Bug 66616 - Kernel panic in ext3
Summary: Kernel panic in ext3
Keywords:
Status: CLOSED ERRATA
Alias: None
Product: Red Hat Linux
Classification: Retired
Component: kernel
Version: 7.3
Hardware: i686
OS: Linux
medium
medium
Target Milestone: ---
Assignee: Stephen Tweedie
QA Contact: Brian Brock
URL:
Whiteboard:
Depends On: 67177
Blocks:
TreeView+ depends on / blocked
 
Reported: 2002-06-12 17:25 UTC by Chris Lange
Modified: 2007-04-18 16:43 UTC (History)
0 users

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2002-08-13 09:47:38 UTC
Embargoed:


Attachments (Terms of Use)

Description Chris Lange 2002-06-12 17:25:26 UTC
From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.9) Gecko/20020513

Description of problem:
Experienced kernel panic.
This happened when a process waas copying several thousand 40KB files to a
single directory on an ext3 partition.

Version-Release number of selected component (if applicable):


How reproducible:
Couldn't Reproduce

Steps to Reproduce:
1.Executed the process that was running when the original panic occured.
2.
3.
	

Actual Results:  No issues.

Expected Results:  No issues.

Additional info:

###Here is the error from /var/log/message

Jun 11 22:39:12 Flanders kernel: Unable to handle kernel NULL pointer
dereference at virtual address 00000008
Jun 11 22:39:12 Flanders kernel:  printing eip:
Jun 11 22:39:12 Flanders kernel: e0860d0c
Jun 11 22:39:12 Flanders kernel: *pde = 00000000
Jun 11 22:39:12 Flanders kernel: Oops: 0000
Jun 11 22:39:12 Flanders kernel: ncpfs ipx autofs loop nfs lockd sunrpc
nls_iso8859-1 nls_cp437 vfat fat sr_mod
Jun 11 22:39:12 Flanders kernel: CPU:    2
Jun 11 22:39:12 Flanders kernel: EIP:    0010:[<e0860d0c>]    Not tainted
Jun 11 22:39:12 Flanders kernel: EFLAGS: 00010202
Jun 11 22:39:12 Flanders kernel:
Jun 11 22:39:12 Flanders kernel: EIP is at ext3_new_block [ext3] 0x9c (2.4.18-4smp)
Jun 11 22:39:12 Flanders kernel: eax: 00000000   ebx: 00001000   ecx: 0000000c 
 edx: cb00c280
Jun 11 22:39:12 Flanders kernel: esi: 00000000   edi: df969000   ebp: cde48000 
 esp: cde49c8c
Jun 11 22:39:12 Flanders automount[10936]: attempting to mount entry
/opt/siarchive/pn2/servers/sipn222
Jun 11 22:39:13 Flanders kernel: ds: 0018   es: 0018   ss: 0018
Jun 11 22:39:13 Flanders kernel: Process cp (pid: 12680, stackpage=cde49000)
Jun 11 22:39:13 Flanders kernel: Stack: cb00c280 00001000 00000000 00000000
c9b195c0 cde49d78 c0e5c384 c020f21e
Jun 11 22:39:13 Flanders kernel:        cde49d78 00000000 00000000 c0e5c374
c0e5c370 c0e5c374 c0e5c370 c9b195c
0
Jun 11 22:39:13 Flanders kernel:        c01f4997 c9b195c0 00000040 00000000
e0b81000 c5e5b420 0000040a 0000040
a
Jun 11 22:39:13 Flanders kernel: Call Trace: [<c020f21e>] udp_getfrag [kernel] 0x4e
Jun 11 22:39:13 Flanders kernel: [<c01f4997>] ip_build_xmit [kernel] 0x2e7
Jun 11 22:39:13 Flanders kernel: [<c011a4f8>] ll_copy_to_user [kernel] 0x38
Jun 11 22:39:13 Flanders kernel: [<e086355d>] ext3_alloc_block [ext3] 0x1d
Jun 11 22:39:13 Flanders kernel: [<e0863869>] ext3_alloc_branch [ext3] 0x49
Jun 11 22:39:13 Flanders kernel: [<c020f8db>] udp_recvmsg [kernel] 0xcb
Jun 11 22:39:13 Flanders kernel: [<c020f997>] udp_recvmsg [kernel] 0x187
Jun 11 22:39:13 Flanders kernel: [<e0863ea7>] ext3_get_block_handle [ext3] 0x217
Jun 11 22:39:13 Flanders kernel: [<c01445f2>] create_buffers [kernel] 0x62
Jun 11 22:39:13 Flanders kernel: [<e0866341>] ext3_do_update_inode [ext3] 0x321
Jun 11 22:39:13 Flanders kernel: [<e0863fd7>] ext3_get_block [ext3] 0x67
Jun 11 22:39:13 Flanders kernel: [<c0144c0f>] __block_prepare_write [kernel] 0x12f
Jun 11 22:39:13 Flanders kernel: [<e085acd0>] .rodata.str1.1 [jbd] 0x30
Jun 11 22:39:13 Flanders kernel: [<e08580c7>] __jbd_kmalloc [jbd] 0x27
Jun 11 22:39:13 Flanders kernel: [<c01455a2>] block_prepare_write [kernel] 0x22
Jun 11 22:39:13 Flanders kernel: [<e0863f70>] ext3_get_block [ext3] 0x0
Jun 11 22:39:13 Flanders kernel: [<e0864512>] ext3_prepare_write [ext3] 0xe2
Jun 11 22:39:13 Flanders kernel: [<e0863f70>] ext3_get_block [ext3] 0x0
Jun 11 22:39:13 Flanders kernel: [<c0139dc8>] __alloc_pages [kernel] 0xa8
Jun 11 22:39:13 Flanders kernel: [<c013285d>] generic_file_write [kernel] 0x4fd
Jun 11 22:39:13 Flanders kernel: [<c0141bf6>] sys_write [kernel] 0x96
Jun 11 22:39:13 Flanders kernel: [<c0108c6b>] system_call [kernel] 0x33
Jun 11 22:39:13 Flanders kernel:
Jun 11 22:39:13 Flanders kernel:
Jun 11 22:39:13 Flanders kernel: Code: ff 50 08 83 c4 10 48 75 70 8b 45 1c 85 c0
79 13 6a 3e 68 e0

