On Fri, Nov 20, 2020 at 11:54 AM Lars-Peter Clausen <lars@xxxxxxxxxx> wrote: > > On 11/20/20 8:17 PM, Matt Ranostay wrote: > > On Fri, Nov 20, 2020 at 5:15 AM rishi gupta <gupt21@xxxxxxxxx> wrote: > >> On Fri, Nov 20, 2020 at 8:31 AM Matt Ranostay > >> <matt.ranostay@xxxxxxxxxxxx> wrote: > >>> Add support for the three 10-bit ADC channels registered via > >>> the IIO subsystem. > >>> > >>> Cc: linux-input@xxxxxxxxxxxxxxx > >>> Cc: linux-iio@xxxxxxxxxxxxxxx > >>> CC: Rishi Gupta <gupt21@xxxxxxxxx> > >>> Signed-off-by: Matt Ranostay <matt.ranostay@xxxxxxxxxxxx> > >>> --- > >>> drivers/hid/Kconfig | 1 + > >>> drivers/hid/hid-mcp2221.c | 65 ++++++++++++++++++++++++++++++++++++++- > >>> 2 files changed, 65 insertions(+), 1 deletion(-) > >>> > >>> diff --git a/drivers/hid/Kconfig b/drivers/hid/Kconfig > >>> index 05315b434276..4795744d9979 100644 > >>> --- a/drivers/hid/Kconfig > >>> +++ b/drivers/hid/Kconfig > >>> @@ -1157,6 +1157,7 @@ config HID_MCP2221 > >>> tristate "Microchip MCP2221 HID USB-to-I2C/SMbus host support" > >>> depends on USB_HID && I2C > >>> depends on GPIOLIB > >>> + depends on IIO > >> I am wondering what will happen on systems which do not enable IIO. > >> This driver can not be used there. > >> Is my understanding correct? > > Actually yeah this should be "select IIO" to avoid that issue. > > No, we should not have a individual driver select a framework. This will > cause all kinds of issues with reverse dependencies. > > It might be worth splitting this driver into a MFD driver, then the MFD > cells could have their own module that depend on the subsystem and if > not enabled the functionality will not be provided. Would it make sense to use IS_REACHABLE(CONFIG_IIO) for the iio blocks? Guessing the weak reference "imply IIO" would still be bad for the driver selecting a framework? Thanks, Matt > > - Lars >
