Slackware :: Accidentally Upgraded Kernel-headers - Should Downgrade ?
May 22, 2011
When I first did my install of Slackware 13.37, I installed the 2.6.38.4 kernel from /testing.
I did a "upgradepkg testing/kernel*.t?z"
Well just now I realized that there was a kernel-headers package in /testing, and I've heard that you should only use the kernel headers that glibc was compiled with.
So did I make a mistake installing the kernel-headers from /testing?
And if I revert back to the stock kernel-headers package, will I have to recompile all the programs I've compiled with the 2.6.38.4 headers?
But the system run into problem with WLAN, and I've search in this site and sombody posted in here: [URL] It's almost the same, the different is the card I use is TP-LINK. Then I decided to remove the newly installed kernel:
I'd like to upgrade a live ISO file before burning a new CD.
So, following the instructions in "How to Customise the Ubuntu Desktop CD", I chrooted to the squashfs and ran "apt-get update/upgrade/dist-upgrade", expecting apt-get to upgrade the chroot kernel... only to find out that apparently, even when being chrooted, this command upgrades the actual kernel. I guess it makes sense, but I was under the - wrong - impression that chroot would build a filesystem totally isolated from the underneath host.
So I rebooted to use this new kernel:
Next, since the Nouveau video driver requires the kernel header files, I ran the following:
Code:
I don't understand why apt-get upgrades the kernel but won't install its header files.
I'd like to upgrade a live ISO file before burning a new CD.
So, following the instructions in "How to Customise the Ubuntu Desktop CD", I chrooted to the squashfs and ran "apt-get update/upgrade/dist-upgrade", expecting apt-get to upgrade the chroot kernel... only to find out that apparently, even when being chrooted, this command upgrades the actual kernel. I guess it makes sense, but I was under the - wrong - impression that chroot would build a filesystem totally isolated from the underneath host.
So I rebooted to use this new kernel:
Code: # uname -r 2.6.32-24-generic
Next, since the Nouveau video driver requires the kernel header files, I ran the following:
Code: # apt-get install linux-headers-$(uname -r) Reading package lists... Done Building dependency tree Reading state information... Done
I have installed and configured Slackware 13.1 and it works perfectly. Now I'm going (actually forced) to downgrade kernel version down to 2.6.32. Is there any caveats? Do I need rebuild any other packages from sources?
I just noticed that I had a sid line on my sources.list, at the end of all squeeze lines, and I've been updating more or less regularly. Does the ordering imply some sort of "preference", so that the first lines are preferred over the subsequent ones, or is it all about the latest version?After removing that line, and taking a look at my installed packages with synaptic, all the "latest versions" of the packages seem to be the ones that are already installed, and I click on properties, and there are no older versions available, only "##.##-### (testing)". Emphasis on "testing", rather than "sid" or "unstable". Of course I didn't check all of them, but about a dozen, and none deviated from that pattern. However, I went to double check on the web pages about the packages, and it seems that many of the squeeze's packages are at the same version of sid. I couldn't yet find one that is not.perhaps the sole fact that the sid line was on sources.list is enough to know that something happened -- as I don't recall having ever set up those configuration files about preferred sources/pins or however it's called), and suggestions on how to proceed? My hunch would be to just downgrade things gradually as they eventually fail (as I did with geeqie, which is why I suspect that there is some mess), but leave everything alone. Not try to fix what is not quite broken. And if it fails too often, go full-blown sid.
I made the terrible mistake of upgrading my live Debian Lenny web server with the the dist-upgrade option in apt-get. I didn't realize this was actually an unstable upgrade and now I have had to make all sorts of choices of what configuration files to keep or upgrade ect. The apache conf files were actually bad after the upgrade and I had to replace them with the backups (phew) and the system is currently still up and running. However my virtualmin installation is no longer working due to a issue with perl ( but thats another question I guess to ask somewhere else maybe ). Anyways... I'm very scared to restart because my server is co-located somewhere else and Im the only one who has ever worked on this server so I would need to go there and fix it myself if it doesn't restart. Basically I have two questions.. is there an easy way to move back to stable packages..If so is this recommended?
And also I'm currently trying to fix some broken dependencies in the package manager but when I run "sudo aptitude -f install" It keeps telling me it is going to remove all of these packages (listed below), some of which I know are very important to the system and I cannot figure out why it would keep trying to do this. I get an error on "phonon-backend-xine" whenever it tries to upgrade just saying this
"(gtk-update-icon-cache:12343): GdkPixbuf-WARNING **: Cannot open pixbuf loader module file '/usr/lib/gdk-pixbuf-2.0/2.10.0/loaders.cache': No such file or directory"
I installed some desktop related packages a while back like gnome-desktop and I know the package is related to this, but all I really care about is making sure the server stays online and not about the desktop packages. I tried just removing kdebase-runtime and anything else that is dependent on it, but it wont let me do anything at all without fixing this broken package.
I really would just like to go back to lenny stable again but I know its probably too late since I already had it install a new kernel and grub 2 (auto configuring my new grub.list)..
