Messages in this thread | | | Date | Thu, 8 Oct 2015 12:11:30 +0900 | From | Jiada Wang <> | Subject | Re: [PATCH 2/2] ASoC: Codec: wm8962: declare ALC Coefficients as 4 separate registers |
| |
Hi
On 10/06/2015 08:01 PM, Mark Brown wrote: > On Tue, Oct 06, 2015 at 04:06:55PM +0900, Jiada Wang wrote: > >> As ALC2 register is volatile, declare it as one of ALC Coefficients >> register together with other non-volatile registers will cause issue, >> in case wm8962 has enter suspend mode, and cache_only flag is set, >> any attempt to read from ALC2 will fail. > >> Instead of declaring one ALC Coefficients register which contains >> ALC1 ~ ALC3 and Noise Gate, this patch declares 4 separate registers, >> so that regmap can handle these registers differently based on their >> classification. > > I don't understand this commit log. Why does regmap care how these > registers are presented to userspace, and how does splitting the > controls up address the problem with one of the registers being volatile? > Surely that register still has the same problem? > .get callback function will call regmap_raw_read() to read register value from these registers, when these 4 regsters are declared as one "ALC coefficient" register, condition check of regmap_volatile_range() will return false, thus regmap will go word by word for the cache from each register of "ALC Coefficient", the failure scenario is, when wm8962 is in suspend mode (cache_only flag is set), as ALC2 doesn't have cached value, then any attempt to read from it fails,
By splitting these registers, regmap can handle ALC2 as a single volatile register, and always read from HW
Thanks, Jiada
| |