Re: [LKP] [sched/fair] 070f5e860e: reaim.jobs_per_min -10.5% regression

From: Xing Zhengjun
Date: Fri Jun 19 2020 - 01:10:45 EST



On 6/18/2020 4:24 PM, Hillf Danton wrote:

On Thu, 18 Jun 2020 10:45:01 +0800 Xing Zhengjun wrote:
On 6/18/2020 12:25 AM, Vincent Guittot wrote:
Le mercredi 17 juin 2020 Ã 16:57:25 (+0200), Vincent Guittot a Ãcrit :
Le mercredi 17 juin 2020 Ã 08:30:21 (+0800), Xing Zhengjun a Ãcrit :


On 6/16/2020 2:54 PM, Vincent Guittot wrote:

Hi Xing,

Le mardi 16 juin 2020 Ã 11:17:16 (+0800), Xing Zhengjun a Ãcrit :


On 6/15/2020 4:10 PM, Vincent Guittot wrote:
Hi Xing,

Le lundi 15 juin 2020 Ã 15:26:59 (+0800), Xing Zhengjun a Ãcrit :


On 6/12/2020 7:06 PM, Hillf Danton wrote:

On Fri, 12 Jun 2020 14:36:49 +0800 Xing Zhengjun wrote:


...

...

I apply the patch based on v5.7, the test result is as the following:

=========================================================================================
tbox_group/testcase/rootfs/kconfig/compiler/runtime/nr_task/debug-setup/test/cpufreq_governor/ucode:

lkp-ivb-d04/reaim/debian-x86_64-20191114.cgz/x86_64-rhel-7.6/gcc-7/300s/100%/test/five_sec/performance/0x21

commit:
9f68395333ad7f5bfe2f83473fed363d4229f11c
070f5e860ee2bf588c99ef7b4c202451faa48236
v5.7
63a5d0fbb5ec62f5148c251c01e709b8358cd0ee (the test patch)

9f68395333ad7f5b 070f5e860ee2bf588c99ef7b4c2 v5.7
63a5d0fbb5ec62f5148c251c01e
---------------- --------------------------- ---------------------------
---------------------------
%stddev %change %stddev %change %stddev %change
%stddev
\ | \ | \
| \
0.69 -10.3% 0.62 -9.1% 0.62
+1.0% 0.69 reaim.child_systime
0.62 -1.0% 0.61 +0.5% 0.62
-0.1% 0.62 reaim.child_utime
66870 -10.0% 60187 -7.6% 61787
+1.1% 67636 reaim.jobs_per_min
16717 -10.0% 15046 -7.6% 15446
+1.1% 16909 reaim.jobs_per_min_child

OK. So the regression disappears when the conditions on runnable_avg are removed.

In the meantime, I have been able to understand more deeply what was happeningi
for this bench and how it is impacted by
commit: 070f5e860ee2 ("sched/fair: Take into account runnable_avg to classify group")

This bench forks a new thread for each and every new step. But a newly forked
threads start with a load_avg and a runnable_avg set to max whereas the threads
are running shortly before exiting. This makes the CPU to be set overloaded in
some case whereas it isn't.

Could you try the patch below ?
It fixes the problem on my setup (I have finally been able to reproduce the problem)

---
kernel/sched/fair.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 0aeffff62807..b33a4a9e1491 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -807,7 +807,7 @@ void post_init_entity_util_avg(struct task_struct *p)
}
}
- sa->runnable_avg = cpu_scale;
+ sa->runnable_avg = sa->util_avg;
if (p->sched_class != &fair_sched_class) {
/*
--
2.17.1


The patch above tries to move back to the group in the same classification as
before but this could harm other benchmarks.

There is another way to fix this by easing the migration of task in the case
of migrate_util imbalance.

Could you also try the patch below instead of the one above ?

---
kernel/sched/fair.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 0aeffff62807..fcaf66c4d086 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -7753,7 +7753,8 @@ static int detach_tasks(struct lb_env *env)
case migrate_util:
util = task_util_est(p);

- if (util > env->imbalance)
+ if (util/2 > env->imbalance &&
+ env->sd->nr_balance_failed <= env->sd->cache_nice_tries)
goto next;

Hmm... this sheds a shaft of light on computing imbalance for
migrate_util, see below.


env->imbalance -= util;
--
2.17.1



I apply the patch based on v5.7, the test result is as the following:

Thanks.


