[Bug middle-end/57503] [4.7/4.8 Regression] Expand uses wrong multiply routine

[gjl at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57503

Georg-Johann Lay <gjl at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |gjl at gcc dot gnu.org

--- Comment #1 from Georg-Johann Lay <gjl at gcc dot gnu.org> ---
Created attachment 30242
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=30242&action=edit
.expand dump

Notice that in rot(), long D.1484_5 is unused and instead the 32-bit value
D.1482_3 is used.

BTW, what does "w*" mean?



;; Function rot (rot, funcdef_no=4, decl_uid=1471, cgraph_uid=4)

rot (unsigned char v)
{
  long int D.1484;
  long int D.1483;
  int D.1482;
  int D.1481;

  # BLOCK 2 freq:10000
  # PRED: ENTRY [100.0%]  (fallthru,exec)
  D.1481_2 = (int) v_1(D);
  D.1482_3 = v_1(D) w* 255;
  D.1484_5 = (long int) D.1482_3;
  D.1483_6 = D.1482_3 w* 254;
  return D.1483_6;
  # SUCC: EXIT [100.0%] 
}