[PATCH] ptp: PTP_1588_CLOCK_PCH depends on x86

[PATCH] ptp: PTP_1588_CLOCK_PCH depends on x86

[Jeff Mahoney]

The EG20T PCH is only compatible with Intel Atom processors so it
should depend on x86.

Cc: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
---
 drivers/ptp/Kconfig |    1 +
 1 file changed, 1 insertion(+)

--- a/drivers/ptp/Kconfig
+++ b/drivers/ptp/Kconfig
@@ -72,6 +72,7 @@ config DP83640_PHY
 
 config PTP_1588_CLOCK_PCH
 	tristate "Intel PCH EG20T as PTP clock"
+	depends on X86
 	select PTP_1588_CLOCK
 	help
 	  This driver adds support for using the PCH EG20T as a PTP

-- 
Jeff Mahoney
SUSE Labs

Re: [PATCH] ptp: PTP_1588_CLOCK_PCH depends on x86

[Richard Cochran]

On Wed, Jan 30, 2013 at 12:31:01PM -0500, Jeff Mahoney wrote:
> The EG20T PCH is only compatible with Intel Atom processors so it
> should depend on x86.

Yes, there is something wrong with PTP_1588_CLOCK_PCH. The last
several times I did 'make oldconfig' for various configs, it asked me
whether to enable this 'new' option. That is really annoying,
especially with non-atom and non-x86 builds.

Ben, you removed the PCH_GBE dependency in 18d359ce. Are you sure that
was the right thing to do?

Thanks,
Richard

 
> Cc: Richard Cochran <richardcochran@gmail.com>
> Signed-off-by: Jeff Mahoney <jeffm@suse.com>
> ---
>  drivers/ptp/Kconfig |    1 +
>  1 file changed, 1 insertion(+)
> 
> --- a/drivers/ptp/Kconfig
> +++ b/drivers/ptp/Kconfig
> @@ -72,6 +72,7 @@ config DP83640_PHY
>  
>  config PTP_1588_CLOCK_PCH
>  	tristate "Intel PCH EG20T as PTP clock"
> +	depends on X86
>  	select PTP_1588_CLOCK
>  	help
>  	  This driver adds support for using the PCH EG20T as a PTP
> 
> -- 
> Jeff Mahoney
> SUSE Labs

Re: [PATCH] ptp: PTP_1588_CLOCK_PCH depends on x86

[Ben Hutchings]

On Wed, 2013-01-30 at 20:02 +0100, Richard Cochran wrote:
> On Wed, Jan 30, 2013 at 12:31:01PM -0500, Jeff Mahoney wrote:
> > The EG20T PCH is only compatible with Intel Atom processors so it
> > should depend on x86.
> 
> Yes, there is something wrong with PTP_1588_CLOCK_PCH. The last
> several times I did 'make oldconfig' for various configs, it asked me
> whether to enable this 'new' option. That is really annoying,
> especially with non-atom and non-x86 builds.
> 
> Ben, you removed the PCH_GBE dependency in 18d359ce. Are you sure that
> was the right thing to do?

The clock is a separate device and the PHC driver doesn't depend on the
net driver.  However, since the clock is a PCI device the driver does
need a dependency on PCI (which was previously an indirect dependency).

> > Cc: Richard Cochran <richardcochran@gmail.com>
> > Signed-off-by: Jeff Mahoney <jeffm@suse.com>
> > ---
> >  drivers/ptp/Kconfig |    1 +
> >  1 file changed, 1 insertion(+)
> > 
> > --- a/drivers/ptp/Kconfig
> > +++ b/drivers/ptp/Kconfig
> > @@ -72,6 +72,7 @@ config DP83640_PHY
> >  
> >  config PTP_1588_CLOCK_PCH
> >       tristate "Intel PCH EG20T as PTP clock"
> > +     depends on X86
> >       select PTP_1588_CLOCK
> >       help
> >         This driver adds support for using the PCH EG20T as a PTP

I agree that the driver is only useful on X86, but some people like to
build-test as many drivers as possible with a single configuration.  The
only build dependency for this seems to be PCI.

(Putting on my distribution hat for a moment: it is useful to know that
a driver is practically architecture-specific even though it can be
built elsewhere.  Perhaps there is a use for a BUILD_TEST or BUILD_ALL
config symbol, so this could be:
	depends on X86 || (BUILD_TEST && PCI)
)

Ben.

-- 
Ben Hutchings, Staff Engineer, Solarflare
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.