nV News Forums

 
 

nV News Forums (http://www.nvnews.net/vbulletin/index.php)
-   NVIDIA Linux (http://www.nvnews.net/vbulletin/forumdisplay.php?f=14)
-   -   odd bug when inserting nvidia module (http://www.nvnews.net/vbulletin/showthread.php?t=87241)

Bloodsurfer 02-28-07 09:31 AM

odd bug when inserting nvidia module
 
2 Attachment(s)
Hello.

My problem is a rather odd one. You can see it in the attached image.

That happens sometimes when booting my system, at the moment when the nvidia module is inserted. It does not happen everytime.

After that error, the system continues to boot, but Xorg won't start, instead I get only a black screen with a (not blinking) cursor, and can only reboot (switching to another virtual tty doesn't work).

When the system boots without that error, it runs absolute flawlessly, without any instability or other errors. It does run for hours/days under full load without problems.

I am using Gentoo Linux on an Intel Core 2 Duo E6600 with Kernel 2.6.20 (using 2.6.19 also produces that error sometimes) and an Asus P5B deluxe motherboard.
Graphics card is an Geforce 7900 GS. Nothing is overclocked, everything is running at standard speed.

netllama 02-28-07 12:38 PM

Re: odd bug when inserting nvidia module
 
Its rather difficult to make out everything in your screenshot, however I do not see any references to the nvidia driver.

Bloodsurfer 03-11-07 06:48 AM

Re: odd bug when inserting nvidia module
 
2 Attachment(s)
You're right, the picture was not the best. I made a new one, it's attached.
In the meantime I updated my BIOS, compiled a new kernel (2.6.20-gentoo-r1) and updated to the new nvidia-drivers.

The problem still persists, same as before.

I also attached a new bug-report file, now made with the new drivers.

WBerlin 03-11-07 07:04 AM

Re: odd bug when inserting nvidia module
 
I would say, that is a hardware problem.
Either your memory chips are broken, or there is something wrong with your sound card. I don't think it has to do with your Nvidia-Card.
Check this:

Try to remove/disable the sound card.

Remove (if you have more then one) one memory module and boot the system again. Try the same with the other memory module(s) you have.

netllama 03-11-07 12:57 PM

Re: odd bug when inserting nvidia module
 
According to the screenshot, the nvidia driver isn't even loaded when the Oops is occurring.

Bloodsurfer 03-11-07 01:21 PM

Re: odd bug when inserting nvidia module
 
Well, I disabled the sound card/driver and will now wait if the error occurs again.
But, to be honest, I don't think the problem is memory or sound related. It's imho a pure coincidence that the sound driver is loaded before the nvidia driver.

Of course the nvidia module is not in that list of linked-in modules - if an error happens while inserting it, it simply will not be inserted, so how could it be in that list?

First the sound modules are inserted - I see an "ok" appearing (ok, that's not visible in the shot because it is on the right side of the screen), then udev tries to load the nvidia module and fails. Trying to insert nvidia is the last thing the kernel does before the error and there is no "ok" message telling the module were successfully inserted.

I also did run several memtests for more than one day, there never was a single error.

Bloodsurfer 03-12-07 07:06 AM

Re: odd bug when inserting nvidia module
 
Well, I was right. The error does still persist, even without any other modules.

I cannot imagine how this could be related to anything else besides the nvidia module.

netllama 03-12-07 10:28 AM

Re: odd bug when inserting nvidia module
 
Please clarify what "without any other modules" means.

Also, the Oops doesn't appear in the bug reports you've attached, and according to the bug reports, X is started & working fine with the nvidia driver. Can you setup a serial console to capture the Oops?

WBerlin 03-12-07 03:23 PM

Re: odd bug when inserting nvidia module
 
odd, really odd.
But whatever, I still think it's a hardware problem. Test your system with Knoppix (so far I know, there is a nvidia driver on the CD).
Or try Windows if it works ($!!%$&"!&%$ hell, I shout my mouth here) a bluescreen would prove my assumption.

Bloodsurfer 03-12-07 06:48 PM

Re: odd bug when inserting nvidia module
 
1 Attachment(s)
Quote:

Originally Posted by netllama
Please clarify what "without any other modules" means.

