You are not logged in.
Pages: 1
Hello, I hope this does not show up as arabic, apparently that happens in steam, I will have to fix that later on, that problem recently came up today after installing KDE... Whatever, going to replace that anyway.
I am trying to install the NVIDIA drivers for Nvidia Geforce GT 440 (http://www.nvidia.com/object/linux-disp … .06-driver)
ERRORS>
Skipping runlevel check utility runlevel failed to run
Noveau error, making /etc/modprobe.d/nvidia-installer-disable-noveau.conf
Trying again...
Skipping runlevel check utility runlevel failed to run
The kernel header file /usr/src/linux/3.8.7-1-ARCH/include/kernel.h does not exist.
The most likely reason for this is that the kernel source path /usr/src/linux/3.8.7-1-ARCH is incorrect. Please make sure you have installed the kernel source files for your kernel and that they are properly configured.
What I believe would be to install that nvidia-hook thing from the AUR (I have read the wiki) but I want to do this correctly. As I am not very familiar with arch, I would he thankful if you would adress the exact commands and things i would have to do, I have gotten familiar with the basic pacman and such but I am not even near being cunning in this, I am learning however, slowly.
Thanks.
Last edited by Laudes (2013-04-23 18:38:21)
Offline
Using the nvidia installer directly is a really bad idea. Especially when you're trying to install a really, *really* old version of the driver.
Just do a
pacman -S nvidia
and reboot. That's all that's needed.
Offline
Why don't you install the nvidia drivers available with pacman ? Nvidia indicates they work with your card.
[Edit] Gusar was faster than me
Last edited by Cthulhu82 (2013-04-23 17:05:56)
« En essayant continuellement, on finit par réussir.
Donc plus ça rate, plus ça a de chance de marcher. »
Devise Shadock
Offline
Don't forget to install "lib32-nvidia-utils" for your 32 bits applications and "libvdpau".
Last edited by Cthulhu82 (2013-04-23 17:10:55)
« En essayant continuellement, on finit par réussir.
Donc plus ça rate, plus ça a de chance de marcher. »
Devise Shadock
Offline
edit: below
Last edited by Laudes (2013-04-23 18:42:30)
Offline
Did not work.... error. modprobe could not insert nvidia invalid argument. fatal server error: no screens found. googled and found that i should try nvidia beta, and i unpacked that and got 3 .run files and another tarball... using the .run files gives me the same error as in original post.
Offline
Err, I said running the nvidia installer is a *bad idea*. Don't do that. And you don't need the beta.
Did you reboot after installing the driver? If not, the nouveau driver is holding the graphic card, so nvidia failing to load is perfectly normal. If you did reboot, we need more info about what went wrong. This isn't a laptop by any chance, is it?
Offline
Skipping runlevel check utility runlevel failed to run
The kernel header file /usr/src/linux/3.8.7-1-ARCH/include/kernel.h does not exist.
The most likely reason for this is that the kernel source path /usr/src/linux/3.8.7-1-ARCH is incorrect. Please make sure you have installed the kernel source files for your kernel and that they are properly configured.
well, it seems that it's not finding kernel.h ... but i think i know what is happening here, as i ran into something like this for VMware, not too long ago.
First, can you locate kernel.h on your machine? try;
locate kernel.h
you should find a few kernel.h files on your system, we only care about one's in /usr/src/linux-XXX. For example;
/usr/src/linux-3.8.4-rt2-1-l-pa/include/uapi/linux/kernel.h
Now, on my machine i also have (below) for the same kernel;
/usr/src/linux-3.8.4-rt2-1-l-pa/include/linux/kernel.h
but it is symlinked from my fist example. ~ this is the part you are missing, i think... So, if you are missing .../include/kernel.h you can try;
ln -s /usr/src/linux-$(uname -r)/include/generated/uapi/linux/version.h /usr/src/linux-$(uname -r)/include/linux/
then try installing nvidia using pacman.
EDIT: and for installing nvidia - you need to make sure you have both nvidia-utils, nvidia AND lib32-nvidia - if you are using a 64bit/multilib system. It shouldn't matter whether you are using the 'stable' or 'beta' series ~ as long as nvidia-utils && nvidia packages are matching versions.... I also say it doesn't matter which you use ~ because i have the same card as you do in one of my machines, which is currently using the beta driver - but was using the stable driver before that. (stable 313.30 / beta 319.12)
cheerz
PS:
references:
a). header location changes on 3.7+ kernels; http://lkml.org/lkml/2012/10/14/281
b). VMware Archwiki; https://wiki.archlinux.org/index.php/VM … stallation
Last edited by triplesquarednine (2013-04-23 19:09:26)
Offline
Did not work.... error. modprobe could not insert nvidia invalid argument. fatal server error: no screens found. googled and found that i should try nvidia beta, and i unpacked that and got 3 .run files and another tarball... using the .run files gives me the same error as in original post.
Stop.
You clearly don't know what you're doing. Before you dig yourself into a deeper hole, read the wiki. Next, understand the wiki. Then try out the suggestions on the wiki which are appropriate for you.
If you still have problems, read this: How to Ask Questions the Smart Way
If your own research doesn't help resolve your issue, post back here asking clear questions, telling us exactly what you tried, with relevant logs, error messages, etc. in properly formatted [ code ] tags.
It'll help us to help you.
Sakura:-
Mobo: MSI MAG X570S TORPEDO MAX // Processor: AMD Ryzen 9 5950X @4.9GHz // GFX: AMD Radeon RX 5700 XT // RAM: 32GB (4x 8GB) Corsair DDR4 (@ 3000MHz) // Storage: 1x 3TB HDD, 6x 1TB SSD, 2x 120GB SSD, 1x 275GB M2 SSD
Making lemonade from lemons since 2015.
Offline
triplesquarednine - fyi, the headers are only required if you're compiling the driver. The OP has been told repeatedly, even by you(!), that he should just install the pre-compiled arch package with pacman, so all that stuff you posted about headers is irrelevant, and potentially confusing.
Offline
@triplesquarednine: Err, the reason for the error is that he doesn't have linux-headers installed. But that's only needed to compile a driver, something he doesn't need to do. Don't encourage people to run the nvidia installer directly, it messes up GL handling. The Arch packages have a special way to deal with GL stuff, something the nvidia installer has no idea about.
PS. Symlinking headers like that is a hack that a) shouldn't really be used, the driver's build system should be patched instead, and b) was only necessary when the driver didn't have support for the newest kernels. It does support them now, so the hack is obsolete.
Last edited by Gusar (2013-04-23 19:11:12)
Offline
Err, I said running the nvidia installer is a *bad idea*. Don't do that. And you don't need the beta.
Did you reboot after installing the driver? If not, the nouveau driver is holding the graphic card, so nvidia failing to load is perfectly normal. If you did reboot, we need more info about what went wrong. This isn't a laptop by any chance, is it?
Alright, I did reboot, and it is a stationary computer. What information do you need?
Offline
The info asked every time there is a problem: dmesg output and /var/log/Xorg.0.log for starters.
Offline
@Gusar - Please take time to re-read my post, since i NEVER encouraged him to manually use the installer, in fact, it was the opposite. Please, take an extra second to read someone's post (if it is needed) rather than making false accusations - it's not appreciated :\ ...but on the rest, duly noted - and thanks for the update.
@Tomk - he did mention using AUR and possibly the beta driver (both of which entail compiling the driver). I had seen that missing kernel.h message (the OP posted) and knew the solution to that particular problem...
sorry for the noise
Last edited by triplesquarednine (2013-04-23 19:29:15)
Offline
Pages: 1