Re: [hg:v4l-dvb] backport I2C changes on several files

[Jean Delvare <khali@xxxxxxxxxxxx>]

On Thu, 5 Jun 2008 15:34:25 +0200 (CEST), Hans Verkuil wrote:
> > Hi Hans,
> >
> > I'm backporting some commits that happened at Linus tree.
> >
> > Most of the changes are trivial. However, there were some changes at
> > saa7115
> > code that might affect some drivers. In particular, I didn't backported
> > this change:
> >
> > diff -upr oldtree/drivers/media/video/ivtv/ivtv-i2c.c
> > git/drivers/media/video/ivtv/ivtv-i2c.c
> > --- oldtree/drivers/media/video/ivtv/ivtv-i2c.c	2008-06-05
> > 10:14:42.000000000 -0300
> > +++ git/drivers/media/video/ivtv/ivtv-i2c.c	2008-05-26 12:29:17.000000000
> > -0300
> > @@ -145,7 +145,7 @@ static const char * const hw_devicenames
> >  	"wm8775",
> >  	"cs53l32a",
> >  	"tveeprom",
> > -	"saa7115",
> > +	"saa7114",
> >  	"upd64031a",
> >  	"upd64083",
> >  	"saa717x",
> >
> > I suspect that this will likely break ivtv support for boards based on
> > saa711x,
> > since it will load the legacy saa7114 module (btw, is there any driver yet
> > using saa7114? we should remove it ;) ).
> 
> Please backport this change! ivtv still loads the saa7115 module, but this
> change tells the saa7115 driver that ivtv expects a saa7114 chip instead
> of an saa7115.
> 
> I've helped khali test these changes, so I know it's OK.
> 
> Without this change ivtv-boards that have an saa7114 chip would fail to work.

If I remember the way I wrote the code correctly, it won't fail, it
will only display a warning in the logs. But still, you're right, this
needs to be backported with the rest, to get rid of this warning.

Mauro, the trick is that the saa7114 driver is not (and will never be)
a new-style i2c driver, so it can't pick new-style i2c devices as
instantiated by the ivtv driver.

-- 
Jean Delvare

_______________________________________________
v4l-dvb-maintainer mailing list
v4l-dvb-maintainer@xxxxxxxxxxx
http://www.linuxtv.org/cgi-bin/mailman/listinfo/v4l-dvb-maintainer