Corrupted metadata and hardfaults

[Catosh]

Hi all; 

I noticed a fancy behavior using S32ds and segger jlink. 

I have a fully working project, sometimes I need to make some edits and reload the project.  Out of the blue, I get an hard fault on the MCU where it should generally execute the systick handler. 

Exploring the bin, I can see that the address of the systick ISR is the one I expect, where I expect it. 

Flashing the MCU using the binary and segger jlink everything works fine.

Flashing the MCU inside Design studio using segger in debug mode it gets the hard fault. 

Many possible fault paths here, but what I found out is that erasing Eclipse metadata and re-importing the project WITHOUT any change fixes the issue. 

Any hint on this point? IT took almost a work week to figure out what was wrong (omitting some verbose detail here) and I think it would be beneficial to have some insight on this point for other developers facing similar issues. 

K.R..

[jiri_kral · Accepted solution]

Sumarry: 

The issue is caused by breakpoints on un-existing code line - typically when you turn optimization and left breakpoints on place which was removed by optimization. Solution is just turn off all breakpoints in this situation. 

Jiri

[Catosh]

Hi Jiri, 

I stored the 7zip of the metadat but I don't think I can share it with everyone (but pelase, be aware that I have really NO IDEA of what the content of a metadata folder is).

BTW, I figured out also that it's enough to erase the current debug configuration and to create a brand new one. 

[jiri_kral]

Hi, 

you can send it to my address - jiri.kral@nxp.com . I'll try figure out what's going on. 

Jiri 

[jiri_kral]

Hi, 

thanks for sharing your issue. In previous S32DS versions sometimes creating new workspace solved some issues, but this should be fixed in current release. 

It is possible get the corrupted meta data and share it? 

Jiri