Old posts from the morph-sources thread

rommel · Posted: Sun Feb 06, 2005 1:24 am Post subject:

i cant get kexec to work properly... it loads the image , then i issue reboot with the edited boot.sh (added -k opt ) and i see all the services being brought down then then a messege starting new kernel and the system locks.... i am not really sure how to troulble shoot this. any ideas?

teutzz · Guru Joined: 22 Apr 2004 Posts: 333 Location: .ro

@yardbird: could you includ generic anticiatory in the next morph?
_________________
Cand nu stii ce sa raspunzi sau ce sa spui un simplu BLA ajunge... lolz

predatorfreak · Posted: Sun Feb 06, 2005 10:26 am Post subject:

Lepaca Kliffoth · Posted: Sun Feb 06, 2005 1:43 pm Post subject:

Remeber my problem with the slow boot process? From a certain point the boot sequence slows down greatly and services take a lot of time to give the [OK]. Seems like the culprit is a certain "Setting terminal blah blah ASCII blah" which usually doesn't appear when booting from a different kernel (nitro and vivid). Everithing works fine until that message appears (btw, it gives the [OK] even if it's obviously causing problems). Any idea? Do you need more detailed informations, like dmesg and my kernel .config? Thanks.
_________________
It isn't enough to win - everyone else must lose, and you also have to rub it in their face (maybe chop off an arm too for good measure).
Animebox!

yardbird · l33t Joined: 20 Apr 2002 Posts: 689 Location: nl.leiden

Hiya all,

2.6.10-morph22 is released. Changelog:

- updated inotify patch
- fixed a compilation error in kexec.

This fixes the compilation problems people have been experiencing without inotify and kexec. Also the inotify patch has been updated to the latest one available. I'm waiting for the softwaresuspend folks to release a 2.6.10 version of the new stable swsusp patch. It will go into next morph-sources.

rommel: you best bet is to try to unload all possible modules before rebooting. Also, double check your configuration for the /etc/init.d/kexec script, just for sure.

teutzz: I'm not too inclined to include genetic-as. cfq seems better to me in a desktop environment, and genetic seems to be suited more for an experimental patchset.

Lepaca Kliffoth: it's pretty weird that script is executed only with certain kernel versions. Are you using fbsplash in morph and maybe not in vivid/nitro, by chance? Maybe in /var/log/messages you can find some debug info from init scripts.
_________________

mayday147 · Posted: Mon Feb 07, 2005 7:02 am Post subject:

genstef · Posted: Mon Feb 07, 2005 8:50 am Post subject:

Did you etc-update everything correctly?

mayday147 · Posted: Mon Feb 07, 2005 9:36 am Post subject:

genstef · Posted: Mon Feb 07, 2005 10:03 am Post subject:

sorry, you have to etc-update to get a working kexec-tools init-script.
Maybe you have a good reason to learn it now

mayday147 · Posted: Mon Feb 07, 2005 11:06 am Post subject:

mayday147 · Posted: Mon Feb 07, 2005 11:24 am Post subject:

It works now. Thanks genstef.
_________________
gentoo.ro

k_max · n00b Joined: 28 Oct 2004 Posts: 20

could someone please post the /etc/init.d/kexec script?
i'm using debian, so i can't emerge it
_________________
2.6.12-rc3-love1 / 2.6.12-rc4-nitro1
static custom DSDT

mayday147 · Posted: Mon Feb 07, 2005 8:20 pm Post subject:

k_max · n00b Joined: 28 Oct 2004 Posts: 20

2 mayday147
thanks, i'll try it.
_________________
2.6.12-rc3-love1 / 2.6.12-rc4-nitro1
static custom DSDT

genstef · Posted: Mon Feb 07, 2005 8:57 pm Post subject:

That is easy to get through our viewcvs:
http://www.gentoo.org/cgi-bin/viewcvs.cgi/*checkout*/sys-apps/kexec-tools/files/kexec.init

But I doubt it will work for debian because of the special conf.d, which is not there and the special functions like depend() in gentoo-init-scripts.

k_max · n00b Joined: 28 Oct 2004 Posts: 20

2 genstef
of course i'll edit those gentoo-specific things before starting

_________________
2.6.12-rc3-love1 / 2.6.12-rc4-nitro1
static custom DSDT

yardbird · l33t Joined: 20 Apr 2002 Posts: 689 Location: nl.leiden

genstef: here's the broken-out tarball:

http://morph-sources.sourceforge.net/broken-out.tar.bz2

One of the kexec patches must be applied with the "-l" switch (I couldn't be bothered to find where the missing/extra space was). Sorry about the delay, I forgot about this till now

_________________

mayday147 · Posted: Tue Feb 08, 2005 5:07 pm Post subject:

A stupid question maybe: What does the CD packet writing patch do? What means packet writing?
_________________
gentoo.ro

yardbird · l33t Joined: 20 Apr 2002 Posts: 689 Location: nl.leiden

nightfrost · Apprentice Joined: 07 Dec 2004 Posts: 293 Location: Sweden

yardbird · l33t Joined: 20 Apr 2002 Posts: 689 Location: nl.leiden

nightfrost · Apprentice Joined: 07 Dec 2004 Posts: 293 Location: Sweden

Actually, I first suggested the patch thinking of people who would take out their xbox-harddrive and put it in their tux-box in order to transfer files and stuff. But thinking over it again it's really gonna be easier to use a crossover cable. However, it would be very cool if your kernel patchset could be used for the xbox-distros as well. According to Michael Steil from the Xbox Linux Project, the changes are rather minimal.

I think this link gives you all the info and further links...
http://www.uwsg.iu.edu/hypermail/linux/kernel/0210.1/1112.html

Tell us what you think about it

Moloch · Posted: Wed Feb 09, 2005 10:55 pm Post subject:

Not to knock your suggestion. I really don't think this patchset should be riddled with drivers and things only one person may use. Mostly it's more work for the maintainer, but can add possible problems in general to the kernel. Unless there is a huge demand for something like this or yardbird is personaly interested, we should keep the number of patches down to a minimum.
_________________
Understanding is a three-edged sword: your side, their side, and the truth. --Kosh
1010011010