ANDROID: sched/fair: add tunable to force selection at cpu g... [chromiumos/third_party/kernel : chromeos-4.4]
213 views
Skip to first unread message
Sonny Rao (Gerrit)
Dec 14, 2016, 4:07:28 AM12/14/16
to Srinath Sridharan, Juri Lelli
Sonny Rao would like Srinath Sridharan and Juri Lelli to review this change.
ANDROID: sched/fair: add tunable to force selection at cpu granularity EAS assumes that clusters with smaller capacity cores are more energy-efficient. This may not be true on non-big-little devices, so EAS can make incorrect cluster selections when finding a CPU to wake. The "sched_is_big_little" hint can be used to cause a cpu-based selection instead of cluster-based selection. This change incorporates the addition of the sync hint enable patch EAS did not honour synchronous wakeup hints, a new sysctl is created to ask EAS to use this information when selecting a CPU. The control is called "sched_sync_hint_enable". Also contains: EAS: sched/fair: for SMP bias toward idle core with capacity For SMP devices, on wakeup bias towards idle cores that have capacity vs busy devices that need a higher OPP eas: favor idle cpus for boosted tasks BUG: 29533997 BUG: 29512132 eas/sched/fair: Favoring busy cpus with low OPPs BUG: 29533997 BUG: 29512132 BUG=chrome-os-partner:58667 TEST=build/boot on kevin Change-Id: I0cc9a1b1b88fb52916f18bf2d25715bdc3634f9c Signed-off-by: Juri Lelli <juri....@arm.com> Signed-off-by: Srinath Sridharan <srin...@google.com> (cherry picked from commit 4a5e890ec60d2e341fa560d8149997f5f0c48d6 from android-msm/android-4.4.y) Signed-off-by: Sonny Rao <sonn...@chromium.org> --- M include/linux/sched/sysctl.h M kernel/sched/fair.c M kernel/sysctl.c 3 files changed, 167 insertions(+), 40 deletions(-)
diff --git a/include/linux/sched/sysctl.h b/include/linux/sched/sysctl.h
index 7d02139..4883dcf 100644
--- a/include/linux/sched/sysctl.h
+++ b/include/linux/sched/sysctl.h
@@ -39,6 +39,8 @@
extern unsigned int sysctl_sched_min_granularity;
extern unsigned int sysctl_sched_wakeup_granularity;
extern unsigned int sysctl_sched_child_runs_first;
+extern unsigned int sysctl_sched_is_big_little;
+extern unsigned int sysctl_sched_sync_hint_enable;
extern unsigned int sysctl_sched_cstate_aware;
enum sched_tunable_scaling {
diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 3964449..6f87780 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -51,7 +51,10 @@
unsigned int sysctl_sched_latency = 6000000ULL;
unsigned int normalized_sysctl_sched_latency = 6000000ULL;
+unsigned int sysctl_sched_is_big_little = 0;
+unsigned int sysctl_sched_sync_hint_enable = 1;
unsigned int sysctl_sched_cstate_aware = 1;
+
/*
* The initial- and re-scaling of tunables is configurable
* (default SCHED_TUNABLESCALING_LOG = *(1+ilog(ncpus))
@@ -5571,13 +5574,111 @@
return target;
}
-static int energy_aware_wake_cpu(struct task_struct *p, int target)
+static inline int find_best_target(struct task_struct *p)
+{
+ int i, boosted;
+ int target_cpu = -1;
+ int target_capacity = 0;
+ int backup_capacity = 0;
+ int idle_cpu = -1;
+ int best_idle_cstate = INT_MAX;
+ int backup_cpu = -1;
+ unsigned long task_util_boosted, new_util;
+
+ /*
+ * Favor 1) busy cpu with most capacity at current OPP
+ * 2) idle_cpu with capacity at current OPP
+ * 3) busy cpu with capacity at higher OPP
+ */
+#ifdef CONFIG_CGROUP_SCHEDTUNE
+ boosted = schedtune_task_boost(p);
+#else
+ boosted = 0;
+#endif
+ task_util_boosted = boosted_task_util(p);
+ for_each_cpu(i, tsk_cpus_allowed(p)) {
+ int cur_capacity = capacity_curr_of(i);
+ struct rq *rq = cpu_rq(i);
+ int idle_idx = idle_get_state_idx(rq);
+
+ /*
+ * p's blocked utilization is still accounted for on prev_cpu
+ * so prev_cpu will receive a negative bias due to the double
+ * accounting. However, the blocked utilization may be zero.
