diff options
-rw-r--r-- | pkg/tcpip/transport/internal/network/endpoint.go | 168 | ||||
-rw-r--r-- | pkg/tcpip/transport/internal/network/endpoint_state.go | 16 |
2 files changed, 120 insertions, 64 deletions
diff --git a/pkg/tcpip/transport/internal/network/endpoint.go b/pkg/tcpip/transport/internal/network/endpoint.go index 3cb821475..e3094f59f 100644 --- a/pkg/tcpip/transport/internal/network/endpoint.go +++ b/pkg/tcpip/transport/internal/network/endpoint.go @@ -18,6 +18,7 @@ package network import ( "fmt" + "sync/atomic" "gvisor.dev/gvisor/pkg/sync" "gvisor.dev/gvisor/pkg/tcpip" @@ -39,11 +40,7 @@ type Endpoint struct { mu sync.RWMutex `state:"nosave"` // +checklocks:mu - state transport.DatagramEndpointState - // +checklocks:mu wasBound bool - // +checklocks:mu - info stack.TransportEndpointInfo // owner is the owner of transmitted packets. // // +checklocks:mu @@ -72,6 +69,34 @@ type Endpoint struct { ipv4TOS uint8 // +checklocks:mu ipv6TClass uint8 + + // Lock ordering: mu > infoMu. + infoMu sync.RWMutex `state:"nosave"` + // info has a dedicated mutex so that we can avoid lock ordering violations + // when reading the endpoint's info. If we used mu, we need to guarantee + // that any lock taken while mu is held is not held when calling Info() + // which is not true as of writing (we hold mu while registering transport + // endpoints (taking the transport demuxer lock but we also hold the demuxer + // lock when delivering packets/errors to endpoints). + // + // Writes must be performed through setInfo. + // + // +checklocks:infoMu + info stack.TransportEndpointInfo + + // state holds a transport.DatagramBasedEndpointState. + // + // state must be accessed with atomics so that we can avoid lock ordering + // violations when reading the state. If we used mu, we need to guarantee + // that any lock taken while mu is held is not held when calling State() + // which is not true as of writing (we hold mu while registering transport + // endpoints (taking the transport demuxer lock but we also hold the demuxer + // lock when delivering packets/errors to endpoints). + // + // Writes must be performed through setEndpointState. + // + // +checkatomics + state uint32 } // +stateify savable @@ -101,7 +126,6 @@ func (e *Endpoint) Init(s *stack.Stack, netProto tcpip.NetworkProtocolNumber, tr netProto: netProto, transProto: transProto, - state: transport.DatagramEndpointStateInitial, info: stack.TransportEndpointInfo{ NetProto: netProto, TransProto: transProto, @@ -111,6 +135,10 @@ func (e *Endpoint) Init(s *stack.Stack, netProto tcpip.NetworkProtocolNumber, tr multicastTTL: 1, multicastMemberships: make(map[multicastMembership]struct{}), } + + e.mu.Lock() + defer e.mu.Unlock() + e.setEndpointState(transport.DatagramEndpointStateInitial) } // NetProto returns the network protocol the endpoint was initialized with. @@ -118,11 +146,19 @@ func (e *Endpoint) NetProto() tcpip.NetworkProtocolNumber { return e.netProto } +// setEndpointState sets the state of the endpoint. +// +// e.mu must be held to synchronize changes to state with the rest of the +// endpoint. +// +// +checklocks:e.mu +func (e *Endpoint) setEndpointState(state transport.DatagramEndpointState) { + atomic.StoreUint32(&e.state, uint32(state)) +} + // State returns the state of the endpoint. func (e *Endpoint) State() transport.DatagramEndpointState { - e.mu.RLock() - defer e.mu.RUnlock() - return e.state + return transport.DatagramEndpointState(atomic.LoadUint32(&e.state)) } // Close cleans the endpoint's resources and leaves the endpoint in a closed @@ -131,7 +167,7 @@ func (e *Endpoint) Close() { e.mu.Lock() defer e.mu.Unlock() - if e.state == transport.DatagramEndpointStateClosed { + if