Thanks for looking at that for us, Mark.

I believe I've fixed up those files with bad EOFs now.

That sounds rather odd, the 724MB thing. I presume you are running with the library from the release i.e. 32 bit linux. I would be interested to see what the Harmony JVM produced - I'll
speak to you and arrange for you to place the dump somewhere.


I expect I'll do another release soon. I'd like to hear from Ant or Robert what we should do next.

Thanks,
    Stuart


Mark Hindess wrote:
In message<[email protected]>, Stuart Monteith writes:
Hello all,
      Steve and I have been working on getting the codebase ready for a
release. This is a non-trivial activity with lots of things to get right.

Given the state of the project as it is just now, we can't claim to be
doing a 1.0 release, instead I'm following the Milestone convention.

I'm not proposing that what I've assembled is submitted as a release
candidate just yet, as there is still work to do on it.

If you look here, you'll find the files that would make up a release:
      http://people.apache.org/~monteith/kato/apache-kato-M1-incubating-RC1/

I've also created a tag:

https://svn.apache.org/repos/asf/incubator/kato/tags/apache-kato-M1-incubating-RC1/


Obvious things that are wrong:
      apache-kato-M1-incubating-tomcat.demos - packages are undocumented
and missing some files.
      The documentation - it needs more work, but I'd be interested in
some cursory feedback.
      The CJVMTI agent needs work under Linux - it's missing some
information.
      The TCK is still under development.
      Some filtering needs to be applied to the rat reports.
      There are some spurious files, like readme.html.

I have found this a very useful exercise, as this is the first that
we've packaged the projects contents and seen what it actually includes,
so this should give everyone a good opportunity to have a look and see.
I had a quick look.  One odd thing I noticed was that several files
have mixed dos and unix line endings:

apache-kato-M1-incubating-src/kato.demo.tomcat/src/main/webapp/OOM.jsp
apache-kato-M1-incubating-src/kato.demo.tomcat/src/main/webapp/OOB.jsp
apache-kato-M1-incubating-src/kato.demo.tomcat/src/main/webapp/deadlock.jsp
apache-kato-M1-incubating-src/kato.demo.tomcat/src/main/webapp/hprofDump.jsp
apache-kato-M1-incubating-src/kato.hprof.api/pom.xml
apache-kato-M1-incubating-src/kato.native/kato.native.vmapi/src/main/native/katovmapi.c
apache-kato-M1-incubating-src/kato.native/kato.native.vmapi/tests/pom.xml
apache-kato-M1-incubating-src/kato.native/kato.native.vmapi/pom.xml

It is probably worth fixing these.

I also tried the cjvmti agent on linux/x86_64 with Harmony.  The first
time I hit break running HelloWorld it created a 724M cdmp file which
caused a crash in the reader.  I tried it again and got a much more
reasonable 5.7M file which could be read okay.  I've still got copies
of these if you want to take a look at them.

Regards,
  Mark.



--
Stuart Monteith
http://blog.stoo.me.uk/

Reply via email to