kfree(fs_devices);
}
+static void btrfs_kobject_uevent(struct block_device *bdev,
+ enum kobject_action action)
+{
+ int ret;
+
+ ret = kobject_uevent(&disk_to_dev(bdev->bd_disk)->kobj, action);
+ if (ret)
+ pr_warn("Sending event '%d' to kobject: '%s' (%p): failed\n",
+ action,
+ kobject_name(&disk_to_dev(bdev->bd_disk)->kobj),
+ &disk_to_dev(bdev->bd_disk)->kobj);
+}
+
void btrfs_cleanup_fs_uuids(void)
{
struct btrfs_fs_devices *fs_devices;
return ret;
}
+/*
+ * Look for a btrfs signature on a device. This may be called out of the mount path
+ * and we are not allowed to call set_blocksize during the scan. The superblock
+ * is read via pagecache
+ */
int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder,
struct btrfs_fs_devices **fs_devices_ret)
{
struct btrfs_super_block *disk_super;
struct block_device *bdev;
- struct buffer_head *bh;
- int ret;
+ struct page *page;
+ void *p;
+ int ret = -EINVAL;
u64 devid;
u64 transid;
u64 total_devices;
+ u64 bytenr;
+ pgoff_t index;
+ /*
+ * we would like to check all the supers, but that would make
+ * a btrfs mount succeed after a mkfs from a different FS.
+ * So, we need to add a special mount option to scan for
+ * later supers, using BTRFS_SUPER_MIRROR_MAX instead
+ */
+ bytenr = btrfs_sb_offset(0);
flags |= FMODE_EXCL;
mutex_lock(&uuid_mutex);
- ret = btrfs_get_bdev_and_sb(path, flags, holder, 0, &bdev, &bh);
- if (ret)
+
+ bdev = blkdev_get_by_path(path, flags, holder);
+
+ if (IS_ERR(bdev)) {
+ ret = PTR_ERR(bdev);
+ printk(KERN_INFO "btrfs: open %s failed\n", path);
goto error;
- disk_super = (struct btrfs_super_block *)bh->b_data;
+ }
+
+ /* make sure our super fits in the device */
+ if (bytenr + PAGE_CACHE_SIZE >= i_size_read(bdev->bd_inode))
+ goto error_bdev_put;
+
+ /* make sure our super fits in the page */
+ if (sizeof(*disk_super) > PAGE_CACHE_SIZE)
+ goto error_bdev_put;
+
+ /* make sure our super doesn't straddle pages on disk */
+ index = bytenr >> PAGE_CACHE_SHIFT;
+ if ((bytenr + sizeof(*disk_super) - 1) >> PAGE_CACHE_SHIFT != index)
+ goto error_bdev_put;
+
+ /* pull in the page with our super */
+ page = read_cache_page_gfp(bdev->bd_inode->i_mapping,
+ index, GFP_NOFS);
+
+ if (IS_ERR_OR_NULL(page))
+ goto error_bdev_put;
+
+ p = kmap(page);
+
+ /* align our pointer to the offset of the super block */
+ disk_super = p + (bytenr & ~PAGE_CACHE_MASK);
+
+ if (btrfs_super_bytenr(disk_super) != bytenr ||
+ strncmp((char *)(&disk_super->magic), BTRFS_MAGIC,
+ sizeof(disk_super->magic)))
+ goto error_unmap;
+
devid = btrfs_stack_device_id(&disk_super->dev_item);
transid = btrfs_super_generation(disk_super);
total_devices = btrfs_super_num_devices(disk_super);
+
if (disk_super->label[0]) {
if (disk_super->label[BTRFS_LABEL_SIZE - 1])
disk_super->label[BTRFS_LABEL_SIZE - 1] = '\0';
} else {
printk(KERN_INFO "device fsid %pU ", disk_super->fsid);
}
+
printk(KERN_CONT "devid %llu transid %llu %s\n",
(unsigned long long)devid, (unsigned long long)transid, path);
+
ret = device_list_add(path, disk_super, devid, fs_devices_ret);
if (!ret && fs_devices_ret)
