Nvidia 177.80 for F9

[blitzo]

I just installed F9 on a new machine and I am having few problems. I have a Nvidia GTX 280, and it looks like hardware support was added in driver version 177 +. The latest version of the driver in Livna is 173. Can someone recommend what I should do? I am thinking of trying the development repo in Livna, it seems this is geared towards F10 so would it be a good idea to use it for the latest nvidia driver? I also could use the official Nvidia driver, but I know it has some issues and Livna builds are recommended. What should I do?

[Hlingler]

You could try downloading the Livna development SRPM(s) (*.src.rpm) and re-build against F9. Would require all the usual build tools, libs, and *-devel packages, etc., etc., etc.

Only other option is to use the NVidia binary installer (not recommended). If you do so, remove all Livna nvidia driver packages first (otherwise you'll have TWO conflicting copies of the driver scattered all over the OS), and that will be a real mess....

V

[leigh123linux]

You could rebuild my SRPM

http://www.dfm.uninsubria.it/compiz/fusion-testing/9/nvidia/SRPM/nvidia-x11-drv-177.80-1.fc9.src.rpm



[root@localhost Desktop]# rpm -qa |grep nvidia
nvidia-x11-drv-32bit-177.80-1.fc9.x86_64
nvidia-x11-drv-177.80-1.fc9.x86_64
[root@localhost Desktop]#


su
yum groupinstall "Development Tools"
yum install rpmdevtools wget
wget http://www.dfm.uninsubria.it/compiz/fusion-testing/9/nvidia/SRPM/nvidia-x11-drv-177.80-1.fc9.src.rpm
yum-builddep nvidia-x11-drv-177.80-1.fc9.src.rpm
exit
rpmdev-setuptree
rpmbuild --rebuild nvidia-x11-drv-177.80-1.fc9.src.rpm

Then install the rpm files.

[leigh123linux]

You could try downloading the Livna development SRPM(s) (*.src.rpm) and re-build against F9. Would require all the usual build tools, libs, and *-devel packages, etc., etc., etc.

Only other option is to use the NVidia binary installer (not recommended). If you do so, remove all Livna nvidia driver packages first (otherwise you'll have TWO conflicting copies of the driver scattered all over the OS), and that will be a real mess....

V

It wont work as the F10 kmod is patched and fails to build ( I believe ) on the 2.6.26 kernel

[Hlingler]

"Ask, and ye shall receive...." :D

That's even better than the Livna stuff. May grab and try it myself.... Thanks, Leigh.

V

[blitzo]

Thanks!

I will try and rebuild that src file, I will let you know how it goes.

[blitzo]

When I install the nvidia-x11-drv rpm, will everything be automatic or do I need tom manually configure X?

I am having alot of issues with Fedora 9 so I am going to give 8 a try.

[leigh123linux]

When I install the nvidia-x11-drv rpm, will everything be automatic or do I need tom manually configure X?

I am having alot of issues with Fedora 9 so I am going to give 8 a try.

I think it does it for you, if not run

su
/usr/sbin/nvidia-xconfig

[blitzo]

Yeah something didn't go right. When I run nvidia-xconfig it tells me that I am not using the nvidia driver.

[blitzo]

I needed to reboot after nvidia-xconfig, at boot it started to make the nvidia module, but it failed. Is there a log of the boot console? dmesg isn't listing anything about nvidia.

[leigh123linux]

I needed to reboot after nvidia-xconfig, at boot it started to make the nvidia module, but it failed. Is there a log of the boot console? dmesg isn't listing anything about nvidia.




There is a build log in

/var/lib/dkms/nvidia/kernel-2.6.26.****

[blitzo]

There is no dkms folder.

Do you think I will be OK to rebuild the src.rpm in Fedora 8?

[devjonfos]

Hello All,

I'm on F9 and I've followed the steps to build the RPM. I have the RPM at /root/rpmbuild/RPMS/i386. I can install the RPM using rpm --install /root/rpmbuild/RPMS/i386/nvidia-x11-drv-177.80-1.fc9.i386.rpm. When I reboot the console flashes several times and I'm stuck at the console prompt.

