zephyr/include/sys_clock.h

249 lines
5.7 KiB
C
Raw Normal View History

/*
* Copyright (c) 2014-2015 Wind River Systems, Inc.
*
* SPDX-License-Identifier: Apache-2.0
*/
/**
* @file
* @brief Variables needed needed for system clock
*
*
* Declare variables used by both system timer device driver and kernel
* components that use timer functionality.
*/
#ifndef ZEPHYR_INCLUDE_SYS_CLOCK_H_
#define ZEPHYR_INCLUDE_SYS_CLOCK_H_
#include <sys/util.h>
#include <sys/dlist.h>
#include <toolchain.h>
#include <zephyr/types.h>
#ifdef __cplusplus
extern "C" {
#endif
kernel: tickless: Add tickless kernel support Adds event based scheduling logic to the kernel. Updates management of timeouts, timers, idling etc. based on time tracked at events rather than periodic ticks. Provides interfaces for timers to announce and get next timer expiry based on kernel scheduling decisions involving time slicing of threads, timeouts and idling. Uses wall time units instead of ticks in all scheduling activities. The implementation involves changes in the following areas 1. Management of time in wall units like ms/us instead of ticks The existing implementation already had an option to configure number of ticks in a second. The new implementation builds on top of that feature and provides option to set the size of the scheduling granurality to mili seconds or micro seconds. This allows most of the current implementation to be reused. Due to this re-use and co-existence with tick based kernel, the names of variables may contain the word "tick". However, in the tickless kernel implementation, it represents the currently configured time unit, which would be be mili seconds or micro seconds. The APIs that take time as a parameter are not impacted and they continue to pass time in mili seconds. 2. Timers would not be programmed in periodic mode generating ticks. Instead they would be programmed in one shot mode to generate events at the time the kernel scheduler needs to gain control for its scheduling activities like timers, timeouts, time slicing, idling etc. 3. The scheduler provides interfaces that the timer drivers use to announce elapsed time and get the next time the scheduler needs a timer event. It is possible that the scheduler may not need another timer event, in which case the system would wait for a non-timer event to wake it up if it is idling. 4. New APIs are defined to be implemented by timer drivers. Also they need to handler timer events differently. These changes have been done in the HPET timer driver. In future other timers that support tickles kernel should implement these APIs as well. These APIs are to re-program the timer, update and announce elapsed time. 5. Philosopher and timer_api applications have been enabled to test tickless kernel. Separate configuration files are created which define the necessary CONFIG flags. Run these apps using following command make pristine && make BOARD=qemu_x86 CONF_FILE=prj_tickless.conf qemu Jira: ZEP-339 ZEP-1946 ZEP-948 Change-Id: I7d950c31bf1ff929a9066fad42c2f0559a2e5983 Signed-off-by: Ramesh Thomas <ramesh.thomas@intel.com>
2017-02-05 19:37:19 -08:00
#ifdef CONFIG_TICKLESS_KERNEL
extern int _sys_clock_always_on;
extern void z_enable_sys_clock(void);
