labels: faster Compare function when using -tags stringlabels (#12451)

Instead of unpacking every individual string, we skip to the point
where there is a difference, going 8 bytes at a time where possible.

Add benchmark for Compare; extend tests too.

---------

Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
Co-authored-by: Oleg Zaytsev <mail@olegzaytsev.com>

model/labels/labels_stringlabels.go

@@ -422,37 +422,49 @@ func FromStrings(ss ...string) Labels {
 
 // Compare compares the two label sets.
 // The result will be 0 if a==b, <0 if a < b, and >0 if a > b.
-// TODO: replace with Less function - Compare is never needed.
-// TODO: just compare the underlying strings when we don't need alphanumeric sorting.
 func Compare(a, b Labels) int {
-	l := len(a.data)
-	if len(b.data) < l {
-		l = len(b.data)
+	// Find the first byte in the string where a and b differ.
+	shorter, longer := a.data, b.data
+	if len(b.data) < len(a.data) {
+		shorter, longer = b.data, a.data
+	}
+	i := 0
+	// First, go 8 bytes at a time. Data strings are expected to be 8-byte aligned.
+	sp := unsafe.Pointer((*reflect.StringHeader)(unsafe.Pointer(&shorter)).Data)
+	lp := unsafe.Pointer((*reflect.StringHeader)(unsafe.Pointer(&longer)).Data)
+	for ; i < len(shorter)-8; i += 8 {
+		if *(*uint64)(unsafe.Add(sp, i)) != *(*uint64)(unsafe.Add(lp, i)) {
+			break
+		}
+	}
+	// Now go 1 byte at a time.
+	for ; i < len(shorter); i++ {
+		if shorter[i] != longer[i] {
+			break
+		}
+	}
+	if i == len(shorter) {
+		// One Labels was a prefix of the other; the set with fewer labels compares lower.
+		return len(a.data) - len(b.data)
 	}
 
-	ia, ib := 0, 0
-	for ia < l {
-		var aName, bName string
-		aName, ia = decodeString(a.data, ia)
-		bName, ib = decodeString(b.data, ib)
-		if aName != bName {
-			if aName < bName {
-				return -1
-			}
-			return 1
-		}
-		var aValue, bValue string
-		aValue, ia = decodeString(a.data, ia)
-		bValue, ib = decodeString(b.data, ib)
-		if aValue != bValue {
-			if aValue < bValue {
-				return -1
-			}
-			return 1
+	// Now we know that there is some difference before the end of a and b.
+	// Go back through the fields and find which field that difference is in.
+	firstCharDifferent := i
+	for i = 0; ; {
+		size, nextI := decodeSize(a.data, i)
+		if nextI+size > firstCharDifferent {
+			break
 		}
+		i = nextI + size
 	}
-	// If all labels so far were in common, the set with fewer labels comes first.
-	return len(a.data) - len(b.data)
+	// Difference is inside this entry.
+	aStr, _ := decodeString(a.data, i)
+	bStr, _ := decodeString(b.data, i)
+	if aStr < bStr {
+		return -1
+	}
+	return +1
 }
 
 // Copy labels from b on top of whatever was in ls previously, reusing memory or expanding if needed.

model/labels/labels_test.go

@@ -361,6 +361,18 @@ func TestLabels_Compare(t *testing.T) {
 				"bbc", "222"),
 			expected: -1,
 		},
+		{
+			compared: FromStrings(
+				"aaa", "111",
+				"bb", "222"),
+			expected: 1,
+		},
+		{
+			compared: FromStrings(
+				"aaa", "111",
+				"bbbb", "222"),
+			expected: -1,
+		},
 		{
 			compared: FromStrings(
 				"aaa", "111"),
@@ -380,6 +392,10 @@ func TestLabels_Compare(t *testing.T) {
 				"bbb", "222"),
 			expected: 0,
 		},
+		{
+			compared: EmptyLabels(),
+			expected: 1,
+		},
 	}
 
 	sign := func(a int) int {
@@ -395,6 +411,8 @@ func TestLabels_Compare(t *testing.T) {
 	for i, test := range tests {
 		got := Compare(labels, test.compared)
 		require.Equal(t, sign(test.expected), sign(got), "unexpected comparison result for test case %d", i)
+		got = Compare(test.compared, labels)
+		require.Equal(t, -sign(test.expected), sign(got), "unexpected comparison result for reverse test case %d", i)
 	}
 }
 
@@ -468,27 +486,34 @@ func BenchmarkLabels_Get(b *testing.B) {
 	}
 }
 
+var comparisonBenchmarkScenarios = []struct {
+	desc        string
+	base, other Labels
+}{
+	{
+		"equal",
+		FromStrings("a_label_name", "a_label_value", "another_label_name", "another_label_value"),
+		FromStrings("a_label_name", "a_label_value", "another_label_name", "another_label_value"),
+	},
+	{
+		"not equal",
+		FromStrings("a_label_name", "a_label_value", "another_label_name", "another_label_value"),
+		FromStrings("a_label_name", "a_label_value", "another_label_name", "a_different_label_value"),
+	},
+	{
+		"different sizes",
+		FromStrings("a_label_name", "a_label_value", "another_label_name", "another_label_value"),
+		FromStrings("a_label_name", "a_label_value"),
+	},
+	{
+		"lots",
+		FromStrings("aaa", "bbb", "ccc", "ddd", "eee", "fff", "ggg", "hhh", "iii", "jjj", "kkk", "lll", "mmm", "nnn", "ooo", "ppp", "qqq", "rrz"),
+		FromStrings("aaa", "bbb", "ccc", "ddd", "eee", "fff", "ggg", "hhh", "iii", "jjj", "kkk", "lll", "mmm", "nnn", "ooo", "ppp", "qqq", "rrr"),
+	},
+}
+
 func BenchmarkLabels_Equals(b *testing.B) {
-	for _, scenario := range []struct {
-		desc        string
-		base, other Labels
-	}{
-		{
-			"equal",
-			FromStrings("a_label_name", "a_label_value", "another_label_name", "another_label_value"),
-			FromStrings("a_label_name", "a_label_value", "another_label_name", "another_label_value"),
-		},
-		{
-			"not equal",
-			FromStrings("a_label_name", "a_label_value", "another_label_name", "another_label_value"),
-			FromStrings("a_label_name", "a_label_value", "another_label_name", "a_different_label_value"),
-		},
-		{
-			"different sizes",
-			FromStrings("a_label_name", "a_label_value", "another_label_name", "another_label_value"),
-			FromStrings("a_label_name", "a_label_value"),
-		},
-	} {
+	for _, scenario := range comparisonBenchmarkScenarios {
 		b.Run(scenario.desc, func(b *testing.B) {
 			b.ResetTimer()
 			for i := 0; i < b.N; i++ {
@@ -498,6 +523,17 @@ func BenchmarkLabels_Equals(b *testing.B) {
 	}
 }
 
+func BenchmarkLabels_Compare(b *testing.B) {
+	for _, scenario := range comparisonBenchmarkScenarios {
+		b.Run(scenario.desc, func(b *testing.B) {
+			b.ResetTimer()
+			for i := 0; i < b.N; i++ {
+				_ = Compare(scenario.base, scenario.other)
+			}
+		})
+	}
+}
+
 func TestLabels_Copy(t *testing.T) {
 	require.Equal(t, FromStrings("aaa", "111", "bbb", "222"), FromStrings("aaa", "111", "bbb", "222").Copy())
 }