Re: [PATCH 1/2] usb: host: Enable compile testing for some of drivers

[Greg Kroah-Hartman]

On Mon, Dec 30, 2019 at 08:42:49PM +0100, Krzysztof Kozlowski wrote:
> On Mon, Dec 30, 2019 at 08:36:21PM +0100, Greg Kroah-Hartman wrote:
> > On Mon, Dec 30, 2019 at 08:29:56PM +0100, Greg Kroah-Hartman wrote:
> > > On Mon, Dec 30, 2019 at 06:22:14PM +0100, Krzysztof Kozlowski wrote:
> > > > Some of the USB host drivers can be compile tested to increase build
> > > > coverage.  Add 'if' conditional to 'default y' so they will not get
> > > > enabled by default on all other architectures.
> > > > 
> > > > Signed-off-by: Krzysztof Kozlowski <krzk@xxxxxxxxxx>
> > > > ---
> > > >  drivers/usb/host/Kconfig | 54 ++++++++++++++++++++--------------------
> > > >  1 file changed, 27 insertions(+), 27 deletions(-)
> > > 
> > > Nice, thanks for these, they should help out a lot with development.
> > 
> > And now I get this build warning with this patch:
> > 
> > WARNING: unmet direct dependencies detected for PHY_DA8XX_USB
> >   Depends on [n]: ARCH_DAVINCI_DA8XX
> >   Selected by [m]:
> >   - USB_OHCI_HCD_DAVINCI [=m] && USB_SUPPORT [=y] && USB [=m] && (ARCH_DAVINCI_DA8XX || COMPILE_TEST [=y]) && USB_OHCI_HCD [=m]
> > 
> > 
> > Care to provide a fix?
> 
> That's my fault. I was testing entire patchset which includes also other
> trees. I sent patch for PHY here:
> https://lore.kernel.org/lkml/20191230172449.17648-2-krzk@xxxxxxxxxx/T/#u
> 
> Probably they should go together or we could ignore the error as it is
> only for build testing.

If that patchset gets merged, all is fine.  If not, let me know and I
can take it through my tree :)

thanks,

greg k-h