Hello,
I'm experiencing an issue when I try to use ETAS INCA to reflash an S32K148 using a custom A2L and Prof file. Briefly, the issue I'm having involves the calibration data. Reflashing with INCA (i.e. flashing from the reference or working page using Flash programming) always returns the code and calibration data on the ECU back to the original build. In other words, any changes I make to calibration variables in INCA are “forgotten” when flashing, and the ECU (and the working page and reference page checksums) returns to its original state. This happens when flashing from both the reference and working pages.
Let me illustrate the issue below. When I start INCA (before any reflash), the checksums all match:
Then I make a change to a calibration variable (it doesn't matter which variable, it's the same behavior regardless) and update the checksums:
Then I copy the working page to the reference page. This step isn't necessary, but it creates a situation where all of the checksums are different from the original, so it's theoretically not possible to return to the original state.
Lastly, I select Flash programming to flash the ECU with either the working or reference page (it doesn't matter which is used, the final result is identical). After the reflash, INCA says flash programming finished successfully, and the checksums are updated again. However, the ECU checksums revert back to what was originally on the ECU, and the INCA checksums remain unchanged. In other words, the calibration change is not reflected in the data that was used to reflash the ECU.
Can you tell what the issue might be here? Is this due to a problem in the A2L file, the prof file, INCA configuration, something else?
Another note: if I change the source memory area for the calibration data from the reference page start address to the working page start address (and leave the destination memory area for the calibration data unchanged), the checksum changes correctly during a Flash programming and it doesn't return back to the original state. In other words, the issue is fixed (however, this has to do with some circumstantial data being all 0xFF's, so it's not a sustainable solution). I thought this might indicate that the reference page being declared RESERVED in the A2L file was the problem, but changing that RESERVED keyword to DATA or OFFLINE_DATA didn't fix anything.
Any help you can provide with respect to this issue is much appreciated.
Thank you very much.
