runtime: avoid division in growslice

Add a special case for power-of-2 sized elements.
We can replace div/mul with left/right shift and avoid expensive operation.
growslice is hotter for short slices of small elements, such as int16, so
add an int16 version for GrowSlice benchmark.

name                   old time/op  new time/op  delta
GrowSlice/Byte-6       61.3ns ± 3%  60.5ns ± 4%  -1.33%  (p=0.002 n=30+30)
GrowSlice/Int16-6      94.0ns ± 4%  84.7ns ± 2%  -9.82%  (p=0.000 n=30+30)
GrowSlice/Int-6         100ns ± 1%    99ns ± 1%  -0.25%  (p=0.032 n=29+28)
GrowSlice/Ptr-6         197ns ± 2%   195ns ± 2%  -0.94%  (p=0.001 n=30+29)
GrowSlice/Struct/24-6   168ns ± 1%   166ns ± 2%  -1.09%  (p=0.000 n=25+30)
GrowSlice/Struct/32-6   187ns ± 2%   180ns ± 1%  -3.59%  (p=0.000 n=30+30)
GrowSlice/Struct/40-6   241ns ± 2%   238ns ± 2%  -1.41%  (p=0.000 n=30+30)

Change-Id: I31e8388d73fd9356e2dcc091d8d92eef3e3ccdbc
Reviewed-on: https://go-review.googlesource.com/102279
Run-TryBot: Ilya Tocar <ilya.tocar@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>

1 files changed, 6 insertions, 0 deletions

diff --git a/src/runtime/slice_test.go b/src/runtime/slice_test.go
index ef1e812c0d..46db071ebe 100644
--- a/src/runtime/slice_test.go
+++ b/src/runtime/slice_test.go
@@ -31,6 +31,12 @@ func BenchmarkGrowSlice(b *testing.B) {
 			_ = append([]byte(nil), x...)
 		}
 	})
+	b.Run("Int16", func(b *testing.B) {
+		x := make([]int16, 9)
+		for i := 0; i < b.N; i++ {
+			_ = append([]int16(nil), x...)
+		}
+	})
 	b.Run("Int", func(b *testing.B) {
 		x := make([]int, 9)
 		for i := 0; i < b.N; i++ {