3 * most of the swift's state machine
5 * Created by Victor Grishchenko on 3/6/09.
6 * Copyright 2009-2012 TECHNISCHE UNIVERSITEIT DELFT. All rights reserved.
11 #include <algorithm> // kill it
17 using namespace swift;
20 struct event_base *Channel::evbase;
21 MessageQueue Channel::messageQueue;
22 struct event Channel::evrecv;
24 #define DEBUGTRAFFIC 0
26 /** Arno: Victor's design allows a sender to choose some data to push to
27 * a receiver, if that receiver is not HINTing at data. Should be disabled
28 * when the receiver has a download rate limit.
30 #define ENABLE_SENDERSIZE_PUSH 0
33 /** Arno, 2011-11-24: When rate limit is on and the download is in progress
34 * we send HINTs for 2 chunks at the moment. This constant can be used to
35 * get greater granularity. Set to 0 for original behaviour.
37 #define HINT_GRANULARITY 16 // chunks
42 - randomized testing of advanced ops (new testcase)
45 void Channel::AddPeakHashes (struct evbuffer *evb) {
46 for(int i=0; i<file().peak_count(); i++) {
47 bin_t peak = file().peak(i);
48 evbuffer_add_8(evb, SWIFT_HASH);
49 evbuffer_add_32be(evb, bin_toUInt32(peak));
50 evbuffer_add_hash(evb, file().peak_hash(i));
51 char bin_name_buf[32];
52 dprintf("%s #%u +phash %s\n",tintstr(),id_,peak.str(bin_name_buf));
57 void Channel::AddUncleHashes (struct evbuffer *evb, bin_t pos) {
59 char bin_name_buf2[32];
60 dprintf("%s #%u +uncle hash for %s\n",tintstr(),id_,pos.str(bin_name_buf2));
62 bin_t peak = file().peak_for(pos);
63 while (pos!=peak && ((NOW&3)==3 || !pos.parent().contains(data_out_cap_)) &&
64 ack_in_.is_empty(pos.parent()) ) {
65 bin_t uncle = pos.sibling();
66 evbuffer_add_8(evb, SWIFT_HASH);
67 evbuffer_add_32be(evb, bin_toUInt32(uncle));
68 evbuffer_add_hash(evb, file().hash(uncle) );
69 char bin_name_buf[32];
70 dprintf("%s #%u +hash %s\n",tintstr(),id_,uncle.str(bin_name_buf));
76 bin_t Channel::ImposeHint () {
77 uint64_t twist = peer_channel_id_; // got no hints, send something randomly
79 twist &= file().peak(0).toUInt(); // FIXME may make it semi-seq here
81 bin_t my_pick = binmap_t::find_complement(ack_in_, file().ack_out(), twist);
83 my_pick.to_twisted(twist);
84 while (my_pick.base_length()>max(1,(int)cwnd_))
85 my_pick = my_pick.left();
87 return my_pick.twisted(twist);
91 bin_t Channel::DequeueHint (bool *retransmitptr) {
92 bin_t send = bin_t::NONE;
94 // Arno, 2012-01-23: Extra protection against channel loss, don't send DATA
95 if (last_recv_time_ < NOW-(3*TINT_SEC))
98 // Arno, 2012-01-18: Reenable Victor's retransmit
99 if (!data_out_tmo_.empty())
101 tintbin tb = data_out_tmo_.front();
103 data_out_tmo_.pop_front();
104 *retransmitptr = true;
107 *retransmitptr = false;
109 if (ENABLE_SENDERSIZE_PUSH && send.is_none() && hint_in_.empty() && last_recv_time_>NOW-rtt_avg_-TINT_SEC) {
110 bin_t my_pick = ImposeHint(); // FIXME move to the loop
111 if (!my_pick.is_none()) {
112 hint_in_.push_back(my_pick);
113 char bin_name_buf[32];
114 dprintf("%s #%u *hint %s\n",tintstr(),id_,my_pick.str(bin_name_buf));
118 while (!hint_in_.empty() && send.is_none()) {
119 bin_t hint = hint_in_.front().bin;
120 tint time = hint_in_.front().time;
121 hint_in_.pop_front();
122 while (!hint.is_base()) { // FIXME optimize; possible attack
123 hint_in_.push_front(tintbin(time,hint.right()));
126 //if (time < NOW-TINT_SEC*3/2 )
127 // continue; bad idea
128 if (!ack_in_.is_filled(hint))
132 // Arno, 2012-03-09: Is mucho expensive on busy server.
