Integrated: 8327452: G1: Improve scalability of Merge Log Buffers
Ivan Walulya
iwalulya at openjdk.org
Wed Mar 13 10:18:19 UTC 2024
On Wed, 6 Mar 2024 11:18:03 GMT, Ivan Walulya <iwalulya at openjdk.org> wrote:
> Hi all,
>
> Please review this change to reduce contention on the DCQS. The parallel work done by the worker threads is very small, therefore contention on the DCQS dominates the Merge Log Buffers phase. In this change, we add a sequential phase to distribute the Log Buffers to the worker threads. This removes the DCQS bottleneck in the highly contended case, at a small cost to the cases with low contention.
>
> Testing Tier 1-3
>
> The graphs below are using Bigramtester at 20gb, "Distribute and Avg Log Buffers" combines the duration of the sequential distribute phase and the avg duration of the parallel `Merge Log Buffers` phase.
> | Time (ms) | # Cards |
> | ------------- | ------------- |
> |  | |
> | | |
> |  | |
> |  |  |
> |  |  |
> |  |  |
This pull request has now been integrated.
Changeset: 49d80089
Author: Ivan Walulya <iwalulya at openjdk.org>
URL: https://git.openjdk.org/jdk/commit/49d8008947534898e2ba36e2d81d87d6ae4b4c02
Stats: 79 lines in 4 files changed: 68 ins; 0 del; 11 mod
8327452: G1: Improve scalability of Merge Log Buffers
Reviewed-by: kbarrett, tschatzl
-------------
PR: https://git.openjdk.org/jdk/pull/18134
More information about the hotspot-gc-dev
mailing list