I see possible issue in using critical sections in function vPortSuppressTicksAndSleep().
~~~
/* Stop the SysTick momentarily. The time the SysTick is stopped for
is accounted for as best it can be, but using the tickless mode will
inevitably result in some tiny drift of the time maintained by the
kernel with respect to calendar time. */
portNVIC_SYSTICK_CTRL_REG &= ~portNVIC_SYSTICK_ENABLE_BIT;
/* Calculate the reload value required to wait xExpectedIdleTime
tick periods. -1 is used because this code will execute part way
through one of the tick periods. */
ulReloadValue = portNVIC_SYSTICK_CURRENT_VALUE_REG + ( ulTimerCountsForOneTick * ( xExpectedIdleTime - 1UL ) );
if( ulReloadValue > ulStoppedTimerCompensation )
{
ulReloadValue -= ulStoppedTimerCompensation;
}
/* Enter a critical section but don't use the taskENTER_CRITICAL()
method as that will mask interrupts that should exit sleep mode. */
__asm volatile( "cpsid i" );
__asm volatile( "dsb" );
__asm volatile( "isb" );
~~~
If interrupt occur after SYSTICK stop and before critical section it will provide inaccuracy in uReloadValue calculation.
Similar issue is present in wakeup case:
~~~
if( xModifiableIdleTime > 0 )
{
__asm volatile( “dsb” );
__asm volatile( “wfi” );
__asm volatile( “isb” );
}
configPOST
SLEEPPROCESSING( xExpectedIdleTime );
/* Stop SysTick. Again, the time the SysTick is stopped for is
accounted for as best it can be, but using the tickless mode will
inevitably result in some tiny drift of the time maintained by the
kernel with respect to calendar time. */
ulSysTickCTRL = portNVIC_SYSTICK_CTRL_REG;
portNVIC_SYSTICK_CTRL_REG = ( ulSysTickCTRL & ~portNVIC_SYSTICK_ENABLE_BIT );
/* Re-enable interrupts - see comments above the cpsid instruction()
above. */
__asm volatile( "cpsie i" );
~~~
The interrupt is enabled and systick is not running if IRQ occur before systick start then time spend in IRQ will cause inaccuracy.
Improvement proposition:
The critical sections should be moved to protect regions when systick reload value is calculated. Do you agree?
Additional possible improvement:
There is a define portMISSED
COUNTSFACTOR used to tune tickles time drift. It would be better to use Program Counter to calculate ‘real’ time when systick was off and use it to compensate time drift
Best regards
Lukasz Antczak