[vtkusers] VTK/OpenGL2 application over VirtualGL/TurboVNC?

Elvis Stansvik elvis.stansvik at orexplore.com
Mon Apr 24 09:31:51 EDT 2017


2017-04-24 15:18 GMT+02:00 Elvis Stansvik <elvis.stansvik at orexplore.com>:
> 2017-04-24 15:00 GMT+02:00 Ken Martin <ken.martin at kitware.com>:
>> And you ran it over virtualGL/TurboVNC?
>
> Yep. See attached screenshot. This was when I ran
>
>     vglrun bin/vtkRenderingOpenGL2CxxTests TestVBOPLYMapper -D
> ExternalData/Testing -I
>
> in a TurboVNC session from my own laptop to the server machine.
>
> Out of curiosity, since we're using Qt 5 in our application, I tried
> running the QVTKRenderWindowInteractor.py script, as a quick test, and
> it gives the same error as when running our app:
>
> ERROR: In /buildbot/vtk7-builder/build/Rendering/OpenGL2/vtkOpenGLRenderWindow.cxx,
> line 628
> vtkXOpenGLRenderWindow (0x25e1670): GLEW could not be initialized.
>
> So maybe the problem is related to Qt 5 + VTK somehow?
>
> In our application (C++), we use a custom QVTKWidget subclass. I'll
> try to make a minimal test case that uses our custom class, but

Instead, I just tried the
http://www.vtk.org/Wiki/VTK/Examples/Cxx/Qt/RenderWindowUISingleInheritance
example, which uses QVTKWidget directly, and it exhibits the same
problem.

So something funny is going on with Qt + VTK under VirtualGL.

I wonder, the reports of success when running ParaView under
VirtualGL/TurboVNC, anyone know if that was with Qt 4 or 5?

