I am new to linux and am trying to get this NVIDIA driver to work so that I can use the 3D acceleration for my games:
I am running SuSE 8.2 Pro with linux kernal 2.4.20.SuSE. I agree to the lic. and then it tells me that NVIDIA's ftp might have the kernal-source, but it comes back saying that it does not. So it then says it will search for the kernal-source RPM that I need on my machine. Well I have it the 2.4.20.SuSE RPM straight from SuSE! It begins to run something (indicated by a progress bar) but then an error occurs. The program finishes and informs me that additional information can be found in the log file pertaining to the error. So I checked that out too. It says that the compiler I used for my current version is not the same compiler used for the kernal-source.
Does anyone know what I need to do to get this to run? It has some
but it recommends that you only use them if you know what your doing, which I don't.:confused:
Easy one :)
# export IGNORE_CC_MISMATCH=y
Then do it again.
Lets pretend that I dont know what that means, which I do. If someone is new to linux, such as myself, how is a random command line going to help them? Do I need to edit the installer, a config file, add this to the term line? I have tried running via .run and the nvidia-installer using the "assumed" command line option that you gave me with no luck. Please be a little more specific as to what;
# export IGNORE_CC_MISMATCH=y
means and what I need to do with this random argument.
Before running the nvidia installer from your command prompt, type and execute the following command line:
This is what he meant.
Details: when the installer will try to compile the nvidia kernel gateway module it will compare the version of the compiler you are using to compile it with the version of the compiler that was used to compile your kernel. It does happen quite a lot with mainstream distros that the compilers are different.
But the compiling process of the nvidia kernel module stops if they are not the same... so this environment variable(i.e. IGNORE_CC_MISMATCH) sets to "y" will bypass the test to allow you to compile the module even if the compiler versions do not match.
Now that I understand!
However I did run into something very weird in my attempts to fix this problem. I did not have a .config file for my 2.4.20-4GB-athlon KERNAL. Does that make any sense to anyone? Does this mean that the KERNAL was copied to my machine via YaST's update, but never compiled under my machine? I figure that this information is not really relevant to NVIDIA Graphics in Linux but found it odd. However this is the only Linux forum that I have had luck with.
Thanks again for the help, this is making my move to Linux much easier!
Binary package and Compiling the source
Mainstream linux distros use most of the time(meaning not all the time) a binary package system. Namely when you install a new or updated binary package(RPM for Redhat and mandrake, SuSE package, DEB for debian....) on your system, everything is already compiled for you. They did the compiling on their system, not yours... So in your case, yes, YaST probably updated your kernel. But then, the pre-compiled nvidia kernel gateway modules provided with the nvidia driver became useless...
Hopefully, all that may be improved with kernel module versioning in the brand new 2.6 kernel...
|All times are GMT -5. The time now is 11:05 AM.|
Powered by vBulletin® Version 3.7.1
Copyright ©2000 - 2015, Jelsoft Enterprises Ltd.
Copyright ©1998 - 2014, nV News.