Refactor software GSO code.

Software GSO implementation currently has a complicated code path with
implicit assumptions that all packets to WritePackets carry same Data
and it does this to avoid allocations on the path etc. But this makes it
hard to reuse the WritePackets API.

This change breaks all such assumptions by introducing a new Vectorised
View API ReadToVV which can be used to cleanly split a VV into multiple
independent VVs. Further this change also makes packet buffers linkable
to form an intrusive list. This allows us to get rid of the array of
packet buffers that are passed in the WritePackets API call and replace
it with a list of packet buffers.

While this code does introduce some more allocations in the benchmarks
it doesn't cause any degradation.

Updates #231

PiperOrigin-RevId: 304731742

28 files changed, 420 insertions, 230 deletions

diff --git a/pkg/ilist/list.go b/pkg/ilist/list.go
index 8f93e4d6d..0d07da3b1 100644
--- a/pkg/ilist/list.go
+++ b/pkg/ilist/list.go
@@ -86,12 +86,21 @@ func (l *List) Back() Element {
 	return l.tail
 }
 
+// Len returns the number of elements in the list.
+//
+// NOTE: This is an O(n) operation.
+func (l *List) Len() (count int) {
+	for e := l.Front(); e != nil; e = e.Next() {
+		count++
+	}
+	return count
+}
+
 // PushFront inserts the element e at the front of list l.
 func (l *List) PushFront(e Element) {
 	linker := ElementMapper{}.linkerFor(e)
 	linker.SetNext(l.head)
 	linker.SetPrev(nil)
-
 	if l.head != nil {
 		ElementMapper{}.linkerFor(l.head).SetPrev(e)
 	} else {
@@ -106,7 +115,6 @@ func (l *List) PushBack(e Element) {
 	linker := ElementMapper{}.linkerFor(e)
 	linker.SetNext(nil)
 	linker.SetPrev(l.tail)
-
 	if l.tail != nil {
 		ElementMapper{}.linkerFor(l.tail).SetNext(e)
 	} else {
@@ -127,7 +135,6 @@ func (l *List) PushBackList(m *List) {
 
 		l.tail = m.tail
 	}
-
 	m.head = nil
 	m.tail = nil
 }
diff --git a/pkg/sentry/kernel/kernel.go b/pkg/sentry/kernel/kernel.go
index 0a448b57c..2e6f42b92 100644
--- a/pkg/sentry/kernel/kernel.go
+++ b/pkg/sentry/kernel/kernel.go
@@ -564,15 +564,25 @@ func (ts *TaskSet) unregisterEpollWaiters() {
 
 	ts.mu.RLock()
 	defer ts.mu.RUnlock()
+
+	// Tasks that belong to the same process could potentially point to the
+	// same FDTable. So we retain a map of processed ones to avoid
+	// processing the same FDTable multiple times.
+	processed := make(map[*FDTable]struct{})
 	for t := range ts.Root.tids {
 		// We can skip locking Task.mu here since the kernel is paused.
-		if t.fdTable != nil {
-			t.fdTable.forEach(func(_ int32, file *fs.File, _ *vfs.FileDescription, _ FDFlags) {
-				if e, ok := file.FileOperations.(*epoll.EventPoll); ok {
-					e.UnregisterEpollWaiters()
-				}
-			})
+		if t.fdTable == nil {
+			continue
+		}
+		if _, ok := processed[t.fdTable]; ok {
+			continue
 		}
+		t.fdTable.forEach(func(_ int32, file *fs.File, _ *vfs.FileDescription, _ FDFlags) {
+			if e, ok := file.FileOperations.(*epoll.EventPoll); ok {
+				e.UnregisterEpollWaiters()
+			}
+		})
+		processed[t.fdTable] = struct{}{}
 	}
 }
 
diff --git a/pkg/tcpip/buffer/view.go b/pkg/tcpip/buffer/view.go
index 8d42cd066..8ec5d5d5c 100644
--- a/pkg/tcpip/buffer/view.go
+++ b/pkg/tcpip/buffer/view.go
@@ -17,6 +17,7 @@ package buffer
 
 import (
 	"bytes"
+	"io"
 )
 
 // View is a slice of a buffer, with convenience methods.
@@ -89,6 +90,47 @@ func (vv *VectorisedView) TrimFront(count int) {
 	}
 }
 
+// Read implements io.Reader.
+func (vv *VectorisedView) Read(v View) (copied int, err error) {
+	count := len(v)
+	for count > 0 && len(vv.views) > 0 {
+		if count < len(vv.views[0]) {
+			vv.size -= count
+			copy(v[copied:], vv.views[0][:count])
+			vv.views[0].TrimFront(count)
+			copied += count
+			return copied, nil
+		}
+		count -= len(vv.views[0])
+		copy(v[copied:], vv.views[0])
+		copied += len(vv.views[0])
+		vv.RemoveFirst()
+	}
+	if copied == 0 {
+		return 0, io.EOF
+	}
+	return copied, nil
+}
+
+// ReadToVV reads up to n bytes from vv to dstVV and removes them from vv. It
+// returns the number of bytes copied.
+func (vv *VectorisedView) ReadToVV(dstVV *VectorisedView, count int) (copied int) {
+	for count > 0 && len(vv.views) > 0 {
+		if count < len(vv.views[0]) {
+			vv.size -= count
+			dstVV.AppendView(vv.views[0][:count])
+			vv.views[0].TrimFront(count)
+			copied += count
+			return
+		}
+		count -= len(vv.views[0])
+		dstVV.AppendView(vv.views[0])
+		copied += len(vv.views[0])
+		vv.RemoveFirst()
+	}
+	return copied
+}
+
 // CapLength irreversibly reduces the length of the vectorised view.
 func (vv *VectorisedView) CapLength(length int) {
 	if length < 0 {
@@ -116,12 +158,12 @@ func (vv *VectorisedView) CapLength(length int) {
 // Clone returns a clone of this VectorisedView.
 // If the buffer argument is large enough to contain all the Views of this VectorisedView,
 // the method will avoid allocations and use the buffer to store the Views of the clone.
-func (vv VectorisedView) Clone(buffer []View) VectorisedView {
+func (vv *VectorisedView) Clone(buffer []View) VectorisedView {
 	return VectorisedView{views: append(buffer[:0], vv.views...), size: vv.size}
 }
 
 // First returns the first view of the vectorised view.
-func (vv VectorisedView) First() View {
+func (vv *VectorisedView) First() View {
 	if len(vv.views) == 0 {
 		return nil
 	}
@@ -134,11 +176,12 @@ func (vv *VectorisedView) RemoveFirst() {
 		return
 	}
 	vv.size -= len(vv.views[0])
+	vv.views[0] = nil
 	vv.views = vv.views[1:]
 }
 
 // Size returns the size in bytes of the entire content stored in the vectorised view.
-func (vv VectorisedView) Size() int {
+func (vv *VectorisedView) Size() int {
 	return vv.size
 }
 
