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
|
package quic
import (
"context"
"fmt"
"sync"
"time"
"github.com/quic-go/quic-go/internal/ackhandler"
"github.com/quic-go/quic-go/internal/flowcontrol"
"github.com/quic-go/quic-go/internal/protocol"
"github.com/quic-go/quic-go/internal/utils"
"github.com/quic-go/quic-go/internal/wire"
)
// A SendStream is a unidirectional Send Stream.
type SendStream struct {
mutex sync.Mutex
numOutstandingFrames int64 // outstanding STREAM and RESET_STREAM frames
retransmissionQueue []*wire.StreamFrame
ctx context.Context
ctxCancel context.CancelCauseFunc
streamID protocol.StreamID
sender streamSender
// reliableSize is the portion of the stream that needs to be transmitted reliably,
// even if the stream is cancelled.
// This requires the peer to support RESET_STREAM_AT.
// This value should not be accessed directly, but only through the reliableOffset method.
// This method returns 0 if the peer doesn't support the RESET_STREAM_AT extension.
reliableSize protocol.ByteCount
writeOffset protocol.ByteCount
shutdownErr error
resetErr *StreamError
queuedResetStreamFrame *wire.ResetStreamFrame
supportsResetStreamAt bool
finishedWriting bool // set once Close() is called
finSent bool // set when a STREAM_FRAME with FIN bit has been sent
// Set when the application knows about the cancellation.
// This can happen because the application called CancelWrite,
// or because Write returned the error (for remote cancellations).
cancellationFlagged bool
completed bool // set when this stream has been reported to the streamSender as completed
dataForWriting []byte // during a Write() call, this slice is the part of p that still needs to be sent out
nextFrame *wire.StreamFrame
writeChan chan struct{}
writeOnce chan struct{}
deadline time.Time
flowController flowcontrol.StreamFlowController
}
var (
_ streamControlFrameGetter = &SendStream{}
_ outgoingStream = &SendStream{}
_ sendStreamFrameHandler = &SendStream{}
)
func newSendStream(
ctx context.Context,
streamID protocol.StreamID,
sender streamSender,
flowController flowcontrol.StreamFlowController,
supportsResetStreamAt bool,
) *SendStream {
s := &SendStream{
streamID: streamID,
sender: sender,
flowController: flowController,
writeChan: make(chan struct{}, 1),
writeOnce: make(chan struct{}, 1), // cap: 1, to protect against concurrent use of Write
supportsResetStreamAt: supportsResetStreamAt,
}
s.ctx, s.ctxCancel = context.WithCancelCause(ctx)
return s
}
// StreamID returns the stream ID.
func (s *SendStream) StreamID() StreamID {
return s.streamID // same for receiveStream and sendStream
}
// Write writes data to the stream.
// Write can be made to time out using [SendStream.SetWriteDeadline].
// If the stream was canceled, the error is a [StreamError].
func (s *SendStream) Write(p []byte) (int, error) {
// Concurrent use of Write is not permitted (and doesn't make any sense),
// but sometimes people do it anyway.
// Make sure that we only execute one call at any given time to avoid hard to debug failures.
s.writeOnce <- struct{}{}
defer func() { <-s.writeOnce }()
isNewlyCompleted, n, err := s.write(p)
if isNewlyCompleted {
s.sender.onStreamCompleted(s.streamID)
}
return n, err
}
func (s *SendStream) write(p []byte) (bool /* is newly completed */, int, error) {
s.mutex.Lock()
defer s.mutex.Unlock()
if s.resetErr != nil {
s.cancellationFlagged = true
return s.isNewlyCompleted(), 0, s.resetErr
}
if s.shutdownErr != nil {
return false, 0, s.shutdownErr
}
if s.finishedWriting {
return false, 0, fmt.Errorf("write on closed stream %d", s.streamID)
}
if !s.deadline.IsZero() && !time.Now().Before(s.deadline) {
return false, 0, errDeadline
}
if len(p) == 0 {
return false, 0, nil
}
s.dataForWriting = p
var (
deadlineTimer *utils.Timer
bytesWritten int
notifiedSender bool
)
for {
var copied bool
var deadline time.Time
// As soon as dataForWriting becomes smaller than a certain size x, we copy all the data to a STREAM frame (s.nextFrame),
// which can then be popped the next time we assemble a packet.
