ASoC HDMI codec location

ASoC HDMI codec location

[Stephen Warren]

Mark,

I'm looking at enabling audio-over-HDMI on Tegra in the ChromeOS kernel.
This will use the SPDIF controller as the CPU DAI. Right now, I have things
working with the stub SPDIF codec (codecs/spdif_transciever.c), but to
support a non-hard-coded sample rate, need a real codec driver that writes
registers in the HDMI controller.

I see that a Samsung chip has such a codec in drivers/video/sh_mobile_hdmi.c.
The equivalent for Tegra would be drivers/video/tegra/dc/hdmi.c. Note that
this location doesn't yet exist upstream. I just wanted to confirm that this
location makes sense to you, vs. implementing a codec in sound/soc/tegra/hdmi.c,
and having that call functions exposed by the Tegra dc/hdmi.c driver?

Note that the SPDIF controller isn't involved in the Tegra Digital Audio
Switch muxing/switching module at all.

Thanks.

-- 
nvpublic

--
To unsubscribe from this list: send the line "unsubscribe linux-tegra" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: ASoC HDMI codec location

[Mark Brown]

On Tue, Feb 15, 2011 at 12:29:59PM -0800, Stephen Warren wrote:

> I see that a Samsung chip has such a codec in drivers/video/sh_mobile_hdmi.c.

That's not Samsung, that's Renesas' SuperH.

> The equivalent for Tegra would be drivers/video/tegra/dc/hdmi.c. Note that
> this location doesn't yet exist upstream. I just wanted to confirm that this
> location makes sense to you, vs. implementing a codec in sound/soc/tegra/hdmi.c,
> and having that call functions exposed by the Tegra dc/hdmi.c driver?

Note that the SH code has code in *both* locations - there's an ASoC
driver that calls into the video driver as well under sound/soc/sh.

> Note that the SPDIF controller isn't involved in the Tegra Digital Audio
> Switch muxing/switching module at all.

Is the DMA shared?
--
To unsubscribe from this list: send the line "unsubscribe linux-tegra" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

RE: ASoC HDMI codec location

[Stephen Warren]

Mark Brown wrote at Tuesday, February 15, 2011 4:59 PM:
> 
> On Tue, Feb 15, 2011 at 12:29:59PM -0800, Stephen Warren wrote:
> 
> > I see that a Samsung chip has such a codec in
> > drivers/video/sh_mobile_hdmi.c.
> 
> That's not Samsung, that's Renesas' SuperH.

Yes, of course. I'd been looking at samsung/smdk_spdif.c and got the
two confused.

> > The equivalent for Tegra would be drivers/video/tegra/dc/hdmi.c. Note
> > that this location doesn't yet exist upstream. I just wanted to confirm
> > that this location makes sense to you, vs. implementing a codec in
> > sound/soc/tegra/hdmi.c, and having that call functions exposed by the
> > Tegra dc/hdmi.c driver?
> 
> Note that the SH code has code in *both* locations - there's an ASoC
> driver that calls into the video driver as well under sound/soc/sh.

I see there's a machine driver there, which refers to the codec that's
implemented in drivers/video/sh_mobile_hdmi.c. I assume that's all you
mean.

> > Note that the SPDIF controller isn't involved in the Tegra Digital Audio
> > Switch muxing/switching module at all.
> 
> Is the DMA shared?

Yes, I'm using the existing DMA driver unmodified (although this allocates
a separate DMA channel through the DMA controller's API), and currently
lumping the two dai_links into the same card.

-- 
nvpublic

--
To unsubscribe from this list: send the line "unsubscribe linux-tegra" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: ASoC HDMI codec location

[Mark Brown]

On Wed, Feb 16, 2011 at 08:41:58AM -0800, Stephen Warren wrote:
> Mark Brown wrote at Tuesday, February 15, 2011 4:59 PM:

> > > The equivalent for Tegra would be drivers/video/tegra/dc/hdmi.c. Note
> > > that this location doesn't yet exist upstream. I just wanted to confirm
> > > that this location makes sense to you, vs. implementing a codec in
> > > sound/soc/tegra/hdmi.c, and having that call functions exposed by the
> > > Tegra dc/hdmi.c driver?

> > Note that the SH code has code in *both* locations - there's an ASoC
> > driver that calls into the video driver as well under sound/soc/sh.

> I see there's a machine driver there, which refers to the codec that's
> implemented in drivers/video/sh_mobile_hdmi.c. I assume that's all you
> mean.

Hrm, I remembered that differently.  I'd prefer it if more of the code
were in ASoC, otherwise we're going to get build breakages when APIs
change.
--
To unsubscribe from this list: send the line "unsubscribe linux-tegra" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html