txscript: add benchmark for IsUnspendable

- create benchmarks to measure allocations
 - add test for benchmark input
 - create a low alloc parseScriptTemplate
 - refactor parsing logic for a single opcode

txscript/opcode.go

@@ -656,6 +656,79 @@ func (pop *parsedOpcode) isDisabled() bool {
 	}
 }
 
+// checkParseableInScript checks whether or not the current opcode is able to be
+// parsed at a certain position in a script.
+// This returns the position of the next opcode to be parsed in the script.
+func (pop *parsedOpcode) checkParseableInScript(script []byte, scriptPos int) (int, error) {
+	// Parse data out of instruction.
+	switch {
+	// No additional data.  Note that some of the opcodes, notably
+	// OP_1NEGATE, OP_0, and OP_[1-16] represent the data
+	// themselves.
+	case pop.opcode.length == 1:
+		scriptPos++
+
+	// Data pushes of specific lengths -- OP_DATA_[1-75].
+	case pop.opcode.length > 1:
+		if len(script[scriptPos:]) < pop.opcode.length {
+			str := fmt.Sprintf("opcode %s requires %d "+
+				"bytes, but script only has %d remaining",
+				pop.opcode.name, pop.opcode.length, len(script[scriptPos:]))
+			return 0, scriptError(ErrMalformedPush, str)
+		}
+
+		// Slice out the data.
+		pop.data = script[scriptPos+1 : scriptPos+pop.opcode.length]
+		scriptPos += pop.opcode.length
+
+	// Data pushes with parsed lengths -- OP_PUSHDATAP{1,2,4}.
+	case pop.opcode.length < 0:
+		var l uint
+		off := scriptPos + 1
+
+		if len(script[off:]) < -pop.opcode.length {
+			str := fmt.Sprintf("opcode %s requires %d "+
+				"bytes, but script only has %d remaining",
+				pop.opcode.name, -pop.opcode.length, len(script[off:]))
+			return 0, scriptError(ErrMalformedPush, str)
+		}
+
+		// Next -length bytes are little endian length of data.
+		switch pop.opcode.length {
+		case -1:
+			l = uint(script[off])
+		case -2:
+			l = ((uint(script[off+1]) << 8) |
+				uint(script[off]))
+		case -4:
+			l = ((uint(script[off+3]) << 24) |
+				(uint(script[off+2]) << 16) |
+				(uint(script[off+1]) << 8) |
+				uint(script[off]))
+		default:
+			str := fmt.Sprintf("invalid opcode length %d",
+				pop.opcode.length)
+			return 0, scriptError(ErrMalformedPush, str)
+		}
+
+		// Move offset to beginning of the data.
+		off += -pop.opcode.length
+
+		// Disallow entries that do not fit script or were
+		// sign extended.
+		if int(l) > len(script[off:]) || int(l) < 0 {
+			str := fmt.Sprintf("opcode %s pushes %d bytes, "+
+				"but script only has %d remaining",
+				pop.opcode.name, int(l), len(script[off:]))
+			return 0, scriptError(ErrMalformedPush, str)
+		}
+
+		pop.data = script[off : off+int(l)]
+		scriptPos += 1 - pop.opcode.length + int(l)
+	}
+	return scriptPos, nil
+}
+
 // alwaysIllegal returns whether or not the opcode is always illegal when passed
 // over by the program counter even if in a non-executed branch (it isn't a
 // coincidence that they are conditionals).

txscript/script.go

@@ -196,80 +196,14 @@ func IsPushOnlyScript(script []byte) bool {
 // the list of parsed opcodes up to the point of failure along with the error.
 func parseScriptTemplate(script []byte, opcodes *[256]opcode) ([]parsedOpcode, error) {
 	retScript := make([]parsedOpcode, 0, len(script))
+	var err error
 	for i := 0; i < len(script); {
 		instr := script[i]
 		op := &opcodes[instr]
 		pop := parsedOpcode{opcode: op}
-
+		i, err = pop.checkParseableInScript(script, i)
-		// Parse data out of instruction.
+		if err != nil {
-		switch {
+			return retScript, err
-		// No additional data.  Note that some of the opcodes, notably
-		// OP_1NEGATE, OP_0, and OP_[1-16] represent the data
-		// themselves.
-		case op.length == 1:
-			i++
-
-		// Data pushes of specific lengths -- OP_DATA_[1-75].
-		case op.length > 1:
-			if len(script[i:]) < op.length {
-				str := fmt.Sprintf("opcode %s requires %d "+
-					"bytes, but script only has %d remaining",
-					op.name, op.length, len(script[i:]))
-				return retScript, scriptError(ErrMalformedPush,
-					str)
-			}
-
-			// Slice out the data.
-			pop.data = script[i+1 : i+op.length]
-			i += op.length
-
-		// Data pushes with parsed lengths -- OP_PUSHDATAP{1,2,4}.
-		case op.length < 0:
-			var l uint
-			off := i + 1
-
-			if len(script[off:]) < -op.length {
-				str := fmt.Sprintf("opcode %s requires %d "+
-					"bytes, but script only has %d remaining",
-					op.name, -op.length, len(script[off:]))
-				return retScript, scriptError(ErrMalformedPush,
-					str)
-			}
-
-			// Next -length bytes are little endian length of data.
-			switch op.length {
-			case -1:
-				l = uint(script[off])
-			case -2:
-				l = ((uint(script[off+1]) << 8) |
-					uint(script[off]))
-			case -4:
-				l = ((uint(script[off+3]) << 24) |
-					(uint(script[off+2]) << 16) |
-					(uint(script[off+1]) << 8) |
-					uint(script[off]))
-			default:
-				str := fmt.Sprintf("invalid opcode length %d",
-					op.length)
-				return retScript, scriptError(ErrMalformedPush,
-					str)
-			}
-
-			// Move offset to beginning of the data.
-			off += -op.length
-
-			// Disallow entries that do not fit script or were
-			// sign extended.
-			if int(l) > len(script[off:]) || int(l) < 0 {
-				str := fmt.Sprintf("opcode %s pushes %d bytes, "+
-					"but script only has %d remaining",
-					op.name, int(l), len(script[off:]))
-				return retScript, scriptError(ErrMalformedPush,
-					str)
-			}
-
-			pop.data = script[off : off+int(l)]
-			i += 1 - op.length + int(l)
 		}
 
