blob: c563842ed8057e1ee960ff0faa75d87c6b797a56 [file] [log] [blame]
Linus Torvalds1da177e2005-04-16 15:20:36 -07001NOTE: ksymoops is useless on 2.6. Please use the Oops in its original format
2(from dmesg, etc). Ignore any references in this or other docs to "decoding
Jesper Juhl62a07e62005-11-07 01:01:03 -08003the Oops" or "running it through ksymoops". If you post an Oops from 2.6 that
Linus Torvalds1da177e2005-04-16 15:20:36 -07004has been run through ksymoops, people will just tell you to repost it.
5
6Quick Summary
7-------------
8
9Find the Oops and send it to the maintainer of the kernel area that seems to be
10involved with the problem. Don't worry too much about getting the wrong person.
11If you are unsure send it to the person responsible for the code relevant to
12what you were doing. If it occurs repeatably try and describe how to recreate
13it. That's worth even more than the oops.
14
15If you are totally stumped as to whom to send the report, send it to
16linux-kernel@vger.kernel.org. Thanks for your help in making Linux as
17stable as humanly possible.
18
19Where is the Oops?
20----------------------
21
22Normally the Oops text is read from the kernel buffers by klogd and
23handed to syslogd which writes it to a syslog file, typically
24/var/log/messages (depends on /etc/syslog.conf). Sometimes klogd dies,
25in which case you can run dmesg > file to read the data from the kernel
26buffers and save it. Or you can cat /proc/kmsg > file, however you
27have to break in to stop the transfer, kmsg is a "never ending file".
28If the machine has crashed so badly that you cannot enter commands or
29the disk is not available then you have three options :-
30
31(1) Hand copy the text from the screen and type it in after the machine
32 has restarted. Messy but it is the only option if you have not
33 planned for a crash.
34
35(2) Boot with a serial console (see Documentation/serial-console.txt),
36 run a null modem to a second machine and capture the output there
37 using your favourite communication program. Minicom works well.
38
39(3) Patch the kernel with one of the crash dump patches. These save
40 data to a floppy disk or video rom or a swap partition. None of
41 these are standard kernel patches so you have to find and apply
42 them yourself. Search kernel archives for kmsgdump, lkcd and
43 oops+smram.
44
45
46Full Information
47----------------
48
49NOTE: the message from Linus below applies to 2.4 kernel. I have preserved it
50for historical reasons, and because some of the information in it still
51applies. Especially, please ignore any references to ksymoops.
52
53From: Linus Torvalds <torvalds@osdl.org>
54
55How to track down an Oops.. [originally a mail to linux-kernel]
56
57The main trick is having 5 years of experience with those pesky oops
58messages ;-)
59
60Actually, there are things you can do that make this easier. I have two
61separate approaches:
62
63 gdb /usr/src/linux/vmlinux
64 gdb> disassemble <offending_function>
65
66That's the easy way to find the problem, at least if the bug-report is
67well made (like this one was - run through ksymoops to get the
68information of which function and the offset in the function that it
69happened in).
70
71Oh, it helps if the report happens on a kernel that is compiled with the
72same compiler and similar setups.
73
74The other thing to do is disassemble the "Code:" part of the bug report:
75ksymoops will do this too with the correct tools, but if you don't have
76the tools you can just do a silly program:
77
78 char str[] = "\xXX\xXX\xXX...";
79 main(){}
80
81and compile it with gcc -g and then do "disassemble str" (where the "XX"
82stuff are the values reported by the Oops - you can just cut-and-paste
83and do a replace of spaces to "\x" - that's what I do, as I'm too lazy
84to write a program to automate this all).
85
86Finally, if you want to see where the code comes from, you can do
87
88 cd /usr/src/linux
89 make fs/buffer.s # or whatever file the bug happened in
90
91and then you get a better idea of what happens than with the gdb
92disassembly.
93
94Now, the trick is just then to combine all the data you have: the C
95sources (and general knowledge of what it _should_ do), the assembly
96listing and the code disassembly (and additionally the register dump you
97also get from the "oops" message - that can be useful to see _what_ the
98corrupted pointers were, and when you have the assembler listing you can
99also match the other registers to whatever C expressions they were used
100for).
101
102Essentially, you just look at what doesn't match (in this case it was the
103"Code" disassembly that didn't match with what the compiler generated).
104Then you need to find out _why_ they don't match. Often it's simple - you
105see that the code uses a NULL pointer and then you look at the code and
106wonder how the NULL pointer got there, and if it's a valid thing to do
107you just check against it..
108
109Now, if somebody gets the idea that this is time-consuming and requires
110some small amount of concentration, you're right. Which is why I will
111mostly just ignore any panic reports that don't have the symbol table
112info etc looked up: it simply gets too hard to look it up (I have some
113programs to search for specific patterns in the kernel code segment, and
114sometimes I have been able to look up those kinds of panics too, but
115that really requires pretty good knowledge of the kernel just to be able
116to pick out the right sequences etc..)
117
118_Sometimes_ it happens that I just see the disassembled code sequence
119from the panic, and I know immediately where it's coming from. That's when
120I get worried that I've been doing this for too long ;-)
121
122 Linus
123
124
125---------------------------------------------------------------------------
126Notes on Oops tracing with klogd:
127
128In order to help Linus and the other kernel developers there has been
129substantial support incorporated into klogd for processing protection
130faults. In order to have full support for address resolution at least
131version 1.3-pl3 of the sysklogd package should be used.
132
133When a protection fault occurs the klogd daemon automatically
134translates important addresses in the kernel log messages to their
135symbolic equivalents. This translated kernel message is then
136forwarded through whatever reporting mechanism klogd is using. The
137protection fault message can be simply cut out of the message files
138and forwarded to the kernel developers.
139
140Two types of address resolution are performed by klogd. The first is
141static translation and the second is dynamic translation. Static
142translation uses the System.map file in much the same manner that
143ksymoops does. In order to do static translation the klogd daemon
144must be able to find a system map file at daemon initialization time.
145See the klogd man page for information on how klogd searches for map
146files.
147
148Dynamic address translation is important when kernel loadable modules
149are being used. Since memory for kernel modules is allocated from the
150kernel's dynamic memory pools there are no fixed locations for either
151the start of the module or for functions and symbols in the module.
152
153The kernel supports system calls which allow a program to determine
154which modules are loaded and their location in memory. Using these
155system calls the klogd daemon builds a symbol table which can be used
156to debug a protection fault which occurs in a loadable kernel module.
157
158At the very minimum klogd will provide the name of the module which
159generated the protection fault. There may be additional symbolic
160information available if the developer of the loadable module chose to
161export symbol information from the module.
162
163Since the kernel module environment can be dynamic there must be a
164mechanism for notifying the klogd daemon when a change in module
165environment occurs. There are command line options available which
166allow klogd to signal the currently executing daemon that symbol
167information should be refreshed. See the klogd manual page for more
168information.
169
170A patch is included with the sysklogd distribution which modifies the
171modules-2.0.0 package to automatically signal klogd whenever a module
172is loaded or unloaded. Applying this patch provides essentially
173seamless support for debugging protection faults which occur with
174kernel loadable modules.
175
176The following is an example of a protection fault in a loadable module
177processed by klogd:
178---------------------------------------------------------------------------
179Aug 29 09:51:01 blizard kernel: Unable to handle kernel paging request at virtual address f15e97cc
180Aug 29 09:51:01 blizard kernel: current->tss.cr3 = 0062d000, %cr3 = 0062d000
181Aug 29 09:51:01 blizard kernel: *pde = 00000000
182Aug 29 09:51:01 blizard kernel: Oops: 0002
183Aug 29 09:51:01 blizard kernel: CPU: 0
184Aug 29 09:51:01 blizard kernel: EIP: 0010:[oops:_oops+16/3868]
185Aug 29 09:51:01 blizard kernel: EFLAGS: 00010212
186Aug 29 09:51:01 blizard kernel: eax: 315e97cc ebx: 003a6f80 ecx: 001be77b edx: 00237c0c
187Aug 29 09:51:01 blizard kernel: esi: 00000000 edi: bffffdb3 ebp: 00589f90 esp: 00589f8c
188Aug 29 09:51:01 blizard kernel: ds: 0018 es: 0018 fs: 002b gs: 002b ss: 0018
189Aug 29 09:51:01 blizard kernel: Process oops_test (pid: 3374, process nr: 21, stackpage=00589000)
190Aug 29 09:51:01 blizard kernel: Stack: 315e97cc 00589f98 0100b0b4 bffffed4 0012e38e 00240c64 003a6f80 00000001
191Aug 29 09:51:01 blizard kernel: 00000000 00237810 bfffff00 0010a7fa 00000003 00000001 00000000 bfffff00
192Aug 29 09:51:01 blizard kernel: bffffdb3 bffffed4 ffffffda 0000002b 0007002b 0000002b 0000002b 00000036
193Aug 29 09:51:01 blizard kernel: Call Trace: [oops:_oops_ioctl+48/80] [_sys_ioctl+254/272] [_system_call+82/128]
194Aug 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
195---------------------------------------------------------------------------
196
197Dr. G.W. Wettstein Oncology Research Div. Computing Facility
198Roger Maris Cancer Center INTERNET: greg@wind.rmcc.com
199820 4th St. N.
200Fargo, ND 58122
201Phone: 701-234-7556
202
203
204---------------------------------------------------------------------------
205Tainted kernels:
206
207Some oops reports contain the string 'Tainted: ' after the program
Randy Dunlap1cc57532005-09-13 01:25:46 -0700208counter. This indicates that the kernel has been tainted by some
209mechanism. The string is followed by a series of position-sensitive
Linus Torvalds1da177e2005-04-16 15:20:36 -0700210characters, each representing a particular tainted value.
211
212 1: 'G' if all modules loaded have a GPL or compatible license, 'P' if
213 any proprietary module has been loaded. Modules without a
214 MODULE_LICENSE or with a MODULE_LICENSE that is not recognised by
215 insmod as GPL compatible are assumed to be proprietary.
216
Randy Dunlap1cc57532005-09-13 01:25:46 -0700217 2: 'F' if any module was force loaded by "insmod -f", ' ' if all
Linus Torvalds1da177e2005-04-16 15:20:36 -0700218 modules were loaded normally.
219
220 3: 'S' if the oops occurred on an SMP kernel running on hardware that
Randy Dunlap1cc57532005-09-13 01:25:46 -0700221 hasn't been certified as safe to run multiprocessor.
222 Currently this occurs only on various Athlons that are not
223 SMP capable.
224
225 4: 'R' if a module was force unloaded by "rmmod -f", ' ' if all
226 modules were unloaded normally.
227
228 5: 'M' if any processor has reported a Machine Check Exception,
229 ' ' if no Machine Check Exceptions have occurred.
230
231 6: 'B' if a page-release function has found a bad page reference or
232 some unexpected page flags.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700233
234The primary reason for the 'Tainted: ' string is to tell kernel
235debuggers if this is a clean kernel or if anything unusual has
Randy Dunlap1cc57532005-09-13 01:25:46 -0700236occurred. Tainting is permanent: even if an offending module is
237unloaded, the tainted value remains to indicate that the kernel is not
Linus Torvalds1da177e2005-04-16 15:20:36 -0700238trustworthy.