Xorg black screen NVidia problem

[KHarvey]

I have been working on this for a couple of days, and I have finally narrowed down the problem, but I haven't quite figured out how to fix the problem.
I just received my new Alienware laptop with a NVidia Geforce GT 650M video card.  I have found that whenever I install the nvidia-glx drivers my screen goes black and no key combination will bring it back to live.  I am able to install Xorg (7.5) but the first time I reboot, the screen comes to a black screen and no matter what I do I can't bring it back to life without powering it off.

After much reading and trial and error I found that if I install Xorg from the CLI, I am able to successfully install Xorg.  But if I try and run startx the screen will go black and never come back.
But if I install Xorg from the CLI and then install nvidia-glx and run startx I get a GUI again, but it is split in half horizontally, and the top 20% of the screen I am unable to access.
Now when I install nvidia-glx or nvidia-settings I receive an error during the install that I haven't found a fix for yet:

Code: [Select]

graphics-3.0.21-tinycore.tcz: OK
Downloading: nvidia-module-3.0.21-tinycore.tcz
[drm:intel_dsm_platform_ux_info] *ERROR* MUX INFO call failed
Connecting to distrob.ibiblio.org(152.19.134.43:80)
nvidi-module-3.0.21 100%

I think it might have to do with the fact that I think my laptop has a Intel video card, and an NVidia video card in it.  But I am not really sure.

A working setting that I have found is if I install Xorg from the CLI, then I installed nvidia-settings (not nvida-glx) from the CLI, then I can run Xorg -configure and then startx.  When I do this the GUI opens up and looks the way that it is supposed to.  But if I try and run the nvidia-settings app, it says that I do not have my xorg configured for nvidia.  If I go back out to a CLI and run nvidia-xconfig then when I try and run startx I receive the error

Code: [Select]

failed in waitforX
When I dig through the xorg.log.0 the error states that no monitors were found.
When I did a comparison of the xorg.conf for both Xorg -configure and nvidia-xconfig I found the one line that appears to be causing the problem.
Here is my Xorg -configure

Code: [Select]

Section "ServerLayout"
Identifier     "X.org Configured"
Screen      0  "Screen0" 0 0
Screen      1  "Screen1" RightOf "Screen0"
InputDevice    "Mouse0" "CorePointer"
InputDevice    "Keyboard0" "CoreKeyboard"
EndSection

Section "Files"
ModulePath   "/usr/local/lib/X11/modules"
FontPath     "/usr/local/lib/X11/fonts/misc/"
FontPath     "/usr/local/lib/X11/fonts/TTF/"
FontPath     "/usr/local/lib/X11/fonts/OTF"
FontPath     "/usr/local/lib/X11/fonts/Type1/"
FontPath     "/usr/local/lib/X11/fonts/100dpi/"
FontPath     "/usr/local/lib/X11/fonts/75dpi/"
EndSection

Section "Module"
Load  "glx"
Load  "extmod"
Load  "dri2"
Load  "dri"
Load  "dbe"
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 "Monitor"
Identifier   "Monitor1"
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     "SWcursor"            # [<bool>]
        #Option     "HWcursor"            # [<bool>]
        #Option     "NoAccel"            # [<bool>]
        #Option     "ShadowFB"            # [<bool>]
        #Option     "UseFBDev"            # [<bool>]
        #Option     "Rotate"              # [<str>]
        #Option     "VideoKey"            # <i>
        #Option     "FlatPanel"          # [<bool>]
        #Option     "FPDither"            # [<bool>]
        #Option     "CrtcNumber"          # <i>
        #Option     "FPScale"            # [<bool>]
        #Option     "FPTweak"            # <i>
        #Option     "DualHead"            # [<bool>]
Identifier  "Card0"
Driver      "nv"
VendorName  "Unknown Vendor"
BoardName   "Unknown Board"
BusID       "PCI:1:0:0"
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     "Rotate"              # <str>
        #Option     "fbdev"              # <str>
        #Option     "debug"              # [<bool>]
Identifier  "Card1"
Driver      "fbdev"
VendorName  "Unknown Vendor"
BoardName   "Unknown Board"
BusID       "PCI:0:2:0"
EndSection

