lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <46A5D619.5030600@imap.cc>
Date:	Tue, 24 Jul 2007 12:36:09 +0200
From:	Tilman Schmidt <tilman@...p.cc>
To:	Olaf Hering <olh@...e.de>
CC:	Jeremy Fitzhardinge <jeremy@...p.org>,
	LKML <linux-kernel@...r.kernel.org>, nix.or.die@...glemail.com,
	andi@...stfloor.org
Subject: Re: 2.6.22-git17 boot failure

Olaf Hering schrieb:
> On Tue, Jul 24, Andi Kleen wrote:
> 
>> > Andi, any thoughts about how to debug the suse boot sequence?
> 
> The SuSE boot sequence is rather long. Which part (and which version) are we talking about?

Quick summary of the problem:

Opensuse 10.2 32-bit install on a Pentium D 940
Kernel 2.6.22-git17 or later (last reproduced with 2.6.23-rc1)
A kernel built with CONFIG_XEN=y fails to detect the SATA disks during
boot, so the root filesystem cannot be mounted and I'm thrown into
the emergency /bin/sh. With CONFIG_XEN=n everything works fine.

Netconsole logs with initcall_debug=1 of a successful and a
failing startup and corresponding .config-s are available at
http://www.phnxsoft.com/~ts/linux/

> As Andi said, putting a 'set -x' into /init script generated by /sbin/mkinitrd
> is one thing.

I'll try. I guess I'll have to unpack/repack the finished initrd
image (gzipped cpio archive, AFAICS) for that - or is there a way
to have mkinitrd inject that statement for me?

> You can also boot with 'linuxrc=trace' to generate a huge
> amount of udev and /init debug.

That looks simple enough, so I'll try that first, although it looks
to me like the trouble happens before udev comes into play.

> Booting with 'confirm' will allow you to step each runlevel script, see
> /etc/init.d/boot, look for DO_CONFIRM=.

Thanks,
Tilman

-- 
Tilman Schmidt                    E-Mail: tilman@...p.cc
Bonn, Germany
Diese Nachricht besteht zu 100% aus wiederverwerteten Bits.
Ungeöffnet mindestens haltbar bis: (siehe Rückseite)


Download attachment "signature.asc" of type "application/pgp-signature" (251 bytes)

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