Brian Wheeler writes: > This diff adds jsr_ic and ret to the interpreter. I don't know if my > way of returning is legal, and I know there's probably issues with 64 > bit machines, but it works...and that's the important part :)
Ok! We definitely need to get some tests into the Parrot tree. Who's going to step forward and do this? You can assume the presence of perl5 and use that to feed various programs to Parrot and test the successful compilation and output if you want. See the current t/ directory for an example. In particular, I really want us to have the ability to test and prove that each opcode behaves correctly (and the questions raised while you create a test that explores every possibility of an opcode will help to define the assembly language, I'm sure). Later on we'll want to have the ability to test and prove datatype behaviour too. The goal is to test everything! Opcode tests would let us easily discover whether there are problems on 64 bit machines, let us know whether it really works, etc. Nat