8275037: Test vmTestbase/nsk/sysdict/vm/stress/btree/btree011/btree011.java crashes with memory exhaustion on Windows

[hseigel]

Please review this fix for JDK-8275037. The .../sysdict/vm/stress/btree tests were intermittently failing on Windows because invocations of javac by the tests were running out of memory. The fix sets the maximum heap to 1G when the tests invoke javac.

Additionally, the fix ignores NoClassDefFoundError exceptions thrown by Class.forName() because those exceptions can be caused by a running out of memory when initializing a class.

The fix was tested by running each .../sysdict/vm/stress test over 2000 times on Windows x64 debug builds. Additionally, each test was run hundreds of times on Linux x64, Linux aarch64, Macosx x64, and Macosx aarch64.

Thanks, Harold

---

Progress

• Change must not contain extraneous whitespace
• Commit message must refer to an issue
• Change must be properly reviewed

Issue

• JDK-8275037: Test vmTestbase/nsk/sysdict/vm/stress/btree/btree011/btree011.java crashes with memory exhaustion on Windows

Reviewers

• Coleen Phillimore (@coleenp - Reviewer)

Reviewing

Using git

Checkout this PR locally:
$ git fetch https://git.openjdk.java.net/jdk pull/6419/head:pull/6419
$ git checkout pull/6419

Update a local copy of the PR:
$ git checkout pull/6419
$ git pull https://git.openjdk.java.net/jdk pull/6419/head

Using Skara CLI tools

Checkout this PR locally:
$ git pr checkout 6419

View PR using the GUI difftool:
$ git pr show -t 6419

Using diff file

Download this PR as a diff file:
https://git.openjdk.java.net/jdk/pull/6419.diff

[bridgekeeper]

👋 Welcome back hseigel! 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]

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

• hotspot-runtime

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]

Webrevs

• 00: Full (91cb10d3)

[coleenp]

This looks good and trivial. I hope this makes these tests more reliable.

[openjdk]

@hseigel 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:

8275037: Test vmTestbase/nsk/sysdict/vm/stress/btree/btree011/btree011.java crashes with memory exhaustion on Windows

Reviewed-by: coleenp

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 8 new commits pushed to the master branch:

• 2af9e59: 8276139: TestJpsHostName.java not reliable, better to expand HostIdentifierCreate.java test
• e9934e1: 8277221: G1: Remove methods without implementations in G1CollectedHeap
• 9aa30de: 8275317: AArch64: Support some type conversion vectorization in SLP
• 08f65a5: 8277313: Validate header failed for test/jdk/java/net/httpclient/HeadTest.java
• 23e5117: 8276559: (httpclient) Consider adding an HttpRequest.Builder.HEAD method to build a HEAD request.
• a77d8dd: 8276787: Improve warning messages for -XX:+RecordDynamicDumpInfo
• 8ed384c: 8276609: Document setting property jdk.serialFilter to an invalid value throws ExceptionInInitializerError
• cddc6ce: 8275811: Incorrect instance to dispose

Please see this link for an up-to-date comparison between the source branch of this pull request and the master branch.
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.

[hseigel]

Thanks Coleen for reviewing this!

/integrate

[openjdk]

Going to push as commit 9f2f46e.
Since your change was applied there have been 8 commits pushed to the master branch:

• 2af9e59: 8276139: TestJpsHostName.java not reliable, better to expand HostIdentifierCreate.java test
• e9934e1: 8277221: G1: Remove methods without implementations in G1CollectedHeap
• 9aa30de: 8275317: AArch64: Support some type conversion vectorization in SLP
• 08f65a5: 8277313: Validate header failed for test/jdk/java/net/httpclient/HeadTest.java
• 23e5117: 8276559: (httpclient) Consider adding an HttpRequest.Builder.HEAD method to build a HEAD request.
• a77d8dd: 8276787: Improve warning messages for -XX:+RecordDynamicDumpInfo
• 8ed384c: 8276609: Document setting property jdk.serialFilter to an invalid value throws ExceptionInInitializerError
• cddc6ce: 8275811: Incorrect instance to dispose

Your commit was automatically rebased without conflicts.

[openjdk]

@hseigel Pushed as commit 9f2f46e.

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

[dholmes-ora]

Hi Harold,
What would the heap max be without -Xmx?
Thanks,
David

[hseigel]

Hi David,
The default heap size on my Linux system is 31G. When I replaced -J-Xmx1G in GenClassesBuilder.java with -J-XX:+PrintFlagsFinal,, and ran the tests using Mach5, it showed max heap sizes of about 12G for Linux 64, Mac OS x64, and Windows x64.
Harold

[dholmes-ora]

Sounds odd that an out-of-memory condition was fixed by giving the VM less memory. :) I can only assume that the default -Xms value was too large for the initial heap. But it still seems a bit odd.

Thanks.