Possible shortcoming in HAL SD library

[MClar.3]

Been trying to get an test application working on a STM32F469NIH Discovery board.

Should be fairly simple - Write a file to the SD Card, with FatFS, the read back:

Try as I might, I couldn't get things working.

After a bit of debugging, I think the issue lies with the DMA callback functions. I'm using FW_F4 V1.27.1, on CubeIDE, and it seems that the supplied SD_DMATransmitCplt() function isn't quite complete, when compared to the SD_DMAReceiveCplt() function.

There seems to be a couple of things missing - Firstly the call to the callback function - This is a must for the SD interface, because there's a queue that is waited on. Without the callback function being called, nothing is ever put on the queue, so a timeout error is received.

The other important bit seems to be the line of code that unlocks the HAL. A sensible thing to do, as we've just finished transferring data. Again, without it, some other bits of code later on fail as the HAL is still locked.

So I copied the code from the RX complete, and that seems to make things work - See attached text file

The question is, am I on the right track here? I'd like to help improve things, but at the same time, I don't want to fix things worse. A proper fix is better than a band-aid.

[I.KHACHINE]

Hello MClar,
I'm analyzing the issue that you mentioned.
Could you pls give me the DMA configuration that you are using the both cases Transmit and Receive?
Best regards,
Ismail

[OTedd.1]

Hi,

Was there any solution to this issue?

I think I have a similar issue where within `SD_write` in `sd_diskio.c`, my program hangs on line `status = osMessageQueueGet(SDQueueID, (void *)&event, NULL, SD_TIMEOUT);` (line 463) because nothing ever gets put into the queue `SDQueueID`.

`BSP_SD_WriteCpltCallback` never gets called.

Like @MClar.3  I see that `SD_DMATransmitCplt` DOES get called.  Is there some code missing there or is there some other configuration error?

I am using an STM32F427GT6, with STM32CubeIDE 1.12.0.
I am using FreeRTOS and FatFS.

Here is my `HAL_SD_MspInit` function

/**
* @brief SD MSP Initialization
* This function configures the hardware resources used in this example
* @param hsd: SD handle pointer
* @retval None
*/
void HAL_SD_MspInit(SD_HandleTypeDef* hsd)
{
  GPIO_InitTypeDef GPIO_InitStruct = {0};
  if(hsd->Instance==SDIO)
  {
  /* USER CODE BEGIN SDIO_MspInit 0 */

  /* USER CODE END SDIO_MspInit 0 */
    /* Peripheral clock enable */
    __HAL_RCC_SDIO_CLK_ENABLE();

    __HAL_RCC_GPIOC_CLK_ENABLE();
    __HAL_RCC_GPIOD_CLK_ENABLE();
    /**SDIO GPIO Configuration
    PC8     ------> SDIO_D0
    PC9     ------> SDIO_D1
    PC10     ------> SDIO_D2
    PC11     ------> SDIO_D3
    PC12     ------> SDIO_CK
    PD2     ------> SDIO_CMD
    */
    GPIO_InitStruct.Pin = GPIO_PIN_8|GPIO_PIN_9|GPIO_PIN_10|GPIO_PIN_11
                          |GPIO_PIN_12;
    GPIO_InitStruct.Mode = GPIO_MODE_AF_PP;
    GPIO_InitStruct.Pull = GPIO_NOPULL;
    GPIO_InitStruct.Speed = GPIO_SPEED_FREQ_VERY_HIGH;
    GPIO_InitStruct.Alternate = GPIO_AF12_SDIO;
    HAL_GPIO_Init(GPIOC, &GPIO_InitStruct);

    GPIO_InitStruct.Pin = GPIO_PIN_2;
    GPIO_InitStruct.Mode = GPIO_MODE_AF_PP;
    GPIO_InitStruct.Pull = GPIO_NOPULL;
    GPIO_InitStruct.Speed = GPIO_SPEED_FREQ_VERY_HIGH;
    GPIO_InitStruct.Alternate = GPIO_AF12_SDIO;
    HAL_GPIO_Init(GPIOD, &GPIO_InitStruct);