+ */
+ new_util = cpu_util(i) + task_util_boosted;
+
+ /*
+ * Ensure minimum capacity to grant the required boost.
+ * The target CPU can be already at a capacity level higher
+ * than the one required to boost the task.
+ */
+
+ if (new_util > capacity_orig_of(i))
+ continue;
+
+ /*
+ * For boosted tasks we favor idle cpus unconditionally to
+ * improve latency.
+ */
+ if (idle_idx >= 0 && boosted) {
+ if (idle_cpu < 0 ||
+ (sysctl_sched_cstate_aware &&
+ best_idle_cstate > idle_idx)) {
+ best_idle_cstate = idle_idx;
+ idle_cpu = i;
+ }
+ continue;
+ }
+
+ if (new_util < cur_capacity) {
+ if (cpu_rq(i)->nr_running) {
+ if (target_capacity == 0 ||
+ target_capacity > cur_capacity) {
+ /* busy CPU with most capacity at current OPP */
+ target_cpu = i;
+ target_capacity = cur_capacity;
+ }
+ } else if (!boosted) {
+ if (idle_cpu < 0 ||
+ (sysctl_sched_cstate_aware &&
+ best_idle_cstate > idle_idx)) {
+ best_idle_cstate = idle_idx;
+ idle_cpu = i;
+ }
+ }
+ } else if (backup_capacity == 0 ||
+ backup_capacity > cur_capacity) {
+ /* first busy CPU with capacity at higher OPP */
+ backup_capacity = cur_capacity;
+ backup_cpu = i;
+ }
+ }
+
+ if (!boosted && target_cpu < 0) {
+ target_cpu = idle_cpu >= 0 ? idle_cpu : backup_cpu;
+ }
+
+ if (boosted && idle_cpu >= 0)
+ target_cpu = idle_cpu;
+ return target_cpu;
+}
+
+static int energy_aware_wake_cpu(struct task_struct *p, int target, int sync)
{
struct sched_domain *sd;
struct sched_group *sg, *sg_target;
int target_max_cap = INT_MAX;
int target_cpu = task_cpu(p);
int i;
+
+ if (sysctl_sched_sync_hint_enable && sync) {
+ int cpu = smp_processor_id();
+ cpumask_t search_cpus;
+ cpumask_and(&search_cpus, tsk_cpus_allowed(p), cpu_online_mask);
+ if (cpumask_test_cpu(cpu, &search_cpus))
+ return cpu;
+ }
sd = rcu_dereference(per_cpu(sd_ea, task_cpu(p)));
@@ -5587,50 +5688,60 @@
sg = sd->groups;
sg_target = sg;
- /*
- * Find group with sufficient capacity. We only get here if no cpu is
- * overutilized. We may end up overutilizing a cpu by adding the task,
- * but that should not be any worse than select_idle_sibling().
- * load_balance() should sort it out later as we get above the tipping
- * point.
- */
- do {
- /* Assuming all cpus are the same in group */
- int max_cap_cpu = group_first_cpu(sg);
+ if (sysctl_sched_is_big_little) {
/*
- * Assume smaller max capacity means more energy-efficient.
- * Ideally we should query the energy model for the right
- * answer but it easily ends up in an exhaustive search.
+ * Find group with sufficient capacity. We only get here if no cpu is
+ * overutilized. We may end up overutilizing a cpu by adding the task,
+ * but that should not be any worse than select_idle_sibling().
+ * load_balance() should sort it out later as we get above the tipping
+ * point.
