From e9e399c25d4fcad2adfe92d73b192b9784774964 Mon Sep 17 00:00:00 2001 From: Bhasker Hariharan Date: Thu, 19 Mar 2020 07:18:47 -0700 Subject: Remove workMu from tcpip.Endpoint. workMu is removed and e.mu is now a mutex that supports TryLock. The packet processing path tries to lock the mutex and if its locked it will just queue the packet and move on. The endpoint.UnlockUser() will process any backlog of packets before unlocking the socket. This simplifies the locking inside tcp endpoints a lot. Further the endpoint.LockUser() implements spinning as long as the lock is not held by another syscall goroutine. This ensures low latency as not spinning leads to the task thread being put to sleep if the lock is held by the packet dispatch path. This is suboptimal as the lower layer rarely holds the lock for long so implementing spinning here helps. If the lock is held by another task goroutine then we just proceed to call LockUser() and the task could be put to sleep. The protocol goroutines themselves just call e.mu.Lock() and block if the lock is currently not available. Updates #231, #357 PiperOrigin-RevId: 301808349 --- pkg/tcpip/transport/tcp/accept.go | 90 +++--- pkg/tcpip/transport/tcp/connect.go | 78 ++--- pkg/tcpip/transport/tcp/dispatcher.go | 8 +- pkg/tcpip/transport/tcp/endpoint.go | 495 ++++++++++++++++-------------- pkg/tcpip/transport/tcp/endpoint_state.go | 5 +- pkg/tcpip/transport/tcp/protocol.go | 38 +-- pkg/tcpip/transport/tcp/rcv.go | 6 - pkg/tcpip/transport/tcp/segment_queue.go | 8 +- pkg/tcpip/transport/tcp/snd.go | 3 - pkg/tcpip/transport/tcp/tcp_test.go | 41 ++- 10 files changed, 402 insertions(+), 370 deletions(-) (limited to 'pkg/tcpip') diff --git a/pkg/tcpip/transport/tcp/accept.go b/pkg/tcpip/transport/tcp/accept.go index 85049e54e..4d7602d54 100644 --- a/pkg/tcpip/transport/tcp/accept.go +++ b/pkg/tcpip/transport/tcp/accept.go @@ -221,7 +221,8 @@ func (l *listenContext) isCookieValid(id stack.TransportEndpointID, cookie seqnu } // createConnectingEndpoint creates a new endpoint in a connecting state, with -// the connection parameters given by the arguments. +// the connection parameters given by the arguments. The endpoint is returned +// with n.mu held. func (l *listenContext) createConnectingEndpoint(s *segment, iss seqnum.Value, irs seqnum.Value, rcvdSynOpts *header.TCPSynOptions, queue *waiter.Queue) (*endpoint, *tcpip.Error) { // Create a new endpoint. netProto := l.netProto @@ -243,21 +244,6 @@ func (l *listenContext) createConnectingEndpoint(s *segment, iss seqnum.Value, i n.initGSO() - // Now inherit any socket options that should be inherited from the - // listening endpoint. - // In case of Forwarder listenEP will be nil and hence this check. - if l.listenEP != nil { - l.listenEP.propagateInheritableOptions(n) - } - - // Register new endpoint so that packets are routed to it. - if err := n.stack.RegisterTransportEndpoint(n.boundNICID, n.effectiveNetProtos, ProtocolNumber, n.ID, n, n.reusePort, n.boundBindToDevice); err != nil { - n.Close() - return nil, err - } - - n.isRegistered = true - // Create sender and receiver. // // The receiver at least temporarily has a zero receive window scale, @@ -269,11 +255,27 @@ func (l *listenContext) createConnectingEndpoint(s *segment, iss seqnum.Value, i // window to grow to a really large value. n.rcvAutoParams.prevCopied = n.initialReceiveWindow() + // Lock the endpoint before registering to ensure that no out of + // band changes are possible due to incoming packets etc till + // the endpoint is done initializing. + n.mu.Lock() + + // Register new endpoint so that packets are routed to it. + if err := n.stack.RegisterTransportEndpoint(n.boundNICID, n.effectiveNetProtos, ProtocolNumber, n.ID, n, n.reusePort, n.boundBindToDevice); err != nil { + n.mu.Unlock() + n.Close() + return nil, err + } + + n.isRegistered = true + return n, nil } // createEndpointAndPerformHandshake creates a new endpoint in connected state // and then performs the TCP 3-way handshake. +// +// The new endpoint is returned with e.mu held. func (l *listenContext) createEndpointAndPerformHandshake(s *segment, opts *header.TCPSynOptions, queue *waiter.Queue) (*endpoint, *tcpip.Error) { // Create new endpoint. irs := s.sequenceNumber @@ -289,9 +291,25 @@ func (l *listenContext) createEndpointAndPerformHandshake(s *segment, opts *head l.listenEP.mu.Lock() if l.listenEP.EndpointState() != StateListen { l.listenEP.mu.Unlock() + // Ensure we release any registrations done by the newly + // created endpoint. + ep.mu.Unlock() + ep.Close() + + // Wake up any waiters. This is strictly not required normally + // as a socket that was never accepted can't really have any + // registered waiters except when stack.Wait() is called which + // waits for all registered endpoints to stop and expects an + // EventHUp. + ep.waiterQueue.Notify(waiter.EventHUp | waiter.EventErr | waiter.EventIn | waiter.EventOut) return nil, tcpip.ErrConnectionAborted } l.addPendingEndpoint(ep) + + // Propagate any inheritable options from the listening endpoint + // to the newly created endpoint. + l.listenEP.propagateInheritableOptionsLocked(ep) + deferAccept = l.listenEP.deferAccept l.listenEP.mu.Unlock() } @@ -299,6 +317,7 @@ func (l *listenContext) createEndpointAndPerformHandshake(s *segment, opts *head // Perform the 3-way handshake. h := newPassiveHandshake(ep, seqnum.Size(ep.initialReceiveWindow()), isn, irs, opts, deferAccept) if err := h.execute(); err != nil { + ep.mu.Unlock() ep.Close() // Wake up any waiters. This is strictly not required normally // as a socket that was never accepted can't really have any @@ -312,9 +331,7 @@ func (l *listenContext) createEndpointAndPerformHandshake(s *segment, opts *head } return nil, err } - ep.mu.Lock() ep.isConnectNotified = true - ep.mu.Unlock() // Update the receive window scaling. We can't do it before the // handshake because it's possible that the peer doesn't support window @@ -366,12 +383,12 @@ func (e *endpoint) deliverAccepted(n *endpoint) { } } -// propagateInheritableOptions propagates any options set on the listening +// propagateInheritableOptionsLocked propagates any options set on the listening // endpoint to the newly created endpoint. -func (e *endpoint) propagateInheritableOptions(n *endpoint) { - e.mu.Lock() +// +// Precondition: e.mu and n.mu must be held. +func (e *endpoint) propagateInheritableOptionsLocked(n *endpoint) { n.userTimeout = e.userTimeout - e.mu.Unlock() } // handleSynSegment is called in its own goroutine once the listening endpoint @@ -382,7 +399,11 @@ func (e *endpoint) propagateInheritableOptions(n *endpoint) { // cookies to accept connections. func (e *endpoint) handleSynSegment(ctx *listenContext, s *segment, opts *header.TCPSynOptions) { defer decSynRcvdCount() - defer e.decSynRcvdCount() + defer func() { + e.mu.Lock() + e.decSynRcvdCount() + e.mu.Unlock() + }() defer s.decRef() n, err := ctx.createEndpointAndPerformHandshake(s, opts, &waiter.Queue{}) @@ -399,29 +420,21 @@ func (e *endpoint) handleSynSegment(ctx *listenContext, s *segment, opts *header } func (e *endpoint) incSynRcvdCount() bool { - e.mu.Lock() - if e.synRcvdCount >= cap(e.acceptedChan) { - e.mu.Unlock() + if e.synRcvdCount >= (cap(e.acceptedChan)) { return false } e.synRcvdCount++ - e.mu.Unlock() return true } func (e *endpoint) decSynRcvdCount() { - e.mu.Lock() e.synRcvdCount-- - e.mu.Unlock() } func (e *endpoint) acceptQueueIsFull() bool { - e.mu.Lock() if l, c := len(e.acceptedChan)+e.synRcvdCount, cap(e.acceptedChan); l >= c { - e.mu.Unlock() return true } - e.mu.Unlock() return false } @@ -559,6 +572,10 @@ func (e *endpoint) handleListenSegment(ctx *listenContext, s *segment) { return } + // Propagate any inheritable options from the listening endpoint + // to the newly created endpoint. + e.propagateInheritableOptionsLocked(n) + // clear the tsOffset for the newly created // endpoint as the Timestamp was already // randomly offset when the original SYN-ACK was @@ -593,14 +610,12 @@ func (e *endpoint) handleListenSegment(ctx *listenContext, s *segment) { func (e *endpoint) protocolListenLoop(rcvWnd seqnum.Size) *tcpip.Error { e.mu.Lock() v6only := e.v6only - e.mu.Unlock() ctx := newListenContext(e.stack, e, rcvWnd, v6only, e.NetProto) defer func() { // Mark endpoint as closed. This will prevent goroutines running // handleSynSegment() from attempting to queue new connections // to the endpoint. - e.mu.Lock() e.setEndpointState(StateClose) // close any endpoints in SYN-RCVD state. @@ -622,7 +637,10 @@ func (e *endpoint) protocolListenLoop(rcvWnd seqnum.Size) *tcpip.Error { s.AddWaker(&e.notificationWaker, wakerForNotification) s.AddWaker(&e.newSegmentWaker, wakerForNewSegment) for { - switch index, _ := s.Fetch(true); index { + e.mu.Unlock() + index, _ := s.Fetch(true) + e.mu.Lock() + switch index { case wakerForNotification: n := e.fetchNotifications() if n¬ifyClose != 0 { @@ -635,7 +653,9 @@ func (e *endpoint) protocolListenLoop(rcvWnd seqnum.Size) *tcpip.Error { s.decRef() } close(e.drainDone) + e.mu.Unlock() <-e.undrain + e.mu.Lock() } case wakerForNewSegment: diff --git a/pkg/tcpip/transport/tcp/connect.go b/pkg/tcpip/transport/tcp/connect.go index be86af502..edb37a549 100644 --- a/pkg/tcpip/transport/tcp/connect.go +++ b/pkg/tcpip/transport/tcp/connect.go @@ -61,6 +61,9 @@ const ( ) // handshake holds the state used during a TCP 3-way handshake. +// +// NOTE: handshake.ep.mu is held during handshake processing. It is released if +// we are going to block and reacquired when we start processing an event. type handshake struct { ep *endpoint state handshakeState @@ -209,9 +212,7 @@ func (h *handshake) resetToSynRcvd(iss seqnum.Value, irs seqnum.Value, opts *hea h.mss = opts.MSS h.sndWndScale = opts.WS h.deferAccept = deferAccept - h.ep.mu.Lock() h.ep.setEndpointState(StateSynRecv) - h.ep.mu.Unlock() } // checkAck checks if the ACK number, if present, of a segment received during @@ -241,9 +242,7 @@ func (h *handshake) synSentState(s *segment) *tcpip.Error { // RFC 793, page 67, states that "If the RST bit is set [and] If the ACK // was acceptable then signal the user "error: connection reset", drop // the segment, enter CLOSED state, delete TCB, and return." - h.ep.mu.Lock() h.ep.workerCleanup = true - h.ep.mu.Unlock() // Although the RFC above calls out ECONNRESET, Linux actually returns // ECONNREFUSED here so we do as well. return tcpip.ErrConnectionRefused @@ -281,9 +280,7 @@ func (h *handshake) synSentState(s *segment) *tcpip.Error { if s.flagIsSet(header.TCPFlagAck) { h.state = handshakeCompleted - h.ep.mu.Lock() h.ep.transitionToStateEstablishedLocked(h) - h.ep.mu.Unlock() h.ep.sendRaw(buffer.VectorisedView{}, header.TCPFlagAck, h.iss+1, h.ackNum, h.rcvWnd>>h.effectiveRcvWndScale()) return nil @@ -293,11 +290,9 @@ func (h *handshake) synSentState(s *segment) *tcpip.Error { // but resend our own SYN and wait for it to be acknowledged in the // SYN-RCVD state. h.state = handshakeSynRcvd - h.ep.mu.Lock() ttl := h.ep.ttl amss := h.ep.amss h.ep.setEndpointState(StateSynRecv) - h.ep.mu.Unlock() synOpts := header.TCPSynOptions{ WS: int(h.effectiveRcvWndScale()), TS: rcvSynOpts.TS, @@ -357,10 +352,6 @@ func (h *handshake) synRcvdState(s *segment) *tcpip.Error { return tcpip.ErrInvalidEndpointState } - h.ep.mu.RLock() - amss := h.ep.amss - h.ep.mu.RUnlock() - h.resetState() synOpts := header.TCPSynOptions{ WS: h.rcvWndScale, @@ -368,7 +359,7 @@ func (h *handshake) synRcvdState(s *segment) *tcpip.Error { TSVal: h.ep.timestamp(), TSEcr: h.ep.recentTimestamp(), SACKPermitted: h.ep.sackPermitted, - MSS: amss, + MSS: h.ep.amss, } h.ep.sendSynTCP(&s.route, h.ep.ID, h.ep.ttl, h.ep.sendTOS, h.flags, h.iss, h.ackNum, h.rcvWnd, synOpts) return nil @@ -399,15 +390,14 @@ func (h *handshake) synRcvdState(s *segment) *tcpip.Error { } h.state = handshakeCompleted - h.ep.mu.Lock() h.ep.transitionToStateEstablishedLocked(h) + // If the segment has data then requeue it for the receiver // to process it again once main loop is started. if s.data.Size() > 0 { s.incRef() h.ep.enqueueSegment(s) } - h.ep.mu.Unlock() return nil } @@ -493,7 +483,9 @@ func (h *handshake) resolveRoute() *tcpip.Error { } if n¬ifyDrain != 0 { close(h.ep.drainDone) + h.ep.mu.Unlock() <-h.ep.undrain + h.ep.mu.Lock() } } @@ -535,7 +527,6 @@ func (h *handshake) execute() *tcpip.Error { // Send the initial SYN segment and loop until the handshake is // completed. - h.ep.mu.Lock() h.ep.amss = calculateAdvertisedMSS(h.ep.userMSS, h.ep.route) synOpts := header.TCPSynOptions{ @@ -546,7 +537,6 @@ func (h *handshake) execute() *tcpip.Error { SACKPermitted: bool(sackEnabled), MSS: h.ep.amss, } - h.ep.mu.Unlock() // Execute is also called in a listen context so we want to make sure we // only send the TS/SACK option when we received the TS/SACK in the @@ -563,7 +553,11 @@ func (h *handshake) execute() *tcpip.Error { h.ep.sendSynTCP(&h.ep.route, h.ep.ID, h.ep.ttl, h.ep.sendTOS, h.flags, h.iss, h.ackNum, h.rcvWnd, synOpts) for h.state != handshakeCompleted { - switch index, _ := s.Fetch(true); index { + h.ep.mu.Unlock() + index, _ := s.Fetch(true) + h.ep.mu.Lock() + switch index { + case wakerForResend: timeOut *= 2 if timeOut > MaxRTO { @@ -600,7 +594,9 @@ func (h *handshake) execute() *tcpip.Error { } } close(h.ep.drainDone) + h.ep.mu.Unlock() <-h.ep.undrain + h.ep.mu.Lock() } case wakerForNewSegment: @@ -1016,7 +1012,6 @@ func (e *endpoint) handleReset(s *segment) (ok bool, err *tcpip.Error) { // except SYN-SENT, all reset (RST) segments are // validated by checking their SEQ-fields." So // we only process it if it's acceptable. - e.mu.Lock() switch e.EndpointState() { // In case of a RST in CLOSE-WAIT linux moves // the socket to closed state with an error set @@ -1040,11 +1035,9 @@ func (e *endpoint) handleReset(s *segment) (ok bool, err *tcpip.Error) { case StateCloseWait: e.transitionToStateCloseLocked() e.HardError = tcpip.ErrAborted - e.mu.Unlock() e.notifyProtocolGoroutine(notifyTickleWorker) return false, nil default: - e.mu.Unlock() // RFC 793, page 37 states that "in all states // except SYN-SENT, all reset (RST) segments are // validated by checking their SEQ-fields." So @@ -1157,9 +1150,7 @@ func (e *endpoint) handleSegment(s *segment) (cont bool, err *tcpip.Error) { // Now check if the received segment has caused us to transition // to a CLOSED state, if yes then terminate processing and do // not invoke the sender. - e.mu.RLock() state := e.state - e.mu.RUnlock() if state == StateClose { // When we get into StateClose while processing from the queue, // return immediately and let the protocolMainloop handle it. @@ -1182,9 +1173,7 @@ func (e *endpoint) handleSegment(s *segment) (cont bool, err *tcpip.Error) { // keepalive packets periodically when the connection is idle. If we don't hear // from the other side after a number of tries, we terminate the connection. func (e *endpoint) keepaliveTimerExpired() *tcpip.Error { - e.mu.RLock() userTimeout := e.userTimeout - e.mu.RUnlock() e.keepalive.Lock() if !e.keepalive.enabled || !e.keepalive.timer.checkExpiration() { @@ -1248,6 +1237,7 @@ func (e *endpoint) disableKeepaliveTimer() { // goroutine and is responsible for sending segments and handling received // segments. func (e *endpoint) protocolMainLoop(handshake bool, wakerInitDone chan<- struct{}) *tcpip.Error { + e.mu.Lock() var closeTimer *time.Timer var closeWaker sleep.Waker @@ -1269,7 +1259,6 @@ func (e *endpoint) protocolMainLoop(handshake bool, wakerInitDone chan<- struct{ } e.mu.Unlock() - e.workMu.Unlock() // When the protocol loop exits we should wake up our waiters. e.waiterQueue.Notify(waiter.EventHUp | waiter.EventErr | waiter.EventIn | waiter.EventOut) } @@ -1280,16 +1269,13 @@ func (e *endpoint) protocolMainLoop(handshake bool, wakerInitDone chan<- struct{ // completion. initialRcvWnd := e.initialReceiveWindow() h := newHandshake(e, seqnum.Size(initialRcvWnd)) - e.mu.Lock() h.ep.setEndpointState(StateSynSent) - e.mu.Unlock() if err := h.execute(); err != nil { e.lastErrorMu.Lock() e.lastError = err e.lastErrorMu.Unlock() - e.mu.Lock() e.setEndpointState(StateError) e.HardError = err @@ -1302,9 +1288,7 @@ func (e *endpoint) protocolMainLoop(handshake bool, wakerInitDone chan<- struct{ e.keepalive.timer.init(&e.keepalive.waker) defer e.keepalive.timer.cleanup() - e.mu.Lock() drained := e.drainDone != nil - e.mu.Unlock() if drained { close(e.drainDone) <-e.undrain @@ -1330,10 +1314,8 @@ func (e *endpoint) protocolMainLoop(handshake bool, wakerInitDone chan<- struct{ // This means the socket is being closed due // to the TCP-FIN-WAIT2 timeout was hit. Just // mark the socket as closed. - e.mu.Lock() e.transitionToStateCloseLocked() e.workerCleanup = true - e.mu.Unlock() return nil }, }, @@ -1388,7 +1370,6 @@ func (e *endpoint) protocolMainLoop(handshake bool, wakerInitDone chan<- struct{ } if n¬ifyClose != 0 && closeTimer == nil { - e.mu.Lock() if e.EndpointState() == StateFinWait2 && e.closed { // The socket has been closed and we are in FIN_WAIT2 // so start the FIN_WAIT2 timer. @@ -1397,7 +1378,6 @@ func (e *endpoint) protocolMainLoop(handshake bool, wakerInitDone chan<- struct{ }) e.waiterQueue.Notify(waiter.EventHUp | waiter.EventErr | waiter.EventIn | waiter.EventOut) } - e.mu.Unlock() } if n¬ifyKeepaliveChanged != 0 { @@ -1417,7 +1397,9 @@ func (e *endpoint) protocolMainLoop(handshake bool, wakerInitDone chan<- struct{ // Only block the worker if the endpoint // is not in closed state or error state. close(e.drainDone) + e.mu.Unlock() <-e.undrain + e.mu.Lock() } } @@ -1460,7 +1442,6 @@ func (e *endpoint) protocolMainLoop(handshake bool, wakerInitDone chan<- struct{ } e.rcvListMu.Unlock() - e.mu.Lock() if e.workerCleanup { e.notifyProtocolGoroutine(notifyClose) } @@ -1468,7 +1449,6 @@ func (e *endpoint) protocolMainLoop(handshake bool, wakerInitDone chan<- struct{ // Main loop. Handle segments until both send and receive ends of the // connection have completed. cleanupOnError := func(err *tcpip.Error) { - e.mu.Lock() e.workerCleanup = true if err != nil { e.resetConnectionLocked(err) @@ -1480,16 +1460,11 @@ func (e *endpoint) protocolMainLoop(handshake bool, wakerInitDone chan<- struct{ loop: for e.EndpointState() != StateTimeWait && e.EndpointState() != StateClose && e.EndpointState() != StateError { e.mu.Unlock() - e.workMu.Unlock() v, _ := s.Fetch(true) - e.workMu.Lock() + e.mu.Lock() - // We need to double check here because the notification maybe + // We need to double check here because the notification may be // stale by the time we got around to processing it. - // - // NOTE: since we now hold the workMu the processors cannot - // change the state of the endpoint so it's safe to proceed - // after this check. switch e.EndpointState() { case StateError: // If the endpoint has already transitioned to an ERROR @@ -1502,21 +1477,17 @@ loop: case StateTimeWait: fallthrough case StateClose: - e.mu.Lock() break loop default: if err := funcs[v].f(); err != nil { cleanupOnError(err) return nil } - e.mu.Lock() } } - state := e.EndpointState() - e.mu.Unlock() var reuseTW func() - if state == StateTimeWait { + if e.EndpointState() == StateTimeWait { // Disable close timer as we now entering real TIME_WAIT. if closeTimer != nil { closeTimer.Stop() @@ -1526,14 +1497,11 @@ loop: s.Done() // Wake up any waiters before we enter TIME_WAIT. e.waiterQueue.Notify(waiter.EventHUp | waiter.EventErr | waiter.EventIn | waiter.EventOut) - e.mu.Lock() e.workerCleanup = true - e.mu.Unlock() reuseTW = e.doTimeWait() } // Mark endpoint as closed. - e.mu.Lock() if e.EndpointState() != StateError { e.transitionToStateCloseLocked() } @@ -1649,9 +1617,9 @@ func (e *endpoint) doTimeWait() (twReuse func()) { defer timeWaitTimer.Stop() for { - e.workMu.Unlock() + e.mu.Unlock() v, _ := s.Fetch(true) - e.workMu.Lock() + e.mu.Lock() switch v { case newSegment: extendTimeWait, reuseTW := e.handleTimeWaitSegments() @@ -1674,7 +1642,9 @@ func (e *endpoint) doTimeWait() (twReuse func()) { e.handleTimeWaitSegments() } close(e.drainDone) + e.mu.Unlock() <-e.undrain + e.mu.Lock() return nil } case timeWaitDone: diff --git a/pkg/tcpip/transport/tcp/dispatcher.go b/pkg/tcpip/transport/tcp/dispatcher.go index d792b07d6..90ac956a9 100644 --- a/pkg/tcpip/transport/tcp/dispatcher.go +++ b/pkg/tcpip/transport/tcp/dispatcher.go @@ -128,7 +128,7 @@ func (p *processor) handleSegments() { continue } - if !ep.workMu.TryLock() { + if !ep.mu.TryLock() { ep.newSegmentWaker.Assert() continue } @@ -138,12 +138,10 @@ func (p *processor) handleSegments() { if err := ep.handleSegments(true /* fastPath */); err != nil || ep.EndpointState() == StateClose { // Send any active resets if required. if err != nil { - ep.mu.Lock() ep.resetConnectionLocked(err) - ep.mu.Unlock() } ep.notifyProtocolGoroutine(notifyTickleWorker) - ep.workMu.Unlock() + ep.mu.Unlock() continue } @@ -151,7 +149,7 @@ func (p *processor) handleSegments() { p.epQ.enqueue(ep) } - ep.workMu.Unlock() + ep.mu.Unlock() } } } diff --git a/pkg/tcpip/transport/tcp/endpoint.go b/pkg/tcpip/transport/tcp/endpoint.go index 5187a5e25..eb8a9d73e 100644 --- a/pkg/tcpip/transport/tcp/endpoint.go +++ b/pkg/tcpip/transport/tcp/endpoint.go @@ -18,6 +18,7 @@ import ( "encoding/binary" "fmt" "math" + "runtime" "strings" "sync/atomic" "time" @@ -33,7 +34,6 @@ import ( "gvisor.dev/gvisor/pkg/tcpip/ports" "gvisor.dev/gvisor/pkg/tcpip/seqnum" "gvisor.dev/gvisor/pkg/tcpip/stack" - "gvisor.dev/gvisor/pkg/tmutex" "gvisor.dev/gvisor/pkg/waiter" ) @@ -283,6 +283,37 @@ func (*EndpointInfo) IsEndpointInfo() {} // synchronized. The protocol implementation, however, runs in a single // goroutine. // +// Each endpoint has a few mutexes: +// +// e.mu -> Primary mutex for an endpoint must be held for all operations except +// in e.Readiness where acquiring it will result in a deadlock in epoll +// implementation. +// +// The following three mutexes can be acquired independent of e.mu but if +// acquired with e.mu then e.mu must be acquired first. +// +// e.rcvListMu -> Protects the rcvList and associated fields. +// e.sndBufMu -> Protects the sndQueue and associated fields. +// e.lastErrorMu -> Protects the lastError field. +// +// LOCKING/UNLOCKING of the endpoint. The locking of an endpoint is different +// based on the context in which the lock is acquired. In the syscall context +// e.LockUser/e.UnlockUser should be used and when doing background processing +// e.mu.Lock/e.mu.Unlock should be used. The distinction is described below +// in brief. +// +// The reason for this locking behaviour is to avoid wakeups to handle packets. +// In cases where the endpoint is already locked the background processor can +// queue the packet up and go its merry way and the lock owner will eventually +// process the backlog when releasing the lock. Similarly when acquiring the +// lock from say a syscall goroutine we can implement a bit of spinning if we +// know that the lock is not held by another syscall goroutine. Background +// processors should never hold the lock for long and we can avoid an expensive +// sleep/wakeup by spinning for a shortwhile. +// +// For more details please see the detailed documentation on +// e.LockUser/e.UnlockUser methods. +// // +stateify savable type endpoint struct { EndpointInfo @@ -299,12 +330,6 @@ type endpoint struct { // Precondition: epQueue.mu must be held to read/write this field.. pendingProcessing bool `state:"nosave"` - // workMu is used to arbitrate which goroutine may perform protocol - // work. Only the main protocol goroutine is expected to call Lock() on - // it, but other goroutines (e.g., send) may call TryLock() to eagerly - // perform work without having to wait for the main one to wake up. - workMu tmutex.Mutex `state:"nosave"` - // The following fields are initialized at creation time and do not // change throughout the lifetime of the endpoint. stack *stack.Stack `state:"manual"` @@ -330,15 +355,11 @@ type endpoint struct { rcvBufSize int rcvBufUsed int rcvAutoParams rcvBufAutoTuneParams - // zeroWindow indicates that the window was closed due to receive buffer - // space being filled up. This is set by the worker goroutine before - // moving a segment to the rcvList. This setting is cleared by the - // endpoint when a Read() call reads enough data for the new window to - // be non-zero. - zeroWindow bool - // The following fields are protected by the mutex. - mu sync.RWMutex `state:"nosave"` + // mu protects all endpoint fields unless documented otherwise. mu must + // be acquired before interacting with the endpoint fields. + mu sync.Mutex `state:"nosave"` + ownedByUser uint32 // state must be read/set using the EndpointState()/setEndpointState() methods. state EndpointState `state:".(EndpointState)"` @@ -583,14 +604,93 @@ func calculateAdvertisedMSS(userMSS uint16, r stack.Route) uint16 { return maxMSS } +// LockUser tries to lock e.mu and if it fails it will check if the lock is held +// by another syscall goroutine. If yes, then it will goto sleep waiting for the +// lock to be released, if not then it will spin till it acquires the lock or +// another syscall goroutine acquires it in which case it will goto sleep as +// described above. +// +// The assumption behind spinning here being that background packet processing +// should not be holding the lock for long and spinning reduces latency as we +// avoid an expensive sleep/wakeup of of the syscall goroutine). +func (e *endpoint) LockUser() { + for { + // Try first if the sock is locked then check if it's owned + // by another user goroutine if not then we spin, otherwise + // we just goto sleep on the Lock() and wait. + if !e.mu.TryLock() { + // If socket is owned by the user then just goto sleep + // as the lock could be held for a reasonably long time. + if atomic.LoadUint32(&e.ownedByUser) == 1 { + e.mu.Lock() + atomic.StoreUint32(&e.ownedByUser, 1) + return + } + // Spin but yield the processor since the lower half + // should yield the lock soon. + runtime.Gosched() + continue + } + atomic.StoreUint32(&e.ownedByUser, 1) + return + } +} + +// UnlockUser will check if there are any segments already queued for processing +// and process any such segments before unlocking e.mu. This is required because +// we when packets arrive and endpoint lock is already held then such packets +// are queued up to be processed. If the lock is held by the endpoint goroutine +// then it will process these packets but if the lock is instead held by the +// syscall goroutine then we can have the syscall goroutine process the backlog +// before unlocking. +// +// This avoids an unnecessary wakeup of the endpoint protocol goroutine for the +// endpoint. It's also required eventually when we get rid of the endpoint +// protocol goroutine altogether. +// +// Precondition: e.LockUser() must have been called before calling e.UnlockUser() +func (e *endpoint) UnlockUser() { + // Lock segment queue before checking so that we avoid a race where + // segments can be queued between the time we check if queue is empty + // and actually unlock the endpoint mutex. + for { + e.segmentQueue.mu.Lock() + if e.segmentQueue.emptyLocked() { + if atomic.SwapUint32(&e.ownedByUser, 0) != 1 { + panic("e.UnlockUser() called without calling e.LockUser()") + } + e.mu.Unlock() + e.segmentQueue.mu.Unlock() + return + } + e.segmentQueue.mu.Unlock() + + switch e.EndpointState() { + case StateEstablished: + if err := e.handleSegments(true /* fastPath */); err != nil { + e.notifyProtocolGoroutine(notifyTickleWorker) + } + default: + // Since we are waking the endpoint goroutine here just unlock + // and let it process the queued segments. + e.newSegmentWaker.Assert() + if atomic.SwapUint32(&e.ownedByUser, 0) != 1 { + panic("e.UnlockUser() called without calling e.LockUser()") + } + e.mu.Unlock() + return + } + } +} + // StopWork halts packet processing. Only to be used in tests. func (e *endpoint) StopWork() { - e.workMu.Lock() + e.mu.Lock() } // ResumeWork resumes packet processing. Only to be used in tests. func (e *endpoint) ResumeWork() { - e.workMu.Unlock() + e.mu.Unlock() } // setEndpointState updates the state of the endpoint to state atomically. This @@ -709,8 +809,6 @@ func newEndpoint(s *stack.Stack, netProto tcpip.NetworkProtocolNumber, waiterQue } e.segmentQueue.setLimit(MaxUnprocessedSegments) - e.workMu.Init() - e.workMu.Lock() e.tsOffset = timeStampOffset() return e @@ -721,9 +819,6 @@ func newEndpoint(s *stack.Stack, netProto tcpip.NetworkProtocolNumber, waiterQue func (e *endpoint) Readiness(mask waiter.EventMask) waiter.EventMask { result := waiter.EventMask(0) - e.mu.RLock() - defer e.mu.RUnlock() - switch e.EndpointState() { case StateInitial, StateBound, StateConnecting, StateSynSent, StateSynRecv: // Ready for nothing. @@ -823,20 +918,22 @@ func (e *endpoint) Abort() { // with it. It must be called only once and with no other concurrent calls to // the endpoint. func (e *endpoint) Close() { - e.mu.Lock() - closed := e.closed - e.closed = true - e.mu.Unlock() - if closed { + e.LockUser() + defer e.UnlockUser() + if e.closed { return } // Issue a shutdown so that the peer knows we won't send any more data // if we're connected, or stop accepting if we're listening. - e.Shutdown(tcpip.ShutdownWrite | tcpip.ShutdownRead) - - e.mu.Lock() + e.shutdownLocked(tcpip.ShutdownWrite | tcpip.ShutdownRead) + e.closeNoShutdownLocked() +} +// closeNoShutdown closes the endpoint without doing a full shutdown. This is +// used when a connection needs to be aborted with a RST and we want to skip +// a full 4 way TCP shutdown. +func (e *endpoint) closeNoShutdownLocked() { // For listening sockets, we always release ports inline so that they // are immediately available for reuse after Close() is called. If also // registered, we unregister