Fedora 16 nVidia Drivers Install Guide (disable nouveau driver)
This is guide, howto install nVidia proprietary drivers on Fedora 16 and disable Nouveau driver. This guide works with GeForce 6/7/8/9/200/300 series cards.
Fedora 16 nVidia driver installation is not much different from previous Fedora versions. I have tested this guide with a couple computers, so let me know, if you have some problems.
Before nVidia drivers installation
Check is your nVidia card supported
lspci |grep -i VGA ## Example output ## 01:00.0 VGA compatible controller: nVidia Corporation GT215 [GeForce GT 240] (rev a2)
List of Supported NVIDIA GPU Products, your card should found before 173.14.xx driver section.
NVIDIA Optimus Technology
If your lspci |grep -i VGA output looks like following:
00:02.0 VGA compatible controller: Intel Corporation 2n Generation Core Processor Family Integrated Graphics Controller (rev 09) 01:00.0 VGA compatible controller: nVidia Corporation GF106 [GeForce GT 555M SDDR3] (rev a1)
Or you know that your computer have NVIDIA Optimus Technology, and it is impossible to turn Intel Graphics / NVDIDIA Optimus off in the BIOS then try If !1 0 user Viger excellent NVIDIA Optimus (Bumblebee) guide for Fedora 16 users.
Install nVidia proprietary drivers on Fedora 16 “Verne” and disable the nouveau driver
1. Change root user
su - ## OR ## sudo -i
2. Make sure that you are running latest kernel and have latest SELinux policy packages
If not then update kernel and reboot
yum update kernel* selinux-policy* reboot
3. Add RPMFusion Repositories (Free and Non-Free)
32-bit and 64-bit
rpm -Uvh http://download1.rpmfusion.org/free/fedora/rpmfusion-free-release-stable.noarch.rpm rpm -Uvh http://download1.rpmfusion.org/nonfree/fedora/rpmfusion-nonfree-release-stable.noarch.rpm
4. Install nVidia proprietary drivers for GeForce 6/7/8/9/200/300/400/500 series cards
Select akmod, kmod or kmod-PAE from following.
akmod-nvidia
yum install akmod-nvidia xorg-x11-drv-nvidia-libsExtra package for kernel-PAE users
yum install kernel-PAE-develor
kmod-nvidia
yum install kmod-nvidia xorg-x11-drv-nvidia-libsor
kmod-nvidia-PAE and kernel-PAE-devel
yum install kernel-PAE-devel kmod-nvidia-PAEakmod is good option and easy way avoid problems on kernel updates and is best and only option if you use:
- self-compiled kernel
- older Fedora kernel
- quickly changing kernels from updates-testing/rawhide
Full spec of kmod and akmod differences, check this.
5. Remove / disable nouveau drivers from kernel initramfs
## Backup old initramfs nouveau image ## mv /boot/initramfs-$(uname -r).img /boot/initramfs-$(uname -r)-nouveau.img ## Create new initramfs image ## dracut /boot/initramfs-$(uname -r).img $(uname -r)
6. All is done and then just reboot
reboot
Troubleshooting
rdblacklist=nouveau nouveau.modeset=0 missing
Some users have reported that the rdblacklist=nouveau nouveau.modeset=0 might be missing on grub2 “linux” row, after NVIDIA drivers installation. Add rdblacklist=nouveau nouveau.modeset=0 to /etc/default/grub file.
## Modify following line ## GRUB_CMDLINE_LINUX="quiet rhgb" ## to ## GRUB_CMDLINE_LINUX="quiet rhgb rdblacklist=nouveau nouveau.modeset=0"
Then make grub2 config with grub2-mkconfig command:
grub2-mkconfig -o /boot/grub2/grub.cfg
Please let me know if you have some problems with nVidia drivers installation. You could also tell you if you got the drivers installed using this guide and what graphics card you have?
Related posts:
- Fedora 15 nVidia Drivers Install Guide (disable nouveau driver)
- Fedora 14 nVidia Drivers Install Guide (disable nouveau driver)
- Fedora 13 nVidia Drivers Install Guide (disable the nouveau driver)
- Fedora 13 ATI Catalyst (fglrx) 10.11 Drivers Install/Uninstall
- Fedora 13 ATI Drivers (Mesa 3D DRI Radeon)
458 Comments
Leave a Comment
Trackbacks/Pingbacks
Sponsors
Subscribe
Popular Posts
Install VirtualBox 4.1.14 on Fedora 16/15, CentOS/Red Hat (RHEL) 6.2/5.8
Posted by JR on 2-18-10
This worked perfectly with Fedora 16, 64-bit system and NVIDIA GeForce GT 240.
lspci |grep -i vga
01:00.0 VGA compatible controller: nVidia Corporation GT215 [GeForce GT 240] (rev a2)
Thanks!
Failure with build KMODS from AKMODS on reboot. System is AMD X4 1055t with nvidia gtx 550ti graphics. System no longer boots.
Hi jonb,
Could you get more specific error message, why akmods fail? You could also try to install kmod directly from RPMFusion repo…
Has akmod worked at all at any point?
Hi JR
Thanks for your interest. Messages go by quickly but say NVIDIA not loaded and drivers not found. Not loaded into kernel.
Fedora splash screen went from the bulb at center of screen before above procedure to blue bars loading at bottom of screen after.
I reverted to original uname img file (restored nouveau) and system splash screen reverted to blue bulb but would not boot anymore.
This is a new install to cope with the new video card, I might just reinstall to save time.
Original 64 bit fedora 16 install booted fine. A kernel update made the system not bootable but system still booted using the former kernel in grub.
