kswapd/reiserfs oopsen in 2.4.18-ac2

DISCUSSION Christian Reis 2 participants 3 messages Jul 12, 2002 1 message not in the archive

Share & tools

Christian Reis

author

2002-07-12 19:51 UTC #0 lore raw

Hey there,

Anybody seen this happen with them? This is a moderately-loaded Athlon1.4G with 512M and 512M swap, which serves NFS to around 20 workstations, and runs some sessions of Mozilla with X forwarded to some clients. The tape drive was erasing at the moment of the crash, but I'm not sure if it has to do with anything.

The box runs ReiserFS over software RAID1, and is currently on 2.4.18-ac2. I'm moving on to ac3 just in case, and will probably look at 2.4.19-rc immediately after if the problem isn't solved.

Looking through the logs, I am seeing a ReiserFS oops about 14 hours earlier, which may or not be related. I haven't had an oops in over 2 years now :-/

Does the signature look familiar to anybody?

Jul 12 16:28:03 anthem kernel: Unable to handle kernel paging request at virtual address 5f796f72
Jul 12 16:28:03 anthem kernel:  printing eip:
Jul 12 16:28:03 anthem kernel: 5f796f72
Jul 12 16:28:03 anthem kernel: *pde = 00000000
Jul 12 16:28:03 anthem kernel: Oops: 0000
Jul 12 16:28:03 anthem kernel: CPU:    0
Jul 12 16:28:03 anthem kernel: EIP:    0010:[<5f796f72>]    Not tainted
Jul 12 16:28:03 anthem kernel: EFLAGS: 00010206
Jul 12 16:28:03 anthem kernel: eax: 5f796f72   ebx: c93f42c0   ecx: c93f42d0   edx: c93f42d0
Jul 12 16:28:03 anthem kernel: esi: 4003b0b8   edi: c1ab3040   ebp: 00002964   esp: c16b5f7c
Jul 12 16:28:03 anthem kernel: ds: 0018   es: 0018   ss: 0018
Jul 12 16:28:03 anthem kernel: Process kswapd (pid: 4, stackpage=c16b5000)
Jul 12 16:28:03 anthem kernel: Stack: c0143727 c93f42c0 c630e6d8 c630e6c0 c93f42c0 c0141766 c93f42c0 000001d0 
Jul 12 16:28:03 anthem kernel:        c02b0900 00000000 000003c6 c0141a4b 00003fae c0129938 00000006 000001d0 
Jul 12 16:28:03 anthem kernel:        000001d0 00000376 c02b0900 00000000 0008e000 c0129c1e 000001d0 00010f00 
Jul 12 16:28:03 anthem kernel: Call Trace: [iput+55/416] [prune_dcache+198/320] [shrink_dcache_memory+27/48] [do_try_to_free_pages+24/384] [kswapd+254/768] 
Jul 12 16:28:03 anthem kernel:    [rest_init+0/48] [kernel_thread+35/48] 
Jul 12 16:28:03 anthem kernel: 
Jul 12 16:28:03 anthem kernel: Code:  Bad EIP value.

And the reiserfs oops:

Jul 12 01:54:55 anthem kernel: Unable to handle kernel NULL pointer dereference at virtual address 00000028
Jul 12 01:54:55 anthem kernel:  printing eip:
Jul 12 01:54:55 anthem kernel: c01431be
Jul 12 01:54:55 anthem kernel: *pde = 00000000
Jul 12 01:54:55 anthem kernel: Oops: 0000
Jul 12 01:54:55 anthem kernel: CPU:    0
Jul 12 01:54:55 anthem kernel: EIP:    0010:[find_inode+30/96]    Not tainted
Jul 12 01:54:55 anthem kernel: EFLAGS: 00010207
Jul 12 01:54:55 anthem kernel: eax: dff8ee00   ebx: 00000000   ecx: 0000000f   edx: 00001327
Jul 12 01:54:55 anthem kernel: esi: 00000000   edi: c018f500   ebp: dff8ee00   esp: ca845e28
Jul 12 01:54:55 anthem kernel: ds: 0018   es: 0018   ss: 0018
Jul 12 01:54:55 anthem kernel: Process tar (pid: 7902, stackpage=ca845000)
Jul 12 01:54:55 anthem kernel: Stack: ca845ea0 c018f500 00001327 df20e800 c01435f7 df20e800 00001327 dff8ee00 
Jul 12 01:54:55 anthem kernel:        c018f500 ca845e80 ca845ea0 ca845f04 ca845ee0 00000000 dff8ee00 c018f547 
Jul 12 01:54:55 anthem kernel:        df20e800 00001327 c018f500 ca845e80 ca845ea0 00000001 00001316 c018b307 
Jul 12 01:54:55 anthem kernel: Call Trace: [reiserfs_find_actor+0/32] [iget4+71/208] [reiserfs_find_actor+0/32] [reiserfs_iget+39/112] [reiserfs_find_actor+0/32] 
Jul 12 01:54:55 anthem kernel:    [reiserfs_lookup+151/224] [real_lookup+83/208] [link_path_walk+1224/1792] [path_walk+26/32] [__user_walk+53/80] [sys_newlstat+25/128] 
Jul 12 01:54:55 anthem kernel:    [sys_close+67/96] [system_call+47/52] 
Jul 12 01:54:55 anthem kernel: 
Jul 12 01:54:55 anthem kernel: Code: 39 53 28 75 24 8b 54 24 14 39 93 9c 00 00 00 75 18 85 ff 74

