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

Newegg Daily Deals

Closed Thread
 
Thread Tools
Old 05-10-05, 11:39 AM   #133
LubosD
Registered User
 
Join Date: Jan 2005
Location: Czech Republic
Posts: 451
Send a message via ICQ to LubosD
Default Re: Trying to make headway into finding the Xid crashes source...

No freeze here. I've tried this (after swapping buffers):

glReadPixels(0, 0, 10, 10, GL_RGB, GL_INT, bigbuf);
LubosD is offline  
Old 05-10-05, 08:30 PM   #134
nukem
Registered User
 
Join Date: Dec 2004
Posts: 226
Default Re: Trying to make headway into finding the Xid crashes source...

I tired LubosD's crash app and nothing happened to my system and nothing was added in dmesg.

cpbotha: Could you please varify that this code is simlar to yours in what it does? If it is not please write an app to reproduce the crash.
nukem is offline  
Old 05-10-05, 10:14 PM   #135
payans
Registered User
 
payans's Avatar
 
Join Date: Apr 2005
Location: Wellington, New Zealand
Posts: 15
Default Re: Trying to make headway into finding the Xid crashes source...

Quote:
Originally Posted by LubosD
Hmm, it seems that strange behavior isn't a nVidia's problem only.

So where is that bug? In X.org?
I had this problem with an ATI card, and it turned out to be caused by using AGPv3. Finding a way in the PC's BIOS to force AGPv2, and disabling FastWrites, fixed the problem for several people in the Rage3d thread referenced earlier.

By the time I confirmed that this was a good fix for the ATI card, I had already bought an NVIDIA card to replace it. X crashed on startup with the NVIDIA card until I went back to using AGPv2.

I got the impression that it was a bug or oversight in the kernel AGP driver for AGPv3, that only trips on some hardware.
payans is offline  
Old 05-10-05, 10:48 PM   #136
nukem
Registered User
 
Join Date: Dec 2004
Posts: 226
Default Re: Trying to make headway into finding the Xid crashes source...

payans: Did you try writing a bug report to the kernel? If so could you provide a link? Do you have any other information? Ill try to confirm this bug.
nukem is offline  
Old 05-10-05, 10:51 PM   #137
deadhamster
Registered User
 
Join Date: Oct 2002
Posts: 10
Default Re: Trying to make headway into finding the Xid crashes source...

I have AGP disabled all together, so I doubt AGP is the problem.

-DH
deadhamster is offline  
Old 05-10-05, 11:56 PM   #138
nukem
Registered User
 
Join Date: Dec 2004
Posts: 226
Default Re: Trying to make headway into finding the Xid crashes source...

My bios does not support changing AGP versions(ASUS A8V Deluxe). It does however support disabling AGP Fast Write and AGP 3.0 Calibration Cycle. I tried disabling both of those to see if that would fix the XiD error, it did not. Im not sure if PCI-X users have the XiD error, if you think its the AGP fault then make sure it dosnt happen on a PCI or PCI-X graphics card.

deadhamster: How did you disable AGP? If you have an AGP card you need to have AGP enabled. Did you switch to a PCI or PCI-X card?
nukem is offline  
Old 05-11-05, 12:02 AM   #139
payans
Registered User
 
payans's Avatar
 
Join Date: Apr 2005
Location: Wellington, New Zealand
Posts: 15
Default Re: Trying to make headway into finding the Xid crashes source...

Quote:
Originally Posted by nukem
payans: Did you try writing a bug report to the kernel? If so could you provide a link? Do you have any other information? Ill try to confirm this bug.
No, I don't think the information I've got is strong enough to claim that there's a definite bug. It may be that the motherboard (GA-8KNXP) is not to spec. Of the four or five posters with the problem on Rage3D, two of us had the same model of this motherboard. I've read elsewhere (I think in the Xorg driver documentation, and maybe even the NVIDIA README) that AGPv3 implementations can be flaky.

How could one confirm it? Would you have to trace AGP activity on a board that exhibited the problem to see where it hung? I've switched out the ATI video card, and the problem I had with NVIDIA and AGPv3 was different, so I can't reproduce the random freeze now.
payans is offline  
Old 05-11-05, 01:33 AM   #140
Soul-Crusher
Leadtek 6800GT
 
Join Date: Aug 2002
Location: Albuquerque, NM, USA
Posts: 195
Send a message via ICQ to Soul-Crusher Send a message via AIM to Soul-Crusher Send a message via Yahoo to Soul-Crusher
Default Re: Trying to make headway into finding the Xid crashes source...

Quote:
Originally Posted by Saulzar
What I noticed is that a simple application (400fps+) will crash _immediately_ within 2 minutes every time. However perhaps not glxgears for some odd reason.
I seem to be having the exact same problem as you. I tried doing what gilboa did at first (strace glxgears) and noticed that when I ran glxgears with strace going, my framerate would drop drastically but everything seemed to run stable. However, running glxgears at normal speed (12,000+ FPS) resulted in glxgears taking up 99.9% of my CPU and causing X to fail to recognize keyboard events (ctrl-alt-backspace to kill it). I could still SSH in from another machine, and when I killed glxgears X would still run just fine.

This is on a Gigabyte K8-Triton (GA-K8VM800M using a VIA K8M800M chipset). I'm running the 7174 series kernel module and GLX libraries and have updated my BIOS to the latest revision. NVAGP and AGPGART/AMD64_AGP act the same. I have not tried reverting drivers or X versions as of yet, but I will try those out some time.
__________________
"Floating in a dream-like state, I am the emporer of a parallel universe." --Arch Enemy
My xorg.conf file
Soul-Crusher is offline  

Old 05-11-05, 01:49 AM   #141
gilboa
Linux addict...
 
Join Date: Jan 2004
Posts: 540
Default Re: Trying to make headway into finding the Xid crashes source...

Something tells me that Xid is a catch alll event: No matter what fails (AGP, driver, kernel, etc) it'll fall into this sink.

My board is a AGP4x board and I've seen reports from people with Intel (!!!) 845 boards, so I doubt that it's only a not-being-up-to-AGP-spec problem.

Cheers,
__________________
DEV-NG: Intel S2600C0, 2xE52658V2, 32GB, 4x2TB, GTX680, F19/x86_64, Dell U2711.
DEV: Intel S5520SC, 2xX5680, 36GB, 5x320GB, GTX550, F19/x86_64, Dell U2711 (^).
SRV: Tyan Tempest i5400XT, 2xE5335, 8GB, 4x2TB, 9800GTX, F19/x86-64, Dell U2412.
LAP: ASUS N56VJ, i7-3630QM, 16GB, 1TB, 635M, F19/x86_64.
gilboa is offline  
Old 05-11-05, 02:09 AM   #142
Soul-Crusher
Leadtek 6800GT
 
Join Date: Aug 2002
Location: Albuquerque, NM, USA
Posts: 195
Send a message via ICQ to Soul-Crusher Send a message via AIM to Soul-Crusher Send a message via Yahoo to Soul-Crusher
Default Re: Trying to make headway into finding the Xid crashes source...

Interesting... I started using APIC as someone else suggested earlier in this thread. That didn't help, but I got a message in my kernel log:
** PCI interrupts are no longer routed automatically. If this
** causes a device to stop working, it is probably because the
** driver failed to call pci_enable_device(). As a temporary
** workaround, the "pci=routeirq" argument restores the old
** behavior. If this argument makes the device work again,
** please email the output of "lspci" to bjorn.helgaas@hp.com
** so I can fix the driver

So I did just that and added pci=routeirq to my kernel line in grub on startup. Now glxgears seems to run just fine at normal speed without wrenching anything.

Edit: It seems that wasn't a permanent fix, but I would suggest trying it out anyway.
__________________
"Floating in a dream-like state, I am the emporer of a parallel universe." --Arch Enemy
My xorg.conf file

Last edited by Soul-Crusher; 05-11-05 at 04:40 AM.
Soul-Crusher is offline  
Old 05-11-05, 06:05 AM   #143
cpbotha
Registered User
 
Join Date: May 2005
Posts: 7
Default Re: Trying to make headway into finding the Xid crashes source...

@LubosD: At the moment, you are probably one of the few people on this forum that can build my example. It requires VTK and uses cmake.

Download my crasher-application source code here:
http://visualisation.tudelft.nl/~cpb...nvcrash.tar.gz

1. untar
2. cd vtkFlRenderWindowInteractor/examples/Cone3
3. ccmake . (and complete all the field as usual for cmake with VTK)
4. make

Now run ./Cone3 and click on the "NV Crash" button. This might not work the first time, so keep on trying.

For the NVIDIA people, if they should ever read this:
This is more than just creating multiple contexts and then destroying them. A gl capable window is created with a specific x11 Display. This windowId is passed to VTK which will subsequently render to this windowId but will create its own Display. Herein lies the problem.

If the windowId AND its x11 Display are passed to VTK, there is no problem. If however only the windowId is passed but no Display, and VTK has to XOpenDisplay its own display, this freeze frequently happens when this window is finally destroyed. It seems that it comes down to using glX calls on a windowId that was created with another Display instance.

This might be considered a misbehaved client application, but no driver should allow such a misbehaved application to freeze the whole machine for minutes on end!
cpbotha is offline  
Old 05-11-05, 06:07 AM   #144
cpbotha
Registered User
 
Join Date: May 2005
Posts: 7
Default Re: Trying to make headway into finding the Xid crashes source...

Quote:
Originally Posted by LubosD
@cpbotha: I tried to create my own crashing application, but it doesn't crash for me. Am I doing it right? Source code is included...

UPDATE: @*: You can compile the file by renaming to main.c and exec'ing this command:
Code:
gcc -o nvcrash main.c -lGL
LubosD, please see my later post with the vtkFlRenderWindowInteractor posting.
cpbotha is offline  
Closed Thread


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
Random crashes, NVRM Xid messages Iesos NVIDIA Linux 90 10-04-12 04:27 AM

All times are GMT -5. The time now is 12:47 PM.


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