View previous topic :: View next topic |
Author |
Message |
yodermk Apprentice
![Apprentice Apprentice](/images/ranks/rank_rect_2.gif)
Joined: 09 Jun 2003 Posts: 189 Location: Quito, Ecuador
|
Posted: Tue Nov 25, 2003 8:22 am Post subject: |
|
|
Cool, I'm running test10 and can confirm that there are no PS/2 mouse problems. Works great!
Also, my 3D is working now, thanks to the tip someone gave me in my other thread -- setting
Code: | export MESA_FORCE_SSE=1 |
(this is the first time I've re-booted to a kernel compiled with DRI since that thread)
glxgears is giving me 430fps, which is IIRC a bit better than it was under 2.4. (I just have a G400, so it's nothing great. Although it gets >700fps with Knoppix!) |
|
Back to top |
|
![](templates/gentoo/images/spacer.gif) |
Ulukay Tux's lil' helper
![Tux's lil' helper Tux's lil' helper](/images/ranks/rank_rect_1.gif)
![](images/avatars/214266036742dbfd90e8520.jpg)
Joined: 08 Oct 2002 Posts: 143
|
Posted: Tue Nov 25, 2003 12:54 pm Post subject: |
|
|
hmm
no AGPGART for me
VIA K8T800 A64 ![Sad :(](images/smiles/icon_sad.gif) _________________ Jeden Tag erhöht sich zwangsläufig die Anzahl derer, die mich
am Arsch lecken können... Das ist heute DEIN Tag! |
|
Back to top |
|
![](templates/gentoo/images/spacer.gif) |
Flo.Lue n00b
![n00b n00b](/images/ranks/rank_rect_0.gif)
Joined: 27 Dec 2002 Posts: 6 Location: Germany
|
Posted: Tue Nov 25, 2003 4:59 pm Post subject: Re: 2.6 kernels hang during boot |
|
|
glennj wrote: |
[snip]
Without fail, the 2.6 kernels fail while starting up processes in /etc/init.d. Here is what is started up (in order): metalog cupsd eth0 portmap fam named netmount nscd ntp-client ntpd postfix samba spamd upsd upsmon vcron xfs gdm It will usually hang while launching postfix but sometimes it will get past postfix and hang while launching samba.
Is anyone besides me seeing this? Does anyone have an idea about what the problem might be?
|
I've got exactly the same problems. I even removed samba from the default runlevel. The effect is that I get to the login prompt which, after entering username and password, just gives the message "Login timed out after 60 seconds".
When I try to reboot my box using ctrl+alt+backspace it hangs when trying to shut down postfix.
A few weeks ago I had a completely working 2.6.0-test4 on my box...
Any ideas?
tia Florian |
|
Back to top |
|
![](templates/gentoo/images/spacer.gif) |
glennj n00b
![n00b n00b](/images/ranks/rank_rect_0.gif)
Joined: 23 Oct 2003 Posts: 18
|
Posted: Tue Nov 25, 2003 5:35 pm Post subject: Re: 2.6 kernels hang during boot |
|
|
Flo.Lue wrote: | glennj wrote: |
[snip]
Without fail, the 2.6 kernels fail while starting up processes in /etc/init.d. Here is what is started up (in order): metalog cupsd eth0 portmap fam named netmount nscd ntp-client ntpd postfix samba spamd upsd upsmon vcron xfs gdm It will usually hang while launching postfix but sometimes it will get past postfix and hang while launching samba.
Is anyone besides me seeing this? Does anyone have an idea about what the problem might be?
|
I've got exactly the same problems. I even removed samba from the default runlevel. The effect is that I get to the login prompt which, after entering username and password, just gives the message "Login timed out after 60 seconds".
When I try to reboot my box using ctrl+alt+backspace it hangs when trying to shut down postfix.
A few weeks ago I had a completely working 2.6.0-test4 on my box...
Any ideas?
tia Florian |
I set up a new empty runlevel and added the services one by one. I see the same thing as you in that I can get a login prompt without postfix, samba, and everything after that in the list above, but I can not login. If I also remove ntp-client and ntpd then I can log in. This seems to indicate that it is some sort of a network problem.
I have tried with two different network cards (different brands, drivers, etc.) but get the same result.
