Re: gtk+ on linux/ia64
- From: Karl Koehler <koehler or uni-bonn de>
- To: Tim Janik <timj gtk org>
- Cc: George <jirka 5z com>, Havoc Pennington <hp redhat com>, Gtk+ Developers <gtk-devel-list gnome org>
- Subject: Re: gtk+ on linux/ia64
- Date: Sun, 6 Jan 2002 20:30:28 +0100 (MEZ)
Thank you for all the input.
> > > 64 bit issues are really common when using varargs functions like
> > > g_signal_new(). Look at the call and see if you are getting a
> > > different number of bits going in than coming out.
Mostly, you will get a SIGILL if you take out more than you put in,
because the return adress on the stack will no longer be ok.
I don't really like va_args ..
> so chances are that changing gtype.h:
>
> -#define G_TYPE_FLAG_RESERVED_ID_BIT (1 << 0)
> +#define G_TYPE_FLAG_RESERVED_ID_BIT ((GType) (1 << 0))
I made the change and recompiled everything. The new crash looks like this:
Program received signal SIGILL, Illegal instruction.
g_type_check_is_value_type (type=2305843009213693976) at gtype.c:2731
2731 {
(gdb) where
#0 g_type_check_is_value_type (type=2305843009213693976) at gtype.c:2731
#1 0x20000000009b2ac0 in g_type_check_is_value_type (type=0) at gtype.c:2731
I think I'll have to dig deeper into this, because the fix does not
seem to be absolute. Ah well, ...
> I have 64 bit hardware so I can help out with these issues. While it
> is not ia64 (alphas are cooler:), issues should be similar.
I don't access to an alpha, :( and I'd like my program to run on this
Itanium here :| .
Karl
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]