Re: [SLE] Kernel upgrade.
On 6/28/00, Brian wrote:
I have tried to upgrade the kernel on my 6.3 to both the 2.2.14 and 2.2.16 from the rpm`s from SuSE. Both of them resulted in: ran out of input data system halted Right after LILO and it started loading the kernel. What does this mean? I`m going to compile my own from the 2.2.14 since there is a problem with 2.2.16 but I`d like to know why I can`t update using the rpm`s. Or am I doing something wrong? I used the k_deflt.rpm with YaST. Is there something else I have to do that I`m missing?
Brian, I just saw your old message and don't know if you got a reply. Today I did a kernel update using the rpm, ran into the same problem, and fixed it. When compiling a new kernel, most people run 'make zlilo'. It would be nice if the kernel update rpm came with a script which does the same thing. When you update with k_deflt.rpm, you must do 2 things. First, run /sbin/lilo Second, update any other bootloader you are using. Missing either of these steps will cause this type of symptom. In my case, I ran lilo. But I forgot to put a copy of my updated lilo boot sector in my Windows NT C partition, for the NT bootloader. I got 'ran out of input data' When I installed a boot sector in Windows, everything worked. It is probably also best to keep a copy of your old /boot/vmlinuz kernel as /boot/vmlinuz.old and make sure lilo is configured for this also -- see the SuSE manual. Suggestion to SuSE - please include a README and scripts in kernel update rpms to cover this kind of issue. dproc -- To unsubscribe send e-mail to suse-linux-e-unsubscribe@suse.com For additional commands send e-mail to suse-linux-e-help@suse.com Also check the FAQ at http://www.suse.com/support/faq
At 21:10 06/08/2000, dproc@dol.net wrote:
When you update with k_deflt.rpm, you must do 2 things.
First, run /sbin/lilo
Second, update any other bootloader you are using.
This is actually mentioned in the support database article "Updating your kernel using RPM packages". However, a related query... Having updated my 6.4 using the 2.2.16 k_deflt.rpm I've noticed that it can no longer find the scsi and usb modules specified on the INITRD parameter in /etc/rc.config. The above article says that you have to re-run mk_initrd after updating the kernel followed by running /sbin/lilo again. I've done this but it is still looking for the 2.2.14 modules. I've checked the lilo configuration and it is definitely using the 2.2.16 kernel and the new initrd image. Any ideas? Thanks, Mike -- Mike Zanker | mike@zanker.org Northampton | United Kingdom | PGP Public Key available -- To unsubscribe send e-mail to suse-linux-e-unsubscribe@suse.com For additional commands send e-mail to suse-linux-e-help@suse.com Also check the FAQ at http://www.suse.com/support/faq
On Mon, 07 Aug 2000, Mike Zanker wrote:
At 21:10 06/08/2000, dproc@dol.net wrote:
When you update with k_deflt.rpm, you must do 2 things.
First, run /sbin/lilo
Second, update any other bootloader you are using.
This is actually mentioned in the support database article "Updating your kernel using RPM packages".
Thank you for pointing this out. I will find the sdb article. I made the mistake of thinking the sdb was only for fixing problems, such as installation on newer hardware or unusual upgrades. I did not think to look there for routine installation instructions. I stand by my suggestion that the RPM include, if not a 'make zlilo' script, at least a copy of the SDB article in the RPM or in the ftp directory. Updating the boot image in NT C:\ for the NT bootloader is not mentioned in the article, but maybe I am the only dual boot user dumb enough to forget. The SuSE manual will always remind me anyway :-)
However, a related query...
Having updated my 6.4 using the 2.2.16 k_deflt.rpm I've noticed that it can no longer find the scsi and usb modules specified on the INITRD parameter in /etc/rc.config.
The above article says that you have to re-run mk_initrd after updating the kernel followed by running /sbin/lilo again. I've done this but it is still looking for the 2.2.14 modules. I've checked the lilo configuration and it is definitely using the 2.2.16 kernel and the new initrd image.
Any ideas?
This article was newly added to the SDB and apparently only applies to 6.4 and later. I use 6.3 and I did not make a new initrd. (Maybe I should have ??) My new kernel is working fine. I have an IDE hard disk, and my boot seems to be loading the 2.2.16 modules (how can I be sure?) There is a mention in the article about OSS failures, but stock OSS/Free is working well for me. My SCSI CD-ROM is erratic but I don't blame the kernel or the boot process. dproc -- To unsubscribe send e-mail to suse-linux-e-unsubscribe@suse.com For additional commands send e-mail to suse-linux-e-help@suse.com Also check the FAQ at http://www.suse.com/support/faq
participants (2)
-
dproc@dol.net
-
mike@zanker.org