JDK-8354517 : Parallel: JDK-8339668 causes up to 3.7x slowdown in openjdk.bench.vm.gc.systemgc
  • Type: Bug
  • Component: hotspot
  • Sub-Component: gc
  • Affected Version: 25
  • Priority: P2
  • Status: Resolved
  • Resolution: Fixed
  • OS: generic
  • CPU: x86_64,aarch64
  • Submitted: 2025-04-14
  • Updated: 2025-06-03
  • Resolved: 2025-05-23
The Version table provides details related to the release that this issue/RFE will be addressed.

Unresolved : Release in which this issue/RFE will be addressed.
Resolved: Release in which this issue/RFE has been resolved.
Fixed : Release in which this issue/RFE has been fixed. The release containing this fix may be available for download as an Early Access Release or a General Availability Release.

To download the current JDK release, click here.
JDK 25
25 b25Fixed
Related Reports
Causes :  
Description
JDK-8339668 adopts PartialArrayState to consolidate marking stack in Full GC. This patch did however introduce performance regressions in openjdk.bench.vm.gc.systemgc.

Listing the worst regressions:
openjdk.bench.vm.gc.systemgc.DifferentObjectSizesArray.gc -98.99% (Linux x64)
openjdk.bench.vm.gc.systemgc.DifferentObjectSizesHashMap.gc  -93.64% (Linux x64)
openjdk.bench.vm.gc.systemgc.OneBigObject.gc -372.00% (Linux aarch64)
Comments
Changeset: 36f6d155 Branch: master Author: Albert Mingkun Yang <ayang@openjdk.org> Date: 2025-05-23 08:22:50 +0000 URL: https://git.openjdk.org/jdk/commit/36f6d155e3b9d0b279be33414573217ea38551ac
23-05-2025

A pull request was submitted for review. Branch: master URL: https://git.openjdk.org/jdk/pull/25382 Date: 2025-05-22 06:34:07 +0000
22-05-2025

One possibility is that the code changes resulted in the compiler changing some inlining decisions, so that some critical paths are no longer being inlined. We've had similar issues before, and they are hard to predict and control.
14-04-2025