Section "Screen"
Identifier "Screen0"
Device     "Card0"
Monitor    "Monitor0"
SubSection "Display"
Viewport   0 0
Depth     1
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

Section "Screen"
Identifier "Screen1"
Device     "Card1"
Monitor    "Monitor1"
SubSection "Display"
Viewport   0 0
Depth     1
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
Here is my nvidia-xconfig

Code: [Select]

# nvidia-xconfig: X configuration file generated by nvidia-xconfig
# nvidia-xconfig:  version 280.13  (buildmeister@swio-display-x86-rhel47-05.nvidia.com)  Wed Jul 27 17:18:55 PDT 2011

Section "ServerLayout"
    Identifier     "X.org Configured"
    Screen      0  "Screen0" 0 0
    Screen      1  "Screen1" RightOf "Screen0"
    InputDevice    "Mouse0" "CorePointer"
    InputDevice    "Keyboard0" "CoreKeyboard"
EndSection

Section "Files"
    ModulePath      "/usr/local/lib/X11/modules"
    FontPath        "/usr/local/lib/X11/fonts/misc/"
    FontPath        "/usr/local/lib/X11/fonts/TTF/"
    FontPath        "/usr/local/lib/X11/fonts/OTF"
    FontPath        "/usr/local/lib/X11/fonts/Type1/"
    FontPath        "/usr/local/lib/X11/fonts/100dpi/"
    FontPath        "/usr/local/lib/X11/fonts/75dpi/"
EndSection

Section "Module"
    Load           "glx"
    Load           "extmod"
    Load           "dri2"
    Load           "dbe"
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 "Monitor"
    Identifier     "Monitor1"
    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     "SWcursor"            # [<bool>]
        #Option     "HWcursor"            # [<bool>]
        #Option     "NoAccel"            # [<bool>]
        #Option     "ShadowFB"            # [<bool>]
        #Option     "UseFBDev"            # [<bool>]
        #Option     "Rotate"              # [<str>]
        #Option     "VideoKey"            # <i>
        #Option     "FlatPanel"          # [<bool>]
        #Option     "FPDither"            # [<bool>]
        #Option     "CrtcNumber"          # <i>
        #Option     "FPScale"            # [<bool>]
        #Option     "FPTweak"            # <i>
        #Option     "DualHead"            # [<bool>]
    Identifier     "Card0"
    Driver         "nvidia"
    VendorName     "Unknown Vendor"
    BoardName      "Unknown Board"
    BusID          "PCI:1:0:0"
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     "Rotate"              # <str>
        #Option     "fbdev"              # <str>
        #Option     "debug"              # [<bool>]
    Identifier     "Card1"
    Driver         "nvidia"
    VendorName     "Unknown Vendor"
    BoardName      "Unknown Board"
    BusID          "PCI:0:2:0"
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

Section "Screen"
    Identifier     "Screen1"
    Device         "Card1"
    Monitor        "Monitor1"
    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

The first thing that I noticed is that in the Screen section I don't have a Depth of 1, so I added that into the nvidia-xconfig but I still received the failed in waitforX.  I then added in the Load "dri" but still no go.
The line that is causing all of my problems is in the second Device section.  The driver says "nvidia" in the nvidia-xconfig, but in the Xorg -configure it says "fbdev".  I found that if I change the nvidia-xconfig version to just have the fbdev then it works, but nvidia-settings still states that I am not using my nvidia driver and need to run nvidia-xconfig

Now on this laptop I have video outputs for:
Laptop screen
VGA
HDMI
DisplayPort

I am really not sure where to go from here.  I could probably manually edit my xorg.conf to make it do what I want in the future, but I would prefer to use the nvidia-settings (I'm kind of lazy).  Also I am not sure if everything is working properly since I am using the fbdev rather than nvidia.

I have a feeling that my video card may be a little too new for these drivers, and maybe I should try and download the Linux drivers from NVidia directly to try and get this to work.  But that may cause even more problems.

I am currently dual booting this laptop (Windows 7 and Backtrack), and that was a pain in itself to get working.  I would really like to use TC as my primary OS though.

