Re: [PATCH 4/6] ASoC: OMAP: Add CPU DAI driver for HDMI

[Liam Girdwood <lrg@ti.com>]

On 08/05/11 11:34, Mark Brown wrote:
> On Sat, May 07, 2011 at 07:05:17PM -0500, Ricardo Neri wrote:
>> Am Freitag, den 06.05.2011, 08:21 -0500 schrieb Mark Brown:
> 
>>> Why do you need to check at all?  The playback won't work when the
>>> output isn't connected but the application layer will find that out when
>>> there's no data transfer anyway.  If you implement the check in startup
>>> I'd imagine there may be some races with hotplug of HDMI cables.
> 
>> I need to check somewhere if the Display Subsystem (and the HDMI IP) is
>> on. Otherwise I will have L3 interconnect errors when trying to access
>> the DSS while it is power-disconnected. I need the HDMI IP on for
>> getting configurations (e.g, deep color mode) and use the HDMI audio
>> FIFO.
> 
> I repeat my question - why do you need to check this?  You'll get errors
> but you're generating an error anyway so it's not clear what the urgency
> is.
> 

I think the problem here is that the errors are fatal data aborts here when L3 DSS access is disabled.

Liam