debug/dwarf: add support for DWARFv5 to (*Data).Ranges

Updates the (*Data).Ranges method to work with DWARFv5 which uses the
new debug_rnglists section instead of debug_ranges.

This does not include supporting DW_FORM_rnglistx.

General support for DWARFv5 was added by CL 175138.

Change-Id: I01f919a865616a3ff12f5bf649c2c9abf89fcf52
Reviewed-on: https://go-review.googlesource.com/c/go/+/236657
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Trust: Emmanuel Odeke <emm.odeke@gmail.com>

5 files changed, 317 insertions, 59 deletions

diff --git a/src/debug/dwarf/const.go b/src/debug/dwarf/const.go
index b11bf90c37..c60709199b 100644
--- a/src/debug/dwarf/const.go
+++ b/src/debug/dwarf/const.go
@@ -461,3 +461,15 @@ const (
 	utSplitCompile = 0x05
 	utSplitType    = 0x06
 )
+
+// Opcodes for DWARFv5 debug_rnglists section.
+const (
+	rleEndOfList    = 0x0
+	rleBaseAddressx = 0x1
+	rleStartxEndx   = 0x2
+	rleStartxLength = 0x3
+	rleOffsetPair   = 0x4
+	rleBaseAddress  = 0x5
+	rleStartEnd     = 0x6
+	rleStartLength  = 0x7
+)
diff --git a/src/debug/dwarf/dwarf5ranges_test.go b/src/debug/dwarf/dwarf5ranges_test.go
new file mode 100644
index 0000000000..2229d439a5
--- /dev/null
+++ b/src/debug/dwarf/dwarf5ranges_test.go
@@ -0,0 +1,36 @@
+// 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.
+
+package dwarf
+
+import (
+	"encoding/binary"
+	"io/ioutil"
+	"reflect"
+	"testing"
+)
+
+func TestDwarf5Ranges(t *testing.T) {
+	rngLists, err := ioutil.ReadFile("testdata/debug_rnglists")
+	if err != nil {
+		t.Fatalf("could not read test data: %v", err)
+	}
+
+	d := &Data{}
+	d.order = binary.LittleEndian
+	if err := d.AddSection(".debug_rnglists", rngLists); err != nil {
+		t.Fatal(err)
+	}
+	ret, err := d.dwarf5Ranges(nil, 0x5fbd, 0xc, [][2]uint64{})
+	if err != nil {
+		t.Fatalf("could not read rnglist: %v", err)
+	}
+	t.Logf("%#v", ret)
+
+	tgt := [][2]uint64{{0x0000000000006712, 0x000000000000679f}, {0x00000000000067af}, {0x00000000000067b3}}
+
+	if reflect.DeepEqual(ret, tgt) {
+		t.Errorf("expected %#v got %#x", tgt, ret)
+	}
+}
diff --git a/src/debug/dwarf/entry.go b/src/debug/dwarf/entry.go
index 88eb56936b..bc05d7ef31 100644
--- a/src/debug/dwarf/entry.go
+++ b/src/debug/dwarf/entry.go
@@ -453,38 +453,28 @@ func (b *buf) entry(cu *Entry, atab abbrevTable, ubase Offset, vers int) *Entry
 			case formAddrx4:
 				off = uint64(b.uint32())
 			}
-			if len(b.dwarf.addr) == 0 {
+			if b.dwarf.addr == nil {
 				b.error("DW_FORM_addrx with no .debug_addr section")
 			}
 			if b.err != nil {
 				return nil
 			}
-			addrsize := b.format.addrsize()
-			if addrsize == 0 {
-				b.error("unknown address size for DW_FORM_addrx")
-			}
-			off *= uint64(addrsize)
 
 			// We have to adjust by the offset of the
 			// compilation unit. This won't work if the
 			// program uses Reader.Seek to skip over the
 			// unit. Not much we can do about that.
+			var addrBase int64
 			if cu != nil {
-				cuOff, ok := cu.Val(AttrAddrBase).(int64)
-				if ok {
-					off += uint64(cuOff)
-				}
+				addrBase, _ = cu.Val(AttrAddrBase).(int64)
 			}
 