I login with root and the X log says, "No devices detected.". I then do an rpm --erase on that package and then reboot to get the default driver and X comes up.

I have two 8800 GT's in SLI and I've tried the Livna steps with the same results. Actually the Livna steps came first, then I tried the steps in this thread after uninstalling the Livna related packages and doing an rpm -qa | grep nvidia to make sure there were no conflicts with other drivers.

Help.

Regards,

Jon Foster

[devjonfos]

I did some more reading in the Xorg log:

(!!) More than one possible primary device found

Which makes sense for SLI, but how do you specify the primary device in your xorg.conf when you have two video cards?

[devjonfos]

Okay, I've got the Livna NVIDIA driver (173) to work. The solution was to specify the BusID as this poster found out: http://forums.fedoraforum.org/showthread.php?t=196539

Section "Device"
Identifier "Videocard0"
Driver "nvidia"
BusID "PCI:1:0:0"
Option "AddARGBGLXVisuals" "True"
EndSection

Section "Device"
Identifier "Videocard1"
Driver "nvidia"
BusID "PCI:2:0:0"
Option "AddARGBGLXVisuals" "True"
EndSection

Section "Screen"
Identifier "Screen0"
Device "Videocard0"
Monitor "Monitor0"
DefaultDepth 24
Option "AddARGBGLXVisuals" "True"
Option "MultiGPU" "on"
Option "SLI" "on"
SubSection "Display"
Viewport 0 0
Depth 24
EndSubSection
EndSection

Now rpm -qa | grep nvidia reports:

[root@localhost ~]# rpm --query --all | grep nvidia
kmod-nvidia-173.14.12-5.lvn9.i686
xorg-x11-drv-nvidia-173.14.12-1.lvn9.i386
xorg-x11-drv-nvidia-libs-173.14.12-1.lvn9.i386
kmod-nvidia-2.6.26.5-45.fc9.i686-173.14.12-5.lvn9.i686
[root@localhost ~]#

So how do I go from the 173 driver to the 178 driver in the RPM that I built?

Do I just yum remove "kmod-nvidia" and then install that RPM build?

[blitzo]

I did a reinstall and was able to get alittle further with the nvidia-x11-drv src.rpm. The nvidia driver is failing to build at boot. I tracked the error in the dkms nvidia make.log.

include/asm/compat.h: In function ‘compat_alloc_user_space’:
include/asm/compat.h:210: warning: pointer of type ‘void *’ used in arithmetic
/var/lib/dkms/nvidia/177.80-1.fc9/build/nv.c: In function ‘__nv_setup_pat_entries’:
/var/lib/dkms/nvidia/177.80-1.fc9/build/nv.c:934: internal compiler error: Segmentation fault
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://bugzilla.redhat.com/bugzilla> for instructions.
Preprocessed source stored into /tmp/ccSGzBQ3.out file, please attach this to your bugreport.
make[3]: *** [/var/lib/dkms/nvidia/177.80-1.fc9/build/nv.o] Error 1
make[2]: *** [_module_/var/lib/dkms/nvidia/177.80-1.fc9/build] Error 2
NVIDIA: left KBUILD.
nvidia.ko failed to build!
make[1]: *** [module] Error 1
make: *** [module] Error 2

Is this an issue with my system or something wrong with the package?

[Hlingler]

Okay, I've got the Livna NVIDIA driver (173) to work. The solution was to specify the BusID as this poster found out: http://forums.fedoraforum.org/showthread.php?t=196539

Now rpm -qa | grep nvidia reports:

[root@localhost ~]# rpm --query --all | grep nvidia
kmod-nvidia-173.14.12-5.lvn9.i686
xorg-x11-drv-nvidia-173.14.12-1.lvn9.i386
xorg-x11-drv-nvidia-libs-173.14.12-1.lvn9.i386
kmod-nvidia-2.6.26.5-45.fc9.i686-173.14.12-5.lvn9.i686
[root@localhost ~]#

So how do I go from the 173 driver to the 178 driver in the RPM that I built?

