blob: b6b03202ec663ea84d6e4c978274cdd4e9495373 (
plain) (
blame)
1
2
3
4
5
|
FPREM/FPREM1: Support rounding mode differences between these instructions
FPREM1 follows IEEE remainder while FPREM does not.
We should support the differences between these two instructions. Currently they map to the same instruction implementation.
In particular: FPREM always truncates the quotient towards zero while FPREM1 rounds towards nearest.
|