I applied your patch, and while it helped for some processes, it failed
on others. I wound up with 60 something tasks in the D state (all on
"wait_on_sup..." according to WCHAN), but only got 2 OOPS.
alt-sysrq-T generated another couple of OOPSes.
This is the second time I've fallen afoul of this problem, and I was
running dpkg whilst doing a compile in the background on both
occasions. Yet it is not regularly repeatable. I've done this on other
occasions without any trouble.
Here is the output from ksymoops. Hope it is useful
Unable to handle kernel NULL pointer dereference at virtual address 00000000
current->tss.cr3 = 02702000, %cr3 = 02702000
*pde = 00000000
Oops: 0002
CPU: 0
EIP: 0010:[<c0110cdb>]
EFLAGS: 00010246
eax: 02702000 ebx: c7babd9c ecx: c71fe000 edx: c1a0d600
esi: c757e000 edi: c757feec ebp: c757ff08 esp: c757fee8
ds: 0018 es: 0018 ss: 0018
Process nacctd (pid: 7235, process nr: 125, stackpage=c757f000)
Stack: c7babd50 c757e000 c7babda8 00000000 00000000 0021fe45 c7babd9c c0110c10
c7abdc60 c01b12dc c7babd9c 5efcbe08 00000000 c012af58 c757ff4c c24c3016
00000001 c012b13e c7abdc60 c757ff4c c24c3000 ffffffe9 00000242 c24c3000
Call Trace: [<c0110c10>] [<c01b12dc>] [<c012af58>] [<c012b13e>] [<c012b2b5>]
[<c01097c4>]
Code: c7 05 00 00 00 00 00 00 00 00 c7 06 02 00 00 00 eb af 8d 76
>>EIP: c0110cdb <__down+ab/fc>
Trace: c0110c10 <generate_oops+0/20>
Trace: c01b12dc <__down_failed+8/c>
Trace: c012af58 <real_lookup+24/7c>
Trace: c012b13e <lookup_dentry+10a/1b8>
Trace: c012b2b5 <open_namei+6d/370>
Trace: c01097c4 <system_call+34/38>
Code: c0110cdb <__down+ab/fc> 00000000 <_EIP>:
Code: c0110cdb <__down+ab/fc> 0: c7 05 00 00 00 movl $0x0,0x0
Code: c0110ce0 <__down+b0/fc> 5: 00 00 00 00 00
Code: c0110ce5 <__down+b5/fc> a: c7 06 02 00 00 movl $0x2,(%esi)
Code: c0110cea <__down+ba/fc> f: 00
Code: c0110ceb <__down+bb/fc> 10: eb af jmp ffffffc1 <_EIP+0xffffffc1> c0110c9c <__down+6c/fc>
Code: c0110ced <__down+bd/fc> 12: 8d 76 00 leal 0x0(%esi),%esi
Cheers,
--Andrew
-
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/