Newsgroups: comp.lang.lisp,comp.os.linux.misc
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!europa.eng.gtefsd.com!howland.reston.ans.net!pipex!dircon!rheged!simon
From: simon@rheged.dircon.co.uk (Simon Brooke)
Subject: Port of CMU Common LISP to Linux (was: Re: Squeezing more speed out of LISP.)
Message-ID: <Cy4B02.4ME@rheged.dircon.co.uk>
Followup-To: sender
Summary: Canvassing for interest in a colaborative porting effort.
Keywords: linux LINUX lisp LISP port
Organization: none. Disorganization: total.
References: <HJSTEIN.94Oct16172905@sunset.huji.ac.il> <HJSTEIN.94Oct18135742@sunset.huji.ac.il> <Cy0GnM.2AG@rheged.dircon.co.uk> <38a3d0$g9j@cantaloupe.srv.cs.cmu.edu>
Date: Sun, 23 Oct 1994 08:45:37 GMT
Lines: 84
Xref: glinda.oz.cs.cmu.edu comp.lang.lisp:15217 comp.os.linux.misc:30635

NOTE: This article is cross posted to comp.lang.lisp and
comp.os.linux.misc, but follows discussion which has taken place in
comp.lang.lisp only. Please follow-up by mail to me, and I'll repost
-- I don't know what the volume of interest will be, and don't want to
swamp the newsgroups.

SUMMARY: In comp.lang.lisp we've been discussing the efficiency of
Common LISP implementations, and there appears to be a strong feeling
that Carnegie Mellon University's CMU CL is the fastest non-commercial
version presently available. However, it is available only for some
RISC based architectures. I wondered how many people would be
interested in a port to Linux, and what would be involved.

ACTION: If you would be seriously interested in using CMU CL on Linux,
please mail me with the subject line 'CMU CL: Interest'

If you believe you have skills relevent to the porting effort, and
would like to be involved, please mail me with the subject line 
'CMU CL: Volunteer', and detail in the message what skills you can
offer, and how much time you could make available.

I will collate responses, and if I feel that there is enough interest,
will set up a mailing list. I shall in any case repost a summarry of
the responses I receive.

BACKGROUND:

In article <38a3d0$g9j@cantaloupe.srv.cs.cmu.edu>,
Scott Fahlman <sef@CS.CMU.EDU> wrote:
>
>In article <Cy0GnM.2AG@rheged.dircon.co.uk> simon@rheged.dircon.co.uk
>(Simon Brooke) writes:
>
>   Hr'hhm. I suspect that, in the near future at least, a lot of serious
>   computer science students (and many professionals at home) will
>   '...only have access to 486s running Linux...'. Linux is a very good
>   system, and cheap :-} The feeling that CMU is the best non-commercial
>   CL around seems widespread, too.
>
>   Has anybody enough knowledge of the CMU source to have a feel of how
>   difficult this would be to port to Linux? Is this something that,
>   perhaps, a group of us could get together on over the net? It seems to
>   me that the goal might well be worth achieveing. If enough people mail
>   me on this I shall work out what it takes to set up a mailing list.
>
>Porting to Linux is not the whole story.  CMU CL is a native-code
>compiler, so you have to port to a specific instruction set as well as
>to a specific OS.  You don't just change the system calls and
>recompile the C.
>
>If porting to the 486 under Linux is the goal, you first have to port
>to the 486 instructions set.  Since this is a CISC instruction set
>with fewer registers than CMU CL uses on the typical RISC machine,
>there are some tricky aspects to the port.  We did some bits of the
>design at one point, when we were thinking about porting to 486/Mach,
>but I'm not sure if this is far enough along to be useful.
>
>I have no idea how much Linux differs from Mach or SunOS in things
>like signals and the mapping of disk files into particular addresses
>in virtual memory.
>
>As a guess, figure a person-year for someone really good who already
>knows his way around Linux.  We would like to see such a port, and
>would offer advice and encouragement, but my people can spare very
>little time for this.
>
>I wonder if a Linux port would be more useful to the world than a
>Windows NT port.  Linux probably would be a good deal easier.
>
>-- Scott
>
>===========================================================================
>Scott E. Fahlman			Internet:  sef+@cs.cmu.edu
>Principal Research Scientist		Phone:     412 268-2575
>School of Computer Science              Fax:       412 268-5576 (new!)
>Carnegie Mellon University		Latitude:  40:26:46 N
>5000 Forbes Avenue			Longitude: 79:56:55 W
>Pittsburgh, PA 15213			Mood:      :-)
>===========================================================================
>


-- 
---------simon@rheged.dircon.co.uk
