[PATCH v4 02/10] sched: Extend scheduler's asym packing

From: Srinivas Pandruvada
Date: Wed Sep 21 2016 - 15:21:39 EST


From: Tim Chen <tim.c.chen@xxxxxxxxxxxxxxx>

We generalize the scheduler's asym packing to provide an ordering
of the cpu beyond just the cpu number. This allows the use of the
ASYM_PACKING scheduler machinery to move loads to preferred CPU in a
sched domain. The preference is defined with the cpu priority
given by arch_asym_cpu_priority(cpu).

We also record the most preferred cpu in a sched group when
we build the cpu's capacity for fast lookup of preferred cpu
during load balancing.

Signed-off-by: Tim Chen <tim.c.chen@xxxxxxxxxxxxxxx>
Signed-off-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@xxxxxxxxxxxxxxx>
---
include/linux/sched.h | 2 ++
kernel/sched/core.c | 18 ++++++++++++++++++
kernel/sched/fair.c | 35 ++++++++++++++++++++++++-----------
kernel/sched/sched.h | 6 ++++++
4 files changed, 50 insertions(+), 11 deletions(-)

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 98fe95f..82ca1e4 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1052,6 +1052,8 @@ static inline int cpu_numa_flags(void)
}
#endif

+int arch_asym_cpu_priority(int cpu);
+
struct sched_domain_attr {
int relax_domain_level;
};
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index e86c4a5..08135ca 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -6237,7 +6237,25 @@ static void init_sched_groups_capacity(int cpu, struct sched_domain *sd)
WARN_ON(!sg);

do {
+ int cpu, max_cpu = -1, prev_cpu = -1;
+
sg->group_weight = cpumask_weight(sched_group_cpus(sg));
+
+ if (!(sd->flags & SD_ASYM_PACKING))
+ goto next;
+
+ for_each_cpu(cpu, sched_group_cpus(sg)) {
+ if (prev_cpu < 0) {
+ prev_cpu = cpu;
+ max_cpu = cpu;
+ } else {
+ if (sched_asym_prefer(cpu, max_cpu))
+ max_cpu = cpu;
+ }
+ }
+ sg->asym_prefer_cpu = max_cpu;
+
+next:
sg = sg->next;
} while (sg != sd->groups);

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index a5cd07b..bb96e1a 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -100,6 +100,16 @@ const_debug unsigned int sysctl_sched_migration_cost = 500000UL;
*/
unsigned int __read_mostly sysctl_sched_shares_window = 10000000UL;

+#ifdef CONFIG_SMP
+/*
+ * For asym packing, by default the lower numbered cpu has higher priority.
+ */
+int __weak arch_asym_cpu_priority(int cpu)
+{
+ return -cpu;
+}
+#endif
+
#ifdef CONFIG_CFS_BANDWIDTH
/*
* Amount of runtime to allocate from global (tg) to local (per-cfs_rq) pool
@@ -6861,16 +6871,18 @@ static bool update_sd_pick_busiest(struct lb_env *env,
if (env->idle == CPU_NOT_IDLE)
return true;
/*
- * ASYM_PACKING needs to move all the work to the lowest
- * numbered CPUs in the group, therefore mark all groups
- * higher than ourself as busy.
+ * ASYM_PACKING needs to move all the work to the highest
+ * prority CPUs in the group, therefore mark all groups
+ * of lower priority than ourself as busy.
*/
- if (sgs->sum_nr_running && env->dst_cpu < group_first_cpu(sg)) {
+ if (sgs->sum_nr_running &&
+ sched_asym_prefer(env->dst_cpu, sg->asym_prefer_cpu)) {
if (!sds->busiest)
return true;

- /* Prefer to move from highest possible cpu's work */
- if (group_first_cpu(sds->busiest) < group_first_cpu(sg))
+ /* Prefer to move from lowest priority cpu's work */
+ if (sched_asym_prefer(sds->busiest->asym_prefer_cpu,
+ sg->asym_prefer_cpu))
return true;
}

@@ -7022,8 +7034,8 @@ static int check_asym_packing(struct lb_env *env, struct sd_lb_stats *sds)
if (!sds->busiest)
return 0;

- busiest_cpu = group_first_cpu(sds->busiest);
- if (env->dst_cpu > busiest_cpu)
+ busiest_cpu = sds->busiest->asym_prefer_cpu;
+ if (sched_asym_prefer(busiest_cpu, env->dst_cpu))
return 0;

env->imbalance = DIV_ROUND_CLOSEST(
@@ -7364,10 +7376,11 @@ static int need_active_balance(struct lb_env *env)

/*
* ASYM_PACKING needs to force migrate tasks from busy but
- * higher numbered CPUs in order to pack all tasks in the
- * lowest numbered CPUs.
+ * lower priority CPUs in order to pack all tasks in the
+ * highest priority CPUs.
*/
- if ((sd->flags & SD_ASYM_PACKING) && env->src_cpu > env->dst_cpu)
+ if ((sd->flags & SD_ASYM_PACKING) &&
+ sched_asym_prefer(env->dst_cpu, env->src_cpu))
return 1;
}

diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h
index b7fc1ce..3f3d04a 100644
--- a/kernel/sched/sched.h
+++ b/kernel/sched/sched.h
@@ -532,6 +532,11 @@ struct dl_rq {

#ifdef CONFIG_SMP

+static inline bool sched_asym_prefer(int a, int b)
+{
+ return arch_asym_cpu_priority(a) > arch_asym_cpu_priority(b);
+}
+
/*
* We add the notion of a root-domain which will be used to define per-domain
* variables. Each exclusive cpuset essentially defines an island domain by
@@ -884,6 +889,7 @@ struct sched_group {

unsigned int group_weight;
struct sched_group_capacity *sgc;
+ int asym_prefer_cpu; /* cpu of highest priority in group */

/*
* The CPUs this group covers.
--
2.7.4