AnsweredAssumed Answered

How to debug this erroneous situation?

Question asked by Yaniv.Sapir on Mar 2, 2015
Latest reply on Mar 27, 2015 by StuartS

using VDSP++ 5.1.2 on EZ-KIT BF518 + HP-USB-ICE to debug an asynchronous program that uses an ADF7023 transceiver daughtercard, utilizing interrupts and DMA processes, my program gets stuck every other run. When halting the run, I see the following message in the output window:

 

Core fault detected before halt at address: 0x00000672.  Attempting to reset processor.

 

and the Call Stack window shows:

 

==> <___wab8 + 0xcec007ee>

    <___wab8 + 0xcec00548>

    <___wab8 + 0xcec0016e>

    <___wab8 + 0xceaeef3a>

    <___wab8 + 0xceaeef3a>

 

How should one approach this kind of error reporting, when the call stack seems to be meaningless (__wab8 is a symbol in the LDF, Address 0x672 seems to be in a place with a function name .P37L4 which I can't find in the sources nor the LDF), and there is no active stack frame so inspecting locals is impossible?

 

----

Update: Re-reading of the MAP file, the .P37L4 seems to be a label in the function named .__primIO starting at 0x64c and ending at 0x676. Not sure where this function is defined.

Outcomes