These benchmarks take the output of AoT and manually tweaks it to explore possible future changes to the compiler to produce this output directly.
These benchmarks take the output of AoT and manually tweaks it to explore possible future changes to the compiler to produce this output directly.