/* ticker.c - microkernel tick event handler */ /* * Copyright (c) 1997-2010, 2012-2015 Wind River Systems, Inc. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions are met: * * 1) Redistributions of source code must retain the above copyright notice, * this list of conditions and the following disclaimer. * * 2) Redistributions in binary form must reproduce the above copyright notice, * this list of conditions and the following disclaimer in the documentation * and/or other materials provided with the distribution. * * 3) Neither the name of Wind River Systems nor the names of its contributors * may be used to endorse or promote products derived from this software without * specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE * POSSIBILITY OF SUCH DAMAGE. */ /* DESCRIPTION This module implements the microkernel's tick event handler. */ #include #include #include #include #include #include #include #include int64_t _k_sys_clock_tick_count = 0; #ifdef CONFIG_TIMESLICING static int32_t slice_count = (int32_t)0; static int32_t slice_time = (int32_t)CONFIG_TIMESLICE_SIZE; static kpriority_t slice_prio = (kpriority_t)CONFIG_TIMESLICE_PRIORITY; #endif /* CONFIG_TIMESLICING */ #ifdef CONFIG_TICKLESS_IDLE /* Number of ticks elapsed that have not been announced to the microkernel */ int32_t _sys_idle_elapsed_ticks = 0; /* Initial value must be 0 */ #endif #ifdef CONFIG_SYS_CLOCK_EXISTS int sys_clock_us_per_tick = 1000000 / sys_clock_ticks_per_sec; int sys_clock_hw_cycles_per_tick = sys_clock_hw_cycles_per_sec / sys_clock_ticks_per_sec; #else /* don't initialize to avoid division-by-zero error */ int sys_clock_us_per_tick; int sys_clock_hw_cycles_per_tick; #endif /******************************************************************************* * * task_cycle_get_32 - read the processor's high precision timer * * This routine reads the processor's high precision timer. It reads the * counter register on the timer device. This counter register increments * at a relatively high rate (e.g. 20 MHz), and thus is considered a * "high resolution" timer. This is in contrast to nano_tick_get_32() and * task_tick_get_32() which return the value of the kernel ticks variable. * * RETURNS: current high precision clock value */ uint32_t task_cycle_get_32(void) { return timer_read(); } /******************************************************************************* * * task_tick_get_32 - read the current system clock value * * This routine returns the lower 32-bits of the current system clock value * as measured in ticks. * * RETURNS: lower 32-bit of the current system clock value */ int32_t task_tick_get_32(void) { return (int32_t)_k_sys_clock_tick_count; } /******************************************************************************* * * task_tick_get - read the current system clock value * * This routine returns the current system clock value as measured in ticks. * * Interrupts are locked while updating clock since some CPUs do not support * native atomic operations on 64 bit values. * * RETURNS: current system clock value */ int64_t task_tick_get(void) { int64_t ticks; int key = irq_lock_inline(); ticks = _k_sys_clock_tick_count; irq_unlock_inline(key); return ticks; } /******************************************************************************* * * sys_clock_increment - increment system clock by "N" ticks * * Interrupts are locked while updating clock since some CPUs do not support * native atomic operations on 64 bit values. * * RETURNS: N/A */ static void sys_clock_increment(int inc) { int key = irq_lock_inline(); _k_sys_clock_tick_count += inc; irq_unlock_inline(key); } /******************************************************************************* * * _TlDebugUpdate - task level debugging tick handler * * If task level debugging is configured this routine updates the low resolution * debugging timer and determines if task level processing should be suspended. * * RETURNS: 0 if task level processing should be halted or 1 if not * * \NOMANUAL */ #ifdef CONFIG_TASK_DEBUG uint32_t __noinit _k_debug_sys_clock_tick_count; static inline int _TlDebugUpdate(int32_t ticks) { _k_debug_sys_clock_tick_count += ticks; return !