Display Resolution issue

English Hardware
1

Hey Guys,
I was playing around today with fstab to get my windows partition to automount during boot and when I rebooted to test the resolution of my monitor went from 1400x1050 to a max of 1024x768. I reinstalled the graphics cards drivers from NVIDIA and still no love. Can someone tell me what’s up and how to fix? Going to post some output that might help:

mike@linux-guhv:~> xrandr
Screen 0: minimum 320 x 240, current 1024 x 768, maximum 1024 x 768
default connected 1024x768+0+0 0mm x 0mm
1024x768 50.0*
800x600 51.0 52.0 53.0
640x480 54.0
512x384 55.0
400x300 56.0 57.0 58.0
320x240 59.0

dmesg snippet:

10.912127] powernow-k8: Found 1 AMD Phenom™ 9550 Quad-Core Processor (4 cpu cores) (version 2.20.00)
10.912180] powernow-k8: 0 : pstate 0 (2200 MHz)
10.912182] powernow-k8: 1 : pstate 1 (1100 MHz)
12.020128] nvidia: module license ‘NVIDIA’ taints kernel.
12.020133] Disabling lock debugging due to kernel taint
12.589814] r8169 0000:04:00.0: eth0: link down
12.590076] ADDRCONF(NETDEV_UP): eth0: link is not ready
12.713316] ACPI: PCI Interrupt Link [LMC9] enabled at IRQ 22
12.713328] nvidia 0000:00:0d.0: PCI INT A -> Link[LMC9] -> GSI 22 (level, low) -> IRQ 22
12.713339] nvidia 0000:00:0d.0: setting latency timer to 64
12.713347] vgaarb: device changed decodes: PCI:0000:00:0d.0,olddecodes=io+mem,decodes=none:owns=io+mem
12.713584] NVRM: loading NVIDIA UNIX x86_64 Kernel Module 260.19.36 Tue Jan 18 16:56:11 PST 2011
12.850949] NET: Registered protocol family 17
13.923376] type=1503 audit(1297714872.685:12): operation=“change_hat” info=“unconfined” error=-1 pid=1987
13.923394] type=1503 audit(1297714872.685:13): operation=“change_hat” info=“unconfined” error=-1 pid=1987
22.776935] CPUFREQ: Per core ondemand sysfs interface is deprecated - up_threshold

Let me know if there’s anything else I can post.

Cheers!

Mike

2

What model is the monitor? How is it connected? This info can be useful

cat /var/log/Xorg.0.log

Cut and paste the output to SUSE Paste

Set the ‘Delete After’ to ‘Never’

Post the link that it generates here.

BTW, if you search these forums, you will many similar threads involving nvidia drivers and EDID problems (which this may be based on your xrandr results).

3

this is a Samsung SyncMaster T220 connected VGA.

This monitor was working just great for several weeks before this reboot.

SUSE Paste

4

As I suspected, its an analogue VGA connection and an EDID issue:

#
    14.932] (WW) NVIDIA(GPU-0): Unable to read EDID for display device CRT-0
#
    14.932] (II) NVIDIA(0): NVIDIA GPU GeForce 7025 / nForce 630a (C61) at PCI:0:13:0
#
    14.932] (II) NVIDIA(0):     (GPU-0)
#
    14.932] (--) NVIDIA(0): Memory: 524288 kBytes
#
    14.932] (--) NVIDIA(0): VideoBIOS: 05.61.32.31.00
#
    14.932] (--) NVIDIA(0): Interlaced video modes are supported on this GPU
#
    14.932] (--) NVIDIA(0): Connected display device(s) on GeForce 7025 / nForce 630a at
#
    14.932] (--) NVIDIA(0):     PCI:0:13:0
#
    14.932] (--) NVIDIA(0):     CRT-0
#
    14.932] (--) NVIDIA(0): CRT-0: 350.0 MHz maximum pixel clock
#
    14.933] (II) NVIDIA(0): Assigned Display Device: CRT-0
#
    14.933] (==) NVIDIA(0):
#
    14.933] (==) NVIDIA(0): No modes were requested; the default mode "nvidia-auto-select"
#
    14.933] (==) NVIDIA(0):     will be used as the requested mode.
#
    14.933] (==) NVIDIA(0):
#
    14.933] (II) NVIDIA(0): Validated modes:
#
    14.933] (II) NVIDIA(0):     "nvidia-auto-select"
#
    14.933] (II) NVIDIA(0): Virtual screen size determined to be 1024 x 768
#
    14.933] (WW) NVIDIA(0): Unable to get display device CRT-0's EDID; cannot compute DPI
#
    14.933] (WW) NVIDIA(0):     from CRT-0's EDID.

!) With the nvidia drivers, it is possible to load EDID from a file, but the trick is to obtain this info in the first place.

Fixing Ugly DVI/HDMI Displays due to EDID bugs on nVidia drivers | analogbit.com

Some users have manage to aquire this EDID binary from Windows, but you’ll have to google for more info on this if you choose this approach.

  1. It may be possible to add a manual modeline to /etc/X11/xorg.conf.d/50-monitor.conf, but this involves adding some other Xorg options to ignore EDID, and I’m not sure how experienced/confident you are with editing system files. There are a lot of old threads dealing with this issue if you care to search.
5

I think it is the best approach anyway, even if he has little experience in modifying system files.

My friend tried to find an EDID file on the net and it wasn’t an easy task either.

I’ll ask him how he managed to do it.

6

Thanks … I’ve tried to find the EDID value and failed on all methods. The software listed is for 32-bit only, NVIDIA settings shows no where to get EDID value and booting into windows XP and running the pheonix software give me an error.

sigh One would think this far along they’d have this fixed out of the box … ahh well. I guess I’ll try and puzzle out the modeline fix. On to option 2.

7

bahahahaha …

Well … I found the issue. After some more reading trying to find a solution I ran across an article where someone had the same problem and it was a KVM issue with recognizing his monitor. Well … the only addition I’ve made to my system in the last while that would have any affect would be a vga extension cable. So I figured what the heck. I removed the monitor extension cable and rebooted … blam!!! The machine came back in all it’s glory and actually recognized the monitor.

So what’s the thoughts … the extension is bad and not allowing the video card to detect the monitor? Monitor extensions are inherently bad? I wonder …ahh well.

Thanks for all your help guys.

Cheers!
Mike

8

Thanks for the update. That makes sense re EDID. (Sometimes you have to go over recent changes/history with a fine tooth comb!)

For VGA connectors, pins 9,12, and 15, are utilised for DDC communication (on which the display capabilities are derived), so maybe your extension cable doesn’t connect these properly.

Extron Electronics - Understanding EDID - Extended Display Identification Data