cmd/addr2line, cmd/nm: factor object reading into cmd/internal/objfile

To do in another CL: make cmd/objdump use cmd/internal/objfile too.

There is a package placement decision in this CL:
cmd/internal/objfile instead of internal/objfile.
I chose to put internal under cmd to make clear (and enforce)
that no standard library packages should use this
(it's a bit dependency-heavy).

LGTM=r
R=r
CC=golang-codereviews
https://golang.org/cl/123910043

7 files changed, 14 insertions, 387 deletions

diff --git a/src/cmd/nm/elf.go b/src/cmd/nm/elf.go
deleted file mode 100644
index 5aaa194dd1..0000000000
--- a/src/cmd/nm/elf.go
+++ /dev/null
@@ -1,57 +0,0 @@
-// Copyright 2013 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.
-
-// Parsing of ELF executables (Linux, FreeBSD, and so on).
-
-package main
-
-import (
-	"debug/elf"
-	"os"
-)
-
-func elfSymbols(f *os.File) []Sym {
-	p, err := elf.NewFile(f)
-	if err != nil {
-		errorf("parsing %s: %v", f.Name(), err)
-		return nil
-	}
-
-	elfSyms, err := p.Symbols()
-	if err != nil {
-		errorf("parsing %s: %v", f.Name(), err)
-		return nil
-	}
-
-	var syms []Sym
-	for _, s := range elfSyms {
-		sym := Sym{Addr: s.Value, Name: s.Name, Size: int64(s.Size), Code: '?'}
-		switch s.Section {
-		case elf.SHN_UNDEF:
-			sym.Code = 'U'
-		case elf.SHN_COMMON:
-			sym.Code = 'B'
-		default:
-			i := int(s.Section)
-			if i < 0 || i >= len(p.Sections) {
-				break
-			}
-			sect := p.Sections[i]
-			switch sect.Flags & (elf.SHF_WRITE | elf.SHF_ALLOC | elf.SHF_EXECINSTR) {
-			case elf.SHF_ALLOC | elf.SHF_EXECINSTR:
-				sym.Code = 'T'
-			case elf.SHF_ALLOC:
-				sym.Code = 'R'
-			case elf.SHF_ALLOC | elf.SHF_WRITE:
-				sym.Code = 'D'
-			}
-		}
-		if elf.ST_BIND(s.Info) == elf.STB_LOCAL {
-			sym.Code += 'a' - 'A'
-		}
-		syms = append(syms, sym)
-	}
-
-	return syms
-}
diff --git a/src/cmd/nm/goobj.go b/src/cmd/nm/goobj.go
deleted file mode 100644
index b0de51db9c..0000000000
--- a/src/cmd/nm/goobj.go
+++ /dev/null
@@ -1,68 +0,0 @@
-// Copyright 2013 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.
-
-// Parsing of Go intermediate object files and archives.
-
-package main
-
-import (
-	"debug/goobj"
-	"fmt"
-	"os"
-)
-
-func goobjName(id goobj.SymID) string {
-	if id.Version == 0 {
-		return id.Name
-	}
-	return fmt.Sprintf("%s<%d>", id.Name, id.Version)
-}
-
-func goobjSymbols(f *os.File) []Sym {
-	pkg, err := goobj.Parse(f, `""`)
-	if err != nil {
-		errorf("parsing %s: %v", f.Name(), err)
-		return nil
-	}
-
-	seen := make(map[goobj.SymID]bool)
-
-	var syms []Sym
-	for _, s := range pkg.Syms {
-		seen[s.SymID] = true
-		sym := Sym{Addr: uint64(s.Data.Offset), Name: goobjName(s.SymID), Size: int64(s.Size), Type: s.Type.Name, Code: '?'}
-		switch s.Kind {
-		case goobj.STEXT, goobj.SELFRXSECT:
-			sym.Code = 'T'