Also, the Oops doesn't appear in the bug reports you've attached, and according to the bug reports, X is started & working fine with the nvidia driver. Can you setup a serial console to capture the Oops?

With "without any other modules" I meant I removed the alsa sound driver modules from my kernel, so the nvidia module is the only one loaded at that point.

That the error didn't appear in my logs was my fault, yes, I made that reports while xorg was running because I always rebooted the machine after the error. I didn't think of logging in and creating the report first.

I made a few reboots and waited till the error appeared again, then logged in (without seeing what I was typing because of the black screen after the bug) and made a new report that is now attached. The error is visible there now.

Quote:

odd, really odd.
But whatever, I still think it's a hardware problem. Test your system with Knoppix (so far I know, there is a nvidia driver on the CD).
Or try Windows if it works ($!!%$&"!&%$ hell, I shout my mouth here) a bluescreen would prove my assumption.
Well, I'm using Win XP on that machine too. I don't like to admit it, but it is running perfectly stable. No single error since I installed it four weeks ago. No crash, no errors while booting, no blue screen, no errors when I play games for hours, nothing. It just works, at the moment better than Linux :(

zander 03-12-07 06:55 PM

Re: odd bug when inserting nvidia module
 
Could you capture the output of `cat /proc/kallsyms` (before loading the module), as well as the exact Oops message (after loading the module)?

Xevious 03-13-07 06:30 AM

Re: odd bug when inserting nvidia module
 
the exact oops was in his second nvidia bugreport:

Code:

Freeing unused kernel memory: 244k freed
nvidia: module license 'NVIDIA' taints kernel.
Unable to handle kernel paging request at 000000007d935800 RIP:
 [<ffffffff8801cf3f>]
PGD 7c468067 PUD 0
Oops: 0002 [1] SMP
CPU 1
Modules linked in: nvidia(P)
Pid: 1584, comm: modprobe Tainted: P      2.6.20-gentoo-r2 #4
RIP: 0010:[<ffffffff8801cf3f>]  [<ffffffff8801cf3f>]
RSP: 0000:ffff81007e29fd60  EFLAGS: 00010246
RAX: 000000007d935800 RBX: ffff81007d935800 RCX: ffff81007b220000
RDX: 000000007d935800 RSI: 0000000000000006 RDI: ffff81007d935800
RBP: 0000000000000001 R08: 0000000000000000 R09: ffff81007d935800
R10: 0000000000000001 R11: 0000000000000004 R12: ffffffff886e5410
R13: 0000000000000000 R14: ffffffff886e53c0 R15: 0000000000000003
FS:  00002b34de7eb730(0000) GS:ffff810002f514c0(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 000000007d935800 CR3: 000000007e55d000 CR4: 00000000000006e0
Process modprobe (pid: 1584, threadinfo ffff81007e29e000, task ffff81007d324000)
Stack:  ffffffff8801cef9 ffff81007d935800 ffffffff8801ce65 ffff81007d935800
 ffffffff8801ce39 ffff81007b220000 ffffffff8801def8 0000000000000788
 ffff81007b220000 0000000000000000 ffffffff8801dcdf ffff81007b220000
Call Trace:
 [<ffffffff8801cef9>]
 [<ffffffff8801ce65>]
 [<ffffffff8801ce39>]
 [<ffffffff8801def8>]
 [<ffffffff8801dcdf>]
 [<ffffffff8801db1c>]
 [<ffffffff8801da09>]
 [<ffffffff8801dbc6>]
 [<ffffffff880389f1>]
 [<ffffffff886eb0e3>]
 [<ffffffff8029d59c>]
 [<ffffffff8020b0fb>]
 [<ffffffff8025c99e>]


Code: 00 00 00 52 07 03 88 48 c7 87 20 05 00 00 06 22 03 88 c3 48
RIP  [<ffffffff8801cf3f>]
 RSP <ffff81007e29fd60>
CR2: 000000007d935800
 <6>EXT3 FS on sda7, internal



All times are GMT -5. The time now is 09:16 PM.

Powered by vBulletin® Version 3.7.1
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Copyright 1998 - 2014, nV News.