(*fs_devices_ret)->total_devices = total_devices;
- brelse(bh);
+
+error_unmap:
+ kunmap(page);
+ page_cache_release(page);
+
+error_bdev_put:
blkdev_put(bdev, flags);
error:
mutex_unlock(&uuid_mutex);
u64 devid;
u64 num_devices;
u8 *dev_uuid;
+ unsigned seq;
int ret = 0;
bool clear_super = false;
mutex_lock(&uuid_mutex);
- all_avail = root->fs_info->avail_data_alloc_bits |
- root->fs_info->avail_system_alloc_bits |
- root->fs_info->avail_metadata_alloc_bits;
+ do {
+ seq = read_seqbegin(&root->fs_info->profiles_lock);
+
+ all_avail = root->fs_info->avail_data_alloc_bits |
+ root->fs_info->avail_system_alloc_bits |
+ root->fs_info->avail_metadata_alloc_bits;
+ } while (read_seqretry(&root->fs_info->profiles_lock, seq));
num_devices = root->fs_info->fs_devices->num_devices;
btrfs_dev_replace_lock(&root->fs_info->dev_replace);
}
} else {
ret = btrfs_get_bdev_and_sb(device_path,
- FMODE_READ | FMODE_EXCL,
+ FMODE_WRITE | FMODE_EXCL,
root->fs_info->bdev_holder, 0,
&bdev, &bh);
if (ret)
ret = 0;
+ /* Notify udev that device has changed */
+ if (bdev)
+ btrfs_kobject_uevent(bdev, KOBJ_CHANGE);
+
error_brelse:
brelse(bh);
-error_close:
if (bdev)
blkdev_put(bdev, FMODE_READ | FMODE_EXCL);
out:
cache = btrfs_lookup_block_group(fs_info, chunk_offset);
chunk_used = btrfs_block_group_used(&cache->item);
- user_thresh = div_factor_fine(cache->key.offset, bargs->usage);
+ if (bargs->usage == 0)
+ user_thresh = 0;
+ else if (bargs->usage > 100)
+ user_thresh = cache->key.offset;
+ else
+ user_thresh = div_factor_fine(cache->key.offset,
+ bargs->usage);
+
if (chunk_used < user_thresh)
ret = 0;
unset_balance_control(fs_info);
ret = del_balance_item(fs_info->tree_root);
BUG_ON(ret);
+
+ atomic_set(&fs_info->mutually_exclusive_operation_running, 0);
}
void update_ioctl_balance_args(struct btrfs_fs_info *fs_info, int lock,
int mixed = 0;
int ret;
u64 num_devices;
+ unsigned seq;
if (btrfs_fs_closing(fs_info) ||
atomic_read(&fs_info->balance_pause_req) ||
/* allow to reduce meta or sys integrity only if force set */
allowed = BTRFS_BLOCK_GROUP_DUP | BTRFS_BLOCK_GROUP_RAID1 |
BTRFS_BLOCK_GROUP_RAID10;
- if (((bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
- (fs_info->avail_system_alloc_bits & allowed) &&
- !(bctl->sys.target & allowed)) ||
- ((bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
- (fs_info->avail_metadata_alloc_bits & allowed) &&
- !(bctl->meta.target & allowed))) {
- if (bctl->flags & BTRFS_BALANCE_FORCE) {
- printk(KERN_INFO "btrfs: force reducing metadata "
- "integrity\n");
- } else {
- printk(KERN_ERR "btrfs: balance will reduce metadata "
- "integrity, use force if you want this\n");
- ret = -EINVAL;
- goto out;
+ do {
+ seq = read_seqbegin(&fs_info->profiles_lock);
+
+ if (((bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
+ (fs_info->avail_system_alloc_bits & allowed) &&
+ !(bctl->sys.target & allowed)) ||
+ ((bctl->meta.flags & BTRFS_BALANCE_ARGS_CONVERT) &&
+ (fs_info->avail_metadata_alloc_bits & allowed) &&
+ !(bctl->meta.target & allowed))) {
+ if (bctl->flags & BTRFS_BALANCE_FORCE) {
