On 2019/12/18 下午1:18, damenly.su@gmail.com wrote:
> From: Su Yue <Damenly_Su@gmx.com>
> 
> The are different behavior of btrfs_lookup_first_block_group() and
> btrfs_lookup_first_block_group_kernel().
> There are many palaces calling the lookup function include extent
> allocation part. It's too complicated to check and change those.
> It will influence many functionalities in progs.
> 
> So here, just make kernel version lookup functions run likely in
> progs behavior.
> 
> Signed-off-by: Su Yue <Damenly_Su@gmx.com>

It should be folded into previous commit, or this will break bisect.

Thanks,
Qu

> ---
>  extent-tree.c | 7 ++++---
>  1 file changed, 4 insertions(+), 3 deletions(-)
> 
> diff --git a/extent-tree.c b/extent-tree.c
> index fdfa29a2409f..3f7b82dc88a2 100644
> --- a/extent-tree.c
> +++ b/extent-tree.c
> @@ -238,12 +238,13 @@ static struct btrfs_block_group_cache *block_group_cache_tree_search(
>  }
>  
>  /*
> - * Return the block group that starts at or after bytenr
> + * Return the block group that contains @bytenr, otherwise return the next one
> + * that starts after @bytenr
>   */
>  struct btrfs_block_group_cache *btrfs_lookup_first_block_group_kernel(
>  		struct btrfs_fs_info *info, u64 bytenr)
>  {
> -	return block_group_cache_tree_search(info, bytenr, 0);
> +	return block_group_cache_tree_search(info, bytenr, 1);
>  }
>  
>  /*
> @@ -252,7 +253,7 @@ struct btrfs_block_group_cache *btrfs_lookup_first_block_group_kernel(
>  struct btrfs_block_group_cache *btrfs_lookup_block_group_kernel(
>  		struct btrfs_fs_info *info, u64 bytenr)
>  {
> -	return block_group_cache_tree_search(info, bytenr, 1);
> +	return block_group_cache_tree_search(info, bytenr, 0);
>  }
>  
>  /*
>