Backporting changeset from 7 to 6 for bugfix (original) (raw)

Joe Darcy joe.darcy at oracle.com
Thu Jun 17 01:02:57 PDT 2010

Andrew John Hughes wrote:

On 17 June 2010 00:59, Joe Darcy <joe.darcy at oracle.com> wrote:

Andrew John Hughes wrote:

On 16 June 2010 01:54, Joe Darcy <joe.darcy at oracle.com> wrote:

Hello. Andrew John Hughes wrote:

On 15 June 2010 16:30, Andrew John Hughes <ahughes at redhat.com> wrote:

On 15 June 2010 13:21, Andrew John Hughes <ahughes at redhat.com> wrote:

On 15 June 2010 01:10, Joe Darcy <joe.darcy at oracle.com> wrote:

Joe Darcy wrote:

Andrew John Hughes wrote:

On 11 June 2010 05:31, Joe Darcy <joe.darcy at oracle.com> wrote:

Joe Darcy wrote:

Andrew John Hughes wrote:

On 10 June 2010 02:02, Joe Darcy <joe.darcy at oracle.com> wrote:

Andrew Su wrote:

[snip] changeset: 2208:0137b5857c63 parent: 2183:90248595ec35 user: okutsu date: Wed Mar 10 14:32:47 2010 +0900 summary: 6932473: (tz) javazic produces incorrect SimpleTimeZone parameters with Sun<=7 Do these look ok?

