RFR (XS) 8244807: Shenandoah: ditch filter in ShenandoahUnload::unload

[Roman Kennke]

Ok, looks good to me.

Roman


> RFE:
>   https://bugs.openjdk.java.net/browse/JDK-8244807
> 
> After JDK-8243323, the filter for is_$phase_in_progress() in ShenandoahUnload::unload() is no longer
> required. Cleanup:
> 
> diff -r ef2f2bc93009 src/hotspot/share/gc/shenandoah/shenandoahUnload.cpp
> --- a/src/hotspot/share/gc/shenandoah/shenandoahUnload.cpp      Tue May 12 08:25:17 2020 +0200
> +++ b/src/hotspot/share/gc/shenandoah/shenandoahUnload.cpp      Tue May 12 11:55:10 2020 +0200
> @@ -144,11 +144,7 @@
> 
>  void ShenandoahUnload::unload() {
> -  assert(ShenandoahConcurrentRoots::can_do_concurrent_class_unloading(), "Why we here?");
> -
>    ShenandoahHeap* heap = ShenandoahHeap::heap();
> -
> -  if (!heap->is_concurrent_weak_root_in_progress()) {
> -    return;
> -  }
> +  assert(ShenandoahConcurrentRoots::can_do_concurrent_class_unloading(), "Filtered by caller");
> +  assert(heap->is_concurrent_weak_root_in_progress(), "Filtered by caller");
> 
>    // Unlink stale metadata and nmethods
> 
> 
> Testing: hotspot_gc_shenandoah
>