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.

How to make example programs run in simulator mode?

I made an earlier post with the same request and somehow it was left un-answered from TI's side.  Is there a list of compile flag / cmd file correction to make these applications run on "simulator"? Is there a list of "no-no" for simulated apps?

Thanks for any input,

Sayee

  • Mokan Kanna said:

    I made an earlier post with the same request and somehow it was left un-answered from TI's side.  Is there a list of compile flag / cmd file correction to make these applications run on "simulator"? Is there a list of "no-no" for simulated apps?

    Thanks for any input,

    Sayee

    Sayee,

    The CCS 4.x simulator for C2000 is a CPU only sim.  Peripherals are not modeled.  With the low cost control STICK hardware we found customers would rather use real hardware over s a simulation environment.  These posts may also help:

    http://processors.wiki.ti.com/index.php/Category:Simulation

    http://e2e.ti.com/support/development_tools/code_composer_studio/f/81/p/99042/346568.aspx#346568

    http://e2e.ti.com/support/development_tools/code_composer_studio/f/81/t/41494.aspx

    Regards,

    Lori

     

  • Lori,

    I can not quite agree with you that low cost boards will be better alternative to simulators. In simulator mode, the loading instantaneous. NO RISK OF FLASH LIFE-CYCLE. NO FEAR OF CSM LOCK-OUTs. Several repeats of fine-tuning/evaluation possible with code. It has so many such benefits (just too good to say no!)

    It is OK if it is a CPU only simulator without peripherals. But still all the example programs should be able to load! Why is it not happening?

    But I have another question: In the simulation mode,  I know, the profile clock is not cycle accurate and does not take into account the instruction pipeline effects.But then, there should be some correlation between what is displayed by profile clock and what is actually realized. Correct? Does it assume the code to run from flash and not from ram? Does it account for flash wait cycles too?

    I did some bench-marking and the profile clock  numbers are typically 3 to 5 times than what you would get when you manually sum-up instruction cycles. Can you give some details on these? In fact, I wanted to open a separate thread to discuss this. But from whatever I read from the forum, I infer that people are being weaned away from simulator from TI's side. May be a decision from the support cost perspective! That is why I almost left it!!

    Sayee

  • Mokan Kanna said:

    I can not quite agree with you that low cost boards will be better alternative to simulators. In simulator mode, the loading instantaneous. NO RISK OF FLASH LIFE-CYCLE. NO FEAR OF CSM LOCK-OUTs. Several repeats of fine-tuning/evaluation possible with code. It has so many such benefits (just too good to say no!)

    Just my two cents here...There should not be a fear of flash issues when running from low cost boards, until you're project is nearing completion.  My recommendation is that you use RAM only until you have to use Flash, then later on convert your RAM code to Flash code.  This alleviates all the possible problems with flash on the parts (well the only real issue is CSM Lock outs and I have noticed they are rarer today then when I first used the TMX F2812s in 2004 - flash life cycles is 100k+ so you shouldnt need to worry about those).  You are able to fine tune your code a lot when you are just using RAM which makes the evaluation boards very appealing.

     

    Tim