two problems with Nvidia_kernel-1.0-3123

adrin · 12-02-02, 07:52 PM

I have two potentially related problems, which I hope gurus on this list will be able to resolve:

I have an AMD Athlon MP-based linux box with a Quadro 4 XGL 700 installed.

With Red Hat 7.3, kernel version 2.4.18-3 and Nvidia_kernel-1.0-2960 and its corresponding GLX files everything works ok.

I've upgraded the kernel to version 2.4.18-18.7.x; and I've downloaded Nvidia_kernel-1.0-3123 for Athlon and the corresponding GLX files. I had difficulty with the rpm approach, so I've used the tar files to install.

Everything compiles (and installs?) without error messages; however:

(1) The NVdriver does not load automatically at boot time, and yes "alias char-major-195 NVdriver" exists in /etc/modules.conf

(2) For some reason the driver does not recognize the monitor and when I use startx, the monitor signals "No Sync" and the linux box reboots itself!

GLcore and dri modules are deleted in XF86Config-4; glx is loaded, and "vesa" is replaced with "nvidia" in the same config file. The monitor I use is Viewsonic VX800, 18" digital flat screen, so as with the previous kernel version that worked perfectly well, I have assigned the horizontal and vertical frequency ranges as well as the resolution in XF86Config-4. However, for some reason startx just won't start properly.

I have checked ALL the files for their existence in the correct directories and with the correct links, as outlined in the NVLinuxR25Notes_3123 notes and have exhausted all tests.

I'd really appreciate any help I can get

adrin

bwkaz · 12-03-02, 08:32 AM

For the automatic loading, you can try changing alias char-major-195 NVdriver to be alias /dev/nvidia* NVdriver instead. Run /sbin/depmod -a afterward.

Then, the easiest way to test is probably to just reboot and try to start X up.

Not sure about the other one, but maybe an Option "ConnectedMonitor" "DFP" in your XF86Config(-4) file?

adrin · 12-03-02, 04:15 PM

I tried "alias /dev/nvidia* NVdriver" + "/sbin/depmod -a" + "reboot" and the NVdriver module still did not load automatically!

Also tried Option "ConnectedMonitor" "DFP" with no luck with X!

Thanks

adrin

bwkaz · 12-03-02, 04:49 PM

Does the module load when you "modprobe NVdriver"?

Are you sure the /dev/nvidia[0123] device files exist?

adrin · 12-20-02, 07:21 PM

Does the module load when you "modprobe NVdriver"?

Yes

Are you sure the /dev/nvidia[0123] device files exist?

/dev/nvidia[0..9 & ctl] device files exist

Andy Mecham · 12-21-02, 01:51 AM

Have you tried 1.0-4191?

--andy

adrin · 12-22-02, 12:31 AM

Right after I submitted my previous post I tried 1.0-4191!

Unfortunately, _exactly_ the same problem(s) again.

I have compiled the source codes (no rpms were used) and there were no errors whatsoever. I have used the correct(?) XF86Config-4; well it's the same as the one that works fine with version 2960 and the earlier linux kernel.

I have checked all the required files and links as per the nvidia document. I have also read the FAQ's etc. I don't see any error messages in the XFree86.0.log file (perhaps I'm not looking at all the files? - I've used verbose 5) My modules.conf has "alias char-major-195 nvidia". The /dev/nvidia* files are there, etc. etc. etc.

I have upgraded my Red Hat 7.3 kernel to 2.4.18-19.7.x and all the required files have higher versions than those required in the nvidia manual.

The problems are, again:

(1) The nvidia module will not load automatically at boot time and I have tried all the recommendations on this thread

(2) subsequent to loading the module manually I run startx but rather than the program crashing and printing a message on the screen (short of doing the right thing, of course!) it automatically reboots the system!

I have tried the process using ViewSonic P95f+ CRT and ViewSonic VX800 DFP, both give the same headache.

I have tried everything I know for well over 2 weeks now and I'm REALLY VERY frustrated.

BTW, I've checked the thread on 1.0-4191 and it seems that others have had sort of similar problems

I would appreciate it if someone would "babysit" me through checking the proper files for possible errors to finally take care of this issue. If there is a particular file that might help resolve this issue please let me know and I'll include it here

Thanks

adrin

12-02-02, 07:52 PM	#1
adrin Registered User Join Date: Dec 2002 Posts: 4	two problems with Nvidia_kernel-1.0-3123 I have two potentially related problems, which I hope gurus on this list will be able to resolve: I have an AMD Athlon MP-based linux box with a Quadro 4 XGL 700 installed. With Red Hat 7.3, kernel version 2.4.18-3 and Nvidia_kernel-1.0-2960 and its corresponding GLX files everything works ok. I've upgraded the kernel to version 2.4.18-18.7.x; and I've downloaded Nvidia_kernel-1.0-3123 for Athlon and the corresponding GLX files. I had difficulty with the rpm approach, so I've used the tar files to install. Everything compiles (and installs?) without error messages; however: (1) The NVdriver does not load automatically at boot time, and yes "alias char-major-195 NVdriver" exists in /etc/modules.conf (2) For some reason the driver does not recognize the monitor and when I use startx, the monitor signals "No Sync" and the linux box reboots itself! GLcore and dri modules are deleted in XF86Config-4; glx is loaded, and "vesa" is replaced with "nvidia" in the same config file. The monitor I use is Viewsonic VX800, 18" digital flat screen, so as with the previous kernel version that worked perfectly well, I have assigned the horizontal and vertical frequency ranges as well as the resolution in XF86Config-4. However, for some reason startx just won't start properly. I have checked ALL the files for their existence in the correct directories and with the correct links, as outlined in the NVLinuxR25Notes_3123 notes and have exhausted all tests. I'd really appreciate any help I can get adrin

12-03-02, 08:32 AM	#2
bwkaz Registered User Join Date: Sep 2002 Posts: 2,262	For the automatic loading, you can try changing alias char-major-195 NVdriver to be *alias /dev/nvidia NVdriver instead. Run /sbin/depmod -a afterward. Then, the easiest way to test is probably to just reboot and try to start X up. Not sure about the other one, but maybe an Option "ConnectedMonitor" "DFP"** in your XF86Config(-4) file? __________________ Registered Linux User #219692

12-03-02, 04:49 PM	#4
bwkaz Registered User Join Date: Sep 2002 Posts: 2,262	Does the module load when you "modprobe NVdriver"? Are you sure the /dev/nvidia[0123] device files exist? __________________ Registered Linux User #219692

12-21-02, 01:51 AM	#6
Andy Mecham l33t master Join Date: Jul 2002 Location: Santa Clara, CA Posts: 1,163	Have you tried 1.0-4191? --andy __________________ Andy Mecham NVIDIA Corporation

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Hengeveld: Big Data Meets HPC to Solve Hard Problems and Improve Lives	News	Latest Tech And Game Headlines	0	09-10-12 09:10 AM
problems with drivers after 295.40 (OpenSuSE 12.1, GeFroce 7900 GS, Phenom II X4)	alawa	NVIDIA Linux	0	06-09-12 06:37 PM
3123 freezing up	jon787	NVIDIA Linux	7	10-08-02 10:27 AM
RH7.3, 3123, was working, now it isn't	phips	NVIDIA Linux	2	10-03-02 02:08 PM
3123 drivers cause system hang	kgf	NVIDIA Linux	5	09-28-02 07:32 AM

12-03-02, 04:15 PM	#3
adrin Registered User Join Date: Dec 2002 Posts: 4	I tried "alias /dev/nvidia* NVdriver" + "/sbin/depmod -a" + "reboot" and the NVdriver module still did not load automatically! Also tried Option "ConnectedMonitor" "DFP" with no luck with X! Thanks adrin

12-20-02, 07:21 PM	#5
adrin Registered User Join Date: Dec 2002 Posts: 4	Does the module load when you "modprobe NVdriver"? Yes Are you sure the /dev/nvidia[0123] device files exist? /dev/nvidia[0..9 & ctl] device files exist

12-22-02, 12:31 AM	#7
adrin Registered User Join Date: Dec 2002 Posts: 4	Right after I submitted my previous post I tried 1.0-4191! Unfortunately, _exactly_ the same problem(s) again. I have compiled the source codes (no rpms were used) and there were no errors whatsoever. I have used the correct(?) XF86Config-4; well it's the same as the one that works fine with version 2960 and the earlier linux kernel. I have checked all the required files and links as per the nvidia document. I have also read the FAQ's etc. I don't see any error messages in the XFree86.0.log file (perhaps I'm not looking at all the files? - I've used verbose 5) My modules.conf has "alias char-major-195 nvidia". The /dev/nvidia* files are there, etc. etc. etc. I have upgraded my Red Hat 7.3 kernel to 2.4.18-19.7.x and all the required files have higher versions than those required in the nvidia manual. The problems are, again: (1) The nvidia module will not load automatically at boot time and I have tried all the recommendations on this thread (2) subsequent to loading the module manually I run startx but rather than the program crashing and printing a message on the screen (short of doing the right thing, of course!) it automatically reboots the system! I have tried the process using ViewSonic P95f+ CRT and ViewSonic VX800 DFP, both give the same headache. I have tried everything I know for well over 2 weeks now and I'm REALLY VERY frustrated. BTW, I've checked the thread on 1.0-4191 and it seems that others have had sort of similar problems I would appreciate it if someone would "babysit" me through checking the proper files for possible errors to finally take care of this issue. If there is a particular file that might help resolve this issue please let me know and I'll include it here Thanks adrin