Loading ...
Sorry, an error occurred while loading the content.

Het ECMP instruction help

Expand Messages
  • Rakesh Kumar
    I was trying to analyze ECMP instruction. Attached is the pdf i am looking to.. In page 49, hr_data is 0x10.But it says HR compare value is 2. Why is that..?
    Message 1 of 3 , Mar 15, 2010
    • 1 Attachment
    • 831 KB
    I was trying to analyze ECMP instruction.
    Attached is the pdf i am looking to..

    In page 49, hr_data is 0x10.But it says HR compare value is 2. Why is that..?

    Can anyone please clarify me?

  • JohnStosh
    Hey Rakesh; Thanks for attaching the PDF -- that made it easy to reply.  It has been a while since I ve used the HET. The HET is pretty amazing, eh?  The
    Message 2 of 3 , Mar 16, 2010
    • 0 Attachment
      Hey Rakesh;

      Thanks for attaching the PDF -- that made it easy to reply.  It has been a while since I've used the HET.

      The HET is pretty amazing, eh?  The manual has many, many details and still we have questions.  Our questions come from background assumptions that the writer has -- assumptions that we don't necessarily share.

      So you've put your finger on one of them -- "How does the counting and comparing occur when the HR isn't 32?"   You can get insights into the answer by looking at the title of Figure 21 on page 51.  The title is "High/Low Resolution Modes for ECMP and PWCNT" -- meaning that the writer was thinking about ECMP when he was writing these paragraphs.  Then on the top of page 51 (page 38 for those following along with a printed manual) you see a Note that says "The HR capture value written into RAM is shifted appropriately depending on the loop resolution prescale divide rate (Lr)."  This note appears to apply to the PCNT instruction, but you must realize that it applies to a broader scope -- and most likely to the ECMP instruction.  I believe that it actually applies to all HR counting and comparing.

      To understand the HET manual and succeed, you must read much and think of how everything connects together.  It doesn't hurt to experiment a little, but you don't want to make a career out of guessing.

      As an aside, make sure you don't use too many time slots.  This is very important.  You must have a firm understanding of LR and HR.   When you understand this, the memory for instructions suddenly becomes much smaller than it appeared to be.

      Also, I remember the oscilloscope being very helpful.

      God Bless,
      -Stosh

      --- On Mon, 3/15/10, Rakesh Kumar <rakesh.kumar58@...> wrote:

      From: Rakesh Kumar <rakesh.kumar58@...>
      Subject: [TMS470_ARM] Het ECMP instruction help [1 Attachment]
      To: "TMS470" <TMS470_ARM@yahoogroups.com>
      Date: Monday, March 15, 2010, 10:27 AM

       

      I was trying to analyze ECMP instruction.
      Attached is the pdf i am looking to..

      In page 49, hr_data is 0x10.But it says HR compare value is 2. Why is that..?

      Can anyone please clarify me?


    • Rakesh Kumar
      Hello John, Thanks for the reply... ... From: JohnStosh Subject: Re: [TMS470_ARM] Het ECMP instruction help To:
      Message 3 of 3 , Mar 18, 2010
      • 0 Attachment

        Hello John,

        Thanks for the reply...

        --- On Tue, 3/16/10, JohnStosh <johnstosh@...> wrote:

        From: JohnStosh <johnstosh@...>
        Subject: Re: [TMS470_ARM] Het ECMP instruction help
        To: TMS470_ARM@yahoogroups.com
        Date: Tuesday, March 16, 2010, 7:58 PM

         

        Hey Rakesh;

        Thanks for attaching the PDF -- that made it easy to reply.  It has been a while since I've used the HET.

        The HET is pretty amazing, eh?  The manual has many, many details and still we have questions.  Our questions come from background assumptions that the writer has -- assumptions that we don't necessarily share.

        So you've put your finger on one of them -- "How does the counting and comparing occur when the HR isn't 32?"   You can get insights into the answer by looking at the title of Figure 21 on page 51.  The title is "High/Low Resolution Modes for ECMP and PWCNT" -- meaning that the writer was thinking about ECMP when he was writing these paragraphs.  Then on the top of page 51 (page 38 for those following along with a printed manual) you see a Note that says "The HR capture value written into RAM is shifted appropriately depending on the loop resolution prescale divide rate (Lr)."  This note appears to apply to the PCNT instruction, but you must realize that it applies to a broader scope -- and most likely to the ECMP instruction.  I believe that it actually applies to all HR counting and comparing.

        To understand the HET manual and succeed, you must read much and think of how everything connects together.  It doesn't hurt to experiment a little, but you don't want to make a career out of guessing.

        As an aside, make sure you don't use too many time slots.  This is very important.  You must have a firm understanding of LR and HR.   When you understand this, the memory for instructions suddenly becomes much smaller than it appeared to be.

        Also, I remember the oscilloscope being very helpful.

        God Bless,
        -Stosh

        --- On Mon, 3/15/10, Rakesh Kumar <rakesh.kumar58@ yahoo.com> wrote:

        From: Rakesh Kumar <rakesh.kumar58@ yahoo.com>
        Subject: [TMS470_ARM] Het ECMP instruction help [1 Attachment]
        To: "TMS470" <TMS470_ARM@yahoogro ups.com>
        Date: Monday, March 15, 2010, 10:27 AM

         

        I was trying to analyze ECMP instruction.
        Attached is the pdf i am looking to..

        In page 49, hr_data is 0x10.But it says HR compare value is 2. Why is that..?

        Can anyone please clarify me?



      Your message has been successfully submitted and would be delivered to recipients shortly.