[Bug middle-end/18641] [4.0 Regression] Another ICE caused by reload of a psuedo reg into f0 for a DImode expr

[uweigand at gcc dot gnu dot org]

------- Additional Comments From uweigand at gcc dot gnu dot org  2004-12-02 16:14 -------
As to 1), this can be achieved by marking the 'f' alternatives
in the *movdi_internal32 pattern as '!*f'; this will prevent 
both regclass and reload from using the alternatives unless
there was a floating-point register involved already before
reload.

As to 2), the proper way to go via memory would be to return
NO_REGS from PREFERRED_RELOAD_CLASS when asked how to load
a constant into (a superclass of) FLOAT_REGS.  Then reload
takes care of force_const_mem and everything related.



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18641