cmd/compile: add column details to export data

This CL updates the export data format to include column details when
writing out position data. cmd/compile is updated to generate and make
use of the new details, but go/internal/gcimporter only knows how to
read the data. It doesn't yet actually make use of it.

Experimentally across a wide range of packages, this increases export
data size by around 4%. However, it has no impact on binary size.
(Notably, it actually shrinks k8s.io/kubernetes/cmd/kubelet's binary
size by 24kB, but it's unclear to me why at this time.)

Updates #28259.

Change-Id: I351fb340839df8d3adced49b3757c4537fb91b3f
Reviewed-on: https://go-review.googlesource.com/c/go/+/196963
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Robert Griesemer <gri@golang.org>
Reviewed-by: Rob Pike <r@golang.org>

src/cmd/compile/internal/gc/iexport.go

@@ -184,8 +184,9 @@
 //     }
 //
 //
-// Pos encodes a file:line pair, incorporating a simple delta encoding
-// scheme within a data object. See exportWriter.pos for details.
+// Pos encodes a file:line:column triple, incorporating a simple delta
+// encoding scheme within a data object. See exportWriter.pos for
+// details.
 //
 //
 // Compiler-specific details.
@@ -212,8 +213,9 @@ import (
 )
 
 // Current indexed export format version. Increase with each format change.
+// 1: added column details to Pos
 // 0: Go1.11 encoding
-const iexportVersion = 0
+const iexportVersion = 1
 
 // predeclReserved is the number of type offsets reserved for types
 // implicitly declared in the universe block.
@@ -405,6 +407,7 @@ type exportWriter struct {
 	currPkg    *types.Pkg
 	prevFile   string
 	prevLine   int64
+	prevColumn int64
 }
 
 func (p *iexporter) doDecl(n *Node) {
@@ -510,29 +513,39 @@ func (w *exportWriter) pos(pos src.XPos) {
 	p := Ctxt.PosTable.Pos(pos)
 	file := p.Base().AbsFilename()
 	line := int64(p.RelLine())
+	column := int64(p.RelCol())
 
-	// When file is the same as the last position (common case),
-	// we can save a few bytes by delta encoding just the line
-	// number.
+	// Encode position relative to the last position: column
+	// delta, then line delta, then file name. We reserve the
+	// bottom bit of the column and line deltas to encode whether
+	// the remaining fields are present.
 	//
 	// Note: Because data objects may be read out of order (or not
 	// at all), we can only apply delta encoding within a single
-	// object. This is handled implicitly by tracking prevFile and
-	// prevLine as fields of exportWriter.
+	// object. This is handled implicitly by tracking prevFile,
+	// prevLine, and prevColumn as fields of exportWriter.
 
-	if file == w.prevFile {
-		delta := line - w.prevLine
-		w.int64(delta)
-		if delta == deltaNewFile {
-			w.int64(-1)
+	deltaColumn := (column - w.prevColumn) << 1
+	deltaLine := (line - w.prevLine) << 1
+
+	if file != w.prevFile {
+		deltaLine |= 1
 	}
-	} else {
-		w.int64(deltaNewFile)
-		w.int64(line) // line >= 0
+	if deltaLine != 0 {
+		deltaColumn |= 1
+	}
+
+	w.int64(deltaColumn)
+	if deltaColumn&1 != 0 {
+		w.int64(deltaLine)
+		if deltaLine&1 != 0 {
 			w.string(file)
-		w.prevFile = file
 		}
+	}
+
+	w.prevFile = file
 	w.prevLine = line
+	w.prevColumn = column
 }
 
 func (w *exportWriter) pkg(pkg *types.Pkg) {

src/cmd/compile/internal/gc/iimport.go

@@ -245,6 +245,7 @@ type importReader struct {
 	currPkg    *types.Pkg
 	prevBase   *src.PosBase
 	prevLine   int64
+	prevColumn int64
 }
 
 func (p *iimporter) newReader(off uint64, pkg *types.Pkg) *importReader {
@@ -446,16 +447,16 @@ func (r *importReader) qualifiedIdent() *types.Sym {
 
 func (r *importReader) pos() src.XPos {
 	delta := r.int64()
-	if delta != deltaNewFile {
-		r.prevLine += delta
-	} else if l := r.int64(); l == -1 {
-		r.prevLine += deltaNewFile
-	} else {
+	r.prevColumn += delta >> 1
+	if delta&1 != 0 {
+		delta = r.int64()
+		r.prevLine += delta >> 1
+		if delta&1 != 0 {
 			r.prevBase = r.posBase()
-		r.prevLine = l
+		}
 	}
 
