[Bug rtl-optimization/123276] [16 regression] ICE when building GNAT (RTL check: expected code 'const_double' and not mode 'VOID', have code 'const_int' and mode 'VOID' in compress_float_constant, at expr.cc:4786)

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=123276

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Assignee|unassigned at gcc dot gnu.org      |pinskia at gcc dot gnu.org
             Status|NEW                         |ASSIGNED

--- Comment #3 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
For float modes, it is not valid to change a:a+b to a+(0:b) except without
fast-math so rejecting them early seems reasonable.