@@ -146,7 +189,7 @@ func (vv VectorisedView) Size() int {
 //
 // If the vectorised view contains a single view, that view will be returned
 // directly.
-func (vv VectorisedView) ToView() View {
+func (vv *VectorisedView) ToView() View {
 	if len(vv.views) == 1 {
 		return vv.views[0]
 	}
@@ -158,7 +201,7 @@ func (vv VectorisedView) ToView() View {
 }
 
 // Views returns the slice containing the all views.
-func (vv VectorisedView) Views() []View {
+func (vv *VectorisedView) Views() []View {
 	return vv.views
 }
 
diff --git a/pkg/tcpip/buffer/view_test.go b/pkg/tcpip/buffer/view_test.go
index ebc3a17b7..106e1994c 100644
--- a/pkg/tcpip/buffer/view_test.go
+++ b/pkg/tcpip/buffer/view_test.go
@@ -233,3 +233,140 @@ func TestToClone(t *testing.T) {
 		})
 	}
 }
+
+func TestVVReadToVV(t *testing.T) {
+	testCases := []struct {
+		comment     string
+		vv          VectorisedView
+		bytesToRead int
+		wantBytes   string
+		leftVV      VectorisedView
+	}{
+		{
+			comment:     "large VV, short read",
+			vv:          vv(30, "012345678901234567890123456789"),
+			bytesToRead: 10,
+			wantBytes:   "0123456789",
+			leftVV:      vv(20, "01234567890123456789"),
+		},
+		{
+			comment:     "largeVV, multiple views, short read",
+			vv:          vv(13, "123", "345", "567", "8910"),
+			bytesToRead: 6,
+			wantBytes:   "123345",
+			leftVV:      vv(7, "567", "8910"),
+		},
+		{
+			comment:     "smallVV (multiple views), large read",
+			vv:          vv(3, "1", "2", "3"),
+			bytesToRead: 10,
+			wantBytes:   "123",
+			leftVV:      vv(0, ""),
+		},
+		{
+			comment:     "smallVV (single view), large read",
+			vv:          vv(1, "1"),
+			bytesToRead: 10,
+			wantBytes:   "1",
+			leftVV:      vv(0, ""),
+		},
+		{
+			comment:     "emptyVV, large read",
+			vv:          vv(0, ""),
+			bytesToRead: 10,
+			wantBytes:   "",
+			leftVV:      vv(0, ""),
+		},
+	}
+
+	for _, tc := range testCases {
+		t.Run(tc.comment, func(t *testing.T) {
+			var readTo VectorisedView
+			inSize := tc.vv.Size()
+			copied := tc.vv.ReadToVV(&readTo, tc.bytesToRead)
+			if got, want := copied, len(tc.wantBytes); got != want {
+				t.Errorf("incorrect number of bytes copied returned in ReadToVV got: %d, want: %d, tc: %+v", got, want, tc)
+			}
+			if got, want := string(readTo.ToView()), tc.wantBytes; got != want {
+				t.Errorf("unexpected content in readTo got: %s, want: %s", got, want)
+			}
+			if got, want := tc.vv.Size(), inSize-copied; got != want {
+				t.Errorf("test VV has incorrect size after reading got: %d, want: %d, tc.vv: %+v", got, want, tc.vv)
+			}
+			if got, want := string(tc.vv.ToView()), string(tc.leftVV.ToView()); got != want {
+				t.Errorf("unexpected data left in vv after read got: %+v, want: %+v", got, want)
+			}
+		})
+	}
+}
+
+func TestVVRead(t *testing.T) {
+	testCases := []struct {
+		comment     string
+		vv          VectorisedView
+		bytesToRead int
+		readBytes   string
+		leftBytes   string
+		wantError   bool
+	}{
+		{
+			comment:     "large VV, short read",
+			vv:          vv(30, "012345678901234567890123456789"),
+			bytesToRead: 10,
+			readBytes:   "0123456789",
+			leftBytes:   "01234567890123456789",
+		},
+		{
+			comment:     "largeVV, multiple buffers, short read",
+			vv:          vv(13, "123", "345", "567", "8910"),
+			bytesToRead: 6,
+			readBytes:   "123345",
+			leftBytes:   "5678910",
+		},
+		{
+			comment:     "smallVV, large read",
+			vv:          vv(3, "1", "2", "3"),
+			bytesToRead: 10,
+			readBytes:   "123",
+			leftBytes:   "",
+		},
+		{
+			comment:     "smallVV, large read",
+			vv:          vv(1, "1"),
+			bytesToRead: 10,
+			readBytes:   "1",
+			leftBytes:   "",
+		},
+		{
+			comment:     "emptyVV, large read",
+			vv:          vv(0, ""),
+			bytesToRead: 10,
+			readBytes:   "",
+			wantError:   true,
+		},
+	}
+
+	for _, tc := range testCases {
+		t.Run(tc.comment, func(t *testing.T) {
+			readTo := NewView(tc.bytesToRead)
+			inSize := tc.vv.Size()
+			copied, err := tc.vv.Read(readTo)
+			if !tc.wantError && err != nil {
+				t.Fatalf("unexpected error in tc.vv.Read(..) = %s", err)
+			}
+			readTo = readTo[:copied]
+			if got, want := copied, len(tc.readBytes); got != want {
+				t.Errorf("incorrect number of bytes copied returned in ReadToVV got: %d, want: %d, tc.vv: %+v", got, want, tc.vv)
+			}
+			if got, want := string(readTo), tc.readBytes; got != want {
+				t.Errorf("unexpected data in readTo got: %s, want: %s", got, want)
+			}
+			if got, want := tc.vv.Size(), inSize-copied; got != want {
+				t.Errorf("test VV has incorrect size after reading got: %d, want: %d, tc.vv: %+v", got, want, tc.vv)
+			}
+			if got, want := string(tc.vv.ToView()), tc.leftBytes; got != want {
+				t.Errorf("vv has incorrect data after Read got: %s, want: %s", got, want)
+			}
+		})
+	}
+}
diff --git a/pkg/tcpip/link/channel/channel.go b/pkg/tcpip/link/channel/channel.go
index a8d6653ce..b4a0ae53d 100644
--- a/pkg/tcpip/link/channel/channel.go
+++ b/pkg/tcpip/link/channel/channel.go
@@ -28,7 +28,7 @@ import (
 
 // PacketInfo holds all the information about an outbound packet.
 type PacketInfo struct {
-	Pkt   stack.PacketBuffer
+	Pkt   *stack.PacketBuffer
 	Proto tcpip.NetworkProtocolNumber
 	GSO   *stack.GSO
 	Route stack.Route
@@ -257,7 +257,7 @@ func (e *Endpoint) WritePacket(r *stack.Route, gso *stack.GSO, protocol tcpip.Ne
 	route := r.Clone()
 	route.Release()
 	p := PacketInfo{
-		Pkt:   pkt,
+		Pkt:   &pkt,
 		Proto: protocol,
 		GSO:   gso,
 		Route: route,
@@ -269,21 +269,15 @@ func (e *Endpoint) WritePacket(r *stack.Route, gso *stack.GSO, protocol tcpip.Ne
 }
 
