extend16_s: Wasm numeric instruction

Try it

(module
  (import "console" "log" (func $log (param i32)))
  (func $main

    i32.const 0xffff
    i32.extend16_s
    call $log

  )
  (start $main)
)

const url = "{%wasm-url%}";
await WebAssembly.instantiateStreaming(fetch(url), { console });

In this example, i32.const 0xffff pushes the value 0xffff (65535) onto the stack. This value represents -1 when read as a signed 16-bit number; i32.extend16_s sign-extends it to 0xffffffff — which is -1 represented as a full 32-bit signed integer.

WAT syntax

value_type.extend16_s <input> <output>

value_type

The type of value the instruction is being run on. The following types support extend16_s:

• i32
• i64

extend16_s

The extend16_s instruction. Must always be included after the value_type and a period (.).

Immediates

None.

Operand stack

[input: i32] -> [output: i32]
[input: i64] -> [output: i64]

input

The input i32 or i64 integer.

output

The output i32 or i64 integer.

Binary encoding

Description

Sign extension is useful because Wasm integers are a fixed width (32- or 64-bit), but you often want to work with smaller values — like an i8 or i16 — stored inside them. If you zero-pad a negative small value to fill the rest of the bits, you get the wrong number: the bit pattern that means -1 represented as a 16-bit value, for example, no longer represents -1 once you've zero-padded it.

The extend16_s instruction fixes this by taking the low 16 bits of the value, treating them as a signed 16-bit integer, and copying the sign bit (bit 15) up through all the remaining bits of the i32 or i64.

This is useful whenever you've loaded or produced an 16-bit value (say, from a two-byte value in memory) and need to use it correctly in arithmetic at the full 32- or 64-bit width.

Specifications

Browser compatibility

See also

• extend8_s
• extend32_s