// This allows us to return Write() when all data but x bytes have been sent out.
// When the user now calls Close(), this is much more likely to happen before we popped that last STREAM frame,
// allowing us to set the FIN bit on that frame (instead of sending an empty STREAM frame with FIN).
if s.canBufferStreamFrame() && len(s.dataForWriting) > 0 {
if s.nextFrame == nil {
f := wire.GetStreamFrame()
f.Offset = s.writeOffset
f.StreamID = s.streamID
f.DataLenPresent = true
f.Data = f.Data[:len(s.dataForWriting)]
copy(f.Data, s.dataForWriting)
s.nextFrame = f
} else {
l := len(s.nextFrame.Data)
s.nextFrame.Data = s.nextFrame.Data[:l+len(s.dataForWriting)]
copy(s.nextFrame.Data[l:], s.dataForWriting)
}
s.dataForWriting = nil
bytesWritten = len(p)
copied = true
} else {
bytesWritten = len(p) - len(s.dataForWriting)
deadline = s.deadline
if !deadline.IsZero() {
if !time.Now().Before(deadline) {
s.dataForWriting = nil
return false, bytesWritten, errDeadline
}
if deadlineTimer == nil {
deadlineTimer = utils.NewTimer()
defer deadlineTimer.Stop()
}
deadlineTimer.Reset(deadline)
}
if s.dataForWriting == nil || s.shutdownErr != nil || s.resetErr != nil {
break
}
}
s.mutex.Unlock()
if !notifiedSender {
s.sender.onHasStreamData(s.streamID, s) // must be called without holding the mutex
notifiedSender = true
}
if copied {
s.mutex.Lock()
break
}
if deadline.IsZero() {
<-s.writeChan
} else {
select {
case <-s.writeChan:
case <-deadlineTimer.Chan():
deadlineTimer.SetRead()
}
}
s.mutex.Lock()
}
if bytesWritten == len(p) {
return false, bytesWritten, nil
}
if s.shutdownErr != nil {
return false, bytesWritten, s.shutdownErr
}
if s.resetErr != nil {
s.cancellationFlagged = true
return s.isNewlyCompleted(), bytesWritten, s.resetErr
}
return false, bytesWritten, nil
}
func (s *SendStream) canBufferStreamFrame() bool {
var l protocol.ByteCount
if s.nextFrame != nil {
l = s.nextFrame.DataLen()
}
return l+protocol.ByteCount(len(s.dataForWriting)) <= protocol.MaxPacketBufferSize
}
// popStreamFrame returns the next STREAM frame that is supposed to be sent on this stream
// maxBytes is the maximum length this frame (including frame header) will have.
func (s *SendStream) popStreamFrame(maxBytes protocol.ByteCount, v protocol.Version) (_ ackhandler.StreamFrame, _ *wire.StreamDataBlockedFrame, hasMore bool) {
s.mutex.Lock()
f, blocked, hasMoreData := s.popNewOrRetransmittedStreamFrame(maxBytes, v)
if f != nil {
s.numOutstandingFrames++
}
s.mutex.Unlock()
if f == nil {
return ackhandler.StreamFrame{}, blocked, hasMoreData
}
return ackhandler.StreamFrame{
Frame: f,
Handler: (*sendStreamAckHandler)(s),
}, blocked, hasMoreData
}
func (s *SendStream) popNewOrRetransmittedStreamFrame(maxBytes protocol.ByteCount, v protocol.Version) (_ *wire.StreamFrame, _ *wire.StreamDataBlockedFrame, hasMoreData bool) {
if s.shutdownErr != nil {
return nil, nil, false
}
if s.resetErr != nil {
reliableOffset := s.reliableOffset()
if reliableOffset == 0 || (s.writeOffset >= reliableOffset && len(s.retransmissionQueue) == 0) {
return nil, nil, false
}
}
if len(s.retransmissionQueue) > 0 {
f, hasMoreRetransmissions := s.maybeGetRetransmission(maxBytes, v)
if f != nil || hasMoreRetransmissions {
if f == nil {
return nil, nil, true
}
// We always claim that we have more data to send.
