libgo: Update to go1.6rc1.

    
    Reviewed-on: https://go-review.googlesource.com/19200


git-svn-id: svn+ssh://gcc.gnu.org/svn/gcc/trunk@233110 138bc75d-0d04-0410-961f-82ee72b054a4

4 files changed, 541 insertions, 200 deletions

diff --git a/libgo/go/unicode/example_test.go b/libgo/go/unicode/example_test.go
new file mode 100644
index 00000000000..50c5b18a48e
--- /dev/null
+++ b/libgo/go/unicode/example_test.go
@@ -0,0 +1,196 @@
+// Copyright 2015 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 unicode_test
+
+import (
+	"fmt"
+	"unicode"
+)
+
+// Functions starting with "Is" can be used to inspect which table of range a
+// rune belongs to. Note that runes may fit into more than one range.
+func Example_is() {
+
+	// constant with mixed type runes
+	const mixed = "\b5Ὂg̀9! ℃ᾭG"
+	for _, c := range mixed {
+		fmt.Printf("For %q:\n", c)
+		if unicode.IsControl(c) {
+			fmt.Println("\tis control rune")
+		}
+		if unicode.IsDigit(c) {
+			fmt.Println("\tis digit rune")
+		}
+		if unicode.IsGraphic(c) {
+			fmt.Println("\tis graphic rune")
+		}
+		if unicode.IsLetter(c) {
+			fmt.Println("\tis letter rune")
+		}
+		if unicode.IsLower(c) {
+			fmt.Println("\tis lower case rune")
+		}
+		if unicode.IsMark(c) {
+			fmt.Println("\tis mark rune")
+		}
+		if unicode.IsNumber(c) {
+			fmt.Println("\tis number rune")
+		}
+		if unicode.IsPrint(c) {
+			fmt.Println("\tis printable rune")
+		}
+		if !unicode.IsPrint(c) {
+			fmt.Println("\tis not printable rune")
+		}
+		if unicode.IsPunct(c) {
+			fmt.Println("\tis punct rune")
+		}
+		if unicode.IsSpace(c) {
+			fmt.Println("\tis space rune")
+		}
+		if unicode.IsSymbol(c) {
+			fmt.Println("\tis symbol rune")
+		}
+		if unicode.IsTitle(c) {
+			fmt.Println("\tis title case rune")
+		}
+		if unicode.IsUpper(c) {
+			fmt.Println("\tis upper case rune")
+		}
+	}
+
+	// Output:
+	// For '\b':
+	// 	is control rune
+	// 	is not printable rune
+	// For '5':
+	// 	is digit rune
+	// 	is graphic rune
+	// 	is number rune
+	// 	is printable rune
+	// For 'Ὂ':
+	// 	is graphic rune
+	// 	is letter rune
+	// 	is printable rune
+	// 	is upper case rune
+	// For 'g':
+	// 	is graphic rune
+	// 	is letter rune
+	// 	is lower case rune
+	// 	is printable rune
+	// For '̀':
+	// 	is graphic rune
+	// 	is mark rune
+	// 	is printable rune
+	// For '9':
+	// 	is digit rune
+	// 	is graphic rune
+	// 	is number rune
+	// 	is printable rune
+	// For '!':
+	// 	is graphic rune
+	// 	is printable rune
+	// 	is punct rune
+	// For ' ':
+	// 	is graphic rune
+	// 	is printable rune
+	// 	is space rune
+	// For '℃':
+	// 	is graphic rune
+	// 	is printable rune
+	// 	is symbol rune
+	// For 'ᾭ':
+	// 	is graphic rune
+	// 	is letter rune
+	// 	is printable rune
+	// 	is title case rune
+	// For 'G':
+	// 	is graphic rune
+	// 	is letter rune
+	// 	is printable rune
+	// 	is upper case rune
+}
+
+func ExampleSimpleFold() {
+	fmt.Printf("%#U\n", unicode.SimpleFold('A'))      // 'a'
+	fmt.Printf("%#U\n", unicode.SimpleFold('a'))      // 'A'
+	fmt.Printf("%#U\n", unicode.SimpleFold('K'))      // 'k'
+	fmt.Printf("%#U\n", unicode.SimpleFold('k'))      // '\u212A' (Kelvin symbol, K)
+	fmt.Printf("%#U\n", unicode.SimpleFold('\u212A')) // 'K'
+	fmt.Printf("%#U\n", unicode.SimpleFold('1'))      // '1'
+
+	// Output:
+	// U+0061 'a'
+	// U+0041 'A'
+	// U+006B 'k'
+	// U+212A 'K'
+	// U+004B 'K'
+	// U+0031 '1'
+}
+
+func ExampleTo() {
+	const lcG = 'g'
+	fmt.Printf("%#U\n", unicode.To(unicode.UpperCase, lcG))
+	fmt.Printf("%#U\n", unicode.To(unicode.LowerCase, lcG))
+	fmt.Printf("%#U\n", unicode.To(unicode.TitleCase, lcG))
+
+	const ucG = 'G'
+	fmt.Printf("%#U\n", unicode.To(unicode.UpperCase, ucG))
+	fmt.Printf("%#U\n", unicode.To(unicode.LowerCase, ucG))
+	fmt.Printf("%#U\n", unicode.To(unicode.TitleCase, ucG))
+
+	// Output:
+	// U+0047 'G'
+	// U+0067 'g'
+	// U+0047 'G'
+	// U+0047 'G'
+	// U+0067 'g'
+	// U+0047 'G'
+}
+
+func ExampleToLower() {
+	const ucG = 'G'
+	fmt.Printf("%#U\n", unicode.ToLower(ucG))
+
+	// Output:
+	// U+0067 'g'
+}
+func ExampleToTitle() {
+	const ucG = 'g'
+	fmt.Printf("%#U\n", unicode.ToTitle(ucG))
+
+	// Output:
+	// U+0047 'G'
+}
+
+func ExampleToUpper() {
+	const ucG = 'g'
+	fmt.Printf("%#U\n", unicode.ToUpper(ucG))
+
+	// Output:
+	// U+0047 'G'
+}
+
+func ExampleSpecialCase() {
+	t := unicode.TurkishCase
+
+	const lci = 'i'
+	fmt.Printf("%#U\n", t.ToLower(lci))
+	fmt.Printf("%#U\n", t.ToTitle(lci))
+	fmt.Printf("%#U\n", t.ToUpper(lci))
+
+	const uci = 'İ'
+	fmt.Printf("%#U\n", t.ToLower(uci))
+	fmt.Printf("%#U\n", t.ToTitle(uci))
+	fmt.Printf("%#U\n", t.ToUpper(uci))
+
+	// Output:
+	// U+0069 'i'
+	// U+0130 'İ'
+	// U+0130 'İ'
+	// U+0069 'i'
+	// U+0130 'İ'
+	// U+0130 'İ'
+}
diff --git a/libgo/go/unicode/tables.go b/libgo/go/unicode/tables.go
index 370a9d1174d..8bb42062f9e 100644
--- a/libgo/go/unicode/tables.go
+++ b/libgo/go/unicode/tables.go
@@ -53,7 +53,7 @@ var Categories = map[string]*RangeTable{
 
