8174 hangs my laptop on starting X

[netllama]

Have you tried this?
http://www.nvnews.net/vbulletin/show...1&postcount=98

-Lonni

[macemoneta]

I have a Toshiba 5105-s607 and I'm having the same problem (black screen; seems to be a CPU loop since the fan kicks on in a few seconds).

The installer appears to leave previous driver pieces (7676). Reverting to the 7676 driver works, but the upgrade to 8174 is incomplete. I've never had a problem with the install/upgrade on this unit before.

I'll try uninstalling the previous driver first (manually, if the --uninstall doesn't get all the bits), and following the suggestions in post #13.

However, if this works it appears to be a packaging/processing error in the installer; it shouldn't be leaving old driver files. Likewise, if new options are required as indicated in the thread pointed to by post #13, they should be configured as part of the installation. Since these are installer issues, not driver bugs, what is the reporting process?

[netllama]

The bug reporting process is always to provide a bug report. If you can't install the driver, the installer log from /var/log should be provided.

Thanks,
Lonni

[macemoneta]

OK, I tried the recommendations in post#13, without success.

I uninstalled the old (7676) driver first. The only bits left after that were /usr/lib/libGLcore.so.1.0.xxxx, /usr/lib/libGL.s0.1.0.xxxx and /usr/lib/tls/libnvidia-tls.so.1.0.xxxx files, which all previous drivers seem to leave. I removed them in any case. I ran the 8174 installer, and it completed without error. I ran a followup --sanity check and there were no problems. I could manually modprobe the nvidia module without error.

Attempting to start X with the suggested changes in post #13 resulted in no usable modes. I never needed options in /etc/modprobe.conf, but I tried cycling through the NVreg_Mobile values (unloading and re-loading the nvidia module each time of course); 1,3,4,5 reported no usable modes, while 2 caused the hang/CPU loop. I tried both with NVreg_SoftEDIDs=0 and without it; no change.

I reverted to 7676 (no /etc/modprobe.conf options), and everything works. The bug report is attached for when I was trying without NVreg_SoftEDIDs=0 and with a NVreg_Mobile set to a value that caused a 'no modes' error. I can't get an xorg.0.log of the hang, because the file is not created (or more likely it gets reverted by ext3 journalling because I have to power-cycle the machine).

[netllama]

Here's the problem:
(--) NVIDIA(0): DFP-0: maximum pixel clock: 0 MHz

That is a completely different problem from what others have reported experiencing in this thread.

I see that you're not specifying any modes in your Screen section. Can you test and verify whether the behavior changes if you specify a known good mode?

Regardless, please start X with the following command:
startx -- -logverbose 6

and then generate and post a new bug report.

Also, which BIOS version are you using?
Where did you obtain the 'mm' kernel that you're using?

Thanks,
Lonni

[joffe]

Originally Posted by macemoneta

OK, I tried the recommendations in post#13, without success.

Can you show us your xorg.conf? I thought I had done everything in the book and given up on the new driver until I added some stuff to that file, among other things. I assume you have seen my post here.

[macemoneta]

Originally Posted by netllama

Here's the problem:
(--) NVIDIA(0): DFP-0: maximum pixel clock: 0 MHz

That is a completely different problem from what others have reported experiencing in this thread.

I see that you're not specifying any modes in your Screen section. Can you test and verify whether the behavior changes if you specify a known good mode?

I added the mode for 1600x1200 (the physical resolution of the laptop LCD), even though this had been previously autodetected.

Regardless, please start X with the following command:
startx -- -logverbose 6

and then generate and post a new bug report.

Done, attached. I noted that the EDID in the verbose log is detecting the 1600x1200 panel and pixel clock, even though they are not being used:

