CRASH(8) System Manager's Manual CRASH(8)

crash
examine and debug system images

crash [
-w
] [
-M core
] [
-N kernel
]

The crash command is used to examine and debug system images.
If run without any arguments, crash operates on the running system.
The options are as follows:
 
 
core
Operate on the specified crash dump instead of the default /dev/mem. Crash dumps should be from the same version of the system and same machine architecture as the running version of crash, and must be uncompressed.
 
 
kernel
Extract the name list from the specified kernel instead of the default /dev/ksyms.
 
 
Enable writing.
The command syntax used by crash is the same as the in-kernel debugger. See the ddb(4) manual page for more information.
Operations and facilities that require a running system, such as breakpoints, are not supported by crash.
crash does not provide pagination. However, by using the pipe symbol, output may be sent to commands available from the shell. For example:
crash> ps | more 
crash> ps | grep ioflush

You can obtain a full list of all kernel stack traces with a shell pipeline like the following:
# echo ps | crash | grep '^[0-9]' | cut -c 30-46 | \ 
  sed 's,^,bt/a ,' | crash

ps(1), vmstat(1), ddb(4), pstat(8)

The crash command appeared in NetBSD 6.0.

The crash command for the sparc port is unstable.
March 22, 2019 NetBSD-current