[PATCH] spi: fsl-lpspi: replace dmaengine_terminate_all with dmaengine_terminate_sync

[PATCH] spi: fsl-lpspi: replace dmaengine_terminate_all with dmaengine_terminate_sync

[Carlos Song (OSS)]

From: Carlos Song <carlos.song@nxp.com>

The terminate API dmaengine_terminate_all() has been deprecated, improve
driver with dmaengine_terminate_sync().

Fixes: 09c04466ce7e ("spi: lpspi: add dma mode support")
Cc: stable@vger.kernel.org
Signed-off-by: Carlos Song <carlos.song@nxp.com>
---
 drivers/spi/spi-fsl-lpspi.c | 18 +++++++++---------
 1 file changed, 9 insertions(+), 9 deletions(-)

diff --git a/drivers/spi/spi-fsl-lpspi.c b/drivers/spi/spi-fsl-lpspi.c
index e201309f8aae..1a94a42fac31 100644
--- a/drivers/spi/spi-fsl-lpspi.c
+++ b/drivers/spi/spi-fsl-lpspi.c
@@ -647,7 +647,7 @@ static int fsl_lpspi_dma_transfer(struct spi_controller *controller,
 				tx->sgl, tx->nents, DMA_MEM_TO_DEV,
 				DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
 	if (!desc_tx) {
-		dmaengine_terminate_all(controller->dma_tx);
+		dmaengine_terminate_sync(controller->dma_tx);
 		return -EINVAL;
 	}
 
@@ -668,8 +668,8 @@ static int fsl_lpspi_dma_transfer(struct spi_controller *controller,
 							transfer_timeout);
 		if (!time_left) {
 			dev_err(fsl_lpspi->dev, "I/O Error in DMA TX\n");
-			dmaengine_terminate_all(controller->dma_tx);
-			dmaengine_terminate_all(controller->dma_rx);
+			dmaengine_terminate_sync(controller->dma_tx);
+			dmaengine_terminate_sync(controller->dma_rx);
 			fsl_lpspi_reset(fsl_lpspi);
 			return -ETIMEDOUT;
 		}
@@ -678,8 +678,8 @@ static int fsl_lpspi_dma_transfer(struct spi_controller *controller,
 							transfer_timeout);
 		if (!time_left) {
 			dev_err(fsl_lpspi->dev, "I/O Error in DMA RX\n");
-			dmaengine_terminate_all(controller->dma_tx);
-			dmaengine_terminate_all(controller->dma_rx);
+			dmaengine_terminate_sync(controller->dma_tx);
+			dmaengine_terminate_sync(controller->dma_rx);
 			fsl_lpspi_reset(fsl_lpspi);
 			return -ETIMEDOUT;
 		}
@@ -688,8 +688,8 @@ static int fsl_lpspi_dma_transfer(struct spi_controller *controller,
 			fsl_lpspi->target_aborted) {
 			dev_dbg(fsl_lpspi->dev,
 				"I/O Error in DMA TX interrupted\n");
-			dmaengine_terminate_all(controller->dma_tx);
-			dmaengine_terminate_all(controller->dma_rx);
+			dmaengine_terminate_sync(controller->dma_tx);
+			dmaengine_terminate_sync(controller->dma_rx);
 			fsl_lpspi_reset(fsl_lpspi);
 			return -EINTR;
 		}
@@ -698,8 +698,8 @@ static int fsl_lpspi_dma_transfer(struct spi_controller *controller,
 			fsl_lpspi->target_aborted) {
 			dev_dbg(fsl_lpspi->dev,
 				"I/O Error in DMA RX interrupted\n");
-			dmaengine_terminate_all(controller->dma_tx);
-			dmaengine_terminate_all(controller->dma_rx);
+			dmaengine_terminate_sync(controller->dma_tx);
+			dmaengine_terminate_sync(controller->dma_rx);
 			fsl_lpspi_reset(fsl_lpspi);
 			return -EINTR;
 		}
-- 
2.43.0

Re: [PATCH] spi: fsl-lpspi: replace dmaengine_terminate_all with dmaengine_terminate_sync

[sashiko-bot]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Critical] Missing termination of the active RX DMA channel when TX DMA preparation fails, leading to potential memory corruption.
--

commit ec7592d39a6caaca48276f43db4b9a0f4fa50bfd
Author: Carlos Song <carlos.song@nxp.com>
spi: fsl-lpspi: replace dmaengine_terminate_all with dmaengine_terminate_sync

This commit replaces the deprecated dmaengine_terminate_all() API with
dmaengine_terminate_sync() in the fsl-lpspi driver.

> diff --git a/drivers/spi/spi-fsl-lpspi.c b/drivers/spi/spi-fsl-lpspi.c
> index e201309f8aae3..1a94a42fac319 100644
> --- a/drivers/spi/spi-fsl-lpspi.c
> +++ b/drivers/spi/spi-fsl-lpspi.c
[ ... ]
> @@ -647,7 +647,7 @@ static int fsl_lpspi_dma_transfer(struct spi_controller *controller,
>  				tx->sgl, tx->nents, DMA_MEM_TO_DEV,
>  				DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
>  	if (!desc_tx) {
> -		dmaengine_terminate_all(controller->dma_tx);
> +		dmaengine_terminate_sync(controller->dma_tx);

Because controller->dma_rx was already submitted and issued just prior to
preparing desc_tx, does this error path also need to terminate the RX
channel?

If dmaengine_prep_slave_sg fails for the TX channel and the driver returns
-EINVAL, the SPI core will abort the transfer and unmap the buffers. Since
controller->dma_rx is still running, could this cause the DMA controller to
write to freed or unmapped memory?

This wasn't introduced by your patch, but modifying this exact error path
might be a good opportunity to fix the missing RX channel cleanup.

>  		return -EINVAL;
>  	}

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260519104517.2794390-1-carlos.song@oss.nxp.com?part=1