RFR (XS): 8076541: Parallel GC registers Java heap twice to NMT

Per Liden per.liden at oracle.com
Tue Apr 21 15:13:39 UTC 2015


Hi Thomas,

On 2015-04-21 16:52, Thomas Schatzl wrote:
> Hi all,
>
>    can I have reviews for the following tiny change that fixes redundant
> method calls in parallel scavenge?
>
> ParallelScavengeHeap::initialize() calls
> MemTracker::record_virtual_memory_type() although
> Universe::reserve_heap() via the ReservedHeapSpace constructor already
> does that.
>
> Also, later it checks whether the the heap has been reserved or not, and
> exits the VM if necessary. Universe::reserve_heap() already does that.
>
> CR:
> https://bugs.openjdk.java.net/browse/JDK-8076541
>
> Webrev:
> http://cr.openjdk.java.net/~tschatzl/8076541/webrev/
>
> Testing:
> jprt, manual checking of NMT output to contain ReservedHeapSpace
> reservation for all gcs. Checked with ulimit -v that all collectors
> print the "Could not reserve enough space for XYKB object heap" message.

Change looks good!

/Per


More information about the hotspot-gc-dev mailing list