This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: 3.3.1/3.4 PATCH: Don't pass --gstabs to Solaris 2/Intel native as
Rainer Orth writes:
> Jakub Jelinek writes:
>
> > Wouldn't it be much better to scan as's stdout/stderr for any warnings
> > in addition to checking error flag.
>
> Possibly: since this test tries to assemble an empty input file, there
> shouldn't be any warnings (or in fact any output whatsoever).
>
> I could implement this if desired.
Though nobody commented, I went ahead and implemented the suggestion.
Bootstrapped without regressions on the 3.3 branch, verified on mainline.
Ok now for 3.3 branch and mainline?
Rainer
-----------------------------------------------------------------------------
Rainer Orth, Faculty of Technology, Bielefeld University
Thu Jun 12 19:17:35 2003 Rainer Orth <ro@TechFak.Uni-Bielefeld.DE>
* configure.in (gcc_cv_as_gstabs_flag): Disable if assembler warns.
* configure: Regenerate.
Fixes PR driver/9362.
Index: gcc/configure.in
===================================================================
RCS file: /cvs/gcc/gcc/gcc/configure.in,v
retrieving revision 1.686
diff -u -p -r1.686 configure.in
--- gcc/configure.in 14 Jun 2003 12:26:30 -0000 1.686
+++ gcc/configure.in 16 Jun 2003 17:12:25 -0000
@@ -2568,6 +2568,12 @@ elif test x$gcc_cv_as != x -a x"$insn" !
if $gcc_cv_as --gstabs -o conftest.o conftest.s > /dev/null 2>&1 ; then
gcc_cv_as_gstabs_flag="yes"
fi
+ # The native Solaris 9/Intel assembler doesn't understand --gstabs
+ # and warns about it, but still exits successfully. So check for this.
+ if $gcc_cv_as --gstabs -o conftest.o conftest.s 2>&1 | \
+ grep -i warning > /dev/null ; then
+ gcc_cv_as_gstabs_flag="no"
+ fi
rm -f conftest.s conftest.o
fi
if test x"$gcc_cv_as_gstabs_flag" = xyes; then