struct btrfs_path *path;
path = btrfs_alloc_path();
- BUG_ON(!path);
+ if (!path)
+ return -ENOMEM;
+
key.objectid = start;
key.offset = len;
btrfs_set_key_type(&key, BTRFS_EXTENT_ITEM_KEY);
for (i = 0; i < multi->num_stripes; i++, stripe++) {
+ if (!stripe->dev->can_discard)
+ continue;
+
ret = btrfs_issue_discard(stripe->dev->bdev,
stripe->physical,
stripe->length);
discarded_bytes += stripe->length;
else if (ret != -EOPNOTSUPP)
break;
+
+ /*
+ * Just in case we get back EOPNOTSUPP for some reason,
+ * just ignore the return value so we don't screw up
+ * people calling discard_extent.
+ */
+ ret = 0;
}
kfree(multi);
}
- if (discarded_bytes && ret == -EOPNOTSUPP)
- ret = 0;
if (actual_bytes)
*actual_bytes = discarded_bytes;
}
ret = btrfs_alloc_chunk(trans, extent_root, flags);
+ if (ret < 0 && ret != -ENOSPC)
+ goto out;
+
spin_lock(&space_info->lock);
if (ret)
space_info->full = 1;
space_info->force_alloc = CHUNK_ALLOC_NO_FORCE;
space_info->chunk_alloc = 0;
spin_unlock(&space_info->lock);
+out:
mutex_unlock(&extent_root->fs_info->chunk_mutex);
return ret;
}
if (ret < 0)
goto out;
+ ret = 0;
+
/*
* So if we were overcommitted it's possible that somebody else flushed
* out enough space and we simply didn't have enough space to reclaim,
goto out;
ret = -EAGAIN;
- if (trans || committed)
+ if (trans)
goto out;
ret = -ENOSPC;
+ if (committed)
+ goto out;
+
trans = btrfs_join_transaction(root);
if (IS_ERR(trans))
goto out;
if (commit_trans) {
if (trans)
return -EAGAIN;
-
trans = btrfs_join_transaction(root);
BUG_ON(IS_ERR(trans));
ret = btrfs_commit_transaction(trans, root);
return block_rsv_migrate_bytes(src_rsv, dst_rsv, num_bytes);
}
+static unsigned drop_outstanding_extent(struct inode *inode)
+{
+ unsigned dropped_extents = 0;
+
+ spin_lock(&BTRFS_I(inode)->lock);
+ BUG_ON(!BTRFS_I(inode)->outstanding_extents);
+ BTRFS_I(inode)->outstanding_extents--;
+
+ /*
+ * If we have more or the same amount of outsanding extents than we have
+ * reserved then we need to leave the reserved extents count alone.
+ */
+ if (BTRFS_I(inode)->outstanding_extents >=
+ BTRFS_I(inode)->reserved_extents)
+ goto out;
+
+ dropped_extents = BTRFS_I(inode)->reserved_extents -
+ BTRFS_I(inode)->outstanding_extents;
+ BTRFS_I(inode)->reserved_extents -= dropped_extents;
+out:
+ spin_unlock(&BTRFS_I(inode)->lock);
+ return dropped_extents;
+}
+
static u64 calc_csum_metadata_size(struct inode *inode, u64 num_bytes)
{
return num_bytes >>= 3;
{
struct btrfs_root *root = BTRFS_I(inode)->root;
struct btrfs_block_rsv *block_rsv = &root->fs_info->delalloc_block_rsv;
- u64 to_reserve;
- int nr_extents;
- int reserved_extents;
+ u64 to_reserve = 0;
+ unsigned nr_extents = 0;
int ret;
if (btrfs_transaction_in_commit(root->fs_info))
num_bytes = ALIGN(num_bytes, root->sectorsize);
- nr_extents = atomic_read(&BTRFS_I(inode)->outstanding_extents) + 1;
- reserved_extents = atomic_read(&BTRFS_I(inode)->reserved_extents);
+ spin_lock(&BTRFS_I(inode)->lock);
+ BTRFS_I(inode)->outstanding_extents++;
+
+ if (BTRFS_I(inode)->outstanding_extents >
+ BTRFS_I(inode)->reserved_extents) {
+ nr_extents = BTRFS_I(inode)->outstanding_extents -
+ BTRFS_I(inode)->reserved_extents;
+ BTRFS_I(inode)->reserved_extents += nr_extents;
- if (nr_extents > reserved_extents) {
- nr_extents -= reserved_extents;
to_reserve = btrfs_calc_trans_metadata_size(root, nr_extents);
- } else {
- nr_extents = 0;
- to_reserve = 0;
}
+ spin_unlock(&BTRFS_I(inode)->lock);
to_reserve += calc_csum_metadata_size(inode, num_bytes);
ret = reserve_metadata_bytes(NULL, root, block_rsv, to_reserve, 1);
- if (ret)
+ if (ret) {
+ unsigned dropped;
+ /*
+ * We don't need the return value since our reservation failed,
+ * we just need to clean up our counter.
