Return Code Bugs and Their Fixes in Function Testing

In this article we discuss two common return‑code bugs encountered during testing, illustrate how an unchanged default return value can mask exceptions, and show how incorrect return‑code checks can make error‑handling branches unreachable.

For the first bug, the function’s ret variable is initialized to QCONF_OK and never updated in error paths, causing the caller to treat an exception as a successful case; the solution is to assign distinct codes for each error branch.

The second bug arises from the caller checking for a literal -1 while the callee never returns -1, making the retry logic dead; the fix is to compare against the defined success code (e.g., QCONF_OK) and use consistent macros for all return values.

A summary table provides checking rules, typical pitfalls, and recommendations such as consistent initialization, exhaustive error‑code assignments, and unified macro usage.