Hi
On Wed, Jun 26, 2019 at 7:56 PM Eric Blake wrote:
>
> On 6/16/19 4:36 PM, Michael S. Tsirkin wrote:
> > From: Marc-André Lureau
> >
> > This check shouldn't be necessary, since _fatal is given as
> > argument and will exit() on failure. However, this change should
> > silence coverity CID 1401761 & 1401705.
> >
> > Signed-off-by: Marc-André Lureau
> > Message-Id: <20190605145829.7674-3-marcandre.lur...@redhat.com>
> > Reviewed-by: Michael S. Tsirkin
> > Signed-off-by: Michael S. Tsirkin
> > ---
> > contrib/vhost-user-gpu/main.c | 4
> > contrib/vhost-user-input/main.c | 4
> > 2 files changed, 8 insertions(+)
> >
> > diff --git a/contrib/vhost-user-gpu/main.c b/contrib/vhost-user-gpu/main.c
> > index 9614c9422c..e0b6df5b4d 100644
> > --- a/contrib/vhost-user-gpu/main.c
> > +++ b/contrib/vhost-user-gpu/main.c
> > @@ -1160,6 +1160,10 @@ main(int argc, char *argv[])
> >
> > if (opt_socket_path) {
> > int lsock = unix_listen(opt_socket_path, _fatal);
> > +if (lsock < 0) {
> > +g_printerr("Failed to listen on %s.\n", opt_socket_path);
> > +exit(EXIT_FAILURE);
> > +}
>
> 4 lines to appease Coverity; wouldn't the following one-liner also do
> the trick?
>
> int lsock = unix_listen(opt_socket_path, _fatal);
> assert (lsock >= 0);
Probably, I didn't think too much about it. However looking at it now,
it would be worthwhile to report the error to the user in a friendlier
way than what error_fatal do. So I guess a follow-up patch should
introduce a local Error.
>
> --
> Eric Blake, Principal Software Engineer
> Red Hat, Inc. +1-919-301-3226
> Virtualization: qemu.org | libvirt.org
>
--
Marc-André Lureau
