commit 0fc9d10 : chroot under kernel 3.4.x core dumps

From: Toralf Förster
Date: Tue Jun 05 2012 - 06:44:00 EST



When I try under kernel versions 3.4.[01] to chroot into a user mode
linux image, I often get this :

+ mount -o loop /home/tfoerste/virtual/uml/gentoo_unstable /mnt/loop
+ mount -t proc none /mnt/loop/proc
+ mount -o bind /dev /mnt/loop/dev
+ mount -o bind /dev/pts /mnt/loop/dev/pts
+ mount -o bind /usr/portage /mnt/loop/usr/portage
+ mount -o bind /usr/local/portage /mnt/loop/usr/local/portage
+ mount -o bind /tmp/uml_portage /mnt/loop/var/tmp/portage
+ chroot /mnt/loop /bin/bash
/home/tfoerste/workspace/bin/chr_uml.sh: line 55: 4143 Segmentation
fault (core dumped) chroot /mnt/loop /bin/bash


The core file itself is only 400 KB in size and the back trace with gbd
is meaningless :

Core was generated by `/bin/bash'.
Program terminated with signal 11, Segmentation fault.
#0 0xb75cac00 in ?? ()
(gdb) bt
#0 0xb75cac00 in ?? ()
#1 0xb76f4ff4 in ?? ()
Cannot access memory at address 0xffffffdc

In the rare cases, where the chroot doesn't crash, all subsequent
commands cores instead. Furthermore it seems, that the boinc grid
software has problems too under kernel 3.4.1.
All kernel 3.3.x works fine.

I bisected it to :

commit 0fc9d1040313047edf6a39fd4d7c7defdca97c62
Author: Konstantin Khlebnikov <khlebnikov@xxxxxxxxxx>
Date: Wed Mar 28 14:42:54 2012 -0700

radix-tree: use iterators in find_get_pages* functions


--
MfG/Sincerely
Toralf Förster
pgp finger print: 7B1A 07F4 EC82 0F90 D4C2 8936 872A E508 7DB6 9DA3
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/