runtime/pprof: add GNU build IDs to Mappings recorded from /proc/self/maps

This helps systems that maintain an external database mapping
build ID to symbol information for the given binary, especially
in the case where /proc/self/maps lists many different files
(for example, many shared libraries).

Avoid importing debug/elf to avoid dragging in that whole
package (and its dependencies like debug/dwarf) into the
build of every program that generates a profile.

Fixes #19431.

Change-Id: I6d4362a79fe23e4f1726dffb0661d20bb57f766f
Reviewed-on: https://go-review.googlesource.com/37855
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>

8 files changed, 126 insertions, 4 deletions

diff --git a/src/runtime/pprof/elf.go b/src/runtime/pprof/elf.go
new file mode 100644
index 0000000000..a8b5ea6817
--- /dev/null
+++ b/src/runtime/pprof/elf.go
@@ -0,0 +1,109 @@
+// Copyright 2017 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.
+
+package pprof
+
+import (
+	"encoding/binary"
+	"errors"
+	"fmt"
+	"os"
+)
+
+var (
+	errBadELF    = errors.New("malformed ELF binary")
+	errNoBuildID = errors.New("no NT_GNU_BUILD_ID found in ELF binary")
+)
+
+// elfBuildID returns the GNU build ID of the named ELF binary,
+// without introducing a dependency on debug/elf and its dependencies.
+func elfBuildID(file string) (string, error) {
+	buf := make([]byte, 256)
+	f, err := os.Open(file)
+	if err != nil {
+		return "", err
+	}
+	defer f.Close()
+
+	if _, err := f.ReadAt(buf[:64], 0); err != nil {
+		return "", err
+	}
+
+	// ELF file begins with \x7F E L F.
+	if buf[0] != 0x7F || buf[1] != 'E' || buf[2] != 'L' || buf[3] != 'F' {
+		return "", errBadELF
+	}
+
+	var byteOrder binary.ByteOrder
+	switch buf[5] {
+	default:
+		return "", errBadELF
+	case 1: // little-endian
+		byteOrder = binary.LittleEndian
+	case 2: // big-endian
+		byteOrder = binary.BigEndian
+	}
+
+	var shnum int
+	var shoff, shentsize int64
+	switch buf[4] {
+	default:
+		return "", errBadELF
+	case 1: // 32-bit file header
+		shoff = int64(byteOrder.Uint32(buf[32:]))
+		shentsize = int64(byteOrder.Uint16(buf[46:]))
+		if shentsize != 40 {
+			return "", errBadELF
+		}
+		shnum = int(byteOrder.Uint16(buf[48:]))
+	case 2: // 64-bit file header
+		shoff = int64(byteOrder.Uint64(buf[40:]))
+		shentsize = int64(byteOrder.Uint16(buf[58:]))
+		if shentsize != 64 {
+			return "", errBadELF
+		}
+		shnum = int(byteOrder.Uint16(buf[60:]))
+	}
+
+	for i := 0; i < shnum; i++ {
+		if _, err := f.ReadAt(buf[:shentsize], shoff+int64(i)*shentsize); err != nil {
+			return "", err
+		}
+		if typ := byteOrder.Uint32(buf[4:]); typ != 7 { // SHT_NOTE
+			continue
+		}
+		var off, size int64
+		if shentsize == 40 {
+			// 32-bit section header
+			off = int64(byteOrder.Uint32(buf[16:]))
+			size = int64(byteOrder.Uint32(buf[20:]))
+		} else {
+			// 64-bit section header
+			off = int64(byteOrder.Uint64(buf[24:]))
+			size = int64(byteOrder.Uint64(buf[32:]))
+		}
+		size += off
+		for off < size {
+			if _, err := f.ReadAt(buf[:16], off); err != nil { // room for header + name GNU\x00
+				return "", err
+			}
+			nameSize := int(byteOrder.Uint32(buf[0:]))
+			descSize := int(byteOrder.Uint32(buf[4:]))
+			noteType := int(byteOrder.Uint32(buf[8:]))
+			descOff := off + int64(12+(nameSize+3)&^3)
+			off = descOff + int64((descSize+3)&^3)
+			if nameSize != 4 || noteType != 3 || buf[12] != 'G' || buf[13] != 'N' || buf[14] != 'U' || buf[15] != '\x00' { // want name GNU\x00 type 3 (NT_GNU_BUILD_ID)
+				continue
+			}
+			if descSize > len(buf) {
+				return "", errBadELF
+			}
+			if _, err := f.ReadAt(buf[:descSize], descOff); err != nil {
+				return "", err
+			}
+			return fmt.Sprintf("%x", buf[:descSize]), nil
+		}
+	}
+	return "", errNoBuildID
+}
diff --git a/src/runtime/pprof/proto.go b/src/runtime/pprof/proto.go
index dd3d5c3b0b..0f74e119b6 100644
--- a/src/runtime/pprof/proto.go
+++ b/src/runtime/pprof/proto.go
@@ -169,13 +169,14 @@ func (b *profileBuilder) pbLine(tag int, funcID uint64, line int64) {
 }
 
