Skip to content

Conversation

@wangchdo
Copy link
Contributor

@wangchdo wangchdo commented Dec 20, 2025

Summary

This PR introduces high-resolution timer (hrtimer) support as a fully independent and optional module for support of the scheduler, without affecting existing scheduler behavior.

Hrtimer is strictly isolated from the current scheduling logic:

  1. All hrtimer code is enabled only when CONFIG_HRTIMER is set.
  2. If CONFIG_HRTIMER is disabled, the scheduler continues to use the existing tick-based or tickless mechanisms with no changes.
  3. When enabled, hrtimer reuses the existing scheduler logic to drive OS ticks via a dedicated hrtimer instance.

The module does not modify any scheduler data structures or timing paths.
Hrtimer acts solely as an alternative time source. Core scheduler functions (nxsched_process_tick(), nxsched_tick_expiration(), etc.) remain unchanged and are reused as-is.

Additional safeguards:

  1. The scheduler hrtimer is initialized lazily and does not impact system startup.
  2. Existing timer infrastructure is not replaced or bypassed unless explicitly configured.
  3. Any failure or imperfection in hrtimer is confined to the module itself and cannot affect legacy scheduling paths.

Integration benefit

This design enables incremental development and review of hrtimer while ensuring that existing NuttX scheduling behavior remains stable even if the hrtimer feature is explicitly enabled.

Development benefit
With this design, developers interested in optimizing the scheduler and those focused on optimizing hrtimer can work independently on their respective improvements.

One other key update
This PR also includes an improvement(also in a seperate PR #17570) to hrtimer by refining its state machine, this is to fix some issues in SMP mode found by @Fix-Point. The refined state-machine is as shown below, and the corresponding diagram is also added in the hrtimer documentation.

image

Impact

Add hrtimer support to nuttx scheduelr, without altering the existing scheduler behavior.

Testing

Test 1 passed (integrated in ostest):

- test implementation:

/****************************************************************************
 * apps/testing/ostest/hrtimer.c
 *
 * SPDX-License-Identifier: Apache-2.0
 *
 * Licensed to the Apache Software Foundation (ASF) under one or more
 * contributor license agreements.  See the NOTICE file distributed with
 * this work for additional information regarding copyright ownership.  The
 * ASF licenses this file to you under the Apache License, Version 2.0 (the
 * "License"); you may not use this file except in compliance with the
 * License.  You may obtain a copy of the License at
 *
 *   http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
 * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.  See the
 * License for the specific language governing permissions and limitations
 * under the License.
 *
 ****************************************************************************/

/****************************************************************************
 * Included Files
 ****************************************************************************/

#include <nuttx/config.h>
#include <nuttx/hrtimer.h>

#include <stdio.h>
#include <sched.h>

#include "ostest.h"

/****************************************************************************
 * Pre-processor Definitions
 ****************************************************************************/

#define NSEC_PER_500MS (500 * NSEC_PER_MSEC)

/* Set a 1ms margin to allow hrtimertest to pass in QEMU.
 *
 * QEMU is a virtual platform, and its timer resolution and scheduling
 * latency may be less precise than on real hardware. Using a larger
 * margin ensures that tests do not fail due to timing inaccuracies.
 *
 * On real hardware (verified on the a2g-tc397-5v-tft board), this
 * margin can be reduced to less than 5 ns because timers are precise
 * and deterministic.
 */

#define NSEC_MARGIN    (NSEC_PER_MSEC)

/* Simple assertion macro for HRTimer test cases */
#define HRTIMER_TEST(expr, value)                                   \
  do                                                                \
    {                                                               \
      ret = (expr);                                                 \
      if (ret != (value))                                           \
        {                                                           \
          printf("ERROR: HRTimer test failed, line=%d ret=%d\n",   \
                 __LINE__, ret);                                    \
          ASSERT(false);                                            \
        }                                                           \
    }                                                               \
  while (0)

/****************************************************************************
 * Private Types
 ****************************************************************************/

/* Structure for HRTimer test tracking */

struct hrtimer_test_s
{
  hrtimer_t timer;    /* HRTimer instance */
  uint64_t  previous; /* Previous timestamp in nanoseconds */
  uint32_t  count;    /* Number of timer expirations */
  uint32_t  period;   /* Expected period between expirations */
  bool      active;   /* True while the test is still running */
};

