Re: [PATCH 6/6] spi: octeon: Add thunderx driver

[David Daney]

On 07/25/2016 09:16 AM, Mark Brown wrote:
> On Mon, Jul 25, 2016 at 05:51:22PM +0200, Jan Glauber wrote:
> > On Sun, Jul 24, 2016 at 10:04:52PM +0100, Mark Brown wrote:
> > > On Sat, Jul 23, 2016 at 12:42:55PM +0200, Jan Glauber wrote:
>
> > > > +	depends on 64BIT && PCI && !CAVIUM_OCTEON_SOC
>
> > > This is a *weird* and most likely broken set of dependencies - why
> > > exclude this if we're on Octeon (or Octeon happens to have been enabled
> > > in a config)?
>
> > I agree that it looks weird, the reasoning is that we would like
> > to avoid making the driver depend on something like ARCH_THUNDER.
>
> Why?
>
> > So I made the driver depend on the things it actually uses
> > (PCI for probing and 64BIT because of readq/writeq) and don't care if it
> > compiles on other platforms too (like x86).
>
> The usual pattern would be something like (ARCH_THUNDER || COMPILE_TEST)
> && PCI && 64BIT (so that people on other platforms where the device will
> never actually appear don't get bothered by the prompt).

ARCH_THUNDER needs to die, so perhaps it should be (ARM64 || 
COMPILE_TEST) && PCI && 64BIT if you really want to hide it from 
non-arm64 kernel configs.


> > That said, I can remove the !CAVIUM_OCTEON_SOC, it compiles without
> > errors on MIPS too. Would that be ok?
>
> Sure.