 var _C = &RangeTable{
 	R16: []Range16{
-		{0x0001, 0x001f, 1},
+		{0x0000, 0x001f, 1},
 		{0x007f, 0x009f, 1},
 		{0x00ad, 0x0600, 1363},
 		{0x0601, 0x0605, 1},
@@ -81,7 +81,7 @@ var _C = &RangeTable{
 
 var _Cc = &RangeTable{
 	R16: []Range16{
-		{0x0001, 0x001f, 1},
+		{0x0000, 0x001f, 1},
 		{0x007f, 0x009f, 1},
 	},
 	LatinOffset: 2,
diff --git a/libgo/go/unicode/utf8/utf8.go b/libgo/go/unicode/utf8/utf8.go
index 9ac37184d69..bbaf14aab8f 100644
--- a/libgo/go/unicode/utf8/utf8.go
+++ b/libgo/go/unicode/utf8/utf8.go
@@ -40,206 +40,208 @@ const (
 	rune1Max = 1<<7 - 1
 	rune2Max = 1<<11 - 1
 	rune3Max = 1<<16 - 1
-)
-
-func decodeRuneInternal(p []byte) (r rune, size int, short bool) {
-	n := len(p)
-	if n < 1 {
-		return RuneError, 0, true
-	}
-	c0 := p[0]
-
-	// 1-byte, 7-bit sequence?
-	if c0 < tx {
-		return rune(c0), 1, false
-	}
-
-	// unexpected continuation byte?
-	if c0 < t2 {
-		return RuneError, 1, false
-	}
-
-	// need first continuation byte
-	if n < 2 {
-		return RuneError, 1, true
-	}
-	c1 := p[1]
-	if c1 < tx || t2 <= c1 {
-		return RuneError, 1, false
-	}
-
-	// 2-byte, 11-bit sequence?
-	if c0 < t3 {
-		r = rune(c0&mask2)<<6 | rune(c1&maskx)
-		if r <= rune1Max {
-			return RuneError, 1, false
-		}
-		return r, 2, false
-	}
 
