1NOTE: 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 3the Oops" or "running it through ksymoops". If you post an Oops from 2.6 that 4has 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. Alternatively, you can take a picture of 34 the screen with a digital camera - not nice, but better than 35 nothing. If the messages scroll off the top of the console, you 36 may find that booting with a higher resolution (eg, vga=791) 37 will allow you to read more of the text. (Caveat: This needs vesafb, 38 so won't help for 'early' oopses) 39 40(2) Boot with a serial console (see Documentation/serial-console.txt), 41 run a null modem to a second machine and capture the output there 42 using your favourite communication program. Minicom works well. 43 44(3) Use Kdump (see Documentation/kdump/kdump.txt), 45 extract the kernel ring buffer from old memory with using dmesg 46 gdbmacro in Documentation/kdump/gdbmacros.txt. 47 48 49Full Information 50---------------- 51 52NOTE: the message from Linus below applies to 2.4 kernel. I have preserved it 53for historical reasons, and because some of the information in it still 54applies. Especially, please ignore any references to ksymoops. 55 56From: Linus Torvalds <torvalds@osdl.org> 57 58How to track down an Oops.. [originally a mail to linux-kernel] 59 60The main trick is having 5 years of experience with those pesky oops 61messages ;-) 62 63Actually, there are things you can do that make this easier. I have two 64separate approaches: 65 66 gdb /usr/src/linux/vmlinux 67 gdb> disassemble <offending_function> 68 69That's the easy way to find the problem, at least if the bug-report is 70well made (like this one was - run through ksymoops to get the 71information of which function and the offset in the function that it 72happened in). 73 74Oh, it helps if the report happens on a kernel that is compiled with the 75same compiler and similar setups. 76 77The other thing to do is disassemble the "Code:" part of the bug report: 78ksymoops will do this too with the correct tools, but if you don't have 79the tools you can just do a silly program: 80 81 char str[] = "\xXX\xXX\xXX..."; 82 main(){} 83 84and compile it with gcc -g and then do "disassemble str" (where the "XX" 85stuff are the values reported by the Oops - you can just cut-and-paste 86and do a replace of spaces to "\x" - that's what I do, as I'm too lazy 87to write a program to automate this all). 88 89Alternatively, you can use the shell script in scripts/decodecode. 90Its usage is: decodecode < oops.txt 91 92The hex bytes that follow "Code:" may (in some architectures) have a series 93of bytes that precede the current instruction pointer as well as bytes at and 94following the current instruction pointer. In some cases, one instruction 95byte or word is surrounded by <> or (), as in "<86>" or "(f00d)". These 96<> or () markings indicate the current instruction pointer. Example from 97i386, split into multiple lines for readability: 98 99Code: f9 0f 8d f9 00 00 00 8d 42 0c e8 dd 26 11 c7 a1 60 ea 2b f9 8b 50 08 a1 10064 ea 2b f9 8d 34 82 8b 1e 85 db 74 6d 8b 15 60 ea 2b f9 <8b> 43 04 39 42 54 1017e 04 40 89 42 54 8b 43 04 3b 05 00 f6 52 c0 102 103Finally, if you want to see where the code comes from, you can do 104 105 cd /usr/src/linux 106 make fs/buffer.s # or whatever file the bug happened in 107 108and then you get a better idea of what happens than with the gdb 109disassembly. 110 111Now, the trick is just then to combine all the data you have: the C 112sources (and general knowledge of what it _should_ do), the assembly 113listing and the code disassembly (and additionally the register dump you 114also get from the "oops" message - that can be useful to see _what_ the 115corrupted pointers were, and when you have the assembler listing you can 116also match the other registers to whatever C expressions they were used 117for). 118 119Essentially, you just look at what doesn't match (in this case it was the 120"Code" disassembly that didn't match with what the compiler generated). 121Then you need to find out _why_ they don't match. Often it's simple - you 122see that the code uses a NULL pointer and then you look at the code and 123wonder how the NULL pointer got there, and if it's a valid thing to do 124you just check against it.. 125 126Now, if somebody gets the idea that this is time-consuming and requires 127some small amount of concentration, you're right. Which is why I will 128mostly just ignore any panic reports that don't have the symbol table 129info etc looked up: it simply gets too hard to look it up (I have some 130programs to search for specific patterns in the kernel code segment, and 131sometimes I have been able to look up those kinds of panics too, but 132that really requires pretty good knowledge of the kernel just to be able 133to pick out the right sequences etc..) 134 135_Sometimes_ it happens that I just see the disassembled code sequence 136from the panic, and I know immediately where it's coming from. That's when 137I get worried that I've been doing this for too long ;-) 138 139 Linus 140 141 142--------------------------------------------------------------------------- 143Notes on Oops tracing with klogd: 144 145In order to help Linus and the other kernel developers there has been 146substantial support incorporated into klogd for processing protection 147faults. In order to have full support for address resolution at least 148version 1.3-pl3 of the sysklogd package should be used. 149 150When a protection fault occurs the klogd daemon automatically 151translates important addresses in the kernel log messages to their 152symbolic equivalents. This translated kernel message is then 153forwarded through whatever reporting mechanism klogd is using. The 154protection fault message can be simply cut out of the message files 155and forwarded to the kernel developers. 