133 //for(int i=0; i<hint_in_.size(); i++)
134 // mass += hint_in_[i].bin.base_length();
135 char bin_name_buf[32];
136 dprintf("%s #%u dequeued %s [%lli]\n",tintstr(),id_,send.str(bin_name_buf),mass);
141 void Channel::AddHandshake (struct evbuffer *evb) {
142 if (!peer_channel_id_) { // initiating
143 evbuffer_add_8(evb, SWIFT_HASH);
144 evbuffer_add_32be(evb, bin_toUInt32(bin_t::ALL));
145 evbuffer_add_hash(evb, file().root_hash());
146 dprintf("%s #%u +hash ALL %s\n",
147 tintstr(),id_,file().root_hash().hex().c_str());
149 evbuffer_add_8(evb, SWIFT_HANDSHAKE);
151 if (send_control_==CLOSE_CONTROL) {
155 encoded = EncodeID(id_);
156 evbuffer_add_32be(evb, encoded);
157 dprintf("%s #%u +hs %x\n",tintstr(),id_,encoded);
162 void Channel::Send () {
164 dprintf("%s #%u Send called \n",tintstr(),id_);
166 struct evbuffer *evb = evbuffer_new();
167 evbuffer_add_32be(evb, peer_channel_id_);
168 bin_t data = bin_t::NONE;
169 int evbnonadplen = 0;
170 if ( is_established() ) {
171 if (send_control_!=CLOSE_CONTROL) {
172 // FIXME: seeder check
175 if (!file().is_complete()) {
177 /* Gertjan fix: 7aeea65f3efbb9013f601b22a57ee4a423f1a94d
178 "Only call Reschedule for 'reverse PEX' if the channel is in keep-alive mode"
186 // Arno: send explicit close
191 AddHave(evb); // Arno, 2011-10-28: from AddHandShake. Why double?
196 lastsendwaskeepalive_ = (evbuffer_get_length(evb) == 4);
198 if (evbuffer_get_length(evb)==4) {// only the channel id; bare keep-alive
201 dprintf("%s #%u sent %ib %s:%x\n",
202 tintstr(),id_,(int)evbuffer_get_length(evb),peer().str(),
205 messageQueue.AddBuffer(socket_, evb, peer(), this);
208 void Channel::Sent(int bytes, evbuffer *evb, bool tofree)
210 raw_bytes_up_ += bytes;
212 last_send_time_ = NOW;
220 void Channel::AddHint (struct evbuffer *evb) {
223 // Policy is to not send hints when we are above speed limit
224 if (transfer().GetCurrentSpeed(DDIR_DOWNLOAD) > transfer().GetMaxSpeed(DDIR_DOWNLOAD)) {
226 fprintf(stderr,"hint: forbidden#");
231 // 1. Calc max of what we are allowed to request, uncongested bandwidth wise
232 tint plan_for = max(TINT_SEC,rtt_avg_*4);
234 tint timed_out = NOW - plan_for*2;
235 while ( !hint_out_.empty() && hint_out_.front().time < timed_out ) {
236 hint_out_size_ -= hint_out_.front().bin.base_length();
237 hint_out_.pop_front();
240 int first_plan_pck = max ( (tint)1, plan_for / dip_avg_ );
242 // Riccardo, 2012-04-04: Actually allowed is max minus what we already asked for
243 int queue_allowed_hints = max(0,first_plan_pck-(int)hint_out_size_);
247 // 2. Calc max of what is allowed by the rate limiter
248 int rate_allowed_hints = LONG_MAX;
249 if (transfer().GetMaxSpeed(DDIR_DOWNLOAD) < DBL_MAX)
251 uint64_t rough_global_hint_out_size = 0; // rough estimate, as hint_out_ clean up is not done for all channels
252 std::set<Channel *>::iterator iter;
253 for (iter=transfer().mychannels_.begin(); iter!=transfer().mychannels_.end(); iter++)
257 rough_global_hint_out_size += c->hint_out_size_;
260 // Policy: this channel is allowed to hint at the limit - global_hinted_at
261 // Handle MaxSpeed = unlimited
262 double rate_hints_limit_float = transfer().GetMaxSpeed(DDIR_DOWNLOAD)/((double)file().chunk_size());
264 int rate_hints_limit = (int)min((double)LONG_MAX,rate_hints_limit_float);
266 // Actually allowed is max minus what we already asked for, globally (=all channels)
267 rate_allowed_hints = max(0,rate_hints_limit-(int)rough_global_hint_out_size);
270 // 3. Take the smallest allowance from rate and queue limit
271 uint64_t plan_pck = (uint64_t)min(rate_allowed_hints,queue_allowed_hints);
273 // 4. Ask allowance in blocks of chunks to get pipelining going from serving peer.
