Bug 543644

Summary: IP6 tables not loaded when booting from kernel 2.6.31.6-145.f12.x86_64
Product: [Fedora] Fedora Reporter: Simon Lewis <simon.lewis>
Component: kernelAssignee: Kernel Maintainer List <kernel-maint>
Status: CLOSED DUPLICATE QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: urgent Docs Contact:
Priority: low    
Version: 12CC: dougsland, gansalmon, itamar, kernel-maint, nando, simon.lewis, zaitcev
Target Milestone: ---   
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2009-12-05 09:03:08 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
Boot log
none
Terminal output when running depmod -aev none

Description Simon Lewis 2009-12-02 19:11:12 UTC
Created attachment 375552 [details]
Boot log

Internet access over eth0 is not available when booting with kernel version 2.6.31.6-145.f12.x86_64

The following 2 errors were noted on booting..

1 modprobe fails

2 Failure to install IP6 tables.

Boot process continues and KDE desktop can be started but many operations are prevented due to missing access rights.

No obvious problems occur on the same installation when booting with kernel  version 22.6.31.5-127.fc12.x86_64.

Comment 1 Marcin 2009-12-03 20:53:52 UTC
Net do not work on ASUS P5W64 Pro mobo with skge/sky2 net driver with 2 GB RAM and NM disabled.

For me problems with iptables are the result of no network initialized on boot.

Errors of no eth0/eth1 device known during boot.

Net with kernel 2.6.31.5-127.fc12.x86_64 works.

Is it releated to Bug 447489 - [x86-64] No network with 4GB RAM support?

BTW. The same kernel-2.6.31.6-145.f12.i686 update on 32-bit IBM Thinkpad T41 effectively disables graphical startup. Only asci one works.

Sorry, but do not have immidiate access to the first mentioned machine. If any logs are required I can send them later. :-)

Comment 2 Simon Lewis 2009-12-03 21:49:24 UTC
Thanks to a tip from Fernando Lopez-Lezcano:

I have been able to determine that kernel installation did not create the "modules.dep", etc.. in the folder "/lib/modules/2.6.31.6-145.fc12.x86_64"

Please advise how to rectify this error?

Simon

Comment 3 Pete Zaitcev 2009-12-03 23:37:19 UTC
It's very easy to re-run depmod -ae, but the issue is, it should be run
automatically from %postin. If that didn't happen, something malfunctioned.

Comment 4 Simon Lewis 2009-12-04 07:07:39 UTC
For reference here is the post install script for the 2.6.31.6-145.fc12.x86_64 kernel:

[root@hp550-01 simon]# rpm -q --scripts kernel
postinstall scriptlet (using /bin/sh):

if [ `uname -i` == "x86_64" -o `uname -i` == "i386" ] &&
   [ -f /etc/sysconfig/kernel ]; then
  /bin/sed -r -i -e 's/^DEFAULTKERNEL=(kernel-smp|kernel-xen)$/DEFAULTKERNEL=kernel/' /etc/sysconfig/kernel || exit $?
fi

/sbin/new-kernel-pkg --package kernel --mkinitrd --dracut --depmod --install 2.6.31.6-145.fc12.x86_64 || exit $?
#if [ -x /sbin/weak-modules ]
#then
#    /sbin/weak-modules --add-kernel 2.6.31.6-145.fc12.x86_64 || exit $?
#fi
preuninstall scriptlet (using /bin/sh):
/sbin/new-kernel-pkg --rminitrd --rmmoddep --remove 2.6.31.6-145.fc12.x86_64 || exit $?
#if [ -x /sbin/weak-modules ]
#then
#    /sbin/weak-modules --remove-kernel 2.6.31.6-145.fc12.x86_64 || exit $?
#fi
posttrans scriptlet (using /bin/sh):
/sbin/new-kernel-pkg --package kernel --rpmposttrans 2.6.31.6-145.fc12.x86_64 || exit $?
[root@hp550-01 simon]#

Comment 5 Simon Lewis 2009-12-04 07:25:12 UTC
Created attachment 376013 [details]
Terminal output when running depmod -aev

Comment 6 Simon Lewis 2009-12-04 07:36:51 UTC
Hello Pete

As you will see in the attached terminal output, depmod only found the 2.6.31.5-127.fc12.x86_64 kernel.

rpm -q kernel confirms that both kernels are correctly installed:

[root@hp550-01 simon]# rpm -V kernel
S.5....T.    /boot/initramfs-2.6.31.5-127.fc12.x86_64.img
S.5....T.    /boot/initramfs-2.6.31.6-145.fc12.x86_64.img
[root@hp550-01 simon]#

What can be missing?

Simon

Comment 7 Fernando Lopez-Lezcano 2009-12-04 18:38:48 UTC
(In reply to comment #6)
> Hello Pete
> 
> As you will see in the attached terminal output, depmod only found the
> 2.6.31.5-127.fc12.x86_64 kernel.

By default depmod only finds the modules for the currently booted kernel, so if you are trying to fix other kernels and can't boot into them....

Well, I think you have to use additional command line options to depmod to tell it what to update. 

Probably something like:
  depmod -a /lib/modules/DIR_OF_KERNEL
(man depmod for more details)

Comment 8 Simon Lewis 2009-12-04 19:49:06 UTC
(In reply to comment #7)
> (In reply to comment #6)
> > Hello Pete
> > 
> > As you will see in the attached terminal output, depmod only found the
> > 2.6.31.5-127.fc12.x86_64 kernel.
> 
> By default depmod only finds the modules for the currently booted kernel, so if
> you are trying to fix other kernels and can't boot into them....
> 
> Well, I think you have to use additional command line options to depmod to tell
> it what to update. 
> 
> Probably something like:
>   depmod -a /lib/modules/DIR_OF_KERNEL
> (man depmod for more details)  

Hello Fernando

I think I need to use depmod's -F switch but I am not sure - It would be nice if Redhat's kernel maintainers would give exact instructions on how repair the faulty installations...

The question also arises why did the rpm post script fail to install the module dependences for both the 2.6.31.5-127.fc12.x86_64 kernel in Fedora's upgrade repo and the 2.6.31.6-1.rt19.2.fc12.ccrma.x86_64.rt kernel in ccrma's repo.

Simon

Comment 9 Simon Lewis 2009-12-05 09:03:08 UTC
Kernel can be installed correctly using yum from the command line - yumex has problems running scripts in rpm packages..

*** This bug has been marked as a duplicate of bug 527528 ***