kernel: tickless: Add tickless kernel support Adds event based scheduling logic to the kernel. Updates management of timeouts, timers, idling etc. based on time tracked at events rather than periodic ticks. Provides interfaces for timers to announce and get next timer expiry based on kernel scheduling decisions involving time slicing of threads, timeouts and idling. Uses wall time units instead of ticks in all scheduling activities. The implementation involves changes in the following areas 1. Management of time in wall units like ms/us instead of ticks The existing implementation already had an option to configure number of ticks in a second. The new implementation builds on top of that feature and provides option to set the size of the scheduling granurality to mili seconds or micro seconds. This allows most of the current implementation to be reused. Due to this re-use and co-existence with tick based kernel, the names of variables may contain the word "tick". However, in the tickless kernel implementation, it represents the currently configured time unit, which would be be mili seconds or micro seconds. The APIs that take time as a parameter are not impacted and they continue to pass time in mili seconds. 2. Timers would not be programmed in periodic mode generating ticks. Instead they would be programmed in one shot mode to generate events at the time the kernel scheduler needs to gain control for its scheduling activities like timers, timeouts, time slicing, idling etc. 3. The scheduler provides interfaces that the timer drivers use to announce elapsed time and get the next time the scheduler needs a timer event. It is possible that the scheduler may not need another timer event, in which case the system would wait for a non-timer event to wake it up if it is idling. 4. New APIs are defined to be implemented by timer drivers. Also they need to handler timer events differently. These changes have been done in the HPET timer driver. In future other timers that support tickles kernel should implement these APIs as well. These APIs are to re-program the timer, update and announce elapsed time. 5. Philosopher and timer_api applications have been enabled to test tickless kernel. Separate configuration files are created which define the necessary CONFIG flags. Run these apps using following command make pristine && make BOARD=qemu_x86 CONF_FILE=prj_tickless.conf qemu Jira: ZEP-339 ZEP-1946 ZEP-948 Change-Id: I7d950c31bf1ff929a9066fad42c2f0559a2e5983 Signed-off-by: Ramesh Thomas <ramesh.thomas@intel.com>
2017-02-05 19:37:19 -08:00
#endif
#if defined(CONFIG_TIMER_READS_ITS_FREQUENCY_AT_RUNTIME)
__syscall int z_clock_hw_cycles_per_sec_runtime_get(void);
static inline int z_impl_z_clock_hw_cycles_per_sec_runtime_get(void)
{
extern int z_clock_hw_cycles_per_sec;
return z_clock_hw_cycles_per_sec;
}
#endif /* CONFIG_TIMER_READS_ITS_FREQUENCY_AT_RUNTIME */
static inline int sys_clock_hw_cycles_per_sec(void)
{
#if defined(CONFIG_TIMER_READS_ITS_FREQUENCY_AT_RUNTIME)
return z_clock_hw_cycles_per_sec_runtime_get();
#else
return CONFIG_SYS_CLOCK_HW_CYCLES_PER_SEC;
#endif
}
/* Note that some systems with comparatively slow cycle counters
* experience precision loss when doing math like this. In the
* general case it is not correct that "cycles" are much faster than
* "ticks".
*/
static inline int sys_clock_hw_cycles_per_tick(void)
{
#ifdef CONFIG_SYS_CLOCK_EXISTS
return sys_clock_hw_cycles_per_sec() / CONFIG_SYS_CLOCK_TICKS_PER_SEC;
#else
return 1; /* Just to avoid a division by zero */
#endif
}
#if defined(CONFIG_SYS_CLOCK_EXISTS) && \
(CONFIG_SYS_CLOCK_HW_CYCLES_PER_SEC == 0)
#error "SYS_CLOCK_HW_CYCLES_PER_SEC must be non-zero!"
#endif
/* number of nsec per usec */
#define NSEC_PER_USEC 1000U
/* number of microseconds per millisecond */
#define USEC_PER_MSEC 1000U
/* number of milliseconds per second */
#define MSEC_PER_SEC 1000U
/* number of microseconds per second */
#define USEC_PER_SEC ((USEC_PER_MSEC) * (MSEC_PER_SEC))
/* number of nanoseconds per second */
#define NSEC_PER_SEC ((NSEC_PER_USEC) * (USEC_PER_MSEC) * (MSEC_PER_SEC))
/* kernel clocks */
/*
* We default to using 64-bit intermediates in timescale conversions,
* but if the HW timer cycles/sec, ticks/sec and ms/sec are all known
* to be nicely related, then we can cheat with 32 bits instead.
