kernel BUG at lib/kernel_lock.c:83! - 2.6.19-1.2895.fc6

[James Wilkinson]

Albert Graham wrote:

> I'm simply reporting this "kernel bug", it seems that whenever NVidia
> drivers are loaded and there's a problem everyone automatically
> "blames" Nvidia for it, and in my experience, Nvidia drivers are very
> very rarely the cause of such problems.

How do you know? Bugs in the Nvidia drivers won't necessarily manifest
in related areas. They could scribble anywhere in memory.

More to the point, how would anyone know? The only people in a position
to do so are Nvidia. They're the only people with all the source code
necessary to hunt down the bug and fix.

The only way to prove that it *isn't* a Nvidia bug is to reproduce it
with an untainted kernel.

> I've been using NVidia drivers
> since FC1 (and actually RH 7.1) with very little problems.

Great. You've presumably been using the Linux kernel for that long with
very few problems, too.

> Also, some of these bugs are really hard to reproduce or completely
> unpredictable, just look out some of my previous posts with and
> without tainted kernels.

> As for your useless "tainted" responses, I'm afraid falls on deaf
> ears, if you cannot see the "usefulness" of my posts, please don't
> respond to them.

Unfortunately, that works the other way around. I might as well say that
if you can't reproduce the problems with untainted kernels, then *your*
posts are useless -- so why make them?

But they're not useless -- they're just in the wrong place. Go to
Nvidia -- they're the only people who can help you.

James
-- 
E-mail:     james@ | "... usually our Plucky Psychic Hero has to get Chosen,
aprilcottage.co.uk | Selected, Sorted, Picked, Weeded, Impressed, Imprinted,
                   | Impaled or whatever (the verb being not nearly so
                   | important as the capitalization)..."   -- Ursula Vernon