Don! Thanks for the Interaction.

Bill Maltby LFS Related lfsbill at wlmlx1.wlmcs.com
Mon Mar 11 19:03:07 PST 2002


Don,

Wow! You are so prescient! Gcc on my 32MB just crapped again in
the semi-auto script. SO, I bit the bullit and run each command
by hand. Guess whatt it did?

It worked! So I guess that means big swap or no, ther's something
about the AMD K5, 32MB, RH 6.0 that it doesn't like.

I guess I'll have to destroy my nice little for-looped, file driven
batch and break it into 3 parts. For-loop-part1, file-part-1 driven,
inlined-do-the-gcc-steps and for-loop-part2, file-part-2 driven.

That sucks. I don't know what it is about that xgcc and bootstrap
stuff that is so special it can't work like the other 95% of the
stuff on this machine.

Oh well. This probably means it'll be even worse when I try and do the
100MHz AMD 5x86-WB (bogomips reported is 66.56). I wonder if that might
be the time to try some of the cross-compiling hints?

I think maybe-positively yes.

On Mon, 11 Mar 2002, Don Smith wrote:

> "Bill Maltby LFS Related" <lfsbill at wlmlx1.wlmcs.com> wrote in message
> news:Pine.LNX.4.10.10203111642530.1435-100000 at wlmlx1.wlmcs.com...
> > Don,
> >
> > Your commnets about swap caused me to say hmmm...
> >
> > You saw that I started a run on my itsy-bitsy. Well,
> > glibc finished in about 1/4 - 1/3 the time it had been
> ><snip>

> You are more than welcome.
> 
> I have lots of experience with small, underpowered machines :-). The
> fastest one I have is a P 120, but it is from Compaq so it is not very
> upgradable (too much proprietary crap) - memory is very limited, so I

Yeah, when I was doing my own business, I used to tell my customer to
avoid them and several others for strictly that reason. Good eqpt, but
when you wanted to upgrade or needed repair, it was just way to expensive.

> haven't bothered put LFS on it yet. Been doing LFS on a 486/66, now with
> 48MB memory (found a 32MB simm behind a 4MB simm).

Sounds a lot like the last machine I will be doing until I get my
new one (whenever that will be). Got to get another job first I think.
Maybe you have some tips that will save me some hair-loss when the time
comes.

> 
> Anyway, reflecting on it all, I wouldn't be surprise if gcc has a bug
> handling the out-of-memory interupt (signal) thrown when doing a new. It
> must be one of those situations that work almost all of the time but
> every-once-in-a-while it returns a non-null, but invalid, pointer when
> the moon lines up with Aldebaran.

Yep, sounds just like my symptoms. Well, I gotta go figure this work-
around for this gcc so I can stay semi-automated. I guess all hope of
"elegance" is in the crapper now.

><snip>
 
Bill Maltby
billm at wlmcs.com

-- 
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe lfs-support' in the subject header of the message



More information about the lfs-support mailing list