[RFC,-next,v2,24/41] md/md-bitmap: merge md_bitmap_start_sync() into bitmap_operations

Message ID	20240814071113.346781-25-yukuai1@huaweicloud.com (mailing list archive)
State	Superseded
Headers	show Received: from dggsgout11.his.huawei.com (dggsgout11.his.huawei.com [45.249.212.51]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 423E51547C4; Wed, 14 Aug 2024 07:15:39 +0000 (UTC) From: Yu Kuai <yukuai1@huaweicloud.com> To: mariusz.tkaczyk@linux.intel.com, hch@infradead.org, song@kernel.org Cc: linux-kernel@vger.kernel.org, linux-raid@vger.kernel.org, yukuai3@huawei.com, yukuai1@huaweicloud.com, yi.zhang@huawei.com, yangerkun@huawei.com Subject: [PATCH RFC -next v2 24/41] md/md-bitmap: merge md_bitmap_start_sync() into bitmap_operations Date: Wed, 14 Aug 2024 15:10:56 +0800 Message-Id: <20240814071113.346781-25-yukuai1@huaweicloud.com> In-Reply-To: <20240814071113.346781-1-yukuai1@huaweicloud.com> References: <20240814071113.346781-1-yukuai1@huaweicloud.com> Precedence: bulk MIME-Version: 1.0 Content-Transfer-Encoding: 8bit
Series	md/md-bitmap: introduce bitmap_operations and make structure internel \| expand [RFC,-next,v2,00/41] md/md-bitmap: introduce bitmap_operations and make structure internel [RFC,-next,v2,01/41] md/raid1: use md_bitmap_wait_behind_writes() in raid1_read_request() [RFC,-next,v2,02/41] md/md-bitmap: replace md_bitmap_status() with a new helper md_bitmap_get_stats… [RFC,-next,v2,03/41] md: use new helper md_bitmap_get_stats() in update_array_info() [RFC,-next,v2,04/41] md/md-bitmap: add 'events_cleared' into struct md_bitmap_stats [RFC,-next,v2,05/41] md/md-bitmap: add 'sync_size' into struct md_bitmap_stats [RFC,-next,v2,06/41] md/md-bitmap: add 'file_pages' into struct md_bitmap_stats [RFC,-next,v2,07/41] md/md-bitmap: add 'behind_writes' and 'behind_wait' into struct md_bitmap_stats [RFC,-next,v2,08/41] md/md-cluster: use helper md_bitmap_get_stats() to get pages in resize_bitmaps… [RFC,-next,v2,09/41] md/md-bitmap: add a new helper md_bitmap_set_pages() [RFC,-next,v2,10/41] md/md-bitmap: introduce struct bitmap_operations [RFC,-next,v2,11/41] md/md-bitmap: simplify md_bitmap_create() + md_bitmap_load() [RFC,-next,v2,12/41] md/md-bitmap: merge md_bitmap_create() into bitmap_operations [RFC,-next,v2,13/41] md/md-bitmap: merge md_bitmap_load() into bitmap_operations [RFC,-next,v2,14/41] md/md-bitmap: merge md_bitmap_destroy() into bitmap_operations [RFC,-next,v2,15/41] md/md-bitmap: merge md_bitmap_flush() into bitmap_operations [RFC,-next,v2,16/41] md/md-bitmap: make md_bitmap_print_sb() internal [RFC,-next,v2,17/41] md/md-bitmap: merge md_bitmap_update_sb() into bitmap_operations [RFC,-next,v2,18/41] md/md-bitmap: merge md_bitmap_status() into bitmap_operations [RFC,-next,v2,19/41] md/md-bitmap: remove md_bitmap_setallbits() [RFC,-next,v2,20/41] md/md-bitmap: merge bitmap_write_all() into bitmap_operations [RFC,-next,v2,21/41] md/md-bitmap: merge md_bitmap_dirty_bits() into bitmap_operations [RFC,-next,v2,22/41] md/md-bitmap: merge md_bitmap_startwrite() into bitmap_operations [RFC,-next,v2,23/41] md/md-bitmap: merge md_bitmap_endwrite() into bitmap_operations [RFC,-next,v2,24/41] md/md-bitmap: merge md_bitmap_start_sync() into bitmap_operations [RFC,-next,v2,25/41] md/md-bitmap: remove the parameter 'aborted' for md_bitmap_end_sync() [RFC,-next,v2,26/41] md/md-bitmap: merge md_bitmap_end_sync() into bitmap_operations [RFC,-next,v2,27/41] md/md-bitmap: merge md_bitmap_close_sync() into bitmap_operations [RFC,-next,v2,28/41] md/md-bitmap: mrege md_bitmap_cond_end_sync() into