-
Notifications
You must be signed in to change notification settings - Fork 5
/
v5_udp.go
919 lines (831 loc) · 26.7 KB
/
v5_udp.go
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
// Copyright 2020 The go-ethereum Authors
// This file is part of the go-ethereum library.
//
// The go-ethereum library is free software: you can redistribute it and/or modify
// it under the terms of the GNU Lesser General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// The go-ethereum library is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU Lesser General Public License for more details.
//
// You should have received a copy of the GNU Lesser General Public License
// along with the go-ethereum library. If not, see <http://www.gnu.org/licenses/>.
package discover
import (
"bytes"
"context"
"crypto/ecdsa"
crand "crypto/rand"
"errors"
"fmt"
"io"
"net"
"sync"
"time"
"github.com/theQRL/go-zond/common/mclock"
"github.com/theQRL/go-zond/log"
"github.com/theQRL/go-zond/p2p/discover/v5wire"
"github.com/theQRL/go-zond/p2p/enode"
"github.com/theQRL/go-zond/p2p/enr"
"github.com/theQRL/go-zond/p2p/netutil"
)
const (
lookupRequestLimit = 3 // max requests against a single node during lookup
findnodeResultLimit = 16 // applies in FINDNODE handler
totalNodesResponseLimit = 5 // applies in waitForNodes
respTimeoutV5 = 700 * time.Millisecond
)
// codecV5 is implemented by v5wire.Codec (and testCodec).
//
// The UDPv5 transport is split into two objects: the codec object deals with
// encoding/decoding and with the handshake; the UDPv5 object handles higher-level concerns.
type codecV5 interface {
// Encode encodes a packet.
Encode(enode.ID, string, v5wire.Packet, *v5wire.Whoareyou) ([]byte, v5wire.Nonce, error)
// Decode decodes a packet. It returns a *v5wire.Unknown packet if decryption fails.
// The *enode.Node return value is non-nil when the input contains a handshake response.
Decode([]byte, string) (enode.ID, *enode.Node, v5wire.Packet, error)
}
// UDPv5 is the implementation of protocol version 5.
type UDPv5 struct {
// static fields
conn UDPConn
tab *Table
netrestrict *netutil.Netlist
priv *ecdsa.PrivateKey
localNode *enode.LocalNode
db *enode.DB
log log.Logger
clock mclock.Clock
validSchemes enr.IdentityScheme
// misc buffers used during message handling
logcontext []interface{}
// talkreq handler registry
talk *talkSystem
// channels into dispatch
packetInCh chan ReadPacket
readNextCh chan struct{}
callCh chan *callV5
callDoneCh chan *callV5
respTimeoutCh chan *callTimeout
sendCh chan sendRequest
unhandled chan<- ReadPacket
// state of dispatch
codec codecV5
activeCallByNode map[enode.ID]*callV5
activeCallByAuth map[v5wire.Nonce]*callV5
callQueue map[enode.ID][]*callV5
// shutdown stuff
closeOnce sync.Once
closeCtx context.Context
cancelCloseCtx context.CancelFunc
wg sync.WaitGroup
}
type sendRequest struct {
destID enode.ID
destAddr *net.UDPAddr
msg v5wire.Packet
}
// callV5 represents a remote procedure call against another node.
type callV5 struct {
id enode.ID
addr *net.UDPAddr
node *enode.Node // This is required to perform handshakes.
packet v5wire.Packet
responseType byte // expected packet type of response
reqid []byte
ch chan v5wire.Packet // responses sent here
err chan error // errors sent here
// Valid for active calls only:
nonce v5wire.Nonce // nonce of request packet
handshakeCount int // # times we attempted handshake for this call
challenge *v5wire.Whoareyou // last sent handshake challenge
timeout mclock.Timer
}
// callTimeout is the response timeout event of a call.
type callTimeout struct {
c *callV5
timer mclock.Timer
}
// ListenV5 listens on the given connection.
