cmd/asm: reject misplaced go:build comments

We are converting from using error-prone ad-hoc syntax // +build lines
to less error-prone, standard boolean syntax //go:build lines.
The timeline is:

Go 1.16: prepare for transition
 - Builds still use // +build for file selection.
 - Source files may not contain //go:build without // +build.
 - Builds fail when a source file contains //go:build lines without // +build lines. <<<

Go 1.17: start transition
 - Builds prefer //go:build for file selection, falling back to // +build
   for files containing only // +build.
 - Source files may contain //go:build without // +build (but they won't build with Go 1.16).
 - Gofmt moves //go:build and // +build lines to proper file locations.
 - Gofmt introduces //go:build lines into files with only // +build lines.
 - Go vet rejects files with mismatched //go:build and // +build lines.

Go 1.18: complete transition
 - Go fix removes // +build lines, leaving behind equivalent // +build lines.

This CL provides part of the <<< marked line above in the Go 1.16 step:
rejecting files containing //go:build but not // +build.

Reject any //go:build comments found after actual assembler code
(include #include etc directives), because the go command itself
doesn't read that far.

For #41184.

Change-Id: Ib460bfd380cce4239993980dd208afd07deff3f1
Reviewed-on: https://go-review.googlesource.com/c/go/+/240602
Trust: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

src/cmd/asm/internal/asm/endtoend_test.go

@@ -257,11 +257,11 @@ func isHexes(s string) bool {
 	return true
 }
 
-// It would be nice if the error messages began with
+// It would be nice if the error messages always began with
 // the standard file:line: prefix,
 // but that's not where we are today.
 // It might be at the beginning but it might be in the middle of the printed instruction.
-var fileLineRE = regexp.MustCompile(`(?:^|\()(testdata[/\\][0-9a-z]+\.s:[0-9]+)(?:$|\))`)
+var fileLineRE = regexp.MustCompile(`(?:^|\()(testdata[/\\][0-9a-z]+\.s:[0-9]+)(?:$|\)|:)`)
 
 // Same as in test/run.go
 var (
@@ -281,6 +281,7 @@ func testErrors(t *testing.T, goarch, file string) {
 	defer ctxt.Bso.Flush()
 	failed := false
 	var errBuf bytes.Buffer
+	parser.errorWriter = &errBuf
 	ctxt.DiagFunc = func(format string, args ...interface{}) {
 		failed = true
 		s := fmt.Sprintf(format, args...)
@@ -292,7 +293,7 @@ func testErrors(t *testing.T, goarch, file string) {
 	pList.Firstpc, ok = parser.Parse()
 	obj.Flushplist(ctxt, pList, nil, "")
 	if ok && !failed {
-		t.Errorf("asm: %s had no errors", goarch)
+		t.Errorf("asm: %s had no errors", file)
 	}
 
 	errors := map[string]string{}
@@ -368,6 +369,10 @@ func TestARMEndToEnd(t *testing.T) {
 	}
 }
 
+func TestGoBuildErrors(t *testing.T) {
+	testErrors(t, "amd64", "buildtagerror")
+}
+
 func TestARMErrors(t *testing.T) {
 	testErrors(t, "arm", "armerror")
 }

src/cmd/asm/internal/asm/parse.go

@@ -29,6 +29,7 @@ type Parser struct {
 	lineNum       int   // Line number in source file.
 	errorLine     int   // Line number of last error.
 	errorCount    int   // Number of errors.
+	sawCode       bool  // saw code in this file (as opposed to comments and blank lines)
 	pc            int64 // virtual PC; count of Progs; doesn't advance for GLOBL or DATA.
 	input         []lex.Token
 	inputPos      int
@@ -132,6 +133,30 @@ func (p *Parser) ParseSymABIs(w io.Writer) bool {
 	return p.errorCount == 0
 }
 
