corrupted kernel module files with 2.6.18-3 on ext3
I've run into this weird filesystem and/or kernel problem where my
kernel module files get corrupted (they live on my root partition).
I shutdown my box every day with "shutdown -h now". Every now and then,
when I restart, some modules (e.g., ALSA) won't have loaded. I do an "ls
-l" in the module directory, and there are question marks ("?") instead
of file size, instead of some file names, etc.
This is my root partition. To fix it (temporarily) I boot into rescue
mode, shutdown networking and various daemons, unmount other
filesystems, and issue "mount -o ro,remount /" to remount / readonly. I
run fsck on it and it fixes the problems, though sometimes the modules
are deleted and I have to reinstall the kernel package.
I've run a badblocks test and smart tests on my drive and it appears
fine, so I think the problem must be a kernel module or filesystem bug.
Has anyone else noticed this? Should I just file a debian bug and
forget about it or should I report it to the ext3 maintainers or...?
Hardware: Dual Opteron 244 on MSI K8T Master2-FAR (Via K8T8000 chipset).
2GB RAM.
Kernel: linux-image-2.6.18-3-amd64 (Debian unstable version)
# /etc/fstab: static file system information.
#
# <file system> <mount point> <type> <options> <dump> <pass>
/dev/hda1 /boot ext3 defaults 0 2
/dev/hda2 /x86 ext3 defaults 0 2
/dev/hda3 / ext3 defaults,errors=remount-ro 0 1
/dev/hda4 /home ext3 defaults 0 2
/dev/hdc /ls120 ext2 defaults,user 0 0
/dev/hdc /floppy vfat defaults,user 0 0
/dev/cdrom /cdrom iso9660 user,ro 0 0
proc /proc proc defaults 0 0
sysfs /sys sysfs none 0 0
devpts /dev/pts devpts none 0 0
tmpfs /dev/shm tmpfs defaults 0 0
tmpfs /tmp tmpfs defaults 0 0
# x86 mounts
/home /x86/home none bind 0 0
/tmp /x86/tmp none bind 0 0
/dev /x86/dev none rbind 0 0
/sys /x86/sys none bind 0 0
proc /x86/proc proc defaults 0 0
Thanks for any help, suggestions, etc.
-s
Reply to: