[aarch64-port-dev ] [RFR] 8u162 Upstream Sync

[Aleksey Shipilev]

On 03/21/2018 07:41 PM, Andrew Hughes wrote:
> On 21 March 2018 at 17:37, Aleksey Shipilev <shade at redhat.com> wrote:
>> On 03/21/2018 05:38 PM, Aleksey Shipilev wrote:
>>> On 03/21/2018 05:19 PM, Andrew Hughes wrote:
>>>>>>
>>>>>> You are doing Shenandoah merges too, right?
>>>>>>
>>>>>
>>>>> That's the plan :-D
>>>>>
>>>>> I don't expect anything as horrific as 8u152, which was a huge update.
>>>>>
>>>>
>>>> This is the Shenandoah merge:
>>>>
>>>> http://cr.openjdk.java.net/~andrew/aarch64-8/u162/hotspot//shenandoah.changeset
>>>>
>>>> Only conflicts were in src/share/vm/opto/phaseX.cpp and
>>>> src/share/vm/compiler/oopMap.cpp
>>>> Let me know if this looks ok and I'll push aarch64-shenandoah-jdk8u162-b12
>>>
>>> It looks okay to me.
>>
>> Double-checking, oopMap should end up like this against upstream:
>>
>> https://builds.shipilev.net/patch-openjdk-shenandoah-jdk8/2018-03-20-v154-vs-e0936bc82c5e/src/share/vm/compiler/oopMap.cpp.sdiff.html
>>
>> -Aleksey
>>
> 
> Yeah, I wondered about that oop cast.
> 
> Is there a reason it is removed from the Shenandoah version?
> It's not obvious from the code.

I think it is to avoid accidental match with oopDesc::operator== that has additional checking in
Shenandoah code.

> What versions are you comparing here? When I looked at
> http://hg.openjdk.java.net/shenandoah/jdk9/hotspot, it had a host
> of other changes too, and http://hg.openjdk.java.net/shenandoah/jdk8u/hotspot
> doesn't yet have 8180855.

I have just picked up your 8u162 sync to http://hg.openjdk.java.net/shenandoah/jdk9/hotspot, and now
it has 8180855, see:
 https://builds.shipilev.net/patch-openjdk-shenandoah-jdk8/2018-03-21-v156-vs-b08b1e9e2963/src/share/vm/compiler/oopMap.cpp.sdiff.html

-Aleksey