Do I just yum remove "kmod-nvidia" and then install that RPM build?I'm not sure why you would want to goof with this if it's currently working. If you do decide to try, I suggest backing up xorg.conf first. Since the two driver RPM formats are different, it's not as simple as updating/upgrading. So:
yum remove livna-config-display `rpm --query --all | grep -i nvidia`
rpm -ivh /<PATH-to-new-RPM>nvidia-x11-drv*

Good Luck,
V

[Hlingler]

I did a reinstall and was able to get alittle further with the nvidia-x11-drv src.rpm. The nvidia driver is failing to build at boot. I tracked the error in the dkms nvidia make.log.

include/asm/compat.h: In function ‘compat_alloc_user_space’:
include/asm/compat.h:210: warning: pointer of type ‘void *’ used in arithmetic
/var/lib/dkms/nvidia/177.80-1.fc9/build/nv.c: In function ‘__nv_setup_pat_entries’:
/var/lib/dkms/nvidia/177.80-1.fc9/build/nv.c:934: internal compiler error: Segmentation fault
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://bugzilla.redhat.com/bugzilla> for instructions.
Preprocessed source stored into /tmp/ccSGzBQ3.out file, please attach this to your bugreport.
make[3]: *** [/var/lib/dkms/nvidia/177.80-1.fc9/build/nv.o] Error 1
make[2]: *** [_module_/var/lib/dkms/nvidia/177.80-1.fc9/build] Error 2
NVIDIA: left KBUILD.
nvidia.ko failed to build!
make[1]: *** [module] Error 1
make: *** [module] Error 2

Is this an issue with my system or something wrong with the package?This appears to be an error in the SRPM. Since Leigh built this SRPM, you just "filed a bug report". :D

V

P.S. I got the SRPM, so I'll try the build too and see if I get the same error. Also, I assume you did the usual RPM rebuild stuff ? As in:
rpmbuild -bb nvidia-x11-drv.spec

or similar ?

EDIT: No, Leigh's SRPM rebuilt fine here![vince@localhost Fri Oct 10 02:04:37 ~/rpmbuild/SRPMS]$ rpmbuild --rebuild nvidia-x11-drv-177.80-1.fc9.src.rpm
Installing nvidia-x11-drv-177.80-1.fc9.src.rpm
[...]
Wrote: /home/vince/rpmbuild/RPMS/i386/nvidia-x11-drv-177.80-1.fc9.i386.rpm
Executing(%clean): /bin/sh -e /var/tmp/rpm-tmp.99170
+ umask 022
+ cd /home/vince/rpmbuild/BUILD
+ cd nvidia-x11-drv-177.80
+ /bin/rm -rf /var/tmp/nvidia-x11-drv-177.80-1.fc9-root
+ exit 0
Executing(--clean): /bin/sh -e /var/tmp/rpm-tmp.99170
+ umask 022
+ cd /home/vince/rpmbuild/BUILD
+ rm -rf nvidia-x11-drv-177.80
+ exit 0
[vince@localhost Fri Oct 10 02:07:10 ~/rpmbuild/SRPMS]$ ls -lha ../RPMS/i386/
total 13M
drwxr-xr-x 2 vince vince 4.0K 2008-10-10 02:07 .
drwxrwxr-x 3 vince vince 4.0K 2008-08-09 21:23 ..
-rw-r--r-- 1 vince vince 12M 2008-10-10 02:07 nvidia-x11-drv-177.80-1.fc9.i386.rpm
-rw-r--r-- 1 vince vince 193K 2008-10-06 07:58 qps-1.10.3-1.fc9.i386.rpm
-rw-r--r-- 1 vince vince 40K 2008-10-06 07:58 qps-debuginfo-1.10.3-1.fc9.i386.rpm
[vince@localhost Fri Oct 10 02:10:55 ~/rpmbuild/SRPMS]$So Blitzo: you need to tell us what you did, because it looks good here.