 // WritePackets stores outbound packets into the channel.
-func (e *Endpoint) WritePackets(r *stack.Route, gso *stack.GSO, pkts []stack.PacketBuffer, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
+func (e *Endpoint) WritePackets(r *stack.Route, gso *stack.GSO, pkts stack.PacketBufferList, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
 	// Clone r then release its resource so we only get the relevant fields from
 	// stack.Route without holding a reference to a NIC's endpoint.
 	route := r.Clone()
 	route.Release()
-	payloadView := pkts[0].Data.ToView()
 	n := 0
-	for _, pkt := range pkts {
-		off := pkt.DataOffset
-		size := pkt.DataSize
+	for pkt := pkts.Front(); pkt != nil; pkt = pkt.Next() {
 		p := PacketInfo{
-			Pkt: stack.PacketBuffer{
-				Header: pkt.Header,
-				Data:   buffer.NewViewFromBytes(payloadView[off : off+size]).ToVectorisedView(),
-			},
+			Pkt:   pkt,
 			Proto: protocol,
 			GSO:   gso,
 			Route: route,
@@ -301,7 +295,7 @@ func (e *Endpoint) WritePackets(r *stack.Route, gso *stack.GSO, pkts []stack.Pac
 // WriteRawPacket implements stack.LinkEndpoint.WriteRawPacket.
 func (e *Endpoint) WriteRawPacket(vv buffer.VectorisedView) *tcpip.Error {
 	p := PacketInfo{
-		Pkt:   stack.PacketBuffer{Data: vv},
+		Pkt:   &stack.PacketBuffer{Data: vv},
 		Proto: 0,
 		GSO:   nil,
 	}
diff --git a/pkg/tcpip/link/fdbased/endpoint.go b/pkg/tcpip/link/fdbased/endpoint.go
index 3b3b6909b..7198742b7 100644
--- a/pkg/tcpip/link/fdbased/endpoint.go
+++ b/pkg/tcpip/link/fdbased/endpoint.go
@@ -441,118 +441,106 @@ func (e *endpoint) WritePacket(r *stack.Route, gso *stack.GSO, protocol tcpip.Ne
 
 // WritePackets writes outbound packets to the file descriptor. If it is not
 // currently writable, the packet is dropped.
-func (e *endpoint) WritePackets(r *stack.Route, gso *stack.GSO, pkts []stack.PacketBuffer, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
-	var ethHdrBuf []byte
-	// hdr + data
-	iovLen := 2
-	if e.hdrSize > 0 {
-		// Add ethernet header if needed.
-		ethHdrBuf = make([]byte, header.EthernetMinimumSize)
-		eth := header.Ethernet(ethHdrBuf)
-		ethHdr := &header.EthernetFields{
-			DstAddr: r.RemoteLinkAddress,
-			Type:    protocol,
-		}
-
-		// Preserve the src address if it's set in the route.
-		if r.LocalLinkAddress != "" {
-			ethHdr.SrcAddr = r.LocalLinkAddress
-		} else {
-			ethHdr.SrcAddr = e.addr
-		}
-		eth.Encode(ethHdr)
-		iovLen++
-	}
+//
+// NOTE: This API uses sendmmsg to batch packets. As a result the underlying FD
+// picked to write the packet out has to be the same for all packets in the
+// list. In other words all packets in the batch should belong to the same
+// flow.
+func (e *endpoint) WritePackets(r *stack.Route, gso *stack.GSO, pkts stack.PacketBufferList, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
+	n := pkts.Len()
 
-	n := len(pkts)
-
-	views := pkts[0].Data.Views()
-	/*
-	 * Each boundary in views can add one more iovec.
-	 *
-	 * payload |      |          |         |
-	 *         -----------------------------
-	 * packets |    |    |    |    |    |  |
-	 *         -----------------------------
-	 * iovecs  |    | |  |    |  | |    |  |
-	 */
-	iovec := make([]syscall.Iovec, n*iovLen+len(views)-1)
 	mmsgHdrs := make([]rawfile.MMsgHdr, n)
+	i := 0
+	for pkt := pkts.Front(); pkt != nil; pkt = pkt.Next() {
+		var ethHdrBuf []byte
+		iovLen := 0
+		if e.hdrSize > 0 {
+			// Add ethernet header if needed.
+			ethHdrBuf = make([]byte, header.EthernetMinimumSize)
+			eth := header.Ethernet(ethHdrBuf)
+			ethHdr := &header.EthernetFields{
+				DstAddr: r.RemoteLinkAddress,
+				Type:    protocol,
+			}
 
-	iovecIdx := 0
-	viewIdx := 0
-	viewOff := 0
-	off := 0
-	nextOff := 0
-	for i := range pkts {
-		// TODO(b/134618279): Different packets may have different data
-		// in the future. We should handle this.
-		if !viewsEqual(pkts[i].Data.Views(), views) {
-			panic("All packets in pkts should have the same Data.")
+			// Preserve the src address if it's set in the route.
+			if r.LocalLinkAddress != "" {
+				ethHdr.SrcAddr = r.LocalLinkAddress
+			} else {
+				ethHdr.SrcAddr = e.addr
+			}
+			eth.Encode(ethHdr)
+			iovLen++
 		}
 
-		prevIovecIdx := iovecIdx
-		mmsgHdr := &mmsgHdrs[i]
-		mmsgHdr.Msg.Iov = &iovec[iovecIdx]
-		packetSize := pkts[i].DataSize
-		hdr := &pkts[i].Header
-
-		off = pkts[i].DataOffset
-		if off != nextOff {
-			// We stop in a different point last time.
-			size := packetSize
-			viewIdx = 0
-			viewOff = 0
-			for size > 0 {
-				if size >= len(views[viewIdx]) {
-					viewIdx++
-					viewOff = 0
-					size -= len(views[viewIdx])
-				} else {
-					viewOff = size
-					size = 0
+		var vnetHdrBuf []byte
+		vnetHdr := virtioNetHdr{}
+		if e.Capabilities()&stack.CapabilityHardwareGSO != 0 {
+			if gso != nil {
+				vnetHdr.hdrLen = uint16(pkt.Header.UsedLength())
+				if gso.NeedsCsum {
+					vnetHdr.flags = _VIRTIO_NET_HDR_F_NEEDS_CSUM
+					vnetHdr.csumStart = header.EthernetMinimumSize + gso.L3HdrLen
+					vnetHdr.csumOffset = gso.CsumOffset
+				}
+				if gso.Type != stack.GSONone && uint16(pkt.Data.Size()) > gso.MSS {
+					switch gso.Type {
+					case stack.GSOTCPv4:
+						vnetHdr.gsoType = _VIRTIO_NET_HDR_GSO_TCPV4
+					case stack.GSOTCPv6:
+						vnetHdr.gsoType = _VIRTIO_NET_HDR_GSO_TCPV6
+					default:
+						panic(fmt.Sprintf("Unknown gso type: %v", gso.Type))
+					}
+					vnetHdr.gsoSize = gso.MSS
 				}
 			}
+			vnetHdrBuf = vnetHdrToByteSlice(&vnetHdr)
+			iovLen++
 		}
-		nextOff = off + packetSize
 
