Irix6 long doubles implemented wrong? (27_io/ostream_inserter_arith)
Kaveh R. Ghazi
ghazi@caip.rutgers.edu
Tue Dec 24 06:34:00 GMT 2002
> From: Alexandre Oliva <aoliva@redhat.com>
>
> On Dec 22, 2002, Alexandre Oliva <aoliva@redhat.com> wrote:
>
> > ? gcc/config/mips/_tilib.c
>
> Here's the file that I had failed to cvs add before taking the diff:
I tried out this stuff on irix6.5 against the trunk. I had to add
this minor patch for it to work there as opposed to 3.3-pre:
--- gcc-CVS20021223/gcc/config/mips/_tilib.c~ 2002-12-23 17:11:54.306377600 -0500
+++ gcc-CVS20021223/gcc/config/mips/_tilib.c 2002-12-23 17:08:53.719848000 -0500
@@ -22,6 +22,8 @@ Boston, MA 02111-1307, USA. */
#include "tconfig.h"
#include "tsystem.h"
+#include "coretypes.h"
+#include "tm.h"
#ifndef LIBGCC2_WORDS_BIG_ENDIAN
#define LIBGCC2_WORDS_BIG_ENDIAN WORDS_BIG_ENDIAN
Bootstrap went ok, however there were a couple of regressions in the
c-torture testsuite:
> FAIL: gcc.c-torture/execute/va-arg-5.c execution, -O0
> FAIL: gcc.c-torture/execute/va-arg-5.c execution, -O1
> FAIL: gcc.c-torture/execute/va-arg-5.c execution, -O2
> FAIL: gcc.c-torture/execute/va-arg-5.c execution, -O3 -fomit-frame-pointer
> FAIL: gcc.c-torture/execute/va-arg-5.c execution, -O3 -g
> FAIL: gcc.c-torture/execute/va-arg-5.c execution, -Os
> FAIL: gcc.c-torture/execute/ieee/20011123-1.c execution, -O0
> FAIL: gcc.c-torture/execute/ieee/20011123-1.c execution, -O1
> FAIL: gcc.c-torture/execute/ieee/20011123-1.c execution, -O2
> FAIL: gcc.c-torture/execute/ieee/20011123-1.c execution, -O3 -fomit-frame-pointer
> FAIL: gcc.c-torture/execute/ieee/20011123-1.c execution, -O3 -g
> FAIL: gcc.c-torture/execute/ieee/20011123-1.c execution, -Os
These tests both manipulate long doubles.
Also the original test 27_io/ostream_inserter_arith still fails.
However it looks as if it's in a different place, I'll check further
and provide details.
--Kaveh
--
Kaveh R. Ghazi ghazi@caip.rutgers.edu
More information about the Gcc-patches
mailing list