u64 num_bytes, int reserve);
static int block_rsv_use_bytes(struct btrfs_block_rsv *block_rsv,
u64 num_bytes);
+int btrfs_pin_extent(struct btrfs_root *root,
+ u64 bytenr, u64 num_bytes, int reserved);
static noinline int
block_group_cache_done(struct btrfs_block_group_cache *cache)
return ret;
while (nr--) {
- cache->bytes_super += stripe_len;
- ret = add_excluded_extent(root, logical[nr],
- stripe_len);
+ u64 start, len;
+
+ if (logical[nr] > cache->key.objectid +
+ cache->key.offset)
+ continue;
+
+ if (logical[nr] + stripe_len <= cache->key.objectid)
+ continue;
+
+ start = logical[nr];
+ if (start < cache->key.objectid) {
+ start = cache->key.objectid;
+ len = (logical[nr] + stripe_len) - start;
+ } else {
+ len = min_t(u64, stripe_len,
+ cache->key.objectid +
+ cache->key.offset - start);
+ }
+
+ cache->bytes_super += len;
+ ret = add_excluded_extent(root, start, len);
if (ret) {
kfree(logical);
return ret;
if (ret)
break;
- if (need_resched() ||
- btrfs_next_leaf(extent_root, path)) {
+ if (need_resched()) {
caching_ctl->progress = last;
btrfs_release_path(path);
up_read(&fs_info->extent_commit_sem);
cond_resched();
goto again;
}
+
+ ret = btrfs_next_leaf(extent_root, path);
+ if (ret < 0)
+ goto err;
+ if (ret)
+ break;
leaf = path->nodes[0];
nritems = btrfs_header_nritems(leaf);
continue;
rcu_read_unlock();
}
-u64 btrfs_find_block_group(struct btrfs_root *root,
- u64 search_start, u64 search_hint, int owner)
-{
- struct btrfs_block_group_cache *cache;
- u64 used;
- u64 last = max(search_hint, search_start);
- u64 group_start = 0;
- int full_search = 0;
- int factor = 9;
- int wrapped = 0;
-again:
- while (1) {
- cache = btrfs_lookup_first_block_group(root->fs_info, last);
- if (!cache)
- break;
-
- spin_lock(&cache->lock);
- last = cache->key.objectid + cache->key.offset;
- used = btrfs_block_group_used(&cache->item);
-
- if ((full_search || !cache->ro) &&
- block_group_bits(cache, BTRFS_BLOCK_GROUP_METADATA)) {
- if (used + cache->pinned + cache->reserved <
- div_factor(cache->key.offset, factor)) {
- group_start = cache->key.objectid;
- spin_unlock(&cache->lock);
- btrfs_put_block_group(cache);
- goto found;
- }
- }
- spin_unlock(&cache->lock);
- btrfs_put_block_group(cache);
- cond_resched();
- }
- if (!wrapped) {
- last = search_start;
- wrapped = 1;
- goto again;
- }
- if (!full_search && factor < 10) {
- last = search_start;
- full_search = 1;
- factor = 10;
- goto again;
- }
-found:
- return group_start;
-}
-
/* simple helper to search for an existing extent at a given offset */
int btrfs_lookup_extent(struct btrfs_root *root, u64 start, u64 len)
{
* helper to add new inline back ref
*/
static noinline_for_stack
-void setup_inline_extent_backref(struct btrfs_trans_handle *trans,
- struct btrfs_root *root,
+void setup_inline_extent_backref(struct btrfs_root *root,
struct btrfs_path *path,
struct btrfs_extent_inline_ref *iref,
u64 parent, u64 root_objectid,
update_inline_extent_backref(root, path, iref,
refs_to_add, extent_op);
} else if (ret == -ENOENT) {
- setup_inline_extent_backref(trans, root, path, iref, parent,
+ setup_inline_extent_backref(root, path, iref, parent,
root_objectid, owner, offset,