Elvis

> QVTKRenderWindowInteractor.py exhibiting the same problem seems to
> suggest it might be related to Qt + VTK.
>
> I also verified that the SmartVolumeMapper example works fine, even
> when modified to use the vtkGPUVolumeRayCastMapper instead (which is
> what we're using).
>
> Elvis
>
>>
>> You can run it with -I added on if you want it to stay up so you can
>> interact with it
>>
>> vtkRenderingOpenGL2CxxTests TestVBOPLYMapper -D ExternalData\Testing -I
>>
>>
>>
>> On Mon, Apr 24, 2017 at 8:57 AM, Elvis Stansvik
>> <elvis.stansvik at orexplore.com> wrote:
>>>
>>> 2017-04-24 14:10 GMT+02:00 Ken Martin <ken.martin at kitware.com>:
>>> > Elvis, could you try running the following test?
>>> >
>>> > ctest -R VBOPLY -VV
>>> >
>>> > or explicitly as
>>> >
>>> > vtkRenderingOpenGL2CxxTests TestVBOPLYMapper -D ExternalData\Testing
>>> >
>>> > and send me the output? That test checks for OpenGL2 support and prints
>>> > out
>>> > an error message if it fails to get a valid context.
>>>
>>> Everything seems fine. The output is
>>>
>>> first render time: 0.381111
>>> opengl version 4.5
>>> interactive render time: 0.0321125
>>> number of triangles: 871414
>>> triangles per second: 2.71363e+07
>>>
>>> and I could see the dragon render quickly before disappearing.
>>>
>>> I'll try to cook up a minimal example that sets up a volume rendering
>>> pipeline similar to what I have in our application, and try to
>>> reproduce the problem I get there.
>>>
>>> Elvis
>>>
>>> >
>>> > Thanks!
>>> > Ken
>>> >
>>> >
>>> >
>>> >
>>> >
>>> >
>>> > On Mon, Apr 24, 2017 at 7:50 AM, Elvis Stansvik
>>> > <elvis.stansvik at orexplore.com> wrote:
>>> >>
>>> >> 2017-04-24 13:35 GMT+02:00 Andreas Buykx <A.Buykx at dianafea.com>:
>>> >> > We have been busy trying to get OpenGL2 to work on our linux backend
>>> >> > machines using TigerVNC and we needed to build our own mesa to get
>>> >> > the
>>> >> > OpenGL support required by the VTK OpenGL2 layer. I'm not sure what
>>> >> > version
>>> >> > of mesa was required, my colleague who worked on this is on holiday.
>>> >> > So far
>>> >> > the transition to VTK OpenGL2  is in an experimental stage, we will
>>> >> > probably
>>> >> > not roll it out until after our next release. We are very interested
>>> >> > in
>>> >> > exchanging experiences though...
>>> >>
>>> >> Alright. My quick test server was a Thinkpad T460p laptop with NVIDIA
>>> >> proprietary driver, so not using Mesa. The laptop is running Ubuntu
>>> >> 16.04.
>>> >>
>>> >> I followed pretty much the instructions at
>>> >> https://github.com/aancel/admin/wiki/VirtualGL-on-Ubuntu to set it up.
>>> >>
>>> >> I'm attaching the output I get from `vglrun glxinfo` in a terminal
>>> >> when connected to the server. It looks good to me.
>>> >>
>>> >> The result when trying to run our VTK application is
>>> >>
>>> >> ERROR: In
>>> >>
>>> >> /buildbot/vtk7-builder/build/Rendering/OpenGL2/vtkOpenGLRenderWindow.cxx,
>>> >> line 628
>>> >> vtkXOpenGLRenderWindow (0x94e5e0): GLEW could not be initialized.
>>> >>
>>> >> followed by a segmentation fault.
>>> >>
>>> >> This is with VTK 7.1.1, configured with the OpenGL2 backend.
>>> >>
>>> >> I'm attaching the backtrace (I had to modify the /usr/bin/vglrun
>>> >> script to run gdb in order to get it).
>>> >>
>>> >> I assume I would get the same problem with a minimal VTK example, but
>>> >> have yet to try this.
>>> >>
>>> >> Cheers,
>>> >> Elvis
>>> >>
>>> >> >
>>> >> > Kind regards,
>>> >> > Andreas
>>> >> >
>>> >> >> -----Original Message-----
>>> >> >> From: vtkusers [mailto:vtkusers-bounces at vtk.org] On Behalf Of Elvis
>>> >> >> Stansvik
>>> >> >> Sent: Monday, April 24, 2017 1:12 PM
>>> >> >> To: Róbert Špir
>>> >> >> Cc: VTK Users
>>> >> >> Subject: Re: [vtkusers] VTK/OpenGL2 application over
>>> >> >> VirtualGL/TurboVNC?
>>> >> >>
>>> >> >> 2017-04-24 12:53 GMT+02:00 Róbert Špir <spir.robert at gmail.com>:
>>> >> >> > Hi Elvis,
>>> >> >> > I tried it without success, but paraview runs without problems, so
>>> >> >> > there's got to be some way...
>>> >> >>
>>> >> >> Ah, that's an interesting piece of info.
