Merge branch 'dev' into add-tx-index

app/protocol/flows/v5/blockrelay/handle_ibd_block_locator.go

@@ -5,7 +5,6 @@ import (
 	"github.com/kaspanet/kaspad/app/protocol/peer"
 	"github.com/kaspanet/kaspad/app/protocol/protocolerrors"
 	"github.com/kaspanet/kaspad/domain"
-	"github.com/kaspanet/kaspad/domain/consensus/model/externalapi"
 	"github.com/kaspanet/kaspad/infrastructure/network/netadapter/router"
 )
 
@@ -34,7 +33,7 @@ func HandleIBDBlockLocator(context HandleIBDBlockLocatorContext, incomingRoute *
 		if err != nil {
 			return err
 		}
-		if !blockInfo.Exists {
+		if !blockInfo.HasHeader() {
 			return protocolerrors.Errorf(true, "received IBDBlockLocator "+
 				"with an unknown targetHash %s", targetHash)
 		}
@@ -47,7 +46,7 @@ func HandleIBDBlockLocator(context HandleIBDBlockLocatorContext, incomingRoute *
 			}
 
 			// The IBD block locator is checking only existing blocks with bodies.
-			if !blockInfo.Exists || blockInfo.BlockStatus == externalapi.StatusHeaderOnly {
+			if !blockInfo.HasBody() {
 				continue
 			}
 

app/protocol/flows/v5/blockrelay/handle_ibd_block_requests.go

@@ -4,7 +4,6 @@ import (
 	"github.com/kaspanet/kaspad/app/appmessage"
 	"github.com/kaspanet/kaspad/app/protocol/protocolerrors"
 	"github.com/kaspanet/kaspad/domain"
-	"github.com/kaspanet/kaspad/domain/consensus/model/externalapi"
 	"github.com/kaspanet/kaspad/infrastructure/network/netadapter/router"
 	"github.com/pkg/errors"
 )
@@ -32,7 +31,7 @@ func HandleIBDBlockRequests(context HandleIBDBlockRequestsContext, incomingRoute
 			if err != nil {
 				return err
 			}
-			if !blockInfo.Exists || blockInfo.BlockStatus == externalapi.StatusHeaderOnly {
+			if !blockInfo.HasBody() {
 				return protocolerrors.Errorf(true, "block %s not found (v5)", hash)
 			}
 			block, err := context.Domain().Consensus().GetBlock(hash)

app/protocol/flows/v5/blockrelay/handle_relay_block_requests.go

@@ -5,7 +5,6 @@ import (
 	peerpkg "github.com/kaspanet/kaspad/app/protocol/peer"
 	"github.com/kaspanet/kaspad/app/protocol/protocolerrors"
 	"github.com/kaspanet/kaspad/domain"
-	"github.com/kaspanet/kaspad/domain/consensus/model/externalapi"
 	"github.com/kaspanet/kaspad/infrastructure/network/netadapter/router"
 	"github.com/pkg/errors"
 )
@@ -33,7 +32,7 @@ func HandleRelayBlockRequests(context RelayBlockRequestsContext, incomingRoute *
 			if err != nil {
 				return err
 			}
-			if !blockInfo.Exists || blockInfo.BlockStatus == externalapi.StatusHeaderOnly {
+			if !blockInfo.HasBody() {
 				return protocolerrors.Errorf(true, "block %s not found", hash)
 			}
 			block, err := context.Domain().Consensus().GetBlock(hash)

app/protocol/flows/v5/blockrelay/handle_request_anticone.go

@@ -47,9 +47,9 @@ func (flow *handleRequestAnticoneFlow) start() error {
 
 		// GetAnticone is expected to be called by the syncee for getting the anticone of the header selected tip
 		// intersected by past of relayed block, and is thus expected to be bounded by mergeset limit since
-		// we relay blocks only if they enter virtual's mergeset. We add 2 for a small margin error.
+		// we relay blocks only if they enter virtual's mergeset. We add a 2 factor for possible sync gaps.
 		blockHashes, err := flow.Domain().Consensus().GetAnticone(blockHash, contextHash,
-			flow.Config().ActiveNetParams.MergeSetSizeLimit+2)
+			flow.Config().ActiveNetParams.MergeSetSizeLimit*2)
 		if err != nil {
 			return protocolerrors.Wrap(true, err, "Failed querying anticone")
 		}

app/protocol/flows/v5/blockrelay/handle_request_headers.go

@@ -46,7 +46,25 @@ func (flow *handleRequestHeadersFlow) start() error {
 		}
 		log.Debugf("Received requestHeaders with lowHash: %s, highHash: %s", lowHash, highHash)
 
