[sheepdog] [PATCH v2] sheepdog: do not blindly memset all read buffers

MORITA Kazutaka morita.kazutaka at lab.ntt.co.jp
Mon Jul 9 10:27:30 CEST 2012


At Sun, 8 Jul 2012 21:33:16 +0200,
Christoph Hellwig wrote:
> 
> Only buffers that map to unallocated blocks need to be zeroed.
> 
> Signed-off-by: Christoph Hellwig <hch at lst.de>
> 
> ---
>  block/sheepdog.c |   28 ++++++++++++++++++----------
>  1 file changed, 18 insertions(+), 10 deletions(-)
> 
> Index: qemu/block/sheepdog.c
> ===================================================================
> --- qemu.orig/block/sheepdog.c	2012-07-08 21:20:12.896597656 +0200
> +++ qemu/block/sheepdog.c	2012-07-08 21:20:56.563264061 +0200
> @@ -1556,18 +1556,26 @@ static int coroutine_fn sd_co_rw_vector(
>  
>          len = MIN(total - done, SD_DATA_OBJ_SIZE - offset);
>  
> -        if (!inode->data_vdi_id[idx]) {
> -            if (acb->aiocb_type == AIOCB_READ_UDATA) {
> +        switch (acb->aiocb_type) {
> +        case AIOCB_READ_UDATA:
> +            if (!inode->data_vdi_id[idx]) {
> +                qemu_iovec_memset_skip(acb->qiov, 0, len, done);
>                  goto done;
>              }
> -
> -            create = 1;
> -        } else if (acb->aiocb_type == AIOCB_WRITE_UDATA
> -                   && !is_data_obj_writable(inode, idx)) {
> -            /* Copy-On-Write */
> -            create = 1;
> -            old_oid = oid;
> -            flags = SD_FLAG_CMD_COW;
> +            break;
> +        case AIOCB_WRITE_UDATA:
> +            if (!inode->data_vdi_id[idx]) {
> +                create = 1;
> +            }
> +            if (!is_data_obj_writable(inode, idx)) {
> +                /* Copy-On-Write */
> +                create = 1;
> +                old_oid = oid;
> +                flags = SD_FLAG_CMD_COW;
> +            }
> +            break;
> +        default:
> +            break;
>          }
>  
>          if (create) {
> -- 

There is no code to remove a blind memset from sd_co_readv, which
existed in the previous patch.

The others look good to me.

Thanks,

Kazutaka



More information about the sheepdog mailing list