Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!usc!jarthur!jarthur.claremont.edu
From: ksmith@jarthur.claremont.edu (Karl J. Smith)
Newsgroups: comp.sys.handhelds
Subject: The Continuing Adventures of Maxwell Smart (stack oddities)
Message-ID: <11874@jarthur.Claremont.EDU>
Date: 24 Apr 91 02:12:32 GMT
Sender: ksmith@jarthur.Claremont.EDU
Organization: Harvey Mudd College, Claremont, CA 91711
Lines: 59

The Continuing Adventures of Maxwell Smart...

A while back I posted an anecdote about discovering that my 48 was
running at roughly half speed, and that an ON-A-F reset fixed it. I was
using this program to test it:

%%HP: T(3)A(R)F(.);
@ SMART, so named because it should produce
@ 86 (or something close) on the stack:
\\<<
  MEM DROP              @ do garbage collection
  TICKS \-> $$$ \<<     @ save current tick count
    @ no EVAL here this time
    TICKS               @ get tick count when done
    $$$ - B\->R         @ compute ticks elapsed
    @ with no EVAL, the actual elapsed
    @ time (86) is just the time it takes to assign $$$ and get here
    @ if we do nothing at this point, we'll have 86 on the stack
  \>>
\>>

Well, once I was running at normal speed again, I started to poke around
with the stack, to see what I could do to it.

I typed this in on the stack, and EVAL'd it:

\<< WHILE 1 REPEAT SMART END \>>		Program #2

I stopped it with ATTN, and ended up with a bunch of numbers on the stack
increasing from 86 to 88, with larger values the longer I let it run.
This was expected. What was not expected was that *sometimes* I also got
an extra binary integer on the stack, that shouldn't be there, such as
this:

# 1D431242660F2h or #1D431244507D0h or #1D431244EC4B7h

Where is it coming from? It's different every time. Maybe it's the
address of the top of the stack? The three #'s above are from three
consecutive runs of program #2, with the stack growing up to a 'time' of
102.

The other thing I discovered was that if you put enough stuff on the
stack, by letting program #2 run for a while, the time it takes SMART
to run alternates between two values, like this:

658
640
658
640
658
641

etc...

(This is with a stack depth of 2844. Last stack, arg, and command are on,
but you get similar results with all three turned off, too.)


-Karl Smith                           ksmith@jarthur.claremont.edu