 // pbMapping encodes a Mapping message to b.pb.
-func (b *profileBuilder) pbMapping(tag int, id, base, limit, offset uint64, file string) {
+func (b *profileBuilder) pbMapping(tag int, id, base, limit, offset uint64, file, buildID string) {
 	start := b.pb.startMessage()
 	b.pb.uint64Opt(tagMapping_ID, id)
 	b.pb.uint64Opt(tagMapping_Start, base)
 	b.pb.uint64Opt(tagMapping_Limit, limit)
 	b.pb.uint64Opt(tagMapping_Offset, offset)
 	b.pb.int64Opt(tagMapping_Filename, b.stringIndex(file))
+	b.pb.int64Opt(tagMapping_BuildID, b.stringIndex(buildID))
 	// TODO: Set any of HasInlineFrames, HasFunctions, HasFilenames, HasLineNumbers?
 	// It seems like they should all be true, but they've never been set.
 	b.pb.endMessage(tag, start)
@@ -438,10 +439,10 @@ func (b *profileBuilder) readMapping() {
 		}
 		next() // dev
 		next() // inode
-		file := line
-		if file == nil {
+		if line == nil {
 			continue
 		}
+		file := string(line)
 
 		// TODO: pprof's remapMappingIDs makes two adjustments:
 		// 1. If there is an /anon_hugepage mapping first and it is
@@ -452,7 +453,8 @@ func (b *profileBuilder) readMapping() {
 		// If we do need them, they would go here, before we
 		// enter the mappings into b.mem in the first place.
 
+		buildID, _ := elfBuildID(file)
 		b.mem = append(b.mem, memMap{uintptr(lo), uintptr(hi)})
-		b.pbMapping(tagProfile_Mapping, uint64(len(b.mem)), lo, hi, offset, string(file))
+		b.pbMapping(tagProfile_Mapping, uint64(len(b.mem)), lo, hi, offset, file, buildID)
 	}
 }
diff --git a/src/runtime/pprof/proto_test.go b/src/runtime/pprof/proto_test.go
index 98f217583b..7b0fa95d71 100644
--- a/src/runtime/pprof/proto_test.go
+++ b/src/runtime/pprof/proto_test.go
@@ -87,8 +87,10 @@ func testPCs(t *testing.T) (addr1, addr2 uint64, map1, map2 *profile.Mapping) {
 		}
 		addr1 = mprof.Mapping[0].Start
 		map1 = mprof.Mapping[0]
+		map1.BuildID, _ = elfBuildID(map1.File)
 		addr2 = mprof.Mapping[1].Start
 		map2 = mprof.Mapping[1]
+		map2.BuildID, _ = elfBuildID(map2.File)
 	} else {
 		addr1 = uint64(funcPC(f1))
 		addr2 = uint64(funcPC(f2))
diff --git a/src/runtime/pprof/testdata/README b/src/runtime/pprof/testdata/README
new file mode 100644
index 0000000000..876538e5d5
--- /dev/null
+++ b/src/runtime/pprof/testdata/README
@@ -0,0 +1,9 @@
+These binaries were generated by:
+
+$ cat empty.s
+.global _start
+_start:
+$ as --32 -o empty.o empty.s && ld  --build-id -m elf_i386 -o test32 empty.o
+$ as --64 -o empty.o empty.s && ld --build-id -o test64 empty.o
+$ powerpc-linux-gnu-as -o empty.o empty.s && powerpc-linux-gnu-ld --build-id -o test32be empty.o
+$ powerpc64-linux-gnu-as -o empty.o empty.s && powerpc64-linux-gnu-ld --build-id -o test64be empty.o
diff --git a/src/runtime/pprof/testdata/test32 b/src/runtime/pprof/testdata/test32
new file mode 100755
index 0000000000..ce594720db
--- /dev/null
+++ b/src/runtime/pprof/testdata/test32
diff --git a/src/runtime/pprof/testdata/test32be b/src/runtime/pprof/testdata/test32be
new file mode 100755
index 0000000000..f13a732203
--- /dev/null
+++ b/src/runtime/pprof/testdata/test32be
diff --git a/src/runtime/pprof/testdata/test64 b/src/runtime/pprof/testdata/test64
new file mode 100755
index 0000000000..3fb42fb3bf
--- /dev/null
+++ b/src/runtime/pprof/testdata/test64
diff --git a/src/runtime/pprof/testdata/test64be b/src/runtime/pprof/testdata/test64be
new file mode 100755
index 0000000000..09b4b01894
--- /dev/null
+++ b/src/runtime/pprof/testdata/test64be