/****************************************************************************
 * Private Functions
 ****************************************************************************/

/****************************************************************************
 * Name: hrtimer_test_init
 *
 * Description:
 *   Initialize a hrtimer_test_s structure for a new test.
 *
 * Input Parameters:
 *   test_hrtimer - Pointer to the test structure to initialize
 *   period       - Expected timer period in nanoseconds
 *
 * Returned Value:
 *   None
 *
 ****************************************************************************/

static void hrtimer_test_init(FAR struct hrtimer_test_s *test_hrtimer,
                              uint32_t period)
{
  test_hrtimer->previous = 0;
  test_hrtimer->count    = 0;
  test_hrtimer->active   = true;
  test_hrtimer->period   = period;
}

/****************************************************************************
 * Name: test_hrtimer_callback
 *
 * Description:
 *   HRTimer callback function for test.
 *
 *   - Verifies the timer interval is exactly 500ms (nanosecond precision)
 *   - Stops the test after 15 expirations
 *   - Re-arms the timer in absolute mode
 *
 * Input Parameters:
 *   hrtimer - Pointer to the expired HRTimer instance
 *
 * Returned Value:
 *   Timer period in nanoseconds (NSEC_PER_500MS)
 *
 ****************************************************************************/

static uint32_t test_hrtimer_callback(FAR hrtimer_t *hrtimer)
{
  struct timespec ts;
  uint32_t diff;
  uint64_t now;
  int ret;

  FAR struct hrtimer_test_s *test =
    (FAR struct hrtimer_test_s *)hrtimer;

  /* Increment expiration count */

  test->count++;

  /* Get current system time */

  clock_systime_timespec(&ts);
  now = clock_time2nsec(&ts);

  /* Skip comparison for first two invocations */

  if (test->count > 2)
    {
      /* Verify the timer interval is exactly
       * 500ms with nsec resolution
       */

      diff = (uint32_t)(now - test->previous);

      HRTIMER_TEST(NSEC_PER_500MS < diff + NSEC_MARGIN, true);
      HRTIMER_TEST(NSEC_PER_500MS > diff - NSEC_MARGIN, true);
    }

  test->previous = now;

  /* Stop the test after 15 expirations */

  if (test->count  >= 15)
    {
      ret = hrtimer_cancel(hrtimer);
      HRTIMER_TEST(ret, 0);

      test->active = false;
    }

  return test->period;
}

/****************************************************************************
 * Public Functions
 ****************************************************************************/

/****************************************************************************
 * Name: hrtimer_test
 *
 * Description:
 *   Entry point for high-resolution timer functional test.
 *
 *   - Initializes a HRTimer
 *   - Starts it with a 500ms relative timeout
 *   - Verifies subsequent expirations occur at 500ms intervals
 *
 * Input Parameters:
 *   None
 *
 * Returned Value:
 *   None
 *
 ****************************************************************************/

void hrtimer_test(void)
{
  int ret;
  struct hrtimer_test_s test_hrtimer_500ms;

  /* Initialize test structure */

  hrtimer_test_init(&test_hrtimer_500ms, NSEC_PER_500MS);

  /* Initialize the high-resolution timer */

  hrtimer_init(&test_hrtimer_500ms.timer,
               test_hrtimer_callback,
               NULL);

  /* Start the timer with 500ms relative timeout */

  ret = hrtimer_start(&test_hrtimer_500ms.timer,
                      test_hrtimer_500ms.period,
                      HRTIMER_MODE_REL);

  HRTIMER_TEST(ret, OK);

  /* Wait until the test completes */

  while (test_hrtimer_500ms.active)
    {
      usleep(500 * USEC_PER_MSEC);
    }
}

test log on rv-virt:smp64:

NuttShell (NSH)
nsh> 
nsh> uname -a
NuttX 0.0.0 6847a0cc95-dirty Dec 20 2025 12:26:39 risc-v rv-virt
nsh> ostest

(...)

