On Wed, Jun 05, 2019 at 04:20:03PM +0200, Michal Koutný <mkoutny@xxxxxxxx> wrote: > I considered relaxing the check to non-root cgroups only, however, as > your example shows, it doesn't prevent reaching the avoided state by > other paths. I'm not that familiar with RT sched to tell whether > RT-priority tasks in different task_groups break any assumptions. So I had another look and the check is bogus. The RT sched with !CONFIG_RT_GROUP_SCHED works only with the struct rt_rq embedded in the generic struct rq -- regardless of the task's membership in the cpu controller hierarchy. Perhaps, the commit message may mention this also prevents enabling cpu controller on unified hierarchy (if there are any (kernel) RT tasks to migrate). Reviewed-by: Michal Koutný <mkoutny@xxxxxxxx>
