> I might as well ask - are these memory problems supposed to be in
> 2.1.42? I've been badblocking disks and running make -kj 16 a
> kernel at the same time, and I've not quite seen what problems
> everyone has been talking about. I have a buffer patch, and the one
> line page_aging patch, but even before that, I never noticed any
> problems. mem=40M, I'll go try it with mem=4M eventually..
This seems to be a very complex problem. Some people apparently can
run the latest development kernels without incident, and some can't. I
can't. Every time I turn around I have memory problems. But they're
not consistent. Sometimes I can run 2.1.43 or 44-pre3 for two or three
hours without a crash. But yesterday, every time I got to the end of
the boot sequence, the kernel would oops and the machine would either
hang with the screen flashing many colors or reboot. Finally I did get
one line of intelligible message: "Unable to handle kernel paging
request at virtual address 18c00970." I looked over every program that
executes during boot, made sure it was the correct version and intact,
and nothing helped. Finally, in desperation, last night I compiled a
2.0.30-pre2 with several patches on another machine and copied it over
here. The machine's run all night without incident and into this
morning.
I really have no idea where else to look for a solution to the
problem.
-- Michael D. Harnois, Redeemer Lutheran Church, Washburn, IA mharnois@sbt.net aa0bt@aa0bt.ampr.org "Few sinners are saved after the first 20 minutes of a sermon." --Mark Twain