as well otherwise the next user would fail @@ -853,6 +950,8 @@ func (e *endpoint) Close() { e.boundPortFlags = ports.Flags{} } + // Mark endpoint as closed. + e.closed = true // Either perform the local cleanup or kick the worker to make sure it // knows it needs to cleanup. switch e.EndpointState() { @@ -873,8 +972,6 @@ func (e *endpoint) Close() { // goroutine terminates. e.notifyProtocolGoroutine(notifyClose) } - - e.mu.Unlock() } // closePendingAcceptableConnections closes all connections that have completed @@ -909,7 +1006,6 @@ func (e *endpoint) closePendingAcceptableConnectionsLocked() { // after Close() is called and the worker goroutine (if any) is done with its // work. func (e *endpoint) cleanupLocked() { - // Close all endpoints that might have been accepted by TCP but not by // the client. if e.acceptedChan != nil { @@ -954,18 +1050,18 @@ func (e *endpoint) initialReceiveWindow() int { // ModerateRecvBuf adjusts the receive buffer and the advertised window // based on the number of bytes copied to user space. func (e *endpoint) ModerateRecvBuf(copied int) { - e.mu.RLock() + e.LockUser() + defer e.UnlockUser() + e.rcvListMu.Lock() if e.rcvAutoParams.disabled { e.rcvListMu.Unlock() - e.mu.RUnlock() return } now := time.Now() if rtt := e.rcvAutoParams.rtt; rtt == 0 || now.Sub(e.rcvAutoParams.measureTime) < rtt { e.rcvAutoParams.copied += copied e.rcvListMu.Unlock() - e.mu.RUnlock() return } prevRTTCopied := e.rcvAutoParams.copied + copied @@ -1021,7 +1117,6 @@ func (e *endpoint) ModerateRecvBuf(copied int) { e.rcvAutoParams.measureTime = now e.rcvAutoParams.copied = 0 e.rcvListMu.Unlock() - e.mu.RUnlock() } // IPTables implements tcpip.Endpoint.IPTables. @@ -1031,7 +1126,7 @@ func (e *endpoint) IPTables() (iptables.IPTables, error) { // Read reads data from the endpoint. func (e *endpoint) Read(*tcpip.FullAddress) (buffer.View, tcpip.ControlMessages, *tcpip.Error) { - e.mu.RLock() + e.LockUser() // The endpoint can be read if it's connected, or if it's already closed // but has some pending unread data. Also note that a RST being received // would cause the state to become StateError so we should allow the @@ -1041,7 +1136,7 @@ func (e *endpoint) Read(*tcpip.FullAddress) (buffer.View, tcpip.ControlMessages, if s := e.EndpointState(); !s.connected() && s != StateClose && bufUsed == 0 { e.rcvListMu.Unlock() he := e.HardError - e.mu.RUnlock() + e.UnlockUser() if s == StateError { return buffer.View{}, tcpip.ControlMessages{}, he } @@ -1051,7 +1146,7 @@ func (e *endpoint) Read(*tcpip.FullAddress) (buffer.View, tcpip.ControlMessages, v, err := e.readLocked() e.rcvListMu.Unlock() - e.mu.RUnlock() + e.UnlockUser() if err == tcpip.ErrClosedForReceive { e.stats.ReadErrors.ReadClosed.Increment() @@ -1124,13 +1219,13 @@ func (e *endpoint) Write(p tcpip.Payloader, opts tcpip.WriteOptions) (int64, <-c // (without the MSG_FASTOPEN flag). Corking is unimplemented, so opts.More // and opts.EndOfRecord are also ignored. - e.mu.RLock() + e.LockUser() e.sndBufMu.Lock() avail, err := e.isEndpointWritableLocked() if err != nil { e.sndBufMu.Unlock() - e.mu.RUnlock() + e.UnlockUser() e.stats.WriteErrors.WriteClosed.Increment() return 0, nil, err } @@ -1142,113 +1237,68 @@ func (e *endpoint) Write(p tcpip.Payloader, opts tcpip.WriteOptions) (int64, <-c // are copying data in. if !opts.Atomic { e.sndBufMu.Unlock() - e.mu.RUnlock() + e.UnlockUser() } // Fetch data. v, perr := p.Payload(avail) if perr != nil || len(v) == 0 { - if opts.Atomic { // See above. + // Note that perr may be nil if len(v) == 0. + if opts.Atomic { e.sndBufMu.Unlock() - e.mu.RUnlock() + e.UnlockUser() } - // Note that perr may be nil if len(v) == 0. return 0, nil, perr } - if opts.Atomic { + queueAndSend := func() (int64, <-chan struct{}, *tcpip.Error) { // Add data to the send queue. s := newSegmentFromView(&e.route, e.ID, v) e.sndBufUsed += len(v) e.sndBufInQueue += seqnum.Size(len(v)) e.sndQueue.PushBack(s) e.sndBufMu.Unlock() - // Release the endpoint lock to prevent deadlocks due to lock - // order inversion when acquiring workMu. - e.mu.RUnlock() - } - if e.workMu.TryLock() { - // Since we released locks in between it's possible that the - // endpoint transitioned to a CLOSED/ERROR states so make - // sure endpoint is still writable before trying to write. - if !opts.Atomic { // See above. - e.mu.RLock() - e.sndBufMu.Lock() - - // Because we released the lock before copying, check state again - // to make sure the endpoint is still in a valid state for a write. - avail, err = e.isEndpointWritableLocked() - if err != nil { - e.sndBufMu.Unlock() - e.mu.RUnlock() - e.stats.WriteErrors.WriteClosed.Increment() - return 0, nil, err - } - - // Discard any excess data copied in due to avail being reduced due - // to a simultaneous write call to the socket. - if avail < len(v) { - v = v[:avail] - } - // Add data to the send queue. - s := newSegmentFromView(&e.route, e.ID, v) - e.sndBufUsed += len(v) - e.sndBufInQueue += seqnum.Size(len(v)) - e.sndQueue.PushBack(s) - e.sndBufMu.Unlock() - // Release the endpoint lock to prevent deadlocks due to lock - // order inversion when acquiring workMu. - e.mu.RUnlock() - - } // Do the work inline. e.handleWrite() - e.workMu.Unlock() - } else { - if !opts.Atomic { // See above. - e.mu.RLock() - e.sndBufMu.Lock() + e.UnlockUser() + return int64(len(v)), nil, nil + } - // Because we released the lock before copying, check state again - // to make sure the endpoint is still in a valid state for a write. - avail, err = e.isEndpointWritableLocked() - if err != nil { - e.sndBufMu.Unlock() - e.mu.RUnlock() - e.stats.WriteErrors.WriteClosed.Increment() - return 0, nil, err - } + if opts.Atomic { + // Locks released in queueAndSend() + return queueAndSend() + } - // Discard any excess data copied in due to avail being reduced due - // to a simultaneous write call to the socket. - if avail < len(v) { - v = v[:avail] - } - // Add data to the send queue. - s := newSegmentFromView(&e.route, e.ID, v) - e.sndBufUsed += len(v) - e.sndBufInQueue += seqnum.Size(len(v)) - e.sndQueue.PushBack(s) - e.sndBufMu.Unlock() - // Release the endpoint lock to prevent deadlocks due to lock - // order inversion when acquiring workMu. - e.mu.RUnlock() + // Since we released locks in between it's possible that the + // endpoint transitioned to a CLOSED/ERROR states so make + // sure endpoint is still writable before trying to write. + e.LockUser() + e.sndBufMu.Lock() + avail, err = e.isEndpointWritableLocked() + if err != nil { + e.sndBufMu.Unlock() + e.UnlockUser() + e.stats.WriteErrors.WriteClosed.Increment() + return 0, nil, err + } - } - // Let the protocol goroutine do the work. - e.sndWaker.Assert() + // Discard any excess data copied in due to avail being reduced due + // to a simultaneous write call to the socket. + if avail < len(v) { + v = v[:avail] } - return int64(len(v)), nil, nil + // Locks released in queueAndSend() + return queueAndSend() } // Peek reads data without consuming it from the endpoint. // // This method does not block if there is no data pending. func (e *endpoint) Peek(vec [][]byte) (int64, tcpip.ControlMessages, *tcpip.Error) { - e.mu.RLock() - defer e.mu.RUnlock() + e.LockUser() + defer e.UnlockUser() // The endpoint can be read if it's connected, or if it's already closed // but has some pending unread data. @@ -1339,6 +1389,9 @@ func (e *endpoint) windowCrossedACKThresholdLocked(deltaBefore int) (crossed boo // SetSockOptBool sets a socket option. func (e *endpoint) SetSockOptBool(opt tcpip.SockOptBool, v bool) *tcpip.Error { + e.LockUser() + defer e.UnlockUser() + switch opt { case tcpip.V6OnlyOption: // We only recognize this option on v6 endpoints. @@ -1346,9 +1399,6 @@ func (e *endpoint) SetSockOptBool(opt tcpip.SockOptBool, v bool) *tcpip.Error { return tcpip.ErrInvalidEndpointState } - e.mu.Lock() - defer e.mu.Unlock() - // We only allow this to be set when we're in the initial state. if e.EndpointState() != StateInitial { return tcpip.ErrInvalidEndpointState @@ -1379,7 +1429,7 @@ func (e *endpoint) SetSockOptInt(opt tcpip.SockOptInt, v int) *tcpip.Error { mask := uint32(notifyReceiveWindowChanged) - e.mu.RLock() + e.LockUser() e.rcvListMu.Lock() // Make sure the receive buffer size allows us to send a @@ -1409,8 +1459,9 @@ func (e *endpoint) SetSockOptInt(opt tcpip.SockOptInt, v int) *tcpip.Error { if crossed, above := e.windowCrossedACKThresholdLocked(availAfter - availBefore); crossed && above { mask |= notifyNonZeroReceiveWindow } + e.rcvListMu.Unlock() - e.mu.RUnlock() + e.UnlockUser() e.notifyProtocolGoroutine(mask) return nil @@ -1466,15 +1517,15 @@ func (e *endpoint) SetSockOpt(opt interface{}) *tcpip.Error { return nil case tcpip.ReuseAddressOption: - e.mu.Lock() + e.LockUser() e.reuseAddr = v != 0 - e.mu.Unlock() + e.UnlockUser() return nil case tcpip.ReusePortOption: - e.mu.Lock() + e.LockUser() e.reusePort = v != 0 - e.mu.Unlock() + e.UnlockUser() return nil case tcpip.BindToDeviceOption: @@ -1482,9 +1533,9 @@ func (e *endpoint) SetSockOpt(opt interface{}) *tcpip.Error { if id != 0 && !e.stack.HasNIC(id) { return tcpip.ErrUnknownDevice } - e.mu.Lock() + e.LockUser() e.bindToDevice = id - e.mu.Unlock() + e.UnlockUser() return nil case tcpip.QuickAckOption: @@ -1500,16 +1551,16 @@ func (e *endpoint) SetSockOpt(opt interface{}) *tcpip.Error { if userMSS < header.TCPMinimumMSS || userMSS > header.TCPMaximumMSS { return tcpip.ErrInvalidOptionValue } - e.mu.Lock() + e.LockUser() e.userMSS = uint16(userMSS) - e.mu.Unlock() + e.UnlockUser() e.notifyProtocolGoroutine(notifyMSSChanged) return nil case tcpip.TTLOption: - e.mu.Lock() + e.LockUser() e.ttl = uint8(v) - e.mu.Unlock() + e.UnlockUser() return nil case tcpip.KeepaliveEnabledOption: @@ -1541,15 +1592,15 @@ func (e *endpoint) SetSockOpt(opt interface{}) *tcpip.Error { return nil case tcpip.TCPUserTimeoutOption: - e.mu.Lock() + e.LockUser() e.userTimeout = time.Duration(v) - e.mu.Unlock() + e.UnlockUser() return nil case tcpip.BroadcastOption: - e.mu.Lock() + e.LockUser() e.broadcast = v != 0 - e.mu.Unlock() + e.UnlockUser() return nil case tcpip.CongestionControlOption: @@ -1563,22 +1614,16 @@ func (e *endpoint) SetSockOpt(opt interface{}) *tcpip.Error { availCC := strings.Split(string(avail), " ") for _, cc := range availCC { if v == tcpip.CongestionControlOption(cc) { - // Acquire the work mutex as we may need to - // reinitialize the congestion control state. - e.mu.Lock() + e.LockUser() state := e.EndpointState() e.cc = v - e.mu.Unlock() switch state { case StateEstablished: - e.workMu.Lock() - e.mu.Lock() if e.EndpointState() == state { e.snd.cc = e.snd.initCongestionControl(e.cc) } - e.mu.Unlock() - e.workMu.Unlock() } + e.UnlockUser() return nil } } @@ -1588,23 +1633,23 @@ func (e *endpoint) SetSockOpt(opt interface{}) *tcpip.Error { return tcpip.ErrNoSuchFile case tcpip.IPv4TOSOption: - e.mu.Lock() + e.LockUser() // TODO(gvisor.dev/issue/995): ECN is not currently supported, // ignore the bits for now. e.sendTOS = uint8(v) & ^uint8(inetECNMask) - e.mu.Unlock() + e.UnlockUser() return nil case tcpip.IPv6TrafficClassOption: - e.mu.Lock() + e.LockUser() // TODO(gvisor.dev/issue/995): ECN is not currently supported, // ignore the bits for now. e.sendTOS = uint8(v) & ^uint8(inetECNMask) - e.mu.Unlock() + e.UnlockUser() return nil case tcpip.TCPLingerTimeoutOption: - e.mu.Lock() + e.LockUser() if v < 0 { // Same as effectively disabling TCPLinger timeout. v = 0 @@ -1622,16 +1667,16 @@ func (e *endpoint) SetSockOpt(opt interface{}) *tcpip.Error { v = stkTCPLingerTimeout } e.tcpLingerTimeout = time.Duration(v) - e.mu.Unlock() + e.UnlockUser() return nil case tcpip.TCPDeferAcceptOption: - e.mu.Lock() + e.LockUser() if time.Duration(v) > MaxRTO { v = tcpip.TCPDeferAcceptOption(MaxRTO) } e.deferAccept = time.Duration(v) - e.mu.Unlock() + e.UnlockUser() return nil default: @@ -1641,8 +1686,8 @@ func (e *endpoint) SetSockOpt(opt interface{}) *tcpip.Error { // readyReceiveSize returns the number of bytes ready to be received. func (e *endpoint) readyReceiveSize() (int, *tcpip.Error) { - e.mu.RLock() - defer e.mu.RUnlock() + e.LockUser() + defer e.UnlockUser() // The endpoint cannot be in listen state. if e.EndpointState() == StateListen { @@ -1664,9 +1709,9 @@ func (e *endpoint) GetSockOptBool(opt tcpip.SockOptBool) (bool, *tcpip.Error) { return false, tcpip.ErrUnknownProtocolOption } - e.mu.Lock() + e.LockUser() v := e.v6only - e.mu.Unlock() + e.UnlockUser() return v, nil } @@ -1730,9 +1775,9 @@ func (e *endpoint) GetSockOpt(opt interface{}) *tcpip.Error { return nil case *tcpip.ReuseAddressOption: - e.mu.RLock() + e.LockUser() v := e.reuseAddr - e.mu.RUnlock() + e.UnlockUser() *o = 0 if v { @@ -1741,9 +1786,9 @@ func (e *endpoint) GetSockOpt(opt interface{}) *tcpip.Error { return nil case *tcpip.ReusePortOption: - e.mu.RLock() + e.LockUser() v := e.reusePort - e.mu.RUnlock() + e.UnlockUser() *o = 0 if v { @@ -1752,9 +1797,9 @@ func (e *endpoint) GetSockOpt(opt interface{}) *tcpip.Error { return nil case *tcpip.BindToDeviceOption: - e.mu.RLock() + e.LockUser() *o = tcpip.BindToDeviceOption(e.bindToDevice) - e.mu.RUnlock() + e.UnlockUser() return nil case *tcpip.QuickAckOption: @@ -1765,16 +1810,16 @@ func (e *endpoint) GetSockOpt(opt interface{}) *tcpip.Error { return nil case *tcpip.TTLOption: - e.mu.Lock() + e.LockUser() *o = tcpip.TTLOption(e.ttl) - e.mu.Unlock() + e.UnlockUser() return nil case *tcpip.TCPInfoOption: *o = tcpip.TCPInfoOption{} - e.mu.RLock() + e.LockUser() snd := e.snd - e.mu.RUnlock() + e.UnlockUser() if snd != nil { snd.rtt.Lock() o.RTT = snd.rtt.srtt @@ -1813,9 +1858,9 @@ func (e *endpoint) GetSockOpt(opt interface{}) *tcpip.Error { return nil case *tcpip.TCPUserTimeoutOption: - e.mu.Lock() + e.LockUser() *o = tcpip.TCPUserTimeoutOption(e.userTimeout) - e.mu.Unlock() + e.UnlockUser() return nil case *tcpip.OutOfBandInlineOption: @@ -1824,9 +1869,9 @@ func (e *endpoint) GetSockOpt(opt interface{}) *tcpip.Error { return nil case *tcpip.BroadcastOption: - e.mu.Lock() + e.LockUser() v := e.broadcast - e.mu.Unlock() + e.UnlockUser() *o = 0 if v { @@ -1835,33 +1880,33 @@ func (e *endpoint) GetSockOpt(opt interface{}) *tcpip.Error { return nil case *tcpip.CongestionControlOption: - e.mu.Lock() + e.LockUser() *o = e.cc - e.mu.Unlock() + e.UnlockUser() return nil case *tcpip.IPv4TOSOption: - e.mu.RLock() + e.LockUser() *o = tcpip.IPv4TOSOption(e.sendTOS) - e.mu.RUnlock() + e.UnlockUser() return nil case *tcpip.IPv6TrafficClassOption: - e.mu.RLock() + e.LockUser() *o = tcpip.IPv6TrafficClassOption(e.sendTOS) - e.mu.RUnlock() + e.UnlockUser() return nil case *tcpip.TCPLingerTimeoutOption: - e.mu.Lock() + e.LockUser() *o = tcpip.TCPLingerTimeoutOption(e.tcpLingerTimeout) - e.mu.Unlock() + e.UnlockUser() return nil case *tcpip.TCPDeferAcceptOption: - e.mu.Lock() + e.LockUser() *o = tcpip.TCPDeferAcceptOption(e.deferAccept) - e.mu.Unlock() + e.UnlockUser() return