From patchwork Fri Jul 22 14:58:05 2011 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Jan Schmidt X-Patchwork-Id: 999772 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by demeter2.kernel.org (8.14.4/8.14.4) with ESMTP id p6MEwMef019648 for ; Fri, 22 Jul 2011 14:58:23 GMT Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754721Ab1GVO6O (ORCPT ); Fri, 22 Jul 2011 10:58:14 -0400 Received: from mort.rzone.de ([81.169.144.234]:9139 "EHLO mort.rzone.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754640Ab1GVO6K (ORCPT ); Fri, 22 Jul 2011 10:58:10 -0400 Received: from gargravarr.store (gargravarr.store [192.168.42.236]) by mort.rzone.de (Postfix) with ESMTP id EB0303E8; Fri, 22 Jul 2011 16:58:08 +0200 (MEST) Received: by gargravarr.store (Postfix, from userid 32566) id E240BC007; Fri, 22 Jul 2011 16:58:08 +0200 (CEST) From: Jan Schmidt To: chris.mason@oracle.com, linux-btrfs@vger.kernel.org Subject: [RFC PATCH 1/4] btrfs: btrfs_multi_bio replaced with btrfs_bio Date: Fri, 22 Jul 2011 16:58:05 +0200 Message-Id: <70f41bb2f34daf5aab8f429170168154c9a89527.1311344751.git.list.btrfs@jan-o-sch.net> X-Mailer: git-send-email 1.7.3.4 In-Reply-To: References: In-Reply-To: References: Sender: linux-btrfs-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-btrfs@vger.kernel.org X-Greylist: IP, sender and recipient auto-whitelisted, not delayed by milter-greylist-4.2.6 (demeter2.kernel.org [140.211.167.43]); Fri, 22 Jul 2011 14:58:23 +0000 (UTC) btrfs_bio is a bio abstraction able to split and not complete after the last bio has returned (like the old btrfs_multi_bio). Additionally, btrfs_bio tracks the mirror_num used to read data which can be used for error correction purposes. Signed-off-by: Jan Schmidt --- fs/btrfs/extent-tree.c | 10 ++-- fs/btrfs/scrub.c | 20 ++++---- fs/btrfs/volumes.c | 128 +++++++++++++++++++++++++---------------------- fs/btrfs/volumes.h | 10 +++- 4 files changed, 90 insertions(+), 78 deletions(-) diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 71cd456..351efb3 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -1772,18 +1772,18 @@ static int btrfs_discard_extent(struct btrfs_root *root, u64 bytenr, { int ret; u64 discarded_bytes = 0; - struct btrfs_multi_bio *multi = NULL; + struct btrfs_bio *bbio = NULL; /* Tell the block device(s) that the sectors can be discarded */ ret = btrfs_map_block(&root->fs_info->mapping_tree, REQ_DISCARD, - bytenr, &num_bytes, &multi, 0); + bytenr, &num_bytes, &bbio, 0); if (!ret) { - struct btrfs_bio_stripe *stripe = multi->stripes; + struct btrfs_bio_stripe *stripe = bbio->stripes; int i; - for (i = 0; i < multi->num_stripes; i++, stripe++) { + for (i = 0; i < bbio->num_stripes; i++, stripe++) { ret = btrfs_issue_discard(stripe->dev->bdev, stripe->physical, stripe->length); @@ -1792,7 +1792,7 @@ static int btrfs_discard_extent(struct btrfs_root *root, u64 bytenr, else if (ret != -EOPNOTSUPP) break; } - kfree(multi); + kfree(bbio); } if (discarded_bytes && ret == -EOPNOTSUPP) ret = 0; diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c index a8d03d5..c04775e 100644 --- a/fs/btrfs/scrub.c +++ b/fs/btrfs/scrub.c @@ -250,7 +250,7 @@ static void scrub_fixup(struct scrub_bio *sbio, int ix) struct scrub_dev *sdev = sbio->sdev; struct btrfs_fs_info *fs_info = sdev->dev->dev_root->fs_info; struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree; - struct btrfs_multi_bio *multi = NULL; + struct btrfs_bio *bbio = NULL; u64 logical = sbio->logical + ix * PAGE_SIZE; u64 length; int i; @@ -269,8 +269,8 @@ static void scrub_fixup(struct scrub_bio *sbio, int ix) length = PAGE_SIZE; ret = btrfs_map_block(map_tree, REQ_WRITE, logical, &length, - &multi, 0); - if (ret || !multi || length < PAGE_SIZE) { + &bbio, 0); + if (ret || !bbio || length < PAGE_SIZE) { printk(KERN_ERR "scrub_fixup: btrfs_map_block failed us for %llu\n", (unsigned