----- Forwarded Message -----
From: Hugh Aguilar <hughaguila...@yahoo.com>
To: Stephen Bloch <bl...@adelphi.edu>
Sent: Wednesday, September 26, 2012 8:07 PM
Subject: Re: [racket] assembly language
You are right that a threaded Forth doesn't need to do any assembly --- it just
"compiles" pointers to functions (code field addresses) in a thread. This can
and has been done with traditional assemblers such as HLA. I'm not interested
in that though.
I want to generate machine-code functions. This involves assembling at run-time
(for the forth system, which is compile-time for the Forth application program).
I don't really understand JIT. I suppose the idea is to save memory by
compiling the program as threaded code (which is much smaller than
machine-code), but to convert this into machine-code "just in time" for it to
run. I suppose this involves pasting together the code from all of those
sub-functions in the threaded stream to create a sequence of machine-code. It
seems like this conversion would take so much time that it would outweigh any
speed advantage of running machine-code rather than threaded code. It must work
somehow though, as it seems to be done quite a lot (the JVM for example).
I'm not planning on JIT (especially as I don't understand it). I just want the
compiler to assemble actual text source-code rather than paste together code
snippets, which I am finding to be a hassle. This will be done at compile-time
for the Forth application --- at run-time the Forth application just runs.
________________________________
From: Stephen Bloch <bl...@adelphi.edu>
To: Hugh Aguilar <hughaguila...@yahoo.com>
Cc: Tony Garnock-Jones <to...@ccs.neu.edu>; "users@racket-lang.org"
<users@racket-lang.org>
Sent: Tuesday, September 25, 2012 5:12 PM
Subject: Re: [racket] assembly language
On Sep 25, 2012, at 7:38 PM, Hugh Aguilar <hughaguila...@yahoo.com> wrote:
I am writing a Forth system. I want it to be interactive in the usual Forth
way. This means assembling a function at run-time and immediately being able to
run the assembled function.
>
>... I'm much better off if I can just assemble the functions at run-time for
>the Forth system (which is compile-time for the user's Forth program).
The traditional way to implement a Forth compiler, IIUC, isn't to generate
executable machine code at all, but rather to generate a sequence of
word-references that are interpreted as procedure calls by the Forth
interpreter (which is in native executable code, but written in advance).
Or are you talking about some kind of JIT compiler?
Stephen Bloch
sbl...@adelphi.edu
who last implemented a Forth system in 1983; I presume things have changed
since then!
____________________
Racket Users list:
http://lists.racket-lang.org/users
