Messages in this thread | | | Date | Mon, 21 Feb 2011 04:32:35 +0100 | From | Andi Kleen <> | Subject | Re: [PATCH 5/6] perf: Optimise topology iteration |
| |
On Mon, Feb 21, 2011 at 11:29:24AM +0800, Lin Ming wrote: > On Mon, 2011-02-21 at 05:15 +0800, Andi Kleen wrote: > > On Mon, Feb 21, 2011 at 12:57:39AM +0800, Lin Ming wrote: > > > Currently we iterate the full machine looking for a matching core_id/nb > > > for the percore and the amd northbridge stuff , using a smaller topology > > > mask makes sense. > > > > This is still wrong for CPU hotplug. The CPU "owning" the per core > > does not necessarily need to be online anymore. > > This is remain issue for hotplug case, no matter we use > for_each_online_cpu or topology_thread_cpumask.
The original code I submitted used for_each_possible_cpu which is correct.
> > > Please drop this patch. > > Re-look at the code, I think for_each_online_cpu is wrong for percore, > we should use topology_thread_cpumask instead.
No, that's also cleared on unplug. You really need the possible map and nothing else.
-Andi
| |