sched,fair: flatten CPU controller runqueues
v4 cover sched,fair: flatten CPU controller runqueues
The current implementation of the CPU controller uses hierarchical runqueues, where on wakeup a task is enqueued on its group's runqueue, the group is enqueued on the runqueue of the group above it, etc.
This increases a fairly large amount of overhead for workloads that do a lot of wakeups a second, especially given that the default systemd hierarchy is 2 or 3 levels deep.
This patch series is an attempt at reducing that overhead, by placing all the tasks on the same runqueue, and scaling the task priority by the priority of the group, which is calculated periodically.
My main TODO items for the next period of time are likely going to be testing, testing, and testing. I hope to find and flush out any corner case I can find, and make sure performance does not regress with any workloads, and hopefully improves some.
Other TODO items:
- More code cleanups.
- Remove some more now unused code.
- Reimplement CONFIG_CFS_BANDWIDTH.
Plan for the CONFIG_CFS_BANDWIDTH reimplementation:
- When a cgroup gets throttled, mark the cgroup and its children as throttled.
- When pick_next_entity finds a task that is on a throttled cgroup, stash it on the cgroup runqueue (which is not used for runnable tasks any more). Leave the vruntime unchanged, and adjust that runqueue's vruntime to be that of the left-most task.
- When a cgroup gets unthrottled, and has tasks on it, place it on a vruntime ordered heap separate from the main runqueue.
- Have pick_next_task_fair grab one task off that heap every time it is called, and the min vruntime of that heap is lower than the vruntime of the CPU's cfs_rq (or the CPU has no other runnable tasks).
- Place that selected task on the CPU's cfs_rq, renormalizing its vruntime with the GENTLE_FAIR_SLEEPERS logic. That should help interleave the already runnable tasks with the recently unthrottled group, and prevent thundering herd issues.
- If the group gets throttled again before all of its task had a chance to run, vruntime sorting ensures all the tasks in the throttled cgroup get a chance to run over time.
Changes from v3:
- replace max_h_load with another hacky idea to ramp up the task_se_h_weight; I believe this new idea is wrong as well, but it will hopefully inspire a better solution (thanks to Peter Zijlstra)
- fix the ordering inside enqueue_task_fair to get task weights set up right (thanks to Peter Zijlstra)
- change wakeup_preempt_entity to reduce the number of task preemptions, hopefully resulting in behavior closer to what people configure in sysctl
- various other small cleanups and fixes
Changes from v2:
- fixed the web server performance regression, in a way vaguely similar to what Josef Bacik suggested (blame me for the implementation)
- removed some code duplication so the diffstat is redder than before
- propagate sum_exec_runtime up the tree, in preparation for CFS_BANDWIDTH
- small cleanups left and right
Changes from v1:
- use task_se_h_weight instead of task_se_h_load in calc_delta_fair and sched_slice, this seems to improve performance a little, but I still have some remaining regression to chase with our web server workload
- implement a number of the changes suggested by Dietmar Eggemann (still holding out for a better name for group_cfs_rq_of_parent)
This series applies on top of 5.2
include/linux/sched.h | 7 kernel/sched/core.c | 3 kernel/sched/debug.c | 15 kernel/sched/fair.c | 803 +++++++++++++++++++++----------------------------- kernel/sched/pelt.c | 68 +--- kernel/sched/pelt.h | 2 kernel/sched/sched.h | 9 7 files changed, 372 insertions(+), 535 deletions(-)
PATCH v4 1/15 sched: introduce task_se_h_load helper
Sometimes the hierarchical load of a sched_entity needs to be calculated. Rename task_h_load to task_se_h_load, and directly pass a sched_entity to that function.
Move the function declaration up above where it will be used later.
No functional changes.
--- kernel/sched/fair.c | 28 ++++++++++++++-------------- 1 file changed, 14 insertions(+), 14 deletions(-)
kernel/sched/fair.c +14 -14
@@ -242,6 +242,7 @@ static u64 __calc_delta(u64 delta_exec, unsigned long weight, struct load_weightconst struct sched_class fair_sched_class; +static unsigned long task_se_h_load (struct sched_entity *se);/************************************************************** * CFS operations on generic schedulable entities:
@@ -706,7 +707,6 @@ static u64 sched_vslice(struct cfs_rq *cfs_rq, struct sched_entity *se)#ifdef CONFIG_SMP static int select_idle_sibling (struct task_struct *p,int prev_cpu,int cpu); -static unsigned long task_h_load (struct task_struct *p);static unsigned long capacity_of (int cpu);/* Give new sched_entity start runnable values to heavy its load in infant time */
@@ -1668,7 +1668,7 @@ static void task_numa_compare(struct task_numa_env *env,/* * In the overloaded case, try and keep the load balanced. */ - load =task_h_load (env->p) -task_h_load (cur); + load =task_se_h_load (env->p->se) -task_se_h_load (cur->se);if (!load) goto assign;
@@ -1706,7 +1706,7 @@ static void task_numa_find_cpu(struct task_numa_env *env,bool maymove = false;int cpu; - load =task_h_load (env->p); + load =task_se_h_load (env->p->se); dst_load = env->dst_stats.load + load; src_load = env->src_stats.load - load;
@@ -3389,7 +3389,7 @@ static inline void add_tg_cfs_propagate(struct cfs_rq *cfs_rq, long runnable_sum* avg. The immediate corollary is that all (fair) tasks must be attached, see * post_init_entity_util_avg(). * -* cfs_rq->avg is used for task_h_load() and update_cfs_share() for example. +* cfs_rq->avg is used for task_se_h_load() and update_cfs_share() for example. * * Returns true if the load decayed or we removed load. *
@@ -3522,7 +3522,7 @@ static inline void update_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *s/* * Track task load average for carrying it to new CPU after migrated, and -* track group sched_entity load average for task_h_load calc in migration +* track group sched_entity load average for task_se_h_load calc in migration */ if (se->avg.last_update_time && !(flags &SKIP_AGE_LOAD ))__update_load_avg_se (now, cfs_rq, se);
@@ -3751,7 +3751,7 @@ static inline void update_misfit_status(struct task_struct *p, struct rq *rq) return; } - rq->misfit_task_load =task_h_load (p); + rq->misfit_task_load =task_se_h_load (&p->se); }#else /* CONFIG_SMP */
@@ -5739,7 +5739,7 @@ wake_affine_weight(struct sched_domain *sd, struct task_struct *p, this_eff_load =target_load (this_cpu, sd->wake_idx);if (sync) { -unsigned long current_load =task_h_load (current); +unsigned long current_load =task_se_h_load (¤t->se);if (current_load > this_eff_load)return this_cpu;
@@ -5747,7 +5747,7 @@ wake_affine_weight(struct sched_domain *sd, struct task_struct *p, this_eff_load -= current_load; } - task_load =task_h_load (p); + task_load =task_se_h_load (&p->se); this_eff_load += task_load;if (sched_feat (WA_BIAS ))
@@ -7600,7 +7600,7 @@ static int detach_tasks(struct lb_env *env)if (!can_migrate_task (p, env)) goto next; - load =task_h_load (p); + load =task_se_h_load (&p->se);if (sched_feat (LB_MIN ) && load <16 && !env->sd->nr_balance_failed) goto next;
@@ -7833,12 +7833,12 @@ static void update_cfs_rq_h_load(struct cfs_rq *cfs_rq) } } -static unsigned long task_h_load (struct task_struct *p) +static unsigned long task_se_h_load (struct sched_entity *se) { -struct cfs_rq *cfs_rq =task_cfs_rq (p); +struct cfs_rq *cfs_rq =cfs_rq_of (se);update_cfs_rq_h_load (cfs_rq); -return div64_ul (p->se.avg.load_avg * cfs_rq->h_load, +return div64_ul (se->avg.load_avg * cfs_rq->h_load,cfs_rq_load_avg (cfs_rq) +1 ); }#else
@@ -7865,9 +7865,9 @@ static inline void update_blocked_averages(int cpu) rq_unlock_irqrestore(rq, &rf); } -static unsigned long task_h_load (struct task_struct *p) +static unsigned long task_se_h_load (struct sched_entity *se) { -return p->se.avg.load_avg; +return se->avg.load_avg; }#endif
Signature · 1 line
2.20.1
Dietmar Eggemann
reviewerRe: [PATCH 01/15] sched: introduce task_se_h_load helper
Quoted 11 lines · Rik van Riel
Sometimes the hierarchical load of a
sched_entityneeds to be calculated. Renametask_h_loadtotask_se_h_load, and directly pass asched_entityto that function.Move the function declaration up above where it will be used later.
No functional changes.
Signed-off-by Rik van Riel <[email protected]>Reviewed-by Josef Bacik <[email protected]>---
[...]
kernel/sched/fair.c static void task_numa_compare(struct task_numa_env *env, · line 1671
Quoted 6 lines
kernel/sched/fair.c +1
-1@@ -1668,7 +1668,7 @@ static void task_numa_compare(struct task_numa_env *env,/* * In the overloaded case, try and keep the load balanced. */ - load =task_h_load (env->p) -task_h_load (cur); + load =task_se_h_load (env->p->se) -task_se_h_load (cur->se);
Shouldn't this be:
load = task_se_h_load(&env->p->se) - task_se_h_load(&cur->se);
kernel/sched/fair.c static void task_numa_find_cpu(struct task_numa_env *env, · line 1709
Quoted 9 lines
kernel/sched/fair.c +1
-1if (!load) goto assign;@@ -1706,7 +1706,7 @@ static void task_numa_find_cpu(struct task_numa_env *env,bool maymove = false;int cpu; - load =task_h_load (env->p); + load =task_se_h_load (env->p->se);
load = task_se_h_load(&env->p->se);
Only visible with CONFIG_NUMA_BALANCING though.
[...]
Re: [PATCH 01/15] sched: introduce task_se_h_load helper
Quoted 12 lines · Dietmar Eggemann
kernel/sched/fair.c static void task_numa_compare(struct · line 1672
Quoted 9 lines
kernel/sched/fair.c +1
-1@@ -1668,7 +1668,7 @@ static void task_numa_compare(structtask_numa_env *env,/* * In the overloaded case, try and keep the load balanced. */ - load =task_h_load (env->p) -task_h_load (cur); + load =task_se_h_load (env->p->se) -task_se_h_load (cur->se);Shouldn't this be:
Quoted 1 line
load =
task_se_h_load(&env->p->se)-task_se_h_load(&cur->se);
Yes indeed. Sorry I forgot to fix these after you
pointed them out last time.
They are now fixed in my tree.
Signature · 1 line
All Rights Reversed.
PATCH v4 2/15 sched: change /proc/sched_debug fields
In reply to Rik van RielRemove some fields from /proc/sched_debug that are removed from sched_entity in a subsequent patch, and add h_load, which comes in very handy to debug CPU controller weight distribution.
--- kernel/sched/debug.c | 11 ++--------- 1 file changed, 2 insertions(+), 9 deletions(-)
kernel/sched/debug.c +2 -9
@@ -416,11 +416,9 @@ static void print_cfs_group_stats(struct seq_file *m, int cpu, struct task_group }P (se->load.weight); -P (se->runnable_weight);#ifdef CONFIG_SMP P (se->avg.load_avg);P (se->avg.util_avg); -P (se->avg.runnable_load_avg);#endif #undef PN_SCHEDSTAT
@@ -538,7 +536,6 @@ void print_cfs_rq(struct seq_file *m, int cpu, struct cfs_rq *cfs_rq)SEQ_printf (m," .%-30s: %d\n" ,"nr_running" , cfs_rq->nr_running);SEQ_printf (m," .%-30s: %ld\n" ,"load" , cfs_rq->load.weight);#ifdef CONFIG_SMP -SEQ_printf (m," .%-30s: %ld\n" ,"runnable_weight" , cfs_rq->runnable_weight);SEQ_printf (m," .%-30s: %lu\n" ,"load_avg" , cfs_rq->avg.load_avg); SEQ_printf(m, " .%-30s: %lu\n", "runnable_load_avg",
@@ -547,17 +544,15 @@ void print_cfs_rq(struct seq_file *m, int cpu, struct cfs_rq *cfs_rq) cfs_rq->avg.util_avg);SEQ_printf (m," .%-30s: %u\n" ,"util_est_enqueued" , cfs_rq->avg.util_est.enqueued); -SEQ_printf (m," .%-30s: %ld\n" ,"removed.load_avg" , - cfs_rq->removed.load_avg);SEQ_printf (m," .%-30s: %ld\n" ,"removed.util_avg" , cfs_rq->removed.util_avg); -SEQ_printf (m," .%-30s: %ld\n" ,"removed.runnable_sum" , - cfs_rq->removed.runnable_sum);#ifdef CONFIG_FAIR_GROUP_SCHED SEQ_printf (m," .%-30s: %lu\n" ,"tg_load_avg_contrib" , cfs_rq->tg_load_avg_contrib);SEQ_printf (m," .%-30s: %ld\n" ,"tg_load_avg" ,atomic_long_read (&cfs_rq->tg->load_avg)); +SEQ_printf (m," .%-30s: %lu\n" ,"h_load" , + cfs_rq->h_load);#endif #endif #ifdef CONFIG_CFS_BANDWIDTH
@@ -961,10 +956,8 @@ void proc_sched_show_task(struct task_struct *p, struct pid_namespace *ns, "nr_involuntary_switches", (long long)p->nivcsw);P (se.load.weight); -P (se.runnable_weight);#ifdef CONFIG_SMP P (se.avg.load_sum); -P (se.avg.runnable_load_sum);P (se.avg.util_sum);P (se.avg.load_avg);P (se.avg.runnable_load_avg);
Signature · 1 line
2.20.1
PATCH v4 3/15 sched,fair: redefine runnable_load_avg as the sum of task_h_load
In reply to Rik van RielThe runnable_load magic is used to quickly propagate information about runnable tasks up the hierarchy of runqueues. The runnable_load_avg is mostly used for the load balancing code, which only examines the value at the root cfs_rq.
Redefine the root cfs_rq runnable_load_avg to be the sum of task_h_loads of the runnable tasks. This works because the hierarchical runnable_load of a task is already equal to the task_se_h_load today. This provides enough information to the load balancer.
The runnable_load_avg of the cgroup cfs_rqs does not appear to be used for anything, so don't bother calculating those.
This removes one of the things that the code currently traverses the cgroup hierarchy for, and getting rid of it brings us one step closer to a flat runqueue for the CPU controller.