I started trying 2.6 kernels with test9 so unlike you, I have never had a working 2.6 kernel. For what it is worth, I have tried turning off acpi and apic but those options did not help. |
|
Back to top |
|
![](templates/gentoo/images/spacer.gif) |
irf2003 Veteran
![Veteran Veteran](/images/ranks/rank_rect_5_vet.gif)
Joined: 10 Sep 2003 Posts: 1078
|
Posted: Tue Nov 25, 2003 8:42 pm Post subject: Re: 2.6 kernels hang during boot |
|
|
glennj wrote: | Flo.Lue wrote: | glennj wrote: |
[snip]
Without fail, the 2.6 kernels fail while starting up processes in /etc/init.d. Here is what is started up (in order): metalog cupsd eth0 portmap fam named netmount nscd ntp-client ntpd postfix samba spamd upsd upsmon vcron xfs gdm It will usually hang while launching postfix but sometimes it will get past postfix and hang while launching samba.
Is anyone besides me seeing this? Does anyone have an idea about what the problem might be?
|
I've got exactly the same problems. I even removed samba from the default runlevel. The effect is that I get to the login prompt which, after entering username and password, just gives the message "Login timed out after 60 seconds".
When I try to reboot my box using ctrl+alt+backspace it hangs when trying to shut down postfix.
A few weeks ago I had a completely working 2.6.0-test4 on my box...
Any ideas?
tia Florian |
I set up a new empty runlevel and added the services one by one. I see the same thing as you in that I can get a login prompt without postfix, samba, and everything after that in the list above, but I can not login. If I also remove ntp-client and ntpd then I can log in. This seems to indicate that it is some sort of a network problem.
I have tried with two different network cards (different brands, drivers, etc.) but get the same result.
I started trying 2.6 kernels with test9 so unlike you, I have never had a working 2.6 kernel. For what it is worth, I have tried turning off acpi and apic but those options did not help. |
does it sounds like
https://bugs.gentoo.org/show_bug.cgi?id=33272
??
what version of baselayout do you have emerged?
/init.d/net.eth0 is broken in the latest version of baselayout
use the one in rc-scripts-1.4.3.11p2.tar.bz2 till they solve the problem.
1) extract this file /init.d/net.eth0 from /usr/portage/distfiles/rc-scripts-1.4.3.11p2.tar.bz2
2) backup /etc/init.d/net.eth0
3) copy the extracted net.eth0 to /etc/init.d/net.eth0
4) reboot
HTH |
|
Back to top |
|
![](templates/gentoo/images/spacer.gif) |
glennj n00b
![n00b n00b](/images/ranks/rank_rect_0.gif)
Joined: 23 Oct 2003 Posts: 18
|
Posted: Tue Nov 25, 2003 11:24 pm Post subject: Re: 2.6 kernels hang during boot |
|
|
irf2003 wrote: | glennj wrote: | Flo.Lue wrote: | glennj wrote: |
[snip]
Without fail, the 2.6 kernels fail while starting up processes in /etc/init.d. Here is what is started up (in order): metalog cupsd eth0 portmap fam named netmount nscd ntp-client ntpd postfix samba spamd upsd upsmon vcron xfs gdm It will usually hang while launching postfix but sometimes it will get past postfix and hang while launching samba.
Is anyone besides me seeing this? Does anyone have an idea about what the problem might be?
|
I've got exactly the same problems. I even removed samba from the default runlevel. The effect is that I get to the login prompt which, after entering username and password, just gives the message "Login timed out after 60 seconds".
When I try to reboot my box using ctrl+alt+backspace it hangs when trying to shut down postfix.
A few weeks ago I had a completely working 2.6.0-test4 on my box...
Any ideas?
tia Florian |
I set up a new empty runlevel and added the services one by one. I see the same thing as you in that I can get a login prompt without postfix, samba, and everything after that in the list above, but I can not login. If I also remove ntp-client and ntpd then I can log in. This seems to indicate that it is some sort of a network problem.
I have tried with two different network cards (different brands, drivers, etc.) but get the same result.
I started trying 2.6 kernels with test9 so unlike you, I have never had a working 2.6 kernel. For what it is worth, I have tried turning off acpi and apic but those options did not help. |
does it sounds like
https://bugs.gentoo.org/show_bug.cgi?id=33272
?? | I do not think so.