func ListenV5(conn UDPConn, ln *enode.LocalNode, cfg Config) (*UDPv5, error) {
t, err := newUDPv5(conn, ln, cfg)
if err != nil {
return nil, err
}
go t.tab.loop()
t.wg.Add(2)
go t.readLoop()
go t.dispatch()
return t, nil
}
// newUDPv5 creates a UDPv5 transport, but doesn't start any goroutines.
func newUDPv5(conn UDPConn, ln *enode.LocalNode, cfg Config) (*UDPv5, error) {
closeCtx, cancelCloseCtx := context.WithCancel(context.Background())
cfg = cfg.withDefaults()
t := &UDPv5{
// static fields
conn: newMeteredConn(conn),
localNode: ln,
db: ln.Database(),
netrestrict: cfg.NetRestrict,
priv: cfg.PrivateKey,
log: cfg.Log,
validSchemes: cfg.ValidSchemes,
clock: cfg.Clock,
// channels into dispatch
packetInCh: make(chan ReadPacket, 1),
readNextCh: make(chan struct{}, 1),
callCh: make(chan *callV5),
callDoneCh: make(chan *callV5),
sendCh: make(chan sendRequest),
respTimeoutCh: make(chan *callTimeout),
unhandled: cfg.Unhandled,
// state of dispatch
codec: v5wire.NewCodec(ln, cfg.PrivateKey, cfg.Clock, cfg.V5ProtocolID),
activeCallByNode: make(map[enode.ID]*callV5),
activeCallByAuth: make(map[v5wire.Nonce]*callV5),
callQueue: make(map[enode.ID][]*callV5),
// shutdown
closeCtx: closeCtx,
cancelCloseCtx: cancelCloseCtx,
}
t.talk = newTalkSystem(t)
tab, err := newMeteredTable(t, t.db, cfg)
if err != nil {
return nil, err
}
t.tab = tab
return t, nil
}
// Self returns the local node record.
func (t *UDPv5) Self() *enode.Node {
return t.localNode.Node()
}
// Close shuts down packet processing.
func (t *UDPv5) Close() {
t.closeOnce.Do(func() {
t.cancelCloseCtx()
t.conn.Close()
t.talk.wait()
t.wg.Wait()
t.tab.close()
})
}
// Ping sends a ping message to the given node.
func (t *UDPv5) Ping(n *enode.Node) error {
_, err := t.ping(n)
return err
}
// Resolve searches for a specific node with the given ID and tries to get the most recent
// version of the node record for it. It returns n if the node could not be resolved.
func (t *UDPv5) Resolve(n *enode.Node) *enode.Node {
if intable := t.tab.getNode(n.ID()); intable != nil && intable.Seq() > n.Seq() {
n = intable
}
// Try asking directly. This works if the node is still responding on the endpoint we have.
if resp, err := t.RequestENR(n); err == nil {
return resp
}
// Otherwise do a network lookup.
result := t.Lookup(n.ID())
for _, rn := range result {
if rn.ID() == n.ID() && rn.Seq() > n.Seq() {
return rn
}
}
return n
}
// AllNodes returns all the nodes stored in the local table.
func (t *UDPv5) AllNodes() []*enode.Node {
t.tab.mutex.Lock()
defer t.tab.mutex.Unlock()
nodes := make([]*enode.Node, 0)
for _, b := range &t.tab.buckets {
for _, n := range b.entries {
nodes = append(nodes, unwrapNode(n))
}
}
return nodes
}
// LocalNode returns the current local node running the
// protocol.
func (t *UDPv5) LocalNode() *enode.LocalNode {
return t.localNode
}
// RegisterTalkHandler adds a handler for 'talk requests'. The handler function is called
// whenever a request for the given protocol is received and should return the response
// data or nil.
func (t *UDPv5) RegisterTalkHandler(protocol string, handler TalkRequestHandler) {
t.talk.register(protocol, handler)
}
// TalkRequest sends a talk request to a node and waits for a response.
func (t *UDPv5) TalkRequest(n *enode.Node, protocol string, request []byte) ([]byte, error) {
req := &v5wire.TalkRequest{Protocol: protocol, Message: request}
resp := t.callToNode(n, v5wire.TalkResponseMsg, req)
defer t.callDone(resp)
select {
case respMsg := <-resp.ch:
return respMsg.(*v5wire.TalkResponse).Message, nil
case err := <-resp.err:
return nil, err
}
}
// TalkRequestToID sends a talk request to a node and waits for a response.
func (t *UDPv5) TalkRequestToID(id enode.ID, addr *net.UDPAddr, protocol string, request []byte) ([]byte, error) {
req := &v5wire.TalkRequest{Protocol: protocol, Message: request}
resp := t.callToID(id, addr, v5wire.TalkResponseMsg, req)
defer t.callDone(resp)
select {
case respMsg := <-resp.ch:
return respMsg.(*v5wire.TalkResponse).Message, nil
case err := <-resp.err:
return nil, err
}
}
// RandomNodes returns an iterator that finds random nodes in the DHT.
func (t *UDPv5) RandomNodes() enode.Iterator {
if t.tab.len() == 0 {
// All nodes were dropped, refresh. The very first query will hit this
// case and run the bootstrapping logic.
<-t.tab.refresh()
}
return newLookupIterator(t.closeCtx, t.newRandomLookup)
}
// Lookup performs a recursive lookup for the given target.
// It returns the closest nodes to target.
func (t *UDPv5) Lookup(target enode.ID) []*enode.Node {
return t.newLookup(t.closeCtx, target).run()
}
// lookupRandom looks up a random target.
// This is needed to satisfy the transport interface.
func (t *UDPv5) lookupRandom() []*enode.Node {
return t.newRandomLookup(t.closeCtx).run()
}
// lookupSelf looks up our own node ID.
// This is needed to satisfy the transport interface.
func (t *UDPv5) lookupSelf() []*enode.Node {
return t.newLookup(t.closeCtx, t.Self().ID()).run()
}
func (t *UDPv5) newRandomLookup(ctx context.Context) *lookup {
var target enode.ID
crand.Read(target[:])
return t.newLookup(ctx, target)
}
func (t *UDPv5) newLookup(ctx context.Context, target enode.ID) *lookup {
return newLookup(ctx, t.tab, target, func(n *node) ([]*node, error) {
return t.lookupWorker(n, target)
})
}
// lookupWorker performs FINDNODE calls against a single node during lookup.
func (t *UDPv5) lookupWorker(destNode *node, target enode.ID) ([]*node, error) {
var (
dists = lookupDistances(target, destNode.ID())
nodes = nodesByDistance{target: target}
err error
)
var r []*enode.Node
r, err = t.findnode(unwrapNode(destNode), dists)
if errors.Is(err, errClosed) {
return nil, err
}
for _, n := range r {
if n.ID() != t.Self().ID() {
nodes.push(wrapNode(n), findnodeResultLimit)
}
}
return nodes.entries, err
}
// lookupDistances computes the distance parameter for FINDNODE calls to dest.
// It chooses distances adjacent to logdist(target, dest), e.g. for a target
// with logdist(target, dest) = 255 the result is [255, 256, 254].
func lookupDistances(target, dest enode.ID) (dists []uint) {
td := enode.LogDist(target, dest)
dists = append(dists, uint(td))
for i := 1; len(dists) < lookupRequestLimit; i++ {
if td+i <= 256 {
dists = append(dists, uint(td+i))
}
if td-i > 0 {
dists = append(dists, uint(td-i))
}
}
return dists
}
// ping calls PING on a node and waits for a PONG response.
func (t *UDPv5) ping(n *enode.Node) (uint64, error) {
req := &v5wire.Ping{ENRSeq: t.localNode.Node().Seq()}
resp := t.callToNode(n, v5wire.PongMsg, req)
defer t.callDone(resp)
select {
case pong := <-resp.ch:
return pong.(*v5wire.Pong).ENRSeq, nil
case err := <-resp.err:
return 0, err
}
}
// RequestENR requests n's record.
func (t *UDPv5) RequestENR(n *enode.Node) (*enode.Node, error) {
nodes, err := t.findnode(n, []uint{0})
if err != nil {
return nil, err
}
if len(nodes) != 1 {
return nil, fmt.Errorf("%d nodes in response for distance zero", len(nodes))
}
return nodes[0], nil
}
// findnode calls FINDNODE on a node and waits for responses.
func (t *UDPv5) findnode(n *enode.Node, distances []uint) ([]*enode.Node, error) {
resp := t.callToNode(n, v5wire.NodesMsg, &v5wire.Findnode{Distances: distances})
return t.waitForNodes(resp, distances)
}
// waitForNodes waits for NODES responses to the given call.
func (t *UDPv5) waitForNodes(c *callV5, distances []uint) ([]*enode.Node, error) {
defer t.callDone(c)
var (
nodes []*enode.Node
seen = make(map[enode.ID]struct{})
received, total = 0, -1
)
for {
select {
case responseP := <-c.ch:
response := responseP.(*v5wire.Nodes)
for _, record := range response.Nodes {
node, err := t.verifyResponseNode(c, record, distances, seen)
if err != nil {
t.log.Debug("Invalid record in "+response.Name(), "id", c.node.ID(), "err", err)
continue
}
nodes = append(nodes, node)
}
if total == -1 {
total = min(int(response.RespCount), totalNodesResponseLimit)
}
if received++; received == total {
return nodes, nil
}
case err := <-c.err:
return nodes, err
}
}
}
// verifyResponseNode checks validity of a record in a NODES response.
func (t *UDPv5) verifyResponseNode(c *callV5, r *enr.Record, distances []uint, seen map[enode.ID]struct{}) (*enode.Node, error) {
node, err := enode.New(t.validSchemes, r)
if err != nil {
return nil, err
}
if err := netutil.CheckRelayIP(c.addr.IP, node.IP()); err != nil {
return nil, err
}
if t.netrestrict != nil && !t.netrestrict.Contains(node.IP()) {
return nil, errors.New("not contained in netrestrict list")
}
if node.UDP() <= 1024 {
return nil, errLowPort
}
if distances != nil {
nd := enode.LogDist(c.id, node.ID())
if !containsUint(uint(nd), distances) {
return nil, errors.New("does not match any requested distance")
}
}
if _, ok := seen[node.ID()]; ok {
return nil, fmt.Errorf("duplicate record")
}
seen[node.ID()] = struct{}{}
return node, nil
}
func containsUint(x uint, xs []uint) bool {
for _, v := range xs {
if x == v {
return true
}
}
return false
}
// callToNode sends the given call and sets up a handler for response packets (of message
// type responseType). Responses are dispatched to the call's response channel.
func (t *UDPv5) callToNode(n *enode.Node, responseType byte, req v5wire.Packet) *callV5 {
addr := &net.UDPAddr{IP: n.IP(), Port: n.UDP()}
c := &callV5{id: n.ID(), addr: addr, node: n}
t.initCall(c, responseType, req)
return c
}
// callToID is like callToNode, but for cases where the node record is not available.
func (t *UDPv5) callToID(id enode.ID, addr *net.UDPAddr, responseType byte, req v5wire.Packet) *callV5 {
c := &callV5{id: id, addr: addr}
t.initCall(c, responseType, req)
return c
}
func (t *UDPv5) initCall(c *callV5, responseType byte, packet v5wire.Packet) {
c.packet = packet
c.responseType = responseType
c.reqid = make([]byte, 8)
c.ch = make(chan v5wire.Packet, 1)
c.err = make(chan error, 1)
// Assign request ID.
