Re: [Patch V3 0/3] StartAllAPs should not use disabled APs

[Laszlo Ersek <lersek@redhat.com>]

On 07/25/18 14:12, Laszlo Ersek wrote:
> On 07/25/18 09:50, Eric Dong wrote:
>> This patch series include changes:
>> 1. StartAllAPs should not use disabled APs, this is required by UEFI spec.
>> 2. Refine the code to remove the redundant definitions.
>>
>> V2 changes:
>>   Use CpuStateReady to distinguish the AP state from CpuStateIdle.
>>
>> V3 Changes:
>>   Only change 3/3 patch. Only not use disabled APs when WakeUpAP called
>> by StartAllAps function. In other cases, also include disabled APs.
>>
>> Eric Dong (3):
>>   UefiCpuPkg/MpInitLib: Remove redundant CpuStateFinished State.
>>   UefiCpuPkg/MpInitLib: Remove StartCount and volatile definition.
>>   UefiCpuPkg/MpInitLib: Not use disabled AP when call StartAllAPs.
>>
>>  UefiCpuPkg/Library/MpInitLib/MpLib.c | 33 +++++++++++++++------------------
>>  UefiCpuPkg/Library/MpInitLib/MpLib.h |  4 +---
>>  2 files changed, 16 insertions(+), 21 deletions(-)
>>
> 
> I requested commit message updates for all three patches. With those
> implemented, please add:
> 
> Reviewed-by: Laszlo Ersek <lersek@redhat.com>
> 
> (No need to repost.)
> 
> Please give me some more time to regression-test this series as well.

This series works nicely on my end. I performed my usual Linux guest
tests, in all my usual guest environments, plus 10 cold boots in each
environment. Everything worked fine.

For patches #1 and #2:

Tested-by: Laszlo Ersek <lersek@redhat.com>

For patch #3:

Regression-tested-by: Laszlo Ersek <lersek@redhat.com>

(I'm giving R-t-b and not T-b for patch #3 because OVMF doesn't exercise
the MP PPI/Protocol member functions that disable APs.)

>From my side, please update the commit messages as agreed, and go ahead
with the pushing.

Thanks!
Laszlo