The humble debugger

Of all the questions a developer may have, the question of why did my application just crash? is probably among the most important. This is also one of the questions which are most easily answered with a debugger. Regardless of whether one is live debugging a process, or analyzing the core dump of a crashed process, the debugger can (hopefully) generate a back trace, also known as a stack trace. This trace contains a chronological list of all the functions which were called since the application was started as one would find them on the stack (see Chapter 2, Multithreading Implementation on the Processor and OS, for details on how a stack works).

The last few entries of this back trace will thus show us in which part of the code things went wrong. If the debug information was compiled into the binary, or provided to the debugger, we can also see the code at that line along with the names of the variables.

Even better, since we're looking at the stack frames, we can also...