--- include/linux/sched.h | 3 +- kernel/sched/core.c | 2 - kernel/sched/debug.c | 1 + kernel/sched/fair.c | 125 +++++++++++++----------------------------- kernel/sched/pelt.c | 64 ++++++--------------- kernel/sched/sched.h | 6 -- 6 files changed, 56 insertions(+), 145 deletions(-)
include/linux/sched.h +1 -2
@@ -391,7 +391,6 @@ struct util_est {struct sched_avg {u64 last_update_time;u64 load_sum; -u64 runnable_load_sum;u32 util_sum;u32 period_contrib;unsigned long load_avg;
@@ -439,7 +438,6 @@ struct sched_statistics {struct sched_entity {/* For load-balancing: */ struct load_weight load; -unsigned long runnable_weight;struct rb_node run_node;struct list_head group_node;unsigned int on_rq;
@@ -455,6 +453,7 @@ struct sched_entity {#ifdef CONFIG_FAIR_GROUP_SCHED int depth; +unsigned long enqueued_h_load;struct sched_entity *parent;/* rq on which this entity is (to be) queued: */ struct cfs_rq *cfs_rq;
kernel/sched/core.c +0 -2
@@ -744,7 +744,6 @@ static void set_load_weight(struct task_struct *p, bool update_load)if (task_has_idle_policy (p)) { load->weight =scale_load (WEIGHT_IDLEPRIO ); load->inv_weight =WMULT_IDLEPRIO ; - p->se.runnable_weight = load->weight;return ; }
@@ -757,7 +756,6 @@ static void set_load_weight(struct task_struct *p, bool update_load) }else { load->weight =scale_load (sched_prio_to_weight[prio]); load->inv_weight = sched_prio_to_wmult[prio]; - p->se.runnable_weight = load->weight; } }
kernel/sched/debug.c +1 -0
@@ -962,6 +962,7 @@ void proc_sched_show_task(struct task_struct *p, struct pid_namespace *ns,P (se.avg.load_avg);P (se.avg.runnable_load_avg);P (se.avg.util_avg); +P (se.enqueued_h_load);P (se.avg.last_update_time);P (se.avg.util_est.ewma);P (se.avg.util_est.enqueued);
kernel/sched/fair.c +38 -87
236 lines not on this page · raw message
kernel/sched/pelt.c +16 -48
190 lines not on this page · raw message
kernel/sched/sched.h +0 -6
@@ -487,7 +487,6 @@ struct cfs_bandwidth { };/* CFS-related fields in a runqueue */ struct cfs_rq {struct load_weight load; -unsigned long runnable_weight;unsigned int nr_running;unsigned int h_nr_running;
@@ -700,11 +699,6 @@ static inline long se_weight(struct sched_entity *se)return scale_load_down (se->load.weight); } -static inline long se_runnable (struct sched_entity *se) -{ -return scale_load_down (se->runnable_weight); -} -static inline bool sched_asym_prefer (int a,int b) {return arch_asym_cpu_priority (a) >arch_asym_cpu_priority (b);
Signature · 1 line
2.20.1
Vincent Guittot
maintainerRe: [PATCH 03/15] sched,fair: redefine runnable_load_avg as the sum of task_h_load
Hi Rik,
Quoted 11 lines · Rik van Riel
The
runnable_loadmagic is used to quickly propagate information about runnable tasks up the hierarchy of runqueues. Therunnable_load_avgis mostly used for the load balancing code, which only examines the value at the rootcfs_rq.Redefine the root
cfs_rqrunnable_load_avgto be the sum oftask_h_loadsof the runnable tasks. This works because the hierarchicalrunnable_loadof a task is already equal to thetask_se_h_loadtoday. This provides enough information to the load balancer.
Quoted 6 lines
The
runnable_load_avgof the cgroupcfs_rqsdoes not appear to be used for anything, so don't bother calculating those.This removes one of the things that the code currently traverses the cgroup hierarchy for, and getting rid of it brings us one step closer to a flat runqueue for the CPU controller.
I like your proposal but just wanted to clarify one thing with this patch. Although you removed the computation of runnable_load_avg of the cgroup cfs_rq, we are still traversing the hierarchy to update the root cfs_rq runnable_load_avg because we are traversing the hierarchy for computing the task_h_loads
That being said, if we manage to remove the need on using runnable_load_avg we will completely skip this traversal. I have a proposal to remove it from load balance and wake up path but i haven't look at numa stats which also use it
Quoted 560 lines
Signed-off-by Rik van Riel <[email protected]>--- include/linux/sched.h | 3 +- kernel/sched/core.c | 2 - kernel/sched/debug.c | 1 + kernel/sched/fair.c | 125 +++++++++++++----------------------------- kernel/sched/pelt.c | 64 ++++++--------------- kernel/sched/sched.h | 6 -- 6 files changed, 56 insertions(+), 145 deletions(-)include/linux/sched.h +1
-2@@ -391,7 +391,6 @@ struct util_est {struct sched_avg {u64 last_update_time;u64 load_sum; -u64 runnable_load_sum;u32 util_sum;u32 period_contrib;unsigned long load_avg;@@ -439,7 +438,6 @@ struct sched_statistics {struct sched_entity {/* For load-balancing: */ struct load_weight load; -unsigned long runnable_weight;struct rb_node run_node;struct list_head group_node;unsigned int on_rq;@@ -455,6 +453,7 @@ struct sched_entity {#ifdef CONFIG_FAIR_GROUP_SCHED int depth; +unsigned long enqueued_h_load;struct sched_entity *parent;/* rq on which this entity is (to be) queued: */ struct cfs_rq *cfs_rq;kernel/sched/core.c +0
-2@@ -744,7 +744,6 @@ static void set_load_weight(struct task_struct *p, bool update_load)if (task_has_idle_policy (p)) { load->weight =scale_load (WEIGHT_IDLEPRIO ); load->inv_weight =WMULT_IDLEPRIO ; - p->se.runnable_weight = load->weight;return ; }@@ -757,7 +756,6 @@ static void set_load_weight(struct task_struct *p, bool update_load) }else { load->weight =scale_load (sched_prio_to_weight[prio]); load->inv_weight = sched_prio_to_wmult[prio]; - p->se.runnable_weight = load->weight; } }kernel/sched/debug.c +1
-0@@ -962,6 +962,7 @@ void proc_sched_show_task(struct task_struct *p, struct pid_namespace *ns,P (se.avg.load_avg);P (se.avg.runnable_load_avg);P (se.avg.util_avg); +P (se.enqueued_h_load);P (se.avg.last_update_time);P (se.avg.util_est.ewma);P (se.avg.util_est.enqueued);kernel/sched/fair.c +38
-87@@ -723,9 +723,7 @@ void init_entity_runnable_average(struct sched_entity *se)* nothing has been attached to the task group yet. */ if (entity_is_task (se)) - sa->runnable_load_avg = sa->load_avg =scale_load_down (se->load.weight); - - se->runnable_weight = se->load.weight; + sa->load_avg =scale_load_down (se->load.weight);/* when this task enqueue'ed, it will contribute to its cfs_rq's load_avg */ }@@ -2766,20 +2764,39 @@ account_entity_dequeue(struct cfs_rq *cfs_rq, struct sched_entity *se)static inline void enqueue_runnable_load_avg (struct cfs_rq *cfs_rq,struct sched_entity *se) { - cfs_rq->runnable_weight += se->runnable_weight; +if (entity_is_task (se)) { +struct cfs_rq *root_cfs_rq = &cfs_rq->rq->cfs; + se->enqueued_h_load =task_se_h_load (se); - cfs_rq->avg.runnable_load_avg += se->avg.runnable_load_avg; - cfs_rq->avg.runnable_load_sum +=se_runnable (se) * se->avg.runnable_load_sum; + root_cfs_rq->avg.runnable_load_avg += se->enqueued_h_load; + } }static inline void dequeue_runnable_load_avg (struct cfs_rq *cfs_rq,struct sched_entity *se) { - cfs_rq->runnable_weight -= se->runnable_weight; +if (entity_is_task (se)) { +struct cfs_rq *root_cfs_rq = &cfs_rq->rq->cfs; +sub_positive (&root_cfs_rq->avg.runnable_load_avg, + se->enqueued_h_load); + } +} + +static inline void +update_runnable_load_avg (struct sched_entity *se) +{ +struct cfs_rq *root_cfs_rq = &cfs_rq_of (se)->rq->cfs; +long new_h_load, delta; -sub_positive (&cfs_rq->avg.runnable_load_avg, se->avg.runnable_load_avg); -sub_positive (&cfs_rq->avg.runnable_load_sum, -se_runnable (se) * se->avg.runnable_load_sum); +SCHED_WARN_ON (!entity_is_task (se)); + +if (!se->on_rq) +return ; + + new_h_load =task_se_h_load (se); + delta = new_h_load - se->enqueued_h_load; + root_cfs_rq->avg.runnable_load_avg += delta; + se->enqueued_h_load = new_h_load; } static inline void@@ -2807,7 +2824,7 @@ dequeue_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *se) { }#endif static void reweight_entity (struct cfs_rq *cfs_rq,struct sched_entity *se, -unsigned long weight,unsigned long runnable) +unsigned long weight) {if (se->on_rq) {/* commit outstanding execution time */ @@ -2818,7 +2835,6 @@ static void reweight_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, }dequeue_load_avg (cfs_rq, se); - se->runnable_weight = runnable;update_load_set (&se->load, weight);#ifdef CONFIG_SMP @@ -2826,8 +2842,6 @@ static void reweight_entity(struct cfs_rq *cfs_rq, struct sched_entity *se,u32 divider =LOAD_AVG_MAX -1024 + se->avg.period_contrib; se->avg.load_avg = div_u64(se_weight(se) * se->avg.load_sum, divider); - se->avg.runnable_load_avg = - div_u64(se_runnable(se) * se->avg.runnable_load_sum, divider); }while (0 );#endif @@ -2845,7 +2859,7 @@ void reweight_task(struct task_struct *p, int prio)struct load_weight *load = &se->load;unsigned long weight =scale_load (sched_prio_to_weight[prio]); -reweight_entity (cfs_rq, se, weight, weight); +reweight_entity (cfs_rq, se, weight); load->inv_weight = sched_prio_to_wmult[prio]; }@@ -2958,49 +2972,6 @@ static long calc_group_shares(struct cfs_rq *cfs_rq) return clamp_t(long, shares, MIN_SHARES, tg_shares); } -/* -* This calculates the effective runnable weight for a group entity based on -* the group entity weight calculated above. -* -* Because of the above approximation (2), our group entity weight is -* an load_avg based ratio (3). This means that it includes blocked load and -* does not represent the runnable weight. -* -* Approximate the group entity's runnable weight per ratio from the group -* runqueue: -* -* grq->avg.runnable_load_avg -* ge->runnable_weight = ge->load.weight * -------------------------- (7) -* grq->avg.load_avg -* -* However, analogous to above, since the avg numbers are slow, this leads to -* transients in the from-idle case. Instead we use: -* -* ge->runnable_weight = ge->load.weight * -* -* max(grq->avg.runnable_load_avg, grq->runnable_weight) -* ----------------------------------------------------- (8) -* max(grq->avg.load_avg, grq->load.weight) -* -* Where these max() serve both to use the 'instant' values to fix the slow -* from-idle and avoid the /0 on to-idle, similar to (6). -*/ -static long calc_group_runnable (struct cfs_rq *cfs_rq,long shares) -{ -long runnable, load_avg; - - load_avg =max (cfs_rq->avg.load_avg, -scale_load_down (cfs_rq->load.weight)); - - runnable =max (cfs_rq->avg.runnable_load_avg, -scale_load_down (cfs_rq->runnable_weight)); - - runnable *= shares; -if (load_avg) - runnable /= load_avg; - -return clamp_t (long, runnable,MIN_SHARES , shares); -}#endif /* CONFIG_SMP */ static inline int throttled_hierarchy (struct cfs_rq *cfs_rq);@@ -3012,25 +2983,26 @@ static inline int throttled_hierarchy(struct cfs_rq *cfs_rq);static void update_cfs_group (struct sched_entity *se) {struct cfs_rq *gcfs_rq =group_cfs_rq (se); -long shares, runnable; +long shares; -if (!gcfs_rq) +if (!gcfs_rq) { +update_runnable_load_avg (se);return ; + }if (throttled_hierarchy (gcfs_rq))return ;#ifndef CONFIG_SMP - runnable = shares =READ_ONCE (gcfs_rq->tg->shares); + shares =READ_ONCE (gcfs_rq->tg->shares);if (likely (se->load.weight == shares))return ;#else shares =calc_group_shares (gcfs_rq); - runnable =calc_group_runnable (gcfs_rq, shares);#endif -reweight_entity (cfs_rq_of (se), se, shares, runnable); +reweight_entity (cfs_rq_of (se), se, shares); }#else /* CONFIG_FAIR_GROUP_SCHED */ @@ -3243,8 +3215,8 @@ static inline void update_tg_cfs_runnable(struct cfs_rq *cfs_rq, struct sched_entity *se, struct cfs_rq *gcfs_rq) { long delta_avg, running_sum, runnable_sum = gcfs_rq->prop_runnable_sum; -unsigned long runnable_load_avg, load_avg; -u64 runnable_load_sum, load_sum =0 ; +unsigned long load_avg; +u64 load_sum =0 ;s64 delta_sum;if (!runnable_sum)@@ -3292,19 +3264,6 @@ update_tg_cfs_runnable(struct cfs_rq *cfs_rq, struct sched_entity *se, struct cf se->avg.load_avg = load_avg;add_positive (&cfs_rq->avg.load_avg, delta_avg);add_positive (&cfs_rq->avg.load_sum, delta_sum); - - runnable_load_sum = (s64 )se_runnable (se) * runnable_sum; - runnable_load_avg =div_s64 (runnable_load_sum,LOAD_AVG_MAX ); - delta_sum = runnable_load_sum -se_weight (se) * se->avg.runnable_load_sum; - delta_avg = runnable_load_avg - se->avg.runnable_load_avg; - - se->avg.runnable_load_sum = runnable_sum; - se->avg.runnable_load_avg = runnable_load_avg; - -if (se->on_rq) { -add_positive (&cfs_rq->avg.runnable_load_avg, delta_avg); -add_positive (&cfs_rq->avg.runnable_load_sum, delta_sum); - } }static inline void add_tg_cfs_propagate (struct cfs_rq *cfs_rq,long runnable_sum)@@ -3399,7 +3358,7 @@ static inline void add_tg_cfs_propagate(struct cfs_rq *cfs_rq, long runnable_sumstatic inline int update_cfs_rq_load_avg (u64 now,struct cfs_rq *cfs_rq) { -unsigned long removed_load =0 , removed_util =0 , removed_runnable_sum =0 ; +unsigned long removed_load =0 , removed_util =0 ;struct sched_avg *sa = &cfs_rq->avg;int decayed =0 ;@@ -3410,7 +3369,6 @@ update_cfs_rq_load_avg(u64 now, struct cfs_rq *cfs_rq)raw_spin_lock (&cfs_rq->removed.lock);swap (cfs_rq->removed.util_avg, removed_util);swap (cfs_rq->removed.load_avg, removed_load); -swap (cfs_rq->removed.runnable_sum, removed_runnable_sum); cfs_rq->removed.nr =0 ;raw_spin_unlock (&cfs_rq->removed.lock);@@ -3422,8 +3380,6 @@ update_cfs_rq_load_avg(u64 now, struct cfs_rq *cfs_rq)sub_positive (&sa->util_avg, r);sub_positive (&sa->util_sum, r * divider); -add_tg_cfs_propagate (cfs_rq, -(long )removed_runnable_sum); - decayed =1 ; }@@ -3477,8 +3433,6 @@ static void attach_entity_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *sdiv_u64 (se->avg.load_avg * se->avg.load_sum,se_weight (se)); } - se->avg.runnable_load_sum = se->avg.load_sum; -enqueue_load_avg (cfs_rq, se); cfs_rq->avg.util_avg += se->avg.util_avg; cfs_rq->avg.util_sum += se->avg.util_sum;@@ -7735,9 +7689,6 @@ static inline bool cfs_rq_is_decayed(struct cfs_rq *cfs_rq)if (cfs_rq->avg.util_sum)return false; -if (cfs_rq->avg.runnable_load_sum) -return false; -return true; }kernel/sched/pelt.c +16