    /* SDIO DMA Init */
    /* SDIO_RX Init */
    hdma_sdio_rx.Instance = DMA2_Stream3;
    hdma_sdio_rx.Init.Channel = DMA_CHANNEL_4;
    hdma_sdio_rx.Init.Direction = DMA_PERIPH_TO_MEMORY;
    hdma_sdio_rx.Init.PeriphInc = DMA_PINC_DISABLE;
    hdma_sdio_rx.Init.MemInc = DMA_MINC_ENABLE;
    hdma_sdio_rx.Init.PeriphDataAlignment = DMA_PDATAALIGN_WORD;
    hdma_sdio_rx.Init.MemDataAlignment = DMA_MDATAALIGN_WORD;
    hdma_sdio_rx.Init.Mode = DMA_PFCTRL;
    hdma_sdio_rx.Init.Priority = DMA_PRIORITY_VERY_HIGH;
    hdma_sdio_rx.Init.FIFOMode = DMA_FIFOMODE_ENABLE;
    hdma_sdio_rx.Init.FIFOThreshold = DMA_FIFO_THRESHOLD_FULL;
    hdma_sdio_rx.Init.MemBurst = DMA_MBURST_INC4;
    hdma_sdio_rx.Init.PeriphBurst = DMA_PBURST_INC4;
    if (HAL_DMA_Init(&hdma_sdio_rx) != HAL_OK)
    {
      Error_Handler();
    }

    __HAL_LINKDMA(hsd,hdmarx,hdma_sdio_rx);

    /* SDIO_TX Init */
    hdma_sdio_tx.Instance = DMA2_Stream6;
    hdma_sdio_tx.Init.Channel = DMA_CHANNEL_4;
    hdma_sdio_tx.Init.Direction = DMA_MEMORY_TO_PERIPH;
    hdma_sdio_tx.Init.PeriphInc = DMA_PINC_DISABLE;
    hdma_sdio_tx.Init.MemInc = DMA_MINC_ENABLE;
    hdma_sdio_tx.Init.PeriphDataAlignment = DMA_PDATAALIGN_WORD;
    hdma_sdio_tx.Init.MemDataAlignment = DMA_MDATAALIGN_WORD;
    hdma_sdio_tx.Init.Mode = DMA_PFCTRL;
    hdma_sdio_tx.Init.Priority = DMA_PRIORITY_VERY_HIGH;
    hdma_sdio_tx.Init.FIFOMode = DMA_FIFOMODE_ENABLE;
    hdma_sdio_tx.Init.FIFOThreshold = DMA_FIFO_THRESHOLD_FULL;
    hdma_sdio_tx.Init.MemBurst = DMA_MBURST_INC4;
    hdma_sdio_tx.Init.PeriphBurst = DMA_PBURST_INC4;
    if (HAL_DMA_Init(&hdma_sdio_tx) != HAL_OK)
    {
      Error_Handler();
    }

    __HAL_LINKDMA(hsd,hdmatx,hdma_sdio_tx);

  /* USER CODE BEGIN SDIO_MspInit 1 */
  // DMA priority above set to VERY_HIGH as suggested by
  // https://community.st.com/t5/stm32-mcus-embedded-software/stm32f4-sdio-fat-write-problem/td-p/406687/page/2
  /* USER CODE END SDIO_MspInit 1 */
  }

}

By modifying the `SD_DMATransmitCplt` function in the `stm32f4xx_hal_sd.c` file so that the `Tx` handler is called (similar but not quite the same as @MClar.3 modification)

static void SD_DMATransmitCplt(DMA_HandleTypeDef *hdma)
{
  SD_HandleTypeDef* hsd = (SD_HandleTypeDef* )(hdma->Parent);

  // /* Enable DATAEND Interrupt */
  // __HAL_SD_ENABLE_IT(hsd, (SDIO_IT_DATAEND));
  
  /* Disable the DMA transfer for transmit request by setting the DMAEN bit
  in the SD DCTRL register */
  hsd->Instance->DCTRL &= (uint32_t)~((uint32_t)SDIO_DCTRL_DMAEN);

  /* Clear all the static flags */
  __HAL_SD_CLEAR_FLAG(hsd, SDIO_STATIC_DATA_FLAGS);

  hsd->State = HAL_SD_STATE_READY;
  hsd->Context = SD_CONTEXT_NONE;

#if (USE_HAL_SD_REGISTER_CALLBACKS == 1)
  hsd->TxCpltCallback(hsd);
#else
  HAL_SD_TxCpltCallback(hsd);
#endif
}

then writes are working.

