Merge 1e48c1007e into ba5b671e14

This commit introduces the following methods:

  - GetRuleByHandle
  - ResetRule
  - ResetRules

It also refactors GetRules and the deprecated GetRule methods to share
a common getRules implementation.

conn.go

@@ -233,6 +233,19 @@ func (cc *Conn) CloseLasting() error {
 
 // Flush sends all buffered commands in a single batch to nftables.
 func (cc *Conn) Flush() error {
+	return cc.flush(0)
+}
+
+// FlushWithGenID sends all buffered commands in a single batch to nftables
+// along with the provided gen ID. If the ruleset has changed since the gen ID
+// was retrieved, an ERESTART error will be returned.
+func (cc *Conn) FlushWithGenID(genID uint32) error {
+	return cc.flush(genID)
+}
+
+// flush sends all buffered commands in a single batch to nftables. If genID is
+// non-zero, it will be included in the batch messages.
+func (cc *Conn) flush(genID uint32) error {
 	cc.mu.Lock()
 	defer func() {
 		cc.messages = nil
@@ -259,7 +272,12 @@ func (cc *Conn) Flush() error {
 		return err
 	}
 
-	messages, err := conn.SendMessages(batch(cc.messages))
+	batch, err := batch(cc.messages, genID)
+	if err != nil {
+		return err
+	}
+
+	messages, err := conn.SendMessages(batch)
 	if err != nil {
 		return fmt.Errorf("SendMessages: %w", err)
 	}
@@ -388,14 +406,30 @@ func (cc *Conn) marshalExpr(fam byte, e expr.Any) []byte {
 	return b
 }
 
-func batch(messages []netlinkMessage) []netlink.Message {
+// batch wraps the given messages in a batch begin and end message, and returns
+// the resulting slice of netlink messages. If the genID is non-zero, it will be
+// included in both batch messages.
+func batch(messages []netlinkMessage, genID uint32) ([]netlink.Message, error) {
 	batch := make([]netlink.Message, len(messages)+2)
+
+	data := extraHeader(0, unix.NFNL_SUBSYS_NFTABLES)
+
+	if genID > 0 {
+		attr, err := netlink.MarshalAttributes([]netlink.Attribute{
+			{Type: unix.NFNL_BATCH_GENID, Data: binaryutil.BigEndian.PutUint32(genID)},
+		})
+		if err != nil {
+			return nil, err
+		}
+		data = append(data, attr...)
+	}
+
 	batch[0] = netlink.Message{
 		Header: netlink.Header{
 			Type:  netlink.HeaderType(unix.NFNL_MSG_BATCH_BEGIN),
 			Flags: netlink.Request,
 		},
-		Data: extraHeader(0, unix.NFNL_SUBSYS_NFTABLES),
+		Data: data,
 	}
 
 	for i, msg := range messages {
@@ -410,10 +444,10 @@ func batch(messages []netlinkMessage) []netlink.Message {
 			Type:  netlink.HeaderType(unix.NFNL_MSG_BATCH_END),
 			Flags: netlink.Request,
 		},
-		Data: extraHeader(0, unix.NFNL_SUBSYS_NFTABLES),
+		Data: data,
 	}
 
-	return batch
+	return batch, nil
 }
 
 // allocateTransactionID allocates an identifier which is only valid in the

gen.go

@@ -8,15 +8,18 @@ import (
 	"golang.org/x/sys/unix"
 )
 
-type GenMsg struct {
+type Gen struct {
 	ID       uint32
 	ProcPID  uint32
 	ProcComm string // [16]byte - max 16bytes - kernel TASK_COMM_LEN
 }
 
+// Deprecated: GenMsg is an inconsistent old name for Gen. Prefer using Gen.
+type GenMsg = Gen
+
 const genHeaderType = netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | unix.NFT_MSG_NEWGEN)
 