-		case goobj.STYPE, goobj.SSTRING, goobj.SGOSTRING, goobj.SGOFUNC, goobj.SRODATA, goobj.SFUNCTAB, goobj.STYPELINK, goobj.SSYMTAB, goobj.SPCLNTAB, goobj.SELFROSECT:
-			sym.Code = 'R'
-		case goobj.SMACHOPLT, goobj.SELFSECT, goobj.SMACHO, goobj.SMACHOGOT, goobj.SNOPTRDATA, goobj.SINITARR, goobj.SDATA, goobj.SWINDOWS:
-			sym.Code = 'D'
-		case goobj.SBSS, goobj.SNOPTRBSS, goobj.STLSBSS:
-			sym.Code = 'B'
-		case goobj.SXREF, goobj.SMACHOSYMSTR, goobj.SMACHOSYMTAB, goobj.SMACHOINDIRECTPLT, goobj.SMACHOINDIRECTGOT, goobj.SFILE, goobj.SFILEPATH, goobj.SCONST, goobj.SDYNIMPORT, goobj.SHOSTOBJ:
-			sym.Code = 'X' // should not see
-		}
-		if s.Version != 0 {
-			sym.Code += 'a' - 'A'
-		}
-		syms = append(syms, sym)
-	}
-
-	for _, s := range pkg.Syms {
-		for _, r := range s.Reloc {
-			if !seen[r.Sym] {
-				seen[r.Sym] = true
-				sym := Sym{Name: goobjName(r.Sym), Code: 'U'}
-				if s.Version != 0 {
-					// should not happen but handle anyway
-					sym.Code = 'u'
-				}
-				syms = append(syms, sym)
-			}
-		}
-	}
-
-	return syms
-}
diff --git a/src/cmd/nm/macho.go b/src/cmd/nm/macho.go
deleted file mode 100644
index c60bde55b4..0000000000
--- a/src/cmd/nm/macho.go
+++ /dev/null
@@ -1,69 +0,0 @@
-// Copyright 2013 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.
-
-// Parsing of Mach-O executables (OS X).
-
-package main
-
-import (
-	"debug/macho"
-	"os"
-	"sort"
-)
-
-func machoSymbols(f *os.File) []Sym {
-	p, err := macho.NewFile(f)
-	if err != nil {
-		errorf("parsing %s: %v", f.Name(), err)
-		return nil
-	}
-
-	if p.Symtab == nil {
-		errorf("%s: no symbol table", f.Name())
-		return nil
-	}
-
-	// Build sorted list of addresses of all symbols.
-	// We infer the size of a symbol by looking at where the next symbol begins.
-	var addrs []uint64
-	for _, s := range p.Symtab.Syms {
-		addrs = append(addrs, s.Value)
-	}
-	sort.Sort(uint64s(addrs))
-
-	var syms []Sym
-	for _, s := range p.Symtab.Syms {
-		sym := Sym{Name: s.Name, Addr: s.Value, Code: '?'}
-		i := sort.Search(len(addrs), func(x int) bool { return addrs[x] > s.Value })
-		if i < len(addrs) {
-			sym.Size = int64(addrs[i] - s.Value)
-		}
-		if s.Sect == 0 {
-			sym.Code = 'U'
-		} else if int(s.Sect) <= len(p.Sections) {
-			sect := p.Sections[s.Sect-1]
-			switch sect.Seg {
-			case "__TEXT":
-				sym.Code = 'R'
-			case "__DATA":
-				sym.Code = 'D'
-			}
-			switch sect.Seg + " " + sect.Name {
-			case "__TEXT __text":
-				sym.Code = 'T'
-			case "__DATA __bss", "__DATA __noptrbss":
-				sym.Code = 'B'
-			}
-		}
-		syms = append(syms, sym)
-	}
-
-	return syms
-}
-
-type uint64s []uint64
-
-func (x uint64s) Len() int           { return len(x) }
-func (x uint64s) Swap(i, j int)      { x[i], x[j] = x[j], x[i] }
-func (x uint64s) Less(i, j int) bool { return x[i] < x[j] }
diff --git a/src/cmd/nm/nm.go b/src/cmd/nm/nm.go
index a4036184e4..3089e481be 100644
--- a/src/cmd/nm/nm.go
+++ b/src/cmd/nm/nm.go
@@ -6,13 +6,13 @@ package main
 