Comment 1 Chris Lange 2002-06-12 17:32:05 UTC
Kernel 2.4.18-4smp

Comment 2 Need Real Name 2002-07-29 16:18:35 UTC
have you try the kernel 2.4.18-5smp?

Comment 3 Chris Lange 2002-07-30 09:43:54 UTC
No, I have not tried the kernel 2.4.18-5smp.  I have not been able to reproduce 
the panic under kernel 2.4.18-4smp.

Comment 4 Stephen Tweedie 2002-07-31 13:39:44 UTC
We've seen this internally, several times, but only on one system and never
under normal testing stress loads.  What happens is that one of the filesystem's
data structures becomes corrupt in a very predictable manner: the filesystem
seems to get marked as having quotas, but there is no quota operations vector in
the superblock and the filesystem OOPSes when it tries to indirect through the
quota vector.

It happens in several different places, and we have seen it affect both ext3 and
NFS filesystems so we have, as yet, no evidence to narrow it down to any one
filesystem being at fault.

It would really help if you could give us as much background information on your
system as you can: what workloads it serves, whether it uses software raid
and/or LVM, what filesystems and device drivers it is using actively.

Thanks.

Comment 5 Colin Tinker 2002-08-12 08:11:00 UTC
I have had this on kernel 2.4.5 and it is reproducable by moving large files 
around the hard disk ie cdrom iso images.  I have an Abit KR7A mb with an IBM 
deskstar 80gb ata 100 hdd.  It makes no difference if I optimize with hdparm 
or not it will still happen.  It does not happen with kernel 2.4.18-3 and I 
never saw as far as I can remember with 2.4.18-4, it does seem to be related 
to a kernel or other upgrade.  Sometimes the system will be fine for 3 or 4 
days sometimes it will happen frequently every few minutes.  It does happen a 
lot when updatedb runs as a cron job.  I have 896 mb ram and a 1.333gb amd 
athlon cpu.  Video is NVidia, net card is intel eepro100, I have a wintv card 
initio scsi card with scsi cdwriter (HP 9200) ide cdrom.  The NVidia is on its 
own irq as are both ide channels.

Comment 6 Stephen Tweedie 2002-11-11 22:01:33 UTC
We can't reproduce this any more internally with the 2.4.18-10 errata kernel. 
Please reopen if you still have a problem with this version.


Note You need to log in before you can comment on or make changes to this bug.