OpenCores
URL https://opencores.org/ocsvn/or1k/or1k/trunk

Subversion Repositories or1k

[/] [or1k/] [trunk/] [rc203soc/] [sw/] [uClinux/] [Documentation/] [oops-tracing.txt] - Blame information for rev 1777

Go to most recent revision | Details | Compare with Previous | View Log

Line No. Rev Author Line
1 1625 jcastillo
From: Linus Torvalds 
2
 
3
How to track down an Oops.. [originally a mail to linux-kernel]
4
 
5
The main trick is having 5 years of experience with those pesky oops
6
messages ;-)
7
 
8
Actually, there are things you can do that make this easier. I have two
9
separate approaches:
10
 
11
        gdb /usr/src/linux/vmlinux
12
        gdb> disassemble 
13
 
14
That's the easy way to find the problem, at least if the bug-report is
15
well made (like this one was - run through ksymoops to get the
16
information of which function and the offset in the function that it
17
happened in).
18
 
19
Oh, it helps if the report happens on a kernel that is compiled with the
20
same compiler and similar setups.
21
 
22
The other thing to do is disassemble the "Code:" part of the bug report:
23
ksymoops will do this too with the correct tools (and new version of
24
ksymoops), but if you don't have the tools you can just do a silly
25
program:
26
 
27
        char str[] = "\xXX\xXX\xXX...";
28
        main(){}
29
 
30
and compile it with gcc -g and then do "disassemble str" (where the "XX"
31
stuff are the values reported by the Oops - you can just cut-and-paste
32
and do a replace of spaces to "\x" - that's what I do, as I'm too lazy
33
to write a program to automate this all).
34
 
35
Finally, if you want to see where the code comes from, you can do
36
 
37
        cd /usr/src/linux
38
        make fs/buffer.s        # or whatever file the bug happened in
39
 
40
and then you get a better idea of what happens than with the gdb
41
disassembly.
42
 
43
Now, the trick is just then to combine all the data you have: the C
44
sources (and general knowledge of what it _should_ do, the assembly
45
listing and the code disassembly (and additionally the register dump you
46
also get from the "oops" message - that can be useful to see _what_ the
47
corrupted pointers were, and when you have the assembler listing you can
48
also match the other registers to whatever C expressions they were used
49
for).
50
 
51
Essentially, you just look at what doesn't match (in this case it was the
52
"Code" disassembly that didn't match with what the compiler generated).
53
Then you need to find out _why_ they don't match. Often it's simple - you
54
see that the code uses a NULL pointer and then you look at the code and
55
wonder how the NULL pointer got there, and if it's a valid thing to do
56
you just check against it..
57
 
58
Now, if somebody gets the idea that this is time-consuming and requires
59
some small amount of concentration, you're right. Which is why I will
60
mostly just ignore any panic reports that don't have the symbol table
61
info etc looked up: it simply gets too hard to look it up (I have some
62
programs to search for specific patterns in the kernel code segment, and
63
sometimes I have been able to look up those kinds of panics too, but
64
that really requires pretty good knowledge of the kernel just to be able
65
to pick out the right sequences etc..)
66
 
67
_Sometimes_ it happens that I just see the disassembled code sequence
68
from the panic, and I know immediately where it's coming from. That's when
69
I get worried that I've been doing this for too long ;-)
70
 
71
                Linus
72
 
73
 
74
---------------------------------------------------------------------------
75
Notes on Oops tracing with klogd:
76
 
77
In order to help Linus and the other kernel developers there has been
78
substantial support incorporated into klogd for processing protection
79
faults.  In order to have full support for address resolution at least
80
version 1.3-pl3 of the sysklogd package should be used.
81
 
82
When a protection fault occurs the klogd daemon automatically
83
translates important addresses in the kernel log messages to their
84
symbolic equivalents.  This translated kernel message is then
85
forwarded through whatever reporting mechanism klogd is using.  The
86
protection fault message can be simply cut out of the message files
87
and forwarded to the kernel developers.
88
 
