EXT2-fs warning: bit already cleared for inode XXXXX ?

EXT2-fs warning: bit already cleared for inode XXXXX ?

Post by Aarron Shaughnes » Wed, 26 Aug 1998 04:00:00



I have just moved one of my Slakware systems from a P100 into a spare
486dx4. I assumed that, unlike WIN95, a Linux setup would run on whatever
motherboard configuration I threw at it.

Two days after the move the following started to appear in the logs
(note: shifting inodes, always the root disk only).....

warning:Aug 24 15:33:47 java kernel: EXT2-fs warning (device 03:02):
ext2_free_inode: bit already cleared for inode 150156

A run of ext2fs said it cleared the problem, but they came back (different
inodes), then came a flood, then came a crash.

Its a modern motherboard with onboard IDE, LBA, 32 bit mode, PNP bios  blah.
The disk setup hasnt changed from the P100; 1Gb Fireball root disk, 6Gb
Seagate UDMA data disk.  

I formatted the drive as dos, and checked the drive for bad sectors but
nothing showed up. I reinstalled linux, new partitions, clean format, fresh
install from CD. All systems came up OK, but the errors started again. I took
the PC apart and checked the master slave jumpers, and changed the cables,
still no joy.

I have now been running a dying system for about a month.  /etc/ld.so.cache
corrupts are common.

What could be causing these errors ? (exactly what are these errors ?)

Please save me from my growing log files.

Aarron

--
Aarron Shaughnessy
ADG Computer Systems Ltd
Tel:   +44 (0)1865 792702
Fax:   +44 (0)1865 792704

 
 
 

EXT2-fs warning: bit already cleared for inode XXXXX ?

Post by Graffit » Wed, 26 Aug 1998 04:00:00



[snip]

Quote:>warning:Aug 24 15:33:47 java kernel: EXT2-fs warning (device 03:02):
>ext2_free_inode: bit already cleared for inode 150156
[snip]
>I formatted the drive as dos, and checked the drive for bad sectors but
>nothing showed up. I reinstalled linux, new partitions, clean format, fresh
>install from CD. All systems came up OK, but the errors started again. I took
>the PC apart and checked the master slave jumpers, and changed the cables,
>still no joy.

>I have now been running a dying system for about a month.  /etc/ld.so.cache
>corrupts are common.

>What could be causing these errors ? (exactly what are these errors ?)

>Please save me from my growing log files.

Try checking the cable lengths.  If the cables aren't too short/long, try moving
the hard drives from primary<->secondary, and boot off a floppy instead.  If
the errors still pop up, try running hdparam and start disabling various stuff.
If the errors still pop up, chances are, you have flaky hardware somewhere.

Is the Fireball your root disk?  Or the UDMA seagate?

Also, try upgrading to the latest kernel.  If you're running 2.0.x, 2.0.35 (or
2.0.36, dunno.)  If you're running a developmental kernel, try asking for help
on linux-kernel.  The address is in the source tree's README file.

And on the off chance it's the controller, try changing that, since this doesn't
look like a kernel problem to me.  More likely hardware (might be memory, who
knows?)

-- DN (Guessing wildly)

 
 
 

1. ext2_free_inode: bit already cleared for inode 191809

Anybody have a clue what this is and what to do about it?

Apr 30 04:24:43 localhost kernel: EXT2-fs warning (device ide0(3,4)):
ext2_free_inode: bit already cleared for inode 191778

Apr 30 04:24:43 localhost kernel: EXT2-fs warning (device ide0(3,4)):
ext2_free_blocks: bit already cleared for block 811296

Apr 30 04:24:43 localhost kernel: EXT2-fs warning (device ide0(3,4)):
ext2_free_inode: bit already cleared for inode 191782

Apr 30 04:24:43 localhost kernel: EXT2-fs warning (device ide0(3,4)):
ext2_free_blocks: bit already cleared for block 811291

Apr 30 04:24:43 localhost kernel: EXT2-fs warning (device ide0(3,4)):
ext2_free_inode: bit already cleared for inode 191777

I have a weekly cron job that copies my root partition to a backup
partition (both on IDE0/hda). I get a slew of these with each run.
Easily reproducible. I first got these last weekend. I ran e2fsck, which
found and fixed some errors. But evidently not permaneantly.

I aslo I am having trouble umounting my root partion for maintenance.
Usually:

 #init 1
 #umount -a
 #mount -n -o remount,ro /

works, but today all I get is 'mount: / device busy'. All other
partitions umount cleanly. Tried it a zillion times. Extremely few
processes actually running (10?). I occasionally get this for various
partitions on shutdown too. Seems to come and go, but now I'd like it to
just go. This is updated RH6.2 with 2.2.15pre9 kernel.

I know I can do emergency umount with SySRq keys (Alt-Printscreen-u). I
feel safe doing this for emergency shutdown purposes, but am not
comfortable using it in this situation.

TIA

--
Hal B

--

2. mail and virtual domains

3. fs/inode.c sync fix and fs/ext2/inode.c tidy

4. Debian woody -> apt-get probleme (cannot find answer else where)

5. EXT2-FS warning for nonexistant FS?

6. XFree86-3.3.3 on 3.0-RELEASE

7. 2.0.31-pre7 et al: EXT2-fs panic bad inode number: 0

8. What happens if Microsoft only gets.......

9. PROBLEM: ext2-fs panic (inode=2, block=16777477)

10. inode crashed root partition/EXT2-fs panic

11. fs: Unattached inode xxxxx

12. two questions SCSI UW and "EXT2-fs warning (device 08:02):"

13. EXT2-fs warning