-	// need second continuation byte
-	if n < 3 {
-		return RuneError, 1, true
-	}
-	c2 := p[2]
-	if c2 < tx || t2 <= c2 {
-		return RuneError, 1, false
-	}
-
-	// 3-byte, 16-bit sequence?
-	if c0 < t4 {
-		r = rune(c0&mask3)<<12 | rune(c1&maskx)<<6 | rune(c2&maskx)
-		if r <= rune2Max {
-			return RuneError, 1, false
-		}
-		if surrogateMin <= r && r <= surrogateMax {
-			return RuneError, 1, false
-		}
-		return r, 3, false
-	}
-
-	// need third continuation byte
-	if n < 4 {
-		return RuneError, 1, true
-	}
-	c3 := p[3]
-	if c3 < tx || t2 <= c3 {
-		return RuneError, 1, false
-	}
-
-	// 4-byte, 21-bit sequence?
-	if c0 < t5 {
-		r = rune(c0&mask4)<<18 | rune(c1&maskx)<<12 | rune(c2&maskx)<<6 | rune(c3&maskx)
-		if r <= rune3Max || MaxRune < r {
-			return RuneError, 1, false
-		}
-		return r, 4, false
-	}
+	// The default lowest and highest continuation byte.
+	locb = 0x80 // 1000 0000
+	hicb = 0xBF // 1011 1111
+
+	// These names of these constants are chosen to give nice alignment in the
+	// table below. The first nibble is an index into acceptRanges or F for
+	// special one-byte cases. The second nibble is the Rune length or the
+	// Status for the special one-byte case.
+	xx = 0xF1 // invalid: size 1
+	as = 0xF0 // ASCII: size 1
+	s1 = 0x02 // accept 0, size 2
+	s2 = 0x13 // accept 1, size 3
+	s3 = 0x03 // accept 0, size 3
+	s4 = 0x23 // accept 2, size 3
+	s5 = 0x34 // accept 3, size 4
+	s6 = 0x04 // accept 0, size 4
+	s7 = 0x44 // accept 4, size 4
+)
 
-	// error
-	return RuneError, 1, false
+// first is information about the first byte in a UTF-8 sequence.
+var first = [256]uint8{
+	//   1   2   3   4   5   6   7   8   9   A   B   C   D   E   F
+	as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x00-0x0F
+	as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x10-0x1F
+	as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x20-0x2F
+	as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x30-0x3F
+	as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x40-0x4F
+	as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x50-0x5F
+	as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x60-0x6F
+	as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x70-0x7F
+	//   1   2   3   4   5   6   7   8   9   A   B   C   D   E   F
+	xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, // 0x80-0x8F
+	xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, // 0x90-0x9F
+	xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, // 0xA0-0xAF
+	xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, // 0xB0-0xBF
+	xx, xx, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, // 0xC0-0xCF
+	s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, // 0xD0-0xDF
+	s2, s3, s3, s3, s3, s3, s3, s3, s3, s3, s3, s3, s3, s4, s3, s3, // 0xE0-0xEF
+	s5, s6, s6, s6, s7, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, // 0xF0-0xFF
 }
 
