mirror of
https://github.com/golang/go
synced 2024-11-17 14:14:56 -07:00
0301c6c351
Use a small python script to consolidate duplicate
ppc64/ppc64le tests into a single ppc64x codegen test.
This makes small assumption that anytime two tests with
for different arch/variant combos exists, those tests
can be combined into a single ppc64x test.
E.x:
// ppc64le: foo
// ppc64le/power9: foo
into
// ppc64x: foo
or
// ppc64: foo
// ppc64le: foo
into
// ppc64x: foo
import glob
import re
files = glob.glob("codegen/*.go")
for file in files:
with open(file) as f:
text = [l for l in f]
i = 0
while i < len(text):
first = re.match("\s*// ?ppc64(le)?(/power[89])?:(.*)", text[i])
if first:
j = i+1
while j < len(text):
second = re.match("\s*// ?ppc64(le)?(/power[89])?:(.*)", text[j])
if not second:
break
if (not first.group(2) or first.group(2) == second.group(2)) and first.group(3) == second.group(3):
text[i] = re.sub(" ?ppc64(le|x)?"," ppc64x",text[i])
text=text[:j] + (text[j+1:])
else:
j += 1
i+=1
with open(file, 'w') as f:
f.write("".join(text))
Change-Id: Ic6b009b54eacaadc5a23db9c5a3bf7331b595821
Reviewed-on: https://go-review.googlesource.com/c/go/+/463220
Reviewed-by: Cherry Mui <cherryyz@google.com>
Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com>
Reviewed-by: Bryan Mills <bcmills@google.com>
Run-TryBot: Paul Murphy <murp@ibm.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
149 lines
3.0 KiB
Go
149 lines
3.0 KiB
Go
// asmcheck
|
|
|
|
// Copyright 2018 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.
|
|
|
|
package codegen
|
|
|
|
// This file contains codegen tests related to arithmetic
|
|
// simplifications and optimizations on float types.
|
|
// For codegen tests on integer types, see arithmetic.go.
|
|
|
|
// --------------------- //
|
|
// Strength-reduce //
|
|
// --------------------- //
|
|
|
|
func Mul2(f float64) float64 {
|
|
// 386/sse2:"ADDSD",-"MULSD"
|
|
// amd64:"ADDSD",-"MULSD"
|
|
// arm/7:"ADDD",-"MULD"
|
|
// arm64:"FADDD",-"FMULD"
|
|
// ppc64x:"FADD",-"FMUL"
|
|
return f * 2.0
|
|
}
|
|
|
|
func DivPow2(f1, f2, f3 float64) (float64, float64, float64) {
|
|
// 386/sse2:"MULSD",-"DIVSD"
|
|
// amd64:"MULSD",-"DIVSD"
|
|
// arm/7:"MULD",-"DIVD"
|
|
// arm64:"FMULD",-"FDIVD"
|
|
// ppc64x:"FMUL",-"FDIV"
|
|
x := f1 / 16.0
|
|
|
|
// 386/sse2:"MULSD",-"DIVSD"
|
|
// amd64:"MULSD",-"DIVSD"
|
|
// arm/7:"MULD",-"DIVD"
|
|
// arm64:"FMULD",-"FDIVD"
|
|
// ppc64x:"FMUL",-"FDIVD"
|
|
y := f2 / 0.125
|
|
|
|
// 386/sse2:"ADDSD",-"DIVSD",-"MULSD"
|
|
// amd64:"ADDSD",-"DIVSD",-"MULSD"
|
|
// arm/7:"ADDD",-"MULD",-"DIVD"
|
|
// arm64:"FADDD",-"FMULD",-"FDIVD"
|
|
// ppc64x:"FADD",-"FMUL",-"FDIV"
|
|
z := f3 / 0.5
|
|
|
|
return x, y, z
|
|
}
|
|
|
|
func indexLoad(b0 []float32, b1 float32, idx int) float32 {
|
|
// arm64:`FMOVS\s\(R[0-9]+\)\(R[0-9]+<<2\),\sF[0-9]+`
|
|
return b0[idx] * b1
|
|
}
|
|
|
|
func indexStore(b0 []float64, b1 float64, idx int) {
|
|
// arm64:`FMOVD\sF[0-9]+,\s\(R[0-9]+\)\(R[0-9]+<<3\)`
|
|
b0[idx] = b1
|
|
}
|
|
|
|
// ----------- //
|
|
// Fused //
|
|
// ----------- //
|
|
|
|
func FusedAdd32(x, y, z float32) float32 {
|
|
// s390x:"FMADDS\t"
|
|
// ppc64x:"FMADDS\t"
|
|
// arm64:"FMADDS"
|
|
return x*y + z
|
|
}
|
|
|
|
func FusedSub32_a(x, y, z float32) float32 {
|
|
// s390x:"FMSUBS\t"
|
|
// ppc64x:"FMSUBS\t"
|
|
return x*y - z
|
|
}
|
|
|
|
func FusedSub32_b(x, y, z float32) float32 {
|
|
// arm64:"FMSUBS"
|
|
return z - x*y
|
|
}
|
|
|
|
func FusedAdd64(x, y, z float64) float64 {
|
|
// s390x:"FMADD\t"
|
|
// ppc64x:"FMADD\t"
|
|
// arm64:"FMADDD"
|
|
return x*y + z
|
|
}
|
|
|
|
func FusedSub64_a(x, y, z float64) float64 {
|
|
// s390x:"FMSUB\t"
|
|
// ppc64x:"FMSUB\t"
|
|
return x*y - z
|
|
}
|
|
|
|
func FusedSub64_b(x, y, z float64) float64 {
|
|
// arm64:"FMSUBD"
|
|
return z - x*y
|
|
}
|
|
|
|
func Cmp(f float64) bool {
|
|
// arm64:"FCMPD","(BGT|BLE|BMI|BPL)",-"CSET\tGT",-"CBZ"
|
|
return f > 4 || f < -4
|
|
}
|
|
|
|
func CmpZero64(f float64) bool {
|
|
// s390x:"LTDBR",-"FCMPU"
|
|
return f <= 0
|
|
}
|
|
|
|
func CmpZero32(f float32) bool {
|
|
// s390x:"LTEBR",-"CEBR"
|
|
return f <= 0
|
|
}
|
|
|
|
func CmpWithSub(a float64, b float64) bool {
|
|
f := a - b
|
|
// s390x:-"LTDBR"
|
|
return f <= 0
|
|
}
|
|
|
|
func CmpWithAdd(a float64, b float64) bool {
|
|
f := a + b
|
|
// s390x:-"LTDBR"
|
|
return f <= 0
|
|
}
|
|
|
|
// ---------------- //
|
|
// Non-floats //
|
|
// ---------------- //
|
|
|
|
// We should make sure that the compiler doesn't generate floating point
|
|
// instructions for non-float operations on Plan 9, because floating point
|
|
// operations are not allowed in the note handler.
|
|
|
|
func ArrayZero() [16]byte {
|
|
// amd64:"MOVUPS"
|
|
// plan9/amd64/:-"MOVUPS"
|
|
var a [16]byte
|
|
return a
|
|
}
|
|
|
|
func ArrayCopy(a [16]byte) (b [16]byte) {
|
|
// amd64:"MOVUPS"
|
|
// plan9/amd64/:-"MOVUPS"
|
|
b = a
|
|
return
|
|
}
|