user_main: hrtimer test

End of test memory usage:
VARIABLE  BEFORE   AFTER
======== ======== ========
arena     1fbdec0  1fbdec0
ordblks         9        9
mxordblk  1f73880  1f73880
uordblks    109b0     ff18
fordblks  1fad510  1fadfa8

Final memory usage:
VARIABLE  BEFORE   AFTER
======== ======== ========
arena     1fbdec0  1fbdec0
ordblks         1        9
mxordblk  1fb2cf8  1f73880
uordblks     b1c8     ff18
fordblks  1fb2cf8  1fadfa8
user_main: Exiting
ostest_main: Exiting with status 0

test 2 passed (provided by @Fix-Point )

test implementation

/****************************************************************************
 * apps/examples/hello/hello_main.c
 *
 * SPDX-License-Identifier: Apache-2.0
 *
 * Licensed to the Apache Software Foundation (ASF) under one or more
 * contributor license agreements.  See the NOTICE file distributed with
 * this work for additional information regarding copyright ownership.  The
 * ASF licenses this file to you under the Apache License, Version 2.0 (the
 * "License"); you may not use this file except in compliance with the
 * License.  You may obtain a copy of the License at
 *
 *   http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
 * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.  See the
 * License for the specific language governing permissions and limitations
 * under the License.
 *
 ****************************************************************************/

/****************************************************************************
 * Included Files
 ****************************************************************************/

#include <nuttx/config.h>
#include <stdio.h>
#include <unistd.h>

#include <nuttx/hrtimer.h>

#define HRTIMER_TEST_THREAD_NR (1)
#define HRTIMER_TEST_NR        (1000000)

/****************************************************************************
 * Public Functions
 ****************************************************************************/

static int volatile tcount = 0;
static volatile uint32_t next = 0;

static uint32_t test_callback(FAR struct hrtimer_s *hrtimer)
{

  tcount++;
  up_ndelay(hrtimer->expired % (10 * NSEC_PER_USEC));

  return 0;
}

static uint32_t test_callback_background(FAR struct hrtimer_s *hrtimer)
{
  up_ndelay(hrtimer->expired % NSEC_PER_USEC);
  return 0;
}


static void test1(int tid)
{
  hrtimer_t   timer;
  int         count = 0;
  irqstate_t flags;
  spinlock_t lock;

  if (tid == 0)
    {
      hrtimer_init(&timer, test_callback, NULL);
    }
  else
    {
      hrtimer_init(&timer, test_callback_background, NULL);
    }

  while (count++ < HRTIMER_TEST_NR)
    {
      int ret;
      if (tid == 0)
        {
          uint64_t delay = rand() % (10 * NSEC_PER_MSEC);

          /* Simulate the periodical hrtimer.. */

          flags = spin_lock_irqsave(&lock);

          /* Use as periodical timer */

          ret = hrtimer_cancel(&timer);
          ret = hrtimer_start(&timer, 1000, HRTIMER_MODE_REL);

          spin_unlock_irqrestore(&lock, flags);

          up_ndelay(NSEC_PER_MSEC);

          flags = spin_lock_irqsave(&lock);

          ret = hrtimer_cancel_sync(&timer);
          ret = hrtimer_start(&timer, 1000, HRTIMER_MODE_REL);
          spin_unlock_irqrestore(&lock, flags);

          up_ndelay(NSEC_PER_MSEC);

          hrtimer_cancel_sync(&timer); // stucked here????
          printf("???\n");
        }
      else
        {
          /* Simulate the background hrtimer.. */

          uint64_t delay = rand() % (10 * NSEC_PER_MSEC);

          ret = hrtimer_cancel(&timer);
          ret = hrtimer_start(&timer, delay, HRTIMER_MODE_REL);
        }

      UNUSED(ret);
    }
}

static void* test_thread(void *arg)
{
  while (1)
    {
      test1((int)arg);
    }
  return NULL;
}
/****************************************************************************
 * hello_main
 ****************************************************************************/

int main(int argc, FAR char *argv[])
{
  unsigned int   thread_id;
  pthread_attr_t attr;
  pthread_t      pthreads[HRTIMER_TEST_THREAD_NR];

  printf("hrtimer_test start...\n");

  ASSERT(pthread_attr_init(&attr) == 0);

