Failing tests on dashlin1

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Failing tests on dashlin1

Andrew Maclean-3
Does anyone know why the following tests regularly fail on dashlin1.kitware?
vtkRenderingCorePython-rendererSource
vtkRenderingOpenGLCxx-TestTilingCxx

Andrew
--
___________________________________________
Andrew J. P. Maclean

___________________________________________

_______________________________________________
Powered by www.kitware.com

Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html

Follow this link to subscribe/unsubscribe:
http://www.vtk.org/mailman/listinfo/vtk-developers

Reply | Threaded
Open this post in threaded view
|

Re: Failing tests on dashlin1

David Cole-2
Those tests are some of the few (I think there are 3 altogether, but
there could be more) that use the front buffer when doing the image
capture for comparison to the baseline. The vast majority of tests
capture the back buffer.

Tests that capture the front buffer are susceptible to accidentally
capturing pixels from overlapping (and completely arbitrary, depending
on what's on screen at the time) windows as part of the image
capture... So you end up with test results that have junk in them like
this:

http://open.cdash.org/testDetails.php?test=259246517&build=3358643

I know what it is, but I'm not sure what to do about it...


HTH,
David C.

_______________________________________________
Powered by www.kitware.com

Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html

Follow this link to subscribe/unsubscribe:
http://www.vtk.org/mailman/listinfo/vtk-developers

Reply | Threaded
Open this post in threaded view
|

Re: Failing tests on dashlin1

Andrew Maclean-3
Thanks for the explanation. I have only ever seen it for three tests and I did wonder about the junk on the images ... I thought it was a new style of branding :-)

Regards
   Andrew


On Fri, Jun 6, 2014 at 12:27 PM, David Cole <[hidden email]> wrote:
Those tests are some of the few (I think there are 3 altogether, but there could be more) that use the front buffer when doing the image capture for comparison to the baseline. The vast majority of tests capture the back buffer.

Tests that capture the front buffer are susceptible to accidentally capturing pixels from overlapping (and completely arbitrary, depending on what's on screen at the time) windows as part of the image capture... So you end up with test results that have junk in them like this:

http://open.cdash.org/testDetails.php?test=259246517&build=3358643

I know what it is, but I'm not sure what to do about it...


HTH,
David C.




--
___________________________________________
Andrew J. P. Maclean

___________________________________________

_______________________________________________
Powered by www.kitware.com

Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html

Follow this link to subscribe/unsubscribe:
http://www.vtk.org/mailman/listinfo/vtk-developers