// This might be incorrect, in which case there'll be a spurious call to popStreamFrame in the future.
return f, nil, true
}
}
if len(s.dataForWriting) == 0 && s.nextFrame == nil {
if s.finishedWriting && !s.finSent {
s.finSent = true
return &wire.StreamFrame{
StreamID: s.streamID,
Offset: s.writeOffset,
DataLenPresent: true,
Fin: true,
}, nil, false
}
return nil, nil, false
}
maxDataLen := s.flowController.SendWindowSize()
if maxDataLen == 0 {
return nil, nil, true
}
// if the stream is canceled, only data up to the reliable size needs to be sent
reliableOffset := s.reliableOffset()
if s.resetErr != nil && reliableOffset > 0 {
maxDataLen = min(maxDataLen, reliableOffset-s.writeOffset)
}
f, hasMoreData := s.popNewStreamFrame(maxBytes, maxDataLen, v)
if f == nil {
return nil, nil, hasMoreData
}
if f.DataLen() > 0 {
s.writeOffset += f.DataLen()
s.flowController.AddBytesSent(f.DataLen())
}
if s.resetErr != nil && s.writeOffset >= reliableOffset {
hasMoreData = false
}
var blocked *wire.StreamDataBlockedFrame
// If the entire send window is used, the stream might have become blocked on stream-level flow control.
// This is not guaranteed though, because the stream might also have been blocked on connection-level flow control.
if f.DataLen() == maxDataLen && s.flowController.IsNewlyBlocked() {
blocked = &wire.StreamDataBlockedFrame{StreamID: s.streamID, MaximumStreamData: s.writeOffset}
}
f.Fin = s.finishedWriting && s.dataForWriting == nil && s.nextFrame == nil && !s.finSent
if f.Fin {
s.finSent = true
}
return f, blocked, hasMoreData
}
// popNewStreamFrame returns a new STREAM frame to send for this stream
// hasMoreData says if there's more data to send, *not* taking into account the reliable size
func (s *SendStream) popNewStreamFrame(maxBytes, maxDataLen protocol.ByteCount, v protocol.Version) (_ *wire.StreamFrame, hasMoreData bool) {
if s.nextFrame != nil {
maxDataLen := min(maxDataLen, s.nextFrame.MaxDataLen(maxBytes, v))
if maxDataLen == 0 {
return nil, true
}
nextFrame := s.nextFrame
s.nextFrame = nil
if nextFrame.DataLen() > maxDataLen {
s.nextFrame = wire.GetStreamFrame()
s.nextFrame.StreamID = s.streamID
s.nextFrame.Offset = s.writeOffset + maxDataLen
s.nextFrame.Data = s.nextFrame.Data[:nextFrame.DataLen()-maxDataLen]
s.nextFrame.DataLenPresent = true
copy(s.nextFrame.Data, nextFrame.Data[maxDataLen:])
nextFrame.Data = nextFrame.Data[:maxDataLen]
} else {
s.signalWrite()
}
return nextFrame, s.nextFrame != nil || s.dataForWriting != nil
}
f := wire.GetStreamFrame()
f.Fin = false
f.StreamID = s.streamID
f.Offset = s.writeOffset
f.DataLenPresent = true
f.Data = f.Data[:0]
hasMoreData = s.popNewStreamFrameWithoutBuffer(f, maxBytes, maxDataLen, v)
if len(f.Data) == 0 && !f.Fin {
f.PutBack()
return nil, hasMoreData
}
return f, hasMoreData
}
func (s *SendStream) popNewStreamFrameWithoutBuffer(f *wire.StreamFrame, maxBytes, sendWindow protocol.ByteCount, v protocol.Version) bool {
maxDataLen := f.MaxDataLen(maxBytes, v)
if maxDataLen == 0 { // a STREAM frame must have at least one byte of data
return s.dataForWriting != nil || s.nextFrame != nil || s.finishedWriting
}
s.getDataForWriting(f, min(maxDataLen, sendWindow))
return s.dataForWriting != nil || s.nextFrame != nil || s.finishedWriting
}
func (s *SendStream) maybeGetRetransmission(maxBytes protocol.ByteCount, v protocol.Version) (*wire.StreamFrame, bool /* has more retransmissions */) {
f := s.retransmissionQueue[0]
newFrame, needsSplit := f.MaybeSplitOffFrame(maxBytes, v)
if needsSplit {
return newFrame, true
}
s.retransmissionQueue = s.retransmissionQueue[1:]
return f, len(s.retransmissionQueue) > 0
}
func (s *SendStream) getDataForWriting(f *wire.StreamFrame, maxBytes protocol.ByteCount) {
if protocol.ByteCount(len(s.dataForWriting)) <= maxBytes {
f.Data = f.Data[:len(s.dataForWriting)]
copy(f.Data, s.dataForWriting)
s.dataForWriting = nil
s.signalWrite()
return
}
f.Data = f.Data[:maxBytes]
copy(f.Data, s.dataForWriting)
s.dataForWriting = s.dataForWriting[maxBytes:]
if s.canBufferStreamFrame() {
s.signalWrite()
}
}
func (s *SendStream) isNewlyCompleted() bool {
if s.completed {
return false
}
if s.nextFrame != nil && s.nextFrame.DataLen() > 0 {
return false
}
// We need to keep the stream around until all frames have been sent and acknowledged.
