Shop OBEX P1 Docs P2 Docs Learn Events
PR0 to PR7 are termed Communication Registers .Registers $1E0 to $1EF appear to be in limbo? — Parallax Forums

PR0 to PR7 are termed Communication Registers .Registers $1E0 to $1EF appear to be in limbo?

Bob DruryBob Drury Posts: 236
in Propeller 2

Below is a list of named registers from the documentation. For some reason $1E0 to $1EF is not named . PR0-PR7 are called communication registers in the documentation. Is $1E0-$1EF reserved for a special function yet to be specified?

PR0 $1D8 IJMP3 $1F0 PTRA $1F8
PR1 $1D9 IRET3 $1F1 PTRB $1F9
PR2 $1DA IJMP2 $1F2 DIRA $1FA
PR3 $1DB IRET2 $1F3 DIRB $1FB
PR4 $1DC IJMP1 $1F4 OUTA $1FC
PR5 $1DD IRET1 $1F5 OUTB $1FD
PR6 $1DE PA $1F6 INA $1FE
PR7 $1DF PB $1F7 INB $1FF

Regards
Bob (WRD)

Comments

  • RaymanRayman Posts: 14,641

    What documentation did you get that PR0..PR7 from?
    I don't see that in the silicon doc...

    Maybe that's something old...

  • Bob DruryBob Drury Posts: 236

    Here is where I got info. PR0-PR7 can be referenced directly (eg, MOV PR0,#1)
    https://docs.google.com/document/d/16qVkmA6Co5fUNKJHF6pBfGfDupuRwDtf-wyieh_fbqw/edit#heading=h.lcty4rb2rm1s
    In document Under "Variables" is a chart.
    Regards
    Bob (WRD)

  • RaymanRayman Posts: 14,641

    Interesting... Spin2 has added some things lately...

    I do see this there: $1E0..$1EF, which are available for scratchpad use, but will likely be rewritten when Spin2 resumes.

    Also, I see this:
    Avoid writing to $130..$1D7 and LUT RAM, since the Spin2 interpreter occupies these areas. You can look in "Spin2_interpreter.spin2" to see the interpreter code.

    Where I think we decided that $130 is wrong by looking at that .spin2 file.

  • ersmithersmith Posts: 6,053

    I believe $1E0-$1EF are used internally by the Spin interpreter.

  • Cluso99Cluso99 Posts: 18,069

    @Rayman said:
    Interesting... Spin2 has added some things lately...

    I do see this there: $1E0..$1EF, which are available for scratchpad use, but will likely be rewritten when Spin2 resumes.

    Also, I see this:
    Avoid writing to $130..$1D7 and LUT RAM, since the Spin2 interpreter occupies these areas. You can look in "Spin2_interpreter.spin2" to see the interpreter code.

    Where I think we decided that $130 is wrong by looking at that .spin2 file.

    @ersmith said:
    I believe $1E0-$1EF are used internally by the Spin interpreter.

    I used $1E0-$1EF and $1D0-$1DF for the monitor and SD in the ROM but Chip used that in spin2 which broke the ROM toolbox so I know for sure these are used by spin2 :(

  • Bob DruryBob Drury Posts: 236

    It looks like $1E0-$1Ef should be listed as not usable when spin is used/
    Thanks for Info
    Bob (WRD)

  • cgraceycgracey Posts: 14,152

    Actually, the Spin2 interpreter starts at $124 now, not $130.

    Some of you noted this in the Spin2 doc and I changed these references today.

  • RaymanRayman Posts: 14,641

    @"Bob Drury" said:
    It looks like $1E0-$1Ef should be listed as not usable when spin is used/
    Thanks for Info
    Bob (WRD)

    I think you can use them when doing pasm2 with Spin2. But, they will be used by Spin2 when not doing pasm2.
    So, they are for temporary use only…

  • Bob DruryBob Drury Posts: 236

    cgracey
    Does this mean that the spin interperter fits between $124 to $1D7 for 172 bytes ?
    Pretty amazing if so.
    Regards
    Bob (WRD)

  • evanhevanh Posts: 15,915

    cogRAM addresses are in longwords, not bytes. My understanding is the core of the engine sits in lutRAM, but hubRAM will be used too. Especially for debug() feature.

  • RaymanRayman Posts: 14,641

    Chip posted this about HUB RAM usage:
    The Spin2 interpreter is now 4,784 bytes long, occupying $00000..$012AF

    So, I think it's mostly in the hub.

  • Bob DruryBob Drury Posts: 236

    When I said 172 bytes I was thinking longs oops. If spin interperter actually exist in HUB ram then the 172 cog bytes are needed to step through the hub spin code. I guess if spin is running in every cog
    the same 172 longs would exist in every cog and hub spin code would probably only exist once with a bunch of stack space in hub ram. So spin is 4k bytes hub space and 172x4 bytes cog space.
    I am in over my head and should shut up.
    Regards
    Bob (WRD)

  • RaymanRayman Posts: 14,641

    There might be some in LUT too

  • cgraceycgracey Posts: 14,152

    @"Bob Drury" said:
    When I said 172 bytes I was thinking longs oops. If spin interperter actually exist in HUB ram then the 172 cog bytes are needed to step through the hub spin code. I guess if spin is running in every cog
    the same 172 longs would exist in every cog and hub spin code would probably only exist once with a bunch of stack space in hub ram. So spin is 4k bytes hub space and 172x4 bytes cog space.
    I am in over my head and should shut up.
    Regards
    Bob (WRD)

    That's right.

    Spin2 uses $124..$1D7 in cog RAM, plus LUT $210..$3FF, then the first 4KB+ of hub 4AM.

Sign In or Register to comment.