9.3 Floating-Point Data Types

The compiler uses the IEEE-754 format. The following table shows floating point data types that are supported. All floating point data types are arithmetic type real.

Table 9-2. Floating Point Data Types
Type	Bits	E Min	E Max	N Min	N Max
`float`	32	-126	127	2^-126	2¹²⁸
`double*`	32	-126	127	2^-126	2¹²⁸
`long double`	64	-1022	1023	2^-1022	2¹⁰²⁴
E = Exponent N = Normalized (approximate) * `double` is equivalent to `long double` if `-fno-short-double` is used.

All floating point values are specified in 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

As an example, the double value of 1.2345678 is stored at address 0x100 as follows:


`0x100`	`0x101`	`0x102`	`0X103`
`0x51`	`0x06`	`0x9E`	`0x3F`

As another example, the double value of 1.2345678 is stored in registers w4 and w5:


w4	w5
`0x0651`	`0x3F9E`

Floating-point types are always signed and the unsigned keyword is illegal when specifying a floating-point type. Preprocessor macros that specify valid ranges are available after including <float.h> in your source code. For information on implementation-defined behavior of floating point numbers, see section 22.6 Floating Point.