[PATCH v6 1/5] server: Allow volume information queries to be asynchronous

[Erich E. Hoover]

On Wed, Feb 10, 2021 at 9:15 PM Zebediah Figura (she/her)
<zfigura at codeweavers.com> wrote:
>
> Sorry I didn't notice these before, but...

That's alright, I always appreciate your feedback - it increases the
chances that AJ will be happy ;)

Best,
Erich

> On 2/6/21 12:24 PM, Erich E. Hoover wrote:
> > From: "Erich E. Hoover" <erich.e.hoover at gmail.com>
> > Subject: [PATCH v6 1/5] server: Allow volume information queries to be asynchronous
> > Message-Id: <CAEU2+vqETWab9BeFry-kO9+9=+6Xvm47Azhi7jumnf3D8ej0Jg at mail.gmail.com>
> > Date: Sat, 6 Feb 2021 11:24:44 -0700
> >...
> > +    if ((async = create_request_async( fd, fd->comp_flags, &req->async )))
> >      {
> > -        fd->fd_ops->get_volume_info( fd, req->info_class );
> > -        release_object( fd );
> > +        reply->wait    = async_handoff( async, fd->fd_ops->get_volume_info( fd, async, req->info_class ), NULL, 0 );
>
> Shouldn't we pass 1 to force_blocking here? From the client's
> perspective I believe this call is always synchronous.

That makes sense to me, I'll go ahead and change it.

> (Also, weird spacing before = in that line...)

Sorry about that, that's from copy-pasting from another instance of
async_handoff and forgetting to delete the alignment spaces.

> > ...
> >  static void message_queue_read( struct pipe_end *pipe_end, struct iosb *iosb )
> > diff --git a/server/protocol.def b/server/protocol.def
> > index 3747a3b63ec..cd7ed35d9a3 100644
> > --- a/server/protocol.def
> > +++ b/server/protocol.def
> > @@ -1400,9 +1400,11 @@ enum server_fd_type
> >
> >  /* Query volume information */
> >  @REQ(get_volume_info)
> > +    async_data_t  async;        /* async I/O parameters */
>
> Another inconsistent spacing nitpick here.

Consider it taken care of.

> > ...