LPC MicrocontrollersのトピックLPC4330 SDIF DMA Starting additional data write transfers on open ended multi-block writes

LPC4330 SDIF DMA Starting additional data write transfers on open ended multi-block writes

nsmith17044 — Thu, 20 Oct 2016 18:34:40 GMT

I'm attempting to perform an open ended multi-block write operation to a SD memory card. Once the first multi-block transfer is finished (not sending auto_stop at the end), the next DMA is setup with Chip_SDIF_SetByteCnt() and Chip_SDIF_DmaSetup().

Once the card is no longer busy, how do I initiate that DMA data transfer withing issuing a command on the CMD line? I've been through the registers a few times and can't seem to see how to do this. Or is there a SD command that would be safe to send that would allow proper data write transfer to continue while the card is in multi-block write mode and start the DMA transfer that has been setup?

Re: LPC4330 SDIF DMA Starting additional data write transfers on open ended multi-block writes

nsmith17044 — Mon, 24 Oct 2016 15:51:57 GMT

One of the solutions I tried was to send a 'send status' command with the MCI_CMD_DAT_WR and MCI_CMD_DAT_EXP bits set in the command register. I had hoped that this will start the DMA transfer. This does issue the command on the CMD bus line but still does NOT start the DMA transfer of data.

I had hoped that simply setting the MCI_CMD_DAT_WR and MCI_CMD_DAT_EXP bits would trick the SDIF DMA into starting the initialized descriptor.

The FIFO is cleared and the DMA initialized with:

    // setup the SDIF DMA to write the data
    Chip_SDIF_SetClearIntFifo(pSDMMC);
    Chip_SDIF_SetByteCnt(pSDMMC, cbWrote);
    Chip_SDIF_DmaSetup(pSDMMC, &g_card_info->sdif_dev, (uint32_t) buffer, cbWrote);
    Chip_SDIF_ClrIntStatus( LPC_SDMMC , 0xFFFFFFFF);

Can anyone tell me how to continue to write data via the SDIF DMA? Or a way to perform the 'continued' open-ended multiblock write process?

Re: LPC4330 SDIF DMA Starting additional data write transfers on open ended multi-block writes

nsmith17044 — Mon, 24 Oct 2016 18:56:25 GMT

PROBLEM SOLVED:

As I watched more carefully I noticed that there was no command response after the SEND STATUS. I realized that I wasn't supplying a proper argument for SEND STATUS and therefore the card wasn't responding with a response. After correcting this oversight, the command response was returned and I saw the DMA start as expected.

Evidently the SDIF needs to see a proper command response (even it it is not expected) for the SDIF DMA to start.