RFA (S) 8175887: C1 value numbering handling of Unsafe.get*Volatile is incorrect

[Aleksey Shipilev]

Thanks Rob!

Vladimir Ivanov, could you sponsor this one too, please?

-Aleksey

On 03/04/2017 12:09 AM, Rob McKenna wrote:
> Approved, please work with the compiler team to find an appropriate
> sponsor.
> 
>     -Rob
> 
> On 03/03/17 06:48, Aleksey Shipilev wrote:
>> Hi,
>>
>> Please approve the backport of this important bugfix to 8u:
>>
>> Bug:
>>   https://bugs.openjdk.java.net/browse/JDK-8175887
>>
>> JDK 9 changeset:
>>   http://hg.openjdk.java.net/jdk9/hs/hotspot/rev/2ff05d967fb2
>>
>> Compiler folks agree with backport, and nightlies seem clean:
>>   http://mail.openjdk.java.net/pipermail/hotspot-compiler-dev/2017-March/025767.html
>>
>> JDK 8u webrev:
>>   http://cr.openjdk.java.net/~shade/8175887/webrev.04.8u/
>>
>> Product change applies cleanly, but test needs to change j.i.m.Unsafe to
>> s.m.Unsafe.
>>
>> Testing: Linux x86_64/release hotspot/test/compiler/c1 before/after the fix.
>>
>> Need a sponsor to push.
>>
>> Thanks,
>> -Aleksey
>>