274 if (hint_out_size_ == 0 || plan_pck > HINT_GRANULARITY)
276 bin_t hint = transfer().picker().Pick(ack_in_,plan_pck,NOW+plan_for*2);
277 if (!hint.is_none()) {
281 fprintf(stderr,"hint c%d: ask %s\n", id(), hint.str(binstr) );
283 evbuffer_add_8(evb, SWIFT_HINT);
284 evbuffer_add_32be(evb, bin_toUInt32(hint));
285 char bin_name_buf[32];
286 dprintf("%s #%u +hint %s [%lli]\n",tintstr(),id_,hint.str(bin_name_buf),hint_out_size_);
287 hint_out_.push_back(hint);
288 hint_out_size_ += hint.base_length();
289 //fprintf(stderr,"send c%d: HINTLEN %i\n", id(), hint.base_length());
290 //fprintf(stderr,"HL %i ", hint.base_length());
293 dprintf("%s #%u Xhint\n",tintstr(),id_);
299 bin_t Channel::AddData (struct evbuffer *evb) {
301 if (transfer().GetCurrentSpeed(DDIR_UPLOAD) > transfer().GetMaxSpeed(DDIR_UPLOAD)) {
302 transfer().OnSendNoData();
306 if (!file().size()) // know nothing
309 bin_t tosend = bin_t::NONE;
310 bool isretransmit = false;
311 tint luft = send_interval_>>4; // may wake up a bit earlier
312 if (data_out_.size()<cwnd_ &&
313 last_data_out_time_+send_interval_<=NOW+luft) {
314 tosend = DequeueHint(&isretransmit);
315 if (tosend.is_none()) {
316 dprintf("%s #%u sendctrl no idea what data to send\n",tintstr(),id_);
317 if (send_control_!=KEEP_ALIVE_CONTROL && send_control_!=CLOSE_CONTROL)
318 SwitchSendControl(KEEP_ALIVE_CONTROL);
321 dprintf("%s #%u sendctrl wait cwnd %f data_out %i next %s\n",
322 tintstr(),id_,cwnd_,(int)data_out_.size(),tintstr(last_data_out_time_+NOW-send_interval_));
324 if (tosend.is_none())// && (last_data_out_time_>NOW-TINT_SEC || data_out_.empty()))
325 return bin_t::NONE; // once in a while, empty data is sent just to check rtt FIXED
327 if (ack_in_.is_empty() && file().size())
331 if (file().get_check_netwvshash())
332 AddUncleHashes(evb,tosend);
334 if (!ack_in_.is_empty()) // TODO: cwnd_>1
335 data_out_cap_ = tosend;
337 // Arno, 2011-11-03: May happen when first data packet is sent to empty
338 // leech, then peak + uncle hashes may be so big that they don't fit in eth
339 // frame with DATA. Send 2 datagrams then, one with peaks so they have
340 // a better chance of arriving. Optimistic violation of atomic datagram
342 if (file().chunk_size() == SWIFT_DEFAULT_CHUNK_SIZE && evbuffer_get_length(evb) > SWIFT_MAX_NONDATA_DGRAM_SIZE) {
343 dprintf("%s #%u fsent %ib %s:%x\n",
344 tintstr(),id_,(int)evbuffer_get_length(evb),peer().str(),
346 messageQueue.AddBuffer(socket_, evb, peer(), this, false);
347 evbuffer_add_32be(evb, peer_channel_id_);
350 if (file().chunk_size() != SWIFT_DEFAULT_CHUNK_SIZE && isretransmit) {
352 * Arno, 2012-01-17: We observe strange behaviour when using
353 * fragmented UDP packets. When ULANC sends a specific datagram ("995"),
354 * the 2nd IP packet carrying it gets lost structurally. When
355 * downloading from the same asset hosted on a Linux 32-bit machine
356 * using a Win7 32-bit client (behind a NAT), one specific full
357 * datagram never gets delivered (6970 one before do). A workaround
358 * is to add some random data to the datagram. Hence we introduce
359 * the SWIFT_RANDOMIZE message, that is added to the datagram carrying
360 * the DATA on a retransmit.
363 fprintf(stderr,"AddData: retransmit of randomized chunk %s\n",tosend.str(binstr) );
364 evbuffer_add_8(evb, SWIFT_RANDOMIZE);
365 evbuffer_add_32be(evb, (int)rand() );
368 evbuffer_add_8(evb, SWIFT_DATA);
369 evbuffer_add_32be(evb, bin_toUInt32(tosend));
371 struct evbuffer_iovec vec;
372 if (evbuffer_reserve_space(evb, file().chunk_size(), &vec, 1) < 0) {
373 print_error("error on evbuffer_reserve_space");
376 size_t r = pread(file().file_descriptor(),(char *)vec.iov_base,
377 file().chunk_size(),tosend.base_offset()*file().chunk_size());
378 // TODO: corrupted data, retries, caching
380 print_error("error on reading");
382 evbuffer_commit_space(evb, &vec, 1);
385 // assert(dgram.space()>=r+4+1);
387 if (evbuffer_commit_space(evb, &vec, 1) < 0) {
388 print_error("error on evbuffer_commit_space");
392 last_data_out_time_ = NOW;
393 data_out_.push_back(tosend);
395 global_bytes_up += r;
397 char bin_name_buf[32];
398 dprintf("%s #%u +data %s\n",tintstr(),id_,tosend.str(bin_name_buf));
401 // ARNOSMPTODO: count overhead bytes too? Move to Send() then.
