All of lore.kernel.org
 help / color / mirror / Atom feed
* [RFC PATCH 0/8] Removing duplicate video/pad ops
@ 2014-12-04  9:54 Hans Verkuil
  2014-12-04  9:54 ` [RFC PATCH 1/8] v4l2 subdevs: replace get/set_crop by get/set_selection Hans Verkuil
                   ` (8 more replies)
  0 siblings, 9 replies; 28+ messages in thread
From: Hans Verkuil @ 2014-12-04  9:54 UTC (permalink / raw)
  To: linux-media; +Cc: g.liakhovetski, laurent.pinchart, prabhakar.csengg

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


^ permalink raw reply	[flat|nested] 28+ messages in thread

end of thread, other threads:[~2014-12-20 18:56 UTC | newest]

Thread overview: 28+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2014-12-04  9:54 [RFC PATCH 0/8] Removing duplicate video/pad ops Hans Verkuil
2014-12-04  9:54 ` [RFC PATCH 1/8] v4l2 subdevs: replace get/set_crop by get/set_selection Hans Verkuil
2014-12-12 13:24   ` Sakari Ailus
2014-12-16 14:49   ` Prabhakar Lad
2014-12-04  9:54 ` [RFC PATCH 2/8] v4l2-subdev: drop get/set_crop pad ops Hans Verkuil
2014-12-12 13:27   ` Sakari Ailus
2014-12-16 14:52   ` Prabhakar Lad
2014-12-04  9:54 ` [RFC PATCH 3/8] v4l2-subdev: drop unused op enum_mbus_fmt Hans Verkuil
2014-12-07 22:57   ` Laurent Pinchart
2014-12-12 13:28   ` Sakari Ailus
2014-12-16 14:54   ` Prabhakar Lad
2014-12-18 22:08   ` Guennadi Liakhovetski
2014-12-19 11:37     ` Hans Verkuil
2014-12-04  9:54 ` [RFC PATCH 4/8] v4l2-subdev: replace v4l2_subdev_fh by v4l2_subdev_pad_config Hans Verkuil
2014-12-07 23:37   ` Laurent Pinchart
2014-12-04  9:54 ` [RFC PATCH 5/8] media/i2c/Kconfig: drop superfluous MEDIA_CONTROLLER Hans Verkuil
2014-12-07 23:38   ` Laurent Pinchart
2014-12-19 11:44     ` Hans Verkuil
2014-12-19 12:18       ` Laurent Pinchart
2014-12-19 12:21         ` Hans Verkuil
2014-12-20 18:56           ` Laurent Pinchart
2014-12-04  9:54 ` [RFC PATCH 6/8] v4l2-subdev: add v4l2_subdev_create_pad_configs Hans Verkuil
2014-12-07 23:50   ` Laurent Pinchart
2014-12-19 13:28     ` Hans Verkuil
2014-12-04  9:54 ` [RFC PATCH 7/8] v4l2-subdev: remove enum_framesizes/intervals Hans Verkuil
2014-12-04  9:54 ` [RFC PATCH 8/8] v4l2-subdev: remove g/s_crop and cropcap from video ops Hans Verkuil
2014-12-12 13:22   ` Sakari Ailus
2014-12-17  6:45 ` [RFC PATCH 0/8] Removing duplicate video/pad ops Prabhakar Lad

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.