Re: [PATCH V2] tg: add sched wait_count of a task group

Re: [PATCH V2] tg: add sched wait_count of a task group

[Peter Zijlstra]

Your subject is wrong; this is 100% a sched patch.

On Fri, Jan 22, 2021 at 04:11:57AM +0800, wu860403-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org wrote:
> From: Liming Wu <wu860403-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
> 
> Now we can rely on PSI to reflect whether there is contention in
> the task group, but it cannot reflect the details of the contention.
> Through this metric, we can get avg latency of task group contention
>  from the dimension of scheduling.
>    delta(wait_usec)/delta(nr_waits)

Only if all tasks have the same weight. So in general this is useless.

> Also change wait_sum to wait_usec.

Also patches should do 1 thing, also you can't do that, you'll break
everybody that expect the current value.

[PATCH V2] tg: add sched wait_count of a task group

[wu860403-Re5JQEeQqe8AvxtiuMwx3w]

From: Liming Wu <wu860403-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>

Now we can rely on PSI to reflect whether there is contention in
the task group, but it cannot reflect the details of the contention.
Through this metric, we can get avg latency of task group contention
 from the dimension of scheduling.
   delta(wait_usec)/delta(nr_waits)
Also change wait_sum to wait_usec.

Signed-off-by Liming Wu <wu860403-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
---
 kernel/sched/core.c | 23 +++++++++++++++++++++--
 1 file changed, 21 insertions(+), 2 deletions(-)

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index e7e453492..45e46114e 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -8146,12 +8146,17 @@ static int cpu_cfs_stat_show(struct seq_file *sf, void *v)
 
 	if (schedstat_enabled() && tg != &root_task_group) {
 		u64 ws = 0;
+		u64 wc = 0;
 		int i;
 
-		for_each_possible_cpu(i)
+		for_each_possible_cpu(i) {
 			ws += schedstat_val(tg->se[i]->statistics.wait_sum);
+			wc += schedstat_val(tg->se[i]->statistics.wait_count);
+		}
 
-		seq_printf(sf, "wait_sum %llu\n", ws);
+		do_div(ws, NSEC_PER_USEC);
+		seq_printf(sf, "wait_usec %llu\n"
+			   "nr_waits %llu\n", ws, wc);
 	}
 
 	return 0;
@@ -8255,6 +8260,20 @@ static int cpu_extra_stat_show(struct seq_file *sf,
 			   "throttled_usec %llu\n",
 			   cfs_b->nr_periods, cfs_b->nr_throttled,
 			   throttled_usec);
+		if (schedstat_enabled() && tg != &root_task_group) {
+			u64 ws = 0;
+			u64 wc = 0;
+			int i;
+
+			for_each_possible_cpu(i) {
+				ws += schedstat_val(tg->se[i]->statistics.wait_sum);
+				wc += schedstat_val(tg->se[i]->statistics.wait_count);
+			}
+
+			do_div(ws, NSEC_PER_USEC);
+			seq_printf(sf, "wait_usec %llu\n"
+				   "nr_waits %llu\n", ws, wc);
+		}
 	}
 #endif
 	return 0;
-- 
2.20.1