402 transfer_->OnSendData(file().chunk_size());
408 void Channel::AddAck (struct evbuffer *evb) {
409 if (data_in_==tintbin())
410 //if (data_in_.bin==bin64_t::NONE)
412 // sometimes, we send a HAVE (e.g. in case the peer did repetitive send)
413 evbuffer_add_8(evb, data_in_.time==TINT_NEVER?SWIFT_HAVE:SWIFT_ACK);
414 evbuffer_add_32be(evb, bin_toUInt32(data_in_.bin));
415 if (data_in_.time!=TINT_NEVER)
416 evbuffer_add_64be(evb, data_in_.time);
420 fprintf(stderr,"send c%d: ACK %i\n", id(), bin_toUInt32(data_in_.bin));
422 have_out_.set(data_in_.bin);
423 char bin_name_buf[32];
424 dprintf("%s #%u +ack %s %s\n",
425 tintstr(),id_,data_in_.bin.str(bin_name_buf),tintstr(data_in_.time));
426 if (data_in_.bin.layer()>2)
427 data_in_dbl_ = data_in_.bin;
429 //fprintf(stderr,"data_in_ c%d\n", id() );
430 data_in_ = tintbin();
431 //data_in_ = tintbin(NOW,bin64_t::NONE);
435 void Channel::AddHave (struct evbuffer *evb) {
436 if (!data_in_dbl_.is_none()) { // TODO: do redundancy better
437 evbuffer_add_8(evb, SWIFT_HAVE);
438 evbuffer_add_32be(evb, bin_toUInt32(data_in_dbl_));
439 data_in_dbl_=bin_t::NONE;
442 fprintf(stderr,"send c%d: HAVE ",id() );
443 for(int count=0; count<4; count++) {
444 bin_t ack = binmap_t::find_complement(have_out_, file().ack_out(), 0); // FIXME: do rotating queue
447 ack = file().ack_out().cover(ack);
449 evbuffer_add_8(evb, SWIFT_HAVE);
450 evbuffer_add_32be(evb, bin_toUInt32(ack));
453 fprintf(stderr," %i", bin_toUInt32(ack));
455 char bin_name_buf[32];
456 dprintf("%s #%u +have %s\n",tintstr(),id_,ack.str(bin_name_buf));
459 fprintf(stderr,"\n");
464 void Channel::Recv (struct evbuffer *evb) {
465 dprintf("%s #%u recvd %ib\n",tintstr(),id_,(int)evbuffer_get_length(evb)+4);
468 lastrecvwaskeepalive_ = (evbuffer_get_length(evb) == 0);
469 if (lastrecvwaskeepalive_)
470 // Update speed measurements such that they decrease when DL stops
471 transfer().OnRecvData(0);
473 if (last_send_time_ && rtt_avg_==TINT_SEC && dev_avg_==0) {
474 rtt_avg_ = NOW - last_send_time_;
477 dprintf("%s #%u sendctrl rtt init %lli\n",tintstr(),id_,rtt_avg_);
480 bin_t data = evbuffer_get_length(evb) ? bin_t::NONE : bin_t::ALL;
483 fprintf(stderr,"recv c%d: size %d ", id(), evbuffer_get_length(evb));
485 while (evbuffer_get_length(evb)) {
486 uint8_t type = evbuffer_remove_8(evb);
489 fprintf(stderr," %d", type);
492 case SWIFT_HANDSHAKE: OnHandshake(evb); break;
493 case SWIFT_DATA: data=OnData(evb); break;
494 case SWIFT_HAVE: OnHave(evb); break;
495 case SWIFT_ACK: OnAck(evb); break;
496 case SWIFT_HASH: OnHash(evb); break;
497 case SWIFT_HINT: OnHint(evb); break;
498 case SWIFT_PEX_ADD: OnPex(evb); break;
499 case SWIFT_PEX_REQ: OnPexReq(); break;
500 case SWIFT_RANDOMIZE: OnRandomize(evb); break; //FRAGRAND
502 dprintf("%s #%u ?msg id unknown %i\n",tintstr(),id_,(int)type);
508 fprintf(stderr,"\n");
512 last_recv_time_ = NOW;
513 sent_since_recv_ = 0;
518 * Arno: FAXME: HASH+DATA should be handled as a transaction: only when the
519 * hashes check out should they be stored in the hashtree, otherwise revert.
521 void Channel::OnHash (struct evbuffer *evb) {
522 bin_t pos = bin_fromUInt32(evbuffer_remove_32be(evb));
523 Sha1Hash hash = evbuffer_remove_hash(evb);
524 file().OfferHash(pos,hash);
525 char bin_name_buf[32];
526 dprintf("%s #%u -hash %s\n",tintstr(),id_,pos.str(bin_name_buf));
528 //fprintf(stderr,"HASH %lli hex %s\n",pos.toUInt(), hash.hex().c_str() );
532 void Channel::CleanHintOut (bin_t pos) {
534 while (hi<hint_out_.size() && !hint_out_[hi].bin.contains(pos))
536 if (hi==hint_out_.size())
537 return; // something not hinted or hinted in far past
538 while (hi--) { // removing likely snubbed hints
539 hint_out_size_ -= hint_out_.front().bin.base_length();
540 hint_out_.pop_front();
542 while (hint_out_.front().bin!=pos) {
543 tintbin f = hint_out_.front();
545 assert (f.bin.contains(pos));
553 hint_out_.front().bin = f.bin.sibling();
554 hint_out_.push_front(f);
556 hint_out_.pop_front();
561 bin_t Channel::OnData (struct evbuffer *evb) { // TODO: HAVE NONE for corrupted data
563 char bin_name_buf[32];
564 bin_t pos = bin_fromUInt32(evbuffer_remove_32be(evb));
566 // Arno: Assuming DATA last message in datagram
567 if (evbuffer_get_length(evb) > file().chunk_size()) {
568 dprintf("%s #%u !data chunk size mismatch %s: exp %lu got " PRISIZET "\n",tintstr(),id_,pos.str(bin_name_buf), file().chunk_size(), evbuffer_get_length(evb));
569 fprintf(stderr,"WARNING: chunk size mismatch: exp %lu got " PRISIZET "\n",file().chunk_size(), evbuffer_get_length(evb));
572 int length = (evbuffer_get_length(evb) < file().chunk_size()) ? evbuffer_get_length(evb) : file().chunk_size();
573 if (!file().ack_out().is_empty(pos)) {
574 // Arno, 2012-01-24: print message for duplicate
575 dprintf("%s #%u Ddata %s\n",tintstr(),id_,pos.str(bin_name_buf));
576 evbuffer_drain(evb, length);
577 data_in_ = tintbin(TINT_NEVER,transfer().ack_out().cover(pos));
579 // Arno, 2012-01-24: Make sure data interarrival periods don't get
580 // screwed up because of these (ignored) duplicates.