_k_debug_halt; } #else #define _TlDebugUpdate(ticks) 1 #endif /******************************************************************************* * * _TimeSliceUpdate - tick handler time slice logic * * This routine checks to see if it is time for the current task * to relinquish control, and yields CPU if so. * * RETURNS: N/A * * \NOMANUAL */ static inline void _TimeSliceUpdate(void) { #ifdef CONFIG_TIMESLICING int yield = slice_time && (_k_current_task->Prio >= slice_prio) && (++slice_count >= slice_time); if (yield) { slice_count = 0; _k_task_yield(NULL); } #else /* do nothing */ #endif /* CONFIG_TIMESLICING */ } /******************************************************************************* * * _SysIdleElapsedTicksGet - get elapsed ticks * * If tickless idle support is configured this routine returns the number * of ticks since going idle and then resets the global elapsed tick counter back * to zero indicating all elapsed ticks have been consumed. This is done with * interrupts locked to prevent the timer ISR from modifying the global elapsed * tick counter. * If tickless idle support is not configured in it simply returns 1. * * RETURNS: number of ticks to process * * \NOMANUAL */ static inline int32_t _SysIdleElapsedTicksGet(void) { #ifdef CONFIG_TICKLESS_IDLE int32_t ticks; int key; key = irq_lock(); ticks = _sys_idle_elapsed_ticks; _sys_idle_elapsed_ticks = 0; irq_unlock(key); return ticks; #else /* A single tick always elapses when not in tickless mode */ return 1; #endif } /******************************************************************************* * * K_ticker - microkernel tick handler * * This routine informs other microkernel subsystems that a tick event has * occurred. * * RETURNS: 1 */ int K_ticker(int event) { (void)event; /* prevent "unused argument" compiler warning */ int32_t ticks; ticks = _SysIdleElapsedTicksGet(); _k_workload_monitor_update(); if (_TlDebugUpdate(ticks)) { _TimeSliceUpdate(); _k_timer_list_update(ticks); sys_clock_increment(ticks); } return 1; } #ifdef CONFIG_TIMESLICING /******************************************************************************* * * scheduler_time_slice_set - set time slicing period and scope * * This routine controls how task time slicing is performed by the task * scheduler, by specifying the maximum time slice length (in ticks) and * the highest priority task level for which time slicing is performed. * * To enable time slicing, a non-zero time slice length must be specified. * The task scheduler then ensures that no executing task runs for more than * the specified number of ticks before giving other tasks of that priority * a chance to execute. (However, any task whose priority is higher than the * specified task priority level is exempted, and may execute as long as * desired without being pre-empted due to time slicing.) * * Time slicing only limits that maximum amount of time a task may continuously * execute. Once the scheduler selects a task for execution, there is no minimum * guaranteed time the task will execute before tasks of greater or equal * priority are scheduled. * * If the currently executing task is the only one of that priority eligible * for execution this routine has no effect, as that task will be immediately * rescheduled once the slice period expires. * * To disable timeslicing, call the API with both parameters set to zero. * * RETURNS: N/A */ void scheduler_time_slice_set(int32_t t, /* time slice in ticks */ kpriority_t p /* beginning priority level to which time slicing applies */ ) { slice_time = t; slice_prio = p; } #endif /* CONFIG_TIMESLICING */ /******************************************************************************* * * _k_time_elapse - handle elapsed ticks calculation request * * This routine, called by K_swapper(), handles the request for calculating the * time elapsed since the specified reference time. * * RETURNS: N/A */ void _k_time_elapse(struct k_args *P) { int64_t now = task_tick_get(); P->Args.c1.time2 = now - P->Args.c1.time1; P->Args.c1.time1 = now; } /******************************************************************************* * * task_tick_delta - return ticks between calls * * This function is meant to be used in contained fragments of code. The first * call to it in a particular code fragment fills in a reference time variable * which then gets passed and updated every time the function is called. From * the second call on, the delta between the value passed to it and the current * tick count is the return value. Since the first call is meant to only fill in * the reference time, its return value should be discarded. * * Since a code fragment that wants to use task_tick_delta() passes in its * own reference time variable, multiple code fragments can make use of this * function concurrently. * * Note that it is not necessary to allocate a timer to use this call. * * RETURNS: elapsed time in system ticks */ int64_t task_tick_delta(int64_t *reftime /* pointer to reference time */ ) { struct k_args A; A.Comm = ELAPSE; A.Args.c1.time1 = *reftime; KERNEL_ENTRY(&A); *reftime = A.Args.c1.time1; return A.Args.c1.time2; }