RFR: 8295158: G1: Increase card-based cost sample reporting threshold

Thomas Schatzl tschatzl at openjdk.org
Tue Oct 11 14:43:58 UTC 2022


Hi all,

  can I have reviews for this change that makes the threshold that determines whether g1 uses that sample for prediction configurable (as diagnostic option) and increases it to 1000 instead of 10 because testing has shown that that's a much better value, filtering out outliers much better.

This change is based on PR#10654, please review that one first if still open.

Testing: Testing: gha, local testing, tier1-5 with other changes in this patch series, perf testing with other changes in this patch series

Thanks,
  Thomas

-------------

Depends on: https://git.openjdk.org/jdk/pull/10654

Commit messages:
 - Initial version cost sample threshold

Changes: https://git.openjdk.org/jdk/pull/10656/files
 Webrev: https://webrevs.openjdk.org/?repo=jdk&pr=10656&range=00
  Issue: https://bugs.openjdk.org/browse/JDK-8295158
  Stats: 11 lines in 2 files changed: 5 ins; 4 del; 2 mod
  Patch: https://git.openjdk.org/jdk/pull/10656.diff
  Fetch: git fetch https://git.openjdk.org/jdk pull/10656/head:pull/10656

PR: https://git.openjdk.org/jdk/pull/10656



More information about the hotspot-gc-dev mailing list