  /* Create wdog test thread */

  for (thread_id = 0; thread_id < HRTIMER_TEST_THREAD_NR; thread_id++)
    {
      ASSERT(pthread_create(&pthreads[thread_id], &attr,
                            test_thread, (void *)thread_id) == 0);
    }

  for (thread_id = 0; thread_id < HRTIMER_TEST_THREAD_NR; thread_id++)
    {
      pthread_join(pthreads[thread_id], NULL);
    }

  ASSERT(pthread_attr_destroy(&attr) == 0);

  printf("hrtimer_test end...\n");
  return 0;
}

test passed log on rv-virt:smp64

nsh> uname -a
NuttX 0.0.0 6847a0cc95-dirty Dec 20 2025 12:26:39 risc-v rv-virt
nsh> 
nsh> hello
???
???
???
(...)

test 3 passed (provided by @Fix-Point )

test implementation

/****************************************************************************
 * apps/examples/hello/hello_main.c
 *
 * SPDX-License-Identifier: Apache-2.0
 *
 * Licensed to the Apache Software Foundation (ASF) under one or more
 * contributor license agreements.  See the NOTICE file distributed with
 * this work for additional information regarding copyright ownership.  The
 * ASF licenses this file to you under the Apache License, Version 2.0 (the
 * "License"); you may not use this file except in compliance with the
 * License.  You may obtain a copy of the License at
 *
 *   http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
 * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.  See the
 * License for the specific language governing permissions and limitations
 * under the License.
 *
 ****************************************************************************/

/****************************************************************************
 * Included Files
 ****************************************************************************/

#include <nuttx/config.h>
#include <stdio.h>

#include <nuttx/hrtimer.h>

#define HRTIMER_TEST_THREAD_NR (CONFIG_SMP_NCPUS * 8)

/****************************************************************************
 * Public Functions
 ****************************************************************************/

static uint32_t test_callback(FAR struct hrtimer_s *hrtimer)
{
  printf("test callback ...\n");
  return 0;
}

static void* test_thread(void *arg)
{
  irqstate_t  flags;
  hrtimer_t   timer;
  spinlock_t  lock = SP_UNLOCKED;
  hrtimer_init(&timer, test_callback, NULL);
  while (1)
    {
      uint64_t delay = rand() % NSEC_PER_MSEC;
      int ret;

      /* Simulate the usage of driver->wait_dog. */

      flags = spin_lock_irqsave(&lock);

      /* The driver lock acquired */

      /* First try, failed. Because hrtimer_start can not ensure the timer being started. */

      ret = hrtimer_cancel(&timer);
      // ret = hrtimer_start(&timer, 10 * NSEC_PER_USEC, HRTIMER_MODE_REL); /* May fail */

      /* This try-loop start should be OK. But it failed again.
       * Besides, we can not sleep or spin in the critical sections.
       */

      while (hrtimer_start(&timer, 10 * NSEC_PER_USEC, HRTIMER_MODE_REL) != OK);
      ret = OK;

      /* Second try, Success, but we can not sleep or spin in the critical section. */

      // ret = hrtimer_cancel_sync(&timer); /* Sleep in critical sections */
      // ret = hrtimer_start(&timer, delay, HRTIMER_MODE_REL);


      spin_unlock_irqrestore(&lock, flags);

      if (ret != OK)
        {
          printf("hrtimer_start failed\n");
        }
      up_ndelay(delay);
    }
  return NULL;
}

/****************************************************************************
 * hello_main
 ****************************************************************************/

int main(int argc, FAR char *argv[])
{
  unsigned int   thread_id;
  pthread_attr_t attr;
  pthread_t      pthreads[HRTIMER_TEST_THREAD_NR];

  printf("hrtimer_test start...\n");

  ASSERT(pthread_attr_init(&attr) == 0);

  /* Create wdog test thread */

  for (thread_id = 0; thread_id < HRTIMER_TEST_THREAD_NR; thread_id++)
    {
      ASSERT(pthread_create(&pthreads[thread_id], &attr,
                            test_thread, NULL) == 0);
    }

  for (thread_id = 0; thread_id < HRTIMER_TEST_THREAD_NR; thread_id++)
    {
      pthread_join(pthreads[thread_id], NULL);
    }

