cmd/objdump: implement objdump of .o files

Update goobj reader so it can provide all the information
necessary to disassemble .o (and .a) files.

Grab architecture of .o files from header.

.o files have relocations in them.  This CL also contains a simple
mechanism to disassemble relocations and add relocation info as an extra
column in the output.

Fixes #13862

Change-Id: I608fd253ff1522ea47f18be650b38d528dae9054
Reviewed-on: https://go-review.googlesource.com/24818Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>

src/cmd/internal/goobj/read.go

@@ -163,7 +163,7 @@ type Data struct {
 // A Reloc describes a relocation applied to a memory image to refer
 // to an address within a particular symbol.
 type Reloc struct {
-	// The bytes at [Offset, Offset+Size) within the memory image
+	// The bytes at [Offset, Offset+Size) within the containing Sym
 	// should be updated to refer to the address Add bytes after the start
 	// of the symbol Sym.
 	Offset int
@@ -174,7 +174,7 @@ type Reloc struct {
 	// The Type records the form of address expected in the bytes
 	// described by the previous fields: absolute, PC-relative, and so on.
 	// TODO(rsc): The interpretation of Type is not exposed by this package.
-	Type int
+	Type obj.RelocType
 }
 
 // A Var describes a variable in a function stack frame: a declared
@@ -220,6 +220,7 @@ type Package struct {
 	SymRefs    []SymID  // list of symbol names and versions referred to by this pack
 	Syms       []*Sym   // symbols defined by this package
 	MaxVersion int      // maximum Version in any SymID in Syms
+	Arch       string   // architecture
 }
 
 var (
@@ -561,14 +562,13 @@ func (r *objReader) parseArchive() error {
 // The format of that part is defined in a comment at the top
 // of src/liblink/objfile.c.
 func (r *objReader) parseObject(prefix []byte) error {
-	// TODO(rsc): Maybe use prefix and the initial input to
-	// record the header line from the file, which would
-	// give the architecture and other version information.
-
 	r.p.MaxVersion++
+	h := make([]byte, 0, 256)
+	h = append(h, prefix...)
 	var c1, c2, c3 byte
 	for {
 		c1, c2, c3 = c2, c3, r.readByte()
+		h = append(h, c3)
 		// The new export format can contain 0 bytes.
 		// Don't consider them errors, only look for r.err != nil.
 		if r.err != nil {
@@ -579,6 +579,12 @@ func (r *objReader) parseObject(prefix []byte) error {
 		}
 	}
 
+	hs := strings.Fields(string(h))
+	if len(hs) >= 4 {
+		r.p.Arch = hs[3]
+	}
+	// TODO: extract OS + build ID if/when we need it
+
 	r.readFull(r.tmp[:8])
 	if !bytes.Equal(r.tmp[:8], []byte("\x00\x00go17ld")) {
 		return r.error(errCorruptObject)
@@ -643,7 +649,7 @@ func (r *objReader) parseObject(prefix []byte) error {
 			rel := &s.Reloc[i]
 			rel.Offset = r.readInt()
 			rel.Size = r.readInt()
-			rel.Type = r.readInt()
+			rel.Type = obj.RelocType(r.readInt())
 			rel.Add = r.readInt()
 			rel.Sym = r.readSymID()
 		}
@@ -693,3 +699,18 @@ func (r *objReader) parseObject(prefix []byte) error {
 
 	return nil
 }
+
+func (r *Reloc) String(insnOffset uint64) string {
+	delta := r.Offset - int(insnOffset)
+	s := fmt.Sprintf("[%d:%d]%s", delta, delta+r.Size, r.Type)
+	if r.Sym.Name != "" {
+		if r.Add != 0 {
+			return fmt.Sprintf("%s:%s+%d", s, r.Sym.Name, r.Add)
+		}
+		return fmt.Sprintf("%s:%s", s, r.Sym.Name)
+	}
+	if r.Add != 0 {
+		return fmt.Sprintf("%s:%d", s, r.Add)
+	}
+	return s
+}

src/cmd/internal/obj/link.go

@@ -442,14 +442,16 @@ const (
 type Reloc struct {
 	Off  int32
 	Siz  uint8
-	Type int32
+	Type RelocType
 	Add  int64
 	Sym  *LSym
 }
 
