I know a year is a long time to pop the stack on this issue, but I still
don't have an adequate setup to reproduce it.  However, it occurred to
me that this issue fits the same description of the one Nathan was
describing recently, whereby an app that attempted to iterate through
the visuals and choose one based on a particular set of attributes would
end up with a StaticGray FB config because VirtualGL wasn't being choosy
enough.

So, can you try the attached patch to VirtualGL along with the latest
TigerVNC pre-release
(http://www.virtualgl.org/DeveloperInfo/TigerVNCPreReleases) and see
what happens?  If it doesn't work, then you might also try tweaking the
attributes in _MatchConfig(), such as adding GLX_ALPHA_SIZE, 8 to the
list to force VirtualGL to obtain a 32-bit FB config.

DRC


On 4/13/10 9:02 AM, Александр Кириллов wrote:
> By the way, I've disabled some checks in KDE4 and it is now able to 
> render itself in OpenGL mode,
> but it shows only gray background for windows (a single rectangle) and 
> black background for
> message boxes (also a rectangle).
>> Visuals are not the same as FB configs in VirtualGL.  Visuals are on the
>> client and FB configs are on the server.  Thus, a visual will always
>> have a 24-bit depth even though the FB config used to create the Pbuffer
>> may have a 32-bit depth.
>>
>> Thus, I don't think that's the cause of the issue.  If it's looking for
>> a visual with GLX_BIND_TO_TEXTURE_RGBA_EXT, then that may be the cause.
>>   How did you determine that it is looking for such a visual?  Please be
>> as verbose as possible.  Don't just say "from debugging", as this does
>> not tell me how you debugged the problem.
>>
>> On 4/11/10 6:49 PM, Александр Кириллов wrote:
>>    
>>> Well, a 32-bit FB config supports rendering alpha channel (24-bit FB
>>> config has 0 bits for alpha and 32-bit has 8 bits for alpha).
>>> Yes, glxinfo repots this and also the debugging - the depth field of a
>>> visual equals 24.
>>>      
>>>> I'm not sure I understand what you mean about 24-bit vs. 32-bit FB
>>>> configs.  And what do you mean by "VirtualGL tells me that those are
>>>> 24-bit configs"?  Are you running /opt/VirtualGL/bin/glxinfo -c to get
>>>> this information?
>>>>
>>>> On 4/10/10 3:01 PM, Александр Кириллов wrote:
>>>>
>>>>        
>>>>> Hello!
>>>>>
>>>>> I'm trying to run KDE4 over VNC with TigerVNC and VirtualGL, but it
>>>>> didn't work out of the box.
>>>>> I used a debugger and found out that KDE4 cannot find a suitable
>>>>> GLXFBConfig - 32-bit with GLX_BIND_TO_TEXTURE_RGBA_EXT (or the same with
>>>>> RGB instead of RGBA) support.
>>>>> On the host system there are plenty of such GLXFBConfigs, but VirtualGL
>>>>> tells me that those are 24-bit configs (I've compared them by number).
>>>>> Can you please suggest me what to do in order to help VirtualGL find a
>>>>> suitable framebuffer configuration?
>>>>>
>>>>> Best regards,
>>>>> Alexander Kirillov
>>>>>
>>>>>
>>>>> ------------------------------------------------------------------------------
>>>>> Download Intel® Parallel Studio Eval
>>>>> Try the new software tools for yourself. Speed compiling, find bugs
>>>>> proactively, and fine-tune applications for parallel performance.
>>>>> See why Intel Parallel Studio got high marks during beta.
>>>>> http://p.sf.net/sfu/intel-sw-dev
>>>>> _______________________________________________
>>>>> VirtualGL-Users mailing list
>>>>> [email protected]
>>>>> https://lists.sourceforge.net/lists/listinfo/virtualgl-users
>>>>>
>>>>>          
>>>> ------------------------------------------------------------------------------
>>>> Download Intel® Parallel Studio Eval
>>>> Try the new software tools for yourself. Speed compiling, find bugs
>>>> proactively, and fine-tune applications for parallel performance.
>>>> See why Intel Parallel Studio got high marks during beta.
>>>> http://p.sf.net/sfu/intel-sw-dev
>>>> _______________________________________________
>>>> VirtualGL-Users mailing list
>>>> [email protected]
>>>> https://lists.sourceforge.net/lists/listinfo/virtualgl-users
>>>>
>>>>        
>>>
>>> ------------------------------------------------------------------------------
>>> Download Intel® Parallel Studio Eval
>>> Try the new software tools for yourself. Speed compiling, find bugs
>>> proactively, and fine-tune applications for parallel performance.
>>> See why Intel Parallel Studio got high marks during beta.
>>> http://p.sf.net/sfu/intel-sw-dev
>>> _______________________________________________
>>> VirtualGL-Users mailing list
>>> [email protected]
>>> https://lists.sourceforge.net/lists/listinfo/virtualgl-users
>>>      
>> ------------------------------------------------------------------------------
>> Download Intel® Parallel Studio Eval
>> Try the new software tools for yourself. Speed compiling, find bugs
>> proactively, and fine-tune applications for parallel performance.
>> See why Intel Parallel Studio got high marks during beta.
>> http://p.sf.net/sfu/intel-sw-dev
>> _______________________________________________
>> VirtualGL-Users mailing list
>> [email protected]
>> https://lists.sourceforge.net/lists/listinfo/virtualgl-users
>>    
> 
> 
> ------------------------------------------------------------------------------
> Download Intel® Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> VirtualGL-Users mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/virtualgl-users
Index: rr/faker-glx.cpp
===================================================================
RCS file: /cvsroot/virtualgl/vgl/rr/faker-glx.cpp,v
retrieving revision 1.52
diff -C2 -r1.52 faker-glx.cpp
*** rr/faker-glx.cpp    23 Mar 2011 05:14:12 -0000      1.52
--- rr/faker-glx.cpp    29 Mar 2011 22:37:29 -0000
***************
*** 53,57 ****
                int attribs[]={GLX_DOUBLEBUFFER, 1, GLX_RED_SIZE, 8, 
GLX_GREEN_SIZE, 8,
                        GLX_BLUE_SIZE, 8, GLX_RENDER_TYPE, GLX_RGBA_BIT, 
GLX_STEREO, 0,
!                       GLX_DRAWABLE_TYPE, GLX_PBUFFER_BIT, None};
                if(__vglClientVisualAttrib(dpy, DefaultScreen(dpy), 
vis->visualid, GLX_STEREO))
                        attribs[11]=1;
--- 53,58 ----
                int attribs[]={GLX_DOUBLEBUFFER, 1, GLX_RED_SIZE, 8, 
GLX_GREEN_SIZE, 8,
                        GLX_BLUE_SIZE, 8, GLX_RENDER_TYPE, GLX_RGBA_BIT, 
GLX_STEREO, 0,
!                       GLX_DRAWABLE_TYPE, GLX_PBUFFER_BIT, GLX_X_VISUAL_TYPE, 
GLX_TRUE_COLOR,
!                       GLX_DEPTH_SIZE, 24, None};
                if(__vglClientVisualAttrib(dpy, DefaultScreen(dpy), 
vis->visualid, GLX_STEREO))
                        attribs[11]=1;
------------------------------------------------------------------------------
Create and publish websites with WebMatrix
Use the most popular FREE web apps or write code yourself; 
WebMatrix provides all the features you need to develop and 
publish your website. http://p.sf.net/sfu/ms-webmatrix-sf
_______________________________________________
VirtualGL-Users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/virtualgl-users

Reply via email to