Hello Scott,
I understand you are referring to that code before entering Stop Mode 2
while( LL_HSEM_1StepLock( HSEM, CFG_HW_RCC_SEMID ) );
if ( ! LL_HSEM_1StepLock( HSEM, CFG_HW_ENTRY_STOP_MODE_SEMID ) )
{
if( LL_PWR_IsActiveFlag_C2DS( ) )
{
/* Release ENTRY_STOP_MODE semaphore */
LL_HSEM_ReleaseLock( HSEM, CFG_HW_ENTRY_STOP_MODE_SEMID, 0 );
/**
* The switch on HSI before entering Stop Mode is required
*/
Switch_On_HSI( );
}
}
else
{
/**
* The switch on HSI before entering Stop Mode is required
*/
Switch_On_HSI( );
}
The purpose of this code is not to decide if we are allowed to enter Stop Mode or not. Whatever we are doing there, the decision has already been taken to enter Stop Mode.
The purpose of this code is to understand if we can switch on HSI or not before entering Stop Mode.
There are two points to check :
1/ We are checking C2DS because we dont want to switch on HSI if CPU2 is running as it requires a 32Mhz Clock.
2/ As both CPU are almost running the same algorithm, if each of them checks at the same time the CxDS bit, none of them will switch on HSI whereas they are both on the way to enter Stop Mode. The "Strange" use of CFG_HW_ENTRY_STOP_MODE_SEMID semaphore is there to handle this corner cases.
Now the only point is the reason why to switch on HSI before entering Stop Mode 2. This was mandatory on Cut2.0 and the bug is fixed on Cut2.1. Unless you get some board at mass market launch during some workshop, you should have a Cut2.1 board.
The workaround has been kept for Cut2.1 because it does not hurt and all our validation have been executed with the workaround. Therefor, it is recommended to keep it for safety.
Regards.
