On Thu, Dec 13, 2007 at 08:45:45AM -0800, Greg KH wrote:
> >> There still is a remaining shutdown problem in 2.6.22 with old APM based
> >> systems, but this fix is not the correct one
> > ..
> >
> > Eh? It may not fix APM systems, but it *does* fix several known examples
> > of ACPI systems.
>
> And I have a large number of real systems that this patch has caused
> problems with. So much so that we had to revert it from a shipping SuSE
> product already, so I figured it was safe to also revert this in
> mainline as it did cause regressions.
Cheesy, but why not something like..
+ if (acpi_enabled) // Or whatever the var name is these days
+ disable_nonboot_cpus();
until we find out more?
Dave
--
http://www.codemonkey.org.uk