(II) NVIDIA(0): Connected display device(s): DFP-0, TV-0
(WW) NVIDIA(0): Multiple displays connected, but only one display allowed;
(WW) NVIDIA(0): using boot display
(II) NVIDIA(0): Enabled display device(s): DFP-0
(II) NVIDIA(0): Mapping display device 1 (DFP-0) to CRTC 1
(--) NVIDIA(0): DFP-0: maximum pixel clock: 0 MHz
(--) NVIDIA(0): DFP-0: Internal Dual Link LVDS
(--) NVIDIA(0):
(--) NVIDIA(0): --- EDID for Display Device DFP-0 ---
(--) NVIDIA(0): EDID Version : 1.3
(--) NVIDIA(0): Manufacturer : NVD
(--) NVIDIA(0): Monitor Name : NVIDIA Nvidia Default Flat
(--) NVIDIA(0): Panel
(--) NVIDIA(0): Product ID : 1024
(--) NVIDIA(0): 32-bit Serial Number : 0
(--) NVIDIA(0): Serial Number String :
(--) NVIDIA(0): Manufacture Date : 2002, week 45
(--) NVIDIA(0): DPMS Capabilities : Standby Suspend Active Off
(--) NVIDIA(0): Prefer first detailed timing : Yes
(--) NVIDIA(0): Supports GTF : No
(--) NVIDIA(0): Maximum Image Size : 320mm x 240mm
(--) NVIDIA(0): Valid HSync Range : 29kHz - 76kHz
(--) NVIDIA(0): Valid VRefresh Range : 0Hz - 60Hz
(--) NVIDIA(0):
(--) NVIDIA(0): Detailed Timings:
(--) NVIDIA(0): 1600 x 1200 @ 60Hz
(--) NVIDIA(0): Pixel Clock : 162.00 MHz
(--) NVIDIA(0): HRes, HSyncStart : 1600, 1664
(--) NVIDIA(0): HSyncEnd, HTotal : 1856, 2160
(--) NVIDIA(0): VRes, VSyncStart : 1200, 1201
(--) NVIDIA(0): VSyncEnd, VTotal : 1204, 1250
(--) NVIDIA(0): H/V Polarity : +/+
(--) NVIDIA(0):
(--) NVIDIA(0):
(--) NVIDIA(0): Raw EDID bytes:
(--) NVIDIA(0):
(--) NVIDIA(0): 00 ff ff ff ff ff ff 00 3a c4 00 04 00 00 00 00
(--) NVIDIA(0): 2d 0c 01 03 80 20 18 00 ea a8 e0 99 57 4b 92 25
(--) NVIDIA(0): 1c 50 54 00 00 00 01 01 01 01 01 01 01 01 01 01
(--) NVIDIA(0): 01 01 01 01 01 01 48 3f 40 30 62 b0 32 40 40 c0
(--) NVIDIA(0): 13 00 42 f1 10 00 00 1e 00 00 00 fc 00 4e 76 69
(--) NVIDIA(0): 64 69 61 20 44 65 66 61 75 6c 00 00 00 fc 00 74
(--) NVIDIA(0): 20 46 6c 61 74 20 50 61 6e 65 6c 00 00 00 00 fd
(--) NVIDIA(0): 00 00 3c 1d 4c 11 00 00 20 20 20 20 20 00 00 aa
(--) NVIDIA(0):
(--) NVIDIA(0): --- End of EDID for Display Device DFP-0 ---
(--) NVIDIA(0):
(**) NVIDIA(0): Digital Vibrance for display 1 set to 5
(II) NVIDIA(0): Frequency information for DFP-0:
(II) NVIDIA(0): HorizSync : 29.000-76.000 kHz
(II) NVIDIA(0): VertRefresh : 0.000-60.000 Hz
(II) NVIDIA(0): (HorizSync from EDID)
(II) NVIDIA(0): (VertRefresh from EDID)
(II) NVIDIA(0): Adding EDID-provided mode "1600x1200" for DFP-0.

Also, which BIOS version are you using?
Where did you obtain the 'mm' kernel that you're using?

Thanks,
Lonni

The laptop BIOS is the Toshiba 1.90 BIOS. The 'mm' kernel is locally compiled from the Fedora Core kernel of the same release. It contains the following additional patches to support the laptop hardware:

- Latest Toshiba ACPI
- Patch for proper IEEE1394 interface initialization
- Patch for Synaptics cPad touchpad

I didn't think that those patches would impact the video driver, but just to be sure I switched to the standard Fedora kernel (same release, unmodified, downloaded from Fedora's updates) when repeating the test for the attached bug report. There was no change in the behavior.

[netllama]

Can you try what joffe suggested here?
http://www.nvnews.net/vbulletin/show...&postcount=100

I'll see if I can locate your laptop to reproduce this.

Thanks,
Lonni

[macemoneta]

Originally Posted by netllama

Can you try what joffe suggested here?
http://www.nvnews.net/vbulletin/show...&postcount=100

I'll see if I can locate your laptop to reproduce this.

Thanks,
Lonni

OK, I made the changes in the link, and get essentially the same result. The pixel clock is still reporting as 0, but no EDID (as expected from the options used in the link). The results in terms on the hang are the same. I'm attaching both the 8174 bug report with this configuration, and a 7676 bug report with the same configuration (the only changes being I rebooted my patched kernel with the 7676 driver) for comparison.

[wschutzer]

Originally Posted by netllama

Have you tried this?
http://www.nvnews.net/vbulletin/show...1&postcount=98

-Lonni

Whoa!

I just followed that guide and now 1.0-8174 works! And not only that, now I too can enjoy some eyecandy (See the screenshot below). Wheeeehe! Please tell this to everybody else out there who is having trouble with the NV17.

Thanks a lot, Lonni

Cheers,
Waldeck

Just a few notes:
1. The DPMS bug is gone. Congratulations! Now my screen is back at the touch of a key.
2. The console corruption bug is also gone. Congratulations again!
3. I'm not sure if it's my eyes but the fonts look different. They seem crisper, but not all of them look quite as good.
4. I nave Composite and RenderAccel turned on for over 4 hours now and so far I have experienced no crash. It seems that stability was greatly improved over 1.0-7676.
5. I also have the impression that 1.0-8174 is substantially faster and uses less memory than 1.0-7676.



PS: 24hrs+ uptime, suspended to ram many times, and no crash!

[Imperito]

There seem to be a few problems at once going on in this thread. My original problem was partially resolved by the NVreg_Mobile=1 bit, but my revised problem remains unsolved. I cannot get full resolution or TV out with 8174. I guess we should keep going in this thread, although I think a lot of the posts are unrelated.

[macemoneta]

The problem is that the initial symptoms are the same; it's only once some triage has been done that the different root causes are observed. We can certainly split the threads, though that may make it harder for new people reporting a problem to find a solution (like wschutzer, above).