oops during normal swap
Christoph Rohland
hans-christoph.rohland en sap.com
Vie Ene 28 14:07:50 CST 2000
Hi,
After some rereading shm swap code I rewrote my swap test program to
use anonymous mmap instead of shm to drive my machine into swapping
and viola:
shrink_mmap: unknown LRU page!
kernel BUG at swap_state.c:49!
invalid operand: 0000
CPU: 0
EIP: 0010:[<c013a994>]
Using defaults from ksymoops -t elf32-i386 -a i386
EFLAGS: 00010296
eax: 0000001f ebx: c403d984 ecx: c028229c edx: c028229c
esi: 00000000 edi: 03014000 ebp: 00000000 esp: f79b5da4
ds: 0018 es: 0018 ss: 0018
Process sendmail (pid: 518, stackpage=f79b5000)
Stack: c022e704 c022e9ef 00000031 c403d984 c013ad7b c403d984 03014000 0180a500
00000000 03014a00 c012c912 03014000 00000000 f79b4000 00000000 00000000
f79b4000 00000010 00030140 c012c9ec 03014a00 f79a0ae0 4017d040 f79b4000
Call Trace: [<c022e704>] [<c022e9ef>] [<c013ad7b>] [<c012c912>] [<c012c9ec>]
[<c0116a99>] [<c010d121>]
Code: 0f 0b 83 c4 0c 8d b4 26 00 00 00 00 83 7b 08 00 74 1a 6a 33
>>EIP; c013a994 <add_to_swap_cache+2c/70> <=====
Trace; c022e704 <tvecs+3360/733c>
Trace; c022e9ef <tvecs+364b/733c>
Trace; c013ad7b <read_swap_cache_async+b3/f8>
Trace; c012c912 <swapin_readahead+9a/e8>
Trace; c012c9ec <do_swap_page+8c/35c>
Trace; c0116a99 <do_page_fault+181/4c8>
Trace; c010d121 <error_code+2d/34>
Code; c013a994 <add_to_swap_cache+2c/70>
00000000 <_EIP>:
Code; c013a994 <add_to_swap_cache+2c/70> <=====
0: 0f 0b ud2a <=====
Code; c013a996 <add_to_swap_cache+2e/70>
2: 83 c4 0c addl $0xc,%esp
Code; c013a999 <add_to_swap_cache+31/70>
5: 8d b4 26 00 00 00 00 leal 0x0(%esi,1),%esi
Code; c013a9a0 <add_to_swap_cache+38/70>
c: 83 7b 08 00 cmpl $0x0,0x8(%ebx)
Code; c013a9a4 <add_to_swap_cache+3c/70>
10: 74 1a je 2c <_EIP+0x2c> c013a9c0 <add_to_swap_cache+58/70>
Code; c013a9a6 <add_to_swap_cache+3e/70>
12: 6a 33 pushl $0x33
kernel BUG at page_alloc.c:83!
invalid operand: 0000
CPU: 6
EIP: 0010:[<c013a0ae>]
EFLAGS: 00010296
eax: 0000001f ebx: c403d984 ecx: c028229c edx: c028229c
esi: c3ea9668 edi: 00000000 ebp: c01469b5 esp: c1147cc4
ds: 0018 es: 0018 ss: 0018
Process klogd (pid: 348, stackpage=c1147000)
Stack: c022e344 c022e5d6 00000053 c3ea9620 c3ea9668 c3ea9620 c01469b5 c0285bb4
00000286 00000023 c0285bac c01469b5 c0146a99 c3ea9620 00001000 c3ea9620
00001000 00000006 00000003 c3ea9620 c0144f12 c403d984 c403d984 c403d984
Call Trace: [<c022e344>] [<c022e5d6>] [<c01469b5>] [<c01469b5>] [<c0146a99>]
[<c012cb7c>] [<c0143e56>] [<c013aa4b>] [<c012d621>] [<c0116a99>]
[<c011e500>] [<c0183c4c>] [<c011ed1d>] [<c01687e5>] [<c0141256>]
Code: 0f 0b 83 c4 0c 8b 43 18 a8 01 74 16 6a 55 68 d6 e5 22 c0 68
>>EIP; c013a0ae <__free_pages_ok+5e/2a4> <=====
Trace; c022e344 <tvecs+2fa0/733c>
Trace; c022e5d6 <tvecs+3232/733c>
Trace; c01469b5 <try_to_free_buffers+1d5/38c>
Trace; c01469b5 <try_to_free_buffers+1d5/38c>
Trace; c0146a99 <try_to_free_buffers+2b9/38c>
Trace; c012cb7c <do_swap_page+21c/35c>
Trace; c0143e56 <end_buffer_io_async+13a/194>
Trace; c013aa4b <__delete_from_swap_cache+73/7c>
Trace; c012d621 <handle_mm_fault+4c1/6ac>
Trace; c0116a99 <do_page_fault+181/4c8>
Trace; c011e500 <do_syslog+410/9f0>
Trace; c0183c4c <sock_write+9c/a8>
Trace; c011ed1d <printk+1f5/204>
Trace; c01687e5 <kmsg_read+11/18>
Trace; c0141256 <sys_read+106/128>
Code; c013a0ae <__free_pages_ok+5e/2a4>
00000000 <_EIP>:
Code; c013a0ae <__free_pages_ok+5e/2a4> <=====
0: 0f 0b ud2a <=====
Code; c013a0b0 <__free_pages_ok+60/2a4>
2: 83 c4 0c addl $0xc,%esp
Code; c013a0b3 <__free_pages_ok+63/2a4>
5: 8b 43 18 movl 0x18(%ebx),%eax
Code; c013a0b6 <__free_pages_ok+66/2a4>
8: a8 01 testb $0x1,%al
Code; c013a0b8 <__free_pages_ok+68/2a4>
a: 74 16 je 22 <_EIP+0x22> c013a0d0 <__free_pages_ok+80/2a4>
Code; c013a0ba <__free_pages_ok+6a/2a4>
c: 6a 55 pushl $0x55
Code; c013a0bc <__free_pages_ok+6c/2a4>
e: 68 d6 e5 22 c0 pushl $0xc022e5d6
Code; c013a0c1 <__free_pages_ok+71/2a4>
13: 68 00 00 00 00 pushl $0x0
The stack is not complete, my serial console truncated the lines.
This is on 2.3.40+smp-41-A2 from mingo with HIGHMEM enabled and 8GB 8way.
I am restarting the tests now with HIGHMEM disabled.
Greetings
Christoph
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo en vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
Más información sobre la lista de distribución Ayuda