Linux 调度器调频学习笔记_操作系统

Linux5.4 Qcom平台

1. per-cpu的 update_util_data 实例是调度器与schedutil调频驱动沟通的桥梁，cpufreq_update_util()函数中访问里面的回调函数进行调频。

2. 触发调频的时机

(1) enqueue_task_fair 时

activate_task 
move_queued_task 
do_set_cpus_allowed 
rt_mutex_setprio 
set_user_nice 
__sched_setscheduler 
sched_setnuma 
sched_move_task 
enqueue_task_core 
    enqueue_task --> enqueue_task_fair --> if (p->in_iowait) cpufreq_update_util(rq, SCHED_CPUFREQ_IOWAIT);

(2) pick_next_task_fair时

balance_fair 
pick_next_task_fair 
    newidle_balance --> nohz_newidle_balance --> _nohz_idle_balance --> rebalance_domains --> load_balance --> find_busiest_group --> update_sd_lb_stats --> update_sg_lb_stats --> update_nohz_stats --> update_blocked_averages --> cpufreq_update_util(rq, 0); 
    newidle_balance --> nohz_newidle_balance --> _nohz_idle_balance --> update_nohz_stats --> update_blocked_averages --> cpufreq_update_util(rq, 0); 
    newidle_balance --> load_balance --> find_busiest_group --> update_sd_lb_stats --> update_sg_lb_stats --> update_nohz_stats --> update_blocked_averages --> cpufreq_update_util(rq, 0);

(3) scheduler_tick时

scheduler_tick() --> trigger_load_balance //只有在 scheduler_tick()中通过 SCHED_SOFTIRQ 触发负载均衡。 
__init init_sched_fair_class --> open_softirq(SCHED_SOFTIRQ, run_rebalance_domains); 
trigger_load_balance --> raise_softirq(SCHED_SOFTIRQ); 
    run_rebalance_domains --> nohz_idle_balance --> _nohz_idle_balance --> update_nohz_stats --> update_blocked_averages --> cpufreq_update_util(rq, 0); 
    run_rebalance_domains --> nohz_idle_balance --> _nohz_idle_balance --> rebalance_domains --> load_balance --> find_busiest_group --> update_sd_lb_stats --> update_sg_lb_stats --> update_nohz_stats --> update_blocked_averages --> cpufreq_update_util(rq, 0); 
    run_rebalance_domains --> update_blocked_averages --> cpufreq_update_util(rq, 0); 
    run_rebalance_domains --> rebalance_domains --> load_balance --> find_busiest_group --> update_sd_lb_stats --> update_sg_lb_stats --> update_nohz_stats --> update_blocked_averages --> cpufreq_update_util(rq, 0);

(4) switched_to回调

switched_to_fair --> attach_task_cfs_rq --> attach_entity_cfs_rq --> attach_entity_load_avg --> cfs_rq_util_change(cfs_rq, flags) --> cpufreq_update_util(rq, flags)

(5) task_change_group回调

task_change_group_fair --> task_move_group_fair --> attach_task_cfs_rq --> attach_entity_cfs_rq --> attach_entity_load_avg --> cfs_rq_util_change(cfs_rq, flags) --> cpufreq_update_util(rq, flags)

(6) 唤醒进程时

各种锁，进程间通信机制调用wake_up_q 
_do_fork 
    wake_up_new_task --> post_init_entity_util_avg --> attach_entity_cfs_rq --> attach_entity_load_avg --> cfs_rq_util_change(cfs_rq, flags) --> cpufreq_update_util(rq, flags)

(7) sched_online_group调用时

sched_online_group --> online_fair_sched_group --> attach_entity_cfs_rq --> attach_entity_load_avg --> cfs_rq_util_change(cfs_rq, flags) --> cpufreq_update_util(rq, flags)

