Message ID | 20190419194649.18467-3-pierre-louis.bossart@linux.intel.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
Series | soundwire: fix Kconfig select/depend issues | expand |
On 19-04-19, 14:46, Pierre-Louis Bossart wrote: > The mechanism should be > > config CODEC_XYX_SDW > depends on SOUNDWIRE > select REGMAP_SOUNDWIRE > > config REGMAP_SOUNDWIRE > select SOUNDWIRE_BUS > > SOUNDWIRE_BUS can be independently selected by the SOC driver. The SOC > driver should not know or care about REGMAP_SOUNDWIRE. lgtm, Mark can I have ypu ack so that I can carry the series thru soundwire tree Thanks > Fixes: 7c22ce6e2184 ('03fc8746f7915b5a391d8227f7e1') > Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> > --- > drivers/base/regmap/Kconfig | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/drivers/base/regmap/Kconfig b/drivers/base/regmap/Kconfig > index 6ad5ef48b61e..c12e3f1fb110 100644 > --- a/drivers/base/regmap/Kconfig > +++ b/drivers/base/regmap/Kconfig > @@ -44,7 +44,7 @@ config REGMAP_IRQ > > config REGMAP_SOUNDWIRE > tristate > - depends on SOUNDWIRE_BUS > + select SOUNDWIRE_BUS > > config REGMAP_SCCB > tristate > -- > 2.17.1
On Fri, Apr 19, 2019 at 02:46:49PM -0500, Pierre-Louis Bossart wrote: > config REGMAP_SOUNDWIRE > tristate > - depends on SOUNDWIRE_BUS > + select SOUNDWIRE_BUS This now makes _SOUNDWIRE different to all the other bus types; if this is a good change then surely the same thing should be done for all the other bus types. It's also not clear to me that this actually does anything, do selects from symbols that are themselves selected actually do anything?
On 5/2/19 11:39 PM, Mark Brown wrote: > On Fri, Apr 19, 2019 at 02:46:49PM -0500, Pierre-Louis Bossart wrote: > >> config REGMAP_SOUNDWIRE >> tristate >> - depends on SOUNDWIRE_BUS >> + select SOUNDWIRE_BUS > > This now makes _SOUNDWIRE different to all the other bus types; if this > is a good change then surely the same thing should be done for all the > other bus types. It's also not clear to me that this actually does > anything, do selects from symbols that are themselves selected actually > do anything? yes, this works, but if you prefer alignment I can follow the SLIMBUS model config SND_SOC_WCD9335 tristate "WCD9335 Codec" depends on SLIMBUS select REGMAP_SLIMBUS select REGMAP_IRQ config REGMAP_SLIMBUS tristate depends on SLIMBUS menuconfig SLIMBUS tristate "SLIMbus support" if SLIMBUS # SLIMbus controllers config SLIM_QCOM_CTRL ... As I mentioned it'll compile the bus even if there is no user for it, but it's your call: alignment or optimization.
On 5/5/19 11:40 PM, Mark Brown wrote: > On Fri, May 03, 2019 at 09:32:53AM -0500, Pierre-Louis Bossart wrote: > >> As I mentioned it'll compile the bus even if there is no user for it, but >> it's your call: alignment or optimization. > > You can have both. Alignment is a requirement. If you want to optimize > this then it'd be better to optimize all the bus types rather than just > having the one weird bus type that does something different for no > documented reason. Fine, I'll align if this is the requirement. Thanks for the feedback.
diff --git a/drivers/base/regmap/Kconfig b/drivers/base/regmap/Kconfig index 6ad5ef48b61e..c12e3f1fb110 100644 --- a/drivers/base/regmap/Kconfig +++ b/drivers/base/regmap/Kconfig @@ -44,7 +44,7 @@ config REGMAP_IRQ config REGMAP_SOUNDWIRE tristate - depends on SOUNDWIRE_BUS + select SOUNDWIRE_BUS config REGMAP_SCCB tristate
The mechanism should be config CODEC_XYX_SDW depends on SOUNDWIRE select REGMAP_SOUNDWIRE config REGMAP_SOUNDWIRE select SOUNDWIRE_BUS SOUNDWIRE_BUS can be independently selected by the SOC driver. The SOC driver should not know or care about REGMAP_SOUNDWIRE. Fixes: 7c22ce6e2184 ('03fc8746f7915b5a391d8227f7e1') Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> --- drivers/base/regmap/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)