nftables/set.go

1001 lines
34 KiB
Go
Raw Normal View History

// Copyright 2018 Google LLC. All Rights Reserved.
//
// 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 nftables
import (
"encoding/binary"
"errors"
"fmt"
"math"
"strings"
"time"
"github.com/google/nftables/binaryutil"
2019-08-28 01:51:13 -05:00
"github.com/google/nftables/expr"
2023-05-20 11:23:52 -05:00
"github.com/google/nftables/internal/parseexprfunc"
"github.com/google/nftables/userdata"
"github.com/mdlayher/netlink"
"golang.org/x/sys/unix"
)
// SetConcatTypeBits defines concatination bits, originally defined in
// https://git.netfilter.org/iptables/tree/iptables/nft.c?id=26753888720d8e7eb422ae4311348347f5a05cb4#n1002
const (
SetConcatTypeBits = 6
SetConcatTypeMask = (1 << SetConcatTypeBits) - 1
// below consts added because not found in go unix package
// https://git.netfilter.org/nftables/tree/include/linux/netfilter/nf_tables.h?id=d1289bff58e1878c3162f574c603da993e29b113#n306
NFT_SET_CONCAT = 0x80
// https://git.netfilter.org/nftables/tree/include/linux/netfilter/nf_tables.h?id=d1289bff58e1878c3162f574c603da993e29b113#n330
NFTA_SET_DESC_CONCAT = 2
// https://git.netfilter.org/nftables/tree/include/linux/netfilter/nf_tables.h?id=d1289bff58e1878c3162f574c603da993e29b113#n428
NFTA_SET_ELEM_KEY_END = 10
2022-10-29 01:33:22 -05:00
// https://git.netfilter.org/nftables/tree/include/linux/netfilter/nf_tables.h?id=d1289bff58e1878c3162f574c603da993e29b113#n429
NFTA_SET_ELEM_EXPRESSIONS = 0x11
)
// SetDatatype represents a datatype declared by nft.
type SetDatatype struct {
Name string
Bytes uint32
// nftMagic represents the magic value that nft uses for
// certain types (ie: IP addresses). We populate SET_KEY_TYPE
// identically, so `nft list ...` commands produce correct output.
nftMagic uint32
}
2019-12-01 03:10:42 -06:00
// GetNFTMagic returns a custom datatype based on user's parameters
func (s *SetDatatype) GetNFTMagic() uint32 {
return s.nftMagic
}
// SetNFTMagic returns a custom datatype based on user's parameters
func (s *SetDatatype) SetNFTMagic(nftMagic uint32) {
s.nftMagic = nftMagic
}
// NFT datatypes. See: https://git.netfilter.org/nftables/tree/include/datatype.h
var (
2019-12-01 03:10:42 -06:00
TypeInvalid = SetDatatype{Name: "invalid", nftMagic: 0}
2019-08-28 01:51:13 -05:00
TypeVerdict = SetDatatype{Name: "verdict", Bytes: 0, nftMagic: 1}
TypeNFProto = SetDatatype{Name: "nf_proto", Bytes: 1, nftMagic: 2}
TypeBitmask = SetDatatype{Name: "bitmask", Bytes: 0, nftMagic: 3}
TypeInteger = SetDatatype{Name: "integer", Bytes: 4, nftMagic: 4}
TypeString = SetDatatype{Name: "string", Bytes: 0, nftMagic: 5}
TypeLLAddr = SetDatatype{Name: "ll_addr", Bytes: 0, nftMagic: 6}
TypeIPAddr = SetDatatype{Name: "ipv4_addr", Bytes: 4, nftMagic: 7}
TypeIP6Addr = SetDatatype{Name: "ipv6_addr", Bytes: 16, nftMagic: 8}
TypeEtherAddr = SetDatatype{Name: "ether_addr", Bytes: 6, nftMagic: 9}
TypeEtherType = SetDatatype{Name: "ether_type", Bytes: 2, nftMagic: 10}
TypeARPOp = SetDatatype{Name: "arp_op", Bytes: 2, nftMagic: 11}
TypeInetProto = SetDatatype{Name: "inet_proto", Bytes: 1, nftMagic: 12}
TypeInetService = SetDatatype{Name: "inet_service", Bytes: 2, nftMagic: 13}
TypeICMPType = SetDatatype{Name: "icmp_type", Bytes: 1, nftMagic: 14}
TypeTCPFlag = SetDatatype{Name: "tcp_flag", Bytes: 1, nftMagic: 15}
TypeDCCPPktType = SetDatatype{Name: "dccp_pkttype", Bytes: 1, nftMagic: 16}
TypeMHType = SetDatatype{Name: "mh_type", Bytes: 1, nftMagic: 17}
TypeTime = SetDatatype{Name: "time", Bytes: 8, nftMagic: 18}
TypeMark = SetDatatype{Name: "mark", Bytes: 4, nftMagic: 19}
TypeIFIndex = SetDatatype{Name: "iface_index", Bytes: 4, nftMagic: 20}
TypeARPHRD = SetDatatype{Name: "iface_type", Bytes: 2, nftMagic: 21}
TypeRealm = SetDatatype{Name: "realm", Bytes: 4, nftMagic: 22}
TypeClassID = SetDatatype{Name: "classid", Bytes: 4, nftMagic: 23}
TypeUID = SetDatatype{Name: "uid", Bytes: sizeOfUIDT, nftMagic: 24}
TypeGID = SetDatatype{Name: "gid", Bytes: sizeOfGIDT, nftMagic: 25}
TypeCTState = SetDatatype{Name: "ct_state", Bytes: 4, nftMagic: 26}
TypeCTDir = SetDatatype{Name: "ct_dir", Bytes: 1, nftMagic: 27}
TypeCTStatus = SetDatatype{Name: "ct_status", Bytes: 4, nftMagic: 28}
TypeICMP6Type = SetDatatype{Name: "icmpv6_type", Bytes: 1, nftMagic: 29}
TypeCTLabel = SetDatatype{Name: "ct_label", Bytes: ctLabelBitSize / 8, nftMagic: 30}
TypePktType = SetDatatype{Name: "pkt_type", Bytes: 1, nftMagic: 31}
TypeICMPCode = SetDatatype{Name: "icmp_code", Bytes: 1, nftMagic: 32}
TypeICMPV6Code = SetDatatype{Name: "icmpv6_code", Bytes: 1, nftMagic: 33}
TypeICMPXCode = SetDatatype{Name: "icmpx_code", Bytes: 1, nftMagic: 34}
TypeDevGroup = SetDatatype{Name: "devgroup", Bytes: 4, nftMagic: 35}
TypeDSCP = SetDatatype{Name: "dscp", Bytes: 1, nftMagic: 36}
TypeECN = SetDatatype{Name: "ecn", Bytes: 1, nftMagic: 37}
TypeFIBAddr = SetDatatype{Name: "fib_addrtype", Bytes: 4, nftMagic: 38}
TypeBoolean = SetDatatype{Name: "boolean", Bytes: 1, nftMagic: 39}