long long)logical); @@ -278,19 +278,19 @@ static void scrub_fixup(struct scrub_bio *sbio, int ix) return; } - if (multi->num_stripes == 1) + if (bbio->num_stripes == 1) /* there aren't any replicas */ goto uncorrectable; /* * first find a good copy */ - for (i = 0; i < multi->num_stripes; ++i) { + for (i = 0; i < bbio->num_stripes; ++i) { if (i == sbio->spag[ix].mirror_num) continue; - if (scrub_fixup_io(READ, multi->stripes[i].dev->bdev, - multi->stripes[i].physical >> 9, + if (scrub_fixup_io(READ, bbio->stripes[i].dev->bdev, + bbio->stripes[i].physical >> 9, sbio->bio->bi_io_vec[ix].bv_page)) { /* I/O-error, this is not a good copy */ continue; @@ -299,7 +299,7 @@ static void scrub_fixup(struct scrub_bio *sbio, int ix) if (scrub_fixup_check(sbio, ix) == 0) break; } - if (i == multi->num_stripes) + if (i == bbio->num_stripes) goto uncorrectable; if (!sdev->readonly) { @@ -314,7 +314,7 @@ static void scrub_fixup(struct scrub_bio *sbio, int ix) } } - kfree(multi); + kfree(bbio); spin_lock(&sdev->stat_lock); ++sdev->stat.corrected_errors; spin_unlock(&sdev->stat_lock); @@ -325,7 +325,7 @@ static void scrub_fixup(struct scrub_bio *sbio, int ix) return; uncorrectable: - kfree(multi); + kfree(bbio); spin_lock(&sdev->stat_lock); ++sdev->stat.uncorrectable_errors; spin_unlock(&sdev->stat_lock); diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 19450bc..e839b72 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2801,7 +2801,7 @@ static int find_live_mirror(struct map_lookup *map, int first, int num, static int __btrfs_map_block(struct btrfs_mapping_tree *map_tree, int rw, u64 logical, u64 *length, - struct btrfs_multi_bio **multi_ret, + struct btrfs_bio **bbio_ret, int mirror_num) { struct extent_map *em; @@ -2819,18 +2819,18 @@ static int __btrfs_map_block(struct btrfs_mapping_tree *map_tree, int rw, int i; int num_stripes; int max_errors = 0; - struct btrfs_multi_bio *multi = NULL; + struct btrfs_bio *bbio = NULL; - if (multi_ret && !(rw & (REQ_WRITE | REQ_DISCARD))) + if (bbio_ret && !(rw & (REQ_WRITE | REQ_DISCARD))) stripes_allocated = 1; again: - if (multi_ret) { - multi = kzalloc(btrfs_multi_bio_size(stripes_allocated), + if (bbio_ret) { + bbio = kzalloc(btrfs_bio_size(stripes_allocated), GFP_NOFS); - if (!multi) + if (!bbio) return -ENOMEM; - atomic_set(&multi->error, 0); + atomic_set(&bbio->error, 0); } read_lock(&em_tree->lock); @@ -2851,7 +2851,7 @@ again: if (mirror_num > map->num_stripes) mirror_num = 0; - /* if our multi bio struct is too small, back off and try again */ + /* if our btrfs_bio struct is too small, back off and try again */ if (rw & REQ_WRITE) { if (map->type & (BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_DUP)) { @@ -2870,11 +2870,11 @@ again: stripes_required = map->num_stripes; } } - if (multi_ret && (rw & (REQ_WRITE | REQ_DISCARD)) && + if (bbio_ret && (rw & (REQ_WRITE | REQ_DISCARD)) && stripes_allocated < stripes_required) { stripes_allocated = map->num_stripes; free_extent_map(em); - kfree(multi); + kfree(bbio); goto again; } stripe_nr = offset; @@ -2903,7 +2903,7 @@ again: *length = em->len - offset; } - if (!multi_ret) + if (!bbio_ret) goto out; num_stripes = 1; @@ -2928,13 +2928,17 @@ again: stripe_index = find_live_mirror(map, 0, map->num_stripes, current->pid % map->num_stripes); + mirror_num = stripe_index + 1; } } else if (map->type & BTRFS_BLOCK_GROUP_DUP) { - if (rw & (REQ_WRITE | REQ_DISCARD)) + if (rw & (REQ_WRITE | REQ_DISCARD)) { num_stripes = map->num_stripes; - else if (mirror_num) + } else if (mirror_num) { stripe_index = mirror_num - 1; + } else { + mirror_num = 1; + } } else if (map->type & BTRFS_BLOCK_GROUP_RAID10) { int factor = map->num_stripes / map->sub_stripes; @@ -2954,6 +2958,7 @@ again: stripe_index = find_live_mirror(map, stripe_index, map->sub_stripes, stripe_index + current->pid % map->sub_stripes); + mirror_num = stripe_index + 1; } } else { /* @@ -2962,15 +2967,16 @@ again: * stripe_index is the number of our device in the stripe array */ stripe_index = do_div(stripe_nr, map->num_stripes); + mirror_num = stripe_index + 1; } BUG_ON(stripe_index >= map->num_stripes); if (rw & REQ_DISCARD) { for (i = 0; i < num_stripes; i++) { - multi->stripes[i].physical = + bbio->stripes[i].physical = map->stripes[stripe_index].physical + stripe_offset + stripe_nr * map->stripe_len; - multi->stripes[i].dev = map->stripes[stripe_index].dev; + bbio->stripes[i].dev = map->stripes[stripe_index].dev; if (map->type & BTRFS_BLOCK_GROUP_RAID0) { u64 stripes; @@ -2991,16 +2997,16 @@ again: } stripes = stripe_nr_end - 1 - j; do_div(stripes, map->num_stripes); - multi->stripes[i].length = map->stripe_len * + bbio->stripes[i].length = map->stripe_len * (stripes - stripe_nr + 1); if (i == 0) { - multi->stripes[i].length -= + bbio->stripes[i].length -= stripe_offset; stripe_offset = 0; } if (stripe_index == last_stripe) - multi->stripes[i].length -= + bbio->stripes[i].length -= stripe_end_offset; } else if (map->type & BTRFS_BLOCK_GROUP_RAID10) { u64 stripes; @@ -3025,11 +3031,11 @@ again: } stripes = stripe_nr_end - 1 - j; do_div(stripes, factor); - multi->stripes[i].length = map->stripe_len * + bbio->stripes[i].length = map->stripe_len * (stripes - stripe_nr + 1); if (i < map->sub_stripes) { - multi->stripes[i].length -= + bbio->stripes[i].length -= stripe_offset; if (i == map->sub_stripes - 1) stripe_offset = 0; @@ -3037,11 +3043,11 @@ again: if (stripe_index >= last_stripe && stripe_index <= (last_stripe + map->sub_stripes - 1)) { - multi->stripes[i].length -= + bbio->stripes[i].length -= stripe_end_offset; } } else - multi->stripes[i].length = *length; + bbio->stripes[i].length = *length; stripe_index++; if (stripe_index == map->num_stripes) { @@ -3052,19 +3058,20 @@ again: } } else { for (i = 0; i < num_stripes; i++) { - multi->stripes[i].physical = + bbio->stripes[i].physical = map->stripes[stripe_index].physical + stripe_offset + stripe_nr * map->stripe_len; - multi->stripes[i].dev = + bbio->stripes[i].dev = map->stripes[stripe_index].dev; stripe_index++; } } - if (multi_ret) { - *multi_ret = multi; - multi->num_stripes = num_stripes; - multi->max_errors = max_errors; + if (bbio_ret) { + *bbio_ret = bbio; + bbio->num_stripes = num_stripes; + bbio->max_errors = max_errors; + bbio->mirror_num = mirror_num; } out: free_extent_map(em); @@ -3073,9 +3080,9 @@ out: int btrfs_map_block(struct btrfs_mapping_tree *map_tree, int rw, u64 logical, u64 *length, - struct btrfs_multi_bio **multi_ret, int mirror_num) + struct btrfs_bio **bbio_ret, int mirror_num) { - return __btrfs_map_block(map_tree, rw, logical, length, multi_ret, + return __btrfs_map_block(map_tree, rw, logical, length, bbio_ret, mirror_num); } @@ -3144,28 +3151,28 @@ int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree, return 0; } -static void end_bio_multi_stripe(struct bio *bio, int err) +static void btrfs_end_bio(struct bio *bio, int err) { - struct btrfs_multi_bio *multi = bio->bi_private; + struct btrfs_bio *bbio = bio->bi_private; int is_orig_bio = 0; if (err) - atomic_inc(&multi->error); + atomic_inc(&bbio->error); - if (bio == multi->orig_bio) + if (bio == bbio->orig_bio) is_orig_bio = 1; - if (atomic_dec_and_test(&multi->stripes_pending)) { + if (atomic_dec_and_test(&bbio->stripes_pending)) { if (!is_orig_bio) { bio_put(bio); - bio = multi->orig_bio; + bio = bbio->orig_bio; } - bio->bi_private = multi->private; - bio->bi_end_io = multi->end_io; + bio->bi_private = bbio->private; + bio->bi_end_io = bbio->end_io; /* only send an error to the higher layers if it is * beyond the tolerance of the multi-bio */ - if (atomic_read(&multi->error) > multi->max_errors) { + if (atomic_read(&bbio->error) > bbio->max_errors) { err = -EIO; } else if (err) { /* @@ -3175,7 +3182,7 @@ static void end_bio_multi_stripe(struct bio *bio, int err) set_bit(BIO_UPTODATE, &bio->bi_flags); err = 0; } - kfree(multi); + kfree(bbio); bio_endio(bio, err); } else if (!is_orig_bio) { @@ -3255,20 +3262,20 @@ int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio, u64 logical = (u64)bio->bi_sector << 9; u64 length = 0; u64 map_length; - struct btrfs_multi_bio *multi = NULL; int ret; int dev_nr = 0; int total_devs = 1; + struct btrfs_bio *bbio = NULL; length = bio->bi_size; map_tree = &root->fs_info->mapping_tree; map_length = length; - ret = btrfs_map_block(map_tree, rw, logical, &map_length, &multi, + ret = btrfs_map_block(map_tree, rw, logical, &map_length, &bbio, mirror_num); BUG_ON(ret); - total_devs = multi->num_stripes; + total_devs = bbio->num_stripes; if (map_length < length) { printk(KERN_CRIT "mapping failed logical %llu bio len %llu " "len %llu\n", (unsigned long long)logical, @@ -3276,25 +3283,28 @@ int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio, (unsigned long long)map_length); BUG(); } - multi->end_io = first_bio->bi_end_io; - multi->private = first_bio->bi_private; - multi->orig_bio = first_bio; - atomic_set(&multi->stripes_pending, multi->num_stripes); + + bbio->orig_bio = first_bio; + bbio->private = first_bio->bi_private; + bbio->end_io = first_bio->bi_end_io; + atomic_set(&bbio->stripes_pending, bbio->num_stripes); while (dev_nr < total_devs) { - if (total_devs > 1) { - if (dev_nr < total_devs - 1) { - bio = bio_clone(first_bio, GFP_NOFS); - BUG_ON(!bio); - } else { - bio = first_bio; - } - bio->bi_private = multi; - bio->bi_end_io = end_bio_multi_stripe; + if (dev_nr < total_devs - 1) { + bio = bio_clone(first_bio, GFP_NOFS); + BUG_ON(!bio); + } else { + bio = first_bio; } - bio->bi_sector = multi->stripes[dev_nr].physical >> 9; - dev = multi->stripes[dev_nr].dev; + bio->bi_private = bbio; + bio->bi_end_io = btrfs_end_bio; + bio->bi_sector = bbio->stripes[dev_nr].physical >> 9; + dev = bbio->stripes[dev_nr].dev; if (dev && dev->bdev && (rw != WRITE || dev->writeable)) { + pr_debug("btrfs_map_bio: rw %d, secor=%llu, dev=%lu " + "(%s id %llu), size=%u\n", rw, + (u64)bio->bi_sector, (u_long)dev->bdev->bd_dev, + dev->name, dev->devid, bio->bi_size); bio->bi_bdev = dev->bdev; if (async_submit) schedule_bio(root, dev, rw, bio); @@ -3307,8 +3317,6 @@ int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio, } dev_nr++; } - if (total_devs == 1) - kfree(multi); return 0; } diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index 7c12d61..f5ce92b 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -134,7 +134,10 @@ struct btrfs_bio_stripe { u64 length; /* only used for discard mappings */ }; -struct btrfs_multi_bio { +struct btrfs_bio; +typedef void (btrfs_bio_end_io_t) (struct btrfs_bio *bio, int err); + +struct btrfs_bio { atomic_t stripes_pending; bio_end_io_t *end_io; struct bio *orig_bio; @@ -142,6 +145,7 @@ struct btrfs_multi_bio { atomic_t error; int max_errors; int num_stripes; + int mirror_num; struct btrfs_bio_stripe stripes[]; }; @@ -169,7 +173,7 @@ struct map_lookup { int btrfs_account_dev_extents_size(struct btrfs_device *device, u64 start, u64 end, u64 *length); -#define btrfs_multi_bio_size(n) (sizeof(struct btrfs_multi_bio) + \ +#define btrfs_bio_size(n) (sizeof(struct btrfs_bio) + \ (sizeof(struct btrfs_bio_stripe) * (n))) int btrfs_alloc_dev_extent(struct btrfs_trans_handle *trans, @@ -178,7 +182,7 @@ int btrfs_alloc_dev_extent(struct btrfs_trans_handle *trans, u64 chunk_offset, u64 start, u64 num_bytes); int btrfs_map_block(struct btrfs_mapping_tree *map_tree, int rw, u64 logical, u64 *length, - struct btrfs_multi_bio **multi_ret, int mirror_num); + struct btrfs_bio **bbio_ret, int mirror_num); int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree, u64 chunk_start, u64 physical, u64 devid, u64 **logical, int *naddrs, int *stripe_len);