[concurrency-interest] Semantics of compareAndSwapX

Doug Lea dl at cs.oswego.edu
Thu Feb 27 06:22:49 EST 2014


On 02/27/2014 04:22 AM, Andrew Haley wrote:
> On 02/26/2014 06:11 PM, Hans Boehm wrote:
>>
>> On ARMv8, I would expect a volatile store to be compiled to a store
>> release, and a volatile load to be compiled to a load acquire.  Period.
>
> ...
>
>> I'm hedging a bit here, because the original Java memory model doesn't say
>> anything about CAS, and I don't fully understand the details of the ARMv8
>> model, particularly the interaction between acquire/release loads and
>> stores and traditional ARM fences.
>
> Doug, is this right?  It sounds reasonable enough.
>

It seems right. As mentioned in hotspot-dev posts by me, you
and SAP, some hotspot internals probably ought to be revised
to better support this.

-Doug




More information about the Concurrency-interest mailing list