{{{#!div class="box"
[[Image(http://xpra.org/icons/opengl.png)]]
= OpenGL Applications =

You can run [https://www.opengl.org/ OpenGL] applications in an xpra session in a number of ways:
* by running the xpra session directly on a "real" X11 server using an accelerated display driver - this is hard to setup and undocumented: you need to use the {{{xvfb}}} switch to start an accelerated X11 server
* by [/wiki/ShadowServer shadowing] an existing accelerated X11 server (sub-optimal)
* by using [http://www.virtualgl.org/ VirtualGL]

[[BR]]

This is not related at all to the [/wiki/ClientRendering/OpenGL OpenGL client rendering]: you can connect with any client you want, accelerated or not.
}}}


{{{#!div class="box"
== VirtualGL setup ==

Usually, you can just {{{vglrun}}} your application and things will "just work".
That is the case for simple application like {{{glxgears}}} / {{{glxspheres}}}.

Other applications may require more complicated workarounds, see the extensive [http://svn.code.sf.net/p/virtualgl/code/tags/2.4.1/doc/index.html VirtualGL documentation] for details, in particular [http://svn.code.sf.net/p/virtualgl/code/tags/2.4.1/doc/index.html#hd0012 12: Using VirtualGL with setuid/setgid Executables].

Caveat: for some unknown reason, you may need to use {{{/etc/ld.so.preload}}} instead of {{{LD_PRELOAD}}} to make some suid applications work through xpra. (#898)
Temporary solution:
{{{
$ mv /etc/ld.so.preload /etc/ld.so.preload.saved >& null
$ echo "/usr/lib64/VirtualGL/libdlfaker.so /usr/lib64/VirtualGL/libgefaker.so /usr/lib64/VirtualGL/librrfaker.so" > /etc/ld.so.preload
$ vglrun yourapp
$ rm /etc/ld.so.preload
$ mv /etc/ld.so.preload.saved /etc/ld.so.preload >& null
}}}
If you application is suid, you may also need to chmod it.
}}}

{{{#!div class="box"
== OpenGL Performance ==

Some rough figures and compatibility tests.

Running [http://www.geeks3d.com/gputest/ GpuTest] 0.7.0, on a Fedora 22 system with a [http://www.geforce.com/ GeForce GTX 760], kernel 4.0.5, [http://www.nvidia.com/Download/index.aspx nvidia driver] 346.59 at 1024x720 resolution.

The [http://www.vmware.com/products/player vmplayer] tests below start the same tests in a Windows 7 Ultimate 64-bit virtual machine.


||= Platform =||||||||||||= Test =||
||= =||||= {{{FurMark}}} =||||= {{{TessMark X8}}} =||||= {{{PixMark JuliaFP32}}} =||||= {{{PixMark Volplosion}}} =||
||= =||= Points =||= FPS =||= Points =||= FPS =||= Points =||= FPS =||= Points =||= FPS =||
||Bare Metal|| 4283|| 71|| 87143|| 1453|| 29994|| 498|| 2969|| 49||
||vglrun|| 3769|| 62|| 8842|| 147|| 10621|| 177|| 2696|| 44||
||vmplayer|| 1736|| 28|||| failed || 3459|| 57|||| failed ||
||vglrun vmplayer|| 1774|| 29|||| failed || 3525|| 58|||| failed ||
||xpra vglrun vmplayer|| 2494|| 41|||| failed || 10927|| 181|||| failed ||

Initial observations: virtualgl adds some overhead which is most noticeable when running on bare metal, hardly noticeable when running through vmplayer as it slows down the opengl performance considerably already. Running through xpra actually gives a speed boost, probably because it does not need to wait for the actual screen refresh since there isn't one.

Note: [https://www.virtualbox.org/ VirtualBox] 4.3.x failed to launch any of the tests and is not included here.
}}}