harmony-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Vladimir Ivanov" <ivavladi...@gmail.com>
Subject Re: [testing][general] CC/CI reports failures
Date Mon, 14 May 2007 05:23:14 GMT
Sorry, forgot to copy test name:
classlib\trunk\build.xml:169: The following error occurred while
executing this line: classlib\trunk\make\build-test.xml:142: There
were test crashes:
classlib\trunk\build\test_report\TEST-tests.api.java.nio.charset.CharsetDecoderTest.xml

On 5/14/07, Vladimir Ivanov <ivavladimir@gmail.com> wrote:
> Thanks!
> Classlib tests passed for me on 3 platforms now. On Win64 classlib
> tests failed due to failure of test:
>
> It is intermittent failure and it is hard to reproduce :(
>
>  thanks, Vladimir
>
> On 5/14/07, Xiao-Feng Li <xiaofeng.li@gmail.com> wrote:
> > On 5/14/07, Vladimir Ivanov <ivavladimir@gmail.com> wrote:
> > > Hello everybody,
> > > the summary of CC/CI status available at
> > > http://www.harmonytest.org/upload/cc1.html and
> > > http://www.harmonytest.org/upload/cc2.html
> > >
> > > is:
> > > 1) the classlib test test_CtorLjava_io_ByteArrayInputStreamClass (from
> > > org.apache.harmony.security.tests.asn1.der.BerInputStreamTest)
> > > failed with java.lang.OutOfMemoryError on DRLVM+GCv5 on 4 platforms.
> > > Note this test passed on DRLVM GCv4.1
> >
> > Vladimir, I've reverted a commit that caused the failure. Now it
> > should have no problem.
> >
> > > 2) DRLVM test test_freeMemoryClass (from java.lang.RuntimeTest)
> > > intermittently failed with "junit.framework.AssertionFailedError:
> > > Runtime.freeMemory method should be sensitive to huge memory
> > > allocating!". Issue 3711 was created to track it.
> >
> > This should pass as well after the revert.
> >
> > > 3) DRLVM kernel test testGetContextClassLoader_MainClass (from
> > > java.lang.ThreadTest) intermittently failed on DRLVM with
> > > java.util.NoSuchElementException. Issue 3838 was created to track it.
> >
> > This one is not related to the reverted commit. It is suspected to be
> > an issue with the association mechanism between native thread and Java
> > thread. Still under investigation.
> >
> > > I suggest excluding these tests up to end of failures evaluation if
> > > this evaluation will take long time (say, more than 1 day).
> > > Is it OK?
> > >  thanks, Vladimir
> >
> > Basically I agree to exclude the third one for the moment.
> >
> > Thanks, xiaofeng
> >
> > --
> > http://xiao-feng.blogspot.com
> >
>

Mime
View raw message