Neither oneiric, mint or ultimate ubuntus will install with the nvidia gtx 550 present. Stuck using windows.
cheers
jonb
cheers
Hi again jonb,
Your nvidia gtx 550 card should be supported, but with this information it’s very hard/impossible to say what is wrong.
If you need further assistance with NVIDIA drivers then please post following information:
Then also post full /var/log/Xorg.0.log file (after boot with NVIDIA drivers), example to pastebin.
Hi JR,
Thanks for these instructions. I used them to set up my 32-bit Fedora 16 system with a GF 8400 GS card. I had to manually change grub2 to add “rdblacklist=nouveau nouveau.modeset=0″. All went smoothly, and the graphics work with Gnome 3.
However sometimes, parts of the screen do not redraw correctly. If I then move the window, it will redraw. Any idea why this happens?
This is the output from lspci |grep -i VGA
01:00.0 VGA compatible controller: nVidia Corporation GT218 [GeForce 8400 GS] (rev a2)
This is the output from uname -a
Linux laria 3.2.5-3.fc16.i686.PAE #1 SMP Thu Feb 9 02:02:37 UTC 2012 i686 i686 i386 GNU/Linux
The problems all appeared after upgrading from Fedora 15 to Fedora 16. Up until then, the default Fedora driver had been sufficient for me. However on moving to Fedora 16, the default driver would not allow me to run Gnome 3.
My upgrade to Fedora 16 was not clean – a problem (apparently known) in migrating to GRUB 2 – so it may be that I have a messed up installation, and will have to just reinstall from scratch.
Any advice much welcomed.
Hi Jeremy,
So your drivers seems to work, but that drawing problem is more difficult. One thing that came to my mind is that the your card is overheating with NVIDIA proprietary drivers. I have see sometimes overheating cause this style problems. Of course you can try to reinstall Fedora 16 from scratch, but if you use same NVIDIA proprietary drivers, then I guess that you will unfortunately see same problems.
Could you post output of following commands:
Does this problem exists just with some programs or with every programs?
Hi JR,
Thanks for the suggestion about overheating. One test I can do is to try to cool the system and see if the problem reduces. Looking at the thermal settings in nvidia-settings, it reports 63C, with the temperature bar being mostly green, with just one yellow bar showing.
The output from glxinfo |grep -i -e render -e opengl is:
direct rendering: Yes OpenGL vendor string: NVIDIA Corporation OpenGL renderer string: GeForce 8400GS/PCI/SSE2 OpenGL version string: 3.3.0 NVIDIA 290.10 OpenGL shading language version string: 3.30 NVIDIA via Cg compiler OpenGL extensions: GL_NV_conditional_render, GL_NV_copy_depth_to_color, GL_NV_copy_image, GL_NV_parameter_buffer_object2, GL_NV_path_rendering, GL_NVX_conditional_render, GL_NVX_gpu_memory_info, GL_OES_depth24, GL_OES_fbo_render_mipmap, GL_OES_get_program_binary, GL_OES_mapbuffer,The output from lsmod |grep -e nouv -e nvidia is:
The problem occurs with a number of programs. It is particularly obvious when playing aisleriot – the cards just don’t redraw on a deal for example. However I also see it when running test commands in xterm – sometimes several lines of output don’t draw, or when typing a command, a character does not properly redraw.
Thanks,
Jeremy
Your NVIDIA drivers are installed perfectly, direct rendering is on, OpenGL works and right modules are loaded so problem really might be overheating.
In fact, 63 C sounds a high temperature, of course it will depend on the card. Some of the cards are always hotter, but you could check from BIOS, if you could adjust your graphics card fan speed to higher?
Hi,
I followed the instructions and installed the driver. Now the system won’t boot. I’m not getting error messages, just that GUI is taking a long time to load. I take a photo of the screen if that helps.
I’m runing an Nvidia GForce 525M (for mobile) graphics card.
Thanks
Hi,
Does Geforce 525M use NVIDIA Optimus technology?
I’m not sure if anyone before me has had this problem:
My setup is a Fedora 16 KDE-spin with a self-compiled kernel (fc16-3.2.6-3) and my graphics card is an nvidia 9600M GT.
I followed the guide step by step, but in step 4 I only did
yum install akmod-nvidia xorg-x11-drv-nvidia-libsbecause that is how I understood the guide (is that correct?).
However, after the reboot in step 6 all I get is a screen, that when fully loaded looks like .
Can anyone help me or give advice?
Thanks in advance.
Somehow the link to my image got lost when posting.
I put it on:
imageshack[dot]us/photo/my-images/828/nvbootfail[dot]png
Hi Malhelo,
Yes you are installed right packages.
Could you try to boot Fedora 16 own kernel?
Change another virtual console and please post first output of following commands, so I get some background information:
First: I was not able to switch into another virtual console, because attempting to do so only showed a blinking cursor. Instead I started the machine in runlevel 3 and ran the commands.
I think you won’t need much of the output to see what is wrong, as there are a couple of errors in the Xorg.0.log
Still, here’s the output of all commands:
(WW) warning, (EE) error, (NI) not implemented, (??) unknown.
[ 84.209] (II) Loading extension MIT-SCREEN-SAVER
[ 84.211] (EE) Failed to load /usr/lib64/xorg/modules/extensions/nvidia/libglx.so: libnvidia-tls.so.290.10: cannot open shared object file: No such file or directory
[ 84.211] (EE) Failed to load module "glx" (loader failed, 7)
[ 84.245] (EE) NVIDIA: Failed to load the NVIDIA kernel module. Please check your
[ 84.245] (EE) NVIDIA: system's kernel log for additional error messages.
