On Wed, Oct 01, 2008 at 05:54:21PM +0100, Andy Green wrote: > It's simply that I can take the patchset and move on. I tried removing > 5/11 but at least 7/11 depends on it and I don't have time to unpick the > intentions.
7/11 is part of the I2C API conversion - patches 5-9 are all part of that. The net effect is to move the registration of the I2C devices for the codec and (on GTA01) external amp to use the new I2C registration API and out of the ASoC drivers and into the arch/arm stuff which is the way this is supposed to be done. The new API makes things much cleaner, especially with multiple I2C buses, and lets you do things like pass platform data to drivers so it's a substantial win all round. It will be a requirement in mainline at some point very soon which is why Jean kindly volunteered to do the work). The current mainline patches only go half way to what Jonas has done ATM - they still do the registration of the I2C devices in the ASoC code. This is partly due to Jean making minimally intrusive changes and partly due to the fact that ASoC core isn't actually able to cope with the new model yet (there is code pending to let it cope but it's not merged). While it will almost certainly work to go the whole way with the current core it was safer not to just in case it tripped up any platforms.
