1.3.72 oops

Michail Brzitwa (mb@icem.de)
Mon, 11 Mar 1996 00:20:19 +0100 (GMT+0100)


3 seconds after closing a ppp-connection under X, 1.3.72 oopsed on me:

Unable to handle kernel paging request at virtual address d8005584
current->tss.cr3 = 00aae000,
*pde = 00000000
Oops: 0000
CPU: 0
EIP: 0010:[<001295ee>]
EFLAGS: 00010297
eax: 18005550 ebx: 00000004 ecx: bffffb94 edx: c0000000
esi: bffffb94 edi: 0800a000 ebp: 0800a000 esp: 008edfb0
ds: 0018 es: 0018 fs: 002b gs: 002b ss: 0018
Process crond (pid: 854, process nr: 40, stackpage=008ed000)
Stack: 009e8810 0800a000 0010a5f9 00000004 bffffb94 00000004 0800a000 0800a000
0800a000 ffffffda bfff002b 0000002b 0000002b 0000002b 0000006c 400179f7
00100023 00000246 bffffb58 0009002b
Call Trace: [<0010a5f9>]
Code: 8b 44 98 24 85 c0 74 07 8b 40 20 85 c0 75 13 b8 f7 ff ff ff

-->

Using `/boot/kernel/System.map.1.3.72' to map addresses to symbols.

>>EIP: 1295ee <sys_newfstat+2e/70>
Trace: 10a5f9 <system_call+59/a0>

Code: 1295ee <sys_newfstat+2e/70> movl 0x24(%eax,%ebx,4),%eax
Code: 1295f2 <sys_newfstat+32/70> testl %eax,%eax
Code: 1295f4 <sys_newfstat+34/70> je 1295fd <sys_newfstat+3d/70>
Code: 1295f6 <sys_newfstat+36/70> movl 0x20(%eax),%eax
Code: 1295f9 <sys_newfstat+39/70> testl %eax,%eax
Code: 1295fb <sys_newfstat+3b/70> jne 129610 <sys_newfstat+50/70>
Code: 1295fd <sys_newfstat+3d/70> movl $0xfffffff7,%eax

This seems to be in fs/stat.c, line 215/216. From then on legions of similar
reports (all refering to sys_newfstat) came in, no processes could be loaded,
so no other information about the system state is available, I had to reset.

No modules except aout, modules 1.3.69, ppp2.2.0e (if that matters), a plain
486/66 32mb, two ide hd, one scsi, root on xiafs, 32mb swap, one reset-button.

Michail

-- 
Michail Brzitwa              <mb@icem.de>                +49-511-343215