-msoft-float on PowerPC

[Michael Meissner]

I believe that the change that you made to egcs on the 18th of
September to pass software floating point in the correct registers
should be done differently.  In particular, I believe that we need a
compatibility switch in order to not break people's current code (I
will be writing one shortly, and the change needs to be highlighted
more -- we also need to have different multilibs).  Also, there was no
ChangeLog entry for this change -- I only discovered it when diffing
egcs with my current sandbox for the ASM_OUTPUT_MI_THUNK change.

-- 
Michael Meissner, Cygnus Solutions (Massachusetts office)
4th floor, 955 Massachusetts Avenue, Cambridge, MA 02139, USA
meissner@cygnus.com,	617-354-5416 (office),	617-354-7161 (fax)