[Bug rtl-optimization/59728] [4.9 Regression] gcc.target/mips/octeon-baddu-1.c is failing

[pinskia at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=59728

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |missed-optimization
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2014-01-08
          Component|target                      |rtl-optimization
   Target Milestone|---                         |4.9.0
            Summary|gcc.target/mips/octeon-badd |[4.9 Regression]
                   |u-1.c is failing            |gcc.target/mips/octeon-badd
                   |                            |u-1.c is failing
     Ever confirmed|0                           |1

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Steve Ellcey from comment #0)
> I am not sure exactly what checkin caused this change or if the change
> in generated code is correct or not.

The generated code is correct just non-optimal.  

The most likely candidate patch which caused this performance regression is:
2013-12-10  Eric Botcazou  <ebotcazou@adacore.com>

        PR rtl-optimization/58295
        * simplify-rtx.c (simplify_truncation): Restrict the distribution for
        WORD_REGISTER_OPERATIONS targets.

Which was partly expected and octeon's baddu instruction was mentioned in the
thread about this patch too.