Bug 199591
| Summary: | [exp-run] openjdk8 as default jdk | ||||||
|---|---|---|---|---|---|---|---|
| Product: | Ports & Packages | Reporter: | Steve Wills <swills> | ||||
| Component: | Ports Framework | Assignee: | Steve Wills <swills> | ||||
| Status: | Closed FIXED | ||||||
| Severity: | Affects Only Me | CC: | glewis, marino, portmgr, truckman | ||||
| Priority: | --- | ||||||
| Version: | Latest | ||||||
| Hardware: | Any | ||||||
| OS: | Any | ||||||
| See Also: | https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=201317 | ||||||
| Bug Depends on: | 203061, 203123, 203124, 203125, 203142, 203143, 203495, 203506, 203511, 203513 | ||||||
| Bug Blocks: | |||||||
| Attachments: |
|
||||||
Take for exp-run Exp-run results: http://package18.nyi.freebsd.org/jail.html?mastername=101amd64-default-PR199591 New failures: + {"origin"=>"databases/libgda5-jdbc", "pkgname"=>"libgda5-jdbc-5.2.2", "phase"=>"build", "errortype"=>"missing_header"} + {"origin"=>"databases/mysql-connector-java", "pkgname"=>"mysql-connector-java-5.1.35", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"devel/hadoop2", "pkgname"=>"hadoop2-2.6.0", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"devel/java-util-concurrent", "pkgname"=>"java-util-concurrent-1.3.4_2", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"devel/luajava", "pkgname"=>"luajava-1.1_2", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"devel/umlgraph", "pkgname"=>"umlgraph-5.6_1", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"java/eclipse-ecj", "pkgname"=>"eclipse-ecj-3.8.2", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"java/jakarta-commons-collections", "pkgname"=>"jakarta-commons-collections-3.2.1", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"java/junit", "pkgname"=>"junit-4.11_1", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"lang/racket", "pkgname"=>"racket-6.1_2", "phase"=>"build", "errortype"=>"process_failed"} + {"origin"=>"lang/racket-minimal", "pkgname"=>"racket-minimal-6.1_1", "phase"=>"build", "errortype"=>"process_failed"} + {"origin"=>"lang/x10", "pkgname"=>"x10-2.2.1_3", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"mail/davmail", "pkgname"=>"davmail-4.5.1", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"math/jakarta-commons-math", "pkgname"=>"jakarta-commons-math-2.2_1", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"net-im/jitsi", "pkgname"=>"jitsi-2.3", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"net/openmq", "pkgname"=>"openmq-4.5.2.p.1,1", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"net/openmq-client", "pkgname"=>"openmq-client-4.5.2.p.1,1", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"print/pdf-renderer", "pkgname"=>"pdf-renderer-0.9.1,1", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"security/keynote", "pkgname"=>"keynote-2.3_1", "phase"=>"patch", "errortype"=>"???"} + {"origin"=>"security/quantis", "pkgname"=>"quantis-v13.11.08", "phase"=>"configure", "errortype"=>"???"} + {"origin"=>"www/helma", "pkgname"=>"helma-1.7.0_1", "phase"=>"build", "errortype"=>"???"} Failure logs: http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/libgda5-jdbc-5.2.2.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/mysql-connector-java-5.1.35.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/hadoop2-2.6.0.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/java-util-concurrent-1.3.4_2.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/luajava-1.1_2.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/umlgraph-5.6_1.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/eclipse-ecj-3.8.2.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/jakarta-commons-collections-3.2.1.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/junit-4.11_1.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/racket-6.1_2.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/racket-minimal-6.1_1.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/x10-2.2.1_3.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/davmail-4.5.1.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/jakarta-commons-math-2.2_1.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/jitsi-2.3.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/openmq-4.5.2.p.1,1.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/openmq-client-4.5.2.p.1,1.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/pdf-renderer-0.9.1,1.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/keynote-2.3_1.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/quantis-v13.11.08.