(8) migrate_task_rq回调时

migrate_task_rq_fair --> detach_entity_cfs_rq --> detach_entity_load_avg --> cfs_rq_util_change(cfs_rq, flags) --> cpufreq_update_util(rq, flags)

(9) switched_from回调时

switched_from_fair --> detach_task_cfs_rq --> detach_entity_cfs_rq --> detach_entity_load_avg --> cfs_rq_util_change(cfs_rq, flags) --> cpufreq_update_util(rq, flags)

(10) switched_to回调时

switched_to_fair --> detach_task_cfs_rq --> detach_entity_cfs_rq --> detach_entity_load_avg --> cfs_rq_util_change(cfs_rq, flags) --> cpufreq_update_util(rq, flags)

(11) 往rq中添加/移除任务时

enqueue_entity 
dequeue_entity 
set_next_entity 
put_prev_entity 
entity_tick 
enqueue_task_fair 
dequeue_task_fair 
__update_blocked_fair 
propagate_entity_cfs_rq 
detach_entity_cfs_rq 
attach_entity_cfs_rq 
sched_group_set_shares 
    update_load_avg --> cfs_rq_util_change(cfs_rq, flags) --> cpufreq_update_util(rq, flags) 
    update_load_avg --> attach_entity_load_avg --> cfs_rq_util_change(cfs_rq, flags) --> cpufreq_update_util(rq, flags)

(12) scheduler_tick 中直接调用调频

scheduler_tick --> cpufreq_update_util

3. 以调度节拍为例介绍调频

(1) 调用路径

时钟中断 --> scheduler_tick --> curr->sched_class->task_tick(CFS：task_tick_fair) --> entity_tick --> update_load_avg --> cfs_rq_util_change --> cpufreq_update_util(rq, flags)

(2) cpufreq_update_util(rq, flags)被Qcom更改为只接受flag中设置SCHED_CPUFREQ_WALT标志的调频了。

static inline void cpufreq_update_util(struct rq *rq, unsigned int flags) 
{ 
    struct update_util_data *data; 
    u64 clock; 
 
#ifdef CONFIG_SCHED_WALT 
    if (!(flags & SCHED_CPUFREQ_WALT)) /*变成只接受设置有WALT标志位的调频*/ 
        return; 
    clock = sched_ktime_clock(); 
#else 
    clock = rq_clock(rq); 
#endif 
 
    data = rcu_dereference_sched(*per_cpu_ptr(&cpufreq_update_util_data, cpu_of(rq))); 
    if (data) 
        data->func(data, clock, flags); 
}

(3) cpufreq_update_util(rq, flags) 中flags支持的调频标志位

//include\linux\sched\cpufreq.h 
#define SCHED_CPUFREQ_IOWAIT            (1U << 0) 
#define SCHED_CPUFREQ_MIGRATION            (1U << 1) 
#define SCHED_CPUFREQ_INTERCLUSTER_MIG    (1U << 3) 
#define SCHED_CPUFREQ_WALT                (1U << 4) 
#define SCHED_CPUFREQ_PL                (1U << 5) 
#define SCHED_CPUFREQ_EARLY_DET            (1U << 6) 
#define SCHED_CPUFREQ_CONTINUE            (1U << 8)

4. cpufreq还与thermal的cooldown有关联

struct cpufreq_policy { 
    ... 
    struct thermal_cooling_device *cdev 
};

5. governor/driver/policy分别是

# cat /sys/devices/system/cpu/cpufreq/policy0/scaling_driver  //driver 
qcom-cpufreq-hw 
# cat /sys/devices/system/cpu/cpufreq/policy0/scaling_governor //governor 
schedutil 
# ls /sys/devices/system/cpu/cpufreq/policy0/schedutil //policy 
above_hispeed_delay  down_rate_limit_us  hispeed_freq  hispeed_load  pl  rtg_boost_freq  target_loads  up_rate_limit_us

本文参考链接：https://www.cnblogs.com/hellokitty2/p/14909197.html