TypeCTEventBit = SetDatatype{Name: "ct_event", Bytes: 4, nftMagic: 40}
TypeIFName = SetDatatype{Name: "ifname", Bytes: ifNameSize, nftMagic: 41}
TypeIGMPType = SetDatatype{Name: "igmp_type", Bytes: 1, nftMagic: 42}
TypeTimeDate = SetDatatype{Name: "time", Bytes: 8, nftMagic: 43}
TypeTimeHour = SetDatatype{Name: "hour", Bytes: 4, nftMagic: 44}
TypeTimeDay = SetDatatype{Name: "day", Bytes: 1, nftMagic: 45}
TypeCGroupV2 = SetDatatype{Name: "cgroupsv2", Bytes: 8, nftMagic: 46}
nftDatatypes = []SetDatatype{
TypeVerdict,
TypeNFProto,
TypeBitmask,
TypeInteger,
TypeString,
TypeLLAddr,
TypeIPAddr,
TypeIP6Addr,
TypeEtherAddr,
TypeEtherType,
TypeARPOp,
TypeInetProto,
TypeInetService,
TypeICMPType,
TypeTCPFlag,
TypeDCCPPktType,
TypeMHType,
TypeTime,
TypeMark,
TypeIFIndex,
TypeARPHRD,
TypeRealm,
TypeClassID,
TypeUID,
TypeGID,
TypeCTState,
TypeCTDir,
TypeCTStatus,
TypeICMP6Type,
TypeCTLabel,
TypePktType,
TypeICMPCode,
TypeICMPV6Code,
TypeICMPXCode,
TypeDevGroup,
TypeDSCP,
TypeECN,
TypeFIBAddr,
TypeBoolean,
TypeCTEventBit,
TypeIFName,
TypeIGMPType,
TypeTimeDate,
TypeTimeHour,
TypeTimeDay,
TypeCGroupV2,
}
)
const (
// ctLabelBitSize is defined in https://git.netfilter.org/nftables/tree/src/ct.c.
ctLabelBitSize uint32 = 128
// ifNameSize is called IFNAMSIZ in linux/if.h.
ifNameSize uint32 = 16
// bits/typesizes.h
sizeOfUIDT uint32 = 4
sizeOfGIDT uint32 = 4
)
var nftDatatypesByName map[string]SetDatatype
var nftDatatypesByMagic map[uint32]SetDatatype
// Create maps for efficient datatype lookup.
func init() {
nftDatatypesByName = make(map[string]SetDatatype, len(nftDatatypes))
nftDatatypesByMagic = make(map[uint32]SetDatatype, len(nftDatatypes))
for _, dt := range nftDatatypes {
nftDatatypesByName[dt.Name] = dt
nftDatatypesByMagic[dt.nftMagic] = dt
}
}
// ErrTooManyTypes is the error returned by ConcatSetType, if nftMagic would overflow.
var ErrTooManyTypes = errors.New("too many types to concat")
// MustConcatSetType does the same as ConcatSetType, but panics instead of an
// error. It simplifies safe initialization of global variables.
func MustConcatSetType(types ...SetDatatype) SetDatatype {
t, err := ConcatSetType(types...)
if err != nil {
panic(err)
}
return t
}
// ConcatSetType constructs a new SetDatatype which consists of a concatenation
// of the passed types. It returns ErrTooManyTypes, if nftMagic would overflow
// (more than 5 types).
func ConcatSetType(types ...SetDatatype) (SetDatatype, error) {
if len(types) > 32/SetConcatTypeBits {
return SetDatatype{}, ErrTooManyTypes
}
var magic, bytes uint32
names := make([]string, len(types))
for i, t := range types {
bytes += t.Bytes
// concatenated types pad the length to multiples of the register size (4 bytes)
// see https://git.netfilter.org/nftables/tree/src/datatype.c?id=488356b895024d0944b20feb1f930558726e0877#n1162
if t.Bytes%4 != 0 {
bytes += 4 - (t.Bytes % 4)
}
names[i] = t.Name
magic <<= SetConcatTypeBits
magic |= t.nftMagic & SetConcatTypeMask
}
return SetDatatype{Name: strings.Join(names, " . "), Bytes: bytes, nftMagic: magic}, nil
}
// ConcatSetTypeElements uses the ConcatSetType name to calculate and return
// a list of base types which were used to construct the concatenated type
func ConcatSetTypeElements(t SetDatatype) []SetDatatype {
names := strings.Split(t.Name, " . ")
types := make([]SetDatatype, len(names))
for i, n := range names {
types[i] = nftDatatypesByName[n]
}
return types
}
// Set represents an nftables set. Anonymous sets are only valid within the
// context of a single batch.
type Set struct {
Table *Table
ID uint32
Name string
Anonymous bool
Constant bool
Interval bool
AutoMerge bool
IsMap bool
HasTimeout bool
2022-10-29 01:33:22 -05:00
Counter bool
// Can be updated per evaluation path, per `nft list ruleset`
// indicates that set contains "flags dynamic"
// https://git.netfilter.org/libnftnl/tree/include/linux/netfilter/nf_tables.h?id=84d12cfacf8ddd857a09435f3d982ab6250d250c#n298
Dynamic bool
// Indicates that the set contains a concatenation
// https://git.netfilter.org/nftables/tree/include/linux/netfilter/nf_tables.h?id=d1289bff58e1878c3162f574c603da993e29b113#n306
Concatenation bool
Timeout time.Duration
KeyType SetDatatype
DataType SetDatatype
// Either host (binaryutil.NativeEndian) or big (binaryutil.BigEndian) endian as per
// https://git.netfilter.org/nftables/tree/include/datatype.h?id=d486c9e626405e829221b82d7355558005b26d8a#n109
KeyByteOrder binaryutil.ByteOrder
2024-12-19 03:24:56 -06:00
Comment string
// Indicates that the set has "size" specifier
Size uint32
}
// SetElement represents a data point within a set.
type SetElement struct {
Key []byte
Val []byte
// Field used for definition of ending interval value in concatenated types
// https://git.netfilter.org/libnftnl/tree/include/set_elem.h?id=e2514c0eff4da7e8e0aabd410f7b7d0b7564c880#n11
KeyEnd []byte
IntervalEnd bool
2019-08-28 01:51:13 -05:00
// To support vmap, a caller must be able to pass Verdict type of data.
