On Thu, Sep 9, 2010 at 10:22 PM, Mark Brown broonie@opensource.wolfsonmicro.com wrote:
On Thu, Sep 09, 2010 at 09:56:23PM +0900, Jassi Brar wrote:
I would like to have points to consider from every stakeholder.
I don't really have anything myself, though I may have some comments on the actual proposal I guess.
Btw, AC97, PCM and DMA were easy kill.
It's I2S that's PITA. I am thinking of leaving 24xx-i2s, i.e pre-64xx(shared with 2412), unchanged and letting them die of old age (obsoleted and dropped duly if no stakeholder keeps up).
I plan to categorize the newer I2S blocks into i2s_v2(s3c2412), i2s_v3(v2 with minor variation in reg field order), i2s_v4 (v3 plus 5.1 channel support) and v5 (v4 plus Secondary Stereo-Channel for h/w mixing support). A new flag 'version' in 'struct s3c_audio_pdata' passed via platform_data can be used to differentiate quirks and features run-time in a single CPU driver.
Ben, Mark and other interested parties:- Any opinions, before I have spent too many hours working on it?
Thanks.