@@ -29,4 +29,14 @@ config INTEL_RAPL
controller, CPU core (Power Plance 0), graphics uncore (Power Plane
1), etc.
+config IDLE_INJECT
+ bool "Idle injection framework"
+ depends on CPU_IDLE
+ default n
+ help
+ This enables support for the idle injection framework. It
+ provides a way to force idle periods on a set of specified
+ CPUs for power capping. Idle period can be injected
+ synchronously on a set of specified CPUs or alternatively
+ on a per CPU basis.
endif
@@ -1,2 +1,3 @@
obj-$(CONFIG_POWERCAP) += powercap_sys.o
obj-$(CONFIG_INTEL_RAPL) += intel_rapl.o
+obj-$(CONFIG_IDLE_INJECT) += idle_inject.o
new file mode 100644
@@ -0,0 +1,359 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright 2018 Linaro Limited
+ *
+ * Author: Daniel Lezcano <daniel.lezcano@linaro.org>
+ *
+ * The idle injection framework proposes a way to force a cpu to enter
+ * an idle state during a specified amount of time for a specified
+ * period.
+ *
+ * It relies on the smpboot kthreads which handles, via its main loop,
+ * the common code for hotplugging and [un]parking.
+ *
+ * At init time, all the kthreads are created.
+ *
+ * A cpumask is specified as parameter for the idle injection
+ * registering function. The kthreads will be synchronized regarding
+ * this cpumask.
+ *
+ * The idle + run duration is specified via the helpers and then the
+ * idle injection can be started at this point.
+ *
+ * A kthread will call play_idle() with the specified idle duration
+ * from above.
+ *
+ * A timer is set after waking up all the tasks, to the next idle
+ * injection cycle.
+ *
+ * The task handling the timer interrupt will wakeup all the kthreads
+ * belonging to the cpumask.
+ *
+ * Stopping the idle injection is synchronous, when the function
+ * returns, there is the guarantee there is no more idle injection
+ * kthread in activity.
+ *
+ * It is up to the user of this framework to provide a lock at an
+ * upper level to prevent stupid things to happen, like starting while
+ * we are unregistering.
+ */
+#define pr_fmt(fmt) "ii_dev: " fmt
+
+#include <linux/cpu.h>
+#include <linux/hrtimer.h>
+#include <linux/kthread.h>
+#include <linux/sched.h>
+#include <linux/slab.h>
+#include <linux/smpboot.h>
+
+#include <uapi/linux/sched/types.h>
+
+/**
+ * struct idle_inject_thread - task on/off switch structure
+ * @tsk: a pointer to a task_struct injecting the idle cycles
+ * @should_run: a integer used as a boolean by the smpboot kthread API
+ */
+struct idle_inject_thread {
+ struct task_struct *tsk;
+ int should_run;
+};
+
+/**
+ * struct idle_inject_device - data for the idle injection
+ * @timer: a hrtimer giving the tempo for the idle injection
+ * @idle_duration_ms: an unsigned int specifying the idle duration
+ * @run_duration_ms: an unsigned int specifying the running duration
+ * @cpumask: a cpumask containing the list of CPUs managed by the device
+ */
+struct idle_inject_device {
+ struct hrtimer timer;
+ unsigned int idle_duration_ms;
+ unsigned int run_duration_ms;
+ unsigned long int cpumask[0];
+};
+
+static DEFINE_PER_CPU(struct idle_inject_thread, idle_inject_thread);
+static DEFINE_PER_CPU(struct idle_inject_device *, idle_inject_device);
+
+/**
+ * idle_inject_wakeup - Wake up all idle injection threads
+ * @ii_dev: the idle injection device
+ *
+ * Every idle injection task belonging to the idle injection device
+ * and running on an online CPU will be wake up by this call.
+ */
+static void idle_inject_wakeup(struct idle_inject_device *ii_dev)
+{
+ struct idle_inject_thread *iit;
+ unsigned int cpu;
+
+ for_each_cpu_and(cpu, to_cpumask(ii_dev->cpumask), cpu_online_mask) {
+ iit = per_cpu_ptr(&idle_inject_thread, cpu);
+ iit->should_run = 1;
+ wake_up_process(iit->tsk);
+ }
+}
+
+/**
+ * idle_inject_timer_fn - idle injection timer callback
+ * @timer: a hrtimer structure
+ *
+ * This function is called when the idle injection timer expires which
+ * will wake up the idle injection tasks and these ones, in turn, play
+ * idle a specified amount of time.
+ *
+ * Return: HRTIMER_RESTART.