refs_to_add, extent_op);
ret = 0;
u32 item_size;
int ret;
int err = 0;
- int metadata = (node->type == BTRFS_TREE_BLOCK_REF_KEY ||
- node->type == BTRFS_SHARED_BLOCK_REF_KEY);
+ int metadata = !extent_op->is_data;
if (trans->aborted)
return 0;
key.objectid = node->bytenr;
if (metadata) {
- struct btrfs_delayed_tree_ref *tree_ref;
-
- tree_ref = btrfs_delayed_node_to_tree_ref(node);
key.type = BTRFS_METADATA_ITEM_KEY;
- key.offset = tree_ref->level;
+ key.offset = extent_op->level;
} else {
key.type = BTRFS_EXTENT_ITEM_KEY;
key.offset = node->num_bytes;
!trans->delayed_ref_elem.seq) {
/* list without seq or seq without list */
btrfs_err(fs_info,
- "qgroup accounting update error, list is%s empty, seq is %llu",
+ "qgroup accounting update error, list is%s empty, seq is %#x.%x",
list_empty(&trans->qgroup_ref_list) ? "" : " not",
- trans->delayed_ref_elem.seq);
+ (u32)(trans->delayed_ref_elem.seq >> 32),
+ (u32)trans->delayed_ref_elem.seq);
BUG();
}
int btrfs_set_disk_extent_flags(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
u64 bytenr, u64 num_bytes, u64 flags,
- int is_data)
+ int level, int is_data)
{
struct btrfs_delayed_extent_op *extent_op;
int ret;
extent_op->update_flags = 1;
extent_op->update_key = 0;
extent_op->is_data = is_data ? 1 : 0;
+ extent_op->level = level;
ret = btrfs_add_delayed_extent_op(root->fs_info, trans, bytenr,
num_bytes, extent_op);
WARN_ON(ret);
if (i_size_read(inode) > 0) {
+ ret = btrfs_check_trunc_cache_free_space(root,
+ &root->fs_info->global_block_rsv);
+ if (ret)
+ goto out_put;
+
ret = btrfs_truncate_free_space_cache(root, trans, path,
inode);
if (ret)
* progress (either running or paused) picks the target profile (if it's
* already available), otherwise falls back to plain reducing.
*/
-u64 btrfs_reduce_alloc_profile(struct btrfs_root *root, u64 flags)
+static u64 btrfs_reduce_alloc_profile(struct btrfs_root *root, u64 flags)
{
/*
* we add in the count of missing devices because we want
rcu_read_unlock();
}
+static inline u64 calc_global_rsv_need_space(struct btrfs_block_rsv *global)
+{
+ return (global->size << 1);
+}
+
static int should_alloc_chunk(struct btrfs_root *root,
struct btrfs_space_info *sinfo, int force)
{
* global_rsv, it doesn't change except when the transaction commits.
*/
if (sinfo->flags & BTRFS_BLOCK_GROUP_METADATA)
- num_allocated += global_rsv->size;
+ num_allocated += calc_global_rsv_need_space(global_rsv);
/*
* in limited mode, we want to have some free space up to
{
struct btrfs_block_rsv *global_rsv = &root->fs_info->global_block_rsv;
u64 profile = btrfs_get_alloc_profile(root, 0);
- u64 rsv_size = 0;
+ u64 space_size;
u64 avail;
u64 used;
u64 to_add;
used = space_info->bytes_used + space_info->bytes_reserved +
space_info->bytes_pinned + space_info->bytes_readonly;
- spin_lock(&global_rsv->lock);
- rsv_size = global_rsv->size;
- spin_unlock(&global_rsv->lock);
-
/*
* We only want to allow over committing if we have lots of actual space
* free, but if we don't have enough space to handle the global reserve
* space then we could end up having a real enospc problem when trying
* to allocate a chunk or some other such important allocation.