+ */
+ dropped = drop_outstanding_extent(inode);
+ WARN_ON(dropped > 1);
return ret;
-
- atomic_add(nr_extents, &BTRFS_I(inode)->reserved_extents);
- atomic_inc(&BTRFS_I(inode)->outstanding_extents);
+ }
block_rsv_add_bytes(block_rsv, to_reserve, 1);
- if (block_rsv->size > 512 * 1024 * 1024)
- shrink_delalloc(NULL, root, to_reserve, 0);
-
return 0;
}
void btrfs_delalloc_release_metadata(struct inode *inode, u64 num_bytes)
{
struct btrfs_root *root = BTRFS_I(inode)->root;
- u64 to_free;
- int nr_extents;
- int reserved_extents;
+ u64 to_free = 0;
+ unsigned dropped;
num_bytes = ALIGN(num_bytes, root->sectorsize);
- atomic_dec(&BTRFS_I(inode)->outstanding_extents);
- WARN_ON(atomic_read(&BTRFS_I(inode)->outstanding_extents) < 0);
-
- reserved_extents = atomic_read(&BTRFS_I(inode)->reserved_extents);
- do {
- int old, new;
-
- nr_extents = atomic_read(&BTRFS_I(inode)->outstanding_extents);
- if (nr_extents >= reserved_extents) {
- nr_extents = 0;
- break;
- }
- old = reserved_extents;
- nr_extents = reserved_extents - nr_extents;
- new = reserved_extents - nr_extents;
- old = atomic_cmpxchg(&BTRFS_I(inode)->reserved_extents,
- reserved_extents, new);
- if (likely(old == reserved_extents))
- break;
- reserved_extents = old;
- } while (1);
+ dropped = drop_outstanding_extent(inode);
to_free = calc_csum_metadata_size(inode, num_bytes);
- if (nr_extents > 0)
- to_free += btrfs_calc_trans_metadata_size(root, nr_extents);
+ if (dropped > 0)
+ to_free += btrfs_calc_trans_metadata_size(root, dropped);
btrfs_block_rsv_release(root, &root->fs_info->delalloc_block_rsv,
to_free);
printk(KERN_ERR "umm, got %d back from search"
", was looking for %llu\n", ret,
(unsigned long long)bytenr);
- btrfs_print_leaf(extent_root, path->nodes[0]);
+ if (ret > 0)
+ btrfs_print_leaf(extent_root,
+ path->nodes[0]);
}
BUG_ON(ret);
extent_slot = path->slots[0];
* group is does point to and try again
*/
if (!last_ptr_loop && last_ptr->block_group &&
- last_ptr->block_group != block_group) {
+ last_ptr->block_group != block_group &&
+ index <=
+ get_block_group_index(last_ptr->block_group)) {
btrfs_put_block_group(block_group);
block_group = last_ptr->block_group;
u32 size = sizeof(*extent_item) + sizeof(*block_info) + sizeof(*iref);
path = btrfs_alloc_path();
- BUG_ON(!path);
+ if (!path)
+ return -ENOMEM;
path->leave_spinning = 1;
ret = btrfs_insert_empty_item(trans, fs_info->extent_root, path,
if (!buf)
return ERR_PTR(-ENOMEM);
btrfs_set_header_generation(buf, trans->transid);
- btrfs_set_buffer_lockdep_class(buf, level);
+ btrfs_set_buffer_lockdep_class(root->root_key.objectid, buf, level);
btrfs_tree_lock(buf);
clean_tree_block(trans, root, buf);
return 1;
if (path->locks[level] && !wc->keep_locks) {
- btrfs_tree_unlock(eb);
+ btrfs_tree_unlock_rw(eb, path->locks[level]);
path->locks[level] = 0;
}
return 0;
* keep the tree lock
*/
if (path->locks[level] && level > 0) {
- btrfs_tree_unlock(eb);
+ btrfs_tree_unlock_rw(eb, path->locks[level]);
path->locks[level] = 0;
}
return 0;
BUG_ON(level != btrfs_header_level(next));
