Review Request 118603: Remove data race on boolean by removing qRegisterMetaType call altogether.
Kevin Ottens
ervin at kde.org
Sat Jun 7 16:46:12 UTC 2014
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/118603/#review59515
-----------------------------------------------------------
Ship it!
Ship It!
- Kevin Ottens
On June 7, 2014, 3:12 p.m., David Faure wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://git.reviewboard.kde.org/r/118603/
> -----------------------------------------------------------
>
> (Updated June 7, 2014, 3:12 p.m.)
>
>
> Review request for KDE Frameworks, Kevin Ottens, Michael Pyne, and Stephen Kelly.
>
>
> Repository: kcoreaddons
>
>
> Description
> -------
>
> Remove data race on boolean by removing qRegisterMetaType call altogether.
>
> Apparently the Q_DECLARE_METATYPE macro is enough to get rid of this runtime warning:
> QWARN: KJobTest::testProgressTracking() QSignalSpy::initArgs: Don't know how to handle 'KJob::Unit', use qRegisterMetaType to register it.
>
>
> Diffs
> -----
>
> src/lib/jobs/kjob.h fb7de853fc306ea938d19aba2729ff7713f9b7bb
> src/lib/jobs/kjob.cpp 33d894ad7233921dee77ab9d403016421a04978d
> src/lib/jobs/kjob_p.h c11a4d8ea8f4380ec8bc4da1cd88e1053042c691
>
> Diff: https://git.reviewboard.kde.org/r/118603/diff/
>
>
> Testing
> -------
>
> This showed up in the first place when starting to write a threaded unittest for kiocore. The first race was on KJobPrivate::_k_kjobUnitEnumRegistered.
>
> I tested that qRegisterMetaType isn't actually necessary anymore in Qt5, using kjobtest, as shown in commit log.
>
>
> Thanks,
>
> David Faure
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kde-frameworks-devel/attachments/20140607/1d4aeb0e/attachment.html>
More information about the Kde-frameworks-devel
mailing list