8287139: aarch64 intrinsic for unsignedMultiplyHigh #8840

rgiulietti · 2022-05-23T11:59:02Z

Adds aarch64 intrinsic support for Math.unsignedMultiplyHigh()

Progress

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

Issue

JDK-8287139: aarch64 intrinsic for unsignedMultiplyHigh

Reviewers

Andrew Haley (@theRealAph - Reviewer) ⚠️ Review applies to 14905156
Nick Gasson (@nick-arm - Reviewer)
@kristylee88 (no known github.com user name / role) ⚠️ Review applies to 14905156

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.java.net/jdk pull/8840/head:pull/8840
$ git checkout pull/8840

Update a local copy of the PR:
$ git checkout pull/8840
$ git pull https://git.openjdk.java.net/jdk pull/8840/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 8840

View PR using the GUI difftool:
$ git pr show -t 8840

Using diff file

Download this PR as a diff file:
https://git.openjdk.java.net/jdk/pull/8840.diff

rgiulietti · 2022-05-23T12:00:29Z

Verified with hsdis

  0x00000001163fa830:   umulh	x10, x10, x10               ;*invokestatic unsignedMultiplyHigh {reexecute=0 rethrow=0 return_oop=0}
                                                            ; - Umul::main@17 (line 5)

bridgekeeper · 2022-05-23T12:00:44Z

👋 Welcome back rgiulietti! 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 · 2022-05-23T12:03:11Z

@rgiulietti The following label will be automatically applied to this pull request:

hotspot-compiler

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 · 2022-05-23T12:07:05Z

Webrevs

01: Full - Incremental (07226edf)
00: Full (14905156)

openjdk · 2022-05-23T12:10:33Z

@rgiulietti 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:

8287139: aarch64 intrinsic for unsignedMultiplyHigh

Reviewed-by: aph, ngasson

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 33 new commits pushed to the master branch:

5974f5f: 8284213: Replace usages of 'a the' in xml
e0d361c: 8284191: Replace usages of 'a the' in hotspot and java.base
9473c38: 8286057: Make javac error on a generic enum friendlier
a0f6dd3: 8287121: Fix typo in jlink's description resource key lookup
bd361bc: 8286376: Wrong condition for using non-immediate oops on AArch64
4042dba: 8287138: Make VerifyOption an enum class
a276cd2: 8287151: Remove unused parameter in G1CollectedHeap::mark_evac_failure_object
cf57d72: 8287174: Remove deprecated configure arguments
6458a56: 8286177: C2: "failed: non-reduction loop contains reduction nodes" assert failure
1cd7850: 8287194: build failure on riscv after JDK-8286825
... and 23 more: https://git.openjdk.java.net/jdk/compare/89a1d055d93ad57bcec7c1accb3f53b4c30f594d...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 (@theRealAph, @nick-arm) 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).

rgiulietti · 2022-05-23T12:12:51Z

@theRealAph Thanks Andrew, will wait 24 hours for all timezones reviewers.

nick-arm

Looks fine apart from the tiny cosmetic issue. Did you try the MathBench.unsignedMultiplyHighLongLong JMH benchmark introduced with the x86 implementation?

nick-arm · 2022-05-23T12:41:25Z

src/hotspot/cpu/aarch64/aarch64.ad

+  match(Set dst (UMulHiL src1 src2));
+
+  ins_cost(INSN_COST * 7);
+  format %{ "umulh   $dst, $src1, $src2, \t# umulhi" %}


There's an unnecessary trailing comma here and in the smulh pattern above.

I didn't run the benchmark.
Isn't a single instruction supposed to be faster anyway, as opposed to the java method?

Hi, I tried the JMH benchmark with this patch, it does generate the "umulh" instruction instead of smulh.

Thanks @Bhavana-Kilambi, as the hsdis output above shows, this is indeed the case.

nsjian · 2022-05-23T12:50:28Z

src/hotspot/cpu/aarch64/aarch64.ad

@@ -11143,6 +11143,22 @@ instruct mulHiL_rReg(iRegLNoSp dst, iRegL src1, iRegL src2, rFlagsReg cr)
  ins_pipe(lmul_reg_reg);
 %}

+instruct umulHiL_rReg(iRegLNoSp dst, iRegL src1, iRegL src2, rFlagsReg cr)


rFlagsReg cr is not used.

Yes, I noticed, but since there are many more unused rFlagsReg cr in the file and since I don't know its semantics, I prefer to just leave it as it is.

rgiulietti · 2022-05-24T11:51:42Z

/integrate

openjdk · 2022-05-24T11:52:51Z

@rgiulietti
Your change (at version 07226ed) is now ready to be sponsored by a Committer.

rgiulietti · 2022-05-24T15:11:53Z

Hi, any committer willing to sponsor the integration? I don't have committer status.

nick-arm · 2022-05-24T15:26:28Z

/sponsor

rgiulietti · 2022-05-24T15:48:06Z

Sorry @nick-arm to bother again, but the /sponsor command provoked nothing, even after waiting for 21 min.
Could you please re-issue it again? Thanks

openjdk · 2022-05-24T15:49:52Z

Going to push as commit fdc147e.
Since your change was applied there have been 35 commits pushed to the master branch:

