diff options
Diffstat (limited to 'drivers/md/md.c')
| -rw-r--r-- | drivers/md/md.c | 415 |
1 files changed, 350 insertions, 65 deletions
diff --git a/drivers/md/md.c b/drivers/md/md.c index 01233d855eb2..d5ab4493c8be 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -402,6 +402,7 @@ void mddev_resume(struct mddev *mddev) wake_up(&mddev->sb_wait); mddev->pers->quiesce(mddev, 0); + set_bit(MD_RECOVERY_NEEDED, &mddev->recovery); md_wakeup_thread(mddev->thread); md_wakeup_thread(mddev->sync_thread); /* possibly kick off a reshape */ } @@ -452,7 +453,7 @@ static void submit_flushes(struct work_struct *ws) atomic_inc(&rdev->nr_pending); atomic_inc(&rdev->nr_pending); rcu_read_unlock(); - bi = bio_alloc_mddev(GFP_KERNEL, 0, mddev); + bi = bio_alloc_mddev(GFP_NOIO, 0, mddev); bi->bi_end_io = md_end_flush; bi->bi_private = rdev; bi->bi_bdev = rdev->bdev; @@ -607,6 +608,7 @@ void mddev_init(struct mddev *mddev) init_waitqueue_head(&mddev->sb_wait); init_waitqueue_head(&mddev->recovery_wait); mddev->reshape_position = MaxSector; + mddev->reshape_backwards = 0; mddev->resync_min = 0; mddev->resync_max = MaxSector; mddev->level = LEVEL_NONE; @@ -802,7 +804,7 @@ static int alloc_disk_sb(struct md_rdev * rdev) return 0; } -static void free_disk_sb(struct md_rdev * rdev) +void md_rdev_clear(struct md_rdev *rdev) { if (rdev->sb_page) { put_page(rdev->sb_page); @@ -815,8 +817,10 @@ static void free_disk_sb(struct md_rdev * rdev) put_page(rdev->bb_page); rdev->bb_page = NULL; } + kfree(rdev->badblocks.page); + rdev->badblocks.page = NULL; } - +EXPORT_SYMBOL_GPL(md_rdev_clear); static void super_written(struct bio *bio, int error) { @@ -887,6 +891,10 @@ int sync_page_io(struct md_rdev *rdev, sector_t sector, int size, rdev->meta_bdev : rdev->bdev; if (metadata_op) bio->bi_sector = sector + rdev->sb_start; + else if (rdev->mddev->reshape_position != MaxSector && + (rdev->mddev->reshape_backwards == + (sector >= rdev->mddev->reshape_position))) + bio->bi_sector = sector + rdev->new_data_offset; else bio->bi_sector = sector + rdev->data_offset; bio_add_page(bio, page, size, 0); @@ -1034,12 +1042,17 @@ static unsigned int calc_sb_csum(mdp_super_t * sb) struct super_type { char *name; struct module *owner; - int (*load_super)(struct md_rdev *rdev, struct md_rdev *refdev, + int (*load_super)(struct md_rdev *rdev, + struct md_rdev *refdev, int minor_version); - int (*validate_super)(struct mddev *mddev, struct md_rdev *rdev); - void (*sync_super)(struct mddev *mddev, struct md_rdev *rdev); + int (*validate_super)(struct mddev *mddev, + struct md_rdev *rdev); + void (*sync_super)(struct mddev *mddev, + struct md_rdev *rdev); unsigned long long (*rdev_size_change)(struct md_rdev *rdev, sector_t num_sectors); + int (*allow_new_offset)(struct md_rdev *rdev, + unsigned long long new_offset); }; /* @@ -1111,6 +1124,7 @@ static int super_90_load(struct md_rdev *rdev, struct md_rdev *refdev, int minor rdev->preferred_minor = sb->md_minor; rdev->data_offset = 0; + rdev->new_data_offset = 0; rdev->sb_size = MD_SB_BYTES; rdev->badblocks.shift = -1; @@ -1184,7 +1198,11 @@ static int super_90_validate(struct mddev *mddev, struct md_rdev *rdev) mddev->dev_sectors = ((sector_t)sb->size) * 2; mddev->events = ev1; mddev->bitmap_info.offset = 0; + mddev->bitmap_info.space = 0; + /* bitmap can use 60 K after the 4K superblocks */ mddev->bitmap_info.default_offset = MD_SB_BYTES >> 9; + mddev->bitmap_info.default_space = 64*2 - (MD_SB_BYTES >> 9); + mddev->reshape_backwards = 0; if (mddev->minor_version >= 91) { mddev->reshape_position = sb->reshape_position; @@ -1192,6 +1210,8 @@ static int super_90_validate(struct mddev *mddev, struct md_rdev *rdev) mddev->new_level = sb->new_level; mddev->new_layout = sb->new_layout; mddev->new_chunk_sectors = sb->new_chunk >> 9; + if (mddev->delta_disks < 0) + mddev->reshape_backwards = 1; } else { mddev->reshape_position = MaxSector; mddev->delta_disks = 0; @@ -1218,9 +1238,12 @@ static int super_90_validate(struct mddev *mddev, struct md_rdev *rdev) mddev->max_disks = MD_SB_DISKS; if (sb->state & (1<<MD_SB_BITMAP_PRESENT) && - mddev->bitmap_info.file == NULL) + mddev->bitmap_info.file == NULL) { mddev->bitmap_info.offset = mddev->bitmap_info.default_offset; + mddev->bitmap_info.space = + mddev->bitmap_info.space; + } } else if (mddev->pers == NULL) { /* Insist on good event counter while assembling, except @@ -1434,6 +1457,12 @@ super_90_rdev_size_change(struct md_rdev *rdev, sector_t num_sectors) return num_sectors; } +static int +super_90_allow_new_offset(struct md_rdev *rdev, unsigned long long new_offset) +{ + /* non-zero offset changes not possible with v0.90 */ + return new_offset == 0; +} /* * version 1 superblock @@ -1469,6 +1498,7 @@ static int super_1_load(struct md_rdev *rdev, struct md_rdev *refdev, int minor_ struct mdp_superblock_1 *sb; int ret; sector_t sb_start; + sector_t sectors; char b[BDEVNAME_SIZE], b2[BDEVNAME_SIZE]; int bmask; @@ -1523,9 +1553,18 @@ static int super_1_load(struct md_rdev *rdev, struct md_rdev *refdev, int minor_ bdevname(rdev->bdev,b)); return -EINVAL; } + if (sb->pad0 || + sb->pad3[0] || + memcmp(sb->pad3, sb->pad3+1, sizeof(sb->pad3) - sizeof(sb->pad3[1]))) + /* Some padding is non-zero, might be a new feature */ + return -EINVAL; rdev->preferred_minor = 0xffff; rdev->data_offset = le64_to_cpu(sb->data_offset); + rdev->new_data_offset = rdev->data_offset; + if ((le32_to_cpu(sb->feature_map) & MD_FEATURE_RESHAPE_ACTIVE) && + (le32_to_cpu(sb->feature_map) & MD_FEATURE_NEW_OFFSET)) + rdev->new_data_offset += (s32)le32_to_cpu(sb->new_offset); atomic_set(&rdev->corrected_errors, le32_to_cpu(sb->cnt_corrected_read)); rdev->sb_size = le32_to_cpu(sb->max_dev) * 2 + 256; @@ -1536,6 +1575,9 @@ static int super_1_load(struct md_rdev *rdev, struct md_rdev *refdev, int minor_ if (minor_version && rdev->data_offset < sb_start + (rdev->sb_size/512)) return -EINVAL; + if (minor_version + && rdev->new_data_offset < sb_start + (rdev->sb_size/512)) + return -EINVAL; if (sb->level == cpu_to_le32(LEVEL_MULTIPATH)) rdev->desc_nr = -1; @@ -1607,16 +1649,14 @@ static int super_1_load(struct md_rdev *rdev, struct md_rdev *refdev, int minor_ else ret = 0; } - if (minor_version) - rdev->sectors = (i_size_read(rdev->bdev->bd_inode) >> 9) - - le64_to_cpu(sb->data_offset); - else - rdev->sectors = rdev->sb_start; - if (rdev->sectors < le64_to_cpu(sb->data_size)) + if (minor_version) { + sectors = (i_size_read(rdev->bdev->bd_inode) >> 9); + sectors -= rdev->data_offset; + } else + sectors = rdev->sb_start; + if (sectors < le64_to_cpu(sb->data_size)) return -EINVAL; rdev->sectors = le64_to_cpu(sb->data_size); - if (le64_to_cpu(sb->size) > rdev->sectors) - return -EINVAL; return ret; } @@ -1644,17 +1684,37 @@ static int super_1_validate(struct mddev *mddev, struct md_rdev *rdev) mddev->dev_sectors = le64_to_cpu(sb->size); mddev->events = ev1; mddev->bitmap_info.offset = 0; + mddev->bitmap_info.space = 0; + /* Default location for bitmap is 1K after superblock + * using 3K - total of 4K + */ mddev->bitmap_info.default_offset = 1024 >> 9; - + mddev->bitmap_info.default_space = (4096-1024) >> 9; + mddev->reshape_backwards = 0; + mddev->recovery_cp = le64_to_cpu(sb->resync_offset); memcpy(mddev->uuid, sb->set_uuid, 16); mddev->max_disks = (4096-256)/2; if ((le32_to_cpu(sb->feature_map) & MD_FEATURE_BITMAP_OFFSET) && - mddev->bitmap_info.file == NULL ) + mddev->bitmap_info.file == NULL) { mddev->bitmap_info.offset = (__s32)le32_to_cpu(sb->bitmap_offset); + /* Metadata doesn't record how much space is available. + * For 1.0, we assume we can use up to the superblock + * if before, else to 4K beyond superblock. + * For others, assume no change is possible. + */ + if (mddev->minor_version > 0) + mddev->bitmap_info.space = 0; + else if (mddev->bitmap_info.offset > 0) + mddev->bitmap_info.space = + 8 - mddev->bitmap_info.offset; + else + mddev->bitmap_info.space = + -mddev->bitmap_info.offset; + } if ((le32_to_cpu(sb->feature_map) & MD_FEATURE_RESHAPE_ACTIVE)) { mddev->reshape_position = le64_to_cpu(sb->reshape_position); @@ -1662,6 +1722,11 @@ static int super_1_validate(struct mddev *mddev, struct md_rdev *rdev) mddev->new_level = le32_to_cpu(sb->new_level); mddev->new_layout = le32_to_cpu(sb->new_layout); mddev->new_chunk_sectors = le32_to_cpu(sb->new_chunk); + if (mddev->delta_disks < 0 || + (mddev->delta_disks == 0 && + (le32_to_cpu(sb->feature_map) + & MD_FEATURE_RESHAPE_BACKWARDS))) + mddev->reshape_backwards = 1; } else { mddev->reshape_position = MaxSector; mddev->delta_disks = 0; @@ -1735,7 +1800,6 @@ static void super_1_sync(struct mddev *mddev, struct md_rdev *rdev) sb->feature_map = 0; sb->pad0 = 0; sb->recovery_offset = cpu_to_le64(0); - memset(sb->pad1, 0, sizeof(sb->pad1)); memset(sb->pad3, 0, sizeof(sb->pad3)); sb->utime = cpu_to_le64((__u64)mddev->utime); @@ -1757,6 +1821,8 @@ static void super_1_sync(struct mddev *mddev, struct md_rdev *rdev) sb->devflags |= WriteMostly1; else sb->devflags &= ~WriteMostly1; + sb->data_offset = cpu_to_le64(rdev->data_offset); + sb->data_size = cpu_to_le64(rdev->sectors); if (mddev->bitmap && mddev->bitmap_info.file == NULL) { sb->bitmap_offset = cpu_to_le32((__u32)mddev->bitmap_info.offset); @@ -1781,6 +1847,16 @@ static void super_1_sync(struct mddev *mddev, struct md_rdev *rdev) sb->delta_disks = cpu_to_le32(mddev->delta_disks); sb->new_level = cpu_to_le32(mddev->new_level); sb->new_chunk = cpu_to_le32(mddev->new_chunk_sectors); + if (mddev->delta_disks == 0 && + mddev->reshape_backwards) + sb->feature_map + |= cpu_to_le32(MD_FEATURE_RESHAPE_BACKWARDS); + if (rdev->new_data_offset != rdev->data_offset) { + sb->feature_map + |= cpu_to_le32(MD_FEATURE_NEW_OFFSET); + sb->new_offset = cpu_to_le32((__u32)(rdev->new_data_offset + - rdev->data_offset)); + } } if (rdev->badblocks.count == 0) @@ -1857,6 +1933,8 @@ super_1_rdev_size_change(struct md_rdev *rdev, sector_t num_sectors) sector_t max_sectors; if (num_sectors && num_sectors < rdev->mddev->dev_sectors) return 0; /* component must fit device */ + if (rdev->data_offset != rdev->new_data_offset) + return 0; /* too confusing */ if (rdev->sb_start < rdev->data_offset) { /* minor versions 1 and 2; superblock before data */ max_sectors = i_size_read(rdev->bdev->bd_inode) >> 9; @@ -1884,6 +1962,40 @@ super_1_rdev_size_change(struct md_rdev *rdev, sector_t num_sectors) rdev->sb_page); md_super_wait(rdev->mddev); return num_sectors; + +} + +static int +super_1_allow_new_offset(struct md_rdev *rdev, + unsigned long long new_offset) +{ + /* All necessary checks on new >= old have been done */ + struct bitmap *bitmap; + if (new_offset >= rdev->data_offset) + return 1; + + /* with 1.0 metadata, there is no metadata to tread on + * so we can always move back */ + if (rdev->mddev->minor_version == 0) + return 1; + + /* otherwise we must be sure not to step on + * any metadata, so stay: + * 36K beyond start of superblock + * beyond end of badblocks + * beyond write-intent bitmap + */ + if (rdev->sb_start + (32+4)*2 > new_offset) + return 0; + bitmap = rdev->mddev->bitmap; + if (bitmap && !rdev->mddev->bitmap_info.file && + rdev->sb_start + rdev->mddev->bitmap_info.offset + + bitmap->storage.file_pages * (PAGE_SIZE>>9) > new_offset) + return 0; + if (rdev->badblocks.sector + rdev->badblocks.size > new_offset) + return 0; + + return 1; } static struct super_type super_types[] = { @@ -1894,6 +2006,7 @@ static struct super_type super_types[] = { .validate_super = super_90_validate, .sync_super = super_90_sync, .rdev_size_change = super_90_rdev_size_change, + .allow_new_offset = super_90_allow_new_offset, }, [1] = { .name = "md-1", @@ -1902,6 +2015,7 @@ static struct super_type super_types[] = { .validate_super = super_1_validate, .sync_super = super_1_sync, .rdev_size_change = super_1_rdev_size_change, + .allow_new_offset = super_1_allow_new_offset, }, }; @@ -2105,9 +2219,7 @@ static void unbind_rdev_from_array(struct md_rdev * rdev) sysfs_remove_link(&rdev->kobj, "block"); sysfs_put(rdev->sysfs_state); rdev->sysfs_state = NULL; - kfree(rdev->badblocks.page); rdev->badblocks.count = 0; - rdev->badblocks.page = NULL; /* We need to delay this, otherwise we can deadlock when * writing to 'remove' to "dev/state". We also need * to delay it due to rcu usage. @@ -2158,7 +2270,7 @@ static void export_rdev(struct md_rdev * rdev) bdevname(rdev->bdev,b)); if (rdev->mddev) MD_BUG(); - free_disk_sb(rdev); + md_rdev_clear(rdev); #ifndef MODULE if (test_bit(AutoDetected, &rdev->flags)) md_autodetect_dev(rdev->bdev->bd_dev); @@ -2809,9 +2921,8 @@ offset_show(struct md_rdev *rdev, char *page) static ssize_t offset_store(struct md_rdev *rdev, const char *buf, size_t len) { - char *e; - unsigned long long offset = simple_strtoull(buf, &e, 10); - if (e==buf || (*e && *e != '\n')) + unsigned long long offset; + if (strict_strtoull(buf, 10, &offset) < 0) return -EINVAL; if (rdev->mddev->pers && rdev->raid_disk >= 0) return -EBUSY; @@ -2820,12 +2931,70 @@ offset_store(struct md_rdev *rdev, const char *buf, size_t len) * can be sane */ return -EBUSY; rdev->data_offset = offset; + rdev->new_data_offset = offset; return len; } static struct rdev_sysfs_entry rdev_offset = __ATTR(offset, S_IRUGO|S_IWUSR, offset_show, offset_store); +static ssize_t new_offset_show(struct md_rdev *rdev, char *page) +{ + return sprintf(page, "%llu\n", + (unsigned long long)rdev->new_data_offset); +} + +static ssize_t new_offset_store(struct md_rdev *rdev, + const char *buf, size_t len) +{ + unsigned long long new_offset; + struct mddev *mddev = rdev->mddev; + + if (strict_strtoull(buf, 10, &new_offset) < 0) + return -EINVAL; + + if (mddev->sync_thread) + return -EBUSY; + if (new_offset == rdev->data_offset) + /* reset is always permitted */ + ; + else if (new_offset > rdev->data_offset) { + /* must not push array size beyond rdev_sectors */ + if (new_offset - rdev->data_offset + + mddev->dev_sectors > rdev->sectors) + return -E2BIG; + } + /* Metadata worries about other space details. */ + + /* decreasing the offset is inconsistent with a backwards + * reshape. + */ + if (new_offset < rdev->data_offset && + mddev->reshape_backwards) + return -EINVAL; + /* Increasing offset is inconsistent with forwards + * reshape. reshape_direction should be set to + * 'backwards' first. + */ + if (new_offset > rdev->data_offset && + !mddev->reshape_backwards) + return -EINVAL; + + if (mddev->pers && mddev->persistent && + !super_types[mddev->major_version] + .allow_new_offset(rdev, new_offset)) + return -E2BIG; + rdev->new_data_offset = new_offset; + if (new_offset > rdev->data_offset) + mddev->reshape_backwards = 1; + else if (new_offset < rdev->data_offset) + mddev->reshape_backwards = 0; + + return len; +} +static struct rdev_sysfs_entry rdev_new_offset = +__ATTR(new_offset, S_IRUGO|S_IWUSR, new_offset_show, new_offset_store); + static ssize_t rdev_size_show(struct md_rdev *rdev, char *page) { @@ -2870,6 +3039,8 @@ rdev_size_store(struct md_rdev *rdev, const char *buf, size_t len) if (strict_blocks_to_sectors(buf, §ors) < 0) return -EINVAL; + if (rdev->data_offset != rdev->new_data_offset) + return -EINVAL; /* too confusing */ if (my_mddev->pers && rdev->raid_disk >= 0) { if (my_mddev->persistent) { sectors = super_types[my_mddev->major_version]. @@ -3006,6 +3177,7 @@ static struct attribute *rdev_default_attrs[] = { &rdev_errors.attr, &rdev_slot.attr, &rdev_offset.attr, + &rdev_new_offset.attr, &rdev_size.attr, &rdev_recovery_start.attr, &rdev_bad_blocks.attr, @@ -3080,6 +3252,7 @@ int md_rdev_init(struct md_rdev *rdev) rdev->raid_disk = -1; rdev->flags = 0; rdev->data_offset = 0; + rdev->new_data_offset = 0; rdev->sb_events = 0; rdev->last_read_error.tv_sec = 0; rdev->last_read_error.tv_nsec = 0; @@ -3178,8 +3351,7 @@ static struct md_rdev *md_import_device(dev_t newdev, int super_format, int supe abort_free: if (rdev->bdev) unlock_rdev(rdev); - free_disk_sb(rdev); - kfree(rdev->badblocks.page); + md_rdev_clear(rdev); kfree(rdev); return