[jdk11u-dev] RFR: 8313816: Accessing jmethodID might lead to spurious crashes [v2]

[Jaroslav Bachorik]

Hi Goetz,

I am working on a fix. It is a test issue when the new method I introduced
to WhiteBox calls
ClassLoaderDataGraph::do_unloading
and that will trigger an assert if not called in safepoint. Unfortunately,
the custom reg tests for tier2 I ran were not run with fastdebug but only
release so this issue escaped.

The fix should be as easy as wrapping the 'do_unloading' call in a
VM_WhiteBoxOperation. Also, I fixed our local GHA definition to run both
release and fastdebug for custom test selection so this does not repeat
again.

I apologize again for the inconvenience. I am going to file a JBS ticket to
track the test failure and propose a PR immediately.

Regards,

Jaroslav

On Sun, Dec 24, 2023 at 12:59 AM Lindenmaier, Goetz <
goetz.lindenmaier at sap.com> wrote:

> Hi Jaroslav,
>
> I backed out the change in our nightly testing and the issue
> is gone. So it's really caused by your change.
>
> Best regards,
>   Goetz.
>
>
> > -----Original Message-----
> > From: jdk-updates-dev <jdk-updates-dev-retn at openjdk.org> On Behalf Of
> > Amos Shi
> > Sent: Samstag, 23. Dezember 2023 00:02
> > To: jdk-updates-dev at openjdk.org
> > Subject: Re: [jdk11u-dev] RFR: 8313816: Accessing jmethodID might lead to
> > spurious crashes [v2]
> >
> > [Some people who received this message don't often get email from
> > ashi at openjdk.org. Learn why this is important at
> > https://aka.ms/LearnAboutSenderIdentification ]
> >
> > On Wed, 6 Dec 2023 10:38:56 GMT, Jaroslav Bachorik
> > <jbachorik at openjdk.org> wrote:
> >
> > >> Looks like a clean backport from the 17u PR, except for replacing
> nullptr
> > with NULL as you note, and the unused declaration of
> > metaspaceSharedRegionAlignment in WhiteBox.java. The latter doesn't seem
> > have a definition in 11u: if so, please remove it.
> > >
> > > @phohensee There are the additional changes in the testing that were
> > necessary. Although they are not affecting any existing code/tests.
> >
> > Hi @jbachorik,
> >
> > In  today's test result, the test case
> >
> `test/hotspot/jtreg/serviceability/jvmti/thread/GetStackTrace/GetStackTraceAn
> > dRetransformTest/GetStackTraceAndRetransformTest.java` is failing with
> the
> > following info:
> >
> > - _Unsupported internal testing APIs have been used._
> >
> >
> > ----------System.out:(20/1185)----------
> > # To suppress the following error report, specify this argument
> > # after -XX: or in .hotspotrc:
> SuppressErrorAt=/metadataOnStackMark.cpp:50
> > #
> > # A fatal error has been detected by the Java Runtime Environment:
> > #
> > #  Internal Error (/priv/jenkins/client-home/workspace/openjdk-11u-dev-
> > linux_x86_64-
> > dbg/jdk/src/hotspot/share/classfile/metadataOnStackMark.cpp:50),
> pid=65261,
> > tid=65413
> > #  assert(SafepointSynchronize::is_at_safepoint()) failed: sanity check
> > #
> > # JRE version: OpenJDK Runtime Environment (11.0.23) (fastdebug build
> > 11.0.23-internal+0-adhoc..jdk)
> > # Java VM: OpenJDK 64-Bit Server VM (fastdebug 11.0.23-internal+0-
> > adhoc..jdk, mixed mode, tiered, compressed oops, g1 gc, linux-amd64)
> > # Core dump will be written. Default location:
> > /priv/jvmtests/output_openjdk11_dev_dbgU_linuxx86_64/jtreg_hotspot_tier2
> > _work/JTwork/scratch/8/core.65261
> > #
> > Unsupported internal testing APIs have been used.
> >
> > # An error report file with more information is saved as:
> > #
> > /priv/jvmtests/output_openjdk11_dev_dbgU_linuxx86_64/jtreg_hotspot_tier2
> > _work/JTwork/scratch/8/hs_err_pid65261.log
> > #
> > # If you would like to submit a bug report, please visit:
> > #
> >
> https://eur03.safelinks.protection.outlook.com/?url=https%3A%2F%2Fbugrepo
> > rt.java.com%2Fbugreport%2Fcrash.jsp&data=05%7C02%7Cgoetz.lindenmaier%
> > 40sap.com%7Cd348179c4a08439f11d208dc03420cea%7C42f7676cf455423c8
> > 2f6dc2d99791af7%7C0%7C0%7C638388829421636554%7CUnknown%7CTWF
> > pbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI
> > 6Mn0%3D%7C3000%7C%7C%7C&sdata=xS617527CbTO5OPMg6uxqMvXEbAfy
> > nXDYVaWQGqbYzA%3D&reserved=0
> >
> >
> > Need your help to identify if this failure is caused by current PR or
> not.
> >
> > Thanks
> > Amos
> > A Happy Developer
> >
> > -------------
> >
> > PR Comment: https://git.openjdk.org/jdk11u-dev/pull/2321#issuecomment-
> > 1868118735
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.openjdk.org/pipermail/jdk-updates-dev/attachments/20231227/9c518c86/attachment-0001.htm>