problem setting up dual head on two separate graphics cards

jenhu · 01-05-2010, 07:27 AM

I have a Benq G2420HDB screen connected to a nVidia GeForce 6200SE card and a Benq FP73G connected to a nVidia MCP61S onboard graphics card.

Code:

/sbin/lspci | grep VGA

00:0d.0 VGA compatible controller: nVidia Corporation GeForce 6100 nForce 405 (rev a2)
02:00.0 VGA compatible controller: nVidia Corporation NV44 [GeForce 6200SE TurboCache (TM)] (rev a1)

I have tried to set up Xinerama, but don't know why it doesn't work: Screen 2 doesn't get any signal.
As expected, the onboard graphics controller was disabled when the pci-e card was attached, but lspci found it after I enabled support for multiple screens in the BIOS.
The relevant parts (I think) of xorg.conf are:

Code:

Section "Monitor"
    Identifier  "Monitor 1"
    HorizSync   31.5-67.5
    VertRefresh 50-70
EndSection

Section "Monitor"
    Identifier  "Monitor 2"
    HorizSync   31.5-50
    VertRefresh 40-90
EndSection


Section "Device"
    Identifier  "Standard VGA"
    VendorName  "Unknown"
    BoardName   "Unknown"
EndSection

Section "Device"
    Identifier  "NVIDIA GeForce 6200"
    Driver      "nv"
    BusID       "PCI:2:0:0"
    Screen      0
EndSection

Section "Device"
    Identifier  "NVIDIA GeForce 6100"
    Driver      "nv"
    Screen      1
    BusID       "PCI:0:13:0"
EndSection


Section "Screen"
    Identifier  "Screen 1"
    Device      "NVIDIA GeForce 6200"
    Monitor     "Monitor 1"
    DefaultDepth 24

    Subsection "Display"
        Depth       24
        Modes       "1920x1080" "1280x768"
        ViewPort    0 0
        Virtual 1920 1080
    EndSubsection
EndSection

Section "Screen"
    Identifier  "Screen 2"
    Device      "NVIDIA GeForce 6100"
    Monitor     "Monitor 2"
    DefaultDepth 24

    Subsection "Display"
        Depth           24
        Modes           "1280x1024" "1024x768" "800x600" "640x480"
        ViewPort        0 0
    EndSubsection
EndSection


Section "ServerLayout"
    Identifier  "Simple Layout"

    Screen 0 "Screen 1"
    Screen 1 "Screen 2" RightOf "Screen 1"
    Option "Xinerama" "on"
    Option "Clone" "off"
EndSection

Xorg.0.log says:

Code:

...
(II) Primary Device is: PCI 02:00:0
(WW) NV: No matching Device section for instance (BusID PCI:0:13:0) found
...
(EE) Screen 1 deleted because of no matching config section.
...

I'm using Slackware 12.0, kernel 2.6.21.5-smp.
Please ask for more information if needed.
Thank you

Elemecca · 01-10-2010, 03:22 PM

I think your problem is the "Screen N" lines in the "Device" sections. These are meant to distinguish between multiple outputs on the same card. When you put "Screen 1" in the "Device" section for the 6100, you are telling it to use the second screen on that card. As it only has one screen attached, there is no such thing and that device definition is dropped.

jenhu · 01-12-2010, 08:23 AM

Thank you for the response!

I tried removing the Screen N option, but now neither of the screens show anything!

the last few lines in the Xorg log from that session is:

Code:

(II) Loading sub module "int10"
(II) LoadModule: "int10"
(II) Reloading /usr/lib/xorg/modules//libint10.so
(II) NV(1): Initializing int10
(II) Attempted to read BIOS 128KB from /sys/bus/pci/devices/0000:00:0d.0/rom: got 0KB

Elemecca · 01-12-2010, 08:27 PM

It looks like X is trying to soft-boot the card at 0:13:0 with the int10 module instead of using the nVidia POST routine. Try adding the line below to the "Device" section for that card.

Code:

Option "UseInt10Module" "false"

You may find the nVidia driver documentation interesting.

Elemecca · 01-12-2010, 08:29 PM

What are the version numbers of your X server and the nVidia driver? They should show up near the top of the X log.

jenhu · 01-13-2010, 10:42 AM

I tried adding 'Option "UseInt10Module" "false"' as you instructed, but nothing seems to change.

X version and driver version:

Code:

X Window System Version 1.3.0
Release Date: 19 April 2007
X Protocol Version 11, Revision 0, Release 1.3
Build Operating System: Slackware 12.0 Slackware Linux Project
Current Operating System: Linux host 2.6.21.5-smp #2 SMP Tue Jun 19 14:58:11 CDT 2007 i686
...
(II) LoadModule: "nv"
(II) Loading /usr/lib/xorg/modules/drivers//nv_drv.so
(II) Module nv: vendor="X.Org Foundation"
        compiled for 1.3.0, module version = 2.0.96
        Module class: X.Org Video Driver
        ABI class: X.Org Video Driver, version 1.2

Elemecca · 01-14-2010, 09:10 AM

Oops. I had thought you were using nvidia, not nv. As you may know, nv is an open-source driver developed in part by nVidia that only supports 2D acceleration whereas nvidia is a proprietary closed-source driver released by nVidia the supports 3D acceleration. There's also Nouveau, which is an open-source 3D driver developed by the community that I've never played with.

