CWE-758

Class Abstraction Level
Pillar — Highest-level weakness category
Class — Abstract, language-independent
Base — Specific enough to detect
Variant — Tied to specific technology
Compound — Requires multiple weaknesses Incomplete MITRE CWE Status
Stable — Fully reviewed and complete
Draft — Under development, may change
Incomplete — Partially defined by MITRE
Deprecated — No longer recommended
Obsolete — Replaced by another CWE
Reliance on Undefined, Unspecified, or Implementation-Defined Behavior

Description

The product uses an API function, data structure, or other entity in a way that relies on properties that are not always guaranteed to hold for that entity.

This can lead to resultant weaknesses when the required properties change, such as when the product is ported to a different platform or if an interaction error (CWE-435) occurs.

Consequences

Other — Reduce Maintainability, Unexpected State, Quality Degradation

Detection

Fuzzing

Fuzz testing (fuzzing) is a powerful technique for generating large numbers of diverse inputs - either randomly or algorithmically - and dynamically invoking the code with those inputs. Even with random inputs, it is often capable of generating unexpected results such as crashes, memory corruption, or resource consumption. Fuzzing effectively produces repeatable test cases that clearly indicate bugs, which helps developers to diagnose the issues.