ERR_PTR(err); } @@ -3419,6 +3591,7 @@ level_store(struct mddev *mddev, const char *buf, size_t len) mddev->new_chunk_sectors = mddev->chunk_sectors; mddev->raid_disks -= mddev->delta_disks; mddev->delta_disks = 0; + mddev->reshape_backwards = 0; module_put(pers->owner); printk(KERN_WARNING "md: %s: %s would not accept array\n", mdname(mddev), clevel); @@ -3492,6 +3665,7 @@ level_store(struct mddev *mddev, const char *buf, size_t len) mddev->layout = mddev->new_layout; mddev->chunk_sectors = mddev->new_chunk_sectors; mddev->delta_disks = 0; + mddev->reshape_backwards = 0; mddev->degraded = 0; if (mddev->pers->sync_request == NULL) { /* this is now an array without redundancy, so @@ -3501,10 +3675,8 @@ level_store(struct mddev *mddev, const char *buf, size_t len) del_timer_sync(&mddev->safemode_timer); } pers->run(mddev); - mddev_resume(mddev); set_bit(MD_CHANGE_DEVS, &mddev->flags); - set_bit(MD_RECOVERY_NEEDED, &mddev->recovery); - md_wakeup_thread(mddev->thread); + mddev_resume(mddev); sysfs_notify(&mddev->kobj, NULL, "level"); md_new_event(mddev); return rv; @@ -3582,9 +3754,20 @@ raid_disks_store(struct mddev *mddev, const char *buf, size_t len) if (mddev->pers) rv = update_raid_disks(mddev, n); else if (mddev->reshape_position != MaxSector) { + struct md_rdev *rdev; int olddisks = mddev->raid_disks - mddev->delta_disks; + + rdev_for_each(rdev, mddev) { + if (olddisks < n && + rdev->data_offset < rdev->new_data_offset) + return -EINVAL; + if (olddisks > n && + rdev->data_offset > rdev->new_data_offset) + return -EINVAL; + } mddev->delta_disks = n - olddisks; mddev->raid_disks = n; + mddev->reshape_backwards = (mddev->delta_disks < 0); } else mddev->raid_disks = n; return rv ? rv : len; @@ -3744,8 +3927,8 @@ array_state_show(struct mddev *mddev, char *page) return sprintf(page, "%s\n", array_states[st]); } -static int do_md_stop(struct mddev * mddev, int ro, int is_open); -static int md_set_readonly(struct mddev * mddev, int is_open); +static int do_md_stop(struct mddev * mddev, int ro, struct block_device *bdev); +static int md_set_readonly(struct mddev * mddev, struct block_device *bdev); static int do_md_run(struct mddev * mddev); static int restart_array(struct mddev *mddev); @@ -3761,14 +3944,14 @@ array_state_store(struct mddev *mddev, const char *buf, size_t len) /* stopping an active array */ if (atomic_read(&mddev->openers) > 0) return -EBUSY; - err = do_md_stop(mddev, 0, 0); + err = do_md_stop(mddev, 0, NULL); break; case inactive: /* stopping an active array */ if (mddev->pers) { if (atomic_read(&mddev->openers) > 0) return -EBUSY; - err = do_md_stop(mddev, 2, 0); + err = do_md_stop(mddev, 2, NULL); } else err = 0; /* already inactive */ break; @@ -3776,7 +3959,7 @@ array_state_store(struct mddev *mddev, const char *buf, size_t len) break; /* not supported yet */ case readonly: if (mddev->pers) - err = md_set_readonly(mddev, 0); + err = md_set_readonly(mddev, NULL); else { mddev->ro = 1; set_disk_ro(mddev->gendisk, 1); @@ -3786,7 +3969,7 @@ array_state_store(struct mddev *mddev, const char *buf, size_t len) case read_auto: if (mddev->pers) { if (mddev->ro == 0) - err = md_set_readonly(mddev, 0); + err = md_set_readonly(mddev, NULL); else if (mddev->ro == 1) err = restart_array(mddev); if (err == 0) { @@ -4266,7 +4449,8 @@ sync_completed_show(struct mddev *mddev, char *page) if (!test_bit(MD_RECOVERY_RUNNING, &mddev->recovery)) return sprintf(page, "none\n"); - if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery)) + if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery) || + test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery)) max_sectors = mddev->resync_max_sectors; else max_sectors = mddev->dev_sectors; @@ -4428,6 +4612,7 @@ reshape_position_show(struct mddev *mddev, char *page) static ssize_t reshape_position_store(struct mddev *mddev, const char *buf, size_t len) { + struct md_rdev *rdev; char *e; unsigned long long new = simple_strtoull(buf, &e, 10); if (mddev->pers) @@ -4436,9 +4621,12 @@ reshape_position_store(struct mddev *mddev, const char *buf, size_t len) return -EINVAL; mddev->reshape_position = new; mddev->delta_disks = 0; + mddev->reshape_backwards = 0; mddev->new_level = mddev->level; mddev->new_layout = mddev->layout; mddev->new_chunk_sectors = mddev->chunk_sectors; + rdev_for_each(rdev, mddev) + rdev->new_data_offset = rdev->data_offset; return len; } @@ -4447,6 +4635,42 @@ __ATTR(reshape_position, S_IRUGO|S_IWUSR, reshape_position_show, reshape_position_store); static ssize_t +reshape_direction_show(struct mddev *mddev, char *page) +{ + return sprintf(page, "%s\n", + mddev->reshape_backwards ? "backwards" : "forwards"); +} + +static ssize_t +reshape_direction_store(struct mddev *mddev, const char *buf, size_t len) +{ + int backwards = 0; + if (cmd_match(buf, "forwards")) + backwards = 0; + else if (cmd_match(buf, "backwards")) + backwards = 1; + else + return -EINVAL; + if (mddev->reshape_backwards == backwards) + return len; + + /* check if we are allowed to change */ + if (mddev->delta_disks) + return -EBUSY; + + if (mddev->persistent && + mddev->major_version == 0) + return -EINVAL; + + mddev->reshape_backwards = backwards; + return len; +} + +static struct md_sysfs_entry md_reshape_direction = +__ATTR(reshape_direction, S_IRUGO|S_IWUSR, reshape_direction_show, + reshape_direction_store); + +static ssize_t array_size_show(struct mddev *mddev, char *page) { if (mddev->external_size) @@ -4501,6 +4725,7 @@ static struct attribute *md_default_attrs[] = { &md_safe_delay.attr, &md_array_state.attr, &md_reshape_position.attr, + &md_reshape_direction.attr, &md_array_size.attr, &max_corr_read_errors.attr, NULL, @@ -4914,7 +5139,8 @@ int md_run(struct mddev *mddev) err = -EINVAL; mddev->pers->stop(mddev); } - if (err == 0 && mddev->pers->sync_request) { + if (err == 0 && mddev->pers->sync_request && + (mddev->bitmap_info.file || mddev->bitmap_info.offset)) { err = bitmap_create(mddev); if (err) { printk(KERN_ERR "%s: failed to create bitmap (%d)\n", @@ -5064,6 +5290,7 @@ static void md_clean(struct mddev *mddev) mddev->events = 0; mddev->can_decrease_events = 0; mddev->delta_disks = 0; + mddev->reshape_backwards = 0; mddev->new_level = LEVEL_NONE; mddev->new_layout = 0; mddev->new_chunk_sectors = 0; @@ -5079,6 +5306,7 @@ static void md_clean(struct mddev *mddev) mddev->merge_check_needed = 0; mddev->bitmap_info.offset = 0; mddev->bitmap_info.default_offset = 0; + mddev->bitmap_info.default_space = 0; mddev->bitmap_info.chunksize = 0; mddev->bitmap_info.daemon_sleep = 0; mddev->bitmap_info.max_write_behind = 0; @@ -5124,15 +5352,17 @@ void md_stop(struct mddev *mddev) } EXPORT_SYMBOL_GPL(md_stop); -static int md_set_readonly(struct mddev *mddev, int is_open) +static int md_set_readonly(struct mddev *mddev, struct block_device *bdev) { int err = 0; mutex_lock(&mddev->open_mutex); - if (atomic_read(&mddev->openers) > is_open) { + if (atomic_read(&mddev->openers) > !!bdev) { printk("md: %s still in use.\n",mdname(mddev)); err = -EBUSY; goto out; } + if (bdev) + sync_blockdev(bdev); if (mddev->pers) { __md_stop_writes(mddev); @@ -5154,18 +5384,26 @@ out: * 0 - completely stop and dis-assemble array * 2 - stop but do not disassemble array */ -static int do_md_stop(struct mddev * mddev, int mode, int is_open) +static int do_md_stop(struct mddev * mddev, int mode, + struct block_device *bdev) { struct gendisk *disk = mddev->gendisk; struct md_rdev *rdev; mutex_lock(&mddev->open_mutex); - if (atomic_read(&mddev->openers) > is_open || + if (atomic_read(&mddev->openers) > !!bdev || mddev->sysfs_active) { printk("md: %s still in use.