crand.Read(c.reqid)
packet.SetRequestID(c.reqid)
// Send call to dispatch.
select {
case t.callCh <- c:
case <-t.closeCtx.Done():
c.err <- errClosed
}
}
// callDone tells dispatch that the active call is done.
func (t *UDPv5) callDone(c *callV5) {
// This needs a loop because further responses may be incoming until the
// send to callDoneCh has completed. Such responses need to be discarded
// in order to avoid blocking the dispatch loop.
for {
select {
case <-c.ch:
// late response, discard.
case <-c.err:
// late error, discard.
case t.callDoneCh <- c:
return
case <-t.closeCtx.Done():
return
}
}
}
// dispatch runs in its own goroutine, handles incoming packets and deals with calls.
//
// For any destination node there is at most one 'active call', stored in the t.activeCall*
// maps. A call is made active when it is sent. The active call can be answered by a
// matching response, in which case c.ch receives the response; or by timing out, in which case
// c.err receives the error. When the function that created the call signals the active
// call is done through callDone, the next call from the call queue is started.
//
// Calls may also be answered by a WHOAREYOU packet referencing the call packet's authTag.
// When that happens the call is simply re-sent to complete the handshake. We allow one
// handshake attempt per call.
func (t *UDPv5) dispatch() {
defer t.wg.Done()
// Arm first read.
t.readNextCh <- struct{}{}
for {
select {
case c := <-t.callCh:
t.callQueue[c.id] = append(t.callQueue[c.id], c)
t.sendNextCall(c.id)
case ct := <-t.respTimeoutCh:
active := t.activeCallByNode[ct.c.id]
if ct.c == active && ct.timer == active.timeout {
ct.c.err <- errTimeout
}
case c := <-t.callDoneCh:
active := t.activeCallByNode[c.id]
if active != c {
panic("BUG: callDone for inactive call")
}
c.timeout.Stop()
delete(t.activeCallByAuth, c.nonce)
delete(t.activeCallByNode, c.id)
t.sendNextCall(c.id)
case r := <-t.sendCh:
t.send(r.destID, r.destAddr, r.msg, nil)
case p := <-t.packetInCh:
t.handlePacket(p.Data, p.Addr)
// Arm next read.
t.readNextCh <- struct{}{}
case <-t.closeCtx.Done():
close(t.readNextCh)
for id, queue := range t.callQueue {
for _, c := range queue {
c.err <- errClosed
}
delete(t.callQueue, id)
}
for id, c := range t.activeCallByNode {
c.err <- errClosed
delete(t.activeCallByNode, id)
delete(t.activeCallByAuth, c.nonce)
}
return
}
}
}
// startResponseTimeout sets the response timer for a call.
func (t *UDPv5) startResponseTimeout(c *callV5) {
if c.timeout != nil {
c.timeout.Stop()
}
var (
timer mclock.Timer
done = make(chan struct{})
)
timer = t.clock.AfterFunc(respTimeoutV5, func() {
<-done
select {
case t.respTimeoutCh <- &callTimeout{c, timer}:
case <-t.closeCtx.Done():
}
})
c.timeout = timer
close(done)
}
// sendNextCall sends the next call in the call queue if there is no active call.
func (t *UDPv5) sendNextCall(id enode.ID) {
queue := t.callQueue[id]
if len(queue) == 0 || t.activeCallByNode[id] != nil {
return
}
t.activeCallByNode[id] = queue[0]
t.sendCall(t.activeCallByNode[id])
if len(queue) == 1 {
delete(t.callQueue, id)
} else {
copy(queue, queue[1:])
t.callQueue[id] = queue[:len(queue)-1]
}
}
// sendCall encodes and sends a request packet to the call's recipient node.
