nV News Forums

 
 

nV News Forums (http://www.nvnews.net/vbulletin/index.php)
-   NVIDIA Linux (http://www.nvnews.net/vbulletin/forumdisplay.php?f=14)
-   -   X freezes with FX5200 (http://www.nvnews.net/vbulletin/showthread.php?t=28716)

elad 05-13-04 05:50 PM

X freezes with FX5200
 
Hi,

I encounter an occasional lockup in X with the nVidia FX5200. This usually happens when I play a video (with any of the players available for MDK, such as totem, mplayer etc.)
It seems that my PC is still running: the mouse pointer moves, and USB LEDs are on. Unfortunately the keyboard doesn't function, except for rebooting with ALT-Print Screen-B.
I am running MDK PPack 10. official, with the stock kernel. The hardware is Athlon 2600+, 512Mb,ASUS A7V8X-X (VIA KT400 chipset.) The driver version is 5336 (kernel 2.6.3.7mdk), and the driver section in XF86Config is as follows:

Section "Device"
Identifier "device1"
VendorName "NVidia"
BoardName "NVIDIA GeForce FX (generic)"
Driver "nvidia"
Option "DPMS"
#Option "IgnoreEDID" "1"
EndSection

Thanks,
Elad

jerickson314 05-15-04 03:56 PM

Re: X freezes with FX5200
 
AGP support is one frequent cause of crashes. Disabling it (option 0) or switching to agpgart mode (option 2) is worth a try. To your device section try adding one of the following:

Option "NvAGP" "0"
(Disabled, more likely to work but slower graphics performance)

Option "NvAGP" "2"
(agpgart, Less likely to work but still worth a try due to the faster graphics performance)

I suggest trying option 0 first and option 2 if that fixes the problem.

Also you may want to try getting the latest BIOS from Asus if the configuration change doesn't help.

For me agpgart works with the 2.6.4 kernel in SuSE 9.1 and the same model of motherboard you have (A7V8X-X). I have not needed to upgrade my BIOS.

Also, I have an FX5200 board but for BoardName I have "0x0322". I think the 0x board names may vary within a model due to AGP speed, Video RAM size, or something like that, but I'm not sure.

elad 05-24-04 10:02 AM

Re: X freezes with FX5200
 
Unfortunately, setting "NvAGP" to 2 hasn't solved the problem. By the way, I thought that using agpgart was the default, if "NvAGP" is not set at all.
I'll lokk at my BIOS, just in case. Removing AGP support is too costly, so I'll only revert to it as a last resort. Does anyone know if this is a common problem, and if someone is working on fixing it?

Elad

jerickson314 05-24-04 04:25 PM

Re: X freezes with FX5200
 
I was mistaken the first time, the default option does try agpgart before it tries NvAGP. You might want to try NvAGP 1, which uses nVidia's proprietary AGP driver. However, the KT400 chipset on the A7V8X-X is not officially supported by that option, and when I tried it on SuSE 9.0 (kernel 2.4) it gave my system frequent crashes.
One thing you may want to try is to disable the framebuffer (remove any "vga=" and/or "splash=" options from your GRUB or LILO configuration for booting Linux). This will revert the consoles to 640x480 with no bootsplash, but it may help with your stability problems.

When the system crashes is there a message on the system log about "Badness in pci_find_subsys"? This is a common issue. I did discover that on my system with NvAGP 2 my system still sometimes crashes, but infrequently. When it does crash, it leaves the error message. You can search the forums for that text to find out more about that problem or for suggestions. I do believe nVidia is working on a fix.

BTW, I have just started testing the system with AGP enabled but the framebuffer disabled; sometimes a bad setup won't crash on me for several days so I don't know if that fixed the problem for me. I did try updating my BIOS to the latest from Asus and it didn't help me; in fact it kept me from being able to enter Setup most of the time so I reverted to the original BIOS I had backed up. So if you try upgrading the BIOS, make sure you make a backup copy of the old version first in case there is trouble. There are instructions in the motherboard manual and on Asus's web site for how to do that. The easiest way to make a DOS boot disk on Linux is to enter the commands given in the instructions into dosemu.

elad 05-24-04 05:32 PM

Re: X freezes with FX5200
 
The system doesn't crash - X simply hangs. I am pretty sure the system is still running because the my USB mouse works, and my USB keyboard is able to send the ALT-PRTSCR-B sequence used for rebooting.

elad 06-07-04 11:55 AM

Re: X freezes with FX5200
 
I've just encountered this lock-up again. Here is what /var/log/messages had to say about it:

Jun 7 12:48:16 localhost kernel: Badness in pci_find_subsys at drivers/pci/search.c:167
Jun 7 12:48:16 localhost kernel: Call Trace:
Jun 7 12:48:16 localhost kernel: [pci_find_subsys+215/224] pci_find_subsys+0xd7/0xe0
Jun 7 12:48:16 localhost kernel: [<c01c86f7>] pci_find_subsys+0xd7/0xe0
Jun 7 12:48:16 localhost kernel: [pci_find_device+24/32] pci_find_device+0x18/0x20
Jun 7 12:48:16 localhost kernel: [<c01c8718>] pci_find_device+0x18/0x20
Jun 7 12:48:16 localhost kernel: [pci_find_slot+26/96] pci_find_slot+0x1a/0x60
Jun 7 12:48:16 localhost kernel: [<c01c855a>] pci_find_slot+0x1a/0x60
Jun 7 12:48:16 localhost kernel: [__crc_open_private_file+4183888/4965071] 0xe1d05b69
Jun 7 12:48:16 localhost kernel: [<e1d05b69>] 0xe1d05b69
Jun 7 12:48:16 localhost kernel: [__crc_open_private_file+4289670/4965071] 0xe1d1f89f
Jun 7 12:48:16 localhost kernel: [<e1d1f89f>] 0xe1d1f89f
Jun 7 12:48:16 localhost kernel: [__crc_ide_register_subdriver+529903/5323321] 0xe1e66155
Jun 7 12:48:16 localhost kernel: [<e1e66155>] 0xe1e66155
Jun 7 12:48:16 localhost kernel: [__crc_ide_destroy_dmatable+41604/68736] 0xe1dce96c
Jun 7 12:48:16 localhost kernel: [<e1dce96c>] 0xe1dce96c
Jun 7 12:48:16 localhost kernel: [__crc_ide_register_subdriver+528681/5323321] 0xe1e65c8f
Jun 7 12:48:16 localhost kernel: [<e1e65c8f>] 0xe1e65c8f
Jun 7 12:48:16 localhost kernel: [__crc_ide_register_subdriver+102434/5323321] 0xe1dfdb88
Jun 7 12:48:16 localhost kernel: [<e1dfdb88>] 0xe1dfdb88
Jun 7 12:48:16 localhost kernel: [__crc_ide_register_subdriver+101266/5323321] 0xe1dfd6f8
Jun 7 12:48:16 localhost kernel: [<e1dfd6f8>] 0xe1dfd6f8
Jun 7 12:48:16 localhost kernel: [__crc_open_private_file+4390542/4965071] 0xe1d382a7
Jun 7 12:48:16 localhost kernel: [<e1d382a7>] 0xe1d382a7
Jun 7 12:48:16 localhost kernel: [__crc_ide_register_subdriver+348125/5323321] 0xe1e39b43
Jun 7 12:48:16 localhost kernel: [<e1e39b43>] 0xe1e39b43
Jun 7 12:48:16 localhost kernel: [__crc_open_private_file+4300737/4965071] 0xe1d223da
Jun 7 12:48:16 localhost kernel: [<e1d223da>] 0xe1d223da
Jun 7 12:48:16 localhost kernel: [__crc_open_private_file+4306907/4965071] 0xe1d23bf4
Jun 7 12:48:16 localhost kernel: [<e1d23bf4>] 0xe1d23bf4
Jun 7 12:48:16 localhost kernel: [__crc_open_private_file+4174883/4965071] 0xe1d0383c
Jun 7 12:48:16 localhost kernel: [<e1d0383c>] 0xe1d0383c
Jun 7 12:48:16 localhost kernel: [__crc_open_private_file+4174859/4965071] 0xe1d03824
Jun 7 12:48:16 localhost kernel: [<e1d03824>] 0xe1d03824
Jun 7 12:48:16 localhost kernel: [tasklet_action+57/112] tasklet_action+0x39/0x70
Jun 7 12:48:16 localhost kernel: [<c0126749>] tasklet_action+0x39/0x70
Jun 7 12:48:16 localhost kernel: [do_softirq+97/192] do_softirq+0x61/0xc0
Jun 7 12:48:16 localhost kernel: [<c0126511>] do_softirq+0x61/0xc0
Jun 7 12:48:16 localhost kernel: [do_IRQ+229/256] do_IRQ+0xe5/0x100
Jun 7 12:48:16 localhost kernel: [<c010d845>] do_IRQ+0xe5/0x100
Jun 7 12:48:16 localhost kernel: [common_interrupt+24/32] common_interrupt+0x18/0x20
Jun 7 12:48:16 localhost kernel: [<c010bf28>] common_interrupt+0x18/0x20
Jun 7 12:48:16 localhost kernel:
Jun 7 12:48:16 localhost kernel: Badness in pci_find_subsys at drivers/pci/search.c:167
Jun 7 12:48:16 localhost kernel: Call Trace:
Jun 7 12:48:16 localhost kernel: [pci_find_subsys+215/224] pci_find_subsys+0xd7/0xe0
Jun 7 12:48:16 localhost kernel: [<c01c86f7>] pci_find_subsys+0xd7/0xe0
Jun 7 12:48:16 localhost kernel: [pci_find_device+24/32] pci_find_device+0x18/0x20
Jun 7 12:48:16 localhost kernel: [<c01c8718>] pci_find_device+0x18/0x20
Jun 7 12:48:16 localhost kernel: [pci_find_slot+26/96] pci_find_slot+0x1a/0x60
Jun 7 12:48:16 localhost kernel: [<c01c855a>] pci_find_slot+0x1a/0x60
Jun 7 12:48:16 localhost kernel: [__crc_open_private_file+4183888/4965071] 0xe1d05b69
Jun 7 12:48:16 localhost kernel: [<e1d05b69>] 0xe1d05b69
Jun 7 12:48:16 localhost kernel: [__crc_ide_register_subdriver+31545/5323321] 0xe1dec69f
Jun 7 12:48:16 localhost kernel: [<e1dec69f>] 0xe1dec69f
Jun 7 12:48:16 localhost kernel: [__crc_open_private_file+4289670/4965071] 0xe1d1f89f
Jun 7 12:48:16 localhost kernel: [<e1d1f89f>] 0xe1d1f89f
Jun 7 12:48:16 localhost kernel: [__crc_ide_destroy_dmatable+50101/68736] 0xe1dd0a9d
Jun 7 12:48:16 localhost kernel: [<e1dd0a9d>] 0xe1dd0a9d
Jun 7 12:48:16 localhost kernel: [__crc_ide_register_subdriver+362875/5323321] 0xe1e3d4e1
Jun 7 12:48:16 localhost kernel: [<e1e3d4e1>] 0xe1e3d4e1
Jun 7 12:48:16 localhost kernel: [__crc_ide_register_subdriver+529966/5323321] 0xe1e66194
Jun 7 12:48:16 localhost kernel: [<e1e66194>] 0xe1e66194
Jun 7 12:48:16 localhost kernel: [__crc_ide_destroy_dmatable+41604/68736] 0xe1dce96c
Jun 7 12:48:16 localhost kernel: [<e1dce96c>] 0xe1dce96c
Jun 7 12:48:16 localhost kernel: [__crc_ide_register_subdriver+528681/5323321] 0xe1e65c8f
Jun 7 12:48:16 localhost kernel: [<e1e65c8f>] 0xe1e65c8f
Jun 7 12:48:16 localhost kernel: [__crc_ide_register_subdriver+102434/5323321] 0xe1dfdb88
Jun 7 12:48:16 localhost kernel: [<e1dfdb88>] 0xe1dfdb88
Jun 7 12:48:16 localhost kernel: [__crc_ide_register_subdriver+101266/5323321] 0xe1dfd6f8
Jun 7 12:48:16 localhost kernel: [<e1dfd6f8>] 0xe1dfd6f8
Jun 7 12:48:16 localhost kernel: [__crc_open_private_file+4390542/4965071] 0xe1d382a7
Jun 7 12:48:16 localhost kernel: [<e1d382a7>] 0xe1d382a7
Jun 7 12:48:16 localhost kernel: [__crc_ide_register_subdriver+348125/5323321] 0xe1e39b43
Jun 7 12:48:16 localhost kernel: [<e1e39b43>] 0xe1e39b43
Jun 7 12:48:16 localhost kernel: [__crc_open_private_file+4300737/4965071] 0xe1d223da
Jun 7 12:48:16 localhost kernel: [<e1d223da>] 0xe1d223da
Jun 7 12:48:16 localhost kernel: [__crc_open_private_file+4306907/4965071] 0xe1d23bf4
Jun 7 12:48:16 localhost kernel: [<e1d23bf4>] 0xe1d23bf4
Jun 7 12:48:16 localhost kernel: [__crc_open_private_file+4174883/4965071] 0xe1d0383c
Jun 7 12:48:16 localhost kernel: [<e1d0383c>] 0xe1d0383c
Jun 7 12:48:16 localhost kernel: [__crc_open_private_file+4174859/4965071] 0xe1d03824
Jun 7 12:48:16 localhost kernel: [<e1d03824>] 0xe1d03824
Jun 7 12:48:16 localhost kernel: [tasklet_action+57/112] tasklet_action+0x39/0x70
Jun 7 12:48:16 localhost kernel: [<c0126749>] tasklet_action+0x39/0x70
Jun 7 12:48:16 localhost kernel: [do_softirq+97/192] do_softirq+0x61/0xc0
Jun 7 12:48:16 localhost kernel: [<c0126511>] do_softirq+0x61/0xc0
Jun 7 12:48:16 localhost kernel: [do_IRQ+229/256] do_IRQ+0xe5/0x100
Jun 7 12:48:16 localhost kernel: [<c010d845>] do_IRQ+0xe5/0x100
Jun 7 12:48:16 localhost kernel: [common_interrupt+24/32] common_interrupt+0x18/0x20
Jun 7 12:48:16 localhost kernel: [<c010bf28>] common_interrupt+0x18/0x20
Jun 7 12:48:16 localhost kernel:

Any clues?

jerickson314 06-07-04 03:43 PM

Re: X freezes with FX5200
 
That's what I thought may be causing it. This is a common issue you can search the forum for (Just search for "Badness in pci_find_subsys"). The biggest causes seem to be AGP problems, bad memory (test with memtest86 on its own boot disk), use of the framebuffer ("vga=" option on bootloader), and BIOS problems; there seems to be a wide variety of causes including ones I haven't listed. In my case it turned out to be a thermal problem, my card actually overheated so many times it eventually broke, so I bought a new one that has a fan and improved my case ventillation. I haven't had any problems but haven't tested thoroughly. The issue does seem to affect only kernel 2.6 users, the function being referenced is deprecated in kernel 2.6.

elad 06-09-04 08:20 AM

Re: X freezes with FX5200
 
I gave up and reverted to the "nv" driver. I don't need 3D at the moment, so I can live with that.
Searching Google for "Badness in pci_find_subsys" reveals that this problem goes back at least to October 2003. What's going on? Is nVidia ignoring Linux?
I had severe problems with their nforce2 chipset on Linux, which were not addressed for several months, until I swtiched to a VIA-based motherboard. Do these cases form a pattern?


All times are GMT -5. The time now is 12:08 AM.

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