*/
- if (capacity_of(max_cap_cpu) < target_max_cap &&
- task_fits_max(p, max_cap_cpu)) {
- sg_target = sg;
- target_max_cap = capacity_of(max_cap_cpu);
- }
- } while (sg = sg->next, sg != sd->groups);
+ do {
+ /* Assuming all cpus are the same in group */
+ int max_cap_cpu = group_first_cpu(sg);
- /* Find cpu with sufficient capacity */
- for_each_cpu_and(i, tsk_cpus_allowed(p), sched_group_cpus(sg_target)) {
+ /*
+ * Assume smaller max capacity means more energy-efficient.
+ * Ideally we should query the energy model for the right
+ * answer but it easily ends up in an exhaustive search.
+ */
+ if (capacity_of(max_cap_cpu) < target_max_cap &&
+ task_fits_max(p, max_cap_cpu)) {
+ sg_target = sg;
+ target_max_cap = capacity_of(max_cap_cpu);
+ }
+ } while (sg = sg->next, sg != sd->groups);
+
+ /* Find cpu with sufficient capacity */
+ for_each_cpu_and(i, tsk_cpus_allowed(p), sched_group_cpus(sg_target)) {
+ /*
+ * p's blocked utilization is still accounted for on prev_cpu
+ * so prev_cpu will receive a negative bias due to the double
+ * accounting. However, the blocked utilization may be zero.
+ */
+ int new_util = cpu_util(i) + boosted_task_util(p);
+
+ if (new_util > capacity_orig_of(i))
+ continue;
+
+ if (new_util < capacity_curr_of(i)) {
+ target_cpu = i;
+ if (cpu_rq(i)->nr_running)
+ break;
+ }
+
+ /* cpu has capacity at higher OPP, keep it as fallback */
+ if (target_cpu == task_cpu(p))
+ target_cpu = i;
+ }
+ } else {
/*
- * p's blocked utilization is still accounted for on prev_cpu
- * so prev_cpu will receive a negative bias due to the double
- * accounting. However, the blocked utilization may be zero.
+ * Find a cpu with sufficient capacity
*/
- int new_util = cpu_util(i) + boosted_task_util(p);
-
- if (new_util > capacity_orig_of(i))
- continue;
-
- if (new_util < capacity_curr_of(i)) {
- target_cpu = i;
- if (cpu_rq(i)->nr_running)
- break;
- }
-
- /* cpu has capacity at higher OPP, keep it as fallback */
- if (target_cpu == task_cpu(p))
- target_cpu = i;
+ int tmp_target = find_best_target(p);
+ if (tmp_target >= 0)
+ target_cpu = tmp_target;
}
if (target_cpu != task_cpu(p)) {
@@ -5707,7 +5818,7 @@
if (!sd) {
if (energy_aware() && !cpu_rq(cpu)->rd->overutilized)
- new_cpu = energy_aware_wake_cpu(p, prev_cpu);
+ new_cpu = energy_aware_wake_cpu(p, prev_cpu, sync);
else if (sd_flag & SD_BALANCE_WAKE) /* XXX always ? */
new_cpu = select_idle_sibling(p, new_cpu);
diff --git a/kernel/sysctl.c b/kernel/sysctl.c
index 090622e..4e50740 100644
--- a/kernel/sysctl.c
+++ b/kernel/sysctl.c
@@ -308,6 +308,20 @@
.extra2 = &max_sched_granularity_ns,
},
{
+ .procname = "sched_is_big_little",
+ .data = &sysctl_sched_is_big_little,
+ .maxlen = sizeof(unsigned int),
+ .mode = 0644,
+ .proc_handler = proc_dointvec,
+ },
+ {
+ .procname = "sched_sync_hint_enable",
+ .data = &sysctl_sched_sync_hint_enable,
+ .maxlen = sizeof(unsigned int),
+ .mode = 0644,
+ .proc_handler = proc_dointvec,
+ },
+ {
.procname = "sched_cstate_aware",
.data = &sysctl_sched_cstate_aware,
.maxlen = sizeof(unsigned int),
To view, visit this change. To unsubscribe, visit settings.
