[pvrusb2] Test : Compile with 2.6.14-rc1

[Mike Isely]

On Tue, 20 Sep 2005 xavier.gnata at free.fr wrote:

>
> > I think I know about this problem.
>
> Sounds interesting ;)
>
> > I've been talking to the current official V4L maintainer
>
> who is now ???
>

He's Mauro Carvalho Chehab <mchehab at brturbo.com.br>.


>  over the past
> > week or so, and just last Saturday I pointed out to him about the fact
> > that the tveeprom code doesn't provide a proper way to retrieve the device
> > serial number (it's in the eeprom and it is being parsed out but there's
> > no official API to get at it).  The pvrusb2 driver has to get this by
> > calling an "internal" function.  He then mentioned that the eeprom parsing
> > code has been considerably updated in 2.6.14 - which is another way to say
> > that it has been changed.
>
> outch...Sounds like quite a lot of changes to do in the driver :(

I suspect the changes might be easy; the problem is backwards
compatibility with tveeprom.c that is still bundled with the driver and
possibly also with whatever the current ivtv is...

I still need to research this.  Actually I still need to pull down the
latest release candidate and at least have a look.

>
> > I need to examine this and determine the nature of the change.  It will
> > probably end up being incompatible with the included tveeprom.c source in
> > the pvrusb2 driver, which means I have to decide how to handle this
> > situation...
>
> Ok so this time the fix is not obvious :(.
> Is is possbile to have a backward compatible fix??

Not sure yet.  See above.


> I do not like to have to add more than 2 or 3 different if(kernel_version > foo)
> statements in a code driver :(

No argument there.

  -Mike

-- 
                        |         Mike Isely          |     PGP fingerprint
     Spammers Die!!     |                             | 03 54 43 4D 75 E5 CC 92
                        |   isely @ pobox (dot) com   | 71 16 01 E2 B5 F5 C1 E8
                        |                             |