RFR: 8365880: Shenandoah: Unify memory usage accounting in ShenandoahFreeSet [v31]
William Kemper
wkemper at openjdk.org
Fri Oct 24 21:13:13 UTC 2025
On Fri, 24 Oct 2025 03:41:50 GMT, Kelvin Nilsen <kdnilsen at openjdk.org> wrote:
>> This PR eliminates redundant bookkeeping that had been carried out by both ShenandoahGeneration and ShenandoahFreeSet. In the new code, we keep a single tally of relevant information within ShenandoahFreeSet.
>> Queries serviced by ShenandoahGeneration are now delegated to ShenandoahFreeSet.
>>
>> This change eliminates rare and troublesome assertion failures that were often raised when the ShenandoahFreeSet tallies did not match the ShenandoahGeneration tallies. These assertion failures resulted because the two sets of books are updated at different times, using different synchronization mechanisms.
>>
>> The other benefit of this change is that we have less synchronization overhead because we only have to maintain a single set of books.
>
> Kelvin Nilsen has updated the pull request incrementally with one additional commit since the last revision:
>
> Rework implementation of CompressedClassSpaceSizeInJmapHeap.java
test/hotspot/jtreg/gc/metaspace/CompressedClassSpaceSizeInJmapHeap.java line 78:
> 76: do {
> 77: exitValue = run(pb);
> 78: } while ((exitValue != 0) && (allowed_retries-- > 0));
Just a nit, but the other variables here use Java's camel case convention, so should probably have `allowedRetries`, not `allowed_retries`.
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/26867#discussion_r2461988687
More information about the serviceability-dev
mailing list