-func genFromMsg(msg netlink.Message) (*GenMsg, error) {
+func genFromMsg(msg netlink.Message) (*Gen, error) {
 	if got, want := msg.Header.Type, genHeaderType; got != want {
 		return nil, fmt.Errorf("unexpected header type: got %v, want %v", got, want)
 	}
@@ -26,7 +29,7 @@ func genFromMsg(msg netlink.Message) (*GenMsg, error) {
 	}
 	ad.ByteOrder = binary.BigEndian
 
-	msgOut := &GenMsg{}
+	msgOut := &Gen{}
 	for ad.Next() {
 		switch ad.Type() {
 		case unix.NFTA_GEN_ID:
@@ -36,7 +39,7 @@ func genFromMsg(msg netlink.Message) (*GenMsg, error) {
 		case unix.NFTA_GEN_PROC_NAME:
 			msgOut.ProcComm = ad.String()
 		default:
-			return nil, fmt.Errorf("Unknown attribute: %d %v\n", ad.Type(), ad.Bytes())
+			return nil, fmt.Errorf("unknown attribute: %d, %v", ad.Type(), ad.Bytes())
 		}
 	}
 	if err := ad.Err(); err != nil {
@@ -44,3 +47,42 @@ func genFromMsg(msg netlink.Message) (*GenMsg, error) {
 	}
 	return msgOut, nil
 }
+
+// GetGen retrieves the current nftables generation ID together with the name
+// and ID of the process that last modified the ruleset.
+// https://docs.kernel.org/networking/netlink_spec/nftables.html#getgen
+func (cc *Conn) GetGen() (*Gen, error) {
+	conn, closer, err := cc.netlinkConn()
+	if err != nil {
+		return nil, err
+	}
+	defer func() { _ = closer() }()
+
+	data, err := netlink.MarshalAttributes([]netlink.Attribute{
+		{Type: unix.NFTA_GEN_ID},
+	})
+	if err != nil {
+		return nil, err
+	}
+
+	message := netlink.Message{
+		Header: netlink.Header{
+			Type:  netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | unix.NFT_MSG_GETGEN),
+			Flags: netlink.Request | netlink.Acknowledge,
+		},
+		Data: append(extraHeader(0, 0), data...),
+	}
+
+	if _, err := conn.SendMessages([]netlink.Message{message}); err != nil {
+		return nil, fmt.Errorf("SendMessages: %v", err)
+	}
+
+	reply, err := receiveAckAware(conn, message.Header.Flags)
+	if err != nil {
+		return nil, fmt.Errorf("receiveAckAware: %v", err)
+	}
+	if len(reply) == 0 {
+		return nil, fmt.Errorf("receiveAckAware: no reply")
+	}
+	return genFromMsg(reply[0])
+}

monitor_test.go

@@ -71,7 +71,7 @@ func TestMonitor(t *testing.T) {
 				return
 			}
 
-			genMsg := event.GeneratedBy.Data.(*nftables.GenMsg)
+			genMsg := event.GeneratedBy.Data.(*nftables.Gen)
 			fileName := filepath.Base(os.Args[0])
 
