8388882: Disable automatic integer spills to floating point registers on Intel platforms

[missa-prime]

When operating on integer variables using supported Intel CPUs, the current approach is to spill onto floating point registers when general purpose registers aren't available. Unfortunately, it appears that the cost of doing this is higher than just leaving those variables on the stack. This change disables the UseFPUForSpilling flag by default on Intel CPUs.

To measure performance, the SpillCode.testSpillForManyInts micro-benchmark was run on the parts listed below.

• Intel® Xeon® Platinum 8252C
• Intel® Xeon® Platinum 8375C
• Intel® Xeon® Platinum 8488C
• Intel® Xeon® Platinum 8559C
• Intel® Xeon® 6975P-C

The table below shows how performance changed using the OpenJDK v28-b6 build. The Baseline column represents runs where the flag was enabled, and the New columns represents runs where the flag was disabled. Each result is the mean of 4 individual runs. Overall, setting UseFPUForSpilling to false provides an uplift between 7% and 10%.

Part ID	Baseline runtime (ns/op)	New runtime (ns/op)	Speedup
Intel® Xeon® Platinum 8252C	551.357	516.526	1.07x
Intel® Xeon® Platinum 8375C	726.729	661.194	1.10x
Intel® Xeon® Platinum 8488C	644.611	597.476	1.08x
Intel® Xeon® Platinum 8559C	613.719	567.634	1.08x
Intel® Xeon® Platinum 6975P-C	627.716	581.841	1.08x

Additionally, other micro-benchmarks show improvements when disabling the flag.

---

• I confirm that I make this contribution in accordance with the OpenJDK Interim AI Policy.

---

Progress

• Change must not contain extraneous whitespace
• Commit message must refer to an issue
• Change must be properly reviewed (2 reviews required, with at least 1 Reviewer, 1 Author)

Issue

• JDK-8388882: Disable automatic integer spills to floating point registers on Intel platforms (Enhancement - P4)

Reviewers

• Sandhya Viswanathan (@sviswa7 - Reviewer)
• Derek White (@dwhite-intel - Committer)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk.git pull/32031/head:pull/32031
$ git checkout pull/32031

Update a local copy of the PR:
$ git checkout pull/32031
$ git pull https://git.openjdk.org/jdk.git pull/32031/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 32031

View PR using the GUI difftool:
$ git pr show -t 32031

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk/pull/32031.diff

Using Webrev

Link to Webrev Comment

[bridgekeeper]

👋 Welcome back missa! 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]

@missa-prime This change now passes all automated pre-integration checks.

ℹ️ This project also has non-automated pre-integration requirements. Please see the file CONTRIBUTING.md for details.

After integration, the commit message for the final commit will be:

8388882: Disable automatic integer spills to floating point registers on Intel platforms

Reviewed-by: sviswanathan, drwhite

You can use pull request commands such as /summary, /contributor and /issue to adjust it as needed.

At the time when this comment was updated there had been 20 new commits pushed to the master branch:

• bb2cdfe: 8387743: [aot] AOT tests fail intermittently with "incompatible CompressedOops::base()"
• e6f5c5d: 8388809: Shenandoah: Control thread may observe incorrect cancellation reason
• a4e39d6: 8388443: C2: Assertion in PhaseRegAlloc::reg2offset() is too strict
• ... and 17 more: https://git.openjdk.org/jdk/compare/611f186cf3a47258116146d70e405ef483e65e2a...master

As there are no conflicts, your changes will automatically be rebased on top of these commits when integrating. If you prefer to avoid this automatic rebasing, please check the documentation for the /integrate command for further details.

As you do not have Committer status in this project an existing Committer must agree to sponsor your change. Possible candidates are the reviewers of this PR (@sviswa7, @dwhite-intel) but any other Committer may sponsor as well.

➡️ To flag this PR as ready for integration with the above commit message, type /integrate in a new comment. (Afterwards, your sponsor types /sponsor in a new comment to perform the integration).

[openjdk]

@missa-prime The following label will be automatically applied to this pull request:

• hotspot

When this pull request is ready to be reviewed, an "RFR" email will be sent to the corresponding mailing list. If you would like to change these labels, use the /label pull request command.

[openjdk]

The total number of required reviews for this PR has been set to 2 based on the presence of this label: hotspot. This can be overridden with the /reviewers command.

[mlbridge]

Webrevs

• 00: Full (31efcb31)

[sviswa7]

Looks good to me.

[sviswa7]

/label hotspot-compiler

[openjdk]

@sviswa7
The hotspot-compiler label was successfully added.

[dwhite-intel]

Yes, looks good!