-func decodeRuneInStringInternal(s string) (r rune, size int, short bool) {
-	n := len(s)
-	if n < 1 {
-		return RuneError, 0, true
-	}
-	c0 := s[0]
-
-	// 1-byte, 7-bit sequence?
-	if c0 < tx {
-		return rune(c0), 1, false
-	}
-
-	// unexpected continuation byte?
-	if c0 < t2 {
-		return RuneError, 1, false
-	}
-
-	// need first continuation byte
-	if n < 2 {
-		return RuneError, 1, true
-	}
-	c1 := s[1]
-	if c1 < tx || t2 <= c1 {
-		return RuneError, 1, false
-	}
-
-	// 2-byte, 11-bit sequence?
-	if c0 < t3 {
-		r = rune(c0&mask2)<<6 | rune(c1&maskx)
-		if r <= rune1Max {
-			return RuneError, 1, false
-		}
-		return r, 2, false
-	}
-
-	// need second continuation byte
-	if n < 3 {
-		return RuneError, 1, true
-	}
-	c2 := s[2]
-	if c2 < tx || t2 <= c2 {
-		return RuneError, 1, false
-	}
-
-	// 3-byte, 16-bit sequence?
-	if c0 < t4 {
-		r = rune(c0&mask3)<<12 | rune(c1&maskx)<<6 | rune(c2&maskx)
-		if r <= rune2Max {
-			return RuneError, 1, false
-		}
-		if surrogateMin <= r && r <= surrogateMax {
-			return RuneError, 1, false
-		}
-		return r, 3, false
-	}
-
-	// need third continuation byte
-	if n < 4 {
-		return RuneError, 1, true
-	}
-	c3 := s[3]
-	if c3 < tx || t2 <= c3 {
-		return RuneError, 1, false
-	}
-
-	// 4-byte, 21-bit sequence?
-	if c0 < t5 {
-		r = rune(c0&mask4)<<18 | rune(c1&maskx)<<12 | rune(c2&maskx)<<6 | rune(c3&maskx)
-		if r <= rune3Max || MaxRune < r {
-			return RuneError, 1, false
-		}
-		return r, 4, false
-	}
+// acceptRange gives the range of valid values for the second byte in a UTF-8
+// sequence.
+type acceptRange struct {
+	lo uint8 // lowest value for second byte.
+	hi uint8 // highest value for second byte.
+}
 
-	// error
-	return RuneError, 1, false
+var acceptRanges = [...]acceptRange{
+	0: {locb, hicb},
+	1: {0xA0, hicb},
+	2: {locb, 0x9F},
+	3: {0x90, hicb},
+	4: {locb, 0x8F},
 }
 
 // FullRune reports whether the bytes in p begin with a full UTF-8 encoding of a rune.
 // An invalid encoding is considered a full Rune since it will convert as a width-1 error rune.
 func FullRune(p []byte) bool {
-	_, _, short := decodeRuneInternal(p)
-	return !short
+	n := len(p)
+	if n == 0 {
+		return false
+	}
+	x := first[p[0]]
+	if n >= int(x&7) {
+		return true // ASCII, invalid or valid.
+	}
+	// Must be short or invalid.
+	accept := acceptRanges[x>>4]
+	if n > 1 {
+		if c := p[1]; c < accept.lo || accept.hi < c {
+			return true
+		} else if n > 2 && (p[2] < locb || hicb < p[2]) {
+			return true
+		}
+	}
+	return false
 }
 
 // FullRuneInString is like FullRune but its input is a string.
 func FullRuneInString(s string) bool {
-	_, _, short := decodeRuneInStringInternal(s)
-	return !short
+	n := len(s)
+	if n == 0 {
+		return false
+	}
+	x := first[s[0]]
+	if n >= int(x&7) {
+		return true // ASCII, invalid, or valid.
+	}
+	// Must be short or invalid.
+	accept := acceptRanges[x>>4]
+	if n > 1 {
+		if c := s[1]; c < accept.lo || accept.hi < c {
+			return true
+		} else if n > 2 && (s[2] < locb || hicb < s[2]) {
+			return true
+		}
+	}
+	return false
 }
 
 // DecodeRune unpacks the first UTF-8 encoding in p and returns the rune and
 // its width in bytes. If p is empty it returns (RuneError, 0). Otherwise, if
 // the encoding is invalid, it returns (RuneError, 1). Both are impossible
-// results for correct UTF-8.
+// results for correct, non-empty UTF-8.
 //
 // An encoding is invalid if it is incorrect UTF-8, encodes a rune that is
 // out of range, or is not the shortest possible UTF-8 encoding for the
 // value. No other validation is performed.
 func DecodeRune(p []byte) (r rune, size int) {
-	r, size, _ = decodeRuneInternal(p)
-	return
+	n := len(p)
+	if n < 1 {
+		return RuneError, 0
+	}
+	p0 := p[0]
+	x := first[p0]
+	if x >= as {
+		// The following code simulates an additional check for x == xx and
+		// handling the ASCII and invalid cases accordingly. This mask-and-or
+		// approach prevents an additional branch.
+		mask := rune(x) << 31 >> 31 // Create 0x0000 or 0xFFFF.
+		return rune(p[0])&^mask | RuneError&mask, 1
+	}
+	sz := x & 7
+	accept := acceptRanges[x>>4]
+	if n < int(sz) {
+		return RuneError, 1
+	}
+	b1 := p[1]
+	if b1 < accept.lo || accept.hi < b1 {
+		return RuneError, 1
+	}
+	if sz == 2 {
+		return rune(p0&mask2)<<6 | rune(b1&maskx), 2
+	}
+	b2 := p[2]
+	if b2 < locb || hicb < b2 {
+		return RuneError, 1
+	}
+	if sz == 3 {
+		return rune(p0&mask3)<<12 | rune(b1&maskx)<<6 | rune(b2&maskx), 3
+	}
+	b3 := p[3]
+	if b3 < locb || hicb < b3 {
+		return RuneError, 1
+	}
+	return rune(p0&mask4)<<18 | rune(b1&maskx)<<12 | rune(b2&maskx)<<6 | rune(b3&maskx), 4
 }
 