EDIT #2: For the record, the nvidia.ko kmod built and loaded fine too. There was some brief goofiness with the keyboard, but a logout/login fixed that.

[blitzo]

I built the src.rpm with no problems the same way you did. I will try doing it again and see what happens.

[devjonfos]

I'm not sure why you would want to goof with this if it's currently working...it's not as simple as updating/upgrading...

You're right! I wanted a working NVIDIA driver so I could load and run FlightGear. Since the Livna NVIDIA Driver (173) gives me a working graphics system that I can use with FlightGear, I've reached my "happy medium", so to speak.

The only reason to upgrade to 178 was to get the latest. But I see from your post that the commands are complicated and there is potential for something to go wrong. So I'll stay with my "happy medium" at 173. I was wondering if you might know (estimate) when Livna would be updated to the NVIDIA 178 driver?

[Hlingler]

The only reason to upgrade to 178 was to get the latest. But I see from your post that the commands are complicated and there is potential for something to go wrong. So I'll stay with my "happy medium" at 173. I was wondering if you might know (estimate) when Livna would be updated to the NVIDIA 178 driver?Yeah, I know - I want the "latest and greatest" too - usually. But (painful) experience has taught me some caution with respect to the NVidia diver: unnecessary updates can break as much as they fix. So, I usually look at the NVidia web site to see what changes have been made, and if it's really necessary and worth the update. In this case, the answer is: yes, very probably so. There are an unusually large number of bug fixes and such since 173.14.x.

I have no idea when Livna may have an update available - typically, it's been within a couple days to a week, but other factors (merge with RPMFusion, switch of build system, etc.) have delayed things recently. In any case, Betas are traditionally NOT supported. But REV. 177.80.x went "official" three days ago (07 Oct 2008), so I expect an update momentarily.

HTH,
V

P.S. @Blitzo: Sorry, no can do on x86_64 here. Maybe Leigh or someone else can help...?

[blitzo]

leigh123 provided me with some x86_64 rpms. After I installed them I got the same error. It must be an issue with my system. I am going to try a few things and see what I can accomplish. Let me know if you have any suggestions.

[blitzo]

The error I am getting is related to some kind of system problem. I am having a ton of headache on this machine with Fedora 9. I tried the Nvidia provided driver and I received the same error. I then installed Ubuntu and compiled the driver with no issues and it is working fine. I don't know what I should do about this Fedora install.

[John5342]

The error I am getting is related to some kind of system problem. I am having a ton of headache on this machine with Fedora 9. I tried the Nvidia provided driver and I received the same error. I then installed Ubuntu and compiled the driver with no issues and it is working fine. I don't know what I should do about this Fedora install.

I have seen a few of the symptoms you are having before and a couple of possibilities spring to mind (there are plenty of others. these are ones that spring to mind):

I have seen similar segmentation faults and drivers failing (even when compiled correctly) countless times as a result of memory errors. (easily checked with memtest86 which i believe is on the install cd as a boot option)

You may also have stumbled on a possible X server compatibility problem. Fedora uses the latest 1.5. Ubuntu on the other hand last time i checked still uses the previous stable version. Although that wouldnt explain the segmentation fault during compile.

[blitzo]

I have seen similar segmentation faults and drivers failing (even when compiled correctly) countless times as a result of memory errors. (easily checked with memtest86 which i believe is on the install cd as a boot option)

You may also have stumbled on a possible X server compatibility problem. Fedora uses the latest 1.5. Ubuntu on the other hand last time i checked still uses the previous stable version. Although that wouldnt explain the segmentation fault during compile.

I ran a memtest it it is showing no errors. I am not receiving and segmentation faults in Ubuntu.

Yeah I doubt it is a X compatibility problem as I can't compile. It must be some kind of error in a library or dev tool.

