"Maciej W. Rozycki" <ma...@linux-mips.org> writes: > In that case I think the HONOR_NANS checks will best be globally removed > first (where applicable of course), with a separate preparatory change.
With a comment though :-) I.e. say that although NEG is the IEEE negate operation, we don't need to honour NaNs in the unfused combiner patterns because the rtx operations with which they're being combined don't preserve the signs of NaNs. Thanks, Richard