[Development] How to track down CI problems

Fält Simo Simo.Falt at theqtcompany.com
Fri Oct 24 08:15:53 CEST 2014

You should test the patch against the target branch, dev not 5.4. Also please notice that CI do set the dependencies to its latest sha, eg. git reset --hard refs/heads/dev . So if you're using init-repo script you might have a bit different src tree than what CI is using.
Unfortunately it is not possible to trigger only single configuration from CI. CI is configured so that it will abort all still running builds when the first one fails. This is just to save time. 


-----Original Message-----
From: development-bounces+simo.falt=theqtcompany.com at qt-project.org [mailto:development-bounces+simo.falt=theqtcompany.com at qt-project.org] On Behalf Of Bernd Lamecker
Sent: 23. lokakuuta 2014 17:12
To: development at qt-project.org
Subject: [Development] How to track down CI problems


i tried to find the reason why https://codereview.qt-project.org/#/c/91748/ CI fails. 
I set up a Windows 7-x86 in a VM using MSVC2010SP1 and compiled Qt 5.4 and QtWebChannel and the test are passing. But off course i have not the complete environment from the CI. 
Is it possible to trigger CI runs for specific environments only? I'm not even sure if the linux versions were finished and passing, because CI seems to stop all runs as soon as one failed.  
Other ideas on how to find the issue?

Development mailing list
Development at qt-project.org

More information about the Development mailing list