On 08/05/2017 09:28, Paolo Bonzini wrote:
>
>
> On 08/05/2017 05:31, Paul Mackerras wrote:
>> On Sat, May 06, 2017 at 05:22:13PM +0200, Paolo Bonzini wrote:
>>>
>>>
>>> On 06/05/2017 12:02, Paul Mackerras wrote:
>>>> On Sat, Apr 29, 2017 at 02:39:11PM +0200, Paolo Bonzini wrote:
>>>>>
>>>>>
>>>>> On 29/04/2017 02:25, Paul Mackerras wrote:
>>>>>>
>>>>>> I would like this to go into 4.12. I know this pull request is
>>>>>> somewhat late, but that is mostly because I had 3 weeks vacation this
>>>>>> month. I have tested this on POWER9 and also tested that it doesn't
>>>>>> break other platforms such as POWER8.
>>>>>
>>>>> No problem, I'm a bit late too and this is linux-next anyway. I'll pull
>>>>> it next Tuesday because I have a public holiday.
>>>>
>>>> I haven't seen it go into kvm/next yet. Is there a problem with it,
>>>> or have you just not got up to it yet, or did you do it and I missed
>>>> it?
>>>
>>> It was in kvm/queue only, now it's there.
>>
>> I'm still not seeing it. Did you forget to push it out or something?
>
> Ok, connected my brain and pulled now. Sorry for the inconvenience.
I'm un-pulling this, because it creates merge conflicts with Linus's
tree due to an unwanted rebase (?) by either you or Michael.
In particular this commit
commit 8bf8f2e8c786f37991bd27332c75edcc524d2232
Author: Nicholas Piggin <npiggin@xxxxxxxxx>
Date: Wed Apr 19 05:12:16 2017 +1000
powerpc/64s: Revert setting of LPCR[LPES] on POWER9
is already present in Linus's tree with a different hash:
commit 8d1b48ef580097e111c2644e6fc6041b9784d218
Author: Nicholas Piggin <npiggin@xxxxxxxxx>
Date: Wed Apr 19 05:12:16 2017 +1000
powerpc/64s: Revert setting of LPCR[LPES] on POWER9
and causes conflicts with this subsequent commit:
commit 700b7eadd5625d22b8235fb21259b3d7d564c000
Author: Nicholas Piggin <npiggin@xxxxxxxxx>
Date: Wed Apr 19 05:12:17 2017 +1000
powerpc/64s: Power9 has no LPCR[VRMASD] field so don't set it
I'll take care of sending a separate pull request to Linus, but please
avoid doing this kind of cherry-pick in the future.
Paolo
>> I don't know if you saw the email exchange between Michael Ellerman
>> and Linus. Michael put into his tree a commit that moves the
>> declaration of powerpc_debugfs_root from asm/debug.h to asm/debugfs.h,
>> and that commit is now in Linus' tree. Since my tree adds a use of
>> powerpc_debugfs_root, it is necessary when merging my tree into Linus'
>> to add #include <asm/debugfs.h> to arch/powerpc/kvm/book3s_xive.c. I
>> can't do that in my tree because asm/debugfs.h doesn't exist there.
>
> Yes, I noted that and will remind Linus today when sending the pull request.
