[PATCH] ARM: clean-up common multi-platform kconfig options

[Rob Herring]

On Sat, Dec 7, 2013 at 12:10 PM, Russell King - ARM Linux
<linux at arm.linux.org.uk> wrote:
> On Fri, Dec 06, 2013 at 02:01:51PM -0600, Rob Herring wrote:
>> On Fri, Dec 6, 2013 at 10:41 AM, Arnd Bergmann <arnd at arndb.de> wrote:
>> > This prevents you from building a plain v6 SMP kernel, and we've
>> > had a variation of that since CONFIG_SMP was first marked non-BROKEN.
>>
>> I'm saying a v6k enabled kernel would not work on pure v6 h/w. But
>> since all implementations appear to support doubleword exclusives and
>> there may not be any pure v6 only h/w, then we may never have a
>> problem with the code below.
>
> I wonder why we have the SMP_ON_UP option.  It's there so that we can
> do exactly this - build a SMP kernel (so, one for V6K or later) _and_
> have it run on non-SMP capable V6 platforms.

Exactly, but __kuser_cmpxchg64 is missing the necessary SMP_ON_UP
patching or correct ifdef logic. I have yet to get my head around what
is the right fix here.

> Now, while it's true that ARMv6 doesn't have the WFI instruction, and
> that was introduced in ARMv6K, when we include support for ARMv6, we
> don't use that instruction - we use its MCR equivalent.
>
> Remember, we've specifically put work into the kernel so that ARMv6
> can operate as part of the single zImage kernel precisely because we
> have mixtures of SoCs which have this spread of architectures.
>
> So, as far as I'm aware, today, a kernel which has V6, V6K and V7
> will work across all those CPUs.
>
> What's rather annoying in this thread is that you and Arnd are running
> around seemingly making decisions on this without bothering to find out
> the details, in persuit of endless cleanups.  Where is this heading?
> Yet more fscking breakage of the ARM kernel, leading to platforms which
> won't boot.  I'm getting rather sick and tired of cleaning up after
> this kind of activity, so I'm going to tell you right now to stop this.
>
> I've had to push a whole load of footbridge patches into stable kernels
> during the last fortnight because of this.  For instance, Rob, your
> changes to the way that VGA was supported on ARM pretty much broke the
> whole thing leading to an oops on boot at 0xb80000.
>
> Please stop this madness.

Highlighting the above issue and selecting the correct cpu option for
various platforms is hardly madness. This doesn't affect just v6
platforms, but also all v7 platforms in v6+v7 builds since less
optimal non-v6K code paths are selected.

Rob