On 22.05.2018 06:51, Gu Jinxiang wrote:
> set_extent_bits may return 0/-EEXIST, so return the result in
> add_excluded_extent. And handle the failures in upper callers.
>
> Caller of add_excluded_extent and failure process currently:
> exclude_super_stripes
> <- btrfs_make_block_group //handles the failure
> <- btrfs_read_block_groups //handles the failure
> __exclude_logged_extent
> <- btrfs_exclude_logged_extents
> <- btrfs_alloc_logged_file_extent //propagate failure to upper
> caller
> Add logic of propagate return value of __exclude_logged_extent to
> btrfs_exclude_logged_extents.
>
> Signed-off-by: Gu Jinxiang <gujx@xxxxxxxxxxxxxx>
> ---
> fs/btrfs/extent-tree.c | 20 +++++++++++++++-----
> 1 file changed, 15 insertions(+), 5 deletions(-)
>
> diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
> index 75cfb80d2551..28fd71579141 100644
> --- a/fs/btrfs/extent-tree.c
> +++ b/fs/btrfs/extent-tree.c
> @@ -215,11 +215,16 @@ static int add_excluded_extent(struct btrfs_fs_info *fs_info,
> u64 start, u64 num_bytes)
> {
> u64 end = start + num_bytes - 1;
> - set_extent_bits(&fs_info->freed_extents[0],
> + int ret = 0;
> +
> + ret = set_extent_bits(&fs_info->freed_extents[0],
> start, end, EXTENT_UPTODATE);
> - set_extent_bits(&fs_info->freed_extents[1],
> + if (ret)
> + goto out;
> + ret = set_extent_bits(&fs_info->freed_extents[1],
> start, end, EXTENT_UPTODATE);
> - return 0;
> +out:
> + return ret;
> }
>
> static void free_excluded_extents(struct btrfs_fs_info *fs_info,
> @@ -6463,6 +6468,7 @@ int btrfs_exclude_logged_extents(struct btrfs_fs_info *fs_info,
> struct btrfs_key key;
> int found_type;
> int i;
> + int ret = 0;
>
> if (!btrfs_fs_incompat(fs_info, MIXED_GROUPS))
> return 0;
> @@ -6479,10 +6485,14 @@ int btrfs_exclude_logged_extents(struct btrfs_fs_info *fs_info,
> continue;
> key.objectid = btrfs_file_extent_disk_bytenr(eb, item);
> key.offset = btrfs_file_extent_disk_num_bytes(eb, item);
> - __exclude_logged_extent(fs_info, key.objectid, key.offset);
> + ret = __exclude_logged_extent(fs_info, key.objectid,
> + key.offset);
> + if (ret)
> + goto out;
> }
>
> - return 0;
> +out:
> + return ret;
> }
The 2nd and 3rd hunks need to go in a separate patch.
__exclude_logged_extent can return an error value for any number of
reasons (btrfs_lookup_block_group failing, btrfs_remove_free_space
failing or add_excluded_extents also failing). SO what you are actually
fixing here is __exclude_logged_extent return value not being handled in
btrfs_exclude_logged_extent which is a separate issue from
add_excluded_extent always returning 0. So put this in a separate patch.
>
> static void
>
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