[leigh123linux]

I did a reinstall and was able to get alittle further with the nvidia-x11-drv src.rpm. The nvidia driver is failing to build at boot. I tracked the error in the dkms nvidia make.log.

include/asm/compat.h: In function ‘compat_alloc_user_space’:
include/asm/compat.h:210: warning: pointer of type ‘void *’ used in arithmetic
/var/lib/dkms/nvidia/177.80-1.fc9/build/nv.c: In function ‘__nv_setup_pat_entries’:
/var/lib/dkms/nvidia/177.80-1.fc9/build/nv.c:934: internal compiler error: Segmentation fault
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://bugzilla.redhat.com/bugzilla> for instructions.
Preprocessed source stored into /tmp/ccSGzBQ3.out file, please attach this to your bugreport.
make[3]: *** [/var/lib/dkms/nvidia/177.80-1.fc9/build/nv.o] Error 1
make[2]: *** [_module_/var/lib/dkms/nvidia/177.80-1.fc9/build] Error 2
NVIDIA: left KBUILD.
nvidia.ko failed to build!
make[1]: *** [module] Error 1
make: *** [module] Error 2

Is this an issue with my system or something wrong with the package?


Can you post this file /tmp/ccSGzBQ3

[leigh123linux]

I have seen a few of the symptoms you are having before and a couple of possibilities spring to mind (there are plenty of others. these are ones that spring to mind):

I have seen similar segmentation faults and drivers failing (even when compiled correctly) countless times as a result of memory errors. (easily checked with memtest86 which i believe is on the install cd as a boot option)

You may also have stumbled on a possible X server compatibility problem. Fedora uses the latest 1.5. Ubuntu on the other hand last time i checked still uses the previous stable version. Although that wouldnt explain the segmentation fault during compile.

I doubt it is a X server issue !

[leigh05@localhost ~]$ rpm -q fedora-release
fedora-release-9-5.transition.noarch
[leigh05@localhost ~]$ uname -r
2.6.26.6-67.fc9.x86_64
[leigh05@localhost ~]$ rpm -qa |grep nvidia
nvidia-x11-drv-32bit-177.80-1.fc9.x86_64
nvidia-x11-drv-177.80-1.fc9.x86_64
[leigh05@localhost ~]$ rpm -q xorg-x11-server-Xorg
xorg-x11-server-Xorg-1.5.2-2.fc9.x86_64
[leigh05@localhost ~]$

[blitzo]

You are going to hate me.


It was the RAM.


I don't fully understand why with the symptoms I had. I ran Memtest for a couple hours and RAm checked out fine. I then installed Ubuntu and was able to compile the Nvidia drivers and experienced no problems. After several hours of gaming firefox 3 started to get random pixel scatters, I first blamed this on firefox 3 and dropped down to firefox 2. Firefox 2 ran fine, well for a few hours, I then experience the same pixel scatters. So I installed Fedora 9 i686, and everything ran fine for a little while, then I started to experience similar problems. I knew it had to be hardware related at this point, and more specially it had to be ram. So running my system and default settings and timings, I decided to under clock my RAM from the manufacturer's specs . I then reinstalled Fedora and everything is running great, I compiled Nvidia drivers and have no other issues. I guess I really don't understand why it ran fine in Ubuntu.

I guess I should have looked at this a little sooner, but everything pointed to a software error. I wouldn't declare it a full waste of time though. In my research I got a little more in-depth with Fedora and Linux in general. Learned a lot about RPM development, got to try a different popular Linux distro and had the Fedora community provide some extra help to me on this issue. I owe you all one.

Thanks!!

[Hlingler]

To be perfectly honest, Blitzo, I probably would have called the RAM good after two hours without error on memtest86+ too. Clearly, we all learn here that memtest86+ must be run overnight for a really thorough test. And I'm going to guess that the differences between Fedora and Ubungtu probably are the result of the different kernel versions, and how they handle RAM.

