Re: [Vala] object leak in async and closure

[Michele Dionisio <michele dionisio gmail com>]

Yes it is true, I'm sorry my fault. I'm rying your last fix


2014/1/4 Luca Bruno <lethalman88 gmail com>

> I see "destroy pippo" with valac-0.22.
>
>
> On Sat, Jan 4, 2014 at 9:28 PM, Michele Dionisio <
> michele dionisio gmail com> wrote:
>
> > I will try, reading the comment on git it seams a problem introduced in
> > vala 0.23.1 but I have test also vala-0.22 and it have the same problem.
> >
> >
> > 2014/1/4 Luca Bruno <lethalman88 gmail com>
> >
> > > I've just fixed the bug in vala git master, can you check?
> > > Commit 550b107896
> > >
> > >
> > > On Sat, Jan 4, 2014 at 7:56 PM, Luca Bruno <lethalman88 gmail com>wrote:
> > >
> > > > Please report a bug at bugzilla.gnome.org .
> > > > So, using "o" in the lambda seems to be the problem.
> > > >
> > > >
> > > > On Sat, Jan 4, 2014 at 7:39 PM, Michele Dionisio <
> > > > michele dionisio gmail com> wrote:
> > > >
> > > > > Hi, I'm new in this mailing list and quite new in vala programming.
> > > > >
> > > > > I have a big software that somethime leak objct. Now I have replicate
> > > > > the
> > > > > issue with a small code:
> > > > >
> > > > >
> > > > > ############################################################################
> > > > >     class Pippo: Object {
> > > > >         public async void foo () {
> > > > >             Object o = new Object ();
> > > > >             stdout.printf("out-- %u\n",o.ref_count);
> > > > >             stdout.printf("this -- %u\n",this.ref_count);
> > > > >
> > > > >             SourceFunc f = () => {
> > > > >                 stdout.printf("infunc-- %u\n",o.ref_count);
> > > > >                 stdout.printf("in this -- %u\n",this.ref_count);
> > > > >                 //this.unref();
> > > > >                 stdout.printf("in this -- %u\n",this.ref_count);
> > > > >                 stdout.printf("end function f\n");
> > > > >                 return false;
> > > > >             };
> > > > >             GLib.Idle.add( (owned) f );
> > > > >
> > > > >             stdout.printf("end function foo\n");
> > > > >         }
> > > > >
> > > > >         ~Pippo() {
> > > > >             stdout.printf("destroy pippo\n");
> > > > >         }
> > > > >     }
> > > > >
> > > > >     void main() {
> > > > >         Pippo ? tmp = new Pippo();
> > > > >
> > > > >         tmp.foo.begin();
> > > > >
> > > > >         MainLoop loop = new MainLoop ();
> > > > >
> > > > >         GLib.Timeout.add(1000, () => {stdout.printf("timeout1\n"); tmp
> > > > > =
> > > > > null; return false;});
> > > > >
> > > > >         GLib.Timeout.add(2000, () => {stdout.printf("timeout2\n");
> > > > > loop.quit(); return false;});
> > > > >
> > > > >
> > > > >         loop.run ();
> > > > > }
> > > > >
> > > > >
> > > > > ############################################################################
> > > > >
> > > > > If in the function
> > > > > SourceFunc f
> > > > >
> > > > > there is any reference to any variable of foo or to any object of this
> > > > > we
> > > > > have the reference counting of this increase.
> > > > >
> > > > > and so after timeout1 the object tmp is not destroy. (And there is no
> > > > > reference on it)
> > > > >
> > > > > The only way to fix the issue is to manually decrease the reference of
> > > > > this. But this is a big workaraound.
> > > > >
> > > > > Someone have any idea where is my fault or if there is a bug in vala
> > > > > itself?
> > > > >
> > > > > regards
> > > > > _______________________________________________
> > > > > vala-list mailing list
> > > > > vala-list gnome org
> > > > > https://mail.gnome.org/mailman/listinfo/vala-list
> > > >
> > > >
> > > >
> > > > --
> > > > www.debian.org - The Universal Operating System
> > >
> > >
> > >
> > > --
> > > www.debian.org - The Universal Operating System
>
>
> --
> www.debian.org - The Universal Operating System