[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Axiom-developer] Re: Axiom memory usage and GCL
From: |
Gabriel Dos Reis |
Subject: |
[Axiom-developer] Re: Axiom memory usage and GCL |
Date: |
27 Sep 2006 01:10:24 +0200 |
"Page, Bill" <address@hidden> writes:
| Gaby,
|
| On Tuesday, September 26, 2006 11:33 AM you wrote:
| >
| > Bill Page writes:
| > | ...
| > | Starting with Camm's newest version of 2.6.8pre it is
| > | possible to set maxpage a non-power of 2. I am not using
| > | --enable-maxpage=196*1024 on MathAction and this seems to
| > | be working so far.
| >
|
| What I meant to write was: I am *now* using
|
| --enable-maxpage=196*1024
OK.
| > Many thanks.
| >
| > The reason I was asking is that I have a successful build of
| > the axiom.build-improvements branch on the two machines I have,
| > but when I tried on the SF Deian machine x86-linux1 I got a
| > failure
|
| How do you access the SF x86-linux1 machine? I used to have
| access but the method I used (via cf.sourceforge.net) stopped
| working a few weeks ago and I haven't been able to access any
| of the compile farm machines since that time.
I first log onto cf-shell.sf.net; then ssh x86-linux1 from there.
[...]
| I think the effects of low memory can be unpredictable. You
| should definitely try using the newest gcl-2.6.8pre and
| changing CGLOPTS to specify a smaller default memory such as
|
| --enable-maxpage=196*1024
the build-improvements branch uses the newest gcl-2.6.8pre.
I've been doing a binary search (which gives me a chance to appreciate
changetset commits); I may be near the commit that caused the problem,
but I really wish Axiom were faster to compile. I hope to be at a
point where I can start parallel make, but I first need to get this
issue out of the way.
-- Gaby