bitmap_operations [RFC,-next,v2,29/41] md/md-bitmap: merge md_bitmap_sync_with_cluster() into bitmap_operations [RFC,-next,v2,30/41] md/md-bitmap: merge md_bitmap_unplug_async() into md_bitmap_unplug() [RFC,-next,v2,31/41] md/md-bitmap: merge bitmap_unplug() into bitmap_operations [RFC,-next,v2,32/41] md/md-bitmap: merge md_bitmap_daemon_work() into bitmap_operations [RFC,-next,v2,33/41] md/md-bitmap: pass in mddev directly for md_bitmap_resize() [RFC,-next,v2,34/41] md/md-bitmap: merge md_bitmap_resize() into bitmap_operations [RFC,-next,v2,35/41] md/md-bitmap: merge get_bitmap_from_slot() into bitmap_operations [RFC,-next,v2,36/41] md/md-bitmap: merge md_bitmap_copy_from_slot() into struct bitmap_operation. [RFC,-next,v2,37/41] md/md-bitmap: merge md_bitmap_set_pages() into struct bitmap_operations [RFC,-next,v2,38/41] md/md-bitmap: merge md_bitmap_free() into bitmap_operations [RFC,-next,v2,39/41] md/md-bitmap: merge md_bitmap_wait_behind_writes() into bitmap_operations [RFC,-next,v2,40/41] md/md-bitmap: merge md_bitmap_enabled() into bitmap_operations [RFC,-next,v2,41/41] md/md-bitmap: make in memory structure internal

diff --git a/drivers/md/md-bitmap.c b/drivers/md/md-bitmap.c index 8da6e889d5f1..e42a0433e926 100644 --- a/drivers/md/md-bitmap.c +++ b/drivers/md/md-bitmap.c @@ -1579,24 +1579,26 @@ static void bitmap_endwrite(struct mddev *mddev, sector_t offset, } } -static int __bitmap_start_sync(struct bitmap *bitmap, sector_t offset, sector_t *blocks, - int degraded) +static bool __bitmap_start_sync(struct bitmap *bitmap, sector_t offset, + sector_t *blocks, bool degraded) { bitmap_counter_t *bmc; - int rv; + bool rv; + if (bitmap == NULL) {/* FIXME or bitmap set as 'failed' */ *blocks = 1024; - return 1; /* always resync if no bitmap */ + return true; /* always resync if no bitmap */ } spin_lock_irq(&bitmap->counts.lock); + + rv = false; bmc = md_bitmap_get_counter(&bitmap->counts, offset, blocks, 0); - rv = 0; if (bmc) { /* locked */ - if (RESYNC(*bmc)) - rv = 1; - else if (NEEDED(*bmc)) { - rv = 1; + if (RESYNC(*bmc)) { + rv = true; + } else if (NEEDED(*bmc)) { + rv = true; if (!degraded) { /* don't set/clear bits if degraded */ *bmc |= RESYNC_MASK; *bmc &= ~NEEDED_MASK; @@ -1604,11 +1606,12 @@ static int __bitmap_start_sync(struct bitmap *bitmap, sector_t offset, sector_t } } spin_unlock_irq(&bitmap->counts.lock); + return rv; } -int md_bitmap_start_sync(struct bitmap *bitmap, sector_t offset, sector_t *blocks, - int degraded) +static bool bitmap_start_sync(struct mddev *mddev, sector_t offset, + sector_t *blocks, bool degraded) { /* bitmap_start_sync must always report on multiples of whole * pages, otherwise resync (which is very PAGE_SIZE based) will @@ -1617,19 +1620,19 @@ int md_bitmap_start_sync(struct bitmap *bitmap, sector_t offset, sector_t *block * At least PAGE_SIZE>>9 blocks are covered. * Return the 'or' of the result. */ - int rv = 0; + bool rv = false; sector_t blocks1; *blocks = 0; while (*blocks < (PAGE_SIZE>>9)) { - rv |= __bitmap_start_sync(bitmap, offset, + rv |= __bitmap_start_sync(mddev->bitmap, offset, &blocks1, degraded); offset += blocks1; *blocks += blocks1; } + return rv; } -EXPORT_SYMBOL(md_bitmap_start_sync); void md_bitmap_end_sync(struct bitmap *bitmap, sector_t offset, sector_t *blocks, int aborted) { @@ -1723,7 +1726,7 @@ void md_bitmap_sync_with_cluster(struct mddev *mddev, WARN((blocks > new_lo) && old_lo, "alignment is not correct for lo\n"); for (sector = old_hi; sector < new_hi; ) { - md_bitmap_start_sync(bitmap, sector, &blocks, 