diff mbox

[v2,04/14] ASoC: sun4i-codec: Increase DMA max burst to 8

Message ID 20161103075556.29018-5-wens@csie.org (mailing list archive)
State Accepted
Commit 730e2dd0cbc7a7ec10174d9d291cdd8e8082a948
Headers show

Commit Message

Chen-Yu Tsai Nov. 3, 2016, 7:55 a.m. UTC
According to the DMA engine API documentation, maxburst denotes the
largest possible size of a single transfer, so as not to overflow
destination FIFOs as explained in this excerpt from dmaengine.h

 * @src_maxburst: the maximum number of words (note: words, as in
 * units of the src_addr_width member, not bytes) that can be sent
 * in one burst to the device. Typically something like half the
 * FIFO depth on I/O peripherals so you don't overflow it. This
 * may or may not be applicable on memory sources.
 * @dst_maxburst: same as src_maxburst but for destination target
 * mutatis mutandis.

The TX FIFO is 64 samples deep for stereo, and the RX FIFO is 16
samples deep. So maxburst could be 32 and 8 for TX and RX respectively.

Unfortunately the sunxi DMA controller driver takes maxburst as
the requested burst size, rather than a limit, and returns an error
for unsupported values. The original value was 4, but some later
SoCs do not officially support this burst size.

This patch increases maxburst on the TX side to 8, which is supported
by all variants of the sunxi DMA controller.

Cc: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Chen-Yu Tsai <wens@csie.org>
---
 sound/soc/sunxi/sun4i-codec.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

Comments

Maxime Ripard Nov. 3, 2016, 8:26 a.m. UTC | #1
On Thu, Nov 03, 2016 at 03:55:46PM +0800, Chen-Yu Tsai wrote:
> According to the DMA engine API documentation, maxburst denotes the
> largest possible size of a single transfer, so as not to overflow
> destination FIFOs as explained in this excerpt from dmaengine.h
> 
>  * @src_maxburst: the maximum number of words (note: words, as in
>  * units of the src_addr_width member, not bytes) that can be sent
>  * in one burst to the device. Typically something like half the
>  * FIFO depth on I/O peripherals so you don't overflow it. This
>  * may or may not be applicable on memory sources.
>  * @dst_maxburst: same as src_maxburst but for destination target
>  * mutatis mutandis.
> 
> The TX FIFO is 64 samples deep for stereo, and the RX FIFO is 16
> samples deep. So maxburst could be 32 and 8 for TX and RX respectively.
> 
> Unfortunately the sunxi DMA controller driver takes maxburst as
> the requested burst size, rather than a limit, and returns an error
> for unsupported values. The original value was 4, but some later
> SoCs do not officially support this burst size.
> 
> This patch increases maxburst on the TX side to 8, which is supported
> by all variants of the sunxi DMA controller.
> 
> Cc: Vinod Koul <vinod.koul@intel.com>
> Signed-off-by: Chen-Yu Tsai <wens@csie.org>

Acked-by: Maxime Ripard <maxime.ripard@free-electrons.com>

Thanks,
Maxime
Priit Laes Nov. 3, 2016, 8:42 a.m. UTC | #2
On Thu, 2016-11-03 at 15:55 +0800, Chen-Yu Tsai wrote:
> According to the DMA engine API documentation, maxburst denotes the
> largest possible size of a single transfer, so as not to overflow
> destination FIFOs as explained in this excerpt from dmaengine.h
> 
>  * @src_maxburst: the maximum number of words (note: words, as in
>  * units of the src_addr_width member, not bytes) that can be sent
>  * in one burst to the device. Typically something like half the
>  * FIFO depth on I/O peripherals so you don't overflow it. This
>  * may or may not be applicable on memory sources.
>  * @dst_maxburst: same as src_maxburst but for destination target

>  * mutatis mutandis.

^^ :)

> 
> The TX FIFO is 64 samples deep for stereo, and the RX FIFO is 16
> samples deep. So maxburst could be 32 and 8 for TX and RX
> respectively.
> 
> Unfortunately the sunxi DMA controller driver takes maxburst as
> the requested burst size, rather than a limit, and returns an error
> for unsupported values. The original value was 4, but some later
> SoCs do not officially support this burst size.
> 
> This patch increases maxburst on the TX side to 8, which is supported
> by all variants of the sunxi DMA controller.
> 
> Cc: Vinod Koul <vinod.koul@intel.com>
> Signed-off-by: Chen-Yu Tsai <wens@csie.org>
> ---
>  sound/soc/sunxi/sun4i-codec.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/sound/soc/sunxi/sun4i-codec.c b/sound/soc/sunxi/sun4i-
> codec.c
> index 61ae502a5061..d867b96d367b 100644
> --- a/sound/soc/sunxi/sun4i-codec.c
> +++ b/sound/soc/sunxi/sun4i-codec.c
> @@ -886,12 +886,12 @@ static int sun4i_codec_probe(struct
> platform_device *pdev)
>  
>  	/* DMA configuration for TX FIFO */
>  	scodec->playback_dma_data.addr = res->start + quirks-
> >reg_dac_txdata;
> -	scodec->playback_dma_data.maxburst = 4;
> +	scodec->playback_dma_data.maxburst = 8;
>  	scodec->playback_dma_data.addr_width =
> DMA_SLAVE_BUSWIDTH_2_BYTES;
>  
>  	/* DMA configuration for RX FIFO */
>  	scodec->capture_dma_data.addr = res->start + quirks-
> >reg_adc_rxdata;
> -	scodec->capture_dma_data.maxburst = 4;
> +	scodec->capture_dma_data.maxburst = 8;
>  	scodec->capture_dma_data.addr_width =
> DMA_SLAVE_BUSWIDTH_2_BYTES;
>  
>  	ret = snd_soc_register_codec(&pdev->dev, quirks->codec,
> -- 
> 2.10.2
>
diff mbox

Patch

diff --git a/sound/soc/sunxi/sun4i-codec.c b/sound/soc/sunxi/sun4i-codec.c
index 61ae502a5061..d867b96d367b 100644
--- a/sound/soc/sunxi/sun4i-codec.c
+++ b/sound/soc/sunxi/sun4i-codec.c
@@ -886,12 +886,12 @@  static int sun4i_codec_probe(struct platform_device *pdev)
 
 	/* DMA configuration for TX FIFO */
 	scodec->playback_dma_data.addr = res->start + quirks->reg_dac_txdata;
-	scodec->playback_dma_data.maxburst = 4;
+	scodec->playback_dma_data.maxburst = 8;
 	scodec->playback_dma_data.addr_width = DMA_SLAVE_BUSWIDTH_2_BYTES;
 
 	/* DMA configuration for RX FIFO */
 	scodec->capture_dma_data.addr = res->start + quirks->reg_adc_rxdata;
-	scodec->capture_dma_data.maxburst = 4;
+	scodec->capture_dma_data.maxburst = 8;
 	scodec->capture_dma_data.addr_width = DMA_SLAVE_BUSWIDTH_2_BYTES;
 
 	ret = snd_soc_register_codec(&pdev->dev, quirks->codec,