+ printk(KERN_INFO "btrfs: force reducing metadata "
+ "integrity\n");
+ } else {
+ printk(KERN_ERR "btrfs: balance will reduce metadata "
+ "integrity, use force if you want this\n");
+ ret = -EINVAL;
+ goto out;
+ }
}
- }
+ } while (read_seqretry(&fs_info->profiles_lock, seq));
if (bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) {
int num_tolerated_disk_barrier_failures;
mutex_lock(&fs_info->balance_mutex);
atomic_dec(&fs_info->balance_running);
+ if (bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) {
+ fs_info->num_tolerated_disk_barrier_failures =
+ btrfs_calc_num_tolerated_disk_barrier_failures(fs_info);
+ }
+
if (bargs) {
memset(bargs, 0, sizeof(*bargs));
update_ioctl_balance_args(fs_info, 0, bargs);
__cancel_balance(fs_info);
}
- if (bctl->sys.flags & BTRFS_BALANCE_ARGS_CONVERT) {
- fs_info->num_tolerated_disk_barrier_failures =
- btrfs_calc_num_tolerated_disk_barrier_failures(fs_info);
- }
-
wake_up(&fs_info->balance_wait_q);
return ret;
out:
if (bctl->flags & BTRFS_BALANCE_RESUME)
__cancel_balance(fs_info);
- else
+ else {
kfree(bctl);
+ atomic_set(&fs_info->mutually_exclusive_operation_running, 0);
+ }
return ret;
}
ret = btrfs_balance(fs_info->balance_ctl, NULL);
}
- atomic_set(&fs_info->mutually_exclusive_operation_running, 0);
mutex_unlock(&fs_info->balance_mutex);
mutex_unlock(&fs_info->volume_mutex);
return 0;
}
- WARN_ON(atomic_xchg(&fs_info->mutually_exclusive_operation_running, 1));
tsk = kthread_run(balance_kthread, fs_info, "btrfs-balance");
if (IS_ERR(tsk))
return PTR_ERR(tsk);
btrfs_balance_sys(leaf, item, &disk_bargs);
btrfs_disk_balance_args_to_cpu(&bctl->sys, &disk_bargs);
+ WARN_ON(atomic_xchg(&fs_info->mutually_exclusive_operation_running, 1));
+
mutex_lock(&fs_info->volume_mutex);
mutex_lock(&fs_info->balance_mutex);
return 0;
}
+struct btrfs_raid_attr btrfs_raid_array[BTRFS_NR_RAID_TYPES] = {
+ [BTRFS_RAID_RAID10] = {
+ .sub_stripes = 2,
+ .dev_stripes = 1,
+ .devs_max = 0, /* 0 == as many as possible */
+ .devs_min = 4,
+ .devs_increment = 2,
+ .ncopies = 2,
+ },
+ [BTRFS_RAID_RAID1] = {
+ .sub_stripes = 1,
+ .dev_stripes = 1,
+ .devs_max = 2,
+ .devs_min = 2,
+ .devs_increment = 2,
+ .ncopies = 2,
+ },
+ [BTRFS_RAID_DUP] = {
+ .sub_stripes = 1,
+ .dev_stripes = 2,
+ .devs_max = 1,
+ .devs_min = 1,
+ .devs_increment = 1,
+ .ncopies = 2,
+ },
+ [BTRFS_RAID_RAID0] = {
+ .sub_stripes = 1,
+ .dev_stripes = 1,
+ .devs_max = 0,
+ .devs_min = 2,
+ .devs_increment = 1,
+ .ncopies = 1,
+ },
+ [BTRFS_RAID_SINGLE] = {
+ .sub_stripes = 1,
+ .dev_stripes = 1,
+ .devs_max = 1,
+ .devs_min = 1,
+ .devs_increment = 1,
+ .ncopies = 1,
+ },
+};
+
static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans,
struct btrfs_root *extent_root,
struct map_lookup **map_ret,
int ndevs;
int i;
int j;
+ int index;
BUG_ON(!alloc_profile_is_valid(type, 0));
if (list_empty(&fs_devices->alloc_list))
return -ENOSPC;
- sub_stripes = 1;
- dev_stripes = 1;
- devs_increment = 1;
- ncopies = 1;
- devs_max = 0; /* 0 == as many as possible */
- devs_min = 1;
+ index = __get_raid_index(type);
- /*
- * define the properties of each RAID type.