584 uint8_t *data = evbuffer_pullup(evb, length);
585 data_in_ = tintbin(NOW,bin_t::NONE);
586 if (!file().OfferData(pos, (char*)data, length)) {
587 evbuffer_drain(evb, length);
588 char bin_name_buf[32];
589 dprintf("%s #%u !data %s\n",tintstr(),id_,pos.str(bin_name_buf));
592 evbuffer_drain(evb, length);
593 dprintf("%s #%u -data %s\n",tintstr(),id_,pos.str(bin_name_buf));
596 fprintf(stderr,"$ ");
598 bin_t cover = transfer().ack_out().cover(pos);
599 for(int i=0; i<transfer().cb_installed; i++)
600 if (cover.layer()>=transfer().cb_agg[i])
601 transfer().callbacks[i](transfer().fd(),cover); // FIXME
602 if (cover.layer() >= 5) // Arno: tested with 32K, presently = 2 ** 5 * chunk_size CHUNKSIZE
603 transfer().OnRecvData( pow((double)2,(double)5)*((double)file().chunk_size()) );
608 bytes_down_ += length;
609 global_bytes_down += length;
614 void Channel::UpdateDIP(bin_t pos)
616 if (!pos.is_none()) {
617 if (last_data_in_time_) {
618 tint dip = NOW - last_data_in_time_;
619 dip_avg_ = ( dip_avg_*3 + dip ) >> 2;
621 last_data_in_time_ = NOW;
626 void Channel::OnAck (struct evbuffer *evb) {
627 bin_t ackd_pos = bin_fromUInt32(evbuffer_remove_32be(evb));
628 tint peer_time = evbuffer_remove_64be(evb); // FIXME 32
629 // FIXME FIXME: wrap around here
630 if (ackd_pos.is_none())
631 return; // likely, broken chunk/ insufficient hashes
632 if (file().size() && ackd_pos.base_offset()>=file().size_in_chunks()) {
633 char bin_name_buf[32];
634 eprintf("invalid ack: %s\n",ackd_pos.str(bin_name_buf));
637 ack_in_.set(ackd_pos);
639 //fprintf(stderr,"OnAck: got bin %s is_complete %d\n", ackd_pos.str(), (int)ack_in_.is_complete_arno( file().ack_out().get_height() ));
642 // find an entry for the send (data out) event
643 while ( di<data_out_.size() && ( data_out_[di]==tintbin() ||
644 !ackd_pos.contains(data_out_[di].bin) ) )
646 // FUTURE: delayed acks
647 // rule out retransmits
648 while ( ri<data_out_tmo_.size() && !ackd_pos.contains(data_out_tmo_[ri].bin) )
650 char bin_name_buf[32];
651 dprintf("%s #%u %cack %s %lli\n",tintstr(),id_,
652 di==data_out_.size()?'?':'-',ackd_pos.str(bin_name_buf),peer_time);
653 if (di!=data_out_.size() && ri==data_out_tmo_.size()) { // not a retransmit
654 // round trip time calculations
655 tint rtt = NOW-data_out_[di].time;
656 rtt_avg_ = (rtt_avg_*7 + rtt) >> 3;
657 dev_avg_ = ( dev_avg_*3 + tintabs(rtt-rtt_avg_) ) >> 2;
658 assert(data_out_[di].time!=TINT_NEVER);
659 // one-way delay calculations
660 tint owd = peer_time - data_out_[di].time;
661 owd_cur_bin_ = 0;//(owd_cur_bin_+1) & 3;
662 owd_current_[owd_cur_bin_] = owd;
663 if ( owd_min_bin_start_+TINT_SEC*30 < NOW ) {
664 owd_min_bin_start_ = NOW;
665 owd_min_bin_ = (owd_min_bin_+1) & 3;
666 owd_min_bins_[owd_min_bin_] = TINT_NEVER;
668 if (owd_min_bins_[owd_min_bin_]>owd)
669 owd_min_bins_[owd_min_bin_] = owd;
670 dprintf("%s #%u sendctrl rtt %lli dev %lli based on %s\n",
671 tintstr(),id_,rtt_avg_,dev_avg_,data_out_[di].bin.str(bin_name_buf));
673 // early loss detection by packet reordering
674 for (int re=0; re<di-MAX_REORDERING; re++) {
675 if (data_out_[re]==tintbin())
677 ack_not_rcvd_recent_++;
678 data_out_tmo_.push_back(data_out_[re].bin);
679 dprintf("%s #%u Rdata %s\n",tintstr(),id_,data_out_.front().bin.str(bin_name_buf));
680 data_out_cap_ = bin_t::ALL;
681 data_out_[re] = tintbin();
684 if (di!=data_out_.size())
685 data_out_[di]=tintbin();
686 // clear zeroed items
687 while (!data_out_.empty() && ( data_out_.front()==tintbin() ||
688 ack_in_.is_filled(data_out_.front().bin) ) )
689 data_out_.pop_front();
690 assert(data_out_.empty() || data_out_.front().time!=TINT_NEVER);
694 void Channel::TimeoutDataOut ( ) {
695 // losses: timeouted packets
696 tint timeout = NOW - ack_timeout();
697 while (!data_out_.empty() &&
698 ( data_out_.front().time<timeout || data_out_.front()==tintbin() ) ) {
