RFR(S): 8154452: Deferred cleanups after split of G1CollectorPolicy code

Stefan Johansson stefan.johansson at oracle.com
Wed Apr 20 12:32:08 UTC 2016

On 2016-04-20 14:14, Erik Helin wrote:
> On 2016-04-20, Mikael Gerdin wrote:
>> Hi all,
>> Please review this small set of cleanups brought up during the review of
>> 8154154.
>> I've decided to simply remove {is,as}_g1_policy since they had no callers
>> and the code instead checks the UseG1GC flag to determine if G1 is running.
>> As Kim suggested, CollectorPolicy::post_heap_initialize does no useful work
>> for neither G1 nor other collectors and I decided to remove it.
>> Removing post_heap_initialize allowed me to remove its only caller,
>> ColelctedHeap::post_initialize and I made that method pure virtual.
>> I also moved the setting of G1HeapRegionSize closer to where the actual
>> value is computed. This allowed me to get rid of
>> G1CollectorPolicy::initialize_flags, thus making G1CollectorPolicy implement
>> a single pure virtual method, initialize_alignments().
>> Bug: https://bugs.openjdk.java.net/browse/JDK-8154452
>> Webrev: http://cr.openjdk.java.net/~mgerdin/8154452/webrev.0
> Looks good, Reviewed.
> Thanks,
> Erik
>> Testing: RBT GC Testing
>> /Mikael

More information about the hotspot-gc-dev mailing list