12.2 Data Types

All 12 of the primary fixed-point types and their aliases, described in Section 4.1 “Overview and principles of the fixed-point data types” of ISO/IEC TR 18037:2008, are supported. Fixed-point data values contain fractional and optional integral parts. The format of fixed-point data in XC32 are as specified in the table below.

In the formats shown, s is the sign bit for a signed type (there is no sign bit for an unsigned type). The period character (.) is the specifier that separates the integral part and the fractional part. The numeric digits represent the number of bits in the integral part or in the fractional part.

The _Sat type modifier may be used with any type in the above table to indicate that values are saturated, as described in ISO/IEC TR 18037:2008. For example, _Sat short _Fract is the saturating form of short _Fract. Signed types saturate at the most negative and positive numbers representable in the corresponding format. Unsigned types saturate at zero and the most positive number representable in the format.

  #include <stdfix.h>
  void main () {
    int i;
    fract a[5] = {0.5,0.4,0.2,0.0,-0.1};
    fract b[5] = {0.1,0.8,0.6,0.5,-0.1};
    accum dp = 0;
    /* Calculate dot product of a[] and b[] */
    for (i=0; i<5; i++) {
        dp += a[i] * b[i];
    }
  }

The default behavior of overflow on signed or unsigned types is saturation. The pragmas described in Section 4.1.3 "Rounding and Overflow" of ISO/IEC TR 18037:2008 to control the rounding and overflow behavior are not supported.

The following table describes the fixed-point literal suffixes supported to form fixed-point literals of each type.