[1/1] dmaengine: fsl-edma: correct calculation of 'nbytes' in multi-fifo scenario

Commit Message

Frank Li Jan. 31, 2024, 4:33 p.m. UTC

From: Joy Zou <[e-mail]>

The 'nbytes' should be equivalent to burst * width in audio multi-fifo
setups. Given that the FIFO width is fixed at 32 bits, adjusts the burst
size for multi-fifo configurations to match the slave maxburst in the
configuration.

Cc: [e-mail]
Fixes: 72f5801a4e2b ("dmaengine: fsl-edma: integrate v3 support")
Signed-off-by: Joy Zou <[e-mail]>
Signed-off-by: Frank Li <[e-mail]>
---
 drivers/dma/fsl-edma-common.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Comments

Vinod Koul Feb. 7, 2024, 8:49 a.m. UTC | #1

On Wed, 31 Jan 2024 11:33:18 -0500, Frank Li wrote:
> The 'nbytes' should be equivalent to burst * width in audio multi-fifo
> setups. Given that the FIFO width is fixed at 32 bits, adjusts the burst
> size for multi-fifo configurations to match the slave maxburst in the
> configuration.
> 
> 

Applied, thanks!

[1/1] dmaengine: fsl-edma: correct calculation of 'nbytes' in multi-fifo scenario
      commit: 9ba17defd9edd87970b701085402bc8ecc3a11d4

Best regards,

Patch

diff --git a/drivers/dma/fsl-edma-common.c b/drivers/dma/fsl-edma-common.c
index b53f46245c377..793f1a7ad5e34 100644
--- a/drivers/dma/fsl-edma-common.c
+++ b/drivers/dma/fsl-edma-common.c
@@ -503,7 +503,7 @@  void fsl_edma_fill_tcd(struct fsl_edma_chan *fsl_chan,
 	if (fsl_chan->is_multi_fifo) {
 		/* set mloff to support multiple fifo */
 		burst = cfg->direction == DMA_DEV_TO_MEM ?
-				cfg->src_addr_width : cfg->dst_addr_width;
+				cfg->src_maxburst : cfg->dst_maxburst;
 		nbytes |= EDMA_V3_TCD_NBYTES_MLOFF(-(burst * 4));
 		/* enable DMLOE/SMLOE */
 		if (cfg->direction == DMA_MEM_TO_DEV) {