The connection between AUTOSAR Network Management sleep/wake-up

[ydygcvjhbjkjnkn]

Hello, I am currently debugging EcuM and Network Management. I have consulted some specification manuals, but I still have some confusion and would like to request guidance.

1. Does the sleep/wake-up of Network Management necessarily require coordination with EcuM (i.e., EcuM validates the wake‑up cause)?

2. During the Network Management sleep/wake‑up process, does the ECU need to go to sleep (i.e., EcuM sleep)?

3. If the answer to question 1 is yes, then EcuM obtains wake‑up events through the interface EcuM_SetWakeupEvent. However, the AUTOSAR EcuM specification SWS_EcuM_04318 states: “Wakeup events that are not associated to the selected sleep mode shall be ignored.” How should I understand this?

4. The AUTOSAR EcuM specification also contains two entries regarding the interface EcuM_ValidateWakeupEvent – SWS_EcuM_02790 and SWS_EcuM_02791. It says that wake‑up sources associated with a ComM channel shall be effective in any phase. Does this mean that in any phase, a wake‑up event can be pended (i.e., successfully pass through EcuM_SetWakeupEvent) and then be validated?

[ydygcvjhbjkjnkn]

Hello, thank you for your answer. I still have some questions regarding your response. I apologize for bothering you repeatedly, and I would like to thank you in advance.

I understand that the sleep/wake-up of Network Management also seems to require EcuM to validate the wake-up source, because in the passive mode of Network Management, EcuM validation of the wake-up source is required. After the wake-up source associated with the ComM channel is successfully validated, EcuM calls the ComM interface, and then the network can wake up. In your first answer, you said that EcuM's wake-up validation is only necessary when the ECU actually enters the EcuM sleep state. So, I interpret this as meaning that after Network Management sleep/wake-up, the ECU also needs to enter the EcuM sleep state, and then go through the EcuM wake-up validation process. However, this seems to conflict somewhat with your second answer. I'm not sure whether my understanding is correct.

In your third and fourth answers, SWS_EcuM_04318 states that EcuM only accepts wake-up events configured for the currently selected sleep mode. In answer 4, you said that ComM-related wake-up sources are a special case. So, can ComM-related wake-up sources simply ignore the constraints of SWS_EcuM_04318, or must I, during the design phase, configure all ComM-related wake-up sources as wake-up events for the sleep mode?

Does the wake-up source need to be enabled first? EcuM calls EcuM_EnableWakeupSource for enabling during the sleep phase. Only after enabling can it be pended (successfully pass the EcuM_SetWakeupEvent interface filtering)? If so, regarding the statement in answer 4 that "ComM-channel-associated sources (e.g., CAN wakeup frame) can be pended via EcuM_SetWakeupEvent," does this pend operation only occur during the EcuM sleep phase, and then validation occurs in any phase of EcuM? Or is the enabling step done by other drivers themselves, or are there other scenarios?

I look forward to your answers.

[PetrS]

Hi,

1. Does NM sleep/wake-up require EcuM coordination?
Not necessarily. NM/ComM manages network communication, while EcuM manages ECU power states. EcuM wakeup validation is mainly relevant when the ECU actually enters an EcuM sleep state. 

2. Does NM sleep mean the ECU must enter EcuM sleep?
No. The network can enter Bus-Sleep while the ECU remains in RUN mode and continues executing local functions.

3. How to understand SWS_EcuM_04318?
It means EcuM only accepts wakeup events that are configured for the currently selected sleep mode. Wakeup events from sources not enabled for that sleep mode are ignored. It prevents spurious hardware events from waking the ECU unintentionally.

4. Can ComM-related wakeup sources be validated in any EcuM phase?
Yes. This is a deliberate special case. Generic wakeup sources are only accepted during SLEEP, but ComM-channel-associated sources (e.g., CAN wakeup frame) can be pended via EcuM_SetWakeupEvent and validated via EcuM_ValidateWakeupEvent in any EcuM phase. This allows a wakeup event to abort an ongoing Go-Sleep sequence — which is the key mechanism linking NM wakeup back to EcuM.

In short: NM sleep ≠ ECU sleep, and ComM-linked wakeup sources receive special handling in EcuM, allowing validation outside the normal wakeup sequence. 

BR, Petr