*/
- rsv_size <<= 1;
- if (used + rsv_size >= space_info->total_bytes)
+ spin_lock(&global_rsv->lock);
+ space_size = calc_global_rsv_need_space(global_rsv);
+ spin_unlock(&global_rsv->lock);
+ if (used + space_size >= space_info->total_bytes)
return 0;
used += space_info->bytes_may_use;
return 0;
}
-void btrfs_writeback_inodes_sb_nr(struct btrfs_root *root,
- unsigned long nr_pages)
+static void btrfs_writeback_inodes_sb_nr(struct btrfs_root *root,
+ unsigned long nr_pages)
{
struct super_block *sb = root->fs_info->sb;
int started;
fs_info->csum_root->block_rsv = &fs_info->global_block_rsv;
fs_info->dev_root->block_rsv = &fs_info->global_block_rsv;
fs_info->tree_root->block_rsv = &fs_info->global_block_rsv;
+ if (fs_info->quota_root)
+ fs_info->quota_root->block_rsv = &fs_info->global_block_rsv;
fs_info->chunk_root->block_rsv = &fs_info->chunk_block_rsv;
update_global_block_rsv(fs_info);
u64 bytenr, u64 num_bytes)
{
struct btrfs_block_group_cache *cache;
+ int ret;
cache = btrfs_lookup_block_group(root->fs_info, bytenr);
- BUG_ON(!cache); /* Logic error */
+ if (!cache)
+ return -EINVAL;
/*
* pull in the free space cache (if any) so that our pin
pin_down_extent(root, cache, bytenr, num_bytes, 0);
/* remove us from the free space cache (if we're there at all) */
- btrfs_remove_free_space(cache, bytenr, num_bytes);
+ ret = btrfs_remove_free_space(cache, bytenr, num_bytes);
btrfs_put_block_group(cache);
- return 0;
+ return ret;
}
/**
}
refs = btrfs_extent_refs(leaf, ei);
- BUG_ON(refs < refs_to_drop);
+ if (refs < refs_to_drop) {
+ btrfs_err(info, "trying to drop %d refs but we only have %Lu "
+ "for bytenr %Lu\n", refs_to_drop, refs, bytenr);
+ ret = -EINVAL;
+ btrfs_abort_transaction(trans, extent_root, ret);
+ goto out;
+ }
refs -= refs_to_drop;
if (refs > 0) {
struct btrfs_root *orig_root,
u64 num_bytes, u64 empty_size,
u64 hint_byte, struct btrfs_key *ins,
- u64 data)
+ u64 flags)
{
int ret = 0;
struct btrfs_root *root = orig_root->fs_info->extent_root;
int empty_cluster = 2 * 1024 * 1024;
struct btrfs_space_info *space_info;
int loop = 0;
- int index = __get_raid_index(data);
- int alloc_type = (data & BTRFS_BLOCK_GROUP_DATA) ?
+ int index = __get_raid_index(flags);
+ int alloc_type = (flags & BTRFS_BLOCK_GROUP_DATA) ?
RESERVE_ALLOC_NO_ACCOUNT : RESERVE_ALLOC;
bool found_uncached_bg = false;
bool failed_cluster_refill = false;
ins->objectid = 0;
ins->offset = 0;
- trace_find_free_extent(orig_root, num_bytes, empty_size, data);
+ trace_find_free_extent(orig_root, num_bytes, empty_size, flags);
- space_info = __find_space_info(root->fs_info, data);
+ space_info = __find_space_info(root->fs_info, flags);
if (!space_info) {
- btrfs_err(root->fs_info, "No space info for %llu", data);
+ btrfs_err(root->fs_info, "No space info for %llu", flags);
return -ENOSPC;
}
if (btrfs_mixed_space_info(space_info))
use_cluster = false;
- if (data & BTRFS_BLOCK_GROUP_METADATA && use_cluster) {
+ if (flags & BTRFS_BLOCK_GROUP_METADATA && use_cluster) {
last_ptr = &root->fs_info->meta_alloc_cluster;
if (!btrfs_test_opt(root, SSD))
empty_cluster = 64 * 1024;
}
- if ((data & BTRFS_BLOCK_GROUP_DATA) && use_cluster &&
+ if ((flags & BTRFS_BLOCK_GROUP_DATA) && use_cluster &&
btrfs_test_opt(root, SSD)) {
last_ptr = &root->fs_info->data_alloc_cluster;
}
* However if we are re-searching with an ideal block group
* picked out then we don't care that the block group is cached.
*/
- if (block_group && block_group_bits(block_group, data) &&
+ if (block_group && block_group_bits(block_group, flags) &&
block_group->cached != BTRFS_CACHE_NO) {
down_read(&space_info->groups_sem);
if (list_empty(&block_group->list) ||
* raid types, but we want to make sure we only allocate
* for the proper type.
*/
- if (!block_group_bits(block_group, data)) {
+ if (!block_group_bits(block_group, flags)) {
u64 extra = BTRFS_BLOCK_GROUP_DUP |
BTRFS_BLOCK_GROUP_RAID1 |
BTRFS_BLOCK_GROUP_RAID5 |
* doesn't provide them, bail. This does allow us to
* fill raid0 from raid1.
