T0b1/wasmtime
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Implement SqmulRoundSat for interpreter

Browse Source 
Implemented `SqmulRoundSat` for the Cranelift interpreter, performing
QN-format fixed point multiplication for 16 and 32-bit integers in
SIMD vectors.

Copyright (c) 2021, Arm Limited
This commit is contained in:
dheaton-arm
2021-09-08 09:18:38 +01:00
parent 1a5a2c7c5d
commit 02ff19f2fc
3 changed files with 53 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
cranelift/filetests/filetests/runtests/simd-sqmulroundsat.clif Normal file
View File

@@ -0,0 +1,13 @@
test interpret
test run
target aarch64
set enable_simd
target x86_64
function %sqmulrs_i16x8(i16x8, i16x8) -> i16x8 {
block0(v0: i16x8, v1: i16x8):
v2 = sqmul_round_sat v0, v1
return v2
}
; run: %sqmulrs_i16x8([1 2 3 4 5 6 7 8], [1 10 100 1000 10000 15000 20000 25000]) == [0 0 0 0 2 3 4 6]
; run: %sqmulrs_i16x8([32767 32767 -32768 -32768 -32768 -32768 0 0], [32767 32767 -32768 -32768 32767 32767 0 0]) == [32766 32766 32767 32767 -32767 -32767 0 0]