CONFIG_OF vs DMA_OF

CONFIG_OF vs DMA_OF

[Kuninori Morimoto]

Hi Vinod

${LINUX}/drivers/dma/of-dma.c will be compiled if .config has DMA_OF

	obj-$(CONFIG_DMA_OF) += of-dma.o

But, ${LINUX}/include/linux/of_dma.h is based on CONFIG_OF

	#ifdef CONFIG_OF
	extern int of_dma_controller_register(struct device_node *np,
	...
	#else
	static inline int of_dma_controller_register(struct device_node *np,
	...
	#endif

I have issue if .config has CONFIG_OF but not have DMA_OF.
Is this correct patch ?

---------------------------------
diff --git a/include/linux/of_dma.h b/include/linux/of_dma.h
index 98ba752..9440622 100644
--- a/include/linux/of_dma.h
+++ b/include/linux/of_dma.h
@@ -34,7 +34,7 @@ struct of_dma_filter_info {
        dma_filter_fn   filter_fn;
 };
 
-#ifdef CONFIG_OF
+#ifdef DMA_OF
 extern int of_dma_controller_register(struct device_node *np,
                struct dma_chan *(*of_dma_xlate)
                (struct of_phandle_args *, struct of_dma *),
---------------------------------


Best regards
---
Kuninori Morimoto

Re: CONFIG_OF vs DMA_OF

[Vinod Koul]

On Fri, Aug 21, 2015 at 05:35:58AM +0000, Kuninori Morimoto wrote:
> 
> Hi Vinod
> 
> ${LINUX}/drivers/dma/of-dma.c will be compiled if .config has DMA_OF
> 
> 	obj-$(CONFIG_DMA_OF) += of-dma.o
> 
> But, ${LINUX}/include/linux/of_dma.h is based on CONFIG_OF

Intially it was OF only and then DMA_OF was added. I think we missed it
during the move, so yes your below patch makes sense

Can you please send as formal patch for me to apply

-- 
~Vinod

> 
> 	#ifdef CONFIG_OF
> 	extern int of_dma_controller_register(struct device_node *np,
> 	...
> 	#else
> 	static inline int of_dma_controller_register(struct device_node *np,
> 	...
> 	#endif
> 
> I have issue if .config has CONFIG_OF but not have DMA_OF.
> Is this correct patch ?
> 
> ---------------------------------
> diff --git a/include/linux/of_dma.h b/include/linux/of_dma.h
> index 98ba752..9440622 100644
> --- a/include/linux/of_dma.h
> +++ b/include/linux/of_dma.h
> @@ -34,7 +34,7 @@ struct of_dma_filter_info {
>         dma_filter_fn   filter_fn;
>  };
>  
> -#ifdef CONFIG_OF
> +#ifdef DMA_OF
>  extern int of_dma_controller_register(struct device_node *np,
>                 struct dma_chan *(*of_dma_xlate)
>                 (struct of_phandle_args *, struct of_dma *),
> ---------------------------------
> 
> 
> Best regards
> ---
> Kuninori Morimoto

-- 
-- 
~Vinod

[PATCH 0/2] CONFIG_OF vs DMA_OF

[Kuninori Morimoto]

Hi Vinod, Mark

Current OF base DMAEngine API will be compiled if .config has
CONFIG_DMA_OF. but of-dma.h based on CONFIG_OF instead of CONFIG_DMA_OF.
These patches tidyup this issue.

Please check 1) patch carefully, since DMAEngine API has many dependency
to other drivers.

> Mark

I think 2) patch is based on 1). Please check it

Kuninori Morimoto (2):
      1) dmaengine: OF DMAEngine API based on CONFIG_DMA_OF instead of CONFIG_OF
      2) ASoC: rsnd: SND_SOC_RCAR doesn't depend on DMA_OF

 include/linux/of_dma.h | 2 +-
 sound/soc/sh/Kconfig   | 1 -
 2 files changed, 1 insertion(+), 2 deletions(-)


Best regards
---
Kuninori Morimoto

[PATCH 1/2] dmaengine: OF DMAEngine API based on CONFIG_DMA_OF instead of CONFIG_OF

[Kuninori Morimoto]

From: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>

5fa422c ("dmaengine: move drivers/of/dma.c -> drivers/dma/of-dma.c")
moved OF base DMAEngine code to of-dma.c, then it based on CONFIG_DMA_OF.
But, OF base DMAEngine API on of_dma.h still based on CONFIG_OF now.
So, current kernel can't find OF base DMAEngine API if .config has CONFIG_OF,
but not have CONFIG_DMA_OF. This patch tidyup it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
---
 include/linux/of_dma.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/include/linux/of_dma.h b/include/linux/of_dma.h
index 98ba752..36112cd 100644
--- a/include/linux/of_dma.h
+++ b/include/linux/of_dma.h
@@ -34,7 +34,7 @@ struct of_dma_filter_info {
 	dma_filter_fn	filter_fn;
 };
 
