Date: Wed, 17 Sep 2003 12:15:03 +0200
To: Olivier Andrieu <andrieu at ijm.jussieu.fr>
Cc: Sven Luther <sven.luther at wanadoo.fr>, lablgtk at kaba.or.jp
Subject: Re: Memory handling bug: cannot access fields in a combo box belonging to a dialog which has been destroyed
Message-ID: <20030917101503.GA6265 at iliana>
References: <20030916103137.GA25181 at redhat.com> <16231.2269.152007.823440 at karryall.dnsalias.org> <20030916162604.GA30706 at redhat.com> <20030917071141A.garrigue at kurims.kyoto-u.ac.jp> <20030917080141.GA2325 at iliana> <16232.13115.936897.972444 at akasha.ijm.jussieu.fr>
Mime-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <16232.13115.936897.972444 at akasha.ijm.jussieu.fr>
From: Sven Luther <sven.luther at wanadoo.fr>

On Wed, Sep 17, 2003 at 12:11:07PM +0200, Olivier Andrieu wrote:
>  Sven Luther [Wednesday 17 September 2003] :
>  >
>  > On Wed, Sep 17, 2003 at 07:11:41AM +0900, Jacques Garrigue wrote:
>  > > The unability to tell whether an object is destroyed or not is a pain.
>  > > Fortunately, GTK makes sure that you don't get a segmentation fault,
>  > > but a reasonable(?) error message.
>  > 
>  > Err, i thought that the gtk+ widget were reference-counted garbage
>  > collected or something such ? 
> 
> They are, but "destroying" a widget doesn't mean "release its
> memory". From the GTK+ documentation : 
> 
>   « The "destroy" signal asks all code owning a reference to an object
>   to release said reference. So, for example, if you call
>   gtk_object_destroy() on a GtkWindow, GTK+ will release the reference
>   count that it owns; if you call gtk_object_destroy() on a GtkButton,
>   then the button will be removed from its parent container and the
>   parent container will release its reference to the button. Because
>   these references are released, calling gtk_object_destroy() should
>   result in freeing all memory associated with an object [ ... ]
>   Freeing memory (referred to as finalization) only happens if the
>   reference count reaches zero. »
> 
> Objects that appear on the caml side have an extra reference, so they
> are not GTK-finalized when they are destroyed, but later, when they
> are caml-finalized.

Err, so a destroyed object would have a refcount of 1 ?

Friendly,

Sven Luther