From 8f8b18645c1124003a53208098febb57b96d935f Mon Sep 17 00:00:00 2001
From: Andy Balholm <andy@balholm.com>
Date: Fri, 15 May 2020 10:43:19 -0700
Subject: [PATCH] Read multiple bytes in findMatchLengthWithLimit

Use 64- or 32-bit loads instead of reading a byte at a time. The
original C source did something like this, in a very C-ish way. It
needed to be simplified to translate it to Go. The exact way this works
was suggested by the assembly code in github.com/golang/snappy.
---
 brotli_test.go       |  3 ++-
 find_match_length.go | 28 ++++++++++++++++++++++++++++
 2 files changed, 30 insertions(+), 1 deletion(-)

diff --git a/brotli_test.go b/brotli_test.go
index 8c87979..c465694 100644
--- a/brotli_test.go
+++ b/brotli_test.go
@@ -461,8 +461,9 @@ func BenchmarkEncodeLevelsReset(b *testing.B) {
 		w := NewWriterLevel(buf, level)
 		w.Write(opticks)
 		w.Close()
-		b.Run(fmt.Sprintf("%d(%.1f%%)", level, float64(buf.Len())/float64(len(opticks))*100), func(b *testing.B) {
+		b.Run(fmt.Sprintf("%d", level), func(b *testing.B) {
 			b.ReportAllocs()
+			b.ReportMetric(float64(len(opticks))/float64(buf.Len()), "ratio")
 			b.SetBytes(int64(len(opticks)))
 			for i := 0; i < b.N; i++ {
 				w.Reset(ioutil.Discard)
diff --git a/find_match_length.go b/find_match_length.go
index 6730119..09d2ae6 100644
--- a/find_match_length.go
+++ b/find_match_length.go
@@ -1,5 +1,11 @@
 package brotli
 
+import (
+	"encoding/binary"
+	"math/bits"
+	"runtime"
+)
+
 /* Copyright 2010 Google Inc. All Rights Reserved.
 
    Distributed under MIT license.
@@ -10,6 +16,28 @@ package brotli
 func findMatchLengthWithLimit(s1 []byte, s2 []byte, limit uint) uint {
 	var matched uint = 0
 	_, _ = s1[limit-1], s2[limit-1] // bounds check
+	switch runtime.GOARCH {
+	case "amd64":
+		// Compare 8 bytes at at time.
+		for matched+8 <= limit {
+			w1 := binary.LittleEndian.Uint64(s1[matched:])
+			w2 := binary.LittleEndian.Uint64(s2[matched:])
+			if w1 != w2 {
+				return matched + uint(bits.TrailingZeros64(w1^w2)>>3)
+			}
+			matched += 8
+		}
+	case "386":
+		// Compare 4 bytes at at time.
+		for matched+4 <= limit {
+			w1 := binary.LittleEndian.Uint32(s1[matched:])
+			w2 := binary.LittleEndian.Uint32(s2[matched:])
+			if w1 != w2 {
+				return matched + uint(bits.TrailingZeros32(w1^w2)>>3)
+			}
+			matched += 4
+		}
+	}
 	for matched < limit && s1[matched] == s2[matched] {
 		matched++
 	}