Skip to Content.
Sympa Menu

coq-club - Re: [Coq-Club] Coq memory usage

coq-club AT inria.fr

Subject: The Coq mailing list

List archive

Re: [Coq-Club] Coq memory usage


chronological Thread 
  • From: Julien Narboux <Julien.Narboux AT inria.fr>
  • To: "Nickolay V. Shmyrev" <nshmyrev AT yandex.ru>
  • Cc: Coq Club <coq-club AT pauillac.inria.fr>
  • Subject: Re: [Coq-Club] Coq memory usage
  • Date: Sun, 06 Nov 2005 11:54:26 +0100
  • List-archive: <http://pauillac.inria.fr/pipermail/coq-club/>

Hello,

Ocaml does not allow dynamic loading of modules (in native code).
But Fabrice Le Fessant has implemented "Asmdynlink: a Virtual Machine for Ocaml bytecode written in Ocaml, to dynlink bytecode modules in native code programs (distributed in the CDK)".

I don't know if it is still maintained.

Regards

Julien Narboux

Nickolay V. Shmyrev a écrit :
Hello.

While working with CoRN repository I've noticed that Coq very
extensively use memory. It requires around 40 Mb for rather simple
proofs and when advanced tactics like "rational" are used, it may take
more then 200 Mb and even fail to compile correct proof. Why is it so
big and is there any way to optimize memory usage?

In CoRN, there is also ml-based implementation of rational tactic, it
much more quick, but it requires top level building. Is it planned to
implement the dynamic loading of ml modules? If so, how dynamic loading
can be implemented?


--------------------------------------------------------
Bug reports: http://coq.inria.fr/bin/coq-bugs
Archives: http://pauillac.inria.fr/pipermail/coq-club
          http://pauillac.inria.fr/bin/wilma/coq-club
Info: http://pauillac.inria.fr/mailman/listinfo/coq-club




Archive powered by MhonArc 2.6.16.

Top of Page