2008-08-26 19:37:34

by Geert Uytterhoeven

[permalink] [raw]
Subject: [PATCH] VIDEO_SH_MOBILE_CEU should depend on HAS_DMA (was: Re: m68k allmodconfig)

Hi Andrew,

On Thu, 14 Aug 2008, Andrew Morton wrote:
> ERROR: "dma_alloc_coherent" [drivers/media/video/videobuf-dma-contig.ko] undefined!
> ERROR: "dma_sync_single_for_cpu" [drivers/media/video/videobuf-dma-contig.ko] undefined!
> ERROR: "dma_free_coherent" [drivers/media/video/videobuf-dma-contig.ko] undefined!

M68k allmodconfig still selects Sun-3, which sets NO_DMA.
I guess you're also seeing this on the other NO_DMA platforms (h8300, m32r,
s390, and PCI-less SPARC)?

Below is a patch.

Shouldn't it also (or instead) depend on SUPERH or some SuperH platform?
Or is this not done to have more compile-coverage?

Subject: [PATCH] VIDEO_SH_MOBILE_CEU should depend on HAS_DMA

commit 0d3244d6439c8c31d2a29efd587c7aca9042c8aa ("V4L/DVB (8342):
sh_mobile_ceu_camera: Add SuperH Mobile CEU driver V3") introduced
VIDEO_SH_MOBILE_CEU, which selects VIDEOBUF_DMA_CONTIG. This circumvents the
dependency on HAS_DMA of VIDEOBUF_DMA_CONTIG.

Add a dependency on HAS_DMA to VIDEO_SH_MOBILE_CEU to fix this.

Signed-off-by: Geert Uytterhoeven <[email protected]>
---
drivers/media/video/Kconfig | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

--- a/drivers/media/video/Kconfig
+++ b/drivers/media/video/Kconfig
@@ -968,7 +968,7 @@ config VIDEO_PXA27x

config VIDEO_SH_MOBILE_CEU
tristate "SuperH Mobile CEU Interface driver"
- depends on VIDEO_DEV
+ depends on VIDEO_DEV && HAS_DMA
select SOC_CAMERA
select VIDEOBUF_DMA_CONTIG
---help---

Gr{oetje,eeting}s,

Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds


2008-08-27 07:01:35

by Paul Mundt

[permalink] [raw]
Subject: Re: [PATCH] VIDEO_SH_MOBILE_CEU should depend on HAS_DMA (was: Re: m68k allmodconfig)

On Tue, Aug 26, 2008 at 09:37:21PM +0200, Geert Uytterhoeven wrote:
> Hi Andrew,
>
> On Thu, 14 Aug 2008, Andrew Morton wrote:
> > ERROR: "dma_alloc_coherent" [drivers/media/video/videobuf-dma-contig.ko] undefined!
> > ERROR: "dma_sync_single_for_cpu" [drivers/media/video/videobuf-dma-contig.ko] undefined!
> > ERROR: "dma_free_coherent" [drivers/media/video/videobuf-dma-contig.ko] undefined!
>
> M68k allmodconfig still selects Sun-3, which sets NO_DMA.
> I guess you're also seeing this on the other NO_DMA platforms (h8300, m32r,
> s390, and PCI-less SPARC)?
>
> Below is a patch.
>
> Shouldn't it also (or instead) depend on SUPERH or some SuperH platform?
> Or is this not done to have more compile-coverage?
>
The SUPERH dependency was there initially, but was dropped for increased
compilation coverage. The HAS_DMA dependence is certainly the right thing
to do here.

> Subject: [PATCH] VIDEO_SH_MOBILE_CEU should depend on HAS_DMA
>
> commit 0d3244d6439c8c31d2a29efd587c7aca9042c8aa ("V4L/DVB (8342):
> sh_mobile_ceu_camera: Add SuperH Mobile CEU driver V3") introduced
> VIDEO_SH_MOBILE_CEU, which selects VIDEOBUF_DMA_CONTIG. This circumvents the
> dependency on HAS_DMA of VIDEOBUF_DMA_CONTIG.
>
> Add a dependency on HAS_DMA to VIDEO_SH_MOBILE_CEU to fix this.
>
> Signed-off-by: Geert Uytterhoeven <[email protected]>

Acked-by: Paul Mundt <[email protected]>

2008-08-27 09:41:24

by Magnus Damm

[permalink] [raw]
Subject: Re: [PATCH] VIDEO_SH_MOBILE_CEU should depend on HAS_DMA (was: Re: m68k allmodconfig)

On Wed, Aug 27, 2008 at 4:37 AM, Geert Uytterhoeven
<[email protected]> wrote:
> Hi Andrew,
>
> On Thu, 14 Aug 2008, Andrew Morton wrote:
>> ERROR: "dma_alloc_coherent" [drivers/media/video/videobuf-dma-contig.ko] undefined!
>> ERROR: "dma_sync_single_for_cpu" [drivers/media/video/videobuf-dma-contig.ko] undefined!
>> ERROR: "dma_free_coherent" [drivers/media/video/videobuf-dma-contig.ko] undefined!
>
> M68k allmodconfig still selects Sun-3, which sets NO_DMA.
> I guess you're also seeing this on the other NO_DMA platforms (h8300, m32r,
> s390, and PCI-less SPARC)?
>
> Below is a patch.
>
> Shouldn't it also (or instead) depend on SUPERH or some SuperH platform?
> Or is this not done to have more compile-coverage?
>
> Subject: [PATCH] VIDEO_SH_MOBILE_CEU should depend on HAS_DMA
>
> commit 0d3244d6439c8c31d2a29efd587c7aca9042c8aa ("V4L/DVB (8342):
> sh_mobile_ceu_camera: Add SuperH Mobile CEU driver V3") introduced
> VIDEO_SH_MOBILE_CEU, which selects VIDEOBUF_DMA_CONTIG. This circumvents the
> dependency on HAS_DMA of VIDEOBUF_DMA_CONTIG.
>
> Add a dependency on HAS_DMA to VIDEO_SH_MOBILE_CEU to fix this.

Thank you!

> Signed-off-by: Geert Uytterhoeven <[email protected]>

Acked-by: Magnus Damm <[email protected]>