Subject: Re: mroe about cc compile errores.
To: Brian Chase <bdc@world.std.com>
From: rj45 <rj45@lart.xmission.com>
List: port-vax
Date: 07/24/2000 01:05:41
I Actually did
limit datasize 64000
the default datasize is 16MB
but seems like ther eis an hard conded kernel limit because the data size
limit cannot exceed 24MB.
Should I Recompile the kernel ??
ther eis like in FreeBSD some configuration file at login process which
can set the resource limit for users ?
in FreeBSD it is a /etc/login.conf file.
THANK
rj45
On Sat, 22 Jul 2000, Brian Chase wrote:
> On Sat, 22 Jul 2000, rj45 wrote:
>
> > I got a cc1,core of 16MB after the cc error.
> > Perhaps the system cannot handle such a big executable creation ?
> >
> > I have 32MB of system memory and 100M of swap space.
>
> You might need to tweak the runtime limit values in order to compile
> certain files which cause large processes. The `limit' command will
> display the current values. See the csh man page for info on the usage of
> the limit command.
>
> I don't know what the current maximum process size is for the NetBSD/vax
> port, but I recall it be relatively small due to the way the page tables
> are setup... I'm thinking it was somewhere around 24-32Megs, but I'd have
> to look at the source to be sure.
>
> Another trick that's handy for dealing with compile processes that grow
> too large is to compile the offending file without optimization. So you
> get some non-optimal object code, but once that file is done you re-run
> the make and it'll move on to finish the compile (unless it hits another
> big file).
>
> Of course you said it died with a signal 4. That's the illegal
> instruction execution trap. That could very well be a bug in the
> compiler, or at least in its object code.
>
> -brian.
> --- Brian Chase | bdc@world.std.com | http://world.std.com/~bdc/ -----
> But enough talk about serious science.
> Let's talk about ELASTIC BACON! -- K.
>
>