-			if uint64(int(off)) != off {
-				b.error("DW_FORM_addrx offset out of range")
-			}
-
-			b1 := makeBuf(b.dwarf, b.format, "addr", 0, b.dwarf.addr)
-			b1.skip(int(off))
-			val = b1.addr()
-			if b1.err != nil {
-				b.err = b1.err
+			var err error
+			val, err = b.dwarf.debugAddr(uint64(addrBase), off)
+			if err != nil {
+				if b.err == nil {
+					b.err = err
+				}
 				return nil
 			}
 
@@ -935,53 +925,187 @@ func (d *Data) Ranges(e *Entry) ([][2]uint64, error) {
 		ret = append(ret, [2]uint64{low, high})
 	}
 
-	ranges, rangesOK := e.Val(AttrRanges).(int64)
-	if rangesOK && d.ranges != nil {
-		// The initial base address is the lowpc attribute
-		// of the enclosing compilation unit.
-		// Although DWARF specifies the lowpc attribute,
-		// comments in gdb/dwarf2read.c say that some versions
-		// of GCC use the entrypc attribute, so we check that too.
-		var cu *Entry
-		if e.Tag == TagCompileUnit {
-			cu = e
-		} else {
-			i := d.offsetToUnit(e.Offset)
-			if i == -1 {
-				return nil, errors.New("no unit for entry")
+	var u *unit
+	if uidx := d.offsetToUnit(e.Offset); uidx >= 0 && uidx < len(d.unit) {
+		u = &d.unit[uidx]
+	}
+
+	if u != nil && u.vers >= 5 && d.rngLists != nil {
+		// DWARF version 5 and later
+		field := e.AttrField(AttrRanges)
+		if field == nil {
+			return ret, nil
+		}
+		switch field.Class {
+		case ClassRangeListPtr:
+			ranges, rangesOK := field.Val.(int64)
+			if !rangesOK {
+				return ret, nil
 			}
-			u := &d.unit[i]
-			b := makeBuf(d, u, "info", u.off, u.data)
-			cu = b.entry(nil, u.atable, u.base, u.vers)
-			if b.err != nil {
-				return nil, b.err
+			cu, base, err := d.baseAddressForEntry(e)
+			if err != nil {
+				return nil, err
 			}
+			return d.dwarf5Ranges(cu, base, ranges, ret)
+
+		case ClassRngList:
+			// TODO: support DW_FORM_rnglistx
+			return ret, nil
+
+		default:
+			return ret, nil
 		}
+	}
 
-		var base uint64
-		if cuEntry, cuEntryOK := cu.Val(AttrEntrypc).(uint64); cuEntryOK {
-			base = cuEntry
-		} else if cuLow, cuLowOK := cu.Val(AttrLowpc).(uint64); cuLowOK {
-			base = cuLow
+	// DWARF version 2 through 4
+	ranges, rangesOK := e.Val(AttrRanges).(int64)
+	if rangesOK && d.ranges != nil {
+		_, base, err := d.baseAddressForEntry(e)
+		if err != nil {
+			return nil, err
 		}
+		return d.dwarf2Ranges(u, base, ranges, ret)
+	}
 
-		u := &d.unit[d.offsetToUnit(e.Offset)]
-		buf := makeBuf(d, u, "ranges", Offset(ranges), d.ranges[ranges:])
-		for len(buf.data) > 0 {
-			low = buf.addr()
-			high = buf.addr()
+	return ret, nil
+}
 