\n",mdname(mddev)); mutex_unlock(&mddev->open_mutex); return -EBUSY; } + if (bdev) + /* It is possible IO was issued on some other + * open file which was closed before we took ->open_mutex. + * As that was not the last close __blkdev_put will not + * have called sync_blockdev, so we must. + */ + sync_blockdev(bdev); if (mddev->pers) { if (mddev->ro) @@ -5239,7 +5477,7 @@ static void autorun_array(struct mddev *mddev) err = do_md_run(mddev); if (err) { printk(KERN_WARNING "md: do_md_run() returned %d\n", err); - do_md_stop(mddev, 0, 0); + do_md_stop(mddev, 0, NULL); } } @@ -5421,7 +5659,7 @@ static int get_bitmap_file(struct mddev * mddev, void __user * arg) goto out; /* bitmap disabled, zero the first byte and copy out */ - if (!mddev->bitmap || !mddev->bitmap->file) { + if (!mddev->bitmap || !mddev->bitmap->storage.file) { file->pathname[0] = '\0'; goto copy_out; } @@ -5430,7 +5668,8 @@ static int get_bitmap_file(struct mddev * mddev, void __user * arg) if (!buf) goto out; - ptr = d_path(&mddev->bitmap->file->f_path, buf, sizeof(file->pathname)); + ptr = d_path(&mddev->bitmap->storage.file->f_path, + buf, sizeof(file->pathname)); if (IS_ERR(ptr)) goto out; @@ -5556,8 +5795,7 @@ static int add_new_disk(struct mddev * mddev, mdu_disk_info_t *info) super_types[mddev->major_version]. validate_super(mddev, rdev); if ((info->state & (1<<MD_DISK_SYNC)) && - (!test_bit(In_sync, &rdev->flags) || - rdev->raid_disk != info->raid_disk)) { + rdev->raid_disk != info->raid_disk) { /* This was a hot-add request, but events doesn't * match, so reject it. */ @@ -5875,6 +6113,7 @@ static int set_array_info(struct mddev * mddev, mdu_array_info_t *info) set_bit(MD_CHANGE_DEVS, &mddev->flags); mddev->bitmap_info.default_offset = MD_SB_BYTES >> 9; + mddev->bitmap_info.default_space = 64*2 - (MD_SB_BYTES >> 9); mddev->bitmap_info.offset = 0; mddev->reshape_position = MaxSector; @@ -5888,6 +6127,7 @@ static int set_array_info(struct mddev * mddev, mdu_array_info_t *info) mddev->new_chunk_sectors = mddev->chunk_sectors; mddev->new_layout = mddev->layout; mddev->delta_disks = 0; + mddev->reshape_backwards = 0; return 0; } @@ -5922,11 +6162,7 @@ static int update_size(struct mddev *mddev, sector_t num_sectors) */ if (mddev->sync_thread) return -EBUSY; - if (mddev->bitmap) - /* Sorry, cannot grow a bitmap yet, just remove it, - * grow, and re-add. - */ - return -EBUSY; + rdev_for_each(rdev, mddev) { sector_t avail = rdev->sectors; @@ -5944,6 +6180,7 @@ static int update_size(struct mddev *mddev, sector_t num_sectors) static int update_raid_disks(struct mddev *mddev, int raid_disks) { int rv; + struct md_rdev *rdev; /* change the number of raid disks */ if (mddev->pers->check_reshape == NULL) return -EINVAL; @@ -5952,11 +6189,27 @@ static int update_raid_disks(struct mddev *mddev, int raid_disks) return -EINVAL; if (mddev->sync_thread || mddev->reshape_position != MaxSector) return -EBUSY; + + rdev_for_each(rdev, mddev) { + if (mddev->raid_disks < raid_disks && + rdev->data_offset < rdev->new_data_offset) + return -EINVAL; + if (mddev->raid_disks > raid_disks && + rdev->data_offset > rdev->new_data_offset) + return -EINVAL; + } + mddev->delta_disks = raid_disks - mddev->raid_disks; + if (mddev->delta_disks < 0) + mddev->reshape_backwards = 1; + else if (mddev->delta_disks > 0) + mddev->reshape_backwards = 0; rv = mddev->pers->check_reshape(mddev); - if (rv < 0) + if (rv < 0) { mddev->delta_disks = 0; + mddev->reshape_backwards = 0; + } return rv; } @@ -6039,6 +6292,8 @@ static int update_array_info(struct mddev *mddev, mdu_array_info_t *info) return -EINVAL; mddev->bitmap_info.offset = mddev->bitmap_info.default_offset; + mddev->bitmap_info.space = + mddev->bitmap_info.default_space; mddev->pers->quiesce(mddev, 1); rv = bitmap_create(mddev); if (!rv) @@ -6050,7 +6305,7 @@ static int update_array_info(struct mddev *mddev, mdu_array_info_t *info) /* remove the bitmap */ if (!mddev->bitmap) return -ENOENT; - if (mddev->bitmap->file) + if (mddev->bitmap->storage.file) return -EINVAL; mddev->pers->quiesce(mddev, 1); bitmap_destroy(mddev); @@ -6237,11 +6492,11 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode, goto done_unlock; case STOP_ARRAY: - err = do_md_stop(mddev, 0, 1); + err = do_md_stop(mddev, 0, bdev); goto done_unlock; case STOP_ARRAY_RO: - err = md_set_readonly(mddev, 