Troubleshooting tips to make troubleshooting a Linux kernel panic easier

Issue Description:

Kernel panics on Linux are hard to identify and troubleshoot. Troubleshooting kernel panics often requires reproducing a situation that occurs rarely and collecting data that is difficult to gather.

Solution Summary:

This document outlines several techniques that can help reduce the amount of time necessary to troubleshoot a kernel panic.


Technical Discussion:

What is a kernel panic? 

As the name implies, it is when the Linux kernel gets into a situation where it doesn't know what to do next. When this happens, the kernel gives as much information as it can about what caused the problem, depending on what led to the panic.

There are two main kinds of kernel panics:
1) Hard Panic (also known as Aieee! )
2) Soft Panic (also known as Oops )

What can cause a kernel panic?

Only modules that are located within kernel space can directly cause the kernel to panic. To see what modules are dynamically loaded, do lsmod this shows all dynamically loaded modules (Dialogic® drivers, LiS, SCSI driver, filesystem, etc.). In addition to these dynamically loaded modules, components that are built into the kernel (memory map, etc.) can cause a panic.

Since hard panics and soft panics are different in nature, this document discusses how to deal with each separately.

How to Troubleshoot a Hard Kernel Panic

Symptoms:
1) Machine is completely locked up and unusable
2) Num Lock / Caps Lock / Scroll Lock keys usually blink
3) If in console mode, dump is displayed on monitor (including the phrase “Aieee!”)
4) Similar to Windows® Blue Screen of Death

Causes:
The most common cause of a hard kernel panic is when a driver crashes within an interrupt handler, usually because it tried to access a null pointer within the interrupt handler. When this happens, that driver cannot handle any new interrupts and eventually the system crashes. This is not exclusive to Dialogic drivers.

Information to collect:
Depending on the nature of the panic, the kernel will log all information it can prior to locking up. Since a kernel panic is a drastic failure, it is uncertain how much information will be logged. Below are key pieces of information to collect. It is important to gather as many of these as possible, but there is no guarantee that all of them will be available, especially the first time a panic is seen.

1) /var/log/messages -- sometimes the entire kernel panic stack trace will be logged there
2) Application / Library logs (RTF, cheetah, etc.) – may show what was happening before the panic
3) Other information about what happened just prior to the panic, or how to reproduce the condition
4) Screen dump from console. Since the OS is locked, you cannot cut and paste from the screen. There are two common ways to get this information:
• Digital photograph of screen (preferred, since it’s quicker and easier)
• Copying screen with pen and paper or typing to another computer

If the dump is not available either in /var/log/message or on the screen, follow these steps to get a dump:

1) If in GUI mode, switch to full console mode – no dump information is passed to the GUI (not even to GUI shell)
2) Make sure screen stays on during full test run – if a screen saver kicks in, the screen won’t return after a kernel panic. Use these settings to keep the screen on:
• setterm -blank 0
• setterm -powerdown 0
• setvesablank off
3) From console, copy dump from screen (see above)

Troubleshooting when a full trace is available

The stack trace is the most important piece of information to use in troubleshooting a kernel panic. It is often crucial to have a full stack trace, something that may not be available if only a screen dump is provided – the top of the stack may scroll off the screen, leaving only a partial stack trace. If a full trace is available, it is usually sufficient to isolate root cause. To identify whether or not you have a large enough stack trace, look for a line with EIP, which will show what function call and module caused the panic. In the example below, this is shown in the following line:
EIP is at _dlgn_setevmask [streams-dlgnDriver] 0xe

If the culprit is a Dialogic driver you will see a module name with:
streams-xxxxDriver (xxxx = dlgn, dvbm, mercd, etc.)

Hard panic – full trace example:  

Unable to handle kernel NULL pointer dereference at virtual address 0000000c 
printing eip: 
f89e568a 
*pde = 32859001 
*pte = 00000000 
Oops: 0000 
Kernel 2.4.9-31enterprise 
CPU: 1 
EIP: 0010:[] Tainted: PF 
EFLAGS: 00010096 
EIP is at _dlgn_setevmask [streams-dlgnDriver] 0xe 
eax: 00000000 ebx: f65f5410 ecx: f5e16710 edx: f65f5410 
esi: 00001ea0 edi: f5e23c30 ebp: f65f5410 esp: f1cf7e78 
ds: 0018 es: 0018 ss: 0018 
Process pwcallmgr (pid: 10334, stackpage=f1cf7000) 
Stack: 00000000 c01067fa 00000086 f1cf7ec0 00001ea0 f5e23c30 f65f5410 f89e53ec 
f89fcd60 f5e16710 f65f5410 f65f5410 f8a54420 f1cf7ec0 f8a4d73a 0000139e 
f5e16710 f89fcd60 00000086 f5e16710 f5e16754 f65f5410 0000034a f894e648 
Call Trace: [setup_sigcontext+218/288] setup_sigcontext [kernel] 0xda 
Call Trace: [] setup_sigcontext [kernel] 0xda 
[] dlgnwput [streams-dlgnDriver] 0xe8 
[] Sm_Handle [streams-dlgnDriver] 0x1ea0 
[] intdrv_lock [streams-dlgnDriver] 0x0 
[] Gn_Maxpm [streams-dlgnDriver] 0x8ba 
[] Sm_Handle [streams-dlgnDriver] 0x1ea0 
[] lis_safe_putnext [streams] 0x168 
[] __insmod_streams-dvbmDriver_S.bss_L117376 [streams-dvbmDriver] 0xab8 
[] dvbmwput [streams-dvbmDriver] 0x6f5 
[] dvwinit [streams-dvbmDriver] 0x2c0 
[] lis_safe_putnext [streams] 0x168 
[] lis_strputpmsg [streams] 0x54c 
[] __insmod_streams_S.rodata_L35552 [streams] 0x182e 
[] sys_putpmsg [streams] 0x6f 
[system_call+51/56] system_call [kernel] 0x33 
[] system_call [kernel] 0x33 
Nov 28 12:17:58 talus kernel: 
Nov 28 12:17:58 talus kernel: 
Code: 8b 70 0c 8b 06 83 f8 20 8b 54 24 20 8b 6c 24 24 76 1c 89 5c