>>> >> >>
>>> >> >> Perhaps there's some ParaView dev here who knows what might set
>>> >> >> ParaView apart in this context?
>>> >> >>
>>> >> >> Elvis
>>> >> >>
>>> >> >> >
>>> >> >> > Robert
>>> >> >> >
>>> >> >> > -----Original Message-----
>>> >> >> > From: vtkusers [mailto:vtkusers-bounces at vtk.org] On Behalf Of
>>> >> >> > Elvis
>>> >> >> > Stansvik
>>> >> >> > Sent: Monday, April 24, 2017 12:51 PM
>>> >> >> > To: VTK Users <vtkusers at vtk.org>
>>> >> >> > Subject: [vtkusers] VTK/OpenGL2 application over
>>> >> >> > VirtualGL/TurboVNC?
>>> >> >> >
>>> >> >> > Has anyone successfully run a VTK application using the OpenGL2
>>> >> >> > backend over VirtualGL/TurboVNC?
>>> >> >> >
>>> >> >> > GLEW fails to initialize when I try to run the application using
>>> >> >> > vglrun. I'm ready to give up, since this was just a casual
>>> >> >> > experiment,
>>> >> >> > but I thought I'd ask first if anyone has it running. The output
>>> >> >> > from
>>> >> >> > `vglrun glxinfo` looks good to me.
>>> >> >> >
>>> >> >> > Elvis
>>> >> >> > _______________________________________________
>>> >> >> > Powered by www.kitware.com
>>> >> >> >
>>> >> >> > Visit other Kitware open-source projects at
>>> >> >> > http://www.kitware.com/opensource/opensource.html
>>> >> >> >
>>> >> >> > Please keep messages on-topic and check the VTK FAQ at:
>>> >> >> > http://www.vtk.org/Wiki/VTK_FAQ
>>> >> >> >
>>> >> >> > Search the list archives at:
>>> >> >> > http://markmail.org/search/?q=vtkusers
>>> >> >> >
>>> >> >> > Follow this link to subscribe/unsubscribe:
>>> >> >> > http://public.kitware.com/mailman/listinfo/vtkusers
>>> >> >> >
>>> >> >> _______________________________________________
>>> >> >> Powered by www.kitware.com
>>> >> >>
>>> >> >> Visit other Kitware open-source projects at
>>> >> >> http://www.kitware.com/opensource/opensource.html
>>> >> >>
>>> >> >> Please keep messages on-topic and check the VTK FAQ at:
>>> >> >> http://www.vtk.org/Wiki/VTK_FAQ
>>> >> >>
>>> >> >> Search the list archives at: http://markmail.org/search/?q=vtkusers
>>> >> >>
>>> >> >> Follow this link to subscribe/unsubscribe:
>>> >> >> http://public.kitware.com/mailman/listinfo/vtkusers
>>> >>
>>> >> _______________________________________________
>>> >> Powered by www.kitware.com
>>> >>
>>> >> Visit other Kitware open-source projects at
>>> >> http://www.kitware.com/opensource/opensource.html
>>> >>
>>> >> Please keep messages on-topic and check the VTK FAQ at:
>>> >> http://www.vtk.org/Wiki/VTK_FAQ
>>> >>
>>> >> Search the list archives at: http://markmail.org/search/?q=vtkusers
>>> >>
>>> >> Follow this link to subscribe/unsubscribe:
>>> >> http://public.kitware.com/mailman/listinfo/vtkusers
>>> >>
>>> >
>>> >
>>> >
>>> > --
>>> > Ken Martin PhD
>>> > Distinguished Engineer
>>> > Kitware Inc.
>>> > 28 Corporate Drive
>>> > Clifton Park NY 12065
>>> >
>>> > This communication, including all attachments, contains confidential and
>>> > legally privileged information, and it is intended only for the use of
>>> > the
>>> > addressee.  Access to this email by anyone else is unauthorized. If you
>>> > are
>>> > not the intended recipient, any disclosure, copying, distribution or any
>>> > action taken in reliance on it is prohibited and may be unlawful. If you
>>> > received this communication in error please notify us immediately and
>>> > destroy the original message.  Thank you.
>>
>>
>>
>>
>> --
>> Ken Martin PhD
>> Distinguished Engineer
>> Kitware Inc.
>> 28 Corporate Drive
>> Clifton Park NY 12065
>>
>> This communication, including all attachments, contains confidential and
>> legally privileged information, and it is intended only for the use of the
>> addressee.  Access to this email by anyone else is unauthorized. If you are
>> not the intended recipient, any disclosure, copying, distribution or any
>> action taken in reliance on it is prohibited and may be unlawful. If you
>> received this communication in error please notify us immediately and
>> destroy the original message.  Thank you.


More information about the vtkusers mailing list