Skip to content
This repository was archived by the owner on Dec 22, 2021. It is now read-only.

Commit 4604d0d

Browse files
Maratyszczatlively
andauthored
Update proposals/simd/SIMD.md
Co-authored-by: Thomas Lively <[email protected]>
1 parent dd43218 commit 4604d0d

File tree

1 file changed

+11
-11
lines changed

1 file changed

+11
-11
lines changed

proposals/simd/SIMD.md

Lines changed: 11 additions & 11 deletions
Original file line numberDiff line numberDiff line change
@@ -900,30 +900,30 @@ Lane-wise IEEE `multiplication`.
900900

901901
Lane-wise IEEE `squareRoot`.
902902

903-
### Round to nearest integer, ties to even
904-
* `f32x4.nearest(a: v128) -> v128`
905-
* `f64x2.nearest(a: v128) -> v128`
903+
### Round to integer above (ceiling)
904+
* `f32x4.ceil(a: v128) -> v128`
905+
* `f64x2.ceil(a: v128) -> v128`
906906

907-
Lane-wise rounding to the nearest integral value; if two values are equally near, rounds to the even one.
907+
Lane-wise rounding to the nearest integral value not smaller than the input.
908908

909909
### Round to integer toward zero (truncate to integer)
910910
* `f32x4.trunc(a: v128) -> v128`
911911
* `f64x2.trunc(a: v128) -> v128`
912912

913-
Lane-wise rounding to the nearest integral value with the magniture not larger than the input.
914-
915-
### Round to integer above (ceiling)
916-
* `f32x4.ceil(a: v128) -> v128`
917-
* `f64x2.ceil(a: v128) -> v128`
918-
919-
Lane-wise rounding to the nearest integral value not smaller than the input.
913+
Lane-wise rounding to the nearest integral value with the magnitude not larger than the input.
920914

921915
### Round to integer below (floor)
922916
* `f32x4.floor(a: v128) -> v128`
923917
* `f64x2.floor(a: v128) -> v128`
924918

925919
Lane-wise rounding to the nearest integral value not greater than the input.
926920

921+
### Round to nearest integer, ties to even
922+
* `f32x4.nearest(a: v128) -> v128`
923+
* `f64x2.nearest(a: v128) -> v128`
924+
925+
Lane-wise rounding to the nearest integral value; if two values are equally near, rounds to the even one.
926+
927927
## Conversions
928928
### Integer to floating point
929929
* `f32x4.convert_i32x4_s(a: v128) -> v128`

0 commit comments

Comments
 (0)