path->nodes[level] = next;
path->slots[level] = 0;
- path->locks[level] = 1;
+ path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
wc->level = level;
if (wc->level == 1)
wc->reada_slot = 0;
BUG_ON(level == 0);
btrfs_tree_lock(eb);
btrfs_set_lock_blocking(eb);
- path->locks[level] = 1;
+ path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
ret = btrfs_lookup_extent_info(trans, root,
eb->start, eb->len,
BUG_ON(ret);
BUG_ON(wc->refs[level] == 0);
if (wc->refs[level] == 1) {
- btrfs_tree_unlock(eb);
- path->locks[level] = 0;
+ btrfs_tree_unlock_rw(eb, path->locks[level]);
return 1;
}
}
btrfs_header_generation(eb) == trans->transid) {
btrfs_tree_lock(eb);
btrfs_set_lock_blocking(eb);
- path->locks[level] = 1;
+ path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
}
clean_tree_block(trans, root, eb);
}
return 0;
if (path->locks[level]) {
- btrfs_tree_unlock(path->nodes[level]);
+ btrfs_tree_unlock_rw(path->nodes[level],
+ path->locks[level]);
path->locks[level] = 0;
}
free_extent_buffer(path->nodes[level]);
* also make sure backrefs for the shared block and all lower level
* blocks are properly updated.
*/
-int btrfs_drop_snapshot(struct btrfs_root *root,
- struct btrfs_block_rsv *block_rsv, int update_ref)
+void btrfs_drop_snapshot(struct btrfs_root *root,
+ struct btrfs_block_rsv *block_rsv, int update_ref)
{
struct btrfs_path *path;
struct btrfs_trans_handle *trans;
int level;
path = btrfs_alloc_path();
- BUG_ON(!path);
+ if (!path) {
+ err = -ENOMEM;
+ goto out;
+ }
wc = kzalloc(sizeof(*wc), GFP_NOFS);
- BUG_ON(!wc);
+ if (!wc) {
+ btrfs_free_path(path);
+ err = -ENOMEM;
+ goto out;
+ }
trans = btrfs_start_transaction(tree_root, 0);
BUG_ON(IS_ERR(trans));
path->nodes[level] = btrfs_lock_root_node(root);
btrfs_set_lock_blocking(path->nodes[level]);
path->slots[level] = 0;
- path->locks[level] = 1;
+ path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
memset(&wc->update_progress, 0,
sizeof(wc->update_progress));
} else {
path->lowest_level = 0;
if (ret < 0) {
err = ret;
- goto out;
+ goto out_free;
}
WARN_ON(ret > 0);
free_extent_buffer(root->commit_root);
kfree(root);
}
-out:
+out_free:
btrfs_end_transaction_throttle(trans, tree_root);
kfree(wc);
btrfs_free_path(path);
- return err;
+out:
+ if (err)
+ btrfs_std_error(root->fs_info, err);
+ return;
}
/*
level = btrfs_header_level(node);
path->nodes[level] = node;
path->slots[level] = 0;
- path->locks[level] = 1;
+ path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING;
wc->refs[parent_level] = 1;
wc->flags[parent_level] = BTRFS_BLOCK_FLAG_FULL_BACKREF;
return flags;
}
-static int set_block_group_ro(struct btrfs_block_group_cache *cache)
+static int set_block_group_ro(struct btrfs_block_group_cache *cache, int force)
{
struct btrfs_space_info *sinfo = cache->space_info;
u64 num_bytes;
+ u64 min_allocable_bytes;
int ret = -ENOSPC;
- if (cache->ro)
- return 0;
+
+ /*
+ * We need some metadata space and system metadata space for
+ * allocating chunks in some corner cases until we force to set
+ * it to be readonly.