- * FIXME: move this to a global table and use it in all RAID
- * calculation code
- */
- if (type & (BTRFS_BLOCK_GROUP_DUP)) {
- dev_stripes = 2;
- ncopies = 2;
- devs_max = 1;
- } else if (type & (BTRFS_BLOCK_GROUP_RAID0)) {
- devs_min = 2;
- } else if (type & (BTRFS_BLOCK_GROUP_RAID1)) {
- devs_increment = 2;
- ncopies = 2;
- devs_max = 2;
- devs_min = 2;
- } else if (type & (BTRFS_BLOCK_GROUP_RAID10)) {
- sub_stripes = 2;
- devs_increment = 2;
- ncopies = 2;
- devs_min = 4;
- } else {
- devs_max = 1;
- }
+ sub_stripes = btrfs_raid_array[index].sub_stripes;
+ dev_stripes = btrfs_raid_array[index].dev_stripes;
+ devs_max = btrfs_raid_array[index].devs_max;
+ devs_min = btrfs_raid_array[index].devs_min;
+ devs_increment = btrfs_raid_array[index].devs_increment;
+ ncopies = btrfs_raid_array[index].ncopies;
if (type & BTRFS_BLOCK_GROUP_DATA) {
max_stripe_size = 1024 * 1024 * 1024;
if (max_avail < BTRFS_STRIPE_LEN * dev_stripes)
continue;
+ if (ndevs == fs_devices->rw_devices) {
+ WARN(1, "%s: found more than %llu devices\n",
+ __func__, fs_devices->rw_devices);
+ break;
+ }
devices_info[ndevs].dev_offset = dev_offset;
devices_info[ndevs].max_avail = max_avail;
devices_info[ndevs].total_avail = total_avail;
devices_info[ndevs].dev = device;
++ndevs;
- WARN_ON(ndevs > fs_devices->rw_devices);
}
/*
write_lock(&em_tree->lock);
ret = add_extent_mapping(em_tree, em);
write_unlock(&em_tree->lock);
- free_extent_map(em);
- if (ret)
- goto error;
-
- ret = btrfs_make_block_group(trans, extent_root, 0, type,
- BTRFS_FIRST_CHUNK_TREE_OBJECTID,
- start, num_bytes);
- if (ret)
+ if (ret) {
+ free_extent_map(em);
goto error;
+ }
for (i = 0; i < map->num_stripes; ++i) {
struct btrfs_device *device;
info->chunk_root->root_key.objectid,
BTRFS_FIRST_CHUNK_TREE_OBJECTID,
start, dev_offset, stripe_size);
- if (ret) {
- btrfs_abort_transaction(trans, extent_root, ret);
- goto error;
- }
+ if (ret)
+ goto error_dev_extent;
}
+ ret = btrfs_make_block_group(trans, extent_root, 0, type,
+ BTRFS_FIRST_CHUNK_TREE_OBJECTID,
+ start, num_bytes);
+ if (ret) {
+ i = map->num_stripes - 1;
+ goto error_dev_extent;
+ }
+
+ free_extent_map(em);
kfree(devices_info);
return 0;
+error_dev_extent:
+ for (; i >= 0; i--) {
+ struct btrfs_device *device;
+ int err;
+
+ device = map->stripes[i].dev;
+ err = btrfs_free_dev_extent(trans, device, start);
+ if (err) {
+ btrfs_abort_transaction(trans, extent_root, err);
+ break;
+ }
+ }
+ write_lock(&em_tree->lock);
+ remove_extent_mapping(em_tree, em);
+ write_unlock(&em_tree->lock);
+
+ /* One for our allocation */
+ free_extent_map(em);
+ /* One for the tree reference */
+ free_extent_map(em);
error:
kfree(map);
kfree(devices_info);
if (ret)
return ret;
- alloc_profile = BTRFS_BLOCK_GROUP_METADATA |
- fs_info->avail_metadata_alloc_bits;
- alloc_profile = btrfs_reduce_alloc_profile(root, alloc_profile);
-
+ alloc_profile = btrfs_get_alloc_profile(extent_root, 0);