0); + bitmap_start_sync(mddev, sector, &blocks, false); sector += blocks; } WARN((blocks > new_hi) && old_hi, "alignment is not correct for hi\n"); @@ -2005,7 +2008,7 @@ static int bitmap_load(struct mddev *mddev) */ while (sector < mddev->resync_max_sectors) { sector_t blocks; - md_bitmap_start_sync(bitmap, sector, &blocks, 0); + bitmap_start_sync(mddev, sector, &blocks, false); sector += blocks; } md_bitmap_close_sync(bitmap); @@ -2731,6 +2734,7 @@ static struct bitmap_operations bitmap_ops = { .startwrite = bitmap_startwrite, .endwrite = bitmap_endwrite, + .start_sync = bitmap_start_sync, .update_sb = bitmap_update_sb, .get_stats = bitmap_get_stats, diff --git a/drivers/md/md-bitmap.h b/drivers/md/md-bitmap.h index 223dc5d6f4e4..6fba3ff2f949 100644 --- a/drivers/md/md-bitmap.h +++ b/drivers/md/md-bitmap.h @@ -258,6 +258,8 @@ struct bitmap_operations { unsigned long sectors, bool behind); void (*endwrite)(struct mddev *mddev, sector_t offset, unsigned long sectors, bool success, bool behind); + bool (*start_sync)(struct mddev *mddev, sector_t offset, + sector_t *blocks, bool degraded); void (*update_sb)(struct bitmap *bitmap); int (*get_stats)(struct bitmap *bitmap, struct md_bitmap_stats *stats); @@ -267,7 +269,6 @@ struct bitmap_operations { void mddev_set_bitmap_ops(struct mddev *mddev); /* these are exported */ -int md_bitmap_start_sync(struct bitmap *bitmap, sector_t offset, sector_t *blocks, int degraded); void md_bitmap_end_sync(struct bitmap *bitmap, sector_t offset, sector_t *blocks, int aborted); void md_bitmap_close_sync(struct bitmap *bitmap); void md_bitmap_cond_end_sync(struct bitmap *bitmap, sector_t sector, bool force); diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c index aeee09965817..555b56d15d92 100644 --- a/drivers/md/raid1.c +++ b/drivers/md/raid1.c @@ -2767,7 +2767,7 @@ static sector_t raid1_sync_request(struct mddev *mddev, sector_t sector_nr, int wonly = -1; int write_targets = 0, read_targets = 0; sector_t sync_blocks; - int still_degraded = 0; + bool still_degraded = false; int good_sectors = RESYNC_SECTORS; int min_bad = 0; /* number of sectors that are bad in all devices */ int idx = sector_to_idx(sector_nr); @@ -2809,7 +2809,7 @@ static sector_t raid1_sync_request(struct mddev *mddev, sector_t sector_nr, /* before building a request, check if we can skip these blocks.. * This call the bitmap_start_sync doesn't actually record anything */ - if (!md_bitmap_start_sync(mddev->bitmap, sector_nr, &sync_blocks, 1) && + if (!mddev->bitmap_ops->start_sync(mddev, sector_nr, &sync_blocks, true) && !conf->fullsync && !test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) { /* We can skip this block, and probably several more */ *skipped = 1; @@ -2860,7 +2860,7 @@ static sector_t raid1_sync_request(struct mddev *mddev, sector_t sector_nr, if (rdev == NULL || test_bit(Faulty, &rdev->flags)) { if (i < conf->raid_disks) - still_degraded = 1; + still_degraded = true; } else if (!test_bit(In_sync, &rdev->flags)) { bio->bi_opf = REQ_OP_WRITE; bio->bi_end_io = end_sync_write; @@ -2984,8 +2984,8 @@ static sector_t raid1_sync_request(struct mddev *mddev, sector_t sector_nr, if (len == 0) break; if (sync_blocks == 0) { - if (!md_bitmap_start_sync(mddev->bitmap, sector_nr, - &sync_blocks, still_degraded) && + if (!mddev->bitmap_ops->start_sync(mddev, sector_nr, + &sync_blocks, still_degraded) && !conf->fullsync && !test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) break; diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c index 52f29b907ab6..2ab641166c8c 100644 --- a/drivers/md/raid10.c +++ b/drivers/md/raid10.c @@ -3289,10 +3289,10 @@ static sector_t raid10_sync_request(struct mddev *mddev, sector_t sector_nr, r10_bio = NULL; for (i = 0 ; i < conf->geo.raid_disks; i++) { - int still_degraded; + bool still_degraded; struct r10bio *rb2; sector_t sect; - int must_sync; + bool must_sync; int any_working; struct raid10_info *mirror = &conf->mirrors[i]; struct md_rdev *mrdev, *mreplace; @@ -3309,7 +3309,7 @@ static sector_t raid10_sync_request(struct mddev *mddev, sector_t sector_nr, if (!mrdev && !mreplace) continue; - still_degraded = 0; + still_degraded = false; /* want to reconstruct this device */ rb2 = r10_bio; sect = raid10_find_virt(conf, sector_nr, i); @@ -3322,8 +3322,9 @@ static sector_t raid10_sync_request(struct mddev *mddev, sector_t sector_nr, * we only need to recover the block if it is set in * the bitmap */ - must_sync = md_bitmap_start_sync(mddev->bitmap, sect, - &sync_blocks, 1); + must_sync = mddev->bitmap_ops->start_sync(mddev, sect, + &sync_blocks, + true); if (sync_blocks < max_sync) max_sync = sync_blocks; if (!must_sync && @@ -3361,13 +3362,13 @@ static sector_t raid10_sync_request(struct mddev *mddev, sector_t sector_nr, struct md_rdev *rdev = conf->mirrors[j].rdev; if (rdev == NULL || test_bit(Faulty, &rdev->flags)) { - still_degraded = 1; + still_degraded = false; break; } } - must_sync = md_bitmap_start_sync(mddev->bitmap, sect, - &sync_blocks, still_degraded); + must_sync = mddev->bitmap_ops->start_sync(mddev, sect, + &sync_blocks, still_degraded); any_working = 0; for (j=0; j<conf->copies;j++) { @@ -3544,8 +3545,9 @@ static sector_t raid10_sync_request(struct mddev *mddev, sector_t sector_nr, mddev_is_clustered(mddev) && (sector_nr + 2 * RESYNC_SECTORS > conf->cluster_sync_high)); - if (!md_bitmap_start_sync(mddev->bitmap, sector_nr, - &sync_blocks, mddev->degraded) && + if (!mddev->bitmap_ops->start_sync(mddev, sector_nr, + &sync_blocks, + mddev->degraded) && !conf->fullsync && !test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery)) { /* We can skip this block */ diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index 93d582b9f922..313904dd6555 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -6485,7 +6485,7 @@ static inline sector_t raid5_sync_request(struct mddev *mddev, sector_t sector_n struct r5conf *conf = mddev->private; struct stripe_head *sh; sector_t sync_blocks; - int still_degraded = 0; + bool still_degraded = false; int i; if (sector_nr >= max_sector) { @@ -6530,7 +6530,8 @@ static inline sector_t raid5_sync_request(struct mddev *mddev, sector_t sector_n } if (!test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery) && !conf->fullsync && - !md_bitmap_start_sync(mddev->bitmap, sector_nr, &sync_blocks, 1) && + !mddev->bitmap_ops->start_sync(mddev, sector_nr, &sync_blocks, + true) && sync_blocks >= RAID5_STRIPE_SECTORS(conf)) { /* we can skip this block, and probably more */ do_div(sync_blocks, RAID5_STRIPE_SECTORS(conf)); @@ -6558,10 +6559,11 @@ static inline sector_t raid5_sync_request(struct mddev *mddev, sector_t sector_n struct md_rdev *rdev = conf->disks[i].rdev; if (rdev == NULL || test_bit(Faulty, &rdev->flags)) - still_degraded = 1; + still_degraded = true; } - md_bitmap_start_sync(mddev->bitmap, sector_nr, &sync_blocks, still_degraded); + mddev->bitmap_ops->start_sync(mddev, sector_nr, &sync_blocks, + still_degraded); set_bit(STRIPE_SYNC_REQUESTED, &sh->state); set_bit(STRIPE_HANDLE, &sh->state);

[RFC,-next,v2,24/41] md/md-bitmap: merge md_bitmap_start_sync() into bitmap_operations

Commit Message

Patch