 import (
 	"bufio"
-	"bytes"
 	"flag"
 	"fmt"
-	"io"
 	"log"
 	"os"
 	"sort"
+
+	"cmd/internal/objfile"
 )
 
 func usage() {
@@ -85,55 +85,22 @@ func errorf(format string, args ...interface{}) {
 	exitCode = 1
 }
 
-type Sym struct {
-	Addr uint64
-	Size int64
-	Code rune
-	Name string
-	Type string
-}
-
-var parsers = []struct {
-	prefix []byte
-	parse  func(*os.File) []Sym
-}{
-	{[]byte("!<arch>\n"), goobjSymbols},
-	{[]byte("go object "), goobjSymbols},
-	{[]byte("\x7FELF"), elfSymbols},
-	{[]byte("\xFE\xED\xFA\xCE"), machoSymbols},
-	{[]byte("\xFE\xED\xFA\xCF"), machoSymbols},
-	{[]byte("\xCE\xFA\xED\xFE"), machoSymbols},
-	{[]byte("\xCF\xFA\xED\xFE"), machoSymbols},
-	{[]byte("MZ"), peSymbols},
-	{[]byte("\x00\x00\x01\xEB"), plan9Symbols}, // 386
-	{[]byte("\x00\x00\x04\x07"), plan9Symbols}, // mips
-	{[]byte("\x00\x00\x06\x47"), plan9Symbols}, // arm
-	{[]byte("\x00\x00\x8A\x97"), plan9Symbols}, // amd64
-}
-
 func nm(file string) {
-	f, err := os.Open(file)
+	f, err := objfile.Open(file)
 	if err != nil {
 		errorf("%v", err)
 		return
 	}
 	defer f.Close()
 
-	buf := make([]byte, 16)
-	io.ReadFull(f, buf)
-	f.Seek(0, 0)
-
-	var syms []Sym
-	for _, p := range parsers {
-		if bytes.HasPrefix(buf, p.prefix) {
-			syms = p.parse(f)
-			goto HaveSyms
-		}
+	syms, err := f.Symbols()
+	if err != nil {
+		errorf("reading %s: %v", file, err)
+	}
+	if len(syms) == 0 {
+		errorf("reading %s: no symbols", file)
 	}
-	errorf("%v: unknown file format", file)
-	return
 
-HaveSyms:
 	switch *sortOrder {
 	case "address":
 		sort.Sort(byAddr(syms))
@@ -165,19 +132,19 @@ HaveSyms:
 	w.Flush()
 }
 
-type byAddr []Sym
+type byAddr []objfile.Sym
 
 func (x byAddr) Len() int           { return len(x) }
 func (x byAddr) Swap(i, j int)      { x[i], x[j] = x[j], x[i] }
 func (x byAddr) Less(i, j int) bool { return x[i].Addr < x[j].Addr }
 
-type byName []Sym
+type byName []objfile.Sym
 
 func (x byName) Len() int           { return len(x) }
 func (x byName) Swap(i, j int)      { x[i], x[j] = x[j], x[i] }
 func (x byName) Less(i, j int) bool { return x[i].Name < x[j].Name }
 
-type bySize []Sym
+type bySize []objfile.Sym
 
 func (x bySize) Len() int           { return len(x) }
 func (x bySize) Swap(i, j int)      { x[i], x[j] = x[j], x[i] }
diff --git a/src/cmd/nm/nm_test.go b/src/cmd/nm/nm_test.go
index 74773877f3..f447e8e491 100644
--- a/src/cmd/nm/nm_test.go
+++ b/src/cmd/nm/nm_test.go
@@ -77,7 +77,7 @@ func TestNM(t *testing.T) {
 		"elf/testdata/gcc-amd64-linux-exec",
 		"macho/testdata/gcc-386-darwin-exec",
 		"macho/testdata/gcc-amd64-darwin-exec",
-		"pe/testdata/gcc-amd64-mingw-exec",
+		// "pe/testdata/gcc-amd64-mingw-exec", // no symbols!
 		"pe/testdata/gcc-386-mingw-exec",
 		"plan9obj/testdata/amd64-plan9-exec",
 		"plan9obj/testdata/386-plan9-exec",
@@ -87,7 +87,7 @@ func TestNM(t *testing.T) {
 		cmd := exec.Command(testnmpath, exepath)
 		out, err := cmd.CombinedOutput()
 		if err != nil {
-			t.Fatalf("go tool nm %v: %v\n%s", exepath, err, string(out))
+			t.Errorf("go tool nm %v: %v\n%s", exepath, err, string(out))
 		}
 	}
 