-48@@ -106,7 +106,7 @@ static u32 __accumulate_pelt_segments(u64 periods, u32 d1, u32 d3)*/ static __always_inline u32accumulate_sum (u64 delta,struct sched_avg *sa, -unsigned long load,unsigned long runnable,int running) +unsigned long load,int running) {u32 contrib = (u32 )delta;/* p == 0 -> delta < 1024 */ u64 periods;@@ -119,8 +119,6 @@ accumulate_sum(u64 delta, struct sched_avg *sa,*/ if (periods) { sa->load_sum =decay_load (sa->load_sum, periods); - sa->runnable_load_sum = -decay_load (sa->runnable_load_sum, periods); sa->util_sum =decay_load ((u64)(sa->util_sum), periods);/* @@ -134,8 +132,6 @@ accumulate_sum(u64 delta, struct sched_avg *sa,if (load) sa->load_sum += load * contrib; -if (runnable) - sa->runnable_load_sum += runnable * contrib;if (running) sa->util_sum += contrib <<SCHED_CAPACITY_SHIFT ;@@ -172,7 +168,7 @@ accumulate_sum(u64 delta, struct sched_avg *sa,*/ static __always_inline int___update_load_sum (u64 now,struct sched_avg *sa, -unsigned long load,unsigned long runnable,int running) +unsigned long load,int running) {u64 delta;@@ -206,7 +202,7 @@ ___update_load_sum(u64 now, struct sched_avg *sa,* update_blocked_averages() */ if (!load) - runnable = running =0 ; + running =0 ;/* * Now we know we crossed measurement unit boundaries. The *_avg @@ -215,14 +211,14 @@ ___update_load_sum(u64 now, struct sched_avg *sa,* Step 1: accumulate *_sum since last_update_time. If we haven't * crossed period boundaries, finish. */ -if (!accumulate_sum (delta, sa, load, runnable, running)) +if (!accumulate_sum (delta, sa, load, running))return 0 ;return 1 ; } static __always_inline void -___update_load_avg (struct sched_avg *sa,unsigned long load,unsigned long runnable) +___update_load_avg (struct sched_avg *sa,unsigned long load) {u32 divider =LOAD_AVG_MAX -1024 + sa->period_contrib;@@ -230,41 +226,25 @@ ___update_load_avg(struct sched_avg *sa, unsigned long load, unsigned long runna* Step 2: update *_avg. */ sa->load_avg =div_u64 (load * sa->load_sum, divider); - sa->runnable_load_avg =div_u64 (runnable * sa->runnable_load_sum, divider);WRITE_ONCE (sa->util_avg, sa->util_sum / divider); }/* * sched_entity: * -* task: -* se_runnable() == se_weight() -* -* group: [ see update_cfs_group() ] -* se_weight() = tg->weight * grq->load_avg / tg->load_avg -* se_runnable() = se_weight(se) * grq->runnable_load_avg / grq->load_avg -* * load_sum := runnable_sum * load_avg = se_weight(se) * runnable_avg * -* runnable_load_sum := runnable_sum -* runnable_load_avg = se_runnable(se) * runnable_avg -* -* XXX collapse load_sum and runnable_load_sum -* * cfq_rq: * * load_sum = \Sum se_weight(se) * se->avg.load_sum * load_avg = \Sum se->avg.load_avg -* -* runnable_load_sum = \Sum se_runnable(se) * se->avg.runnable_load_sum -* runnable_load_avg = \Sum se->avg.runable_load_avg */ int __update_load_avg_blocked_se (u64 now,struct sched_entity *se) { -if (___update_load_sum (now, &se->avg,0 ,0 ,0 )) { -___update_load_avg (&se->avg,se_weight (se),se_runnable (se)); +if (___update_load_sum (now, &se->avg,0 ,0 )) { +___update_load_avg (&se->avg,se_weight (se));return 1 ; }@@ -273,10 +253,10 @@ int __update_load_avg_blocked_se(u64 now, struct sched_entity *se)int __update_load_avg_se (u64 now,struct cfs_rq *cfs_rq,struct sched_entity *se) { -if (___update_load_sum (now, &se->avg, !!se->on_rq, !!se->on_rq, +if (___update_load_sum (now, &se->avg, !!se->on_rq, cfs_rq->curr == se)) { -___update_load_avg (&se->avg,se_weight (se),se_runnable (se)); +___update_load_avg (&se->avg,se_weight (se));cfs_se_util_change (&se->avg);return 1 ; }@@ -288,10 +268,9 @@ int __update_load_avg_cfs_rq(u64 now, struct cfs_rq *cfs_rq) {if (___update_load_sum (now, &cfs_rq->avg,scale_load_down (cfs_rq->load.weight), -scale_load_down (cfs_rq->runnable_weight), cfs_rq->curr !=NULL )) { -___update_load_avg (&cfs_rq->avg,1 ,1 ); +___update_load_avg (&cfs_rq->avg,1 );return 1 ; }@@ -303,20 +282,18 @@ int __update_load_avg_cfs_rq(u64 now, struct cfs_rq *cfs_rq)* * util_sum = \Sum se->avg.util_sum but se->avg.util_sum is not tracked * util_sum = cpu_scale * load_sum -* runnable_load_sum = load_sum * -* load_avg and runnable_load_avg are not supported and meaningless. +* load_avg is not supported and meaningless. * */ int update_rt_rq_load_avg (u64 now,struct rq *rq,int running) {if (___update_load_sum (now, &rq->avg_rt, - running, running, running)) { -___update_load_avg (&rq->avg_rt,1 ,1 ); +___update_load_avg (&rq->avg_rt,1 );return 1 ; }@@ -328,18 +305,16 @@ int update_rt_rq_load_avg(u64 now, struct rq *rq, int running)* * util_sum = \Sum se->avg.util_sum but se->avg.util_sum is not tracked * util_sum = cpu_scale * load_sum -* runnable_load_sum = load_sum * */ int update_dl_rq_load_avg (u64 now,struct rq *rq,int running) {if (___update_load_sum (now, &rq->avg_dl, - running, running, running)) { -___update_load_avg (&rq->avg_dl,1 ,1 ); +___update_load_avg (&rq->avg_dl,1 );return 1 ; }@@ -352,7 +327,6 @@ int update_dl_rq_load_avg(u64 now, struct rq *rq, int running)* * util_sum = \Sum se->avg.util_sum but se->avg.util_sum is not tracked * util_sum = cpu_scale * load_sum -* runnable_load_sum = load_sum * */ @@ -379,17 +353,11 @@ int update_irq_load_avg(struct rq *rq, u64 running)* We can safely remove running from rq->clock because * rq->clock += delta with delta >= running */ - ret =___update_load_sum (rq->clock - running, &rq->avg_irq, -0 , -0 , -0 ); - ret +=___update_load_sum (rq->clock, &rq->avg_irq, -1 , -1 , -1 ); + ret =___update_load_sum (rq->clock - running, &rq->avg_irq,0 ,0 ); + ret +=___update_load_sum (rq->clock, &rq->avg_irq,1 ,1 );if (ret) -___update_load_avg (&rq->avg_irq,1 ,1 ); +___update_load_avg (&rq->avg_irq,1 );return ret; }kernel/sched/sched.h +0
-7@@ -487,7 +487,6 @@ struct cfs_bandwidth { };/* CFS-related fields in a runqueue */ struct cfs_rq {struct load_weight load; -unsigned long runnable_weight;unsigned int nr_running;unsigned int h_nr_running;@@ -700,11 +699,6 @@ static inline long se_weight(struct sched_entity *se)return scale_load_down (se->load.weight); } -static inline long se_runnable (struct sched_entity *se) -{ -return scale_load_down (se->runnable_weight); -} -static inline bool sched_asym_prefer (int a,int b) {return arch_asym_cpu_priority (a) >arch_asym_cpu_priority (b); --2.20.1
Re: [PATCH 03/15] sched,fair: redefine runnable_load_avg as the sum of task_h_load
Quoted 28 lines · Vincent Guittot
Hi Rik,
Quoted 16 lines · Rik van Riel
The
runnable_loadmagic is used to quickly propagate information about
runnable tasks up the hierarchy of runqueues. Therunnable_load_avgis
mostly used for the load balancing code, which only examines the value at
the rootcfs_rq.Redefine the root
cfs_rqrunnable_load_avgto be the sum oftask_h_loads
of the runnable tasks. This works because the hierarchicalrunnable_loadof
a task is already equal to thetask_se_h_loadtoday. This provides enough
information to the load balancer.Quoted 8 lines
The
runnable_load_avgof the cgroupcfs_rqsdoes not appear to be used for anything, so don't bother calculating those.This removes one of the things that the code currently traverses the
cgroup hierarchy for, and getting rid of it brings us one step closer
to a flat runqueue for the CPU controller.
Quoted 6 lines
I like your proposal but just wanted to clarify one thing with this patch.
Although you removed the computation ofrunnable_load_avgof the cgroupcfs_rq, we are still traversing the hierarchy to update the rootcfs_rqrunnable_load_avgbecause we are traversing the hierarchy for computing thetask_h_loads
The task_h_load hierarchy traversal in update_cfs_rq_h_load is rate limited to once a jiffy, though. Rate limiting the hierarchy traversal significantly reduces overhead.
Quoted 5 lines
That being said, if we manage to remove the need on using
runnable_load_avgwe will completely skip this traversal. I have a proposal to remove it from load balance and wake up path but i haven't
look at numa stats which also use it
Signature · 1 line
All Rights Reversed.
Vincent Guittot
maintainerRe: [PATCH 03/15] sched,fair: redefine runnable_load_avg as the sum of task_h_load
Quoted 37 lines · Rik van Riel
Quoted 28 lines · Vincent Guittot
Hi Rik,
Quoted 16 lines · Rik van Riel
The
runnable_loadmagic is used to quickly propagate information about
runnable tasks up the hierarchy of runqueues. Therunnable_load_avgis
mostly used for the load balancing code, which only examines the value at
the rootcfs_rq.Redefine the root
cfs_rqrunnable_load_avgto be the sum oftask_h_loads
of the runnable tasks. This works because the hierarchicalrunnable_loadof
a task is already equal to thetask_se_h_loadtoday. This provides enough
information to the load balancer.Quoted 8 lines
The
runnable_load_avgof the cgroupcfs_rqsdoes not appear to be used for anything, so don't bother calculating those.This removes one of the things that the code currently traverses the
cgroup hierarchy for, and getting rid of it brings us one step closer
to a flat runqueue for the CPU controller.Quoted 6 lines
I like your proposal but just wanted to clarify one thing with this patch.
Although you removed the computation ofrunnable_load_avgof the cgroupcfs_rq, we are still traversing the hierarchy to update the rootcfs_rqrunnable_load_avgbecause we are traversing the hierarchy for computing thetask_h_loads
Quoted 2 lines
The
task_h_loadhierarchy traversal inupdate_cfs_rq_h_loadis rate limited to once a jiffy, though. Rate limiting the
Ah yes. I forgot that it was jiffies and not clock_task that is used for limiting the update
Quoted 10 lines
hierarchy traversal significantly reduces overhead.
Quoted 5 lines
That being said, if we manage to remove the need on using
runnable_load_avgwe will completely skip this traversal. I have a proposal to remove it from load balance and wake up path but i haven't
look at numa stats which also use it--
All Rights Reversed.
PATCH v4 4/15 sched,fair: move runnable_load_avg to cfs_rq
In reply to Rik van RielSince only the root cfs_rq runnable_load_avg field is used any more, we can move the field from struct sched_avg, which every sched_entity has one of, directly into the struct cfs_rq, of which we have way fewer.
No functional changes.
--- include/linux/sched.h | 1 - kernel/sched/debug.c | 3 +-- kernel/sched/fair.c | 8 ++++---- kernel/sched/sched.h | 1 + 4 files changed, 6 insertions(+), 7 deletions(-)
include/linux/sched.h +0 -1
@@ -394,7 +394,6 @@ struct sched_avg {u32 util_sum;u32 period_contrib;unsigned long load_avg; -unsigned long runnable_load_avg;unsigned long util_avg;struct util_est util_est; }____cacheline_aligned ;
kernel/sched/debug.c +1 -2
@@ -539,7 +539,7 @@ void print_cfs_rq(struct seq_file *m, int cpu, struct cfs_rq *cfs_rq)SEQ_printf (m," .%-30s: %lu\n" ,"load_avg" , cfs_rq->avg.load_avg);SEQ_printf (m," .%-30s: %lu\n" ,"runnable_load_avg" , - cfs_rq->avg.runnable_load_avg); + cfs_rq->runnable_load_avg);SEQ_printf (m," .%-30s: %lu\n" ,"util_avg" , cfs_rq->avg.util_avg); SEQ_printf(m, " .%-30s: %u\n", "util_est_enqueued",
@@ -960,7 +960,6 @@ void proc_sched_show_task(struct task_struct *p, struct pid_namespace *ns,P (se.avg.load_sum);P (se.avg.util_sum);P (se.avg.load_avg); -P (se.avg.runnable_load_avg);P (se.avg.util_avg);P (se.enqueued_h_load);P (se.avg.last_update_time);
kernel/sched/fair.c +4 -4
@@ -2768,7 +2768,7 @@ enqueue_runnable_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *se)struct cfs_rq *root_cfs_rq = &cfs_rq->rq->cfs; se->enqueued_h_load =task_se_h_load (se); - root_cfs_rq->avg.runnable_load_avg += se->enqueued_h_load; + root_cfs_rq->runnable_load_avg += se->enqueued_h_load; } }
@@ -2777,7 +2777,7 @@ dequeue_runnable_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *se) {if (entity_is_task (se)) {struct cfs_rq *root_cfs_rq = &cfs_rq->rq->cfs; -sub_positive (&root_cfs_rq->avg.runnable_load_avg, +sub_positive (&root_cfs_rq->runnable_load_avg, se->enqueued_h_load); } }
@@ -2795,7 +2795,7 @@ update_runnable_load_avg(struct sched_entity *se) new_h_load =task_se_h_load (se); delta = new_h_load - se->enqueued_h_load; - root_cfs_rq->avg.runnable_load_avg += delta; + root_cfs_rq->runnable_load_avg += delta; se->enqueued_h_load = new_h_load; }
@@ -3561,7 +3561,7 @@ static void remove_entity_load_avg(struct sched_entity *se)static inline unsigned long cfs_rq_runnable_load_avg (struct cfs_rq *cfs_rq) { -return cfs_rq->avg.runnable_load_avg; +return cfs_rq->runnable_load_avg; }static inline unsigned long cfs_rq_load_avg (struct cfs_rq *cfs_rq)
kernel/sched/sched.h +1 -0
@@ -516,6 +516,7 @@ struct cfs_rq {* CFS load tracking */ struct sched_avg avg; +unsigned long runnable_load_avg;#ifndef CONFIG_64BIT u64 load_last_update_time_copy;#endif
Signature · 1 line
2.20.1
PATCH v4 5/15 sched,fair: remove cfs_rqs from leaf_cfs_rq_list bottom up
In reply to Rik van RielReducing the overhead of the CPU controller is achieved by not walking all the sched_entities every time a task is enqueued or dequeued.
One of the things being checked every single time is whether the cfs_rq is on the rq->leaf_cfs_rq_list.
By only removing a cfs_rq from the list once it no longer has children on the list, we can avoid walking the sched_entity hierarchy if the bottom cfs_rq is on the list, once the runqueues have been flattened.