e.State() == transport.DatagramEndpointStateClosed { return } @@ -145,7 +181,7 @@ func (e *Endpoint) Close() { e.connectedRoute = nil } - e.state = transport.DatagramEndpointStateClosed + e.setEndpointState(transport.DatagramEndpointStateClosed) } // SetOwner sets the owner of transmitted packets. @@ -226,7 +262,7 @@ func (e *Endpoint) AcquireContextForWrite(opts tcpip.WriteOptions) (WriteContext return WriteContext{}, &tcpip.ErrInvalidOptionValue{} } - if e.state == transport.DatagramEndpointStateClosed { + if e.State() == transport.DatagramEndpointStateClosed { return WriteContext{}, &tcpip.ErrInvalidEndpointState{} } @@ -238,7 +274,7 @@ func (e *Endpoint) AcquireContextForWrite(opts tcpip.WriteOptions) (WriteContext if opts.To == nil { // If the user doesn't specify a destination, they should have // connected to another address. - if e.state != transport.DatagramEndpointStateConnected { + if e.State() != transport.DatagramEndpointStateConnected { return WriteContext{}, &tcpip.ErrDestinationRequired{} } @@ -250,18 +286,19 @@ func (e *Endpoint) AcquireContextForWrite(opts tcpip.WriteOptions) (WriteContext if nicID == 0 { nicID = tcpip.NICID(e.ops.GetBindToDevice()) } - if e.info.BindNICID != 0 { - if nicID != 0 && nicID != e.info.BindNICID { + info := e.Info() + if info.BindNICID != 0 { + if nicID != 0 && nicID != info.BindNICID { return WriteContext{}, &tcpip.ErrNoRoute{} } - nicID = e.info.BindNICID + nicID = info.BindNICID } if nicID == 0 { - nicID = e.info.RegisterNICID + nicID = info.RegisterNICID } - dst, netProto, err := e.checkV4MappedRLocked(*opts.To) + dst, netProto, err := e.checkV4Mapped(*opts.To) if err != nil { return WriteContext{}, err } @@ -301,20 +338,22 @@ func (e *Endpoint) Disconnect() { e.mu.Lock() defer e.mu.Unlock() - if e.state != transport.DatagramEndpointStateConnected { + if e.State() != transport.DatagramEndpointStateConnected { return } + info := e.Info() // Exclude ephemerally bound endpoints. if e.wasBound { - e.info.ID = stack.TransportEndpointID{ - LocalAddress: e.info.BindAddr, + info.ID = stack.TransportEndpointID{ + LocalAddress: info.BindAddr, } - e.state = transport.DatagramEndpointStateBound + e.setEndpointState(transport.DatagramEndpointStateBound) } else { - e.info.ID = stack.TransportEndpointID{} - e.state = transport.DatagramEndpointStateInitial + info.ID = stack.TransportEndpointID{} + e.setEndpointState(transport.DatagramEndpointStateInitial) } + e.setInfo(info) e.connectedRoute.Release() e.connectedRoute = nil @@ -327,7 +366,7 @@ func (e *Endpoint) Disconnect() { // TODO(https://gvisor.dev/issue/6590): Annotate read lock requirement. // +checklocks:e.mu func (e *Endpoint) connectRouteRLocked(nicID tcpip.NICID, addr tcpip.FullAddress, netProto tcpip.NetworkProtocolNumber) (*stack.Route, tcpip.NICID, tcpip.Error) { - localAddr := e.info.ID.LocalAddress + localAddr := e.Info().ID.LocalAddress if e.isBroadcastOrMulticast(nicID, netProto, localAddr) { // A packet can only originate from a unicast address (i.e., an interface). localAddr = "" @@ -370,24 +409,25 @@ func (e *Endpoint) ConnectAndThen(addr tcpip.FullAddress, f func(netProto tcpip. e.mu.Lock() defer e.mu.Unlock() + info := e.Info() nicID := addr.NIC - switch e.state { + switch e.State() { case transport.DatagramEndpointStateInitial: case transport.DatagramEndpointStateBound, transport.DatagramEndpointStateConnected: - if e.info.BindNICID == 0 { + if info.BindNICID == 0 { break } - if nicID != 0 && nicID != e.info.BindNICID { + if nicID != 0 && nicID != info.BindNICID { return &tcpip.ErrInvalidEndpointState{} } - nicID = e.info.BindNICID + nicID = info.BindNICID default: return &tcpip.ErrInvalidEndpointState{} } - addr, netProto, err := e.checkV4MappedRLocked(addr) + addr, netProto, err := e.checkV4Mapped(addr) if err != nil { return err } @@ -398,14 +438,14 @@ func (e *Endpoint) ConnectAndThen(addr tcpip.FullAddress, f func(netProto tcpip. } id := stack.TransportEndpointID{ - LocalAddress: e.info.ID.LocalAddress, + LocalAddress: info.ID.LocalAddress, RemoteAddress: r.RemoteAddress(), } - if e.state == transport.DatagramEndpointStateInitial { + if e.State() == transport.DatagramEndpointStateInitial { id.LocalAddress = r.LocalAddress() } - if err := f(r.NetProto(), e.info.ID, id); err != nil { + if err := f(r.NetProto(), info.ID, id); err != nil { return err } @@ -414,10 +454,11 @@ func (e *Endpoint) ConnectAndThen(addr tcpip.FullAddress, f func(netProto tcpip. e.connectedRoute.Release() } e.connectedRoute = r - e.info.ID = id - e.info.RegisterNICID = nicID + info.ID = id + info.RegisterNICID = nicID + e.setInfo(info) e.effectiveNetProto = netProto - e.state = transport.DatagramEndpointStateConnected + e.setEndpointState(transport.DatagramEndpointStateConnected) return nil } @@ -426,7 +467,7 @@ func (e *Endpoint) Shutdown() tcpip.Error { e.mu.Lock() defer e.mu.Unlock() - switch state := e.state; state { + switch state := e.State(); state { case transport.DatagramEndpointStateInitial, transport.DatagramEndpointStateClosed: return &tcpip.ErrNotConnected{} case transport.DatagramEndpointStateBound, transport.DatagramEndpointStateConnected: @@ -439,11 +480,9 @@ func (e *Endpoint) Shutdown() tcpip.Error { // checkV4MappedRLocked determines the effective network protocol and converts // addr to its canonical form. -// -// TODO(https://gvisor.dev/issue/6590): Annotate read lock requirement. -// +checklocks:e.mu -func (e *Endpoint) checkV4MappedRLocked(addr tcpip.FullAddress) (tcpip.FullAddress, tcpip.NetworkProtocolNumber, tcpip.Error) { - unwrapped, netProto, err := e.info.AddrNetProtoLocked(addr, e.ops.GetV6Only()) +func (e *Endpoint) checkV4Mapped(addr tcpip.FullAddress) (tcpip.FullAddress, tcpip.NetworkProtocolNumber, tcpip.Error) { + info := e.Info() + unwrapped, netProto, err := info.AddrNetProtoLocked(addr, e.ops.GetV6Only()) if err != nil { return tcpip.FullAddress{}, 0, err } @@ -474,11 +513,11 @@ func (e *Endpoint) BindAndThen(addr tcpip.FullAddress, f func(tcpip.NetworkProto // Don't allow binding once endpoint is not in the initial state // anymore. - if e.state != transport.DatagramEndpointStateInitial { + if e.State() != transport.DatagramEndpointStateInitial { return &tcpip.ErrInvalidEndpointState{} } - addr, netProto, err := e.checkV4MappedRLocked(addr) + addr, netProto, err := e.checkV4Mapped(addr) if err != nil { return err } @@ -497,14 +536,16 @@ func (e *Endpoint) BindAndThen(addr tcpip.FullAddress, f func(tcpip.NetworkProto e.wasBound = true - e.info.ID = stack.TransportEndpointID{ + info := e.Info() + info.ID = stack.TransportEndpointID{ LocalAddress: addr.Addr, } - e.info.BindNICID = addr.NIC - e.info.RegisterNICID = nicID - e.info.BindAddr = addr.Addr + info.BindNICID = addr.NIC + info.RegisterNICID = nicID + info.BindAddr = addr.Addr + e.setInfo(info) e.effectiveNetProto = netProto - e.state = transport.DatagramEndpointStateBound + e.setEndpointState(transport.DatagramEndpointStateBound) return nil } @@ -520,13 +561,14 @@ func (e *Endpoint) GetLocalAddress() tcpip.FullAddress { e.mu.RLock() defer e.mu.RUnlock() - addr := e.info.BindAddr - if e.state == transport.DatagramEndpointStateConnected { + info := e.Info() + addr := info.BindAddr + if e.State() == transport.DatagramEndpointStateConnected { addr = e.connectedRoute.LocalAddress() } return tcpip.FullAddress{ - NIC: e.info.RegisterNICID, + NIC: info.RegisterNICID, Addr: addr, } } @@ -536,13 +578,13 @@ func (e *Endpoint) GetRemoteAddress() (tcpip.FullAddress, bool) { e.mu.RLock() defer e.mu.RUnlock() - if e.state != transport.DatagramEndpointStateConnected { + if e.State() != transport.DatagramEndpointStateConnected { return tcpip.FullAddress{}, false } return tcpip.FullAddress{ Addr: e.connectedRoute.RemoteAddress(), - NIC: e.info.RegisterNICID, + NIC: e.Info().RegisterNICID, }, true } @@ -624,7 +666,7 @@ func (e *Endpoint) SetSockOpt(opt tcpip.SettableSocketOption) tcpip.Error { defer e.mu.Unlock() fa := tcpip.FullAddress{Addr: v.InterfaceAddr} - fa, netProto, err := e.checkV4MappedRLocked(fa) + fa, netProto, err := e.checkV4Mapped(fa) if err != nil { return err } @@ -648,7 +690,7 @@ func (e *Endpoint) SetSockOpt(opt tcpip.SettableSocketOption) tcpip.Error { } } - if e.info.BindNICID != 0 && e.info.BindNICID != nic { + if info := e.Info(); info.BindNICID != 0 && info.BindNICID != nic { return &tcpip.ErrInvalidEndpointState{} } @@ -751,7 +793,19 @@ func (e *Endpoint) GetSockOpt(opt tcpip.GettableSocketOption) tcpip.Error { // Info returns a copy of the endpoint info. func (e *Endpoint) Info() stack.TransportEndpointInfo { - e.mu.RLock() - defer e.mu.RUnlock() + e.infoMu.RLock() + defer e.infoMu.RUnlock() return e.info } + +// setInfo sets the endpoint's info. +// +// e.mu must be held to synchronize changes to info with the rest of the +// endpoint. +// +// +checklocks:e.mu +func (e *Endpoint) setInfo(info stack.TransportEndpointInfo) { + e.infoMu.Lock() + defer e.infoMu.Unlock() + e.info = info +} diff --git a/pkg/tcpip/transport/internal/network/endpoint_state.go b/pkg/tcpip/transport/internal/network/endpoint_state.go index 173197512..68bd1fbf6 100644 --- a/pkg/tcpip/transport/internal/network/endpoint_state.go +++ b/pkg/tcpip/transport/internal/network/endpoint_state.go @@ -35,22 +35,24 @@ func (e *Endpoint) Resume(s *stack.Stack) { } } - switch e.state { + info := e.Info() + + switch state := e.State(); state { case transport.DatagramEndpointStateInitial, transport.DatagramEndpointStateClosed: case transport.DatagramEndpointStateBound: - if len(e.info.ID.LocalAddress) != 0 && !e.isBroadcastOrMulticast(e.info.RegisterNICID, e.effectiveNetProto, e.info.ID.LocalAddress) { - if e.stack.CheckLocalAddress(e.info.RegisterNICID, e.effectiveNetProto, e.info.ID.LocalAddress) == 0 { - panic(fmt.Sprintf("got e.stack.CheckLocalAddress(%d, %d, %s) = 0, want != 0", e.info.RegisterNICID, e.effectiveNetProto, e.info.ID.LocalAddress)) + if len(info.ID.LocalAddress) != 0 && !e.isBroadcastOrMulticast(info.RegisterNICID, e.effectiveNetProto, info.ID.LocalAddress) { + if e.stack.CheckLocalAddress(info.RegisterNICID, e.effectiveNetProto, info.ID.LocalAddress) == 0 { + panic(fmt.Sprintf("got e.stack.CheckLocalAddress(%d, %d, %s) = 0, want != 0", info.RegisterNICID, e.effectiveNetProto, info.ID.LocalAddress)) } } case transport.DatagramEndpointStateConnected: var err tcpip.Error multicastLoop := e.ops.GetMulticastLoop() - e.connectedRoute, err = e.stack.FindRoute(e.info.RegisterNICID, e.info.ID.LocalAddress, e.info.ID.RemoteAddress, e.effectiveNetProto, multicastLoop) + e.connectedRoute, err = e.stack.FindRoute(info.RegisterNICID, info.ID.LocalAddress, info.ID.RemoteAddress, e.effectiveNetProto, multicastLoop) if err != nil { - panic(fmt.Sprintf("e.stack.FindRoute(%d, %s, %s, %d, %t): %s", e.info.RegisterNICID, e.info.ID.LocalAddress, e.info.ID.RemoteAddress, e.effectiveNetProto, multicastLoop, err)) + panic(fmt.Sprintf("e.stack.FindRoute(%d, %s, %s, %d, %t): %s", info.RegisterNICID, info.ID.LocalAddress, info.ID.RemoteAddress, e.effectiveNetProto, multicastLoop, err)) } default: - panic(fmt.Sprintf("unhandled state = %s", e.state)) + panic(fmt.Sprintf("unhandled state = %s", state)) } } |