[Spice-devel] [PATCH v2 15/19] replay: Free spice server to detect leaks

[Pavel Grunt]

On Tue, 2016-11-29 at 04:37 -0500, Frediano Ziglio wrote:
> > 
> > On Tue, 2016-11-29 at 04:03 -0500, Frediano Ziglio wrote:
> > > > 
> > > > On Fri, 2016-11-25 at 14:52 +0000, Frediano Ziglio wrote:
> > > > > Signed-off-by: Frediano Ziglio <fziglio at redhat.com>
> > > > 
> > > > Acked-by: Pavel Grunt <pgrunt at redhat.com>
> > > > > ---
> > > > >  server/tests/replay.c | 2 +-
> > > > >  1 file changed, 1 insertion(+), 1 deletion(-)
> > > > > 
> > > > > diff --git a/server/tests/replay.c b/server/tests/replay.c
> > > > > index f32fa2f..23d4125 100644
> > > > > --- a/server/tests/replay.c
> > > > > +++ b/server/tests/replay.c
> > > > > @@ -221,7 +221,6 @@ static void end_replay(void)
> > > > >          kill(client_pid, SIGINT);
> > > > >          waitpid(client_pid, &child_status, 0);
> > > > >      }
> > > > > -    exit(0);
> > > > >  }
> > > > >  
> > > > >  static void release_resource(QXLInstance *qin, struct
> > > > > QXLReleaseInfoExt release_info)
> > > > > @@ -440,6 +439,7 @@ int main(int argc, char **argv)
> > > > >      if (print_count)
> > > > >          g_print("Counted %d commands\n", ncommands);
> > > > >  
> > > > > +    spice_server_destroy(server);
> > > > >      end_replay();
> > > > >      g_async_queue_unref(display_queue);
> > > > >      g_async_queue_unref(cursor_queue);
> > > 
> > > Are you suggesting I merge these patches (15/19 and 18/19)
> > > before
> > > the patches that implements the cleaning?
> > 
> > if the patches are sent intentionally in series than they should
> > be
> > merged in the series - author must know :) I will review rest of
> > the
> > patches today
> > 
> > About these two i think merging them is safe - it may actually
> > show
> > the difference in the leak
> > 
> > Pavel
> > 
> 
> Yes, I was thinking about your previous comment on another patch
> about merging a function without being used.

Here it is used, no ?
> 
> On a first though I was thinking these patch could be merged before
> so to see the improvements. But on a second though at least we
> should
> merge before the patches that close the worker thread or you'll get
> crashes here and there.

If there is a risk of crashing than rather wait with merging to avoid
potential future issues when bisecting 

Pavel

> 
> Frediano