Yesterday, I think I did something stupid: I removed kernel-headers, gcc, glibc-devel and glibc-headers. My box is a CentOS 5.4 webserver (it has loads of packages installed, but that was done through Virtualmin config, so it's quite coherent all in all). The thing is that now I need to reinstall at least the headers and glibc, but hey! this is what I get :
I'm just installed OpenSuse 11.3 (64) on a 30gb SSD, hoping to get virtualbox 4.0 running to virtualize an instance of Windows 7.I went through some pain with my Nvidia video card and actually getting vb to install, but through lots of searching and tinkering got here.I created a vm in the vb control panel, but when I go to start it I get:
Code: Failed to open a session for the virtual machine Win7Main. The virtual machine 'Win7Main' has terminated unexpectedly during startup with exit code 1.
As a follow-on to something Telemachos said in another post:
Quote:
Originally Posted by Telemachos
You can see what kernels you have installed - to check if you have a virtual kernel and to clean up - by running this command:
Code:
If you've been installing kernel-headers along with the kernels (say to build modules for graphics or wireless), you should remove those when you remove the corresponding kernel. The command to search for those is parallel:
Code:
I would have thought that removing a given kernel package would trigger the removal of the older kernel headers. Can someone confirm that is, or is not, the behavior? I ask this because it seemed to me that the older kernel header packages were indeed removed when I removed some older kernel packages.
For example, the linux kernels I have installed are:
Code:
Also, the linux-headers packages I have installed are:
Code:
So, when I get around to removing the linux-image-2.6.25-2-amd64 package like this:
Code:
I would expect apt-get to automatically also remove linux-headers-2.6.25-2-amd64 and linux-headers-2.6.25-2-common. Is that what will happen, or do I need to explicitly state all three packages on the apt-get remove command?
I'm attempting to install the driver for my atheros AR8131 PCI-E Gigabit Ethernet adapter (in my Lenovo laptop) on my newly installed RHEL5 system (it's not currently being recognized).
I tried using: 'make install' but hit an error "Makefile:61: *** Linux kernel source not found."
After this, I tried: 'sudo yum install kernel-devel kernel-headers'
To rectify this, but hit this error "No package kernel-devel available" (and the same for the headers). What should I do?
I get the following error message trying to install dazuko on xubuntu 10.04: "headers for target kernel version could not be found" But when I run sudo apt-get install linux-headers-$(uname -r), I get the message that I already installed the headers. My current kernel is 2.6.34-020634-generic
How can I install dazuko withouth having this problem??
Is there a way to get the matching Linux kernel headers automatic on a regular kernel update via the Ubuntu packed manager? Every time I get a new kernel I must do an aptitude install linux-headers-`uname -r`
A recent kernel update seems to have misplaced the Kernel Headers. VMWare needs these headers and cannot find them. Attempting to run VMWARE gets the message: Kernel headers for version 2.6.31.12-0.2-desktop were not found.
I'm not sure if I am posting this in the right place, if not im sorry. I want to downgrade the kernel to 2.6.32 or a little older. I see the if I add the 11.2 repo I can install 2.6.31. Will that work though? Id rather not tank my system. I need to the older kernel for better support in a app I regularly use.
It looks like the latest kernel update has borked the sound on my Macbook Pro 5 4. Is it possible to downgrade to the previous version (which i believe is -19) where my sound worked? If so, how do I go about this?
I was removing the latest two kernels installed because of some issues with compiz and stuffs. And I accidentally removed the only kernel left too. Now, grub has only one entry for Windows. I used "yum remove kernel-.....". Now, I'm stuck, I don't want to re-install the system. It FC11 using LVM. Is there any way I can install re-install the kernel only. What I'm thinking of is copying the kernel image from a live fc11 or installing via rpm on maintenance mode if its possible.
downgrade my squeeze installation to kernel 2.6.32-29 (I think I have -30 now).
- what is the proper way to downgrade to that version of the kernel (I got my current kernel version when I installed squeeze few days ago and the official repo does not seem to have the -29 version anymore)?
- how to prevent that future automatic updates "upgrade" me again to the non-working version?
I've installed Vmware Workstation and tried to run it only to have the following appear;
Kernel Headers 2.6.27.25-170.2.72.fc10.i586
Kernel headers for version 2.6.27.25-170.2.72.fc10.i586 were not found. If you installed them in a non-default path you can specify the path below etc.......
I've tried to search and find the kernel headers but can't find them.
Today, on my 11.3 machine. the kernel was updated. When I started my vmware 7 workstation, it came up with a message "kernel-headers for 26.34.7-0.7 were not found. enter an alternative location"
Does anyone know the kernel-headers location, or how to determine that location, in Fedora 13? I'm installing vmware-tools and it's prompting for it. /usr/include/ and /usr/include/linux/ were revealed to have many header files, as shown by doing rpm -ql kernel-headers
However the installer rejected these locations. My only guess as to why is because they're not where the currently-running kernel has them. I also tried /usr/src/kernels/(kernelversion).fc13.i686/include/ with no luck...