// If IsMap is true and VerdictData is not nil, then Val of SetElement will be ignored
// and VerdictData will be wrapped into Attribute data.
VerdictData *expr.Verdict
// To support aging of set elements
Timeout time.Duration
2023-04-21 01:26:50 -05:00
// Life left of the "timeout" elements
Expires time.Duration
2023-05-20 11:23:52 -05:00
Counter *expr.Counter
Comment string
}
2023-05-20 11:23:52 -05:00
func (s *SetElement) decode(fam byte) func(b []byte) error {
2019-08-13 15:19:49 -05:00
return func(b []byte) error {
ad, err := netlink.NewAttributeDecoder(b)
if err != nil {
return fmt.Errorf("failed to create nested attribute decoder: %v", err)
}
ad.ByteOrder = binary.BigEndian
for ad.Next() {
switch ad.Type() {
case unix.NFTA_SET_ELEM_KEY:
s.Key, err = decodeElement(ad.Bytes())
if err != nil {
return err
}
case NFTA_SET_ELEM_KEY_END:
s.KeyEnd, err = decodeElement(ad.Bytes())
if err != nil {
return err
}
2019-08-13 15:19:49 -05:00
case unix.NFTA_SET_ELEM_DATA:
s.Val, err = decodeElement(ad.Bytes())
if err != nil {
return err
}
case unix.NFTA_SET_ELEM_FLAGS:
flags := ad.Uint32()
s.IntervalEnd = (flags & unix.NFT_SET_ELEM_INTERVAL_END) != 0
case unix.NFTA_SET_ELEM_TIMEOUT:
2023-04-21 01:26:50 -05:00
s.Timeout = time.Millisecond * time.Duration(ad.Uint64())
case unix.NFTA_SET_ELEM_EXPIRATION:
s.Expires = time.Millisecond * time.Duration(ad.Uint64())
case unix.NFTA_SET_ELEM_USERDATA:
userData := ad.Bytes()
// Try to extract comment from userdata if present
if comment, ok := userdata.GetString(userData, userdata.NFTNL_UDATA_SET_ELEM_COMMENT); ok {
s.Comment = comment
}
2023-05-20 11:23:52 -05:00
case unix.NFTA_SET_ELEM_EXPR:
elems, err := parseexprfunc.ParseExprBytesFunc(fam, ad)
2023-05-20 11:23:52 -05:00
if err != nil {
return err
}
for _, elem := range elems {
switch item := elem.(type) {
case *expr.Counter:
s.Counter = item
}
}
2019-08-13 15:19:49 -05:00
}
}
return ad.Err()
}
}
func decodeElement(d []byte) ([]byte, error) {
ad, err := netlink.NewAttributeDecoder(d)
if err != nil {
return nil, fmt.Errorf("failed to create nested attribute decoder: %v", err)
}
ad.ByteOrder = binary.BigEndian
var b []byte
for ad.Next() {
switch ad.Type() {
case unix.NFTA_SET_ELEM_KEY:
fallthrough
case unix.NFTA_SET_ELEM_DATA:
b = ad.Bytes()
}
}
if err := ad.Err(); err != nil {
return nil, err
}
return b, nil
}
// SetAddElements applies data points to an nftables set.
func (cc *Conn) SetAddElements(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, unix.NFT_MSG_NEWSETELEM)
}
// SetDeleteElements deletes data points from an nftables set.
func (cc *Conn) SetDeleteElements(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, unix.NFT_MSG_DELSETELEM)
}
// 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.
const maxElemBatchSize = math.MaxUint16 - 1024
func (cc *Conn) appendElemList(s *Set, vals []SetElement, hdrType uint16) error {
if len(vals) == 0 {
return nil
}
var elements []netlink.Attribute
batchSize := 0
var batches [][]netlink.Attribute
for i, v := range vals {
item := make([]netlink.Attribute, 0)
var flags uint32
if v.IntervalEnd {
flags |= unix.NFT_SET_ELEM_INTERVAL_END
2019-08-13 15:19:49 -05:00
item = append(item, netlink.Attribute{Type: unix.NFTA_SET_ELEM_FLAGS | unix.NLA_F_NESTED, Data: binaryutil.BigEndian.PutUint32(flags)})
}
2019-08-27 10:52:21 -05:00
encodedKey, err := netlink.MarshalAttributes([]netlink.Attribute{{Type: unix.NFTA_DATA_VALUE, Data: v.Key}})
if err != nil {
return fmt.Errorf("marshal key %d: %v", i, err)
}
item = append(item, netlink.Attribute{Type: unix.NFTA_SET_ELEM_KEY | unix.NLA_F_NESTED, Data: encodedKey})
if len(v.KeyEnd) > 0 {
encodedKeyEnd, err := netlink.MarshalAttributes([]netlink.Attribute{{Type: unix.NFTA_DATA_VALUE, Data: v.KeyEnd}})
if err != nil {
return fmt.Errorf("marshal key end %d: %v", i, err)
}
item = append(item, netlink.Attribute{Type: NFTA_SET_ELEM_KEY_END | unix.NLA_F_NESTED, Data: encodedKeyEnd})
}
if s.HasTimeout && v.Timeout != 0 {
// Set has Timeout flag set, which means an individual element can specify its own timeout.
item = append(item, netlink.Attribute{Type: unix.NFTA_SET_ELEM_TIMEOUT, Data: binaryutil.BigEndian.PutUint64(uint64(v.Timeout.Milliseconds()))})
}
2019-08-28 01:51:13 -05:00
// The following switch statement deal with 3 different types of elements.