 		retScript = append(retScript, pop)
@@ -278,6 +212,44 @@ func parseScriptTemplate(script []byte, opcodes *[256]opcode) ([]parsedOpcode, e
 	return retScript, nil
 }
 
+// checkScriptTemplateParseable is the same as parseScriptTemplate but does not
+// return the list of opcodes up until the point of failure so that this can be
+// used in functions which do not necessarily have a need for the failed list of
+// opcodes, such as IsUnspendable.
+//
+// This function returns a pointer to a byte. This byte is nil if the parsing
+// has an error, or if the script length is zero. If the script length is not
+// zero and parsing succeeds, then the first opcode parsed will be returned.
+//
+// Not returning the full opcode list up until failure also has the benefit of
+// reducing GC pressure, as the list would get immediately thrown away.
+func checkScriptTemplateParseable(script []byte, opcodes *[256]opcode) (*byte, error) {
+	var err error
+
+	// A script of length zero is an unspendable script but it is parseable.
+	var firstOpcode byte
+	var numParsedInstr uint = 0
+
+	for i := 0; i < len(script); {
+		instr := script[i]
+		op := &opcodes[instr]
+		pop := parsedOpcode{opcode: op}
+		i, err = pop.checkParseableInScript(script, i)
+		if err != nil {
+			return nil, err
+		}
+
+		// if this is a op_return then it is unspendable so we set the first
+		// parsed instruction in case it's an op_return
+		if numParsedInstr == 0 {
+			firstOpcode = pop.opcode.value
+		}
+		numParsedInstr++
+	}
+
+	return &firstOpcode, nil
+}
+
 // parseScript preparses the script in bytes into a list of parsedOpcodes while
 // applying a number of sanity checks.
 func parseScript(script []byte) ([]parsedOpcode, error) {
@@ -851,10 +823,14 @@ func getWitnessSigOps(pkScript []byte, witness wire.TxWitness) int {
 // guaranteed to fail at execution.  This allows inputs to be pruned instantly
 // when entering the UTXO set.
 func IsUnspendable(pkScript []byte) bool {
-	pops, err := parseScript(pkScript)
+	// Not provably unspendable
+	if len(pkScript) == 0 {
+		return false
+	}
+	firstOpcode, err := checkScriptTemplateParseable(pkScript, &opcodeArray)
 	if err != nil {
 		return true
 	}
 
-	return len(pops) > 0 && pops[0].opcode.value == OP_RETURN
+	return firstOpcode != nil && *firstOpcode == OP_RETURN
 }

txscript/script_test.go

@@ -4301,6 +4301,28 @@ func TestIsUnspendable(t *testing.T) {
 				0xfa, 0x0b, 0x5c, 0x88, 0xac},
 			expected: false,
 		},
+		{
+			// Spendable
+			pkScript: []byte{0xa9, 0x14, 0x82, 0x1d, 0xba, 0x94, 0xbc, 0xfb,
+				0xa2, 0x57, 0x36, 0xa3, 0x9e, 0x5d, 0x14, 0x5d, 0x69, 0x75,
+				0xba, 0x8c, 0x0b, 0x42, 0x87},
+			expected: false,
+		},
+		{
+			// Not Necessarily Unspendable
+			pkScript: []byte{},
+			expected: false,
+		},
+		{
+			// Spendable
+			pkScript: []byte{OP_TRUE},
+			expected: false,
+		},
+		{
+			// Unspendable
+			pkScript: []byte{OP_RETURN},
+			expected: true,
+		},
 	}
 
 	for i, test := range tests {
@@ -4312,3 +4334,16 @@ func TestIsUnspendable(t *testing.T) {
 		}
 	}
 }
+
+// BenchmarkIsUnspendable adds a benchmark to compare the time and allocations
+// necessary for the IsUnspendable function.
+func BenchmarkIsUnspendable(b *testing.B) {
+	pkScriptToUse := []byte{0xa9, 0x14, 0x82, 0x1d, 0xba, 0x94, 0xbc, 0xfb, 0xa2, 0x57, 0x36, 0xa3, 0x9e, 0x5d, 0x14, 0x5d, 0x69, 0x75, 0xba, 0x8c, 0x0b, 0x42, 0x87}
+	var res bool = false
+	for i := 0; i < b.N; i++ {
+		res = IsUnspendable(pkScriptToUse)
+	}
+	if res {
+		b.Fatalf("Benchmark should never have res be %t\n", res)
+	}
+}