+// nextToken returns the next non-build-comment token from the lexer.
+// It reports misplaced //go:build comments but otherwise discards them.
+func (p *Parser) nextToken() lex.ScanToken {
+	for {
+		tok := p.lex.Next()
+		if tok == lex.BuildComment {
+			if p.sawCode {
+				p.errorf("misplaced //go:build comment")
+			}
+			continue
+		}
+		if tok != '\n' {
+			p.sawCode = true
+		}
+		if tok == '#' {
+			// A leftover wisp of a #include/#define/etc,
+			// to let us know that p.sawCode should be true now.
+			// Otherwise ignored.
+			continue
+		}
+		return tok
+	}
+}
+
 // line consumes a single assembly line from p.lex of the form
 //
 //   {label:} WORD[.cond] [ arg {, arg} ] (';' | '\n')
@@ -146,7 +171,7 @@ next:
 	// Skip newlines.
 	var tok lex.ScanToken
 	for {
-		tok = p.lex.Next()
+		tok = p.nextToken()
 		// We save the line number here so error messages from this instruction
 		// are labeled with this line. Otherwise we complain after we've absorbed
 		// the terminating newline and the line numbers are off by one in errors.
@@ -179,11 +204,11 @@ next:
 			items = make([]lex.Token, 0, 3)
 		}
 		for {
-			tok = p.lex.Next()
+			tok = p.nextToken()
 			if len(operands) == 0 && len(items) == 0 {
 				if p.arch.InFamily(sys.ARM, sys.ARM64, sys.AMD64, sys.I386) && tok == '.' {
 					// Suffixes: ARM conditionals or x86 modifiers.
-					tok = p.lex.Next()
+					tok = p.nextToken()
 					str := p.lex.Text()
 					if tok != scanner.Ident {
 						p.errorf("instruction suffix expected identifier, found %s", str)

src/cmd/asm/internal/asm/testdata/buildtagerror.s

@@ -0,0 +1,8 @@
+// Copyright 2020 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.
+
+#define X 1
+
+//go:build x // ERROR "misplaced //go:build comment"
+

src/cmd/asm/internal/lex/input.go

@@ -109,6 +109,9 @@ func (in *Input) Next() ScanToken {
 				in.Error("'#' must be first item on line")
 			}
 			in.beginningOfLine = in.hash()
+			in.text = "#"
+			return '#'
+
 		case scanner.Ident:
 			// Is it a macro name?
 			name := in.Stack.Text()

src/cmd/asm/internal/lex/lex.go

@@ -22,11 +22,13 @@ type ScanToken rune
 const (
 	// Asm defines some two-character lexemes. We make up
 	// a rune/ScanToken value for them - ugly but simple.
-	LSH       ScanToken = -1000 - iota // << Left shift.
-	RSH                                // >> Logical right shift.
-	ARR                                // -> Used on ARM for shift type 3, arithmetic right shift.
-	ROT                                // @> Used on ARM for shift type 4, rotate right.
-	macroName                          // name of macro that should not be expanded
+	LSH          ScanToken = -1000 - iota // << Left shift.
+	RSH                                   // >> Logical right shift.
+	ARR                                   // -> Used on ARM for shift type 3, arithmetic right shift.
+	ROT                                   // @> Used on ARM for shift type 4, rotate right.
+	Include                               // included file started here
+	BuildComment                          // //go:build or +build comment
+	macroName                             // name of macro that should not be expanded
 )
 
 // IsRegisterShift reports whether the token is one of the ARM register shift operators.

src/cmd/asm/internal/lex/lex_test.go

@@ -281,6 +281,9 @@ func drain(input *Input) string {
 		if tok == scanner.EOF {
 			return buf.String()
 		}
+		if tok == '#' {
+			continue
+		}
 		if buf.Len() > 0 {
 			buf.WriteByte('.')
 		}

src/cmd/asm/internal/lex/tokenizer.go

@@ -107,10 +107,13 @@ func (t *Tokenizer) Next() ScanToken {
 		if t.tok != scanner.Comment {
 			break
 		}
-		length := strings.Count(s.TokenText(), "\n")
-		t.line += length
-		// TODO: If we ever have //go: comments in assembly, will need to keep them here.
-		// For now, just discard all comments.
+		text := s.TokenText()
+		t.line += strings.Count(text, "\n")
+		// TODO: Use constraint.IsGoBuild once it exists.
+		if strings.HasPrefix(text, "//go:build") {
+			t.tok = BuildComment
+			break
+		}
 	}
 	switch t.tok {
 	case '\n':