699 if (data_out_.front()!=tintbin() && ack_in_.is_empty(data_out_.front().bin)) {
700 ack_not_rcvd_recent_++;
701 data_out_cap_ = bin_t::ALL;
702 data_out_tmo_.push_back(data_out_.front().bin);
703 char bin_name_buf[32];
704 dprintf("%s #%u Tdata %s\n",tintstr(),id_,data_out_.front().bin.str(bin_name_buf));
706 data_out_.pop_front();
708 // clear retransmit queue of older items
709 while (!data_out_tmo_.empty() && data_out_tmo_.front().time<NOW-MAX_POSSIBLE_RTT)
710 data_out_tmo_.pop_front();
714 void Channel::OnHave (struct evbuffer *evb) {
715 bin_t ackd_pos = bin_fromUInt32(evbuffer_remove_32be(evb));
716 if (ackd_pos.is_none())
717 return; // wow, peer has hashes
720 if (ENABLE_VOD_PIECEPICKER) {
721 // Ric: check if we should set the size in the file transfer
722 if (transfer().availability().size() <= 0 && file().size() > 0)
724 transfer().availability().setSize(file().size_in_chunks());
726 // Ric: update the availability if needed
727 transfer().availability().set(id_, ack_in_, ackd_pos);
730 ack_in_.set(ackd_pos);
731 char bin_name_buf[32];
732 dprintf("%s #%u -have %s\n",tintstr(),id_,ackd_pos.str(bin_name_buf));
734 //fprintf(stderr,"OnHave: got bin %s is_complete %d\n", ackd_pos.str(), IsComplete() );
739 void Channel::OnHint (struct evbuffer *evb) {
740 bin_t hint = bin_fromUInt32(evbuffer_remove_32be(evb));
741 // FIXME: wake up here
742 hint_in_.push_back(hint);
743 char bin_name_buf[32];
744 dprintf("%s #%u -hint %s\n",tintstr(),id_,hint.str(bin_name_buf));
748 void Channel::OnHandshake (struct evbuffer *evb) {
750 uint32_t pcid = evbuffer_remove_32be(evb);
751 dprintf("%s #%u -hs %x\n",tintstr(),id_,pcid);
753 if (is_established() && pcid == 0) {
754 // Arno: received explicit close
755 peer_channel_id_ = 0; // == established -> false
760 peer_channel_id_ = pcid;
761 // self-connection check
763 uint32_t try_id = DecodeID(peer_channel_id_);
764 if (channel(try_id) && !channel(try_id)->peer_channel_id_) {
765 peer_channel_id_ = 0;
767 return; // this is a self-connection
771 // FUTURE: channel forking
772 if (is_established())
773 dprintf("%s #%u established %s\n", tintstr(), id_, peer().str());
777 void Channel::OnPex (struct evbuffer *evb) {
778 uint32_t ipv4 = evbuffer_remove_32be(evb);
779 uint16_t port = evbuffer_remove_16be(evb);
780 Address addr(ipv4,port);
781 dprintf("%s #%u -pex %s\n",tintstr(),id_,addr.str());
782 if (transfer().OnPexIn(addr))
783 useless_pex_count_ = 0;
786 dprintf("%s #%u already channel to %s\n", tintstr(),id_,addr.str());
787 useless_pex_count_++;
789 pex_request_outstanding_ = false;
794 void Channel::OnRandomize (struct evbuffer *evb) {
795 dprintf("%s #%u -rand\n",tintstr(),id_ );
796 // Payload is 4 random bytes
797 uint32_t r = evbuffer_remove_32be(evb);
801 void Channel::AddPex (struct evbuffer *evb) {
802 // Gertjan fix: Reverse PEX
803 // PEX messages sent to facilitate NAT/FW puncturing get priority
804 if (!reverse_pex_out_.empty()) {
806 tintbin pex_peer = reverse_pex_out_.front();
807 reverse_pex_out_.pop_front();
808 if (channels[(int) pex_peer.bin.toUInt()] == NULL)
810 Address a = channels[(int) pex_peer.bin.toUInt()]->peer();
811 // Arno, 2012-02-28: Don't send private addresses to non-private peers.
812 if (!a.is_private() || (a.is_private() && peer().is_private()))
814 evbuffer_add_8(evb, SWIFT_PEX_ADD);
815 evbuffer_add_32be(evb, a.ipv4());
816 evbuffer_add_16be(evb, a.port());
817 dprintf("%s #%u +pex (reverse) %s\n",tintstr(),id_,a.str());
819 } while (!reverse_pex_out_.empty() && (SWIFT_MAX_NONDATA_DGRAM_SIZE-evbuffer_get_length(evb)) >= 7);
821 // Arno: 2012-02-23: Don't think this is right. Bit of DoS thing,
822 // that you only get back the addr of people that got your addr.
830 // Arno, 2012-02-28: Don't send private addresses to non-private peers.
831 int chid = 0, tries=0;
835 // Arno, 2011-10-03: Choosing Gertjan's RandomChannel over RevealChannel here.
