String-dedup test failures

[Roman Kennke]

You either need to mark all the stuff in the strdeduptable and thus keep 
everything (normal phases), or not mark it (i.e. treat weakly) and clean 
up unreachable stuff *after marking* (cleanup phases). This is very 
similar to how we treat unload classes and similar stuff. Never should 
you keep unreachable stuff in the strdedup-table (I think this is what 
causes the issues that we see).

Roman
> Thought about overnight, it makes sense to treat it as weak root and 
> deal with it the same way as other weak roots.
>
> -Zhengyu
>
> On 11/16/2017 04:34 AM, Aleksey Shipilev wrote:
>> On 11/16/2017 10:25 AM, Roman Kennke wrote:
>>> I think we need to clean up the string dedup table *after marking*. 
>>> If a byte[] is not reachable
>>> after marking, we need to clean all references to it, because *it 
>>> may become trashed* during evac.
>>> If we want to keep it alive, then we need to mark it. Evac'ing an 
>>> unmarked byte[] after marking is
>>> calling for troubles, because we probably access a pointer that is 
>>> about to be invalidated.
>>
>> Yes, that's what I was suggesting:
>>
>> It seems that we want two separate modes: regular, and cleanup. Much 
>> like we do with "Parallel
>> Cleanup" during Final Mark. In regular mode, we mark all StrDedup 
>> entries always, maybe even
>> concurrently during mark? In cleanup mode, we don't mark StrDedup as 
>> roots, but at Final Mark we
>> purge all stale entries from there. Then if any String cannot 
>> "revive" the stale unmarked entry, and
>> have to insert new, reachable one. I think the mechanism above has 
>> the nice symmetry with class
>> unloading, etc., which makes it easier to reason about. Also makes 
>> cleanup during pause optional.
>>
>> -Aleksey
>>