Re: [PATCH for 3.14.y stable 47/47] arm/arm64: KVM: Keep elrsr/aisr in sync with software model

[Christoffer Dall <christoffer.dall@linaro.org>]

Hi Greg,

On Mon, May 11, 2015 at 08:17:07AM -0700, Greg KH wrote:
> On Mon, May 04, 2015 at 09:52:42AM +0800, shannon.zhao@linaro.org wrote:
> > From: Christoffer Dall <christoffer.dall@linaro.org>
> > 
> > commit ae705930fca6322600690df9dc1c7d0516145a93 upstream.
> 
> No, that's not what the patch below really is.
> 
> Do I have to go back by hand and verify each one of these really is the
> patch you say it is?  That's a major pain...
> 
This is a backport of the referenced commit, but it couldn't be applied
directly because of the churn in the vgic code.

I believed that the commit X upstream notation would indicate the
equivalent fix upstream, not the *exact* commit for the relevant stable
kernel.

Apologies if that was an incorrect assumption.  I believe this is the
only patch which was significantly rewritten because of the churn in the
vgic code and enough users are seeing this in various distro kernels
that I figured it was important to refactor and backport.

What would you like me to do with this patch?

Note that I didn't understand that this is wrong from reading
Documentation/stable_kernel_rules.txt, which may just be because I'm
being stupid.  Is the procedure that we're violating documented
somewhere?

Thanks,
-Christoffer