cmd/internal/gc: use hardware instruction for math.Sqrt (amd64/arm)
I first prototyped this change in Sept 2011, and I discarded it
because it made no difference in the obvious benchmark loop.
It still makes no difference in the obvious benchmark loop,
but in a less obvious one, doing some extra computation
around the calls to Sqrt, not making the call does have a
significant effect.
benchmark old ns/op new ns/op delta
BenchmarkSqrt 4.56 4.57 +0.22%
BenchmarkSqrtIndirect 4.56 4.56 +0.00%
BenchmarkSqrtGo 69.4 69.4 +0.00%
BenchmarkSqrtPrime 4417 3647 -17.43%
This is a warmup for using hardware expansions for some
calls to 1-line assembly routines in the runtime (for example getg).
Change-Id: Ie66be23f8c09d0f7dc4ddd7ca8a93cfce28f55a4
Reviewed-on: https://go-review.googlesource.com/8356
Reviewed-by: Rob Pike <r@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
diff --git a/src/math/all_test.go b/src/math/all_test.go
index c07ac74..84061be 100644
--- a/src/math/all_test.go
+++ b/src/math/all_test.go
@@ -2977,15 +2977,56 @@
}
}
+var Global float64
+
func BenchmarkSqrt(b *testing.B) {
+ x, y := 0.0, 10.0
for i := 0; i < b.N; i++ {
- Sqrt(10)
+ x += Sqrt(y)
}
+ Global = x
+}
+
+func BenchmarkSqrtIndirect(b *testing.B) {
+ x, y := 0.0, 10.0
+ f := Sqrt
+ for i := 0; i < b.N; i++ {
+ x += f(y)
+ }
+ Global = x
}
func BenchmarkSqrtGo(b *testing.B) {
+ x, y := 0.0, 10.0
for i := 0; i < b.N; i++ {
- SqrtGo(10)
+ x += SqrtGo(y)
+ }
+ Global = x
+}
+
+func isPrime(i int) bool {
+ // Yes, this is a dumb way to write this code,
+ // but calling Sqrt repeatedly in this way demonstrates
+ // the benefit of using a direct SQRT instruction on systems
+ // that have one, whereas the obvious loop seems not to
+ // demonstrate such a benefit.
+ for j := 2; float64(j) <= Sqrt(float64(i)); j++ {
+ if i%j == 0 {
+ return false
+ }
+ }
+ return true
+}
+
+func BenchmarkSqrtPrime(b *testing.B) {
+ any := false
+ for i := 0; i < b.N; i++ {
+ if isPrime(100003) {
+ any = true
+ }
+ }
+ if any {
+ Global = 1
}
}