nil default: @@ -1901,8 +1946,8 @@ func (e *endpoint) Connect(addr tcpip.FullAddress) *tcpip.Error { // yet accepted by the app, they are restored without running the main goroutine // here. func (e *endpoint) connect(addr tcpip.FullAddress, handshake bool, run bool) *tcpip.Error { - e.mu.Lock() - defer e.mu.Unlock() + e.LockUser() + defer e.UnlockUser() connectingAddr := addr.Addr @@ -2071,9 +2116,13 @@ func (*endpoint) ConnectEndpoint(tcpip.Endpoint) *tcpip.Error { // Shutdown closes the read and/or write end of the endpoint connection to its // peer. func (e *endpoint) Shutdown(flags tcpip.ShutdownFlags) *tcpip.Error { - e.mu.Lock() + e.LockUser() + defer e.UnlockUser() + return e.shutdownLocked(flags) +} + +func (e *endpoint) shutdownLocked(flags tcpip.ShutdownFlags) *tcpip.Error { e.shutdownFlags |= flags - finQueued := false switch { case e.EndpointState().connected(): // Close for read. @@ -2087,24 +2136,9 @@ func (e *endpoint) Shutdown(flags tcpip.ShutdownFlags) *tcpip.Error { // If we're fully closed and we have unread data we need to abort // the connection with a RST. if (e.shutdownFlags&tcpip.ShutdownWrite) != 0 && rcvBufUsed > 0 { - e.mu.Unlock() - // Try to send an active reset immediately if the - // work mutex is available. - if e.workMu.TryLock() { - e.mu.Lock() - // We need to double check here to make - // sure worker has not transitioned the - // endpoint out of a connected state - // before trying to send a reset. - if e.EndpointState().connected() { - e.resetConnectionLocked(tcpip.ErrConnectionAborted) - e.notifyProtocolGoroutine(notifyTickleWorker) - } - e.mu.Unlock() - e.workMu.Unlock() - } else { - e.notifyProtocolGoroutine(notifyReset) - } + e.resetConnectionLocked(tcpip.ErrConnectionAborted) + // Wake up worker to terminate loop. + e.notifyProtocolGoroutine(notifyTickleWorker) return nil } } @@ -2116,42 +2150,32 @@ func (e *endpoint) Shutdown(flags tcpip.ShutdownFlags) *tcpip.Error { // Already closed. e.sndBufMu.Unlock() if e.EndpointState() == StateTimeWait { - e.mu.Unlock() return tcpip.ErrNotConnected } - break + return nil } // Queue fin segment. s := newSegmentFromView(&e.route, e.ID, nil) e.sndQueue.PushBack(s) e.sndBufInQueue++ - finQueued = true // Mark endpoint as closed. e.sndClosed = true e.sndBufMu.Unlock() + e.handleClose() } + return nil case e.EndpointState() == StateListen: // Tell protocolListenLoop to stop. if flags&tcpip.ShutdownRead != 0 { e.notifyProtocolGoroutine(notifyClose) } + return nil + default: - e.mu.Unlock() return tcpip.ErrNotConnected } - e.mu.Unlock() - if finQueued { - if e.workMu.TryLock() { - e.handleClose() - e.workMu.Unlock() - } else { - // Tell protocol goroutine to close. - e.sndCloseWaker.Assert() - } - } - return nil } // Listen puts the endpoint in "listen" mode, which allows it to accept @@ -2166,8 +2190,8 @@ func (e *endpoint) Listen(backlog int) *tcpip.Error { } func (e *endpoint) listen(backlog int) *tcpip.Error { - e.mu.Lock() - defer e.mu.Unlock() + e.LockUser() + defer e.UnlockUser() // Allow the backlog to be adjusted if the endpoint is not shutting down. // When the endpoint shuts down, it sets workerCleanup to true, and from @@ -2229,7 +2253,6 @@ func (e *endpoint) listen(backlog int) *tcpip.Error { // startAcceptedLoop sets up required state and starts a goroutine with the // main loop for accepted connections. func (e *endpoint) startAcceptedLoop() { - e.mu.Lock() e.workerRunning = true e.mu.Unlock() wakerInitDone := make(chan struct{}) @@ -2240,8 +2263,8 @@ func (e *endpoint) startAcceptedLoop() { // Accept returns a new endpoint if a peer has established a connection // to an endpoint previously set to listen mode. func (e *endpoint) Accept() (tcpip.Endpoint, *waiter.Queue, *tcpip.Error) { - e.mu.RLock() - defer e.mu.RUnlock() + e.LockUser() + defer e.UnlockUser() // Endpoint must be in listen state before it can accept connections. if e.EndpointState() != StateListen { @@ -2260,8 +2283,8 @@ func (e *endpoint) Accept() (tcpip.Endpoint, *waiter.Queue, *tcpip.Error) { // Bind binds the endpoint to a specific local port and optionally address. func (e *endpoint) Bind(addr tcpip.FullAddress) (err *tcpip.Error) { - e.mu.Lock() - defer e.mu.Unlock() + e.LockUser() + defer e.UnlockUser() return e.bindLocked(addr) } @@ -2339,8 +2362,8 @@ func (e *endpoint) bindLocked(addr tcpip.FullAddress) (err *tcpip.Error) { // GetLocalAddress returns the address to which the endpoint is bound. func (e *endpoint) GetLocalAddress() (tcpip.FullAddress, *tcpip.Error) { - e.mu.RLock() - defer e.mu.RUnlock() + e.LockUser() + defer e.UnlockUser() return tcpip.FullAddress{ Addr: e.ID.LocalAddress, @@ -2351,8 +2374,8 @@ func (e *endpoint) GetLocalAddress() (tcpip.FullAddress, *tcpip.Error) { // GetRemoteAddress returns the address to which the endpoint is connected. func (e *endpoint) GetRemoteAddress() (tcpip.FullAddress, *tcpip.Error) { - e.mu.RLock() - defer e.mu.RUnlock() + e.LockUser() + defer e.UnlockUser() if !e.EndpointState().connected() { return tcpip.FullAddress{}, tcpip.ErrNotConnected @@ -2419,7 +2442,6 @@ func (e *endpoint) updateSndBufferUsage(v int) { // to be read, or when the connection is closed for receiving (in which case // s will be nil). func (e *endpoint) readyToRead(s *segment) { - e.mu.RLock() e.rcvListMu.Lock() if s != nil { s.incRef() @@ -2434,7 +2456,6 @@ func (e *endpoint) readyToRead(s *segment) { e.rcvClosed = true } e.rcvListMu.Unlock() - e.mu.RUnlock() e.waiterQueue.Notify(waiter.EventIn) } @@ -2578,9 +2599,7 @@ func (e *endpoint) completeState() stack.TCPEndpointState { s.SegTime = time.Now() // Copy EndpointID. - e.mu.Lock() s.ID = stack.TCPEndpointID(e.ID) - e.mu.Unlock() // Copy endpoint rcv state. e.rcvListMu.Lock() @@ -2710,10 +2729,10 @@ func (e *endpoint) State() uint32 { // Info returns a copy of the endpoint info. func (e *endpoint) Info() tcpip.EndpointInfo { - e.mu.RLock() + e.LockUser() // Make a copy of the endpoint info. ret := e.EndpointInfo - e.mu.RUnlock() + e.UnlockUser() return &ret } @@ -2728,9 +2747,9 @@ func (e *endpoint) Wait() { e.waiterQueue.EventRegister(&waitEntry, waiter.EventHUp) defer e.waiterQueue.EventUnregister(&waitEntry) for { - e.mu.Lock() + e.LockUser() running := e.workerRunning - e.mu.Unlock() + e.UnlockUser() if !running { break } diff --git a/pkg/tcpip/transport/tcp/endpoint_state.go b/pkg/tcpip/transport/tcp/endpoint_state.go index 4a46f0ec5..9175de441 100644 --- a/pkg/tcpip/transport/tcp/endpoint_state.go +++ b/pkg/tcpip/transport/tcp/endpoint_state.go @@ -162,8 +162,8 @@ func (e *endpoint) loadState(state EndpointState) { connectingLoading.Add(1) } // Directly update the state here rather than using e.setEndpointState - // as the endpoint is still being loaded and the stack reference to increment - // metrics is not yet initialized. + // as the endpoint is still being loaded and the stack reference is not + // yet initialized. atomic.StoreUint32((*uint32)(&e.state), uint32(state)) } @@ -180,7 +180,6 @@ func (e *endpoint) afterLoad() { func (e *endpoint) Resume(s *stack.Stack) { e.stack = s e.segmentQueue.setLimit(MaxUnprocessedSegments) - e.workMu.Init() state := e.origEndpointState switch state { case StateInitial, StateBound, StateListen, StateConnecting, StateEstablished: diff --git a/pkg/tcpip/transport/tcp/protocol.go b/pkg/tcpip/transport/tcp/protocol.go index 73098d904..b0f918bb4 100644 --- a/pkg/tcpip/transport/tcp/protocol.go +++ b/pkg/tcpip/transport/tcp/protocol.go @@ -95,7 +95,7 @@ const ( ) type protocol struct { - mu sync.Mutex + mu sync.RWMutex sackEnabled bool delayEnabled bool sendBufferSize SendBufferSizeOption @@ -273,57 +273,57 @@ func (p *protocol) SetOption(option interface{}) *tcpip.Error { func (p *protocol) Option(option interface{}) *tcpip.Error { switch v := option.