[linux-yocto] [PATCH 0/7] Intel Axxia updates to linux-yocto-4.9

[Daniel Dragomir]

On 06.08.2018 02:52, Bruce Ashfield wrote:
>
>
> On 8/3/2018 11:39 AM, Daniel Dragomir wrote:
>> On 26.07.2018 18:15, Bruce Ashfield wrote:
>>> On 2018-07-26 6:58 AM, Daniel Dragomir wrote:
>>>> Hello Bruce!
>>>>
>>>> Please review and merge this series of patches in Axxia branches from
>>>> linux-yocto-4.9:
>>>>   - standard/axxia/base
>>>>   - standard/preempt-rt/axxia/base
>>>>
>>>> NOTE 1: First patch (kernel/time: Fix Compile Warning) is only for the
>>>> preempt-rt branch. It seems that there was a merge error in the past
>>>> and an unused variable left behind causing a build warning.
>>>>
>>>> NOTE 2: There is another merge issue in all preempt-rt branches 
>>>> starting
>>>> with commit 6661efc5c2e9d70ff94ed7561b524a5bf9653a1b "Merge branch
>>>> 'standard/base' into standard/preempt-rt/base". An unsolved merge 
>>>> conflict
>>>> left some version controll markers in kernel/locking/rtmutex.c:
>>>
>>> How annoying. The previous git resolution claimed it took care of it,
>>> and I trusted it.
>>>
>>> I actually merged this -rt patch to standard/preempt-rt/base and then
>>> to all the branches, so the axxia branches have it as well.
>>>
>>> All the changes are now merged.
>>>
>>> Bruce
>>
>> Thank you, Bruce!
>>
>> About the control markers issue, there will be merged a patch to fix 
>> this in all
>> preempt branches, or should we send one for our branch?
>
> I did the resolution for the main branch and just merged it
> out. The resolution was to put things back the way they
> were before on the -rt branch.
>
> I'm out of the office, so I couldn't test it much. So have
> a look and send any follow ups that may be required.
>
> Cheers,
>
> Bruce

Hi Bruce,

There is another hook with some version control markers in the same file:
kernel/locking/rtmutex.c:1152:1: error: version control conflict marker 
in file

I'll send a short commit to fix this to be easier for you. You can 
combine it with yours.

Except this, my build on axxia rt branch works fine.

Regards,
Daniel

>
>>
>> Regards,
>> Daniel
>>
>>>
>>>>
>>>> kernel-source/kernel/locking/rtmutex.c:989:1: error: version 
>>>> control conflict marker in file
>>>> <<<<<<< HEAD
>>>>     if (do_mig_dis)
>>>>         migrate_disable();
>>>> ||||||| merged common ancestors
>>>>
>>>>
>>>> If you add more patches to standard/(preempt-rt)/axxia/base beforehand
>>>> and patches didn't apply, please notify me. I will rebase our changes.
>>>>
>>>> Changes included in this series (axxia tag 1.17 to 1.23):
>>>>
>>>> Linux 4.9, Axxia 1.23 Changes
>>>>      EDAC fix for interrupt storm issue from the SMEM controller.
>>>>
>>>> Linux 4.9, Axxia 1.22 Changes
>>>>      Allow non-secure access to the CCN registers. This is required,
>>>>      for example, to implement some work arounds.
>>>>
>>>> Linux 4.9, Axxia 1.21 Changes
>>>>      Update to MSI/MSIx interrupt support. As only one is supported
>>>>      at a time, select which one is supported based on the first
>>>>      interrupt allocation.
>>>>
>>>> Linux 4.9, Axxia 1.20 Changes
>>>>      Fix a bug in the EDAC driver. Don't set interrupt thread affinity
>>>>      when interrupts are not in use.
>>>>
>>>> Linux 4.9, Axxia 1.19 Changes
>>>>      Use level sensitive interrupts in the MSI interrupt handling 
>>>> code.
>>>>      Extend EDAC CMEM/SMEM debug config for 5600 and 6700.
>>>>      EDAC updates
>>>>          CMEM/SMEM driver trace capabilities added
>>>>          CMEM/SMEM cpu0 affinity used
>>>>          SMEM code maintainability improved
>>>>          CMEM multiple ca parity error event added
>>>>          CMEM mpr dump on signal bit [21]
>>>>          CMEM clear mr5 ca parity error flag added
>>>>          CMEM irq storm fix
>>>>
>>>> Thank you,
>>>> Daniel Dragomir
>>>>
>>>>
>>>> John Jacques (2):
>>>>    kernel/time: Fix Compile Warning
>>>>    drivers/misc: Allow CCN Register Access
>>>>
>>>> Marek Majtyka (3):
>>>>    drivers/edac: include/trace: EDAC improvements
>>>>    drivers/edac: Fixed bug in EDAC L3 driver
>>>>    drivers/edac: Fixed bug in smem driver
>>>>
>>>> Palani (1):
>>>>    drivers/pci/host: PCIe MSI/MSIx Interrupt Fix
>>>>
>>>> palaniap (1):
>>>>    drivers/pci/host: PCIe MSI Interrupt Fix
>>>>
>>>>   drivers/edac/Kconfig                  |   7 ++
>>>>   drivers/edac/axxia_edac-cmc_56xx.c    | 222 
>>>> ++++++++++++++++++++++++++++------
>>>>   drivers/edac/axxia_edac-l2_cpu_56xx.c |   2 +-
>>>>   drivers/edac/axxia_edac-l3_56xx.c     |  44 +++++--
>>>>   drivers/edac/axxia_edac-mc_56xx.c     | 176 
>>>> ++++++++++++++++++++++-----
>>>>   drivers/misc/axxia-oem.c              | 209 
>>>> +++++++++++++++++++++++++++-----
>>>>   drivers/pci/host/pcie-axxia.c         |  76 +++++++++---
>>>>   include/linux/axxia-oem.h             |  10 ++
>>>>   include/trace/events/edac_cmc.h       | 101 ++++++++++++++++
>>>>   include/trace/events/edac_mc.h        | 104 ++++++++++++++++
>>>>   kernel/time/posix-timers.c            |   1 -
>>>>   11 files changed, 829 insertions(+), 123 deletions(-)
>>>>   create mode 100644 include/trace/events/edac_cmc.h
>>>>   create mode 100644 include/trace/events/edac_mc.h
>>>>
>>>
>>