Take care,
--
Christian Reis, Senior Engineer, Async Open Source, Brazil.
http://async.com.br/~kiko/ | [+55 16] 261 2331 | NMFL

Thunder from the hill

2002-07-12 20:12 UTC #1 lore raw

Hi,

Quoted 1 line · Christian Reis

Does the signature look familiar to anybody?

Could you please send the two BUG()s through ksymoops? Just cut and paste. The output is a lot more helpful.

Regards, Thunder

Signature · 7 lines

(Use http://www.ebb.org/ungeek if you can't decode)
------BEGIN GEEK CODE BLOCK------
Version: 3.12
GCS/E/G/S/AT d- s++:-- a? C++$ ULAVHI++++$ P++$ L++++(+++++)$ E W-$
N--- o?  K? w-- O- M V$ PS+ PE- Y- PGP+ t+ 5+ X+ R- !tv b++ DI? !D G
e++++ h* r--- y- 
------END GEEK CODE BLOCK------

A message that is not in the archive (sent off-list or not yet received): Pine.LNX.4.33.0207121607470.8654-100000@coffee.psychology.mcmaster.ca

Christian Reis

author

2002-07-12 20:28 UTC #3 lore raw

Quoted 9 lines · Mark Hahn

Quoted 2 lines
Anybody seen this happen with them? This is a moderately-loaded Athlon1.4G with 512M and 512M swap, which serves NFS to around 20
any reason to think this isn't bad power, hot cpu, bad ram? athlons of that era certainly had those problems a lot, and your two oopses look suspicious like hardware corruption (the first, someone has tried to load eax with a pointer, but got text instead. and the second someone probably did 28(ebx), which was not supposed to be null...)

This box has been online for a couple of months now, and this is the first time. The only thing new has been the new tape drive. I do agree it may be the power supply, however. I've attached the output run through ksymoops anyway, just in case somebody's seen this before.

Quoted 2 lines

try running memtest86 overnight, and if that works, compiling the kernel in a loop for a few hours...

Yep, will do tonight.

Ksymoops follow:

ksymoops 2.4.5 on i686 2.4.18-ac3. Options used
-V (default)
-k /proc/ksyms (default)
-l /proc/modules (default)
-o /lib/modules/2.4.18-ac3/ (default)
-m /boot/System.map (specified)

Warning (compare_maps): mismatch on symbol partition_name , ksyms_base says c0214ba0, System.map says c0150400. Ignoring ksyms_base entry

Jul 12 01:54:55 anthem kernel: Unable to handle kernel NULL pointer dereference at virtual address 00000028
Jul 12 01:54:55 anthem kernel: c01431be
Jul 12 01:54:55 anthem kernel: *pde = 00000000
Jul 12 01:54:55 anthem kernel: Oops: 0000
Jul 12 01:54:55 anthem kernel: CPU:    0
Jul 12 01:54:55 anthem kernel: EIP:    0010:[find_inode+30/96]    Not tainted
Jul 12 01:54:55 anthem kernel: EFLAGS: 00010207
Jul 12 01:54:55 anthem kernel: eax: dff8ee00   ebx: 00000000   ecx: 0000000f   edx: 00001327
Jul 12 01:54:55 anthem kernel: esi: 00000000   edi: c018f500   ebp: dff8ee00   esp: ca845e28
Jul 12 01:54:55 anthem kernel: ds: 0018   es: 0018   ss: 0018
Jul 12 01:54:55 anthem kernel: Process tar (pid: 7902, stackpage=ca845000)
Jul 12 01:54:55 anthem kernel: Stack: ca845ea0 c018f500 00001327 df20e800 c01435f7 df20e800 00001327 dff8ee00 
Jul 12 01:54:55 anthem kernel:        c018f500 ca845e80 ca845ea0 ca845f04 ca845ee0 00000000 dff8ee00 c018f547 
Jul 12 01:54:55 anthem kernel:        df20e800 00001327 c018f500 ca845e80 ca845ea0 00000001 00001316 c018b307 
Jul 12 01:54:55 anthem kernel: Call Trace: [reiserfs_find_actor+0/32] [iget4+71/208] [reiserfs_find_actor+0/32] [reiserfs_iget+39/112] [reiserfs_find_actor+0/32] 
Jul 12 01:54:55 anthem kernel: Code: 39 53 28 75 24 8b 54 24 14 39 93 9c 00 00 00 75 18 85 ff 74