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-04-22_19h31m00s/logs/errors/helma-1.7.0_1.log Note that the following ports make some other ports skipped: java/junit: 13 skipped java/jakarta-commons-collections: 5 skipped devel/hadoop2: 5 skipped print/pdf-renderer: 3 skipped databases/mysql-connector-java: 1 skipped The racket failures may be random failures that are unrelated to openjdk so is the next step to limit most of those ports to openjdk 7? By the way, junit is failing on openjdk7 currently for a different reason, and DF also is having trouble with libgda5-jdbc on openjdk7 is the effort to switch to openjdk8 by default dead? if not, what is the status? Who is leading the charge? This needs to get done. I'll start looking at it later this week, but obviously the more the merrier. Created a patch for mysql-connector-java (see related bug) A commit references this bug: Author: glewis Date: Fri Jul 3 23:34:08 UTC 2015 New revision: 391286 URL: https://svnweb.freebsd.org/changeset/ports/391286 Log: . Mark this for deprecation. These are early versions of what current JDKs have included in the java.util.concurrent package by default since Java 5. PR: 199591 Changes: head/devel/java-util-concurrent/Makefile The error for databases/libgda5-jdbc makes it look like it would be broken for any version of openjdk. I don't think this is worth worrying about. Marked devel/java-util-concurrent for deprecation since it should just die. java/junit has been updated to 4.12 and it's method of building has been changed. It now works. devel/hadoop2 fails even if I restrict it to Java 7. However, I think that is because Maven will run using Java 8 with the switch to that as the default version and its execution of javadoc (where the build failure occurs) will use Java 8 too, regardless of what version of Java the Hadoop build wants to use. I haven't absolutely confirmed that, but the error is identical and that is how Maven looks to work to me. There is a newer version of Hadoop (2.7.0) which may support Java 8 and be a viable path forward there. However, the hadoop2 port looks non-trivial to update (at least to me), so this will likely take some effort. https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=201335 submitted for jakarta-commons-collections https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=201347 submitted for mail/davmail Several of these failures are due to doclint being on by default for Java 8 when running javadoc. I think we'd get quite a few ports fixed if we were able to turn that off during port builds. The way to do so is to add -Xdoclint:none to the command line when invoking javadoc. What I'm considering is a flag that ports could set that would turn off doclint that we'd pass through to javadoc. This likely means changes to javavmwrapper since I don't believe a lot of these ports use $JAVADOC that bsd.java.mk sets but instead directly invoke javadoc themselves. https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=201352 submitted for math/jakarta-commons-math The failure for security/keynote looks to be either transient or have nothing to do with the Java 8 change. Take again Exp-run results on 10.1 amd64: http://package18.nyi.freebsd.org/build.html?mastername=101amd64-default-PR199591&build=2015-09-10_05h33m30s There are 15 new failures: + {"origin"=>"databases/hbase", "pkgname"=>"hbase-0.98.9", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"databases/libgda5-jdbc", "pkgname"=>"libgda5-jdbc-5.2.4", "phase"=>"build", "errortype"=>"missing_header"} + {"origin"=>"devel/luajava", "pkgname"=>"luajava-1.1_2", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"devel/umlgraph", "pkgname"=>"umlgraph-5.6_1", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"editors/openoffice-4", "pkgname"=>"apache-openoffice-4.1.1_13", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"graphics/jogl", "pkgname"=>"jogl-1.1.1_5", "phase"=>"package", "errortype"=>"???"} + {"origin"=>"java/eclipse-ecj", "pkgname"=>"eclipse-ecj-3.8.2", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"lang/x10", "pkgname"=>"x10-2.2.1_3", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"math/scilab", "pkgname"=>"scilab-5.5.2", "phase"=>"configure", "errortype"=>"configure_error"} + {"origin"=>"net-im/jitsi", "pkgname"=>"jitsi-2.3", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"net/openmq", "pkgname"=>"openmq-4.5.2.p.1,1", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"net/openmq-client", "pkgname"=>"openmq-client-4.5.2.p.1,1", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"print/pdf-renderer", "pkgname"=>"pdf-renderer-0.9.1,1", "phase"=>"build", "errortype"=>"???"} + {"origin"=>"security/quantis", "pkgname"=>"quantis-v13.11.08", "phase"=>"configure", "errortype"=>"???"} + {"origin"=>"www/helma", "pkgname"=>"helma-1.7.0_1", "phase"=>"build", "errortype"=>"???"} Failure logs: http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-09-10_05h33m30s/logs/errors/hbase-0.98.9.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-09-10_05h33m30s/logs/errors/libgda5-jdbc-5.2.4.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-09-10_05h33m30s/logs/errors/luajava-1.1_2.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-09-10_05h33m30s/logs/errors/umlgraph-5.6_1.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-09-10_05h33m30s/logs/errors/apache-openoffice-4.1.1_13.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-09-10_05h33m30s/logs/errors/jogl-1.1.1_5.