 			if genMsg.ProcComm != fileName {

nftables_test.go

@@ -128,6 +128,47 @@ func ifname(n string) []byte {
 	return b
 }
 
+func TestTableCreateDestroy(t *testing.T) {
+	c, newNS := nftest.OpenSystemConn(t, *enableSysTests)
+	defer nftest.CleanupSystemConn(t, newNS)
+	defer c.FlushRuleset()
+
+	filter := &nftables.Table{
+		Family: nftables.TableFamilyIPv4,
+		Name:   "filter",
+	}
+	c.DestroyTable(filter)
+	c.AddTable(filter)
+	err := c.Flush()
+	if err != nil {
+		t.Fatalf("on Flush: %q", err.Error())
+	}
+
+	lookupMyTable := func() bool {
+		ts, err := c.ListTables()
+		if err != nil {
+			t.Fatalf("on ListTables: %q", err.Error())
+		}
+		return slices.ContainsFunc(ts, func(t *nftables.Table) bool {
+			return t.Name == filter.Name && t.Family == filter.Family
+		})
+	}
+	if !lookupMyTable() {
+		t.Fatal("AddTable doesn't create my table!")
+	}
+
+	c.DestroyTable(filter)
+	if err = c.Flush(); err != nil {
+		t.Fatalf("on Flush: %q", err.Error())
+	}
+
+	if lookupMyTable() {
+		t.Fatal("DestroyTable doesn't delete my table!")
+	}
+
+	c.DestroyTable(filter) // just for test that 'destroy' ignore error 'not found'
+}
+
 func TestRuleOperations(t *testing.T) {
 	// Create a new network namespace to test these operations,
 	// and tear down the namespace at test completion.
@@ -3777,7 +3818,7 @@ func TestDeleteElementNamedSet(t *testing.T) {
 		Name:    "test",
 		KeyType: nftables.TypeInetService,
 	}
-	if err := c.AddSet(portSet, []nftables.SetElement{{Key: []byte{0, 22}}, {Key: []byte{0, 23}}}); err != nil {
+	if err := c.AddSet(portSet, []nftables.SetElement{{Key: []byte{0, 22}}, {Key: []byte{0, 23}}, {Key: []byte{0, 24}}}); err != nil {
 		t.Errorf("c.AddSet(portSet) failed: %v", err)
 	}
 	if err := c.Flush(); err != nil {
@@ -3794,6 +3835,22 @@ func TestDeleteElementNamedSet(t *testing.T) {
 	if err != nil {
 		t.Errorf("c.GetSets() failed: %v", err)
 	}
+	if len(elems) != 2 {
+		t.Fatalf("len(elems) = %d, want 2", len(elems))
+	}
+
+	c.SetDestroyElements(portSet, []nftables.SetElement{{Key: []byte{0, 24}}})
+	c.SetDestroyElements(portSet, []nftables.SetElement{{Key: []byte{0, 24}}})
+	c.SetDestroyElements(portSet, []nftables.SetElement{{Key: []byte{0, 99}}})
+
+	if err := c.Flush(); err != nil {
+		t.Errorf("Third c.Flush() failed: %v", err)
+	}
+
+	elems, err = c.GetSetElements(portSet)
+	if err != nil {
+		t.Errorf("c.GetSets() failed: %v", err)
+	}
 	if len(elems) != 1 {
 		t.Fatalf("len(elems) = %d, want 1", len(elems))
 	}
@@ -7429,3 +7486,303 @@ func TestAutoBufferSize(t *testing.T) {
 		t.Fatalf("failed to flush: %v", err)
 	}
 }
+
+func TestGetGen(t *testing.T) {
+	conn, newNS := nftest.OpenSystemConn(t, *enableSysTests)
+	defer nftest.CleanupSystemConn(t, newNS)
+	defer conn.FlushRuleset()
+
+	gen, err := conn.GetGen()
+	if err != nil {
+		t.Fatalf("failed to get gen: %v", err)
+	}
+
+	conn.AddTable(&nftables.Table{
+		Name:   "test-table",
+		Family: nftables.TableFamilyIPv4,
+	})
+
+	// Flush to increment the generation ID.
+	if err := conn.Flush(); err != nil {
+		t.Fatalf("failed to flush: %v", err)
+	}
+
+	newGen, err := conn.GetGen()