[ 84.245] (EE) Failed to load module "nvidia" (module-specific error, 0)
[ 84.245] (EE) No drivers available.
# RPM Fusion - nvidia-xorg.conf
#
Section "Device"
Identifier "Videocard0"
Driver "nvidia"
EndSection
Linux chaos 3.2.6-3.malhelo.fc16.x86_64 #1 SMP Sun Feb 19 21:50:20 CET 2012 x86_64 x86_64 x86_64 GNU/Linux
01:00.0 VGA compatible controller: nVidia Corporation G96 [GeForce 9600M GT] (rev a1)Interesting part is following error:
Could you post output of following commands:
Here’s the output. Thanks already for helping.
akmod-nvidia-290.10-1.fc16.4.x86_64nvidia-settings-1.0-13.fc16.x86_64
nvidia-xconfig-1.0-11.fc16.x86_64
xorg-x11-drv-nvidia-290.10-1.fc16.x86_64
xorg-x11-drv-nvidia-libs-290.10-1.fc16.x86_64
total 72900drwxr-xr-x. 3 root root 4096 Feb 21 17:20 .
dr-xr-xr-x. 114 root root 86016 Feb 21 17:10 ..
lrwxrwxrwx. 1 root root 17 Feb 21 17:20 libcuda.so -> libcuda.so.290.10
lrwxrwxrwx. 1 root root 17 Feb 21 17:20 libcuda.so.1 -> libcuda.so.290.10
-rwxr-xr-x. 1 root root 8509349 Nov 17 03:12 libcuda.so.290.10
lrwxrwxrwx. 1 root root 15 Feb 21 17:20 libGL.so.1 -> libGL.so.290.10
-rwxr-xr-x. 1 root root 1046608 Nov 17 03:01 libGL.so.290.10
lrwxrwxrwx. 1 root root 20 Feb 21 17:20 libnvcuvid.so.1 -> libnvcuvid.so.290.10
-rwxr-xr-x. 1 root root 2199008 Nov 17 03:49 libnvcuvid.so.290.10
lrwxrwxrwx. 1 root root 23 Feb 21 17:20 libnvidia-cfg.so.1 -> libnvidia-cfg.so.290.10
-rwxr-xr-x. 1 root root 136616 Nov 17 03:47 libnvidia-cfg.so.290.10
lrwxrwxrwx. 1 root root 28 Feb 21 17:20 libnvidia-compiler.so.1 -> libnvidia-compiler.so.290.10
-rwxr-xr-x. 1 root root 27685208 Nov 17 03:52 libnvidia-compiler.so.290.10
lrwxrwxrwx. 1 root root 26 Feb 21 17:20 libnvidia-glcore.so -> libnvidia-glcore.so.290.10
lrwxrwxrwx. 1 root root 26 Feb 21 17:20 libnvidia-glcore.so.1 -> libnvidia-glcore.so.290.10
-rwxr-xr-x. 1 root root 34523280 Nov 17 02:59 libnvidia-glcore.so.290.10
lrwxrwxrwx. 1 root root 22 Feb 21 17:20 libnvidia-ml.so.1 -> libnvidia-ml.so.290.10
-rwxr-xr-x. 1 root root 239184 Nov 17 03:48 libnvidia-ml.so.290.10
lrwxrwxrwx. 1 root root 23 Feb 21 17:20 libnvidia-tls.so.1 -> libnvidia-tls.so.290.10
-rwxr-xr-x. 1 root root 11000 Nov 17 03:06 libnvidia-tls.so.290.10
lrwxrwxrwx. 1 root root 18 Feb 21 17:20 libOpenCL.so.1 -> libOpenCL.so.1.0.0
-rwxr-xr-x. 1 root root 21296 Nov 17 03:12 libOpenCL.so.1.0.0
lrwxrwxrwx. 1 root root 23 Feb 21 10:59 libXvMCNVIDIA_dynamic.so.1 -> libXvMCNVIDIA.so.290.10
lrwxrwxrwx. 1 root root 23 Feb 21 17:20 libXvMCNVIDIA.so.1 -> libXvMCNVIDIA.so.290.10
-rwxr-xr-x. 1 root root 156632 Nov 17 02:45 libXvMCNVIDIA.so.290.10
drwxr-xr-x. 2 root root 4096 Feb 21 17:20 tls
total 20drwxr-xr-x. 2 root root 4096 Feb 21 17:20 .
drwxr-xr-x. 3 root root 4096 Feb 21 17:20 ..
lrwxrwxrwx. 1 root root 23 Feb 21 17:20 libnvidia-tls.so -> libnvidia-tls.so.290.10
lrwxrwxrwx. 1 root root 23 Feb 21 17:20 libnvidia-tls.so.1 -> libnvidia-tls.so.290.10
-rwxr-xr-x. 1 root root 11344 Nov 17 03:06 libnvidia-tls.so.290.10
All files found, but could you then post following output:
Something broken with libglx.so, because it can’t find libnvidia-tls.so.290.10 and libnvidia-glcore.so.290.10.
I think you should try to reainstall nvidia libraries, like:
Is it working then, or could you post ldd … command output?
xorg-x11-drv-nvidia-libs.i686 was not even installed yet.
So first I installed that and tried to (re)boot then. No success.
Then I reinstalled using your command & reboot. No success.
I thought it might have been an SELinux issue, because the logs looked odd to me. Also, all SELinux ever did, was make my life harder. So I disabled SELinux and rebooted. No success either.