  ASSERT(pthread_attr_destroy(&attr) == 0);

  printf("hrtimer_test end...\n");
  return 0;
}

test passed log on rv-virt:smp64

NuttShell (NSH)
nsh> 
nsh> 
nsh> uname -a
NuttX 0.0.0 6847a0cc95-dirty Dec 20 2025 12:33:59 risc-v rv-virt
nsh> hello
test callback ...
test callback ...
test callback ...
test callback ...
test callback ...
test callback ...
test callback ...
test callback ...
test callback ...
test callback ...
test callback ...

(....)

@github-actions github-actions bot added Area: Documentation Improvements or additions to documentation Area: Drivers Drivers issues Area: OS Components OS Components issues Size: M The size of the change in this PR is medium labels Dec 20, 2025
@wangchdo wangchdo force-pushed the add_hrtimer_sched_support branch from 6847a0c to c073a95 Compare December 20, 2025 04:54
@wangchdo wangchdo changed the title sched: add hrtimer support to os scheduler sched/hrtimer: Part2: add hrtimer support to os scheduler Dec 20, 2025
@wangchdo wangchdo changed the title sched/hrtimer: Part2: add hrtimer support to os scheduler sched/hrtimer: Part2: introduce hrtimer support to os scheduler Dec 20, 2025
@wangchdo wangchdo force-pushed the add_hrtimer_sched_support branch from c073a95 to f1e9f9a Compare December 20, 2025 05:52
@wangchdo wangchdo force-pushed the add_hrtimer_sched_support branch from f1e9f9a to e0f1573 Compare December 20, 2025 05:55
@wangchdo wangchdo force-pushed the add_hrtimer_sched_support branch from e0f1573 to 424634c Compare December 20, 2025 06:04
@wangchdo wangchdo force-pushed the add_hrtimer_sched_support branch from 424634c to ba4a61f Compare December 20, 2025 06:11
@wangchdo wangchdo force-pushed the add_hrtimer_sched_support branch from ba4a61f to 9bcc6ae Compare December 20, 2025 06:53
@wangchdo wangchdo changed the title sched/hrtimer: Part2: introduce hrtimer support to os scheduler sched/hrtimer: Part2: introduce hrtimer support for os scheduler Dec 20, 2025
@wangchdo wangchdo changed the title sched/hrtimer: Part2: introduce hrtimer support for os scheduler sched/hrtimer: Part2: introduce hrtimer support for scheduler Dec 20, 2025
@wangchdo wangchdo force-pushed the add_hrtimer_sched_support branch from 9bcc6ae to 2c0c6b0 Compare December 20, 2025 11:21
Allow running/armed hrtimer to be restarted to
fix hrtimer bug: apache#17567

Signed-off-by: Chengdong Wang <wangchengdong@lixiang.com>
  Update the hrtimer documentation to describe the hrtimer state machine,
  which is introduced to handle safe cancellation and execution in SMP
  environments.

Signed-off-by: Chengdong Wang <wangchengdong@lixiang.com>
  Improve hrtimer to only enable SMP related logic when in SMP mode to
  improve performance in non-SMP mode

Signed-off-by: Chengdong Wang <wangchengdong@lixiang.com>
@wangchdo wangchdo force-pushed the add_hrtimer_sched_support branch from 32bc91f to 4f3dcdd Compare December 24, 2025 12:26
@wangchdo
Copy link
Contributor Author

wangchdo commented Dec 24, 2025

@cederom @acassis @anchao @xiaoxiang781216

Conflicts resolved, please double check

@wangchdo wangchdo force-pushed the add_hrtimer_sched_support branch from 4f3dcdd to 19097a7 Compare December 24, 2025 12:31
acassis
acassis previously approved these changes Dec 24, 2025
@acassis
Copy link
Contributor

acassis commented Dec 24, 2025

Thank you @wangchdo amazing work and perfect reporting! :-)
I have one issue, as there is a rename of very old and important file and functions inside (sched/sched/sched_processtimer.c → sched/sched/sched_processtick.c) this seems to be a breaking change, thus both PR and related git commit topic should contain ! as the first character, plus description body should contain BREAKING CHANGE: description on how API is changed plus quick fix instructions. Maybe such invasive change is not really necessary? This will help existing users of that API to clearly notice a change and adapt their code. Thanks! :-)

Hi @cederom,

Thanks for your comments.

