Messages in this thread | | | Date | Thu, 10 Oct 2013 18:53:37 +0200 | From | Peter Zijlstra <> | Subject | Re: [PATCH 0/6] Optimize the cpu hotplug locking -v2 |
| |
On Thu, Oct 10, 2013 at 09:43:55AM -0700, Andrew Morton wrote: > > But we would like to remove stomp machine from > > CPU hotplug. > > We do? That's news. It wasn't mentioned in the changelog and should > have been. Why?
It would be an unrelated change to this and unrelated to the reasons as to why I want a faster get_online_cpus().
> > Why prevent all CPUs from running when we want to remove > > one? > > So get_online_cpus() goes away. Nothing is more scalable than nothing!
Very much agreed; now stop_machine() wouldn't actually work for hotplug because it will instantly preempt everybody, including someone who might be in the middle of using per-cpu state of the cpu we're about to remove.
The freeze suggestion from Ingo would actually work because we freeze tasks at known good points (userspace and kthread_freeze() points) where we know they're not fiddling with per-cpu state.
| |