On Thu, Feb 23, 2017 at 06:58:12PM +0200, Andy Shevchenko wrote: > On Thu, Feb 23, 2017 at 6:47 PM, Lars-Peter Clausen <lars@xxxxxxxxxx> wrote: > > On 02/23/2017 05:43 PM, Andy Shevchenko wrote: > >> On Wed, Feb 22, 2017 at 12:23 PM, Eva Rachel Retuya <eraretuya@xxxxxxxxx> wrote: > >>> Add SPI driver that initializes SPI regmap for the adxl345 core driver. > >>> The driver supports the same functionality as I2C namely the x, y, z and > >>> scale readings. > > >>> config ADXL345_I2C > >>> tristate > >>> select REGMAP_I2C > >>> > >>> +config ADXL345_SPI > >>> + tristate > >>> + select REGMAP_SPI > >> > >> Hmm... > >> I saw another pattern > >> > >> Library / core part is non-visible to user, while > >> SPI and I2C parts are selectable by user. > >> > >> Why do you use inverted pattern? What did I miss? > > > > The first version of the patch used the other pattern SPI/I2C visible. > > Jonathan suggested this other pattern. I prefer the explicit SPI/I2C visible > > pattern, but in the end it doesn't really matter as long as both work. > > Yes, but this pattern makes extra footprint of the kernel and > basically dead code when I would like, for example, to have SPI bus > enabled, I2C module available, but SPI module not compiled. > > Other one is when I want to have one compiled in, one as a module by > whatever reason. > > At the end I have no strong opinion, though rationale for the opposite is above. > Hello Lars and Andy, I'll revert to the explicit SPI/I2C pattern in order to give more freedom in configuring as per the scenarios previously stated. Thanks, Eva > -- > With Best Regards, > Andy Shevchenko -- To unsubscribe from this list: send the line "unsubscribe devicetree" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