=========================================================================================
tbox_group/testcase/rootfs/kconfig/compiler/runtime/nr_task/debug-setup/test/cpufreq_governor/ucode:
lkp-ivb-d04/reaim/debian-x86_64-20191114.cgz/x86_64-rhel-7.6/gcc-7/300s/100%/test/five_sec/performance/0x21

commit:
9f68395333ad7f5bfe2f83473fed363d4229f11c
070f5e860ee2bf588c99ef7b4c202451faa48236
v5.7
69c81543653bf5f2c7105086502889fa019c15cb (the test patch)

9f68395333ad7f5b 070f5e860ee2bf588c99ef7b4c2 v5.7
69c81543653bf5f2c7105086502
---------------- --------------------------- ---------------------------
---------------------------
%stddev %change %stddev %change
%stddev %change %stddev
\ | \ | \
| \
0.69 -10.3% 0.62 -9.1% 0.62
-7.6% 0.63 reaim.child_systime
0.62 -1.0% 0.61 +0.5% 0.62
+1.9% 0.63 reaim.child_utime
66870 -10.0% 60187 -7.6% 61787
-5.9% 62947 reaim.jobs_per_min
16717 -10.0% 15046 -7.6% 15446
-5.9% 15736 reaim.jobs_per_min_child
97.84 -1.1% 96.75 -0.4% 97.43
-0.4% 97.47 reaim.jti
72000 -10.8% 64216 -8.3% 66000
-5.7% 67885 reaim.max_jobs_per_min
0.36 +10.6% 0.40 +7.8% 0.39
+6.0% 0.38 reaim.parent_time
1.58 Â 2% +71.0% 2.70 Â 2% +26.9% 2.01 Â
2% +23.6% 1.95 Â 3% reaim.std_dev_percent
0.00 Â 5% +110.4% 0.01 Â 3% +48.8% 0.01 Â
7% +43.2% 0.01 Â 5% reaim.std_dev_time
50800 -2.4% 49600 -1.6% 50000
-0.8% 50400 reaim.workload




97.84 -1.1% 96.75 -0.4% 97.43
+0.3% 98.09 reaim.jti
72000 -10.8% 64216 -8.3% 66000
+0.0% 72000 reaim.max_jobs_per_min
0.36 +10.6% 0.40 +7.8% 0.39
-1.1% 0.36 reaim.parent_time
1.58 Â 2% +71.0% 2.70 Â 2% +26.9% 2.01 Â 2%
-11.9% 1.39 Â 4% reaim.std_dev_percent
0.00 Â 5% +110.4% 0.01 Â 3% +48.8% 0.01 Â 7%
-27.3% 0.00 Â 15% reaim.std_dev_time
50800 -2.4% 49600 -1.6% 50000
+0.0% 50800 reaim.workload



=========================================================================================
tbox_group/testcase/rootfs/kconfig/compiler/runtime/nr_task/debug-setup/test/cpufreq_governor/ucode:

lkp-ivb-d04/reaim/debian-x86_64-20191114.cgz/x86_64-rhel-7.6/gcc-7/300s/100%/test/five_sec/performance/0x21

commit:
9f68395333ad7f5bfe2f83473fed363d4229f11c
070f5e860ee2bf588c99ef7b4c202451faa48236
v5.7
3e1643da53f3fc7414cfa3ad2a16ab2a164b7f4d (the test patch)

9f68395333ad7f5b 070f5e860ee2bf588c99ef7b4c2 v5.7
3e1643da53f3fc7414cfa3ad2a1
---------------- --------------------------- ---------------------------
---------------------------
%stddev %change %stddev %change %stddev %change
%stddev
\ | \ | \
| \
0.69 -10.3% 0.62 -9.1% 0.62
-7.1% 0.64 reaim.child_systime
0.62 -1.0% 0.61 +0.5% 0.62
+1.3% 0.63 reaim.child_utime
66870 -10.0% 60187 -7.6% 61787
-6.1% 62807 reaim.jobs_per_min
16717 -10.0% 15046 -7.6% 15446
-6.1% 15701 reaim.jobs_per_min_child
97.84 -1.1% 96.75 -0.4% 97.43
-0.5% 97.34 reaim.jti
72000 -10.8% 64216 -8.3% 66000
-5.7% 67885 reaim.max_jobs_per_min
0.36 +10.6% 0.40 +7.8% 0.39
+6.9% 0.38 reaim.parent_time
1.58 Â 2% +71.0% 2.70 Â 2% +26.9% 2.01 Â 2%
+32.5% 2.09 Â 6% reaim.std_dev_percent
0.00 Â 5% +110.4% 0.01 Â 3% +48.8% 0.01 Â 7%
+61.7% 0.01 Â 8% reaim.std_dev_time
50800 -2.4% 49600 -1.6% 50000
-1.3% 50133 reaim.workload



=========================================================================================
tbox_group/testcase/rootfs/kconfig/compiler/runtime/nr_task/debug-setup/test/cpufreq_governor/ucode:

lkp-ivb-d04/reaim/debian-x86_64-20191114.cgz/x86_64-rhel-7.6/gcc-7/300s/100%/test/five_sec/performance/0x21

commit:
9f68395333ad7f5bfe2f83473fed363d4229f11c
070f5e860ee2bf588c99ef7b4c202451faa48236
v5.7
6b33257768b8dd3982054885ea310871be2cfe0b (Hillf's patch)

