RFR: 8261492: Shenandoah: reconsider forwardee accesses memory ordering [v2]

[Zhengyu Gu]

On Thu, 11 Feb 2021 14:07:58 GMT, Aleksey Shipilev <shade at openjdk.org> wrote:

>> Shenandoah carries forwardee information in object's mark word. Installing the new mark word is effectively "releasing" the object copy, and reading from the new mark word is "acquiring" that object copy.
>> 
>> For the forwardee update side, Hotspot's default for atomic operations is memory_order_conservative, which emits two-way memory fences around the CASes at least on AArch64 and PPC64. This seems to be excessive for Shenandoah forwardee updates, and "release" is enough.
>> 
>> For the forwardee load side, we need to guarantee "acquire". We do not do it now, reading the markword without memory semantics. It does not seem to pose a practical problem today, because GC does not access the object contents in the new copy, and mutators get this from the JRT-called stub that separates the fwdptr access and object contents access by a lot. It still should be cleaner to "acquire" the mark on load to avoid surprises.
>> 
>> Additional testing:
>>  - [x] Linux x86_64 `hotspot_gc_shenandoah`
>>  - [x] Linux AArch64 `hotspot_gc_shenandoah`
>>  - [x] Linux AArch64 `tier1` with Shenandoah
>
> Aleksey Shipilev has updated the pull request incrementally with one additional commit since the last revision:
> 
>   Make sure to access fwdptr with acquire semantics in assembler code

src/hotspot/share/gc/shenandoah/shenandoahForwarding.inline.hpp line 58:

> 56:   assert(Thread::current()->is_Java_thread(), "Must be a mutator thread");
> 57: 
> 58:   markWord mark = obj->mark_acquire();

Actually, I would argue that you don't need acquire here, since you don't touch anything other than mark word, so that there is no order that needs to be enforced.

src/hotspot/share/gc/shenandoah/shenandoahForwarding.inline.hpp line 89:

> 87:   // (would be paired with acquire in forwardee accessors). Acquire on failed update
> 88:   // would get the updated object after the forwardee load.
> 89:   markWord prev_mark = obj->cas_set_mark(new_mark, old_mark, memory_order_acq_rel);

Same here, CAS guarantees you to see latest mark word. memory_order_release should be sufficient here, paired with memory_order_acquire in resolve_forward barrier to ensure safe publishing the new object.

-------------

PR: https://git.openjdk.java.net/jdk/pull/2496