Spin2

1121314151618»

Comments

  • RaymanRayman Posts: 10,435
    edited 2020-02-16 - 16:23:31
    Roy Eltham wrote: »
    Eric's compiler produces PASM to be run in hubexec. It's a whole different animal compared to how Chip's compiler works.

    Actually, I suppose Spin2 would have to produce PASM meant to run in hubexec in a dat section with assembly instructions following an "ORGH", right?

    Since that is something new, maybe there's room for a middle ground? Maybe this kind of code could resolve absolute addresses, same way in Spin2 and FastSpin?

  • @Rayman

    Chip's SPIN2 is interpreted bytecode, while Fastspin is compiled.
    Have you read Chip's code for the interpreter? It's not particularly long.
    The interpreter is split between cog and hub.

  • AJL, Spin2 can also launch assembly cogs using "coginit".
    Those cogs can presumable call hubexec assembly code....
  • Rayman wrote: »
    AJL, Spin2 can also launch assembly cogs using "coginit".
    Those cogs can presumable call hubexec assembly code....

    Perhaps I misunderstood your original comment, or I'm nit-picking.

    While PNut is the current tool to produce SPIN2 code, Chip's aim is for SPIN2 to be self-hosted; PNut produces PASM code for execution, SPIN2 currently does not.

    Given SPIN2 is interpreted, then any assembly cogs would have their code produced by some other tool, e.g. PNut, or a self-hosted PASM2 Assembler.


  • I am going to work tomorrow on the documentation. That will clear the air quite a bit.
Sign In or Register to comment.