8363980: [macos] Add JDK specific keys/values to Info.plist of embedded runtime #28033

sashamatveev · 2025-10-29T01:32:50Z

Added JDK specific keys/values to Info.plist of embedded runtime.
Modified setFakeRuntime() not to include bin folder. By default it was always included, but generated embedded runtime by default does not have bin folder. As a result CustomInfoPListTest failed.
Updated CustomInfoPListTest to test Info.plist with bin folder.

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-8363980: [macos] Add JDK specific keys/values to Info.plist of embedded runtime (Enhancement - P4)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.org/jdk.git pull/28033/head:pull/28033
$ git checkout pull/28033

Update a local copy of the PR:
$ git checkout pull/28033
$ git pull https://git.openjdk.org/jdk.git pull/28033/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 28033

View PR using the GUI difftool:
$ git pr show -t 28033

Using diff file

Download this PR as a diff file:
https://git.openjdk.org/jdk/pull/28033.diff

Using Webrev

Link to Webrev Comment

…ed runtime

…ed runtime [v2]

bridgekeeper · 2025-10-29T01:33:54Z

👋 Welcome back almatvee! 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 · 2025-10-29T01:34:54Z

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

openjdk · 2025-10-29T01:35:44Z

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

core-libs

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.

sashamatveev · 2025-10-29T01:39:36Z

@alexeysemenyukoracle Please review.

mlbridge · 2025-10-29T01:39:40Z

Webrevs

00: Full (04c9a0ea)

alexeysemenyukoracle · 2025-10-29T02:29:24Z

src/jdk.jpackage/macosx/classes/jdk/jpackage/internal/MacPackagingPipeline.java


 final var app = env.app();

+ // We should use full runtime info plist for standalone runtime and for


I guess the comment explains why jpackage should use the same info plist template for a runtime if it has the "bin" subdirectory. But what is the "full" runtime?

"full" runtime info plist -> refers to info plist from standalone runtime. I will re-write it to:

// We should use info plist from standalone runtime for embedded runtime // if embedded runtime contains "bin" folder, so embedded runtime // can act as standalone runtime. Standalone runtime always uses same // info plist which has "JavaVM" dictionary.

so embedded runtime can act as standalone runtime

This can imply many things. I suggest more specific wording:

If the embedded runtime contains executable(s) in the "bin" subdirectory, we should use the standalone runtime info plist template. Otherwise, the user may be unable to run the "java" or other executables in the "bin" subdirectory of the embedded runtime.

alexeysemenyukoracle · 2025-10-29T03:30:34Z

How do we test the defaults? We need a new test that will verify that if the runtime has a "bin" subdirectory, the runtime plist file has "standalone runtime"-specific properties and vice versa.
The test should be used with the followign runtimes:

A runtime with "bin" subdirectory created by jpackage.
A runtime without "bin" subdirectory created by jpackage.
A runtime with "bin" subdirectory passed to jpackage through --runtime-image.
A runtime without "bin" subdirectory passed to jpackage through --runtime-image.

alexeysemenyukoracle · 2025-10-29T03:36:19Z

test/jdk/tools/jpackage/helpers/jdk/jpackage/test/JPackageCommand.java

 // Need to make the code in rpm spec happy as it assumes there is
 // always something in application image.
 fakeRuntimeDir.resolve("bin").toFile().mkdir();
+ createBulkFile.accept(fakeRuntimeDir.resolve(Path.of("bin", "bulk")));


This change makes the above fakeRuntimeDir.resolve("bin").toFile().mkdir(); redundant. Overall, the whole

if (TKit.isLinux()) { // Need to make the code in rpm spec happy as it assumes there is // always something in application image. fakeRuntimeDir.resolve("bin").toFile().mkdir(); }

construction is redundant and can be replaced with:

if (includeBin) { createBulkFile.accept(fakeRuntimeDir.resolve(Path.of("bin", "bulk"))); }

We still need to add file to "bin" for Linux. I will change it to:

// Need to make the code in rpm spec happy as it assumes there is // always something in application image. includeBin |= TKit.isLinux(); if (includeBin) { createBulkFile.accept(fakeRuntimeDir.resolve(Path.of("bin", "bulk"))); }

We still need to add file to "bin" for Linux. I will change it to:

I simplified the body of the setFakeRuntime() function as follows:

addPrerequisiteAction(cmd -> { Path fakeRuntimeDir = TKit.createTempDirectory("fake_runtime"); TKit.trace(String.format("Init fake runtime in [%s] directory", fakeRuntimeDir)); if (TKit.isOSX()) { // Make MacAppImageBuilder happy createBulkFile.accept(fakeRuntimeDir.resolve(Path.of( "lib/jli/libjli.dylib"))); } // Make sure fake runtime takes some disk space. // Package bundles with 0KB size are unexpected and considered // an error by PackageTest. createBulkFile.accept(fakeRuntimeDir.resolve(Path.of("lib", "bulk"))); cmd.setArgumentValue("--runtime-image", fakeRuntimeDir); });

And ran Windows and Linux tests that use it. All passed.