-#ifdef CONFIG_OF
+#ifdef CONFIG_DMA_OF
 extern int of_dma_controller_register(struct device_node *np,
 		struct dma_chan *(*of_dma_xlate)
 		(struct of_phandle_args *, struct of_dma *),
-- 
1.9.1

Re: [PATCH 1/2] dmaengine: OF DMAEngine API based on CONFIG_DMA_OF instead of CONFIG_OF

[Vinod Koul]

On Fri, Aug 21, 2015 at 11:48:37AM +0000, Kuninori Morimoto wrote:
> From: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
> 
> 5fa422c ("dmaengine: move drivers/of/dma.c -> drivers/dma/of-dma.c")
> moved OF base DMAEngine code to of-dma.c, then it based on CONFIG_DMA_OF.
> But, OF base DMAEngine API on of_dma.h still based on CONFIG_OF now.
> So, current kernel can't find OF base DMAEngine API if .config has CONFIG_OF,
> but not have CONFIG_DMA_OF. This patch tidyup it.

I did a quick build with arm config, but didn't see any failures. But still
am worried about random config and other builds may find.

So I think it would be safer to merge this patch after merge window so that
we have ample time to fix any issue

-- 
~Vinod

> Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
> ---
>  include/linux/of_dma.h | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/include/linux/of_dma.h b/include/linux/of_dma.h
> index 98ba752..36112cd 100644
> --- a/include/linux/of_dma.h
> +++ b/include/linux/of_dma.h
> @@ -34,7 +34,7 @@ struct of_dma_filter_info {
>  	dma_filter_fn	filter_fn;
>  };
>  
> -#ifdef CONFIG_OF
> +#ifdef CONFIG_DMA_OF
>  extern int of_dma_controller_register(struct device_node *np,
>  		struct dma_chan *(*of_dma_xlate)
>  		(struct of_phandle_args *, struct of_dma *),
> -- 
> 1.9.1
> 

Re: [PATCH 1/2] dmaengine: OF DMAEngine API based on CONFIG_DMA_OF instead of CONFIG_OF

[Kuninori Morimoto]

Hi Vinod

> > 5fa422c ("dmaengine: move drivers/of/dma.c -> drivers/dma/of-dma.c")
> > moved OF base DMAEngine code to of-dma.c, then it based on CONFIG_DMA_OF.
> > But, OF base DMAEngine API on of_dma.h still based on CONFIG_OF now.
> > So, current kernel can't find OF base DMAEngine API if .config has CONFIG_OF,
> > but not have CONFIG_DMA_OF. This patch tidyup it.
> 
> I did a quick build with arm config, but didn't see any failures. But still
> am worried about random config and other builds may find.
> 
> So I think it would be safer to merge this patch after merge window so that
> we have ample time to fix any issue

OK, thanks

Re: [alsa-devel] [PATCH 1/2] dmaengine: OF DMAEngine API based on CONFIG_DMA_OF instead of CONFIG_OF

[Kuninori Morimoto]

Hi Vinod

> > > 5fa422c ("dmaengine: move drivers/of/dma.c -> drivers/dma/of-dma.c")
> > > moved OF base DMAEngine code to of-dma.c, then it based on CONFIG_DMA_OF.
> > > But, OF base DMAEngine API on of_dma.h still based on CONFIG_OF now.
> > > So, current kernel can't find OF base DMAEngine API if .config has CONFIG_OF,
> > > but not have CONFIG_DMA_OF. This patch tidyup it.
> > 
> > I did a quick build with arm config, but didn't see any failures. But still
> > am worried about random config and other builds may find.
> > 
> > So I think it would be safer to merge this patch after merge window so that
> > we have ample time to fix any issue

Which branch can I find this patch ?

Re: [alsa-devel] [PATCH 1/2] dmaengine: OF DMAEngine API based on CONFIG_DMA_OF instead of CONFIG_OF

[Kuninori Morimoto]

Hi Vinod, again

> > > > 5fa422c ("dmaengine: move drivers/of/dma.c -> drivers/dma/of-dma.c")
> > > > moved OF base DMAEngine code to of-dma.c, then it based on CONFIG_DMA_OF.
> > > > But, OF base DMAEngine API on of_dma.h still based on CONFIG_OF now.
> > > > So, current kernel can't find OF base DMAEngine API if .config has CONFIG_OF,
> > > > but not have CONFIG_DMA_OF. This patch tidyup it.
> > > 
> > > I did a quick build with arm config, but didn't see any failures. But still
> > > am worried about random config and other builds may find.
> > > 
> > > So I think it would be safer to merge this patch after merge window so that
> > > we have ample time to fix any issue
> 
> Which branch can I find this patch ?

