Re: [Qemu-devel] [PATCH v2 4/7] block/nbd-client: drop reply field from NBDClientSession

[Eric Blake <eblake@redhat.com>]

[-- Attachment #1: Type: text/plain, Size: 1656 bytes --]

On 09/18/2017 08:59 AM, Vladimir Sementsov-Ogievskiy wrote:
> Drop 'reply' from NBDClientSession. It's used to only deliver request
> return code to receiving coroutine. Instead introduce per-request ret
> variable to simplify the scheme and make further refactoring possible.
> 
> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
> ---
>  block/nbd-client.h |  2 +-
>  block/nbd-client.c | 22 +++++++++-------------
>  2 files changed, 10 insertions(+), 14 deletions(-)
> 
> diff --git a/block/nbd-client.h b/block/nbd-client.h
> index 3f97d31013..4bc8fe3993 100644
> --- a/block/nbd-client.h
> +++ b/block/nbd-client.h
> @@ -22,6 +22,7 @@ typedef struct {
>      bool receiving;         /* waiting for read_reply_co? */
>      NBDRequest *request;
>      QEMUIOVector *qiov;
> +    int ret;
>  } NBDClientRequest;

I like this idea.  However, I note that:

> @@ -211,11 +211,7 @@ static int nbd_co_receive_reply(NBDClientSession *s, NBDRequest *request)
>      if (!s->ioc || s->quit) {
>          ret = -EIO;
>      } else {
> -        assert(s->reply.handle == request->handle);
> -        ret = -s->reply.error;
> -
> -        /* Tell the read handler to read another header.  */
> -        s->reply.handle = 0;
> +        ret = s->requests[i].ret;

you are removing the assert you added in 2/7, where I questioned whether
we needed NBDClientRequest.request in the first place.  So there may be
some rebase churn, depending on how that conversation pans out.

-- 
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3266
Virtualization:  qemu.org | libvirt.org


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 619 bytes --]