// 1. v is an element of vmap
// 2. v is an element of a regular map
// 3. v is an element of a regular set (default)
switch {
case v.VerdictData != nil:
// Since VerdictData is not nil, v is vmap element, need to add to the attributes
encodedVal := []byte{}
encodedKind, err := netlink.MarshalAttributes([]netlink.Attribute{
{Type: unix.NFTA_DATA_VALUE, Data: binaryutil.BigEndian.PutUint32(uint32(v.VerdictData.Kind))},
})
if err != nil {
return fmt.Errorf("marshal item %d: %v", i, err)
2019-08-28 01:51:13 -05:00
}
encodedVal = append(encodedVal, encodedKind...)
if len(v.VerdictData.Chain) != 0 {
encodedChain, err := netlink.MarshalAttributes([]netlink.Attribute{
{Type: unix.NFTA_SET_ELEM_DATA, Data: []byte(v.VerdictData.Chain + "\x00")},
})
if err != nil {
return fmt.Errorf("marshal item %d: %v", i, err)
2019-08-28 01:51:13 -05:00
}
encodedVal = append(encodedVal, encodedChain...)
}
encodedVerdict, err := netlink.MarshalAttributes([]netlink.Attribute{
{Type: unix.NFTA_SET_ELEM_DATA | unix.NLA_F_NESTED, Data: encodedVal}})
if err != nil {
return fmt.Errorf("marshal item %d: %v", i, err)
2019-08-28 01:51:13 -05:00
}
item = append(item, netlink.Attribute{Type: unix.NFTA_SET_ELEM_DATA | unix.NLA_F_NESTED, Data: encodedVerdict})
case len(v.Val) > 0:
// Since v.Val's length is not 0 then, v is a regular map element, need to add to the attributes
2019-08-27 10:52:21 -05:00
encodedVal, err := netlink.MarshalAttributes([]netlink.Attribute{{Type: unix.NFTA_DATA_VALUE, Data: v.Val}})
if err != nil {
return fmt.Errorf("marshal item %d: %v", i, err)
}
2019-08-28 01:51:13 -05:00
item = append(item, netlink.Attribute{Type: unix.NFTA_SET_ELEM_DATA | unix.NLA_F_NESTED, Data: encodedVal})
2019-08-28 01:51:13 -05:00
default:
// If niether of previous cases matche, it means 'e' is an element of a regular Set, no need to add to the attributes
}
// Add comment to userdata if present
if len(v.Comment) > 0 {
userData := userdata.AppendString(nil, userdata.NFTNL_UDATA_SET_ELEM_COMMENT, v.Comment)
item = append(item, netlink.Attribute{Type: unix.NFTA_SET_ELEM_USERDATA, Data: userData})
}
encodedItem, err := netlink.MarshalAttributes(item)
if err != nil {
return fmt.Errorf("marshal item %d: %v", i, err)
}
itemSize := unix.NLA_HDRLEN + len(encodedItem)
if batchSize+itemSize > maxElemBatchSize {
batches = append(batches, elements)
elements = nil
batchSize = 0
}
elements = append(elements, netlink.Attribute{Type: uint16(i+1) | unix.NLA_F_NESTED, Data: encodedItem})
batchSize += itemSize
}
batches = append(batches, elements)
for _, batch := range batches {
encodedElem, err := netlink.MarshalAttributes(batch)
if err != nil {
return fmt.Errorf("marshal elements: %v", err)
}
message := []netlink.Attribute{
{Type: unix.NFTA_SET_ELEM_LIST_SET, Data: []byte(s.Name + "\x00")},
{Type: unix.NFTA_SET_ELEM_LIST_SET_ID, Data: binaryutil.BigEndian.PutUint32(s.ID)},
{Type: unix.NFTA_SET_ELEM_LIST_TABLE, Data: []byte(s.Table.Name + "\x00")},
{Type: unix.NFTA_SET_ELEM_LIST_ELEMENTS | unix.NLA_F_NESTED, Data: encodedElem},
}
Set rule handle during flush This change makes it possible to delete rules after inserting them, without needing to query the rules first. Rules can be deleted both before and after they are flushed. Additionally, this allows positioning a new rule next to an existing rule, both before and after the existing rule is flushed. There are two ways to refer to a rule: Either by ID or by handle. The ID is assigned by userspace, and is only valid within a transaction, so it can only be used before the flush. The handle is assigned by the kernel when the transaction is committed, and can thus only be used after the flush. We thus need to set an ID on each newly created rule, and retrieve the handle of the rule during the flush. There was an existing mechanism to allocate IDs for sets, but this was using a global counter without any synchronization to prevent data races. I replaced this by a new mechanism which uses a connection-scoped counter. I implemented a new mechanism for retrieving replies in Flush, and handling these replies by adding a callback to netlink messages. There was some existing code to handle "overrun", which I deleted, because it was nonsensical and just worked by accident. NLMSG_OVERRUN is in fact not a flag, but a complete message type, so the (re&netlink.Overrun) masking makes no sense. Even better, NLMSG_OVERRUN is never actually used by Linux. What this code was actually doing was skipping over the NFT_MSG_NEWRULE replies, and possibly a NFT_MSG_NEWGEN reply. I had to update all existing tests which compared generated netlink messages against a reference, by inserting the newly added ID attribute. We also need to generate replies for the NFT_MSG_NEWRULE messages with a handle added.
2025-02-20 13:12:30 -06:00
cc.messages = append(cc.messages, netlinkMessage{
Header: netlink.Header{
Type: netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | hdrType),
Flags: netlink.Request | netlink.Acknowledge | netlink.Create,
},
Data: append(extraHeader(uint8(s.Table.Family), 0), cc.marshalAttr(message)...),
})
}
return nil
}
// AddSet adds the specified Set.
func (cc *Conn) AddSet(s *Set, vals []SetElement) error {
cc.mu.Lock()
defer cc.mu.Unlock()
// Based on nft implementation & linux source.