--- kernel/sched/fair.c | 35 ++++++++++++++++++++++++++++++++++- 1 file changed, 34 insertions(+), 1 deletion(-)
kernel/sched/fair.c +34 -1
@@ -369,6 +369,39 @@ static inline void assert_list_leaf_cfs_rq(struct rq *rq)SCHED_WARN_ON (rq->tmp_alone_branch != &rq->leaf_cfs_rq_list); } +/* +* Because list_add_leaf_cfs_rq always places a child cfs_rq on the list +* immediately before a parent cfs_rq, and cfs_rqs are removed from the list +* bottom-up, we only have to test whether the cfs_rq before us on the list +* is our child. +*/ +static inline bool child_cfs_rq_on_list (struct cfs_rq *cfs_rq) +{ +struct cfs_rq *prev_cfs_rq; +struct list_head *prev; + + prev = cfs_rq->leaf_cfs_rq_list.prev; + prev_cfs_rq = container_of(prev, struct cfs_rq, leaf_cfs_rq_list); + +return (prev_cfs_rq->tg->parent == cfs_rq->tg); +} + +/* +* Remove a cfs_rq from the list if it has no children on the list. +* The scheduler iterates over the list regularly; if conditions for +* removal are still true, we'll get to this cfs_rq in the future. +*/ +static inline void list_del_leaf_cfs_rq_bottom (struct cfs_rq *cfs_rq) +{ +if (!cfs_rq->on_list) +return ; + +if (child_cfs_rq_on_list (cfs_rq)) +return ; + +list_del_leaf_cfs_rq (cfs_rq); +} +/* Iterate thr' all leaf cfs_rq's on a runqueue */ #define for_each_leaf_cfs_rq_safe (rq, cfs_rq, pos) \ list_for_each_entry_safe(cfs_rq, pos, &rq->leaf_cfs_rq_list, \
@@ -7723,7 +7756,7 @@ static void update_blocked_averages(int cpu)* decayed cfs_rqs linger on the list. */ if (cfs_rq_is_decayed (cfs_rq)) -list_del_leaf_cfs_rq (cfs_rq); +list_del_leaf_cfs_rq_bottom (cfs_rq);/* Don't need periodic decay once load/util_avg are null */ if (cfs_rq_has_blocked (cfs_rq))
Signature · 1 line
2.20.1
Re: [PATCH 05/15] sched,fair: remove cfs_rqs from leaf_cfs_rq_list bottom up
Quoted 7 lines · Rik van Riel
Reducing the overhead of the CPU controller is achieved by not walking all the
sched_entitiesevery time a task is enqueued or dequeued.One of the things being checked every single time is whether the
cfs_rqis on therq->leaf_cfs_rq_list.
Quoted 6 lines
By only removing a
cfs_rqfrom the list once it no longer has children on the list, we can avoid walking thesched_entityhierarchy if the bottomcfs_rqis on the list, once the runqueues have been flattened.Signed-off-by Rik van Riel <[email protected]>Suggested-by Vincent Guittot <[email protected]>
Quoted 60 lines
--- kernel/sched/fair.c | 35 ++++++++++++++++++++++++++++++++++- 1 file changed, 34 insertions(+), 1 deletion(-)kernel/sched/fair.c +34
-2@@ -369,6 +369,39 @@ static inline void assert_list_leaf_cfs_rq(struct rq *rq)SCHED_WARN_ON (rq->tmp_alone_branch != &rq->leaf_cfs_rq_list); } +/* +* Because list_add_leaf_cfs_rq always places a child cfs_rq on the list +* immediately before a parent cfs_rq, and cfs_rqs are removed from the list +* bottom-up, we only have to test whether the cfs_rq before us on the list +* is our child. +*/ +static inline bool child_cfs_rq_on_list (struct cfs_rq *cfs_rq) +{ +struct cfs_rq *prev_cfs_rq; +struct list_head *prev; + + prev = cfs_rq->leaf_cfs_rq_list.prev; + prev_cfs_rq = container_of(prev, struct cfs_rq, leaf_cfs_rq_list); + +return (prev_cfs_rq->tg->parent == cfs_rq->tg); +} + +/* +* Remove a cfs_rq from the list if it has no children on the list. +* The scheduler iterates over the list regularly; if conditions for +* removal are still true, we'll get to this cfs_rq in the future. +*/ +static inline void list_del_leaf_cfs_rq_bottom (struct cfs_rq *cfs_rq) +{ +if (!cfs_rq->on_list) +return ; + +if (child_cfs_rq_on_list (cfs_rq)) +return ; + +list_del_leaf_cfs_rq (cfs_rq); +} +/* Iterate thr' all leaf cfs_rq's on a runqueue */ #define for_each_leaf_cfs_rq_safe (rq, cfs_rq, pos) \ list_for_each_entry_safe(cfs_rq, pos, &rq->leaf_cfs_rq_list, \@@ -7723,7 +7756,7 @@ static void update_blocked_averages(int cpu)* decayed cfs_rqs linger on the list. */ if (cfs_rq_is_decayed (cfs_rq)) -list_del_leaf_cfs_rq (cfs_rq); +list_del_leaf_cfs_rq_bottom (cfs_rq);/* Don't need periodic decay once load/util_avg are null */ if (cfs_rq_has_blocked (cfs_rq)) --2.20.1
PATCH v4 6/15 sched,cfs: use explicit cfs_rq of parent se helper
In reply to Rik van RielUse an explicit "cfs_rq of parent sched_entity" helper in a few strategic places, where cfs_rq_of(se) may no longer point at the right runqueue once we flatten the hierarchical cgroup runqueues.
No functional change.
--- kernel/sched/fair.c | 17 +++++++++++++---- 1 file changed, 13 insertions(+), 4 deletions(-)
kernel/sched/fair.c +13 -4
@@ -276,6 +276,15 @@ static inline struct cfs_rq *group_cfs_rq(struct sched_entity *grp)return grp->my_q; } +/* runqueue owned by the parent entity; the root cfs_rq for a top level se */ +static inline struct cfs_rq *group_cfs_rq_of_parent (struct sched_entity *se) +{ +if (se->parent) +return group_cfs_rq (se->parent); + +return cfs_rq_of (se); +} +static inline bool list_add_leaf_cfs_rq (struct cfs_rq *cfs_rq) {struct rq *rq =rq_of (cfs_rq);
@@ -3319,7 +3328,7 @@ static inline int propagate_entity_load_avg(struct sched_entity *se) gcfs_rq->propagate =0 ; - cfs_rq =cfs_rq_of (se); + cfs_rq =group_cfs_rq_of_parent (se);add_tg_cfs_propagate (cfs_rq, gcfs_rq->prop_runnable_sum);
@@ -7796,7 +7805,7 @@ static void update_cfs_rq_h_load(struct cfs_rq *cfs_rq)WRITE_ONCE (cfs_rq->h_load_next,NULL );for_each_sched_entity (se) { - cfs_rq =cfs_rq_of (se); + cfs_rq =group_cfs_rq_of_parent (se);WRITE_ONCE (cfs_rq->h_load_next, se);if (cfs_rq->last_h_load_update == now)break ;
@@ -7819,7 +7828,7 @@ static void update_cfs_rq_h_load(struct cfs_rq *cfs_rq)static unsigned long task_se_h_load (struct sched_entity *se) { -struct cfs_rq *cfs_rq =cfs_rq_of (se); +struct cfs_rq *cfs_rq =group_cfs_rq_of_parent (se);update_cfs_rq_h_load (cfs_rq); return div64_ul(se->avg.load_avg * cfs_rq->h_load,
@@ -10166,7 +10175,7 @@ static void task_tick_fair(struct rq *rq, struct task_struct *curr, int queued)struct sched_entity *se = &curr->se;for_each_sched_entity (se) { - cfs_rq =cfs_rq_of (se); + cfs_rq =group_cfs_rq_of_parent (se);entity_tick (cfs_rq, se, queued); }
Signature · 1 line
2.20.1
Vincent Guittot
maintainerRe: [PATCH 06/15] sched,cfs: use explicit cfs_rq of parent se helper
Quoted 3 lines · Rik van Riel
Use an explicit "
cfs_rqof parentsched_entity" helper in a few strategic places, wherecfs_rq_of(se)may no longer point at the
The only case is the sched_entity of a task which will point to root cfs, isn't it ?
Quoted 68 lines
right runqueue once we flatten the hierarchical cgroup runqueues.
No functional change.
Signed-off-by Rik van Riel <[email protected]>--- kernel/sched/fair.c | 17 +++++++++++++---- 1 file changed, 13 insertions(+), 4 deletions(-)kernel/sched/fair.c +13
-5@@ -276,6 +276,15 @@ static inline struct cfs_rq *group_cfs_rq(struct sched_entity *grp)return grp->my_q; } +/* runqueue owned by the parent entity; the root cfs_rq for a top level se */ +static inline struct cfs_rq *group_cfs_rq_of_parent (struct sched_entity *se) +{ +if (se->parent) +return group_cfs_rq (se->parent); + +return cfs_rq_of (se); +} +static inline bool list_add_leaf_cfs_rq (struct cfs_rq *cfs_rq) {struct rq *rq =rq_of (cfs_rq);@@ -3319,7 +3328,7 @@ static inline int propagate_entity_load_avg(struct sched_entity *se) gcfs_rq->propagate =0 ; - cfs_rq =cfs_rq_of (se); + cfs_rq =group_cfs_rq_of_parent (se);add_tg_cfs_propagate (cfs_rq, gcfs_rq->prop_runnable_sum);@@ -7796,7 +7805,7 @@ static void update_cfs_rq_h_load(struct cfs_rq *cfs_rq)WRITE_ONCE (cfs_rq->h_load_next,NULL );for_each_sched_entity (se) { - cfs_rq =cfs_rq_of (se); + cfs_rq =group_cfs_rq_of_parent (se);WRITE_ONCE (cfs_rq->h_load_next, se);if (cfs_rq->last_h_load_update == now)break ;@@ -7819,7 +7828,7 @@ static void update_cfs_rq_h_load(struct cfs_rq *cfs_rq)static unsigned long task_se_h_load (struct sched_entity *se) { -struct cfs_rq *cfs_rq =cfs_rq_of (se); +struct cfs_rq *cfs_rq =group_cfs_rq_of_parent (se);update_cfs_rq_h_load (cfs_rq); return div64_ul(se->avg.load_avg * cfs_rq->h_load,@@ -10166,7 +10175,7 @@ static void task_tick_fair(struct rq *rq, struct task_struct *curr, int queued)struct sched_entity *se = &curr->se;for_each_sched_entity (se) { - cfs_rq =cfs_rq_of (se); + cfs_rq =group_cfs_rq_of_parent (se);entity_tick (cfs_rq, se, queued); } --2.20.1
Re: [PATCH 06/15] sched,cfs: use explicit cfs_rq of parent se helper
Quoted 6 lines · Vincent Guittot
Quoted 2 lines · Rik van Riel
Use an explicit "
cfs_rqof parentsched_entity" helper in a few strategic places, wherecfs_rq_of(se)may no longer point at theThe only case is the
sched_entityof a task which will point to root cfs, isn't it ?
True. I am more than open to ideas to simplify this part of the code, hopefully in some way that makes it easier to read.
Signature · 1 line
All Rights Reversed.
Vincent Guittot
maintainerRe: [PATCH 06/15] sched,cfs: use explicit cfs_rq of parent se helper
Quoted 6 lines · Rik van Riel
Quoted 4 lines · Vincent Guittot
Quoted 2 lines · Rik van Riel
Use an explicit "
cfs_rqof parentsched_entity" helper in a few strategic places, wherecfs_rq_of(se)may no longer point at the
Quoted 6 lines
Quoted 2 lines
The only case is the
sched_entityof a task which will point to root cfs, isn't it ?True. I am more than open to ideas to simplify this part of the code, hopefully in some way that makes it easier to read.
my question was mainly informative to make sure that i didn't miss a corner case
Quoted 3 lines
--
All Rights Reversed.
PATCH v4 7/15 sched,cfs: fix zero length timeslice calculation
In reply to Rik van RielThe way the time slice length is currently calculated, not only do high priority tasks get longer time slices than low priority tasks, but due to fixed point math, low priority tasks could end up with a zero length time slice. This can lead to cache thrashing and other inefficiencies.
Cap the minimum time slice length to sysctl_sched_min_granularity.
Tasks that end up getting a time slice length too long for their relative priority will simply end up having their vruntime advanced much faster than other tasks, resulting in them receiving time slices less frequently.
--- kernel/sched/fair.c | 7 +++++++ 1 file changed, 7 insertions(+)
kernel/sched/fair.c +7 -0
@@ -732,6 +732,13 @@ static u64 sched_slice(struct cfs_rq *cfs_rq, struct sched_entity *se) } slice =__calc_delta (slice, se->load.weight, load); } + +/* +* To avoid cache thrashing, run at least sysctl_sched_min_granularity. +* The vruntime of a low priority task advances faster; those tasks +* will simply get time slices less frequently. +*/ + slice =max_t (u64, slice, sysctl_sched_min_granularity);return slice; }
Signature · 1 line
2.20.1
Vincent Guittot
maintainerRe: [PATCH 07/15] sched,cfs: fix zero length timeslice calculation
Quoted 5 lines · Rik van Riel
The way the time slice length is currently calculated, not only do high priority tasks get longer time slices than low priority tasks, but due to fixed point math, low priority tasks could end up with a zero length time slice. This can lead to cache thrashing and other inefficiencies.
Have you got more details of those UCs ?
Quoted 6 lines
Cap the minimum time slice length to
sysctl_sched_min_granularity.Tasks that end up getting a time slice length too long for their relative priority will simply end up having their vruntime advanced much faster than other tasks, resulting in them receiving time slices less frequently.
In fact that already happen as we wait for the tick to preempt a task (unless you enable sched_feat(HRTICK))
so it sounds reasonable
Quoted 27 lines
Signed-off-by Rik van Riel <[email protected]>--- kernel/sched/fair.c | 7 +++++++ 1 file changed, 7 insertions(+)kernel/sched/fair.c +7
-1@@ -732,6 +732,13 @@ static u64 sched_slice(struct cfs_rq *cfs_rq, struct sched_entity *se) } slice =__calc_delta (slice, se->load.weight, load); } + +/* +* To avoid cache thrashing, run at least sysctl_sched_min_granularity. +* The vruntime of a low priority task advances faster; those tasks +* will simply get time slices less frequently. +*/ + slice =max_t (u64, slice, sysctl_sched_min_granularity);return slice; } --2.20.1
PATCH v4 8/15 sched,fair: simplify timeslice length code
In reply to Rik van RielThe idea behind __sched_period makes sense, but the results do not always.
When a CPU has one high priority task and a large number of low priority tasks, __sched_period will return a value larger than sysctl_sched_latency, and the one high priority task may end up getting a timeslice all for itself that is also much larger than sysctl_sched_latency.
The low priority tasks will have their time slices rounded up to sysctl_sched_min_granularity, resulting in an even larger scheduling latency than targeted by __sched_period.
Simplify the code by simply ripping out __sched_period and always taking fractions of sysctl_sched_latency.
If a high priority task ends up getting a "too small" time slice compared to low priority tasks, the vruntime scaling ensures that it will simply get scheduled more frequently than low priority tasks.
--- kernel/sched/fair.c | 18 +----------------- 1 file changed, 1 insertion(+), 17 deletions(-)
kernel/sched/fair.c +1 -17
@@ -691,22 +691,6 @@ static inline u64 calc_delta_fair(u64 delta, struct sched_entity *se)return delta; } -/* -* The idea is to set a period in which each task runs once. -* -* When there are too many tasks (sched_nr_latency) we have to stretch -* this period because otherwise the slices get too small. -* -* p = (nr <= nl) ? l : l*nr/nl -*/ -static u64 __sched_period (unsigned long nr_running) -{ -if (unlikely (nr_running > sched_nr_latency)) -return nr_running * sysctl_sched_min_granularity; -else -return sysctl_sched_latency; -} -/* * We calculate the wall-time slice from the period by taking a part * proportional to the weight.
@@ -715,7 +699,7 @@ static u64 __sched_period(unsigned long nr_running)*/ static u64 sched_slice (struct cfs_rq *cfs_rq,struct sched_entity *se) { -u64 slice =__sched_period (cfs_rq->nr_running + !se->on_rq); +u64 slice = sysctl_sched_latency;for_each_sched_entity (se) {struct load_weight *load;
Signature · 1 line
2.20.1
Vincent Guittot
maintainerRe: [PATCH 08/15] sched,fair: simplify timeslice length code
Quoted 7 lines · Rik van Riel
The idea behind
__sched_periodmakes sense, but the results do not always.When a CPU has one high priority task and a large number of low priority tasks,
__sched_periodwill return a value larger thansysctl_sched_latency, and the one high priority task may end up getting a timeslice all for itself that is also much larger thansysctl_sched_latency.
note that unless you enable sched_feat(HRTICK), the sched_slice is mainly used to decide how fast we preempt running task at tick but a newly wake up task can preempt it before
Quoted 4 lines
The low priority tasks will have their time slices rounded up to
sysctl_sched_min_granularity, resulting in an even larger scheduling latency than targeted by__sched_period.