Quote: | what version of baselayout do you have emerged? | 1.8.6.10-r1
From reading the thread it seems like this version is not affected. Should I try to upgrade baselayout to 1.8.6.11?
Thanks. |
|
Back to top |
|
![](templates/gentoo/images/spacer.gif) |
pwk.linuxfan n00b
![n00b n00b](/images/ranks/rank_rect_0.gif)
Joined: 09 Jan 2003 Posts: 25
|
Posted: Tue Nov 25, 2003 11:38 pm Post subject: hmmm |
|
|
just guessing, but it sounds a bit like this http://kerneltrap.org/node/view/1679 - so maybe you should try compiling your kernel without preempion and see how that goes... |
|
Back to top |
|
![](templates/gentoo/images/spacer.gif) |
glennj n00b
![n00b n00b](/images/ranks/rank_rect_0.gif)
Joined: 23 Oct 2003 Posts: 18
|
Posted: Wed Nov 26, 2003 12:25 am Post subject: Re: hmmm |
|
|
pwk.linuxfan wrote: | just guessing, but it sounds a bit like this http://kerneltrap.org/node/view/1679 - so maybe you should try compiling your kernel without preempion and see how that goes... |
I have built kernels without pre-emption enabled but get the same resulting failure. |
|
Back to top |
|
![](templates/gentoo/images/spacer.gif) |
TenPin Guru
![Guru Guru](/images/ranks/rank_rect_3.gif)
![](images/avatars/9744109354383ec3f0b7b6.png)
Joined: 26 Aug 2002 Posts: 500 Location: Kansas City
|
Posted: Wed Nov 26, 2003 11:43 pm Post subject: |
|
|
When I changed from 2.6-test10 to 2.6-test10-mm1 I had to edit /sbin/rc:
Code: |
try mount -n -t tmpfs tmpfs ${svcdir} \
-o rw,mode=0644,size=${svcsize}k; \
|
changes to
Code: |
try mount -n -t tmpfs tmpfs ${svcdir} \
-o rw,mode=0755,size=${svcsize}k; \
|
For some reason in test10 and every other kernel I've run except for test10-mm1 it doesn't matter if /mnt/.init.d is rw-r--r--. With test10-mm1 I need explicit execute permissions to access the directory as you would have thought.
Just thought I'd mention it as otherwise it can't run any boot up scripts at all. |
|
Back to top |
|
![](templates/gentoo/images/spacer.gif) |
Flo.Lue n00b
![n00b n00b](/images/ranks/rank_rect_0.gif)
Joined: 27 Dec 2002 Posts: 6 Location: Germany
|
Posted: Fri Nov 28, 2003 7:12 am Post subject: Re: 2.6 kernels hang during boot |
|
|
glennj wrote: |
[snip]
I set up a new empty runlevel and added the services one by one. I see the same thing as you in that I can get a login prompt without postfix, samba, and everything after that in the list above, but I can not login. If I also remove ntp-client and ntpd then I can log in. This seems to indicate that it is some sort of a network problem.
I have tried with two different network cards (different brands, drivers, etc.) but get the same result.
[snap]
|
Hi glennj, hi all
I'm writing this using my shiny new 2.6 kernel
I tried just like glennj to boot without any network related services running and I was able to successfully log in. After playing around with various services like samba and postfix, I restartet metalog and , tadaaa!, everything worked fine.
It was just pure luck that made me restart metalog because the problem seemed to be network related...
After unmerging metalog, emerging sysklogd and restoring my previous runlevel configuration I rebooted once more and as expected there were no errors
Is this a bug in metalog or in the kernel? Where should a bug like this be filed? |
|
Back to top |
|
![](templates/gentoo/images/spacer.gif) |
glennj n00b
![n00b n00b](/images/ranks/rank_rect_0.gif)
Joined: 23 Oct 2003 Posts: 18
|
Posted: Fri Nov 28, 2003 5:32 pm Post subject: Re: 2.6 kernels hang during boot |
|
|
Flo.Lue wrote: | glennj wrote: |
[snip]
I set up a new empty runlevel and added the services one by one. I see the same thing as you in that I can get a login prompt without postfix, samba, and everything after that in the list above, but I can not login. If I also remove ntp-client and ntpd then I can log in. This seems to indicate that it is some sort of a network problem.
I have tried with two different network cards (different brands, drivers, etc.) but get the same result.