+ */
+static enum hrtimer_restart idle_inject_timer_fn(struct hrtimer *timer)
+{
+ unsigned int duration_ms;
+ struct idle_inject_device *ii_dev =
+ container_of(timer, struct idle_inject_device, timer);
+
+ duration_ms = READ_ONCE(ii_dev->run_duration_ms);
+ duration_ms += READ_ONCE(ii_dev->idle_duration_ms);
+
+ idle_inject_wakeup(ii_dev);
+
+ hrtimer_forward_now(timer, ms_to_ktime(duration_ms));
+
+ return HRTIMER_RESTART;
+}
+
+/**
+ * idle_inject_fn - idle injection routine
+ * @cpu: the CPU number the task belongs to
+ *
+ * The idle injection routine will stay idle the specified amount of
+ * time
+ */
+static void idle_inject_fn(unsigned int cpu)
+{
+ struct idle_inject_device *ii_dev;
+ struct idle_inject_thread *iit;
+
+ ii_dev = per_cpu(idle_inject_device, cpu);
+ iit = per_cpu_ptr(&idle_inject_thread, cpu);
+
+ /*
+ * Boolean used by the smpboot main loop and used as a
+ * flip-flop in this function.
+ */
+ iit->should_run = 0;
+
+ play_idle(READ_ONCE(ii_dev->idle_duration_ms));
+}
+
+/**
+ * idle_inject_set_duration - idle and run duration helper
+ * @run_duration_ms: an unsigned int giving the running time in milliseconds
+ * @idle_duration_ms: an unsigned int giving the idle time in milliseconds
+ */
+void idle_inject_set_duration(struct idle_inject_device *ii_dev,
+ unsigned int run_duration_ms,
+ unsigned int idle_duration_ms)
+{
+ if (run_duration_ms && idle_duration_ms) {
+ WRITE_ONCE(ii_dev->run_duration_ms, run_duration_ms);
+ WRITE_ONCE(ii_dev->idle_duration_ms, idle_duration_ms);
+ }
+}
+
+/**
+ * idle_inject_get_duration - idle and run duration helper
+ * @run_duration_ms: a pointer to an unsigned int to store the running time
+ * @idle_duration_ms: a pointer to an unsigned int to store the idle time
+ */
+void idle_inject_get_duration(struct idle_inject_device *ii_dev,
+ unsigned int *run_duration_ms,
+ unsigned int *idle_duration_ms)
+{
+ *run_duration_ms = READ_ONCE(ii_dev->run_duration_ms);
+ *idle_duration_ms = READ_ONCE(ii_dev->idle_duration_ms);
+}
+
+/**
+ * idle_inject_start - starts the idle injections
+ * @ii_dev: a pointer to an idle_inject_device structure
+ *
+ * The function starts the idle injection cycles by first waking up
+ * all the tasks the ii_dev is attached to and let them handle the
+ * idle-run periods.
+ *
+ * Return: -EINVAL if the idle or the running durations are not set.
+ */
+int idle_inject_start(struct idle_inject_device *ii_dev)
+{
+ unsigned int idle_duration_ms = READ_ONCE(ii_dev->idle_duration_ms);
+ unsigned int run_duration_ms = READ_ONCE(ii_dev->run_duration_ms);
+
+ if (!idle_duration_ms || !run_duration_ms)
+ return -EINVAL;
+
+ pr_debug("Starting injecting idle cycles on CPUs '%*pbl'\n",
+ cpumask_pr_args(to_cpumask(ii_dev->cpumask)));
+
+ idle_inject_wakeup(ii_dev);
+
+ hrtimer_start(&ii_dev->timer,
+ ms_to_ktime(idle_duration_ms + run_duration_ms),
+ HRTIMER_MODE_REL);
+
+ return 0;
+}
+
+/**
+ * idle_inject_stop - stops the idle injections
+ * @ii_dev: a pointer to an idle injection_device structure
+ *
+ * The function stops the idle injection and waits for the threads to
+ * complete. If we are in the process of injecting an idle cycle, then
+ * this will wait the end of the cycle.
+ *
+ * When the function returns there is no more idle injection
+ * activity. The kthreads are scheduled out and the periodic timer is
+ * off.
+ */
+void idle_inject_stop(struct idle_inject_device *ii_dev)
+{
+ struct idle_inject_thread *iit;
+ unsigned int cpu;
+
+ pr_debug("Stopping injecting idle cycles on CPUs '%*pbl'\n",
+ cpumask_pr_args(to_cpumask(ii_dev->cpumask)));
+
+ hrtimer_cancel(&ii_dev->timer);
+
+ /*
+ * We want the guarantee we have a quiescent point where
+ * parked threads stay in there state while we are stopping
+ * the idle injection. After exiting the loop, if any CPU is
+ * plugged in, the 'should_run' boolean being false, the
+ * smpboot main loop schedules the task out.
+ */
+ cpu_hotplug_disable();
+
+ /*
+ * Iterate over all (online + offline) CPUs here in case a CPU
+ * got hotplugged out with "should_run" set to "1" and the
+ * thread may start running again after the ii_dev is freed
+ * and the CPU gets hotplugged in.
