This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Compiler/TMS320F28377D: Issue with TI C2000 compiler: segfaulting

Genius 12760 points
Part Number: TMS320F28377D
Other Parts Discussed in Thread: CONTROLSUITE

Tool/software: TI C/C++ Compiler

Customer has an issue with the TI C2000 16.6.0.STS compiler, where the optimiser is segfaulting in a particular function, but only at optimisation level 3 or 4. Attempting to produce a minimal test case does not reproduce the problem. I have the pre-processed source and the compiler console output but cannot post it here.

Who can support offline?

Regards, Bernd

  • BO said:
    I have the pre-processed source and the compiler console output but cannot post it here.

    Since you are a TI employee, you can find my e-mail address.  Please send it to me that way.

    Thanks and regards,

    -George

  • Hi George,

    offline mail sent.

    Regards, Bernd
  • Thank you for submitting a test case.  I can reproduce the problem.

    The problem is fixed in later versions of the compiler.  I recommend you upgrade to the latest compiler version from the 16.9.x.LTS stream.  At this time, that is version 16.9.2.LTS.  To learn more about the differences among these compiler versions, please see the wiki article Compiler Version Numbers and What They Mean.  To see how to upgrade the compiler, please see Compiler Releases

    After making that compiler upgrade, you will see another problem.  This is due to an error in the controlSUITE header files.  The problem is discussed in this thread.  The fix is to upgrade controlSUITE, as described in this thread.

    Thanks and regards,

    -George

  • George,

    customer cannot fix the problem by upgrading the compiler and controlSUITE header files as suggested. The optimizer still segfaults. Pre-processed code and compiler output I will send offline again. There are minor differences to the application but the same problem presents around the area where I expect RPT || MACF32 instructions to be emitted.

    Regards,
    Bernd
  • Thank you for submitting a second test case.  I can reproduce the problem.  I filed CODEGEN-2182 in the SDOWP system to have this investigated.  You are welcome to follow it with the SDOWP link below in my signature.

    Thanks and regards,

    -George