Another mystery solved. :)

V

[Hlingler]

<somewhat-off-topic>
@Leigh:

Your SRPM rebuilt just fine here on F8 - except that once again, I noticed (by accident) that *libwfb.so is missing. I checked the spec file and all, and it appears to be included, so not sure why it's not included in the rebuilt RPM. Doesn't appear to be important/show-stopper, but odd that it's missing. Thoughts? </somewhat-off-topic>

Thanks and Regards,
V

[leigh123linux]

<somewhat-off-topic>
@Leigh:

Your SRPM rebuilt just fine here on F8 - except that once again, I noticed (by accident) that *libwfb.so is missing. I checked the spec file and all, and it appears to be included, so not sure why it's not included in the rebuilt RPM. Doesn't appear to be important/show-stopper, but odd that it's missing. Thoughts? </somewhat-off-topic>

Thanks and Regards,
V


libwfb.so is commented out in the F9 spec as the xorg server provides the lib.



# Install X driver and extension (is the nvidia_drv.o useful?)
%{__mkdir_p} %{buildroot}%{_libdir}/xorg/modules/drivers/
%{__install} -p -m 0755 usr/X11R6/lib/modules/drivers/nvidia_drv.so \
%{buildroot}%{_libdir}/xorg/modules/drivers/
%{__mkdir_p} %{buildroot}%{_libdir}/xorg/modules/extensions/nvidia/
%{__install} -p -m 0755 usr/X11R6/lib/modules/extensions/libglx.so.%{version} \
%{buildroot}%{_libdir}/xorg/modules/extensions/nvidia/libglx.so
#{__install} -p -m 0755 usr/X11R6/lib/modules/libnvidia-wfb.so.%{version} \
# %{buildroot}%{_libdir}/xorg/modules/libwfb.so




%files
%defattr(-,root,root,-)
%doc LICENSE usr/share/doc/*
# Kernel and dkms related bits
%config %{_sysconfdir}/modprobe.d/nvidia
%{_usrsrc}/%{dkms_name}-%{dkms_vers}/
# udev "configuration"
%config %{_sysconfdir}/udev/makedev.d/60-nvidia.nodes
# Libraries and X modules
%config %{_sysconfdir}/ld.so.conf.d/nvidia.conf
%dir %{nvidialibdir}/
%{nvidialibdir}/*.so.*
%{nvidialibdir}/tls/
%{_libdir}/xorg/modules/drivers/nvidia_drv.so
%dir %{_libdir}/xorg/modules/extensions/nvidia/
%{_libdir}/xorg/modules/extensions/nvidia/libglx.so
#{_libdir}/xorg/modules/libwfb.so
# Tools and utilities
%{_sysconfdir}/profile.d/*
%{_bindir}/*
%{_sbindir}/*
%{_datadir}/applications/*
%{_datadir}/pixmaps/*
%{_mandir}/man1/*
# Not needed devel but would violate the license not to include them
#files devel
#defattr(-,root,root,-)
%{nvidialibdir}/*.a
%{nvidialibdir}/*.so


just remove the comments to build F8 packages .

[Hlingler]

Um... sorry, don't know how I overlooked that....

Thanks,
V

[wvn]

any special changes in the 177 driver from the 173?

[Hlingler]

any special changes in the 177 driver from the 173?http://www.nvidia.com/object/linux_display_ia32_177.80.html

[wvn]

http://www.nvidia.com/object/linux_display_ia32_177.80.html

Thanx Mr. H. :)

[dado]

You could rebuild my SRPM

http://www.dfm.uninsubria.it/compiz/fusion-testing/9/nvidia/SRPM/nvidia-x11-drv-177.80-1.fc9.src.rpm



[root@localhost Desktop]# rpm -qa |grep nvidia
nvidia-x11-drv-32bit-177.80-1.fc9.x86_64
nvidia-x11-drv-177.80-1.fc9.x86_64
[root@localhost Desktop]#


su
yum groupinstall "Development Tools"
yum install rpmdevtools wget
wget http://www.dfm.uninsubria.it/compiz/fusion-testing/9/nvidia/SRPM/nvidia-x11-drv-177.80-1.fc9.src.rpm
yum-buildeps nvidia-x11-drv-177.80-1.fc9.src.rpm
exit
rpmdev-setuptree
rpmbuild --rebuild nvidia-x11-drv-177.80-1.fc9.src.rpm

Then install the rpm files.