89
Two types of address resolution are performed by klogd.  The first is
90
static translation and the second is dynamic translation.  Static
91
translation uses the System.map file in much the same manner that
92
ksymoops does.  In order to do static translation the klogd daemon
93
must be able to find a system map file at daemon initialization time.
94
See the klogd man page for information on how klogd searches for map
95
files.
96
 
97
Dynamic address translation is important when kernel loadable modules
98
are being used.  Since memory for kernel modules is allocated from the
99
kernel's dynamic memory pools there are no fixed locations for either
100
the start of the module or for functions and symbols in the module.
101
 
102
The kernel supports system calls which allow a program to determine
103
which modules are loaded and their location in memory.  Using these
104
system calls the klogd daemon builds a symbol table which can be used
105
to debug a protection fault which occurs in a loadable kernel module.
106
 
107
At the very minimum klogd will provide the name of the module which
108
generated the protection fault.  There may be additional symbolic
109
information available if the developer of the loadable module chose to
110
export symbol information from the module.
111
 
112
Since the kernel module environment can be dynamic there must be a
113
mechanism for notifying the klogd daemon when a change in module
114
environment occurs.  There are command line options available which
115
allow klogd to signal the currently executing daemon that symbol
116
information should be refreshed.  See the klogd manual page for more
117
information.
118
 
119
A patch is included with the sysklogd distribution which modifies the
120
modules-2.0.0 package to automatically signal klogd whenever a module
121
is loaded or unloaded.  Applying this patch provides essentially
122
seamless support for debugging protection faults which occur with
123
kernel loadable modules.
124
 
125
The following is an example of a protection fault in a loadable module
126
processed by klogd:
127
---------------------------------------------------------------------------
128
Aug 29 09:51:01 blizard kernel: Unable to handle kernel paging request at virtual address f15e97cc
129
Aug 29 09:51:01 blizard kernel: current->tss.cr3 = 0062d000, %cr3 = 0062d000
130
Aug 29 09:51:01 blizard kernel: *pde = 00000000
131
Aug 29 09:51:01 blizard kernel: Oops: 0002
132
Aug 29 09:51:01 blizard kernel: CPU:    0
133
Aug 29 09:51:01 blizard kernel: EIP:    0010:[oops:_oops+16/3868]
134
Aug 29 09:51:01 blizard kernel: EFLAGS: 00010212
135
Aug 29 09:51:01 blizard kernel: eax: 315e97cc   ebx: 003a6f80   ecx: 001be77b   edx: 00237c0c
136
Aug 29 09:51:01 blizard kernel: esi: 00000000   edi: bffffdb3   ebp: 00589f90   esp: 00589f8c
137
Aug 29 09:51:01 blizard kernel: ds: 0018   es: 0018   fs: 002b   gs: 002b   ss: 0018
138
Aug 29 09:51:01 blizard kernel: Process oops_test (pid: 3374, process nr: 21, stackpage=00589000)
139
Aug 29 09:51:01 blizard kernel: Stack: 315e97cc 00589f98 0100b0b4 bffffed4 0012e38e 00240c64 003a6f80 00000001
140
Aug 29 09:51:01 blizard kernel:        00000000 00237810 bfffff00 0010a7fa 00000003 00000001 00000000 bfffff00
141
Aug 29 09:51:01 blizard kernel:        bffffdb3 bffffed4 ffffffda 0000002b 0007002b 0000002b 0000002b 00000036
142
Aug 29 09:51:01 blizard kernel: Call Trace: [oops:_oops_ioctl+48/80] [_sys_ioctl+254/272] [_system_call+82/128]
143
Aug 29 09:51:01 blizard kernel: Code: c7 00 05 00 00 00 eb 08 90 90 90 90 90 90 90 90 89 ec 5d c3
144
---------------------------------------------------------------------------
145
 
146
Dr. G.W. Wettstein           Oncology Research Div. Computing Facility
147
Roger Maris Cancer Center    INTERNET: greg@wind.rmcc.com
148
820 4th St. N.
149
Fargo, ND  58122
150
Phone: 701-234-7556

powered by: WebSVN 2.1.0

© copyright 1999-2024 OpenCores.org, equivalent to Oliscience, all rights reserved. OpenCores®, registered trademark.