// Link: https://github.com/torvalds/linux/blob/49a57857aeea06ca831043acbb0fa5e0f50602fd/net/netfilter/nf_tables_api.c#L3395
// Another reference: https://git.netfilter.org/nftables/tree/src
if s.Anonymous && !s.Constant {
return errors.New("anonymous structs must be constant")
}
if s.ID == 0 {
Set rule handle during flush This change makes it possible to delete rules after inserting them, without needing to query the rules first. Rules can be deleted both before and after they are flushed. Additionally, this allows positioning a new rule next to an existing rule, both before and after the existing rule is flushed. There are two ways to refer to a rule: Either by ID or by handle. The ID is assigned by userspace, and is only valid within a transaction, so it can only be used before the flush. The handle is assigned by the kernel when the transaction is committed, and can thus only be used after the flush. We thus need to set an ID on each newly created rule, and retrieve the handle of the rule during the flush. There was an existing mechanism to allocate IDs for sets, but this was using a global counter without any synchronization to prevent data races. I replaced this by a new mechanism which uses a connection-scoped counter. I implemented a new mechanism for retrieving replies in Flush, and handling these replies by adding a callback to netlink messages. There was some existing code to handle "overrun", which I deleted, because it was nonsensical and just worked by accident. NLMSG_OVERRUN is in fact not a flag, but a complete message type, so the (re&netlink.Overrun) masking makes no sense. Even better, NLMSG_OVERRUN is never actually used by Linux. What this code was actually doing was skipping over the NFT_MSG_NEWRULE replies, and possibly a NFT_MSG_NEWGEN reply. I had to update all existing tests which compared generated netlink messages against a reference, by inserting the newly added ID attribute. We also need to generate replies for the NFT_MSG_NEWRULE messages with a handle added.
2025-02-20 13:12:30 -06:00
s.ID = cc.allocateTransactionID()
if s.Anonymous {
s.Name = "__set%d"
2019-08-27 10:52:21 -05:00
if s.IsMap {
s.Name = "__map%d"
}
}
}
var flags uint32
if s.Anonymous {
flags |= unix.NFT_SET_ANONYMOUS
}
if s.Constant {
flags |= unix.NFT_SET_CONSTANT
}
if s.Interval {
flags |= unix.NFT_SET_INTERVAL
}
2019-08-27 10:52:21 -05:00
if s.IsMap {
flags |= unix.NFT_SET_MAP
}
if s.HasTimeout {
flags |= unix.NFT_SET_TIMEOUT
}
if s.Dynamic {
flags |= unix.NFT_SET_EVAL
}
if s.Concatenation {
flags |= NFT_SET_CONCAT
}
tableInfo := []netlink.Attribute{
{Type: unix.NFTA_SET_TABLE, Data: []byte(s.Table.Name + "\x00")},
{Type: unix.NFTA_SET_NAME, Data: []byte(s.Name + "\x00")},
{Type: unix.NFTA_SET_FLAGS, Data: binaryutil.BigEndian.PutUint32(flags)},
{Type: unix.NFTA_SET_KEY_TYPE, Data: binaryutil.BigEndian.PutUint32(s.KeyType.nftMagic)},
{Type: unix.NFTA_SET_KEY_LEN, Data: binaryutil.BigEndian.PutUint32(s.KeyType.Bytes)},
{Type: unix.NFTA_SET_ID, Data: binaryutil.BigEndian.PutUint32(s.ID)},
}
2019-08-27 10:52:21 -05:00
if s.IsMap {
2019-08-28 01:51:13 -05:00
// Check if it is vmap case
if s.DataType.nftMagic == 1 {
// For Verdict data type, the expected magic is 0xfffff0
tableInfo = append(tableInfo, netlink.Attribute{Type: unix.NFTA_SET_DATA_TYPE, Data: binaryutil.BigEndian.PutUint32(uint32(unix.NFT_DATA_VERDICT))},
netlink.Attribute{Type: unix.NFTA_SET_DATA_LEN, Data: binaryutil.BigEndian.PutUint32(s.DataType.Bytes)})
} else {
tableInfo = append(tableInfo, netlink.Attribute{Type: unix.NFTA_SET_DATA_TYPE, Data: binaryutil.BigEndian.PutUint32(s.DataType.nftMagic)},
netlink.Attribute{Type: unix.NFTA_SET_DATA_LEN, Data: binaryutil.BigEndian.PutUint32(s.DataType.Bytes)})
}
2019-08-27 10:52:21 -05:00
}
if s.HasTimeout && s.Timeout != 0 {
// If Set's global timeout is specified, add it to set's attributes
tableInfo = append(tableInfo, netlink.Attribute{Type: unix.NFTA_SET_TIMEOUT, Data: binaryutil.BigEndian.PutUint64(uint64(s.Timeout.Milliseconds()))})
}
2019-08-27 10:52:21 -05:00
if s.Constant {
// nft cli tool adds the number of elements to set/map's descriptor
// It make sense to do only if a set or map are constant, otherwise skip NFTA_SET_DESC attribute
numberOfElements, err := netlink.MarshalAttributes([]netlink.Attribute{
{Type: unix.NFTA_DATA_VALUE, Data: binaryutil.BigEndian.PutUint32(uint32(len(vals)))},
})
if err != nil {
return fmt.Errorf("fail to marshal number of elements %d: %v", len(vals), err)
}
tableInfo = append(tableInfo, netlink.Attribute{Type: unix.NLA_F_NESTED | unix.NFTA_SET_DESC, Data: numberOfElements})
}
var descBytes []byte
if s.Size > 0 {
// Marshal set size description
descSizeBytes, err := netlink.MarshalAttributes([]netlink.Attribute{
{Type: unix.NFTA_SET_DESC_SIZE, Data: binaryutil.BigEndian.PutUint32(s.Size)},
})
if err != nil {
return fmt.Errorf("fail to marshal set size description: %w", err)
}
descBytes = append(descBytes, descSizeBytes...)
}
if s.Concatenation {
// Length of concatenated types is a must, otherwise segfaults when executing nft list ruleset
var concatDefinition []byte
elements := ConcatSetTypeElements(s.KeyType)
for i, v := range elements {
// Marshal base type size value
valData, err := netlink.MarshalAttributes([]netlink.Attribute{
{Type: unix.NFTA_DATA_VALUE, Data: binaryutil.BigEndian.PutUint32(v.Bytes)},
})
if err != nil {
return fmt.Errorf("fail to marshal element key size %d: %v", i, err)
}
// Marshal base type size description
descSize, err := netlink.MarshalAttributes([]netlink.Attribute{
{Type: unix.NFTA_SET_DESC_SIZE, Data: valData},
})
if err != nil {
return fmt.Errorf("fail to marshal base type size description: %w", err)
}
concatDefinition = append(concatDefinition, descSize...)