*/
#ifdef CONFIG_SYS_CLOCK_EXISTS
#if defined(CONFIG_TIMER_READS_ITS_FREQUENCY_AT_RUNTIME) || \
(MSEC_PER_SEC % CONFIG_SYS_CLOCK_TICKS_PER_SEC) || \
(CONFIG_SYS_CLOCK_HW_CYCLES_PER_SEC % CONFIG_SYS_CLOCK_TICKS_PER_SEC)
#define _NEED_PRECISE_TICK_MS_CONVERSION
#endif
#endif
static ALWAYS_INLINE s32_t z_ms_to_ticks(s32_t ms)
{
#ifdef CONFIG_SYS_CLOCK_EXISTS
#ifdef _NEED_PRECISE_TICK_MS_CONVERSION
int cyc = sys_clock_hw_cycles_per_sec();
/* use 64-bit math to keep precision */
return (s32_t)ceiling_fraction((s64_t)ms * cyc,
((s64_t)MSEC_PER_SEC * cyc) / CONFIG_SYS_CLOCK_TICKS_PER_SEC);
#else
/* simple division keeps precision */
s32_t ms_per_tick = MSEC_PER_SEC / CONFIG_SYS_CLOCK_TICKS_PER_SEC;
return (s32_t)ceiling_fraction(ms, ms_per_tick);
#endif
#else
__ASSERT(ms == 0, "ms not zero");
return 0;
#endif
}
static inline u64_t __ticks_to_ms(s64_t ticks)
{
#ifdef CONFIG_SYS_CLOCK_EXISTS
return (u64_t)ticks * MSEC_PER_SEC /
(u64_t)CONFIG_SYS_CLOCK_TICKS_PER_SEC;
#else
__ASSERT(ticks == 0, "ticks not zero");
return 0ULL;
#endif
}
/*
* These are only currently used by k_usleep(), but they are
* defined here for parity with their ms analogs above. Note:
* we don't bother trying the 32-bit intermediate shortcuts
* possible with ms, because of the magnitudes involved.
*/
static inline s32_t z_us_to_ticks(s32_t us)
{
#ifdef CONFIG_SYS_CLOCK_EXISTS
return (s32_t) ceiling_fraction(
(s64_t)us * sys_clock_hw_cycles_per_sec(),
((s64_t)USEC_PER_SEC * sys_clock_hw_cycles_per_sec()) /
CONFIG_SYS_CLOCK_TICKS_PER_SEC);
#else
__ASSERT(us == 0, "us not zero");
return 0;
#endif
}
static inline s32_t __ticks_to_us(s32_t ticks)
{
#ifdef CONFIG_SYS_CLOCK_EXISTS
return (s32_t) ((s64_t)ticks * USEC_PER_SEC /
(s64_t)CONFIG_SYS_CLOCK_TICKS_PER_SEC);
#else
__ASSERT(ticks == 0, "ticks not zero");
return 0;
#endif
}
/* added tick needed to account for tick in progress */
#define _TICK_ALIGN 1
/* SYS_CLOCK_HW_CYCLES_TO_NS64 converts CPU clock cycles to nanoseconds */
#define SYS_CLOCK_HW_CYCLES_TO_NS64(X) \
(((u64_t)(X) * NSEC_PER_SEC) / sys_clock_hw_cycles_per_sec())
/*
* SYS_CLOCK_HW_CYCLES_TO_NS_AVG converts CPU clock cycles to nanoseconds
* and calculates the average cycle time
*/
#define SYS_CLOCK_HW_CYCLES_TO_NS_AVG(X, NCYCLES) \
(u32_t)(SYS_CLOCK_HW_CYCLES_TO_NS64(X) / NCYCLES)
/**
* @defgroup clock_apis Kernel Clock APIs
* @ingroup kernel_apis
* @{
*/
/**
* @brief Compute nanoseconds from hardware clock cycles.
*
* This macro converts a time duration expressed in hardware clock cycles
* to the equivalent duration expressed in nanoseconds.
*
* @param X Duration in hardware clock cycles.
*
* @return Duration in nanoseconds.
*/
#define SYS_CLOCK_HW_CYCLES_TO_NS(X) (u32_t)(SYS_CLOCK_HW_CYCLES_TO_NS64(X))
/**
* @} end defgroup clock_apis
*/
/**
*
* @brief Return the lower part of the current system tick count
*
* @return the current system tick count
*
*/
u32_t z_tick_get_32(void);
/**
*
* @brief Return the current system tick count
*
* @return the current system tick count
*
*/
s64_t z_tick_get(void);
#ifndef CONFIG_SYS_CLOCK_EXISTS
#define z_tick_get() (0)
#define z_tick_get_32() (0)
#endif
/* timeouts */
struct _timeout;
typedef void (*_timeout_func_t)(struct _timeout *t);
struct _timeout {
sys_dnode_t node;
s32_t dticks;
_timeout_func_t fn;
};
#ifdef __cplusplus
}
#endif
#include <syscalls/sys_clock.h>
#endif /* ZEPHYR_INCLUDE_SYS_CLOCK_H_ */