All of this is running the TC 4.5.5 booted from the CD.  I have the same problem when I do the install, but testing from the CD is way faster to wipe and start over.

Does anyone have any suggestions?

[curaga]

Ouch, an Optimus laptop? Nvidia does not support Optimus on Linux.

In short, you can't control the Nvidia card properly (power saving etc), and can only run things on it with third-party hacks that may or may not work.

[Rich]

Hi KHarvey
If you can disable the video devices (maybe in the BIOS?) you could try setting them up one at a time and saving
the config files to new names. Then use the two config files to construct  a hopefully working one by hand.
The Nvidia config utility is misidentifying the Intel card as another Nvidia card.

[KHarvey]

Ah crap, I completely forgot about Optimus.

I will continue to do some more reading on Optimus.

So basically the second device in the Xorg is the Intel video card.  When I run the nvidia-xconfig it is misinterpreting the Intel GPU as an NVidia GPU and assigning it an NVidia driver which is incorrect.  This is why I am unable to startx.
Now when I run the Xorg -configure it is still kind of misinterpreting the Intel GPU as it is adding it as a FrameBuffer (FBDev) driver rather than the Intel driver. 

I don't remember a setting in the BIOS for this, but I haven't rechecked it as of yet. 

While if I continue to run this way I will be running in a degraded state.  The question is how much of a degraded state will I be running in?  I would much rather use the NVidia GPU rather than the Intel GPU.
Power switching isn't that important to me.  My main goals will be standard desktop usage.  I am not really considered with 3D modeling, but I would very much like to use at least dual monitors (excluding laptop screen).

Eventually I would like to get into some CUDA development (one of the reasons I chose this card.....mistakenly) but I can always move that development into Windows if I must.

[Rich]

Hi KHarvey
Though it can be long and a bit tedious to look through, I would suggest you go through the /var/log/Xorg.0.log
file to get a better idea of what xorg thinks it sees when it makes decisions. If you can't disable a video card
any other way, removing references to the card you wish to disable in the xorg.conf file may be sufficient. That
way you figure out what it takes to get one card functioning correctly without the other one getting in the way.
Don't forget to modify the ServerLayout section to indicate only one screen instead of two.

[KHarvey]

I figured out a little bit more.  When I run startx I am using the second screen section which is using Card0 which is the fbdev driver on Monitor1.  If I try to use Card0 for Monitor1 then I receive an error that no monitors could be found.

So it appears that I may not be able to use my nvidia card at all.  I am running at my max resolution on TC though, even though I am using fbdev.  I guess this means that my battery life will be longer.

Is there anyway for me to switch over to the NVidia GPU?

I will keep reading and trying more things to see if I can figure this out.

[Rich]

Hi KHarvey
According to the nvidia-xconfig

# nvidia-xconfig:  version 280.13 ...

But if you check:

http://www.nvidia.com/object/linux-display-amd64-280.13-driver.html

It does not list the GT650M as one of the supported cards. According to Nvidia, you want:
http://www.nvidia.com/object/linux-display-ia32-295.59-driver.html

[KHarvey]

Okay I have combed through every line in my Xorg.0.log (thanks Rich).  I think reading each line put me on the right track.  I believe I am farther away than what I originally thought.

According to the Xorg.0.log it failed to load the NVIDIA kernel module.  So I then checked DMESG and read through every line on there.  If I would have been smart I would have just tailed DMESG and found what I needed, but oh well.
Inside DMESG I show that the Intel GPU is loaded without issue:  (well minus the MUX error)

Code: [Select]