// This performs a handshake if needed.
func (t *UDPv5) sendCall(c *callV5) {
// The call might have a nonce from a previous handshake attempt. Remove the entry for
// the old nonce because we're about to generate a new nonce for this call.
if c.nonce != (v5wire.Nonce{}) {
delete(t.activeCallByAuth, c.nonce)
}
newNonce, _ := t.send(c.id, c.addr, c.packet, c.challenge)
c.nonce = newNonce
t.activeCallByAuth[newNonce] = c
t.startResponseTimeout(c)
}
// sendResponse sends a response packet to the given node.
// This doesn't trigger a handshake even if no keys are available.
func (t *UDPv5) sendResponse(toID enode.ID, toAddr *net.UDPAddr, packet v5wire.Packet) error {
_, err := t.send(toID, toAddr, packet, nil)
return err
}
func (t *UDPv5) sendFromAnotherThread(toID enode.ID, toAddr *net.UDPAddr, packet v5wire.Packet) {
select {
case t.sendCh <- sendRequest{toID, toAddr, packet}:
case <-t.closeCtx.Done():
}
}
// send sends a packet to the given node.
func (t *UDPv5) send(toID enode.ID, toAddr *net.UDPAddr, packet v5wire.Packet, c *v5wire.Whoareyou) (v5wire.Nonce, error) {
addr := toAddr.String()
t.logcontext = append(t.logcontext[:0], "id", toID, "addr", addr)
t.logcontext = packet.AppendLogInfo(t.logcontext)
enc, nonce, err := t.codec.Encode(toID, addr, packet, c)
if err != nil {
t.logcontext = append(t.logcontext, "err", err)
t.log.Warn(">> "+packet.Name(), t.logcontext...)
return nonce, err
}
_, err = t.conn.WriteToUDP(enc, toAddr)
t.log.Trace(">> "+packet.Name(), t.logcontext...)
return nonce, err
}
// readLoop runs in its own goroutine and reads packets from the network.
func (t *UDPv5) readLoop() {
defer t.wg.Done()
buf := make([]byte, maxPacketSize)
for range t.readNextCh {
nbytes, from, err := t.conn.ReadFromUDP(buf)
if netutil.IsTemporaryError(err) {
// Ignore temporary read errors.
t.log.Debug("Temporary UDP read error", "err", err)
continue
} else if err != nil {
// Shut down the loop for permanent errors.
if !errors.Is(err, io.EOF) {
t.log.Debug("UDP read error", "err", err)
}
return
}
t.dispatchReadPacket(from, buf[:nbytes])
}
}
// dispatchReadPacket sends a packet into the dispatch loop.
func (t *UDPv5) dispatchReadPacket(from *net.UDPAddr, content []byte) bool {
select {
case t.packetInCh <- ReadPacket{content, from}:
return true
case <-t.closeCtx.Done():
return false
}
}
// handlePacket decodes and processes an incoming packet from the network.
func (t *UDPv5) handlePacket(rawpacket []byte, fromAddr *net.UDPAddr) error {
addr := fromAddr.String()
fromID, fromNode, packet, err := t.codec.Decode(rawpacket, addr)
if err != nil {
if t.unhandled != nil && v5wire.IsInvalidHeader(err) {
// The packet seems unrelated to discv5, send it to the next protocol.
// t.log.Trace("Unhandled discv5 packet", "id", fromID, "addr", addr, "err", err)
up := ReadPacket{Data: make([]byte, len(rawpacket)), Addr: fromAddr}
copy(up.Data, rawpacket)
t.unhandled <- up
return nil
}
t.log.Debug("Bad discv5 packet", "id", fromID, "addr", addr, "err", err)
return err
}
if fromNode != nil {
// Handshake succeeded, add to table.
t.tab.addSeenNode(wrapNode(fromNode))
}
if packet.Kind() != v5wire.WhoareyouPacket {
// WHOAREYOU logged separately to report errors.
t.logcontext = append(t.logcontext[:0], "id", fromID, "addr", addr)
t.logcontext = packet.AppendLogInfo(t.logcontext)
t.log.Trace("<< "+packet.Name(), t.logcontext...)
