参考文章：
https://www.jb51.net/article/133579.htm
https://blog.csdn.net/flaoter/article/details/77509553
https://www.cnblogs.com/arnoldlu/p/7078204.html 中时间子系统相关系列blog，讲的比较详细。
主要文件所在目录：
kernel/msm-4.9/kernel/time/tick-common.c、tick-dchrf.c、timer.c、hrtimer.c、clockevent.c等相关源文件
kernel/msm-4.9/drivers/clocksource.c
kernel/msm-4.9/drivers/clocksource文件夹下的一些源文件
kernel/sched/core.c
简述
作为进程调度中，最关键的函数：scheduler_tick()。它也是大多数调度函数的源，那么它自身又被谁调用的呢？
scheduler_tick()是所有调度子函数的父函数，而其是由Linux时间子系统的tick_device调用。tick_device是一个周期性定时器，定时时间为1个tick，当触发中断后，会在中断处理函数中，调用scheduler_tick()。
而打开了tickless，即动态tick后，那么就会切换至oneshot模式，并负责调用scheduler_tick()。
这篇文章会简要地解释这个原理。
NO_HZ动态时钟 & hrtimer高精度时钟
因为是tick是由tick device周期性触发，所以当系统在idle时，为了减少系统功耗，应该关闭周期性tick。所以，NO_HZ的动态时钟应运而生。它会在系统空闲，仅有idle进程时，关闭周期性tick；而当跳出idle进程时，会重新再开启周期性tick。
关于NO_HZ的详细资料，可以参考：https://www.kernel.org/doc/Documentation/timers/NO_HZ.txt
SDM845平台的时间子系统就是基于NO_HZ动态时钟， 以及高精度定时时钟的，对应.config配置如下：
# # Timers subsystem # CONFIG_TICK_ONESHOT=y CONFIG_NO_HZ_COMMON=y # CONFIG_HZ_PERIODIC is not set CONFIG_NO_HZ_IDLE=y # CONFIG_NO_HZ_FULL is not set CONFIG_NO_HZ=y CONFIG_HIGH_RES_TIMERS=y
－－－CONFIG_NO_HZ_IDLE(处于idle状态没有tick，非idle状态正常tick)------------当前平台处于这个config
－－－CONFIG_NO_HZ_FULL(处于idle或者cpu仅有一个进程运行，停止tick；其他情况正常)
NO_HZ的情况下，有3种模式：
- 系统动态时钟尚未激活模式
- 高精度工作模式
- 低精度工作模式
enum tick_nohz_mode { NOHZ_MODE_INACTIVE, NOHZ_MODE_LOWRES, NOHZ_MODE_HIGHRES, }; /** * struct tick_sched - sched tick emulation and no idle tick control/stats * @sched_timer: hrtimer to schedule the periodic tick in high * resolution mode * @last_tick: Store the last tick expiry time when the tick * timer is modified for nohz sleeps. This is necessary * to resume the tick timer operation in the timeline * when the CPU returns from nohz sleep. * @tick_stopped: Indicator that the idle tick has been stopped * @idle_jiffies: jiffies at the entry to idle for idle time accounting * @idle_calls: Total number of idle calls * @idle_sleeps: Number of idle calls, where the sched tick was stopped * @idle_entrytime: Time when the idle call was entered * @idle_waketime: Time when the idle was interrupted * @idle_exittime: Time when the idle state was left * @idle_sleeptime: Sum of the time slept in idle with sched tick stopped * @iowait_sleeptime: Sum of the time slept in idle with sched tick stopped, with IO outstanding * @sleep_length: Duration of the current idle sleep * @do_timer_lst: CPU was the last one doing do_timer before going idle */ struct tick_sched { struct hrtimer sched_timer; unsigned long check_clocks; enum tick_nohz_mode nohz_mode; ktime_t last_tick; int inidle; int tick_stopped; unsigned long idle_jiffies; unsigned long idle_calls; unsigned long idle_sleeps; int idle_active; ktime_t idle_entrytime; ktime_t idle_waketime; ktime_t idle_exittime; ktime_t idle_sleeptime; ktime_t iowait_sleeptime; ktime_t sleep_length; unsigned long last_jiffies; u64 next_timer; ktime_t idle_expires; int do_timer_last; atomic_t tick_dep_mask; };
tick_device
tick_device相关数据结构如下：
1、tick_device的工作模式，支持两种：一种是周期性periodic，另一种是一次性oneshot。
enum tick_device_mode { TICKDEV_MODE_PERIODIC, TICKDEV_MODE_ONESHOT, }; struct tick_device { struct clock_event_device *evtdev; enum tick_device_mode mode; };
tick device是通过tick_check_new_device函数进行创建
/* * Check, if the new registered device should be used. Called with * clockevents_lock held and interrupts disabled. */ void tick_check_new_device(struct clock_event_device *newdev) { struct clock_event_device *curdev; struct tick_device *td; int cpu; cpu = smp_processor_id(); //获取当前CPU id td = &per_cpu(tick_cpu_device, cpu); //获取当前CPU的tick device结构体 curdev = td->evtdev; /* cpu local device ? */ if (!tick_check_percpu(curdev, newdev, cpu)) //判断是否是只服务local CPU，否则就会注册broadcast，走下面的分支。 goto out_bc; /* Preference decision */ if (!tick_check_preferred(curdev, newdev)) //如果是onshot模式，并且已有一个tick device，那么就选用其中高rate的。但是如果高rate的是non-CPU local device，那仍然会选用低rate的local tick device goto out_bc; if (!try_module_get(newdev->owner)) return; /* * Replace the eventually existing device by the new * device. If the current device is the broadcast device, do * not give it back to the clockevents layer ! */ if (tick_is_broadcast_device(curdev)) { clockevents_shutdown(curdev); curdev = NULL; } clockevents_exchange_device(curdev, newdev); //更新clock_event_device tick_setup_device(td, newdev, cpu, cpumask_of(cpu)); //setup，下面详细解析 if (newdev->features & CLOCK_EVT_FEAT_ONESHOT) tick_oneshot_notify(); return; out_bc: /* * Can the new device be used as a broadcast device ? */ tick_install_broadcast_device(newdev); }
而在tick_setup_device函数，会进一步初始化tick device。如果是第一次setup，那么模式只能是periodic周期性的tick device。
/* * Setup the tick device */ static void tick_setup_device(struct tick_device *td, struct clock_event_device *newdev, int cpu, const struct cpumask *cpumask) { ktime_t next_event; void (*handler)(struct clock_event_device *) = NULL; /* * First device setup ? */ if (!td->evtdev) { //如果是当前CPU第一个注册的tick device /* * If no cpu took the do_timer update, assign it to * this cpu: */ if (tick_do_timer_cpu == TICK_DO_TIMER_BOOT) { //此tick device将会让其管理全局jiffies等时间信息 if (!tick_nohz_full_cpu(cpu)) tick_do_timer_cpu = cpu; else tick_do_timer_cpu = TICK_DO_TIMER_NONE; tick_next_period = ktime_get(); tick_period = ktime_set(0, NSEC_PER_SEC / HZ); //HZ为1秒内需要有多少的脉冲，基于此来设定定时时间 } /* * Startup in periodic mode first. */ td->mode = TICKDEV_MODE_PERIODIC; //当前cpu第一次设定tick device的时候，缺省设定为周期性的tick } else { handler = td->evtdev->event_handler; next_event = td->evtdev->next_event; td->evtdev->event_handler = clockevents_handle_noop; } td->evtdev = newdev; //将系统clock_event_device赋值给对应tick device的evtdev指针。这是比较关键的一步，代表了tick device找到合适挂载 /* * When the device is not per cpu, pin the interrupt to the * current cpu: */ if (!cpumask_equal(newdev->cpumask, cpumask)) irq_set_affinity(newdev->irq, cpumask); /* * When global broadcasting is active, check if the current * device is registered as a placeholder for broadcast mode. * This allows us to handle this x86 misfeature in a generic * way. This function also returns !=0 when we keep the * current active broadcast state for this CPU. */ if (tick_device_uses_broadcast(newdev, cpu)) return; if (td->mode == TICKDEV_MODE_PERIODIC) tick_setup_periodic(newdev, 0); else tick_setup_oneshot(newdev, handler, next_event); }
在tick_setup_periodic中，
先设置中断handler，再开启定时器。
