The Friday 07 Mar 2014 à 23:55:51 (+0100), Max Reitz wrote :
> Implement this function in the same way as raw_bsd does: Acknowledge
> that this is a passthrough driver (always return BDRV_BLOCK_OFFSET_VALID
> and BDRV_BLOCK_DATA and derive the offset directly from the sector
> index) and add BDRV_BLOCK_RAW to the returned value.
>
> Signed-off-by: Max Reitz <mre...@redhat.com>
> ---
> block/json.c | 10 ++++++++++
> 1 file changed, 10 insertions(+)
>
> diff --git a/block/json.c b/block/json.c
> index e4cdb68..966a5f5 100644
> --- a/block/json.c
> +++ b/block/json.c
> @@ -110,6 +110,15 @@ static coroutine_fn int
> json_co_write_zeroes(BlockDriverState *bs,
> return bdrv_co_write_zeroes(bs->file, sector_num, nb_sectors, flags);
> }
>
> +static coroutine_fn int64_t json_co_get_block_status(BlockDriverState *bs,
> + int64_t sector_num,
> + int nb_sectors, int
> *pnum)
> +{
> + *pnum = nb_sectors;
> + return BDRV_BLOCK_RAW | BDRV_BLOCK_OFFSET_VALID | BDRV_BLOCK_DATA |
> + (sector_num << BDRV_SECTOR_BITS);
> +}
> +
> static void json_invalidate_cache(BlockDriverState *bs)
> {
> return bdrv_invalidate_cache(bs->file);
> @@ -156,6 +165,7 @@ static BlockDriver bdrv_json = {
> .bdrv_aio_discard = json_aio_discard,
>
> .bdrv_co_write_zeroes = json_co_write_zeroes,
> + .bdrv_co_get_block_status = json_co_get_block_status,
>
> .bdrv_invalidate_cache = json_invalidate_cache,
>
> --
> 1.9.0
>
If this filter is stacked on top of qcow2 even the simple BDRV_BLOCK_RAW feel
weird.
I think the best thing we can do is to ask to Peter Lieven if this code has the
intended meaning in a block filter.
Peter:
You wrote the inspiration for this code in raw-posix.c.
What do you think of this piece of code designed to be stacked as a block filter
on top of regular block driver (qcow2 etc) ?
Does it makes sense ? Or would it be better to forward the call to the lower
level ?
Best regards
Benoît