if s.numOutstandingFrames > 0 || len(s.retransmissionQueue) > 0 || s.queuedResetStreamFrame != nil {
return false
}
// The stream is completed if we sent the FIN.
if s.finSent {
s.completed = true
return true
}
// The stream is also completed if:
// 1. the application called CancelWrite, or
// 2. we received a STOP_SENDING, and
// * the application consumed the error via Write, or
// * the application called Close
if s.resetErr != nil && (s.cancellationFlagged || s.finishedWriting) {
s.completed = true
return true
}
return false
}
// Close closes the write-direction of the stream.
// Future calls to Write are not permitted after calling Close.
// It must not be called concurrently with Write.
// It must not be called after calling CancelWrite.
func (s *SendStream) Close() error {
s.mutex.Lock()
if s.shutdownErr != nil || s.finishedWriting {
s.mutex.Unlock()
return nil
}
s.finishedWriting = true
cancelled := s.resetErr != nil
if cancelled {
s.cancellationFlagged = true
}
completed := s.isNewlyCompleted()
s.mutex.Unlock()
if completed {
s.sender.onStreamCompleted(s.streamID)
}
if cancelled {
return fmt.Errorf("close called for canceled stream %d", s.streamID)
}
s.sender.onHasStreamData(s.streamID, s) // need to send the FIN, must be called without holding the mutex
s.ctxCancel(nil)
return nil
}
// SetReliableBoundary marks the data written to this stream so far as reliable.
// It is valid to call this function multiple times, thereby increasing the reliable size.
// It only has an effect if the peer enabled support for the RESET_STREAM_AT extension,
// otherwise, it is a no-op.
func (s *SendStream) SetReliableBoundary() {
s.mutex.Lock()
defer s.mutex.Unlock()
s.reliableSize = s.writeOffset
if s.nextFrame != nil {
s.reliableSize += s.nextFrame.DataLen()
}
}
// CancelWrite aborts sending on this stream.
// Data already written, but not yet delivered to the peer is not guaranteed to be delivered reliably.
// Write will unblock immediately, and future calls to Write will fail.
// When called multiple times it is a no-op.
// When called after Close, it aborts reliable delivery of outstanding stream data.
// Note that there is no guarantee if the peer will receive the FIN or the cancellation error first.
func (s *SendStream) CancelWrite(errorCode StreamErrorCode) {
s.mutex.Lock()
if s.shutdownErr != nil {
s.mutex.Unlock()
return
}
s.cancellationFlagged = true
if s.resetErr != nil {
completed := s.isNewlyCompleted()
s.mutex.Unlock()
// The user has called CancelWrite. If the previous cancellation was because of a
// STOP_SENDING, we don't need to flag the error to the user anymore.
