8300235: Use VarHandle access in Image(Input | Output)StreamImpl classes #12204

minborg · 2023-01-25T16:39:04Z

This PR suggests improving performance by using the newly introduced class jdk.internal.util.ByteArray to improve packing/unpacking operations.

The PR also proposes adding a ByteArrayLittleEndian class for support for little endian packing/unpacking.

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-8300235: Use VarHandle access in Image(Input | Output)StreamImpl classes

Reviewers

Roger Riggs (@RogerRiggs - Reviewer) ⚠️ Review applies to 68d72946

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk pull/12204/head:pull/12204
$ git checkout pull/12204

Update a local copy of the PR:
$ git checkout pull/12204
$ git pull https://git.openjdk.org/jdk pull/12204/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 12204

View PR using the GUI difftool:
$ git pr show -t 12204

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk/pull/12204.diff

bridgekeeper · 2023-01-25T16:40:35Z

👋 Welcome back pminborg! 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 · 2023-01-25T16:43:34Z

@minborg The following labels will be automatically applied to this pull request:

client
core-libs

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

minborg · 2023-01-25T18:59:50Z

src/java.base/share/classes/module-info.java

@@ -272,6 +272,8 @@
        jdk.jfr;
    exports jdk.internal.util.random to
        jdk.random;
+    exports jdk.internal.util to


Added coupling I am afraid.

mlbridge · 2023-01-25T19:05:27Z

Webrevs

bplb · 2023-01-25T22:54:17Z

Looks good overall. Did you measure any performance changes in the image streams? (Note that this PR and its corresponding issue are missing "Stream" in the title.)

minborg · 2023-01-26T08:56:24Z

Here are some performance figures (TLDR: about 2% performance increase):

Baseline (J21 master)

     [java] All test results:
     [java] Test(imageio.input.stream.tests.readFully(int[])) averaged 2.0089243986475606E9 bytes/sec
     [java]     with 250x250, photo, byteArray, !useCache
     [java] Test(imageio.input.stream.tests.readLong) averaged 6.087925258517034E8 bytes/sec
     [java]     with 250x250, photo, byteArray, !useCache

Patch

     [java] All test results:
     [java] Test(imageio.input.stream.tests.readFully(int[]) averaged 2.0537614726652384E9 bytes/sec
     [java]     with 250x250, photo, byteArray, !useCache
     [java] Test(imageio.input.stream.tests.readLong) averaged 6.227265050569915E8 bytes/sec
     [java]     with 250x250, photo, byteArray, !useCache

the readFully(int[]) is a custom benchmark:

                iis.mark();
                int[] array = new int[8];
                do {
                    if (pos + 4*8 > length) {
                        iis.reset();
                        iis.mark();
                        pos = 0;
                    }
                    iis.readFully(array, 0, 8);
                    pos += 4*8;
                } while (--numReps >= 0);

Tests were run on a Mac M1 aarch64 machine.

minborg · 2023-01-26T09:13:26Z

Given the way the benchmarks are written, I suspect the actual performance gains are higher than indicated by the figures above. The benchmark likely spends more time in preparing for the call (e.g. checking, looping) than in the actual target method.

RogerRiggs · 2023-01-26T16:05:23Z

Can the benchmark be moved to the JMH framework; it deals with the setup and measurements in a more predictable way.
See test/micro/org/openjdk/bench/jajva/nio/... for examples.

minborg · 2023-02-02T14:57:40Z

Here are some benchmarks:

Baseline (JDK21, master (called Java 20 below))
Benchmark                       Mode  Cnt  Score   Error  Units
ImageInputStreamBench.readInt   avgt   15  1.289 ± 0.029  ns/op
ImageInputStreamBench.readLong  avgt   15  1.308 ± 0.017  ns/op

Patch (called Java 21 below)
Benchmark                       Mode  Cnt  Score   Error  Units
ImageInputStreamBench.readInt   avgt   15  0.946 ± 0.046  ns/op
ImageInputStreamBench.readLong  avgt   15  0.954 ± 0.010  ns/op

RogerRiggs

Looks good.
Thanks for the JMH benchmarks.

RogerRiggs · 2023-02-02T19:11:03Z

src/java.base/share/classes/jdk/internal/util/ByteArrayLittleEndian.java

+        DOUBLE.set(array, offset, value);
+    }
+
+    private static VarHandle create(Class<?> viewArrayClass) {


Renaming the method to createLittleEndian would make the static assignments at the top unambiguous.

openjdk · 2023-02-02T19:17:26Z

@minborg 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:

8300235: Use VarHandle access in Image(Input | Output)StreamImpl classes

Reviewed-by: rriggs

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 156 new commits pushed to the master branch:

f696785: 8300869: Make use of the Double.toString(double) algorithm in java.util.Formatter
cf6b9eb: 8301637: ThreadLocalRandom.current().doubles().parallel() contention
c647ae6: 8301149: Parallel: Refactor MutableNUMASpace::update_layout
de57733: 8301644: com/sun/jdi/JdbStopThreadTest.java fails after JDK-8300811
930ec00: 8301636: Minor cleanup in CommentHelper and DocPretty
725d57b: 8301659: Resolve initialization reordering issues on Windows for libawt and libsaproc
2d50c7d: 8298979: Remove duplicated serviceability/jvmti/thread/GetAllThreads/allthr01/allthr01.java
59b7fb1: 8300273: [IR framework] Handle message instead of bailing out
5b1584b: 8298880: VectorLogicalOpIdentityTest.java IR test incorrectly use avx3 instead of avx512
21c1afb: 8301612: OopLoadProxy constructor should be explicit
... and 146 more: https://git.openjdk.org/jdk/compare/f279c751a5b2132c11aa3a31084ee37f1a349150...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.

