diff mbox series

[v2,06/10] drm/ingenic: Set DMA descriptor chain address in probe

Message ID 20200629235210.441709-6-paul@crapouillou.net (mailing list archive)
State Not Applicable
Headers show
Series [v2,01/10] dt-bindings: display: Convert ingenic,lcd.txt to YAML | expand

Commit Message

Paul Cercueil June 29, 2020, 11:52 p.m. UTC
The address of the DMA descriptor never changes. It can therefore be set
in the probe function.

Signed-off-by: Paul Cercueil <paul@crapouillou.net>
---

Notes:
    v2: No change

 drivers/gpu/drm/ingenic/ingenic-drm-drv.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

Comments

Sam Ravnborg June 30, 2020, 11:44 a.m. UTC | #1
Hi Paul.

On Tue, Jun 30, 2020 at 01:52:06AM +0200, Paul Cercueil wrote:
> The address of the DMA descriptor never changes. It can therefore be set
> in the probe function.
> 
> Signed-off-by: Paul Cercueil <paul@crapouillou.net>
> ---
> 
> Notes:
>     v2: No change
> 
>  drivers/gpu/drm/ingenic/ingenic-drm-drv.c | 5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/gpu/drm/ingenic/ingenic-drm-drv.c b/drivers/gpu/drm/ingenic/ingenic-drm-drv.c
> index 924c8daf071a..f7b0c5dc8cd8 100644
> --- a/drivers/gpu/drm/ingenic/ingenic-drm-drv.c
> +++ b/drivers/gpu/drm/ingenic/ingenic-drm-drv.c
> @@ -358,8 +358,6 @@ static void ingenic_drm_crtc_atomic_flush(struct drm_crtc *crtc,
>  		ingenic_drm_crtc_update_ctrl(priv, finfo);
>  
>  		clk_set_rate(priv->pix_clk, state->adjusted_mode.clock * 1000);
> -
> -		regmap_write(priv->map, JZ_REG_LCD_DA0, priv->dma_hwdesc->next);
>  	}
>  
>  	if (event) {
> @@ -757,6 +755,9 @@ static int ingenic_drm_probe(struct platform_device *pdev)
>  		}
>  	}
>  
> +	/* Set address of our DMA descriptor chain */
> +	regmap_write(priv->map, JZ_REG_LCD_DA0, priv->dma_hwdesc_phys);
> +


What is the purpose of this code after the assignment was moved:

static int ingenic_drm_probe(struct platform_device *pdev)
{
...
	priv->dma_hwdesc->next = priv->dma_hwdesc_phys;

I do not see ->next used anymore.
Could this assignment be dropped?


And the following line:
	priv->dma_hwdesc->id = 0xdeafbead;

I can only see ->id be assingned, it is never referenced??

Maybe this is all related to the structure assumed by the data pointed
to by JZ_REG_LCD_DA0?

	Sam

>  	ret = drm_dev_register(drm, 0);
>  	if (ret) {
>  		dev_err(dev, "Failed to register DRM driver\n");
> -- 
> 2.27.0
> 
> _______________________________________________
> dri-devel mailing list
> dri-devel@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/dri-devel
Paul Cercueil June 30, 2020, 11:54 a.m. UTC | #2
Hi Sam,