-			if low == 0 && high == 0 {
-				break
+// baseAddressForEntry returns the initial base address to be used when
+// looking up the range list of entry e.
+// DWARF specifies that this should be the lowpc attribute of the enclosing
+// compilation unit, however comments in gdb/dwarf2read.c say that some
+// versions of GCC use the entrypc attribute, so we check that too.
+func (d *Data) baseAddressForEntry(e *Entry) (*Entry, uint64, error) {
+	var cu *Entry
+	if e.Tag == TagCompileUnit {
+		cu = e
+	} else {
+		i := d.offsetToUnit(e.Offset)
+		if i == -1 {
+			return nil, 0, errors.New("no unit for entry")
+		}
+		u := &d.unit[i]
+		b := makeBuf(d, u, "info", u.off, u.data)
+		cu = b.entry(nil, u.atable, u.base, u.vers)
+		if b.err != nil {
+			return nil, 0, b.err
+		}
+	}
+
+	if cuEntry, cuEntryOK := cu.Val(AttrEntrypc).(uint64); cuEntryOK {
+		return cu, cuEntry, nil
+	} else if cuLow, cuLowOK := cu.Val(AttrLowpc).(uint64); cuLowOK {
+		return cu, cuLow, nil
+	}
+
+	return cu, 0, nil
+}
+
+func (d *Data) dwarf2Ranges(u *unit, base uint64, ranges int64, ret [][2]uint64) ([][2]uint64, error) {
+	buf := makeBuf(d, u, "ranges", Offset(ranges), d.ranges[ranges:])
+	for len(buf.data) > 0 {
+		low := buf.addr()
+		high := buf.addr()
+
+		if low == 0 && high == 0 {
+			break
+		}
+
+		if low == ^uint64(0)>>uint((8-u.addrsize())*8) {
+			base = high
+		} else {
+			ret = append(ret, [2]uint64{base + low, base + high})
+		}
+	}
+
+	return ret, nil
+}
+
+// dwarf5Ranges interpets a debug_rnglists sequence, see DWARFv5 section
+// 2.17.3 (page 53).
+func (d *Data) dwarf5Ranges(cu *Entry, base uint64, ranges int64, ret [][2]uint64) ([][2]uint64, error) {
+	var addrBase int64
+	if cu != nil {
+		addrBase, _ = cu.Val(AttrAddrBase).(int64)
+	}
+
+	buf := makeBuf(d, d.rngLists, "rnglists", 0, d.rngLists.data)
+	buf.skip(int(ranges))
+	for {
+		opcode := buf.uint8()
+		switch opcode {
+		case rleEndOfList:
+			if buf.err != nil {
+				return nil, buf.err
+			}
+			return ret, nil
+
+		case rleBaseAddressx:
+			baseIdx := buf.uint()
+			var err error
+			base, err = d.debugAddr(uint64(addrBase), baseIdx)
+			if err != nil {
+				return nil, err
 			}
 
-			if low == ^uint64(0)>>uint((8-u.addrsize())*8) {
-				base = high
-			} else {
-				ret = append(ret, [2]uint64{base + low, base + high})
+		case rleStartxEndx:
+			startIdx := buf.uint()
+			endIdx := buf.uint()
+
+			start, err := d.debugAddr(uint64(addrBase), startIdx)
+			if err != nil {
+				return nil, err
+			}
+			end, err := d.debugAddr(uint64(addrBase), endIdx)
+			if err != nil {
+				return nil, err
+			}
+			ret = append(ret, [2]uint64{start, end})
+
+		case rleStartxLength:
+			startIdx := buf.uint()
+			len := buf.uint()
+			start, err := d.debugAddr(uint64(addrBase), startIdx)
+			if err != nil {
+				return nil, err
 			}
+			ret = append(ret, [2]uint64{start, start + len})
+
+		case rleOffsetPair:
+			off1 := buf.uint()
+			off2 := buf.uint()
+			ret = append(ret, [2]uint64{base + off1, base + off2})
+
+		case rleBaseAddress:
+			base = buf.addr()
+
+		case rleStartEnd:
+			start := buf.addr()
+			end := buf.addr()
+			ret = append(ret, [2]uint64{start, end})
+
+		case rleStartLength:
+			start := buf.addr()
+			len := buf.uint()
+			ret = append(ret, [2]uint64{start, start + len})
 		}
 	}
+}
 
-	return ret, nil
+// debugAddr returns the address at idx in debug_addr
+func (d *Data) debugAddr(addrBase, idx uint64) (uint64, error) {
+	off := idx*uint64(d.addr.addrsize()) + addrBase
+
+	if uint64(int(off)) != off {
+		return 0, errors.New("offset out of range")
+	}
+
+	b := makeBuf(d, d.addr, "addr", 0, d.addr.data)
+	b.skip(int(off))
+	val := b.addr()
+	if b.err != nil {
+		return 0, b.err
+	}
+
+	return val, nil
 }
diff --git a/src/debug/dwarf/open.go b/src/debug/dwarf/open.go
index 72ee64d558..617b8c56dd 100644
--- a/src/debug/dwarf/open.go
+++ b/src/debug/dwarf/open.go
@@ -7,7 +7,10 @@
 // http://dwarfstd.org/doc/dwarf-2.0.0.pdf
 package dwarf
 
