Re: [RFCv2 PATCH 00/14] davinci: clean up input/output/subdev config

2012-09-22 Thread Prabhakar Lad
Hi Hans,

Thanks for the patchset.

On Thu, Sep 20, 2012 at 5:36 PM, Hans Verkuil hverk...@xs4all.nl wrote:
 Hi Prabhakar,

 This is the second patch series for a vpif driver cleanup.

 The first version can be found here:

 http://www.mail-archive.com/linux-media@vger.kernel.org/msg52136.html

 Changes since RFCv1:

 - rebased to a newer git repo:
   
 http://git.linuxtv.org/mhadli/v4l-dvb-davinci_devices.git/shortlog/refs/heads/da850_vpif_machine

 - fixed probe() cleanup code in both display and capture that was seriously
   broken.

 - fixed a broken s_routing implementation in the tvp514x driver: if there
   is no incoming video signal, then s_routing would return EINVAL and
   leave the driver with an inconsistent internal state. This has always
   been a problem, but with this patch series it suddenly became really
   noticable. s_routing shouldn't try to wait for a valid signal, that's
   not what s_routing should do.

 This patch series does some driver cleanup and reorganizes the config
 structs that are used to set up subdevices.

 The current driver associates an input or output with a subdev, but multiple
 inputs may use the same subdev and some inputs may not use a subdev at all
 (this is the case for our hardware).

 Several other things were also configured in the wrong structure. For
 example the vpif_interface struct is really part of the channel config
 and has nothing to do with the subdev.


Acked-by: Prabhakar Lad prabhakar@ti.com
Tested-by: Prabhakar Lad prabhakar@ti.com

 What is missing here is that the output doesn't have the same flexibility
 as the input when it comes to configuration. It would be good if someone
 can pick this up as a follow-up since it's unlikely I'll be working on
 that.

Working on it.

Regards,
--Prabhakar Lad

 What would also be nice is that by leaving the inputs or outputs for the
 second channel empty (NULL) in the configuration you can disable the second
 video node, e.g. trying to use it will always result in ENODEV or something.

 This patch series will at least make things more flexible.

 Regards,

 Hans

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


[RFCv2 PATCH 00/14] davinci: clean up input/output/subdev config

2012-09-20 Thread Hans Verkuil
Hi Prabhakar,

This is the second patch series for a vpif driver cleanup.

The first version can be found here:

http://www.mail-archive.com/linux-media@vger.kernel.org/msg52136.html

Changes since RFCv1:

- rebased to a newer git repo:
  
http://git.linuxtv.org/mhadli/v4l-dvb-davinci_devices.git/shortlog/refs/heads/da850_vpif_machine

- fixed probe() cleanup code in both display and capture that was seriously
  broken.

- fixed a broken s_routing implementation in the tvp514x driver: if there
  is no incoming video signal, then s_routing would return EINVAL and
  leave the driver with an inconsistent internal state. This has always
  been a problem, but with this patch series it suddenly became really
  noticable. s_routing shouldn't try to wait for a valid signal, that's
  not what s_routing should do.

This patch series does some driver cleanup and reorganizes the config
structs that are used to set up subdevices.

The current driver associates an input or output with a subdev, but multiple
inputs may use the same subdev and some inputs may not use a subdev at all
(this is the case for our hardware).

Several other things were also configured in the wrong structure. For
example the vpif_interface struct is really part of the channel config
and has nothing to do with the subdev.

What is missing here is that the output doesn't have the same flexibility
as the input when it comes to configuration. It would be good if someone
can pick this up as a follow-up since it's unlikely I'll be working on
that.

What would also be nice is that by leaving the inputs or outputs for the
second channel empty (NULL) in the configuration you can disable the second
video node, e.g. trying to use it will always result in ENODEV or something.

This patch series will at least make things more flexible.

Regards,

Hans

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