Le mar. 30 juin 2020 à 13:44, Sam Ravnborg <sam@ravnborg.org> a écrit 
:
> Hi Paul.
> 
> On Tue, Jun 30, 2020 at 01:52:06AM +0200, Paul Cercueil wrote:
>>  The address of the DMA descriptor never changes. It can therefore 
>> be set
>>  in the probe function.
>> 
>>  Signed-off-by: Paul Cercueil <paul@crapouillou.net>
>>  ---
>> 
>>  Notes:
>>      v2: No change
>> 
>>   drivers/gpu/drm/ingenic/ingenic-drm-drv.c | 5 +++--
>>   1 file changed, 3 insertions(+), 2 deletions(-)
>> 
>>  diff --git a/drivers/gpu/drm/ingenic/ingenic-drm-drv.c 
>> b/drivers/gpu/drm/ingenic/ingenic-drm-drv.c
>>  index 924c8daf071a..f7b0c5dc8cd8 100644
>>  --- a/drivers/gpu/drm/ingenic/ingenic-drm-drv.c
>>  +++ b/drivers/gpu/drm/ingenic/ingenic-drm-drv.c
>>  @@ -358,8 +358,6 @@ static void 
>> ingenic_drm_crtc_atomic_flush(struct drm_crtc *crtc,
>>   		ingenic_drm_crtc_update_ctrl(priv, finfo);
>> 
>>   		clk_set_rate(priv->pix_clk, state->adjusted_mode.clock * 1000);
>>  -
>>  -		regmap_write(priv->map, JZ_REG_LCD_DA0, priv->dma_hwdesc->next);
>>   	}
>> 
>>   	if (event) {
>>  @@ -757,6 +755,9 @@ static int ingenic_drm_probe(struct 
>> platform_device *pdev)
>>   		}
>>   	}
>> 
>>  +	/* Set address of our DMA descriptor chain */
>>  +	regmap_write(priv->map, JZ_REG_LCD_DA0, priv->dma_hwdesc_phys);
>>  +
> 
> 
> What is the purpose of this code after the assignment was moved:
> 
> static int ingenic_drm_probe(struct platform_device *pdev)
> {
> ...
> 	priv->dma_hwdesc->next = priv->dma_hwdesc_phys;
> 
> I do not see ->next used anymore.
> Could this assignment be dropped?
> 
> 
> And the following line:
> 	priv->dma_hwdesc->id = 0xdeafbead;
> 
> I can only see ->id be assingned, it is never referenced??
> 
> Maybe this is all related to the structure assumed by the data pointed
> to by JZ_REG_LCD_DA0?

As its name suggests, 'dma_hwdesc' points to a DMA hardware descriptor. 
Setting ->next sets the address of the next descriptor (which is itself 
since we only use one descriptor). The 'id' field will be copied by the 
hardware to a register when the DMA descriptor is loaded, so we can 
know from userspace (debugfs) what DMA descriptors are loaded by 
looking up the IDs.

-Paul

> 	Sam
> 
>>   	ret = drm_dev_register(drm, 0);
>>   	if (ret) {
>>   		dev_err(dev, "Failed to register DRM driver\n");
>>  --
>>  2.27.0
>> 
>>  _______________________________________________
>>  dri-devel mailing list
>>  dri-devel@lists.freedesktop.org
>>  https://lists.freedesktop.org/mailman/listinfo/dri-devel
diff mbox series

Patch

diff --git a/drivers/gpu/drm/ingenic/ingenic-drm-drv.c b/drivers/gpu/drm/ingenic/ingenic-drm-drv.c
index 924c8daf071a..f7b0c5dc8cd8 100644
--- a/drivers/gpu/drm/ingenic/ingenic-drm-drv.c
+++ b/drivers/gpu/drm/ingenic/ingenic-drm-drv.c
@@ -358,8 +358,6 @@  static void ingenic_drm_crtc_atomic_flush(struct drm_crtc *crtc,
 		ingenic_drm_crtc_update_ctrl(priv, finfo);
 
 		clk_set_rate(priv->pix_clk, state->adjusted_mode.clock * 1000);
-
-		regmap_write(priv->map, JZ_REG_LCD_DA0, priv->dma_hwdesc->next);
 	}
 
 	if (event) {
@@ -757,6 +755,9 @@  static int ingenic_drm_probe(struct platform_device *pdev)
 		}
 	}
 
+	/* Set address of our DMA descriptor chain */
+	regmap_write(priv->map, JZ_REG_LCD_DA0, priv->dma_hwdesc_phys);
+
 	ret = drm_dev_register(drm, 0);
 	if (ret) {
 		dev_err(dev, "Failed to register DRM driver\n");