Messages in this thread | | | Date | Thu, 17 Sep 2015 16:35:27 +0200 | From | Peter Zijlstra <> | Subject | Re: [PATCH 3/3] sched: Implement interface for cgroup unified hierarchy |
| |
On Sat, Sep 12, 2015 at 10:40:07AM -0400, Tejun Heo wrote: > So, one of the problems is that the kernel can't have tasks w/o > runnable CPUs, so we have to some workaround when, for whatever > reason, a task ends up with no CPU that it can run on.
No, just refuse that configuration.
> You say cpuset isn't a layering thing but that simply isn't true. > It's a cgroup-scope CPU mask. It layers atop task affinities > restricting what they can be configured to, limiting the effective > cpumask to the intersection of actually existing CPUs and overriding > individual affinity setting when the intersection doesn't exist.
No, just fail.
> The kernel does not update all CPU affinity masks when a CPU goes down > or comes up.
I'd be happy to fail a CPU down for user tasks where this is the last runnable CPU of.
| |