8f0eb5d: 8287158: Explicitly reject unsupported call shapes on macos-aarch64 in mainline
0a82c4e: 8287137: Problemlist failing x86_32 tests after Loom integration
5974f5f: 8284213: Replace usages of 'a the' in xml
e0d361c: 8284191: Replace usages of 'a the' in hotspot and java.base
9473c38: 8286057: Make javac error on a generic enum friendlier
a0f6dd3: 8287121: Fix typo in jlink's description resource key lookup
bd361bc: 8286376: Wrong condition for using non-immediate oops on AArch64
4042dba: 8287138: Make VerifyOption an enum class
a276cd2: 8287151: Remove unused parameter in G1CollectedHeap::mark_evac_failure_object
cf57d72: 8287174: Remove deprecated configure arguments
... and 25 more: https://git.openjdk.java.net/jdk/compare/89a1d055d93ad57bcec7c1accb3f53b4c30f594d...master

Your commit was automatically rebased without conflicts.

openjdk · 2022-05-24T15:51:07Z

@nick-arm @rgiulietti Pushed as commit fdc147e.

💡 You may see a message that your pull request was closed with unmerged commits. This can be safely ignored.

nick-arm · 2022-05-24T16:04:29Z

Sorry @nick-arm to bother again, but the /sponsor command provoked nothing, even after waiting for 21 min. Could you please re-issue it again? Thanks

Looks like the bot was just running slow :-)

8287139: aarch64 intrinsic for unsignedMultiplyHigh

1490515

openjdk bot added the rfr Pull request is ready for review label May 23, 2022

openjdk bot added the hotspot-compiler hotspot-compiler-dev@openjdk.org label May 23, 2022

theRealAph approved these changes May 23, 2022

View reviewed changes

openjdk bot added the ready Pull request is ready to be integrated label May 23, 2022

nick-arm reviewed May 23, 2022

View reviewed changes

nsjian reviewed May 23, 2022

View reviewed changes

8287139: aarch64 intrinsic for unsignedMultiplyHigh

07226ed

nick-arm approved these changes May 23, 2022

View reviewed changes

openjdk bot added the sponsor Pull request is ready to be sponsored label May 24, 2022

openjdk bot added the integrated Pull request has been integrated label May 24, 2022

openjdk bot closed this May 24, 2022

openjdk bot removed ready Pull request is ready to be integrated rfr Pull request is ready for review sponsor Pull request is ready to be sponsored labels May 24, 2022

rgiulietti deleted the JDK-8287139 branch May 25, 2022 10:27

ketanv3 mentioned this pull request Aug 12, 2023

Performance improvements for BytesRefHash opensearch-project/OpenSearch#8788

Merged

6 tasks

8287139: aarch64 intrinsic for unsignedMultiplyHigh #8840

8287139: aarch64 intrinsic for unsignedMultiplyHigh #8840

rgiulietti commented May 23, 2022 •

edited by openjdk bot

rgiulietti commented May 23, 2022

bridgekeeper bot commented May 23, 2022

openjdk bot commented May 23, 2022

mlbridge bot commented May 23, 2022 •

edited

openjdk bot commented May 23, 2022 •

edited

rgiulietti commented May 23, 2022

nick-arm left a comment

nick-arm May 23, 2022

rgiulietti May 23, 2022

Bhavana-Kilambi May 23, 2022

rgiulietti May 23, 2022

nsjian May 23, 2022

rgiulietti May 23, 2022

rgiulietti commented May 24, 2022

openjdk bot commented May 24, 2022

rgiulietti commented May 24, 2022

nick-arm commented May 24, 2022

rgiulietti commented May 24, 2022

openjdk bot commented May 24, 2022

openjdk bot commented May 24, 2022

nick-arm commented May 24, 2022

8287139: aarch64 intrinsic for unsignedMultiplyHigh #8840

8287139: aarch64 intrinsic for unsignedMultiplyHigh #8840

Conversation

rgiulietti commented May 23, 2022 • edited by openjdk bot

Progress

Issue

Reviewers

Reviewing

rgiulietti commented May 23, 2022

bridgekeeper bot commented May 23, 2022

openjdk bot commented May 23, 2022

mlbridge bot commented May 23, 2022 • edited

Webrevs

openjdk bot commented May 23, 2022 • edited

rgiulietti commented May 23, 2022

nick-arm left a comment

Choose a reason for hiding this comment

nick-arm May 23, 2022

Choose a reason for hiding this comment

rgiulietti May 23, 2022

Choose a reason for hiding this comment

Bhavana-Kilambi May 23, 2022

Choose a reason for hiding this comment

rgiulietti May 23, 2022

Choose a reason for hiding this comment

nsjian May 23, 2022

Choose a reason for hiding this comment

rgiulietti May 23, 2022

Choose a reason for hiding this comment

rgiulietti commented May 24, 2022

openjdk bot commented May 24, 2022

rgiulietti commented May 24, 2022

nick-arm commented May 24, 2022

rgiulietti commented May 24, 2022

openjdk bot commented May 24, 2022

openjdk bot commented May 24, 2022

nick-arm commented May 24, 2022

rgiulietti commented May 23, 2022 •

edited by openjdk bot

mlbridge bot commented May 23, 2022 •

edited

openjdk bot commented May 23, 2022 •

edited