}
// Marshal all base type descriptions into concatenation size description
concatBytes, err := netlink.MarshalAttributes([]netlink.Attribute{{Type: unix.NLA_F_NESTED | NFTA_SET_DESC_CONCAT, Data: concatDefinition}})
if err != nil {
return fmt.Errorf("fail to marshal concat definition %v", err)
}
descBytes = append(descBytes, concatBytes...)
}
if len(descBytes) > 0 {
// Marshal set description
tableInfo = append(tableInfo, netlink.Attribute{Type: unix.NLA_F_NESTED | unix.NFTA_SET_DESC, Data: descBytes})
}
// https://git.netfilter.org/libnftnl/tree/include/udata.h#n17
var userData []byte
if s.Anonymous || s.Constant || s.Interval || s.KeyByteOrder == binaryutil.BigEndian {
// Semantically useless - kept for binary compatability with nft
userData = userdata.AppendUint32(userData, userdata.NFTNL_UDATA_SET_KEYBYTEORDER, 2)
} else if s.KeyByteOrder == binaryutil.NativeEndian {
// Per https://git.netfilter.org/nftables/tree/src/mnl.c?id=187c6d01d35722618c2711bbc49262c286472c8f#n1165
userData = userdata.AppendUint32(userData, userdata.NFTNL_UDATA_SET_KEYBYTEORDER, 1)
}
if s.Interval && s.AutoMerge {
// https://git.netfilter.org/nftables/tree/src/mnl.c?id=187c6d01d35722618c2711bbc49262c286472c8f#n1174
userData = userdata.AppendUint32(userData, userdata.NFTNL_UDATA_SET_MERGE_ELEMENTS, 1)
}
2024-12-19 03:24:56 -06:00
if len(s.Comment) != 0 {
userData = userdata.AppendString(userData, userdata.NFTNL_UDATA_SET_COMMENT, s.Comment)
}
if len(userData) > 0 {
tableInfo = append(tableInfo, netlink.Attribute{Type: unix.NFTA_SET_USERDATA, Data: userData})
}
2022-10-29 01:33:22 -05:00
if s.Counter {
data, err := netlink.MarshalAttributes([]netlink.Attribute{
{Type: unix.NFTA_LIST_ELEM, Data: []byte("counter\x00")},
{Type: unix.NFTA_SET_ELEM_PAD | unix.NFTA_SET_ELEM_DATA, Data: []byte{}},
})
if err != nil {
return err
}
tableInfo = append(tableInfo, netlink.Attribute{Type: unix.NLA_F_NESTED | NFTA_SET_ELEM_EXPRESSIONS, Data: data})
}
Set rule handle during flush This change makes it possible to delete rules after inserting them, without needing to query the rules first. Rules can be deleted both before and after they are flushed. Additionally, this allows positioning a new rule next to an existing rule, both before and after the existing rule is flushed. There are two ways to refer to a rule: Either by ID or by handle. The ID is assigned by userspace, and is only valid within a transaction, so it can only be used before the flush. The handle is assigned by the kernel when the transaction is committed, and can thus only be used after the flush. We thus need to set an ID on each newly created rule, and retrieve the handle of the rule during the flush. There was an existing mechanism to allocate IDs for sets, but this was using a global counter without any synchronization to prevent data races. I replaced this by a new mechanism which uses a connection-scoped counter. I implemented a new mechanism for retrieving replies in Flush, and handling these replies by adding a callback to netlink messages. There was some existing code to handle "overrun", which I deleted, because it was nonsensical and just worked by accident. NLMSG_OVERRUN is in fact not a flag, but a complete message type, so the (re&netlink.Overrun) masking makes no sense. Even better, NLMSG_OVERRUN is never actually used by Linux. What this code was actually doing was skipping over the NFT_MSG_NEWRULE replies, and possibly a NFT_MSG_NEWGEN reply. I had to update all existing tests which compared generated netlink messages against a reference, by inserting the newly added ID attribute. We also need to generate replies for the NFT_MSG_NEWRULE messages with a handle added.
2025-02-20 13:12:30 -06:00
cc.messages = append(cc.messages, netlinkMessage{
Header: netlink.Header{
Type: netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | unix.NFT_MSG_NEWSET),
Flags: netlink.Request | netlink.Acknowledge | netlink.Create,
},
Data: append(extraHeader(uint8(s.Table.Family), 0), cc.marshalAttr(tableInfo)...),
})
// Set the values of the set if initial values were provided.
return cc.appendElemList(s, vals, unix.NFT_MSG_NEWSETELEM)
}
// DelSet deletes a specific set, along with all elements it contains.
func (cc *Conn) DelSet(s *Set) {
cc.mu.Lock()
defer cc.mu.Unlock()
data := cc.marshalAttr([]netlink.Attribute{
{Type: unix.NFTA_SET_TABLE, Data: []byte(s.Table.Name + "\x00")},
{Type: unix.NFTA_SET_NAME, Data: []byte(s.Name + "\x00")},
})
Set rule handle during flush This change makes it possible to delete rules after inserting them, without needing to query the rules first. Rules can be deleted both before and after they are flushed. Additionally, this allows positioning a new rule next to an existing rule, both before and after the existing rule is flushed. There are two ways to refer to a rule: Either by ID or by handle. The ID is assigned by userspace, and is only valid within a transaction, so it can only be used before the flush. The handle is assigned by the kernel when the transaction is committed, and can thus only be used after the flush. We thus need to set an ID on each newly created rule, and retrieve the handle of the rule during the flush. There was an existing mechanism to allocate IDs for sets, but this was using a global counter without any synchronization to prevent data races. I replaced this by a new mechanism which uses a connection-scoped counter. I implemented a new mechanism for retrieving replies in Flush, and handling these replies by adding a callback to netlink messages. There was some existing code to handle "overrun", which I deleted, because it was nonsensical and just worked by accident. NLMSG_OVERRUN is in fact not a flag, but a complete message type, so the (re&netlink.Overrun) masking makes no sense. Even better, NLMSG_OVERRUN is never actually used by Linux. What this code was actually doing was skipping over the NFT_MSG_NEWRULE replies, and possibly a NFT_MSG_NEWGEN reply. I had to update all existing tests which compared generated netlink messages against a reference, by inserting the newly added ID attribute. We also need to generate replies for the NFT_MSG_NEWRULE messages with a handle added.