+		iovecs := make([]syscall.Iovec, iovLen+1+len(pkt.Data.Views()))
+		mmsgHdr := &mmsgHdrs[i]
+		mmsgHdr.Msg.Iov = &iovecs[0]
+		iovecIdx := 0
+		if vnetHdrBuf != nil {
+			v := &iovecs[iovecIdx]
+			v.Base = &vnetHdrBuf[0]
+			v.Len = uint64(len(vnetHdrBuf))
+			iovecIdx++
+		}
 		if ethHdrBuf != nil {
-			v := &iovec[iovecIdx]
+			v := &iovecs[iovecIdx]
 			v.Base = &ethHdrBuf[0]
 			v.Len = uint64(len(ethHdrBuf))
 			iovecIdx++
 		}
-
-		v := &iovec[iovecIdx]
+		pktSize := uint64(0)
+		// Encode L3 Header
+		v := &iovecs[iovecIdx]
+		hdr := &pkt.Header
 		hdrView := hdr.View()
 		v.Base = &hdrView[0]
 		v.Len = uint64(len(hdrView))
+		pktSize += v.Len
 		iovecIdx++
 
-		for packetSize > 0 {
-			vec := &iovec[iovecIdx]
+		// Now encode the Transport Payload.
+		pktViews := pkt.Data.Views()
+		for i := range pktViews {
+			vec := &iovecs[iovecIdx]
 			iovecIdx++
-
-			v := views[viewIdx]
-			vec.Base = &v[viewOff]
-			s := len(v) - viewOff
-			if s <= packetSize {
-				viewIdx++
-				viewOff = 0
-			} else {
-				s = packetSize
-				viewOff += s
-			}
-			vec.Len = uint64(s)
-			packetSize -= s
+			vec.Base = &pktViews[i][0]
+			vec.Len = uint64(len(pktViews[i]))
+			pktSize += vec.Len
 		}
-
-		mmsgHdr.Msg.Iovlen = uint64(iovecIdx - prevIovecIdx)
+		mmsgHdr.Msg.Iovlen = uint64(iovecIdx)
+		i++
 	}
 
 	packets := 0
 	for packets < n {
-		fd := e.fds[pkts[packets].Hash%uint32(len(e.fds))]
+		fd := e.fds[pkts.Front().Hash%uint32(len(e.fds))]
 		sent, err := rawfile.NonBlockingSendMMsg(fd, mmsgHdrs)
 		if err != nil {
 			return packets, err
diff --git a/pkg/tcpip/link/loopback/loopback.go b/pkg/tcpip/link/loopback/loopback.go
index 4039753b7..1e2255bfa 100644
--- a/pkg/tcpip/link/loopback/loopback.go
+++ b/pkg/tcpip/link/loopback/loopback.go
@@ -92,7 +92,7 @@ func (e *endpoint) WritePacket(_ *stack.Route, _ *stack.GSO, protocol tcpip.Netw
 }
 
 // WritePackets implements stack.LinkEndpoint.WritePackets.
-func (e *endpoint) WritePackets(*stack.Route, *stack.GSO, []stack.PacketBuffer, tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
+func (e *endpoint) WritePackets(*stack.Route, *stack.GSO, stack.PacketBufferList, tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
 	panic("not implemented")
 }
 
diff --git a/pkg/tcpip/link/muxed/injectable.go b/pkg/tcpip/link/muxed/injectable.go
index f5973066d..a5478ce17 100644
--- a/pkg/tcpip/link/muxed/injectable.go
+++ b/pkg/tcpip/link/muxed/injectable.go
@@ -87,7 +87,7 @@ func (m *InjectableEndpoint) InjectInbound(protocol tcpip.NetworkProtocolNumber,
 // WritePackets writes outbound packets to the appropriate
 // LinkInjectableEndpoint based on the RemoteAddress. HandleLocal only works if
 // r.RemoteAddress has a route registered in this endpoint.
-func (m *InjectableEndpoint) WritePackets(r *stack.Route, gso *stack.GSO, pkts []stack.PacketBuffer, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
+func (m *InjectableEndpoint) WritePackets(r *stack.Route, gso *stack.GSO, pkts stack.PacketBufferList, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
 	endpoint, ok := m.routes[r.RemoteAddress]
 	if !ok {
 		return 0, tcpip.ErrNoRoute
diff --git a/pkg/tcpip/link/sharedmem/sharedmem.go b/pkg/tcpip/link/sharedmem/sharedmem.go
index 6461d0108..0796d717e 100644
--- a/pkg/tcpip/link/sharedmem/sharedmem.go
+++ b/pkg/tcpip/link/sharedmem/sharedmem.go
@@ -214,7 +214,7 @@ func (e *endpoint) WritePacket(r *stack.Route, _ *stack.GSO, protocol tcpip.Netw
 }
 
 // WritePackets implements stack.LinkEndpoint.WritePackets.
-func (e *endpoint) WritePackets(r *stack.Route, _ *stack.GSO, pkts []stack.PacketBuffer, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
+func (e *endpoint) WritePackets(r *stack.Route, _ *stack.GSO, pkts stack.PacketBufferList, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
 	panic("not implemented")
 }
 
diff --git a/pkg/tcpip/link/sniffer/sniffer.go b/pkg/tcpip/link/sniffer/sniffer.go
index 0a6b8945c..062388f4d 100644
--- a/pkg/tcpip/link/sniffer/sniffer.go
+++ b/pkg/tcpip/link/sniffer/sniffer.go
@@ -200,7 +200,7 @@ func (e *endpoint) GSOMaxSize() uint32 {
 	return 0
 }
 
-func (e *endpoint) dumpPacket(gso *stack.GSO, protocol tcpip.NetworkProtocolNumber, pkt stack.PacketBuffer) {
+func (e *endpoint) dumpPacket(gso *stack.GSO, protocol tcpip.NetworkProtocolNumber, pkt *stack.PacketBuffer) {
 	if atomic.LoadUint32(&LogPackets) == 1 && e.file == nil {
 		logPacket("send", protocol, pkt.Header.View(), gso)
 	}
@@ -233,20 +233,16 @@ func (e *endpoint) dumpPacket(gso *stack.GSO, protocol tcpip.NetworkProtocolNumb
 // higher-level protocols to write packets; it just logs the packet and
 // forwards the request to the lower endpoint.
 func (e *endpoint) WritePacket(r *stack.Route, gso *stack.GSO, protocol tcpip.NetworkProtocolNumber, pkt stack.PacketBuffer) *tcpip.Error {
-	e.dumpPacket(gso, protocol, pkt)
+	e.dumpPacket(gso, protocol, &pkt)
 	return e.lower.WritePacket(r, gso, protocol, pkt)
 }
 
