Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [papyrus-rt-dev] Why is the Java8 JRE in Hudson a 32-bit build?

Hi,

 

I’m still a little in the « jet lag », so my bad for not been clear.

My only (useful) point is: There are a bunch of errors logged (included the one you reference) in the working tests log (we all agree that green tests with exception in log is a bad thing, but just stating a fact here) [1].

 

So it’s probably not the root of your current problem.

 

In any case, the best is to comment the rcptt tests until the merge of this neon branch in master

 

Regards,

Benoit

1 : https://hudson.eclipse.org/papyrus-rt/job/Papyrus-RT-Gerrit-master/ws/source/releng/org.eclipse.papyrusrt.rcptt/target/runner-workspace/.metadata/.log

 

 

De : papyrus-rt-dev-bounces@xxxxxxxxxxx [mailto:papyrus-rt-dev-bounces@xxxxxxxxxxx] De la part de Christian Damus
Envoyé : lundi 4 avril 2016 17:28
À : papyrus-rt developer discussions <papyrus-rt-dev@xxxxxxxxxxx>
Objet : Re: [papyrus-rt-dev] Why is the Java8 JRE in Hudson a 32-bit build?

 

Hi, Benoit,

 

See some replies in-line, below.

 

Thanks,

 

Christian

 

 

 

On 4 April, 2016 at 11:15:22, MAGGI Benoit (benoit.maggi@xxxxxx) wrote:

Hi Christian,

 

My point was that it’s NOT the consequence of the log you previously referenced.

 

What makes you say that?  I traced through the RCPTT source code, and it seems pretty clear that the version-map isn’t getting the “platform version” inferred from the target’s installed SWT bundle and that the validation of this entry is what causes the test runner to abort.

 

There is exactly the same log in the working tests [1] [2] (which is probably not good)

 

Then how is it that those test builds pass?  The builds all look green to me on Hudson.

 

Another strange thing, the 2 builds are using different runner:

Master org.eclipse.rcptt.runner:rcptt.runner:zip:2.0.2
Neon : org.eclipse.rcptt.runner:rcptt.runner:zip:2.1.0-M4b

 

Yes, because the 2.0.x RCPTT doesn’t support Eclipse Neon platform, from what I understood in their discussion forum, so I configured the Neon patch to use the latest available Neon-compatible test runner milestone.  But I could be wrong about that; maybe the 2.0.2 runner would work just as well (or poorly, as seems to be the case)

 

In SysML 1.4 we chose to put p2/product/rcptt test in the “product” maven profile to be able to call them “on demand” [3]

ð  You can probably comment the rcptt tests for the moment and reapply them before merging the branch

 

Do you mean before merging the neon branch to master?  That makes sense.  But, then we will have to have solved the RCPTT test execution problem.  What has changed in RCPTT or the maven configuration recently? The tests did execute successfully in the Papyrus-RT builds based on Mars, no?  What could have changed that they are now failing?

 

 

Regards,

Benoit

1:https://hudson.eclipse.org/papyrus-rt/job/Papyrus-RT-Gerrit-master/ws/source/releng/org.eclipse.papyrusrt.rcptt/target/runner-workspace/.metadata/.log

2: !MESSAGE Detected potential problems in target platform No Name

        Installation /jobs/genie.papyrus-rt/Papyrus-RT-Gerrit-master/workspace/source/releng/org.eclipse.papyrusrt.rcptt/target/aut/PapyrusRT Default Configuration

        Directory /jobs/genie.papyrus-rt/Papyrus-RT-Gerrit-master/workspace/source/releng/org.eclipse.papyrusrt.rcptt/target/aut/PapyrusRT/plugins

Env: null/null/null/null

JRE: null

Args: null/null

Implicit: null

Handle: 1459775718420.target

3: https://git.eclipse.org/c/papyrus/org.eclipse.papyrus-sysml.git/tree/releng/pom.xml

 

 

 

De : papyrus-rt-dev-bounces@xxxxxxxxxxx [mailto:papyrus-rt-dev-bounces@xxxxxxxxxxx] De la part de Christian Damus
Envoyé : lundi 4 avril 2016 16:52
À : papyrus-rt developer discussions <papyrus-rt-dev@xxxxxxxxxxx>
Objet : Re: [papyrus-rt-dev] Why is the Java8 JRE in Hudson a 32-bit build?

 

Hi, Benoit,

 

Yes, that last exception is the validation of the inference of the “platform version” that failed to happen at the point in the log that I referenced.

 

Looking through the RCPTT source (class AUTInformation), I see that the “platform version” is supposed to be inferred from the version of the org.eclipse.swt bundle that’s installed in the PDE Target.  In the build workspace, I see the 3.104.1 version installed and, according to the putSwtVersion method in the AUTInformation class, that should result in the “platform version” being recorded as 4.5.  But, for some reason, that’s not happening.  So, the validation later complains that it “Failed to detect platform version”.

