8359947: GenShen: use smaller TLABs by default

[earthling-amzn]

Clean backport. This is a performance improvement for the generational mode of Shenandoah.

---

Progress

• Change must not contain extraneous whitespace
• Commit message must refer to an issue
• JDK-8359947 needs maintainer approval

Issue

• JDK-8359947: GenShen: use smaller TLABs by default (Task - P4 - Approved)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk25u.git pull/99/head:pull/99
$ git checkout pull/99

Update a local copy of the PR:
$ git checkout pull/99
$ git pull https://git.openjdk.org/jdk25u.git pull/99/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 99

View PR using the GUI difftool:
$ git pr show -t 99

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk25u/pull/99.diff

Using Webrev

Link to Webrev Comment

[bridgekeeper]

👋 Welcome back wkemper! A progress list of the required criteria for merging this PR into master will be added to the body of your pull request. There are additional pull request commands available for use with this pull request.

[openjdk]

@earthling-amzn This change is no longer ready for integration - check the PR body for details.

[openjdk]

This backport pull request has now been updated with issue from the original commit.

[openjdk]

⚠️ @earthling-amzn This change is now ready for you to apply for maintainer approval. This can be done directly in each associated issue or by using the /approval command.

[mlbridge]

Webrevs

• 00: Full (ffa56c40)

[earthling-amzn]

/approval request This change addresses a performance issue which may cause Shenandoah to allocate LABs which are too large and may leave excess memory unused. We have tested this with GHA and additional jtreg and other performance and stress test suites.

[openjdk]

@earthling-amzn
8359947: The approval request has been created successfully.

[earthling-amzn]

/integrate

[openjdk]

@earthling-amzn
Your change (at version ffa56c4) is now ready to be sponsored by a Committer.

[shipilev]

On further thought and reading through the set of mainline experiments, I suspect we should really do this after the CAS allocation path is done? This would mitigate any problems with contention on heap lock with smaller TLABs.

(The synopsis is also misleading, since it also affects non-generational Shenandoah -- but there is nothing to be done at this point.)