}
t.handle(packet, fromID, fromAddr)
return nil
}
// handleCallResponse dispatches a response packet to the call waiting for it.
func (t *UDPv5) handleCallResponse(fromID enode.ID, fromAddr *net.UDPAddr, p v5wire.Packet) bool {
ac := t.activeCallByNode[fromID]
if ac == nil || !bytes.Equal(p.RequestID(), ac.reqid) {
t.log.Debug(fmt.Sprintf("Unsolicited/late %s response", p.Name()), "id", fromID, "addr", fromAddr)
return false
}
if !fromAddr.IP.Equal(ac.addr.IP) || fromAddr.Port != ac.addr.Port {
t.log.Debug(fmt.Sprintf("%s from wrong endpoint", p.Name()), "id", fromID, "addr", fromAddr)
return false
}
if p.Kind() != ac.responseType {
t.log.Debug(fmt.Sprintf("Wrong discv5 response type %s", p.Name()), "id", fromID, "addr", fromAddr)
return false
}
t.startResponseTimeout(ac)
ac.ch <- p
return true
}
// getNode looks for a node record in table and database.
func (t *UDPv5) getNode(id enode.ID) *enode.Node {
if n := t.tab.getNode(id); n != nil {
return n
}
if n := t.localNode.Database().Node(id); n != nil {
return n
}
return nil
}
// handle processes incoming packets according to their message type.
func (t *UDPv5) handle(p v5wire.Packet, fromID enode.ID, fromAddr *net.UDPAddr) {
switch p := p.(type) {
case *v5wire.Unknown:
t.handleUnknown(p, fromID, fromAddr)
case *v5wire.Whoareyou:
t.handleWhoareyou(p, fromID, fromAddr)
case *v5wire.Ping:
t.handlePing(p, fromID, fromAddr)
case *v5wire.Pong:
if t.handleCallResponse(fromID, fromAddr, p) {
t.localNode.UDPEndpointStatement(fromAddr, &net.UDPAddr{IP: p.ToIP, Port: int(p.ToPort)})
}
case *v5wire.Findnode:
t.handleFindnode(p, fromID, fromAddr)
case *v5wire.Nodes:
t.handleCallResponse(fromID, fromAddr, p)
case *v5wire.TalkRequest:
t.talk.handleRequest(fromID, fromAddr, p)
case *v5wire.TalkResponse:
t.handleCallResponse(fromID, fromAddr, p)
}
}
// handleUnknown initiates a handshake by responding with WHOAREYOU.
func (t *UDPv5) handleUnknown(p *v5wire.Unknown, fromID enode.ID, fromAddr *net.UDPAddr) {
challenge := &v5wire.Whoareyou{Nonce: p.Nonce}
crand.Read(challenge.IDNonce[:])
if n := t.getNode(fromID); n != nil {
challenge.Node = n
challenge.RecordSeq = n.Seq()
}
t.sendResponse(fromID, fromAddr, challenge)
}
var (
errChallengeNoCall = errors.New("no matching call")
errChallengeTwice = errors.New("second handshake")
)
// handleWhoareyou resends the active call as a handshake packet.
func (t *UDPv5) handleWhoareyou(p *v5wire.Whoareyou, fromID enode.ID, fromAddr *net.UDPAddr) {
c, err := t.matchWithCall(fromID, p.Nonce)
if err != nil {
t.log.Debug("Invalid "+p.Name(), "addr", fromAddr, "err", err)
return
}
if c.node == nil {
// Can't perform handshake because we don't have the ENR.
