int128

WARNING

int128 is not implemented in the current release.

Property	Value
Size	16 bytes (128 bits)
Minimum value	−2127 ≈ −1.701411 × 1038
Maximum value	2127−1 ≈ 1.701411 × 1038
Literal suffix	i128
Representation	Two's complement
Hardware support	Emulated as pairs of 64-bit instructions on x86-64 / AArch64 (≈2–4× slower)

int128 is an extended-precision signed integer — a fixed-width type whose size is identical on every target platform. Like all signed types it uses two's complement: the most significant bit is the sign bit. No mainstream CPU has native 128-bit arithmetic, so operations are lowered to pairs of 64-bit instructions and run several times slower than native types — reserve int128 for code that genuinely needs the extra range.

Literals

let a: int128 = 170141183460469231731687303715884105727;
let b = 42i128;  // type suffix

If a literal value does not fit in int128, the compiler emits an error at compile time. See Literals for decimal, hexadecimal, octal, and binary forms.

Typical Use Cases

• Extended-precision intermediate results
• UUID arithmetic
• Overflow-safe accumulation of 64-bit products

Arithmetic

Operator	Description	Compound
+	Addition	+=
-	Subtraction	-=
*	Multiplication	*=
/	Division	/=
%	Remainder	%=
**	Exponentiation	n/a

let a: int128 = 10;
let b: int128 = 3;

let sum  = a + b;   // 13
let quot = a / b;   // 3   truncates toward zero
let rem  = a % b;   // 1   carries the sign of the dividend

Division and remainder truncate toward zero, so the remainder takes the sign of the dividend: -7 % 2 is -1.

Overflow. In debug builds, signed overflow raises a fatal error. In release builds it wraps modulo 2¹²⁸ in two's-complement fashion. Wraparound is well-defined in Rux (unlike C/C++), but relying on it is almost always a logic error.

Comparison

Operator	Description	Result
==	Equal	bool
!=	Not equal	bool
<	Less than	bool
<=	Less than or equal	bool
>	Greater than	bool
>=	Greater than or equal	bool

Both operands must have the same type. Comparing int128 with another integer type — or with an unsigned type such as uint128 — is a compile-time error; cast one operand explicitly first.

Shift and Bitwise

Right shift on a signed value is arithmetic: vacated bits are filled with the sign bit, preserving the sign.

let s: int128 = -8;
let r = s >> 2;     // -2   sign-filled (arithmetic) right shift

The bitwise operators &, |, ^, and ~ are also defined (~x equals −(x + 1)), but bitmask work usually reads more clearly on unsigned types.

Conversion

Rux performs no implicit numeric conversions — every conversion uses the as operator. Widening to a wider signed type preserves the value through sign extension; narrowing keeps only the low-order bits; a same-width signed↔unsigned cast reinterprets the bit pattern.

let x: int128 = -1;
let small = x as int64;    // -1   narrowed, low 64 bits kept
let u     = x as uint128;  // 2^128 − 1   same width, bits reinterpreted

See Also

• uint128 — unsigned counterpart of the same width
• Unsigned Integer Types — the unsigned integer family