• State
  • Locked Locked
  • Replies 4 replies
  • Subscribers 97 subscribers
  • Views 900 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

spi slave dma probem

Lusheng zhang
Intellectual 440 points

1. spi slave transfer: data num < 64K, bitwidth =16, no link and no chain. fifo enable.

How to set ACNT ? For example, I want to transmit and receive 1000 bytes, can I set register like this:

ACNT = 1000; BCNT = 1; CCNT =1;Async mode

why spi-omap2-spi.c in kernel set ACNT = 2; BCNT = 500?

2. At first, but I did not find the file testdma.c in sdk directory:

http://processors.wiki.ti.com/index.php/AM335x_EDMA_Driver%27s_Guide

"The sample EDMA application is available along with other Module examples as a tar archive in src directory of PSP release package (Directory Structure of PSP release package can be seen here AM335x PSP Package Contents)"

Thanks a lot

  • 0
    TI__Guru**** 393215 points

    What processor are you using and which SDK?

  • 0 in reply to Biser Gatchev-XID
    Intellectual 440 points

    AM335x ti-sdk-am335x-evm-06.00.00.00

    I am studing the code spi master, I have doubt about this code :

     if (tx != NULL) {
            int a_cnt, b_cnt, c_cnt, b_cntrld;

            a_cnt    = 1 << data_type;     
            b_cnt    = 1;
            c_cnt    = element_count / 256;       //why divide 256?
            b_cntrld = SZ_64K - 1;                   // why set this value?

            param.opt          = TCINTEN |
                EDMA_TCC(mcspi_dma->dma_tx_channel) | SYNCDIM ; //ABsync
    ......

    }

        if (rx != NULL) {
            int a_cnt, b_cnt, c_cnt, b_cntrld;

            a_cnt    = 1 << data_type;
            c_cnt    = element_count / (SZ_64K - 1); //
            b_cnt    = element_count - c_cnt * (SZ_64K - 1);
            b_cntrld = SZ_64K - 1;
           param.opt          = TCINTEN |  EDMA_TCC(mcspi_dma->dma_rx_channel); //Async
    ......

    //--------------------------- ---------------------------- -------------------------- ------------------------------ ----------------------------

    My question:

    1. why spi_tx_dma use ABsyn mode, but spi_rx_dma use Async?

    2. relation between data aount and the setting of acnt bcnt ccnt?

  • 0 in reply to Lusheng zhang
    Intellectual 440 points

    Another question:

    There are fore event queues for the device for am335x. Is there something wrong with queue3?

    In edma_test.c, I found when event_queue=3, the dma goes wrong.

    result = edma_alloc_channel (EDMA_CHANNEL_ANY, callback1, NULL, event_queue);

     

    Can you tell me the relationship between evnet_queue and channel?

    If I use spi tx and spi rx dma at the same time, tx_dma and rx_dma can use same evnet_queue?

    (In physical layer, tx_dma and rx_dma will work simultaneously)