Using defaults from ksymoops -t elf32-i386 -a i386

Quoted 5 lines

Quoted 5 lines
eax; dff8ee00 <_end+1fc58b1c/20c87d1c>
edx; 00001327 Before first symbol
edi; c018f500 <reiserfs_find_actor+0/20>
ebp; dff8ee00 <_end+1fc58b1c/20c87d1c>
esp; ca845e28 <_end+a50fb44/20c87d1c>

Code; 00000000 Before first symbol
00000000 <_EIP>:
Code; 00000000 Before first symbol
0: 39 53 28 cmpl %edx,0x28(%ebx)
Code; 00000003 Before first symbol
3: 75 24 jne 29 <_EIP+0x29> 00000029 Before first symbol
Code; 00000005 Before first symbol
5: 8b 54 24 14 movl 0x14(%esp,1),%edx
Code; 00000009 Before first symbol
9: 39 93 9c 00 00 00 cmpl %edx,0x9c(%ebx)
Code; 0000000f Before first symbol
f: 75 18 jne 29 <_EIP+0x29> 00000029 Before first symbol
Code; 00000011 Before first symbol
11: 85 ff testl %edi,%edi
Code; 00000013 Before first symbol
13: 74 00 je 15 <_EIP+0x15> 00000015 Before first symbol

1 warning issued. Results may not be reliable.

ksymoops 2.4.5 on i686 2.4.18-ac3. Options used
-V (default)
-k /proc/ksyms (default)
-l /proc/modules (default)
-o /lib/modules/2.4.18-ac3/ (default)
-m /boot/System.map (specified)

Warning (compare_maps): mismatch on symbol partition_name , ksyms_base says c0214ba0, System.map says c0150400. Ignoring ksyms_base entry

Jul 12 16:28:03 anthem kernel: Unable to handle kernel paging request at virtual address 5f796f72
Jul 12 16:28:03 anthem kernel: 5f796f72
Jul 12 16:28:03 anthem kernel: *pde = 00000000
Jul 12 16:28:03 anthem kernel: Oops: 0000
Jul 12 16:28:03 anthem kernel: CPU:    0
Jul 12 16:28:03 anthem kernel: EIP:    0010:[<5f796f72>]    Not tainted
Using defaults from ksymoops -t elf32-i386 -a i386
Jul 12 16:28:03 anthem kernel: EFLAGS: 00010206
Jul 12 16:28:03 anthem kernel: eax: 5f796f72   ebx: c93f42c0   ecx: c93f42d0   edx: c93f42d0
Jul 12 16:28:03 anthem kernel: esi: 4003b0b8   edi: c1ab3040   ebp: 00002964   esp: c16b5f7c
Jul 12 16:28:03 anthem kernel: ds: 0018   es: 0018   ss: 0018
Jul 12 16:28:03 anthem kernel: Process kswapd (pid: 4, stackpage=c16b5000)
Jul 12 16:28:03 anthem kernel: Stack: c0143727 c93f42c0 c630e6d8 c630e6c0 c93f42c0 c0141766 c93f42c0 000001d0 
Jul 12 16:28:03 anthem kernel:        c02b0900 00000000 000003c6 c0141a4b 00003fae c0129938 00000006 000001d0 
Jul 12 16:28:03 anthem kernel:        000001d0 00000376 c02b0900 00000000 0008e000 c0129c1e 000001d0 00010f00 
Jul 12 16:28:03 anthem kernel: Call Trace: [iput+55/416] [prune_dcache+198/320] [shrink_dcache_memory+27/48] [do_try_to_free_pages+24/384] [kswapd+254/768] 
Jul 12 16:28:03 anthem kernel: Code:  Bad EIP value.

Quoted 10 lines

Quoted 10 lines
EIP; 5f796f72 Before first symbol <=====
eax; 5f796f72 Before first symbol
ebx; c93f42c0 <_end+90bdfdc/20c87d1c>
ecx; c93f42d0 <_end+90bdfec/20c87d1c>
edx; c93f42d0 <_end+90bdfec/20c87d1c>
esi; 4003b0b8 Before first symbol
edi; c1ab3040 <_end+177cd5c/20c87d1c>
ebp; 00002964 Before first symbol
esp; c16b5f7c <_end+137fc98/20c87d1c>

1 warning issued. Results may not be reliable.

Take care,
--
Christian Reis, Senior Engineer, Async Open Source, Brazil.
http://async.com.br/~kiko/ | [+55 16] 261 2331 | NMFL