When enter VLPS, system reset

[huisheng_huang]

Hi,

I am facing a tough issue, and have been working on it for a week. But issue not solved yet. so come here for some help.

I am using s32k148, and MCAL (S32K14X_MCAL4_3_RTM_HF2_1_0_1). The issue is when i call function Mcu_SetMode() to enter VLPS, system will reset, with the reset reason SACKERR. 

I have call Os_DisableInterruptSource()(Autosar interface) to disable all the interrupt source but except the PTE5 interrupt which is connected to Rx pin of CAN0.

Reset is occured after execute EXCUTE_WAIT.

asdf.png

[huisheng_huang]

Thank you, Daniel.

Yes, you are great. When I disable the CAN module, no reset  happens. 

But after that, I have another problem.

MCU can't be waked up by a external interrupt, which is PORTE, pin 5.

Before enter VLPS, I enable the interrupt.

asdf.png

When in debugger mode, I checked the PCR register, as below, the ISF is set.

asdfdsf6-11.png

My question is, why didn't enter the ISR and wake up the MCU?

Thank you in advance.

[danielmartynek]

Have you tested the interrupt in the RUN mode?

Does it work?

Do you use Sleep-Now or Sleep-On-Exit mode?

Because in the Sleep-on-exit mode, the MCU goes back to VLPS immediately once the ISR is complete.

So you might not notice the ISR was executed.

BR, Daniel

[huisheng_huang]

Yes, it works in RUN mode.

We don't use the sleep on exit mode. as below:

Snipsdafsdf38.png

A strange phenomenon is when in debugger mode, after the external interrupt occurs, MCU didn't enter ISR. But if i pause the debug session , MCU enters the ISR.

sdfasdfte_2020-11-02_18-32-27.png

[danielmartynek]

Hi,

It looks like an issue of the tool not the MCU.

Please try running the MCU stand-alone without the debugger.

I mean, disconnect the debugger, power cycle the MCU and toggle a GPIO (for example) from the ISR.

Also, you can output BUS_CLK on a CLKOUT pin so that you can see when the MCU is VLPS and when it is running.

BR, Daniel

[huisheng_huang]

Hi Daniel,

Maybe I know the reason.

After below modification, all is normal.

huisheng_huang_0-1604379924033.png

Before the code EXECUTE_WAIT, SW call DisableAllInterrupts(). So I think we should use the corresponding call function EnableAllInterrupt(). If so, then we should modify the MCAL code. I don't know whether ok or not?

[danielmartynek]

Hi,

In the newest MCAL4.2 1.0.4 release, the EXECUTE_WAIT macro contains only the WFI instruction.

Both SuspendAllInterrupts() and ResumeAllInterrupts() have been removed and you can removed it as well.

But anyway, the interrupts must not be disabled/masked (PRIMASK) at the time the WFI instruction is executed, you need to enable them.

Regards,

Daniel

[huisheng_huang]

Hi Daniel,

I have tried your suggestion.

I disconnect the debugger, power off the board and power on again. when MCU enters to VLPS, trigger an  interrupt, but no changes on the GPIO. Besides, the current I measured doesn't change.

I have output the BUS_CLK to the CLKOUT pin. At run mode, the frequency is 2.5MHZ. When enters to VLPS, the CLKOUT is a high level.

So, I think the MCU is not able to be waked up.

[danielmartynek]

Hi,

Can you please disable the FlexCAN module before the transition to VLPS?

FlexCAN is not supported in VLPS and should be disabled.

Thank you,

BR, Daniel