On Sun, 15 Mar 2009, Andy Walls wrote: > On Sun, 2009-03-15 at 10:28 -0700, Trent Piepho wrote: > > > Why are the i2c addresses from various i2c chips moved into the bttv > > driver? Doesn't it make more sense that the addresses for chip X should be > > in the driver for chip X? > > One reason that this may be undesirable is that the devices can be set > to slightly different addresses via external straps (probably a corner > case, I know). The bridge driver has the best chance of knowing what > chips are where with certainty. If one knows that some card has a certain chip at a certain address, then it makes the most sense to store that with the rest of the card's data, i.e. in the bridge driver. But the bttv code looks like (I don't know what Hans intended, since he didn't feel anyone else needed to know what, why, or how his code does whatever it is that it does) it is probing all known address that a given chip could be at to see if the chip is there. In that case what's really being used is an address list from the I2C chip's datasheet and any bridge driver that wants to probe like that will use the same list, which IMHO, makes more sense to put in the driver of the I2C chip rather than every driver that uses the I2C chip. -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
