mirror of
https://github.com/golang/go.git
synced 2025-12-08 06:10:04 +00:00
771a1dc216
For 512-bits they are unchanged. This CL adds the optimization rules for 128/256-bits under feature check. This CL also fixed a bug for masked load variant of instructions and make them zeroing by default as well. Change-Id: I6fe395541c0cd509984a81841420e71c3af732f2 Reviewed-on: https://go-review.googlesource.com/c/go/+/717822 Reviewed-by: David Chase <drchase@google.com> LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
69 lines
1.5 KiB
Go
69 lines
1.5 KiB
Go
// asmcheck
|
|
|
|
// Copyright 2025 The Go Authors. All rights reserved.
|
|
// Use of this source code is governed by a BSD-style
|
|
// license that can be found in the LICENSE file.
|
|
|
|
// These tests check code generation of simd peephole optimizations.
|
|
|
|
//go:build goexperiment.simd
|
|
|
|
package codegen
|
|
|
|
import "simd"
|
|
|
|
func vptest1() bool {
|
|
v1 := simd.LoadUint64x2Slice([]uint64{0, 1})
|
|
v2 := simd.LoadUint64x2Slice([]uint64{0, 0})
|
|
// amd64:`VPTEST\s(.*)(.*)$`
|
|
// amd64:`SETCS\s(.*)$`
|
|
return v1.AndNot(v2).IsZero()
|
|
}
|
|
|
|
func vptest2() bool {
|
|
v1 := simd.LoadUint64x2Slice([]uint64{0, 1})
|
|
v2 := simd.LoadUint64x2Slice([]uint64{0, 0})
|
|
// amd64:`VPTEST\s(.*)(.*)$`
|
|
// amd64:`SETEQ\s(.*)$`
|
|
return v1.And(v2).IsZero()
|
|
}
|
|
|
|
type Args2 struct {
|
|
V0 simd.Uint8x32
|
|
V1 simd.Uint8x32
|
|
x string
|
|
}
|
|
|
|
//go:noinline
|
|
func simdStructNoSpill(a Args2) simd.Uint8x32 {
|
|
// amd64:-`VMOVDQU\s.*$`
|
|
return a.V0.Xor(a.V1)
|
|
}
|
|
|
|
func simdStructWrapperNoSpill(a Args2) simd.Uint8x32 {
|
|
// amd64:-`VMOVDQU\s.*$`
|
|
a.x = "test"
|
|
return simdStructNoSpill(a)
|
|
}
|
|
|
|
//go:noinline
|
|
func simdArrayNoSpill(a [1]Args2) simd.Uint8x32 {
|
|
// amd64:-`VMOVDQU\s.*$`
|
|
return a[0].V0.Xor(a[0].V1)
|
|
}
|
|
|
|
func simdArrayWrapperNoSpill(a [1]Args2) simd.Uint8x32 {
|
|
// amd64:-`VMOVDQU\s.*$`
|
|
a[0].x = "test"
|
|
return simdArrayNoSpill(a)
|
|
}
|
|
|
|
func simdFeatureGuardedMaskOpt() simd.Int16x16 {
|
|
var x, y simd.Int16x16
|
|
if simd.HasAVX512() {
|
|
mask := simd.Mask16x16FromBits(5)
|
|
return x.Add(y).Masked(mask) // amd64:`VPADDW.Z\s.*$`
|
|
}
|
|
mask := simd.Mask16x16FromBits(5)
|
|
return x.Add(y).Masked(mask) // amd64:`VPAND\s.*$`
|
|
}
|