RFR: Fix assert in ShenandoahTraversalGC

[Aleksey Shipilev]

On 06/26/2018 07:25 PM, Roman Kennke wrote:
> We've got one assert wrong.
> 
> # HG changeset patch
> # User rkennke
> # Date 1530033876 -7200
> #      Tue Jun 26 19:24:36 2018 +0200
> # Node ID 44f9dc08c064abe8cc91650961fdabcafd45c993
> # Parent  b1bc743e67843a53b6c77f058bbbceff139883d6
> [mq]: fixassert.patch
> 
> diff --git a/src/hotspot/share/gc/shenandoah/shenandoahTraversalGC.cpp
> b/src/hotspot/share/gc/shenandoah/shenandoahTraversalGC.cpp
> --- a/src/hotspot/share/gc/shenandoah/shenandoahTraversalGC.cpp
> +++ b/src/hotspot/share/gc/shenandoah/shenandoahTraversalGC.cpp
> @@ -957,7 +957,7 @@
>                                         &complete_gc, &yield,
>                                         NULL);
>    }
> -  assert(!sh->cancelled_gc() || task_queues()->is_empty(), "Should be
> empty");
> +  assert(sh->cancelled_gc() || task_queues()->is_empty(), "Should be
> empty");
>  }
> 
>  // Weak Reference Closures
> 
> 
> If approved, I'll push it to both sh/jdk and sh/jdk10
> 
> Ok?

Yes, please do.

-Aleksey