blob: de497e2879c7e957545531a43e8422558f333dd6 [file] [log] [blame]
/*
* Copyright (c) 2018 Intel Corporation
*
* SPDX-License-Identifier: Apache-2.0
*/
#include <zephyr/kernel.h>
#include <zephyr/timing/timing.h>
#include <ksched.h>
#include <zephyr/spinlock.h>
#include <zephyr/sys/check.h>
/* Need one of these for this to work */
#if !defined(CONFIG_USE_SWITCH) && !defined(CONFIG_INSTRUMENT_THREAD_SWITCHING)
#error "No data backend configured for CONFIG_SCHED_THREAD_USAGE"
#endif
static struct k_spinlock usage_lock;
static uint32_t usage_now(void)
{
uint32_t now;
#ifdef CONFIG_THREAD_RUNTIME_STATS_USE_TIMING_FUNCTIONS
now = (uint32_t)timing_counter_get();
#else
now = k_cycle_get_32();
#endif
/* Edge case: we use a zero as a null ("stop() already called") */
return (now == 0) ? 1 : now;
}
#ifdef CONFIG_SCHED_THREAD_USAGE_ALL
static void sched_cpu_update_usage(struct _cpu *cpu, uint32_t cycles)
{
if (!cpu->usage->track_usage) {
return;
}
if (cpu->current != cpu->idle_thread) {
cpu->usage->total += cycles;
#ifdef CONFIG_SCHED_THREAD_USAGE_ANALYSIS
cpu->usage->current += cycles;
if (cpu->usage->longest < cpu->usage->current) {
cpu->usage->longest = cpu->usage->current;
}
} else {
cpu->usage->current = 0;
cpu->usage->num_windows++;
#endif
}
}
#else
#define sched_cpu_update_usage(cpu, cycles) do { } while (0)
#endif
static void sched_thread_update_usage(struct k_thread *thread, uint32_t cycles)
{
thread->base.usage.total += cycles;
#ifdef CONFIG_SCHED_THREAD_USAGE_ANALYSIS
thread->base.usage.current += cycles;
if (thread->base.usage.longest < thread->base.usage.current) {
thread->base.usage.longest = thread->base.usage.current;
}
#endif
}
void z_sched_usage_start(struct k_thread *thread)
{
#ifdef CONFIG_SCHED_THREAD_USAGE_ANALYSIS
k_spinlock_key_t key;
key = k_spin_lock(&usage_lock);
_current_cpu->usage0 = usage_now(); /* Always update */
if (thread->base.usage.track_usage) {
thread->base.usage.num_windows++;
thread->base.usage.current = 0;
}
k_spin_unlock(&usage_lock, key);
#else
/* One write through a volatile pointer doesn't require
* synchronization as long as _usage() treats it as volatile
* (we can't race with _stop() by design).
*/
_current_cpu->usage0 = usage_now();
#endif
}
void z_sched_usage_stop(void)
{
k_spinlock_key_t k = k_spin_lock(&usage_lock);
struct _cpu *cpu = _current_cpu;
uint32_t u0 = cpu->usage0;
if (u0 != 0) {
uint32_t cycles = usage_now() - u0;
if (cpu->current->base.usage.track_usage) {
sched_thread_update_usage(cpu->current, cycles);
}
sched_cpu_update_usage(cpu, cycles);
}
cpu->usage0 = 0;
k_spin_unlock(&usage_lock, k);
}
#ifdef CONFIG_SCHED_THREAD_USAGE_ALL
void z_sched_cpu_usage(uint8_t cpu_id, struct k_thread_runtime_stats *stats)
{
k_spinlock_key_t key;
struct _cpu *cpu;
key = k_spin_lock(&usage_lock);
cpu = _current_cpu;
if (&_kernel.cpus[cpu_id] == cpu) {
uint32_t now = usage_now();
uint32_t cycles = now - cpu->usage0;
/*
* Getting stats for the current CPU. Update both its
* current thread stats and the CPU stats as the CPU's
* [usage0] field will also get updated. This keeps all
* that information up-to-date.
