caml-list - the Caml user's mailing list
 help / color / mirror / Atom feed
From: Thomas Fischbacher <Thomas.Fischbacher@Physik.Uni-Muenchen.DE>
To: Xavier Leroy <Xavier.Leroy@inria.fr>
Cc: Michael Alexander Hamburg <hamburg@fas.harvard.edu>, caml-list@inria.fr
Subject: Re: [Caml-list] How to do this properly with OCaml?
Date: Sat, 23 Jul 2005 21:19:44 +0200 (CEST)	[thread overview]
Message-ID: <Pine.LNX.4.61.0507232100420.3727@katrin.cip.physik.uni-muenchen.de> (raw)
In-Reply-To: <42E2393B.5030209@inria.fr>


On Sat, 23 Jul 2005, Xavier Leroy wrote:

> For instance, the following implementation of "magic" arrays will
> eventually cause the GC to crash:
> 
> type 'a t = int array
> let get (a: 'a t) i = (Obj.magic a.(i) : 'a)
> let set (a: 'a t) i (x: 'a) = a.(i) <- (Obj.magic x : int)
> 
> while the same code with "string" instead of "int" will not.  You
> don't understand why?  Then, don't use Obj.magic.

Could this be for precisely the same reason why I used Obj.magic (1,2) 
where Michael suggested using Obj.magic 0? I suppose Obj.magic () 
also would not work, right?

> Coming back to the initial question, I would first warn against
> premature optimization:

I do not think this is premature optimization in that particular case. The 
code I am talking about is quite mature and performed very well for years 
in its Lisp incarnation - I am just right now transliterating it to 
OCaml, as I need it there.

Another low-level question: it seems to me as if it were a piece of cake 
to build a C-linkable shared object library that in all respects looks 
like any old C .so library, but internally uses code that was compiled 
from OCaml.

However, in order to get there, I had to ar x libcamlrun.a and 
ld a new libcamlbytecode.so shared object from the pieces. It may well be 
that I just missed something in the documentation due to my own 
stupidity/ignorance, but so far it seems to me as if this were 
tremendously useful, but somehow not documented very well.

Anyway, as I clearly do appreciate the option to build C shared objects in 
something better than C, I wonder whether I can rely on being able to do 
that trick with future versions of the OCaml compiler as well.

> quite possibly the overhead of the "option" solution is negligible.

For a lot of applications this is indeed right. However, for the 
particular one I have in mind right now, I am more or less competing
with C code, so I am indeed a bit worried about that extra indirection.

> If not, just ask the user to pass an initial
> value of the heap element type to the "create heap" function.

Well, that's also not that beautiful, but at least, it would work.

But coming back to:

> while the same code with "string" instead of "int" will not.

Can I take this as a guarantee that will also hold for later versions of 
the compiler? Just for my particular private application - not that I 
would advise or teach anyone to use such a technique, or distribute any 
code in compiled or source form that makes use of it.

-- 
regards,               tf@cip.physik.uni-muenchen.de              (o_
 Thomas Fischbacher -  http://www.cip.physik.uni-muenchen.de/~tf  //\
(lambda (n) ((lambda (p q r) (p p q r)) (lambda (g x y)           V_/_
(if (= x 0) y (g g (- x 1) (* x y)))) n 1))                  (Debian GNU)


  parent reply	other threads:[~2005-07-23 19:19 UTC|newest]