+	if err != nil {
+		t.Fatalf("failed to get gen: %v", err)
+	}
+
+	if newGen.ID <= gen.ID {
+		t.Fatalf("gen ID did not increase, got %d, want > %d", newGen.ID, gen.ID)
+	}
+
+	if newGen.ProcComm != gen.ProcComm {
+		t.Errorf("gen ProcComm changed, got %s, want %s", newGen.ProcComm, gen.ProcComm)
+	}
+
+	if newGen.ProcPID != gen.ProcPID {
+		t.Errorf("gen ProcPID changed, got %d, want %d", newGen.ProcPID, gen.ProcPID)
+	}
+}
+
+func TestFlushWithGenID(t *testing.T) {
+	conn, newNS := nftest.OpenSystemConn(t, *enableSysTests)
+	defer nftest.CleanupSystemConn(t, newNS)
+	defer conn.FlushRuleset()
+
+	gen, err := conn.GetGen()
+	if err != nil {
+		t.Fatalf("failed to get gen: %v", err)
+	}
+
+	conn.AddTable(&nftables.Table{
+		Name:   "test-table",
+		Family: nftables.TableFamilyIPv4,
+	})
+
+	// Flush to increment the generation ID.
+	if err := conn.Flush(); err != nil {
+		t.Fatalf("failed to flush: %v", err)
+	}
+
+	conn.AddTable(&nftables.Table{
+		Name:   "test-table-2",
+		Family: nftables.TableFamilyIPv4,
+	})
+
+	err = conn.FlushWithGenID(gen.ID)
+	if err == nil || !errors.Is(err, syscall.ERESTART) {
+		t.Errorf("expected error to be ERESTART, got: %v", err)
+	}
+
+	table, err := conn.ListTable("test-table-2")
+	if table != nil && !errors.Is(err, syscall.ENOENT) {
+		t.Errorf("expected table to not exist, got: %v", table)
+	}
+}
+
+func TestGetRuleByHandle(t *testing.T) {
+	conn, newNS := nftest.OpenSystemConn(t, *enableSysTests)
+	defer nftest.CleanupSystemConn(t, newNS)
+	defer conn.FlushRuleset()
+
+	table := conn.AddTable(&nftables.Table{
+		Name:   "test-table",
+		Family: nftables.TableFamilyIPv4,
+	})
+
+	chain := conn.AddChain(&nftables.Chain{
+		Name:  "test-chain",
+		Table: table,
+	})
+
+	for i := range 3 {
+		conn.AddRule(&nftables.Rule{
+			Table:    table,
+			Chain:    chain,
+			UserData: fmt.Appendf([]byte{}, "rule-%d", i+1),
+			Exprs: []expr.Any{
+				&expr.Verdict{
+					Kind: expr.VerdictAccept,
+				},
+			},
+		})
+	}
+
+	if err := conn.Flush(); err != nil {
+		t.Fatalf("failed to flush: %v", err)
+	}
+
+	rules, err := conn.GetRules(table, chain)
+	if err != nil {
+		t.Fatalf("GetRules failed: %v", err)
+	}
+
+	want := rules[1]
+
+	got, err := conn.GetRuleByHandle(table, chain, want.Handle)
+	if err != nil {
+		t.Fatalf("GetRuleByHandle failed: %v", err)
+	}
+	if !bytes.Equal(got.UserData, want.UserData) {
+		t.Fatalf("expected userdata %q, got %q", got.UserData, want.UserData)
+	}
+}
+
+func TestResetRule(t *testing.T) {
+	conn, newNS := nftest.OpenSystemConn(t, *enableSysTests)
+	defer nftest.CleanupSystemConn(t, newNS)
+	defer conn.FlushRuleset()
+
+	table := conn.AddTable(&nftables.Table{
+		Name:   "test-table",
+		Family: nftables.TableFamilyIPv4,
+	})
+
+	chain := conn.AddChain(&nftables.Chain{
+		Name:  "test-chain",
+		Table: table,
+	})
+
+	tests := [...]struct {
+		Bytes   uint64
+		Packets uint64
+		Reset   bool
+	}{
+		{
+			Bytes:   1024,
+			Packets: 1,
+			Reset:   false,
+		},
+		{
+			Bytes:   2048,
+			Packets: 2,
+			Reset:   true,
+		},
+		{
+			Bytes:   4096,
+			Packets: 4,
+			Reset:   false,
