[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

"Whee.. inode changed from under us. Tell Linus"



I just got the following from a Linux mcahine here, running 2.0.33 (yes
I know 2.0.34 is out).

The machine has a normal cofiguration (64 Megs of RAM, P133) and has
worked flawlesly for more then 2 years.

The following error message was recorded in the logs (and was spewed to
the console too):

Jun  6 09:00:03 ns kernel: general protection: 0000
Jun  6 09:00:03 ns kernel: CPU:    0
Jun  6 09:00:03 ns kernel: EIP:    0010:[<0012ce5e>]
Jun  6 09:00:03 ns kernel: EFLAGS: 00010246
Jun  6 09:00:03 ns kernel: eax: 02db7700   ebx: 002d74c8   ecx:
0804c160   edx: bfdffefc
Jun  6 09:00:03 ns kernel: esi: 00001000   edi: 0804c160   ebp:
bffffae8   esp: 03ed7f8c
Jun  6 09:00:03 ns kernel: ds: 0018   es: 0018   fs: 002b   gs: 002b  
ss: 0018
Jun  6 09:00:03 ns kernel: Process chgrp (pid: 19600, process nr: 43,
stackpage=03ed7000)
Jun  6 09:00:03 ns kernel: Stack: 02db7700 002d74c8 03ed7fa8 0012cd28
03905810 00000003 0804a7b8
0804c160 
Jun  6 09:00:03 ns kernel:        00000000 00001000 00000000 0010a645
00000003 0804c160 00001000
00000003 
Jun  6 09:00:03 ns kernel:        0804a7b8 bffffae8 ffffffda 0000002b
0000002b 0000002b 0000002b
0000008d 
Jun  6 09:00:03 ns kernel: Call Trace: [<0012cd28>] [<0010a645>] 
Jun  6 09:00:03 ns kernel: Code: 8b 42 0c ff d0 83 c4 10 85 c0 7c 22 8b
44 24 10 85 c0 74 16 
Jun  6 09:00:04 ns kernel: general protection: 0000
Jun  6 09:00:04 ns kernel: CPU:    0
Jun  6 09:00:04 ns kernel: EIP:    0010:[<0012ce5e>]
Jun  6 09:00:04 ns kernel: EFLAGS: 00010246
Jun  6 09:00:04 ns kernel: eax: 03a5c700   ebx: 002d72a8   ecx:
0804c2c8   edx: bffffed4
Jun  6 09:00:04 ns kernel: esi: 00001000   edi: 0804c2c8   ebp:
bffffad0   esp: 03ed7f8c
Jun  6 09:00:04 ns kernel: ds: 0018   es: 0018   fs: 002b   gs: 002b  
ss: 0018
Jun  6 09:00:04 ns kernel: Process chmod (pid: 19601, process nr: 43,
stackpage=03ed7000)
Jun  6 09:00:04 ns kernel: Stack: 03a5c700 002d72a8 03ed7fa8 0012cd28
03905810 00000003 0804aee0
0804c2c8 
Jun  6 09:00:04 ns kernel:        00000000 00001000 00000000 0010a645
00000003 0804c2c8 00001000
00000003 
Jun  6 09:00:04 ns kernel:        0804aee0 bffffad0 ffffffda 0000002b
0000002b 0000002b 0000002b
0000008d 
Jun  6 09:00:04 ns kernel: Call Trace: [<0012cd28>] [<0010a645>] 
Jun  6 09:00:04 ns kernel: Code: 8b 42 0c ff d0 83 c4 10 85 c0 7c 22 8b
44 24 10 85 c0 74 16 
Jun  6 09:00:05 ns kernel: general protection: 0000
Jun  6 09:00:05 ns kernel: CPU:    0
Jun  6 09:00:05 ns kernel: EIP:    0010:[<0012ce5e>]
Jun  6 09:00:05 ns kernel: EFLAGS: 00010246
Jun  6 09:00:05 ns kernel: eax: 0268c400   ebx: 002d7d48   ecx:
0804c2c8   edx: 919b4cd4
Jun  6 09:00:05 ns kernel: esi: 00001000   edi: 0804c2c8   ebp:
bffffad0   esp: 03ed7f8c
Jun  6 09:00:05 ns kernel: ds: 0018   es: 0018   fs: 002b   gs: 002b  
ss: 0018
Jun  6 09:00:05 ns kernel: Process chmod (pid: 19602, process nr: 43,
stackpage=03ed7000)
Jun  6 09:00:05 ns kernel: Stack: 0268c400 002d7d48 03ed7fa8 0012cd28
03905810 00000003 0804aee0
0804c2c8 
Jun  6 09:00:05 ns kernel:        00000000 00001000 00000000 0010a645
00000003 0804c2c8 00001000
00000003 
Jun  6 09:00:05 ns kernel:        0804aee0 bffffad0 ffffffda 0000002b
0000002b 0000002b 0000002b
0000008d 
Jun  6 09:00:05 ns kernel: Call Trace: [<0012cd28>] [<0010a645>] 
Jun  6 09:00:05 ns kernel: Code: 8b 42 0c ff d0 83 c4 10 85 c0 7c 22 8b
44 24 10 85 c0 74 16 
Jun  6 09:29:35 ns getty[19617]: exiting on HANGUP signal
Jun  6 09:30:10 ns getty[19618]: exiting on HANGUP signal
Jun  6 19:20:01 ns kernel: Whee.. inode changed from under us. Tell
Linus

This sort of thing has happend one time before, and then we have found
out the  CPU fan has died, this is not the case now, though.

The last change we did was add more RAM (fronm 32 to 64) but this
happend two weeks ago and the machien was running smoothly since (and it
is a very busy machine)

As far as I can tell the chmod and chgrp commands that "caused" this are
part of a script which runs every hour which checks and changes
premissions on a bunch of files - nothing special.

Now, the machine was running all these two years without it's cover, and
for some reason someone with over good intention has put a cover on it
last week - can there be a conncetion? (overheating?) or should I take
Linus on his proposal and mail him? (I dont wont to unless I must, being
a semi-deity is a lot of work as it is... ;-)

TIA,
Gilad.
-- 
Gilad Ben-Yossef <gby@gtek.co.il>     URL:   http://www.gtek.co.il
G.Tek Technologies ltd.               Phone: 972-97740484 ext. 222 
117 Ahuza St. Ra'anana 43373 ISRAEL   Fax:   972-07742790

"Linux law: Any sufficiently advanced Operating System is
indistinguishable from Unix(tm) "