Review Request 122232: KConfig: fix using KSharedConfig in global object destructor.

[David Faure]

> On Feb. 21, 2015, 8:19 p.m., Matthew Dawson wrote:
> > Everything looked good, but when I tried to rerun the tests to make sure everything is ok the kconfig_in_global_object test I'm getting an abort with the same error message.  Backtrace:
> > ```
> > #5  0x00007ffff791b630 in qAppName () at kernel/qcoreapplication.cpp:541
> > #6  0x00007ffff78a9c76 in QLockFilePrivate::tryLock_sys (this=this at entry=0x608ec0) at io/qlockfile_unix.cpp:140
> > #7  0x00007ffff783c1ec in QLockFile::tryLock (this=<optimized out>, timeout=-1) at io/qlockfile.cpp:208
> > #8  0x00007ffff7ba17db in KConfigIniBackend::lock (this=0x608fe0) at /home/matthew/kde/kconfig/src/core/kconfigini.cpp:634
> > #9  0x00007ffff7b84ce0 in KConfigPrivate::lockLocal (this=0x608b40) at /home/matthew/kde/kconfig/src/core/kconfig.cpp:104
> > #10 0x00007ffff7b86aa9 in KConfig::sync (this=0x606ae0) at /home/matthew/kde/kconfig/src/core/kconfig.cpp:415
> > #11 0x00007ffff7b85c28 in KConfig::~KConfig (this=0x606ae0, __in_chrg=<optimized out>) at /home/matthew/kde/kconfig/src/core/kconfig.cpp:267
> > #12 0x00007ffff7b85c82 in KConfig::~KConfig (this=0x606ae0, __in_chrg=<optimized out>) at /home/matthew/kde/kconfig/src/core/kconfig.cpp:270
> > #13 0x0000000000401844 in Tester::~Tester (this=0x6030e8 <(anonymous namespace)::Q_QGS_globalTestObject::innerFunction()::holder>, __in_chrg=<optimized out>)
> >     at /home/matthew/kde/kconfig/autotests/kconfig_in_global_object.cpp:56
> > ```
> > Which seems to be second issue we found.  I'm not sure how to solve this issue yet, but I'd be happy to just document the issue and strike this test.
> > 
> > I only get this when the configuration file does not exist before the program executes, which may be why we missed it from the earlier patches.

Indeed. Good catch.

Fixed in Qt. https://codereview.qt-project.org/106953.


- David


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122232/#review76407
-----------------------------------------------------------


On Feb. 21, 2015, 7:58 p.m., David Faure wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://git.reviewboard.kde.org/r/122232/
> -----------------------------------------------------------
> 
> (Updated Feb. 21, 2015, 7:58 p.m.)
> 
> 
> Review request for KDE Frameworks and Matthew Dawson.
> 
> 
> Repository: kconfig
> 
> 
> Description
> -------
> 
> kconfig_in_global_object.cpp comes from kdelibs4support
> (after porting to Q_GLOBAL_STATIC)
> 
> ksharedconfig_in_global_object.cpp is now in kdelibs4 too
> (where it works) and reproduces Albert's KgDifficulty testcase.
> 
> 
> Diffs
> -----
> 
>   autotests/CMakeLists.txt b91f754b705fc87bb8b729bea72fbb5f7d427ace 
>   autotests/kconfig_in_global_object.cpp PRE-CREATION 
>   autotests/ksharedconfig_in_global_object.cpp PRE-CREATION 
>   src/core/kconfig.cpp 1da816faea5548d2f529755d6538a142a2193720 
>   src/core/ksharedconfig.h d0791e461804d421010bc5f82873371a4d3bc992 
> 
> Diff: https://git.reviewboard.kde.org/r/122232/diff/
> 
> 
> Testing
> -------
> 
> Both tests pass and the QCoreApplication::arguments warning (because called after qApp is destroyed) is gone.
> 
> 
> Thanks,
> 
> David Faure
> 
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kde-frameworks-devel/attachments/20150222/5c96452a/attachment.html>