* [PATCH] cgroup: Fix rootcg cpu.stat guest double counting
@ 2021-10-28 22:15 ` Dan Schatzberg
0 siblings, 0 replies; 7+ messages in thread
From: Dan Schatzberg @ 2021-10-28 22:15 UTC (permalink / raw)
To: Tejun Heo
Cc: Zefan Li, Johannes Weiner, Boris Burkov,
open list:CONTROL GROUP (CGROUP),
open list
In account_guest_time in kernel/sched/cputime.c guest time is
attributed to both CPUTIME_NICE and CPUTIME_USER in addition to
CPUTIME_GUEST_NICE and CPUTIME_GUEST respectively. Therefore, adding
both to calculate usage results in double counting any guest time at
the rootcg.
Fixes: 936f2a70f207 ("cgroup: add cpu.stat file to root cgroup")
Signed-off-by: Dan Schatzberg <schatzberg.dan@gmail.com>
---
kernel/cgroup/rstat.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/kernel/cgroup/rstat.c b/kernel/cgroup/rstat.c
index b264ab5652ba..1486768f2318 100644
--- a/kernel/cgroup/rstat.c
+++ b/kernel/cgroup/rstat.c
@@ -433,8 +433,6 @@ static void root_cgroup_cputime(struct task_cputime *cputime)
cputime->sum_exec_runtime += user;
cputime->sum_exec_runtime += sys;
cputime->sum_exec_runtime += cpustat[CPUTIME_STEAL];
- cputime->sum_exec_runtime += cpustat[CPUTIME_GUEST];
- cputime->sum_exec_runtime += cpustat[CPUTIME_GUEST_NICE];
}
}
--
2.30.2
^ permalink raw reply related [flat|nested] 7+ messages in thread
* [PATCH] cgroup: Fix rootcg cpu.stat guest double counting
@ 2021-10-28 22:15 ` Dan Schatzberg
0 siblings, 0 replies; 7+ messages in thread
From: Dan Schatzberg @ 2021-10-28 22:15 UTC (permalink / raw)
To: Tejun Heo
Cc: Zefan Li, Johannes Weiner, Boris Burkov,
open list:CONTROL GROUP CGROUP, open list
In account_guest_time in kernel/sched/cputime.c guest time is
attributed to both CPUTIME_NICE and CPUTIME_USER in addition to
CPUTIME_GUEST_NICE and CPUTIME_GUEST respectively. Therefore, adding
both to calculate usage results in double counting any guest time at
the rootcg.
Fixes: 936f2a70f207 ("cgroup: add cpu.stat file to root cgroup")
Signed-off-by: Dan Schatzberg <schatzberg.dan-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
---
kernel/cgroup/rstat.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/kernel/cgroup/rstat.c b/kernel/cgroup/rstat.c
index b264ab5652ba..1486768f2318 100644
--- a/kernel/cgroup/rstat.c
+++ b/kernel/cgroup/rstat.c
@@ -433,8 +433,6 @@ static void root_cgroup_cputime(struct task_cputime *cputime)
cputime->sum_exec_runtime += user;
cputime->sum_exec_runtime += sys;
cputime->sum_exec_runtime += cpustat[CPUTIME_STEAL];
- cputime->sum_exec_runtime += cpustat[CPUTIME_GUEST];
- cputime->sum_exec_runtime += cpustat[CPUTIME_GUEST_NICE];
}
}
--
2.30.2
^ permalink raw reply related [flat|nested] 7+ messages in thread
* Re: [PATCH] cgroup: Fix rootcg cpu.stat guest double counting
@ 2021-10-29 17:21 ` Daniel Jordan
0 siblings, 0 replies; 7+ messages in thread
From: Daniel Jordan @ 2021-10-29 17:21 UTC (permalink / raw)
To: Dan Schatzberg
Cc: Tejun Heo, Zefan Li, Johannes Weiner, Boris Burkov,
open list:CONTROL GROUP (CGROUP),
open list, Ingo Molnar, Peter Zijlstra, Andrey Ryabinin
On Thu, Oct 28, 2021 at 03:15:27PM -0700, Dan Schatzberg wrote:
> In account_guest_time in kernel/sched/cputime.c guest time is
> attributed to both CPUTIME_NICE and CPUTIME_USER in addition to
> CPUTIME_GUEST_NICE and CPUTIME_GUEST respectively. Therefore, adding
> both to calculate usage results in double counting any guest time at
> the rootcg.
Yes, definitely a bug. The same fix was posted a couple months ago as
part of a series but never picked up:
https://x-lore.kernel.org/all/20210820094005.20596-3-arbn@yandex-team.com/
The series also touched cputime, so we might be waiting for scheduler
folks? +Ingo +Peter +Andrey
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] cgroup: Fix rootcg cpu.stat guest double counting
@ 2021-10-29 17:21 ` Daniel Jordan
0 siblings, 0 replies; 7+ messages in thread
From: Daniel Jordan @ 2021-10-29 17:21 UTC (permalink / raw)
To: Dan Schatzberg
Cc: Tejun Heo, Zefan Li, Johannes Weiner, Boris Burkov,
open list:CONTROL GROUP (CGROUP),
open list, Ingo Molnar, Peter Zijlstra, Andrey Ryabinin
On Thu, Oct 28, 2021 at 03:15:27PM -0700, Dan Schatzberg wrote:
> In account_guest_time in kernel/sched/cputime.c guest time is
> attributed to both CPUTIME_NICE and CPUTIME_USER in addition to
> CPUTIME_GUEST_NICE and CPUTIME_GUEST respectively. Therefore, adding
> both to calculate usage results in double counting any guest time at
> the rootcg.