 // WritePackets implements the stack.LinkEndpoint interface. It is called by
 // higher-level protocols to write packets; it just logs the packet and
 // forwards the request to the lower endpoint.
-func (e *endpoint) WritePackets(r *stack.Route, gso *stack.GSO, pkts []stack.PacketBuffer, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
-	view := pkts[0].Data.ToView()
-	for _, pkt := range pkts {
-		e.dumpPacket(gso, protocol, stack.PacketBuffer{
-			Header: pkt.Header,
-			Data:   view[pkt.DataOffset:][:pkt.DataSize].ToVectorisedView(),
-		})
+func (e *endpoint) WritePackets(r *stack.Route, gso *stack.GSO, pkts stack.PacketBufferList, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
+	for pkt := pkts.Front(); pkt != nil; pkt = pkt.Next() {
+		e.dumpPacket(gso, protocol, pkt)
 	}
 	return e.lower.WritePackets(r, gso, pkts, protocol)
 }
diff --git a/pkg/tcpip/link/waitable/waitable.go b/pkg/tcpip/link/waitable/waitable.go
index 52fe397bf..2b3741276 100644
--- a/pkg/tcpip/link/waitable/waitable.go
+++ b/pkg/tcpip/link/waitable/waitable.go
@@ -112,9 +112,9 @@ func (e *Endpoint) WritePacket(r *stack.Route, gso *stack.GSO, protocol tcpip.Ne
 // WritePackets implements stack.LinkEndpoint.WritePackets. It is called by
 // higher-level protocols to write packets. It only forwards packets to the
 // lower endpoint if Wait or WaitWrite haven't been called.
-func (e *Endpoint) WritePackets(r *stack.Route, gso *stack.GSO, pkts []stack.PacketBuffer, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
+func (e *Endpoint) WritePackets(r *stack.Route, gso *stack.GSO, pkts stack.PacketBufferList, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
 	if !e.writeGate.Enter() {
-		return len(pkts), nil
+		return pkts.Len(), nil
 	}
 
 	n, err := e.lower.WritePackets(r, gso, pkts, protocol)
diff --git a/pkg/tcpip/link/waitable/waitable_test.go b/pkg/tcpip/link/waitable/waitable_test.go
index 88224e494..54eb5322b 100644
--- a/pkg/tcpip/link/waitable/waitable_test.go
+++ b/pkg/tcpip/link/waitable/waitable_test.go
@@ -71,9 +71,9 @@ func (e *countedEndpoint) WritePacket(r *stack.Route, _ *stack.GSO, protocol tcp
 }
 
 // WritePackets implements stack.LinkEndpoint.WritePackets.
-func (e *countedEndpoint) WritePackets(r *stack.Route, _ *stack.GSO, pkts []stack.PacketBuffer, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
-	e.writeCount += len(pkts)
-	return len(pkts), nil
+func (e *countedEndpoint) WritePackets(r *stack.Route, _ *stack.GSO, pkts stack.PacketBufferList, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
+	e.writeCount += pkts.Len()
+	return pkts.Len(), nil
 }
 
 func (e *countedEndpoint) WriteRawPacket(buffer.VectorisedView) *tcpip.Error {
diff --git a/pkg/tcpip/network/arp/arp.go b/pkg/tcpip/network/arp/arp.go
index 255098372..7acbfa0a8 100644
--- a/pkg/tcpip/network/arp/arp.go
+++ b/pkg/tcpip/network/arp/arp.go
@@ -84,7 +84,7 @@ func (e *endpoint) WritePacket(*stack.Route, *stack.GSO, stack.NetworkHeaderPara
 }
 
 // WritePackets implements stack.NetworkEndpoint.WritePackets.
-func (e *endpoint) WritePackets(*stack.Route, *stack.GSO, []stack.PacketBuffer, stack.NetworkHeaderParams) (int, *tcpip.Error) {
+func (e *endpoint) WritePackets(*stack.Route, *stack.GSO, stack.PacketBufferList, stack.NetworkHeaderParams) (int, *tcpip.Error) {
 	return 0, tcpip.ErrNotSupported
 }
 
diff --git a/pkg/tcpip/network/ip_test.go b/pkg/tcpip/network/ip_test.go
index 4950d69fc..4c20301c6 100644
--- a/pkg/tcpip/network/ip_test.go
+++ b/pkg/tcpip/network/ip_test.go
@@ -172,7 +172,7 @@ func (t *testObject) WritePacket(_ *stack.Route, _ *stack.GSO, protocol tcpip.Ne
 }
 
 // WritePackets implements stack.LinkEndpoint.WritePackets.
-func (t *testObject) WritePackets(_ *stack.Route, _ *stack.GSO, pkt []stack.PacketBuffer, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
+func (t *testObject) WritePackets(_ *stack.Route, _ *stack.GSO, pkt stack.PacketBufferList, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
 	panic("not implemented")
 }
 
diff --git a/pkg/tcpip/network/ipv4/ipv4.go b/pkg/tcpip/network/ipv4/ipv4.go
index a7d9a8b25..104aafbed 100644
--- a/pkg/tcpip/network/ipv4/ipv4.go
+++ b/pkg/tcpip/network/ipv4/ipv4.go
@@ -280,28 +280,47 @@ func (e *endpoint) WritePacket(r *stack.Route, gso *stack.GSO, params stack.Netw
 }
 
 // WritePackets implements stack.NetworkEndpoint.WritePackets.
-func (e *endpoint) WritePackets(r *stack.Route, gso *stack.GSO, pkts []stack.PacketBuffer, params stack.NetworkHeaderParams) (int, *tcpip.Error) {
+func (e *endpoint) WritePackets(r *stack.Route, gso *stack.GSO, pkts stack.PacketBufferList, params stack.NetworkHeaderParams) (int, *tcpip.Error) {
 	if r.Loop&stack.PacketLoop != 0 {
 		panic("multiple packets in local loop")
 	}
 	if r.Loop&stack.PacketOut == 0 {
-		return len(pkts), nil
+		return pkts.Len(), nil
+	}
+
+	for pkt := pkts.Front(); pkt != nil; {
+		ip := e.addIPHeader(r, &pkt.Header, pkt.Data.Size(), params)
+		pkt.NetworkHeader = buffer.View(ip)
+		pkt = pkt.Next()
 	}
 
 	// iptables filtering. All packets that reach here are locally
 	// generated.
 	ipt := e.stack.IPTables()
-	for i := range pkts {
-		if ok := ipt.Check(stack.Output, pkts[i]); !ok {
-			// iptables is telling us to drop the packet.
+	dropped := ipt.CheckPackets(stack.Output, pkts)
+	if len(dropped) == 0 {
+		// Fast path: If no packets are to be dropped then we can just invoke the
+		// faster WritePackets API directly.
+		n, err := e.linkEP.WritePackets(r, gso, pkts, ProtocolNumber)
+		r.Stats().IP.PacketsSent.IncrementBy(uint64(n))
+		return n, err
+	}
+
+	// Slow Path as we are dropping some packets in the batch degrade to
+	// emitting one packet at a time.
+	n := 0
+	for pkt := pkts.Front(); pkt != nil; pkt = pkt.Next() {
+		if _, ok := dropped[pkt]; ok {
 			continue
 		}
-		ip := e.addIPHeader(r, &pkts[i].Header, pkts[i].DataSize, params)
-		pkts[i].NetworkHeader = buffer.View(ip)
+		if err := e.linkEP.WritePacket(r, gso, ProtocolNumber, *pkt); err != nil {
+			r.Stats().IP.PacketsSent.IncrementBy(uint64(n))
+			return n, err
+		}
+		n++
 	}
-	n, err := e.linkEP.WritePackets(r, gso, pkts, ProtocolNumber)
 	r.Stats().IP.PacketsSent.IncrementBy(uint64(n))
-	return n, err
+	return n, nil
 }
 
