[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] [PATCH 06/24] vhost-user: check vhost_user_write() retu
From: |
Michael S. Tsirkin |
Subject: |
Re: [Qemu-devel] [PATCH 06/24] vhost-user: check vhost_user_write() return value |
Date: |
Mon, 4 Jul 2016 18:46:46 +0300 |
On Fri, Jun 24, 2016 at 02:49:22PM +0200, Marc-André Lureau wrote:
> Hi
>
> On Thu, Jun 23, 2016 at 7:08 PM, Michael S. Tsirkin <address@hidden> wrote:
> >>
> >> Callers do not always ignore it (and in general it should not, should
> >> it?), this helps breaking the execution at right moment, help debugging,
> >> code consistency, good practices etc (perhaps it's too obvious to me and I
> >> am missing something?)
> >
> > Reporting error up the stack is helpful if it's handled in some way.
> > If we just keep guest going on this error, then we could
> > maybe log it for debug build but that's all.
>
> Reporting up to guest somehow would be a good thing at some point,
I'm not so sure. If backend will reconnect shortly, we can
handle everything transparently.
> so
> I think we should start from the bottom. vhost-user lacks error
> handling, let's add it.
>
> Regarding debug build messages, I don't think it's enough. As long as
> we don't have an official supported way to handle disconnect. It's
> better to report an error than be silent.
Let's just work on handling it. If we need debug messages to help us
reach that goal fine. But I don't see many reasons to propagate
return codes back and forth if caller just prints and ignores it.
Print it where it's detected :)
>
> --
> Marc-André Lureau
- Re: [Qemu-devel] [PATCH 06/24] vhost-user: check vhost_user_write() return value,
Michael S. Tsirkin <=