*/
if (cpu->current->base.usage.track_usage) {
sched_thread_update_usage(cpu->current, cycles);
}
sched_cpu_update_usage(cpu, cycles);
cpu->usage0 = now;
}
stats->total_cycles = cpu->usage->total;
#ifdef CONFIG_SCHED_THREAD_USAGE_ANALYSIS
stats->current_cycles = cpu->usage->current;
stats->peak_cycles = cpu->usage->longest;
if (cpu->usage->num_windows == 0) {
stats->average_cycles = 0;
} else {
stats->average_cycles = stats->total_cycles /
cpu->usage->num_windows;
}
#endif
stats->idle_cycles =
_kernel.cpus[cpu_id].idle_thread->base.usage.total;
stats->execution_cycles = stats->total_cycles + stats->idle_cycles;
k_spin_unlock(&usage_lock, key);
}
#endif
void z_sched_thread_usage(struct k_thread *thread,
struct k_thread_runtime_stats *stats)
{
struct _cpu *cpu;
k_spinlock_key_t key;
key = k_spin_lock(&usage_lock);
cpu = _current_cpu;
if (thread == cpu->current) {
uint32_t now = usage_now();
uint32_t cycles = now - cpu->usage0;
/*
* Getting stats for the current thread. Update both the
* current thread stats and its CPU stats as the CPU's
* [usage0] field will also get updated. This keeps all
* that information up-to-date.
*/
if (thread->base.usage.track_usage) {
sched_thread_update_usage(thread, cycles);
}
sched_cpu_update_usage(cpu, cycles);
cpu->usage0 = now;
}
stats->execution_cycles = thread->base.usage.total;
stats->total_cycles = thread->base.usage.total;
/* Copy-out the thread's usage stats */
#ifdef CONFIG_SCHED_THREAD_USAGE_ANALYSIS
stats->current_cycles = thread->base.usage.current;
stats->peak_cycles = thread->base.usage.longest;
if (thread->base.usage.num_windows == 0) {
stats->average_cycles = 0;
} else {
stats->average_cycles = stats->total_cycles /
thread->base.usage.num_windows;
}
#endif
#ifdef CONFIG_SCHED_THREAD_USAGE_ALL
stats->idle_cycles = 0;
#endif
stats->execution_cycles = thread->base.usage.total;
k_spin_unlock(&usage_lock, key);
}
#ifdef CONFIG_SCHED_THREAD_USAGE_ANALYSIS
int k_thread_runtime_stats_enable(k_tid_t thread)
{
k_spinlock_key_t key;
CHECKIF(thread == NULL) {
return -EINVAL;
}
key = k_spin_lock(&usage_lock);
if (!thread->base.usage.track_usage) {
thread->base.usage.track_usage = true;
thread->base.usage.num_windows++;
thread->base.usage.current = 0;
}
k_spin_unlock(&usage_lock, key);
return 0;
}
int k_thread_runtime_stats_disable(k_tid_t thread)
{
k_spinlock_key_t key;
CHECKIF(thread == NULL) {
return -EINVAL;
}
key = k_spin_lock(&usage_lock);
struct _cpu *cpu = _current_cpu;
if (thread->base.usage.track_usage) {
thread->base.usage.track_usage = false;
if (thread == cpu->current) {
uint32_t cycles = usage_now() - cpu->usage0;
sched_thread_update_usage(thread, cycles);
sched_cpu_update_usage(cpu, cycles);
}
}
k_spin_unlock(&usage_lock, key);
return 0;
}
#endif
#ifdef CONFIG_SCHED_THREAD_USAGE_ALL
void k_sys_runtime_stats_enable(void)
{
k_spinlock_key_t key;
key = k_spin_lock(&usage_lock);
if (_current_cpu->usage->track_usage) {
/*
* Usage tracking is already enabled on the current CPU
* and thus on all other CPUs (if applicable). There is
* nothing left to do.
*/
k_spin_unlock(&usage_lock, key);
return;
}
/* Enable gathering of runtime stats on each CPU */
unsigned int num_cpus = arch_num_cpus();
for (uint8_t i = 0; i < num_cpus; i++) {
_kernel.cpus[i].usage->track_usage = true;
#ifdef CONFIG_SCHED_THREAD_USAGE_ANALYSIS
_kernel.cpus[i].usage->num_windows++;
_kernel.cpus[i].usage->current = 0;
#endif
}
k_spin_unlock(&usage_lock, key);
}
void k_sys_runtime_stats_disable(void)
{
struct _cpu *cpu;
k_spinlock_key_t key;
key = k_spin_lock(&usage_lock);
if (!_current_cpu->usage->track_usage) {
/*
* Usage tracking is already disabled on the current CPU
* and thus on all other CPUs (if applicable). There is
* nothing left to do.