WinOSConditionTest:

[18:44:19.893] TRACE: assertStringListEquals(): Check jpackage didn't modify ${RUNTIME_IMAGE}=[C:\jpackage-tests\WinOSConditionTest\test\fake_runtime-0] [18:44:19.893] TRACE: assertStringListEquals(1, #) [18:44:19.894] TRACE: assertStringListEquals(2, lib#) [18:44:19.894] TRACE: assertStringListEquals(3, lib\bulk#2025-10-30T22:44:17.6191874Z) [18:44:19.894] TRACE: assertTrue(): Check [WinOSConditionTest\test\output\WinOSConditionTest-1.0.msi] path exists

AppAboutUrlTest.testDefaults (DEB):

[18:49:22.115] TRACE: assertStringListEquals(): Check jpackage didn't modify ${RUNTIME_IMAGE}=[/jpackage-tests/AppAboutUrlTest/testDefaults/fake_runtime] [18:49:22.120] TRACE: assertStringListEquals(1, #) [18:49:22.120] TRACE: assertStringListEquals(2, lib#) [18:49:22.120] TRACE: assertStringListEquals(3, lib/bulk#2025-10-30T22:49:20.128409727Z) [18:49:22.120] TRACE: assertTrue(): Check [AppAboutUrlTest/testDefaults/output/defaultsappabouturltest_1.0_amd64.deb] path exists

AppAboutUrlTest.testDefaults (RPM):

[18:56:37.897] TRACE: assertStringListEquals(): Check jpackage didn't modify ${RUNTIME_IMAGE}=[/jpackage-tests/AppAboutUrlTest/testDefaults/fake_runtime] [18:56:37.898] TRACE: assertStringListEquals(1, #) [18:56:37.898] TRACE: assertStringListEquals(2, lib#) [18:56:37.898] TRACE: assertStringListEquals(3, lib/bulk#2025-10-30T22:56:37.204554597Z) [18:56:37.898] TRACE: assertTrue(): Check [AppAboutUrlTest/testDefaults/output/defaultsappabouturltest-1.0-1.x86_64.rpm] path exists

The diff:

diff --git a/test/jdk/tools/jpackage/helpers/jdk/jpackage/test/JPackageCommand.java b/test/jdk/tools/jpackage/helpers/jdk/jpackage/test/JPackageCommand.java index fcd940e725e..73af859b6ac 100644 --- a/test/jdk/tools/jpackage/helpers/jdk/jpackage/test/JPackageCommand.java +++ b/test/jdk/tools/jpackage/helpers/jdk/jpackage/test/JPackageCommand.java @@ -307,24 +307,16 @@ public JPackageCommand setFakeRuntime() { TKit.trace(String.format("Init fake runtime in [%s] directory", fakeRuntimeDir)); - Files.createDirectories(fakeRuntimeDir); - - if (TKit.isLinux()) { - // Need to make the code in rpm spec happy as it assumes there is - // always something in application image. - fakeRuntimeDir.resolve("bin").toFile().mkdir(); - } - if (TKit.isOSX()) { // Make MacAppImageBuilder happy createBulkFile.accept(fakeRuntimeDir.resolve(Path.of( "lib/jli/libjli.dylib"))); } - // Mak sure fake runtime takes some disk space. + // Make sure fake runtime takes some disk space. // Package bundles with 0KB size are unexpected and considered // an error by PackageTest. - createBulkFile.accept(fakeRuntimeDir.resolve(Path.of("bin", "bulk"))); + createBulkFile.accept(fakeRuntimeDir.resolve(Path.of("lib", "bulk"))); cmd.setArgumentValue("--runtime-image", fakeRuntimeDir); });

alexeysemenyukoracle · 2025-10-29T04:07:14Z

Updated CustomInfoPListTest to test Info.plist with bin folder.

Why? jpackage picks the same file from the resource directory regardless of whether the "bin" subdirectory is present in the runtime. The "bin" subdirectory in the runtime only affects the default plist resource that jpackage uses.

alexeysemenyukoracle · 2025-10-29T14:42:18Z

src/jdk.jpackage/macosx/classes/jdk/jpackage/internal/MacPackagingPipeline.java

+ final var useRuntimeInfoPlist =
+ Files.isDirectory(env.resolvedLayout().runtimeDirectory().resolve("bin")) ||
+ app.isRuntime();
+


The writeRuntimeInfoPlist() action is executed for the MacBuildApplicationTaskID.RUNTIME_INFO_PLIST and MacBuildApplicationTaskID.COPY_RUNTIME_INFO_PLIST tasks. MacBuildApplicationTaskID.RUNTIME_INFO_PLIST doesn't have dependencies. This means it can be executed before the runtime files are available in the output app image. Currently, this is not a problem because jpackage executes tasks synchronously, but eventually it will run them asynchronously, and the Files.isDirectory(...) test will fail.

