Use of floating-point data in Common C library functions can be inconsistent between different compilers and supporting libraries. For example, trigonometric functions in the C runtime library may return different values depending on the Linux version or distribution.