Linux agpgart interface v0.103
agpgart-intel 0000:00:00.0:  Intel Ivybridge Chipset
agpgart-intel 0000:00:00.0:  detected gtt size:  2097152K total, 262144K mappable
agpgart-intel 0000:00:00.0:  detected 65536K stolen memory
agpgart-intel 0000:00:00.0:  AGP aperture is 256M @ 0xc0000000
[drm] Initialized drm 1.1.0 20060810
i915 0000:00:02.0:  PCI INT A -> GSI 16 (level, low) -> IRQ 16
i915 0000:00:02.0:  setting latency timer to 64
mtrr:  no more MTRRs available
[drm] MTRR allocation failed.  Graphics performance may suffer.
i915 0000:00:02.0:  irq 44 for MSI/MSI-X
[drm] Supports vblank timestamp caching Rev 1 (10.10.2010).
[drm] Driver supports precise vblank timestamp query.
[drm:intel_dsm_platform_mux_info] *ERROR* MUX INFO call failed
vgaarb:  device changed decodes:  PCI:0000:00:02.0,olddecodes=io*mem,decodes=none:owns=io+mem
vgaarb:  transferring owner from PCI:0000:00:02.0 to PCI:0000:01:00.0
fbcon:  inteldrmfb (fb0) is primary device
Console:  switching to colour frame buffer device 200x56
fb0:  inteldrmfb frame buffer device
drm:  registered panic notifier
[For,ware Big]:  ACP(PEGP) defines _DOD but not _DOS
acpi device:3c:  registered as cooling_device13
input:  Video Bus as /devices/LNXSYSTM:00/device:00/PNP0A08:00/device:3a/LNXVIDEO:00/input/input7
ACPI:  Video Device [PEGP] (multi-head:  yeas  rom:  yes  post:  no)
acpi device:48:  registered as cooling_device14
input:  Video Bus as /devices/LNXSYSTM:00/device:00/PNP0A08:00/LNXVIEDEO:01/input/input8
ACPI:  Video Device [GFX0] (multi-head:  yes  rom:  no post:  no)
[drm] Initialized i915 1.6.0 20080730 for 0000:00:02.0 on minor 0
nvidia 0000:01:00.0:  power state changed by ACPI to D0
nvidia 0000:01:00.0:  power state changed by ACPI to D0
nvidia 0000:01:00.0:  enabling device (0006 -> 0007)
nvidia 0000:01:00.0:  PCI INT A -> GSI 16 (level, low) -> IRQ 16
nvidia 0000:01:00.0:  setting latency timer to 64

But here is where the problem lies.  I am failing to load the NVDIA GPU:

Code: [Select]

[drm] Initialized i915 1.6.0 20080730 for 0000:00:02.0 on minor 0
nvidia 0000:01:00.0:  power state changed by ACPI to D0
nvidia 0000:01:00.0:  power state changed by ACPI to D0
nvidia 0000:01:00.0:  enabling device (0006 -> 0007)
nvidia 0000:01:00.0:  PCI INT A -> GSI 16 (level, low) -> IRQ 16
nvidia 0000:01:00.0:  setting latency timer to 64
vgaarb:  device changed decodes:  PCI:0000:01:00.0,olddecodes=io+mem,decodes=none:owns=none
NRM:  The NVIDIA GPU 0000:01:00.0 (PCI ID:  10de:0fd1) installed
NVRM:  in this system is not supported by the 280.13 NVIDIA Linux
NVRM:  graphics driver release.  Please see 'Appendix A -
NVRM:  Supported NVIDIA GPU Products' in this release's README,
NVRM:  available on the Linux graphics driver download page at
NVRM:  www.nvidia.com
nvidia 0000:01:00.0:  PCI INT A disabled
nvidia:  probe of 0000:01:00.0 failed with error -1
NVRM:  The NVIDIA probe routine failed for 1 device(s)
NVRM:  None of the NVIDIA graphics adapters were initialized!

So my thought is that I don't have the proper NVIDIA driver for this GPU.  I am going to try and download the NVIDIA driver direct and see if loading that might solve my current issue. 
Unless someone else has a better suggestion?   

Worst case I may just build this using the Intel GPU and see how it performs.  But I will try to leave that as a worst case option.

[KHarvey]

Hi KHarvey
According to the nvidia-xconfig

# nvidia-xconfig:  version 280.13 ...

But if you check:

http://www.nvidia.com/object/linux-display-amd64-280.13-driver.html

It does not list the GT650M as one of the supported cards. According to Nvidia, you want:
http://www.nvidia.com/object/linux-display-ia32-295.59-driver.html

LOL, thanks for the reply.  You seem to only reply while I am typing up my reply 
I guess I need to learn to type faster.