+		},
+	}
+
+	for _, tt := range tests {
+		conn.AddRule(&nftables.Rule{
+			Table: table,
+			Chain: chain,
+			Exprs: []expr.Any{
+				&expr.Counter{
+					Bytes:   tt.Bytes,
+					Packets: tt.Packets,
+				},
+				&expr.Verdict{
+					Kind: expr.VerdictAccept,
+				},
+			},
+		})
+	}
+
+	if err := conn.Flush(); err != nil {
+		t.Fatalf("flush failed: %v", err)
+	}
+
+	rules, err := conn.GetRules(table, chain)
+	if err != nil {
+		t.Fatalf("GetRules failed: %v", err)
+	}
+
+	if len(rules) != len(tests) {
+		t.Fatalf("expected %d rules, got %d", len(tests), len(rules))
+	}
+
+	for i, r := range rules {
+		if !tests[i].Reset {
+			continue
+		}
+		_, err := conn.ResetRule(table, chain, r.Handle)
+		if err != nil {
+			t.Fatalf("ResetRule failed: %v", err)
+		}
+	}
+
+	rules, err = conn.GetRules(table, chain)
+	if err != nil {
+		t.Fatalf("GetRules failed: %v", err)
+	}
+
+	for i, r := range rules {
+		counter, ok := r.Exprs[0].(*expr.Counter)
+		if !ok {
+			t.Errorf("expected first expr to be Counter, got %T", r.Exprs[0])
+		}
+
+		if tests[i].Reset {
+			if counter.Bytes != 0 || counter.Packets != 0 {
+				t.Errorf(
+					"expected counter values to be reset to zero, got Bytes=%d, Packets=%d",
+					counter.Bytes,
+					counter.Packets,
+				)
+			}
+		} else {
+			// Making sure that only the selected rules were reset
+			if counter.Bytes != tests[i].Bytes || counter.Packets != tests[i].Packets {
+				t.Errorf(
+					"unexpected counter values: got Bytes=%d, Packets=%d, want Bytes=%d, Packets=%d",
+					counter.Bytes,
+					counter.Packets,
+					tests[i].Bytes,
+					tests[i].Packets)
+			}
+		}
+	}
+}
+
+func TestResetRules(t *testing.T) {
+	conn, newNS := nftest.OpenSystemConn(t, *enableSysTests)
+	defer nftest.CleanupSystemConn(t, newNS)
+	defer conn.FlushRuleset()
+
+	table := conn.AddTable(&nftables.Table{
+		Name:   "test-table",
+		Family: nftables.TableFamilyIPv4,
+	})
+
+	chain := conn.AddChain(&nftables.Chain{
+		Name:  "test-chain",
+		Table: table,
+	})
+
+	for range 3 {
+		conn.AddRule(&nftables.Rule{
+			Table: table,
+			Chain: chain,
+			Exprs: []expr.Any{
+				&expr.Counter{
+					Bytes:   1,
+					Packets: 1,
+				},
+				&expr.Verdict{
+					Kind: expr.VerdictAccept,
+				},
+			},
+		})
+	}
+
+	if err := conn.Flush(); err != nil {
+		t.Fatalf("flush failed: %v", err)
+	}
+
+	rules, err := conn.GetRules(table, chain)
+	if err != nil {
+		t.Fatalf("GetRules failed: %v", err)
+	}
+
+	if len(rules) != 3 {
+		t.Fatalf("expected %d rules, got %d", 3, len(rules))
+	}
+
+	if _, err := conn.ResetRules(table, chain); err != nil {
+		t.Fatalf("ResetRules failed: %v", err)
+	}
+
+	rules, err = conn.GetRules(table, chain)
+	if err != nil {
+		t.Fatalf("GetRules failed: %v", err)
+	}
+
+	for _, r := range rules {
+		counter, ok := r.Exprs[0].(*expr.Counter)
+		if !ok {
+			t.Errorf("expected first expr to be Counter, got %T", r.Exprs[0])
+		}
+
+		if counter.Bytes != 0 || counter.Packets != 0 {
+			t.Errorf(
+				"expected counter values to be reset to zero, got Bytes=%d, Packets=%d",
+				counter.Bytes,
+				counter.Packets,
+			)
+		}
+	}
+}

