This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: ObjC bug on 64-bit architectures?
- To: Stan Shebs <shebs at apple dot com>
- Subject: Re: ObjC bug on 64-bit architectures?
- From: Richard Henderson <rth at redhat dot com>
- Date: Tue, 16 Oct 2001 12:53:08 -0700
- Cc: Ulrich Weigand <Ulrich dot Weigand at de dot ibm dot com>, gcc at gcc dot gnu dot org, Hartmut Penner <HPENNER at de dot ibm dot com>
- References: <OFE8A06FEB.6E791E38-ONC1256AE6.00586780@de.ibm.com> <3BCB7943.32BA90C1@apple.com>
On Mon, Oct 15, 2001 at 05:11:42PM -0700, Stan Shebs wrote:
> I'll take a look at it. I would have expected to see more 64-bit
> problems actually, since ObjC doesn't get much exercise on the
> 64-bit targets, but I'm not going to complain. :-)
64-bit little-endian tends to "happen to work" much more often than
big-endian. It's one of the reasons why I've always believed that
Little is the One True Endian. ;-)
Anyway, the best solution to this problem is to create proper record
types, create proper var_decls, and call the data output routines in
final.c and varasm.c. Instead of what we do at the moment, which is
to emit each data element by hand.
r~