You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@harmony.apache.org by Vladimir Beliaev <vl...@gmail.com> on 2007/12/21 15:31:24 UTC

[buildtest][eut] regression status for r605636

FYI: I've checked if there are regressions in EUT since M3.

The latest report I see is:
http://people.apache.org/~mloenko/snapshot_testing/script/r605636/index.html

*  windows x86: 440 issues in jdtcoremodel - can not reproduced locally
*  linux x86 / linux x86_64: looks like invalid run - all of GUI suites
failed - it looks like something wrong was with X-connection for these run.

The previous report is M4 relese one:
http://people.apache.org/~mloenko/snapshot_testing/script/r603534/index.html

* windows x86: 3 failures which are expected and out to EFL (works fine
already in r605636 testing)
* linux x86:
  - 1257 issues in jdtcorecompiler are not reproduced locally
  - 1 issue in jdttext is not reproduced locally
  - 6 issues in jdtuirefactoring are real ones caused by TMP/HOME
redefinition - put to EFL as Invalid Tests

So I see *no EUT regressions to work on*. There are still issues on Linux
x86_64 - with jdtdebug and coreresources suites which hang.

BTW, I've updated EUT related information (mainly its run intsruction) at:
http://wiki.apache.org/harmony/Eclipse_Unit_Tests_Pass_on_DRLVM - so you are
welcome to use it.

Thanks
Vladimir

Re: [buildtest][eut] regression status for r605636

Posted by Vladimir Beliaev <vl...@gmail.com>.

FYI:
http://people.apache.org/~mloenko/snapshot_testing/script/r606592/index.html


reports EUT Windows/x86 and Linux x86 as GREEN (with pass rate over 99.7%)
which is good.

reports EUT Linux x86_64 as RED (95.39% pass rate) with set of crashes. I've
double checked these crashes locally:
* org.eclipse.core.tests.resources.AutomatedTests - it is HARMONY-5303
* org.eclipse.jdt.core.tests.builder.BuilderTests - it is not reproduced
locally
* org.eclipse.jdt.debug.tests.AutomatedSuite - it is HARMONY-5305,
workaround is committed, must be "no crash" next testing
* org.eclipse.team.tests.ccvs.core.AllTests - it is not reproducible locally
(might be CVS access collision by several EUTs run by CC)
* org.eclipse.ui.tests.session.SessionTests - it is not reproducible locally

Thanks
Vladimir Beliaev



2007/12/21, Vladimir Beliaev <vl...@gmail.com>:
>
> FYI: I've checked if there are regressions in EUT since M3.
>
> The latest report I see is:
> http://people.apache.org/~mloenko/snapshot_testing/script/r605636/index.html
>
> *  windows x86: 440 issues in jdtcoremodel - can not reproduced locally
> *  linux x86 / linux x86_64: looks like invalid run - all of GUI suites
> failed - it looks like something wrong was with X-connection for these run.
>
> The previous report is M4 relese one:
> http://people.apache.org/~mloenko/snapshot_testing/script/r603534/index.html
>
> * windows x86: 3 failures which are expected and out to EFL (works fine
> already in r605636 testing)
> * linux x86:
>   - 1257 issues in jdtcorecompiler are not reproduced locally
>   - 1 issue in jdttext is not reproduced locally
>   - 6 issues in jdtuirefactoring are real ones caused by TMP/HOME
> redefinition - put to EFL as Invalid Tests
>
> So I see *no EUT regressions to work on*. There are still issues on Linux
> x86_64 - with jdtdebug and coreresources suites which hang.
>
> BTW, I've updated EUT related information (mainly its run intsruction) at:
> http://wiki.apache.org/harmony/Eclipse_Unit_Tests_Pass_on_DRLVM  - so you
> are welcome to use it.
>
> Thanks
> Vladimir
>