Stepping in debugger switches to interpretation mode

[S A]

On Fri, 2024-05-31 at 14:44 +0200, Maksim Zuev wrote:
> Dear Sir/Madam,
>
> I encountered a problem while debugging the code. I am attaching the
> reproducer to this email in the Main.java file.
>
> When running it with the debugger without stepping, the application
> runs in less than a second (see jdb output in the jdb_run.txt file).
> However, after performing a single step, the application is running
> in interpretation mode, becoming very slow (see jdb output in the
> jdb_step.txt file).
>
> I assume running in the interpreter mode, as I
> see InterpreterRuntime::post_method_exit
> calls in the profiler.
>
> Could you please help me figure out what causes the application to
> run in the interpreter mode? Is this a bug or an expected behavior?
> Are there any ways to work around this issue?

There is also a RHEL ticket, maybe more people can comment in the RHEL
tracker, than on the JDK ticket:

https://issues.redhat.com/browse/RHEL-3498

There is more related discussion too.

Best regards,
Simeon

On Fri, 31 May 2024 at 18:49, Severin Gehwolf <sgehwolf at redhat.com> wrote:

> On Fri, 2024-05-31 at 14:44 +0200, Maksim Zuev wrote:
> > Dear Sir/Madam,
> >
> > I encountered a problem while debugging the code. I am attaching the
> > reproducer to this email in the Main.java file.
> >
> > When running it with the debugger without stepping, the application
> > runs in less than a second (see jdb output in the jdb_run.txt file).
> > However, after performing a single step, the application is running
> > in interpretation mode, becoming very slow (see jdb output in the
> > jdb_step.txt file).
> >
> > I assume running in the interpreter mode, as I
> > see InterpreterRuntime::post_method_exit
> > calls in the profiler.
> >
> > Could you please help me figure out what causes the application to
> > run in the interpreter mode? Is this a bug or an expected behavior?
> > Are there any ways to work around this issue?
>
> I think this is:
> https://bugs.openjdk.org/browse/JDK-8229012
>
> Thanks,
> Severin
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.openjdk.org/pipermail/serviceability-dev/attachments/20240531/12603cdd/attachment.htm>