Export the nr_running for each cpu cgroup could help us monitor the
container conveniently.
The total threads of cpu cgroup could be got from the tasks file, and it
could also be got from pids subsystem.
But we still donot know how many processes are running in a container,
only if we traversal the status of all processes, that's a little
expensive.
Hence export the nr_running.
Signed-off-by: Yafang Shao <[email protected]>
---
kernel/sched/core.c | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 644fa2e..926575a 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -6648,10 +6648,17 @@ static int cpu_cfs_stat_show(struct seq_file *sf, void *v)
{
struct task_group *tg = css_tg(seq_css(sf));
struct cfs_bandwidth *cfs_b = &tg->cfs_bandwidth;
+ unsigned int nr_running = 0;
+ int cpu;
+
+ for_each_online_cpu(cpu)
+ if (tg->cfs_rq[cpu])
+ nr_running += tg->cfs_rq[cpu]->h_nr_running;
seq_printf(sf, "nr_periods %d\n", cfs_b->nr_periods);
seq_printf(sf, "nr_throttled %d\n", cfs_b->nr_throttled);
seq_printf(sf, "throttled_time %llu\n", cfs_b->throttled_time);
+ seq_printf(sf, "nr_running %u\n", nr_running);
return 0;
}
--
1.8.3.1