RFR: JDK-8221766: Load-reference barriers for Shenandoah

[Roman Kennke]

>> Does all that answer your question? :-)
> 
> Sure. One thing that always bothered me about the write barrier was that
> it looked fantastically expensive: all of the CPU state was pushed, and
> there is a lot of CPU state on AArch64. I never did understand why we can't
> have a fast path for that.

The write-barrier/LRB slowpath is not actually very hot, and with LRB it 
is even less hot. We did have a fast-path for that, you actually wrote 
it yourself, but it didn't make much difference (or any difference at 
all) in overall throughput. We ditched it when we introduced our 'safety 
net' mechanism to deal with OOM during evacuation, which would have 
required to slap even more stuff into that 'fastpath'. It didn't seem 
useful, and it only added to maintenance...

Roman