/* * Setup the device for a periodic tick */ void tick_setup_periodic(struct clock_event_device *dev, int broadcast) { tick_set_periodic_handler(dev, broadcast); //设置handler /* Broadcast setup ? */ if (!tick_device_is_functional(dev)) return; if ((dev->features & CLOCK_EVT_FEAT_PERIODIC) && !tick_broadcast_oneshot_active()) { clockevents_switch_state(dev, CLOCK_EVT_STATE_PERIODIC); //设置clock工作状态 } else { unsigned long seq; ktime_t next; do { seq = read_seqbegin(&jiffies_lock); next = tick_next_period; } while (read_seqretry(&jiffies_lock, seq)); clockevents_switch_state(dev, CLOCK_EVT_STATE_ONESHOT); for (;;) { if (!clockevents_program_event(dev, next, false)) return; next = ktime_add(next, tick_period); } } }
中断handler，使用的是不支持broadcast的。
/* * Set the periodic handler depending on broadcast on/off */ void tick_set_periodic_handler(struct clock_event_device *dev, int broadcast) { if (!broadcast) dev->event_handler = tick_handle_periodic; //非broadcast else dev->event_handler = tick_handle_periodic_broadcast; }
/** * clockevents_switch_state - set the operating state of a clock event device * @dev: device to modify * @state: new state * * Must be called with interrupts disabled ! */ void clockevents_switch_state(struct clock_event_device *dev, enum clock_event_state state) { if (clockevent_get_state(dev) != state) { if (__clockevents_switch_state(dev, state)) //设置工作状态 return; clockevent_set_state(dev, state); /* * A nsec2cyc multiplicator of 0 is invalid and we'd crash * on it, so fix it up and emit a warning: */ if (clockevent_state_oneshot(dev)) { if (unlikely(!dev->mult)) { dev->mult = 1; WARN_ON(1); } } } }
最后会调用device特定的periodic工作函数
static int __clockevents_switch_state(struct clock_event_device *dev, enum clock_event_state state) { if (dev->features & CLOCK_EVT_FEAT_DUMMY) return 0; /* Transition with new state-specific callbacks */ switch (state) { case CLOCK_EVT_STATE_DETACHED: /* The clockevent device is getting replaced. Shut it down. */ case CLOCK_EVT_STATE_SHUTDOWN: if (dev->set_state_shutdown) return dev->set_state_shutdown(dev); return 0; case CLOCK_EVT_STATE_PERIODIC: /* Core internal bug */ if (!(dev->features & CLOCK_EVT_FEAT_PERIODIC)) return -ENOSYS; if (dev->set_state_periodic) return dev->set_state_periodic(dev); //调用device set_state_periodic工作函数，但实际当前平台没有这个函数。直接return 0 return 0; case CLOCK_EVT_STATE_ONESHOT: /* Core internal bug */ if (!(dev->features & CLOCK_EVT_FEAT_ONESHOT)) return -ENOSYS; if (dev->set_state_oneshot) return dev->set_state_oneshot(dev); return 0; case CLOCK_EVT_STATE_ONESHOT_STOPPED: /* Core internal bug */ if (WARN_ONCE(!clockevent_state_oneshot(dev), "Current state: %d ", clockevent_get_state(dev))) return -EINVAL; if (dev->set_state_oneshot_stopped) return dev->set_state_oneshot_stopped(dev); else return -ENOSYS; default: return -ENOSYS; } }
上面提到set_state_periodic并未定义。那么应该走到哪里呢？我们看前面在 tick_setup_periodic 函数中，会判断：
if ((dev->features & CLOCK_EVT_FEAT_PERIODIC) &&
!tick_broadcast_oneshot_active())
实际当前平台的clock event device不支持CLOCK_EVT_FEAT_PERIODIC模式，所以代码会走到else中，模拟周期tick：
if ((dev->features & CLOCK_EVT_FEAT_PERIODIC) && !tick_broadcast_oneshot_active()) { clockevents_switch_state(dev, CLOCK_EVT_STATE_PERIODIC); } else { //走到else中，模拟周期tick unsigned long seq; ktime_t next; do { seq = read_seqbegin(&jiffies_lock); next = tick_next_period; } while (read_seqretry(&jiffies_lock, seq)); clockevents_switch_state(dev, CLOCK_EVT_STATE_ONESHOT); //这里同样因为没有set_state_onshot的接口函数，直接return 0。 for (;;) { if (!clockevents_program_event(dev, next, false)) //这里是一个无限循环，但是正常情况下，由于return 0，直接跳出循环 return; next = ktime_add(next, tick_period); } }
/** * clockevents_program_event - Reprogram the clock event device. * @dev: device to program * @expires: absolute expiry time (monotonic clock) * @force: program minimum delay if expires can not be set * * Returns 0 on success, -ETIME when the event is in the past. */ int clockevents_program_event(struct clock_event_device *dev, ktime_t expires, bool force) { unsigned long long clc; int64_t delta; int rc; if (unlikely(expires.tv64 < 0)) { WARN_ON_ONCE(1); return -ETIME; } dev->next_event = expires; if (clockevent_state_shutdown(dev)) return 0; /* We must be in ONESHOT state here */ WARN_ONCE(!clockevent_state_oneshot(dev), "Current state: %d ", clockevent_get_state(dev)); /* Shortcut for clockevent devices that can deal with ktime. */ if (dev->features & CLOCK_EVT_FEAT_KTIME) return dev->set_next_ktime(expires, dev); delta = ktime_to_ns(ktime_sub(expires, ktime_get())); if (delta <= 0) return force ? clockevents_program_min_delta(dev) : -ETIME; delta = min(delta, (int64_t) dev->max_delta_ns); delta = max(delta, (int64_t) dev->min_delta_ns); clc = ((unsigned long long) delta * dev->mult) >> dev->shift; rc = dev->set_next_event((unsigned long) clc, dev); return (rc && force) ? clockevents_program_min_delta(dev) : rc; }
/** * clockevents_program_min_delta - Set clock event device to the minimum delay. * @dev: device to program * * Returns 0 on success, -ETIME when the retry loop failed. */ static int clockevents_program_min_delta(struct clock_event_device *dev) { unsigned long long clc; int64_t delta; int i; for (i = 0;;) { delta = dev->min_delta_ns; dev->next_event = ktime_add_ns(ktime_get(), delta); if (clockevent_state_shutdown(dev)) return 0; dev->retries++; clc = ((unsigned long long) delta * dev->mult) >> dev->shift; if (dev->set_next_event((unsigned long) clc, dev) == 0) return 0; if (++i > 2) { /* * We tried 3 times to program the device with the * given min_delta_ns. Try to increase the minimum * delta, if that fails as well get out of here. */ if (clockevents_increase_min_delta(dev)) return -ETIME; i = 0; } } }
当触发了定时器之后，会调用终端handler函数：tick_handle_periodic，函数下半部分会重新设置timer触发的时间。下次触发，仍然进入中断handler函数。如此往复，模拟周期tick。
/* * Event handler for periodic ticks */ void tick_handle_periodic(struct clock_event_device *dev) { int cpu = smp_processor_id(); ktime_t next = dev->next_event; tick_periodic(cpu); //更新wall time等操作，调用update_process() #if defined(CONFIG_HIGH_RES_TIMERS) || defined(CONFIG_NO_HZ_COMMON) /* * The cpu might have transitioned to HIGHRES or NOHZ mode via * update_process_times() -> run_local_timers() -> * hrtimer_run_queues(). */ if (dev->event_handler != tick_handle_periodic) return; #endif if (!clockevent_state_oneshot(dev)) return; for (;;) { /* * Setup the next period for devices, which do not have * periodic mode: */ next = ktime_add(next, tick_period); if (!clockevents_program_event(dev, next, false)) return; /* * Have to be careful here. If we're in oneshot mode, * before we call tick_periodic() in a loop, we need * to be sure we're using a real hardware clocksource. * Otherwise we could get trapped in an infinite * loop, as the tick_periodic() increments jiffies, * which then will increment time, possibly causing * the loop to trigger again and again. */ if (timekeeping_valid_for_hres()) tick_periodic(cpu); } }
在tick_periodic中，
/* * Periodic tick */ static void tick_periodic(int cpu) { if (tick_do_timer_cpu == cpu) { write_seqlock(&jiffies_lock); /* Keep track of the next tick event */ tick_next_period = ktime_add(tick_next_period, tick_period); do_timer(1); write_sequnlock(&jiffies_lock); update_wall_time(); //更新wall time } update_process_times(user_mode(get_irq_regs())); profile_tick(CPU_PROFILING); //代码采集器 }
/* * Called from the timer interrupt handler to charge one tick to the current * process. user_tick is 1 if the tick is user time, 0 for system. */ void update_process_times(int user_tick) { struct task_struct *p = current; /* Note: this timer irq context must be accounted for as well. */ account_process_tick(p, user_tick); run_local_timers(); rcu_check_callbacks(user_tick); #ifdef CONFIG_IRQ_WORK if (in_irq()) irq_work_tick(); #endif scheduler_tick(); //调用scheculer_tick() run_posix_cpu_timers(p); }
关于代码采集器profile_tick的简要知识：
profile_tick()函数为代码监管器采集数据。这个函数在单处理器系统上是由do_timer_interrupt()调用的（即全局时钟中断处理程序调用的），在多处理器系统上是由smp_local_timer_interrupt()函数调用的（即本地时钟中断处理程序调用的） 为了激活代码监管器，在Linux内核启动时必须传递字符串参数"profile=N" ,这里2的N次方，表示要监管的代码段的大小。采集的数据可以从/proc/profile文件中读取。可以通过修改这个文件来重置计数器；在多处理器系统上，修改这个文件还可以改变抽样频率。不过，内核开发者并不直接访问/proc/profile文件，而是用readprofile系统命令 Linux2.6内核还包含了另一个监管器，叫做oprofile .比起readprofile,oprofile除了更灵活、更可定制外，还能用于发现内核代码、用户态应用程序以及系统库中的热点。当使用oprofile时，profile_tick()调用timer_notify()函数来收集这个新监管器所使用的数据。
回归原题，scheduler_tick()具体被调用流程：tick中断->tick_periodic()->update_process_times()->scheduler_tick()或者tick中断->tick_sched_handle()->update_process_times()->scheduler_tick()。本文分析了前者，后者有兴趣可以自行读代码了解。
下面为补充linux时间子系统的相关知识，dev这个结构体是如何初始化，并填充的（包括clock_event_device->feature在哪里定义为CLOCK_EVT_FEAT_ONESHOT）。
clock source & clock event device
我们可以看到是在DTS中有2个timer配置。
DTS配置：
timer { compatible = "arm,armv8-timer"; interrupts = <1 1 0xf08>, <1 2 0xf08>, <1 3 0xf08>, <1 0 0xf08>; clock-frequency = <19200000>; }; timer@0x17C90000{ #address-cells = <1>; #size-cells = <1>; ranges; compatible = "arm,armv7-timer-mem"; reg = <0x17C90000 0x1000>; clock-frequency = <19200000>; ..... };
./kernel/msm-4.9/drivers/clocksource/arm_arch_timer.c中，会根据这2个timer进行clock source初始化。
CLOCKSOURCE_OF_DECLARE(armv8_arch_timer, "arm,armv8-timer", arch_timer_of_init); static int __init arch_timer_of_init(struct device_node *np) { int i; if (arch_timers_present & ARCH_CP15_TIMER) { pr_warn("arch_timer: multiple nodes in dt, skipping "); return 0; } arch_timers_present |= ARCH_CP15_TIMER; //CP15 timer for (i = PHYS_SECURE_PPI; i < MAX_TIMER_PPI; i++) arch_timer_ppi[i] = irq_of_parse_and_map(np, i); arch_timer_detect_rate(NULL, np); //从dts获取频率：19.2M Hz arch_timer_c3stop = !of_property_read_bool(np, "always-on"); #ifdef CONFIG_FSL_ERRATUM_A008585 if (fsl_a008585_enable < 0) fsl_a008585_enable = of_property_read_bool(np, "fsl,erratum-a008585"); if (fsl_a008585_enable) { static_branch_enable(&arch_timer_read_ool_enabled); pr_info("Enabling workaround for FSL erratum A-008585 "); } #endif /* * If we cannot rely on firmware initializing the timer registers then * we should use the physical timers instead. */ if (IS_ENABLED(CONFIG_ARM) && of_property_read_bool(np, "arm,cpu-registers-not-fw-configured")) arch_timer_uses_ppi = PHYS_SECURE_PPI; /* On some systems, the counter stops ticking when in suspend. */ arch_counter_suspend_stop = of_property_read_bool(np, "arm,no-tick-in-suspend"); return arch_timer_init(); //继续进行后续初始化 }
static int __init arch_timer_init(void) { int ret; /* * If HYP mode is available, we know that the physical timer * has been configured to be accessible from PL1. Use it, so * that a guest can use the virtual timer instead. * * If no interrupt provided for virtual timer, we'll have to * stick to the physical timer. It'd better be accessible... * * On ARMv8.1 with VH extensions, the kernel runs in HYP. VHE * accesses to CNTP_*_EL1 registers are silently redirected to * their CNTHP_*_EL2 counterparts, and use a different PPI * number. */ if (is_hyp_mode_available() || !arch_timer_ppi[VIRT_PPI]) { bool has_ppi; if (is_kernel_in_hyp_mode()) { arch_timer_uses_ppi = HYP_PPI; has_ppi = !!arch_timer_ppi[HYP_PPI]; } else { arch_timer_uses_ppi = PHYS_SECURE_PPI; has_ppi = (!!arch_timer_ppi[PHYS_SECURE_PPI] || !!arch_timer_ppi[PHYS_NONSECURE_PPI]); } if (!has_ppi) { pr_warn("arch_timer: No interrupt available, giving up "); return -EINVAL; } } ret = arch_timer_register(); //（1）注册timer if (ret) return ret; ret = arch_timer_common_init(); //（2）timer相关初始化 if (ret) return ret; arch_timer_kvm_info.virtual_irq = arch_timer_ppi[VIRT_PPI]; return 0; }
（1）注册timer：
static int __init arch_timer_register(void) { int err; int ppi; arch_timer_evt = alloc_percpu(struct clock_event_device); if (!arch_timer_evt) { err = -ENOMEM; goto out; } ppi = arch_timer_ppi[arch_timer_uses_ppi]; switch (arch_timer_uses_ppi) { case VIRT_PPI: err = request_percpu_irq(ppi, arch_timer_handler_virt, //仅注册percpu的irq中断（单个cpu独享，非多cpu共享），没有enable irq（真正enable在startup接口中），arch_timer_handler_virt为中断处理函数 "arch_timer", arch_timer_evt); break; case PHYS_SECURE_PPI: case PHYS_NONSECURE_PPI: err = request_percpu_irq(ppi, arch_timer_handler_phys, "arch_timer", arch_timer_evt); if (!err && arch_timer_ppi[PHYS_NONSECURE_PPI]) { ppi = arch_timer_ppi[PHYS_NONSECURE_PPI]; err = request_percpu_irq(ppi, arch_timer_handler_phys, "arch_timer", arch_timer_evt); if (err) free_percpu_irq(arch_timer_ppi[PHYS_SECURE_PPI], arch_timer_evt); } break; case HYP_PPI: err = request_percpu_irq(ppi, arch_timer_handler_phys, "arch_timer", arch_timer_evt); break; default: BUG(); } if (err) { pr_err("arch_timer: can't register interrupt %d (%d) ", ppi, err); goto out_free; } err = arch_timer_cpu_pm_init(); //注册cpu和cpu cluster进入/退出low power的notify if (err) goto out_unreg_notify; /* Register and immediately configure the timer on the boot CPU */ err = cpuhp_setup_state(CPUHP_AP_ARM_ARCH_TIMER_STARTING, //（1.1）设置cpu状态为TIMER_STARTING，注册并马上在boot cpu上配置timer。后2个函数为对应 开启/关闭cpu的callback函数， "AP_ARM_ARCH_TIMER_STARTING", arch_timer_starting_cpu, arch_timer_dying_cpu); if (err) goto out_unreg_cpupm; return 0; out_unreg_cpupm: arch_timer_cpu_pm_deinit(); out_unreg_notify: free_percpu_irq(arch_timer_ppi[arch_timer_uses_ppi], arch_timer_evt); if (arch_timer_has_nonsecure_ppi()) free_percpu_irq(arch_timer_ppi[PHYS_NONSECURE_PPI], arch_timer_evt); out_free: free_percpu(arch_timer_evt); out: return err; }
（1.1）通过__cpuhp_setup_state，注册并调用arch_timer_starting_cpu