if completed {
s.sender.onStreamCompleted(s.streamID)
}
return
}
s.resetErr = &StreamError{StreamID: s.streamID, ErrorCode: errorCode, Remote: false}
s.ctxCancel(s.resetErr)
reliableOffset := s.reliableOffset()
if reliableOffset == 0 {
s.numOutstandingFrames = 0
s.retransmissionQueue = nil
}
s.queuedResetStreamFrame = &wire.ResetStreamFrame{
StreamID: s.streamID,
FinalSize: max(s.writeOffset, reliableOffset),
ErrorCode: errorCode,
// if the peer doesn't support the extension, the reliable offset will always be 0
ReliableSize: reliableOffset,
}
if reliableOffset > 0 {
if s.nextFrame != nil {
if s.nextFrame.Offset >= reliableOffset {
s.nextFrame.PutBack()
s.nextFrame = nil
} else if s.nextFrame.Offset+s.nextFrame.DataLen() > reliableOffset {
s.nextFrame.Data = s.nextFrame.Data[:reliableOffset-s.nextFrame.Offset]
}
}
if len(s.retransmissionQueue) > 0 {
retransmissionQueue := make([]*wire.StreamFrame, 0, len(s.retransmissionQueue))
for _, f := range s.retransmissionQueue {
if f.Offset >= reliableOffset {
f.PutBack()
continue
}
if f.Offset+f.DataLen() <= reliableOffset {
retransmissionQueue = append(retransmissionQueue, f)
} else {
f.Data = f.Data[:reliableOffset-f.Offset]
retransmissionQueue = append(retransmissionQueue, f)
}
}
s.retransmissionQueue = retransmissionQueue
}
}
s.mutex.Unlock()
s.signalWrite()
s.sender.onHasStreamControlFrame(s.streamID, s)
}
func (s *SendStream) enableResetStreamAt() {
s.mutex.Lock()
s.supportsResetStreamAt = true
s.mutex.Unlock()
}
func (s *SendStream) updateSendWindow(limit protocol.ByteCount) {
updated := s.flowController.UpdateSendWindow(limit)
if !updated { // duplicate or reordered MAX_STREAM_DATA frame
return
}
s.mutex.Lock()
hasStreamData := s.dataForWriting != nil || s.nextFrame != nil
s.mutex.Unlock()
if hasStreamData {
s.sender.onHasStreamData(s.streamID, s)
}
}
func (s *SendStream) handleStopSendingFrame(f *wire.StopSendingFrame) {
s.mutex.Lock()
if s.shutdownErr != nil {
s.mutex.Unlock()
return
}
// If the stream was already cancelled (either locally, or due to a previous STOP_SENDING frame),
// there's nothing else to do.
if s.resetErr != nil && s.reliableOffset() == 0 {
s.mutex.Unlock()
return
}
// if the peer stopped reading from the stream, there's no need to transmit any data reliably
s.reliableSize = 0
s.numOutstandingFrames = 0
s.retransmissionQueue = nil
if s.resetErr == nil {
s.resetErr = &StreamError{StreamID: s.streamID, ErrorCode: f.ErrorCode, Remote: true}
s.ctxCancel(s.resetErr)
}
s.queuedResetStreamFrame = &wire.ResetStreamFrame{
StreamID: s.streamID,
FinalSize: s.writeOffset,
ErrorCode: s.resetErr.ErrorCode,
}
s.mutex.Unlock()
s.signalWrite()
s.sender.onHasStreamControlFrame(s.streamID, s)
}
func (s *SendStream) getControlFrame(time.Time) (_ ackhandler.Frame, ok, hasMore bool) {
s.mutex.Lock()
defer s.mutex.Unlock()
if s.queuedResetStreamFrame == nil {
return ackhandler.Frame{}, false, false
}
s.numOutstandingFrames++
f := ackhandler.Frame{
Frame: s.queuedResetStreamFrame,
Handler: (*sendStreamResetStreamHandler)(s),
}
s.queuedResetStreamFrame = nil
return f, true, false
}
func (s *SendStream) reliableOffset() protocol.ByteCount {
if !s.supportsResetStreamAt {
return 0
}
return s.reliableSize
}
// The Context is canceled as soon as the write-side of the stream is closed.
// This happens when Close() or CancelWrite() is called, or when the peer
// cancels the read-side of their stream.
// The cancellation cause is set to the error that caused the stream to
// close, or `context.Canceled` in case the stream is closed without error.
func (s *SendStream) Context() context.Context {
return s.ctx
}
// SetWriteDeadline sets the deadline for future Write calls
// and any currently-blocked Write call.
// Even if write times out, it may return n > 0, indicating that
// some data was successfully written.
// A zero value for t means Write will not time out.
func (s *SendStream) SetWriteDeadline(t time.Time) error {
s.mutex.Lock()
s.deadline = t
s.mutex.Unlock()
s.signalWrite()
return nil
}
// CloseForShutdown closes a stream abruptly.
// It makes Write unblock (and return the error) immediately.