-// Reloc.type
+type RelocType int32
+
+//go:generate stringer -type=RelocType
 const (
-	R_ADDR = 1 + iota
+	R_ADDR RelocType = 1 + iota
 	// R_ADDRPOWER relocates a pair of "D-form" instructions (instructions with 16-bit
 	// immediates in the low half of the instruction word), usually addis followed by
 	// another add or a load, inserting the "high adjusted" 16 bits of the address of

src/cmd/internal/obj/reloctype_string.go

+// Code generated by "stringer -type=RelocType"; DO NOT EDIT
+
+package obj
+
+import "fmt"
+
+const _RelocType_name = "R_ADDRR_ADDRPOWERR_ADDRARM64R_ADDRMIPSR_ADDROFFR_SIZER_CALLR_CALLARMR_CALLARM64R_CALLINDR_CALLPOWERR_CALLMIPSR_CONSTR_PCRELR_TLS_LER_TLS_IER_GOTOFFR_PLT0R_PLT1R_PLT2R_USEFIELDR_USETYPER_METHODOFFR_POWER_TOCR_GOTPCRELR_JMPMIPSR_DWARFREFR_ARM64_TLS_LER_ARM64_TLS_IER_ARM64_GOTPCRELR_POWER_TLS_LER_POWER_TLS_IER_POWER_TLSR_ADDRPOWER_DSR_ADDRPOWER_GOTR_ADDRPOWER_PCRELR_ADDRPOWER_TOCRELR_ADDRPOWER_TOCREL_DSR_PCRELDBLR_ADDRMIPSUR_ADDRMIPSTLS"
+
+var _RelocType_index = [...]uint16{0, 6, 17, 28, 38, 47, 53, 59, 68, 79, 88, 99, 109, 116, 123, 131, 139, 147, 153, 159, 165, 175, 184, 195, 206, 216, 225, 235, 249, 263, 279, 293, 307, 318, 332, 347, 364, 382, 403, 413, 424, 437}
+
+func (i RelocType) String() string {
+	i -= 1
+	if i < 0 || i >= RelocType(len(_RelocType_index)-1) {
+		return fmt.Sprintf("RelocType(%d)", i+1)
+	}
+	return _RelocType_name[_RelocType_index[i]:_RelocType_index[i+1]]
+}

src/cmd/internal/objfile/disasm.go

@@ -22,7 +22,7 @@ import (
 // Disasm is a disassembler for a given File.
 type Disasm struct {
 	syms      []Sym            //symbols in file, sorted by address
-	pcln      *gosym.Table     // pcln table
+	pcln      Liner            // pcln table
 	text      []byte           // bytes of text segment (actual instructions)
 	textStart uint64           // start PC of text
 	textEnd   uint64           // end PC of text
@@ -116,6 +116,7 @@ func (d *Disasm) Print(w io.Writer, filter *regexp.Regexp, start, end uint64) {
 	for _, sym := range d.syms {
 		symStart := sym.Addr
 		symEnd := sym.Addr + uint64(sym.Size)
+		relocs := sym.Relocs
 		if sym.Code != 'T' && sym.Code != 't' ||
 			symStart < d.textStart ||
 			symEnd <= start || end <= symStart ||
@@ -135,7 +136,7 @@ func (d *Disasm) Print(w io.Writer, filter *regexp.Regexp, start, end uint64) {
 			symEnd = end
 		}
 		code := d.text[:end-d.textStart]
-		d.Decode(symStart, symEnd, func(pc, size uint64, file string, line int, text string) {
+		d.Decode(symStart, symEnd, relocs, func(pc, size uint64, file string, line int, text string) {
 			i := pc - d.textStart
 			fmt.Fprintf(tw, "\t%s:%d\t%#x\t", base(file), line, pc)
 			if size%4 != 0 || d.goarch == "386" || d.goarch == "amd64" {
@@ -158,7 +159,7 @@ func (d *Disasm) Print(w io.Writer, filter *regexp.Regexp, start, end uint64) {
 }
 