Will this not break the fairness between a always running task and a short sleeping one with this changes ?
Quoted 7 lines
Simplify the code by simply ripping out
__sched_periodand always taking fractions ofsysctl_sched_latency.If a high priority task ends up getting a "too small" time slice compared to low priority tasks, the vruntime scaling ensures that it will simply get scheduled more frequently than low priority tasks.
Will you not increase the number of context switch ?
Quoted 45 lines
Signed-off-by Rik van Riel <[email protected]>--- kernel/sched/fair.c | 18 +----------------- 1 file changed, 1 insertion(+), 17 deletions(-)kernel/sched/fair.c +1
-18@@ -691,22 +691,6 @@ static inline u64 calc_delta_fair(u64 delta, struct sched_entity *se)return delta; } -/* -* The idea is to set a period in which each task runs once. -* -* When there are too many tasks (sched_nr_latency) we have to stretch -* this period because otherwise the slices get too small. -* -* p = (nr <= nl) ? l : l*nr/nl -*/ -static u64 __sched_period (unsigned long nr_running) -{ -if (unlikely (nr_running > sched_nr_latency)) -return nr_running * sysctl_sched_min_granularity; -else -return sysctl_sched_latency; -} -/* * We calculate the wall-time slice from the period by taking a part * proportional to the weight. @@ -715,7 +699,7 @@ static u64 __sched_period(unsigned long nr_running)*/ static u64 sched_slice (struct cfs_rq *cfs_rq,struct sched_entity *se) { -u64 slice =__sched_period (cfs_rq->nr_running + !se->on_rq); +u64 slice = sysctl_sched_latency;for_each_sched_entity (se) {struct load_weight *load; --2.20.1
Re: [PATCH 08/15] sched,fair: simplify timeslice length code
Quoted 16 lines · Vincent Guittot
Quoted 10 lines · Rik van Riel
The idea behind
__sched_periodmakes sense, but the results do not always.When a CPU has one high priority task and a large number of low priority
tasks,__sched_periodwill return a value larger thansysctl_sched_latency,
and the one high priority task may end up getting a timeslice all for itself
that is also much larger thansysctl_sched_latency.note that unless you enable
sched_feat(HRTICK), thesched_sliceis mainly used to decide how fast we preempt running task at tick but a newly wake up task can preempt it before
Quoted 7 lines
Quoted 4 lines
The low priority tasks will have their time slices rounded up to
sysctl_sched_min_granularity, resulting in an even larger scheduling
latency than targeted by__sched_period.Will this not break the fairness between a always running task and a short sleeping one with this changes ?
In what way?
The vruntime for the always running task will continue to advance the same way it always has.
Quoted 11 lines
Quoted 9 lines
Simplify the code by simply ripping out
__sched_periodand always taking
fractions ofsysctl_sched_latency.If a high priority task ends up getting a "too small" time slice compared
to low priority tasks, the vruntime scaling ensures that it will simply
get scheduled more frequently than low priority tasks.Will you not increase the number of context switch ?
It should actually decrease the number of context switches. If a nice +19 task gets a longer time slice than it would today, its vruntime will be advanced by more than sysctl_sched_latency, and it will not get to run again until another task has caught up with its vruntime.
That means the regular (or high) priority task that shares the CPU with that nice +19 task might get several time slices in a row until the nice +19 task gets to run again.
What am I overlooking?
Signature · 1 line
All Rights Reversed.
Vincent Guittot
maintainerRe: [PATCH 08/15] sched,fair: simplify timeslice length code
Quoted 23 lines · Rik van Riel
Quoted 21 lines · Vincent Guittot
Quoted 10 lines · Rik van Riel
The idea behind
__sched_periodmakes sense, but the results do not always.When a CPU has one high priority task and a large number of low priority
tasks,__sched_periodwill return a value larger thansysctl_sched_latency,
and the one high priority task may end up getting a timeslice all for itself
that is also much larger thansysctl_sched_latency.note that unless you enable
sched_feat(HRTICK), thesched_sliceis mainly used to decide how fast we preempt running task at tick but a newly wake up task can preempt it beforeQuoted 4 lines
The low priority tasks will have their time slices rounded up to
sysctl_sched_min_granularity, resulting in an even larger scheduling
latency than targeted by__sched_period.
Quoted 7 lines
Quoted 2 lines
Will this not break the fairness between a always running task and a short sleeping one with this changes ?
In what way?
The vruntime for the always running task will continue to advance the same way it always has.
Ok so 1st, my brain is probably not yet fully back from vacations as I have read sysctl_sched_min_granularity instead of sysctl_sched_latency
and wrongly thought that you were settingsysctl_sched_min_granularity for all tasks.
That being said, sched_slice is used to prevent other tasks to preempt the running task before it get a chances to run its ideal time compared to others and before new tasks modify the ideal sched_slice of each. By capping this max value, the task can be preempted earlier than before by newly wake up task and don't get the amount of running time it could have expect before the situation is changing
Quoted 11 lines
Quoted 11 lines
Quoted 10 lines
Simplify the code by simply ripping out
__sched_periodand always taking
fractions ofsysctl_sched_latency.If a high priority task ends up getting a "too small" time slice compared
to low priority tasks, the vruntime scaling ensures that it will simply
get scheduled more frequently than low priority tasks.
Quoted 5 lines
Quoted 1 line
Will you not increase the number of context switch ?
It should actually decrease the number of context switches. If a nice +19 task gets a longer time slice than it would today, its vruntime will be advanced by
In fact that's already the case today, when a task is scheduled, it runs a full jiffy even if its sched_slice is smaller than a jiffy (unless you have enabled sched_feat(HRTICK)).
Quoted 10 lines
more than
sysctl_sched_latency, and it will not get to run again until another task has caught up with its vruntime.That means the regular (or high) priority task that shares the CPU with that nice +19 task might get several time slices in a row until the nice +19 task gets to run again.
What am I overlooking?
My point is more for task that runs several ticks in a row. Their sched_slice will be shorter in some cases with your changes so they can be preempted earlier by other runnable tasks with a lower vruntime and there will be more context switch
Quoted 3 lines
--
All Rights Reversed.
Re: [PATCH 08/15] sched,fair: simplify timeslice length code
Quoted 9 lines · Vincent Guittot
Quoted 1 line · Rik van Riel
What am I overlooking?
My point is more for task that runs several ticks in a row. Their
sched_slicewill be shorter in some cases with your changes so they can be preempted earlier by other runnable tasks with a lower vruntime
and there will be more context switch
I can think of exactly one case where the time slice
will be shorter with my new code than with the old code,
and that is the case where:
- A CPU has nr_running > sched_nr_latency
- __sched_period returns a value larger than sysctl_sched_latency
- one of the tasks is much higher priority than the others
- that one task alone gets a timeslice larger than sysctl_sched_latency
With the new code, that high priority task will get a time slice that is a (large) fraction of sysctl_sched_latency, while the other (lower priority) tasks get their time slices rounded up to sysctl_sched_min_granularity.
When tasks get their timeslice rounded up, that will increase the total sched period in a similar way the old code did by returning a longer period from __sched_period.
If a CPU is faced with a large number of equal priority tasks, both the old code and the new code would end up giving each task a timeslice length of sysctl_sched_min_granularity.
What am I missing?
Signature · 1 line
All Rights Reversed.
Vincent Guittot
maintainerRe: [PATCH 08/15] sched,fair: simplify timeslice length code
Quoted 12 lines · Rik van Riel
Quoted 9 lines · Vincent Guittot
Quoted 1 line · Rik van Riel
What am I overlooking?
My point is more for task that runs several ticks in a row. Their
sched_slicewill be shorter in some cases with your changes so they can be preempted earlier by other runnable tasks with a lower vruntime
and there will be more context switch
Quoted 4 lines
I can think of exactly one case where the time slice will be shorter with my new code than with the old code, and that is the case where:
- A CPU hasnr_running>sched_nr_latency
yes nr_running must be higher than sched_nr_latency
Quoted 2 lines
-
__sched_periodreturns a value larger thansysctl_sched_latency
- one of the tasks is much higher priority than the others
it's not only one, that can be several. It depends of the number of running tasks
Quoted 4 lines
- that one task alone gets a timeslice larger than
sysctl_sched_latencyWith the new code, that high priority task will get a time slice that is a (large) fraction of
sysctl_sched_latency,
yes
Quoted 2 lines
while the other (lower priority) tasks get their time slices rounded up to
sysctl_sched_min_granularity.
yes and if the jify period is higher than sysctl_sched_min_granularity they will get a full jiffy period
Quoted 4 lines
When tasks get their timeslice rounded up, that will increase the total sched period in a similar way the old code did by returning a longer period from
__sched_period.
sched_slice is not a strict value and scheduler will not schedule out the task after the sched_slice (unless you enable HRTICK which is disable by default). Instead it will wait for next tick to change the running task
sched_slice is mainly use to ensure a minimum running time in a row. With this change, the running time of the high priority task will most probably be split in several slice instead of one
Quoted 9 lines
If a CPU is faced with a large number of equal priority tasks, both the old code and the new code would end up giving each task a timeslice length of
sysctl_sched_min_granularity.What am I missing?
--
All Rights Reversed.
Re: [PATCH 08/15] sched,fair: simplify timeslice length code
Quoted 9 lines · Vincent Guittot
Quoted 3 lines
When tasks get their timeslice rounded up, that will increase the total sched period in a similar way the old code did by returning a longer period from
__sched_period.
sched_sliceis not a strict value and scheduler will not schedule out the task after thesched_slice(unless you enable HRTICK which is disable by default). Instead it will wait for next tick to change the running task
Quoted 4 lines
sched_sliceis mainly use to ensure a minimum running time in a row. With this change, the running time of the high priority task will most
probably be split in several slice instead of one
I would be more than happy to drop this patch if you prefer. Just let me know.
Signature · 1 line
All Rights Reversed.
Vincent Guittot
maintainerRe: [PATCH 08/15] sched,fair: simplify timeslice length code
Quoted 12 lines · Rik van Riel
Quoted 9 lines · Vincent Guittot
Quoted 3 lines
When tasks get their timeslice rounded up, that will increase the total sched period in a similar way the old code did by returning a longer period from
__sched_period.
sched_sliceis not a strict value and scheduler will not schedule out the task after thesched_slice(unless you enable HRTICK which is disable by default). Instead it will wait for next tick to change the running task
Quoted 7 lines
Quoted 4 lines
sched_sliceis mainly use to ensure a minimum running time in a row. With this change, the running time of the high priority task will most
probably be split in several slice instead of oneI would be more than happy to drop this patch if you prefer. Just let me know.
If i'm not wrong, this change is not mandatory to flatten the runqueue and because of the possible impact if you would prefer to drop it from this serie
Regards,
Vincent
Quoted 3 lines
--
All Rights Reversed.
Re: [PATCH 08/15] sched,fair: simplify timeslice length code
Quoted 8 lines · Vincent Guittot
Quoted 2 lines · Rik van Riel
I would be more than happy to drop this patch if you prefer. Just let me know.
If i'm not wrong, this change is not mandatory to flatten the runqueue and because of the possible impact if you would prefer to drop it from this serie
OK, will do.
Signature · 1 line
All Rights Reversed.
PATCH v4 9/15 sched,fair: refactor enqueue/dequeue_entity
In reply to Rik van RielRefactor enqueue_entity, dequeue_entity, and update_load_avg, in order to split out the things we still want to happen at every level in the cgroup hierarchy with a flat runqueue from the things we only need to happen once.
No functional changes.
--- kernel/sched/fair.c | 64 +++++++++++++++++++++++++++++---------------- 1 file changed, 42 insertions(+), 22 deletions(-)
kernel/sched/fair.c +42 -22
@@ -3502,7 +3502,7 @@ static void detach_entity_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *s#define DO_ATTACH 0x4/* Update task and its cfs_rq load average */ -static inline void update_load_avg (struct cfs_rq *cfs_rq,struct sched_entity *se,int flags) +static inline bool update_load_avg (struct cfs_rq *cfs_rq,struct sched_entity *se,int flags) {u64 now =cfs_rq_clock_pelt (cfs_rq);int decayed;
@@ -3531,6 +3531,8 @@ static inline void update_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *s }else if (decayed && (flags &UPDATE_TG ))update_tg_load_avg (cfs_rq,0 ); + +return decayed; }#ifndef CONFIG_64BIT
@@ -3747,9 +3749,10 @@ static inline void update_misfit_status(struct task_struct *p, struct rq *rq)#define SKIP_AGE_LOAD 0x0#define DO_ATTACH 0x0 -static inline void update_load_avg (struct cfs_rq *cfs_rq,struct sched_entity *se,int not_used1) +static inline bool update_load_avg (struct cfs_rq *cfs_rq,struct sched_entity *se,int not_used1) {cfs_rq_util_change (cfs_rq,0 ); +return false; }static inline void remove_entity_load_avg (struct sched_entity *se) {}
@@ -3872,6 +3875,24 @@ static inline void check_schedstat_required(void)* CPU and an up-to-date min_vruntime on the destination CPU. */ +static bool +enqueue_entity_groups (struct cfs_rq *cfs_rq,struct sched_entity *se,int flags) +{ +/* +* When enqueuing a sched_entity, we must: +* - Update loads to have both entity and cfs_rq synced with now. +* - Add its load to cfs_rq->runnable_avg +* - For group_entity, update its weight to reflect the new share of +* its group cfs_rq +* - Add its new weight to cfs_rq->load.weight +*/ +if (!update_load_avg (cfs_rq, se,UPDATE_TG |DO_ATTACH )) +return false; + +update_cfs_group (se); +return true; +} +static void enqueue_entity (struct cfs_rq *cfs_rq,struct sched_entity *se,int flags) {
@@ -3896,16 +3917,6 @@ enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags)if (renorm && !curr) se->vruntime += cfs_rq->min_vruntime; -/* -* When enqueuing a sched_entity, we must: -* - Update loads to have both entity and cfs_rq synced with now. -* - Add its load to cfs_rq->runnable_avg -* - For group_entity, update its weight to reflect the new share of -* its group cfs_rq -* - Add its new weight to cfs_rq->load.weight -*/ -update_load_avg (cfs_rq, se,UPDATE_TG |DO_ATTACH ); -update_cfs_group (se);enqueue_runnable_load_avg (cfs_rq, se);account_entity_enqueue (cfs_rq, se);
@@ -3972,14 +3983,9 @@ static void clear_buddies(struct cfs_rq *cfs_rq, struct sched_entity *se) static __always_inline void return_cfs_rq_runtime(struct cfs_rq *cfs_rq); -static void -dequeue_entity (struct cfs_rq *cfs_rq,struct sched_entity *se,int flags) +static bool +dequeue_entity_groups (struct cfs_rq *cfs_rq,struct sched_entity *se,int flags) { -/* -* Update run-time statistics of the 'current'. -*/ -update_curr (cfs_rq); -/* * When dequeuing a sched_entity, we must: * - Update loads to have both entity and cfs_rq synced with now.
