[evolution-patches] Re: [patch] another 64-bit bug-fix

[Not Zed <notzed ximian com>]

David,

I'm curious, why does this lead to a crash?  That value is never used other than to test against zero, I can't see how it could cause a crash since the cast only changes the type size.

Thanks,
 Michael

On Tue, 2004-12-14 at 18:36 -0800, David Mosberger wrote:
> Hi Z,
>
> I'm trying to get evolution on Debian/ia64 (unstable) to work and
> while tracking down a crash there, I found that the underlying bug
> still exists in the CVS tree.  I can't build the CVS tree easily at
> the moment, but I hope you accept a verbal description of what needs
> fixing.
>
> What's wrong is that in e-data-cal-factor.c, there is one place where
> a pointer is converted to "int" instead of GType, which of course
> leads to instant crashes on 64-bit platforms.
>
> In the CVS tree, the offending cast occurs in:
>
>  e_data_cal_factory_register_backend()
>
> Just look for the line:
>
>  type = GPOINTER_TO_INT (g_hash_table_lookup (kinds, GINT_TO_POINTER (kind)));
>
> and change it to:
>
>  type = (GType) (g_hash_table_lookup (kinds, GINT_TO_POINTER (kind)));
>
> and things will be fine.
>
> Could this change be checked in?
>
> Thanks!
>
> 	--david

--

Michael Zucchi <notzed ximian com> "Ride, Work, Sleep. Beer." Novell's Evolution and Free Software Developer