The plist template should be selected at the configuration phase, not the bundling phase. However, this will require knowing/guessing if jlink will create native commands in the "bin" directory from its command line. It is straightforward if it boils down to testing whether it has the "--strip-native-commands" option. This is what we do when we test if jlink options are correct for the Apple App Store, right? Test for "--strip-native-commands" in DeployParams.java

I'd suggest adding boolean jdk.jpackage.internal.model.RuntimeBuilder.withNativeCommands() method. Then the code will be:

final var useRuntimeInfoPlist = app.isRuntime() || app.runtimeBuilder().orElseThrow().withNativeCommands();

Alternatively, you may add the BuildApplicationTaskID.RUNTIME task as a dependency for the MacBuildApplicationTaskID.RUNTIME_INFO_PLIST task. However, this is a less preferable solution.

sashamatveev · 2025-10-30T23:27:06Z

Why? jpackage picks the same file from the resource directory regardless of whether the "bin" subdirectory is present in the runtime. The "bin" subdirectory in the runtime only affects the default plist resource that jpackage uses.

Not sure. I agree we should not mix testing this fix with CustomInfoPListTest.

sashamatveev added 5 commits October 20, 2025 19:26

8363980: [macos] Add JDK specific keys/values to Info.plist of embedd…

a5f4cba

…ed runtime

Merge remote-tracking branch 'upstream/master' into JDK-8363980

ca63e69

Merge remote-tracking branch 'upstream/master' into JDK-8363980

535c442

Merge remote-tracking branch 'upstream/master' into JDK-8363980

f3721a3

8363980: [macos] Add JDK specific keys/values to Info.plist of embedd…

04c9a0e

…ed runtime [v2]

openjdk bot added the core-libs core-libs-dev@openjdk.org label Oct 29, 2025

openjdk bot added the rfr Pull request is ready for review label Oct 29, 2025

alexeysemenyukoracle reviewed Oct 29, 2025

View reviewed changes

alexeysemenyukoracle suggested changes Oct 29, 2025

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

8363980: [macos] Add JDK specific keys/values to Info.plist of embedded runtime #28033

8363980: [macos] Add JDK specific keys/values to Info.plist of embedded runtime #28033

sashamatveev commented Oct 29, 2025 •

edited by openjdk bot

Loading

bridgekeeper bot commented Oct 29, 2025

openjdk bot commented Oct 29, 2025

openjdk bot commented Oct 29, 2025

sashamatveev commented Oct 29, 2025

mlbridge bot commented Oct 29, 2025

alexeysemenyukoracle Oct 29, 2025

sashamatveev Oct 30, 2025

alexeysemenyukoracle Oct 30, 2025 •

edited

Loading

alexeysemenyukoracle commented Oct 29, 2025

alexeysemenyukoracle Oct 29, 2025

sashamatveev Oct 30, 2025

alexeysemenyukoracle Oct 30, 2025

alexeysemenyukoracle Oct 30, 2025 •

edited

Loading

alexeysemenyukoracle commented Oct 29, 2025

alexeysemenyukoracle Oct 29, 2025

sashamatveev commented Oct 30, 2025

Labels

2 participants


		final var app = env.app();

		// We should use full runtime info plist for standalone runtime and for

8363980: [macos] Add JDK specific keys/values to Info.plist of embedded runtime #28033

Are you sure you want to change the base?

8363980: [macos] Add JDK specific keys/values to Info.plist of embedded runtime #28033

Conversation

sashamatveev commented Oct 29, 2025 • edited by openjdk bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Progress

Issue

Reviewing

bridgekeeper bot commented Oct 29, 2025

openjdk bot commented Oct 29, 2025

openjdk bot commented Oct 29, 2025

sashamatveev commented Oct 29, 2025

mlbridge bot commented Oct 29, 2025

Webrevs

alexeysemenyukoracle Oct 29, 2025

Choose a reason for hiding this comment

sashamatveev Oct 30, 2025

Choose a reason for hiding this comment

alexeysemenyukoracle Oct 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

alexeysemenyukoracle commented Oct 29, 2025

alexeysemenyukoracle Oct 29, 2025

Choose a reason for hiding this comment

sashamatveev Oct 30, 2025

Choose a reason for hiding this comment

alexeysemenyukoracle Oct 30, 2025

Choose a reason for hiding this comment

alexeysemenyukoracle Oct 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

alexeysemenyukoracle commented Oct 29, 2025

alexeysemenyukoracle Oct 29, 2025

Choose a reason for hiding this comment

sashamatveev commented Oct 30, 2025

Labels

2 participants

sashamatveev commented Oct 29, 2025 •

edited by openjdk bot

Loading

alexeysemenyukoracle Oct 30, 2025 •

edited

Loading

alexeysemenyukoracle Oct 30, 2025 •

edited

Loading