 // WriteHeaderIncludedPacket writes a packet already containing a network
diff --git a/pkg/tcpip/network/ipv6/icmp.go b/pkg/tcpip/network/ipv6/icmp.go
index 6d2d2c034..f91180aa3 100644
--- a/pkg/tcpip/network/ipv6/icmp.go
+++ b/pkg/tcpip/network/ipv6/icmp.go
@@ -79,7 +79,7 @@ func (e *endpoint) handleICMP(r *stack.Route, netHeader buffer.View, pkt stack.P
 	// Only the first view in vv is accounted for by h. To account for the
 	// rest of vv, a shallow copy is made and the first view is removed.
 	// This copy is used as extra payload during the checksum calculation.
-	payload := pkt.Data
+	payload := pkt.Data.Clone(nil)
 	payload.RemoveFirst()
 	if got, want := h.Checksum(), header.ICMPv6Checksum(h, iph.SourceAddress(), iph.DestinationAddress(), payload); got != want {
 		received.Invalid.Increment()
diff --git a/pkg/tcpip/network/ipv6/ipv6.go b/pkg/tcpip/network/ipv6/ipv6.go
index b462b8604..a815b4d9b 100644
--- a/pkg/tcpip/network/ipv6/ipv6.go
+++ b/pkg/tcpip/network/ipv6/ipv6.go
@@ -143,19 +143,17 @@ func (e *endpoint) WritePacket(r *stack.Route, gso *stack.GSO, params stack.Netw
 }
 
 // WritePackets implements stack.LinkEndpoint.WritePackets.
-func (e *endpoint) WritePackets(r *stack.Route, gso *stack.GSO, pkts []stack.PacketBuffer, params stack.NetworkHeaderParams) (int, *tcpip.Error) {
+func (e *endpoint) WritePackets(r *stack.Route, gso *stack.GSO, pkts stack.PacketBufferList, params stack.NetworkHeaderParams) (int, *tcpip.Error) {
 	if r.Loop&stack.PacketLoop != 0 {
 		panic("not implemented")
 	}
 	if r.Loop&stack.PacketOut == 0 {
-		return len(pkts), nil
+		return pkts.Len(), nil
 	}
 
-	for i := range pkts {
-		hdr := &pkts[i].Header
-		size := pkts[i].DataSize
-		ip := e.addIPHeader(r, hdr, size, params)
-		pkts[i].NetworkHeader = buffer.View(ip)
+	for pb := pkts.Front(); pb != nil; pb = pb.Next() {
+		ip := e.addIPHeader(r, &pb.Header, pb.Data.Size(), params)
+		pb.NetworkHeader = buffer.View(ip)
 	}
 
 	n, err := e.linkEP.WritePackets(r, gso, pkts, ProtocolNumber)
diff --git a/pkg/tcpip/stack/BUILD b/pkg/tcpip/stack/BUILD
index 8d80e9cee..5e963a4af 100644
--- a/pkg/tcpip/stack/BUILD
+++ b/pkg/tcpip/stack/BUILD
@@ -15,6 +15,18 @@ go_template_instance(
     },
 )
 
+go_template_instance(
+    name = "packet_buffer_list",
+    out = "packet_buffer_list.go",
+    package = "stack",
+    prefix = "PacketBuffer",
+    template = "//pkg/ilist:generic_list",
+    types = {
+        "Element": "*PacketBuffer",
+        "Linker": "*PacketBuffer",
+    },
+)
+
 go_library(
     name = "stack",
     srcs = [
@@ -29,7 +41,7 @@ go_library(
         "ndp.go",
         "nic.go",
         "packet_buffer.go",
-        "packet_buffer_state.go",
+        "packet_buffer_list.go",
         "rand.go",
         "registration.go",
         "route.go",
diff --git a/pkg/tcpip/stack/forwarder_test.go b/pkg/tcpip/stack/forwarder_test.go
index c45c43d21..e9c652042 100644
--- a/pkg/tcpip/stack/forwarder_test.go
+++ b/pkg/tcpip/stack/forwarder_test.go
@@ -101,7 +101,7 @@ func (f *fwdTestNetworkEndpoint) WritePacket(r *Route, gso *GSO, params NetworkH
 }
 
 // WritePackets implements LinkEndpoint.WritePackets.
-func (f *fwdTestNetworkEndpoint) WritePackets(r *Route, gso *GSO, pkts []PacketBuffer, params NetworkHeaderParams) (int, *tcpip.Error) {
+func (f *fwdTestNetworkEndpoint) WritePackets(r *Route, gso *GSO, pkts PacketBufferList, params NetworkHeaderParams) (int, *tcpip.Error) {
 	panic("not implemented")
 }
 
@@ -260,10 +260,10 @@ func (e fwdTestLinkEndpoint) WritePacket(r *Route, gso *GSO, protocol tcpip.Netw
 }
 
 // WritePackets stores outbound packets into the channel.
-func (e *fwdTestLinkEndpoint) WritePackets(r *Route, gso *GSO, pkts []PacketBuffer, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
+func (e *fwdTestLinkEndpoint) WritePackets(r *Route, gso *GSO, pkts PacketBufferList, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error) {
 	n := 0
-	for _, pkt := range pkts {
-		e.WritePacket(r, gso, protocol, pkt)
+	for pkt := pkts.Front(); pkt != nil; pkt = pkt.Next() {
+		e.WritePacket(r, gso, protocol, *pkt)
 		n++
 	}
 
diff --git a/pkg/tcpip/stack/iptables.go b/pkg/tcpip/stack/iptables.go
index 37907ae24..6c0a4b24d 100644
--- a/pkg/tcpip/stack/iptables.go
+++ b/pkg/tcpip/stack/iptables.go
@@ -209,6 +209,23 @@ func (it *IPTables) Check(hook Hook, pkt PacketBuffer) bool {
 	return true
 }
 
