On Monday 06 July 2015 11:32:54 Edmund Grimley Evans wrote: > > Ideally we should tests the tests on an arm* machine running X11 and see > > what happens there (and that means not calling xvfb too). > > Do you mean running an X server? Why would that be necessary? Could > the test be making use of an X server other than in the usual way, > through DISPLAY? > > It is possible to run the test which is failing on arm* like this on amd64: > > cd obj-*/autotests > DISPLAY=<host-running-vncserver>:1 ./kcolumnresizertest > > Same thing doesn't work on arm64. > > (The vncserver had "xhost +". Other tests in the same directory worked > on both systems when run like that. There were lots of differences > between the two systems besides the architecture, but both were Debian > unstable chroots.) No, I mean running an Xserver and using a VGA/HDMI connected monitor/screen to check it. AFAIU (and I would *love* to be probed wrong here) the difference is at the GPU. If one uses xvfb/vnc the GPU doesn't gets involved and the OpenGL stuff is done trough software. Now if the output is trough the GPU (hdmi/vga/whatever) then the hardware is what it will get used. My idea is to know if the tests keep failing in that case or not. If they are then we have a bigger issue, if they don't, we can simply say we can't run the tests in the normal way. Thanks for trying non the less :) -- UNIX is basically a simple operating system, but you have to be a genius to understand the simplicity. Dennis Ritchie Lisandro Damián Nicanor Pérez Meyer http://perezmeyer.com.ar/ http://perezmeyer.blogspot.com/
Attachment:
signature.asc
Description: This is a digitally signed message part.