Yes, definitely a bug. The same fix was posted a couple months ago as
part of a series but never picked up:
https://x-lore.kernel.org/all/20210820094005.20596-3-arbn-XoJtRXgx1JtgGdypvdXH1w@public.gmane.org/
The series also touched cputime, so we might be waiting for scheduler
folks? +Ingo +Peter +Andrey
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] cgroup: Fix rootcg cpu.stat guest double counting
@ 2021-11-01 16:48 ` Tejun Heo
0 siblings, 0 replies; 7+ messages in thread
From: Tejun Heo @ 2021-11-01 16:48 UTC (permalink / raw)
To: Dan Schatzberg
Cc: Zefan Li, Johannes Weiner, Boris Burkov,
open list:CONTROL GROUP (CGROUP),
open list
On Thu, Oct 28, 2021 at 03:15:27PM -0700, Dan Schatzberg wrote:
> In account_guest_time in kernel/sched/cputime.c guest time is
> attributed to both CPUTIME_NICE and CPUTIME_USER in addition to
> CPUTIME_GUEST_NICE and CPUTIME_GUEST respectively. Therefore, adding
> both to calculate usage results in double counting any guest time at
> the rootcg.
>
> Fixes: 936f2a70f207 ("cgroup: add cpu.stat file to root cgroup")
> Signed-off-by: Dan Schatzberg <schatzberg.dan@gmail.com>
Applying to cgroup/for-5.16.
Thanks.
--
tejun
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] cgroup: Fix rootcg cpu.stat guest double counting
@ 2021-11-01 16:48 ` Tejun Heo
0 siblings, 0 replies; 7+ messages in thread
From: Tejun Heo @ 2021-11-01 16:48 UTC (permalink / raw)
To: Dan Schatzberg
Cc: Zefan Li, Johannes Weiner, Boris Burkov,
open list:CONTROL GROUP (CGROUP),
open list
On Thu, Oct 28, 2021 at 03:15:27PM -0700, Dan Schatzberg wrote:
> In account_guest_time in kernel/sched/cputime.c guest time is
> attributed to both CPUTIME_NICE and CPUTIME_USER in addition to
> CPUTIME_GUEST_NICE and CPUTIME_GUEST respectively. Therefore, adding
> both to calculate usage results in double counting any guest time at
> the rootcg.
>
> Fixes: 936f2a70f207 ("cgroup: add cpu.stat file to root cgroup")
> Signed-off-by: Dan Schatzberg <schatzberg.dan-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Applying to cgroup/for-5.16.
Thanks.
--
tejun
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] cgroup: Fix rootcg cpu.stat guest double counting
2021-10-29 17:21 ` Daniel Jordan
(?)
@ 2021-11-15 16:41 ` Andrey Ryabinin
-1 siblings, 0 replies; 7+ messages in thread
From: Andrey Ryabinin @ 2021-11-15 16:41 UTC (permalink / raw)
To: Daniel Jordan, Dan Schatzberg
Cc: Tejun Heo, Zefan Li, Johannes Weiner, Boris Burkov,
open list:CONTROL GROUP (CGROUP),
open list, Ingo Molnar, Peter Zijlstra
On 10/29/21 8:21 PM, Daniel Jordan wrote:
> On Thu, Oct 28, 2021 at 03:15:27PM -0700, Dan Schatzberg wrote:
>> In account_guest_time in kernel/sched/cputime.c guest time is
>> attributed to both CPUTIME_NICE and CPUTIME_USER in addition to
>> CPUTIME_GUEST_NICE and CPUTIME_GUEST respectively. Therefore, adding
>> both to calculate usage results in double counting any guest time at
>> the rootcg.
>
> Yes, definitely a bug. The same fix was posted a couple months ago as
> part of a series but never picked up:
>
> https://x-lore.kernel.org/all/20210820094005.20596-3-arbn@yandex-team.com/
>
> The series also touched cputime, so we might be waiting for scheduler
> folks? +Ingo +Peter +Andrey
>
Yeah, I'll do rebase and send v3
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2021-11-15 16:40 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-10-28 22:15 [PATCH] cgroup: Fix rootcg cpu.stat guest double counting Dan Schatzberg
2021-10-28 22:15 ` Dan Schatzberg
2021-10-29 17:21 ` Daniel Jordan
2021-10-29 17:21 ` Daniel Jordan
2021-11-15 16:41 ` Andrey Ryabinin
2021-11-01 16:48 ` Tejun Heo
2021-11-01 16:48 ` Tejun Heo
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.