diff options
Diffstat (limited to 'include')
| -rw-r--r-- | include/linux/cgroup_subsys.h | 4 | ||||
| -rw-r--r-- | include/linux/cpufreq.h | 16 | ||||
| -rw-r--r-- | include/linux/cpuidle.h | 2 | ||||
| -rw-r--r-- | include/linux/sched.h | 86 | ||||
| -rw-r--r-- | include/linux/sched/sysctl.h | 26 | ||||
| -rw-r--r-- | include/linux/sched_energy.h | 44 | ||||
| -rw-r--r-- | include/linux/vmstat.h | 2 | ||||
| -rw-r--r-- | include/trace/events/cpufreq_sched.h | 87 | ||||
| -rw-r--r-- | include/trace/events/power.h | 7 | ||||
| -rw-r--r-- | include/trace/events/sched.h | 498 |
10 files changed, 771 insertions, 1 deletions
diff --git a/include/linux/cgroup_subsys.h b/include/linux/cgroup_subsys.h index 1a96fdaa33d5..e133705d794a 100644 --- a/include/linux/cgroup_subsys.h +++ b/include/linux/cgroup_subsys.h @@ -26,6 +26,10 @@ SUBSYS(cpu) SUBSYS(cpuacct) #endif +#if IS_ENABLED(CONFIG_CGROUP_SCHEDTUNE) +SUBSYS(schedtune) +#endif + #if IS_ENABLED(CONFIG_BLK_CGROUP) SUBSYS(io) #endif diff --git a/include/linux/cpufreq.h b/include/linux/cpufreq.h index 3ac01621cd1f..60571292a802 100644 --- a/include/linux/cpufreq.h +++ b/include/linux/cpufreq.h @@ -160,6 +160,7 @@ u64 get_cpu_idle_time(unsigned int cpu, u64 *wall, int io_busy); int cpufreq_get_policy(struct cpufreq_policy *policy, unsigned int cpu); int cpufreq_update_policy(unsigned int cpu); bool have_governor_per_policy(void); +bool cpufreq_driver_is_slow(void); struct kobject *get_governor_parent_kobj(struct cpufreq_policy *policy); #else static inline unsigned int cpufreq_get(unsigned int cpu) @@ -317,6 +318,14 @@ struct cpufreq_driver { */ #define CPUFREQ_NEED_INITIAL_FREQ_CHECK (1 << 5) +/* + * Indicates that it is safe to call cpufreq_driver_target from + * non-interruptable context in scheduler hot paths. Drivers must + * opt-in to this flag, as the safe default is that they might sleep + * or be too slow for hot path use. + */ +#define CPUFREQ_DRIVER_FAST (1 << 6) + int cpufreq_register_driver(struct cpufreq_driver *driver_data); int cpufreq_unregister_driver(struct cpufreq_driver *driver_data); @@ -490,6 +499,9 @@ extern struct cpufreq_governor cpufreq_gov_conservative; #elif defined(CONFIG_CPU_FREQ_DEFAULT_GOV_INTERACTIVE) extern struct cpufreq_governor cpufreq_gov_interactive; #define CPUFREQ_DEFAULT_GOVERNOR (&cpufreq_gov_interactive) +#elif defined(CONFIG_CPU_FREQ_DEFAULT_GOV_SCHED) +extern struct cpufreq_governor cpufreq_gov_sched; +#define CPUFREQ_DEFAULT_GOVERNOR (&cpufreq_gov_sched) #endif /********************************************************************* @@ -619,4 +631,8 @@ unsigned int cpufreq_generic_get(unsigned int cpu); int cpufreq_generic_init(struct cpufreq_policy *policy, struct cpufreq_frequency_table *table, unsigned int transition_latency); + +struct sched_domain; +unsigned long cpufreq_scale_freq_capacity(struct sched_domain *sd, int cpu); +unsigned long cpufreq_scale_max_freq_capacity(int cpu); #endif /* _LINUX_CPUFREQ_H */ diff --git a/include/linux/cpuidle.h b/include/linux/cpuidle.h index 786ad32631a6..6eae1576499e 100644 --- a/include/linux/cpuidle.h +++ b/include/linux/cpuidle.h @@ -204,7 +204,7 @@ static inline int cpuidle_enter_freeze(struct cpuidle_driver *drv, #endif /* kernel/sched/idle.c */ -extern void sched_idle_set_state(struct cpuidle_state *idle_state); +extern void sched_idle_set_state(struct cpuidle_state *idle_state, int index); extern void default_idle_call(void); #ifdef CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED diff --git a/include/linux/sched.h b/include/linux/sched.h index 145c34cb106e..ede29e8db82d 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -173,6 +173,9 @@ extern bool single_task_running(void); extern unsigned long nr_iowait(void); extern unsigned long nr_iowait_cpu(int cpu); extern void get_iowait_load(unsigned long *nr_waiters, unsigned long *load); +#ifdef CONFIG_CPU_QUIET +extern u64 nr_running_integral(unsigned int cpu); +#endif extern void calc_global_load(unsigned long ticks); @@ -314,6 +317,15 @@ extern char ___assert_task_state[1 - 2*!!( /* Task command name length */ #define TASK_COMM_LEN 16 +enum task_event { + PUT_PREV_TASK = 0, + PICK_NEXT_TASK = 1, + TASK_WAKE = 2, + TASK_MIGRATE = 3, + TASK_UPDATE = 4, + IRQ_UPDATE = 5, +}; + #include <linux/spinlock.h> /* @@ -929,6 +941,14 @@ enum cpu_idle_type { #define SCHED_CAPACITY_SHIFT 10 #define SCHED_CAPACITY_SCALE (1L << SCHED_CAPACITY_SHIFT) +struct sched_capacity_reqs { + unsigned long cfs; + unsigned long rt; + unsigned long dl; + + unsigned long total; +}; + /* * Wake-queues are lists of tasks with a pending wakeup, whose * callers have already marked the task as woken internally, @@ -991,6 +1011,7 @@ extern void wake_up_q(struct wake_q_head *head); #define SD_PREFER_SIBLING 0x1000 /* Prefer to place tasks in a sibling domain */ #define SD_OVERLAP 0x2000 /* sched_domains of this level overlap */ #define SD_NUMA 0x4000 /* cross-node balancing */ +#define SD_SHARE_CAP_STATES 0x8000 /* Domain members share capacity state */ #ifdef CONFIG_SCHED_SMT static inline int cpu_smt_flags(void) @@ -1023,6 +1044,24 @@ struct sched_domain_attr { extern int sched_domain_level_max; +struct capacity_state { + unsigned long cap; /* compute capacity */ + unsigned long power; /* power consumption at this compute capacity */ +}; + +struct idle_state { + unsigned long power; /* power consumption in this idle state */ +}; + +struct sched_group_energy { + unsigned int nr_idle_states; /* number of idle states */ + struct idle_state *idle_states; /* ptr to idle state array */ + unsigned int nr_cap_states; /* number of capacity states */ + struct capacity_state *cap_states; /* ptr to capacity state array */ +}; + +unsigned long capacity_curr_of(int cpu); + struct sched_group; struct sched_domain { @@ -1121,6 +1160,8 @@ bool cpus_share_cache(int this_cpu, int that_cpu); typedef const struct cpumask *(*sched_domain_mask_f)(int cpu); typedef int (*sched_domain_flags_f)(void); +typedef +const struct sched_group_energy * const(*sched_domain_energy_f)(int cpu); #define SDTL_OVERLAP 0x01 @@ -1133,6 +1174,7 @@ struct sd_data { struct sched_domain_topology_level { sched_domain_mask_f mask; sched_domain_flags_f sd_flags; + sched_domain_energy_f energy; int flags; int numa_level; struct sd_data data; @@ -1243,6 +1285,41 @@ struct sched_statistics { }; #endif +#ifdef CONFIG_SCHED_WALT +#define RAVG_HIST_SIZE_MAX 5 + +/* ravg represents frequency scaled cpu-demand of tasks */ +struct ravg { + /* + * 'mark_start' marks the beginning of an event (task waking up, task + * starting to execute, task being preempted) within a window + * + * 'sum' represents how runnable a task has been within current + * window. It incorporates both running time and wait time and is + * frequency scaled. + * + * 'sum_history' keeps track of history of 'sum' seen over previous + * RAVG_HIST_SIZE windows. Windows where task was entirely sleeping are + * ignored. + * + * 'demand' represents maximum sum seen over previous + * sysctl_sched_ravg_hist_size windows. 