 // DecodeRuneInString is like DecodeRune but its input is a string. If s is
 // empty it returns (RuneError, 0). Otherwise, if the encoding is invalid, it
-// returns (RuneError, 1). Both are impossible results for correct UTF-8.
+// returns (RuneError, 1). Both are impossible results for correct, non-empty
+// UTF-8.
 //
 // An encoding is invalid if it is incorrect UTF-8, encodes a rune that is
 // out of range, or is not the shortest possible UTF-8 encoding for the
 // value. No other validation is performed.
 func DecodeRuneInString(s string) (r rune, size int) {
-	r, size, _ = decodeRuneInStringInternal(s)
-	return
+	n := len(s)
+	if n < 1 {
+		return RuneError, 0
+	}
+	s0 := s[0]
+	x := first[s0]
+	if x >= as {
+		// The following code simulates an additional check for x == xx and
+		// handling the ASCII and invalid cases accordingly. This mask-and-or
+		// approach prevents an additional branch.
+		mask := rune(x) << 31 >> 31 // Create 0x0000 or 0xFFFF.
+		return rune(s[0])&^mask | RuneError&mask, 1
+	}
+	sz := x & 7
+	accept := acceptRanges[x>>4]
+	if n < int(sz) {
+		return RuneError, 1
+	}
+	s1 := s[1]
+	if s1 < accept.lo || accept.hi < s1 {
+		return RuneError, 1
+	}
+	if sz == 2 {
+		return rune(s0&mask2)<<6 | rune(s1&maskx), 2
+	}
+	s2 := s[2]
+	if s2 < locb || hicb < s2 {
+		return RuneError, 1
+	}
+	if sz == 3 {
+		return rune(s0&mask3)<<12 | rune(s1&maskx)<<6 | rune(s2&maskx), 3
+	}
+	s3 := s[3]
+	if s3 < locb || hicb < s3 {
+		return RuneError, 1
+	}
+	return rune(s0&mask4)<<18 | rune(s1&maskx)<<12 | rune(s2&maskx)<<6 | rune(s3&maskx), 4
 }
 
 // DecodeLastRune unpacks the last UTF-8 encoding in p and returns the rune and
 // its width in bytes. If p is empty it returns (RuneError, 0). Otherwise, if
 // the encoding is invalid, it returns (RuneError, 1). Both are impossible
-// results for correct UTF-8.
+// results for correct, non-empty UTF-8.
 //
 // An encoding is invalid if it is incorrect UTF-8, encodes a rune that is
 // out of range, or is not the shortest possible UTF-8 encoding for the
@@ -278,7 +280,8 @@ func DecodeLastRune(p []byte) (r rune, size int) {
 