 // Decode disassembles the text segment range [start, end), calling f for each instruction.
-func (d *Disasm) Decode(start, end uint64, f func(pc, size uint64, file string, line int, text string)) {
+func (d *Disasm) Decode(start, end uint64, relocs []Reloc, f func(pc, size uint64, file string, line int, text string)) {
 	if start < d.textStart {
 		start = d.textStart
 	}
@@ -171,6 +172,17 @@ func (d *Disasm) Decode(start, end uint64, f func(pc, size uint64, file string,
 		i := pc - d.textStart
 		text, size := d.disasm(code[i:], pc, lookup)
 		file, line, _ := d.pcln.PCToLine(pc)
+		text += "\t"
+		first := true
+		for len(relocs) > 0 && relocs[0].Addr < i+uint64(size) {
+			if first {
+				first = false
+			} else {
+				text += " "
+			}
+			text += relocs[0].Stringer.String(pc - start)
+			relocs = relocs[1:]
+		}
 		f(pc, uint64(size), file, line, text)
 		pc += uint64(size)
 	}
@@ -247,3 +259,9 @@ var byteOrders = map[string]binary.ByteOrder{
 	"ppc64le": binary.LittleEndian,
 	"s390x":   binary.BigEndian,
 }
+
+type Liner interface {
+	// Given a pc, returns the corresponding file, line, and function data.
+	// If unknown, returns "",0,nil.
+	PCToLine(uint64) (string, int, *gosym.Func)
+}

src/cmd/internal/objfile/goobj.go

@@ -8,7 +8,9 @@ package objfile
 
 import (
 	"cmd/internal/goobj"
+	"cmd/internal/sys"
 	"debug/dwarf"
+	"debug/gosym"
 	"errors"
 	"fmt"
 	"os"
@@ -16,6 +18,7 @@ import (
 
 type goobjFile struct {
 	goobj *goobj.Package
+	f     *os.File // the underlying .o or .a file
 }
 
 func openGoobj(r *os.File) (rawFile, error) {
@@ -23,7 +26,7 @@ func openGoobj(r *os.File) (rawFile, error) {
 	if err != nil {
 		return nil, err
 	}
-	return &goobjFile{f}, nil
+	return &goobjFile{goobj: f, f: r}, nil
 }
 
 func goobjName(id goobj.SymID) string {
@@ -55,6 +58,9 @@ func (f *goobjFile) symbols() ([]Sym, error) {
 		if s.Version != 0 {
 			sym.Code += 'a' - 'A'
 		}
+		for i, r := range s.Reloc {
+			sym.Relocs = append(sym.Relocs, Reloc{Addr: uint64(s.Data.Offset) + uint64(r.Offset), Size: uint64(r.Size), Stringer: &s.Reloc[i]})
+		}
 		syms = append(syms, sym)
 	}
 
@@ -75,23 +81,68 @@ func (f *goobjFile) symbols() ([]Sym, error) {
 	return syms, nil
 }
 
-// pcln does not make sense for Go object files, because each
-// symbol has its own individual pcln table, so there is no global
-// space of addresses to map.
 func (f *goobjFile) pcln() (textStart uint64, symtab, pclntab []byte, err error) {
+	// Should never be called.  We implement Liner below, callers
+	// should use that instead.
 	return 0, nil, nil, fmt.Errorf("pcln not available in go object file")
 }
 
