Re: [RFD] Freezing of kernel threads

[Linus Torvalds <torvalds@linux-foundation.org>]

On Sat, 12 May 2007, Linus Torvalds wrote:
> 
>  - make the rule be that you cannot sleep in the above macro, and make the 
>    rule be that CPU hotplug uses the same mechanisms that module unload 
>    already does!

Side note: we obviously already do the stop_machine_run thing for CPU 
down, so doing it for CPU bringup too would seem to just be a good thing. 
And it means that the locking for CPU's disappearing is the same as the 
locking rules for CPU's appearing: you just want to make sure to disable 
preemption (which you already have to do, to make the "CPU went away" case 
work out).

So how about just making sure "__cpu_up()" gets called with the same 
stop_machine_run() logic? Maybe it's not *quite* a one-liner change, but 
it should come fairly close...

		Linus