Re: [PATCH v2 07/12] btrfs-progs: add the ability to corrupt fields of the super block

[David Sterba <dsterba@suse.cz>]

On Wed, Aug 18, 2021 at 05:33:19PM -0400, Josef Bacik wrote:
> Doing the extent tree v2 work I generated an invalid super block with
> the wrong bytes_used set, and only noticed because it affected the block
> groups as well.  Neither modes of fsck check for a valid bytes_used, so
> add the ability to corrupt this field so I can generate a testcase for
> fsck.
> 
> Signed-off-by: Josef Bacik <josef@toxicpanda.com>
> ---
>  btrfs-corrupt-block.c | 66 ++++++++++++++++++++++++++++++++++++++++++-
>  1 file changed, 65 insertions(+), 1 deletion(-)
> 
> diff --git a/btrfs-corrupt-block.c b/btrfs-corrupt-block.c
> index 80622f29..7e576897 100644
> --- a/btrfs-corrupt-block.c
> +++ b/btrfs-corrupt-block.c
> @@ -355,6 +355,24 @@ enum btrfs_block_group_field {
>  	BTRFS_BLOCK_GROUP_ITEM_BAD,
>  };
>  
> +enum btrfs_super_field {
> +	BTRFS_SUPER_FIELD_FLAGS,
> +	BTRFS_SUPER_FIELD_TOTAL_BYTES,
> +	BTRFS_SUPER_FIELD_BYTES_USED,
> +	BTRFS_SUPER_FIELD_BAD,
> +};
> +
> +static enum btrfs_super_field convert_super_field(char *field)
> +{
> +	if (!strncmp(field, "flags", FIELD_BUF_LEN))
> +		return BTRFS_SUPER_FIELD_FLAGS;
> +	if (!strncmp(field, "total_bytes", FIELD_BUF_LEN))
> +		return BTRFS_SUPER_FIELD_TOTAL_BYTES;
> +	if (!strncmp(field, "bytes_used", FIELD_BUF_LEN))
> +		return BTRFS_SUPER_FIELD_BYTES_USED;
> +	return BTRFS_SUPER_FIELD_BAD;

There's a more feature-complete utility to modify super block fields
btrfs-sb-mod, this would be duplicated in corrupt-block.