[Bug testsuite/99685] New: gcc.target/powerpc/divkc3-1.c and mulkc3-1.c fails for 32 bits

[Bug testsuite/99685] New: gcc.target/powerpc/divkc3-1.c and mulkc3-1.c fails for 32 bits

[seurer at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99685

            Bug ID: 99685
           Summary: gcc.target/powerpc/divkc3-1.c and mulkc3-1.c fails for
                    32 bits
           Product: gcc
           Version: 11.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: testsuite
          Assignee: unassigned at gcc dot gnu.org
          Reporter: seurer at gcc dot gnu.org
  Target Milestone: ---

make  -k check-gcc RUNTESTFLAGS="--target_board=unix'{-m32}'
powerpc.exp=gcc.target/powerpc/divkc3-1.c"
FAIL: gcc.target/powerpc/divkc3-1.c execution test
# of expected passes            1
# of unexpected failures        1

make  -k check-gcc RUNTESTFLAGS="--target_board=unix'{-m32}'
powerpc.exp=gcc.target/powerpc/mulkc3-1.c"
FAIL: gcc.target/powerpc/mulkc3-1.c execution test
# of expected passes            1
# of unexpected failures        1


(gdb) run
Starting program: /home/seurer/gcc/git/build/gcc-10-test/./divkc3-1.exe 

Program received signal SIGABRT, Aborted.
0x0fd4b7c8 in __GI_raise (sig=sig@entry=6) at
../nptl/sysdeps/unix/sysv/linux/raise.c:55
55        return INLINE_SYSCALL (tgkill, 3, pid, selftid, sig);
(gdb) where
#0  0x0fd4b7c8 in __GI_raise (sig=sig@entry=6) at
../nptl/sysdeps/unix/sysv/linux/raise.c:55
#1  0x0fd4d584 in __GI_abort () at abort.c:90
#2  0x10000650 in main () at
/home/seurer/gcc/git/gcc-10-test/gcc/testsuite/gcc.target/powerpc/divkc3-1.c:20


  z = divide (5.0q + 5.0jq, 2.0q + 1.0jq);
  a = 3.0q + 1.0jq;
  if (z != a)
    abort ();  // line 20

These two tests both fail when compiled for 32 bits but work fine for 64 bits. 
Are they valid for 32 bits?

They also fail in gcc 10 but back in gcc 9 they were marked as unsupported.

[Bug testsuite/99685] gcc.target/powerpc/divkc3-1.c and mulkc3-1.c fail for 32 bits

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99685

--- Comment #1 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Pat Haugen <pthaugen@gcc.gnu.org>:

https://gcc.gnu.org/g:2ee68beee709e48fce85b8892ff9985acc6a91a8

commit r13-574-g2ee68beee709e48fce85b8892ff9985acc6a91a8
Author: Pat Haugen <pthaugen@linux.ibm.com>
Date:   Tue May 17 15:53:24 2022 -0500

    Fix register count when not splitting Complex IEEE 128-bit args.

    For ABI_V4, we do not split complex args. This created a problem because
    even though an arg would be passed in two VSX regs, we were only advancing
the
    function arg counter by one VSX register. Fixed with this patch.

            PR target/99685

    gcc/
            * config/rs6000/rs6000-call.cc (rs6000_function_arg_advance_1):
Bump
            register count when not splitting IEEE 128-bit Complex.

[Bug testsuite/99685] gcc.target/powerpc/divkc3-1.c and mulkc3-1.c fail for 32 bits

[pthaugen at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99685

pthaugen at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|---                         |FIXED

--- Comment #2 from pthaugen at gcc dot gnu.org ---
Fixed.

[Bug testsuite/99685] gcc.target/powerpc/divkc3-1.c and mulkc3-1.c fail for 32 bits

[bergner at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99685

--- Comment #3 from Peter Bergner <bergner at gcc dot gnu.org> ---
(In reply to pthaugen from comment #2)
> Fixed.

Does this need backporting?

[Bug testsuite/99685] gcc.target/powerpc/divkc3-1.c and mulkc3-1.c fail for 32 bits

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99685

--- Comment #4 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-12 branch has been updated by Pat Haugen
<pthaugen@gcc.gnu.org>:

https://gcc.gnu.org/g:0bc43f9ae1384787b0efcf9a41f00c1eeaf8bb8b

commit r12-8842-g0bc43f9ae1384787b0efcf9a41f00c1eeaf8bb8b
Author: Pat Haugen <pthaugen@linux.ibm.com>
Date:   Tue May 17 15:53:24 2022 -0500

    Fix register count when not splitting Complex IEEE 128-bit args.

    For ABI_V4, we do not split complex args. This created a problem because
    even though an arg would be passed in two VSX regs, we were only advancing
the
    function arg counter by one VSX register. Fixed with this patch.

            PR target/99685

    gcc/
            * config/rs6000/rs6000-call.cc (rs6000_function_arg_advance_1):
Bump
            register count when not splitting IEEE 128-bit Complex.

    (cherry picked from commit 2ee68beee709e48fce85b8892ff9985acc6a91a8)

[Bug testsuite/99685] gcc.target/powerpc/divkc3-1.c and mulkc3-1.c fail for 32 bits

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99685

--- Comment #5 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-10 branch has been updated by Pat Haugen
<pthaugen@gcc.gnu.org>:

https://gcc.gnu.org/g:7af52b39059b4496b17337d0efc7575f17febb26

commit r10-11043-g7af52b39059b4496b17337d0efc7575f17febb26
Author: Pat Haugen <pthaugen@linux.ibm.com>
Date:   Mon Oct 17 14:53:11 2022 -0500

        Fix register count when not splitting Complex IEEE 128-bit args.

        For ABI_V4, we do not split complex args. This created a problem
because
        even though an arg would be passed in two VSX regs, we were only
advancing the
        function arg counter by one VSX register. Fixed with this patch.

        (backported commit 2ee68beee709e48fce85b8892ff9985acc6a91a8)

    gcc/
            PR target/99685
            * config/rs6000/rs6000-call.c (rs6000_function_arg_advance_1): Bump
            register count when not splitting IEEE 128-bit Complex.

[Bug testsuite/99685] gcc.target/powerpc/divkc3-1.c and mulkc3-1.c fail for 32 bits

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99685

--- Comment #6 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-11 branch has been updated by Pat Haugen
<pthaugen@gcc.gnu.org>:

https://gcc.gnu.org/g:e7f0d923dc29965ff92f27245ea63c5726edcacd

commit r11-10319-ge7f0d923dc29965ff92f27245ea63c5726edcacd
Author: Pat Haugen <pthaugen@linux.ibm.com>
Date:   Mon Oct 17 15:11:42 2022 -0500

    Fix register count when not splitting Complex IEEE 128-bit args.

    For ABI_V4, we do not split complex args. This created a problem because
    even though an arg would be passed in two VSX regs, we were only advancing
the
    function arg counter by one VSX register. Fixed with this patch.

    (backported commit 2ee68beee709e48fce85b8892ff9985acc6a91a8)

    gcc/
            PR target/99685
            * config/rs6000/rs6000-call.c (rs6000_function_arg_advance_1): Bump
            register count when not splitting IEEE 128-bit Complex.

[Bug testsuite/99685] gcc.target/powerpc/divkc3-1.c and mulkc3-1.c fail for 32 bits

[pthaugen at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99685

pthaugen at gcc dot gnu.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |CLOSED

--- Comment #7 from pthaugen at gcc dot gnu.org ---
Backports complete.