On Mon, 18 May 2009 16:33:18 +0100, Mark Brown broonie@opensource.wolfsonmicro.com wrote:
On Tue, May 19, 2009 at 12:25:42AM +0900, Atsushi Nemoto wrote:
Thank you for elaboration. Then, this untested patch (on top of the last patch) is what you mean? If yes, I will split this into arch part and driver part, and then update both patches.
Partly - I'd also expect to see the resources moved around in the architecture code so that they are on the platform devices that are being used to probe the DAIs (unless that's what the architecture side of your patch was doing, patch wasn't quite giving me enough context to be 100% clear).
Now the resources are provided with "txx9aclc-ac97" platform device which used to probe DAIs. I will send whole patchset again after some testing.
You could also roll the resource requesting and unrequesting into the probe functions for the DAI & DMA platform devices, though that is not essential.
OK, I will move them into txx9aclc_ac97_dev_probe(). Then I can use devres functions to make the resource management much simpler.
--- Atsushi Nemoto