Yes; also approved to go back. Since some Sun copyrighted files have leaked back in, we'll done another pass to clean up any such files before b20 goes final. Thanks, -Joe I think only a few more changes are needed to wrap up b20: * The various backports identified in the earlier message in this thread Done now. * Backport of 6957375 "java/net/ResponseCache getResponseCode and ResponseCacheTest fail after rebranding" I see this come in while I was backporting; had to merge. * One more pass of the Sun -> Oracle copyright script There are some new Zero bugfixes that could be backported for b20 (and would allow us to drop the local copy in IcedTea6). I've CCed Gary Benson who should be able to provide a list of bug IDs for backport. Other than that, we should be good to go with b20. Being able to drop the copy of Zero in IcedTea sounds worthwhile. After looking over the bugids and a webrev, I'd expect to approve those fixes going back. The Zero fixes are the last currently planned fixes for b20. -Joe Hello. I'd like to get b20 wrapped up in the next day or two; please provide a list of the Zeros bug and I webrev so I can look them over. Thanks, -Joe I was waiting for a response from Gary with the bug IDs, but looks like one isn't forthcoming. I'll compile a list from recent OpenJDK7 build drops and post that shortly. There's also http://mail.openjdk.java.net/pipermail/jdk6-dev/2010-June/001750.html which causes build breakage with IcedTea. -- Andrew :-) Free Java Software Engineer Red Hat, Inc. (http://www.redhat.com) Support Free Java! Contribute to GNU Classpath and the OpenJDK http://www.gnu.org/software/classpath http://openjdk.java.net PGP Key: 94EFD9D8 (http://subkeys.pgp.net) Fingerprint: F8EF F1EA 401E 2E60 15FA 7927 142C 2591 94EF D9D8 b87 == 6934089 Zero 32-bit/64kb page fix http://hg.openjdk.java.net/jdk7/jdk7/hotspot/rev/da06d1795d84 b90 === 6939180 Zero locking fix http://hg.openjdk.java.net/jdk7/jdk7/hotspot/rev/747d26efc5fa b92 === 6941224: Improved stack overflow handling for Zero http://hg.openjdk.java.net/jdk7/jdk7/hotspot/rev/f9271ff9d324 6944503: Improved Zero crash dump http://hg.openjdk.java.net/jdk7/jdk7/hotspot/rev/c544d979f886 6944473: 6941224 misses new files http://hg.openjdk.java.net/jdk7/jdk7/hotspot/rev/aa9c266de52a These are hs18 Zero fixes. hs18 also includes 6944028 "6940701 broke Zero," but that fix is not needed in OpenJDK 6 yet since we don't have the fix for 6940701 from hs18, as you note below. Approved to go back. b94 === 6939182: Zero JNI handles fix http://hg.openjdk.java.net/jdk7/jdk7/hotspot/rev/0c5b3cf3c1f5 6950178: Zero stack improvements http://hg.openjdk.java.net/jdk7/jdk7/hotspot/rev/348346af6676 6950617: Zero/Shark interface updates http://hg.openjdk.java.net/jdk7/jdk7/hotspot/rev/6cfbdb113e52 6951784: Zero deoptimizer changes http://hg.openjdk.java.net/jdk7/jdk7/hotspot/rev/22af4ce8dba1 All of these are in hs19 and are approved to go back. The bug 6960550 Missing semicolon in Zero is "fixed" in hs19, but not yet integrated. At your option, you may apply this fix to OpenJDK 6 now too. Ok, I only looked at integrated fixes. But this fix looks simple enough and as you approve, I've backported it too :-) I'm surprised it even builds without the semi-colon but I guess that's C++ for you ;-) b96 === 6951686: Using large pages on Linux prevents zero based compressed oops http://hg.openjdk.java.net/jdk7/jdk7/hotspot/rev/79bf863697eb Here's the webrev for those: http://cr.openjdk.java.net/~andrew/zero/webrev.05/ and the previous Nimbus one: http://cr.openjdk.java.net/~andrew/nimbus/webrev.07/ I approve the -source 6/-target 6 changes for Nimbus; I don't think a wholesale change to -source 6/-target 6 is necessary. Please use 6961536 "Nimbus code using @Override break OpenJDK/IcedBuild on ecj" for the push. The following were skipped due to the referenced changesets not being in OpenJDK6: 6939731: JSR 292 Zero build fix after 6934494 6944028: 6940701 broke Zero 6949830: 6939134 broke Zero 6949423: remove tagged stack interpreter for Zero http://hg.openjdk.java.net/jdk7/jdk7/hotspot/rev/68d6683eaef7 Depends on: changeset: 1426:2338d41fbd81 user: twisti date: Fri Apr 30 08:37:24 2010 -0700 summary: 6943304: remove tagged stack interpreter I've checked the bug database and all the Zero fixes in hs18 and hs19 are somehow addressed in this email thread. -- Andrew :-) Free Java Software Engineer Red Hat, Inc. (http://www.redhat.com) Support Free Java! Contribute to GNU Classpath and the OpenJDK http://www.gnu.org/software/classpath http://openjdk.java.net PGP Key: 94EFD9D8 (http://subkeys.pgp.net) Fingerprint: F8EF F1EA 401E 2E60 15FA 7927 142C 2591 94EF D9D8 Scratch 6951686; the mention of 'zero' mislead me. It's not a Zero fix. Let me know when I can push the rest, Cheers, Good to push! All pushed! Thanks, -Joe Thanks, Thanks for the pushes! With the integration of 6961870 "More rebranding fixes for templates/gpl-*-header files," which I just pushed, I believe that is the last change for b20. I will tag the repos accordingly pending successful results from a round of sanity building and testing. -Joe I don't want to drag b20 out any longer either, but I think it's probably worth doing a final sync of HotSpot to hs17b16, which was just announced: http://mail.openjdk.java.net/pipermail/jdk6-dev/2010-June/001781.html If you recall, we brought in 6955813: Fix incorrect Oracle rebranding headers from 6941466 already so we now have hs17b15 + that fix. I think it would be better to have OpenJDK6 builds use a known HotSpot tag, which syncing with hs17b16 would give us. It actually only means bringing in one more real fix: 6953058: G1: A bigapp crashes with SIGSEGV in compiled code http://hg.openjdk.java.net/hsx/hsx17/master/rev/d53d91cb378c which sounds important. What do you think?

Well, I suppose we can have one more sync with HotSpot for b20; you're approved to sync hs17b15 into OpenJDK 6.

Thanks,

-Joe

More information about the jdk6-dev mailing list