*/
- if ((data & extra) && !(block_group->flags & extra))
+ if ((flags & extra) && !(block_group->flags & extra))
goto loop;
}
if (used_block_group != block_group &&
(!used_block_group ||
used_block_group->ro ||
- !block_group_bits(used_block_group, data))) {
+ !block_group_bits(used_block_group, flags))) {
used_block_group = block_group;
goto refill_cluster;
}
index = 0;
loop++;
if (loop == LOOP_ALLOC_CHUNK) {
- ret = do_chunk_alloc(trans, root, data,
+ ret = do_chunk_alloc(trans, root, flags,
CHUNK_ALLOC_FORCE);
/*
* Do not bail out on ENOSPC since we
struct btrfs_root *root,
u64 num_bytes, u64 min_alloc_size,
u64 empty_size, u64 hint_byte,
- struct btrfs_key *ins, u64 data)
+ struct btrfs_key *ins, int is_data)
{
bool final_tried = false;
+ u64 flags;
int ret;
- data = btrfs_get_alloc_profile(root, data);
+ flags = btrfs_get_alloc_profile(root, is_data);
again:
WARN_ON(num_bytes < root->sectorsize);
ret = find_free_extent(trans, root, num_bytes, empty_size,
- hint_byte, ins, data);
+ hint_byte, ins, flags);
if (ret == -ENOSPC) {
if (!final_tried) {
} else if (btrfs_test_opt(root, ENOSPC_DEBUG)) {
struct btrfs_space_info *sinfo;
- sinfo = __find_space_info(root->fs_info, data);
+ sinfo = __find_space_info(root->fs_info, flags);
btrfs_err(root->fs_info, "allocation failed flags %llu, wanted %llu",
- (unsigned long long)data,
+ (unsigned long long)flags,
(unsigned long long)num_bytes);
if (sinfo)
dump_space_info(sinfo, num_bytes, 1);
if (!caching_ctl) {
BUG_ON(!block_group_cache_done(block_group));
ret = btrfs_remove_free_space(block_group, start, num_bytes);
- BUG_ON(ret); /* -ENOMEM */
+ if (ret)
+ goto out;
} else {
mutex_lock(&caching_ctl->mutex);
if (start >= caching_ctl->progress) {
ret = add_excluded_extent(root, start, num_bytes);
- BUG_ON(ret); /* -ENOMEM */
} else if (start + num_bytes <= caching_ctl->progress) {
ret = btrfs_remove_free_space(block_group,
start, num_bytes);
- BUG_ON(ret); /* -ENOMEM */
} else {
num_bytes = caching_ctl->progress - start;
ret = btrfs_remove_free_space(block_group,
start, num_bytes);
- BUG_ON(ret); /* -ENOMEM */
+ if (ret)
+ goto out_lock;
start = caching_ctl->progress;
num_bytes = ins->objectid + ins->offset -
caching_ctl->progress;
ret = add_excluded_extent(root, start, num_bytes);
- BUG_ON(ret); /* -ENOMEM */
}
-
+out_lock:
mutex_unlock(&caching_ctl->mutex);
put_caching_control(caching_ctl);
+ if (ret)
+ goto out;
}
ret = btrfs_update_reserved_bytes(block_group, ins->offset,
RESERVE_ALLOC_NO_ACCOUNT);
BUG_ON(ret); /* logic error */
- btrfs_put_block_group(block_group);
ret = alloc_reserved_file_extent(trans, root, 0, root_objectid,
0, owner, offset, ins, 1);
+out:
+ btrfs_put_block_group(block_group);
return ret;
}
-struct extent_buffer *btrfs_init_new_buffer(struct btrfs_trans_handle *trans,
- struct btrfs_root *root,
- u64 bytenr, u32 blocksize,
- int level)
+static struct extent_buffer *
+btrfs_init_new_buffer(struct btrfs_trans_handle *trans, struct btrfs_root *root,
+ u64 bytenr, u32 blocksize, int level)
{
struct extent_buffer *buf;
struct btrfs_block_rsv *block_rsv;
struct btrfs_block_rsv *global_rsv = &root->fs_info->global_block_rsv;
int ret;
+ bool global_updated = false;
block_rsv = get_block_rsv(trans, root);
- if (block_rsv->size == 0) {
- ret = reserve_metadata_bytes(root, block_rsv, blocksize,
- BTRFS_RESERVE_NO_FLUSH);
- /*
- * If we couldn't reserve metadata bytes try and use some from
- * the global reserve.
