> Mike had an idea for what was causing this problem, but no fix to my
> knowledge.
> [However the CPU detection does switch suspend_on_halt on as a side
> effect, so the melting CPU problem is fixed.]
> I think I'd still leave this out of the main kernel development line
> for 2.0.x. [if you have a Cyrix aware BIOS it should have set
> suspend_on_halt]
but all the more reason to add it to the 2.1.x range of kernels
otherwise people will not use it and find out what the problems are
i have to make a consious effort to patch a new kernel to add the 6x86 stuff
to it - most people won't bother (or know about the patch)
it therefore won't get tested
-- simon running 2.1.20 (with cyrix patch) on a 6x86 166+ with step 1.6 and the wb cache enabled