lk/linux-kernel-test
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Btrfs: use bytes_may_use for all ENOSPC reservations

Browse Source 
We have been using bytes_reserved for metadata reservations, which is wrong
since we use that to keep track of outstanding reservations from the allocator.
This resulted in us doing a lot of silly things to make sure we don't allocate a
bunch of metadata chunks since we never had a real view of how much space was
actually in use by metadata.

This passes Arne's enospc test and xfstests as well as my own enospc tests.
Hopefully this will get us moving in the right direction.  Thanks,

Signed-off-by: Josef Bacik <josef@redhat.com>
This commit is contained in:
Josef Bacik
2011-07-26 17:00:46 -04:00
parent 830c4adbd0
commit fb25e9141a
3 changed files with 112 additions and 82 deletions
Download Patch File Download Diff File Expand all files Collapse all files

29
fs/btrfs/free-space-cache.c
View File

@@ -2472,9 +2472,19 @@ int btrfs_trim_block_group(struct btrfs_block_group_cache *block_group,
spin_unlock(&ctl->tree_lock);
if (bytes >= minlen) {
int update_ret;
update_ret = btrfs_update_reserved_bytes(block_group,
bytes, 1, 1);
struct btrfs_space_info *space_info;
int update = 0;
space_info = block_group->space_info;
spin_lock(&space_info->lock);
spin_lock(&block_group->lock);
if (!block_group->ro) {
block_group->reserved += bytes;
space_info->bytes_reserved += bytes;
update = 1;
}
spin_unlock(&block_group->lock);
spin_unlock(&space_info->lock);
ret = btrfs_error_discard_extent(fs_info->extent_root,
start,
@@ -2482,9 +2492,16 @@ int btrfs_trim_block_group(struct btrfs_block_group_cache *block_group,
&actually_trimmed);
btrfs_add_free_space(block_group, start, bytes);
if (!update_ret)
btrfs_update_reserved_bytes(block_group,
bytes, 0, 1);
if (update) {
spin_lock(&space_info->lock);
spin_lock(&block_group->lock);
if (block_group->ro)
space_info->bytes_readonly += bytes;
block_group->reserved -= bytes;
space_info->bytes_reserved -= bytes;
spin_unlock(&space_info->lock);
spin_unlock(&block_group->lock);
}
if (ret)
break;