10.4 Fixed-Point Data Types
The following table shows fixed-point data types that are supported by the
compiler when the -menable-fixed command line option is specified. See
Fixed-Point Arithmetic Support for more details
on the compiler's support for the fixed-point C language dialect. If the signed or unsigned
type specifier is not present, the type is assumed to be signed.
| Type | Bits | Min | Max |
|---|---|---|---|
_Fract | 16 | -1.0 | 1.0 - 2^-15 |
short _Fract | 16 | -1.0 | 1.0 - 2^-15 |
signed _Fract | 16 | -1.0 | 1.0 - 2^-15 |
signed short _Fract | 16 | -1.0 | 1.0 - 2^-15 |
unsigned _Fract | 16 | 0.0 | 1.0 - 2^-15 |
unsigned short
_Fract | 16 | 0.0 | 1.0 - 2^-15 |
long _Fract | 32 | -1.0 | 1.0 - 2^-31 |
signed long _Fract | 32 | -1.0 | 1.0 - 2^-31 |
unsigned long _Fract | 32 | 0.0 | 1.0 - 2^-31 |
_Accum | 40 | -256.0 | 256.0 - 2^-31 |
short _Accum | 40 | -256.0 | 256.0 - 2^-31 |
long _Accum | 40 | -256.0 | 256.0 - 2^-31 |
signed _Accum | 40 | -256.0 | 256.0 - 2^-31 |
signed short _Accum | 40 | -256.0 | 256.0 - 2^-31 |
signed long _Accum | 40 | -256.0 | 256.0 - 2^-31 |
unsigned _Accum | 40 | 0.0 | 256.0 - 2^-31 |
unsigned short
_Accum | 40 | 0.0 | 256.0 - 2^-31 |
unsigned long _Accum | 40 | 0.0 | 256.0 - 2^-31 |
As with integer and floating point data types, all fixed-point values are represented in a little endian format, which means:
- The Least Significant Byte (LSB) is stored at the lowest address
- The Least Significant bit (LSb) is stored at the lowest-numbered bit position