ret = __btrfs_alloc_chunk(trans, extent_root, &map, &chunk_size,
&stripe_size, chunk_offset, alloc_profile);
if (ret)
sys_chunk_offset = chunk_offset + chunk_size;
- alloc_profile = BTRFS_BLOCK_GROUP_SYSTEM |
- fs_info->avail_system_alloc_bits;
- alloc_profile = btrfs_reduce_alloc_profile(root, alloc_profile);
-
+ alloc_profile = btrfs_get_alloc_profile(fs_info->chunk_root, 0);
ret = __btrfs_alloc_chunk(trans, extent_root, &sys_map,
&sys_chunk_size, &sys_stripe_size,
sys_chunk_offset, alloc_profile);
else
ret = 1;
free_extent_map(em);
+
+ btrfs_dev_replace_lock(&fs_info->dev_replace);
+ if (btrfs_dev_replace_is_ongoing(&fs_info->dev_replace))
+ ret++;
+ btrfs_dev_replace_unlock(&fs_info->dev_replace);
+
return ret;
}
struct btrfs_dev_replace *dev_replace = &fs_info->dev_replace;
int dev_replace_is_ongoing = 0;
int num_alloc_stripes;
+ int patch_the_first_stripe_for_dev_replace = 0;
+ u64 physical_to_patch_in_first_stripe = 0;
read_lock(&em_tree->lock);
em = lookup_extent_mapping(em_tree, logical, *length);
map = (struct map_lookup *)em->bdev;
offset = logical - em->start;
- if (mirror_num > map->num_stripes)
- mirror_num = 0;
-
stripe_nr = offset;
/*
* stripe_nr counts the total number of stripes we have to stride
if (!dev_replace_is_ongoing)
btrfs_dev_replace_unlock(dev_replace);
+ if (dev_replace_is_ongoing && mirror_num == map->num_stripes + 1 &&
+ !(rw & (REQ_WRITE | REQ_DISCARD | REQ_GET_READ_MIRRORS)) &&
+ dev_replace->tgtdev != NULL) {
+ /*
+ * in dev-replace case, for repair case (that's the only
+ * case where the mirror is selected explicitly when
+ * calling btrfs_map_block), blocks left of the left cursor
+ * can also be read from the target drive.
+ * For REQ_GET_READ_MIRRORS, the target drive is added as
+ * the last one to the array of stripes. For READ, it also
+ * needs to be supported using the same mirror number.
+ * If the requested block is not left of the left cursor,
+ * EIO is returned. This can happen because btrfs_num_copies()
+ * returns one more in the dev-replace case.
+ */
+ u64 tmp_length = *length;
+ struct btrfs_bio *tmp_bbio = NULL;
+ int tmp_num_stripes;
+ u64 srcdev_devid = dev_replace->srcdev->devid;
+ int index_srcdev = 0;
+ int found = 0;
+ u64 physical_of_found = 0;
+
+ ret = __btrfs_map_block(fs_info, REQ_GET_READ_MIRRORS,
+ logical, &tmp_length, &tmp_bbio, 0);
+ if (ret) {
+ WARN_ON(tmp_bbio != NULL);
+ goto out;
+ }
+
+ tmp_num_stripes = tmp_bbio->num_stripes;
+ if (mirror_num > tmp_num_stripes) {
+ /*
+ * REQ_GET_READ_MIRRORS does not contain this
+ * mirror, that means that the requested area
+ * is not left of the left cursor
+ */
+ ret = -EIO;
+ kfree(tmp_bbio);
+ goto out;
+ }
+
+ /*
+ * process the rest of the function using the mirror_num
+ * of the source drive. Therefore look it up first.
+ * At the end, patch the device pointer to the one of the
+ * target drive.