rule.go

@@ -71,31 +71,98 @@ type Rule struct {
 
 // GetRule returns the rules in the specified table and chain.
 //
-// Deprecated: use GetRules instead.
+// Deprecated: use GetRuleByHandle instead.
 func (cc *Conn) GetRule(t *Table, c *Chain) ([]*Rule, error) {
 	return cc.GetRules(t, c)
 }
 
+// GetRuleByHandle returns the rule in the specified table and chain by its
+// handle.
+// https://docs.kernel.org/networking/netlink_spec/nftables.html#getrule
+func (cc *Conn) GetRuleByHandle(t *Table, c *Chain, handle uint64) (*Rule, error) {
+	rules, err := cc.getRules(t, c, unix.NFT_MSG_GETRULE, handle)
+	if err != nil {
+		return nil, err
+	}
+
+	if got, want := len(rules), 1; got != want {
+		return nil, fmt.Errorf("expected rule count %d, got %d", want, got)
+	}
+
+	return rules[0], nil
+}
+
 // GetRules returns the rules in the specified table and chain.
 func (cc *Conn) GetRules(t *Table, c *Chain) ([]*Rule, error) {
+	return cc.getRules(t, c, unix.NFT_MSG_GETRULE, 0)
+}
+
+// ResetRule resets the stateful expressions (e.g., counters) of the given
+// rule. The reset is applied immediately (no Flush is required). The returned
+// rule reflects its state prior to the reset. The provided rule must have a
+// valid Handle.
+// https://docs.kernel.org/networking/netlink_spec/nftables.html#getrule-reset
+func (cc *Conn) ResetRule(t *Table, c *Chain, handle uint64) (*Rule, error) {
+	if handle == 0 {
+		return nil, fmt.Errorf("rule must have a valid handle")
+	}
+
+	rules, err := cc.getRules(t, c, unix.NFT_MSG_GETRULE_RESET, handle)
+	if err != nil {
+		return nil, err
+	}
+
+	if got, want := len(rules), 1; got != want {
+		return nil, fmt.Errorf("expected rule count %d, got %d", want, got)
+	}
+
+	return rules[0], nil
+}
+
+// ResetRules resets the stateful expressions (e.g., counters) of all rules
+// in the given table and chain. The reset is applied immediately (no Flush
+// is required). The returned rules reflect their state prior to the reset.
+// state.
+// https://docs.kernel.org/networking/netlink_spec/nftables.html#getrule-reset
+func (cc *Conn) ResetRules(t *Table, c *Chain) ([]*Rule, error) {
+	return cc.getRules(t, c, unix.NFT_MSG_GETRULE_RESET, 0)
+}
+
+// getRules retrieves rules from the given table and chain, using the provided
+// msgType (either unix.NFT_MSG_GETRULE or unix.NFT_MSG_GETRULE_RESET). If the
+// handle is non-zero, the operation applies only to the rule with that handle.
+func (cc *Conn) getRules(t *Table, c *Chain, msgType int, handle uint64) ([]*Rule, error) {
 	conn, closer, err := cc.netlinkConn()
 	if err != nil {
 		return nil, err
 	}
 	defer func() { _ = closer() }()
 
-	data, err := netlink.MarshalAttributes([]netlink.Attribute{
+	attrs := []netlink.Attribute{
 		{Type: unix.NFTA_RULE_TABLE, Data: []byte(t.Name + "\x00")},
 		{Type: unix.NFTA_RULE_CHAIN, Data: []byte(c.Name + "\x00")},
-	})
+	}
+
+	var flags netlink.HeaderFlags = netlink.Request | netlink.Acknowledge | netlink.Dump
+
+	if handle != 0 {
+		attrs = append(attrs, netlink.Attribute{
+			Type: unix.NFTA_RULE_HANDLE,
+			Data: binaryutil.BigEndian.PutUint64(handle),
+		})
+
+		flags = netlink.Request | netlink.Acknowledge
+	}
+
+	data, err := netlink.MarshalAttributes(attrs)
 	if err != nil {
 		return nil, err
 	}
 
 	message := netlink.Message{
 		Header: netlink.Header{
-			Type:  netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | unix.NFT_MSG_GETRULE),
-			Flags: netlink.Request | netlink.Acknowledge | netlink.Dump,
+			Type:  netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | msgType),
+			Flags: flags,
 		},
 		Data: append(extraHeader(uint8(t.Family), 0), data...),
 	}

set.go

@@ -44,6 +44,9 @@ const (
 	NFTA_SET_ELEM_KEY_END = 10
 	// https://git.netfilter.org/nftables/tree/include/linux/netfilter/nf_tables.h?id=d1289bff58e1878c3162f574c603da993e29b113#n429
 	NFTA_SET_ELEM_EXPRESSIONS = 0x11
+	// FIXME: in sys@v0.34.0 no unix.NFT_MSG_DESTROYSETELEM const yet.
+	// See nf_tables_msg_types enum in https://git.netfilter.org/nftables/tree/include/linux/netfilter/nf_tables.h
+	NFT_MSG_DESTROYSETELEM = 0x1e
 )
 
