On Fri, 8 Nov 2024 09:44:01 GMT, Alan Bateman <al...@openjdk.org> wrote:
>> Severin Gehwolf has updated the pull request incrementally with six >> additional commits since the last revision: >> >> - Mandy's feedback >> - Remove "jlink.runtime.linkable" property from VMProps >> >> It's no longer used in any tests. >> - Fix JLinkDedupTestBatchSizeOne.java >> - Fix plugins/IncludeLocalesPluginTest.java >> >> Similar to GenerateJLIClassesPluginTest.java, the default module path >> doesn't need to be explicitly specified. >> - Fix plugins/GenerateJLIClassesPluginTest.java test >> >> The test doesn't need the default module path and default jmods >> generated from the helper. Using this approach makes it work for >> linkable run-time images as well. >> - Simplify runtimeImage tests > > test/jtreg-ext/requires/VMProps.java line 28: > >> 26: import java.lang.module.ModuleFinder; >> 27: import java.lang.module.ModuleReader; >> 28: import java.lang.module.ModuleReference; > > I assume they aren't needed now. True. > test/langtools/tools/javac/plugin/AutostartPlugins.java line 33: > >> 31: * jdk.jlink >> 32: * @build toolbox.ToolBox toolbox.JavacTask toolbox.JarTask >> 33: * @run main/othervm AutostartPlugins > > Why do the two javac plugin tests changed to /othervm? I assume they fail > with a JDK built with --generate-linkable-runtime but I can't immediately see > why. jtreg patches the `java.base` module and those langtools tests *link* from the run-time image (in such a config). Having a patched JDK module is not supported (and I don't think we ever should). Running in a separate JVM doesn't have this problem. See https://github.com/openjdk/jdk/pull/14787/commits/d8e1e834508589725adb8d70acb862a1270678ca. Does that make sense? ------------- PR Review Comment: https://git.openjdk.org/jdk/pull/14787#discussion_r1834057484 PR Review Comment: https://git.openjdk.org/jdk/pull/14787#discussion_r1834056602