Thread overview: 105+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2005-07-22 14:26 Thomas Fischbacher
2005-07-22 14:52 ` [Caml-list] " Eric Cooper
2005-07-22 15:26 ` [SPAM_PROBABLE] - [Caml-list] How to do this properly with OCaml? - Bayesian Filter detected spam Christophe Dehlinger
2005-07-22 18:58   ` [Caml-list] How to do this properly with OCaml? Stephane Glondu
2005-07-22 15:50 ` Berke Durak
2005-07-22 16:47   ` brogoff
2005-07-22 15:54 ` Michel Quercia
2005-07-23  5:00 ` Michael Alexander Hamburg
2005-07-23 12:34   ` Xavier Leroy
2005-07-23 13:16     ` Berke Durak
2005-07-23 16:36       ` Stephane Glondu
2005-07-23 18:27         ` Berke Durak
2005-07-23 18:50           ` Matthieu Sozeau
2005-07-24  8:35             ` Berke Durak
2005-07-23 19:18           ` Stephane Glondu
2005-07-23 19:35             ` Thomas Fischbacher
2005-07-23 19:50               ` Stephane Glondu
2005-07-23 19:59                 ` Thomas Fischbacher
2005-07-23 20:15                   ` Brian Hurt
2005-07-23 23:16                     ` james woodyatt
2005-07-23 23:27                   ` Stephane Glondu
2005-07-23 18:37     ` Michael Alexander Hamburg
2005-07-23 18:52       ` Bardur Arantsson
2005-07-23 21:35         ` [Caml-list] " Michael Alexander Hamburg
2005-07-23 19:19     ` Thomas Fischbacher [this message]
2005-07-24  7:27     ` [Caml-list] "Just say no!" campaign against Obj [was: How to do this properly with OCaml?] Alex Baretta
2005-07-24  8:02       ` [Caml-list] "Just say no!" campaign against Obj james woodyatt
2005-07-24 17:27       ` Stephane Glondu
2005-07-25  8:43         ` Alex Baretta
2005-07-24 21:37       ` [Caml-list] "Just say no!" campaign against Obj [was: How to do this properly with OCaml?] brogoff
2005-07-25  8:15         ` Alex Baretta
2005-07-25 17:08           ` brogoff
2005-07-25  8:57         ` Alex Baretta
2005-07-24 17:33     ` [Caml-list] How to do this properly with OCaml? skaller
2005-07-24 18:13       ` Stephane Glondu
2005-07-24 18:48         ` skaller
2005-07-24 19:14           ` Stephane Glondu
2005-07-24 20:29             ` skaller
2005-07-24 20:49               ` skaller
2005-07-24 21:08                 ` Stephane Glondu
2005-07-24 21:55                   ` skaller
2005-07-24 23:23                     ` Stephane Glondu
2005-07-25  0:32                       ` skaller
2005-07-25  6:45                         ` Stephane Glondu
2005-07-25 11:35                           ` skaller
2005-07-26  0:47                             ` Brian Hurt
2005-07-26  0:56                               ` Jere Sanisalo
2005-07-26  1:10                                 ` Brian Hurt
2005-07-26  1:34                                   ` Jere Sanisalo
2005-07-26  9:03                                     ` Richard Jones
2005-07-27 17:21                                     ` skaller
2005-07-27 19:44                                       ` [Caml-list] Games Jon Harrop
2005-07-27 20:35                                         ` Jere Sanisalo
2005-07-28  0:13                                           ` Jon Harrop
2005-07-28  1:12                                             ` Jere Sanisalo
2005-07-28  2:44                                               ` Jon Harrop
2005-07-28  4:49                                                 ` skaller
2005-07-28 19:48                                                   ` Jon Harrop
2005-07-28 21:32                                                     ` David Thomas
2005-07-28 22:31                                                       ` Jon Harrop
2005-07-29  1:44                                                         ` Michael Walter
2005-07-29  2:32                                                         ` David Thomas
2005-07-29  3:52                                                           ` skaller
2005-07-29 12:57                                                             ` David Thomas
2005-07-28 10:58                                               ` Gerd Stolpmann
2005-07-28 17:19                                         ` David Thomas
2005-07-28 19:22                                           ` Jon Harrop
2005-07-27 21:13                                       ` [Caml-list] How to do this properly with OCaml? Pal-Kristian Engstad
2005-07-27 22:28                                         ` skaller
2005-07-28  1:47                                           ` Michael Walter
2005-07-27 23:17                                         ` [Caml-list] Games Jon Harrop
2005-07-28  0:03                                         ` [Caml-list] How to do this properly with OCaml? Paul Snively
2005-07-28 18:26                                           ` Jonathan Bryant
2005-07-28 23:10                                             ` Paul Snively
2005-07-27 16:03                                   ` skaller
2005-07-26  1:01                               ` Stephane Glondu
2005-07-26  1:15                                 ` Brian Hurt
2005-07-27 15:33                                 ` skaller
2005-07-30 23:24                                   ` Thomas Fischbacher
2005-07-31  0:06                                     ` Jon Harrop
2005-07-31  3:10                                       ` skaller
2005-07-31  2:54                                     ` skaller
2005-07-26 20:32                               ` David Thomas
2005-07-27 15:05                               ` skaller
2005-07-27 15:29                                 ` Jon Harrop
2005-07-27 15:35                                   ` David Thomas
2005-07-27 20:11                                     ` skaller
2005-07-28 16:35                                       ` David Thomas
2005-07-30 23:33                                     ` Thomas Fischbacher
2005-07-31  0:39                                       ` james woodyatt
2005-07-27 19:59                                   ` skaller
2005-07-26  1:22                             ` Jon Harrop
2005-07-27 17:23                               ` skaller
2005-07-26  1:05                         ` Jon Harrop
2005-07-26  1:20                           ` Brian Hurt
2005-07-26  1:28                             ` Jon Harrop
2005-07-27 17:03                             ` skaller
2005-07-27 16:09                           ` skaller
2005-07-24 23:26               ` Brian Hurt
2005-07-25 17:21       ` Ken Rose
2005-07-25 19:19         ` skaller
2005-07-26  7:10           ` Alex Baretta
2005-07-23 18:58   ` Thomas Fischbacher
2005-07-26  1:32 Jon Harrop
     [not found] <200507271635.28931.jon@ffconsultancy.com>
2005-07-27 16:31 ` David Thomas

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=Pine.LNX.4.61.0507232100420.3727@katrin.cip.physik.uni-muenchen.de \
    --to=thomas.fischbacher@physik.uni-muenchen.de \
    --cc=Xavier.Leroy@inria.fr \
    --cc=caml-list@inria.fr \
    --cc=hamburg@fas.harvard.edu \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).