1.13.0 Release Imminent - JamVM update to 2013-11-18 revision

[Andrew]

----- Original Message -----
> 2013-12-19 13:36, Andrew skrev:
> > The changes proposed for 1.13.0 are now nearly all complete:
> > 
> > http://icedtea.classpath.org/bugzilla/show_bug.cgi?id=1272
> > 
> > so the plan is to make a release over the next few days, hopefully
> > before the new year break.  Any later and we risk hitting the next round
> > of security updates planned for mid-January [0]
> > 
> > Please make any final changes ASAP.
> > 
> > [0] http://www.oracle.com/technetwork/topics/security/alerts-086861.html
> > 
> > Thanks,
> > 
> 
> I would like to update JamVM for the icedtea 1.13 release to the 2013-11-18
> revision, patch attached.
> For icedtea 1.13 this update will bring in the following JamVM bug-fixes and
> improvements:
> 
> - JVM_NewMultiArray: element class may be an array
>   http://git.berlios.de/cgi-bin/cgit.cgi/jamvm/commit/?id=ec18fb9e49e62dce16c5094ef1527eed619463aa
> 
> - jtreg test java/lang/reflect/Array/ExceedMaxDim
>   http://git.berlios.de/cgi-bin/cgit.cgi/jamvm/commit/?id=0c1e419fa29feee086e157a15eb24763ae92e52f
> 
> - OpenJDK: Fix Local/Anonymous class checks
>   http://git.berlios.de/cgi-bin/cgit.cgi/jamvm/commit/?id=123e1cf425f01fd3a7d9ced3d4ff032850b98dc2
> 
> - classlibMarkThreadTerminated should return Object*
>   http://git.berlios.de/cgi-bin/cgit.cgi/jamvm/commit/?id=e5eba542ddc649ef387a7ab580c6a9575b9510c5
> 
> - Fix race-condition in constant-pool resolution code
>   http://git.berlios.de/cgi-bin/cgit.cgi/jamvm/commit/?id=56e3c750558281127202289e38cac0aadf874845
> 
> - Race condition in setting up imethod table
>   http://git.berlios.de/cgi-bin/cgit.cgi/jamvm/commit/?id=9ec80dac5bf7551b1c95649d2f9b4b6edca59755
> 
> - GC: Minor performance improvement
>   http://git.berlios.de/cgi-bin/cgit.cgi/jamvm/commit/?id=8f9f9b7b55f7d9b647cb8c733837d2c547851f70
> 
> Test results:
> https://gist.github.com/xranby/817621fa14862158d8a3 32bit X86 Ubuntu 12.04
> No regressions observed for hotspot and langtools.
> The JDK test still stops in the middle of the make check-jdk run although it
> is now able to run more than 1000 tests
> compared to previous record of ~200 before stopping which is a major
> improvement.
> 
> Cheers
> Xerxes
> 

I think the new entries should go at the bottom of the JamVM NEWS section (as they were added last).
I also don't see a need to repeat them all in the ChangeLog.

Fine with those changes.
-- 
Andrew :)

Free Java Software Engineer
Red Hat, Inc. (http://www.redhat.com)

PGP Key: 248BDC07 (https://keys.indymedia.org/)
Fingerprint = EC5A 1F5E C0AD 1D15 8F1F  8F91 3B96 A578 248B DC07