 

I’m baffled at this point, and feel like just commenting-out the RCPTT Runner in my patch so that we can finally make progress on porting to Neon.  Because there are a lot of glitches in the UI that need to be worked on to complete the Neon port, which we can’t start until this Gerrit patch is merged, including at least:

  • weird positioning of ports in composite diagrams
  • Model Explorer customizations for protocols not working at all
  • Model Explorer showing RT diagrams nested within themselves

Also, switching everybody over to the Neon platform might be easier if we can provide an Oomph setup for developers:

Cheers,

 

Christian

 

 

On 4 April, 2016 at 10:37:27, MAGGI Benoit (benoit.maggi@xxxxxx) wrote:

Hi Christian,

 

You can find RcpTT logs here[1]

I think the problematic exception is the last one [2]

Which is strange because the target seems present [3].

ð I don’t know what is happening here.

ð Did you manage to run the tests on your mac?(Add macosx in the platform)

 

Some remarks:

-       Product Name papyrusRt (neon) vs PapyrusRT (master)

-       Even working RcpTT tests have still a bunch of warnings (see [4])

-       Moving to eclipse target Plaform would maybe help (see [5], failing due to a wrong localization of pom.xml?)

 

Regards,

Benoit

 

1 : https://hudson.eclipse.org/papyrus-rt/job/PapyrusRT-Gerrit-for-Neon/ws/source/releng/org.eclipse.papyrusrt.rcptt/target/runner-workspace/.metadata/.log

2 : !ENTRY org.eclipse.osgi 4 0 2016-04-04 10:01:23.033

!MESSAGE Application error

!STACK 1

org.eclipse.core.runtime.CoreException: Invalid eclipse target platform: No name /jobs/genie.papyrus-rt/PapyrusRT-Gerrit-for-Neon/workspace/source/releng/org.eclipse.papyrusrt.rcptt/target/aut/papyrusRT

        at org.eclipse.rcptt.internal.launching.ext.Q7TargetPlatformInitializer.getInfo(Q7TargetPlatformInitializer.java:239)

        at org.eclipse.rcptt.internal.launching.ext.Q7TargetPlatformInitializer.initialize(Q7TargetPlatformInitializer.java:100)

       at org.eclipse.rcptt.internal.launching.ext.Q7TargetPlatformManager.createTargetPlatform(Q7TargetPlatformManager.java:129)

        at org.eclipse.rcptt.runner.util.TargetPlatformChecker.initializeTargetPlatform(TargetPlatformChecker.java:112)

        at org.eclipse.rcptt.runner.util.TargetPlatformChecker.initAndCheckTargetPlatform(TargetPlatformChecker.java:70)

        at org.eclipse.rcptt.runner.HeadlessRunner.performCoolThings(HeadlessRunner.java:48)

        at org.eclipse.rcptt.runner.HeadlessRunnerApp.start(HeadlessRunnerApp.java:54)

        at org.eclipse.equinox.internal.app.EclipseAppHandle.run(EclipseAppHandle.java:196)

        at org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.runApplication(EclipseAppLauncher.java:134)

        at org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.start(EclipseAppLauncher.java:104)

        at org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:380)

        at org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:235)

        at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)

        at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)

        at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)

        at java.lang.reflect.Method.invoke(Method.java:497)

        at org.eclipse.equinox.launcher.Main.invokeFramework(Main.java:669)

        at org.eclipse.equinox.launcher.Main.basicRun(Main.java:608)

        at org.eclipse.equinox.launcher.Main.run(Main.java:1515)

        at org.eclipse.equinox.launcher.Main.main(Main.java:1488)

Contains: Failed to detect platform version

3: https://hudson.eclipse.org/papyrus-rt/job/PapyrusRT-Gerrit-for-Neon/ws/source/releng/org.eclipse.papyrusrt.rcptt/target/aut/papyrusRT/

4: https://hudson.eclipse.org/papyrus-rt/job/Papyrus-RT-Gerrit-master/ws/source/releng/org.eclipse.papyrusrt.rcptt/target/runner-workspace/.metadata/.log

5: https://git.eclipse.org/r/#/c/65272/15

 

De : papyrus-rt-dev-bounces@xxxxxxxxxxx [mailto:papyrus-rt-dev-bounces@xxxxxxxxxxx] De la part de Christian Damus
Envoyé : lundi 4 avril 2016 16:06
À : papyrus-rt developer discussions <papyrus-rt-dev@xxxxxxxxxxx>; Céline JANSSENS <celine.janssens@xxxxxxxxxxx>
Objet : Re: [papyrus-rt-dev] Why is the Java8 JRE in Hudson a 32-bit build?

 

Thanks, Céline.