[snap]
|
Hi glennj, hi all
I'm writing this using my shiny new 2.6 kernel
I tried just like glennj to boot without any network related services running and I was able to successfully log in. After playing around with various services like samba and postfix, I restartet metalog and , tadaaa!, everything worked fine.
It was just pure luck that made me restart metalog because the problem seemed to be network related...
After unmerging metalog, emerging sysklogd and restoring my previous runlevel configuration I rebooted once more and as expected there were no errors
Is this a bug in metalog or in the kernel? Where should a bug like this be filed? |
Wow, I never would have suspected metalog. After reading your message I thought about what might cause that. I had enabled console logging in metalog.conf so I commented out that section of the configuration file and booted the 2.6 kernel. It booted without any problem.
So, it looks like it is a problem with the console logging feature of metalog. I wonder if anyone using the console logging feature of syslog-ng is also having this boot up problem. Maybe it is not actually metalog but the enabling of console logging on 2.6 kernels.
Thanks. |
|
Back to top |
|
![](templates/gentoo/images/spacer.gif) |
Flo.Lue n00b
![n00b n00b](/images/ranks/rank_rect_0.gif)
Joined: 27 Dec 2002 Posts: 6 Location: Germany
|
Posted: Fri Nov 28, 2003 6:46 pm Post subject: Re: 2.6 kernels hang during boot |
|
|
glennj wrote: |
Wow, I never would have suspected metalog. After reading your message I thought about what might cause that. I had enabled console logging in metalog.conf so I commented out that section of the configuration file and booted the 2.6 kernel. It booted without any problem.
So, it looks like it is a problem with the console logging feature of metalog. I wonder if anyone using the console logging feature of syslog-ng is also having this boot up problem. Maybe it is not actually metalog but the enabling of console logging on 2.6 kernels.
|
The 'consolelog.sh' script just redirects the output from metalog to a vc. If I do something like Code: | echo "Hello" > /dev/vc/10 | it works just perfectly. Wheres is the difference between doing it manually and the way metalog does it? |
|
Back to top |
|
![](templates/gentoo/images/spacer.gif) |
glennj n00b
![n00b n00b](/images/ranks/rank_rect_0.gif)
Joined: 23 Oct 2003 Posts: 18
|
Posted: Fri Nov 28, 2003 7:12 pm Post subject: Re: 2.6 kernels hang during boot |
|
|
Flo.Lue wrote: | glennj wrote: |
Wow, I never would have suspected metalog. After reading your message I thought about what might cause that. I had enabled console logging in metalog.conf so I commented out that section of the configuration file and booted the 2.6 kernel. It booted without any problem.
So, it looks like it is a problem with the console logging feature of metalog. I wonder if anyone using the console logging feature of syslog-ng is also having this boot up problem. Maybe it is not actually metalog but the enabling of console logging on 2.6 kernels.
|
The 'consolelog.sh' script just redirects the output from metalog to a vc. If I do something like Code: | echo "Hello" > /dev/vc/10 | it works just perfectly. Wheres is the difference between doing it manually and the way metalog does it? |
I do not know but if I enable the console logging feature in metalog.conf, a 2.6 kernel will not boot. Does metalog lock the vc? Could this possibly be something with pts and devfs, which is different in 2.6 kernels from 2.4 kernels? |
|
Back to top |
|
![](templates/gentoo/images/spacer.gif) |
Flo.Lue n00b
![n00b n00b](/images/ranks/rank_rect_0.gif)
Joined: 27 Dec 2002 Posts: 6 Location: Germany
|
Posted: Fri Nov 28, 2003 7:35 pm Post subject: Re: 2.6 kernels hang during boot |
|
|
glennj wrote: |
I do not know but if I enable the console logging feature in metalog.conf, a 2.6 kernel will not boot. Does metalog lock the vc? Could this possibly be something with pts and devfs, which is different in 2.6 kernels from 2.4 kernels? |
I don't think that metalog is locking the vc. AFAIK if metalog finds a 'command' line for a logging facility it just runs the specified script with 'time facility message' as parameters. Metalog itself doesn't know what the script will be doing so it can't lock anything.
I just have no idea what could cause our problems but thx again for your help. Let me know if you find out anything new. |
|
Back to top |
|
![](templates/gentoo/images/spacer.gif) |
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
|