1); + err = md_set_readonly(mddev, bdev); goto done_unlock; case BLKROSET: @@ -6373,6 +6628,9 @@ static int md_open(struct block_device *bdev, fmode_t mode) struct mddev *mddev = mddev_find(bdev->bd_dev); int err; + if (!mddev) + return -ENODEV; + if (mddev->gendisk != bdev->bd_disk) { /* we are racing with mddev_put which is discarding this * bd_disk. @@ -6503,7 +6761,7 @@ struct md_thread *md_register_thread(void (*run) (struct mddev *), struct mddev thread->tsk = kthread_run(md_thread, thread, "%s_%s", mdname(thread->mddev), - name ?: mddev->pers->name); + name); if (IS_ERR(thread->tsk)) { kfree(thread); return NULL; @@ -6584,7 +6842,8 @@ static void status_resync(struct seq_file *seq, struct mddev * mddev) resync = mddev->curr_resync - atomic_read(&mddev->recovery_active); - if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery)) + if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery) || + test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery)) max_sectors = mddev->resync_max_sectors; else max_sectors = mddev->dev_sectors; @@ -7049,6 +7308,7 @@ void md_do_sync(struct mddev *mddev) int skipped = 0; struct md_rdev *rdev; char *desc; + struct blk_plug plug; /* just incase thread restarts... */ if (test_bit(MD_RECOVERY_DONE, &mddev->recovery)) @@ -7147,7 +7407,7 @@ void md_do_sync(struct mddev *mddev) j = mddev->recovery_cp; } else if (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery)) - max_sectors = mddev->dev_sectors; + max_sectors = mddev->resync_max_sectors; else { /* recovery follows the physical size of devices */ max_sectors = mddev->dev_sectors; @@ -7198,6 +7458,7 @@ void md_do_sync(struct mddev *mddev) } mddev->curr_resync_completed = j; + blk_start_plug(&plug); while (j < max_sectors) { sector_t sectors; @@ -7303,6 +7564,7 @@ void md_do_sync(struct mddev *mddev) * this also signals 'finished resyncing' to md_stop */ out: + blk_finish_plug(&plug); wait_event(mddev->recovery_wait, !atomic_read(&mddev->recovery_active)); /* tell personality that we are finished */ @@ -7598,7 +7860,7 @@ void md_check_recovery(struct mddev *mddev) goto unlock; if (mddev->pers->sync_request) { - if (spares && mddev->bitmap && ! mddev->bitmap->file) { + if (spares) { /* We are adding a device or devices to an array * which has the bitmap stored on all devices. * So make sure all bitmap pages get written @@ -7646,6 +7908,20 @@ void md_wait_for_blocked_rdev(struct md_rdev *rdev, struct mddev *mddev) } EXPORT_SYMBOL(md_wait_for_blocked_rdev); +void md_finish_reshape(struct mddev *mddev) +{ + /* called be personality module when reshape completes. */ + struct md_rdev *rdev; + + rdev_for_each(rdev, mddev) { + if (rdev->data_offset > rdev->new_data_offset) + rdev->sectors += rdev->data_offset - rdev->new_data_offset; + else + rdev->sectors -= rdev->new_data_offset - rdev->data_offset; + rdev->data_offset = rdev->new_data_offset; + } +} +EXPORT_SYMBOL(md_finish_reshape); /* Bad block management. * We can record which blocks on each device are 'bad' and so just @@ -7894,10 +8170,15 @@ static int md_set_badblocks(struct badblocks *bb, sector_t s, int sectors, } int rdev_set_badblocks(struct md_rdev *rdev, sector_t s, int sectors, - int acknowledged) + int is_new) { - int rv = md_set_badblocks(&rdev->badblocks, - s + rdev->data_offset, sectors, acknowledged); + int rv; + if (is_new) + s += rdev->new_data_offset; + else + s += rdev->data_offset; + rv = md_set_badblocks(&rdev->badblocks, + s, sectors, 0); if (rv) { /* Make sure they get written out promptly */ sysfs_notify_dirent_safe(rdev->sysfs_state); @@ -8003,11 +8284,15 @@ out: return rv; } -int rdev_clear_badblocks(struct md_rdev *rdev, sector_t s, int sectors) +int rdev_clear_badblocks(struct md_rdev *rdev, sector_t s, int sectors, + int is_new) { + if (is_new) + s += rdev->new_data_offset; + else + s += rdev->data_offset; return md_clear_badblocks(&rdev->badblocks, - s + rdev->data_offset, - sectors); + s, sectors); } EXPORT_SYMBOL_GPL(rdev_clear_badblocks); |