Are there any official, almost-official or maybe even reasonably stable kernel repositories with new kernels for opensuse 11.2? I recently changed my laptop for a new one which has some issues, which then again are solved in 2.6.33 (and even in 2.6.32 with some tweaking). I would by any means try to avoid kernel recompilation.
openSUSE does not provide major kernel upgrades to a distribution’s version. You will only see updates to 2.6.31 in the “official repos”, i.e. Update. If you want another kernel, either use the above at your own risk, or compile your own kernel.
Until 2 hours ago, the HEAD kernel was the same as for M6. Now 2.6.34-rc7 is in the HEAD repo. I have it running on my laptop. No issues so far.
Mind again: forgot to mention that installing this kernel will remove some installed packages, IIRC preload-kmp. I have not found any delay or slowdown after doing so.
The problem I’m experiencing is related to USB ports, being dead after the machine wakes up from sleep mode, so I guess a live CD wouldn’t prove much.
Additional problem which maybe waits for me is related to NVidia drivers, which I’ll obviously need to compile manually, even more I’ll have to use beta ones…
Just to complete the story, this new laptop is Lenovo ThinkPad T510, which is really great and better than the old one, ThinkPad R61, but on R61 I had some crazy uptimes for a laptop, actually I used to reboot the machine only on kernel updates, everything worked flawlessly.
Unfortunatelly, computer won’t even boot with the latest, 2.6.34-rc7 kernel. It looks like that kernel doesn’t have compiled support for ext4, which is quite strange. No luck here…
Something went wrong during your attempts to install the kernel. Don’t know what, but IMHO it can be fixed. You will need an 11.2 LiveCD. Download one, burn the ISO to disk and boot from it. Get back here as soon as you have that, I can help you through.
Hi Knurpht, and thank you very much for your kind offer.
Fortunatelly, I managed to revert the stable kernel and right now I’m tweaking other packages, nvidia driver and so on. 10+ years in linux make you kernel-panic-proof
To install, start in runlevel 3
sh NVIDIA…run -q, accept all,
init 5
10+ in linux should make this mean something to you . Nice you got it sorted though.
Tested this tonight: Reinstalled 11.2’s 2.6.31 kernel-packages, incl. sources, added Kernel:HEAD repo, started software installer, selected the Kernel:HEAD repo in the repo view, clicked the “Switch system packages to…”, accepted, installed, rebooted, all OK, back on 2.6.34-rc7 again, NVIDIA driver installs without problem.
Thanks for your detailed description. I’ve followed the steps, installed the Linux Kernel Development pattern and upgraded my kernel:
davor@linux:~> uname -a
Linux linux 2.6.34-35-desktop #1 SMP PREEMPT 2010-05-20 21:31:23 +0200 i686 i686 i386 GNU/Linux
However, NVidia driver fails to load. It seems to be compiled correctly, but not installed and loaded. I’ve tried both 195.36.24 and 195.36.15 versions, with the same results. This is the error log message (from 195.36.15):
...snip...
/usr/src/linux-2.6.34-35/arch/x86/include/asm/mmzone_32.h: In function ‘pf
n_valid’:
/usr/src/linux-2.6.34-35/arch/x86/include/asm/mmzone_32.h:87: warning: compa
rison between signed and unsigned integer expressions
ld -r -m elf_i386 -T /usr/src/linux-2.6.34-35/scripts/module-common.lds --
build-id -o /tmp/selfgz14424/NVIDIA-Linux-x86-195.36.15-pkg1/usr/src/nv/nvid
ia.ko /tmp/selfgz14424/NVIDIA-Linux-x86-195.36.15-pkg1/usr/src/nv/nvidia.o /
tmp/selfgz14424/NVIDIA-Linux-x86-195.36.15-pkg1/usr/src/nv/nvidia.mod.o
NVIDIA: left KBUILD.
-> done.
-> Kernel module compilation complete.
ERROR: Unable to load the kernel module 'nvidia.ko'. This happens most
frequently when this kernel module was built against the wrong or
improperly configured kernel sources, with a version of gcc that differs
from the one used to build the target kernel, or if a driver such as
rivafb/nvidiafb is present and prevents the NVIDIA kernel module from
obtaining ownership of the NVIDIA graphics device(s), or NVIDIA GPU
installed in this system is not supported by this NVIDIA Linux graphics
driver release.
Please see the log entries 'Kernel module load error' and 'Kernel
messages' at the end of the file '/var/log/nvidia-installer.log' for
more information.
-> Kernel module load error: insmod: error inserting './usr/src/nv/nvidia.ko':
-1 No such device
-> Kernel messages:
2712.852192] NVRM: Try unloading the rivafb, nvidiafb or rivatv kernel
module
2712.852192] NVRM: (and/or reconfigure your kernel without rivafb/nvidiafb
2712.852193] NVRM: support), then try loading the NVIDIA kernel module
again.
2712.852198] NVRM: No NVIDIA graphics adapter probed!
2902.163757] CPUFREQ: Per core ondemand sysfs interface is deprecated -
...snip...
no nvidiafb, rivafb nor rivatv are loaded, graphic card is supported. gcc looks suspicious, although I don’t where’s the problem:
Eh… I remembered that “native” nvidia module is now available in kernel sources, nouveau, which NVidia installation procedure didn’t mention. I had to blacklist it in /etc/modprobe.d/50-blacklist.conf
After restart, nvidia driver compilation procedure finally succeeded and now I’m happily running 2.6.34 with the lastest nvidia driver!
Right now I have only one kernel (the newest one, 2.6.34, check my older posts to see which packages are installed) which has been installed over the original one, 2.6.31.12. Other, non-kernel related packages are the same and everything works just fine.
However, I remember my older days when I had enough patience to poke my system here and there: having many kernels on the same system is quite possible and could be done without any problems, especially when the kernels are of the same major version (e.g. 2.6.x) and reasonably compatible.