+// CheckPackets runs pkts through the rules for hook and returns a map of packets that
+// should not go forward.
+//
+// NOTE: unlike the Check API the returned map contains packets that should be
+// dropped.
+func (it *IPTables) CheckPackets(hook Hook, pkts PacketBufferList) (drop map[*PacketBuffer]struct{}) {
+	for pkt := pkts.Front(); pkt != nil; pkt = pkt.Next() {
+		if ok := it.Check(hook, *pkt); !ok {
+			if drop == nil {
+				drop = make(map[*PacketBuffer]struct{})
+			}
+			drop[pkt] = struct{}{}
+		}
+	}
+	return drop
+}
+
 // Precondition: pkt.NetworkHeader is set.
 func (it *IPTables) checkChain(hook Hook, pkt PacketBuffer, table Table, ruleIdx int) chainVerdict {
 	// Start from ruleIdx and walk the list of rules until a rule gives us
diff --git a/pkg/tcpip/stack/ndp_test.go b/pkg/tcpip/stack/ndp_test.go
index 598468bdd..27dc8baf9 100644
--- a/pkg/tcpip/stack/ndp_test.go
+++ b/pkg/tcpip/stack/ndp_test.go
@@ -468,7 +468,7 @@ func TestDADResolve(t *testing.T) {
 				// As per RFC 4861 section 4.3, a possible option is the Source Link
 				// Layer option, but this option MUST NOT be included when the source
 				// address of the packet is the unspecified address.
-				checker.IPv6(t, p.Pkt.Header.View().ToVectorisedView().First(),
+				checker.IPv6(t, p.Pkt.Header.View(),
 					checker.SrcAddr(header.IPv6Any),
 					checker.DstAddr(snmc),
 					checker.TTL(header.NDPHopLimit),
diff --git a/pkg/tcpip/stack/packet_buffer.go b/pkg/tcpip/stack/packet_buffer.go
index 9367de180..dc125f25e 100644
--- a/pkg/tcpip/stack/packet_buffer.go
+++ b/pkg/tcpip/stack/packet_buffer.go
@@ -23,9 +23,11 @@ import (
 // As a PacketBuffer traverses up the stack, it may be necessary to pass it to
 // multiple endpoints. Clone() should be called in such cases so that
 // modifications to the Data field do not affect other copies.
-//
-// +stateify savable
 type PacketBuffer struct {
+	// PacketBufferEntry is used to build an intrusive list of
+	// PacketBuffers.
+	PacketBufferEntry
+
 	// Data holds the payload of the packet. For inbound packets, it also
 	// holds the headers, which are consumed as the packet moves up the
 	// stack. Headers are guaranteed not to be split across views.
@@ -34,14 +36,6 @@ type PacketBuffer struct {
 	// or otherwise modified.
 	Data buffer.VectorisedView
 
-	// DataOffset is used for GSO output. It is the offset into the Data
-	// field where the payload of this packet starts.
-	DataOffset int
-
-	// DataSize is used for GSO output. It is the size of this packet's
-	// payload.
-	DataSize int
-
 	// Header holds the headers of outbound packets. As a packet is passed
 	// down the stack, each layer adds to Header.
 	Header buffer.Prependable
diff --git a/pkg/tcpip/stack/packet_buffer_state.go b/pkg/tcpip/stack/packet_buffer_state.go
deleted file mode 100644
index 0c6b7924c..000000000
--- a/pkg/tcpip/stack/packet_buffer_state.go
+++ /dev/null
@@ -1,27 +0,0 @@
-// Copyright 2019 The gVisor Authors.
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-//
-//     http://www.apache.org/licenses/LICENSE-2.0
-//
-// Unless required by applicable law or agreed to in writing, software
-// distributed under the License is distributed on an "AS IS" BASIS,
-// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-// See the License for the specific language governing permissions and
-// limitations under the License.
-
-package stack
-
-import "gvisor.dev/gvisor/pkg/tcpip/buffer"
-
-// beforeSave is invoked by stateify.
-func (pk *PacketBuffer) beforeSave() {
-	// Non-Data fields may be slices of the Data field. This causes
-	// problems for SR, so during save we make each header independent.
-	pk.Header = pk.Header.DeepCopy()
-	pk.LinkHeader = append(buffer.View(nil), pk.LinkHeader...)
-	pk.NetworkHeader = append(buffer.View(nil), pk.NetworkHeader...)
-	pk.TransportHeader = append(buffer.View(nil), pk.TransportHeader...)
-}
diff --git a/pkg/tcpip/stack/registration.go b/pkg/tcpip/stack/registration.go
index ac043b722..23ca9ee03 100644
--- a/pkg/tcpip/stack/registration.go
+++ b/pkg/tcpip/stack/registration.go
@@ -246,7 +246,7 @@ type NetworkEndpoint interface {
 
 	// WritePackets writes packets to the given destination address and
 	// protocol. pkts must not be zero length.
-	WritePackets(r *Route, gso *GSO, pkts []PacketBuffer, params NetworkHeaderParams) (int, *tcpip.Error)
+	WritePackets(r *Route, gso *GSO, pkts PacketBufferList, params NetworkHeaderParams) (int, *tcpip.Error)
 
 	// WriteHeaderIncludedPacket writes a packet that includes a network
 	// header to the given destination address.
@@ -393,7 +393,7 @@ type LinkEndpoint interface {
 	// Right now, WritePackets is used only when the software segmentation
 	// offload is enabled. If it will be used for something else, it may
 	// require to change syscall filters.
-	WritePackets(r *Route, gso *GSO, pkts []PacketBuffer, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error)
+	WritePackets(r *Route, gso *GSO, pkts PacketBufferList, protocol tcpip.NetworkProtocolNumber) (int, *tcpip.Error)
 
 	// WriteRawPacket writes a packet directly to the link. The packet
 	// should already have an ethernet header.
diff --git a/pkg/tcpip/stack/route.go b/pkg/tcpip/stack/route.go
index 9fbe8a411..a0e5e0300 100644
--- a/pkg/tcpip/stack/route.go
+++ b/pkg/tcpip/stack/route.go
@@ -168,23 +168,26 @@ func (r *Route) WritePacket(gso *GSO, params NetworkHeaderParams, pkt PacketBuff
 	return err
 }
 
-// WritePackets writes the set of packets through the given route.
-func (r *Route) WritePackets(gso *GSO, pkts []PacketBuffer, params NetworkHeaderParams) (int, *tcpip.Error) {
+// WritePackets writes a list of n packets through the given route and returns
+// the number of packets written.
+func (r *Route) WritePackets(gso *GSO, pkts PacketBufferList, params NetworkHeaderParams) (int, *tcpip.Error) {
 	if !r.ref.isValidForOutgoing() {
 		return 0, tcpip.ErrInvalidEndpointState
 	}
 
 	n, err := r.ref.ep.WritePackets(r, gso, pkts, params)
 	if err != nil {
-		r.Stats().IP.OutgoingPacketErrors.IncrementBy(uint64(len(pkts) - n))
+		r.Stats().IP.OutgoingPacketErrors.IncrementBy(uint64(pkts.Len() - n))
 	}
 	r.ref.nic.stats.Tx.Packets.IncrementBy(uint64(n))
-	payloadSize := 0
-	for i := 0; i < n; i++ {
-		r.ref.nic.stats.Tx.Bytes.IncrementBy(uint64(pkts[i].Header.UsedLength()))
-		payloadSize += pkts[i].DataSize
+
+	writtenBytes := 0
+	for i, pb := 0, pkts.Front(); i < n && pb != nil; i, pb = i+1, pb.Next() {
+		writtenBytes += pb.Header.UsedLength()
+		writtenBytes += pb.Data.Size()
 	}
-	r.ref.nic.stats.Tx.Bytes.IncrementBy(uint64(payloadSize))
+
+	r.ref.nic.stats.Tx.Bytes.IncrementBy(uint64(writtenBytes))
 	return n, err
 }
 
diff --git a/pkg/tcpip/stack/stack_test.go b/pkg/tcpip/stack/stack_test.go
index b8543b71e..3f8a2a095 100644
--- a/pkg/tcpip/stack/stack_test.go
+++ b/pkg/tcpip/stack/stack_test.go
@@ -153,7 +153,7 @@ func (f *fakeNetworkEndpoint) WritePacket(r *stack.Route, gso *stack.GSO, params
 }
 
