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.

EDMA transfer measure time should be similar or not regardless the queue used?

EDMA transfer time measure.

Hi,

I'm trying to measure the time for a EDMA transfer of 8K bytes from DDR3 to L2D.

I used the EDMA 1, channel 0 and I performed the same test with the queues 0, 1, 2 and 3. (function CSL_edma3HwChannelSetupQue() ).

I used the same transfer routine modifying only the Transfer controller mentioned above.

I received the same result  (arround 2.1 micro seconds) for the four measures every one with its own queue using the same EDMA..

Is ok to receive approximately the same result for CSL_EDMA3_QUE_0, CSL_EDMA3_QUE_1, CSL_EDMA3_QUE_2 and CSL_EDMA3_QUE_3 using the same EDMA or the result must be very different?

In case that the result needs to be very different, please, could somebody explain me why?

Thans in advance.

Shmuel.

 

  • Shmulik,

    The datasheet and the EDMA User's Guide are where you would find any details on specific difference between the different EDMA instances and between the different Queues. The real distinction is between the different Transfer Controllers, but it is likely from your measurements that all four of the transfer controllers are identical for you to get these identical results.

    You may get faster transfers if you use TPCC0 since it runs at a faster clock rate and is more closely tied to the DDR3 interface. Some of these details are not easy to see in the current documentation. But try TPCC0 and see how the speed varies.

    If you would, please, post your test project as it currently stands. This will help me and others to follow your tests.

    Regards,
    RandyP