@@ -3988,7 +3994,21 @@ dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags)* - For group entity, update its weight to reflect the new share * of its group cfs_rq. */ -update_load_avg (cfs_rq, se,UPDATE_TG ); +if (!update_load_avg (cfs_rq, se,UPDATE_TG )) +return false; +update_cfs_group (se); + +return true; +} + +static void +dequeue_entity (struct cfs_rq *cfs_rq,struct sched_entity *se,int flags) +{ +/* +* Update run-time statistics of the 'current'. +*/ +update_curr (cfs_rq); +dequeue_runnable_load_avg (cfs_rq, se);update_stats_dequeue (cfs_rq, se, flags);
@@ -4012,8 +4032,6 @@ dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags)/* return excess runtime on last dequeue */ return_cfs_rq_runtime (cfs_rq); -update_cfs_group (se); -/* * Now advance min_vruntime if @se was the entity holding it back, * except when: DEQUEUE_SAVE && !DEQUEUE_MOVE, in this case we'll be
@@ -5178,6 +5196,7 @@ enqueue_task_fair(struct rq *rq, struct task_struct *p, int flags)if (se->on_rq)break ; cfs_rq =cfs_rq_of (se); +enqueue_entity_groups (cfs_rq, se, flags);enqueue_entity (cfs_rq, se, flags);/*
@@ -5260,6 +5279,7 @@ static void dequeue_task_fair(struct rq *rq, struct task_struct *p, int flags)for_each_sched_entity (se) { cfs_rq =cfs_rq_of (se); +dequeue_entity_groups (cfs_rq, se, flags);dequeue_entity (cfs_rq, se, flags);/*
Signature · 1 line
2.20.1
Vincent Guittot
maintainerRe: [PATCH 09/15] sched,fair: refactor enqueue/dequeue_entity
Hi Rik,
Quoted 27 lines · Rik van Riel
Refactor
enqueue_entity,dequeue_entity, andupdate_load_avg, in order to split out the things we still want to happen at every level in the cgroup hierarchy with a flat runqueue from the things we only need to happen once.No functional changes.
Signed-off-by Rik van Riel <[email protected]>--- kernel/sched/fair.c | 64 +++++++++++++++++++++++++++++---------------- 1 file changed, 42 insertions(+), 22 deletions(-)kernel/sched/fair.c +1
-1@@ -3502,7 +3502,7 @@ static void detach_entity_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *s#define DO_ATTACH 0x4/* Update task and its cfs_rq load average */ -static inline void update_load_avg (struct cfs_rq *cfs_rq,struct sched_entity *se,int flags) +static inline bool update_load_avg (struct cfs_rq *cfs_rq,struct sched_entity *se,int flags) {u64 now =cfs_rq_clock_pelt (cfs_rq);int decayed;
kernel/sched/fair.c static inline void update_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *s · line 3535
Quoted 5 lines
kernel/sched/fair.c +2
-0@@ -3531,6 +3531,8 @@ static inline void update_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *s }else if (decayed && (flags &UPDATE_TG ))update_tg_load_avg (cfs_rq,0 ); + +return decayed;
This is a functional change, isn't it ?update_cfs_group is now called only if decayed but we can we attach a task during the enqueue and there is no decay
Quoted 127 lines
kernel/sched/fair.c +20
-1}#ifndef CONFIG_64BIT @@ -3747,9 +3749,10 @@ static inline void update_misfit_status(struct task_struct *p, struct rq *rq)#define SKIP_AGE_LOAD 0x0#define DO_ATTACH 0x0 -static inline void update_load_avg (struct cfs_rq *cfs_rq,struct sched_entity *se,int not_used1) +static inline bool update_load_avg (struct cfs_rq *cfs_rq,struct sched_entity *se,int not_used1) {cfs_rq_util_change (cfs_rq,0 ); +return false; }static inline void remove_entity_load_avg (struct sched_entity *se) {}@@ -3872,6 +3875,24 @@ static inline void check_schedstat_required(void)* CPU and an up-to-date min_vruntime on the destination CPU. */ +static bool +enqueue_entity_groups (struct cfs_rq *cfs_rq,struct sched_entity *se,int flags) +{ +/* +* When enqueuing a sched_entity, we must: +* - Update loads to have both entity and cfs_rq synced with now. +* - Add its load to cfs_rq->runnable_avg +* - For group_entity, update its weight to reflect the new share of +* its group cfs_rq +* - Add its new weight to cfs_rq->load.weight +*/ +if (!update_load_avg (cfs_rq, se,UPDATE_TG |DO_ATTACH )) +return false; + +update_cfs_group (se); +return true; +} +static void enqueue_entity (struct cfs_rq *cfs_rq,struct sched_entity *se,int flags){
kernel/sched/fair.c +19
-21@@ -3896,16 +3917,6 @@ enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags)if (renorm && !curr) se->vruntime += cfs_rq->min_vruntime; -/* -* When enqueuing a sched_entity, we must: -* - Update loads to have both entity and cfs_rq synced with now. -* - Add its load to cfs_rq->runnable_avg -* - For group_entity, update its weight to reflect the new share of -* its group cfs_rq -* - Add its new weight to cfs_rq->load.weight -*/ -update_load_avg (cfs_rq, se,UPDATE_TG |DO_ATTACH ); -update_cfs_group (se);enqueue_runnable_load_avg (cfs_rq, se);account_entity_enqueue (cfs_rq, se);@@ -3972,14 +3983,9 @@ static void clear_buddies(struct cfs_rq *cfs_rq, struct sched_entity *se) static __always_inline void return_cfs_rq_runtime(struct cfs_rq *cfs_rq); -static void -dequeue_entity (struct cfs_rq *cfs_rq,struct sched_entity *se,int flags) +static bool +dequeue_entity_groups (struct cfs_rq *cfs_rq,struct sched_entity *se,int flags) { -/* -* Update run-time statistics of the 'current'. -*/ -update_curr (cfs_rq); -/* * When dequeuing a sched_entity, we must: * - Update loads to have both entity and cfs_rq synced with now. @@ -3988,7 +3994,21 @@ dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags)* - For group entity, update its weight to reflect the new share * of its group cfs_rq. */ -update_load_avg (cfs_rq, se,UPDATE_TG ); +if (!update_load_avg (cfs_rq, se,UPDATE_TG )) +return false; +update_cfs_group (se); + +return true; +} + +static void +dequeue_entity (struct cfs_rq *cfs_rq,struct sched_entity *se,int flags) +{ +/* +* Update run-time statistics of the 'current'. +*/ +update_curr (cfs_rq); +dequeue_runnable_load_avg (cfs_rq, se);update_stats_dequeue (cfs_rq, se, flags);@@ -4012,8 +4032,6 @@ dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags)/* return excess runtime on last dequeue */ return_cfs_rq_runtime (cfs_rq); -update_cfs_group (se); -/* * Now advance min_vruntime if @se was the entity holding it back, * except when: DEQUEUE_SAVE && !DEQUEUE_MOVE, in this case we'll be @@ -5178,6 +5196,7 @@ enqueue_task_fair(struct rq *rq, struct task_struct *p, int flags)if (se->on_rq)break ; cfs_rq =cfs_rq_of (se); +enqueue_entity_groups (cfs_rq, se, flags);enqueue_entity (cfs_rq, se, flags);/* @@ -5260,6 +5279,7 @@ static void dequeue_task_fair(struct rq *rq, struct task_struct *p, int flags)for_each_sched_entity (se) { cfs_rq =cfs_rq_of (se); +dequeue_entity_groups (cfs_rq, se, flags);dequeue_entity (cfs_rq, se, flags);/* -- 2.20.1
Re: [PATCH 09/15] sched,fair: refactor enqueue/dequeue_entity
Quoted 38 lines · Vincent Guittot
Hi Rik,
Quoted 35 lines · Rik van Riel
Refactor
enqueue_entity,dequeue_entity, andupdate_load_avg, in order
to split out the things we still want to happen at every level in the
cgroup hierarchy with a flat runqueue from the things we only need to
happen once.No functional changes.
Signed-off-by Rik van Riel <[email protected]>--- kernel/sched/fair.c | 64 +++++++++++++++++++++++++++++---------- ------ 1 file changed, 42 insertions(+), 22 deletions(-)kernel/sched/fair.c +0
-0@@ -3502,7 +3502,7 @@ static void detach_entity_load_avg(struct
cfs_rq*cfs_rq, structsched_entity*s
#defineDO_ATTACH0x4/* Update task and its
cfs_rqload average */
-static inline voidupdate_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags)
+static inline boolupdate_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags){ u64 now = cfs_rq_clock_pelt(cfs_rq); int decayed;kernel/sched/fair.c +0
-0@@ -3531,6 +3531,8 @@ static inline void update_load_avg(struct
cfs_rq*cfs_rq, structsched_entity*s
Quoted 8 lines
kernel/sched/fair.c static inline void update_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *s · line 3535
Quoted 4 lines
} else if (decayed && (flags & UPDATE_TG)) update_tg_load_avg(cfs_rq, 0);kernel/sched/fair.c +2
-0+ +return decayed;This is a functional change, isn't it ?
update_cfs_groupis now called only if decayed but we can we attach a task during the enqueue and there is no decay
Yes, it is, and patch 11 changes the way this functional change is done.
If you want, I can change this patch to not have the functional change, though in the end it should not make any difference.
Signature · 1 line
All Rights Reversed.
Vincent Guittot
maintainerRe: [PATCH 09/15] sched,fair: refactor enqueue/dequeue_entity
Quoted 49 lines · Rik van Riel
Quoted 38 lines · Vincent Guittot
Hi Rik,
Quoted 35 lines · Rik van Riel
Refactor
enqueue_entity,dequeue_entity, andupdate_load_avg, in order
to split out the things we still want to happen at every level in the
cgroup hierarchy with a flat runqueue from the things we only need to
happen once.No functional changes.
Signed-off-by Rik van Riel <[email protected]>--- kernel/sched/fair.c | 64 +++++++++++++++++++++++++++++---------- ------ 1 file changed, 42 insertions(+), 22 deletions(-)kernel/sched/fair.c +0
-0@@ -3502,7 +3502,7 @@ static void detach_entity_load_avg(struct
cfs_rq*cfs_rq, structsched_entity*s
#defineDO_ATTACH0x4/* Update task and its
cfs_rqload average */
-static inline voidupdate_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags)
+static inline boolupdate_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags){ u64 now = cfs_rq_clock_pelt(cfs_rq); int decayed;kernel/sched/fair.c +0
-0@@ -3531,6 +3531,8 @@ static inline void update_load_avg(struct
cfs_rq*cfs_rq, structsched_entity*sQuoted 8 lines
kernel/sched/fair.c static inline void update_load_avg(struct cfs_rq *cfs_rq, struct sched_entity *s · line 3535
Quoted 4 lines
} else if (decayed && (flags & UPDATE_TG)) update_tg_load_avg(cfs_rq, 0);kernel/sched/fair.c +2
-0+ +return decayed;This is a functional change, isn't it ?
update_cfs_groupis now called only if decayed but we can we attach a task during the enqueue and there is no decay
Quoted 6 lines
Yes, it is, and patch 11 changes the way this functional change is done.
If you want, I can change this patch to not have the functional change, though in the end it should not make any difference.
It's mainly that the code is not aligned with the commit message
I will continue to review other patch and will come back to this one after reviewing patch 11
Thanks
Vincent
Quoted 3 lines
--
All Rights Reversed.
PATCH v4 10/15 sched,fair: add helper functions for flattened runqueue
In reply to Rik van RielAdd helper functions to make the flattened runqueue patch a little smaller.
The task_se_h_weight function is similar to task_se_h_load, but scales the task weight by the group weight, without taking the task's duty cycle into account.
The task_se_in_cgroup helper is functionally identical to parent_entity, but directly calling a function with that name obscures what the other code is trying to use it for, and would make the code harder to understand.
--- kernel/sched/fair.c | 31 +++++++++++++++++++++++++++++++ 1 file changed, 31 insertions(+)
kernel/sched/fair.c +31 -0
@@ -243,6 +243,7 @@ static u64 __calc_delta(u64 delta_exec, unsigned long weight, struct load_weightconst struct sched_class fair_sched_class;static unsigned long task_se_h_load (struct sched_entity *se); +static unsigned long task_se_h_weight (struct sched_entity *se);/************************************************************** * CFS operations on generic schedulable entities:
@@ -431,6 +432,12 @@ static inline struct sched_entity *parent_entity(struct sched_entity *se)return se->parent; } +/* Is this (task) sched_entity in a non-root cgroup? */ +static inline bool task_se_in_cgroup (struct sched_entity *se) +{ +return parent_entity (se); +} +static void find_matching_se (struct sched_entity **se,struct sched_entity **pse) {
@@ -513,6 +520,11 @@ static inline struct sched_entity *parent_entity(struct sched_entity *se)return NULL ; } +static inline bool task_se_in_cgroup (struct sched_entity *se) +{ +return false; +} +static inline void find_matching_se (struct sched_entity **se,struct sched_entity **pse) {
@@ -7837,6 +7849,20 @@ static void update_cfs_rq_h_load(struct cfs_rq *cfs_rq) } } +static unsigned long task_se_h_weight (struct sched_entity *se) +{ +struct cfs_rq *cfs_rq; + +if (!task_se_in_cgroup (se)) +return se->load.weight; + + cfs_rq =group_cfs_rq_of_parent (se); +update_cfs_rq_h_load (cfs_rq); + +/* Reduce the load.weight by the h_load of the group the task is in. */ +return (cfs_rq->h_load * se->load.weight) >>SCHED_FIXEDPOINT_SHIFT ; +} +static unsigned long task_se_h_load (struct sched_entity *se) {struct cfs_rq *cfs_rq =group_cfs_rq_of_parent (se);
@@ -7873,6 +7899,11 @@ static unsigned long task_se_h_load(struct sched_entity *se) {return se->avg.load_avg; } + +static unsigned long task_se_h_weight (struct sched_entity *se) +{ +return se->load.weight; +}#endif /********** Helpers for find_busiest_group ************************/
Signature · 1 line
2.20.1
PATCH v4 11/15 sched,fair: flatten hierarchical runqueues
In reply to Rik van RielFlatten the hierarchical runqueues into just the per CPU rq.cfs runqueue.
Iteration of the sched_entity hierarchy is rate limited to once per jiffy per sched_entity, which is a smaller change than it seems, because load average adjustments were already rate limited to once per jiffy before this patch series.
This patch breaks CONFIG_CFS_BANDWIDTH. The plan for that is to park tasks from throttled cgroups onto their cgroup runqueues, and slowly (using the GENTLE_FAIR_SLEEPERS) wake them back up, in vruntime order, once the cgroup gets unthrottled, to prevent thundering herd issues.
Header from folded patch 'fix-attach-detach_enticy_cfs_rq.patch~':
Subject: sched,fair: fix attach/detach_entity_cfs_rq
While attach_entity_cfs_rq and detach_entity_cfs_rq should iterate over the hierarchy, they do not need to so that twice.
Passing flags into propagate_entity_cfs_rq allows us to reuse that same loop from other functions.
Header from folded patch 'enqueue-order.patch':
Subject: sched,fair: better ordering at enqueue_task_fair time
In order to get useful numbers for the task's hierarchical weight, task priority, etc things need to be done in a certain order at task enqueue time.
Specifically:
1) static load/weight to "local" cfs_rq
2) propagate load/weight up the tree
3) add runnable load avg to root cfs_rq
The reason is that each step depends on the things done by the step beforehand, and we can end up with nonsense numbers if we do not do things right.
Also, make sure that we walk all the way up the hierarchy at enqueue_task_fair time in order to get the benefit from the ramp-up logic in update_cfs_group.