Dmitry Torokhov (Gerrit)
Dec 21, 2016, 2:11:48 AM12/21/16
to Sonny Rao, Juri Lelli, Srinath Sridharan
Dmitry Torokhov posted comments on this change.
Patch Set 4: Code-Review+2
To view, visit this change. To unsubscribe, visit settings.
Sonny Rao (Gerrit)
Dec 21, 2016, 2:27:15 AM12/21/16
to Juri Lelli, Srinath Sridharan
Sonny Rao uploaded patch set #3 to this change.
To view, visit this change. To unsubscribe, visit settings.
Sonny Rao (Gerrit)
Dec 21, 2016, 2:48:35 PM12/21/16
to Juri Lelli, Dmitry Torokhov, Srinath Sridharan
Sonny Rao posted comments on this change.
Patch Set 5: Verified+1
To view, visit this change. To unsubscribe, visit settings.
Sonny Rao (Gerrit)
Dec 21, 2016, 2:54:00 PM12/21/16
to Juri Lelli, Dmitry Torokhov, Srinath Sridharan
Sonny Rao posted comments on this change.
Patch Set 5: Commit-Queue+1
To view, visit this change. To unsubscribe, visit settings.
Sonny Rao (Gerrit)
Dec 21, 2016, 8:40:55 PM12/21/16
to Juri Lelli, ChromeOS Commit Bot, Dmitry Torokhov, Srinath Sridharan
Sonny Rao posted comments on this change.
Patch Set 6: Commit-Queue+1
To view, visit this change. To unsubscribe, visit settings.
Sonny Rao (Gerrit)
Dec 21, 2016, 8:57:28 PM12/21/16
to Juri Lelli, Srinath Sridharan, Dmitry Torokhov
Sonny Rao uploaded patch set #6 to this change.
To view, visit this change. To unsubscribe, visit settings.
Sonny Rao (Gerrit)
Dec 22, 2016, 2:45:47 AM12/22/16
to Juri Lelli, ChromeOS Commit Bot, Dmitry Torokhov, Srinath Sridharan
Sonny Rao (Gerrit)
Dec 22, 2016, 3:15:38 AM12/22/16
to Juri Lelli, ChromeOS Commit Bot, Dmitry Torokhov, Srinath Sridharan
Sonny Rao posted comments on this change.
Patch Set 7: Commit-Queue+1
To view, visit this change. To unsubscribe, visit settings.
Sonny Rao (Gerrit)
Dec 22, 2016, 5:31:25 AM12/22/16
to Juri Lelli, ChromeOS Commit Bot, Dmitry Torokhov, Srinath Sridharan
Sonny Rao posted comments on this change.
Patch Set 7: Verified+1
To view, visit this change. To unsubscribe, visit settings.
ChromeOS Commit Bot (Gerrit)
Dec 22, 2016, 9:44:09 AM12/22/16
to Sonny Rao, Juri Lelli, Srinath Sridharan, Dmitry Torokhov
ChromeOS Commit Bot uploaded patch set #8 to this change.
> Reviewed-on: https://chromium-review.googlesource.com/420053 Reviewed-by: Dmitry Torokhov <dt...@chromium.org> --- M include/linux/sched/sysctl.h M kernel/sched/fair.c M kernel/sysctl.c 3 files changed, 167 insertions(+), 40 deletions(-)
To view, visit this change. To unsubscribe, visit settings.
ChromeOS Commit Bot (Gerrit)
Dec 22, 2016, 9:46:13 AM12/22/16
to Sonny Rao, Juri Lelli, Dmitry Torokhov, Srinath Sridharan
ChromeOS Commit Bot merged this change.
> Reviewed-on: https://chromium-review.googlesource.com/420053 Reviewed-by: Dmitry Torokhov <dt...@chromium.org> --- M include/linux/sched/sysctl.h M kernel/sched/fair.c M kernel/sysctl.c 3 files changed, 167 insertions(+), 40 deletions(-)
Reply all
Reply to author
Forward
0 new messages