156 157Two types of address resolution are performed by klogd. The first is 158static translation and the second is dynamic translation. Static 159translation uses the System.map file in much the same manner that 160ksymoops does. In order to do static translation the klogd daemon 161must be able to find a system map file at daemon initialization time. 162See the klogd man page for information on how klogd searches for map 163files. 164 165Dynamic address translation is important when kernel loadable modules 166are being used. Since memory for kernel modules is allocated from the 167kernel's dynamic memory pools there are no fixed locations for either 168the start of the module or for functions and symbols in the module. 169 170The kernel supports system calls which allow a program to determine 171which modules are loaded and their location in memory. Using these 172system calls the klogd daemon builds a symbol table which can be used 173to debug a protection fault which occurs in a loadable kernel module. 174 175At the very minimum klogd will provide the name of the module which 176generated the protection fault. There may be additional symbolic 177information available if the developer of the loadable module chose to 178export symbol information from the module. 179 180Since the kernel module environment can be dynamic there must be a 181mechanism for notifying the klogd daemon when a change in module 182environment occurs. There are command line options available which 183allow klogd to signal the currently executing daemon that symbol 184information should be refreshed. See the klogd manual page for more 185information. 186 187A patch is included with the sysklogd distribution which modifies the 188modules-2.0.0 package to automatically signal klogd whenever a module 189is loaded or unloaded. Applying this patch provides essentially 190seamless support for debugging protection faults which occur with 191kernel loadable modules. 192 193The following is an example of a protection fault in a loadable module 194processed by klogd: 195--------------------------------------------------------------------------- 196Aug 29 09:51:01 blizard kernel: Unable to handle kernel paging request at virtual address f15e97cc 197Aug 29 09:51:01 blizard kernel: current->tss.cr3 = 0062d000, %cr3 = 0062d000 198Aug 29 09:51:01 blizard kernel: *pde = 00000000 199Aug 29 09:51:01 blizard kernel: Oops: 0002 200Aug 29 09:51:01 blizard kernel: CPU: 0 201Aug 29 09:51:01 blizard kernel: EIP: 0010:[oops:_oops+16/3868] 202Aug 29 09:51:01 blizard kernel: EFLAGS: 00010212 203Aug 29 09:51:01 blizard kernel: eax: 315e97cc ebx: 003a6f80 ecx: 001be77b edx: 00237c0c 204Aug 29 09:51:01 blizard kernel: esi: 00000000 edi: bffffdb3 ebp: 00589f90 esp: 00589f8c 205Aug 29 09:51:01 blizard kernel: ds: 0018 es: 0018 fs: 002b gs: 002b ss: 0018 206Aug 29 09:51:01 blizard kernel: Process oops_test (pid: 3374, process nr: 21, stackpage=00589000) 207Aug 29 09:51:01 blizard kernel: Stack: 315e97cc 00589f98 0100b0b4 bffffed4 0012e38e 00240c64 003a6f80 00000001 208Aug 29 09:51:01 blizard kernel: 00000000 00237810 bfffff00 0010a7fa 00000003 00000001 00000000 bfffff00 209Aug 29 09:51:01 blizard kernel: bffffdb3 bffffed4 ffffffda 0000002b 0007002b 0000002b 0000002b 00000036 210Aug 29 09:51:01 blizard kernel: Call Trace: [oops:_oops_ioctl+48/80] [_sys_ioctl+254/272] [_system_call+82/128] 211Aug 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 212--------------------------------------------------------------------------- 213 214Dr. G.W. Wettstein Oncology Research Div. Computing Facility 215Roger Maris Cancer Center INTERNET: greg@wind.rmcc.com 216820 4th St. N. 217Fargo, ND 58122 218Phone: 701-234-7556 219 220 221--------------------------------------------------------------------------- 222Tainted kernels: 223 224Some oops reports contain the string 'Tainted: ' after the program 225counter. This indicates that the kernel has been tainted by some 226mechanism. The string is followed by a series of position-sensitive 227characters, each representing a particular tainted value. 228 229 1: 'G' if all modules loaded have a GPL or compatible license, 'P' if 230 any proprietary module has been loaded. Modules without a 231 MODULE_LICENSE or with a MODULE_LICENSE that is not recognised by 232 insmod as GPL compatible are assumed to be proprietary. 233 234 2: 'F' if any module was force loaded by "insmod -f", ' ' if all 235 modules were loaded normally. 236 237 3: 'S' if the oops occurred on an SMP kernel running on hardware that 238 hasn't been certified as safe to run multiprocessor. 239 Currently this occurs only on various Athlons that are not 240 SMP capable. 241 242 4: 'R' if a module was force unloaded by "rmmod -f", ' ' if all 243 modules were unloaded normally. 244 245 5: 'M' if any processor has reported a Machine Check Exception, 246 ' ' if no Machine Check Exceptions have occurred. 247 248 6: 'B' if a page-release function has found a bad page reference or 249 some unexpected page flags. 250 251 7: 'U' if a user or user application specifically requested that the 252 Tainted flag be set, ' ' otherwise. 253 254 8: 'D' if the kernel has died recently, i.e. there was an OOPS or BUG. 255 256 9: 'A' if the ACPI table has been overridden. 257 258 10: 'W' if a warning has previously been issued by the kernel. 259 260The primary reason for the 'Tainted: ' string is to tell kernel 261debuggers if this is a clean kernel or if anything unusual has 262occurred. Tainting is permanent: even if an offending module is 263unloaded, the tainted value remains to indicate that the kernel is not 264trustworthy. 265