We have one 4xSMP machine that isn't quite stable. As long as it's
not loaded, it stays up but under load it tends to crash. We started
seti@home on it and there it went ..
It was running 2.2.13pre11. Because it crashed in select_dcache, and
I haven't seen any references to that in Alan's 2.2.13preX changelog
files I thought it might be useful to post the decoded oops here.
This Oops was repeatedly printed on the serial console. The machine did
still reply to pings, but not much else, probably because it spent all
its time in the serial console printing routines ..
I'm now going to compile & install 2.2.13pre17. We'll see how it
holds up.
Unable to handle kernel paging request at virtual address 6bccbce8
current->tss.cr3 = 54bd2000, %cr3 = 54bd2000
*pde = 00000000
Oops: 0000
CPU: 2
EIP: 0010:[<8013367a>]
Using defaults from ksymoops -t elf32-i386 -a i386
EFLAGS: 00010217
eax: 6bccbce0 ebx: ebc92550 ecx: 6bccbd00 edx: 0000401a
esi: 6bccbd00 edi: 00000004 ebp: 00000000 esp: fb069e90
ds: 0018 es: 0018 ss: 0018
Process apache (pid: 16263, process nr: 34, stackpage=fb069000)
Stack: 08075614 7ffffc10 00000005 00000004 000017c8 0000085b ebccb200 80134791
000007ab 00000000 00000004 801347f8 00001006 00000004 fac5ba1c 08075614
7ffffc10 faa43198 80149e99 fae1fc70 80134acd 00000004 801479cf 00000004
Call Trace: [<80134791>] [<801347f8>] [<80149e99>] [<80134acd>] [<801479cf>] [<8
Code: 8b 5e e8 c7 44 24 10 00 00 00 00 8b 76 04 83 38 00 74 17 ff
>>EIP; 8013367a <select_dcache+62/10c> <=====
Trace; 80134791 <try_to_free_inodes+25/4c>
Trace; 801347f8 <grow_inodes+20/1bc>
Trace; 80149e99 <__release_sock+29/80>
Trace; 80134acd <get_empty_inode+9d/a8>
Trace; 801479cf <sock_alloc+7/a4>
Code; 8013367a <select_dcache+62/10c>
00000000 <_EIP>:
Code; 8013367a <select_dcache+62/10c> <=====
0: 8b 5e e8 movl 0xffffffe8(%esi),%ebx <=====
Code; 8013367d <select_dcache+65/10c>
3: c7 44 24 10 00 00 00 movl $0x0,0x10(%esp,1)
Code; 80133684 <select_dcache+6c/10c>
a: 00
Code; 80133685 <select_dcache+6d/10c>
b: 8b 76 04 movl 0x4(%esi),%esi
Code; 80133688 <select_dcache+70/10c>
e: 83 38 00 cmpl $0x0,(%eax)
Code; 8013368b <select_dcache+73/10c>
11: 74 17 je 2a <_EIP+0x2a> 801336a4 <select_dcache+
8c/10c>
Code; 8013368d <select_dcache+75/10c>
13: ff 00 incl (%eax)
Mike.
-- First things first, but not necessarily in that order.- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/