/** * __cpuhp_setup_state - Setup the callbacks for an hotplug machine state * @state: The state to setup * @invoke: If true, the startup function is invoked for cpus where * cpu state >= @state * @startup: startup callback function * @teardown: teardown callback function * * Returns 0 if successful, otherwise a proper error code */ int __cpuhp_setup_state(enum cpuhp_state state, const char *name, bool invoke, int (*startup)(unsigned int cpu), int (*teardown)(unsigned int cpu), bool multi_instance) { int cpu, ret = 0; int dyn_state = 0; if (cpuhp_cb_check(state) || !name) return -EINVAL; get_online_cpus(); mutex_lock(&cpuhp_state_mutex); /* currently assignments for the ONLINE state are possible */ if (state == CPUHP_AP_ONLINE_DYN) { dyn_state = 1; ret = cpuhp_reserve_state(state); if (ret < 0) goto out; state = ret; } cpuhp_store_callbacks(state, name, startup, teardown, multi_instance); //配置并保存接口sp->startup.single = startup; sp->teardown.single = teardown; if (!invoke || !startup) goto out; /* * Try to call the startup callback for each present cpu * depending on the hotplug state of the cpu. */ for_each_present_cpu(cpu) { struct cpuhp_cpu_state *st = per_cpu_ptr(&cpuhp_state, cpu); int cpustate = st->state; if (cpustate < state) continue; ret = cpuhp_issue_call(cpu, state, true, NULL); //（1.1.1）调用各个处于online cpu的startup if (ret) { if (teardown) cpuhp_rollback_install(cpu, state, NULL); cpuhp_store_callbacks(state, NULL, NULL, NULL, false); goto out; } } out: mutex_unlock(&cpuhp_state_mutex); put_online_cpus(); if (!ret && dyn_state) return state; return ret; }
（1.1.1）调用startup接口，配置clock event device
... cb = bringup ? step->startup.single : step->teardown.single; if (!cb) return 0; ret = cb(cpu); ...
static int arch_timer_starting_cpu(unsigned int cpu) { struct clock_event_device *clk = this_cpu_ptr(arch_timer_evt); u32 flags; __arch_timer_setup(ARCH_CP15_TIMER, clk); //（1.1.1.1）setup和配置clock event device flags = check_ppi_trigger(arch_timer_ppi[arch_timer_uses_ppi]); enable_percpu_irq(arch_timer_ppi[arch_timer_uses_ppi], flags); //这里真正enable timer的per cpu irq if (arch_timer_has_nonsecure_ppi()) { flags = check_ppi_trigger(arch_timer_ppi[PHYS_NONSECURE_PPI]); enable_percpu_irq(arch_timer_ppi[PHYS_NONSECURE_PPI], flags); } arch_counter_set_user_access(); //设置user上层无法access timer和 physical counter，只能access virtual counter if (evtstrm_enable) arch_timer_configure_evtstream(); return 0; }
1.1.1.1 配置clock event device
其中arch_sys_timer为tick_device，arch_mem_timer为boardcast device。
当前这里是arch_sys_timer
static void __arch_timer_setup(unsigned type, struct clock_event_device *clk) { clk->features = CLOCK_EVT_FEAT_ONESHOT; if (type == ARCH_CP15_TIMER) { if (arch_timer_c3stop) clk->features |= CLOCK_EVT_FEAT_C3STOP; clk->name = "arch_sys_timer"; clk->rating = 450; clk->cpumask = cpumask_of(smp_processor_id()); clk->irq = arch_timer_ppi[arch_timer_uses_ppi]; switch (arch_timer_uses_ppi) { case VIRT_PPI: clk->set_state_shutdown = arch_timer_shutdown_virt; //这里就是配置clock event device的api，确实并没有set_state_periodic clk->set_state_oneshot_stopped = arch_timer_shutdown_virt; clk->set_next_event = arch_timer_set_next_event_virt; break; case PHYS_SECURE_PPI: case PHYS_NONSECURE_PPI: case HYP_PPI: clk->set_state_shutdown = arch_timer_shutdown_phys; clk->set_state_oneshot_stopped = arch_timer_shutdown_phys; clk->set_next_event = arch_timer_set_next_event_phys; break; default: BUG(); } fsl_a008585_set_sne(clk); } else { clk->features |= CLOCK_EVT_FEAT_DYNIRQ; clk->name = "arch_mem_timer"; clk->rating = 400; clk->cpumask = cpu_all_mask; if (arch_timer_mem_use_virtual) { clk->set_state_shutdown = arch_timer_shutdown_virt_mem; clk->set_state_oneshot_stopped = arch_timer_shutdown_virt_mem; clk->set_next_event = arch_timer_set_next_event_virt_mem; } else { clk->set_state_shutdown = arch_timer_shutdown_phys_mem; clk->set_state_oneshot_stopped = arch_timer_shutdown_phys_mem; clk->set_next_event = arch_timer_set_next_event_phys_mem; } } clk->set_state_shutdown(clk); //先关闭该clock event device clockevents_config_and_register(clk, arch_timer_rate, 0xf, 0x7fffffff); //将配置好的clock event device进一步配置并注册到系统中 }
/** * clockevents_config_and_register - Configure and register a clock event device * @dev: device to register * @freq: The clock frequency * @min_delta: The minimum clock ticks to program in oneshot mode * @max_delta: The maximum clock ticks to program in oneshot mode * * min/max_delta can be 0 for devices which do not support oneshot mode. */ void clockevents_config_and_register(struct clock_event_device *dev, u32 freq, unsigned long min_delta, unsigned long max_delta) { dev->min_delta_ticks = min_delta; dev->max_delta_ticks = max_delta; clockevents_config(dev, freq); //对应19.2MHz的clk，并根据max ticks配置最长的sleep时间 clockevents_register_device(dev); //注册device } EXPORT_SYMBOL_GPL(clockevents_config_and_register);
/** * clockevents_register_device - register a clock event device * @dev: device to register */ void clockevents_register_device(struct clock_event_device *dev) { unsigned long flags; /* Initialize state to DETACHED */ clockevent_set_state(dev, CLOCK_EVT_STATE_DETACHED); //初始化state if (!dev->cpumask) { WARN_ON(num_possible_cpus() > 1); dev->cpumask = cpumask_of(smp_processor_id()); } raw_spin_lock_irqsave(&clockevents_lock, flags); list_add(&dev->list, &clockevent_devices); //加入链表 tick_check_new_device(dev); //回到最开始分析的tick device创建 clockevents_notify_released(); raw_spin_unlock_irqrestore(&clockevents_lock, flags); } EXPORT_SYMBOL_GPL(clockevents_register_device);
（2）timer相关初始化：
static int __init arch_timer_common_init(void) { unsigned mask = ARCH_CP15_TIMER | ARCH_MEM_TIMER; /* Wait until both nodes are probed if we have two timers */ if ((arch_timers_present & mask) != mask) { //这里会等待"arm,armv7-timer-mem"（下面会分析） 和 "arm,armv8-timer"都probe完成，才进行下一步。
if (arch_timer_needs_probing(ARCH_MEM_TIMER, arch_timer_mem_of_match)) return 0; if (arch_timer_needs_probing(ARCH_CP15_TIMER, arch_timer_of_match)) return 0; } arch_timer_banner(arch_timers_present); //打印timer相关重要debug信息，LOG：03-09 04:17:39.725 root 0 0 I arm_arch_timer: Architected cp15 and mmio timer(s) running at 19.20MHz (virt/virt). arch_counter_register(arch_timers_present); //（2.1）counter注册和初始化 clocksource_select_force(); //选择clock source，即上一步注册进clock list中的arch_sys_counter return arch_timer_arch_init(); //配置并注册delay timer
（2.1）计时器注册和初始化
static struct clocksource clocksource_counter = { .name = "arch_sys_counter", .rating = 400, .read = arch_counter_read, .mask = CLOCKSOURCE_MASK(56), .flags = CLOCK_SOURCE_IS_CONTINUOUS, };
