8257876: Avoid Reference.isEnqueued in tests #1691

kimbarrett · 2020-12-08T09:52:51Z

Please review this change that eliminates the use of Reference.isEnqueued by
tests. There were three tests using it:

vmTestbase/gc/gctests/ReferencesGC/ReferencesGC.java
vmTestbase/gc/gctests/WeakReferenceGC/WeakReferenceGC.java
jdk/java/lang/ref/ReferenceEnqueue.java

In each of them, some combination of using Reference.refersTo and
ReferenceQueue.remove with a timeout were used to eliminate the use of
Reference.isEnqueued.

I also cleaned up ReferencesGC.java in various respects. It contained
several bits of dead code, and the failure checks were made stronger.

Testing:
mach5 tier1
Locally (linux-x64) ran all three tests with each GC (including Shenandoah).

/label core-libs, hotspot-gc

Progress

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

Issue

JDK-8257876: Avoid Reference.isEnqueued in tests

Reviewers

Mandy Chung (@mlchung - Reviewer) ⚠️ Review applies to e87206a
Thomas Schatzl (@tschatzl - Reviewer) ⚠️ Review applies to 0171056

Download

$ git fetch https://git.openjdk.java.net/jdk pull/1691/head:pull/1691
$ git checkout pull/1691

bridgekeeper · 2020-12-08T09:57:30Z

👋 Welcome back kbarrett! 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 · 2020-12-08T09:58:16Z

@kimbarrett Unknown command labels - for a list of valid commands use /help.

openjdk · 2020-12-08T09:58:35Z

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

core-libs
hotspot-gc

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.

mlbridge · 2020-12-08T10:13:17Z

Webrevs

openjdk · 2020-12-08T17:30:33Z

@kimbarrett 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:

8257876: Avoid Reference.isEnqueued in tests

Reviewed-by: mchung, tschatzl

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 1 new commit pushed to the master branch:

4a839e9: 8256459: java/net/httpclient/ManyRequests.java and java/net/httpclient/LineBodyHandlerTest.java fail infrequently with java.net.ConnectException: Connection timed out: no further information

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.

tschatzl · 2020-12-09T13:22:32Z

I'm not able to put this in the appropriate place using the github UI:

[pre-existing] The topWeakReferenceGC.java description at the top describes that the test calls System.gc() explicitly to trigger garbage collections at the end. It does not. Maybe this could be weasel-worded around like in the other cases in that text.

tschatzl · 2020-12-09T13:23:47Z

test/jdk/java/lang/ref/ReferenceEnqueue.java

-                }
-
-                Thread.sleep(100);
+                enqueued = (queue.remove(100) == ref);


The code does not catch InterruptedException like it does in the other files.

I understand that the test code previously just forwarded the InterruptedException if it happened in the Thread.sleep() call too. So this may only be an exiting issue and please ignore this comment.
Not catching InterruptedException here only seems to be a cause for unnecessary failure. Then again, it probably does not happen a lot.

Nothing in the test calls Thread.interrupt(), so there isn't a risk of
failure due to not handling that exception in some "interesting" way. But
InterruptedException must be "handled" somehow, because it's a checked
exception. That's already dealt with by the run() method declaring that it
throws that type, and main declaring that it throws Exception. The other
tests modified in this change don't take that approach (just let it
propagate out through main), instead wrapping the interruptable calls in
try/catch, though again just to satisfy the requirement that a checked
exception must be statically verified to be handled, even though there
aren't going to be any thrown.

tschatzl · 2020-12-09T13:26:04Z

test/hotspot/jtreg/vmTestbase/gc/gctests/ReferencesGC/ReferencesGC.java

                }

-                for (int i = 0; i < (int) (RANGE * RATIO); i++) {
+                int REMOVE = (int) (RANGE * RATIO);


These two constants could be factored out as static finals to match the casing.

I'm making REMOVE and RETAIN statics, near RANGE and RATIO. (Meant to do that before, but forgot.) They can't be final though, because RANGE and RATIO aren't final, and can be set from command line arguments. So they'll get initialized in parseArgs.

kimbarrett · 2020-12-10T08:35:03Z

[pre-existing] The topWeakReferenceGC.java description at the top describes that the test calls System.gc() explicitly to trigger garbage collections at the end. It does not. Maybe this could be weasel-worded around like in the other cases in that text.

