[Bug rtl-optimization/125308] [16/17 Regression] i386: ICE in extract_insn during IRA pass with -m32 -msse3 -mfpmath=sse -ffloat-store -O2

[jakub at gcc dot gnu.org via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=125308

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P2
                 CC|                            |jakub at gcc dot gnu.org
           Assignee|unassigned at gcc dot gnu.org      |jakub at gcc dot gnu.org
   Last reconfirmed|                            |2026-05-14
             Status|UNCONFIRMED                 |ASSIGNED
     Ever confirmed|0                           |1
            Summary|i386: ICE in extract_insn   |[16/17 Regression] i386:
                   |during IRA pass with -m32   |ICE in extract_insn during
                   |-msse3 -mfpmath=sse         |IRA pass with -m32 -msse3
                   |-ffloat-store -O2           |-mfpmath=sse -ffloat-store
                   |                            |-O2
   Target Milestone|---                         |16.2

--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Reduced testcase -m32 -msse2  -O2 -ffloat-store:
float a[4];

void
foo ()
{
  for (int i = 0; i < 4; i++)
    a[i] = a[i] > 0 ? a[i] : 0;
}

Started with r16-5846-g106e001b4bb5313a8d59b858e80d5633f90fcec0