[Dspforum] Lab 6 Questions
Alan Ding
alan33d at gmail.com
Mon Oct 19 14:14:21 MDT 2009
Thanks. I knew about the macros to change the csl register values but I was
curious where I could find a definition of the registers or what they are
defined to be (I wanted to get more information about them beyond the scope
of what is shown on the FFT EDMA lecture notes). SPRU401 really helps. :)
~Alan
On Sun, Oct 18, 2009 at 8:43 PM, Joseph Haber <josephhaber at gmail.com> wrote:
> Alan,
>
> Check out slide 33 in the lecture notes. What you will be using is the
> "Chip Support Library," or the CSL. Remember in the second lecture when I
> mentioned that in the "old days" you would get a header file that defined
> all the registers, but now you operate through the CSL. This defines a whole
> set of macros to set registers, fields, etc.
>
> For example, there is the macro EDMA_RSET that allows you to directly
> access the QOPT, QSRC, etc. registers.
>
> EDMA_RSET(QOPT,0x21300001);//32bitshighpriorityTCC0
>
> See Section 7 EDMA Module in Spru401 (which is available in the help menu
> from CCS 3.1).
>
> Joseph
>
>
> On Sun, Oct 18, 2009 at 7:48 PM, Alan Ding <alan33d at gmail.com> wrote:
>
>> Thanks Doug.
>>
>> I'm also having a tough time locating register definitions from the
>> lecture notes:
>>
>> CIPR
>> QOPT...etc.
>>
>> I found the macros in csl_edmahal.h which was not part of the example
>> project.
>>
>> ~Alan
>>
>>
>> On Sun, Oct 18, 2009 at 7:39 PM, Doug Wenstrand <doug at echelonembedded.com
>> > wrote:
>>
>>> Yea, I forgot to time it and fill in the requirement. Your presumption
>>> of course is correct -- We want you to use QDMA to do it efficiently, but
>>> you need not worry about overlapping the transfers with computations.
>>> So -- the following pseudocode is what would meet the timing requirement:
>>>
>>> loop:
>>> dma data in
>>> wait for it
>>> fft
>>> dma data out
>>>
>>>
>>> On Sun, Oct 18, 2009 at 6:46 PM, Alan Ding <alan33d at gmail.com> wrote:
>>>
>>>> For the requirements on the back of the lab, it states that our STFT
>>>> "must take less than xx ms". What exactly is xx? ~1.7ms*100 or ~48ms*100?
>>>> I take it you guys want us to get roughly the first timing requirement using
>>>> QDMA.
>>>>
>>>> When we first capture the data in ext ram from CCS. The instructions
>>>> wants us to create a plot of this data in CCS. How do you do that? Should
>>>> we use the Graph function under the view tab? What options should we use
>>>> for it?
>>>>
>>>> Thanks,
>>>>
>>>> --
>>>> -------------------------------------------------------
>>>> Alan Ding
>>>> Master of Sci, Electrical Engineering
>>>> Johns Hopkins University
>>>>
>>>> Email: alan33d at gmail.com
>>>> Contact #: 4844593290
>>>>
>>>> _______________________________________________
>>>> Dspforum mailing list
>>>> Dspforum at echelonembedded.com
>>>> http://echelonembedded.com/mailman/listinfo/dspforum_echelonembedded.com
>>>>
>>>>
>>>
>>> _______________________________________________
>>> Dspforum mailing list
>>> Dspforum at echelonembedded.com
>>> http://echelonembedded.com/mailman/listinfo/dspforum_echelonembedded.com
>>>
>>>
>>
>>
>> --
>> -------------------------------------------------------
>> Alan Ding
>> Master of Sci, Electrical Engineering
>> Johns Hopkins University
>>
>> Email: alan33d at gmail.com
>> Contact #: 4844593290
>>
>> _______________________________________________
>> Dspforum mailing list
>> Dspforum at echelonembedded.com
>> http://echelonembedded.com/mailman/listinfo/dspforum_echelonembedded.com
>>
>>
>
> _______________________________________________
> Dspforum mailing list
> Dspforum at echelonembedded.com
> http://echelonembedded.com/mailman/listinfo/dspforum_echelonembedded.com
>
>
--
-------------------------------------------------------
Alan Ding
Master of Sci, Electrical Engineering
Johns Hopkins University
Email: alan33d at gmail.com
Contact #: 4844593290
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://echelonembedded.com/pipermail/dspforum_echelonembedded.com/attachments/20091019/438e7986/attachment.html>
More information about the Dspforum
mailing list