 // WritePackets implements stack.LinkEndpoint.WritePackets.
-func (f *fakeNetworkEndpoint) WritePackets(r *stack.Route, gso *stack.GSO, pkts []stack.PacketBuffer, params stack.NetworkHeaderParams) (int, *tcpip.Error) {
+func (f *fakeNetworkEndpoint) WritePackets(r *stack.Route, gso *stack.GSO, pkts stack.PacketBufferList, params stack.NetworkHeaderParams) (int, *tcpip.Error) {
 	panic("not implemented")
 }
 
diff --git a/pkg/tcpip/transport/tcp/connect.go b/pkg/tcpip/transport/tcp/connect.go
index 3239a5911..2ca3fb809 100644
--- a/pkg/tcpip/transport/tcp/connect.go
+++ b/pkg/tcpip/transport/tcp/connect.go
@@ -756,8 +756,7 @@ func (e *endpoint) sendTCP(r *stack.Route, tf tcpFields, data buffer.VectorisedV
 func buildTCPHdr(r *stack.Route, tf tcpFields, pkt *stack.PacketBuffer, gso *stack.GSO) {
 	optLen := len(tf.opts)
 	hdr := &pkt.Header
-	packetSize := pkt.DataSize
-	off := pkt.DataOffset
+	packetSize := pkt.Data.Size()
 	// Initialize the header.
 	tcp := header.TCP(hdr.Prepend(header.TCPMinimumSize + optLen))
 	pkt.TransportHeader = buffer.View(tcp)
@@ -782,12 +781,18 @@ func buildTCPHdr(r *stack.Route, tf tcpFields, pkt *stack.PacketBuffer, gso *sta
 		// header and data and get the right sum of the TCP packet.
 		tcp.SetChecksum(xsum)
 	} else if r.Capabilities()&stack.CapabilityTXChecksumOffload == 0 {
-		xsum = header.ChecksumVVWithOffset(pkt.Data, xsum, off, packetSize)
+		xsum = header.ChecksumVV(pkt.Data, xsum)
 		tcp.SetChecksum(^tcp.CalculateChecksum(xsum))
 	}
 }
 
 func sendTCPBatch(r *stack.Route, tf tcpFields, data buffer.VectorisedView, gso *stack.GSO, owner tcpip.PacketOwner) *tcpip.Error {
+	// We need to shallow clone the VectorisedView here as ReadToView will
+	// split the VectorisedView and Trim underlying views as it splits. Not
+	// doing the clone here will cause the underlying views of data itself
+	// to be altered.
+	data = data.Clone(nil)
+
 	optLen := len(tf.opts)
 	if tf.rcvWnd > 0xffff {
 		tf.rcvWnd = 0xffff
@@ -796,31 +801,25 @@ func sendTCPBatch(r *stack.Route, tf tcpFields, data buffer.VectorisedView, gso
 	mss := int(gso.MSS)
 	n := (data.Size() + mss - 1) / mss
 
-	// Allocate one big slice for all the headers.
-	hdrSize := header.TCPMinimumSize + int(r.MaxHeaderLength()) + optLen
-	buf := make([]byte, n*hdrSize)
-	pkts := make([]stack.PacketBuffer, n)
-	for i := range pkts {
-		pkts[i].Header = buffer.NewEmptyPrependableFromView(buf[i*hdrSize:][:hdrSize])
-	}
-
 	size := data.Size()
-	off := 0
+	hdrSize := header.TCPMinimumSize + int(r.MaxHeaderLength()) + optLen
+	var pkts stack.PacketBufferList
 	for i := 0; i < n; i++ {
 		packetSize := mss
 		if packetSize > size {
 			packetSize = size
 		}
 		size -= packetSize
-		pkts[i].DataOffset = off
-		pkts[i].DataSize = packetSize
-		pkts[i].Data = data
-		pkts[i].Hash = tf.txHash
-		pkts[i].Owner = owner
-		buildTCPHdr(r, tf, &pkts[i], gso)
-		off += packetSize
+		var pkt stack.PacketBuffer
+		pkt.Header = buffer.NewPrependable(hdrSize)
+		pkt.Hash = tf.txHash
+		pkt.Owner = owner
+		data.ReadToVV(&pkt.Data, packetSize)
+		buildTCPHdr(r, tf, &pkt, gso)
 		tf.seq = tf.seq.Add(seqnum.Size(packetSize))
+		pkts.PushBack(&pkt)
 	}
+
 	if tf.ttl == 0 {
 		tf.ttl = r.DefaultTTL()
 	}
@@ -845,12 +844,10 @@ func sendTCP(r *stack.Route, tf tcpFields, data buffer.VectorisedView, gso *stac
 	}
 
 	pkt := stack.PacketBuffer{
-		Header:     buffer.NewPrependable(header.TCPMinimumSize + int(r.MaxHeaderLength()) + optLen),
-		DataOffset: 0,
-		DataSize:   data.Size(),
-		Data:       data,
-		Hash:       tf.txHash,
-		Owner:      owner,
+		Header: buffer.NewPrependable(header.TCPMinimumSize + int(r.MaxHeaderLength()) + optLen),
+		Data:   data,
+		Hash:   tf.txHash,
+		Owner:  owner,
 	}
 	buildTCPHdr(r, tf, &pkt, gso)
 
diff --git a/pkg/tcpip/transport/tcp/segment.go b/pkg/tcpip/transport/tcp/segment.go
index e6fe7985d..40461fd31 100644
--- a/pkg/tcpip/transport/tcp/segment.go
+++ b/pkg/tcpip/transport/tcp/segment.go
@@ -77,9 +77,11 @@ func newSegmentFromView(r *stack.Route, id stack.TransportEndpointID, v buffer.V
 		id:     id,
 		route:  r.Clone(),
 	}
-	s.views[0] = v
-	s.data = buffer.NewVectorisedView(len(v), s.views[:1])
 	s.rcvdTime = time.Now()
+	if len(v) != 0 {
+		s.views[0] = v
+		s.data = buffer.NewVectorisedView(len(v), s.views[:1])
+	}
 	return s
 }