On Fri, 17 Jan 2014 17:20:25 +0000, Mark Brown wrote:
it's really common for device vendors to say that their device or code is special in some way and can't do the standard things but this rarely turns out to be true. It is therefore really important to do things like highlight specific technical things that mean new approaches are needed (like the fact that the addresses are independently controllable for the functions on this device) otherwise it is very easy for it to look like a common pattern is being repeated.
Certainly wasn't trying to con anyone here. Do wish I'd brought that feature up a lot earlier though as it would've saved us both time.
What I would suggest doing is writing a binding document for the device (which seems to be missing anyway) and sending that along with patches changing both the MFD and CODEC compatible strings and IDs and adding comments saying that this is intentional to ensure that nobody cleans this up again.
Ok, I'll look into that. I guess it makes sense to write one document for the PMIC and one for the CODEC as well, given their individual nature?