Mailing list for all users of the OCaml language and system.
 help / color / mirror / Atom feed
From: Christophe Raffalli <christophe.raffalli@univ-savoie.fr>
To: Caml List <caml-list@yquem.inria.fr>,
	Pierre Hyvernat <Pierre.Hyvernat@univ-savoie.fr>,
	Guillaume Theyssier <guillaume.theyssier@univ-savoie.fr>,
	Tom Hirschowitz <tom.hirschowitz@univ-savoie.fr>,
	Florian Hatat <florian.hatat@univ-savoie.fr>
Subject: Automated GC tweaking for OCaml
Date: Wed, 24 Nov 2010 18:02:58 +0100	[thread overview]
Message-ID: <4CED4542.7050805@univ-savoie.fr> (raw)

[-- Attachment #1: Type: text/plain, Size: 1684 bytes --]


Hello,

Here is a temptative of automated GC tweaking.

It mainly tries to guess a minor_heap_size that will both increase speed and reactivity !!!
It has a description in it ! It also move to a proportional increment of major_heap instead
of a constant one, but this is not very usefull in my tests.

To use it, adjust the parameter at the beginning of the file and just link with it !

The main questions are

- more testing : I am interested on the behavior on your programs especially
  * interactive ones !
  * seems not to do oscillation very much ... do you have a program where
    minor heap size oscillates
  * does it reaches the maximum minor_heap_size for you (with not a too low
    reactivity parameter)

- measuring the value of the "gamma" constant uses in the program on real OCaml program:

let gamma = 3.0 (* time in major slice attached to a minor GC  / time for minor GC
                : should use a real estimation, here just a guess !!! *)

Is this a universal constant that does not depend too much on the program and architecture ?
Probably not, because of the grey_val for instance ...

Cheers,
Christophe

-- 
Christophe Raffalli
Universite de Savoie
Batiment Le Chablais, bureau 21
73376 Le Bourget-du-Lac Cedex

tel: (33) 4 79 75 81 03
fax: (33) 4 79 75 87 42
mail: Christophe.Raffalli@univ-savoie.fr
www: http://www.lama.univ-savoie.fr/~RAFFALLI
---------------------------------------------
IMPORTANT: this mail is signed using PGP/MIME
At least Enigmail/Mozilla, mutt or evolution
can check this signature. The public key is
stored on www.keyserver.net
---------------------------------------------


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 262 bytes --]

             reply	other threads:[~2010-11-24 17:03 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-11-24 17:02 Christophe Raffalli [this message]
2010-11-24 17:05 ` Christophe Raffalli

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=4CED4542.7050805@univ-savoie.fr \
    --to=christophe.raffalli@univ-savoie.fr \
    --cc=Pierre.Hyvernat@univ-savoie.fr \
    --cc=caml-list@yquem.inria.fr \
    --cc=florian.hatat@univ-savoie.fr \
    --cc=guillaume.theyssier@univ-savoie.fr \
    --cc=tom.hirschowitz@univ-savoie.fr \
    /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