 // DecodeLastRuneInString is like DecodeLastRune but its input is a string. If
 // s is empty it returns (RuneError, 0). Otherwise, if the encoding is invalid,
-// it returns (RuneError, 1). Both are impossible results for correct UTF-8.
+// it returns (RuneError, 1). Both are impossible results for correct,
+// non-empty UTF-8.
 //
 // An encoding is invalid if it is incorrect UTF-8, encodes a rune that is
 // out of range, or is not the shortest possible UTF-8 encoding for the
@@ -367,65 +370,141 @@ func EncodeRune(p []byte, r rune) int {
 // RuneCount returns the number of runes in p.  Erroneous and short
 // encodings are treated as single runes of width 1 byte.
 func RuneCount(p []byte) int {
-	i := 0
+	np := len(p)
 	var n int
-	for n = 0; i < len(p); n++ {
-		if p[i] < RuneSelf {
+	for i := 0; i < np; {
+		n++
+		c := p[i]
+		if c < RuneSelf {
+			// ASCII fast path
 			i++
-		} else {
-			_, size := DecodeRune(p[i:])
-			i += size
+			continue
+		}
+		x := first[c]
+		if x == xx {
+			i++ // invalid.
+			continue
+		}
+		size := int(x & 7)
+		if i+size > np {
+			i++ // Short or invalid.
+			continue
 		}
+		accept := acceptRanges[x>>4]
+		if c := p[i+1]; c < accept.lo || accept.hi < c {
+			size = 1
+		} else if size == 2 {
+		} else if c := p[i+2]; c < locb || hicb < c {
+			size = 1
+		} else if size == 3 {
+		} else if c := p[i+3]; c < locb || hicb < c {
+			size = 1
+		}
+		i += size
 	}
 	return n
 }
 
 // RuneCountInString is like RuneCount but its input is a string.
 func RuneCountInString(s string) (n int) {
-	for range s {
-		n++
+	ns := len(s)
+	for i := 0; i < ns; n++ {
+		c := s[i]
+		if c < RuneSelf {
+			// ASCII fast path
+			i++
+			continue
+		}
+		x := first[c]
+		if x == xx {
+			i++ // invalid.
+			continue
+		}
+		size := int(x & 7)
+		if i+size > ns {
+			i++ // Short or invalid.
+			continue
+		}
+		accept := acceptRanges[x>>4]
+		if c := s[i+1]; c < accept.lo || accept.hi < c {
+			size = 1
+		} else if size == 2 {
+		} else if c := s[i+2]; c < locb || hicb < c {
+			size = 1
+		} else if size == 3 {
+		} else if c := s[i+3]; c < locb || hicb < c {
+			size = 1
+		}
+		i += size
 	}
-	return
+	return n
 }
 
-// RuneStart reports whether the byte could be the first byte of
-// an encoded rune.  Second and subsequent bytes always have the top
-// two bits set to 10.
+// RuneStart reports whether the byte could be the first byte of an encoded,
+// possibly invalid rune.  Second and subsequent bytes always have the top two
+// bits set to 10.
 func RuneStart(b byte) bool { return b&0xC0 != 0x80 }
 
 // Valid reports whether p consists entirely of valid UTF-8-encoded runes.
 func Valid(p []byte) bool {
-	i := 0
-	for i < len(p) {
-		if p[i] < RuneSelf {
+	n := len(p)
+	for i := 0; i < n; {
+		pi := p[i]
+		if pi < RuneSelf {
 			i++
-		} else {
-			_, size := DecodeRune(p[i:])
-			if size == 1 {
-				// All valid runes of size 1 (those
-				// below RuneSelf) were handled above.
-				// This must be a RuneError.
-				return false
-			}
-			i += size
+			continue
+		}
+		x := first[pi]
+		if x == xx {
+			return false // Illegal starter byte.
 		}
+		size := int(x & 7)
+		if i+size > n {
+			return false // Short or invalid.
+		}
+		accept := acceptRanges[x>>4]
+		if c := p[i+1]; c < accept.lo || accept.hi < c {
+			return false
+		} else if size == 2 {
+		} else if c := p[i+2]; c < locb || hicb < c {
+			return false
+		} else if size == 3 {
+		} else if c := p[i+3]; c < locb || hicb < c {
+			return false
+		}
+		i += size
 	}
 	return true
 }
 