+ */
+ if ((sinfo->flags &
+ (BTRFS_BLOCK_GROUP_SYSTEM | BTRFS_BLOCK_GROUP_METADATA)) &&
+ !force)
+ min_allocable_bytes = 1 * 1024 * 1024;
+ else
+ min_allocable_bytes = 0;
spin_lock(&sinfo->lock);
spin_lock(&cache->lock);
+
+ if (cache->ro) {
+ ret = 0;
+ goto out;
+ }
+
num_bytes = cache->key.offset - cache->reserved - cache->pinned -
cache->bytes_super - btrfs_block_group_used(&cache->item);
if (sinfo->bytes_used + sinfo->bytes_reserved + sinfo->bytes_pinned +
sinfo->bytes_may_use + sinfo->bytes_readonly +
- cache->reserved_pinned + num_bytes <= sinfo->total_bytes) {
+ cache->reserved_pinned + num_bytes + min_allocable_bytes <=
+ sinfo->total_bytes) {
sinfo->bytes_readonly += num_bytes;
sinfo->bytes_reserved += cache->reserved_pinned;
cache->reserved_pinned = 0;
cache->ro = 1;
ret = 0;
}
-
+out:
spin_unlock(&cache->lock);
spin_unlock(&sinfo->lock);
return ret;
do_chunk_alloc(trans, root, 2 * 1024 * 1024, alloc_flags,
CHUNK_ALLOC_FORCE);
- ret = set_block_group_ro(cache);
+ ret = set_block_group_ro(cache, 0);
if (!ret)
goto out;
alloc_flags = get_alloc_profile(root, cache->space_info->flags);
CHUNK_ALLOC_FORCE);
if (ret < 0)
goto out;
- ret = set_block_group_ro(cache);
+ ret = set_block_group_ro(cache, 0);
out:
btrfs_end_transaction(trans, root);
return ret;
struct btrfs_space_info *space_info;
struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
struct btrfs_device *device;
+ u64 min_free;
+ int index;
+ int dev_nr = 0;
+ int dev_min = 1;
int full = 0;
int ret = 0;
if (!block_group)
return -1;
+ min_free = btrfs_block_group_used(&block_group->item);
+
/* no bytes used, we're good */
- if (!btrfs_block_group_used(&block_group->item))
+ if (!min_free)
goto out;
space_info = block_group->space_info;
* all of the extents from this block group. If we can, we're good
*/
if ((space_info->total_bytes != block_group->key.offset) &&
- (space_info->bytes_used + space_info->bytes_reserved +
- space_info->bytes_pinned + space_info->bytes_readonly +
- btrfs_block_group_used(&block_group->item) <
- space_info->total_bytes)) {
+ (space_info->bytes_used + space_info->bytes_reserved +
+ space_info->bytes_pinned + space_info->bytes_readonly +
+ min_free < space_info->total_bytes)) {
spin_unlock(&space_info->lock);
goto out;
}
if (full)
goto out;
+ /*
+ * index:
+ * 0: raid10
+ * 1: raid1
+ * 2: dup
+ * 3: raid0
+ * 4: single
+ */
+ index = get_block_group_index(block_group);
+ if (index == 0) {
+ dev_min = 4;
+ min_free /= 2;
+ } else if (index == 1) {
+ dev_min = 2;
+ } else if (index == 2) {
+ min_free *= 2;
+ } else if (index == 3) {
+ dev_min = fs_devices->rw_devices;
+ min_free /= dev_min;
+ }
+
mutex_lock(&root->fs_info->chunk_mutex);
list_for_each_entry(device, &fs_devices->alloc_list, dev_alloc_list) {
- u64 min_free = btrfs_block_group_used(&block_group->item);
u64 dev_offset;
/*
ret = find_free_dev_extent(NULL, device, min_free,
&dev_offset, NULL);
if (!ret)
+ dev_nr++;
+
+ if (dev_nr >= dev_min)
break;
+
ret = -1;
}
}
set_avail_alloc_bits(root->fs_info, cache->flags);
if (btrfs_chunk_readonly(root, cache->key.objectid))
- set_block_group_ro(cache);
+ set_block_group_ro(cache, 1);
}
list_for_each_entry_rcu(space_info, &root->fs_info->space_info, list) {
* mirrored block groups.
*/
list_for_each_entry(cache, &space_info->block_groups[3], list)
- set_block_group_ro(cache);
+ set_block_group_ro(cache, 1);
list_for_each_entry(cache, &space_info->block_groups[4], list)
- set_block_group_ro(cache);
+ set_block_group_ro(cache, 1);
}
init_global_block_rsv(info);
spin_unlock(&cluster->refill_lock);
path = btrfs_alloc_path();
- BUG_ON(!path);
+ if (!path) {
+ ret = -ENOMEM;
+ goto out;
+ }
inode = lookup_free_space_inode(root, block_group, path);
if (!IS_ERR(inode)) {
- btrfs_orphan_add(trans, inode);
+ ret = btrfs_orphan_add(trans, inode);
+ BUG_ON(ret);
clear_nlink(inode);
/* One for the block groups ref */
spin_lock(&block_group->lock);
git.openpandora.org / pandora-kernel.git / blobdiff