[Bug rtl-optimization/63340] [5.0 regression] FAIL: gcc.dg/atomic/c11-atomic-exec-2.c -O2 execution test

[rsandifo at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63340

rsandifo at gcc dot gnu.org <rsandifo at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |ASSIGNED
   Last reconfirmed|                            |2014-09-29
           Assignee|unassigned at gcc dot gnu.org      |rsandifo at gcc dot gnu.org
     Ever confirmed|0                           |1

--- Comment #4 from rsandifo at gcc dot gnu.org <rsandifo at gcc dot gnu.org> ---
Sorry for the slow response to this.  Been looking for a way
out that looks reasonable, think I finally have one.

The problem is to do with the union class GR_AND_FR_REGS.
Before my patch this was rejected if a register underwant
a mode change that is forbidden by FR_REGS, whereas after
the patch it is accepted.

Testing a patch now, but I expect it will need a couple
of revisions.