2025-02-20 13:12:30 -06:00
cc.messages = append(cc.messages, netlinkMessage{
Header: netlink.Header{
Type: netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | unix.NFT_MSG_DELSET),
Flags: netlink.Request | netlink.Acknowledge,
},
Data: append(extraHeader(uint8(s.Table.Family), 0), data...),
})
}
2019-10-23 08:10:52 -05:00
// FlushSet deletes all data points from an nftables set.
func (cc *Conn) FlushSet(s *Set) {
cc.mu.Lock()
defer cc.mu.Unlock()
2019-10-23 08:10:52 -05:00
data := cc.marshalAttr([]netlink.Attribute{
{Type: unix.NFTA_SET_TABLE, Data: []byte(s.Table.Name + "\x00")},
{Type: unix.NFTA_SET_NAME, Data: []byte(s.Name + "\x00")},
})
Set rule handle during flush This change makes it possible to delete rules after inserting them, without needing to query the rules first. Rules can be deleted both before and after they are flushed. Additionally, this allows positioning a new rule next to an existing rule, both before and after the existing rule is flushed. There are two ways to refer to a rule: Either by ID or by handle. The ID is assigned by userspace, and is only valid within a transaction, so it can only be used before the flush. The handle is assigned by the kernel when the transaction is committed, and can thus only be used after the flush. We thus need to set an ID on each newly created rule, and retrieve the handle of the rule during the flush. There was an existing mechanism to allocate IDs for sets, but this was using a global counter without any synchronization to prevent data races. I replaced this by a new mechanism which uses a connection-scoped counter. I implemented a new mechanism for retrieving replies in Flush, and handling these replies by adding a callback to netlink messages. There was some existing code to handle "overrun", which I deleted, because it was nonsensical and just worked by accident. NLMSG_OVERRUN is in fact not a flag, but a complete message type, so the (re&netlink.Overrun) masking makes no sense. Even better, NLMSG_OVERRUN is never actually used by Linux. What this code was actually doing was skipping over the NFT_MSG_NEWRULE replies, and possibly a NFT_MSG_NEWGEN reply. I had to update all existing tests which compared generated netlink messages against a reference, by inserting the newly added ID attribute. We also need to generate replies for the NFT_MSG_NEWRULE messages with a handle added.
2025-02-20 13:12:30 -06:00
cc.messages = append(cc.messages, netlinkMessage{
2019-10-23 08:10:52 -05:00
Header: netlink.Header{
Type: netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | unix.NFT_MSG_DELSETELEM),
Flags: netlink.Request | netlink.Acknowledge,
},
Data: append(extraHeader(uint8(s.Table.Family), 0), data...),
})
}
const (
newSetHeaderType = netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | unix.NFT_MSG_NEWSET)
delSetHeaderType = netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | unix.NFT_MSG_DELSET)
)
func setsFromMsg(msg netlink.Message) (*Set, error) {
if got, want1, want2 := msg.Header.Type, newSetHeaderType, delSetHeaderType; got != want1 && got != want2 {
return nil, fmt.Errorf("unexpected header type: got %v, want %v or %v", got, want1, want2)
}
ad, err := netlink.NewAttributeDecoder(msg.Data[4:])
if err != nil {
return nil, err
}
ad.ByteOrder = binary.BigEndian
var set Set
for ad.Next() {
switch ad.Type() {
case unix.NFTA_SET_NAME:
set.Name = ad.String()
case unix.NFTA_SET_ID:
set.ID = binary.BigEndian.Uint32(ad.Bytes())
case unix.NFTA_SET_TIMEOUT:
set.Timeout = time.Duration(time.Millisecond * time.Duration(binary.BigEndian.Uint64(ad.Bytes())))
set.HasTimeout = true
case unix.NFTA_SET_FLAGS:
flags := ad.Uint32()
set.Constant = (flags & unix.NFT_SET_CONSTANT) != 0
set.Anonymous = (flags & unix.NFT_SET_ANONYMOUS) != 0
set.Interval = (flags & unix.NFT_SET_INTERVAL) != 0
set.IsMap = (flags & unix.NFT_SET_MAP) != 0
set.HasTimeout = (flags & unix.NFT_SET_TIMEOUT) != 0
set.Dynamic = (flags & unix.NFT_SET_EVAL) != 0
set.Concatenation = (flags & NFT_SET_CONCAT) != 0
case unix.NFTA_SET_KEY_TYPE:
nftMagic := ad.Uint32()
dt, err := parseSetDatatype(nftMagic)
if err != nil {
return nil, fmt.Errorf("could not determine data type: %w", err)
}
set.KeyType = dt
case unix.NFTA_SET_KEY_LEN:
set.KeyType.Bytes = binary.BigEndian.Uint32(ad.Bytes())
2019-08-27 10:52:21 -05:00
case unix.NFTA_SET_DATA_TYPE:
nftMagic := ad.Uint32()
// Special case for the data type verdict, in the message it is stored as 0xffffff00 but it is defined as 1
if nftMagic == 0xffffff00 {
set.KeyType = TypeVerdict
break
}
dt, err := parseSetDatatype(nftMagic)
if err != nil {
return nil, fmt.Errorf("could not determine data type: %w", err)
}
set.DataType = dt
case unix.NFTA_SET_DATA_LEN:
set.DataType.Bytes = binary.BigEndian.Uint32(ad.Bytes())
case unix.NFTA_SET_USERDATA:
data := ad.Bytes()
value, ok := userdata.GetUint32(data, userdata.NFTNL_UDATA_SET_MERGE_ELEMENTS)
set.AutoMerge = ok && value == 1
case unix.NFTA_SET_DESC:
nestedAD, err := netlink.NewAttributeDecoder(ad.Bytes())
if err != nil {
return nil, fmt.Errorf("nested NewAttributeDecoder() failed: %w", err)
}
for nestedAD.Next() {
switch nestedAD.Type() {
case unix.NFTA_SET_DESC_SIZE:
set.Size = binary.BigEndian.Uint32(nestedAD.Bytes())
}
}
if nestedAD.Err() != nil {
return nil, fmt.Errorf("decoding set description: %w", nestedAD.Err())
}
}
}
return &set, nil
}
func parseSetDatatype(magic uint32) (SetDatatype, error) {
types := make([]SetDatatype, 0, 32/SetConcatTypeBits)
for magic != 0 {
t := magic & SetConcatTypeMask
magic = magic >> SetConcatTypeBits
dt, ok := nftDatatypesByMagic[t]
if !ok {
return TypeInvalid, fmt.Errorf("could not determine data type %+v", dt)
}
// Because we start with the last type, we insert the later types at the front.
types = append([]SetDatatype{dt}, types...)
