This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Weird stuff in xm-proc.h files ?
- To: egcs at egcs dot cygnus dot com
- Subject: Weird stuff in xm-proc.h files ?
- From: Marc Espie <Marc dot Espie at liafa dot jussieu dot fr>
- Date: Wed, 10 Feb 1999 00:57:42 +0100
- Cc: "Steve Murphree, Jr." <smurph at airnet dot net>, Dale Rahn <drahn at urbana dot css dot mot dot com>, niklas at petra dot appli dot se
I've been fixing a few minor details over the few days with xm-MACHINE.h
files.
I don't like some of these, point in case:
rs6000/xm-rs6000.h, or m88k/xm-m88k.h
I was naively hoping to find proc-dependent information in that and ONLY that,
but it seem that the OS somehow got hard-coded in there...
it does not look like what the documentation says...
I've got basically several choices:
- cp rs6000/xm-rs6000.h rs6000/xm-rs6000.h, edit and remove the icky parts.
- write a stubby rs6000/xm-openbsd.h that will
#include <rs6000/xm-rs6000.h>
#undef USG
#undef COLLECT_EXPORT_LIST
- fix rs6000/xm-rs6000.h to *only* cover rs6000, then write an rs6000/xm-aix.h
that deals with the proper aix parts, and fix configure.in accordingly.
Basically, deal's the same for m88k.
I think the proper course is the last option, but I can live with a kludge
if I have to :)
Seeking advice, please tell me what I should do...
--
Marc Espie
|anime, sf, juggling, unicycle, acrobatics, comics...
|AmigaOS, OpenBSD, C++, perl, Icon, PostScript...
| `real programmers don't die, they just get out of beta'