-// text does not make sense for Go object files, because
-// each function has a separate section.
+// Find returns the file name, line, and function data for the given pc.
+// Returns "",0,nil if unknown.
+// This function implements the Liner interface in preference to pcln() above.
+func (f *goobjFile) PCToLine(pc uint64) (string, int, *gosym.Func) {
+	// TODO: this is really inefficient.  Binary search?  Memoize last result?
+	var arch *sys.Arch
+	for _, a := range sys.Archs {
+		if a.Name == f.goobj.Arch {
+			arch = a
+			break
+		}
+	}
+	if arch == nil {
+		return "", 0, nil
+	}
+	for _, s := range f.goobj.Syms {
+		if pc < uint64(s.Data.Offset) || pc >= uint64(s.Data.Offset+s.Data.Size) {
+			continue
+		}
+		if s.Func == nil {
+			return "", 0, nil
+		}
+		pcfile := make([]byte, s.Func.PCFile.Size)
+		_, err := f.f.ReadAt(pcfile, s.Func.PCFile.Offset)
+		if err != nil {
+			return "", 0, nil
+		}
+		fileID := gosym.PCValue(pcfile, pc-uint64(s.Data.Offset), arch.MinLC)
+		fileName := s.Func.File[fileID]
+		pcline := make([]byte, s.Func.PCLine.Size)
+		_, err = f.f.ReadAt(pcline, s.Func.PCLine.Offset)
+		if err != nil {
+			return "", 0, nil
+		}
+		line := gosym.PCValue(pcline, pc-uint64(s.Data.Offset), arch.MinLC)
+		// Note: we provide only the name in the Func structure.
+		// We could provide more if needed.
+		return fileName, line, &gosym.Func{Sym: &gosym.Sym{Name: s.Name}}
+	}
+	return "", 0, nil
+}
+
+// We treat the whole object file as the text section.
 func (f *goobjFile) text() (textStart uint64, text []byte, err error) {
-	return 0, nil, fmt.Errorf("text not available in go object file")
+	var info os.FileInfo
+	info, err = f.f.Stat()
+	if err != nil {
+		return
+	}
+	text = make([]byte, info.Size())
+	_, err = f.f.ReadAt(text, 0)
+	return
 }
 
-// goarch makes sense but is not exposed in debug/goobj's API,
-// and we don't need it yet for any users of internal/objfile.
 func (f *goobjFile) goarch() string {
-	return "GOARCH unimplemented for debug/goobj files"
+	return f.goobj.Arch
 }
 
 func (f *goobjFile) loadAddress() (uint64, error) {

src/cmd/internal/objfile/objfile.go

@@ -35,6 +35,19 @@ type Sym struct {
 	Size   int64   // size in bytes
 	Code   rune    // nm code (T for text, D for data, and so on)
 	Type   string  // XXX?
+	Relocs []Reloc // in increasing Addr order
+}
+
+type Reloc struct {
+	Addr     uint64 // Address of first byte that reloc applies to.
+	Size     uint64 // Number of bytes
+	Stringer RelocStringer
+}
+
+type RelocStringer interface {
+	// insnOffset is the offset of the instruction containing the relocation
+	// from the start of the symbol containing the relocation.
+	String(insnOffset uint64) string
 }
 
 var openers = []func(*os.File) (rawFile, error){
@@ -80,7 +93,13 @@ func (x byAddr) Less(i, j int) bool { return x[i].Addr < x[j].Addr }
 func (x byAddr) Len() int           { return len(x) }
 func (x byAddr) Swap(i, j int)      { x[i], x[j] = x[j], x[i] }
 
-func (f *File) PCLineTable() (*gosym.Table, error) {
+func (f *File) PCLineTable() (Liner, error) {
+	// If the raw file implements Liner directly, use that.
+	// Currently, only Go intermediate objects and archives (goobj) use this path.
+	if pcln, ok := f.raw.(Liner); ok {
+		return pcln, nil
+	}
+	// Otherwise, read the pcln tables and build a Liner out of that.
 	textStart, symtab, pclntab, err := f.raw.pcln()
 	if err != nil {
 		return nil, err

src/cmd/internal/sys/arch.go

@@ -146,3 +146,16 @@ var ArchS390X = &Arch{
 	RegSize:   8,
 	MinLC:     2,
 }
+
+var Archs = [...]*Arch{
+	Arch386,
+	ArchAMD64,
+	ArchAMD64P32,
+	ArchARM,
+	ArchARM64,
+	ArchMIPS64,
+	ArchMIPS64LE,
+	ArchPPC64,
+	ArchPPC64LE,
+	ArchS390X,
+}

src/cmd/link/internal/arm/asm.go

@@ -492,12 +492,12 @@ func archrelocvariant(ctxt *ld.Link, r *ld.Reloc, s *ld.Symbol, t int64) int64 {
 	return t
 }
 