Now I reverted the whole process as described in JR’s post on November 28, 2011.
I guess I’ll report back when I tried the whole process again.
Ok, first off: It’s working now.
The solution was pretty simple: I didn’t install the kernel-devel packages for my self-compiled kernel.
If anyone ever makes the same mistake, here’s how I found out / how to resolve it:
1. Boot without rhgb (basically, that means verbose). Then there was a “FAILED” entry telling to check
'systemctl status akmods.service'.2. Reboot into init 3, and execute that command. It then showed a message that it could not find the files it needs in
and asked if I had the correct kernel-devel packages installed.
3. Went to my home dir and the rpmbuild sub-folder where I had my kernel compiled and
Otherwise it will most probably skip the kernel, because it failed before.
First of all, excellent to hear that you got it working and thank you for the explanation and instructions, how you solve this problem! :) Awesome thanks!
I also had a similar problem, and I did what Malhelo did and it also worked for me.
yum install kernel-devel-$(uname -r)
/etc/rc.d/init.d/akmods start
maybe it should be added to the guide, or to the Troubleshooting section of the guide.
Thank you :)
Hi Arthur,
Did you have this same problem with self-compiled kernel or Fedora’s own kernel?
Fedora’s own kernel. (3.1.0-7.fc16.x86_64)
maybe executing “yum install kernel-devel-$(uname -r)” wasn’t needed, but I did it anyway, and then I ran “/etc/rc.d/init.d/akmods start” witch probably did the trick.
Okay. :) I have noticed something odd behavior with akmod-nvidia, maybe kmod-nvidia is currently better choice.
Great info.
Used it to fix my noisy fan on the GPU card.
Thanks
Not sure if anyone else has had this problem, but I have found that sometimes there is not enough kernel memory to load the nvidia kernel module. In this case the system just hangs when trying to load X even when all of the nvidia configuration is correct.
In this situation the xorg log (/var/log/Xorg.0.log) shows:
(EE) NVIDIA(0): Failed to initialize the NVIDIA GPU at PCI:1:0:0. Please
(EE) NVIDIA(0): check your systm's kernel log for additional error
(EE) NVIDIA(0): messages and refer to Chapter 8: Common Problems in the
(EE) NVIDIA(0): README for additional information.
(EE) NVIDIA(0): Failed to initialize the NVIDIA graphics device!
(EE) Screens(s) found, but none have a usable configuration.
I have a TV card installed on my machine and the combination of nvidia and that card seems to cause problems with kernel memory (this has been a problem on other systems too like my Thinkpad T410). I was able to get my machine to boot properly with nvidia by adding vmalloc=256M to end of my boot line:
To test if this is the problem edit your boot command from grub on boot, e.g.:
linux /vmlinu-3.2.7-1.fc16.i686 root=UUID=2f68b130-5fb1-4196-af45-6c44c4229863 ro rd.md=0 rd.lvm=0 rd.dm=0 KEYTABLE=us quiet SYSFONT=latarcyrheb-sun16 rhgb rd.luks=0 LANG=en_US.UTF-8 nouveau.modeset=0 rd.driver.blacklist=nouveau vmalloc=256M
To permanently add this parameter edit /etc/default/grub, e.g.:
GRUB_CMDLINE_LINUX="rd.md=0 rd.lvm=0 rd.dm=0 KEYTABLE=us quiet SYSFONT=latarcyrheb-sun16 rhgb rd.luks=0 LANG=en_US.UTF-8 nouveau.modeset=0 rd.driver.blacklist=nouveau vmalloc=256M"
Then save and re-generate /boot/grub2/grub.cfg:
# cp /boot/grub2/grub.cfg /boot/grub2/grub.cfg.old
# grub2-mkconfig -o /boot/grub2/grub.cfg
Looks like my attempt to add emphasis to the new parameter in the previous comment is making it invisible! The parameter you need to add is:
vmalloc=256M
Thanks Josh, this is good tip! :)
I updated your comment and make vmalloc=256M visible again… ;D
JR and others,
I am trying to install Fedora 16 on my Dell XPS 8300 (with a Nvidia GeoForce GT420) that’s already running Windows 7.
After a lot of reading up, I could get Fedora 16 to install using the basic video mode / troubleshooting mode. After the install, GNOME wont work and I get to see this basic desktop.
I also tried installing by hitting the TAB key and appending “nouveau.modeset=0 rdblacklist=nouveau text” before the install. No respite! I saw a black screen with some “nouveau” text and the install just stalls.
Then I went to Nvidia’s website where they have some drivers listed for Linux x86_64. The file is named ‘NVIDIA-Linux-x86_64-295.20.run’. Can I try installing thus driver after installing Fedora 16?
Thank you.
Suresh
Hi Suresh,
So can you boot to basic desktop?
Boot to runlevel 3 and get following information:
Hey JR,
Thank you for offering help.
Yes, I can boot to basic desktop. Here is the data you asked for.
[root@INFINITY ~]# grep “EE” /var/log/Xorg.0.log
(WW) warning, (EE) error, (NI) not implemented, (??) unknown.
[ 14.559] (II) Loading extension MIT-SCREEN-SAVER
[root@INFINITY ~]# cat /etc/X11/xorg.conf
Section “Device”
Identifier “Videocard0″
Driver “vesa”
EndSection
[root@INFINITY ~]# uname -a
Linux INFINITY 3.1.0-7.fc16.x86_64 #1 SMP Tue Nov 1 21:10:48 UTC 2011 x86_64 x86_64 x86_64 GNU/Linux
[root@INFINITY ~]# lspci | grep -i VGA
01:00.0 VGA compatible controller: nVidia Corporation Device 0de2 (rev a1)
Thank you.
