On Fri, Dec 12, 2014 at 03:48:51PM -0800, Ben Zhang wrote: > On Fri, Dec 12, 2014 at 5:27 AM, Mark Brown <broo...@kernel.org> wrote: > > On Wed, Dec 10, 2014 at 08:15:26PM -0800, Ben Zhang wrote:
> > Why is this being specified as some magic number rather than using the > > voltage (or at least providing defines for the voltage) - this is going > > to do little to make the DT legible and... > >> +enum rt5677_micbias { > >> + RT5677_MICBIAS_1_476V = 0, > >> + RT5677_MICBIAS_2_970V = 1, > >> + RT5677_MICBIAS_1_242V = 2, > >> + RT5677_MICBIAS_2_475V = 3, > >> +}; > > ...I see there are defined for platform data. > This patch adds both an entry to the platform data and a DT binding > for MICBIAS level selection. The 4 voltage options > (1.476V/2.970V/1.242V/2.475V) are the only ones supported by the codec > hardware, so it seems an enum is better than specifying the exact It's not that clear that an enum *is* better, and a magic numbers enum is definitely worse for anyone who has to read the resulting DT. > voltage directly. I was following the two examples below: > Documentation/devicetree/bindings/sound/cs42l52.txt (cirrus,micbias-lvl) > Documentation/devicetree/bindings/sound/tlv320aic3x.txt (ai3x-micbias-vg) Old DT bindings are old and often not best practice. > I'm new to devicetree bindings. Is there something like an enum in DT? include/dt-bindings
signature.asc
Description: Digital signature