 

Now, after configuring the Neon gerrit job to use the 64-bit Java8, we get past the problem of not being able to load the SWT native library.  With another configuration change to enable Xvnc for the test execution so that the X11/Gtk environment is available to SWT, I can report that my Gerrit patch [1] for porting Papyrus-RT to Papyrus Neon M6 now only has the problem that I originally got stuck on in my local test builds:  RCPTT cannot detect the target platform:

 

!MESSAGE Detected potential problems in target platform No Name
        Installation /jobs/genie.papyrus-rt/PapyrusRT-Gerrit-for-Neon/workspace/source/releng/org.eclipse.papyrusrt.rcptt/target/aut/papyrusRT Default Configuration
        Directory /jobs/genie.papyrus-rt/PapyrusRT-Gerrit-for-Neon/workspace/source/releng/org.eclipse.papyrusrt.rcptt/target/aut/papyrusRT/plugins
Env: null/null/null/null
JRE: null
Args: null/null
Implicit: null
Handle: 1459778481221.target

 

I understand that you are doing some work on the RCPTT execution.  Do you recognize this?  I haven’t even been able to find the source code in the RCPTT Git repository for the component that attempts to infer the target platform, to guess at how to give it what it needs.  I can find the code that bombs the execution on finding that the target platform wasn’t resolved, but not the code that is responsible for that resolution step.

 

Cheers,

 

Christian

 

 

 

On 4 April, 2016 at 03:35:48, Céline JANSSENS (celine.janssens@xxxxxxxxxxx) wrote:

Hi Everyone,

The JDK 64bits has been added to the installed JDKs. 

I hope it will help for the Neon installation.

Regards

Céline

Le 31/03/2016 11:27, Céline JANSSENS a écrit :

Hello everyone,

I have created the ticket :

https://bugs.eclipse.org/bugs/show_bug.cgi?id=490763

To ask a JRE 8 on a 64bit HIPP instance.

Regards

Céline

Le 31/03/2016 10:09, Céline JANSSENS a écrit :

I do not have such access.

As far as I know Remi doesn't have it as well.
The quickest would be to open a ticket.

Regards
Céline

Le 30/03/2016 20:10, Christian Damus a écrit :

Thanks, Camille,

 

Surely somebody on this mailing list has the requisite admin access to the HIPP.  I agree that we can have no need for the 32-bit VM.

 

Christian

 

On 30 March, 2016 at 12:14:37, LETAVERNIER Camille (camille.letavernier@xxxxxx) wrote:

FWIW, on the Papyrus Hudson instance, I had to manually configure the JDK-8 variable to the following:

jdk1.8.0-latest = /shared/common/jdk1.8.0_x64-latest

I had to do this because, at the time, it wasn't pre-configured by webmasters. I don't know how much this has changed now. If someone has admin rights on the Papyrus-RT HIPP, then he can configure something similar. Otherwise, you need to open a ticket on Bugzilla so that Webmasters can take care of that for you

HTH,
Camille

De :papyrus-rt-dev-bounces@xxxxxxxxxxx [papyrus-rt-dev-bounces@xxxxxxxxxxx] de la part de Christian Damus [give.a.damus@xxxxxxxxx]
Envoyé : mercredi 30 mars 2016 17:26
À : papyrus-rt developer discussions
Objet : Re: [papyrus-rt-dev] Why is the Java8 JRE in Hudson a 32-bit build?

hi, Peter,

 

I don’t know what the XULRunner is for. But in any case my work-around doesn’t work because 32-bit SWT fails to resolve  libpthread.  So I’m stuck until I can get a 64-bit Java 8.

 

Christian

_______________________________________________
papyrus-rt-dev mailing list
papyrus-rt-dev@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/papyrus-rt-dev

 

_______________________________________________
papyrus-rt-dev mailing list
papyrus-rt-dev@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/papyrus-rt-dev

 

--

 

 

 

Céline JANSSENS
Software Engineer
+33 (0)2 44 47 23 23

 

Mail : cej@xxxxxxxxxxx

6 rue Léonard De Vinci - BP 0119 - 53001 LAVAL Cedex - FRANCE
www.all4tec.net

 

_______________________________________________
papyrus-rt-dev mailing list
papyrus-rt-dev@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/papyrus-rt-dev

 

--

 

 

 

Céline JANSSENS
Software Engineer
+33 (0)2 44 47 23 23

 

Mail : cej@xxxxxxxxxxx

6 rue Léonard De Vinci - BP 0119 - 53001 LAVAL Cedex - FRANCE
www.all4tec.net

 

_______________________________________________
papyrus-rt-dev mailing list
papyrus-rt-dev@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/papyrus-rt-dev

 

--

 

 

 

Céline JANSSENS
Software Engineer
+33 (0)2 44 47 23 23

 

Mail : cej@xxxxxxxxxxx

6 rue Léonard De Vinci - BP 0119 - 53001 LAVAL Cedex - FRANCE
www.all4tec.net

_______________________________________________
papyrus-rt-dev mailing list
papyrus-rt-dev@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/papyrus-rt-dev

_______________________________________________
papyrus-rt-dev mailing list
papyrus-rt-dev@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/papyrus-rt-dev

_______________________________________________ 
papyrus-rt-dev mailing list 
papyrus-rt-dev@xxxxxxxxxxx 
To change your delivery options, retrieve your password, or unsubscribe from this list, visit 
https://dev.eclipse.org/mailman/listinfo/papyrus-rt-dev 


Back to the top