'demand' could drive frequency + * demand for tasks. + * + * 'curr_window' represents task's contribution to cpu busy time + * statistics (rq->curr_runnable_sum) in current window + * + * 'prev_window' represents task's contribution to cpu busy time + * statistics (rq->prev_runnable_sum) in previous window + */ + u64 mark_start; + u32 sum, demand; + u32 sum_history[RAVG_HIST_SIZE_MAX]; + u32 curr_window, prev_window; + u16 active_windows; +}; +#endif + struct sched_entity { struct load_weight load; /* for load-balancing */ struct rb_node run_node; @@ -1400,6 +1477,15 @@ struct task_struct { const struct sched_class *sched_class; struct sched_entity se; struct sched_rt_entity rt; +#ifdef CONFIG_SCHED_WALT + struct ravg ravg; + /* + * 'init_load_pct' represents the initial task load assigned to children + * of this task + */ + u32 init_load_pct; +#endif + #ifdef CONFIG_CGROUP_SCHED struct task_group *sched_task_group; #endif diff --git a/include/linux/sched/sysctl.h b/include/linux/sched/sysctl.h index c9e4731cf10b..d68e88c9d4d7 100644 --- a/include/linux/sched/sysctl.h +++ b/include/linux/sched/sysctl.h @@ -39,6 +39,16 @@ extern unsigned int sysctl_sched_latency; extern unsigned int sysctl_sched_min_granularity; extern unsigned int sysctl_sched_wakeup_granularity; extern unsigned int sysctl_sched_child_runs_first; +extern unsigned int sysctl_sched_is_big_little; +extern unsigned int sysctl_sched_sync_hint_enable; +extern unsigned int sysctl_sched_initial_task_util; +extern unsigned int sysctl_sched_cstate_aware; +#ifdef CONFIG_SCHED_WALT +extern unsigned int sysctl_sched_use_walt_cpu_util; +extern unsigned int sysctl_sched_use_walt_task_util; +extern unsigned int sysctl_sched_walt_init_task_load_pct; +extern unsigned int sysctl_sched_walt_cpu_high_irqload; +#endif enum sched_tunable_scaling { SCHED_TUNABLESCALING_NONE, @@ -77,6 +87,22 @@ extern int sysctl_sched_rt_runtime; extern unsigned int sysctl_sched_cfs_bandwidth_slice; #endif +#ifdef CONFIG_SCHED_TUNE +extern unsigned int sysctl_sched_cfs_boost; +int sysctl_sched_cfs_boost_handler(struct ctl_table *table, int write, + void __user *buffer, size_t *length, + loff_t *ppos); +static inline unsigned int get_sysctl_sched_cfs_boost(void) +{ + return sysctl_sched_cfs_boost; +} +#else +static inline unsigned int get_sysctl_sched_cfs_boost(void) +{ + return 0; +} +#endif + #ifdef CONFIG_SCHED_AUTOGROUP extern unsigned int sysctl_sched_autogroup_enabled; #endif diff --git a/include/linux/sched_energy.h b/include/linux/sched_energy.h new file mode 100644 index 000000000000..1daf3e1f98a7 --- /dev/null +++ b/include/linux/sched_energy.h @@ -0,0 +1,44 @@ +#ifndef _LINUX_SCHED_ENERGY_H +#define _LINUX_SCHED_ENERGY_H + +#include <linux/sched.h> +#include <linux/slab.h> + +/* + * There doesn't seem to be an NR_CPUS style max number of sched domain + * levels so here's an arbitrary constant one for the moment. + * + * The levels alluded to here correspond to entries in struct + * sched_domain_topology_level that are meant to be populated by arch + * specific code (topology.c). + */ +#define NR_SD_LEVELS 8 + +#define SD_LEVEL0 0 +#define SD_LEVEL1 1 +#define SD_LEVEL2 2 +#define SD_LEVEL3 3 +#define SD_LEVEL4 4 +#define SD_LEVEL5 5 +#define SD_LEVEL6 6 +#define SD_LEVEL7 7 + +/* + * Convenience macro for iterating through said sd levels. + */ +#define for_each_possible_sd_level(level) \ + for (level = 0; level < NR_SD_LEVELS; level++) + +#ifdef CONFIG_SMP + +extern struct sched_group_energy *sge_array[NR_CPUS][NR_SD_LEVELS]; + +void init_sched_energy_costs(void); + +#else + +#define init_sched_energy_costs() do { } while (0) + +#endif /* CONFIG_SMP */ + +#endif diff --git a/include/linux/vmstat.h b/include/linux/vmstat.h index 3e5d9075960f..73fae8c4a5fb 100644 --- a/include/linux/vmstat.h +++ b/include/linux/vmstat.h @@ -189,6 +189,7 @@ extern void __inc_zone_state(struct zone *, enum zone_stat_item); extern void dec_zone_state(struct zone *, enum zone_stat_item); extern void __dec_zone_state(struct zone *, enum zone_stat_item); +void quiet_vmstat(void); void cpu_vm_stats_fold(int cpu); void refresh_zone_stat_thresholds(void); @@ -249,6 +250,7 @@ static inline void __dec_zone_page_state(struct page *page, static inline void refresh_zone_stat_thresholds(void) { } static inline void cpu_vm_stats_fold(int cpu) { } +static inline void quiet_vmstat(void) { } static inline void drain_zonestat(struct zone *zone, struct per_cpu_pageset *pset) { } diff --git a/include/trace/events/cpufreq_sched.h b/include/trace/events/cpufreq_sched.h new file mode 100644 index 000000000000..a46cd088e969 --- /dev/null +++ b/include/trace/events/cpufreq_sched.h @@ -0,0 +1,87 @@ +/* + * Copyright (C) 2015 Steve Muckle <smuckle@linaro.org> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#undef TRACE_SYSTEM +#define TRACE_SYSTEM cpufreq_sched + +#if !defined(_TRACE_CPUFREQ_SCHED_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_CPUFREQ_SCHED_H + +#include <linux/sched.h> +#include <linux/tracepoint.h> + +TRACE_EVENT(cpufreq_sched_throttled, + TP_PROTO(unsigned int rem), + TP_ARGS(rem), + TP_STRUCT__entry( + __field( unsigned int, rem) + ), + TP_fast_assign( + __entry->rem = rem; + ), + TP_printk("throttled - %d usec remaining", __entry->rem) +); + +TRACE_EVENT(cpufreq_sched_request_opp, + TP_PROTO(int cpu, + unsigned long capacity, + unsigned int freq_new, + unsigned int requested_freq), + TP_ARGS(cpu, capacity, freq_new, requested_freq), + TP_STRUCT__entry( + __field( int, cpu) + __field( unsigned long, capacity) + __field( unsigned int, freq_new) + __field( unsigned int, requested_freq) + ), + TP_fast_assign( + __entry->cpu = cpu; + __entry->capacity = capacity; + __entry->freq_new = freq_new; + __entry->requested_freq = requested_freq; + ), + TP_printk("cpu %d cap change, cluster cap request %ld => OPP %d " + "(cur %d)", + __entry->cpu, __entry->capacity, __entry->freq_new, + __entry->requested_freq) +); + +TRACE_EVENT(cpufreq_sched_update_capacity, + TP_PROTO(int cpu, + bool request, + struct sched_capacity_reqs *scr, + unsigned long new_capacity), + TP_ARGS(cpu, request, scr, new_capacity), + TP_STRUCT__entry( + __field( int, cpu) + __field( bool, request) + __field( unsigned long, cfs) + __field( unsigned long, rt) + __field( unsigned long, dl) + __field( unsigned long, total) + __field( unsigned long, new_total) + ), + TP_fast_assign( + __entry->cpu = cpu; + __entry->request = request; + __entry->cfs = scr->cfs; + __entry->rt = scr->rt; + __entry->dl = scr->dl; + __entry->total = scr->total; + __entry->new_total = new_capacity; + ), + TP_printk("cpu=%d set_cap=%d cfs=%ld rt=%ld dl=%ld old_tot=%ld " + "new_tot=%ld", + __entry->cpu, __entry->request, __entry->cfs, __entry->rt, + __entry->dl, __entry->total, __entry->new_total) +); + +#endif /* _TRACE_CPUFREQ_SCHED_H */ + +/* This part must be outside protection */ +#include <trace/define_trace.h> diff --git a/include/trace/events/power.h b/include/trace/events/power.h index 9af0d898016a..8924cc2b4ca8 100644 --- a/include/trace/events/power.h +++ b/include/trace/events/power.h @@ -145,6 +145,13 @@ TRACE_EVENT(cpu_frequency_limits, (unsigned long)__entry->cpu_id) ); +DEFINE_EVENT(cpu, cpu_capacity, + + TP_PROTO(unsigned int capacity, unsigned int cpu_id), + + TP_ARGS(capacity, cpu_id) +); + TRACE_EVENT(device_pm_callback_start, TP_PROTO(struct device *dev, const char *pm_ops, int event), diff --git a/include/trace/events/sched.h b/include/trace/events/sched.h index d34eba74af27..dffaffab4bc8 100644 --- a/include/trace/events/sched.h +++ b/include/trace/events/sched.h @@ -611,6 +611,504 @@ TRACE_EVENT(sched_wake_idle_without_ipi, TP_printk("cpu=%d", __entry->cpu) ); + +TRACE_EVENT(sched_contrib_scale_f, + + TP_PROTO(int cpu, unsigned long freq_scale_factor, + unsigned long cpu_scale_factor), + + TP_ARGS(cpu, freq_scale_factor, cpu_scale_factor), + + TP_STRUCT__entry( + __field(int, cpu) + __field(unsigned long, freq_scale_factor) + __field(unsigned long, cpu_scale_factor) + ), + + TP_fast_assign( + __entry->cpu = cpu; + __entry->freq_scale_factor = freq_scale_factor; + __entry->cpu_scale_factor = cpu_scale_factor; + ), + + TP_printk("cpu=%d freq_scale_factor=%lu cpu_scale_factor=%lu", + __entry->cpu, __entry->freq_scale_factor, + __entry->cpu_scale_factor) +); + +#ifdef CONFIG_SMP + +/* + * Tracepoint for accounting sched averages for tasks. + */ +TRACE_EVENT(sched_load_avg_task, + + TP_PROTO(struct task_struct *tsk, struct sched_avg *avg), + + TP_ARGS(tsk, avg), + + TP_STRUCT__entry( + __array( char, comm, TASK_COMM_LEN ) + __field( pid_t, pid ) + __field( int, cpu ) + __field( unsigned long, load_avg ) + __field( unsigned long, util_avg ) + __field( u64, load_sum ) + __field( u32, util_sum ) + __field( u32, period_contrib ) + ), + + TP_fast_assign( + memcpy(__entry->comm, tsk->comm, TASK_COMM_LEN); + __entry->pid = tsk->pid; + __entry->cpu = task_cpu(tsk); + __entry->load_avg = avg->load_avg; + __entry->util_avg = avg->util_avg; + __entry->load_sum = avg->load_sum; + __entry->util_sum = avg->util_sum; + __entry->period_contrib = avg->period_contrib; + ), + + TP_printk("comm=%s pid=%d cpu=%d load_avg=%lu util_avg=%lu load_sum=%llu" + " util_sum=%u period_contrib=%u", + __entry->comm, + __entry->pid, + __entry->cpu, + __entry->load_avg, + __entry->util_avg, + (u64)__entry->load_sum, + (u32)__entry->util_sum, + (u32)__entry->period_contrib) +); + +/* + * Tracepoint for accounting sched averages for cpus. + */ +TRACE_EVENT(sched_load_avg_cpu, + + TP_PROTO(int cpu, struct cfs_rq *cfs_rq), + + TP_ARGS(cpu, cfs_rq), + + TP_STRUCT__entry( + __field( int, cpu ) + __field( unsigned long, load_avg ) + __field( unsigned long, util_avg ) + ), + + TP_fast_assign( + __entry->cpu = cpu; + __entry->load_avg = cfs_rq->avg.load_avg; + __entry->util_avg = cfs_rq->avg.util_avg; + ), + + TP_printk("cpu=%d load_avg=%lu util_avg=%lu", + __entry->cpu, __entry->load_avg, __entry->util_avg) +); + +/* + * Tracepoint for sched_tune_config settings + */ +TRACE_EVENT(sched_tune_config, + + TP_PROTO(int boost), + + TP_ARGS(boost), + + TP_STRUCT__entry( + __field( int, boost ) + ), + + TP_fast_assign( + __entry->boost = boost; + ), + + TP_printk("boost=%d ", __entry->boost) +); + +/* + * Tracepoint for accounting CPU boosted utilization + */ +TRACE_EVENT(sched_boost_cpu, + + TP_PROTO(int cpu, unsigned long util, long margin), + + TP_ARGS(cpu, util, margin), + + TP_STRUCT__entry( + __field( int, cpu ) + __field( unsigned long, util ) + __field(long, margin ) + ), + + TP_fast_assign( + __entry->cpu = cpu; + __entry->util = util; + __entry->margin = margin; + ), + + TP_printk("cpu=%d util=%lu margin=%ld", + __entry->cpu, + __entry->util, + __entry->margin) +); + +/* + * Tracepoint for schedtune_tasks_update + */ +TRACE_EVENT(sched_tune_tasks_update, + + TP_PROTO(struct task_struct *tsk, int cpu, int tasks, int idx, + int boost, int max_boost), + + TP_ARGS(tsk, cpu, tasks, idx, boost, max_boost), + + TP_STRUCT__entry( + __array( char, comm, TASK_COMM_LEN ) + __field( pid_t, pid ) + __field( int, cpu ) + __field( int, tasks ) + __field( int, idx ) + __field( int, boost ) + __field( int, max_boost ) + ), + + TP_fast_assign( + memcpy(__entry->comm, tsk->comm, TASK_COMM_LEN); + __entry->pid = tsk->pid; + __entry->cpu = cpu; + __entry->tasks = tasks; + __entry->idx = idx; + __entry->boost = boost; + __entry->max_boost = max_boost; + ), + + TP_printk("pid=%d comm=%s " + "cpu=%d tasks=%d idx=%d boost=%d max_boost=%d", + __entry->pid, __entry->comm, + __entry->cpu, __entry->tasks, __entry->idx, + __entry->boost, __entry->max_boost) +); + +/* + * Tracepoint for schedtune_boostgroup_update + */ +TRACE_EVENT(sched_tune_boostgroup_update, + + TP_PROTO(int cpu, int variation, int max_boost), + + TP_ARGS(cpu, variation, max_boost), + + TP_STRUCT__entry( + __field( int, cpu ) + __field( int, variation ) + __field( int, max_boost ) + ), + + TP_fast_assign( + __entry->cpu = cpu; + __entry->variation = variation; + __entry->max_boost = max_boost; + ), + + TP_printk("cpu=%d variation=%d max_boost=%d", + __entry->cpu, __entry->variation, __entry->max_boost) +); + +/* + * Tracepoint for accounting task boosted utilization + */ +TRACE_EVENT(sched_boost_task, + + TP_PROTO(struct task_struct *tsk, unsigned long util, long margin), + + TP_ARGS(tsk, util, margin), + + TP_STRUCT__entry( + __array( char, comm, TASK_COMM_LEN ) + __field( pid_t, pid ) + __field( unsigned long, util ) + __field( long, margin ) + + ), + + TP_fast_assign( + memcpy(__entry->comm, tsk->comm, TASK_COMM_LEN); + __entry->pid = tsk->pid; + __entry->util = util; + __entry->margin = margin; + ), + + TP_printk("comm=%s pid=%d util=%lu margin=%ld", + __entry->comm, __entry->pid, + __entry->util, + __entry->margin) +); + +/* + * Tracepoint for accounting sched group energy + */ +TRACE_EVENT(sched_energy_diff, + + TP_PROTO(struct task_struct *tsk, int scpu, int dcpu, int udelta, + int nrgb, int nrga, int nrgd, int capb, int capa, int capd, + int nrgn, int nrgp), + + TP_ARGS(tsk, scpu, dcpu, udelta, + nrgb, nrga, nrgd, capb, capa, capd, + nrgn, nrgp), + + TP_STRUCT__entry( + __array( char, comm, TASK_COMM_LEN ) + __field( pid_t, pid ) + __field( int, scpu ) + __field( int, dcpu ) + __field( int, udelta ) + __field( int, nrgb ) + __field( int, nrga ) + __field( int, nrgd ) + __field( int, capb ) + __field( int, capa ) + __field( int, capd ) + __field( int, nrgn ) + __field( int, nrgp ) + ), + + TP_fast_assign( + memcpy(__entry->comm, tsk->comm, TASK_COMM_LEN); + __entry->pid = tsk->pid; + __entry->scpu = scpu; + __entry->dcpu = dcpu; + __entry->udelta = udelta; + __entry->nrgb = nrgb; + __entry->nrga = nrga; + __entry->nrgd = nrgd; + __entry->capb = capb; + __entry->capa = capa; + __entry->capd = capd; + __entry->nrgn = nrgn; + __entry->nrgp = nrgp; + ), + + TP_printk("pid=%d comm=%s " + "src_cpu=%d dst_cpu=%d usage_delta=%d " + "nrg_before=%d nrg_after=%d nrg_diff=%d " + "cap_before=%d cap_after=%d cap_delta=%d " + "nrg_delta=%d nrg_payoff=%d", + __entry->pid, __entry->comm, + __entry->scpu, __entry->dcpu, __entry->udelta, + __entry->nrgb, __entry->nrga, __entry->nrgd, + __entry->capb, __entry->capa, __entry->capd, + __entry->nrgn, __entry->nrgp) +); + +/* + * Tracepoint for schedtune_tasks_update + */ +TRACE_EVENT(sched_tune_filter, + + TP_PROTO(int nrg_delta, int cap_delta, + int nrg_gain, int cap_gain, + int payoff, int region), + + TP_ARGS(nrg_delta, cap_delta, nrg_gain, cap_gain, payoff, region), + + TP_STRUCT__entry( + __field( int, nrg_delta ) + __field( int, cap_delta ) + __field( int, nrg_gain ) + __field( int, cap_gain ) + __field( int, payoff ) + __field( int, region ) + ), + + TP_fast_assign( + __entry->nrg_delta = nrg_delta; + __entry->cap_delta = cap_delta; + __entry->nrg_gain = nrg_gain; + __entry->cap_gain = cap_gain; + __entry->payoff = payoff; + __entry->region = region; + ), + + TP_printk("nrg_delta=%d cap_delta=%d nrg_gain=%d cap_gain=%d payoff=%d region=%d", + __entry->nrg_delta, __entry->cap_delta, + __entry->nrg_gain, __entry->cap_gain, + __entry->payoff, __entry->region) +); + +/* + * Tracepoint for system overutilized flag + */ +TRACE_EVENT(sched_overutilized, + + TP_PROTO(bool overutilized), + + TP_ARGS(overutilized), + + TP_STRUCT__entry( + __field( bool, overutilized ) + ), + + TP_fast_assign( + __entry->overutilized = overutilized; + ), + + TP_printk("overutilized=%d", + __entry->overutilized ? 1 : 0) +); +#ifdef CONFIG_SCHED_WALT +struct rq; + +TRACE_EVENT(walt_update_task_ravg, + + TP_PROTO(struct task_struct *p, struct rq *rq, int evt, + u64 wallclock, u64 irqtime), + + TP_ARGS(p, rq, evt, wallclock, irqtime), + + TP_STRUCT__entry( + __array( char, comm, TASK_COMM_LEN ) + __field( pid_t, pid ) + __field( pid_t, cur_pid ) + __field(unsigned int, cur_freq ) + __field( u64, wallclock ) + __field( u64, mark_start ) + __field( u64, delta_m ) + __field( u64, win_start ) + __field( u64, delta ) + __field( u64, irqtime ) + __field( int, evt ) + __field(unsigned int, demand ) + __field(unsigned int, sum ) + __field( int, cpu ) + __field( u64, cs ) + __field( u64, ps ) + __field( u32, curr_window ) + __field( u32, prev_window ) + __field( u64, nt_cs ) + __field( u64, nt_ps ) + __field( u32, active_windows ) + ), + + TP_fast_assign( + __entry->wallclock = wallclock; + __entry->win_start = rq->window_start; + __entry->delta = (wallclock - rq->window_start); + __entry->evt = evt; + __entry->cpu = rq->cpu; + __entry->cur_pid = rq->curr->pid; + __entry->cur_freq = rq->cur_freq; + memcpy(__entry->comm, p->comm, TASK_COMM_LEN); + __entry->pid = p->pid; + __entry->mark_start = p->ravg.mark_start; + __entry->delta_m = (wallclock - p->ravg.mark_start); + __entry->demand = p->ravg.demand; + __entry->sum = p->ravg.sum; + __entry->irqtime = irqtime; + __entry->cs = rq->curr_runnable_sum; + __entry->ps = rq->prev_runnable_sum; + __entry->curr_window = p->ravg.curr_window; + __entry->prev_window = p->ravg.prev_window; + __entry->nt_cs = rq->nt_curr_runnable_sum; + __entry->nt_ps = rq->nt_prev_runnable_sum; + __entry->active_windows = p->ravg.active_windows; + ), + + TP_printk("wc %llu ws %llu delta %llu event %d cpu %d cur_freq %u cur_pid %d task %d (%s) ms %llu delta %llu demand %u sum %u irqtime %llu" + " cs %llu ps %llu cur_window %u prev_window %u nt_cs %llu nt_ps %llu active_wins %u" + , __entry->wallclock, __entry->win_start, __entry->delta, + __entry->evt, __entry->cpu, + __entry->cur_freq, __entry->cur_pid, + __entry->pid, __entry->comm, __entry->mark_start, + __entry->delta_m, __entry->demand, + __entry->sum, __entry->irqtime, + __entry->cs, __entry->ps, + __entry->curr_window, __entry->prev_window, + __entry->nt_cs, __entry->nt_ps, + __entry->active_windows + ) +); + +TRACE_EVENT(walt_update_history, + + TP_PROTO(struct rq *rq, struct task_struct *p, u32 runtime, int samples, + int evt), + + TP_ARGS(rq, p, runtime, samples, evt), + + TP_STRUCT__entry( + __array( char, comm, TASK_COMM_LEN ) + __field( pid_t, pid ) + __field(unsigned int, runtime ) + __field( int, samples ) + __field( int, evt ) + __field( u64, demand ) + __field(unsigned int, walt_avg ) + __field(unsigned int, pelt_avg ) + __array( u32, hist, RAVG_HIST_SIZE_MAX) + __field( int, cpu ) + ), + + TP_fast_assign( + memcpy(__entry->comm, p->comm, TASK_COMM_LEN); + __entry->pid = p->pid; + __entry->runtime = runtime; + __entry->samples = samples; + __entry->evt = evt; + __entry->demand = p->ravg.demand; + __entry->walt_avg = (__entry->demand << 10); + do_div(__entry->walt_avg, walt_ravg_window); + __entry->pelt_avg = p->se.avg.util_avg; + memcpy(__entry->hist, p->ravg.sum_history, + RAVG_HIST_SIZE_MAX * sizeof(u32)); + __entry->cpu = rq->cpu; + ), + + TP_printk("%d (%s): runtime %u samples %d event %d demand %llu" + " walt %u pelt %u (hist: %u %u %u %u %u) cpu %d", + __entry->pid, __entry->comm, + __entry->runtime, __entry->samples, __entry->evt, + __entry->demand, + __entry->walt_avg, + __entry->pelt_avg, + __entry->hist[0], __entry->hist[1], + __entry->hist[2], __entry->hist[3], + __entry->hist[4], __entry->cpu) +); + +TRACE_EVENT(walt_migration_update_sum, + + TP_PROTO(struct rq *rq, struct task_struct *p), + + TP_ARGS(rq, p), + + TP_STRUCT__entry( + __field(int, cpu ) + __field(int, pid ) + __field( u64, cs ) + __field( u64, ps ) + __field( s64, nt_cs ) + __field( s64, nt_ps ) + ), + + TP_fast_assign( + __entry->cpu = cpu_of(rq); + __entry->cs = rq->curr_runnable_sum; + __entry->ps = rq->prev_runnable_sum; + __entry->nt_cs = (s64)rq->nt_curr_runnable_sum; + __entry->nt_ps = (s64)rq->nt_prev_runnable_sum; + __entry->pid = p->pid; + ), + + TP_printk("cpu %d: cs %llu ps %llu nt_cs %lld nt_ps %lld pid %d", + __entry->cpu, __entry->cs, __entry->ps, + __entry->nt_cs, __entry->nt_ps, __entry->pid) +); +#endif /* CONFIG_SCHED_WALT */ + +#endif /* CONFIG_SMP */ + #endif /* _TRACE_SCHED_H */ /* This part must be outside protection */ |