Re: [PATCH 0/5] intel_pstate updates for v3.14-rcX

[Dirk Brandewie <dirk.brandewie@gmail.com>]

On 02/18/2014 04:43 PM, Rafael J. Wysocki wrote:
> On Tuesday, February 18, 2014 04:24:02 PM Dirk Brandewie wrote:
>> On Tuesday, February 18, 2014, Rafael J. Wysocki <rjw@rjwysocki.net> wrote:
>>
>>> On Tuesday, February 18, 2014 03:53:48 PM Dirk Brandewie wrote:
>>>> On 02/18/2014 02:27 PM, Rafael J. Wysocki wrote:
>>>>> On Tuesday, February 18, 2014 12:29:54 PM Dirk Brandewie wrote:
>>>>>> Hi Rafael,
>>>>>
>>>>> Hi,
>>>>>
>>>>>> On 02/12/2014 10:01 AM, dirk.brandewie@gmail.com <javascript:;>wrote:
>>>>>>> From: Dirk Brandewie <dirk.j.brandewie@intel.com <javascript:;>>
>>>>>>>
>>>>>>> Based on v3.14-rc2
>>>>>>>
>>>>>>> Patch 1 removes energy reporting the patch from Maurizio Lambardi
>>>>>>>     intel_pstate: fix race condition in intel_pstate_init() can be
>>> dropped.
>>>>>>>
>>>>>>
>>>>>> Any reason why patches 2-5 did not make rc3 other than timing?
>>>>>>
>>>>>> Patches 2/3 can easily wait for v3.15.x
>>>>>>
>>>>>> Patches 4/5 fix bugs that are in the wild.
>>>>>
>>>>> I asked you about them, but you didn't reply:
>>>>>
>>>>> http://marc.info/?l=linux-pm&m=139225158531023&w=4
>>>>>
>>>>> Again, do patches [4-5/5] depend on [2-3/5]?
>>>>>
>>>>> If not, I can queue them up for -rc4.
>>>>
>>>> All the patches are independent of one another.
>>>>
>>>> Patch 2 is straight cleanup no functional change but reduces the memory
>>>> footprint slightly.
>>>>
>>>> Patch 3 is a bug that will only be seen when the PID is reset at init
>>> time
>>>> or when a change is made to PID params via debugfs.  The problem will
>>> only
>>>> exist for one sample time since it is setting last_err in the PID.
>>>>
>>>> Patch 4-5 are bugs found during Baytrail-T testing
>>>
>>> Are there any pointers to bug reports that may be included in the
>>> changelogs
>>> of these?
>>
>>
>> No.  I got the reports via email.  I could probably get the reporters to
>> file bugzillas.
>
> It would be good to add information about what machines are affected
> and what the user-visible problems are to the changelogs for future
> reference.
>
> And do we want these two patches in -stable?  If so, what -stable series should
> they go into?

Patch 2 v3.10+

Patch 3 v3.12+

Patch 4/5 v3.13+

--Dirk
>
> Rafael
>