Hi! On Mon, Jul 06, 2020 at 10:48:25AM +0100, Richard Sandiford wrote: > Originally combine always produced shorter sequences, so by the
(Shorter in # insns, not # bytes). > above reasoning, combining for == was correct. These days we allow > N-to-N replacements too, which are obviously a good thing when Only 2-2 so far (not 1-1 yet, there are some target problems to overcome first). > the new cost is lower, but are more of a wash when the costs > are the same. But even then, the combination should have a > “canonicalisation” effect. (Unfortunately that effect includes > the result of expand_compound_operation/make_compound_operation.) 2-2 is always reducing latency if the costs are equal (and sane ;-) ), that is a large part of what makes 2-2 combinations useful. Originally the output of i2 is input to i3, but not anymore in the new insns. Segher
