FAQ Search Today's Posts Mark Forums Read
» Video Reviews

» Linux Archive

Linux-archive is a website aiming to archive linux email lists and to make them easily accessible for linux users/developers.


» Sponsor

» Partners

» Sponsor

Go Back   Linux Archive > Debian > Debian Kernel

 
 
LinkBack Thread Tools
 
Old 09-03-2011, 05:39 AM
Jonathan Nieder
 
Default Bug#611493: Failure on boot during initrd script execution

retitle 611493 Failure on boot during initrd script execution
reassign 611493 initramfs-tools 0.98.7
merge 616689 611493
affects 616689 + linux-2.6 src:linux-2.6
quit

Pascal BERNARD wrote:

> By adding rootdelay=20 on the command line, the problem does not show.
> It is not very satisfying though, since the casual user will not dive
> into the initrd to read the script init and see that you can set this
> variable on the command line.

Sounds like <http://bugs.debian.org/616689>. Merging but please feel
free to unmerge if I have missed something.



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20110903053933.GA1892@elie">http://lists.debian.org/20110903053933.GA1892@elie
 
Old 09-03-2011, 11:31 AM
Pascal BERNARD
 
Default Bug#611493: Failure on boot during initrd script execution

616689 is probably part of the problem. In my case, LVM runs on top of a
raid0. Sometimes, the kernel freezes several minutes after an apparently
normal boot. I have tried to improve the situation with the rootdelay
and also timeout from grub. The rootdelay parameter helped, but the
problem re-happened, unfortunately. I wonder if there is not some
interference with grub.

As I am running two different systems (Mint not Window$ !) on the same
computer, I could notice that:
- the problem is less likely to happen if I reboot (the machine is warm)
- the problem does not happen with Mint, kernel 2.6.24-28-generic, with
boot&root partition is a native partition and home is on a raid1/LVM
partition.

I am sorry, but I do not know how I could be more helpful.

Le samedi 03 septembre 2011 à 00:39 -0500, Jonathan Nieder a écrit :
> retitle 611493 Failure on boot during initrd script execution
> reassign 611493 initramfs-tools 0.98.7
> merge 616689 611493
> affects 616689 + linux-2.6 src:linux-2.6
> quit
>
> Pascal BERNARD wrote:
>
> > By adding rootdelay=20 on the command line, the problem does not show.
> > It is not very satisfying though, since the casual user will not dive
> > into the initrd to read the script init and see that you can set this
> > variable on the command line.
>
> Sounds like <http://bugs.debian.org/616689>. Merging but please feel
> free to unmerge if I have missed something.





--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 1315049504.2917.31.camel@moraes.localdomain">http://lists.debian.org/1315049504.2917.31.camel@moraes.localdomain
 
Old 09-03-2011, 12:37 PM
Jonathan Nieder
 
Default Bug#611493: Failure on boot during initrd script execution

Hi Pascal,

Pascal BERNARD wrote:

> In my case, LVM runs on top of a
[... lots of nice details snipped ...]
> I am sorry, but I do not know how I could be more helpful.

No, that looks very helpful!

What needs to happen next is probably something like what is described
at bugs.debian.org/616689:

| 3. Come up with some way of asynchronous notification of volume groups
| being available for activation, and invoke vgchange on this
| notification. I clearly don't know about how LVM works internally to
| accomplish something like this, an udev hook, perhaps?

Meanwhile, some documentation or other mechanism to help people find
scsi_mod.scan=sync and rootdelay=20 would be helpful as a stopgap, if
you have ideas for where it should go.

Thanks again.
Jonathan



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20110903123717.GA9494@elie">http://lists.debian.org/20110903123717.GA9494@elie
 
Old 09-03-2011, 06:07 PM
maximilian attems
 
Default Bug#611493: Failure on boot during initrd script execution

On Sat, Sep 03, 2011 at 07:37:17AM -0500, Jonathan Nieder wrote:
>
> | 3. Come up with some way of asynchronous notification of volume groups
> | being available for activation, and invoke vgchange on this
> | notification. I clearly don't know about how LVM works internally to
> | accomplish something like this, an udev hook, perhaps?

this can be done via udev rules.
Ubuntu does so, but it *hardcodes* udev into initramfs and has
it's own races that are no fun either.

> Meanwhile, some documentation or other mechanism to help people find
> scsi_mod.scan=sync and rootdelay=20 would be helpful as a stopgap, if
> you have ideas for where it should go.

documented here:
http://wiki.debian.org/InitramfsDebug

thanks.

--
maks



--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20110903180701.GA29640@vostochny.stro.at">http://lists.debian.org/20110903180701.GA29640@vostochny.stro.at
 
Old 09-03-2011, 08:07 PM
Pascal BERNARD
 
Default Bug#611493: Failure on boot during initrd script execution

We have the linux-docs-2.6.32 package which could hold the
kernel-parameters.txt file (part of linux kernel sources). But this file
is quite big and if the user has to adjust parameters for his own
installation, these parameters could be highlighted in some Debian file
of the package. Scsi_mod.scan does not appear in this file, maybe it is
for use by the initramfs scripts ? It looks like the respective roles of
the loader, the kernel, the initrd, and the init scripts have to be
clarified. At least, it is quite obscure to me still !

I will try things from http://wiki.debian.org/InitramfsDebug if I get
the chance to end up with the initrd sheel. But this is not always the
case.

Thanks,

Pascal




--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 1315080430.3677.206.camel@moraes.localdomain">http ://lists.debian.org/1315080430.3677.206.camel@moraes.localdomain
 
Old 09-03-2011, 09:36 PM
Ben Hutchings
 
Default Bug#611493: Failure on boot during initrd script execution

On Sat, 2011-09-03 at 22:07 +0200, Pascal BERNARD wrote:
> We have the linux-docs-2.6.32 package which could hold the
> kernel-parameters.txt file (part of linux kernel sources). But this file
> is quite big and if the user has to adjust parameters for his own
> installation, these parameters could be highlighted in some Debian file
> of the package. Scsi_mod.scan does not appear in this file, maybe it is
> for use by the initramfs scripts ? It looks like the respective roles of
> the loader, the kernel, the initrd, and the init scripts have to be
> clarified. At least, it is quite obscure to me still !
[...]

The kernel ignores any unrecognised any unrecognised arguments on the
command line, except that if they do follow the format name=value they
are passed as arguments to the init process (i.e. the initramfs).

*Any* process may read and interpret the kernel command line as it
wishes, so the set of possible parameters is effectively infinite.

Ben.
 
Old 09-03-2011, 11:37 PM
Ben Hutchings
 
Default Bug#611493: Failure on boot during initrd script execution

Let me try that again:

The kernel ignores any unrecognised arguments on the command line,
except that if they do not follow the format name=value they are
passed as arguments to the init process (i.e. the initramfs).

Ben.
 

Thread Tools




All times are GMT. The time now is 08:24 AM.

VBulletin, Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO ©2007, Crawlability, Inc.
Copyright ©2007 - 2008, www.linux-archive.org