RFR: Report how much we have failed to allocate during Allocation Failure

Roman Kennke rkennke at redhat.com
Thu Apr 12 10:24:57 UTC 2018


Am 12.04.2018 um 11:39 schrieb Aleksey Shipilev:
> http://cr.openjdk.java.net/~shade/shenandoah/alloc-report-fail-size/webrev.01/
> 
> This adds the message about the failed allocation size into the GC log. With free set state
> printout, this makes an easy way to spot fragmentation problems. Sample output with completely
> depleted free set:
> 
> [4.884s][info][gc      ] GC(1) Failed to allocate 4K for evacuation
> [4.884s][info][gc      ] GC(1) Cancelling concurrent GC: Allocation Failure During Evac
> [4.886s][info][gc      ] GC(1) Concurrent evacuation 8749M->10239M(10240M) 97.007ms
> [4.886s][info][gc,ergo ] Free set: Used: 1499M of 1499M, Regions: 0 mutator, 0 collector
> [4.886s][info][gc,ergo ] Free set: Mutator view: Max regular: 0K, Max humongous: 0K, External frag:
> 0%, Internal frag: 0%
> [4.886s][info][gc,ergo ] Free set: Collector view: Max regular: 0K
> 
> Testing: hotspot_gc_shenandoah
> 
> Thanks,
> -Aleksey
> 

Looks good!

Thanks,
Roman



More information about the shenandoah-dev mailing list