Messages in this thread Patch in this message | | | Date | Mon, 14 Nov 2011 17:42:25 +0300 | Subject | Re: [PATCH v2 12/14] Keep number of running processes per-cgroup | From | Andrew Wagin <> |
| |
> This relies on the fact that the scheduler classes keep the number > of running processes internally in their entities. > > Signed-off-by: Glauber Costa <glommer@parallels.com> > --- ....
> +#ifdef CONFIG_CGROUP_SCHED > + /* root task group has autogrouping, so this doesn't hold */ > + if (tg != &root_task_group) { > +#ifdef CONFIG_FAIR_GROUP_SCHED > + tg_nr_running += cpu_rq(i)->cfs.nr_running;
I think you want to say tg->cfs_rq[i]->nr_running, otherwise it will show nr_running for root_task_group.
# mount -t cgroup -o cpu xxx /mnt/ # echo $$ > /mnt/1/tasks # while :; do :; done & [1] 1679 # while :; do :; done & [2] 1682 # while :; do :; done & [3] 1684 # while :; do :; done & [4] 1686 # while :; do :; done & [5] 1688 # cat /mnt/1/cpu.proc.stat | grep procs_running procs_running 2
> +#endif > +#ifdef CONFIG_RT_GROUP_SCHED > + tg_nr_running += cpu_rq(i)->rt.rt_nr_running; > +#endif > + } else > +#endif > + tg_nr_running += cpu_rq(i)->nr_running; > > for (j = 0; j < NR_SOFTIRQS; j++) { > unsigned int softirq_stat = kstat_softirqs_cpu(j, i); > @@ -9911,7 +9924,7 @@ int cpu_cgroup_proc_stat(struct cgroup *cgrp, struct cftype *cft, > tg_nr_switches, > (unsigned long)jif, > total_forks, > - nr_running(), > + tg_nr_running, > tg_iowait), > > seq_printf(p, "softirq %llu", (unsigned long long)sum_softirq); -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |