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

Newegg Daily Deals

Reply
 
Thread Tools
Old 02-02-09, 04:19 AM   #1
Xilman
Registered User
 
Join Date: Feb 2009
Posts: 2
Default "NVRM: API mismatch" installing CUDA 2.1 on RHEL5

I am attempting to install CUDA 2.1 on a RHEL5 X86_64 box.

Messages like these keep appearing in /var/log/messages
Code:
Feb  1 22:35:23 localhost kernel: NVRM: loading NVIDIA UNIX x86_64 Kernel Module  169.12  Thu Feb 14 17:51:09 PST 2008
Feb  1 22:35:23 localhost kernel: NVRM: API mismatch: the client has the version 180.22, but
Feb  1 22:35:23 localhost kernel: NVRM: this kernel module has the version 169.12.  Please
Feb  1 22:35:23 localhost kernel: NVRM: make sure that this kernel module and all NVIDIA driver
Feb  1 22:35:23 localhost kernel: NVRM: components have the same version.
and X11 fails to start.

Google and the nvnews search each indicate that driver version mismatch has been a common problem over many years. Nothing I have yet found has helped me resolve it. Most solutions appear to be for Ubuntu rather than RedHat. According to uname -a my system is

Linux localhost.localdomain 2.6.18-128.el5 #1 SMP Wed Dec 17 11:41:38 EST 2008 x86_64 x86_64 x86_64 GNU/Linux

The process, which I have run several times now with various small changes to see whether they help, has been:

1) Boot the machine
2) telinit 3
3) Login in as root
4) rmmod nvidia
5) sh NVIDIA-Linux-x86_64-180.22-pkg2.run
6) Reboot the machine and discover that X does not start.
7) sh NVIDIA-Linux-x86_64-180.22-pkg2.run --uninstall
8) Reboot the machine and X runs perfectly.

Attached are gzipped versions of various log files as per the instructions.


Any assistance would be much appreciated.


Paul
Attached Files
File Type: gz messages.gz (81.2 KB, 72 views)
File Type: gz nvidia-installer.log.gz (4.8 KB, 64 views)
File Type: gz nvidia-bug-report.log.gz (27.4 KB, 73 views)
File Type: gz Xorg.0.log.gz (4.0 KB, 72 views)
Xilman is offline   Reply With Quote
Old 02-02-09, 05:39 AM   #2
leigh123linux
Registered User
 
leigh123linux's Avatar
 
Join Date: Feb 2008
Posts: 163
Default Re: "NVRM: API mismatch" installing CUDA 2.1 on RHEL5

You have a stale nvidia.ko left over from a previous nvidia driver


Try

Code:
su
updatedb
locate nvidia.ko
Then delete the old nvidia.ko

Code:
su
rm -f  /lib/modules/`uname -r`/extra/nvidia.ko
You may need to change the path

Then run the installer again.
__________________
leigh123linux
leigh123linux is offline   Reply With Quote
Old 02-02-09, 06:37 AM   #3
Xilman
Registered User
 
Join Date: Feb 2009
Posts: 2
Default Re: "NVRM: API mismatch" installing CUDA 2.1 on RHEL5

Quote:
Originally Posted by leigh123@linux View Post
You have a stale nvidia.ko left over from a previous nvidia driver


Try

Code:
su
updatedb
locate nvidia.ko
Then delete the old nvidia.ko

Code:
su
rm -f  /lib/modules/`uname -r`/extra/nvidia.ko
You may need to change the path

Then run the installer again.
That worked a treat! Many thanks.

For the benefit of anyone else using a search engine to resolve the same problem, the nvidia.ko files were found in /lib/modules/2.6.18-128.el5/extra/nvidia.ko and /var/lib/dkms/nvidia/169.12/2.6.18-128.el5/x86_64/module/nvidia.ko

Paranoid chap that I am, I moved them somewhere safe rather than deleting them to allow for them to be restored if everything went terribly wrong.

Paul
Xilman 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
Need Help Installing NVIDIA Tesla M2070Q in Linux RHEL5 Ferianto85 NVIDIA Linux 0 05-18-12 08:35 PM

All times are GMT -5. The time now is 10:53 AM.


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