- */
- if (ret && block_rsv != global_rsv) {
- ret = block_rsv_use_bytes(global_rsv, blocksize);
- if (!ret)
- return global_rsv;
- return ERR_PTR(ret);
- } else if (ret) {
- return ERR_PTR(ret);
- }
+ if (unlikely(block_rsv->size == 0))
+ goto try_reserve;
+again:
+ ret = block_rsv_use_bytes(block_rsv, blocksize);
+ if (!ret)
return block_rsv;
+
+ if (block_rsv->failfast)
+ return ERR_PTR(ret);
+
+ if (block_rsv->type == BTRFS_BLOCK_RSV_GLOBAL && !global_updated) {
+ global_updated = true;
+ update_global_block_rsv(root->fs_info);
+ goto again;
}
- ret = block_rsv_use_bytes(block_rsv, blocksize);
+ if (btrfs_test_opt(root, ENOSPC_DEBUG)) {
+ static DEFINE_RATELIMIT_STATE(_rs,
+ DEFAULT_RATELIMIT_INTERVAL * 10,
+ /*DEFAULT_RATELIMIT_BURST*/ 1);
+ if (__ratelimit(&_rs))
+ WARN(1, KERN_DEBUG
+ "btrfs: block rsv returned %d\n", ret);
+ }
+try_reserve:
+ ret = reserve_metadata_bytes(root, block_rsv, blocksize,
+ BTRFS_RESERVE_NO_FLUSH);
if (!ret)
return block_rsv;
- if (ret && !block_rsv->failfast) {
- if (btrfs_test_opt(root, ENOSPC_DEBUG)) {
- static DEFINE_RATELIMIT_STATE(_rs,
- DEFAULT_RATELIMIT_INTERVAL * 10,
- /*DEFAULT_RATELIMIT_BURST*/ 1);
- if (__ratelimit(&_rs))
- WARN(1, KERN_DEBUG
- "btrfs: block rsv returned %d\n", ret);
- }
- ret = reserve_metadata_bytes(root, block_rsv, blocksize,
- BTRFS_RESERVE_NO_FLUSH);
- if (!ret) {
- return block_rsv;
- } else if (ret && block_rsv != global_rsv) {
- ret = block_rsv_use_bytes(global_rsv, blocksize);
- if (!ret)
- return global_rsv;
- }
+ /*
+ * If we couldn't reserve metadata bytes try and use some from
+ * the global reserve if its space type is the same as the global
+ * reservation.
+ */
+ if (block_rsv->type != BTRFS_BLOCK_RSV_GLOBAL &&
+ block_rsv->space_info == global_rsv->space_info) {
+ ret = block_rsv_use_bytes(global_rsv, blocksize);
+ if (!ret)
+ return global_rsv;
}
-
- return ERR_PTR(-ENOSPC);
+ return ERR_PTR(ret);
}
static void unuse_block_rsv(struct btrfs_fs_info *fs_info,
extent_op->update_key = 1;
extent_op->update_flags = 1;
extent_op->is_data = 0;
+ extent_op->level = level;
ret = btrfs_add_delayed_tree_ref(root->fs_info, trans,
ins.objectid,
ret = btrfs_dec_ref(trans, root, eb, 0, wc->for_reloc);
BUG_ON(ret); /* -ENOMEM */
ret = btrfs_set_disk_extent_flags(trans, root, eb->start,
- eb->len, flag, 0);
+ eb->len, flag,
+ btrfs_header_level(eb), 0);
BUG_ON(ret); /* -ENOMEM */
wc->flags[level] |= flag;
}
if (reada && level == 1)
reada_walk_down(trans, root, wc, path);
next = read_tree_block(root, bytenr, blocksize, generation);
- if (!next)
+ if (!next || !extent_buffer_uptodate(next)) {
+ free_extent_buffer(next);
return -EIO;
+ }
btrfs_tree_lock(next);
btrfs_set_lock_blocking(next);
}