I believe this change is not invasive. But to avoid any confusion, I have reverted the introduction of the new nxsched_timer.c file and kept using sched_processtimer.c without renaming it.

Could you please help double-check the latest update of this PR?

Below is an explanation of why I consider this change to be non-invasive:

A new internal function, sched_process_tick(), has been added in sched/sched/sched_processtimer.c. All of the original logic previously implemented in sched_process_timer() has been moved into this new internal function, without any functional changes.

sched_process_timer() has now been updated to act as a thin wrapper around sched_process_tick(). Its responsibility is limited to handling the differences in how OS ticks are driven—either directly by a hardware timer or via an hrtimer instance.

The motivation for this refactoring is to clearly separate OS tick timer source selection from OS tick processing. The new internal function sched_process_tick() focuses solely on OS tick processing and preserves exactly the same behavior and logic as before.

As a result:

* When hrtimer is not enabled, sched_process_timer() behaves exactly as it did previously, simply delegating to sched_process_tick().

* When hrtimer is enabled, sched_process_timer() creates and uses an hrtimer instance to drive sched_process_tick().

In summary, no existing behavior is changed. This refactoring only introduces a clearer separation of responsibilities while keeping all original functionality intact. The only addition is the internal sched_process_tick() function, which contains the same OS tick processing logic that was previously implemented directly in sched_process_timer().

@acassis, could you also help to double-check this update?

@cederom just renaming an internal C file shouldn't be considered a breaking change if it didn't change the API. In other hand renaming a header file could be considered a breaking change, because now drivers needs to point to a new file.

@wangchdo wangchdo force-pushed the add_hrtimer_sched_support branch from 19097a7 to 98beb06 Compare December 25, 2025 00:54
@wangchdo wangchdo force-pushed the add_hrtimer_sched_support branch 2 times, most recently from d88ae85 to 26f4bbc Compare December 25, 2025 04:47
@wangchdo wangchdo force-pushed the add_hrtimer_sched_support branch 2 times, most recently from 5ec7cf1 to dc73f9e Compare December 25, 2025 05:14
    This commit add hrtimer support to scheduler
    tick without altering the existing scheduler behavior.

Signed-off-by: Chengdong Wang <wangchengdong@lixiang.com>
…t issue

    When hrtimer is enabled, the tickless scheduler should call
    nxsched_hrtimer_start to start the timer, this is because
    the tick system is support by hrtimer

Signed-off-by: Chengdong Wang <wangchengdong@lixiang.com>
@wangchdo wangchdo force-pushed the add_hrtimer_sched_support branch from dc73f9e to 765aa3b Compare December 25, 2025 05:16
@wangchdo wangchdo requested a review from acassis December 25, 2025 09:30
@wangchdo
Copy link
Contributor Author

Just fixed CI issue, please help to double check @acassis @anchao @cederom

Comment on lines +145 to +152
#ifdef CONFIG_SMP
if (hrtimer->cpus == 0)
{
#endif
hrtimer->state = HRTIMER_STATE_INACTIVE;
#ifdef CONFIG_SMP
}
#endif
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
#ifdef CONFIG_SMP
if (hrtimer->cpus == 0)
{
#endif
hrtimer->state = HRTIMER_STATE_INACTIVE;
#ifdef CONFIG_SMP
}
#endif
#ifdef CONFIG_SMP
if (hrtimer->cpus == 0)
#endif
{
hrtimer->state = HRTIMER_STATE_INACTIVE;
}

{
/* Timer was canceled during callback execution */

#ifdef CONFIG_SMP
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ditto


static void nxsched_process_hrtimer(void)
{
uint64_t now = hrtimer_gettime();
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

move hrtimer_gettime() before hrtimer_process()

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Area: Documentation Improvements or additions to documentation Area: Drivers Drivers issues Area: OS Components OS Components issues Size: L The size of the change in this PR is large Size: M The size of the change in this PR is medium

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants