Go Back   nV News Forums > Linux Support Forums > NVIDIA Linux

Newegg Daily Deals

Reply
 
Thread Tools
Old 07-15-07, 11:23 PM   #49
epoxi
Registered User
 
Join Date: Jul 2007
Posts: 2
Default Re: Random freezes on Fedora 7

Hi Team!

Hi have a similar situation. NV module crashes, making my whole system to freeze.

Motherboard: Asus M2N-E
CPU: AMD Athlon(tm) 64 X2 Dual Core Processor 4200+
GPU: GeForce 7300 GS (XFX PV-T72G-WANG)
RAM: 2 x 512 OCZ

OS: Debian Lenny / 2.6.21-2-amd64 #1 SMP Mon Jun 25 22:01:38 CEST 2007 x86_64 GNU/Linux

Here's what DMESG says when the module crashes:

NVRM: Xid (0007:00): 26, Ch 000001ff M 00001ffc D ffffffff intr ffffffff
NVRM: Xid (0007:00): 1, Ch 000001ff M 00001ffc D ffffffff intr ffffffff
NVRM: Xid (0007:00): 16, Head 00000000 Count 00082597
NVRM: Xid (0007:00): 16, Head 00000000 Count 000825a2
NVRM: Xid (0007:00): 16, Head 00000000 Count 000825ab
NVRM: Xid (0007:00): 16, Head 00000000 Count 000825b2
NVRM: Xid (0007:00): 16, Head 00000000 Count 000825c7
NVRM: Xid (0007:00): 16, Head 00000000 Count 000825d2
BUG: soft lockup detected on CPU#0!

Call Trace:
<IRQ> [<ffffffff802a24f8>] softlockup_tick+0xd5/0xe7
[<ffffffff80284137>] update_process_times+0x42/0x68
[<ffffffff802693aa>] smp_local_timer_interrupt+0x34/0x55
[<ffffffff80269ab6>] smp_apic_timer_interrupt+0x44/0x5c
[<ffffffff80255d56>] apic_timer_interrupt+0x66/0x70
<EOI> [<ffffffff884fd046>] :nvidia:_nv003521rm+0x20/0x22
[<ffffffff88702241>] :nvidia:_nv007344rm+0x7c/0xa3
[<ffffffff8880920f>] :nvidia:_nv001531rm+0x2f/0x106
[<ffffffff88807ca7>] :nvidia:_nv001553rm+0xd8/0x1aa
[<ffffffff887047e0>] :nvidia:_nv007714rm+0xde/0x15d
[<ffffffff886e6543>] :nvidia:_nv010297rm+0x12d/0x1c9
[<ffffffff886e6911>] :nvidia:_nv008698rm+0x10a/0x1aa
[<ffffffff886cd960>] :nvidia:_nv008558rm+0x77/0x191
[<ffffffff884ddea7>] :nvidia:_nv002734rm+0x3e0/0x40c
[<ffffffff884dd123>] :nvidia:_nv002735rm+0x110/0x2ac
[<ffffffff884d90e8>] :nvidia:_nv004588rm+0x68/0x95
[<ffffffff88507ed5>] :nvidia:_nv002683rm+0x1c5/0x6d9
[<ffffffff88504bc4>] :nvidia:rm_ioctl+0x2f/0x67
[<ffffffff888433a9>] :nvidia:nv_kern_ioctl+0x347/0x3c3
[<ffffffff88843464>] :nvidia:nv_kern_unlocked_ioctl+0x1c/0x23
[<ffffffff8023c1dc>] do_ioctl+0x21/0x6b
[<ffffffff8022cdc7>] vfs_ioctl+0x24e/0x267
[<ffffffff8022ce01>] __up_write+0x21/0x10d
[<ffffffff80245e24>] sys_ioctl+0x59/0x78
[<ffffffff8025511e>] system_call+0x7e/0x83



NVIDIA bug report log files is attached below.

I've tested almost all the sticky posts workarounds, but no one was useful... system still crashing.

Please, any suggestion will be well received

BRegards.-
ePoxi
Attached Files
File Type: gz nvidia-bug-report.log.gz (28.1 KB, 225 views)
epoxi is offline   Reply With Quote
Old 07-16-07, 12:36 AM   #50
MeSat
Registered User
 
Join Date: Jul 2007
Posts: 6
Default Re: Random freezes on Fedora 7

Quote:
Originally Posted by billydv
Lonnie, with all due respect, I have always bought Nvidia, and for a long time the original linux drivers worked relatively well, a few oddball problems but not the problems of today, I had been very happy but honestly, If someone is to spend several hundred dollars for a video card as I have on several, We should have some assistance from Nvidia with regards to Linux drivers, not only the driver but also the implementation of the driver on the system. Admittedly, I believe Nvidia has given the linux community alot of help with what theyve done till now but In my opinion ,Nvidia has to do more with regards to providing recommendations as to the implementation of the driver, xorg.conf guidelines and kernel config guidelines to get systems working. I have now had my new kernel running with the anticipatory scheduler instead of cfq for almost 24 hrs with no freezeups or delays whatsoever and I am becoming surer and surer that this was my problem. Nvidia should spend more time running linux systems in order to try to get these glitches worked out.
I'm not saying I'm mad at Nvidia as I will continue to support Nvidia over ATI but it would sure be nice if Nvidia could support us linux users just a little bit more.

Thanks
Billydv, I fully agree.

From this thread, and your test, it is showing that the bug is associated to more than just the Nvidia driver. I was thinking about this as I have tried various drivers with various versions of lockups. This is surely looking like an issue with a change to Xorg or as your test is showing, the kernel.

Now, if we all posted information into individual bug reports, many would miss this information as there would be to many bugs to read through. What this thread is showing is the issue is across platforms and due to something recent. Not isolated to a single application or piece of hardware. The initial indication is the video driver as the driver is changed and then the problem occurs. Naturally, many want to dump on Nvidia for this issue.

Now, in regards to bug reporting, many users won't take the time to complain. The demands on many to follow the instructions for file a bug report are to hard. Details that may be useful won't be submitted so Nvidia then misses a chance for information.

Your post has shown that a change to the kernel may be the real cause. It may be worth looking at in all cases to see if it is a common issue. I tend to agree.

I also agree that Nvidia's support for Linux and from a friend that uses Windows, better than ATI's support. I am now stuck because I want to upgrade my systems but they are all Nvidia based systems and I cannot deal with all the freezes.

The question, is this information supplied to RedHat and the kernel development team for troubleshooting on their part?
MeSat is offline   Reply With Quote
Old 08-10-07, 06:18 AM   #51
kahuna0k
Registered User
 
Join Date: Mar 2003
Posts: 26
Default Re: Random freezes on Fedora 7

I have a Dell Precision M65 with debian Etch (x86_64), running latest 100.04.11 and was suffering the same kind of lockups. I solved it in two different ways, first the only thing that solved it for the standard 2.6.18 debian kernel was adding the idle=poll as a kernel parameter. Later I updated my system to the latest 2.6.22.1 kernel in order to get s2ram and s2disk fully working with the nvidia module loaded and X running and now I have no need to add the idle=poll parameter. Finally I can say that EVERYTHING in my laptop, even the fingerprint reader and the smartcard is fully working in debian, s2ram and s2disk is also working, even with beryl running. And finally with a little help of the powertop tool I can squeeze about seven hours from a fully charged battery. Really Happy!

When I get some spare time I will make a web explaining the full process.

Regards,

- german
kahuna0k is offline   Reply With Quote
Old 08-10-07, 07:47 AM   #52
slavagt
Registered User
 
Join Date: Aug 2007
Posts: 39
Default Re: Random freezes on Fedora 7

As I see this a big problem for all, who have dual core CPU. All "random lookup" problems which have many peoples with different distros and different hardware, have only one common things - Dual Core CPU (AMD Athlon X2 or Intel Core DUO).
Please, who have such problem, try to boot kernel in sinfgle CPU mode and check system stability. For booting kernel in single CPU mode, you need only to add "maxcpus=1" to kernel startup options in /boot/grub/grub.conf. Or edit boot string then startup ("e") then you see grub menu for temporary adding this kernel boot options.
Please, all who have such problem, check this solution. For me, it make system "rock solid" . This not real solution, because with this option the system work at one core only, but it will be help to find problem, and can be temporary solution up to problem was finding.
=============
Also, for testing it's better to use SPECviewperf (SPECViewperf9.0.3.tar.gz from http://www.spec.org/gwpg/downloadindex.html). This test make heavy load at GPU, and if something wrong with video subsystem, system will be lookup in 3-10 minutes. Also, you can test system stability by mprime -t + mprime -t -a 1 (in console mode) for checking CPU, memory, motherboard and power supply stability before testing video subsystem stability.
slavagt is offline   Reply With Quote
Old 08-10-07, 10:14 AM   #53
sambo57u
Registered User
 
Join Date: Jan 2006
Posts: 52
Default Re: Random freezes on Fedora 7

For all that matters, I am running 4 dual and quad core systems,
DELL Precision 390 and 380 with latest bios, running Fedora 7 + updates, and
Nvidia 100.14.11 installed directly (not livna rpms). I have not yet
encountered a single freeze. Is everyone here using livna? Livna is
usually fine but better make sure you clean up well if you have an
older nvidia installation.
sambo57u is offline   Reply With Quote
Old 08-10-07, 04:53 PM   #54
slavagt
Registered User
 
Join Date: Aug 2007
Posts: 39
Default Re: Random freezes on Fedora 7

Quote:
Originally Posted by sambo57u
For all that matters, I am running 4 dual and quad core systems,
DELL Precision 390 and 380 with latest bios, running Fedora 7 + updates, and
Nvidia 100.14.11 installed directly (not livna rpms). I have not yet
encountered a single freeze. Is everyone here using livna? Livna is
usually fine but better make sure you clean up well if you have an
older nvidia installation.
Did you can try Specviewperf test (SPECViewperf9.0.3.tar.gz from http://www.spec.org/gwpg/downloadindex.html) ? Did it pass ?
slavagt is offline   Reply With Quote
Old 08-10-07, 09:02 PM   #55
MichaelFischer
kingfischer289
 
Join Date: Aug 2007
Posts: 13
Default Re: Random freezes on Fedora 7

I'm also experiencing lockups, generally followed a short time later by kernel crashes:
Processor: AMD 64 X2 Dual Core Processor 4400+
Motherboard: ASUS A8N-E, bios 1013 (latest)
OS: Fedora 7
Linux Kernel: 2.6.22.1-41.fc7
Nvidia driver: 100.14.11
Video card: GeForce 7300 GS rev 161

Unfortunately, the kernel crashes generally prevent me from running nvidia-bug-report.sh. Once I was able to log in remotely and run it, but the bug reporter itself hung after writing the line "/proc/driver/nvidia/cards/0" to the log file.

I've tried several workarounds, and none of them solved my problem: idle=poll, pci=nommconf, elevator=as. I'm currently running with maxcpus=1, and everything seems rock solid (but I miss the second core).

I've been having lockups once or twice a day, but when I switched to a dual monitor configuration a couple of days ago, they began happening frequently -- within 5 or 10 minutes after rebooting.

Lockups definitely occurred with the 100.14.09 driver. I'm not sure whether they occurred with the 9xxx drivers and/or 2.6.21 and before kernels.

I can supply more information if that would be useful, but now I'm just confirming what others have observed.
MichaelFischer is offline   Reply With Quote
Old 08-11-07, 02:07 AM   #56
roshi
Registered User
 
Join Date: Mar 2007
Posts: 19
Cool Re: Random freezes on Fedora 7

Hi there.
Same problem here, although not so frequently (I'm glad).

My system is a Dell m1210 laptop running Sabayon (Gentoo based) GNU/Linux
Dual Core T7200 2GHz 4MB L2 cache
Geforce 7400 TC (64MB + 192MB shared with the main memory 2GB of DDR)

Apparently common among the people experiencing this problem.
Running Beryl and KDE.

My system will gradually degrade performance as days go by (until the next forced reboot).
It might be a couple of days before the first flick of the screen.
This has an associated
Aug 11 13:31:29 patino [93844.374000] NVRM: Xid (0001:00): 6, PE0000 0374 00000000 0000fc80 ff000000 03200000
Aug 11 13:31:29 patino [93844.391000] NVRM: Xid (0001:00): 30, L0 -> L0
as you can see in the log file. Apparently the Xid most frequent numbers change over time/drivers. Or maybe it's just my feeling. But lately the pair 6+30 has been the most common.

No problem here. The system keeps on working, but it flickers more often.
I changed Nvidia drivers to 100.14.11 actually to try to get rid of the black window bug that is more problematic for TC boards.

I can get my system to stay up for about 1 month max (until now).
As I said, as days go by I get more black windows, the more the screen flickers and the bigger X memory footprint gets. (more that 1GB)

Then its time for a ctrl+alt+backspace. I can tell you that it does wonders.
It's almost like rebooting. At least it cleans the main memory. Not sure if it cleans the GPU memory because the black window happens sooner than after a reboot. (or maybe its just my impression)

I also experience X's random locks. 99% of the time after a few seconds the system gets back to the "normal", and yap, it leaves a trace of 100% CPU1 usage (CPU0 apparently is not affected).

I can say that about 100% of this kind of lockup happened while using the touch pad to scroll down some web page in Firefox.

But even when the system doesn't recovers it continues happily playing music in the background and doing downloads etc. Therefore, although I haven't tried to do a remote login on such occasions, I'm sure that it would be possible because I'm always listening to web radio...

Back in the past I used to try some different solutions to try to solve these problems, but from what I can tell from my own experience and from what I read from other's usually it doesn't solve the problem.
I believe that it might be drivers related...
So I'll be waiting for the drivers that solve the back windows. Maybe the other problems will go away at the same time.

So people at Nvidia do your best. (nice to see that job opening for a linux CG eng)
Attached Files
File Type: bz2 nvidia-bug-report.log.bz2 (27.7 KB, 175 views)
roshi is offline   Reply With Quote

Old 08-11-07, 11:54 AM   #57
JoachimF
Registered User
 
Join Date: Aug 2007
Posts: 17
Default Re: Random freezes on Fedora 7

Quote:
Originally Posted by billydv
I am still testing this but I think I may have found the solution and it has nothing to do with module options or xorg.conf options. I have an amd64 system with multiprocessor cpu running gentoo. As soon as I got to thinking this might be the problem I compiled two kernels identically with just a difference in schedulers, One with the commonly used cfq and the other with the old default anticipatory. The cfq kernel definitely freezes if left alone, return from screensavers and either no return to desktop or mouse movement with frozen screen, the anticipatory since yesterday afternoon does not, absolutely not a single freeze or delay. I am still testing but you should try to see if it changes anything on your computer. To boot a stock kernel with the anticipatory scheduler, add elevator=as to your kernel boot line.
I got my hopes up when I saw your post about the scheduler, and immediately tried it, but no luck. It froze up solid after about 20 minutes playing GWeled (not my fav game, but if I'd triggered it with Warcraft III, someone might have tried blaming Wine. Using WC III as trigger, the machine freezes in under a minute.)

My problem seems similiar to what others mention;
* 7600GS
* Latest MB BIOS
* Fedora 7 2.6.22.1-41 (currently)
* Driver version 100.14.11 (currently)
* Freezes only occur when the nvidia driver is enabled. Machine is solid when the nvidia driver is disabled.
* Freezes are HARD - (magic SysRq-key doesn't work) - and no log entries are generated insofar as I can tell.
* Freezes quicker when in 3D, therefore I have disabled Compiz/Beryl
* Machine is rock solid in 3D when using Windows XP.
* Card temp is not an issue (In linux, freezes can be triggered even in 2D and while card temp is in the 40s. In XP the same machine is solid while cycling 3DMark on an OVERCLOCKED video card (=card temp much higher)... And I'm not overclocking the card in Linux.)
* Tried numerous kernels (FC6 as well as several versions of Fedora 7)
* Tried numerous driver versions (recompiled for each kernel version)

EDIT: Since the machine hard-freezes, I obviously can't run 'nvidia-bug-report.sh' when the bug has occurred. Is it any use running it anyway? /EDIT
JoachimF is offline   Reply With Quote
Old 08-11-07, 03:29 PM   #58
slavagt
Registered User
 
Join Date: Aug 2007
Posts: 39
Default Re: Random freezes on Fedora 7

Quote:
Originally Posted by kyleki
Man, no more than 30 seconds went by after sending this when it froze again! So, nevermind on the kernel parameter being a fix although it did seem to prolong the freeze a little bit. However, due to the random nature of these freezes, it could just simply be my perception of time before this last freeze.

Back to the drawing boards. At this point I believe I've tried everything suggested in the sticky posts and am still having this problem. Any suggestions?
Did you have dual core CPU ? If dual core, please try "maxcpus=1" bootup kernel options. Did it help ?
slavagt is offline   Reply With Quote
Old 08-11-07, 06:21 PM   #59
JoachimF
Registered User
 
Join Date: Aug 2007
Posts: 17
Default Re: Random freezes on Fedora 7

Quote:
Originally Posted by slavagt
Did you have dual core CPU ? If dual core, please try "maxcpus=1" bootup kernel options. Did it help ?
Can't speak for kyleki, but it didn't work for me.

I checked against the sticky posts in this forum; http://www.nvnews.net/vbulletin/showthread.php?t=58498
- dmesg already reporting 'PCI:'Not using MMCONFIG'
- tried 'idle=poll'. Still freezes
- motherboard already at latest BIOS.

Also tried 'elevator=as' as per billydv:s suggestion. Still freezes.

Latest attempt was your suggestion; 'maxcpus=1'. (actually, I left the 'idle=poll' in there too). Still freezes.
JoachimF is offline   Reply With Quote
Old 08-11-07, 08:55 PM   #60
MichaelFischer
kingfischer289
 
Join Date: Aug 2007
Posts: 13
Default Re: Random freezes on Fedora 7

Quote:
Originally Posted by JoachimF
Can't speak for kyleki, but it didn't work for me.

I checked against the sticky posts in this forum; http://www.nvnews.net/vbulletin/showthread.php?t=58498
- dmesg already reporting 'PCI:'Not using MMCONFIG'
- tried 'idle=poll'. Still freezes
- motherboard already at latest BIOS.

Also tried 'elevator=as' as per billydv:s suggestion. Still freezes.

Latest attempt was your suggestion; 'maxcpus=1'. (actually, I left the 'idle=poll' in there too). Still freezes.
maxcpus=1 works for me. I've been up for over 20 hours without a single problem, and the log file doesn't show any of those nasty
NVRM: Xid (0001:00): 16, Head 00000001 Count 00011617
messages. Try it without the "idle=poll".
MichaelFischer is offline   Reply With Quote
Reply


Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


Similar Threads
Thread Thread Starter Forum Replies Last Post
Fedora 16, GTX 550Ti, random X hangs, "Attempted to yield the CPU..." MquwIBUO NVIDIA Linux 7 06-12-12 04:39 PM
Tex Is Back: Project Fedora Receives Its Funding News Archived News Items 0 06-07-12 10:50 AM
Random loot packs are the best-selling item in free-to-play City of Heroes News Archived News Items 0 05-06-12 12:00 PM
crashes and freezes Lethal Weapon NVIDIA Linux 9 09-27-02 09:23 PM
Recompiled Kernel freezes H3g3m0n NVIDIA Linux 4 08-17-02 11:09 AM

All times are GMT -5. The time now is 03:47 AM.


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