836 chid = transfer().RandomChannel(id_);
837 if (chid==-1 || chid==id_ || tries > 5) {
838 pex_requested_ = false;
841 a = channels[chid]->peer();
842 if (!a.is_private() || (a.is_private() && peer().is_private()))
847 evbuffer_add_8(evb, SWIFT_PEX_ADD);
848 evbuffer_add_32be(evb, a.ipv4());
849 evbuffer_add_16be(evb, a.port());
850 dprintf("%s #%u +pex %s\n",tintstr(),id_,a.str());
852 pex_requested_ = false;
853 /* Ensure that we don't add the same id to the reverse_pex_out_ queue
855 for (tbqueue::iterator i = channels[chid]->reverse_pex_out_.begin();
856 i != channels[chid]->reverse_pex_out_.end(); i++)
857 if ((int) (i->bin.toUInt()) == id_)
860 dprintf("%s #%u adding pex for channel %u at time %s\n", tintstr(), chid,
861 id_, tintstr(NOW + 2 * TINT_SEC));
862 // Arno, 2011-10-03: should really be a queue of (tint,channel id(= uint32_t)) pairs.
863 channels[chid]->reverse_pex_out_.push_back(tintbin(NOW + 2 * TINT_SEC, bin_t(id_)));
864 if (channels[chid]->send_control_ == KEEP_ALIVE_CONTROL &&
865 channels[chid]->next_send_time_ > NOW + 2 * TINT_SEC)
866 channels[chid]->Reschedule();
869 void Channel::OnPexReq(void) {
870 dprintf("%s #%u -pex req\n", tintstr(), id_);
871 if (NOW > MIN_PEX_REQUEST_INTERVAL + last_pex_request_time_)
872 pex_requested_ = true;
875 void Channel::AddPexReq(struct evbuffer *evb) {
876 // Rate limit the number of PEX requests
877 if (NOW < next_pex_request_time_)
880 // If no answer has been received from a previous request, count it as useless
881 if (pex_request_outstanding_)
882 useless_pex_count_++;
884 pex_request_outstanding_ = false;
886 // Initiate at most SWIFT_MAX_CONNECTIONS connections
887 if (transfer().hs_in_.size() >= SWIFT_MAX_CONNECTIONS ||
888 // Check whether this channel has been providing useful peer information
889 useless_pex_count_ > 2)
891 // Arno, 2012-02-23: Fix: Code doesn't recover from useless_pex_count_ > 2,
892 // let's just try again in 30s
893 useless_pex_count_ = 0;
894 next_pex_request_time_ = NOW + 30 * TINT_SEC;
899 dprintf("%s #%u +pex req\n", tintstr(), id_);
900 evbuffer_add_8(evb, SWIFT_PEX_REQ);
901 /* Add a little more than the minimum interval, such that the other party is
902 less likely to drop it due to too high rate */
903 next_pex_request_time_ = NOW + MIN_PEX_REQUEST_INTERVAL * 1.1;
904 pex_request_outstanding_ = true;
910 * Channel class methods
913 void Channel::LibeventReceiveCallback(evutil_socket_t fd, short event, void *arg) {
914 // Called by libevent when a datagram is received on the socket
917 event_add(&evrecv, NULL);
920 #define NUM_DATAGRAMS 10
922 void Channel::RecvDatagram (evutil_socket_t socket) {
923 struct evbuffer *pevb[NUM_DATAGRAMS];
924 for (int i=0; i<NUM_DATAGRAMS; ++i)
925 pevb[i] = evbuffer_new();
927 //FIXME: make this more readable
929 addr.addr = (struct sockaddr_mptp *) calloc(1, sizeof(struct sockaddr_mptp) + NUM_DATAGRAMS * sizeof(struct mptp_dest));
930 addr.addr->count = NUM_DATAGRAMS;
931 RecvFrom(socket, addr, pevb);
933 for (; i<addr.addr->count; ++i) {
934 struct evbuffer *evb = pevb[i];
936 fromi.addr->dests[0].addr = addr.addr->dests[i].addr;
937 fromi.addr->dests[0].port = addr.addr->dests[i].port;
938 size_t evboriglen = evbuffer_get_length(evb);
939 #define return_log(...) { fprintf(stderr,__VA_ARGS__); evbuffer_free(evb); return; }
940 if (evbuffer_get_length(evb)<4)
941 return_log("socket layer weird: datagram shorter than 4 bytes from %s (prob ICMP unreach)\n",fromi.str());
942 uint32_t mych = evbuffer_remove_32be(evb);
944 Channel* channel = NULL;
945 if (mych==0) { // peer initiates handshake
946 if (evbuffer_get_length(evb)<1+4+1+4+Sha1Hash::SIZE)
947 return_log ("%s #0 incorrect size %i initial handshake packet %s\n",
948 tintstr(),(int)evbuffer_get_length(evb),fromi.str());
949 uint8_t hashid = evbuffer_remove_8(evb);
950 if (hashid!=SWIFT_HASH)
951 return_log ("%s #0 no hash in the initial handshake %s\n",
952 tintstr(),fromi.str());
953 bin_t pos = bin_fromUInt32(evbuffer_remove_32be(evb));
955 return_log ("%s #0 that is not the root hash %s\n",tintstr(),fromi.str());
956 hash = evbuffer_remove_hash(evb);
957 FileTransfer* ft = FileTransfer::Find(hash);
959 return_log ("%s #0 hash %s unknown, requested by %s\n",tintstr(),hash.hex().c_str(),fromi.str());
960 dprintf("%s #0 -hash ALL %s\n",tintstr(),hash.hex().c_str());
962 // Arno, 2012-02-27: Check for duplicate channel
963 Channel* existchannel = ft->FindChannel(fromi,NULL);
966 // Arno: 2011-10-13: Ignore if established, otherwise consider
967 // it a concurrent connection attempt.
