8367611: Enable vblendvp[sd] on Future ECore

[missa-prime]

The upcoming ECore platforms will benefit from using vblendvps and vblendvpd instructions when the destination register is the same as the source register. This change takes that situation into account.

The JTREG test shown below was used to verify correctness against the OpenJDK v26-b15 baseline build. This test frequently calls vblendvps() and vblendvpd() in the macro assembler. On Darkmont and non-Darkmont cores, the right code paths are followed when checking with asserts.

1. jtreg:test/hotspot/jtreg/compiler/intrinsics/math/TestSignumIntrinsic.java

---

Progress

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

Issue

• JDK-8367611: Enable vblendvp[sd] on Future ECore (Enhancement - P4)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk.git pull/27354/head:pull/27354
$ git checkout pull/27354

Update a local copy of the PR:
$ git checkout pull/27354
$ git pull https://git.openjdk.org/jdk.git pull/27354/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 27354

View PR using the GUI difftool:
$ git pr show -t 27354

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk/pull/27354.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]

❗ This change is not yet ready to be integrated.
See the Progress checklist in the description for automated requirements.

[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.

[mlbridge]

Webrevs

• 02: Full - Incremental (5e5c0275)
• 01: Full - Incremental (88ac7733)
• 00: Full (078a295d)

[mhaessig]

Thank you for this improvement, @missa-prime! The change looks good to me, but I have a question.

The JTREG test shown below was used to verify correctness against the OpenJDK v26-b15 baseline build.

• jtreg:test/hotspot/jtreg/compiler/intrinsics/math/TestSignumIntrinsic.java

Can you elaborate, what this verification entails? Why this test?

[missa-prime]

Thank you for this improvement, @missa-prime! The change looks good to me, but I have a question.

The JTREG test shown below was used to verify correctness against the OpenJDK v26-b15 baseline build.

• jtreg:test/hotspot/jtreg/compiler/intrinsics/math/TestSignumIntrinsic.java

Can you elaborate, what this verification entails? Why this test?

@mhaessig I used this test because it frequently calls vblendvps() and vblendvpd() in the macro assembler. On Darkmont and non-Darkmont cores, I checked the right code paths were followed with asserts. Also, I wrote some sample code and checked the instructions generated at runtime.

[vpaprotsk]

My only real comment is a nit-pick about the variable name.. use_blend_instr implies a range of darkmont and above and 'EnableX86CoreOpts' and 'below'.. which is not the value of this variable (i.e. just darkmont and above). Perhaps blend_fixed? blend_partly_fixed? Or 'just' add it to the blend_emulation...

bool blend_emulation = EnableX86ECoreOpts && UseAVX > 1 
                    && VM_Version::is_intel_darkmont() && (dst == src1); // partly fixed on Darkmont

[missa-prime]

I added it to blend_emulation as that seems to make the most sense.