* config/rs6000/rs6000.c (struct processor_costs): Add new fields

	* config/rs6000/rs6000.c (struct processor_costs): Add new fields
	for simple floating point operations "fp", double precision
	multiplication "dmul", and single and double precision division
	"sdiv" and "ddiv".  Update all CPU variant tables as appropriate.
	(ppc630_cost): New table split from ppc620_cost, to distinguish
	differences in floating point latencies.
	(rs6000_override_options): Use ppc630_cost for PROCESSOR_PPC630.
	(rs6000_rtx_costs): Add support for single and double precision
	floating point addition, subtraction, multiplication, division,
	fused-multiply-add, fused-multiply-sub, negation, absolute value
	and negative absolute value.  Tweak MEM case to use prefered idiom.

git-svn-id: svn+ssh://gcc.gnu.org/svn/gcc/trunk@84226 138bc75d-0d04-0410-961f-82ee72b054a4