968 if (existchannel->is_established()) {
969 // ARNOTODO: Read complete handshake here so we know whether
970 // attempt is to new channel or to existing. Currently read
973 return_log("%s #0 have a channel already to %s\n",tintstr(),fromi.str());
975 channel = existchannel;
976 //fprintf(stderr,"Channel::RecvDatagram: HANDSHAKE: reuse channel %s\n", channel->peer_.str() );
979 if (channel == NULL) {
980 //fprintf(stderr,"Channel::RecvDatagram: HANDSHAKE: create new channel %s\n", addr.str() );
981 channel = new Channel(ft, socket, fromi);
983 //fprintf(stderr,"CHANNEL INCOMING DEF hass %s is id %d\n",hash.hex().c_str(),channel->id());
985 } else { // peer responds to my handshake (and other messages)
986 mych = DecodeID(mych);
987 if (mych>=channels.size())
988 return_log("%s invalid channel #%u, %s\n",tintstr(),mych,fromi.str());
989 channel = channels[mych];
991 return_log ("%s #%u is already closed\n",tintstr(),mych);
992 if (channel->IsDiffSenderOrDuplicate(fromi,mych)) {
993 channel->Schedule4Close();
996 channel->own_id_mentioned_ = true;
998 channel->raw_bytes_down_ += evboriglen;
999 //dprintf("recvd %i bytes for %i\n",data.size(),channel->id);
1000 bool wasestablished = channel->is_established();
1002 dprintf("%s #%u peer %s recv_peer %s addr %s\n", tintstr(),mych, channel->peer().str(), channel->recv_peer().str(), fromi.str() );
1008 if (wasestablished && !channel->is_established()) {
1009 // Arno, 2012-01-26: Received an explict close, clean up channel, safely.
1010 channel->Schedule4Close();
1013 for (; i<NUM_DATAGRAMS; ++i)
1014 evbuffer_free(pevb[i]);
1020 * Channel instance methods
1023 void Channel::CloseChannelByAddress(const Address &addr)
1025 // fprintf(stderr,"CloseChannelByAddress: address is %s\n", addr.str() );
1026 std::vector<Channel *>::iterator iter;
1027 for (iter = channels.begin(); iter != channels.end(); iter++)
1030 if (c != NULL && c->peer_ == addr)
1032 // ARNOSMPTODO: will do another send attempt before not being
1034 c->peer_channel_id_ = 0; // established->false, do no more sending
1035 c->Schedule4Close();
1042 void Channel::Close () {
1044 this->SwitchSendControl(CLOSE_CONTROL);
1046 if (is_established())
1047 this->Send(); // Arno: send explicit close
1049 if (ENABLE_VOD_PIECEPICKER) {
1050 // Ric: remove it's binmap from the availability
1051 transfer().availability().remove(id_, ack_in_);
1055 // Arno: ensure LibeventSendCallback is no longer called with ptr to this Channel
1060 void Channel::Reschedule () {
1062 // Arno: CAREFUL: direct send depends on diff between next_send_time_ and
1063 // NOW to be 0, so any calls to Time in between may put things off. Sigh.
1065 next_send_time_ = NextSendTime();
1066 if (next_send_time_!=TINT_NEVER) {
1068 assert(next_send_time_<NOW+TINT_MIN);
1069 tint duein = next_send_time_-NOW;
1071 // Arno, 2011-10-18: libevent's timer implementation appears to be
1072 // really slow, i.e., timers set for 100 usec from now get called
1073 // at least two times later :-( Hence, for sends after receives
1074 // perform them directly.
1075 dprintf("%s #%u requeue direct send\n",tintstr(),id_);
1076 LibeventSendCallback(-1,EV_TIMEOUT,this);
1079 if (evsend_ptr_ != NULL) {
1080 struct timeval duetv = *tint2tv(duein);
1081 evtimer_add(evsend_ptr_,&duetv);
1082 dprintf("%s #%u requeue for %s in %lli\n",tintstr(),id_,tintstr(next_send_time_), duein);
1085 dprintf("%s #%u cannot requeue for %s, closed\n",tintstr(),id_,tintstr(next_send_time_));
1089 dprintf("%s #%u resched, will close\n",tintstr(),id_);
1090 this->Schedule4Close();
1096 * Channel class methods
1098 void Channel::LibeventSendCallback(int fd, short event, void *arg) {
1100 // Called by libevent when it is the requested send time.
1102 Channel * sender = (Channel*) arg;
1103 if (NOW<sender->next_send_time_-TINT_MSEC)
1104 dprintf("%s #%u suspicious send %s<%s\n",tintstr(),
1105 sender->id(),tintstr(NOW),tintstr(sender->next_send_time_));
1106 if (sender->next_send_time_ != TINT_NEVER)