-		isLowSelectedAncestorOfHigh, err := flow.Domain().Consensus().IsInSelectedParentChainOf(lowHash, highHash)
+		consensus := flow.Domain().Consensus()
+
+		lowHashInfo, err := consensus.GetBlockInfo(lowHash)
+		if err != nil {
+			return err
+		}
+		if !lowHashInfo.HasHeader() {
+			return protocolerrors.Errorf(true, "Block %s does not exist", lowHash)
+		}
+
+		highHashInfo, err := consensus.GetBlockInfo(highHash)
+		if err != nil {
+			return err
+		}
+		if !highHashInfo.HasHeader() {
+			return protocolerrors.Errorf(true, "Block %s does not exist", highHash)
+		}
+
+		isLowSelectedAncestorOfHigh, err := consensus.IsInSelectedParentChainOf(lowHash, highHash)
 		if err != nil {
 			return err
 		}
@@ -62,7 +80,7 @@ func (flow *handleRequestHeadersFlow) start() error {
 			// in order to avoid locking the consensus for too long
 			// maxBlocks MUST be >= MergeSetSizeLimit + 1
 			const maxBlocks = 1 << 10
-			blockHashes, _, err := flow.Domain().Consensus().GetHashesBetween(lowHash, highHash, maxBlocks)
+			blockHashes, _, err := consensus.GetHashesBetween(lowHash, highHash, maxBlocks)
 			if err != nil {
 				return err
 			}
@@ -70,7 +88,7 @@ func (flow *handleRequestHeadersFlow) start() error {
 
 			blockHeaders := make([]*appmessage.MsgBlockHeader, len(blockHashes))
 			for i, blockHash := range blockHashes {
-				blockHeader, err := flow.Domain().Consensus().GetBlockHeader(blockHash)
+				blockHeader, err := consensus.GetBlockHeader(blockHash)
 				if err != nil {
 					return err
 				}

app/rpc/rpchandlers/add_peer.go

@@ -9,6 +9,14 @@ import (
 
 // HandleAddPeer handles the respectively named RPC command
 func HandleAddPeer(context *rpccontext.Context, _ *router.Router, request appmessage.Message) (appmessage.Message, error) {
+	if context.Config.SafeRPC {
+		log.Warn("AddPeer RPC command called while node in safe RPC mode -- ignoring.")
+		response := appmessage.NewAddPeerResponseMessage()
+		response.Error =
+			appmessage.RPCErrorf("AddPeer RPC command called while node in safe RPC mode")
+		return response, nil
+	}
+
 	AddPeerRequest := request.(*appmessage.AddPeerRequestMessage)
 	address, err := network.NormalizeAddress(AddPeerRequest.Address, context.Config.ActiveNetParams.DefaultPort)
 	if err != nil {

app/rpc/rpchandlers/ban.go

@@ -9,6 +9,14 @@ import (
 
 // HandleBan handles the respectively named RPC command
 func HandleBan(context *rpccontext.Context, _ *router.Router, request appmessage.Message) (appmessage.Message, error) {
+	if context.Config.SafeRPC {
+		log.Warn("Ban RPC command called while node in safe RPC mode -- ignoring.")
+		response := appmessage.NewBanResponseMessage()
+		response.Error =
+			appmessage.RPCErrorf("Ban RPC command called while node in safe RPC mode")
+		return response, nil
+	}
+
 	banRequest := request.(*appmessage.BanRequestMessage)
 	ip := net.ParseIP(banRequest.IP)
 	if ip == nil {

app/rpc/rpchandlers/estimate_network_hashes_per_second.go

@@ -27,6 +27,27 @@ func HandleEstimateNetworkHashesPerSecond(
 		}
 	}
 
+	if context.Config.SafeRPC {
+		const windowSizeLimit = 10000
+		if windowSize > windowSizeLimit {
+			response := &appmessage.EstimateNetworkHashesPerSecondResponseMessage{}
+			response.Error =
+				appmessage.RPCErrorf(
+					"Requested window size %d is larger than max allowed in RPC safe mode (%d)",
+					windowSize, windowSizeLimit)
+			return response, nil
+		}
+	}
+
+	if uint64(windowSize) > context.Config.ActiveNetParams.PruningDepth() {
+		response := &appmessage.EstimateNetworkHashesPerSecondResponseMessage{}
+		response.Error =
+			appmessage.RPCErrorf(
+				"Requested window size %d is larger than pruning point depth %d",
+				windowSize, context.Config.ActiveNetParams.PruningDepth())
+		return response, nil
+	}
+
 	networkHashesPerSecond, err := context.Domain.Consensus().EstimateNetworkHashesPerSecond(startHash, windowSize)
 	if err != nil {
 		response := &appmessage.EstimateNetworkHashesPerSecondResponseMessage{}

