Re: [Qemu-devel] [PATCH] block: document semanatics of bdrv_co_preadv|pwritev

[Kevin Wolf <kwolf@redhat.com>]

Am 04.08.2017 um 12:50 hat Daniel P. Berrange geschrieben:
> Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
> ---
>  include/block/block_int.h | 29 +++++++++++++++++++++++++++++
>  1 file changed, 29 insertions(+)
> 
> diff --git a/include/block/block_int.h b/include/block/block_int.h
> index d4f4ea7584..deb81a58bd 100644
> --- a/include/block/block_int.h
> +++ b/include/block/block_int.h
> @@ -147,12 +147,41 @@ struct BlockDriver {
>  
>      int coroutine_fn (*bdrv_co_readv)(BlockDriverState *bs,
>          int64_t sector_num, int nb_sectors, QEMUIOVector *qiov);
> +
> +    /**
> +     * @offset: position in bytes to read at
> +     * @bytes: number of bytes to read
> +     * @qiov: the buffers to fill with read data
> +     *
> +     * @offset and @bytes will be a multiple of 'request_alignment',
> +     * but the length of individual @qiov elements does not have to
> +     * be a multiple.
> +     *
> +     * @bytes may be less than the total sizeof @iov, and will be
> +     * no larger than 'max_transfer'.

Really? We are asserting that they match in bdrv_aligned_preadv():

    assert(!qiov || bytes == qiov->size);

Also, s/sizeof @iov/size of @qiov/

> +     *
> +     * The buffer in @qiov may point directly to guest memory.
> +     */
>      int coroutine_fn (*bdrv_co_preadv)(BlockDriverState *bs,
>          uint64_t offset, uint64_t bytes, QEMUIOVector *qiov, int flags);
>      int coroutine_fn (*bdrv_co_writev)(BlockDriverState *bs,
>          int64_t sector_num, int nb_sectors, QEMUIOVector *qiov);
>      int coroutine_fn (*bdrv_co_writev_flags)(BlockDriverState *bs,
>          int64_t sector_num, int nb_sectors, QEMUIOVector *qiov, int flags);
> +    /**
> +     * @offset: position in bytes to write at
> +     * @bytes: number of bytes to write
> +     * @qiov: the buffers containing data to write
> +     *
> +     * @offset and @bytes will be a multiple of 'request_alignment',
> +     * but the length of individual @qiov elements does not have to
> +     * be a multiple.
> +     *
> +     * @bytes may be less than the total sizeof @iov, and will be
> +     * no larger than 'max_transfer'.

The same assertion exists in bdrv_aligned_pwritev() (and the same typo
in your comment).

> +     * The buffer in @qiov may point directly to guest memory.
> +     */
>      int coroutine_fn (*bdrv_co_pwritev)(BlockDriverState *bs,
>          uint64_t offset, uint64_t bytes, QEMUIOVector *qiov, int flags);

Kevin