Re: [PATCH 3/6] Documentation/ioctl/ioctl-number.txt: Remove some now freed ioctl ranges

[Mauro Carvalho Chehab <mchehab@redhat.com>]

Em 27-12-2010 11:23, Hans Verkuil escreveu:
> On Monday, December 27, 2010 14:03:03 Mauro Carvalho Chehab wrote:
>> Em 27-12-2010 10:01, Hans Verkuil escreveu:
>>> On Monday, December 27, 2010 12:38:39 Mauro Carvalho Chehab wrote:
>>>> The V4L1 removal patches removed a few ioctls. Update it at the docspace.
>>>>
>>>> Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
>>>>
>>>> diff --git a/Documentation/ioctl/ioctl-number.txt b/Documentation/ioctl/ioctl-number.txt
>>>> index 63ffd78..49d7f00 100644
>>>> --- a/Documentation/ioctl/ioctl-number.txt
>>>> +++ b/Documentation/ioctl/ioctl-number.txt
>>>> @@ -260,14 +260,11 @@ Code  Seq#(hex)	Include File		Comments
>>>>  't'	80-8F	linux/isdn_ppp.h
>>>>  't'	90	linux/toshiba.h
>>>>  'u'	00-1F	linux/smb_fs.h		gone
>>>> -'v'	all	linux/videodev.h	conflict!
>>>>  'v'	00-1F	linux/ext2_fs.h		conflict!
>>>>  'v'	00-1F	linux/fs.h		conflict!
>>>>  'v'	00-0F	linux/sonypi.h		conflict!
>>>> -'v'	C0-CF	drivers/media/video/ov511.h	conflict!
>>>>  'v'	C0-DF	media/pwc-ioctl.h	conflict!
>>>>  'v'	C0-FF	linux/meye.h		conflict!
>>>> -'v'	C0-CF	drivers/media/video/zoran/zoran.h	conflict!
>>>>  'v'	D0-DF	drivers/media/video/cpia2/cpia2dev.h	conflict!
>>>>  'w'	all				CERN SCI driver
>>>>  'y'	00-1F				packet based user level communications
>>>>
>>>
>>> There is also a line for media/ovcamchip.h in this file that can be removed.
>>
>> Ok, I'll do that.
>>
>>> The media/rds.h line can also be removed (this is kernel internal only).
>>
>> There are two rds.h, related to V4L:
>> ./include/linux/rds.h
> 
> Not related to V4L, this is something from Oracle. It is this header that is public,
> not the media/rds.h header.

Ah, ok.

>> ./include/media/rds.h
>>
>> One of them is at the public api:
>>
>> include/linux/Kbuild:header-y += rds.h
>>
>> Btw, that's weird:
>>
>> $ git grep RDS_CMD_OPEN
>> drivers/media/video/saa6588.c:    case RDS_CMD_OPEN:
>> include/media/rds.h:#define RDS_CMD_OPEN  _IOW('R',1,int)
>>
>> as saa6588 is a subdev.
>>
>> IMO, we should remove or rename the internal header first.
> 
> media/rds.h should be renamed to media/saa6588.h. It is also included in
> drivers/media/radio/si470x/radio-si470x.h, but that's obsolete and can be
> removed.

The rds file were the old RDS API, before we add it at V4L2. We should, instead,
convert saa6588 to use the new way, and remove the legacy stuff.

>>> Ditto for media/bt819.h.
>>
>> There are also some issues there related to videodev2 stuff.
>>
>> I prefer to apply the path as-is (just removing the ovcamchip.h) and,
>> on some later cleanup, check and fix the remaining stuff.
> 
> I can make a patch fixing the rds.h header usage. It's all internal stuff
> and the weird naming is just historical and should be changed.

It would be nice if you can do it.

>>>
>>> All other patches in this series:
>>>
>>> Reviewed-by: Hans Verkuil <hverkuil@xs4all.nl>
>>
>> Thanks!
>>>
>>> BTW, it is probably also a good idea to move the dabusb driver to staging and
>>> mark it for removal in 2.6.39.
>>
>> Not sure about that. I don't see any good reason to remove dabusb driver, as
>> nobody reported that it is broken.
> 
> Nobody has the hardware :-)

This is too strong :) Are you absolutely sure that there's absolutely nobody in
the World with that hardware? ;)
> 
> I know you have asked the authors about a possible removal of this driver a few
> months ago. Did you get any reply from them?

Nope.

> It seems to be a demonstration driver only and I've never seen anyone with the
> hardware.

It seems so, but I can't see any technical reason for its removal. The BKL fix were
applied on it, as someone wrote a patch for it.

Cheers,
Mauro