[Bug tree-optimization/57162] Ofast does not make use of avx while O3 does

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=57162



Richard Biener <rguenth at gcc dot gnu.org> changed:



           What    |Removed                     |Added

----------------------------------------------------------------------------

           Keywords|                            |missed-optimization

             Status|UNCONFIRMED                 |NEW

   Last reconfirmed|                            |2013-05-06

             Blocks|                            |53947

     Ever Confirmed|0                           |1



--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> 2013-05-06 
11:46:24 UTC ---

-ffast-math reassociates the expressions which then causes us to fail

the SLP build because of



-t.C:6: note: Load permutation 0 0 0 0 1 1 1 1 2 2 2 2 3 3 3 3 

-t.C:6: note: vect_model_load_cost: aligned.

-t.C:6: note: vect_model_load_cost: inside_cost = 1, prologue_cost = 0 .

+t.C:6: note: Load permutation 0 1 1 1 1 0 0 0 3 2 2 2 2 3 3 3 

+t.C:6: note: Build SLP failed: unsupported load permutation c[i_37][0] =

sum_44;