static void __init arch_counter_register(unsigned type) { u64 start_count; /* Register the CP15 based counter if we have one */ if (type & ARCH_CP15_TIMER) { if (IS_ENABLED(CONFIG_ARM64) || arch_timer_uses_ppi == VIRT_PPI) arch_timer_read_counter = arch_counter_get_cntvct; //提供read接口 else arch_timer_read_counter = arch_counter_get_cntpct; clocksource_counter.archdata.vdso_direct = true; #ifdef CONFIG_FSL_ERRATUM_A008585 /* * Don't use the vdso fastpath if errata require using * the out-of-line counter accessor. */ if (static_branch_unlikely(&arch_timer_read_ool_enabled)) clocksource_counter.archdata.vdso_direct = false; #endif } else { arch_timer_read_counter = arch_counter_get_cntvct_mem; } if (!arch_counter_suspend_stop) clocksource_counter.flags |= CLOCK_SOURCE_SUSPEND_NONSTOP; start_count = arch_timer_read_counter(); clocksource_register_hz(&clocksource_counter, arch_timer_rate); //install clocksource（19.2MHz），将其加入clocksource list，计算mult，shift。 LOG： 03-09 04:17:39.725 root 0 0 I clocksource: arch_sys_counter: mask: 0xffffffffffffff max_cycles: 0x46d987e47, max_idle_ns: 440795202767 ns cyclecounter.mult = clocksource_counter.mult; cyclecounter.shift = clocksource_counter.shift; timecounter_init(&arch_timer_kvm_info.timecounter, //计算出来的mult，shift到计时器进行初始化配置。 &cyclecounter, start_count); /* 56 bits minimum, so we assume worst case rollover */ sched_clock_register(arch_timer_read_counter, 56, arch_timer_rate); //（2.1.1）注册sched clock source }
（2.1.1）注册sched clock source
void __init sched_clock_register(u64 (*read)(void), int bits, unsigned long rate) { u64 res, wrap, new_mask, new_epoch, cyc, ns; u32 new_mult, new_shift; unsigned long r; char r_unit; struct clock_read_data rd; if (cd.rate > rate) return; WARN_ON(!irqs_disabled()); /* Calculate the mult/shift to convert counter ticks to ns. */ clocks_calc_mult_shift(&new_mult, &new_shift, rate, NSEC_PER_SEC, 3600); //计算mult，shift.转换tick数到ns单位 new_mask = CLOCKSOURCE_MASK(bits); cd.rate = rate; /* Calculate how many nanosecs until we risk wrapping */ wrap = clocks_calc_max_nsecs(new_mult, new_shift, 0, new_mask, NULL); //计算：多少ns，可能会溢出 cd.wrap_kt = ns_to_ktime(wrap); rd = cd.read_data[0]; /* Update epoch for new counter and update 'epoch_ns' from old counter*/ new_epoch = read(); cyc = cd.actual_read_sched_clock(); ns = rd.epoch_ns + cyc_to_ns((cyc - rd.epoch_cyc) & rd.sched_clock_mask, rd.mult, rd.shift); cd.actual_read_sched_clock = read; rd.read_sched_clock = read; rd.sched_clock_mask = new_mask; rd.mult = new_mult; rd.shift = new_shift; rd.epoch_cyc = new_epoch; rd.epoch_ns = ns; update_clock_read_data(&rd); //配置read clock data接口 if (sched_clock_timer.function != NULL) { /* update timeout for clock wrap */ hrtimer_start(&sched_clock_timer, cd.wrap_kt, HRTIMER_MODE_REL); } r = rate; if (r >= 4000000) { r /= 1000000; r_unit = 'M'; } else { if (r >= 1000) { r /= 1000; r_unit = 'k'; } else { r_unit = ' '; } } /* Calculate the ns resolution of this counter */ res = cyc_to_ns(1ULL, new_mult, new_shift); pr_info("sched_clock: %u bits at %lu%cHz, resolution %lluns, wraps every %lluns ", //LOG：03-09 04:17:39.725 root 0 0 I sched_clock: 56 bits at 19MHz, resolution 52ns, wraps every 4398046511078ns bits, r, r_unit, res, wrap); /* Enable IRQ time accounting if we have a fast enough sched_clock() */ if (irqtime > 0 || (irqtime == -1 && rate >= 1000000)) enable_sched_clock_irqtime(); pr_debug("Registered %pF as sched_clock source ", read); }
下面为mem_timer的部分相关流程，读者有兴趣可以自行跟踪代码。
CLOCKSOURCE_OF_DECLARE(armv7_arch_timer_mem, "arm,armv7-timer-mem", arch_timer_mem_init); static int __init arch_timer_mem_init(struct device_node *np) { struct device_node *frame, *best_frame = NULL; void __iomem *cntctlbase, *base; unsigned int irq, ret = -EINVAL; u32 cnttidr; arch_timers_present |= ARCH_MEM_TIMER; cntctlbase = of_iomap(np, 0); if (!cntctlbase) { pr_err("arch_timer: Can't find CNTCTLBase "); return -ENXIO; } cnttidr = readl_relaxed_no_log(cntctlbase + CNTTIDR); /* * Try to find a virtual capable frame. Otherwise fall back to a * physical capable frame. */ for_each_available_child_of_node(np, frame) { int n; u32 cntacr; if (of_property_read_u32(frame, "frame-number", &n)) { pr_err("arch_timer: Missing frame-number "); of_node_put(frame); goto out; } /* Try enabling everything, and see what sticks */ cntacr = CNTACR_RFRQ | CNTACR_RWPT | CNTACR_RPCT | CNTACR_RWVT | CNTACR_RVOFF | CNTACR_RVCT; writel_relaxed(cntacr, cntctlbase + CNTACR(n)); cntacr = readl_relaxed(cntctlbase + CNTACR(n)); if ((cnttidr & CNTTIDR_VIRT(n)) && !(~cntacr & (CNTACR_RWVT | CNTACR_RVCT))) { of_node_put(best_frame); best_frame = frame; arch_timer_mem_use_virtual = true; break; } if (~cntacr & (CNTACR_RWPT | CNTACR_RPCT)) continue; of_node_put(best_frame); best_frame = of_node_get(frame); } ret= -ENXIO; base = arch_counter_base = of_iomap(best_frame, 0); if (!base) { pr_err("arch_timer: Can't map frame's registers "); goto out; } if (arch_timer_mem_use_virtual) irq = irq_of_parse_and_map(best_frame, 1); else irq = irq_of_parse_and_map(best_frame, 0); ret = -EINVAL; if (!irq) { pr_err("arch_timer: Frame missing %s irq", arch_timer_mem_use_virtual ? "virt" : "phys"); goto out; } arch_timer_detect_rate(base, np); ret = arch_timer_mem_register(base, irq); if (ret) goto out; return arch_timer_common_init(); out: iounmap(cntctlbase); of_node_put(best_frame); return ret; }
adb下Debug 信息
通过adb可以确认，其中arch_sys_timer为tick_device，arch_mem_timer为boardcast device：
cat /sys/devices/system/clocksource/clocksource0/available_clocksource
arch_sys_counter
cat /sys/devices/system/clockevents/clockevent*/current_device
arch_sys_timer
arch_sys_timer
arch_sys_timer
arch_sys_timer
arch_sys_timer
arch_sys_timer
arch_sys_timer
arch_sys_timer
Tick Device list也可以通过adb确认：
