Hello, folks. There is some optimization activity in MIPS target, and we have some simple micro-benchmarks for measuring speed of emulation of a single target instruction, but I am not particulary happy with them. They are just a single instruction assembly within a “for“ loop. I tried some other approaches, but couldnt get results that made sense.
Do you have some suggestion or idea what would be a good way for such measurements? Grateful, Aleksandar
