This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH] another obvious altivec testcase fix
On Fri, Aug 13, 2004 at 11:19:02AM -0700, Janis Johnson wrote:
> On Fri, Aug 13, 2004 at 02:11:50PM -0400, Daniel Jacobowitz wrote:
> > On Thu, Aug 12, 2004 at 11:11:04AM -0700, Janis Johnson wrote:
> > > This patch makes an AltiVec execution test check for the existence
> > > of AltiVec hardware before using any AltiVec instructions. Committed
> > > to mainline as obvious after testing on powerpc64-linux systems with
> > > and without VMX hardware.
> >
> > You may need to mark main1 as noinline for this to work...
>
> Ah, that might explain why it works for me sometimes but not others.
This test checks compile-time functionality so I'm reverting the previous
changes and turning it into a compile-only test.
2004-08-13 Janis Johnson <janis187@us.ibm.com>
* g++.dg/ext/altivec-2.C: Revert previous changes and don't execute.
--- altivec-2.C.orig 2004-08-13 13:14:47.000000000 -0700
+++ altivec-2.C 2004-08-13 13:13:57.000000000 -0700
@@ -1,4 +1,4 @@
-/* { dg-do run { target powerpc*-*-* } } */
+/* { dg-do compile { target powerpc*-*-* } } */
/* { dg-xfail-if "" { "powerpc-ibm-aix*" } { "-maltivec" } { "" } } */
/* { dg-options "-maltivec -Wall" } */
@@ -8,13 +8,15 @@
#include <altivec.h>
#include "altivec_check.h"
-int main1 (void)
+int main (int argc, const char * argv[])
{
int i;
const float cf = 1.0;
vector float v;
const vector float cv = (vector float){1.0, 2.0, 3.0, 4.0};
+ altivec_check ();
+
vec_dst(&cv, i, 0);
v = vec_ld(0, &cv);
v = vec_lde(0, &cf);
@@ -22,9 +24,3 @@ int main1 (void)
return 0;
}
-
-int main (int argc, const char * argv[])
-{
- altivec_check ();
- return main1 ();
-}