*/
k_spin_unlock(&usage_lock, key);
return;
}
uint32_t now = usage_now();
unsigned int num_cpus = arch_num_cpus();
for (uint8_t i = 0; i < num_cpus; i++) {
cpu = &_kernel.cpus[i];
if (cpu->usage0 != 0) {
sched_cpu_update_usage(cpu, now - cpu->usage0);
}
cpu->usage->track_usage = false;
}
k_spin_unlock(&usage_lock, key);
}
#endif
#ifdef CONFIG_OBJ_CORE_STATS_THREAD
int z_thread_stats_raw(struct k_obj_core *obj_core, void *stats)
{
k_spinlock_key_t key;
key = k_spin_lock(&usage_lock);
memcpy(stats, obj_core->stats, sizeof(struct k_cycle_stats));
k_spin_unlock(&usage_lock, key);
return 0;
}
int z_thread_stats_query(struct k_obj_core *obj_core, void *stats)
{
struct k_thread *thread;
thread = CONTAINER_OF(obj_core, struct k_thread, obj_core);
z_sched_thread_usage(thread, stats);
return 0;
}
int z_thread_stats_reset(struct k_obj_core *obj_core)
{
k_spinlock_key_t key;
struct k_cycle_stats *stats;
struct k_thread *thread;
thread = CONTAINER_OF(obj_core, struct k_thread, obj_core);
key = k_spin_lock(&usage_lock);
stats = obj_core->stats;
stats->total = 0ULL;
#ifdef CONFIG_SCHED_THREAD_USAGE_ANALYSIS
stats->current = 0ULL;
stats->longest = 0ULL;
stats->num_windows = (thread->base.usage.track_usage) ? 1U : 0U;
#endif
if (thread != _current_cpu->current) {
/*
* If the thread is not running, there is nothing else to do.
* If the thread is running on another core, then it is not
* safe to do anything else but unlock and return (and pretend
* that its stats were reset at the start of its execution
* window.
*/
k_spin_unlock(&usage_lock, key);
return 0;
}
/* Update the current CPU stats. */
uint32_t now = usage_now();
uint32_t cycles = now - _current_cpu->usage0;
sched_cpu_update_usage(_current_cpu, cycles);
_current_cpu->usage0 = now;
k_spin_unlock(&usage_lock, key);
return 0;
}
int z_thread_stats_disable(struct k_obj_core *obj_core)
{
#ifdef CONFIG_SCHED_THREAD_USAGE_ANALYSIS
struct k_thread *thread;
thread = CONTAINER_OF(obj_core, struct k_thread, obj_core);
return k_thread_runtime_stats_disable(thread);
#else
return -ENOTSUP;
#endif
}
int z_thread_stats_enable(struct k_obj_core *obj_core)
{
#ifdef CONFIG_SCHED_THREAD_USAGE_ANALYSIS
struct k_thread *thread;
thread = CONTAINER_OF(obj_core, struct k_thread, obj_core);
return k_thread_runtime_stats_enable(thread);
#else
return -ENOTSUP;
#endif
}
#endif
#ifdef CONFIG_OBJ_CORE_STATS_SYSTEM
int z_cpu_stats_raw(struct k_obj_core *obj_core, void *stats)
{
k_spinlock_key_t key;
key = k_spin_lock(&usage_lock);
memcpy(stats, obj_core->stats, sizeof(struct k_cycle_stats));
k_spin_unlock(&usage_lock, key);
return 0;
}
int z_cpu_stats_query(struct k_obj_core *obj_core, void *stats)
{
struct _cpu *cpu;
cpu = CONTAINER_OF(obj_core, struct _cpu, obj_core);
z_sched_cpu_usage(cpu->id, stats);
return 0;
}
#endif
#ifdef CONFIG_OBJ_CORE_STATS_SYSTEM
int z_kernel_stats_raw(struct k_obj_core *obj_core, void *stats)
{
k_spinlock_key_t key;
key = k_spin_lock(&usage_lock);
memcpy(stats, obj_core->stats,
CONFIG_MP_MAX_NUM_CPUS * sizeof(struct k_cycle_stats));
k_spin_unlock(&usage_lock, key);
return 0;
}
int z_kernel_stats_query(struct k_obj_core *obj_core, void *stats)
{
ARG_UNUSED(obj_core);
return k_thread_runtime_stats_all_get(stats);
}
#endif