------- Comment #5 from rearnsha at gcc dot gnu dot org 2005-11-18 20:05
-------
Sometimes when the reload needs to reload an expression that is a subreg
of a wider register, X, into a different class than X it will call
find_valid_class. This routine simply checks whether the class has any
registers that *could* hold the value in X, it does not check that there
*are* any such registers available. This was causing a problem on ARM in
some circumstances because HARD_REGNO_MODE_OK was returning true for a
class that would be suitable *if* such registers were available on the
machine, but because of compilation options in force the set was currently
empty.
--
rearnsha at gcc dot gnu dot org changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |RESOLVED
Resolution| |FIXED
Target Milestone|--- |4.0.4
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24914