I must have missed something, but the package doesn't build the kmod which is then missing.

[leigh123linux]

I must have missed something, but the package doesn't build the kmod which is then missing.

The nvidia-x11-drv is DKMS based so it builds the kernel module on bootup.


Post


rpm -qa |grep nvidia

[dado]

The nvidia-x11-drv is DKMS based so it builds the kernel module on bootup.


Yeah, I didn't figure that one out and haven't even tried to reboot, it works now. Thank you very much.

[Hlingler]

Yeah, I didn't figure that one out and haven't even tried to reboot, it works now. Thank you very much.You can manually force DKMS to re-build a new kmod, but a re-boot is actually easier, unless for some reason you can't re-boot: http://fedoraforum.org/forum/showpost.php?p=975945&postcount=67.

V

[carpetfeller]

My advice would be to use the binary installer from the nvidia site that's what I always do and I've never had any problems with it...

Until F9. No matter what I did in F9 I could not get SLI to work. So, I switched to Mandriva 2009.0 and it works flawlessly.

Just my two cents, I hope it works out for you. But by my experience: Good luck.

[growl]

http://mirror.yandex.ru/fedora/tigro/9/i386/ you can use drivers from these repos

[a_small_cake]

Are there any sings of 177.80 for Fedora 9 in Livna repo?

[M4rc0]

Oh I should have found this topic earlier..

Tonight Ive decided to update my nvidia driver to the latest 177.82 to be able to get sound on my HDMI output (no sound on the current livna driver)

I got the package from the site, ran it, and it all went well. I opened nvidia settings and it was showing that 177.82 for my surprise. But then when I checked on yumex for nvidia ive found the livna drivers still installed, so i uninstalled them

So i reboot and nothing works anymore :D
It says it cant find the module or something, I also added the BusID to the xorg to try but still no success

The only way to get startx to work is removing xorg.conf

But now comes the worse part, when i type yum search nvidia or install, it doesnt find the package!
I have livna repo enabled and everything, but no nvidia for me to install

So what do i do??

EDIT: Funny, I had 177.82 installed here, i ran nvidia-installer to try to figure it out how to uninstall, then it prompts a message saying it will upgrade to 177.80 because the drive that I have installed, 177.82 is old and needs to be uninstalled lol

I say yes, nothing to lose here, and so it creates a new xorg and I have my nvidia graphic back

What really bothers me is that i have no nvidia-settings here and I still can't find any nvidia packages with yum anymore.

Nothing installed and no nvidia-settings or old drivers 173 to install, nothing.

Any ideas?

[leigh123linux]

Oh I should have found this topic earlier..

Tonight Ive decided to update my nvidia driver to the latest 177.82 to be able to get sound on my HDMI output (no sound on the current livna driver)

I got the package from the site, ran it, and it all went well. I opened nvidia settings and it was showing that 177.82 for my surprise. But then when I checked on yumex for nvidia ive found the livna drivers still installed, so i uninstalled them

So i reboot and nothing works anymore :D
It says it cant find the module or something, I also added the BusID to the xorg to try but still no success

The only way to get startx to work is removing xorg.conf

But now comes the worse part, when i type yum search nvidia or install, it doesnt find the package!
I have livna repo enabled and everything, but no nvidia for me to install

So what do i do??


Try


su
yum erase *nvidia*
yum install kernel-devel kernel-headers gcc


Then re-run the nvidia 177.82 installer

[M4rc0]

Leigh thank you for your quick reply!

Like the edit on my last post I made it work by running nvidia-installer again, it downgraded to 177.80 saying it's newer then 177.82 (?????) and it works now.
For the nvidia-settings, it was just a .rc file with the previous settings..

I'm sure I will be able to see nvidia packages later after performing your command

And I might do that later since there is no reason to update to this 177.80 IMO

No hdmi sound output and no advantage so far

Thanks.