tc_ocla1_sprout:/ # cat /proc/timer_list Timer List Version: v0.8 HRTIMER_MAX_CLOCK_BASES: 4 now at 102343778233457 nsecs cpu: 0 clock 0: .base: 0000000000000000 .index: 0 .resolution: 1 nsecs .get_time: ktime_get .offset: 0 nsecs active timers: #0: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102343782447797-102343782497797 nsecs [in 4214340 to 4264340 nsecs] #1: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102343784131582-102343784181582 nsecs [in 5898125 to 5948125 nsecs] #2: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102343746625332-102343786625332 nsecs [in -31608125 to 8391875 nsecs] #3: <0000000000000000>, tick_sched_timer, S:03 # expires at 102343790000000-102343790000000 nsecs [in 11766543 to 11766543 nsecs] #4: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102344559131374-102344559181374 nsecs [in 780897917 to 780947917 nsecs] #5: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102344624344916-102344625344906 nsecs [in 846111459 to 847111449 nsecs] #6: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102344865701377-102344865751377 nsecs [in 1087467920 to 1087517920 nsecs] #7: <0000000000000000>, timerfd_tmrproc, S:01 # expires at 102346297934997-102346297934997 nsecs [in 2519701540 to 2519701540 nsecs] #8: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102350106997217-102350146997217 nsecs [in 6328763760 to 6368763760 nsecs] #9: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102362935108406-102362975108406 nsecs [in 19156874949 to 19196874949 nsecs] #10: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102399338762364-102399394359362 nsecs [in 55560528907 to 55616125905 nsecs] #11: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102585376718491-102585416718491 nsecs [in 241598485034 to 241638485034 nsecs] #12: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102608578134731-102608578184731 nsecs [in 264799901274 to 264799951274 nsecs] #13: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102627061046224-102627161046224 nsecs [in 283282812767 to 283382812767 nsecs] #14: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102634070817794-102634170817794 nsecs [in 290292584337 to 290392584337 nsecs] #15: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102705307211360-102705407211360 nsecs [in 361528977903 to 361628977903 nsecs] #16: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102733142273329-102733242273329 nsecs [in 389364039872 to 389464039872 nsecs] #17: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102917415334665-102917515334665 nsecs [in 573637101208 to 573737101208 nsecs] #18: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 105191125652957-105191225652957 nsecs [in 2847347419500 to 2847447419500 nsecs] #19: <0000000000000000>, sched_clock_poll, S:01 # expires at 105359635153912-105359635153912 nsecs [in 3015856920455 to 3015856920455 nsecs] #20: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 150008206178411-150008206228411 nsecs [in 47664427944954 to 47664427994954 nsecs] #21: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 172812354468858-172812354518858 nsecs [in 70468576235401 to 70468576285401 nsecs] #22: <0000000000000000>, it_real_fn, S:01 # expires at 172812376385680-172812376385680 nsecs [in 70468598152223 to 70468598152223 nsecs] #23: <0000000000000000>, timerfd_tmrproc, S:01 # expires at 173057145132000-173057145132000 nsecs [in 70713366898543 to 70713366898543 nsecs] #24: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 500008591686350-500008591736350 nsecs [in 397664813452893 to 397664813502893 nsecs] clock 1: .base: 0000000000000000 .index: 1 .resolution: 1 nsecs .get_time: ktime_get_real .offset: 1577592197245305630 nsecs active timers: clock 2: .base: 0000000000000000 .index: 2 .resolution: 1 nsecs .get_time: ktime_get_boottime .offset: 5348254632 nsecs active timers: #0: <0000000000000000>, alarmtimer_fired, S:01 # expires at 102670651000000-102670651000000 nsecs [in 321524511911 to 321524511911 nsecs] #1: <0000000000000000>, timerfd_tmrproc, S:01 # expires at 107028105000000-107028105000000 nsecs [in 4678978511911 to 4678978511911 nsecs] clock 3: .base: 0000000000000000 .index: 3 .resolution: 1 nsecs .get_time: ktime_get_clocktai .offset: 1577592197245305630 nsecs active timers: .expires_next : 102343782497797 nsecs .hres_active : 1 .nr_events : 9716538 .nr_retries : 9612 .nr_hangs : 0 .max_hang_time : 0 .nohz_mode : 2 .last_tick : 102343760000000 nsecs .tick_stopped : 0 .idle_jiffies : 4305171671 .idle_calls : 15389066 .idle_sleeps : 11043177 .idle_entrytime : 102343772001947 nsecs .idle_waketime : 102343743066739 nsecs .idle_exittime : 102343760567884 nsecs .idle_sleeptime : 91073881832478 nsecs .iowait_sleeptime: 126261797989 nsecs .last_jiffies : 4305171673 .next_timer : 102343780000000 .idle_expires : 102344740000000 nsecs jiffies: 4305171674 cpu: 1 clock 0: .base: 0000000000000000 .index: 0 .resolution: 1 nsecs .get_time: ktime_get .offset: 0 nsecs active timers: #0: <0000000000000000>, tick_sched_timer, S:03 # expires at 102343790000000-102343790000000 nsecs [in 11766543 to 11766543 nsecs] #1: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102345616605957-102345616655957 nsecs [in 1838372500 to 1838422500 nsecs] #2: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102348461282998-102348491282997 nsecs [in 4683049541 to 4713049540 nsecs] #3: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102352914732997-102352914782997 nsecs [in 9136499540 to 9136549540 nsecs] #4: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102368297707318-102368397707318 nsecs [in 24519473861 to 24619473861 nsecs] #5: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102369592027095-102369632027095 nsecs [in 25813793638 to 25853793638 nsecs] #6: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102602692588558-102602692638558 nsecs [in 258914355101 to 258914405101 nsecs] #7: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 120558704386589-120558804386589 nsecs [in 18214926153132 to 18215026153132 nsecs] #8: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 172847398703681-172847498703681 nsecs [in 70503620470224 to 70503720470224 nsecs] clock 1: .base: 0000000000000000 .index: 1 .resolution: 1 nsecs .get_time: ktime_get_real .offset: 1577592197245305630 nsecs active timers: clock 2: .base: 0000000000000000 .index: 2 .resolution: 1 nsecs .get_time: ktime_get_boottime .offset: 5348254632 nsecs active timers: clock 3: .base: 0000000000000000 .index: 3 .resolution: 1 nsecs .get_time: ktime_get_clocktai .offset: 1577592197245305630 nsecs active timers: .expires_next : 102343790000000 nsecs .hres_active : 1 .nr_events : 8525252 .nr_retries : 6736 .nr_hangs : 0 .max_hang_time : 0 .nohz_mode : 2 .last_tick : 102343750000000 nsecs .tick_stopped : 0 .idle_jiffies : 4305171670 .idle_calls : 714821 .idle_sleeps : 515264 .idle_entrytime : 102343745435957 nsecs .idle_waketime : 102343745435957 nsecs .idle_exittime : 102343745472051 nsecs .idle_sleeptime : 5177799957603 nsecs .iowait_sleeptime: 12700959475 nsecs .last_jiffies : 4305171670 .next_timer : 102343830000000 .idle_expires : 102343830000000 nsecs jiffies: 4305171674 cpu: 2 clock 0: .base: 0000000000000000 .index: 0 .resolution: 1 nsecs .get_time: ktime_get .offset: 0 nsecs active timers: #0: <0000000000000000>, tick_sched_timer, S:03 # expires at 102343790000000-102343790000000 nsecs [in 11766543 to 11766543 nsecs] #1: <0000000000000000>, sched_rt_period_timer, S:03 # expires at 102344000000000-102344000000000 nsecs [in 221766543 to 221766543 nsecs] #2: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102349155736740-102349165736738 nsecs [in 5377503283 to 5387503281 nsecs] #3: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102353327602474-102353427602474 nsecs [in 9549369017 to 9649369017 nsecs] #4: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102354056718424-102354056768424 nsecs [in 10278484967 to 10278534967 nsecs] #5: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102356502485854-102356517402853 nsecs [in 12724252397 to 12739169396 nsecs] #6: <0000000000000000>, posix_timer_fn, S:01 # expires at 102380890693049-102380890693049 nsecs [in 37112459592 to 37112459592 nsecs] #7: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102396512824866-102396512874866 nsecs [in 52734591409 to 52734641409 nsecs] #8: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 104019335142591-104019335192591 nsecs [in 1675556909134 to 1675556959134 nsecs] clock 1: .base: 0000000000000000 .index: 1 .resolution: 1 nsecs .get_time: ktime_get_real .offset: 1577592197245305630 nsecs active timers: clock 2: .base: 0000000000000000 .index: 2 .resolution: 1 nsecs .get_time: ktime_get_boottime .offset: 5348254632 nsecs active timers: #0: <0000000000000000>, alarmtimer_fired, S:01 # expires at 102392489700397-102392489700397 nsecs [in 43363212308 to 43363212308 nsecs] #1: <0000000000000000>, alarmtimer_fired, S:01 # expires at 102624969000000-102624969000000 nsecs [in 275842511911 to 275842511911 nsecs] clock 3: .base: 0000000000000000 .index: 3 .resolution: 1 nsecs .get_time: ktime_get_clocktai .offset: 1577592197245305630 nsecs active timers: .expires_next : 102343790000000 nsecs .hres_active : 1 .nr_events : 8791822 .nr_retries : 27696 .nr_hangs : 0 .max_hang_time : 0 .nohz_mode : 2 .last_tick : 102343680000000 nsecs .tick_stopped : 0 .idle_jiffies : 4305171663 .idle_calls : 806233 .idle_sleeps : 590526 .idle_entrytime : 102343780041634 nsecs .idle_waketime : 102343653006739 nsecs .idle_exittime : 102343673157676 nsecs .idle_sleeptime : 5080392137502 nsecs .iowait_sleeptime: 5705048172 nsecs .last_jiffies : 4305171673 .next_timer : 102343780000000 .idle_expires : 102344990000000 nsecs jiffies: 4305171674 cpu: 3 clock 0: .base: 0000000000000000 .index: 0 .resolution: 1 nsecs .get_time: ktime_get .offset: 0 nsecs active timers: #0: <0000000000000000>, tick_sched_timer, S:03 # expires at 102343790000000-102343790000000 nsecs [in 11766543 to 11766543 nsecs] #1: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102344414214447-102344415214442 nsecs [in 635980990 to 636980985 nsecs] #2: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102397177072627-102397277072627 nsecs [in 53398839170 to 53498839170 nsecs] #3: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 102632303944714-102632403944714 nsecs [in 288525711257 to 288625711257 nsecs] #4: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 104414961740811-104415061740811 nsecs [in 2071183507354 to 2071283507354 nsecs] #5: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 108022052013874-108022152013874 nsecs [in 5678273780417 to 5678373780417 nsecs] #6: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 110014003359763-110014103359763 nsecs [in 7670225126306 to 7670325126306 nsecs] #7: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 9223372036854775807-9223372036854775807 nsecs [in 9223269693076542350 to 9223269693076542350 nsecs] clock 1: .base: 0000000000000000 .index: 1 .resolution: 1 nsecs .get_time: ktime_get_real .offset: 1577592197245305630 nsecs active timers: clock 2: .base: 0000000000000000 .index: 2 .resolution: 1 nsecs .get_time: ktime_get_boottime .offset: 5348254632 nsecs active timers: clock 3: .base: 0000000000000000 .index: 3 .resolution: 1 nsecs .get_time: ktime_get_clocktai .offset: 1577592197245305630 nsecs active timers: .expires_next : 102343790000000 nsecs .hres_active : 1 .nr_events : 8028797 .nr_retries : 16288 .nr_hangs : 0 .max_hang_time : 0 .nohz_mode : 2 .last_tick : 102343780000000 nsecs .tick_stopped : 0 .idle_jiffies : 4305171673 .idle_calls : 776457 .idle_sleeps : 558645 .idle_entrytime : 102343780298822 nsecs .idle_waketime : 102343709265801 nsecs .idle_exittime : 102343780298822 nsecs .idle_sleeptime : 5070930572889 nsecs .iowait_sleeptime: 8522453661 nsecs .last_jiffies : 4305171673 .next_timer : 102344650000000 .idle_expires : 102344650000000 nsecs jiffies: 4305171674 cpu: 4 clock 0: .base: 0000000000000000 .index: 0 .resolution: 1 nsecs .get_time: ktime_get .offset: 0 nsecs active timers: #0: <0000000000000000>, hrtimer_wakeup, S:01 # expires at 111408145011166-111408245011166 nsecs [in 9064366777709 to 9064466777709 nsecs] clock 1: .base: 0000000000000000 .index: 1 .resolution: 1 nsecs .get_time: ktime_get_real .offset: 1577592197245305630 nsecs active timers: clock 2: .base: 0000000000000000 .index: 2 .resolution: 1 nsecs .get_time: ktime_get_boottime .offset: 5348254632 nsecs active timers: clock 3: .base: 0000000000000000 .index: 3 .resolution: 1 nsecs .get_time: ktime_get_clocktai .offset: 1577592197245305630 nsecs active timers: .expires_next : 111408245011166 nsecs .hres_active : 1 .nr_events : 51287 .nr_retries : 39 .nr_hangs : 0 .max_hang_time : 0 .nohz_mode : 2 .last_tick : 102343760000000 nsecs .tick_stopped : 1 .idle_jiffies : 4305171671 .idle_calls : 16077 .idle_sleeps : 15389 .idle_entrytime : 102343780268353 nsecs .idle_waketime : 102343751119343 nsecs .idle_exittime : 102343749631009 nsecs .idle_sleeptime : 5771415901159 nsecs .iowait_sleeptime: 91151360 nsecs .last_jiffies : 4305171671 .next_timer : 9223372036854775807 .idle_expires : 9223372036854775807 nsecs jiffies: 4305171674 cpu: 5 clock 0: .base: 0000000000000000 .index: 0 .resolution: 1 nsecs .get_time: ktime_get .offset: 0 nsecs active timers: clock 1: .base: 0000000000000000 .index: 1 .resolution: 1 nsecs .get_time: ktime_get_real .offset: 1577592197245305630 nsecs active timers: clock 2: .base: 0000000000000000 .index: 2 .resolution: 1 nsecs .get_time: ktime_get_boottime .offset: 5348254632 nsecs active timers: clock 3: .base: 0000000000000000 .index: 3 .resolution: 1 nsecs .get_time: ktime_get_clocktai .offset: 1577592197245305630 nsecs active timers: .expires_next : 9223372036854775807 nsecs .hres_active : 1 .nr_events : 32597 .nr_retries : 26 .nr_hangs : 0 .max_hang_time : 