diff --git a/src/cmd/nm/pe.go b/src/cmd/nm/pe.go
deleted file mode 100644
index 52d05e51d0..0000000000
--- a/src/cmd/nm/pe.go
+++ /dev/null
@@ -1,98 +0,0 @@
-// Copyright 2013 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.
-
-// Parsing of PE executables (Microsoft Windows).
-
-package main
-
-import (
-	"debug/pe"
-	"os"
-	"sort"
-)
-
-func peSymbols(f *os.File) []Sym {
-	p, err := pe.NewFile(f)
-	if err != nil {
-		errorf("parsing %s: %v", f.Name(), err)
-		return nil
-	}
-
-	// Build sorted list of addresses of all symbols.
-	// We infer the size of a symbol by looking at where the next symbol begins.
-	var addrs []uint64
-
-	var imageBase uint64
-	switch oh := p.OptionalHeader.(type) {
-	case *pe.OptionalHeader32:
-		imageBase = uint64(oh.ImageBase)
-	case *pe.OptionalHeader64:
-		imageBase = oh.ImageBase
-	default:
-		errorf("parsing %s: file format not recognized", f.Name())
-		return nil
-	}
-
-	var syms []Sym
-	for _, s := range p.Symbols {
-		const (
-			N_UNDEF = 0  // An undefined (extern) symbol
-			N_ABS   = -1 // An absolute symbol (e_value is a constant, not an address)
-			N_DEBUG = -2 // A debugging symbol
-		)
-		sym := Sym{Name: s.Name, Addr: uint64(s.Value), Code: '?'}
-		switch s.SectionNumber {
-		case N_UNDEF:
-			sym.Code = 'U'
-		case N_ABS:
-			sym.Code = 'C'
-		case N_DEBUG:
-			sym.Code = '?'
-		default:
-			if s.SectionNumber < 0 {
-				errorf("parsing %s: invalid section number %d", f.Name(), s.SectionNumber)
-				return nil
-			}
-			if len(p.Sections) < int(s.SectionNumber) {
-				errorf("parsing %s: section number %d is large then max %d", f.Name(), s.SectionNumber, len(p.Sections))
-				return nil
-			}
-			sect := p.Sections[s.SectionNumber-1]
-			const (
-				text  = 0x20
-				data  = 0x40
-				bss   = 0x80
-				permX = 0x20000000
-				permR = 0x40000000
-				permW = 0x80000000
-			)
-			ch := sect.Characteristics
-			switch {
-			case ch&text != 0:
-				sym.Code = 'T'
-			case ch&data != 0:
-				if ch&permW == 0 {
-					sym.Code = 'R'
-				} else {
-					sym.Code = 'D'
-				}
-			case ch&bss != 0:
-				sym.Code = 'B'
-			}
-			sym.Addr += imageBase + uint64(sect.VirtualAddress)
-		}
-		syms = append(syms, sym)
-		addrs = append(addrs, sym.Addr)
-	}
-
-	sort.Sort(uint64s(addrs))
-	for i := range syms {
-		j := sort.Search(len(addrs), func(x int) bool { return addrs[x] > syms[i].Addr })
-		if j < len(addrs) {
-			syms[i].Size = int64(addrs[j] - syms[i].Addr)
-		}
-	}
-
-	return syms
-}
diff --git a/src/cmd/nm/plan9obj.go b/src/cmd/nm/plan9obj.go
deleted file mode 100644
index 006c66ebfd..0000000000
--- a/src/cmd/nm/plan9obj.go
+++ /dev/null
@@ -1,48 +0,0 @@
-// Copyright 2014 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.
-
-// Parsing of Plan 9 a.out executables.
-
-package main
-
-import (
-	"debug/plan9obj"
-	"os"
-	"sort"
-)
-
-func plan9Symbols(f *os.File) []Sym {
-	p, err := plan9obj.NewFile(f)
-	if err != nil {
-		errorf("parsing %s: %v", f.Name(), err)
-		return nil
-	}
-
-	plan9Syms, err := p.Symbols()
-	if err != nil {
-		errorf("parsing %s: %v", f.Name(), err)
-		return nil
-	}
-
-	// Build sorted list of addresses of all symbols.
-	// We infer the size of a symbol by looking at where the next symbol begins.
-	var addrs []uint64
-	for _, s := range plan9Syms {
-		addrs = append(addrs, s.Value)
-	}
-	sort.Sort(uint64s(addrs))
-
-	var syms []Sym
-
-	for _, s := range plan9Syms {
-		sym := Sym{Addr: s.Value, Name: s.Name, Code: rune(s.Type)}
-		i := sort.Search(len(addrs), func(x int) bool { return addrs[x] > s.Value })
-		if i < len(addrs) {
-			sym.Size = int64(addrs[i] - s.Value)
-		}
-		syms = append(syms, sym)
-	}
-
-	return syms
-}