The shr_u numeric instruction, short for shift-right unsigned, is used for performing a bitwise right-shift on unsigned integers, similar to the >>> operator in other languages.
(module
(func (export "shift_right") (param $num i32) (param $by i32) (result i32)
;; load the number to shift and the by how many spots
local.get $num
local.get $by
;; shift and return the result
i32.shr_u
)
)const url = "{%wasm-url%}";
await WebAssembly.instantiateStreaming(fetch(url), { console }).then(
(result) => {
const shift_right = result.instance.exports.shift_right;
const res = shift_right(0b00000000_00000000_00000000_00000111, 1);
console.log(numToBin(res));
// Expected output: "00000000_00000000_00000000_00000011"
},
);
function numToBin(num) {
return (num >>> 0)
.toString(2)
.padStart(32, "0")
.match(/.{1,8}/g)
.join("_");
}value_type.shr_u
value_typeThe type of value the instruction is being run on. The following types support shr_u:
i32i64v128 interpretations:i8x16i16x8i32x4i64x2shr_uThe shr_u instruction. Must always be included after the value_type and a period (.).
[input, shift_value] -> [output]
inputThe input value.
shift_valueThe value that you want to shift the value by.
outputThe output value.
For a non-SIMD shr_u, the input and output will be basic numeric values such as 3 or 12.
For a SIMD shr_u, the input and output will be v128 value interpretations, for example i32x4 2 30 86 120. Each lane of the output pushed to the stack contains the corresponding lane in the input value shifted to the right by the specified shift_value.
| Instruction | Binary format | Example text => binary |
|---|---|---|
i32.shr_u | 0x76 | i32.shr_u => 0x76 |
i64.shr_u | 0x88 | i64.shr_u => 0x88 |
i8x16.shr_u | 0xfd 109:u32 | i8x16.shr_u => 0xfd 0x6d |
i16x8.shr_u | 0xfd 141:u32 | i16x8.shr_u => 0xfd 0x8d 0x01 |
i32x4.shr_u | 0xfd 173:u32 | i32x4.shr_u => 0xfd 0xad 0x01 |
i64x2.shr_u | 0xfd 205:u32 | i64x2.shr_u => 0xfd 0xcd 0x01 |
In this example, we demonstrate performing a right-shift on a SIMD value and outputting one of the lane values.
In our script, we grab a reference to a <p> element that we will output our result to, then define an object for import into Wasm containing a single function that writes a value to the output <p>. We then compile and instantiate our Wasm module using the WebAssembly.instantiateStreaming() method, importing the object in the process.
const outputElem = document.querySelector("p");
const obj = {
output(val) {
outputElem.textContent += val;
},
};
WebAssembly.instantiateStreaming(fetch("{%wasm-url%}"), {
obj,
});In our Wasm module, we first import the JavaScript output() function, making sure to declare that it has an i32 parameter. We then declare a SIMD i16x8 value, then right-shift it by 2 using i16x8.shr_u. Finally we extract the value stored in lane 6 of the output SIMD value using the extract_lane_u instruction, and output it to the DOM by calling the imported output() function.
(module
;; Import output function
(import "obj" "output" (func $output (param i32)))
(func $main
;; load two SIMD values onto the stack
v128.const i16x8 9 10 11 12 89 90 91 92
i32.const 2
i16x8.shr_u ;; shift-right by 2
i16x8.extract_lane_u 6 ;; Extract a value from the result
call $output
)
(start $main)
)The output is as follows:
The result is 22, because the value stored in lane 6 of the input value is 91. Once shifted right by two positions, the output value's lane 6 will contain the value 22.