+ */
+ for_each_cpu(cpu, to_cpumask(ii_dev->cpumask)) {
+ iit = per_cpu_ptr(&idle_inject_thread, cpu);
+ iit->should_run = 0;
+
+ wait_task_inactive(iit->tsk, 0);
+ }
+
+ cpu_hotplug_enable();
+}
+
+/**
+ * idle_inject_setup - initialize the current task as a RT task
+ * @cpu: the CPU number where the kthread is running on (not used)
+ *
+ * Called one time, this function is in charge of setting the task
+ * scheduler parameters.
+ */
+static void idle_inject_setup(unsigned int cpu)
+{
+ struct sched_param param = { .sched_priority = MAX_USER_RT_PRIO / 2 };
+
+ sched_setscheduler(current, SCHED_FIFO, ¶m);
+}
+
+/**
+ * idle_inject_should_run - function helper for the smpboot API
+ * @cpu: the CPU number where the kthread is running on
+ *
+ * Return: a boolean telling if the thread can run.
+ */
+static int idle_inject_should_run(unsigned int cpu)
+{
+ struct idle_inject_thread *iit =
+ per_cpu_ptr(&idle_inject_thread, cpu);
+
+ return iit->should_run;
+}
+
+/**
+ * idle_inject_register - idle injection init routine
+ * @cpumask: the list of CPUs managed by the idle injection device
+ *
+ * This is the initialization function in charge of creating the
+ * initializing of the timer and allocate the structures. It does not
+ * starts the idle injection cycles.
+ *
+ * Return: NULL if an allocation fails.
+ */
+struct idle_inject_device *idle_inject_register(struct cpumask *cpumask)
+{
+ struct idle_inject_device *ii_dev;
+ int cpu, cpu_rb;
+
+ ii_dev = kzalloc(sizeof(*ii_dev) + cpumask_size(), GFP_KERNEL);
+ if (!ii_dev)
+ return NULL;
+
+ cpumask_copy(to_cpumask(ii_dev->cpumask), cpumask);
+ hrtimer_init(&ii_dev->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
+ ii_dev->timer.function = idle_inject_timer_fn;
+
+ for_each_cpu(cpu, to_cpumask(ii_dev->cpumask)) {
+
+ if (per_cpu(idle_inject_device, cpu)) {
+ pr_err("cpu%d is already registered\n", cpu);
+ goto out_rollback;
+ }
+
+ per_cpu(idle_inject_device, cpu) = ii_dev;
+ }
+
+ return ii_dev;
+
+out_rollback:
+ for_each_cpu(cpu_rb, to_cpumask(ii_dev->cpumask)) {
+ if (cpu == cpu_rb)
+ break;
+ per_cpu(idle_inject_device, cpu_rb) = NULL;
+ }
+
+ kfree(ii_dev);
+
+ return NULL;
+}
+
+/**
+ * idle_inject_unregister - Unregister the idle injection device
+ * @ii_dev: a pointer to an idle injection device
+ *
+ * The function is in charge of stopping the idle injections,
+ * unregister the kthreads and free the allocated memory in the
+ * register function.
+ */
+void idle_inject_unregister(struct idle_inject_device *ii_dev)
+{
+ unsigned int cpu;
+
+ idle_inject_stop(ii_dev);
+
+ for_each_cpu(cpu, to_cpumask(ii_dev->cpumask))
+ per_cpu(idle_inject_device, cpu) = NULL;
+
+ kfree(ii_dev);
+}
+
+static struct smp_hotplug_thread idle_inject_threads = {
+ .store = &idle_inject_thread.tsk,
+ .setup = idle_inject_setup,
+ .thread_fn = idle_inject_fn,
+ .thread_comm = "idle_inject/%u",
+ .thread_should_run = idle_inject_should_run,
+};
+
+static int __init idle_inject_init(void)
+{
+ return smpboot_register_percpu_thread(&idle_inject_threads);
+}
+early_initcall(idle_inject_init);
new file mode 100644
@@ -0,0 +1,29 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright (C) 2018 Linaro Ltd
+ *
+ * Author: Daniel Lezcano <daniel.lezcano@linaro.org>
+ *
+ */
+#ifndef __IDLE_INJECT_H__
+#define __IDLE_INJECT_H__
+
+/* private idle injection device structure */
+struct idle_inject_device;
+
+struct idle_inject_device *idle_inject_register(struct cpumask *cpumask);
+
+void idle_inject_unregister(struct idle_inject_device *ii_dev);
+
+int idle_inject_start(struct idle_inject_device *ii_dev);
+
+void idle_inject_stop(struct idle_inject_device *ii_dev);
+
+void idle_inject_set_duration(struct idle_inject_device *ii_dev,
+ unsigned int run_duration_ms,
+ unsigned int idle_duration_ms);
+
+void idle_inject_get_duration(struct idle_inject_device *ii_dev,
+ unsigned int *run_duration_ms,
+ unsigned int *idle_duration_ms);
+#endif /* __IDLE_INJECT_H__ */