-import "encoding/binary"
+import (
+	"encoding/binary"
+	"errors"
+)
 
 // Data represents the DWARF debugging information
 // loaded from an executable file (for example, an ELF or Mach-O executable).
@@ -23,9 +26,10 @@ type Data struct {
 	str      []byte
 
 	// New sections added in DWARF 5.
-	addr       []byte
+	addr       *debugAddr
 	lineStr    []byte
 	strOffsets []byte
+	rngLists   *rngLists
 
 	// parsed data
 	abbrevCache map[uint64]abbrevTable
@@ -36,6 +40,23 @@ type Data struct {
 	unit        []unit
 }
 
+// rngLists represents the contents of a debug_rnglists section (DWARFv5).
+type rngLists struct {
+	is64  bool
+	asize uint8
+	data  []byte
+	ver   uint16
+}
+
+// debugAddr represents the contents of a debug_addr section (DWARFv5).
+type debugAddr struct {
+	is64  bool
+	asize uint8
+	data  []byte
+}
+
+var errSegmentSelector = errors.New("non-zero segment_selector size not supported")
+
 // New returns a new Data object initialized from the given parameters.
 // Rather than calling this function directly, clients should typically use
 // the DWARF method of the File type of the appropriate package debug/elf,
@@ -108,14 +129,79 @@ func (d *Data) AddTypes(name string, types []byte) error {
 // so forth. This approach is used for new DWARF sections added in
 // DWARF 5 and later.
 func (d *Data) AddSection(name string, contents []byte) error {
+	var err error
 	switch name {
 	case ".debug_addr":
-		d.addr = contents
+		d.addr, err = d.parseAddrHeader(contents)
 	case ".debug_line_str":
 		d.lineStr = contents
 	case ".debug_str_offsets":
 		d.strOffsets = contents
+	case ".debug_rnglists":
+		d.rngLists, err = d.parseRngListsHeader(contents)
 	}
 	// Just ignore names that we don't yet support.
-	return nil
+	return err
+}
+
+// parseRngListsHeader reads the header of a debug_rnglists section, see
+// DWARFv5 section 7.28 (page 242).
+func (d *Data) parseRngListsHeader(bytes []byte) (*rngLists, error) {
+	rngLists := &rngLists{data: bytes}
+
+	buf := makeBuf(d, unknownFormat{}, "rnglists", 0, bytes)
+	_, rngLists.is64 = buf.unitLength()
+
+	rngLists.ver = buf.uint16() // version
+
+	rngLists.asize = buf.uint8()
+	segsize := buf.uint8()
+	if segsize != 0 {
+		return nil, errSegmentSelector
+	}
+
+	// Header fields not read: offset_entry_count, offset table
+
+	return rngLists, nil
+}
+
+func (rngLists *rngLists) version() int {
+	return int(rngLists.ver)
+}
+
+func (rngLists *rngLists) dwarf64() (bool, bool) {
+	return rngLists.is64, true
+}
+
+func (rngLists *rngLists) addrsize() int {
+	return int(rngLists.asize)
+}
+
+// parseAddrHeader reads the header of a debug_addr section, see DWARFv5
+// section 7.27 (page 241).
+func (d *Data) parseAddrHeader(bytes []byte) (*debugAddr, error) {
+	addr := &debugAddr{data: bytes}
+
+	buf := makeBuf(d, unknownFormat{}, "addr", 0, bytes)
+	_, addr.is64 = buf.unitLength()
+
+	addr.asize = buf.uint8()
+	segsize := buf.uint8()
+	if segsize != 0 {
+		return nil, errSegmentSelector
+	}
+
+	return addr, nil
+}
+
+func (addr *debugAddr) version() int {
+	return 5
+}
+
+func (addr *debugAddr) dwarf64() (bool, bool) {
+	return addr.is64, true
+}
+
+func (addr *debugAddr) addrsize() int {
+	return int(addr.asize)
 }
diff --git a/src/debug/dwarf/testdata/debug_rnglists b/src/debug/dwarf/testdata/debug_rnglists
new file mode 100644
index 0000000000..985ec6c9f2
--- /dev/null
+++ b/src/debug/dwarf/testdata/debug_rnglists