}
dt, err := ConcatSetType(types...)
if err != nil {
return TypeInvalid, fmt.Errorf("could not create data type: %w", err)
}
return dt, nil
}
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)
)
2023-05-20 11:23:52 -05:00
func elementsFromMsg(fam byte, msg netlink.Message) ([]SetElement, error) {
if got, want1, want2 := msg.Header.Type, newElemHeaderType, delElemHeaderType; got != want1 && got != want2 {
return nil, fmt.Errorf("unexpected header type: got %v, want %v or %v", got, want1, want2)
}
ad, err := netlink.NewAttributeDecoder(msg.Data[4:])
if err != nil {
return nil, err
}
ad.ByteOrder = binary.BigEndian
var elements []SetElement
for ad.Next() {
b := ad.Bytes()
2019-08-13 15:19:49 -05:00
if ad.Type() == unix.NFTA_SET_ELEM_LIST_ELEMENTS {
ad, err := netlink.NewAttributeDecoder(b)
if err != nil {
return nil, err
}
ad.ByteOrder = binary.BigEndian
for ad.Next() {
2019-08-13 15:19:49 -05:00
var elem SetElement
switch ad.Type() {
2019-08-13 15:19:49 -05:00
case unix.NFTA_LIST_ELEM:
2023-05-20 11:23:52 -05:00
ad.Do(elem.decode(fam))
}
2019-08-13 15:19:49 -05:00
elements = append(elements, elem)
}
}
}
return elements, nil
}
// GetSets returns the sets in the specified table.
func (cc *Conn) GetSets(t *Table) ([]*Set, error) {
conn, closer, err := cc.netlinkConn()
if err != nil {
return nil, err
}
defer func() { _ = closer() }()
data, err := netlink.MarshalAttributes([]netlink.Attribute{
{Type: unix.NFTA_SET_TABLE, Data: []byte(t.Name + "\x00")},
})
if err != nil {
return nil, err
}
message := netlink.Message{
Header: netlink.Header{
Type: netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | unix.NFT_MSG_GETSET),
Flags: netlink.Request | netlink.Acknowledge | netlink.Dump,
},
Data: append(extraHeader(uint8(t.Family), 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 {
2024-08-11 03:00:48 -05:00
return nil, fmt.Errorf("receiveAckAware: %v", err)
}
var sets []*Set
for _, msg := range reply {
s, err := setsFromMsg(msg)
if err != nil {
return nil, err
}
s.Table = &Table{Name: t.Name, Use: t.Use, Flags: t.Flags, Family: t.Family}
sets = append(sets, s)
}
return sets, nil
}
// GetSetByName returns the set in the specified table if matching name is found.
func (cc *Conn) GetSetByName(t *Table, name string) (*Set, error) {
conn, closer, err := cc.netlinkConn()
if err != nil {
return nil, err
}
defer func() { _ = closer() }()
data, err := netlink.MarshalAttributes([]netlink.Attribute{
{Type: unix.NFTA_SET_TABLE, Data: []byte(t.Name + "\x00")},
{Type: unix.NFTA_SET_NAME, Data: []byte(name + "\x00")},
})
if err != nil {
return nil, err
}
message := netlink.Message{
Header: netlink.Header{
Type: netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | unix.NFT_MSG_GETSET),
Flags: netlink.Request | netlink.Acknowledge,
},
Data: append(extraHeader(uint8(t.Family), 0), data...),
}
if _, err := conn.SendMessages([]netlink.Message{message}); err != nil {
return nil, fmt.Errorf("SendMessages: %w", err)
}
reply, err := receiveAckAware(conn, message.Header.Flags)
if err != nil {
2024-08-11 03:00:48 -05:00
return nil, fmt.Errorf("receiveAckAware: %w", err)
}
if len(reply) != 1 {
2024-08-11 03:00:48 -05:00
return nil, fmt.Errorf("receiveAckAware: expected to receive 1 message but got %d", len(reply))
}
rs, err := setsFromMsg(reply[0])
if err != nil {
return nil, err
}
rs.Table = &Table{Name: t.Name, Use: t.Use, Flags: t.Flags, Family: t.Family}
return rs, nil
}
// GetSetElements returns the elements in the specified set.
func (cc *Conn) GetSetElements(s *Set) ([]SetElement, error) {
conn, closer, err := cc.netlinkConn()
if err != nil {
return nil, err
}
defer func() { _ = closer() }()
data, err := netlink.MarshalAttributes([]netlink.Attribute{
{Type: unix.NFTA_SET_ELEM_LIST_TABLE, Data: []byte(s.Table.Name + "\x00")},
{Type: unix.NFTA_SET_ELEM_LIST_SET, Data: []byte(s.Name + "\x00")},
})
if err != nil {
return nil, err
}
message := netlink.Message{
Header: netlink.Header{
Type: netlink.HeaderType((unix.NFNL_SUBSYS_NFTABLES << 8) | unix.NFT_MSG_GETSETELEM),
Flags: netlink.Request | netlink.Acknowledge | netlink.Dump,
},
Data: append(extraHeader(uint8(s.Table.Family), 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 {
2024-08-11 03:00:48 -05:00
return nil, fmt.Errorf("receiveAckAware: %v", err)
}
var elems []SetElement
for _, msg := range reply {
2023-05-20 11:23:52 -05:00
s, err := elementsFromMsg(uint8(s.Table.Family), msg)
if err != nil {
return nil, err
}
elems = append(elems, s...)
}
return elems, nil
}