[Bug target/19780] New: Floating point computation far slower for -mfpmath=sse

[uros at kss-loka dot si]

The testcase from PR 8126 runs ~20% slower when compiled with -mfpmath=sse:

--cut here--
#include <stdio.h>

typedef float real;

int
main (int argc, char *argv[])
{
  int i;

  real v1x, v1y, v1z;
  real v2x, v2y, v2z;
  real v3x, v3y, v3z;

  printf ("Start?\n");

  v1x = 1.;
  v1y = 0.;
  v1z = 0.;

  v2x = 0.;
  v2y = 1.;
  v2z = 0.;

  for (i = 0; i < 100000000; i++)
    {
      v3x = v1y * v2z - v1z * v2y;
      v3y = v1z * v2x - v1x * v2z;
      v3z = v1x * v2y - v1y * v2x;

      v1x = v2x;
      v1y = v2y;
      v1z = v2z;

      v2x = v3x;
      v2y = v3y;
      v2z = v3z;
    }

  printf ("Stop!\n");
  printf ("Result = %f, %f, %f\n", v3x, v3y, v3z);

  return 0;
}
--cut here--

gcc -O3 -march=pentium4
real    0m0.603s
user    0m0.602s
sys     0m0.002s

gcc -O3 -march=pentium4 -mfpmath=sse
real    0m0.726s
user    0m0.727s
sys     0m0.000s

-- 
           Summary: Floating point computation far slower for -mfpmath=sse
           Product: gcc
           Version: 4.0.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: target
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: uros at kss-loka dot si
                CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: i686-pc-linux-gnu
  GCC host triplet: i686-pc-linux-gnu
GCC target triplet: i686-pc-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19780