Unless you have problems with using proprietary software, I would recommend trying the nvidia driver. It generally seems to work better than nv, at least for me. Your current X configuration should work fine with nvidia, just change the Driver line from "nv" to "nvidia". If you don't have the driver installed, it's probably in your package manager as "nvidia-drivers".

jenhu · 01-15-2010, 10:13 AM

I installed the nvidia drivers, did some experimenting in xorg.conf and now it works!

Thank you very much for your help!

jenhu · 01-16-2010, 05:26 AM

Well.. It did work.

Now, the next time I booted, the second GPU (GeForce 6100) can't find the screen (FP73G).

Code:

(II) NVIDIA(GPU-1): No display devices connected; falling back to: CRT-0
(WW) NVIDIA(GPU-1): Unable to read EDID for display device CRT-0
(II) NVIDIA(1): NVIDIA GPU GeForce 6100 nForce 405 (C61) at PCI:0:13:0
(II) NVIDIA(1):     (GPU-1)
(--) NVIDIA(1): Memory: 524288 kBytes
(--) NVIDIA(1): VideoBIOS: 05.61.32.25.04
(--) NVIDIA(1): Interlaced video modes are supported on this GPU
(--) NVIDIA(1): Connected display device(s) on GeForce 6100 nForce 405 at
(--) NVIDIA(1):     PCI:0:13:0:
(--) NVIDIA(1):     CRT-0
(--) NVIDIA(1): CRT-0: 350.0 MHz maximum pixel clock
(II) NVIDIA(1): Assigned Display Device: CRT-0
(WW) NVIDIA(1): No valid modes for "1280x1024"; removing.
(EE) NVIDIA(1): The requested configuration of display devices (CRT-0) in
(EE) NVIDIA(1):     MetaMode "1024x768" is not supported on this GPU.
(EE) NVIDIA(1): The requested configuration of display devices (CRT-0) in
(EE) NVIDIA(1):     MetaMode "800x600" is not supported on this GPU.
(EE) NVIDIA(1): The requested configuration of display devices (CRT-0) in
(EE) NVIDIA(1):     MetaMode "640x480" is not supported on this GPU.
(WW) NVIDIA(1):
(WW) NVIDIA(1): Unable to validate any modes; falling back to the default mode
(WW) NVIDIA(1):     "nvidia-auto-select".
(WW) NVIDIA(1):
(EE) NVIDIA(1): The requested configuration of display devices (CRT-0) in
(EE) NVIDIA(1):     MetaMode "nvidia-auto-select" is not supported on this
(EE) NVIDIA(1):     GPU.
(EE) NVIDIA(1): Unable to use default mode "nvidia-auto-select".
(II) UnloadModule: "nvidia"
(II) UnloadModule: "wfb"
(II) UnloadModule: "fb"
(--) Depth 24 pixmap format is 32 bpp
(II) do I need RAC?  No, I don't.

When it worked, that same section looked like this:

Code:

(II) NVIDIA(1): NVIDIA GPU GeForce 6100 nForce 405 (C61) at PCI:0:13:0
(II) NVIDIA(1):     (GPU-1)
(--) NVIDIA(1): Memory: 524288 kBytes
(--) NVIDIA(1): VideoBIOS: 05.61.32.25.04
(--) NVIDIA(1): Interlaced video modes are supported on this GPU
(--) NVIDIA(1): Connected display device(s) on GeForce 6100 nForce 405 at
(--) NVIDIA(1):     PCI:0:13:0:
(--) NVIDIA(1):     BenQ FP73G (CRT-0)
(--) NVIDIA(1): BenQ FP73G (CRT-0): 350.0 MHz maximum pixel clock
(II) NVIDIA(1): Assigned Display Device: CRT-0
(II) NVIDIA(1): Validated modes:
(II) NVIDIA(1):     "1280x1024"
(II) NVIDIA(1):     "1024x768"
(II) NVIDIA(1):     "800x600"
(II) NVIDIA(1):     "640x480"
(II) NVIDIA(1): Virtual screen size determined to be 1280 x 1024
(--) NVIDIA(1): DPI set to (95, 96); computed from "UseEdidDpi" X config
(--) NVIDIA(1):     option
(==) NVIDIA(1): Disabling 32-bit ARGB GLX visuals.
(--) Depth 24 pixmap format is 32 bpp
(II) do I need RAC?  Yes, I do.

I found others with the same problem, but on cold boot. I tried rebooting, but the GPU still can't find the screen.