8230392: Define AArch64 as MULTI_COPY_ATOMIC

[Andrew Dinn]

On 24/01/2020 09:01, Andrew Dinn wrote:
> On 23/01/2020 22:41, David Holmes wrote:
>> On 24/01/2020 2:47 am, Andrew Haley wrote:
>>> http://cr.openjdk.java.net/~aph/8230392-2/
>>>
>>> Better?
>>
>> That seems fine to me.
> 
> Yes, indeed, it looks good.
> 
>> This does make me wonder whether other lock-free code in the VM needs
>> special handling for non-CPU_MULTI_COPY_ATOMIC ??
> The only other place I could find a mention (not a use) of symbol
> CPU_MULTI_COPY_ATOMIC -- excluding the point where it is defined -- was
> in a comment in parse1.cpp. That comment talks about barriers for final
> fields which I believe is handled via definition of
> support_IRIW_for_not_multiple_copy_atomic_cpu.
Ah sorry, David, having reread I see now your concern may be whether
other lock-free code /should/ be considering the setting of
CPU_MULTI_COPY_ATOMIC but currently may /not/ be. That's an interesting
question . . . :-)

regards,


Andrew Dinn
-----------
Senior Principal Software Engineer
Red Hat UK Ltd
Registered in England and Wales under Company Registration No. 03798903
Directors: Michael Cunningham, Michael ("Mike") O'Neill