I still can't find this patch.
Where can I find ?

Re: [alsa-devel] [PATCH 1/2] dmaengine: OF DMAEngine API based on CONFIG_DMA_OF instead of CONFIG_OF

[Kuninori Morimoto]

Hi Vinod

> > > > 5fa422c ("dmaengine: move drivers/of/dma.c -> drivers/dma/of-dma.c")
> > > > moved OF base DMAEngine code to of-dma.c, then it based on CONFIG_DMA_OF.
> > > > But, OF base DMAEngine API on of_dma.h still based on CONFIG_OF now.
> > > > So, current kernel can't find OF base DMAEngine API if .config has CONFIG_OF,
> > > > but not have CONFIG_DMA_OF. This patch tidyup it.
> > > 
> > > I did a quick build with arm config, but didn't see any failures. But still
> > > am worried about random config and other builds may find.
> > > 
> > > So I think it would be safer to merge this patch after merge window so that
> > > we have ample time to fix any issue
> 
> Which branch can I find this patch ?

Which branch can I find this patch ?

Re: [alsa-devel] [PATCH 1/2] dmaengine: OF DMAEngine API based on CONFIG_DMA_OF instead of CONFIG_OF

[Vinod Koul]

On Thu, Sep 10, 2015 at 08:26:11AM +0000, Kuninori Morimoto wrote:
> Hi Vinod
> 
> > > > 5fa422c ("dmaengine: move drivers/of/dma.c -> drivers/dma/of-dma.c")
> > > > moved OF base DMAEngine code to of-dma.c, then it based on CONFIG_DMA_OF.
> > > > But, OF base DMAEngine API on of_dma.h still based on CONFIG_OF now.
> > > > So, current kernel can't find OF base DMAEngine API if .config has CONFIG_OF,
> > > > but not have CONFIG_DMA_OF. This patch tidyup it.
> > > 
> > > I did a quick build with arm config, but didn't see any failures. But still
> > > am worried about random config and other builds may find.
> > > 
> > > So I think it would be safer to merge this patch after merge window so that
> > > we have ample time to fix any issue
> 
> Which branch can I find this patch ?

Sorry I missed to do this after merge window. Applied now

-- 
~Vinod

Re: [alsa-devel] [PATCH 1/2] dmaengine: OF DMAEngine API based on CONFIG_DMA_OF instead of CONFIG_OF

[Kuninori Morimoto]

Hi Vinod, Mark

> > > > > 5fa422c ("dmaengine: move drivers/of/dma.c -> drivers/dma/of-dma.c")
> > > > > moved OF base DMAEngine code to of-dma.c, then it based on CONFIG_DMA_OF.
> > > > > But, OF base DMAEngine API on of_dma.h still based on CONFIG_OF now.
> > > > > So, current kernel can't find OF base DMAEngine API if .config has CONFIG_OF,
> > > > > but not have CONFIG_DMA_OF. This patch tidyup it.
> > > > 
> > > > I did a quick build with arm config, but didn't see any failures. But still
> > > > am worried about random config and other builds may find.
> > > > 
> > > > So I think it would be safer to merge this patch after merge window so that
> > > > we have ample time to fix any issue
> > 
> > Which branch can I find this patch ?
> 
> Sorry I missed to do this after merge window. Applied now

Thank you

>> Mark

Now Vinod's branch (= vinod/topic/of) has this patch
Can you please check [2/2] patch which is related to it

	[PATCH 2/2] ASoC: rsnd: SND_SOC_RCAR doesn't depend on	DMA_OF

[PATCH 2/2] ASoC: rsnd: SND_SOC_RCAR doesn't depend on DMA_OF

[Kuninori Morimoto]

From: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>

8616774("ASoC: rnsd: fix build regression without CONFIG_OF")
added "depends on DMA_OF" in SND_SOC_RCAR to avoid compile error of

    sound/built-in.o: In function `rsnd_dma_request_channel':
    :(.text+0x9fb84): undefined reference to `of_dma_request_slave_channel'

But, it was OF base DMAEngine API definition issue, not SND_SOC_RCAR
issue. This patch remove DMA_OF dependence.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
---
 sound/soc/sh/Kconfig | 1 -
 1 file changed, 1 deletion(-)

diff --git a/sound/soc/sh/Kconfig b/sound/soc/sh/Kconfig
index 07114b0..2b30304 100644
--- a/sound/soc/sh/Kconfig
+++ b/sound/soc/sh/Kconfig
@@ -36,7 +36,6 @@ config SND_SOC_SH4_SIU
 
 config SND_SOC_RCAR
 	tristate "R-Car series SRU/SCU/SSIU/SSI support"
-	depends on DMA_OF
 	select SND_SIMPLE_CARD
 	select REGMAP_MMIO
 	help
-- 
1.9.1