Re: [RFC PATCH 0/8] Removing duplicate video/pad ops

2014-12-16 Thread Prabhakar Lad
On Thu, Dec 4, 2014 at 3:24 PM, Hans Verkuil  wrote:
> This patch series attempts to remove some of the duplicate video/pad ops.
> The first two patches have been posted before. The only thing changed is
> that the subdevs no longer add checks for pad values != 0 as suggested.
>
> The third patch removes an unused subdev op. Somehow we must have missed
> that one.
>
> The fourth replaces v4l2_subdev_fh by v4l2_subdev_pad_config. No other
> changes other than that. This patch paves the way for bridge drivers to
> call pad ops that need v4l2_subdev_pad_config since bridge drivers do
> not have a v4l2_subdev_fh struct.
>
> The fifth patch is a small Kconfig cleanup.
>
> The sixth patch causes v4l2_device_register_subdev() to initialize a
> v4l2_subdev_pad_config array for use in bridge drivers. Bridge drivers
> can then use sd->pad_configs as argument to the pad ops. It's done
> behind the scenes, so this requires no driver changes.
>
> One disadvantage, though: bridge drivers that never call those pad ops
> since they expect that userspace will call them are now allocating memory
> for this when they will never need it.
>
> Should I add a V4L2_FL_DONT_CREATE_PAD_CONFIGS flag that such drivers
> can set, prohibiting allocating this memory?
>
> Is the code I use to initialize the pad_configs correct? Is it something
> that subdev_fh_init() in v4l2-subdev.c should do as well?
>
> The seventh patch removes the video enum_framesizes/intervals ops.
>
> The last patch removes the video g/s_crop and cropcap ops. This especially
> affects soc_camera. Note that I only tested if it compiles, I have not
> tried this on my soc_camera board. I will try to get my renesas board up
> and running with this kernel to test it but more help with this would
> be much appreciated.
>
FYI, I am on a holiday and out of reach of the hardware, will test it once I get
from holiday after Christmas.

Thanks,
--Prabhakar Lad
--
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


[RFC PATCH 0/8] Removing duplicate video/pad ops

2014-12-04 Thread Hans Verkuil
This patch series attempts to remove some of the duplicate video/pad ops.
The first two patches have been posted before. The only thing changed is
that the subdevs no longer add checks for pad values != 0 as suggested.

The third patch removes an unused subdev op. Somehow we must have missed
that one.

The fourth replaces v4l2_subdev_fh by v4l2_subdev_pad_config. No other
changes other than that. This patch paves the way for bridge drivers to
call pad ops that need v4l2_subdev_pad_config since bridge drivers do
not have a v4l2_subdev_fh struct.

The fifth patch is a small Kconfig cleanup.

The sixth patch causes v4l2_device_register_subdev() to initialize a
v4l2_subdev_pad_config array for use in bridge drivers. Bridge drivers
can then use sd->pad_configs as argument to the pad ops. It's done
behind the scenes, so this requires no driver changes.

One disadvantage, though: bridge drivers that never call those pad ops
since they expect that userspace will call them are now allocating memory
for this when they will never need it.

Should I add a V4L2_FL_DONT_CREATE_PAD_CONFIGS flag that such drivers
can set, prohibiting allocating this memory?

Is the code I use to initialize the pad_configs correct? Is it something
that subdev_fh_init() in v4l2-subdev.c should do as well?

The seventh patch removes the video enum_framesizes/intervals ops.

The last patch removes the video g/s_crop and cropcap ops. This especially
affects soc_camera. Note that I only tested if it compiles, I have not
tried this on my soc_camera board. I will try to get my renesas board up
and running with this kernel to test it but more help with this would
be much appreciated.

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