Date: Sun, 9 Jun 2002 11:04:45 -0700 (PDT) From: Matthew Dillon <dillon@apollo.backplane.com> To: Martin Blapp <mb@imp.ch> Cc: Nuno Teixeira <nunotex@pt-quorum.com>, Jens Rehsack <rehsack@liwing.de>, <freebsd-stable@FreeBSD.ORG> Subject: Re: openoffice stack and datasize kernel problems Message-ID: <200206091804.g59I4jo0087411@apollo.backplane.com> References: <20020609195005.T11797-100000@levais.imp.ch>
next in thread | previous in thread | raw e-mail | index | archive | help
:
:
:Hi Matt,
:
:> I'm still playing with this. Actually, I'm still playing with this
:> for the first time! I started building the port on my workstation
:> yesterday afternoon and, well, it's *still* building 24 hours later!
:> Holy cow!
:
:Hehe.
:
:It takes 6 hours here with a fast Raid here, 512M Ram and
:a 2Ghz P IV :-)
:
:Martin
:
:PS: What bulding system you have ?
I wanted to run it on my workstation so I made the mistake of building
it on my workstation, A 650MHz Celeron with 256M of memory. It's
entirely cpu bound. Someone pass the dunce hat... I should have
moved the build to a real box the moment I noticed that the distribution
file it was downloading was 116MB :-)
I *think* it's almost gotten to the point where the link issues
crop up. I already have some approaches in mind:
* Ld has a no-keep-memory option which may reduce VM use. I
am going to try that first.
* Reordering object modules and -l libraries may reduce
stacking issues.
And if all else fails:
* Pre-linking (ld -r) objects in smaller sets then doing a
final link will greatly reduce the final link stage footprint
by pre-resolving static symbol targets.
But first I have to get the build to the point where ld fails.
-Matt
Matthew Dillon
<dillon@backplane.com>
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200206091804.g59I4jo0087411>