This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: C++ PATCH: Fix PRs 16518, 16337
Richard Henderson wrote:
On Tue, Jul 13, 2004 at 03:39:59PM -0700, Mark Mitchell wrote:
Do we now support them, even on machines that don't have
vector hardware?
Yes, for a year or two now. They get decomposed.
But seem to result in warnings on some platforms. :-)
I checked in the patch attached as a test case for PR 16337.
--
Mark Mitchell
CodeSourcery, LLC
(916) 791-8304
mark@codesourcery.com
2004-07-17 Mark Mitchell <mark@codesourcery.com>
PR c++/16337
* g++.dg/parse/attr2.C: New test.
Index: testsuite/g++.dg/parse/attr2.C
===================================================================
RCS file: testsuite/g++.dg/parse/attr2.C
diff -N testsuite/g++.dg/parse/attr2.C
*** /dev/null 1 Jan 1970 00:00:00 -0000
--- testsuite/g++.dg/parse/attr2.C 17 Jul 2004 18:05:44 -0000
***************
*** 0 ****
--- 1,15 ----
+ // PR c++/16337
+ // On i686-pc-linux-gnu, without options, we get:
+ // warning: SSE vector return without SSE enabled changes the ABI
+ // { dg-options "-w" }
+
+ #define vector __attribute__((vector_size(16)))
+ vector signed int foo (void)
+ {
+ vector signed int v1 = { 1, 2, 3, 4 };
+ vector signed int v2, v3;
+ vector signed int v4;
+ v2 = v1;
+ v3 = v1;
+ v4 = v1;
+ }