (!is_power_of_2(new_chunk)) return -EINVAL; - if (new_chunk < PAGE_SIZE) + if (new_chunk < (PAGE_SIZE >> 9)) return -EINVAL; - if (mddev->array_sectors & ((new_chunk>>9)-1)) + if (mddev->array_sectors & (new_chunk-1)) /* not factor of array size */ return -EINVAL; } /* They look valid */ - - if (new_layout >= 0) - mddev->new_layout = new_layout; - if (new_chunk > 0) - mddev->new_chunk = new_chunk; - - return 0; + return check_reshape(mddev); } static void *raid5_takeover(mddev_t *mddev) @@ -5570,8 +5741,6 @@ static void *raid5_takeover(mddev_t *mddev) * raid1 - if there are two drives. We need to know the chunk size * raid4 - trivial - just use a raid4 layout. * raid6 - Providing it is a *_6 layout - * - * For now, just do raid1 */ if (mddev->level == 1) @@ -5653,12 +5822,11 @@ static struct mdk_personality raid6_personality = .sync_request = sync_request, .resize = raid5_resize, .size = raid5_size, - .check_reshape = raid5_check_reshape, + .check_reshape = raid6_check_reshape, .start_reshape = raid5_start_reshape, .finish_reshape = raid5_finish_reshape, .quiesce = raid5_quiesce, .takeover = raid6_takeover, - .reconfig = raid6_reconfig, }; static struct mdk_personality raid5_personality = { @@ -5681,7 +5849,6 @@ static struct mdk_personality raid5_personality = .finish_reshape = raid5_finish_reshape, .quiesce = raid5_quiesce, .takeover = raid5_takeover, - .reconfig = raid5_reconfig, }; static struct mdk_personality raid4_personality = @@ -5724,6 +5891,7 @@ static void raid5_exit(void) module_init(raid5_init); module_exit(raid5_exit); MODULE_LICENSE("GPL"); +MODULE_DESCRIPTION("RAID4/5/6 (striping with parity) personality for MD"); MODULE_ALIAS("md-personality-4"); /* RAID5 */ MODULE_ALIAS("md-raid5"); MODULE_ALIAS("md-raid4");