There are a lot of things much more wrong with that comment. Doing more GCs
doesn't cause more enqueues to happen. The "non-deterministic" enqueuing is
just a race. The GC adds references to the pending list. The reference
processing thread transfers references from the pending list to their
associated queue (if any). The test code is racing with that. The change
to use Reference.remove with a timeout eliminates all that, and one GC
should be. Addressing all that would be a substantial rewrite of this
test though. Mind if I defer that to a new RFE?

tschatzl

Also good with deferring the changes to the comments and the move of the statics.

tschatzl · 2020-12-10T09:22:57Z

test/jdk/java/lang/ref/ReferenceEnqueue.java

-                }
-
-                Thread.sleep(100);
+                enqueued = (queue.remove(100) == ref);


kimbarrett · 2020-12-10T10:34:28Z

/integrate

openjdk · 2020-12-10T10:35:13Z

@kimbarrett Since your change was applied there has been 1 commit pushed to the master branch:

4a839e9: 8256459: java/net/httpclient/ManyRequests.java and java/net/httpclient/LineBodyHandlerTest.java fail infrequently with java.net.ConnectException: Connection timed out: no further information

Your commit was automatically rebased without conflicts.

Pushed as commit db5da96.

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

kimbarrett · 2020-12-10T12:06:38Z

Thanks for reviews @mlchung and @tschatzl

openjdk bot added core-libs core-libs-dev@openjdk.org hotspot-gc hotspot-gc-dev@openjdk.org labels Dec 8, 2020

kimbarrett added 3 commits December 8, 2020 05:01

update ReferencesGC test

db03ad5

update ReferenceQueue test

f954421

update WeakReferenceGC test

e87206a

kimbarrett force-pushed the no_isenqueued branch from 376a8f5 to e87206a Compare December 8, 2020 10:04

kimbarrett marked this pull request as ready for review December 8, 2020 10:10

openjdk bot added the rfr Pull request is ready for review label Dec 8, 2020

mlchung approved these changes Dec 8, 2020

View reviewed changes

openjdk bot added the ready Pull request is ready to be integrated label Dec 8, 2020

tschatzl suggested changes Dec 9, 2020

View reviewed changes

move REMOVE and RETAIN decls and init

0171056

tschatzl approved these changes Dec 10, 2020

View reviewed changes

test/jdk/java/lang/ref/ReferenceEnqueue.java

}

Thread.sleep(100);

enqueued = (queue.remove(100) == ref);

Copy link

Contributor

tschatzl Dec 10, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Okay.

Merge branch 'master' into no_isenqueued

d535534

openjdk bot closed this Dec 10, 2020

openjdk bot added integrated Pull request has been integrated and removed ready Pull request is ready to be integrated rfr Pull request is ready for review labels Dec 10, 2020

kimbarrett deleted the no_isenqueued branch December 10, 2020 12:07

8257876: Avoid Reference.isEnqueued in tests #1691

8257876: Avoid Reference.isEnqueued in tests #1691

kimbarrett commented Dec 8, 2020 •

edited by openjdk bot

bridgekeeper bot commented Dec 8, 2020

openjdk bot commented Dec 8, 2020

openjdk bot commented Dec 8, 2020

mlbridge bot commented Dec 8, 2020 •

edited

openjdk bot commented Dec 8, 2020 •

edited

tschatzl commented Dec 9, 2020

tschatzl Dec 9, 2020

tschatzl Dec 9, 2020

kimbarrett Dec 10, 2020

tschatzl Dec 10, 2020

tschatzl Dec 9, 2020

kimbarrett Dec 10, 2020

kimbarrett commented Dec 10, 2020

tschatzl left a comment

tschatzl Dec 10, 2020

kimbarrett commented Dec 10, 2020

openjdk bot commented Dec 10, 2020

kimbarrett commented Dec 10, 2020

8257876: Avoid Reference.isEnqueued in tests #1691

8257876: Avoid Reference.isEnqueued in tests #1691

Conversation

kimbarrett commented Dec 8, 2020 • edited by openjdk bot

Progress

Issue

Reviewers

Download

bridgekeeper bot commented Dec 8, 2020

openjdk bot commented Dec 8, 2020

openjdk bot commented Dec 8, 2020

mlbridge bot commented Dec 8, 2020 • edited

Webrevs

openjdk bot commented Dec 8, 2020 • edited

tschatzl commented Dec 9, 2020

tschatzl Dec 9, 2020

Choose a reason for hiding this comment

tschatzl Dec 9, 2020

Choose a reason for hiding this comment

kimbarrett Dec 10, 2020

Choose a reason for hiding this comment

tschatzl Dec 10, 2020

Choose a reason for hiding this comment

tschatzl Dec 9, 2020

Choose a reason for hiding this comment

kimbarrett Dec 10, 2020

Choose a reason for hiding this comment

kimbarrett commented Dec 10, 2020

tschatzl left a comment

Choose a reason for hiding this comment

tschatzl Dec 10, 2020

Choose a reason for hiding this comment

kimbarrett commented Dec 10, 2020

openjdk bot commented Dec 10, 2020

kimbarrett commented Dec 10, 2020

kimbarrett commented Dec 8, 2020 •

edited by openjdk bot

mlbridge bot commented Dec 8, 2020 •

edited

openjdk bot commented Dec 8, 2020 •

edited