Re: Intermittent failure for PR18699 test case

[TJ Laurenzo <tlaurenzo at gmail dot com>]

It's a Debian Sarge install on a dual hyperthreaded Xeon server.  With
four effective processors, this is certainly the type of configuration
that would bring out latent race conditions.

When I get a chance I'll try to reproduce this and figure out what is going on.

On 9/22/05, Bryce McKinlay <mckinlay@redhat.com> wrote:
> TJ Laurenzo wrote:
>
> >FYI -
> >I was just running the testsuite to verify some patches and
> >experienced an intermittent failure:
> >   FAIL: PR18699 execution - gij test
> >The log shows a "WARNING: Timed out" just prior to this.
> >The bug in question is marked resolved but is related to threading
> >issues which would explain the intermittent results.
> >
> >So far, I've run the libjava testsuite three times and experienced the
> >problem once.  The machine didn't have any load on it other than
> >running the testsuite, so I think it was legitimately hung for that
> >one time.
> >
> >
> What hardware/software configuration is this? I don't recall seeing this
> fail, but you're correct in that this is the type of test that can fail
> intermittently if there is some kind of race condition somewhere.
>
> If you are able to reproduce the hang (perhaps by running the test case
> manually), the best way to diagnose it would be to attach in gdb and get
> a backtrace for each thread.
>
> Bryce
>
>