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-09-10_05h33m30s/logs/errors/eclipse-ecj-3.8.2.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-09-10_05h33m30s/logs/errors/x10-2.2.1_3.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-09-10_05h33m30s/logs/errors/scilab-5.5.2.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-09-10_05h33m30s/logs/errors/jitsi-2.3.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-09-10_05h33m30s/logs/errors/openmq-4.5.2.p.1,1.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-09-10_05h33m30s/logs/errors/openmq-client-4.5.2.p.1,1.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-09-10_05h33m30s/logs/errors/pdf-renderer-0.9.1,1.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-09-10_05h33m30s/logs/errors/quantis-v13.11.08.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-09-10_05h33m30s/logs/errors/helma-1.7.0_1.log Fix for luajava is in https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=203061 Fix for print/pdf-renderer is in PR 203123 fix for devel/umlgraph is in bug 203124 fix for jitsi is in bug 203125 A commit references this bug: Author: kwm Date: Wed Sep 16 11:10:24 UTC 2015 New revision: 397054 URL: https://svnweb.freebsd.org/changeset/ports/397054 Log: Allow libgda5-jdbc to build with openjdk8. PR: 199591 (openjdk8 by default exp-run) Submitted by: swills@ Changes: head/databases/libgda5/Makefile I believe all the issues are solved now. Shall we have one more exp-run or should we just do it? Exp-run results on 10.1 i386: http://package18.nyi.freebsd.org/build.html?mastername=101i386-default-PR199591&build=2015-10-16_18h43m01s There are 5 failures, but the devel/spark one is frequently seen The 4 new ones: http://package18.nyi.freebsd.org/data/101i386-default-PR199591/2015-10-16_18h43m01s/logs/errors/x10-2.2.1_3.log http://package18.nyi.freebsd.org/data/101i386-default-PR199591/2015-10-16_18h43m01s/logs/errors/tn5250j-0.6.0_3.log http://package18.nyi.freebsd.org/data/101i386-default-PR199591/2015-10-16_18h43m01s/logs/errors/apache-openoffice-4.1.1_16.log http://package18.nyi.freebsd.org/data/101i386-default-PR199591/2015-10-16_18h43m01s/logs/errors/apache-openoffice-devel-4.1.1707648_1,4.log I believe the openoffice-devel error was not there a few weeks ago but it seems it came back from openoffice. When I wrote "came back from openoffice-4" I meant this commit: r398790 | truckman | 2015-10-08 01:09:38 +0000 (Thu, 08 Oct 2015) | 10 lines Temporarily downgrade the openoffice-devel port from the upstream trunk branch to 4.1.2-RC1 in order to give the latter wider exposure before 4.1.2 is released. Exp-run results on 10.1 amd64: http://package18.nyi.freebsd.org/build.html?mastername=101amd64-default-PR199591&build=2015-10-17_05h36m38s 4 failures: http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-10-17_05h36m38s/logs/errors/x10-2.2.1_3.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-10-17_05h36m38s/logs/errors/tn5250j-0.6.0_3.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-10-17_05h36m38s/logs/errors/apache-openoffice-4.1.1_16.log http://package18.nyi.freebsd.org/data/101amd64-default-PR199591/2015-10-17_05h36m38s/logs/errors/apache-openoffice-devel-4.1.1707648_1,4.log The remaining broken ones have been marked broken, I don't know how to fix them. If anyone else does, that would be great. So should we let any of those hold up the switch of the default? Or should we go ahead? I'm currently testing a fix for OpenOffice. I hope to have something to commit by tomorrow. (In reply to Don Lewis from comment #30) Awesome, thanks for the update! A commit references this bug: Author: truckman Date: Sun Oct 18 08:18:26 UTC 2015 New revision: 399605 URL: https://svnweb.freebsd.org/changeset/ports/399605 Log: Patch the source with r1697228, r1697237, r1697247, r1697306, and r1697312 from upstream trunk (with a few tweaks to compensate for some other upstream changes not included here) to unbreak the build with OpenJDK8. Revert r399536 to unmark broken. PR: 199591 (openjdk8 by default exp-run) Changes: head/editors/openoffice-4/Makefile head/editors/openoffice-4/files/patch-openjdk8 A commit references this bug: Author: truckman Date: Sun Oct 18 08:20:07 UTC 2015 New revision: 399606 URL: https://svnweb.freebsd.org/changeset/ports/399606 Log: Patch the source with r1697228, r1697237, r1697247, r1697306, and r1697312 from upstream trunk (with a few tweaks to compensate for some other upstream changes not included here) to unbreak the build with OpenJDK8. Revert r399537 to unmark broken. PR: 199591 (openjdk8 by default exp-run) Changes: head/editors/openoffice-devel/Makefile head/editors/openoffice-devel/files/patch-openjdk8 (In reply to Steve Wills from comment #31) editors/openoffice-4 and editors/openoffice-devel should be fixed now. I confirm openoffice is fixed (no more failures at http://package18.nyi.freebsd.org/build.html?mastername=101amd64-default-PR199591&build=2015-10-18_08h57m25s ) Committed in r399693. |
Created attachment 155820 [details] default jdk change patch please run exp-run with openjdk8 as default