➡️ To integrate this PR with the above commit message to the master branch, type /integrate in a new comment.

minborg · 2023-02-03T07:22:34Z

/integrate

openjdk · 2023-02-03T07:24:29Z

Going to push as commit b504c94.
Since your change was applied there have been 163 commits pushed to the master branch:

406021a: 8300929: Avoid unnecessary array fill after creation in java.awt.image
7f313b0: 8180266: Convert sun/security/provider/KeyStore/DKSTest.sh to Java Jtreg Test
3ad6aef: 8301313: RISC-V: C2: assert(false) failed: bad AD file due to missing match rule
4c9de87: 8301655: Problemlist jdk/jdk/nio/zipfs/TestLocOffsetFromZip64EF.java on Linux
04278e6: 8301564: Non-C-heap allocated ResourceHashtable keys and values must have trivial destructor
b00b70c: 8286907: keytool should warn about weak PBE algorithms
ee0f5b5: 8301392: Port fdlibm log1p to Java
f696785: 8300869: Make use of the Double.toString(double) algorithm in java.util.Formatter
cf6b9eb: 8301637: ThreadLocalRandom.current().doubles().parallel() contention
c647ae6: 8301149: Parallel: Refactor MutableNUMASpace::update_layout
... and 153 more: https://git.openjdk.org/jdk/compare/f279c751a5b2132c11aa3a31084ee37f1a349150...master

Your commit was automatically rebased without conflicts.

openjdk · 2023-02-03T07:24:40Z

@minborg Pushed as commit b504c94.

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

AlanBateman · 2023-02-03T07:37:21Z

Do I read this correctly that the proposal is to have ByteArray and ByteArrayLittleEndian?

Use jdk.internal.util.ByteArray for pack/unpack

ded43f6

openjdk bot added client client-libs-dev@openjdk.org core-libs core-libs-dev@openjdk.org labels Jan 25, 2023

minborg added 2 commits January 25, 2023 19:43

Add little endian performance improvements

7f01fcd

Remove comments

3fb7133

minborg commented Jan 25, 2023

View reviewed changes

minborg marked this pull request as ready for review January 25, 2023 19:00

openjdk bot added the rfr Pull request is ready for review label Jan 25, 2023

minborg changed the title ~~8300235: Use VarHandle access in Image(Input | Output)Impl classes~~ 8300235: Use VarHandle access in Image(Input | Output)StreamImpl classes Jan 26, 2023

minborg closed this Jan 26, 2023

minborg reopened this Jan 26, 2023

RogerRiggs mentioned this pull request Jan 28, 2023

(Withdrawn) Update field access methods in ZipUtils to use the VarHandle API #12273

Closed

3 tasks

Add benchmark

3b3873d

Fix problem in benchmark

68d7294

RogerRiggs approved these changes Feb 2, 2023

View reviewed changes

openjdk bot added the ready Pull request is ready to be integrated label Feb 2, 2023

Rename to createLittleEndian

d58e22f

openjdk bot added the integrated Pull request has been integrated label Feb 3, 2023

openjdk bot closed this Feb 3, 2023

openjdk bot removed ready Pull request is ready to be integrated rfr Pull request is ready for review labels Feb 3, 2023

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

8300235: Use VarHandle access in Image(Input | Output)StreamImpl classes #12204

8300235: Use VarHandle access in Image(Input | Output)StreamImpl classes #12204

minborg commented Jan 25, 2023 •

edited by openjdk bot

Loading

bridgekeeper bot commented Jan 25, 2023

openjdk bot commented Jan 25, 2023

minborg Jan 25, 2023

mlbridge bot commented Jan 25, 2023 •

edited

Loading

bplb commented Jan 25, 2023

minborg commented Jan 26, 2023 •

edited

Loading

minborg commented Jan 26, 2023

RogerRiggs commented Jan 26, 2023

minborg commented Feb 2, 2023 •

edited

Loading

RogerRiggs left a comment

RogerRiggs Feb 2, 2023

openjdk bot commented Feb 2, 2023

minborg commented Feb 3, 2023

openjdk bot commented Feb 3, 2023

openjdk bot commented Feb 3, 2023

AlanBateman commented Feb 3, 2023

8300235: Use VarHandle access in Image(Input | Output)StreamImpl classes #12204

8300235: Use VarHandle access in Image(Input | Output)StreamImpl classes #12204

Conversation

minborg commented Jan 25, 2023 • edited by openjdk bot Loading

Progress

Issue

Reviewers

Reviewing

bridgekeeper bot commented Jan 25, 2023

openjdk bot commented Jan 25, 2023

minborg Jan 25, 2023

Choose a reason for hiding this comment

mlbridge bot commented Jan 25, 2023 • edited Loading

Webrevs

bplb commented Jan 25, 2023

minborg commented Jan 26, 2023 • edited Loading

minborg commented Jan 26, 2023

RogerRiggs commented Jan 26, 2023

minborg commented Feb 2, 2023 • edited Loading

RogerRiggs left a comment

Choose a reason for hiding this comment

RogerRiggs Feb 2, 2023

Choose a reason for hiding this comment

openjdk bot commented Feb 2, 2023

minborg commented Feb 3, 2023

openjdk bot commented Feb 3, 2023

openjdk bot commented Feb 3, 2023

AlanBateman commented Feb 3, 2023

minborg commented Jan 25, 2023 •

edited by openjdk bot

Loading

mlbridge bot commented Jan 25, 2023 •

edited

Loading

minborg commented Jan 26, 2023 •

edited

Loading

minborg commented Feb 2, 2023 •

edited

Loading