--- include/linux/sched.h | 2 + kernel/sched/fair.c | 502 ++++++++++++++---------------------------- kernel/sched/pelt.c | 6 +- kernel/sched/pelt.h | 2 +- kernel/sched/sched.h | 2 +- 5 files changed, 170 insertions(+), 344 deletions(-)
include/linux/sched.h +2 -0
@@ -453,6 +453,8 @@ struct sched_entity {#ifdef CONFIG_FAIR_GROUP_SCHED int depth;unsigned long enqueued_h_load; +unsigned long enqueued_h_weight; +struct load_weight h_load;struct sched_entity *parent;/* rq on which this entity is (to be) queued: */ struct cfs_rq *cfs_rq;
kernel/sched/fair.c +163 -339
844 lines not on this page · raw message
kernel/sched/pelt.c +3 -3
@@ -251,10 +251,10 @@ int __update_load_avg_blocked_se(u64 now, struct sched_entity *se)return 0 ; } -int __update_load_avg_se (u64 now,struct cfs_rq *cfs_rq,struct sched_entity *se) +int __update_load_avg_se (u64 now,struct cfs_rq *cfs_rq,struct sched_entity *se,bool load,bool running) { -if (___update_load_sum (now, &se->avg, !!se->on_rq, - cfs_rq->curr == se)) { +if (___update_load_sum (now, &se->avg, (!!se->on_rq || load), + (cfs_rq->curr == se) || running)) {___update_load_avg (&se->avg,se_weight (se));cfs_se_util_change (&se->avg);
kernel/sched/pelt.h +1 -1
@@ -2,7 +2,7 @@#include "sched-pelt.h" int __update_load_avg_blocked_se (u64 now,struct sched_entity *se); -int __update_load_avg_se (u64 now,struct cfs_rq *cfs_rq,struct sched_entity *se); +int __update_load_avg_se (u64 now,struct cfs_rq *cfs_rq,struct sched_entity *se,bool load,bool running);int __update_load_avg_cfs_rq (u64 now,struct cfs_rq *cfs_rq);int update_rt_rq_load_avg (u64 now,struct rq *rq,int running);int update_dl_rq_load_avg (u64 now,struct rq *rq,int running);
kernel/sched/sched.h +1 -1
@@ -1443,7 +1443,7 @@ static inline void set_task_rq(struct task_struct *p, unsigned int cpu)#ifdef CONFIG_FAIR_GROUP_SCHED set_task_rq_fair (&p->se, p->se.cfs_rq, tg->cfs_rq[cpu]); - p->se.cfs_rq = tg->cfs_rq[cpu]; + p->se.cfs_rq = &cpu_rq (cpu)->cfs; p->se.parent = tg->se[cpu];#endif
Signature · 1 line
2.20.1
Dietmar Eggemann
reviewerRe: [PATCH 11/15] sched,fair: flatten hierarchical runqueues
Quoted 47 lines · Rik van Riel
Flatten the hierarchical runqueues into just the per CPU rq.cfs runqueue.
Iteration of the
sched_entityhierarchy is rate limited to once per jiffy persched_entity, which is a smaller change than it seems, because load average adjustments were already rate limited to once per jiffy before this patch series.This patch breaks
CONFIG_CFS_BANDWIDTH. The plan for that is to park tasks from throttled cgroups onto their cgroup runqueues, and slowly (using theGENTLE_FAIR_SLEEPERS) wake them back up, in vruntime order, once the cgroup gets unthrottled, to prevent thundering herd issues.Signed-off-by Rik van Riel <[email protected]>Header from folded patch 'fix-attach-detach_enticy_cfs_rq.patch~':
Subject: sched,fair: fix attach/detach_entity_cfs_rq
While
attach_entity_cfs_rqanddetach_entity_cfs_rqshould iterate over the hierarchy, they do not need to so that twice.Passing flags into
propagate_entity_cfs_rqallows us to reuse that same loop from other functions.Signed-off-by Rik van Riel <[email protected]>Header from folded patch 'enqueue-order.patch':
Subject: sched,fair: better ordering at
enqueue_task_fairtimeIn order to get useful numbers for the task's hierarchical weight, task priority, etc things need to be done in a certain order at task enqueue time.
Specifically:
1) static load/weight to "local"cfs_rq
2) propagate load/weight up the tree
3) add runnable load avg to rootcfs_rqThe reason is that each step depends on the things done by the step beforehand, and we can end up with nonsense numbers if we do not do things right.
Also, make sure that we walk all the way up the hierarchy at
enqueue_task_fairtime in order to get the benefit from the ramp-up logic inupdate_cfs_group.
[...]
Quoted 77 lines
/*
kernel/sched/fair.c +1
-60@@ -6953,7 +6849,6 @@ static void check_preempt_wakeup(struct rq *rq, struct task_struct *p, int wake_if (unlikely (p->policy !=SCHED_NORMAL ) || !sched_feat (WAKEUP_PREEMPTION ))return ; -find_matching_se (&se, &pse);update_curr (cfs_rq_of (se));BUG_ON (!pse);if (wakeup_preempt_entity (se, pse) ==1 ) {@@ -6994,100 +6889,18 @@ pick_next_task_fair(struct rq *rq, struct task_struct *prev, struct rq_flags *rfstruct task_struct *p;int new_tasks; +put_prev_task (rq, prev); again:if (!cfs_rq->nr_running) goto idle; -#ifdef CONFIG_FAIR_GROUP_SCHED -if (prev->sched_class != &fair_sched_class) - goto simple; - -/* -* Because of the set_next_buddy() in dequeue_task_fair() it is rather -* likely that a next task is from the same cgroup as the current. -* -* Therefore attempt to avoid putting and setting the entire cgroup -* hierarchy, only change the part that actually changes. -*/ - -do { -struct sched_entity *curr = cfs_rq->curr; - -/* -* Since we got here without doing put_prev_entity() we also -* have to consider cfs_rq->curr. If it is still a runnable -* entity, update_curr() will update its vruntime, otherwise -* forget we've ever seen it. -*/ -if (curr) { -if (curr->on_rq) -update_curr (cfs_rq); -else - curr =NULL ; - -/* -* This call to check_cfs_rq_runtime() will do the -* throttle and dequeue its entity in the parent(s). -* Therefore the nr_running test will indeed -* be correct. -*/ -if (unlikely (check_cfs_rq_runtime (cfs_rq))) { - cfs_rq = &rq->cfs; - -if (!cfs_rq->nr_running) - goto idle; - - goto simple; - } - } - - se =pick_next_entity (cfs_rq, curr); - cfs_rq =group_cfs_rq (se); - }while (cfs_rq); - - p =task_of (se); - -/* -* Since we haven't yet done put_prev_entity and if the selected task -* is a different task than we started out with, try and touch the -* least amount of cfs_rqs. -*/ -if (prev != p) { -struct sched_entity *pse = &prev->se; - -while (!(cfs_rq =is_same_group (se, pse))) { -int se_depth = se->depth;
kernel/sched/fair.c pick_next_task_fair(struct rq *rq, struct task_struct *prev, struct rq_flags *rf · line 7069
Quoted 10 lines
kernel/sched/fair.c +0
-10-int pse_depth = pse->depth; - -if (se_depth <= pse_depth) { -put_prev_entity (cfs_rq_of (pse), pse); - pse =parent_entity (pse); - } -if (se_depth >= pse_depth) { -set_next_entity (cfs_rq_of (se), se); - se =parent_entity (se); - }
Looks like with the se->depth related code gone here in pick_next_task_fair() and the call to find_matching_se() in check_preempt_wakeup() you could remove se->depth entirely.
[...]
Re: [PATCH 11/15] sched,fair: flatten hierarchical runqueues
Quoted 6 lines · Dietmar Eggemann
Looks like with the
se->depthrelated code gone here inpick_next_task_fair()and the call tofind_matching_se()incheck_preempt_wakeup()you could removese->depthentirely.[...]
I've just done that in a separate patch in my series, in case we need it again. If we don't, diffstat tells us we're getting this:
include/linux/sched.h | 1 - kernel/sched/fair.c | 50 ++----------------------------------------- ------- 2 files changed, 2 insertions(+), 49 deletions(-)
Thank you!
Signature · 1 line
All Rights Reversed.
PATCH v4 12/15 sched,fair: flatten update_curr functionality
In reply to Rik van RielMake it clear that update_curr only works on tasks any more.
There is no need for task_tick_fair to call it on every sched entity up the hierarchy, so move the call out of entity_tick.
--- kernel/sched/fair.c | 24 +++++++++++------------- 1 file changed, 11 insertions(+), 13 deletions(-)
kernel/sched/fair.c +11 -13
@@ -872,10 +872,11 @@ static void update_tg_load_avg(struct cfs_rq *cfs_rq, int force)static void update_curr (struct cfs_rq *cfs_rq) {struct sched_entity *curr = cfs_rq->curr; +struct task_struct *curtask;u64 now =rq_clock_task (rq_of (cfs_rq));u64 delta_exec; -if (unlikely (!curr)) +if (unlikely (!curr) || !entity_is_task (curr))return ; delta_exec = now - curr->exec_start;
@@ -893,13 +894,10 @@ static void update_curr(struct cfs_rq *cfs_rq) curr->vruntime +=calc_delta_fair (delta_exec, curr);update_min_vruntime (cfs_rq); -if (entity_is_task (curr)) { -struct task_struct *curtask =task_of (curr); - -trace_sched_stat_runtime (curtask, delta_exec, curr->vruntime); -cgroup_account_cputime (curtask, delta_exec); -account_group_exec_runtime (curtask, delta_exec); - } + curtask =task_of (curr); +trace_sched_stat_runtime (curtask, delta_exec, curr->vruntime); +cgroup_account_cputime (curtask, delta_exec); +account_group_exec_runtime (curtask, delta_exec);account_cfs_rq_runtime (cfs_rq, delta_exec); }
@@ -4196,11 +4194,6 @@ static void put_prev_entity(struct cfs_rq *cfs_rq, struct sched_entity *prev)static void entity_tick (struct cfs_rq *cfs_rq,struct sched_entity *curr,int queued) { -/* -* Update run-time statistics of the 'current'. -*/ -update_curr (cfs_rq); -/* * Ensure that runnable average is periodically updated. */
@@ -10032,6 +10025,11 @@ static void task_tick_fair(struct rq *rq, struct task_struct *curr, int queued)struct cfs_rq *cfs_rq;struct sched_entity *se = &curr->se; +/* +* Update run-time statistics of the 'current'. +*/ +update_curr (&rq->cfs); +for_each_sched_entity (se) { cfs_rq =group_cfs_rq_of_parent (se);entity_tick (cfs_rq, se, queued);
Signature · 1 line
2.20.1
Dietmar Eggemann
reviewerRe: [PATCH 12/15] sched,fair: flatten update_curr functionality
Quoted 16 lines · Rik van Riel
Make it clear that
update_curronly works on tasks any more.There is no need for
task_tick_fairto call it on every sched entity up the hierarchy, so move the call out ofentity_tick.Signed-off-by Rik van Riel <[email protected]>`Signed-off-by Rik van Riel <[email protected]>--- kernel/sched/fair.c | 24 +++++++++++------------- 1 file changed, 11 insertions(+), 13 deletions(-)kernel/sched/fair.c +0
-0
kernel/sched/fair.c static void update_tg_load_avg(struct cfs_rq *cfs_rq, int force) · line 880
Quoted 10 lines
kernel/sched/fair.c +2
-1@@ -872,10 +872,11 @@ static void update_tg_load_avg(struct cfs_rq *cfs_rq, int force)static void update_curr (struct cfs_rq *cfs_rq) {struct sched_entity *curr = cfs_rq->curr; +struct task_struct *curtask;u64 now =rq_clock_task (rq_of (cfs_rq));u64 delta_exec; -if (unlikely (!curr)) +if (unlikely (!curr) || !entity_is_task (curr))return ;
Shouldn't this be
- if (unlikely(!curr) || !entity_is_task(curr))
+ if (unlikely(!curr))
return;
+ BUG_ON(!entity_is_task(curr));
IMHO, cfs_rq->curr can only be a task with your changes.
PATCH v4 13/15 sched,fair: propagate sum_exec_runtime up the hierarchy
In reply to Rik van RielNow that enqueue_task_fair and dequeue_task_fair no longer iterate up the hierarchy all the time, a method to lazily propagate sum_exec_runtime up the hierarchy is necessary.
Once a tick, propagate the newly accumulated exec_runtime up the hierarchy, and feed it into CFS bandwidth control.
Remove the pointless call to account_cfs_rq_runtime from update_curr, which is always called with a root cfs_rq.
--- include/linux/sched.h | 1 + kernel/sched/core.c | 1 + kernel/sched/fair.c | 22 ++++++++++++++++++++-- 3 files changed, 22 insertions(+), 2 deletions(-)
include/linux/sched.h +1 -0
@@ -454,6 +454,7 @@ struct sched_entity {int depth;unsigned long enqueued_h_load;unsigned long enqueued_h_weight; +u64 propagated_exec_runtime;struct load_weight h_load;struct sched_entity *parent;/* rq on which this entity is (to be) queued: */
kernel/sched/core.c +1 -0
@@ -2137,6 +2137,7 @@ static void __sched_fork(unsigned long clone_flags, struct task_struct *p)INIT_LIST_HEAD (&p->se.group_node);#ifdef CONFIG_FAIR_GROUP_SCHED + p->se.propagated_exec_runtime =0 ; p->se.cfs_rq =NULL ;#endif
kernel/sched/fair.c +20 -2
@@ -898,8 +898,6 @@ static void update_curr(struct cfs_rq *cfs_rq)trace_sched_stat_runtime (curtask, delta_exec, curr->vruntime);cgroup_account_cputime (curtask, delta_exec);account_group_exec_runtime (curtask, delta_exec); - -account_cfs_rq_runtime (cfs_rq, delta_exec); }static void update_curr_fair (struct rq *rq)
@@ -3412,6 +3410,20 @@ static inline bool skip_blocked_update(struct sched_entity *se) return true; } +static void propagate_exec_runtime (struct cfs_rq *cfs_rq, +struct sched_entity *se) +{ +struct sched_entity *parent = se->parent; +u64 diff = se->sum_exec_runtime - se->propagated_exec_runtime; + +if (parent) { + parent->sum_exec_runtime += diff; +account_cfs_rq_runtime (cfs_rq, diff); + } + + se->propagated_exec_runtime = se->sum_exec_runtime; +} +#else /* CONFIG_FAIR_GROUP_SCHED */ static inline void update_tg_load_avg (struct cfs_rq *cfs_rq,int force) {}
@@ -3423,6 +3435,11 @@ static inline int propagate_entity_load_avg(struct sched_entity *se)static inline void add_tg_cfs_propagate (struct cfs_rq *cfs_rq,long runnable_sum) {} +static void propagate_exec_runtime (struct cfs_rq *cfs_rq, +struct sched_entity *se); +{ +} +#endif /* CONFIG_FAIR_GROUP_SCHED */ /**
@@ -10157,6 +10174,7 @@ static void propagate_entity_cfs_rq(struct sched_entity *se, int flags)if (!(flags &DO_ATTACH ))break ; +propagate_exec_runtime (cfs_rq, se);update_cfs_group (se); } }
Signature · 1 line
2.20.1
Dietmar Eggemann
reviewerRe: [PATCH 13/15] sched,fair: propagate sum_exec_runtime up the hierarchy
Quoted 9 lines · Rik van Riel
Now that
enqueue_task_fairanddequeue_task_fairno longer iterate up the hierarchy all the time, a method to lazily propagatesum_exec_runtimeup the hierarchy is necessary.Once a tick, propagate the newly accumulated
exec_runtimeup the hierarchy, and feed it into CFS bandwidth control.Remove the pointless call to
account_cfs_rq_runtimefromupdate_curr, which is always called with a rootcfs_rq.
But what about the call to account_cfs_rq_runtime() in set_curr_task_fair()? Here you always call it with the root cfs_rq. Shouldn't this be called also in a loop over all se's until !se->parent (like in propagate_exec_runtime() further below).
