Am 12.09.2011 16:59, schrieb Devin Nakamura:
> Signed-off-by: Devin Nakamura <devin...@gmail.com>
> ---
> block/qed.c | 15 +++++++++++++++
> 1 files changed, 15 insertions(+), 0 deletions(-)
>
> diff --git a/block/qed.c b/block/qed.c
> index 341cf9d..caecdff 100644
> --- a/block/qed.c
> +++ b/block/qed.c
> @@ -1586,6 +1586,20 @@ static int bdrv_qed_map(BlockDriverState *bs, uint64_t
> guest_offset,
> return 0;
> }
>
> +static int bdrv_qed_copy_header(BlockDriverState *bs, char* filename)
> +{
> + BlockDriverState *backup;
> + uint8_t buffer[512];
> + bdrv_create_file(filename, NULL);
> + bdrv_file_open(&backup, filename, BDRV_O_RDWR);
> + bdrv_read(bs->file, 0, buffer, 1); /*TODO: check return code*/
> + bdrv_write(backup, 0, buffer, 1); /*TODO: check return code*/
Come on, checking return values isn't much more work than adding a TODO
comment.
> + bdrv_close(backup);
> +
> + qed_write_header_sync(bs->opaque);
> + return 0;
> +}
This is probably right, but it makes me wonder if we are sure that QED
never tries to write out the header before bdrv_qed_copy_header() is
called? In that case we would have destroyed the original format before
completing the conversion.
I would feel more comfortable if BDRVQEDState had a field header_offset
that would be set to a temporary location by
bdrv_qed_open_conversion_target and reset to 0 by bdrv_qed_copy_header.
Kevin
