Carlos E. R. wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
The Wednesday 2007-12-05 at 14:37 -0500, Richard Creighton wrote:
Shortly after rebooting, the updater said it had a security patch on the new .13 kernel and (I know, I should know better than to trust anything by now), it was small, a patch and I said, go ahead, it isn't installing a new kernel, just requires a reboot to load it into memory after the update.....Yeah, right!....
Patch or no patch, a kernel patch replaces the whole kernel, even if only a small part of it actually changes.
I know it replaces the kernel, it has to, but I wasn't expecting it to REPLACE 2 earlier versions with the patched version which seems to have the same version number....yup....just checked, the version number did not change and the patch didn't mention anything about a version change
It not only ate my GRUB configuration files and replaced them, it also destroyed (by erasure) all of the other linux kernels in /boot, their syms AND all of the modules AND sources in /lib for those versions!!!!! Dammitalltohellanyway!!!!
All the kernels with different version numbers that the one it was replacing? It should only replace the previous kernel, no more. If it removes other kernels, open a bugzilla.
It erased 2.6.22.12-0.1-bigsmp 2.6.22.9-0.4 bigsmp and reinstalled 2.6.22.13-0.3-bigsmp (I assumed patched) which was already online unpatched with that version.
And, if you say you had compiled your own kernel, that one would not be touched - provided you compiled it with a different name. Furthermore, once you give the kernel a name (inside the make), the /lib/modules/ tree receives also a different name, and that one is not replaced:
1505880 Nov 22 22:21 /boot/vmlinuz-2.6.22.12-0.1-cer <-- mine 1559220 Nov 12 04:13 /boot/vmlinuz-2.6.22.12-0.1-cer.old <-- mine, old 1593968 Nov 7 17:09 /boot/vmlinuz-2.6.22.12-0.1-default <-- theirs
Only "theirs" is replaced. Plus:
I start with a make mrproper but I am not enough of a kernel expert to answer the thousands of questions in the .config so I compile using a make O=<my source area> oldconfig and the only thing I have to do when done is to make O=<my source area> modules_install install insmod /lib/modules/`uname -r`/kernel/drivers/scsi/rr174x.ko depmod mkinitrd
/lib/modules/2.6.22.12-0.1-cer/ <=== mine /lib/modules/2.6.22.12-0.1-default/ /lib/modules/2.6.22.12-0.1-debug/ /lib/modules/2.6.22.12-0.1-xen/ /lib/modules/2.6.22.12-0.1-xenpae/
Those I have named your kernel, prior to compile, mind! are not replaced by Yast. They don't belong to any rpm, thus they are not touched.
The sources, yes, that would be lost unless I copied them over. That is known.
But yes, for something as important as a kernel, it should ask. But I think you can go into YOU (never, ever, do an automatic update) and select the previous kernel to be maintained - but... no, as it has the same rpm name, the option is not given. You can't.
For many weeks, going on months, I studiously ignored the red triangle of the updater. I weakened this one time. I will NEVER trust the updater again. All updates had been done via the YAST online update where I had some control over what and when. Kinda defeats the purpose of the automatic updater, doesn't it? I have filed bugreports until I am tired of being ignored. One nice thing about being a retired college professor after retiring from being a programmer for many years, is that I now have a lot of time to annoy Novel senior managment by trying to write lucid letters to them in the hopes that writing business case (read dollars and cents) reasons why it is to their interests to pay attention to these little details. I live and breathe Linux, love SuSE and invoke Gates' name in vain as often as I can, but when stuff like this happens due to poor management decisions being made, Linux, SuSE and Novell (where the money comes from) all suffer in the long run. We want the 'unwashed masses' that currently run MS to come and try out SuSE and abandon monopolistic offerings of Gates and Co., but when stuff like this is allowed to fester unchallenged, it is really unlikely that openSuSE, et al., we ever make that much of an impact on those we want most to switch. If you have to be an ex college professor that taught OS theory in order to have a chance to succeed with openSuSE, what chance does the average Windoze schmuck have at success? Only by great luck will he have any. Cheers back atcha Richard
- -- Cheers, Carlos E. R.
-----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.4-svn0 (GNU/Linux)
iD8DBQFHVwLvtTMYHG2NR9URAr6jAJ9hpzL0ZGxPsUSpk3AmoDCH/yYoMwCgkYG7 0K09hp35c8RTelJ/ErbTWk0= =4Xya -----END PGP SIGNATURE-----
-- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org