Unreachable code is programming code that cannot be executed in any context. There are a variety of reasons this could occur, and many computer languages do not allow execution of a program if a section of code is unreachable. Problems generated by unreachable code can be fixed by either removing the problematic section of code or rewriting part of the code to make the section reachable again.
When engineering software, it can be relatively easy to inadvertently create unreachable code. Any significant disruption in the flow of the program has the potential to cause a code section that is unreachable. For example, accidental creation of an infinite loop may render all code that comes after it unreachable because there would be no way to ever break out of the loop and continue the program. Another way this may happen is if a function is coded to return before all its code is executed. In this case, the code after the return statement would be unreachable.
There are two ways in which unreachable code may be generated. First, improperly handling control structures when writing raw code may result in sections of code that are unreachable. Second, errors in control structures that aren't obvious during the coding process, but show up when a program is run, may also cause code that is unreachable. The first type of error goes by a variety of names, including coding error or syntax error, and the second type of error is a runtime error. Coding errors are usually caught by parsers or compilers, but runtime errors can only be detected by testing and running the program after completion.
Unreachable code is often confused with dead code, but the two are significantly different things. Dead code is code that executes a particular calculation, but then does nothing with the result of the calculation. This is not the same as code that is unreachable, which is never executed at all. Both unreachable and dead code can often be removed from a program without detriment, or they can be integrated back into the program by paying careful attention to the program's control flow. In addition, both unreachable and dead code can usually be detected during coding and before running a program.
Despite its seemingly dire definition, unreachable code is often relatively easy to fix. The presence of unreachable code always indicates a control flow problem, so careful analysis of a program's control flow will usually identify the offending code in short order. If a program's control flow is carefully monitored throughout the coding process, the likelihood of generating unreachable code is very low.