[Bug tree-optimization/46085] New: [4..6 Regression] gcc.dg/vect/fast-math-vect-reduc-[57].c failed with -mavx -ffast-math -O3

[Bug tree-optimization/46085] New: [4..6 Regression] gcc.dg/vect/fast-math-vect-reduc-[57].c failed with -mavx -ffast-math -O3

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46085

           Summary: [4..6 Regression]
                    gcc.dg/vect/fast-math-vect-reduc-[57].c failed with
                    -mavx -ffast-math -O3
           Product: gcc
           Version: 4.6.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: hjl.tools@gmail.com


[hjl@gnu-18 gcc]$ ./xgcc -B./
/export/gnu/import/git/gcc-avx/gcc/testsuite/gcc.dg/vect/fast-math-vect-reduc-5.c
-mavx  -ffast-math  -O3[hjl@gnu-18 gcc]$ ./a.out 
Aborted
[hjl@gnu-18 gcc]$ ./xgcc -B./
/export/gnu/import/git/gcc-avx/gcc/testsuite/gcc.dg/vect/fast-math-vect-reduc-7.c
-mavx  -ffast-math  -O3
[hjl@gnu-18 gcc]$ ./a.out 
Aborted
[hjl@gnu-18 gcc]$ 

They failed due to 256bit vectorizer.

[Bug tree-optimization/46085] [4..6 Regression] gcc.dg/vect/fast-math-vect-reduc-[57].c failed with -mavx -ffast-math -O3

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46085

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |rguenth at gcc dot gnu.org
   Target Milestone|---                         |4.6.0

[Bug tree-optimization/46085] [4..6 Regression] gcc.dg/vect/fast-math-vect-reduc-[57].c failed with -mavx -ffast-math -O3

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46085

--- Comment #1 from H.J. Lu <hjl.tools at gmail dot com> 2010-10-19 19:48:35 UTC ---
[hjl@gnu-18 gcc]$ cat x.c
extern void abort ();

float b[16] = {0,3,6,9,12,15,18,21,24,27,30,33,36,39,42,45};
float c[16] = {0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15};

int main ()
{
  int i;
  float diff = 2;

  for (i = 0; i < 16; i++) {
    diff += (b[i] - c[i]);
  }

  if (diff != 242)
    abort ();

  return 0;
}
[hjl@gnu-18 gcc]$  ./xgcc -B./ x.c -ftree-vectorize -g -mavx -O  -ffast-math 
[hjl@gnu-18 gcc]$ ./a.out 
Aborted
[hjl@gnu-18 gcc]$  ./xgcc -B./ x.c -ftree-vectorize -g -mavx -O -ffast-math    
-mno-avx -msse4
[hjl@gnu-18 gcc]$ ./a.out 
[hjl@gnu-18 gcc]$

[Bug tree-optimization/46085] [4..6 Regression] gcc.dg/vect/fast-math-vect-reduc-[57].c failed with -mavx -ffast-math -O3

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46085

--- Comment #2 from H.J. Lu <hjl.tools at gmail dot com> 2010-10-19 21:32:05 UTC ---
(define_expand "reduc_splus_v8sf"
  [(match_operand:V8SF 0 "register_operand" "")
   (match_operand:V8SF 1 "register_operand" "")]
  "TARGET_AVX"
{
  rtx tmp = gen_reg_rtx (V8SFmode);
  rtx tmp2 = gen_reg_rtx (V8SFmode);
  emit_insn (gen_avx_haddv8sf3 (tmp, operands[1], operands[1]));
  emit_insn (gen_avx_haddv8sf3 (tmp2, operands[1], operands[1]));
  emit_insn (gen_avx_haddv8sf3 (operands[0], tmp2, tmp2));
  DONE;
})

and

(define_expand "reduc_splus_v4df"
  [(match_operand:V4DF 0 "register_operand" "")
   (match_operand:V4DF 1 "register_operand" "")]
  "TARGET_AVX"
{
  rtx tmp = gen_reg_rtx (V4DFmode);
  emit_insn (gen_avx_haddv4df3 (tmp, operands[1], operands[1]));
  emit_insn (gen_avx_haddv4df3 (operands[0], tmp, tmp));
  DONE;
})

are incorrect.

[Bug tree-optimization/46085] [4..6 Regression] gcc.dg/vect/fast-math-vect-reduc-[57].c failed with -mavx -ffast-math -O3

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46085

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                URL|                            |http://gcc.gnu.org/ml/gcc-p
                   |                            |atches/2010-10/msg01681.htm
                   |                            |l

--- Comment #3 from H.J. Lu <hjl.tools at gmail dot com> 2010-10-19 22:15:07 UTC ---
A patch is posted at

http://gcc.gnu.org/ml/gcc-patches/2010-10/msg01681.html

[Bug tree-optimization/46085] [4..6 Regression] gcc.dg/vect/fast-math-vect-reduc-[57].c failed with -mavx -ffast-math -O3

[hjl at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46085

--- Comment #4 from hjl at gcc dot gnu.org <hjl at gcc dot gnu.org> 2010-10-20 11:21:22 UTC ---
Author: hjl
Date: Wed Oct 20 11:21:19 2010
New Revision: 165719

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=165719
Log:
Correct reduc_splus_v8sf and reduc_splus_v4df.

gcc/

2010-10-20  H.J. Lu  <hongjiu.lu@intel.com>

    PR target/46085
    * config/i386/sse.md (reduc_splus_v8sf): Updated.
    (reduc_splus_v4df): Likewise.

gcc/testsuite/

2010-10-20  H.J. Lu  <hongjiu.lu@intel.com>

    PR target/46085
    * gcc.target/i386/pr46085-1.c: New.
    * gcc.target/i386/pr46085-2.c: Likewise.

Added:
    trunk/gcc/testsuite/gcc.target/i386/pr46085-1.c
    trunk/gcc/testsuite/gcc.target/i386/pr46085-2.c
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/config/i386/sse.md
    trunk/gcc/testsuite/ChangeLog

[Bug tree-optimization/46085] [4..6 Regression] gcc.dg/vect/fast-math-vect-reduc-[57].c failed with -mavx -ffast-math -O3

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46085

H.J. Lu <hjl.tools at gmail dot com> changed:

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

--- Comment #5 from H.J. Lu <hjl.tools at gmail dot com> 2010-10-20 11:26:39 UTC ---
Fixed.