(type) { case *SACKEnabled: - p.mu.Lock() + p.mu.RLock() *v = SACKEnabled(p.sackEnabled) - p.mu.Unlock() + p.mu.RUnlock() return nil case *DelayEnabled: - p.mu.Lock() + p.mu.RLock() *v = DelayEnabled(p.delayEnabled) - p.mu.Unlock() + p.mu.RUnlock() return nil case *SendBufferSizeOption: - p.mu.Lock() + p.mu.RLock() *v = p.sendBufferSize - p.mu.Unlock() + p.mu.RUnlock() return nil case *ReceiveBufferSizeOption: - p.mu.Lock() + p.mu.RLock() *v = p.recvBufferSize - p.mu.Unlock() + p.mu.RUnlock() return nil case *tcpip.CongestionControlOption: - p.mu.Lock() + p.mu.RLock() *v = tcpip.CongestionControlOption(p.congestionControl) - p.mu.Unlock() + p.mu.RUnlock() return nil case *tcpip.AvailableCongestionControlOption: - p.mu.Lock() + p.mu.RLock() *v = tcpip.AvailableCongestionControlOption(strings.Join(p.availableCongestionControl, " ")) - p.mu.Unlock() + p.mu.RUnlock() return nil case *tcpip.ModerateReceiveBufferOption: - p.mu.Lock() + p.mu.RLock() *v = tcpip.ModerateReceiveBufferOption(p.moderateReceiveBuffer) - p.mu.Unlock() + p.mu.RUnlock() return nil case *tcpip.TCPLingerTimeoutOption: - p.mu.Lock() + p.mu.RLock() *v = tcpip.TCPLingerTimeoutOption(p.tcpLingerTimeout) - p.mu.Unlock() + p.mu.RUnlock() return nil case *tcpip.TCPTimeWaitTimeoutOption: - p.mu.Lock() + p.mu.RLock() *v = tcpip.TCPTimeWaitTimeoutOption(p.tcpTimeWaitTimeout) - p.mu.Unlock() + p.mu.RUnlock() return nil default: diff --git a/pkg/tcpip/transport/tcp/rcv.go b/pkg/tcpip/transport/tcp/rcv.go index d80aff1b6..caf8977b3 100644 --- a/pkg/tcpip/transport/tcp/rcv.go +++ b/pkg/tcpip/transport/tcp/rcv.go @@ -168,7 +168,6 @@ func (r *receiver) consumeSegment(s *segment, segSeq seqnum.Value, segLen seqnum // We just received a FIN, our next state depends on whether we sent a // FIN already or not. - r.ep.mu.Lock() switch r.ep.EndpointState() { case StateEstablished: r.ep.setEndpointState(StateCloseWait) @@ -183,7 +182,6 @@ func (r *receiver) consumeSegment(s *segment, segSeq seqnum.Value, segLen seqnum case StateFinWait2: r.ep.setEndpointState(StateTimeWait) } - r.ep.mu.Unlock() // Flush out any pending segments, except the very first one if // it happens to be the one we're handling now because the @@ -208,7 +206,6 @@ func (r *receiver) consumeSegment(s *segment, segSeq seqnum.Value, segLen seqnum // Handle ACK (not FIN-ACK, which we handled above) during one of the // shutdown states. if s.flagIsSet(header.TCPFlagAck) && s.ackNumber == r.ep.snd.sndNxt { - r.ep.mu.Lock() switch r.ep.EndpointState() { case StateFinWait1: r.ep.setEndpointState(StateFinWait2) @@ -222,7 +219,6 @@ func (r *receiver) consumeSegment(s *segment, segSeq seqnum.Value, segLen seqnum case StateLastAck: r.ep.transitionToStateCloseLocked() } - r.ep.mu.Unlock() } return true @@ -336,10 +332,8 @@ func (r *receiver) handleRcvdSegmentClosing(s *segment, state EndpointState, clo // handleRcvdSegment handles TCP segments directed at the connection managed by // r as they arrive. It is called by the protocol main loop. func (r *receiver) handleRcvdSegment(s *segment) (drop bool, err *tcpip.Error) { - r.ep.mu.RLock() state := r.ep.EndpointState() closed := r.ep.closed - r.ep.mu.RUnlock() if state != StateEstablished { drop, err := r.handleRcvdSegmentClosing(s, state, closed) diff --git a/pkg/tcpip/transport/tcp/segment_queue.go b/pkg/tcpip/transport/tcp/segment_queue.go index bd20a7ee9..48a257137 100644 --- a/pkg/tcpip/transport/tcp/segment_queue.go +++ b/pkg/tcpip/transport/tcp/segment_queue.go @@ -28,10 +28,16 @@ type segmentQueue struct { used int } +// emptyLocked determines if the queue is empty. +// Preconditions: q.mu must be held. +func (q *segmentQueue) emptyLocked() bool { + return q.used == 0 +} + // empty determines if the queue is empty. func (q *segmentQueue) empty() bool { q.mu.Lock() - r := q.used == 0 + r := q.emptyLocked() q.mu.Unlock() return r diff --git a/pkg/tcpip/transport/tcp/snd.go b/pkg/tcpip/transport/tcp/snd.go index 657c3146e..17fed4ec5 100644 --- a/pkg/tcpip/transport/tcp/snd.go +++ b/pkg/tcpip/transport/tcp/snd.go @@ -455,9 +455,7 @@ func (s *sender) retransmitTimerExpired() bool { // Give up if we've waited more than a minute since the last resend or // if a user time out is set and we have exceeded the user specified // timeout since the first retransmission. - s.ep.mu.RLock() uto := s.ep.userTimeout - s.ep.mu.RUnlock() if s.firstRetransmittedSegXmitTime.IsZero() { // We store the original xmitTime of the segment that we are @@ -713,7 +711,6 @@ func (s *sender) maybeSendSegment(seg *segment, limit int, end seqnum.Value) (se default: s.ep.setEndpointState(StateFinWait1) } - } else { // We're sending a non-FIN segment. if seg.flags&header.TCPFlagFin != 0 { diff --git a/pkg/tcpip/transport/tcp/tcp_test.go b/pkg/tcpip/transport/tcp/tcp_test.go index 5b2b16afa..39d36d2ba 100644 --- a/pkg/tcpip/transport/tcp/tcp_test.go +++ b/pkg/tcpip/transport/tcp/tcp_test.go @@ -2236,9 +2236,17 @@ func TestSegmentMerging(t *testing.T) { c.CreateConnected(789, 30000, -1 /* epRcvBuf */) - // Prevent the endpoint from processing packets. - test.stop(c.EP) + // Send 10 1 byte segments to fill up InitialWindow but don't + // ACK. That should prevent anymore packets from going out. + for i := 0; i < 10; i++ { + view := buffer.NewViewFromBytes([]byte{0}) + if _, _, err := c.EP.Write(tcpip.SlicePayload(view), tcpip.WriteOptions{}); err != nil { + t.Fatalf("Write #%d failed: %v", i+1, err) + } + } + // Now send the segments that should get merged as the congestion + // window is full and we won't be able to send any more packets. var allData []byte for i, data := range [][]byte{{1, 2, 3, 4}, {5, 6, 7}, {8, 9}, {10}, {11}} { allData = append(allData, data...) @@ -2248,8 +2256,29 @@ func TestSegmentMerging(t *testing.T) { } } - // Let the endpoint process the segments that we just sent. - test.resume(c.EP) + // Check that we get 10 packets of 1 byte each. + for i := 0; i < 10; i++ { + b := c.GetPacket() + checker.IPv4(t, b, + checker.PayloadLen(header.TCPMinimumSize+1), + checker.TCP( + checker.DstPort(context.TestPort), + checker.SeqNum(uint32(c.IRS)+uint32(i)+1), + checker.AckNum(790), + checker.TCPFlagsMatch(header.TCPFlagAck, ^uint8(header.TCPFlagPsh)), + ), + ) + } + + // Acknowledge the data. + c.SendPacket(nil, &context.Headers{ + SrcPort: context.TestPort, + DstPort: c.Port, + Flags: header.TCPFlagAck, + SeqNum: 790, + AckNum: c.IRS.Add(1 + 10), // 10 for the 10 bytes of payload. + RcvWnd: 30000, + }) // Check that data is received. b := c.GetPacket() @@ -2257,7 +2286,7 @@ func TestSegmentMerging(t *testing.T) { checker.PayloadLen(len(allData)+header.TCPMinimumSize), checker.TCP( checker.DstPort(context.TestPort), - checker.SeqNum(uint32(c.IRS)+1), + checker.SeqNum(uint32(c.IRS)+11), checker.AckNum(790), checker.TCPFlagsMatch(header.TCPFlagAck, ^uint8(header.TCPFlagPsh)), ), @@ -2273,7 +2302,7 @@ func TestSegmentMerging(t *testing.T) { DstPort: c.Port, Flags: header.TCPFlagAck, SeqNum: 790, - AckNum: c.IRS.Add(1 + seqnum.Size(len(allData))), + AckNum: c.IRS.Add(11 + seqnum.Size(len(allData))), RcvWnd: 30000, }) }) -- cgit v1.2.3