Thanks for the pointer DRC,
I've been looking around and I've come up with the following solution to
get gdm to start 2 X servers at boot up on this machine by modifying
/etc/gdm/custom.conf as follows:
[servers]
0=/usr/bin/X -config /etc/X11/xorg.0.conf
1=/usr/bin/X -config /etc/X11/xorg.1.conf
where I have modified the default nvida configured xorg.conf to make the
two new xorg.conf files as below (for clarity I'll only show a diff for
the second file):
/etc/X11/xorg.0.conf
-----------------------------------------------------------------------------------------------------------
# nvidia-xconfig: X configuration file generated by nvidia-xconfig
# nvidia-xconfig: version 1.0 (buildmeis...@builder58) Wed Dec 9
16:34:26 PST 2009
Section "DRI"
Mode 0666
EndSection
Section "ServerLayout"
Identifier "X.org Configured"
Screen 0 "Screen0" 0 0
InputDevice "Mouse0" "CorePointer"
InputDevice "Keyboard0" "CoreKeyboard"
EndSection
Section "Files"
RgbPath "/usr/share/X11/rgb"
ModulePath "/usr/lib64/xorg/modules"
FontPath "unix/:7100"
FontPath "built-ins"
EndSection
Section "Module"
Load "record"
Load "dbe"
Load "extmod"
Load "glx"
Load "xtrap"
EndSection
Section "InputDevice"
Identifier "Keyboard0"
Driver "kbd"
EndSection
Section "InputDevice"
Identifier "Mouse0"
Driver "mouse"
Option "Protocol" "auto"
Option "Device" "/dev/input/mice"
Option "ZAxisMapping" "4 5 6 7"
EndSection
Section "Monitor"
Identifier "Monitor0"
VendorName "Monitor Vendor"
ModelName "Monitor Model"
EndSection
Section "Device"
### Available Driver options are:-
### Values: <i>: integer, <f>: float, <bool>: "True"/"False",
### <string>: "String", <freq>: "<f> Hz/kHz/MHz"
### [arg]: arg optional
#Option "ShadowFB" # [<bool>]
#Option "DefaultRefresh" # [<bool>]
#Option "ModeSetClearScreen" # [<bool>]
Identifier "Card0"
Driver "nvidia"
VendorName "nVidia Corporation"
BoardName "Unknown Board"
BusID "PCI:3:0:0"
EndSection
Section "Screen"
Identifier "Screen0"
Device "Card0"
Monitor "Monitor0"
SubSection "Display"
Viewport 0 0
EndSubSection
SubSection "Display"
Viewport 0 0
Depth 4
EndSubSection
SubSection "Display"
Viewport 0 0
Depth 8
EndSubSection
SubSection "Display"
Viewport 0 0
Depth 15
EndSubSection
SubSection "Display"
Viewport 0 0
Depth 16
EndSubSection
SubSection "Display"
Viewport 0 0
Depth 24
EndSubSection
EndSection
-----------------------------------------------------------------------------
diff /etc/X11/xorg.*.conf
58c58
< Identifier "Card0"
---
> Identifier "Card1"
62c62
< BusID "PCI:3:0:0"
---
> BusID "PCI:4:0:0"
67c67
< Device "Card0"
---
> Device "Card1"
Both X servers start successfully and both have sensible output for
xdpyinfo and /opt/VirtualGL/bin/glxinfo.
I get sensible ouput for both
# vglrun +v -d :0 /opt/VirtualGL/bin/glxinfo &> glxinfo0
and
# vglrun +v -d :1 /opt/VirtualGL/bin/glxinfo &> glxinfo1
# diff glxinfo0 glxinfo1
1,2c1,2
< [VGL] Shared memory segment ID for vglconfig: 5210123
< [VGL] Opening local display :0
---
> [VGL] Shared memory segment ID for vglconfig: 5177355
> [VGL] Opening local display :1
but when I try glxspheres I get an error on display 0 but it works
perfectly on display 1
# vglrun +v -d :0 /opt/VirtualGL/bin/glxspheres &> glxspheres0
# cat glxspheres0
Polygons in scene: 62464
[VGL] Shared memory segment ID for vglconfig: 5242891
[VGL] Opening local display :0
Visual ID of window: 0x22
[VGL] ERROR: OpenGL error 0x0502
[VGL] ERROR: in readpixels--
[VGL] 624: Could not Read Pixels
# vglrun +v -d :0 /opt/VirtualGL/bin/glxspheres &> glxspheres1
# cat glxspheres1
Polygons in scene: 62464
[VGL] Shared memory segment ID for vglconfig: 5308427
[VGL] Opening local display :1
Visual ID of window: 0x22
282.851470 frames/sec - 289.357054 Mpixels/sec
287.179087 frames/sec - 293.784206 Mpixels/sec
286.921263 frames/sec - 293.520452 Mpixels/sec
287.996523 frames/sec - 294.620443 Mpixels/sec
am I doing something fundamentally wrong here or is my approach correct
and I'm missing something?
Antony
DRC wrote:
> Yeah, there isn't anything in the docs about it because I haven't ever
> personally done it. :) But many others have, and I think it's fairly
> straightforward. You just set up the second 3D card on display :1 with
> a second X server. There are many sites that describe how to set up
> multiple X servers on multiple graphics cards.
>
> Antony Cleave wrote:
>
>> Hi all,
>>
>> I am about to get a machine with two Quadro FX5800 cards in it that I
>> want to configure to allow multiple users to visualise their data. I
>> have a working testing box setup with a single graphics card configured
>> using TurboVNC and VirtualGL 2.1.4 and it is amazing. How much more
>> complicated would a multiple graphics card install be? What extra steps
>> would I have to go through? There does not appear to be much on this in
>> the documentation except a quick mention in the advanced configuration
>> settings at the bottom where it says I can use vglrun -d <display> or
>> the VGL_DISPLAY variable to select which X display should do the
>> rendering but there is nothing about how to setup the server. Is this
>> because it "just works" or are there some more things I have to do to
>> make it work?
>>
>> thanks for your time
>>
>> Antony
>>
>>
>>
>
> ------------------------------------------------------------------------------
> 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
>
--
Antony Cleave
Systems Architect
ClusterVision
12 Westgate House
The Island
Gloucester
GL1 2RU
United Kingdom
Office: +44 1452 260024
skype: antony.cleave
------------------------------------------------------------------------------
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
