[deliverable/linux.git] / drivers / staging / media / bcm2048 / TODO

TODO:

From the initial code review:

The main thing you need to do is to implement all the controls using the
control framework (see Documentation/video4linux/v4l2-controls.txt).
Most drivers are by now converted to the control framework, so you will
find many examples of how to do this in drivers/media/radio.

The sysfs stuff should be replaced by controls as well. A lot of the RDS
support is now available as controls (although there may well be some
missing features, but that is easy enough to add). Since the RDS data is
actually read() from the device I am not sure whether the RDS
properties/controls should be there at all.

Correct Coding Style, as this driver also violates several Style
rules, and do evil tricks, like returning from a function inside a
macro.

Finally this driver should probably be split up into two parts: one
v4l2_subdev-based core driver and one platform driver. See e.g.
radio-si4713/si4713-i2c.c as a good example. But I would wait with that
until the rest of the driver is cleaned up. Then I have a better idea of
whether this is necessary or not.
Commit	Line	Data
d44250fd HV	1	TODO:
	2
	3	From the initial code review:
	4
	5	The main thing you need to do is to implement all the controls using the
	6	control framework (see Documentation/video4linux/v4l2-controls.txt).
	7	Most drivers are by now converted to the control framework, so you will
	8	find many examples of how to do this in drivers/media/radio.
	9
	10	The sysfs stuff should be replaced by controls as well. A lot of the RDS
	11	support is now available as controls (although there may well be some
	12	missing features, but that is easy enough to add). Since the RDS data is
	13	actually read() from the device I am not sure whether the RDS
	14	properties/controls should be there at all.
	15
	16	Correct Coding Style, as this driver also violates several Style
	17	rules, and do evil tricks, like returning from a function inside a
	18	macro.
	19
	20	Finally this driver should probably be split up into two parts: one
	21	v4l2_subdev-based core driver and one platform driver. See e.g.
	22	radio-si4713/si4713-i2c.c as a good example. But I would wait with that
	23	until the rest of the driver is cleaned up. Then I have a better idea of
	24	whether this is necessary or not.