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.

TMS570LC4357 STC Global Control Register 0 (STCGCR0)

I have noticed the bit definition of the STC Global Control Register 0 (STCGCR0) contained in Technical Reference Manual (Literature Number: SPNU563) defines bits 1-0 as RS_CNT.  Shouldn't this be just bit 0.

We are having difficulty in getting the self test to pass. Below is a screen shot of the results we are seeing with bit 0 set to 1:

The test still fails when we set both bits 0 and 1 but the CORE2_ICOUNT and CORE1_1COUNT are 0002.

Regards,

Tom

  • The internal spec says
    00 ]
    01 ] as described in SPNU563
    1x ] restarts from a user specified interval.

    So I don't know why 1x isn't documented [it may be that it doesn't work well on this silicon or may be an oversight] but I believe we can say that the field is a 2-bit field.

    The Self Test of the CPU will break your debug connection and clear any breakpoints you had set in the process so it's not the simplest thing to 'debug' - how are you doing the debug?
  • Your response explains why setting bit 1 causes the CORE2_ICOUNT and CORE1_1COUNT to increment.
    Is this going to be corrected in the next version of the reference manual?
  • Tom,

    You're a little ahead of me but yes if it is an error I'll file a literature bug report on this so it's tracked & corrected.
  • Tom,

    So discussing this more with colleagues, the value 1x isn't exactly as described.
    Yes it lets you start at a specific interval but the interval isn't arbitrary, there are only a few choices
    and they are the starting intervals of the segments 0 and 1.

    Segment 0 is the CPU and segment 1 is the uSCU.

    But there is an errata that prevents you from just running the STC on uSCU; it says you always have to run the STC on intervals 0 and 1 together.

    So this makes the 1x option for RS_CNT not useful .. we think this is why someone undocumented it.

    In any case I just filed several literature bugs on this topic because of the inconsistencies and I'm not sure which will be worked & which will be rejected. It depends on the philosophy of the chapter owner and whether the whole segment concept should just be removed because it's not working [due to errata] or left in and clearly explained. Sorry for the confusion.