Bug 71513

Summary:	e2fsck endless loop
Product:	Gentoo Linux	Reporter:	YbbY <ybby12>
Component:	[OLD] baselayout	Assignee:	Gentoo's Team for Core System packages <base-system>
Status:	VERIFIED INVALID
Severity:	enhancement
Priority:	High
Version:	unspecified
Hardware:	All
OS:	Linux
Whiteboard:
Package list:		Runtime testing required:	---

Description YbbY 2004-11-17 04:44:38 UTC

i was trying to setup RAID1 on my system, but i made mistakes, i forgot to set partition type to fd and i changed the type with fdisk. not sure if it is linked but since then , i'm desperatly trying to get back my hdd working, even without RAID.

the actual situation is that e2fsck enters an endless loop :

# e2fsck /dev/hdf1
e2fsck 1.35 (28-Feb-2004)
Group descriptors look bad... trying backup blocks...
Inode table for group 2 is not in group. (block 8454146)
WARNING: SEVERE DATA LOSS POSSIBLE.
Relocate<y>? yes

Inode table for group 5 is not in group. (block 8552452)
WARNING: SEVERE DATA LOSS POSSIBLE.
Relocate<y>? yes

/dev/hdf1 was not cleanly unmounted, check forced.
Pass 1: Checking inodes, blocks, and sizes
Relocating group 2's inode table to 65538...
Relocating group 5's inode table to 163844...
Restarting e2fsck from the beginning...
Group descriptors look bad... trying backup blocks...
Inode table for group 2 is not in group. (block 8454146)
WARNING: SEVERE DATA LOSS POSSIBLE.
Relocate<y>? yes

Inode table for group 5 is not in group. (block 8552452)
WARNING: SEVERE DATA LOSS POSSIBLE.
Relocate<y>? yes

Inode table for group 5 is not in group. (block 8552452)
WARNING: SEVERE DATA LOSS POSSIBLE.
Relocate<y>?

not sure, but it looks like it can be considered as bug as nothing is repaired and if fsck -y is launched it can never ends.

thx,

YbbY

Comment 1 SpanKY gentoo-dev

2004-11-17 06:10:20 UTC

please report this at http://e2fsprogs.sf.net/ ... it's your machine and will be able to give additional information about it when requested

the way you describe the situation, you dont have any real data on /dev/hdf do you ?  why not just run `mke2fs` on it again ?

Comment 2 YbbY 2004-11-17 06:38:29 UTC

thx, 

yes, there is no data on hdf yet but i already runned mke2fs on it many times, actually, fsck fail on repairing hdf1 just after having the fs created !

my plan was to run a raid device made of hde + hdf,  but quickly had the problem on hde+hdf :-(

my experiments led me to try to run "dd if=/dev/zero of=/dev/hde" , after that the problem disapeared on hde. but it changed the geometry of hde from "2 heads, 4 sectors/track, 29305206 cylinders" to "255 heads, 63 sectors/track, 14593 cylinders"! 

considering that all this could be due to the default geometry of my seagate disk, right now, i hesitate between keeping hdf unusable and write a bug report to e2fsprogs, or wipe out hdf with "dd" and forget about this...

Regards, YbbY

Comment 3 YbbY 2004-11-17 08:11:05 UTC

probably disk geometry pb, dd if=/dev/zero of=/dev/hdf in progress