Suresh
Thanks,
If you have this guide packages installed, then move/backup your xorg.conf with following command and reboot:
Do you get Gnome-Shell running or do you get some other problems?
Worked like a charm. Graphical bootloader gone though.
Many thanks for this.
lspci | grep -i VGA
01:00.0 VGA compatible controller: nVidia Corporation G98 [GeForce 9300M GS] (rev a1)
Hello Jr.
First off, I am very new to Linux. Thank you for the great guide. I am having the very same issues as mahalo but I do not have a custom kernel as I am too green to do that. My card is a GTX 460M with the latest 3.3.04.fc16.x86_64 kernel. I get the same “failed to start LSB: Builds and install new kmods from akmods packages…see systemctl status akmods.service” the next line says “failed to start LSB: Starts and stops login and scan… iSCSI devices…see systemctl status iSCSI.service ( I think I was getting this one before I tried installing the drivers so don’t think it’s relevant but included it just in case). Then the system continues to:
Starting Display manager…
Started Display Manager [ OK. ]
then it freezes…..I’ve let it sit for hours see if it doing something as the hard drive light flashes periodically to no avail. My question is do you think I should attemp mahalos fix and if so, I am a ill fuzzy in step 4?
Thanks in advance.
Hi Rico,
First you can boot to runlevel 3. Then start akmods with following command:
Then you can see more info, what is happening, do you get some errors?
Another way is try to update all packages with yum update command and try to install kmod-nvidia.
with no modification from your info, works perfectly under Fedora 17 alpha (Beta test-candidate 1). you Rock!
Hi deanO,
Excellent to hear that this works perfectly on Fedora 17. :)
You saved my night and a fresh install. Many thanks. Worked perfectly with my Quadro FX card.
Hi Amit,
Excellent, you are very welcome! :)
Many thanks! Drivers work as expected unlike nouveau ones.
[user@localhost ~]$ lspci | grep -i VGA
01:00.0 VGA compatible controller: nVidia Corporation G72M [Quadro NVS 110M/GeForce Go 7300] (rev a1)
[user@localhost ~]$
Hi skywalkie,
You are welcome! Nice to hear that you got nVidia drivers working! :)
I was looking for something which would make it easier to install Nvidia’s own drivers on my friends computer. Nvidia’s own script required too much hands on development knowledge for my friend. Video card is Asus GT440
uname -a
Linux SoupDuJour 3.3.0-4.fc16.i686 #1 SMP Tue Mar 20 18:45:14 UTC 2012 i686 i686 i386 GNU/Linux
After following your instructions and using akmod we just now have the Nividia splash screen appearing before the login screen. I’m assuming all went well. Certainly we have more available display resolutions than before.
Whether or not this upgrade to manufactures own drivers will deal with some of the intermittent corruptions of the background image on the login screen he is experiencing after using frame buffering devices, remains to be seen. We are switching between gdm3 and plasma to see which work space we like better. Its a bit of a conceptual switch – switchable full screen workspaces, rather than a cluttered “pinned” desktop, but so far so good.
Thanks for your guide, it was a great help and much easier than setting up a development box with the appropriate headers. My friend is new to linux and while I have been using Linux for many years as a desktop user I often don’t have the time to explore the edge anymore.
Regards
Russell
I am having a strange problem: I can get the nvidia drivers to work by following the instructions here, and the graphics display works fine, but there are no text consoles, just a flashing cursor. Any help would be appreciated.
Hi John,
I have heard same style no tty(s) problems with Ubuntu and nVidia, but this is the first time, when I hear the same problems with Fedora.
Let’s see what we can do, could you first post output of following command:
Sure. I added the last 2 in the last line:
————–
GRUB_TIMEOUT=5
GRUB_DISTRIBUTOR=”Fedora”
GRUB_DEFAULT=saved
GRUB_CMDLINE_LINUX=”rd.md=0 rd.lvm=0 rd.dm=0 KEYTABLE=us quiet SYSFONT=latarcyrheb-sun16 rhgb rd.luks=0 LANG=en_US.UTF-8 rdblacklist=nouveau nouveau.modeset=0″
A couple of other notes:
1) I tried switching the “rdblacklist=nouveau” to “rd.driver.blacklist=nouveau” (like one of the examples above, but the behavior is the same
2) I tried adding “vmalloc=256M” and again no change, but
interestingly:
3) If I do a Ctl-Alt-F2, then log in as if there was a text console, then do a “w” it shows that I have successfully logged in on tty2. So it’s not that the consoles are not active, it’s that I can’t see them.
Ok, I have a fix. Your clue about Ubuntu having the same problem set me searching in the right direction. From an Ubuntu forum I found that turning off Kernel Mode Setting entirely with “nomodeset” did the trick. Any idea what collateral damage I’m causing by doing this?
Excellent to hear that you got it working! :)
Kernel Mode Setting
Kernel Mode Setting is a feature that allows the kernel to switch resolution without reinitializing the graphics card/driver.
It is especially nice when using a framebuffer, which enables the kernel to display graphics on an otherwise text only console. This resolves the need for a separate framebuffer driver (such as the ones detailed in Framebuffer) to enable a high resolution console.
Currently, the Intel GMA, nouveau and radeon drivers support Kernel Mode Setting.
So it’s not very harmful to disable it… :)
01:00.0 VGA compatible controller: nVidia Corporation GF108 [Quadro NVS 4200M] (rev a1)
I have 3 monitors and want display on 2 of them (assuming the card only supports 2).
After I install the driver Gnome3 comes up fine on one monitor, and when I run nvidia-settings to configure the second monitor, everything looks great. However, when I save the xorg.conf to a new file and slam it over /etc/X11/xorg.conf, Gnome3 boots up in fallback mode. The following message is in the Xorg log:
[ 22.002] (II) LoadModule: “glx”
[ 22.002] (II) Loading /usr/lib64/xorg/modules/extensions/libglx.so
[ 22.002] (II) Module glx: vendor=”X.Org Foundation”
[ 22.002] compiled for 1.11.4, module version = 1.0.0
[ 22.002] ABI class: X.Org Server Extension, version 6.0
[ 22.002] (==) AIGLX enabled
[ 22.002] (II) Loading extension GLX
…
[ 22.014] (**) NVIDIA(0): Depth 24, (–) framebuffer bpp 32
[ 22.014] (==) NVIDIA(0): RGB weight 888
[ 22.014] (==) NVIDIA(0): Default visual is TrueColor
[ 22.014] (==) NVIDIA(0): Using gamma correction (1.0, 1.0, 1.0)
[ 22.014] (**) NVIDIA(0): Option “TwinView” “1″
[ 22.014] (**) NVIDIA(0): Option “MetaModes” “CRT: 1680x1050_60 +0+0, DFP-2: 1680x1050_60 +1680+0, DFP-0: NULL”
[ 22.014] (**) NVIDIA(0): Option “TwinViewXineramaInfoOrder” “CRT-0″
[ 22.014] (**) NVIDIA(0): Enabling 2D acceleration
[ 22.014] (EE) NVIDIA(0): Failed to initialize the GLX module; please check in your X
[ 22.014] (EE) NVIDIA(0): log file that the GLX module has been loaded in your X
[ 22.014] (EE) NVIDIA(0): server, and that the module is the NVIDIA GLX module. If
[ 22.014] (EE) NVIDIA(0): you continue to encounter problems, Please try
[ 22.014] (EE) NVIDIA(0): reinstalling the NVIDIA driver.
[ 23.256] (II) NVIDIA(GPU-0): Display (DELL 2208WFP (CRT-0)) does not support NVIDIA 3D
[ 23.256] (II) NVIDIA(GPU-0): Vision stereo.
[ 23.257] (II) NVIDIA(GPU-0): Display (Seiko/Epson (DFP-0)) does not support NVIDIA 3D
[ 23.257] (II) NVIDIA(GPU-0): Vision stereo.
[ 23.288] (II) NVIDIA(GPU-0): Display (DELL 2208WFP (DFP-2)) does not support NVIDIA 3D
[ 23.288] (II) NVIDIA(GPU-0): Vision stereo.
This is related to the following bug: https://bugzilla.redhat.com/show_bug.cgi?id=679227
A hack work-around:
Hey,
thanks for the installation guide! It works almost fine. The driver works perfectly but I get an error while booting:
Failed to start LSB: Builds and install new kmods from akmod packages [FAILED]
See ‘systemctl status akmods.service’ for details.
So:
$ systemctl status akmods.service
akmods.service – LSB: Builds and install new kmods from akmod packages
Loaded: loaded (/etc/rc.d/init.d/akmods)
Active: failed since Wed, 11 Apr 2012 16:02:48 +0200; 37s ago
Process: 894 ExecStart=/etc/rc.d/init.d/akmods start (code=exited, status=0/SUCCESS)
CGroup: name=systemd:/system/akmods.service
If I start akmods manually, it works perfect:
# /etc/rc.d/init.d/akmods start
Checking kmods exist for 3.3.1-3.fc16.x86_64 [ OK ]
Does anybody know how to fix this?
Thank you! :)
Hi Bob,
Do you still have same problem?
Do you have latest updates and Fedora’s original kernel?
I got a problem with loading the window myself. I can get the window to startup but after it comes up, just before the login screen, I get an error: “Oh no Something has gone wrong. A problem has occurred and the system can’t recover. Please contact a system adminstrator”.
In the system log: /var/log/messages, I find the following error: SYSTEM kernel: [ 26.376361] gnome-shell[1427]: segfault at 0 ip (null) sp 00007fffa64649e8 error 14 in gnome-shell[400000+5000]
I have Fedora 16 with all updates and nvidia gt430.
Please Help
Hi Andy,
Could you post output of following commands:
akmod-nvidia-295.33-2.fc16.x86_64
glibc-2.14.90-24.fc16.6.x86_64
glibc-common-2.14.90-24.fc16.6.x86_64
glibc-devel-2.14.90-24.fc16.6.x86_64
glibc-headers-2.14.90-24.fc16.6.x86_64
gnome-shell-3.2.2.1-1.fc16.x86_64
kmod-nvidia-3.3.1-3.fc16.x86_64-295.33-2.fc16.3.x86_64
kmod-nvidia-3.3.1-5.fc16.x86_64-295.33-2.fc16.x86_64
nvidia-graphics285.05.09-libs-285.05.09-139.fc16.x86_64
nvidia-settings-1.0-15.fc16.x86_64
nvidia-xconfig-1.0-13.fc16.x86_64
selinux-policy-3.10.0-80.fc16.noarch
selinux-policy-targeted-3.10.0-80.fc16.noarch
xorg-x11-drv-nvidia-295.33-3.fc16.x86_64
xorg-x11-drv-nvidia-libs-295.33-3.fc16.x86_64
system info:
Linux Wul 3.3.1-5.fc16.x86_64 #1 SMP Tue Apr 10 19:56:52 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux
nouveau:
nvidia 12281762 30
i2c_core 37991 2 nvidia,i2c_i801
Nothing comes back from the last command. even with sudo in front of it.
