Add tests (#1)

.github/workflows/test.yml

@@ -0,0 +1,27 @@
+name: test
+
+on:
+  push:
+    branches: [ main ]
+  pull_request:
+    branches: [ main ]
+
+jobs:
+  example:
+    name: example
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v2
+      - uses: actions/setup-go@v4
+      - name: Install clang
+        uses: ConorMacBride/install-package@v1
+        with:
+          apt: clang libc6-dev-i386
+      - name: Install GOAT
+        run: go install .
+      - name: Run GOAT
+        run: goat src/avx_mul_to.c -O3 -mavx -mfma
+        working-directory: example
+      - name: Run tests
+        run: go test -v ./...
+        working-directory: example

README.md

@@ -6,7 +6,18 @@ It help to utilize optimization from C compiler in Go projects. For example, gen
 
 ## Install
 
+1. Install prerequisites
+
+```bash
+sudo apt update
+sudo apt install clang libc6-dev-i386
+```
+
+2. Install GoAT
+
+```bash
 go install github.com/gorse-io/goat@latest
+```
 
 ## Usage
 

example/avx_mul_to_test.go

@@ -0,0 +1,24 @@
+package example
+
+import (
+	"reflect"
+	"testing"
+	"unsafe"
+)
+
+func AVXMulTo(a, b, c []float32) {
+	if len(a) != len(b) || len(a) != len(c) {
+		panic("floats: slice lengths do not match")
+	}
+	avx_mul_to(unsafe.Pointer(&a[0]), unsafe.Pointer(&b[0]), unsafe.Pointer(&c[0]), unsafe.Pointer(uintptr(len(a))))
+}
+
+func TestAVXMulTo(t *testing.T) {
+	a := []float32{1, 2, 3, 4}
+	b := []float32{5, 6, 7, 8}
+	c := make([]float32, len(a))
+	AVXMulTo(a, b, c)
+	if !reflect.DeepEqual(c, []float32{5, 12, 21, 32}) {
+		t.Errorf("AVXMulTo(%v, %v, %v) = %v, want %v", a, b, c, c, []float32{5, 12, 21, 32})
+	}
+}

example/src/avx_mul_to.c

@@ -0,0 +1,22 @@
+#include <immintrin.h>
+#include <stdint.h>
+
+void avx_mul_to(float *a, float *b, float *c, int64_t n)
+{
+    int epoch = n / 8;
+    int remain = n % 8;
+    for (int i = 0; i < epoch; i++)
+    {
+        __m256 v1 = _mm256_loadu_ps(a);
+        __m256 v2 = _mm256_loadu_ps(b);
+        __m256 v = _mm256_mul_ps(v1, v2);
+        _mm256_storeu_ps(c, v);
+        a += 8;
+        b += 8;
+        c += 8;
+    }
+    for (int i = 0; i < remain; i++)
+    {
+        c[i] = a[i] * b[i];
+    }
+}