-func addpltreloc(ctxt *ld.Link, plt *ld.Symbol, got *ld.Symbol, sym *ld.Symbol, typ int) *ld.Reloc {
+func addpltreloc(ctxt *ld.Link, plt *ld.Symbol, got *ld.Symbol, sym *ld.Symbol, typ obj.RelocType) *ld.Reloc {
 	r := ld.Addrel(plt)
 	r.Sym = got
 	r.Off = int32(plt.Size)
 	r.Siz = 4
-	r.Type = int32(typ)
+	r.Type = typ
 	r.Add = int64(sym.Got) - 8
 
 	plt.Attr |= ld.AttrReachable

src/cmd/link/internal/ld/ldelf.go

@@ -912,7 +912,7 @@ func ldelf(ctxt *Link, f *bio.Reader, pkg string, length int64, pn string) {
 				rp.Sym = sym.sym
 			}
 
-			rp.Type = 256 + int32(info)
+			rp.Type = 256 + obj.RelocType(info)
 			rp.Siz = relSize(ctxt, pn, uint32(info))
 			if rela != 0 {
 				rp.Add = int64(add)

src/cmd/link/internal/ld/ldmacho.go

@@ -828,7 +828,7 @@ func ldmacho(ctxt *Link, f *bio.Reader, pkg string, length int64, pn string) {
 			}
 
 			rp.Siz = rel.length
-			rp.Type = 512 + (int32(rel.type_) << 1) + int32(rel.pcrel)
+			rp.Type = 512 + (obj.RelocType(rel.type_) << 1) + obj.RelocType(rel.pcrel)
 			rp.Off = int32(rel.addr)
 
 			// Handle X86_64_RELOC_SIGNED referencing a section (rel->extrn == 0).

src/cmd/link/internal/ld/link.go

@@ -32,6 +32,7 @@ package ld
 
 import (
 	"bufio"
+	"cmd/internal/obj"
 	"cmd/internal/sys"
 	"debug/elf"
 	"fmt"
@@ -135,7 +136,7 @@ type Reloc struct {
 	Off     int32
 	Siz     uint8
 	Done    uint8
-	Type    int32
+	Type    obj.RelocType
 	Variant int32
 	Add     int64
 	Xadd    int64

src/cmd/link/internal/ld/objfile.go

@@ -326,7 +326,7 @@ overwrite:
 			s.R[i] = Reloc{
 				Off:  r.readInt32(),
 				Siz:  r.readUint8(),
-				Type: r.readInt32(),
+				Type: obj.RelocType(r.readInt32()),
 				Add:  r.readInt64(),
 				Sym:  r.readSymIndex(),
 			}

src/cmd/pprof/pprof.go

@@ -6,7 +6,6 @@ package main
 
 import (
 	"debug/dwarf"
-	"debug/gosym"
 	"flag"
 	"fmt"
 	"net/url"
@@ -161,7 +160,7 @@ func (t *objTool) Disasm(file string, start, end uint64) ([]plugin.Inst, error)
 		return nil, err
 	}
 	var asm []plugin.Inst
-	d.Decode(start, end, func(pc, size uint64, file string, line int, text string) {
+	d.Decode(start, end, nil, func(pc, size uint64, file string, line int, text string) {
 		asm = append(asm, plugin.Inst{Addr: pc, File: file, Line: line, Text: text})
 	})
 	return asm, nil
@@ -203,7 +202,7 @@ type file struct {
 	offset uint64
 	sym    []objfile.Sym
 	file   *objfile.File
-	pcln   *gosym.Table
+	pcln   objfile.Liner
 
 	triedDwarf bool
 	dwarf      *dwarf.Data

src/debug/gosym/pclntab.go

@@ -291,13 +291,17 @@ func (t *LineTable) step(p *[]byte, pc *uint64, val *int32, first bool) bool {
 	return true
 }
 
+// PCValue looks up the given PC in a pc value table. target is the
+// offset of the pc from the entry point.
+func PCValue(tab []byte, target uint64, quantum int) int {
+	t := LineTable{Data: tab, quantum: uint32(quantum)}
+	return int(t.pcvalue(0, 0, target))
+}
+
 // pcvalue reports the value associated with the target pc.
 // off is the offset to the beginning of the pc-value table,
 // and entry is the start PC for the corresponding function.
 func (t *LineTable) pcvalue(off uint32, entry, targetpc uint64) int32 {
-	if off == 0 {
-		return -1
-	}
 	p := t.Data[off:]
 
 	val := int32(-1)