On 6/26/07, Stepan Mishura wrote:
On 6/25/07, Ivan Popov wrote:
> I see two failures of jdktools tests, which are not new ones.
>
Ivan,
Thanks for prompt failure analysis. I'm going to add HoldEventsTest to
exclude list for intermittent failures on Linux, increase timeout and
re-run tests.
Ops, sorry ... I've updated the exclude list without getting agreement
from another committer. Please let me know if there any objections for
updating the exclude list.
Thanks,
Stepan.
> Test org.apache.harmony.jpda.tests.jdwp.VirtualMachine.HoldEventsTest
> is known as intemittently failing on Windows (HARMONY-3508) and now it
> fails on Linux too. It make sense to move this test to common exclude
> list.
>
> Test mentioned as vmcarsh
> (org.apache.harmony.jpda.tests.jdwp.ThreadReference.FramesTest) failed
> due to exceeded timeout for tests run. You may want to increase
> default timeout (900000 milliseconds), e.g, specifying
> -Dhy.test.timeout=1200000 .
>
> Thanks.
> Ivan
>
> On 6/25/07, Stepan Mishura <[EMAIL PROTECTED]> wrote:
> > On 6/25/07, Mikhail Loenko <[EMAIL PROTECTED]> wrote:
> > > 2007/6/25, Stepan Mishura <[EMAIL PROTECTED]>:
> > > > On 6/25/07, Mark Hindess wrote:
> > > > >
> > > > > On 25 June 2007 at 13:59, "Stepan Mishura" <[EMAIL PROTECTED]> wrote:
> > > > > > On 6/24/07, Mikhail Loenko <[EMAIL PROTECTED]> wrote:
> > > > > > >
> > > > > > > We have passed our code freeze date for M2
> > > > > > >
> > > > > >
> > > > > > Mikhail,
> > > > > >
> > > > > > Just to be clear - M1 milestone published snapshots include build
for
> > > > > > Windows x86, Linux (libstdc++ v5 and libstdc++ v6) x86 and Windows
> > > > > > x86_64.
> > > > >
> > > > > Do you think it would be possible to produce source snapshots? The
> > > > > Apache release FAQ (at http://www.apache.org/dev/release.html ) says:
> > > > >
> > > > > The Apache Software Foundation produces open source software. All
> > > > > releases are in the form of the source materials needed to make
> > > > > changes to the software being released. In some cases,
binary/bytecode
> > > > > packages are also produced as a convenience to users that might
> > > > > not have the appropriate tools to build a compiled version of the
> > > > > source. In all such cases, the binary/bytecode package must have
> > > > > the same version number as the source release and may only add
> > > > > binary/bytecode files that are the result of compiling that version
of
> > > > > the source code release.
> > > > >
> > > > > Currently binaries are our primary artifact. I appreciated that it
may
> > > > > be a little late to try to correct this for this release, but I think
it
> > > > > is important that we try to correct this before we get too comfortable
> > > > > with the current release process.
> > > > >
> > > >
> > > > I'd suggest to switch to using source snapshots right after M2.
> > > >
> > > > > It should now be possible to do:
> > > > >
> > > > > ant bundle_src
> > > > > mkdir /tmp/build
> > > > > tar -C /tmp/build -xzf
target/apache-harmony-src-r550411-snapshot.tar.gz
> > > > > cd /tmp/build/harmony-src-550411
> > > > > ant -Dauto.fetch=true
> > > > >
> > > > > which would seem to me to be more in-keeping with the Apache release
> > > > > guidelines.
> > > > >
> > > > > On this subject, I'd like to permission to commit a patch to correct
the
> > > > > top-level directory name in the source tar.gz/zip files from:
> > > > >
> > > > > harmony-src-550411
> > > > >
> > > > > to:
> > > > >
> > > > > apache-harmony-src-r550411
> > > > >
> > > >
> > > > I'm OK with it.
> > >
> > > will it require rebuild of the release candidate?
> > >
> >
> > I think it is not required.
> >
> > OK, I've built and uploaded milestone candidates (r550333) for
> > - Linux x86/x86_64 libstdc++ v5 (and v6 for x86 is in progress)
> > - Windows x86/x86_64
> >
> > The snapshots are available from "snapshots v5" page[1] (or can be
> > taken from dir [2])
> >
> > [1] http://harmony.apache.org/snapshots_v5.html
> > [2] http://people.apache.org/builds/harmony/snapshots/r550333/
> >
> > -Stepan.
> >
> > > Thanks,
> > > Mikhail
> > >
> > > >
> > > > Thanks,
> > > > Stepan.
> > > >
> > > > > The format of the archive names changed over time and these have
become
> > > > > inconsistent.
> > > > >
> > > > > Regards,
> > > > > Mark.
> > > > >
> > > > > > I assume that we still aimed to x86 architecture and I need to build
> > > > > > milestone candidates for:
> > > > > > - Windows x86
> > > > > > - Linux x86. BWT, again for both libstdc++ versions?
> > > > > >
> > > > > > And what about x86_64?
> > > > > >
> > > > > > As I said M1 includes Windows x86_64. Should we publish them to let
> > > > > > the community test them to see how good they are?
> > > > > >
> > > > > > Thanks,
> > > > > > Stepan.