// The peer will NOT be informed about this: the stream is closed without sending a FIN or RST.
func (s *SendStream) closeForShutdown(err error) {
s.mutex.Lock()
if s.shutdownErr == nil && !s.finishedWriting {
s.shutdownErr = err
}
s.mutex.Unlock()
s.signalWrite()
}
// signalWrite performs a non-blocking send on the writeChan
func (s *SendStream) signalWrite() {
select {
case s.writeChan <- struct{}{}:
default:
}
}
type sendStreamAckHandler SendStream
var _ ackhandler.FrameHandler = &sendStreamAckHandler{}
func (s *sendStreamAckHandler) OnAcked(f wire.Frame) {
sf := f.(*wire.StreamFrame)
sf.PutBack()
s.mutex.Lock()
if s.resetErr != nil && (*SendStream)(s).reliableOffset() == 0 {
s.mutex.Unlock()
return
}
s.numOutstandingFrames--
if s.numOutstandingFrames < 0 {
panic("numOutStandingFrames negative")
}
completed := (*SendStream)(s).isNewlyCompleted()
s.mutex.Unlock()
if completed {
s.sender.onStreamCompleted(s.streamID)
}
}
func (s *sendStreamAckHandler) OnLost(f wire.Frame) {
sf := f.(*wire.StreamFrame)
s.mutex.Lock()
// If the reliable size was 0 when the stream was cancelled,
// the number of outstanding frames was immediately set to 0, and the retransmission queue was dropped.
if s.resetErr != nil && (*SendStream)(s).reliableOffset() == 0 {
s.mutex.Unlock()
return
}
s.numOutstandingFrames--
if s.numOutstandingFrames < 0 {
panic("numOutStandingFrames negative")
}
if s.resetErr != nil && (*SendStream)(s).reliableOffset() > 0 {
// If the stream was reset, and this frame is beyond the reliable offset,
// it doesn't need to be retransmitted.
if sf.Offset >= (*SendStream)(s).reliableOffset() {
sf.PutBack()
// If this frame was the last one tracked, losing it might cause the stream to be completed.
completed := (*SendStream)(s).isNewlyCompleted()
s.mutex.Unlock()
if completed {
s.sender.onStreamCompleted(s.streamID)
}
return
}
// If the payload of the frame extends beyond the reliable size,
// truncate the frame to the reliable size.
if sf.Offset+sf.DataLen() > (*SendStream)(s).reliableOffset() {
sf.Data = sf.Data[:(*SendStream)(s).reliableOffset()-sf.Offset]
}
}
sf.DataLenPresent = true
s.retransmissionQueue = append(s.retransmissionQueue, sf)
s.mutex.Unlock()
s.sender.onHasStreamData(s.streamID, (*SendStream)(s))
}
type sendStreamResetStreamHandler SendStream
var _ ackhandler.FrameHandler = &sendStreamResetStreamHandler{}
func (s *sendStreamResetStreamHandler) OnAcked(f wire.Frame) {
rsf := f.(*wire.ResetStreamFrame)
s.mutex.Lock()
// If the peer sent a STOP_SENDING after we sent a RESET_STREAM_AT frame,
// we sent 1. reduced the reliable size to 0 and 2. sent a RESET_STREAM frame.
// In this case, we don't care about the acknowledgment of this frame.
if rsf.ReliableSize != (*SendStream)(s).reliableOffset() {
s.mutex.Unlock()
return
}
s.numOutstandingFrames--
if s.numOutstandingFrames < 0 {
panic("numOutStandingFrames negative")
}
completed := (*SendStream)(s).isNewlyCompleted()
s.mutex.Unlock()
if completed {
s.sender.onStreamCompleted(s.streamID)
}
}
func (s *sendStreamResetStreamHandler) OnLost(f wire.Frame) {
rsf := f.(*wire.ResetStreamFrame)
s.mutex.Lock()
// If the peer sent a STOP_SENDING after we sent a RESET_STREAM_AT frame,
// we sent 1. reduced the reliable size to 0 and 2. sent a RESET_STREAM frame.
// In this case, the loss of the RESET_STREAM_AT frame can be ignored.
if rsf.ReliableSize != (*SendStream)(s).reliableOffset() {
s.mutex.Unlock()
return
}
s.queuedResetStreamFrame = rsf
s.numOutstandingFrames--
s.mutex.Unlock()
s.sender.onHasStreamControlFrame(s.streamID, (*SendStream)(s))
}
|