Hi, > OK, so this is a very simple case and it's very surprising that it > fails. > > If you reboot back to the old kernel version, does that still work? > > Can you capture the kernel log from a failed boot? If you add > 'break=bottom' to the kernel command line, you should get to a shell in > the initramfs. You can then mount a USB flash drive or similar, and > copy the boot messages to it, e.g.: > > (initramfs) ls -l /dev/sda2 # did the SATA drive appear
This case could have a hardware (harddisk problem ?) component. I have noticed Grub to issue error messages before loading the boot option screen in some of the trial boots. The message was: Grub: error: hd0 out of disk. In one case, it even entered rescue mode. In other cases, boot screen appeared. In ~2/4 cases no problem happened despite the warning by Grub. In other 2 cases kernel reported at the start of boot: loading initramfs..... No such file or directory. Then it starts loading kernel. In the next screen it panics as reported before. I did try your trick 'break=bottom' in kernel boot option line. However, this works when kernel has mostly booted, and it gives me a busybox shell. However, when kernel panic happened it happened much earlier, and no commands would work (including ALT-Sysrq-B to reboot). I could cause the old kernel to also panic in one such occasion. Hope these give some clues. Regards, Subhashis On Wed, May 1, 2013, at 10:47 PM, Subhashis Roy wrote: > > > ----- Original message ----- > From: Ben Hutchings <b...@decadent.org.uk> > To: Subhashis Roy <subhash...@fastmail.fm> > Cc: 706...@bugs.debian.org > Subject: Re: Bug#706355: linux-image-3.2.0-4-amd64: Kernel Panic with > the new kernel during booting > Date: Tue, 30 Apr 2013 14:35:25 +0100 > > On Mon, 2013-04-29 at 13:57 +0500, Subhashis Roy wrote: > > Hi, > > > > > This panic is expected if the root filesystem could not be mounted. So > > > the question is, why did that fail? Is the root device a simple > > > partition or logical volume? Is the physical device attached by SATA, > > > USB, or other means? > > > > The root device is a simple partition '/dev/sda2' (the details of it was > > already part of the mail as captured by > > 'reportbug' when it booted from the kernel on which I flled the bug > > report). > > The physical volume is attached by SATA. > > OK, so this is a very simple case and it's very surprising that it > fails. > > If you reboot back to the old kernel version, does that still work? > > Can you capture the kernel log from a failed boot? If you add > 'break=bottom' to the kernel command line, you should get to a shell in > the initramfs. You can then mount a USB flash drive or similar, and > copy the boot messages to it, e.g.: > > (initramfs) ls -l /dev/sda2 # did the SATA drive appear? > ls: /dev/sda2: No such file or directory > (initramfs) mkdir /mnt > (initramfs) mount /dev/sda1 /mnt # removable drive should be sda > (initramfs) dmesg > /mnt/dmesg > (initramfs) umount /mnt > (initramfs) reboot -f > > Ben. > > -- > Ben Hutchings > Knowledge is power. France is bacon. > > > -- > -- > > -- > http://www.fastmail.fm - Send your email first class > > Email had 1 attachment: > + signature.asc > 1k (application/pgp-signature) -- -- -- http://www.fastmail.fm - Faster than the air-speed velocity of an unladen european swallow -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
