On Sun, 2004-03-14 at 08:54, Pierre De Boeck wrote: > Ok Mark, > > I joined the file in attachement. Note that > I took the one in /etc/X11 and not the one in /etc. > > And here are my "progress": > > As no setting was working, I have reinstall everything with > Dell 1024*780 LCD and it was not working either. > > But I used the rescue mode to edit the inittab and now > I can boot in level 3 with an NIC working. :-) > > I also used wget to download the last driver > (NVIDIA-Linux-x86-1.0-5336-pkg1.run) from nvidia > and executed the sh script. > > Then it complained about some compiler mismatch version. > > I joined the log file also. > > Anyway, thanks a lot for all your help. I already installed > RH9 on our servers, configured apache, sendmail,.. and I never > had such problems. But obviously our servers hardware were more > common I suppose than the one found on my laptop. Ok...There are several things that you need to do: First, the issue of the compiler errors with NVidia is well known and has been discussed on this list at length. As I understand unfortunately, the archives are down right now, which would preclude a direct search. This issue stems from the fact that the FC kernel is compiled with a different version of gcc than the rest of the system. Thus when compiling kernel modules you get the error, since the "system" gcc is used by default. There are multiple workarounds for this, but the one that I use is the following TWO steps: ./NVIDIA-Linux-x86-1.0-5336-pkg1.run --add-this-kernel ./NVIDIA-Linux-x86-1.0-5336-pkg1-custom.run The first command creates the second file, which you can then run. This will get around the compiler error. You will need to REDO the above, any time you update the kernel. Once you then have the NVidia module installed, you need to make the following changes in your XF86Config file, which are documented in the NVidia README file at: ftp://download.nvidia.com/XFree86/Linux-x86/1.0-5336/README 1. In the module section: Section "Module" Load "dbe" Load "extmod" Load "fbdevhw" Load "glx" Load "record" Load "freetype" Load "type1" Load "dri" EndSection You need to remove the line: Load "dri" 2. In the device section: Section "Device" Identifier "Videocard0" Driver "nv" VendorName "Videocard vendor" BoardName "NVIDIA GeForce FX (generic)" EndSection Replace the line: Driver "nv" With: Driver "nvidia" 3. In the Monitor section, change: Section "Monitor" Identifier "Monitor0" VendorName "Monitor Vendor" ModelName "0" HorizSync 31.5 - 48.5 VertRefresh 59.0 - 75.0 Option "dpms" EndSection To: Section "Monitor" Identifier "Monitor0" VendorName "Monitor Vendor" ModelName "Dell 1600x1200 Laptop Display Panel" HorizSync 31.5 - 90.0 VertRefresh 59.0 - 85.0 Option "dpms" EndSection Note the different Sync/Refresh settings and Name. 4. In the Display section, change: Section "Screen" Identifier "Screen0" Device "Videocard0" Monitor "Monitor0" DefaultDepth 24 SubSection "Display" Depth 24 Modes "1024x768" "800x600" "640x480" EndSubSection To: Section "Screen" Identifier "Screen0" Device "Videocard0" Monitor "Monitor0" DefaultDepth 24 SubSection "Display" Depth 24 Modes "1600x1200" "1280x1024" "1024x768" "800x600" "640x480" EndSubSection EndSection Note the addition of "1600x1200" and "1280x1024" to the list of resolutions. 5. In order to get the 3D screen savers to work, you need to remove the following RPM: XFree86-Mesa-libGL-4.3.0-55 Using: rpm -e --nodeps XFree86-Mesa-libGL If you have not yet done any system updates, the version of the RPM you have installed may be different than what I have above. This is something that you will need to do again when there are XFree86 updates. In order to avoid conflicts during up2date, what I have done is to use up2date to include the above in the RPM database, but not to actually install the RPM using: up2date --justdb XFree86-Mesa-libGL The above should get you going. HTH, Marc Schwartz