+ */
+ for (i = 0; i < tmp_num_stripes; i++) {
+ if (tmp_bbio->stripes[i].dev->devid == srcdev_devid) {
+ /*
+ * In case of DUP, in order to keep it
+ * simple, only add the mirror with the
+ * lowest physical address
+ */
+ if (found &&
+ physical_of_found <=
+ tmp_bbio->stripes[i].physical)
+ continue;
+ index_srcdev = i;
+ found = 1;
+ physical_of_found =
+ tmp_bbio->stripes[i].physical;
+ }
+ }
+
+ if (found) {
+ mirror_num = index_srcdev + 1;
+ patch_the_first_stripe_for_dev_replace = 1;
+ physical_to_patch_in_first_stripe = physical_of_found;
+ } else {
+ WARN_ON(1);
+ ret = -EIO;
+ kfree(tmp_bbio);
+ goto out;
+ }
+
+ kfree(tmp_bbio);
+ } else if (mirror_num > map->num_stripes) {
+ mirror_num = 0;
+ }
+
num_stripes = 1;
stripe_index = 0;
stripe_nr_orig = stripe_nr;
BUG_ON(stripe_index >= map->num_stripes);
num_alloc_stripes = num_stripes;
- if (dev_replace_is_ongoing && (rw & (REQ_WRITE | REQ_DISCARD)))
- num_alloc_stripes <<= 1;
+ if (dev_replace_is_ongoing) {
+ if (rw & (REQ_WRITE | REQ_DISCARD))
+ num_alloc_stripes <<= 1;
+ if (rw & REQ_GET_READ_MIRRORS)
+ num_alloc_stripes++;
+ }
bbio = kzalloc(btrfs_bio_size(num_alloc_stripes), GFP_NOFS);
if (!bbio) {
ret = -ENOMEM;
}
}
num_stripes = index_where_to_add;
+ } else if (dev_replace_is_ongoing && (rw & REQ_GET_READ_MIRRORS) &&
+ dev_replace->tgtdev != NULL) {
+ u64 srcdev_devid = dev_replace->srcdev->devid;
+ int index_srcdev = 0;
+ int found = 0;
+ u64 physical_of_found = 0;
+
+ /*
+ * During the dev-replace procedure, the target drive can
+ * also be used to read data in case it is needed to repair
+ * a corrupt block elsewhere. This is possible if the
+ * requested area is left of the left cursor. In this area,
+ * the target drive is a full copy of the source drive.
+ */
+ for (i = 0; i < num_stripes; i++) {
+ if (bbio->stripes[i].dev->devid == srcdev_devid) {
+ /*
+ * In case of DUP, in order to keep it
+ * simple, only add the mirror with the
+ * lowest physical address
+ */
+ if (found &&
+ physical_of_found <=
+ bbio->stripes[i].physical)
+ continue;
+ index_srcdev = i;
+ found = 1;
+ physical_of_found = bbio->stripes[i].physical;
+ }
+ }
+ if (found) {
+ u64 length = map->stripe_len;
+
+ if (physical_of_found + length <=
+ dev_replace->cursor_left) {
+ struct btrfs_bio_stripe *tgtdev_stripe =
+ bbio->stripes + num_stripes;
+
+ tgtdev_stripe->physical = physical_of_found;
+ tgtdev_stripe->length =
+ bbio->stripes[index_srcdev].length;
+ tgtdev_stripe->dev = dev_replace->tgtdev;
+
+ num_stripes++;
+ }
+ }
}
*bbio_ret = bbio;
bbio->num_stripes = num_stripes;
bbio->max_errors = max_errors;
bbio->mirror_num = mirror_num;
+
+ /*
+ * this is the case that REQ_READ && dev_replace_is_ongoing &&
+ * mirror_num == num_stripes + 1 && dev_replace target drive is
+ * available as a mirror
+ */
+ if (patch_the_first_stripe_for_dev_replace && num_stripes > 0) {
+ WARN_ON(num_stripes > 1);
+ bbio->stripes[0].dev = dev_replace->tgtdev;
+ bbio->stripes[0].physical = physical_to_patch_in_first_stripe;
+ bbio->mirror_num = map->num_stripes + 1;
+ }
out:
if (dev_replace_is_ongoing)
btrfs_dev_replace_unlock(dev_replace);
em->bdev = (struct block_device *)map;
em->start = logical;
em->len = length;
+ em->orig_start = 0;
em->block_start = 0;
em->block_len = em->len;
projects / deliverable / linux.git / blobdiff