[15] RFR 8245124: Shenandoah: optimize code root evacuation/update during concurrent class unloading

[Zhengyu Gu]

>> Updated: http://cr.openjdk.java.net/~zgu/JDK-8245124/webrev.01/index.html
> 
> Good. Any reason why ShenandoahEvacOOMScope is not inside heal_nmethod_metadata()? Is there any
> problem to call _bs->disarm(nm) when holding that scope?
This is something I want to get rid of, especially on code root 
unlinking path. Moving ShenandoahEvacOOMScope up is another noticeable 
improvement, unfortunately, unload_nmethod_caches() triggers barrier.

-Zhengyu

>