Do you have some reason why nvidia-graphics285.05.09-libs-285.05.09-139.fc16.x86_64 package is installed?
Could you also post output of following command:
I first tried to install that version of Nvidia from an rpm.
grep -i nvidia /var/log/Xorg.0.log |grep -i -e glx -e drive
[ 22.799] (II) Loading /usr/lib64/xorg/modules/extensions/nvidia/libglx.so
[ 23.288] (II) Module glx: vendor=”NVIDIA Corporation”
[ 23.288] (II) NVIDIA GLX Module 295.33 Sat Mar 17 15:16:17 PDT 2012
[ 23.294] (II) Loading /usr/lib64/xorg/modules/drivers/nvidia_drv.so
[ 23.295] (II) NVIDIA dlloader X Driver 295.33 Sat Mar 17 14:57:21 PDT 2012
[ 23.295] (II) NVIDIA Unified Driver for all Supported NVIDIA GPUs
[ 23.298] (II) Loading /usr/lib64/xorg/modules/drivers/nvidia_drv.so
Try to remove nvidia-graphics285.05.09-libs-285.05.09-139.fc16.x86_64 package just in case and run yum update. NVIDIA drivers is now updated to 295.40 version. Your problem actually sounds old glibc problem, but it’s fixed long time ago…
Hi JR,
I am trying to install nvidia on my desktop, 3.3.1-5.fc16.i686 fedora OS (32 bit).
After installing nVidia, loging screen comes for a moment and goes back to black screen with a desktop picture by giving a message
“oh there is something wrong contact your system admin”
Then I check that the following error while booting:
Failed to start LSB: Builds and install new kmods from akmod packages [FAILED]
See ‘systemctl status akmods.service’ for details.
So:
$ systemctl status akmods.service
akmods.service – LSB: Builds and install new kmods from akmod packages
Loaded: loaded (/etc/rc.d/init.d/akmods)
Active: failed since Wed, 11 Apr 2012 16:02:48 +0200; 37s ago
Process: 894 ExecStart=/etc/rc.d/init.d/akmods start (code=exited, status=0/SUCCESS)
CGroup: name=systemd:/system/akmods.service
If I start akmods manually, it works perfect:
# /etc/rc.d/init.d/akmods start
Checking kmods exist for 3.3.1-3.fc16.x86_64 [ OK ]
Does anybody know how to fix this?
Here is my system specification.
$ lspci |grep -i vga
00:0d.0 VGA compatible controller: nVidia Corporation C61 [GeForce 7025 / nForce 630a] (rev a2)
Hi swapan,
I actually see this same problem also with akmods.service.
/etc/rc.d/init.d/akmods start – works normally
systemctl start akmods.service – fails
Only difference is that the nVidia drivers work normally for me and also gdm and gnome-shell works (this is 64-bit system)…
Could you try to remove and reinstall whole gnome:
Build akdmods:
Then reboot. What happens then?
Hi JR,
By doing that I am getting same error while booting:
Failed to start LSB: Builds and install new kmods from akmod packages [FAILED]
See ‘systemctl status akmods.service’ for details.
If you ignore this error, is your Gnome working then?
Hi, thanks for your time.
I’ve just installed F16 and tried to install the nvidia drivers with your guide (i followed every step) but it didn’t work, after reboot it shows a message “oh! something went wrong …”
lspci |grep -i VGA00:0d.0 VGA compatible controller: nVidia Corporation C61 [GeForce 6150SE nForce 430] (rev a2)
rpm -qa *nvidia* *kernel*|sort;uname -r;lsmod |grep -e nvidia -e nouveau;cat /etc/X11/xorg.conf
abrt-addon-kerneloops-2.0.7-2.fc16.x86_64
kernel-3.1.0-7.fc16.x86_64
kernel-3.3.2-1.fc16.x86_64
kernel-devel-3.3.2-1.fc16.x86_64
kernel-headers-3.3.2-1.fc16.x86_64
libreport-plugin-kerneloops-2.0.8-4.fc16.x86_64
3.3.2-1.fc16.x86_64
nvidia 12319418 26
i2c_core 37991 2 nvidia,i2c_nforce2
# RPM Fusion – nvidia-xorg.conf
#
Section “Device”
Identifier “Videocard0″
Driver “nvidia”
EndSection
I’ve also tried adding the “rdblacklist=nouveau nouveau.modeset=0″ line in the /etc/default/grub file but nothing happened. Any ideas?
Thanks in advance.
Hi Fabián,
Some reason you are missing akmod-nvidia and/or kmod-nvidia packages? And however you have NVIDIA module loaded…
This worked alright for two nVidia gForce 6800 cards on Fedora 16 64bit
yum install akmod-nvidia xorg-x11-drv-nvidia-libs
was the install option that is fine, don’t forget the rest like generating nvidia-xconfig and rebuilding the initramfs, no PAE needed and no troubleshooting. Thanks. People here might like some links to change their resolution during boot for GRUB GRUB2 and kernel start.
http://harrison3001.blogspot.ca/2009/09/grub-2-graphical-boot-tips-to-set.html
http://superuser.com/questions/66428/how-can-i-change-console-shells-resolution-in-ubuntu-9-10
https://wiki.archlinux.org/index.php/GRUB2#Visual_Configuration
Ubuntu also lets you do a lot of fun things with the boot loader.
https://help.ubuntu.com/community/Grub2
Absolutely brilliant guide.