 // ValidString reports whether s consists entirely of valid UTF-8-encoded runes.
 func ValidString(s string) bool {
-	for i, r := range s {
-		if r == RuneError {
-			// The RuneError value can be an error
-			// sentinel value (if it's size 1) or the same
-			// value encoded properly. Decode it to see if
-			// it's the 1 byte sentinel value.
-			_, size := DecodeRuneInString(s[i:])
-			if size == 1 {
-				return false
-			}
+	n := len(s)
+	for i := 0; i < n; {
+		si := s[i]
+		if si < RuneSelf {
+			i++
+			continue
+		}
+		x := first[si]
+		if x == xx {
+			return false // Illegal starter byte.
+		}
+		size := int(x & 7)
+		if i+size > n {
+			return false // Short or invalid.
+		}
+		accept := acceptRanges[x>>4]
+		if c := s[i+1]; c < accept.lo || accept.hi < c {
+			return false
+		} else if size == 2 {
+		} else if c := s[i+2]; c < locb || hicb < c {
+			return false
+		} else if size == 3 {
+		} else if c := s[i+3]; c < locb || hicb < c {
+			return false
 		}
+		i += size
 	}
 	return true
 }
diff --git a/libgo/go/unicode/utf8/utf8_test.go b/libgo/go/unicode/utf8/utf8_test.go
index 758d7a0f8e9..51571b61eb9 100644
--- a/libgo/go/unicode/utf8/utf8_test.go
+++ b/libgo/go/unicode/utf8/utf8_test.go
@@ -100,6 +100,15 @@ func TestFullRune(t *testing.T) {
 			t.Errorf("FullRune(%q) = true, want false", s1)
 		}
 	}
+	for _, s := range []string{"\xc0", "\xc1"} {
+		b := []byte(s)
+		if !FullRune(b) {
+			t.Errorf("FullRune(%q) = false, want true", s)
+		}
+		if !FullRuneInString(s) {
+			t.Errorf("FullRuneInString(%q) = false, want true", s)
+		}
+	}
 }
 
 func TestEncodeRune(t *testing.T) {
@@ -300,6 +309,8 @@ var runecounttests = []RuneCountTest{
 	{"☺☻☹", 3},
 	{"1,2,3,4", 7},
 	{"\xe2\x00", 2},
+	{"\xe2\x80", 2},
+	{"a\xe2\x80", 3},
 }
 
 func TestRuneCount(t *testing.T) {
@@ -352,6 +363,7 @@ var validTests = []ValidTest{
 	{"ЖЖ", true},
 	{"брэд-ЛГТМ", true},
 	{"☺☻☹", true},
+	{"aa\xe2", false},
 	{string([]byte{66, 250}), false},
 	{string([]byte{66, 250, 67}), false},
 	{"a\uFFFDb", true},
@@ -404,17 +416,57 @@ func TestValidRune(t *testing.T) {
 }
 
 func BenchmarkRuneCountTenASCIIChars(b *testing.B) {
+	s := []byte("0123456789")
 	for i := 0; i < b.N; i++ {
-		RuneCountInString("0123456789")
+		RuneCount(s)
 	}
 }
 
 func BenchmarkRuneCountTenJapaneseChars(b *testing.B) {
+	s := []byte("日本語日本語日本語日")
+	for i := 0; i < b.N; i++ {
+		RuneCount(s)
+	}
+}
+
+func BenchmarkRuneCountInStringTenASCIIChars(b *testing.B) {
+	for i := 0; i < b.N; i++ {
+		RuneCountInString("0123456789")
+	}
+}
+
+func BenchmarkRuneCountInStringTenJapaneseChars(b *testing.B) {
 	for i := 0; i < b.N; i++ {
 		RuneCountInString("日本語日本語日本語日")
 	}
 }
 
+func BenchmarkValidTenASCIIChars(b *testing.B) {
+	s := []byte("0123456789")
+	for i := 0; i < b.N; i++ {
+		Valid(s)
+	}
+}
+
+func BenchmarkValidTenJapaneseChars(b *testing.B) {
+	s := []byte("日本語日本語日本語日")
+	for i := 0; i < b.N; i++ {
+		Valid(s)
+	}
+}
+
+func BenchmarkValidStringTenASCIIChars(b *testing.B) {
+	for i := 0; i < b.N; i++ {
+		ValidString("0123456789")
+	}
+}
+
+func BenchmarkValidStringTenJapaneseChars(b *testing.B) {
+	for i := 0; i < b.N; i++ {
+		ValidString("日本語日本語日本語日")
+	}
+}
+
 func BenchmarkEncodeASCIIRune(b *testing.B) {
 	buf := make([]byte, UTFMax)
 	for i := 0; i < b.N; i++ {
@@ -442,3 +494,17 @@ func BenchmarkDecodeJapaneseRune(b *testing.B) {
 		DecodeRune(nihon)
 	}
 }
+
+func BenchmarkFullASCIIRune(b *testing.B) {
+	a := []byte{'a'}
+	for i := 0; i < b.N; i++ {
+		FullRune(a)
+	}
+}
+
+func BenchmarkFullJapaneseRune(b *testing.B) {
+	nihon := []byte("本")
+	for i := 0; i < b.N; i++ {
+		FullRune(nihon)
+	}
+}