failed gerrit builds

classic Classic list List threaded Threaded
5 messages Options
Terrence Enger Terrence Enger
Reply | Threaded
Open this post in threaded view
|

failed gerrit builds

Greetings all,

In addition to my failure to run Junit tests locally, as I have
described in a previous message, my patch in gerrit
<https://gerrit.libreoffice.org/60048> for bug 119625 is failing
jenkins builds:

(*) patch set 5, gerrit_windows, build # 16069
    <https://ci.libreoffice.org/job/gerrit_windows/16069/>

    That page summarizes:

        Aborted by user anonymous

    The console output ends with:

        [build SLC] sc
        [build SLC] sd
        Build was aborted
        Aborted by unknown
        Finished: ABORTED

(*) patch set 6, gerrit_windows, build # 16088
    <https://ci.libreoffice.org/job/gerrit_windows/16088/>

    That page summarizes:

        Identified problems
            No identified problem
            Uncategorized build failure. If it is generic enough, please add a suitable Cause for it.

    The console output contains:

        [build CHK] registry
        /usr/bin/cp: cannot remove 'C:/cygwin/home/tdf/lode/jenkins/workspace/gerrit_windows/instdir/program/types.rdb': Device or resource busy
        make[1]: *** [C:/cygwin/home/tdf/lode/jenkins/workspace/gerrit_windows/solenv/gbuild/Package.mk:35: C:/cygwin/home/tdf/lode/jenkins/workspace/gerrit_windows/instdir/program/types.rdb] Error 1

I have time available, but I have no idea about how to dig into these
problems.  Moreover, I do not know how my changes since patch set 4,
which built successfully on jenkins, can be making the problem.
Guidance welcome.

Thank you, all, for your attention.
Terry.

_______________________________________________
LibreOffice mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/libreoffice
Christian Lohmaier-3 Christian Lohmaier-3
Reply | Threaded
Open this post in threaded view
|

Re: failed gerrit builds



Terrence Enger <[hidden email]> schrieb am Fr., 21. Sep. 2018, 13:16:


        Aborted by user anonymous

Not related to your changes. The build was aborted manually, likely because of the other issues listed further down (also not related to your changes)
 
        [build CHK] registry
        /usr/bin/cp: cannot remove 'C:/cygwin/home/tdf/lode/jenkins/workspace/gerrit_windows/instdir/program/types.rdb': Device or resource busy

Device or resource busy is an issue with the build machine, issue on the OS/filesystem level and not related to your changes.
Likely a previous build didn't finish properly and left junk (leftover processes) behind.

Ciao
Christian

_______________________________________________
LibreOffice mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/libreoffice
Terrence Enger Terrence Enger
Reply | Threaded
Open this post in threaded view
|

Re: local build not running java unit tests

In reply to this post by Terrence Enger
Stephen's advice here solves the problem I presented.  I am trying to
move this email to another
thread, which discusses difficulties more
like the next difficulty I face here.

On Fri, 2018-09-21 at 09:35 +0200, Stephan Bergmann wrote:
> On 20/09/2018 20:56, Terrence Enger wrote
... that his local builds did not run Junit tests ...
>
> Is it just a question of plain `make` (running just a few tests, mostly
> CppUnit ones) vs. `make check` (running all tests)?  Some of the
> Gerrit/Jenkins bots run the latter.

Exactly so!  Thank you.

On my machine, the test ran successfully.  I do not know how to dig
deeper.

Meanwhile, I shall try again.  Maybe the problem will go away.  (I
hate it when a problem just goes away.  Almost, I hope that the
problem persists.  Almost.  Sigh!  The test *did* run successfully for
the previous patch set.)

Thank you, Stephen.
Terry.


_______________________________________________
LibreOffice mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/libreoffice
sberg sberg
Reply | Threaded
Open this post in threaded view
|

Re: local build not running java unit tests

On 21/09/2018 16:00, Terrence Enger wrote:
> On my machine, the test ran successfully.  I do not know how to dig
> deeper.
>
> Meanwhile, I shall try again.  Maybe the problem will go away.  (I
> hate it when a problem just goes away.  Almost, I hope that the
> problem persists.  Almost.  Sigh!  The test *did* run successfully for
> the previous patch set.)

<https://ci.libreoffice.org/job/gerrit_linux_clang_dbgutil/15160/> is
JunitTest_starmath_unoapi failing due to

> Creating: sm.SmEditAccessible
> LOG> Log started 15.08.2018 - 20:30:40
> warn:vcl.opengl:15069:16426:vcl/opengl/x11/X11DeviceInfo.cxx:356: unknown vendor => blocked
> warn:vcl.gdi:15069:15069:vcl/headless/svpgdi.cxx:256: non default depth bitmap, slow convert, upscale the input
> LOG> ImplementationName SmEditAccessible
> Environment created
> LOG> Log started 15.08.2018 - 20:30:42
> checking: [sm.SmEditAccessible::com::sun::star::accessibility::XAccessibleComponent] is iface: [com.sun.star.accessibility.XAccessibleComponent] testcode: [ifc.accessibility._XAccessibleComponent]
> LOG> Execute: containsPoint()
> LOG> Upper bound of box containsPoint point (0,0) - OK
> LOG> Lower bound of box containsPoint point (0,138) - OK
> LOG> Left bound of box containsPoint point (0,0) - OK
> LOG> Right bound of box containsPoint point (476,0) - OK
> LOG> Outer upper and lower bounds contain no component points - OK
> LOG> Outer left and right bounds contain no component points - OK
> Method containsPoint() finished with state OK
> LOG> containsPoint(): COMPLETED.OK
>
> LOG> Execute: getAccessibleAtPoint()
> LOG> Checking child with bounds (0,0),(2,18): 41,(Paragraph: 0 ):
> LOG> StateType containsPoint SHOWING: true
> LOG> finding the point which lies on the component
> LOG> Child found at point (1,9) - OK
> LOG> The children found is not the same
> LOG> Expected:
> LOG> Description:  Paragraph: 0 sum hat a
> LOG> Found:
> LOG> Description:  Paragraph: 0 sum hat a
> LOG> ... FAILED
> LOG> No children found at point (-1,-1) - OK
> Method getAccessibleAtPoint() finished with state FAILED
> LOG> getAccessibleAtPoint(): COMPLETED.FAILED

Some of the tests are known to be fragile and unfortunately failing
occasionally, for one reason or another.  This sm.SmEditAccessible test
is in that infamous set (and nobody knows why it occasionally fails, I
think; maybe some timing issue).
_______________________________________________
LibreOffice mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/libreoffice
Terrence Enger Terrence Enger
Reply | Threaded
Open this post in threaded view
|

[solved] failed gerrit builds

In reply to this post by Christian Lohmaier-3
On Fri, 2018-09-21 at 13:24 +0200, Christian Lohmaier wrote:

Device or resource busy is an issue with the build machine, issue on the
OS/filesystem level and not related to your changes.
Likely a previous build didn't finish properly and left junk (leftover
processes) behind.

Thank you. My next attempt (patch set 7) built successfully.


_______________________________________________
LibreOffice mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/libreoffice