Quoted 86 lines
Signed-off-by Rik van Riel <[email protected]>--- include/linux/sched.h | 1 + kernel/sched/core.c | 1 + kernel/sched/fair.c | 22 ++++++++++++++++++++-- 3 files changed, 22 insertions(+), 2 deletions(-)include/linux/sched.h +1
-0@@ -454,6 +454,7 @@ struct sched_entity {int depth;unsigned long enqueued_h_load;unsigned long enqueued_h_weight; +u64 propagated_exec_runtime;struct load_weight h_load;struct sched_entity *parent;/* rq on which this entity is (to be) queued: */ kernel/sched/core.c +1
-0@@ -2137,6 +2137,7 @@ static void __sched_fork(unsigned long clone_flags, struct task_struct *p)INIT_LIST_HEAD (&p->se.group_node);#ifdef CONFIG_FAIR_GROUP_SCHED + p->se.propagated_exec_runtime =0 ; p->se.cfs_rq =NULL ;#endif kernel/sched/fair.c +20
-2@@ -898,8 +898,6 @@ static void update_curr(struct cfs_rq *cfs_rq)trace_sched_stat_runtime (curtask, delta_exec, curr->vruntime);cgroup_account_cputime (curtask, delta_exec);account_group_exec_runtime (curtask, delta_exec); - -account_cfs_rq_runtime (cfs_rq, delta_exec); }static void update_curr_fair (struct rq *rq)@@ -3412,6 +3410,20 @@ static inline bool skip_blocked_update(struct sched_entity *se) return true; } +static void propagate_exec_runtime (struct cfs_rq *cfs_rq, +struct sched_entity *se) +{ +struct sched_entity *parent = se->parent; +u64 diff = se->sum_exec_runtime - se->propagated_exec_runtime; + +if (parent) { + parent->sum_exec_runtime += diff; +account_cfs_rq_runtime (cfs_rq, diff); + } + + se->propagated_exec_runtime = se->sum_exec_runtime; +} +#else /* CONFIG_FAIR_GROUP_SCHED */ static inline void update_tg_load_avg (struct cfs_rq *cfs_rq,int force) {}@@ -3423,6 +3435,11 @@ static inline int propagate_entity_load_avg(struct sched_entity *se)static inline void add_tg_cfs_propagate (struct cfs_rq *cfs_rq,long runnable_sum) {} +static void propagate_exec_runtime (struct cfs_rq *cfs_rq, +struct sched_entity *se); +{ +} +#endif /* CONFIG_FAIR_GROUP_SCHED */ /** @@ -10157,6 +10174,7 @@ static void propagate_entity_cfs_rq(struct sched_entity *se, int flags)if (!(flags &DO_ATTACH ))break ; +propagate_exec_runtime (cfs_rq, se);update_cfs_group (se); } }
Re: [PATCH 13/15] sched,fair: propagate sum_exec_runtime up the hierarchy
Quoted 15 lines · Dietmar Eggemann
Quoted 6 lines · Rik van Riel
Now that
enqueue_task_fairanddequeue_task_fairno longer iterate up
the hierarchy all the time, a method to lazily propagatesum_exec_runtime
up the hierarchy is necessary.Quoted 7 lines
Once a tick, propagate the newly accumulated
exec_runtimeup the hierarchy,
and feed it into CFS bandwidth control.Remove the pointless call to
account_cfs_rq_runtimefromupdate_curr,
which is always called with a rootcfs_rq.
Quoted 5 lines
But what about the call to
account_cfs_rq_runtime()inset_curr_task_fair()? Here you always call it with the rootcfs_rq. Shouldn't this be called also in a loop over all se's until !se-Quoted 1 line
parent
(like in
propagate_exec_runtime()further below).
I believe that call should be only on the cgroup cfs_rq, with account_cfs_rq_runtime figuring out whether more runtime needs to be obtained from further up in the hierarchy.
By default we should probably work under the assumption that account_cfs_rq_runtime() will succeed at the current level, and no gymnastics are required to obtain CPU time.
Signature · 1 line
All Rights Reversed.
Dietmar Eggemann
reviewerRe: [PATCH 13/15] sched,fair: propagate sum_exec_runtime up the hierarchy
Quoted 19 lines · Rik van Riel
Quoted 18 lines · Dietmar Eggemann
Quoted 6 lines · Rik van Riel
Now that
enqueue_task_fairanddequeue_task_fairno longer iterate up
the hierarchy all the time, a method to lazily propagatesum_exec_runtime
up the hierarchy is necessary.Quoted 7 lines
Once a tick, propagate the newly accumulated
exec_runtimeup the hierarchy,
and feed it into CFS bandwidth control.Remove the pointless call to
account_cfs_rq_runtimefromupdate_curr,
which is always called with a rootcfs_rq.But what about the call to
account_cfs_rq_runtime()inset_curr_task_fair()? Here you always call it with the rootcfs_rq. Shouldn't this be called also in a loop over all se's until !se-
Quoted 7 lines
Quoted 2 lines
Quoted 1 line
parent
(like in
propagate_exec_runtime()further below).I believe that call should be only on the cgroup
cfs_rq, withaccount_cfs_rq_runtimefiguring out whether more runtime needs to be obtained from further up in the hierarchy.
So like this?
@@ -10248,7 +10248,8 @@ static void set_curr_task_fair(struct rq *rq)
set_next_entity(cfs_rq, se);
/* ensure bandwidth has been allocated on our new cfs_rq */- account_cfs_rq_runtime(cfs_rq, 0);
+ if (task_se_in_cgroup(se))
+ account_cfs_rq_runtime(group_cfs_rq_of_parent(se), 0);
}
I fail to understand the second part of your sentence, and how is this related to the code in propagate_exec_runtime():
for_each_sched_entity(se) {
propagate_exec_runtime() {
if (parent)
account_cfs_rq_runtime(cfs_rq, diff);
}
}Quoted 3 lines
By default we should probably work under the assumption that
account_cfs_rq_runtime()will succeed at the current level, and no gymnastics are required to obtain CPU time.
Maybe this all will become clearer when the reworked CFS Bandwidth support is ready ;-) I see this patch as the first part of it.
Re: [PATCH 13/15] sched,fair: propagate sum_exec_runtime up the hierarchy
Quoted 46 lines · Dietmar Eggemann
Quoted 22 lines · Rik van Riel
Quoted 21 lines · Dietmar Eggemann
Quoted 7 lines · Rik van Riel
Now that
enqueue_task_fairanddequeue_task_fairno longer iterate
up
the hierarchy all the time, a method to lazily propagatesum_exec_runtime
up the hierarchy is necessary.Quoted 8 lines
Once a tick, propagate the newly accumulated
exec_runtimeup the
hierarchy,
and feed it into CFS bandwidth control.Remove the pointless call to
account_cfs_rq_runtimefromupdate_curr,
which is always called with a rootcfs_rq.But what about the call to
account_cfs_rq_runtime()inset_curr_task_fair()? Here you always call it with the rootcfs_rq.
Shouldn't this be called also in a loop over all se's until !se-Quoted 7 lines
Quoted 2 lines
Quoted 1 line
parent
(like in
propagate_exec_runtime()further below).I believe that call should be only on the cgroup
cfs_rq, withaccount_cfs_rq_runtimefiguring out whether more runtime needs to be obtained from further up in the hierarchy.So like this?
(quoted diff) +2
-1@@ -10248,7 +10248,8 @@ static void set_curr_task_fair(struct rq *rq) set_next_entity(cfs_rq, se); /* ensure bandwidth has been allocated on our new cfs_rq */ - account_cfs_rq_runtime(cfs_rq, 0); + if (task_se_in_cgroup(se)) + account_cfs_rq_runtime(group_cfs_rq_of_parent(se), 0); }I fail to understand the second part of your sentence, and how is this related to the code in
propagate_exec_runtime():
Quoted 8 lines
for_each_sched_entity(se){propagate_exec_runtime() { if (parent) account_cfs_rq_runtime(cfs_rq, diff); } }
I am not sure how that would work for distributing runtime, since runtime would have to be distributed downwards and on demand, no?
That seems like a very different code path than "upwards, and periodically".
Then again, I have not worked out all the details of reimplementing CFS bandwidth yet...
Quoted 6 lines
Quoted 3 lines
By default we should probably work under the assumption that
account_cfs_rq_runtime()will succeed at the current level, and no gymnastics are required to obtain CPU time.Maybe this all will become clearer when the reworked CFS Bandwidth support is ready ;-) I see this patch as the first part of it.
That is one of the reasons I have not been "fixing" CFS bandwidth related code in the current patch series.
Having all of those changes in one location seems like it would be best.
Signature · 1 line
All Rights Reversed.
PATCH v4 14/15 sched,fair: ramp up task_se_h_weight quickly
In reply to Rik van RielThe code in update_cfs_group / calc_group_shares has some logic to quickly ramp up the load when a task has just started running in a cgroup, in order to get sane values for the cgroup se->load.weight.
This code adds a similar hack to task_se_h_weight.
However, THIS CODE IS WRONG, since it does not do things hierarchically.
I am wondering a few things here:
1) Should I have something similar to the logic in calc_group_shares in update_cfs_rq_h_load?
2) If so, should I also use that fast-ramp-up value for task_h_load, to prevent the load balancer from thinking it is moving zero weight tasks around?
3) If update_cfs_rq_h_load is the wrong place, where should I be calculating a hierarchical group weight value, instead?
--- kernel/sched/fair.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
kernel/sched/fair.c +6 -1
@@ -7672,6 +7672,7 @@ static void update_cfs_rq_h_load(struct cfs_rq *cfs_rq)static unsigned long task_se_h_weight (struct sched_entity *se) { +unsigned long group_load;struct cfs_rq *cfs_rq;if (!task_se_in_cgroup (se))
@@ -7680,8 +7681,12 @@ static unsigned long task_se_h_weight(struct sched_entity *se) cfs_rq =group_cfs_rq_of_parent (se);update_cfs_rq_h_load (cfs_rq); +/* Ramp up quickly to keep h_weight sane. */ + group_load =max (scale_load_down (se->parent->load.weight), + cfs_rq->h_load); +/* Reduce the load.weight by the h_load of the group the task is in. */ -return (cfs_rq->h_load * se->load.weight) >>SCHED_FIXEDPOINT_SHIFT ; +return (group_load * se->load.weight) >>SCHED_FIXEDPOINT_SHIFT ; }static unsigned long task_se_h_load (struct sched_entity *se)
Signature · 1 line
2.20.1
PATCH v4 15/15 sched,fair: scale vdiff in wakeup_preempt_entity
In reply to Rik van RielWhen a task wakes back up after having gone to sleep, place_entity will limit the vruntime difference between min_vruntime and the woken up task to half of sysctl_sched_latency.
The code in wakeup_preempt_entity calculates how much vruntime a time slice for the woken up task represents, in wakeup_gran.
It then assumes that all the vruntime used since the task went to sleep was used by the currently running task (which has its vruntime scaled by calc_delta_fair, as well).
However, that assumption is not necessarily true, and the vruntime may have advanced at different rates, pushed ahead by different tasks on the CPU. This becomes more visible when the CPU controller is enabled.
This leads to the symptom that a high priority woken up task is likely to preempt whatever is running, even if the currently running task is of equal or higher priority than the woken up task!
Scaling the vdiff down if the currently running task is also high priority solves that symptom.
This is not the correct thing to do if all of the vruntime was accumulated by the current task, or other tasks at similar priority, and already scaled by the same priority, but I do not have any better ideas on how to tackle the "task X got preempted by task Y of the same priority" issue that system administrators try to resolve by setting the sched_wakeup_granularity sysctl variable to a larger value than half of sysctl_sched_latency...
--- kernel/sched/fair.c | 1 + 1 file changed, 1 insertion(+)
kernel/sched/fair.c +1 -0
@@ -6774,6 +6774,7 @@ wakeup_preempt_entity(struct sched_entity *curr, struct sched_entity *se)if (vdiff <=0 )return -1 ; + vdiff =min ((u64 )vdiff,calc_delta_fair (vdiff, curr)); gran =wakeup_gran (se);if (vdiff > gran)return 1 ;
Signature · 1 line
2.20.1
Dietmar Eggemann
reviewerQuoted 8 lines · Rik van Riel
The current implementation of the CPU controller uses hierarchical runqueues, where on wakeup a task is enqueued on its group's runqueue, the group is enqueued on the runqueue of the group above it, etc.
This increases a fairly large amount of overhead for workloads that do a lot of wakeups a second, especially given that the default systemd hierarchy is 2 or 3 levels deep.
Quoted 8 lines
This patch series is an attempt at reducing that overhead, by placing all the tasks on the same runqueue, and scaling the task priority by the priority of the group, which is calculated periodically.
My main TODO items for the next period of time are likely going to be testing, testing, and testing. I hope to find and flush out any corner case I can find, and make sure performance does not regress with any workloads, and hopefully improves some.
I did some testing with a small & simple rt-app based test-case:
2 CPUs (rq->cpu_capacity_orig=1024), CPUfreq performance governor
2 taskgroups /tg0 and /tg1
6 CFS tasks (periodic, 8/16ms (runtime/period))
/tg0 (cpu.shares=1024) ran 4 tasks and /tg1 (cpu.shares=1024) ran 2 tasks
(arm64 defconfig with !CONFIG_NUMA_BALANCING, !CONFIG_SCHED_AUTOGROUP)
---
v5.2:
The 2 /tg1 tasks ran 8/16ms. The 4 /tg0 tasks ran 4/16ms in the beginning and then 8/16ms after the 2 /tg1 tasks did finish.
---
v5.2 + v4:
There is no runtime/period pattern visible anymore. I see a lot of extra wakeup latency for those tasks though.
v5.2 + (v4 without 07/15, 08/15, 15/15) didn't change much.
---
I could try to reduce the stack even further (e.g. without 13/15).
IMHO it's a good idea to have a set of these small & simple test-cases handy to verify that the base-functionality is still in place. This might be hard to achieve with benchmarks.
Quoted 30 lines · Dietmar Eggemann
Quoted 4 lines · Rik van Riel
My main TODO items for the next period of time are likely going to be testing, testing, and testing. I hope to find and flush out any corner case I can find, and make sure performance does not regress with any workloads, and hopefully improves some.
I did some testing with a small & simple rt-app based test-case:
2 CPUs (rq->cpu_capacity_orig=1024), CPUfreq performance governor
2 taskgroups /tg0 and /tg1
6 CFS tasks (periodic, 8/16ms (runtime/period))
/tg0 (cpu.shares=1024) ran 4 tasks and /tg1 (cpu.shares=1024) ran 2 tasks
(arm64 defconfig with !CONFIG_NUMA_BALANCING,
!CONFIG_SCHED_AUTOGROUP)---
v5.2:
The 2 /tg1 tasks ran 8/16ms. The 4 /tg0 tasks ran 4/16ms in the beginning and then 8/16ms after the 2 /tg1 tasks did finish.
---
Quoted 7 lines
v5.2 + v4:
There is no runtime/period pattern visible anymore. I see a lot of extra
wakeup latency for those tasks though.v5.2 + (v4 without 07/15, 08/15, 15/15) didn't change much.
One thing to keep in mind is that with the hierarchical CPU controller code, you are always either comparing tg0 and tg1 (of equal priority), or tasks of equal priority, once the load balancer has equalized the load between CPUs.
With the flat CPU controller, the preemption code is comparing sched_entities with other sched_entities that have 2x the priority, similar to a nice level 0 entity compared against a nice level ~3 task.
I do not know whether the code has ever given a predictable scheduling pattern when the CPU is fully loaded with a mix of different priority tasks that each want a 50% duty cycle.
But maybe it has, and I need to look into that :)
Figuring out exactly what the preemption code should do might be a good discussion for Plumbers, too.
Signature · 1 line
All Rights Reversed.