backport "powerpc/powernv: Restore non-volatile CRs after nap" to stable

backport "powerpc/powernv: Restore non-volatile CRs after nap" to stable

[Greg Kurz]

Hi,

Please consider backporting the following upstream commit:

commit 0aab3747091db309b8a484cfd382a41644552aa3
Author: Sam Bobroff <sam.bobroff@au1.ibm.com>
Date:   Fri May 1 16:50:34 2015 +1000

    powerpc/powernv: Restore non-volatile CRs after nap

It is needed on 3.19 and 4.0 to be able to run KVM guests on POWER7 hosts

FYI, it currently prevents use of KVM for fedora 20, 21 and 22 !

Thanks.

--
Greg

Re: backport "powerpc/powernv: Restore non-volatile CRs after nap" to stable

[Michael Ellerman]

On Wed, 2015-06-03 at 19:21 +0200, Greg Kurz wrote:
> Hi,
> 
> Please consider backporting the following upstream commit:

Hi Greg,

The stable folks don't take requests for backports, they just take
backports that you've done for them :)

I also don't take requests for backports, because I'm too busy! :)

So if you're hitting this then you should do the backport and send it to
stable@vger.kernel.org and CC me and Sam (the original author).

cheers

Re: backport "powerpc/powernv: Restore non-volatile CRs after nap" to stable

[Greg Kurz]

On Sun, 07 Jun 2015 19:08:43 +1000
Michael Ellerman <mpe@ellerman.id.au> wrote:

> On Wed, 2015-06-03 at 19:21 +0200, Greg Kurz wrote:
> > Hi,
> > 
> > Please consider backporting the following upstream commit:
> 
> Hi Greg,
> 
> The stable folks don't take requests for backports, they just take
> backports that you've done for them :)
> 
> I also don't take requests for backports, because I'm too busy! :)
> 
> So if you're hitting this then you should do the backport and send it to
> stable@vger.kernel.org and CC me and Sam (the original author).
> 
> cheers
> 

Hi Michael,

Since the patch is already in Linus's tree and applies flawlessly on 3.19 and
4.0, I did as explained in Documentation/stable_kernel_rules.txt:

   --- Option 2 ---

   After the patch has been merged to Linus' tree, send an email to
   stable@vger.kernel.org containing the subject of the patch, the commit ID,
   why you think it should be applied, and what kernel version you wish it to
   be applied to.

I guess the confusion comes from the "backport" word, where "apply" would
have been more appropriate.

Stable folks,

Please apply the following upstream commit to 3.19 and 4.0 stable trees:


commit 0aab3747091db309b8a484cfd382a41644552aa3
Author: Sam Bobroff <sam.bobroff@au1.ibm.com>
Date:   Fri May 1 16:50:34 2015 +1000

powerpc/powernv: Restore non-volatile CRs after nap


It fixes guest hangs on POWER7 platforms that were introduced in 3.19 by
the following commits:

7cba160ad789 ("powernv/cpuidle: Redesign idle states management")
77b54e9f213f ("powernv/powerpc: Add winkle support for offline cpus")

Thanks.

--
Greg

Re: backport "powerpc/powernv: Restore non-volatile CRs after nap" to stable

[Greg KH]

On Sun, Jun 07, 2015 at 07:08:43PM +1000, Michael Ellerman wrote:
> On Wed, 2015-06-03 at 19:21 +0200, Greg Kurz wrote:
> > Hi,
> > 
> > Please consider backporting the following upstream commit:
> 
> Hi Greg,
> 
> The stable folks don't take requests for backports, they just take
> backports that you've done for them :)
> 
> I also don't take requests for backports, because I'm too busy! :)
> 
> So if you're hitting this then you should do the backport and send it to
> stable@vger.kernel.org and CC me and Sam (the original author).

Patches that apply cleanly as cherry-picked, like this one, don't need
to be sent as a patch, this is just fine.

thanks,

greg k-h