linux-next regression on ARM926

[Dave Martin]

On Thu, Jul 21, 2011 at 5:52 PM, Russell King - ARM Linux
<linux at arm.linux.org.uk> wrote:
> On Thu, Jul 21, 2011 at 05:39:06PM +0100, Dave Martin wrote:
>> Did you miss proc-v6, or is that touched by some other patch?  I
>> already have the relevant patch, so I'm happy to post that if you
>> don't already have it.
>
> Well spotted, I've just corrected that.

I think that should be all of them -- I had hacked up a quick script
to find and change all instances, but proc-v6 was the only one it
touched which wasn't in your series.

I also build-tested the change in several configurations just to be
sure.  It looks sound, so far as I can tell.

>
>> My proc-maros.S patch is slightly different in that it avoids defining
>> the default values twice (i.e., once for !CONFIG_PM_SUSPEND and once
>> for suspend=0); but it's otherwise equivalent.
>>
>> I also took the opportunity to remove the definitions of
>> cpu_<name>_suspend_size for the !CONFIG_PM_SUSPEND case, since this
>> helps to ensure they are not used accidentally.  That's a more
>> cosmetic change.
>
> The missing symbols for do_suspend/do_resume should be enough on their
> own to catch people using it, so I don't think that's necessary.

OK, agreed.

>> Do you want me to post my whole series anyway, or otherwise can you
>> point me to where you applied the patches so I can filter my series?
>
> I'll push it out later this evening, it'll be part of the devel-stable
> branch so once pushed it becomes immutable.

in that that case I guess I don't have anything to add, so I'll drop
my patches.  Apologies for the last-minute churn...

Thanks
---Dave