On 25 September 2014 21:05, Dietmar Eggemann <dietmar.eggemann@xxxxxxx> wrote:
On 23/09/14 17:08, Vincent Guittot wrote:The usage can't be higher than the full capacity of the CPU because
Monitor the usage level of each group of each sched_domain level. The usage isWhy you are returning rq->cpu_capacity_orig + 1 (1025) in case
the amount of cpu_capacity that is currently used on a CPU or group of CPUs.
We use the utilization_load_avg to evaluate the usage level of each group.
Signed-off-by: Vincent Guittot <vincent.guittot@xxxxxxxxxx>
---
kernel/sched/fair.c | 13 +++++++++++++
1 file changed, 13 insertions(+)
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 2cf153d..4097e3f 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -4523,6 +4523,17 @@ static int select_idle_sibling(struct task_struct *p, int target)
return target;
}
+static int get_cpu_usage(int cpu)
+{
+ unsigned long usage = cpu_rq(cpu)->cfs.utilization_load_avg;
+ unsigned long capacity = capacity_orig_of(cpu);
+
+ if (usage >= SCHED_LOAD_SCALE)
+ return capacity + 1;
utilization_load_avg is greater or equal than 1024 and not usage or
(usage * capacity) >> SCHED_LOAD_SHIFT too?
it's about the running time on this CPU. Nevertheless, usage can be
higher than SCHED_LOAD_SCALE because of unfortunate rounding in
avg_period and running_load_avg or just after migrating tasks until
the average stabilizes with the new running time.
In case the weight of a sched group is greater than 1, you might loosethat's exactly for sched_group with more than 1 CPU that we need to
the information that the whole sched group is over-utilized too.
cap the usage of a CPU to 100%. Otherwise, the group could be seen as
overloaded (CPU0 usage at 121% + CPU1 usage at 80%) whereas CPU1 has
20% of available capacity
You add up the individual cpu usage values for a group bywe want to compare the output of the function with some capacity
sgs->group_usage += get_cpu_usage(i) in update_sg_lb_stats and later use
sgs->group_usage in group_is_overloaded to compare it against
sgs->group_capacity (taking imbalance_pct into consideration).
+Nit-pick: Since you're multiplying by a capacity value
+ return (usage * capacity) >> SCHED_LOAD_SHIFT;
(rq->cpu_capacity_orig) you should shift by SCHED_CAPACITY_SHIFT.
figures so i think that >> SCHED_LOAD_SHIFT is the right operation.
Just to make sure: You do this scaling of usage by cpu_capacity_origI do this for any system with CPUs that have an original capacity that
here only to cater for the fact that cpu_capacity_orig might be uarch
scaled (by arch_scale_cpu_capacity, !SMT) in update_cpu_capacity while
is different from SCHED_CAPACITY_SCALE so it's for both uArch and SMT.
utilization_load_avg is currently not.sorry i don't catch your point
We don't even uArch scale on ARM TC2 big.LITTLE platform in mainline
today due to the missing clock-frequency property in the device tree.
I think it's hard for people to grasp that your patch-set takes uArch--
scaling of capacity into consideration but not frequency scaling of
capacity (via arch_scale_freq_capacity, not used at the moment).
+}
+
/*
* select_task_rq_fair: Select target runqueue for the waking task in domains
* that have the 'sd_flag' flag set. In practice, this is SD_BALANCE_WAKE,
@@ -5663,6 +5674,7 @@ struct sg_lb_stats {
unsigned long sum_weighted_load; /* Weighted load of group's tasks */
unsigned long load_per_task;
unsigned long group_capacity;
+ unsigned long group_usage; /* Total usage of the group */
unsigned int sum_nr_running; /* Nr tasks running in the group */
unsigned int group_capacity_factor;
unsigned int idle_cpus;
@@ -6037,6 +6049,7 @@ static inline void update_sg_lb_stats(struct lb_env *env,
load = source_load(i, load_idx);
sgs->group_load += load;
+ sgs->group_usage += get_cpu_usage(i);
sgs->sum_nr_running += rq->cfs.h_nr_running;
if (rq->nr_running > 1)
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/