-	if (r.prevBase == nil || r.prevBase.AbsFilename() == "") && r.prevLine == 0 {
+	if (r.prevBase == nil || r.prevBase.AbsFilename() == "") && r.prevLine == 0 && r.prevColumn == 0 {
 		// TODO(mdempsky): Remove once we reliably write
 		// position information for all nodes.
 		return src.NoXPos
@@ -464,7 +465,7 @@ func (r *importReader) pos() src.XPos {
 	if r.prevBase == nil {
 		Fatalf("missing posbase")
 	}
-	pos := src.MakePos(r.prevBase, uint(r.prevLine), 0)
+	pos := src.MakePos(r.prevBase, uint(r.prevLine), uint(r.prevColumn))
 	return Ctxt.PosTable.XPos(pos)
 }
 

src/go/internal/gcimporter/bimport.go

@@ -328,7 +328,7 @@ func (p *importer) pos() token.Pos {
 	p.prevFile = file
 	p.prevLine = line
 
-	return p.fake.pos(file, line)
+	return p.fake.pos(file, line, 0)
 }
 
 // Synthesize a token.Pos
@@ -337,7 +337,9 @@ type fakeFileSet struct {
 	files map[string]*token.File
 }
 
-func (s *fakeFileSet) pos(file string, line int) token.Pos {
+func (s *fakeFileSet) pos(file string, line, column int) token.Pos {
+	// TODO(mdempsky): Make use of column.
+
 	// Since we don't know the set of needed file positions, we
 	// reserve maxlines positions per file.
 	const maxlines = 64 * 1024

src/go/internal/gcimporter/iimport.go

@@ -61,8 +61,8 @@ const (
 // If the export data version is not recognized or the format is otherwise
 // compromised, an error is returned.
 func iImportData(fset *token.FileSet, imports map[string]*types.Package, data []byte, path string) (_ int, pkg *types.Package, err error) {
-	const currentVersion = 0
-	version := -1
+	const currentVersion = 1
+	version := int64(-1)
 	defer func() {
 		if e := recover(); e != nil {
 			if version > currentVersion {
@@ -75,9 +75,9 @@ func iImportData(fset *token.FileSet, imports map[string]*types.Package, data []
 
 	r := &intReader{bytes.NewReader(data), path}
 
-	version = int(r.uint64())
+	version = int64(r.uint64())
 	switch version {
-	case currentVersion:
+	case currentVersion, 0:
 	default:
 		errorf("unknown iexport format version %d", version)
 	}
@@ -92,6 +92,7 @@ func iImportData(fset *token.FileSet, imports map[string]*types.Package, data []
 
 	p := iimporter{
 		ipath:   path,
+		version: int(version),
 
 		stringData:  stringData,
 		stringCache: make(map[uint64]string),
@@ -170,6 +171,7 @@ func iImportData(fset *token.FileSet, imports map[string]*types.Package, data []
 
 type iimporter struct {
 	ipath   string
+	version int
 
 	stringData  []byte
 	stringCache map[uint64]string
@@ -249,6 +251,7 @@ type importReader struct {
 	currPkg    *types.Package
 	prevFile   string
 	prevLine   int64
+	prevColumn int64
 }
 
 func (r *importReader) obj(name string) {
@@ -438,6 +441,19 @@ func (r *importReader) qualifiedIdent() (*types.Package, string) {
 }
 
 func (r *importReader) pos() token.Pos {
+	if r.p.version >= 1 {
+		r.posv1()
+	} else {
+		r.posv0()
+	}
+
+	if r.prevFile == "" && r.prevLine == 0 && r.prevColumn == 0 {
+		return token.NoPos
+	}
+	return r.p.fake.pos(r.prevFile, int(r.prevLine), int(r.prevColumn))
+}
+
+func (r *importReader) posv0() {
 	delta := r.int64()
 	if delta != deltaNewFile {
 		r.prevLine += delta
@@ -447,12 +463,18 @@ func (r *importReader) pos() token.Pos {
 		r.prevFile = r.string()
 		r.prevLine = l
 	}
+}
 
-	if r.prevFile == "" && r.prevLine == 0 {
-		return token.NoPos
+func (r *importReader) posv1() {
+	delta := r.int64()
+	r.prevColumn += delta >> 1
+	if delta&1 != 0 {
+		delta = r.int64()
+		r.prevLine += delta >> 1
+		if delta&1 != 0 {
+			r.prevFile = r.string()
+		}
 	}
-
-	return r.p.fake.pos(r.prevFile, int(r.prevLine))
 }
 
 func (r *importReader) typ() types.Type {