01:55fdobridge: <karolherbst🐧🦀> would be interesting to use cbindgen at some point as well
19:47fdobridge: <gfxstrand> I'm REALLY tempted to rewrite at least the turing sim in rust.
19:48fdobridge: <marysaka> then I guess I should rewrite it in Rust later :AkkoDerp:
19:48fdobridge: <gfxstrand> Rewriting the builder would be more work because C calling into rust is a bit of a pain.
19:48fdobridge: <marysaka> I think I found multiple inconsistency on it btw but need to confirm those
20:17fdobridge: <marysaka> I think I should start the tester next considering it's pretty much done.. the only missing bits are:
20:17fdobridge: <marysaka> - Implement a tester for the mme_fermi_builder
20:17fdobridge: <marysaka> - Decide if we want to emulate mme_loop
20:17fdobridge: <marysaka> - Decide if we want to emulate missing branch instruction (it only support BEQ and BNE)
20:17fdobridge: <marysaka> - Figure out UNK6 operation (probably FIFO related)
23:17fdobridge: <gfxstrand> Nice!
23:19fdobridge: <gfxstrand> As for `mme_loop` and branches... Neither are really necessary AFAIK. We can emulate them if needed but I'm happy for you to not bother for now. I think all loops we use today have exact bounds so we can probably replace most instances of `lt` with `ne`.