Re: [PATCH] tests/qtest: netdev: test stream and dgram backends

["Daniel P. Berrangé" <berrange@redhat.com>]

On Fri, Nov 04, 2022 at 11:58:29AM +0100, Laurent Vivier wrote:
> On 11/4/22 10:41, Daniel P. Berrangé wrote:
> ...
> > > +static void test_stream_unix(void)
> > > +{
> > > +    QTestState *qts0, *qts1;
> > > +    char *expect;
> > > +    gchar *path;
> > > +    int ret;
> > > +
> > > +    ret = g_file_open_tmp("netdev-XXXXXX", &path, NULL);
> > > +    g_assert_true(ret >= 0);
> > > +    close(ret);
> > 
> > This is creating a zero length plain file, and then paassing
> > that as a path for the UNIX socket.
> > 
> > This is pretty dubious and only works because the code will
> > be doing 'unlink' on the path. Just delete this as there's
> > no reason to pre-create anything on disk for UNIX sockets.
> > 
> 
> The idea here is to generate a path for the socket and to be sure this path
> is actually not already in use.

Create a temporary directory, and let it create a socket inside
that dir ?

> The same for the abstract one, how to be sure we are not running the same
> test concurrently and select a different unix name?

I guess creating the temp file gives you an indirect guarantee, because
we know the tempfile path will be transformed to an abstract socket path
by twiddling byte 0 to NUL.

With regards,
Daniel
-- 
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|