 

Troubleshooting when a full trace is not available
If only a partial stack trace is available, it can be tricky to isolate the root cause, since there is no explicit information about what module of function call caused the panic. Instead, only commands leading up to the final command will be seen in a partial stack trace. In this case, it is very important to collect as much information as possible about what happened leading up to the kernel panic (application logs, library traces, steps to reproduce, etc).    

Hard panic – partial trace example (note there is no line with EIP information) 

[] ip_rcv [kernel] 0x357 
[] sramintr [streams_dlgnDriver] 0x32d 
[] lis_spin_lock_irqsave_fcn [streams] 0x7d 
[] inthw_lock [streams_dlgnDriver] 0x1c 
[] pwswtbl [streams_dlgnDriver] 0x0 
[] dlgnintr [streams_dlgnDriver] 0x4b 
[] Gn_Maxpm [streams_dlgnDriver] 0x7ae 
[] __run_timers [kernel] 0xd1 
[] handle_IRQ_event [kernel] 0x5e 
[] do_IRQ [kernel] 0xa4 
[] default_idle [kernel] 0x0 
[] default_idle [kernel] 0x0 
[] call_do_IRQ [kernel] 0x5 
[] default_idle [kernel] 0x0 
[] default_idle [kernel] 0x0 
[] default_idle [kernel] 0x2d 
[] cpu_idle [kernel] 0x2d 
[] __call_console_drivers [kernel] 0x4b 
[] call_console_drivers [kernel] 0xeb 
Code: 8b 50 0c 85 d2 74 31 f6 42 0a 02 74 04 89 44 24 08 31 f6 0f 
<0> Kernel panic: Aiee, killing interrupt handler! 
In interrupt handler - not syncing

 

Using kernel debugger (KDB)
If only a partial trace is available and the supporting information is not sufficient to isolate root cause, it may be useful to use kernel debugger (KDB). KDB is a tool that is compiled into the kernel to cause the kernel to break into a shell rather than lock up when a panic occurs. This enables you to collect additional information about the panic, which is often useful in determining root cause.

Some important things to note about using KDB:
1) If this is a potential Dialogic issue, then Dialogic technical support should be contacted prior to the to use of KDB
2) Must use base kernel; i.e. 2.4.18 kernel instead of 2.4.18-5 from RedHat. This is because KDB is only available for the base kernels, and not the builds created by RedHat. Although this does create a slight deviation from the original configuration, it usually does not interfere with root cause analysis.
3) Needs different Dialogic® drivers compiled to handle the specific kernel.  

How to Troubleshoot a Soft Kernel Panic

Symptoms
1) Much less severe than hard panic
2) Usually results in a segmentation fault
3) Can see an oops message – search /var/log/messages for string ‘Oops’
4) Machine still somewhat usable (but should be rebooted after information is collected)

Causes
Almost anything that causes a module to crash when it is not within an interrupt handler can cause a soft panic. In this case, the driver itself will crash but will not cause catastrophic system failure since it was not locked in the interrupt handler. The same possible causes exist for soft panics as for hard panics (i.e. accessing a null pointer during runtime).

Information to collect
When a soft panic occurs, the kernel will generate a dump that contains kernel symbols – this information is logged in /var/log/messages. To begin troubleshooting, use the ksymoops utility to turn kernel symbols into meaningful data.

To generate a ksymoops file:
1) Create new file from text of stack trace found in /var/log/messages. Make sure to strip off timestamps, otherwise ksymoops will fail.
2) Run ksymoops on new stack trace file:
Generic: ksymoops -o [location of Dialogic drivers] filename
Example: ksymoops -o /lib/modules/2.4.18-5/misc ksymoops.log
All other defaults should work fine



Soft panic – oops trace example

Code: 8b 70 0c 50 e8 69 f9 f8 ff 83 c4 10 83 f8 08 74 35 66 c7 47 
EIP; f89ba71e <[streams-dlgnDriver]_dlgn_setidlestate+1e/8c> 
Trace; f8951bd6 <[streams]lis_wakeup_close+86/110> 
Trace; f8a2705c <[streams-dlgnDriver]__module_parm_r4_feature+280/1453> 
Trace; f8a27040 <[streams-dlgnDriver]__module_parm_r4_feature+264/1453> 
Trace; f89b9198 <[streams-dlgnDriver]dlgnwput+e8/204>

Product List

Dialogic® System Release Software for Linux, all versions

Glossary of Acronyms / Terms

LiS – Linux Streams
SCSI – Small Computer Systems Interface
RTF – Runtime Tracing Facility
KDB – Kernel Debugger