Re: [PATCH V3 07/12] ata/sata_mv: Remove conditional compilation of clk code

[Russell King - ARM Linux <linux@arm.linux.org.uk>]

On Wed, Apr 25, 2012 at 11:31:09AM +0200, Andrew Lunn wrote:
> > If an arch has HAVE_CLK enabled it must provide valid clocks (be it
> > dummy clocks) for all devices it supports.
> 
> So, lets take the theoretical exaple of a unicore32 PUV3
> 
> config ARCH_PUV3
>         def_bool y
>         select CPU_UCV2
>         select GENERIC_CLOCKEVENTS
>         select HAVE_CLK
>         select ARCH_REQUIRE_GPIOLIB
>         select ARCH_HAS_CPUFREQ
> 
> Seems like this somewhat unknown, to me at least, architecture, also
> supports PCI. So i plug in an HP Adaptec AIC8120 SATA host bus adapter
> into a spare PCI slot. This uses the Marvell 88SX6041, which the
> SATA_MV driver supports.
> 
> Should i expect that the unicore32 PUV3 has created a dummy clk for
> this case?

Are you going to bother answering the detailed email I sent describing
my POV on this case?