 // SetDatatype represents a datatype declared by nft.
@@ -391,6 +394,16 @@ func (cc *Conn) SetDeleteElements(s *Set, vals []SetElement) error {
 	return cc.appendElemList(s, vals, unix.NFT_MSG_DELSETELEM)
 }
 
+// SetDestroyElements like SetDeleteElements, but not an error if setelement doesn't exists
+func (cc *Conn) SetDestroyElements(s *Set, vals []SetElement) error {
+	cc.mu.Lock()
+	defer cc.mu.Unlock()
+	if s.Anonymous {
+		return errors.New("anonymous sets cannot be updated")
+	}
+	return cc.appendElemList(s, vals, NFT_MSG_DESTROYSETELEM)
+}
+
 // maxElemBatchSize is the maximum size in bytes of encoded set elements which
 // are sent in one netlink message. The size field of a netlink attribute is a
 // uint16, and 1024 bytes is more than enough for the per-message headers.
@@ -826,8 +839,9 @@ func parseSetDatatype(magic uint32) (SetDatatype, error) {
 }
 
 const (
-	newElemHeaderType = netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | unix.NFT_MSG_NEWSETELEM)
-	delElemHeaderType = netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | unix.NFT_MSG_DELSETELEM)
+	newElemHeaderType     = netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | unix.NFT_MSG_NEWSETELEM)
+	delElemHeaderType     = netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | unix.NFT_MSG_DELSETELEM)
+	destroyElemHeaderType = netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | NFT_MSG_DESTROYSETELEM)
 )
 
 func elementsFromMsg(fam byte, msg netlink.Message) ([]SetElement, error) {

table.go

@@ -24,6 +24,10 @@ import (
 const (
 	newTableHeaderType = netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | unix.NFT_MSG_NEWTABLE)
 	delTableHeaderType = netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | unix.NFT_MSG_DELTABLE)
+
+	// FIXME: in sys@v0.34.0 no unix.NFT_MSG_DESTROYTABLE const yet.
+	// See nf_tables_msg_types enum in https://git.netfilter.org/nftables/tree/include/linux/netfilter/nf_tables.h
+	destroyTableHeaderType = netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | 0x1a)
 )
 
 // TableFamily specifies the address family for this table.
@@ -51,15 +55,25 @@ type Table struct {
 
 // DelTable deletes a specific table, along with all chains/rules it contains.
 func (cc *Conn) DelTable(t *Table) {
+	cc.delTable(t, delTableHeaderType)
+}
+
+// DestroyTable is like DelTable, but not an error if table doesn't exists
+func (cc *Conn) DestroyTable(t *Table) {
+	cc.delTable(t, destroyTableHeaderType)
+}
+
+func (cc *Conn) delTable(t *Table, hdrType netlink.HeaderType) {
 	cc.mu.Lock()
 	defer cc.mu.Unlock()
 	data := cc.marshalAttr([]netlink.Attribute{
 		{Type: unix.NFTA_TABLE_NAME, Data: []byte(t.Name + "\x00")},
 		{Type: unix.NFTA_TABLE_FLAGS, Data: []byte{0, 0, 0, 0}},
 	})
+
 	cc.messages = append(cc.messages, netlinkMessage{
 		Header: netlink.Header{
-			Type:  netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | unix.NFT_MSG_DELTABLE),
+			Type:  hdrType,
 			Flags: netlink.Request | netlink.Acknowledge,
 		},
 		Data: append(extraHeader(uint8(t.Family), 0), data...),