t.log.Debug("Can't handle "+p.Name(), "addr", fromAddr, "err", "call has no ENR")
c.err <- errors.New("remote wants handshake, but call has no ENR")
return
}
// Resend the call that was answered by WHOAREYOU.
t.log.Trace("<< "+p.Name(), "id", c.node.ID(), "addr", fromAddr)
c.handshakeCount++
c.challenge = p
p.Node = c.node
t.sendCall(c)
}
// matchWithCall checks whether a handshake attempt matches the active call.
func (t *UDPv5) matchWithCall(fromID enode.ID, nonce v5wire.Nonce) (*callV5, error) {
c := t.activeCallByAuth[nonce]
if c == nil {
return nil, errChallengeNoCall
}
if c.handshakeCount > 0 {
return nil, errChallengeTwice
}
return c, nil
}
// handlePing sends a PONG response.
func (t *UDPv5) handlePing(p *v5wire.Ping, fromID enode.ID, fromAddr *net.UDPAddr) {
remoteIP := fromAddr.IP
// Handle IPv4 mapped IPv6 addresses in the
// event the local node is binded to an
// ipv6 interface.
if remoteIP.To4() != nil {
remoteIP = remoteIP.To4()
}
t.sendResponse(fromID, fromAddr, &v5wire.Pong{
ReqID: p.ReqID,
ToIP: remoteIP,
ToPort: uint16(fromAddr.Port),
ENRSeq: t.localNode.Node().Seq(),
})
}
// handleFindnode returns nodes to the requester.
func (t *UDPv5) handleFindnode(p *v5wire.Findnode, fromID enode.ID, fromAddr *net.UDPAddr) {
nodes := t.collectTableNodes(fromAddr.IP, p.Distances, findnodeResultLimit)
for _, resp := range packNodes(p.ReqID, nodes) {
t.sendResponse(fromID, fromAddr, resp)
}
}
// collectTableNodes creates a FINDNODE result set for the given distances.
func (t *UDPv5) collectTableNodes(rip net.IP, distances []uint, limit int) []*enode.Node {
var nodes []*enode.Node
var processed = make(map[uint]struct{})
for _, dist := range distances {
// Reject duplicate / invalid distances.
_, seen := processed[dist]
if seen || dist > 256 {
continue
}
// Get the nodes.
var bn []*enode.Node
if dist == 0 {
bn = []*enode.Node{t.Self()}
} else if dist <= 256 {
t.tab.mutex.Lock()
bn = unwrapNodes(t.tab.bucketAtDistance(int(dist)).entries)
t.tab.mutex.Unlock()
}
processed[dist] = struct{}{}
// Apply some pre-checks to avoid sending invalid nodes.
for _, n := range bn {
// TODO livenessChecks > 1
if netutil.CheckRelayIP(rip, n.IP()) != nil {
continue
}
nodes = append(nodes, n)
if len(nodes) >= limit {
return nodes
}
}
}
return nodes
}
// packNodes creates NODES response packets for the given node list.
func packNodes(reqid []byte, nodes []*enode.Node) []*v5wire.Nodes {
if len(nodes) == 0 {
return []*v5wire.Nodes{{ReqID: reqid, RespCount: 1}}
}
// This limit represents the available space for nodes in output packets. Maximum
// packet size is 1280, and out of this ~80 bytes will be taken up by the packet
// frame. So limiting to 1000 bytes here leaves 200 bytes for other fields of the
// NODES message, which is a lot.
const sizeLimit = 1000
var resp []*v5wire.Nodes
for len(nodes) > 0 {
p := &v5wire.Nodes{ReqID: reqid}
size := uint64(0)
for len(nodes) > 0 {
r := nodes[0].Record()
if size += r.Size(); size > sizeLimit {
break
}
p.Nodes = append(p.Nodes, r)
nodes = nodes[1:]
}
resp = append(resp, p)
}
for _, msg := range resp {
msg.RespCount = uint8(len(resp))
}
return resp
}