"Matt D. Robinson" <yakker@xxxxxxxxxxxxxx> wrote:
> Guy Edjlali wrote:
>>
>> Hello,
>>
>> I was trying to run a 2.4.0 kernel with lkcd and kdb.
>> Has anyone tryed this? The stack for the process that
>> crashed the kernel (crashnow) is empty.
> A couple of questions:
> - Did the system die due to a panic() or Oops?
> - Did you modify any of the incoming registers with kdb?
> - Can we get a look at the crash dump to validate it?
i did the same here and was wondering about the missing stack traces
too ... i always used a simple module calling panic() for testing
but yesterday i had a real crash and was able to get a stacktrace
- maybe the reason is the "artificial" nature of the crashing
method with a panic module?
ok - but the reason why i write this here is that i also had another
problem: the system crashed, wrote the crashdump but on trying to
analyse it on bootup it hung on the process table which you can
see very good in the analysis output from that bootup:
c1790000 0 2635 1 1 0 sa1
c20e2000 0 2637 2635 1 0 sadc
c22d6000 0 2640 1344 1 0 sh
c29d6000 26 2655 1 0 40 postmaster
c22d6000 0 2640 1344 1 0 sh
c29d6000 26 2655 1 0 40 postmaster
c22d6000 0 2640 1344 1 0 sh
c29d6000 26 2655 1 0 40 postmaster
c22d6000 0 2640 1344 1 0 sh
...
any idea what happened here? ... maybe some more relevant output
from the analysis:
================
COREFILE SUMMARY
================
The system died due to a software failure.
===================
UTSNAME INFORMATION
===================
sysname : Linux
nodename : oxygen.bln.innominate.de
release : 2.4.1-XFS
version : #4 SMP Sat Feb 17 11:13:08 CET 2001
machine : i586
domainname :
===============
LOG BUFFER DUMP
===============
...
<2>EXT2-fs error (device ide0(3,6)): ext2_free_blocks: Freeing blocks not in
datazone - block = 85861, count = 1
<2>EXT2-fs error (device ide0(3,6)): free_inode: reserved inode or
nonexistent inode
<4>kernel BUG at inode.c:911!
<4>invalid operand: 0000
<4>CPU: 0
<4>EIP: 0010:[<c014c44f>]
<4>EFLAGS: 00210286
<4>eax: 0000001b ebx: c36d74a0 ecx: 00000001 edx: 00000001
<4>esi: c040d5e0 edi: c200fda0 ebp: c29d7f50 esp: c29d7f3c
<4>ds: 0018 es: 0018 ss: 0018
<4>Process postmaster (pid: 2655, stackpage=c29d7000)
<4>Stack: c0309f29 c0309fe9 0000038f c2282d60 c36d74a0 c29d7f64 c014ab76
c36d74a0
<4> 00000000 c29d6000 c29d7f80 c0143ad7 c2282d60 c2282d60 c30fa000
c29d7fa0
<4> c200fe10 c29d7fbc c0143ba8 c200fda0 c2282d60 c29d6000 081f69d0
081f6800
<4>Call Trace: [<c014ab76>] [<c0143ad7>] [<c0143ba8>] [<c0109197>]
<4>
<4>Code: 0f 0b 83 c4 0c eb 7a 39 1b 74 3e f6 83 ec 00 00 00 07 75 26
<6>Dumping to device 0x302 [ide0(3,2)] ...
<4>Writing dump header ...
<4>Writing dump pages ...<2>EXT2-fs error (device ide0(3,6)):
ext2_new_inode: reserved inode or inode > inodes count - block_group =
4,inode=60292
<2>EXT2-fs error (device ide0(3,6)): ext2_find_entry: bad entry in
directory #30146: inode out of bounds - offset=68, inode=30217, rec_len=20,
name_len=10
ng blocks not in datazone - block = 89400, count = 16
<2>EXT2-fs error (device ide0(3,6)): ext2_free_blocks: Freeing blocks not
in datazone - block = 89440, count = 16
<2>EXT2-fs error (device ide0(3,6)): ext2_free_blocks: Freeing blocks not
in datazone - block = 89464, count = 16
<2>EXT2-fs error (device ide0(3,6)): ext2_free_blocks: Freeing blocks not
in datazone - block = 89488, count = 16
<2>EXT2-fs error (device ide0(3,6)): ext2_free_blocks: Freeing blocks not
in datazone - block = 89512, count = 16
<2>EXT2-fs error (device ide0(3,6)): ext2_free_blocks: Freeing blocks not
in datazone - block = 89544, count = 16
<2>EXT2-f
====================
CURRENT SYSTEM TASKS
...
a lot of thanks in advance
t
p.s.: i still have the crashdump here if there is interest in it
--
thomas.graichen@xxxxxxxxxxxxxx
innominate AG
the linux architects
tel: +49-30-308806-13 fax: -77 http://www.innominate.com
|