Worked like a charm: NVS Quadro 450 – 3 monitors – with FC16.
Thank you so much.
Thanks! This worked like a charm for an nVidia GeForce 6800 Ultra.
Worked perfect for me thankyou so much !!!
When I type the first command after installing the NVIDIA drivers
lspci | grep -i VGA
I see the following output
02:00.0 VGA compatible controller: nVidia Corporation Device 1201 (rev a1)
Can somebody point me to what’s going wrong?
Hi Praveen,
Could you post output of following commands (example to http://pastebin.com):
Could you also post more info about your computer? Do you know what nVidia card you should have on your computer?
lspci -k gives me
http://pastebin.com/LqkE0UJF#
lspci -v gives me
http://pastebin.com/kLWnJubG#
I have a Geforce GTX 560 card
I have never seen before just that lspci output, but your output maybe depends on manufacturer ZOTAC. So I guess that nVidia drivers should work normally if your card is GeForce GTX 560…
So try installation to see what happens, if you get some problems then just revert back to nouveau drivers. :)
So try installation to see what happens?
Does that mean I should install it from the nVidia website?
I have already gone through the installation steps. The problem is my monitor has a 1024×768 resolution. I want a higher resolution, but since the 560 drivers are not getting installed, it is pretty painful to look at the screen :-(
Oh, now I understood, I thought that you consider that the should you install drivers or not. :)
Please boot with installed nVidia driver and post output of following commands:
The thing is I have installed the drivers and this is booting after the drivers have been supposedly installed. Is there something wrong that I am doing?
glxinfo |grep -i -e render -e opengl
Output
Xlib: extension “GLX” missing on display “:0.0″.
Xlib: extension “GLX” missing on display “:0.0″.
Xlib: extension “GLX” missing on display “:0.0″.
Xlib: extension “GLX” missing on display “:0.0″.
Xlib: extension “GLX” missing on display “:0.0″.
Error: couldn’t find RGB GLX visual or fbconfig
Xlib: extension “GLX” missing on display “:0.0″.
Xlib: extension “GLX” missing on display “:0.0″.
Xlib: extension “GLX” missing on display “:0.0″.
Xlib: extension “GLX” missing on display “:0.0″.
Xlib: extension “GLX” missing on display “:0.0″.
lsmod |grep -e nouv -e nvidia
nouveau 580599 0
ttm 55029 1 nouveau
drm_kms_helper 26490 1 nouveau
drm 194532 3 nouveau,ttm,drm_kms_helper
i2c_algo_bit 4958 1 nouveau
i2c_core 25728 5 i2c_i801,nouveau,drm_kms_helper,drm,i2c_algo_bit
mxm_wmi 1727 1 nouveau
video 12388 1 nouveau
rpm -qa \*nvidia\* kernel\* | sort
kernel-3.1.0-7.fc16.x86_64
kernel-3.3.2-6.fc16.x86_64
kmod-nvidia-295.40-2.fc16.2.x86_64
kmod-nvidia-3.3.2-6.fc16.x86_64-295.40-2.fc16.2.x86_64
nvidia-settings-1.0-16.fc16.x86_64
nvidia-xconfig-1.0-14.fc16.x86_64
xorg-x11-drv-nvidia-295.40-1.fc16.x86_64
xorg-x11-drv-nvidia-libs-295.40-1.fc16.x86_64
uname -a
Linux praveen-fedora 3.1.0-7.fc16.x86_64 #1 SMP Tue Nov 1 21:10:48 UTC 2011 x86_64 x86_64 x86_64 GNU/Linux
I went back and reverted my drivers to nouveau and magically, my monitor displays high res data. Any idea what could have gone wrong?
Hi again Praveen,
Your problem with nVidia drivers looks simple. You was running kernel version 3.1.0-7.fc16.x86_64 and you had only kmod-nvidia for kernel 3.3.2-6.fc16.x86_64 installed. So that’s why it was not working.
You could use akmod-nvidia or boot with lastest kernel version 3.3.2-6.fc16.x86_64.
oh ok..Got it. But I guess for the time being I’ll stick on to the nouveau drivers. Will install the Nvidia drivers when I get the time.
In any case, thanks so much for the help.
Muito bom seu sitio sobre instalação de placas NVIDEA. Eu, pessoalmente esta quebrando muita a cabeça em instalar esse budega, rsss
Grato pela sua atenção! T+
Thanks to you, I now have my dual monitor system based on a Club 3D nVidia GeForce 9500GT board and a nVidia NV4 Riva TNT board, running on Fedora 16.
On first installing Fedora 16, one of the screens displayed X, but the interface froze at irregular, but short intervals. I faithfully followed your instructions f
or installing the nvidia drivers on Fedora 16, including the additions to grub2/grub.conf. On reboot to runlevel 3, the nouveau driver was not loaded (great
news!), but executing “startx” was unsuccessful. I inspected the Xorg.0.log and noticed 2 things:-
First, depressingly, the error message “The NVIDIA RIVA TNT GPU installed in this system is supported through the NVIDIA 71.86.xx Legacy drivers. Please vis
it……….”, which I assumed wrongly meant that the installation of the nvidia drivers had failed.
Secondly, I noticed that an xorg.conf had been found, which referred to a “videocard”, when there was no such file before I carried out your instructions.
This xorg.conf identified itself as coming from rpmfusion and was particularly short. I replaced this with an xorg.conf referencing the nvidia drivers that I had constructed for my set up, ran startx again, and wonder of wonders (and great joy!) I had a dual monitor system up and running: the error message in the log file was a red-herring.
Anyway, thank you for the very clear instructions!