adafruit-mirror/zephyr
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 9

kernel/timer: fix period argument clamp handling

Browse source 
Commit 3e729b2b1c ("kernel/timer: Correctly clamp period argument")
increased the lower limit to 1 so that it wouldn't conflict with a
K_NO_WAIT. But in doing so it enforced a minimum period of 2 ticks.
And the subtraction must obviously be avoided if the period is zero, etc.

Instead of doing this masquerade in k_timer_start(), let's move the
subtraction and clamping in z_timer_expiration_handler() right before
registering a new timeout. It makes the code cleaner, and then it is
possible to have single-tick periods again.

Whith this, timer_jitter_drift in tests/kernel/timer/timer_behavior does
pass with any CONFIG_SYS_CLOCK_TICKS_PER_SEC value, even when the tick
period is equal or larger than the specified timer period for the test
which failed the test before.

Signed-off-by: Nicolas Pitre <npitre@baylibre.com>
This commit is contained in:
Nicolas Pitre 2023-02-13 00:43:47 -05:00 committed by Fabio Baltieri
parent f21051b728
commit 6a51a10dec
1 changed files with 6 additions and 8 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

14
kernel/timer.c
View file

@ -52,6 +52,9 @@ void z_timer_expiration_handler(struct _timeout *t)
!K_TIMEOUT_EQ(timer->period, K_FOREVER)) { !K_TIMEOUT_EQ(timer->period, K_FOREVER)) {
k_timeout_t next = timer->period; k_timeout_t next = timer->period;
/* see note about z_add_timeout() in z_impl_k_timer_start() */
next.ticks = MAX(next.ticks - 1, 0);
#ifdef CONFIG_TIMEOUT_64BIT #ifdef CONFIG_TIMEOUT_64BIT
/* Exploit the fact that uptime during a kernel /* Exploit the fact that uptime during a kernel
* timeout handler reflects the time of the scheduled * timeout handler reflects the time of the scheduled
@ -64,8 +67,7 @@ void z_timer_expiration_handler(struct _timeout *t)
* beginning of a tick, so need to defeat the "round * beginning of a tick, so need to defeat the "round
* down" behavior on timeout addition). * down" behavior on timeout addition).
*/ */
next = K_TIMEOUT_ABS_TICKS(k_uptime_ticks() + 1 next = K_TIMEOUT_ABS_TICKS(k_uptime_ticks() + 1 + next.ticks);
+ timer->period.ticks);
#endif #endif
z_add_timeout(&timer->timeout, z_timer_expiration_handler, z_add_timeout(&timer->timeout, z_timer_expiration_handler,
next); next);
@ -139,8 +141,8 @@ void z_impl_k_timer_start(struct k_timer *timer, k_timeout_t duration,
* to round up to the next tick (by convention it waits for * to round up to the next tick (by convention it waits for
* "at least as long as the specified timeout"), but the * "at least as long as the specified timeout"), but the
* period interval is always guaranteed to be reset from * period interval is always guaranteed to be reset from
* within the timer ISR, so no round up is desired. Subtract * within the timer ISR, so no round up is desired and 1 is
* one. * subtracted in there.
* *
* Note that the duration (!) value gets the same treatment * Note that the duration (!) value gets the same treatment
* for backwards compatibility. This is unfortunate * for backwards compatibility. This is unfortunate
@ -148,10 +150,6 @@ void z_impl_k_timer_start(struct k_timer *timer, k_timeout_t duration,
* argument the same way k_sleep() does), but historical. The * argument the same way k_sleep() does), but historical. The
* timer_api test relies on this behavior. * timer_api test relies on this behavior.
*/ */
if (!K_TIMEOUT_EQ(period, K_FOREVER) && period.ticks != 0 &&
Z_TICK_ABS(period.ticks) < 0) {
period.ticks = MAX(period.ticks - 1, 1);
}
if (Z_TICK_ABS(duration.ticks) < 0) { if (Z_TICK_ABS(duration.ticks) < 0) {
duration.ticks = MAX(duration.ticks - 1, 0); duration.ticks = MAX(duration.ticks - 1, 0);
} }