5.1.3.2 Floating-point Types

The specification of the floating-point types, especially the double and long double types can vary from that mandated by the C standard. For example, the DBL_DIG/LDBL_DIG macro values specified by <float.h> are smaller than that indicated by the standard. Similarly the DBL_EPSILON/LDBL_EPSILON macro values are larger than that indicated by the standard.

These macros are listed in the Microchip Unified Standard Library Reference Guide.