luckily i saw this while i was reading your reply to my intel/alsa thread since i was also trying to figure out a way to do this.
luckily i saw this while i was reading your reply to my intel/alsa thread since i was also trying to figure out a way to do this.
I actually just updated the HOWTO to use a more generic method to install the driver. There is also another example of installing the virtualbox driver using the same method here: http://ubuntuforums.org/showpost.php...94&postcount=7
Don't try to make something "fast" until you are able to quantify "slow".
Thanks! I think I'll integrate this method into the release after KernelCheck HopeStar!
MK
The one and only, Master Kernel Thread
KernelCheck - Finally. A way to easily download, compile, and install the latest kernel. Release 1.2.5 available now.
You could probably also check for a bunch of other things like virtualbox, vmware, ndiswrapper drivers, firmware, etc and use the same basic method to have them auto-install into new kernels. I'm using this method for both nvidia and virtualbox with custom compiled kernels and it works great. Once you know where the hooks are, there are a lot of possibilities.
Don't try to make something "fast" until you are able to quantify "slow".
Vor,
To be completely honest I do not even know if I entered the script correctly. Here is what I did I followed you instruction to
Then I attempted to install the script by a copy and paste of the entire entryThe first thing I recommend doing this is to move the driver you are using to /usr/src and make a symlink to it. For example:
Code:
sudo mv NVIDIA-Linux-x86-173.14.05-pkg1.run /usr/src
sudo ln -s /usr/src/NVIDIA-Linux-x86-173.14.05-pkg1.run /usr/src/nvidia-driver
all at once except for "exit" then I was lost at how to name the script and install it. When I tried to google the answer for "installing a script Ubuntu" I ran into a wall. I am a complete noobe here at the Ubuntu wheel. Thank you for your reply.#!/bin/bash
#
# Set this to the exact path of the nvidia driver you plan to use
# It is recommended to use a symlink here so that this script doesn't
# have to be modified when you change driver versions.
DRIVER=/usr/src/nvidia-driver
# Build new driver if it doesn't exist
if [ -e /lib/modules/$1/kernel/drivers/video/nvidia.ko ] ; then
echo "NVIDIA driver already exists for this kernel." >&2
else
echo "Building NVIDIA driver for kernel $1" >&2
sh $DRIVER -K -k $1 -s -n 2>1 > /dev/null
fi
Ktechman
Coould you repost the old way to do it, I want to clean and do the new methode but I have forget what was the old..; thanks
I think this should be sufficient to revert the old way and let you follow the instructions for the new way.
First:
Then, ensure that the file /etc/kernel-img.conf looks exactly like this:Code:sudo rm /usr/sbin/update-nvidia-grub
After that, you should be able to follow the instructions for the new method.Code:do_symlinks = yes relative_links = yes do_bootloader = no do_bootfloppy = no do_initrd = yes link_in_boot = no postinst_hook = update-grub postrm_hook = update-grub
Don't try to make something "fast" until you are able to quantify "slow".
Are you certain that you need to use this tutorial? Have you downloaded and installed the nvidia drivers directly from nvidia?
If you are sure you need to use it, just use Applications->Accessories->Text Editor, paste the script into it (including the "exit" part) and save it as update-nvidia in your home directory. You should be able to follow the rest of the guide after that.
Don't try to make something "fast" until you are able to quantify "slow".
Vor,
I have not been at this very long but when I updated from kernel 18 to 19 none I mean none of the settings for my driver were in tact. So for now I reinstalled the driver the manual way and everything is back to normal until another kernel upgrade. I believe that some things that Ubuntu veterans take for granted should be put in the instructions just for the simple reason that even with all of my windows experience I spent little or no time with the command line or adding scripts. MS has babied us with so many user friendly self explaining programs that getting a handle on something new such as Ubuntu takes a bit of a kick in the seat of the pants, so to speak. And with all of my attention being spent on the newest threat from MS who has time to focus on anything else. Please endure with this new generation of users who have ventured over to the other side of the fence where the grass is truly greener. I am eternally grateful for being pulled from the clutches of MS even if they do seem benign, the alternative of going back gives me great pause so I will muddle through over here learning what should already be second nature if not for the constant lulling of a company who's only concerned is with their bottom line, hiding behind the guise of being benevolent with the highly publicized facts of Bills philanthropy. With all that said I appreciate your response and the help. BTW went without a hitch.
Best Regards
Last edited by ktechman; July 4th, 2008 at 07:19 PM.
Ktechman
Why are we piping output to /dev/null here? This strikes me as a bad idea. If something goes wrong in this step, the script dies silently, making debugging a real pain. I know this because it just happened to me.
I would suggest not hiding the output from the sh command - it may make your log file look 'ugly', but usually ugly is preferred to invisible in case of failure.
Many thanks for this btw; I use a number of patched modules in my systems and was unaware of this hook. Now I don't need to wait to upgrade my kernel until I have 'enough time'.
/drdaz
Last edited by drdaz; July 6th, 2008 at 06:50 PM. Reason: Forgot to say thanks.
Bookmarks