Truncate (float to float)
The trunc instructions, short for truncate, are used for getting the value of a number without its fractional part.
trunc differs from floor when used on negative numbers, floor will round down in those cases while trunc will round up.
There's another trunc instruction that truncates the fractional part of a floating point and converts it to an integer.
Try it
Syntax
wasm
;; load a number onto the stack
f32.const 2.7
;; discard the fractional part (.7)
f32.trunc
;; the top item on the stack will now be 2
| Instruction | Binary opcode |
|---|---|
f32.trunc |
0x8f |
f64.trunc |
0x9d |