RFR: Improve/trim register saving/restoring around WB slowpath call in stub

[Roman Kennke]

Hi Aleksey,

>> I don't think it's doing much performance wise because it's the
>> slow-path anyway, and vector saving/restoring (32*8 words) would drown
>> it anway, but it's slightly nicer too I think. I'll not push for it, if
>> not considered useful.
> 
> It might be useful if it simplifies and streamlines things. Death by a thousand cuts is real.
> 
>> Testing: tier3_gc_shenandoah, specjvm+aggressive
>>
>> http://cr.openjdk.java.net/~rkennke/improve-wb-stub/webrev.00/
> 
> Can you rebase it to current Shenandoah, please?

It is already based on current Shenandoah.

Roman