So from my last post it does appear that I need to try the driver directly from NVIDIA.  I will attempt to load that now, and I will post back my findings.  Once again thank you for your help with this.  I swear I am getting smarter, and am able to solve some of my own issues 

[Rich]

Hi KHarvey
Please read the following thread for an exciting preview of what you may be in for:
http://forum.tinycorelinux.net/index.php/topic,12965.msg71206.html#msg71206

[KHarvey]

I read that thread a little bit earlier.  I'm currently re-reading it, thanks.

I was trying to follow the directions from Arslan S. in this thread:
http://forum.tinycorelinux.net/index.php/topic,2635.msg19712.html#msg19712
But the first problem that I ran into appears that the Linux Headers in the repository are from an older version of TC.  Or at least I am receiving mismatch errors when I try to run the NVIDIAxxx.run
I tried it using the linux-headers-3.0.3-tinycore.tcz

I just rebooted, and I am attempting again using the linux-headers-3.0.21-tinycore.tcz to see if that works.  If that doesn't work, I will reboot again and try again with the linux-3.0.1_api_headers.tcz

Is there a way for me to uninstall using "ab" that way I don't have to reboot each time and reinstall everything?

[KHarvey]

I guess I should have ran a uname first.  uname -r shoes 3.0.21-tinycore, so I should have started with linux-headers-3.0.21-tinycore.tcz

Installing everything again now.  I should be able to post my results in another 10 - 15 minutes.

[Rich]

Hi KHarvey

Is there a way for me to uninstall using "ab" that way I don't have to reboot each time and reinstall everything?

No, AppBrowser marks extensions for removal but does not actually remove them until you reboot. You might
want to plug in a thumb drive to use as temporary persistent storage.

[KHarvey]

I apologize for the late response.  After 12 straight hours I needed a break.

So I was correct, when I used linux-headers-3.0.21-tinycore.tcz I was able to make it past the headers error.  Although it failed again, because of a similar error with the GCC not matching up.  But I found the fix for that, I had to run the command:

Code: [Select]

sudo sh /.NVIDIAxxxx.run -k $(uname -r)

Once I ran that it compiled and installed without issue.

My xorg.conf is quite a bit smaller this time, I only have the NVidia stuff in it.  But when I try to startx it pauses for several seconds and then says failed in waitforx. 
When I check Xorg.0.log it appears it still can't load the driver:

Code: [Select]

(EE) NVIDIA:  Failed to load the NVIDIA kernel module
(II) UnloadModule:  "nvidia"
(II) Unloading /usr/local/lib/X11/modules/drivers/nvidia_drv.so
(EE) Failed to load module "nvidia" (module-specific error, 0)
(EE) No drivers available
Fatal server error:  no screens found

When I check dmesg it appears the driver loaded properly this go around.  But at the bottom I have numerous errors that say:

Code: [Select]

nvidia:  Unknown symbol agp_bridges (err 0)
nvidia:  Unknown symbol agp_free_memory (err 0)
nvidia:  Unknown symbol agp_allocate_memory (err 0)
nvidia:  Unknown symbol agp_unbind_memory (err 0)
nvidia:  Unknown symbol agp_copy_info (err 0)
nvidia:  Unknown symbol agp_backend_release (err 0)
nvidia:  Unknown symbol agp_bind_memory (err 0)
nvidia:  Unknown symbol agp_enable (err 0)
nvidia:  Unknown symbol agp_backend_acquire (err 0)

Well, back to Googling some more.

*Edit*
Sorry, I forgot to put that I can't modprobe agpgart

Code: [Select]

modprobe:  module agpgart not found in modules.dep
Also I couldn't modprobe nvidia:

Code: [Select]

modprobe:  can't load module nvidia (kernel/drivers/video/nvidia.ko):  unknown symbol in module, or unknown parameter


[Juanito]

When I check dmesg it appears the driver loaded properly this go around.  But at the bottom I have numerous errors that say...

Those errors look like they occur because the agp driver is not loaded