9f68395333ad7f5b 070f5e860ee2bf588c99ef7b4c2 v5.7
6b33257768b8dd3982054885ea3
---------------- --------------------------- ---------------------------
---------------------------
%stddev %change %stddev %change %stddev %change
%stddev
\ | \ | \
| \
0.69 -10.3% 0.62 -9.1% 0.62
-10.1% 0.62 reaim.child_systime
0.62 -1.0% 0.61 +0.5% 0.62
+0.3% 0.62 reaim.child_utime
66870 -10.0% 60187 -7.6% 61787
-8.3% 61305 reaim.jobs_per_min
16717 -10.0% 15046 -7.6% 15446
-8.3% 15326 reaim.jobs_per_min_child
97.84 -1.1% 96.75 -0.4% 97.43
-0.5% 97.37 reaim.jti
72000 -10.8% 64216 -8.3% 66000
-8.3% 66000 reaim.max_jobs_per_min
0.36 +10.6% 0.40 +7.8% 0.39
+9.4% 0.39 reaim.parent_time
1.58 Â 2% +71.0% 2.70 Â 2% +26.9% 2.01 Â 2%
+33.2% 2.11 reaim.std_dev_percent
0.00 Â 5% +110.4% 0.01 Â 3% +48.8% 0.01 Â 7%
+65.3% 0.01 Â 3% reaim.std_dev_time
50800 -2.4% 49600 -1.6% 50000
-1.8% 49866 reaim.workload



--
Zhengjun Xing

--
Zhengjun Xing

--
Zhengjun Xing

--
Zhengjun Xing


This is not for testing. Compute imbalance for migrate_util with both
groups taken into account, aiming at making every one fully busy, a
state which cannot be measured without the help of imbalance_pct.


--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -8995,17 +8995,32 @@ static inline void calculate_imbalance(s
*/
if (local->group_type == group_has_spare) {
if (busiest->group_type > group_fully_busy) {
+ unsigned long cap, util;
+ unsigned int imbalance_pct = env->sd->imbalance_pct;
+
+ if (local->group_capacity *100 >
+ local->group_util * imbalance_pct)
+ cap = (local->group_capacity *100 -
+ local->group_util * imbalance_pct) /100;
+ else
+ cap = 0;
+
+ if (cap != 0 && busiest->group_util * imbalance_pct >
+ busiest->group_capacity *100)
+ util = (busiest->group_util * imbalance_pct -
+ busiest->group_capacity *100) /100;
+ else
+ util = 0;
/*
* If busiest is overloaded, try to fill spare
* capacity. This might end up creating spare capacity
* in busiest or busiest still being overloaded but
- * there is no simple way to directly compute the
+ * there is no simpler way to directly compute the
* amount of load to migrate in order to balance the
- * system.
+ * system than trying to make both groups fully busy.
*/
env->migration_type = migrate_util;
- env->imbalance = max(local->group_capacity, local->group_util) -
- local->group_util;
+ env->imbalance = min(cap, util);
/*
* In some cases, the group's utilization is max or even


I also test the above patch, the test result is as the following:
=========================================================================================
tbox_group/testcase/rootfs/kconfig/compiler/runtime/nr_task/debug-setup/test/cpufreq_governor/ucode:

lkp-ivb-d04/reaim/debian-x86_64-20191114.cgz/x86_64-rhel-7.6/gcc-7/300s/100%/test/five_sec/performance/0x21

commit:
9f68395333ad7f5bfe2f83473fed363d4229f11c
070f5e860ee2bf588c99ef7b4c202451faa48236
v5.7
dc2401de2e8cdc180f7d55cc1cdaee4136054680 (the above test patch)

9f68395333ad7f5b 070f5e860ee2bf588c99ef7b4c2 v5.7 dc2401de2e8cdc180f7d55cc1cd
---------------- --------------------------- --------------------------- ---------------------------
%stddev %change %stddev %change %stddev %change %stddev
\ | \ | \ | \
0.69 -10.3% 0.62 -9.1% 0.62 -8.2% 0.63 reaim.child_systime
0.62 -1.0% 0.61 +0.5% 0.62 +1.1% 0.63 reaim.child_utime
66870 -10.0% 60187 -7.6% 61787 -7.4% 61897 reaim.jobs_per_min
16717 -10.0% 15046 -7.6% 15446 -7.4% 15474 reaim.jobs_per_min_child
97.84 -1.1% 96.75 -0.4% 97.43 -0.5% 97.31 reaim.jti
72000 -10.8% 64216 -8.3% 66000 -8.3% 66000 reaim.max_jobs_per_min
0.36 +10.6% 0.40 +7.8% 0.39 +8.8% 0.39 reaim.parent_time
1.58 Â 2% +71.0% 2.70 Â 2% +26.9% 2.01 Â 2% +33.7% 2.11 Â 6% reaim.std_dev_percent
0.00 Â 5% +110.4% 0.01 Â 3% +48.8% 0.01 Â 7% +61.6% 0.01 Â 5% reaim.std_dev_time
50800 -2.4% 49600 -1.6% 50000 -1.6% 50000 reaim.workload



--
Zhengjun Xing