• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 30 subscribers
  • Views 267 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Thumb state

Mark Blay
Intellectual 370 points
Other Parts Discussed in Thread: TMS570LS20206

Hi, I am using TMS570LS20206 and reviewing the errata for Silicon Revision A (spnz176d).

For the following errata it is stated that the problem only exists when the processor is in thumb state

CORTEX-R4#25

CORTEX-R4#26

When we compile we specify --code_state=32. Does this mean that the processor will not be in thumb state thus the errata will not impact the application, or am I not correctly understanding this compiler option.

Thanks,

Mark.

  • 0
    TI__Guru**** 197126 points

    Hi Mark,

    The compiler generates 32-bit code by default or using --code_state=32. I suggest to use the recommended workaround in the errata for #25, and use 32-bit code (--code_state=32) for #32.

    Regards,

    QJ

  • 0 in reply to QJ Wang
    Intellectual 370 points

    Thanks QJ.

    I think I should have explained the situation clearer in my original post.

    For the project I am working on the errate have not really been considered early enough, so the situation is that I now have verified software that I don't want to change in any way unless I have to.

    I realise that wherever there is a workaround that is the recommended solution, but I am trying where possible to demonstrate that for each erratum that there is no impact on my software, either due to the use of a compiler switch or the way the processor is being used. The situation is the same for the other post you responded to in relation to CORTEX-R4#46 and CORTEX-R4#55

    I had notices thar for #25 and #26 the erratum was only applicable if the processor was in thumb state so I was hoping that usei of the --code_state=32 compiler option would ensure this.

    Sorry for any confusion,

    Mark.