2.3. Math bit helpers
The MATH_BITS module provides bit manipulation functions for floating point
numbers, including type punning between integer and float representations,
and efficient integer math operations like int_bits_to_float and
float_bits_to_int.
All functions and symbols are in “math_bits” module, use require to get access to it.
require daslib/math_bits
Example:
require daslib/math_bits
[export]
def main() {
let f = uint_bits_to_float(0x3F800000u)
print("uint_bits_to_float(0x3F800000) = {f}\n")
let back = float_bits_to_uint(1.0)
print("float_bits_to_uint(1.0) = {back}\n")
}
// output:
// uint_bits_to_float(0x3F800000) = 1
// float_bits_to_uint(1.0) = 0x3f800000
2.3.1. float in int,uint
2.3.1.1. int_bits_to_float
- int_bits_to_float(x: int ): float
bit representation of x is interpreted as a float
- Arguments:
x : int
- int_bits_to_float(x: int2 ): float2
- int_bits_to_float(x: int3 ): float3
- int_bits_to_float(x: int4 ): float4
2.3.1.2. uint_bits_to_float
- uint_bits_to_float(x: uint ): float
bit representation of x is interpreted as a float
- Arguments:
x : uint
- uint_bits_to_float(x: uint2 ): float2
- uint_bits_to_float(x: uint3 ): float3
- uint_bits_to_float(x: uint4 ): float4
2.3.2. int,uint in float
2.3.2.1. float_bits_to_int
- float_bits_to_int(x: float ): int
bit representation of x is interpreted as a int
- Arguments:
x : float
- float_bits_to_int(x: float2 ): int2
- float_bits_to_int(x: float3 ): int3
- float_bits_to_int(x: float4 ): int4
2.3.2.2. float_bits_to_uint
- float_bits_to_uint(x: float ): uint
bit representation of x is interpreted as a uint
- Arguments:
x : float
- float_bits_to_uint(x: float2 ): uint2
- float_bits_to_uint(x: float3 ): uint3
- float_bits_to_uint(x: float4 ): uint4
2.3.3. int64,uint64 in double
- double_bits_to_int64(x: double ): int64
bit representation of x is interpreted as a int64
- Arguments:
x : double
- double_bits_to_uint64(x: double ): uint64
bit representation of x is interpreted as a uint64
- Arguments:
x : double
- int64_bits_to_double(x: int64 ): double
bit representation of x is interpreted as a double
- Arguments:
x : int64
- uint64_bits_to_double(x: uint64 ): double
bit representation of x is interpreted as a double
- Arguments:
x : uint64
2.3.4. bit-cast vec4f
- cast_to_int16(data: float4 ): int16
return an int16 which was bit-cast from x
- Arguments:
data : float4
- cast_to_int32(data: float4 ): int
return an int32 which was bit-cast from x
- Arguments:
data : float4
- cast_to_int64(data: float4 ): int64
return an int64 which was bit-cast from x
- Arguments:
data : float4
- cast_to_int8(data: float4 ): int8
return an int8 which was bit-cast from x
- Arguments:
data : float4
- cast_to_pointer(data: float4 ): void?
return a pointer which was bit-cast from x
- Arguments:
data : float4
- cast_to_string(data: float4 ): string
return a string which pointer was bit-cast from x
- Arguments:
data : float4
2.3.4.1. cast_to_vec4f
- cast_to_vec4f(x: bool ): float4
return a float4 which stores bit-cast version of x
- Arguments:
x : bool
- cast_to_vec4f(x: int64 ): float4