0 .nohz_mode : 2 .last_tick : 102343760000000 nsecs .tick_stopped : 1 .idle_jiffies : 4305171671 .idle_calls : 11180 .idle_sleeps : 10973 .idle_entrytime : 102343780269447 nsecs .idle_waketime : 102343751066895 nsecs .idle_exittime : 102343749803353 nsecs .idle_sleeptime : 5773496930467 nsecs .iowait_sleeptime: 25364688 nsecs .last_jiffies : 4305171671 .next_timer : 9223372036854775807 .idle_expires : 9223372036854775807 nsecs jiffies: 4305171674 cpu: 6 clock 0: .base: 0000000000000000 .index: 0 .resolution: 1 nsecs .get_time: ktime_get .offset: 0 nsecs active timers: clock 1: .base: 0000000000000000 .index: 1 .resolution: 1 nsecs .get_time: ktime_get_real .offset: 1577592197245305630 nsecs active timers: clock 2: .base: 0000000000000000 .index: 2 .resolution: 1 nsecs .get_time: ktime_get_boottime .offset: 5348254632 nsecs active timers: clock 3: .base: 0000000000000000 .index: 3 .resolution: 1 nsecs .get_time: ktime_get_clocktai .offset: 1577592197245305630 nsecs active timers: .expires_next : 9223372036854775807 nsecs .hres_active : 1 .nr_events : 77050 .nr_retries : 18 .nr_hangs : 0 .max_hang_time : 0 .nohz_mode : 2 .last_tick : 102343760000000 nsecs .tick_stopped : 1 .idle_jiffies : 4305171671 .idle_calls : 9996 .idle_sleeps : 9851 .idle_entrytime : 102343780270280 nsecs .idle_waketime : 102343761084603 nsecs .idle_exittime : 102343750332936 nsecs .idle_sleeptime : 5772812686518 nsecs .iowait_sleeptime: 10439268 nsecs .last_jiffies : 4305171672 .next_timer : 9223372036854775807 .idle_expires : 9223372036854775807 nsecs jiffies: 4305171674 cpu: 7 clock 0: .base: 0000000000000000 .index: 0 .resolution: 1 nsecs .get_time: ktime_get .offset: 0 nsecs active timers: clock 1: .base: 0000000000000000 .index: 1 .resolution: 1 nsecs .get_time: ktime_get_real .offset: 1577592197245305630 nsecs active timers: clock 2: .base: 0000000000000000 .index: 2 .resolution: 1 nsecs .get_time: ktime_get_boottime .offset: 5348254632 nsecs active timers: clock 3: .base: 0000000000000000 .index: 3 .resolution: 1 nsecs .get_time: ktime_get_clocktai .offset: 1577592197245305630 nsecs active timers: .expires_next : 9223372036854775807 nsecs .hres_active : 1 .nr_events : 44091 .nr_retries : 13 .nr_hangs : 0 .max_hang_time : 0 .nohz_mode : 2 .last_tick : 102343680000000 nsecs .tick_stopped : 1 .idle_jiffies : 4305171663 .idle_calls : 7542 .idle_sleeps : 7490 .idle_entrytime : 102343780271114 nsecs .idle_waketime : 102343680010332 nsecs .idle_exittime : 102343679531114 nsecs .idle_sleeptime : 5773215745442 nsecs .iowait_sleeptime: 18214117 nsecs .last_jiffies : 4305171664 .next_timer : 9223372036854775807 .idle_expires : 9223372036854775807 nsecs jiffies: 4305171674 Tick Device: mode: 1 Broadcast device Clock Event Device: arch_mem_timer max_delta_ns: 111848106728 min_delta_ns: 1000 mult: 82463372 shift: 32 mode: 3 next_event: 102343880000000 nsecs set_next_event: arch_timer_set_next_event_virt_mem shutdown: arch_timer_shutdown_virt_mem event_handler: tick_handle_oneshot_broadcast retries: 52148 tick_broadcast_mask: 00 tick_broadcast_oneshot_mask: fc Tick Device: mode: 1 Per CPU device: 0 Clock Event Device: arch_sys_timer max_delta_ns: 111848106728 min_delta_ns: 1000 mult: 82463372 shift: 32 mode: 3 next_event: 102343782497797 nsecs set_next_event: arch_timer_set_next_event_virt shutdown: arch_timer_shutdown_virt event_handler: hrtimer_interrupt retries: 4949 Tick Device: mode: 1 Per CPU device: 1 Clock Event Device: arch_sys_timer max_delta_ns: 111848106728 min_delta_ns: 1000 mult: 82463372 shift: 32 mode: 3 next_event: 102343790000000 nsecs set_next_event: arch_timer_set_next_event_virt shutdown: arch_timer_shutdown_virt event_handler: hrtimer_interrupt retries: 4068 Tick Device: mode: 1 Per CPU device: 2 Clock Event Device: arch_sys_timer max_delta_ns: 111848106728 min_delta_ns: 1000 mult: 82463372 shift: 32 mode: 1 next_event: 102343880000000 nsecs set_next_event: arch_timer_set_next_event_virt shutdown: arch_timer_shutdown_virt event_handler: hrtimer_interrupt retries: 3241 Tick Device: mode: 1 Per CPU device: 3 Clock Event Device: arch_sys_timer max_delta_ns: 111848106728 min_delta_ns: 1000 mult: 82463372 shift: 32 mode: 1 next_event: 102344415214442 nsecs set_next_event: arch_timer_set_next_event_virt shutdown: arch_timer_shutdown_virt event_handler: hrtimer_interrupt retries: 2861 Tick Device: mode: 1 Per CPU device: 4 Clock Event Device: arch_sys_timer max_delta_ns: 111848106728 min_delta_ns: 1000 mult: 82463372 shift: 32 mode: 1 next_event: 111408245011166 nsecs set_next_event: arch_timer_set_next_event_virt shutdown: arch_timer_shutdown_virt event_handler: hrtimer_interrupt retries: 67 Tick Device: mode: 1 Per CPU device: 5 Clock Event Device: arch_sys_timer max_delta_ns: 111848106728 min_delta_ns: 1000 mult: 82463372 shift: 32 mode: 1 next_event: 9223372036854775807 nsecs set_next_event: arch_timer_set_next_event_virt shutdown: arch_timer_shutdown_virt event_handler: hrtimer_interrupt retries: 54 Tick Device: mode: 1 Per CPU device: 6 Clock Event Device: arch_sys_timer max_delta_ns: 111848106728 min_delta_ns: 1000 mult: 82463372 shift: 32 mode: 1 next_event: 9223372036854775807 nsecs set_next_event: arch_timer_set_next_event_virt shutdown: arch_timer_shutdown_virt event_handler: hrtimer_interrupt retries: 54 Tick Device: mode: 1 Per CPU device: 7 Clock Event Device: arch_sys_timer max_delta_ns: 111848106728 min_delta_ns: 1000 mult: 82463372 shift: 32 mode: 1 next_event: 9223372036854775807 nsecs set_next_event: arch_timer_set_next_event_virt shutdown: arch_timer_shutdown_virt event_handler: hrtimer_interrupt retries: 60
Ftrace抓取的call stack：
<idle>-0 [003] d.h2 98348.520507: scheduler_tick <-update_process_times <idle>-0 [003] d.h2 98348.520534: <stack trace> => tick_sched_timer => __hrtimer_run_queues => hrtimer_interrupt => tick_handle_oneshot_broadcast => arch_timer_handler_virt_mem => handle_irq_event_percpu => handle_irq_event => handle_fasteoi_irq => generic_handle_irq => __handle_domain_irq => gic_handle_irq => el1_irq => lpm_cpuidle_enter => cpuidle_enter_state => cpuidle_enter => cpu_startup_entry => secondary_start_kernel => <idle>-0 [003] d.h2 98348.540237: scheduler_tick <-update_process_times <idle>-0 [003] d.h2 98348.540267: <stack trace> => tick_sched_timer => __hrtimer_run_queues => hrtimer_interrupt => arch_timer_handler_virt => handle_percpu_devid_irq => generic_handle_irq => __handle_domain_irq => gic_handle_irq => el1_irq => lpm_cpuidle_enter => cpuidle_enter_state => cpuidle_enter => cpu_startup_entry => secondary_start_kernel =>