S32K31x eMcem_Init problems

[Tiberiu_Dumitrescu]

Hello,

Environement: uC S32k31x, MCAL SW32K3_S32M27x_RTD_R21-11_6.0.0, SPD SW32K3_SPD_1.0.5_D2503

TLDR: After calling eMcem_Init, eMcem_DriverState remains EMCEM_S_UNINIT.

Long story:
I configured the eMcem, call eMcem_Init, and after the driver remains uninitialized, I did some debugging and it seems that it does not enter in the CONFIGURE mode. After writing the CTRLK register with the key and the CTRL to enter in CONFIGURE mode, nothing seems to happen, it remains in NORMAL mode.
I found some info on forum and verified MC_ME PRTN1_COFB0_STAT register whether the ERM clock is running or not before the eMCEM and it is ok.

But I found out on another forum post that in the release notes from this SPD specifies that it is compatible with MCAL 5.0.0 and 4.0.0. Therefore, using RTD 6.0.0 may lead to compatibility issues.

Questions:
1.Are there any specific things that I could further check?
2.Is there a newer SPD which is compatible with the newer MCALs 6 and 7? Or which are the limitations and problems in using the above mentioned pair of MCAL and SPD?
3.Would it be possible to have a teams call/meeting to do a hands-on debugging session or at least to show you the behavior?

[Tiberiu_Dumitrescu]

"Version 1.0.6 for RTD 7.0.0 should be available at the end of the year. According to the 1.0.5 release notes, the SW has been developed and tested with RTD 4.0.0 and 5.0.0. We cannot guarantee compatibility with RTD 6.0.0 or newer"
Currently we have 6.0.0, will SPD 1.0.6 cover also 6.0.0 or we will need to migrate to RTD 7.0.0?

[danielmartynek]

With both RTD versions (specifically with patch 6.0.0 QLP04).

Regards,

Daniel

[danielmartynek]

Hi @Tiberiu_Dumitrescu,

1.

This issue might be caused by active NCF faults. There is a known hardware issue with the NCF[0] lockstep error, which can be triggered when attaching a debugger.
Workaround: Do not reset the S32K device using the debugger. Instead, allow the application to start without the debugger after POR and insert a while(1) loop in the code before eMcem_Init, and then attach the debugger without causing a reset. This approach prevents the NCF[0] fault from being raised.
You can verify this by checking the NCF status registers to see if any faults are present when you observe the failure during eMcem_Init().

2.

Version 1.0.6 for RTD 7.0.0 should be available at the end of the year. According to the 1.0.5 release notes, the SW has been developed and tested with RTD 4.0.0 and 5.0.0. We cannot guarantee compatibility with RTD 6.0.0 or newer,

3.

Our online technical support does not provide assistance through phone calls or live debug sessions. If your company has NXP Field Application Engineers (FAEs) assigned, they should be your primary point of contact for support.

Regards,

Daniel