RFR: 8297936: Use reachabilityFence to manage liveness in ClassUnload tests

[David Holmes]

On Tue, 14 Feb 2023 09:02:09 GMT, Afshin Zafari <duke at openjdk.org> wrote:

> ### Description
> Instead of using static instances or using an object in print to keep it alive, `java.lang.ref.Reference.reachabilityFence(Object)` is used to keep an object alive and avoid class unloading.
> 
> ### Tests
> linux-x64-debug, macosx-aarch64-debug, mach5 tier1-5

Changes requested by dholmes (Reviewer).

test/hotspot/jtreg/runtime/ClassUnload/ConstantPoolDependsTest.java line 35:

> 33:  * @compile p2/c2.java MyDiffClassLoader.java
> 34:  * @run driver jdk.test.lib.helpers.ClassFileInstaller jdk.test.whitebox.WhiteBox
> 35:  * @run main/othervm -Xbootclasspath/a:. -Xmn8m -XX:+UnlockDiagnosticVMOptions -Xlog:class+unload -Xcomp -XX:+WhiteBoxAPI ConstantPoolDependsTest

As Coleen states, do not force the tests to run in Xcomp mode.

test/hotspot/jtreg/runtime/ClassUnload/KeepAliveObject.java line 68:

> 66:         ClassUnloadCommon.failIf(!isAlive, "should be alive");
> 67:     }
> 68:     Reference.reachabilityFence(o);

Please add comment:

// Don't let `o` get prematurely reclaimed by the GC.

test/hotspot/jtreg/runtime/ClassUnload/UnloadTest.java line 52:

> 50:  */
> 51: public class UnloadTest {
> 52:     // Using a global static field to keep the object live in -Xcomp mode.

This comment needs deleting now.

test/hotspot/jtreg/runtime/ClassUnload/UnloadTest.java line 79:

> 77:         ClassUnloadCommon.failIf(!wb.isClassAlive(className), "should still be live");
> 78: 
> 79:         Reference.reachabilityFence(o);

Again add a comment please.

test/hotspot/jtreg/runtime/ClassUnload/UnloadTest.java line 108:

> 106:         ClassUnloadCommon.failIf(!wb.isClassAlive(className), "should still be live");
> 107: 
> 108:         Reference.reachabilityFence(o);

Again please add comment.

-------------

PR: https://git.openjdk.org/jdk/pull/12552