Re: [Vm-dev] Does Cog Interpreter state and cycle differ from the    standard VM ?

gettimothy gettimothy at
Tue Nov 26 22:55:59 UTC 2013

That's very good, thanks.

I will give it a try.



---- On Tue, 26 Nov 2013 14:30:53 -0800 Nicolas Cellier<nicolas.cellier.aka.nice at> wrote ---- 

 Of course, a nickname is something personal, but I suggest tty, if you have affinity with bare os and low level stuff, it's a good match ;)

 2013/11/26 gettimothy <gettimothy at>
Thank you Eliot.

'Spaghetti Stack' I love it...

I want to get the standard VM under my belt first--walk before run and all that.
t, tim, or timothy is fine. 't' is easier to type and will not conflict with Tim Rowledge



---- On Tue, 26 Nov 2013 13:26:33 -0800 Eliot Miranda<eliot.miranda at> wrote ---- 
 Hi Tim(othy?).

On Tue, Nov 26, 2013 at 1:22 PM, gettimothy <gettimothy at> wrote:
 Does Cog Interpreter state and cycle differ from the standard VM ?

Yes.  Quite a lot.  The best overview is  Rationale for closures is in the posts named "Closure Overview" & "Closure Bytecodes".  HTH.

Details are not needed, just a heads up so that when I do get to Cog work, I make it a point to study them.

>From the Blue Book, here is the definition of Interpreter State and Cycle:

       Interpreter State:
      1. The CompiledMethod whose bytecodes ar being executed.
     2. The location of the next bytecode to be executed in that CompiledMethod: i.e its /instruction pointer/
     3. The receiver and arguments of the message that invoked the CompiledMethod.
      4. Any temporary variables needed by the CompiledMethod.
     5. a stack.

       Interpreter Cycle:
     1. fetch the bytecode from the CompiledMethod indicated by the instruction pointer.
     2. Increment the instruction pointer.
      3. Perform the function specified by the bytecode.







-------------- next part --------------
An HTML attachment was scrubbed...

More information about the Vm-dev mailing list