nV News Forums

 
 

nV News Forums (http://www.nvnews.net/vbulletin/index.php)
-   NVIDIA Linux (http://www.nvnews.net/vbulletin/forumdisplay.php?f=14)
-   -   Debian sid w/ nForce - X won't start (http://www.nvnews.net/vbulletin/showthread.php?t=357)

questioncolon 08-03-02 02:43 PM

Debian sid w/ nForce - X won't start
 
After installing the kernel and glx drivers on NVidia's website, X no longer starts up. I get a black screen and nothing happens until I switch back to the terminal, and I get the following error:

XIO: fatal IO error 104 (Connection reset by peer) on X server ":0.0" after 0 requests (0 known processed) with 0 events remaining.

Starting up X by just typing "X" at the command line does the same thing, except that I get a seg fault instead of that error message.

The NVDriver module is loaded, and my XF86Config-4 is set to use the "nvidia" driver. I commented out GLcore and dri in the Module section. Anyone have this problem before? I'm running Debian sid with XFree 4.1 on a nForcee 420D motherboard.

vforge37 08-03-02 05:01 PM

Weird
 
Sorry to just state the obvious, but that is really weird.

Can you just upgrade to XFree86 4.2 ?

questioncolon 08-03-02 05:43 PM

I tried doing this on Gentoo also (using XFree 4.2). Same result..

vforge37 08-03-02 06:01 PM

What about another distro?
 
I wouldn't trust a problem with Gentoo as meaning that it shouldn't work. With Gentoo, too many other things can go wrong. What about trying another distro like RH or Mandrake or Suse?

Or, I still think upgrading to XFree86 4.2.0 on your debian partition wouldn't (shouldn't) hurt...?

Thunderbird 08-04-02 03:37 AM

Upgrading to 4.2 on debian doesn't have to be good. (I upgraded but had my reasons for it.) 4.1 should work fine. For debian will still take some time before official 4.2 packages show up. They will be released when xfree86 4.2 works fine on all 11 (or were it 12 already) officially supported architectures (i386, ppp, sparc, arm, alpha ...) . (most aren't even tested by the xfree86 developers)

There are unofficial 4.2 debs for some of the architectures but those aren't that reliable yet. Check here if you want them: http://people.debian.org/~branden/

questioncolon 08-04-02 01:22 PM

1 Attachment(s)
Quick update on the issue - I've discovered that the logger is able to get me some error messages (although I have absolutely no idea what they mean). Here is the error that I get after trying to start X:

Unable to handle kernel NULL pointer dereference at virtual address 00000000
printing eip:
d6a4e456
*pde = 00000000
Oops: 0000
CPU: 0
EIP: 0010:[<d6a4e456>] Tainted: P
EFLAGS: 00010246
eax: 00000000 ebx: 00000000 ecx: d4880000 edx: 00000000
esi: 00000000 edi: 00000000 ebp: c1cabd0c esp: c1cab488
ds: 0018 es: 0018 ss: 0018
Process XFree86 (pid: 7907, stackpage=c1cab000)
Stack: c1cabd7c c1810004 c1810004 00000000 00000000 d4880000 6d756372 78656c66
d8b329bc 00000000 40080200 c1cab4fc 00000200 00000200 00000200 16000000
00000080 d5fca400 d8b32168 c1810134 c1810294 c1817110 0909090a 6365205b
Call Trace: [<d6a7ae9e>] [<d6a4c552>] [<d6a4cdc0>] [<d6a806c5>] [<d6aaa4c3>]
[<d6ac9b60>] [<d6a03e2c>] [<d6a042d3>] [<d6ac9b60>] [<d6ac9b60>] [<d69f3d39>]
[<d6ac9b60>] [<c013f844>] [<c013de9d>] [<c013ddb2>] [<c013e147>] [<c0108a9b>]

Code: 8b 03 8b bd 90 f7 ff ff 89 07 83 c3 04 89 9d 88 f7 ff ff 83
<3>XFree86[7907] exited with preempt_count 2


My XF86Config-4 is attached.

Thunderbird 08-04-02 02:10 PM

Perhaps an upgrade to 4.2 isn't a bad idea for you after all. The error you got is not something normal. It is what a blue screen of death is on windows. Linux really crashed. (Or a switch back to some woody packages)

questioncolon 08-04-02 02:18 PM

Yikes, same thing happens in 4.2 :mad:

Thunderbird 08-04-02 03:29 PM

Perhaps something in the kernel. Try to switch back to a more stable distro. (woody ..)

questioncolon 08-04-02 04:46 PM

I've tried this in Debian woody, sid, and Gentoo 1.2, and it's all giving me the same thing. :mad:

dmw400 08-04-02 08:14 PM

I'm getting this with RedHat 7.3. Truly strange.

dmw400 08-04-02 08:33 PM

Ok, a little more investigation has led me to try strace, and this is what I got:

[root@tiamat root]# strace X 2>/tmp/Xlog

... blah blah, failure ...

[root@tiamat root]# cat /tmp/Xlog

... blah, blah ...

write(0, "(II) ", 5) = 5
write(0, "Module vgahw: vendor=\"The XFree8"..., 43) = 43
write(0, "\tcompiled for 4.2.0", 19) = 19
write(0, ", module version = 0.1.0\n", 25) = 25
write(0, "\tABI class: XFree86 Video Driver"..., 46) = 46
write(0, "(**) ", 5) = 5
write(0, "NVIDIA(0): ", 11) = 11
write(0, "Depth 8, ", 9) = 9
write(0, "(--) ", 5) = 5
write(0, "framebuffer bpp 8\n", 18) = 18
write(0, "(==) ", 5) = 5
write(0, "NVIDIA(0): ", 11) = 11
write(0, "Default visual is PseudoColor\n", 30) = 30
write(0, "(==) ", 5) = 5
write(0, "NVIDIA(0): ", 11) = 11
write(0, "Using gamma correction (1.0, 1.0"..., 39) = 39
write(0, "(--) ", 5) = 5
write(0, "NVIDIA(0): ", 11) = 11
write(0, "Linear framebuffer at 0xE0000000"..., 33) = 33
write(0, "(--) ", 5) = 5
write(0, "NVIDIA(0): ", 11) = 11
write(0, "MMIO registers at 0xEC000000\n", 29) = 29
open("/dev/nvidiactl", O_RDWR) = 6
ioctl(6, 0xc0044602, 0x8241ac0) = 0
ioctl(6, 0xc0044622, 0xbffffa20) = 0
ioctl(6, 0xc0044604, 0xbffff988) = 0
ioctl(6, 0xc0044623, 0xbffffa10) = 0
open("/dev/nvidia0", O_RDWR <unfinished ...>
+++ killed by SIGSEGV +++


Aha! Is this what you get when you try that, as well?


All times are GMT -5. The time now is 11:40 AM.

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