gcl-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Gcl-devel] Re: sockets, continued


From: Camm Maguire
Subject: [Gcl-devel] Re: sockets, continued
Date: 29 Oct 2005 22:22:25 -0400
User-agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2

Greetings!

Robert Boyer <address@hidden> writes:

> > In principle you are right, but in practice things are far safer.
> 
> Hmmmm.
> 
> > most typically this is done at the kernel level by the system
> > administrator.
> 
> It is really cool that GCL can support services to the Internet via sockets
> and that system administrators can restrict access to such sockets.
> 
> However, I do doubt that one wants to use something as public as sockets at
> all for ordinary interprocess communication in GCL in those cases in which
> all one wants is the power of parallelism.  What about all the thousands of
> other users here at UTCS who may be running jobs on the machine I am running
> on?  I don't want to begin to think about giving other users access to my
> processes via a socket merely to fork off a few processes to take advantages
> of the possible presence of multiple-cores.  GCL needs a very private method
> for forks to communicate, besides sockets, I suspect.  It probably has one
> that I just don't know about.  Which is not saying much.
> 

Indeed -- totally agree.  This is what I meant in my earlier post
about preconnected pipe streams.

Here is what I have locally now -- hope to commit soon:

=============================================================================
(defun fib (n) (if (<= n 2) 1 (+ (fib (1- n)) (fib (- n 2)))))

FIB

>(compile 'fib)

;; Compiling ./gazonk1.lsp.
;; End of Pass 1.  
;; End of Pass 2.  
;; OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0, Speed=3, 
(Debug quality ignored)
;; Finished compiling ./gazonk1.o.
Loading /fix/t1/camm/debian/gcl/tmp/tmp/foo/unixport/gazonk1.o
start address -T 0x13e8568 Finished loading 
/fix/t1/camm/debian/gcl/tmp/tmp/foo/unixport/gazonk1.o
#<compiled-function FIB>
NIL
NIL

>(time (values (fib 31) (fib 31)))

real time       :     14.050 secs
run-gbc time    :      5.820 secs
child run time  :      0.000 secs
gbc time        :      0.000 secs
1346269
1346269

>(time (let ((x (si::fork))) (if (= (car x) 0) (progn (print (fib 31) (cadr 
>x))(bye)) (values (fib 31) (read (cadr x))))))

real time       :      8.170 secs
run-gbc time    :      2.980 secs
child run time  :      0.000 secs
gbc time        :      0.000 secs
1346269
1346269

>
=============================================================================

Take care,


> Bob
> 
> 
> 
> 
> 
> 

-- 
Camm Maguire                                            address@hidden
==========================================================================
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah




reply via email to

[Prev in Thread] Current Thread [Next in Thread]