However is this the proper solution?

Many thanks,
Ollie

[MClar.3]

Hi @I.KHACHINE 

Apologies for the (very long) delay in a reply...

Here's the SD DMA init from the sdio.c that CubeMX generates:

/* SDIO DMA Init */
/* SDIO_RX Init */
hdma_sdio_rx.Instance = DMA2_Stream3;
hdma_sdio_rx.Init.Channel = DMA_CHANNEL_4;
hdma_sdio_rx.Init.Direction = DMA_PERIPH_TO_MEMORY;
hdma_sdio_rx.Init.PeriphInc = DMA_PINC_DISABLE;
hdma_sdio_rx.Init.MemInc = DMA_MINC_ENABLE;
hdma_sdio_rx.Init.PeriphDataAlignment = DMA_PDATAALIGN_WORD;
hdma_sdio_rx.Init.MemDataAlignment = DMA_MDATAALIGN_WORD;
hdma_sdio_rx.Init.Mode = DMA_PFCTRL;
hdma_sdio_rx.Init.Priority = DMA_PRIORITY_LOW;
hdma_sdio_rx.Init.FIFOMode = DMA_FIFOMODE_ENABLE;
hdma_sdio_rx.Init.FIFOThreshold = DMA_FIFO_THRESHOLD_FULL;
hdma_sdio_rx.Init.MemBurst = DMA_MBURST_INC4;
hdma_sdio_rx.Init.PeriphBurst = DMA_PBURST_INC4;
if (HAL_DMA_Init(&hdma_sdio_rx) != HAL_OK)
{
  Error_Handler();
}

__HAL_LINKDMA(sdHandle,hdmarx,hdma_sdio_rx);

/* SDIO_TX Init */
hdma_sdio_tx.Instance = DMA2_Stream6;
hdma_sdio_tx.Init.Channel = DMA_CHANNEL_4;
hdma_sdio_tx.Init.Direction = DMA_MEMORY_TO_PERIPH;
hdma_sdio_tx.Init.PeriphInc = DMA_PINC_DISABLE;
hdma_sdio_tx.Init.MemInc = DMA_MINC_ENABLE;
hdma_sdio_tx.Init.PeriphDataAlignment = DMA_PDATAALIGN_WORD;
hdma_sdio_tx.Init.MemDataAlignment = DMA_MDATAALIGN_WORD;
hdma_sdio_tx.Init.Mode = DMA_PFCTRL;
hdma_sdio_tx.Init.Priority = DMA_PRIORITY_LOW;
hdma_sdio_tx.Init.FIFOMode = DMA_FIFOMODE_ENABLE;
hdma_sdio_tx.Init.FIFOThreshold = DMA_FIFO_THRESHOLD_FULL;
hdma_sdio_tx.Init.MemBurst = DMA_MBURST_INC4;
hdma_sdio_tx.Init.PeriphBurst = DMA_PBURST_INC4;
if (HAL_DMA_Init(&hdma_sdio_tx) != HAL_OK)
{
  Error_Handler();
}

__HAL_LINKDMA(sdHandle,hdmatx,hdma_sdio_tx);

I've just updated the code to the latest I have on CubeIDE 1.15.1, which is STM32F4 MCU package 1.28.0 (previously 1.27.1) and FATFs R0.12c. The SD_DMATransmitCplt() function still looks a bit bare. I've had other issues on the project (with a lithium battery power path / charger IC misbehaving and randomly powering on when it should be off), so it's been back burner for a while.

@OTedd.1, I'll reload the new MCU package onto the dev board, and see if I can get a better solution - Maybe we can get a proper fix in place. 

Is it possible to submit changes / pull requests to ST's GIT repo to include a fix?