app/rpc/rpchandlers/get_mempool_entries_by_addresses.go

@@ -43,7 +43,7 @@ func HandleGetMempoolEntriesByAddresses(context *rpccontext.Context, _ *router.R
 	if getMempoolEntriesByAddressesRequest.IncludeOrphanPool {
 
 		orphanPoolTransactions := context.Domain.MiningManager().AllOrphanTransactions()
-		orphanPoolEntriesByAddresse, err := extractMempoolEntriesByAddressesFromTransactions(
+		orphanPoolEntriesByAddress, err := extractMempoolEntriesByAddressesFromTransactions(
 			context,
 			getMempoolEntriesByAddressesRequest.Addresses,
 			orphanPoolTransactions,
@@ -59,7 +59,7 @@ func HandleGetMempoolEntriesByAddresses(context *rpccontext.Context, _ *router.R
 			return errorMessage, nil
 		}
 
-		mempoolEntriesByAddresses = append(mempoolEntriesByAddresses, orphanPoolEntriesByAddresse...)
+		mempoolEntriesByAddresses = append(mempoolEntriesByAddresses, orphanPoolEntriesByAddress...)
 	}
 
 	return appmessage.NewGetMempoolEntriesByAddressesResponseMessage(mempoolEntriesByAddresses), nil
@@ -80,9 +80,13 @@ func extractMempoolEntriesByAddressesFromTransactions(context *rpccontext.Contex
 		for _, transaction := range transactions {
 
 			for i, input := range transaction.Inputs {
-				// TODO: Fix this
 				if input.UTXOEntry == nil {
-					log.Errorf("Couldn't find UTXO entry for input %d in mempool transaction %s. This is a bug and should be fixed.", i, consensushashing.TransactionID(transaction))
+					if !areOrphans { // Orphans can legitimately have `input.UTXOEntry == nil`
+						// TODO: Fix the underlying cause of the bug for non-orphan entries
+						log.Debugf(
+							"Couldn't find UTXO entry for input %d in mempool transaction %s. This is a bug and should be fixed.",
+							i, consensushashing.TransactionID(transaction))
+					}
 					continue
 				}
 

app/rpc/rpchandlers/resolve_finality_conflict.go

@@ -8,6 +8,14 @@ import (
 
 // HandleResolveFinalityConflict handles the respectively named RPC command
 func HandleResolveFinalityConflict(context *rpccontext.Context, _ *router.Router, request appmessage.Message) (appmessage.Message, error) {
+	if context.Config.SafeRPC {
+		log.Warn("ResolveFinalityConflict RPC command called while node in safe RPC mode -- ignoring.")
+		response := &appmessage.ResolveFinalityConflictResponseMessage{}
+		response.Error =
+			appmessage.RPCErrorf("ResolveFinalityConflict RPC command called while node in safe RPC mode")
+		return response, nil
+	}
+
 	response := &appmessage.ResolveFinalityConflictResponseMessage{}
 	response.Error = appmessage.RPCErrorf("not implemented")
 	return response, nil

app/rpc/rpchandlers/shut_down.go

@@ -12,6 +12,14 @@ const pauseBeforeShutDown = time.Second
 
 // HandleShutDown handles the respectively named RPC command
 func HandleShutDown(context *rpccontext.Context, _ *router.Router, _ appmessage.Message) (appmessage.Message, error) {
+	if context.Config.SafeRPC {
+		log.Warn("ShutDown RPC command called while node in safe RPC mode -- ignoring.")
+		response := appmessage.NewShutDownResponseMessage()
+		response.Error =
+			appmessage.RPCErrorf("ShutDown RPC command called while node in safe RPC mode")
+		return response, nil
+	}
+
 	log.Warn("ShutDown RPC called.")
 
 	// Wait a second before shutting down, to allow time to return the response to the caller

app/rpc/rpchandlers/unban.go

@@ -9,6 +9,14 @@ import (
 
 // HandleUnban handles the respectively named RPC command
 func HandleUnban(context *rpccontext.Context, _ *router.Router, request appmessage.Message) (appmessage.Message, error) {
+	if context.Config.SafeRPC {
+		log.Warn("Unban RPC command called while node in safe RPC mode -- ignoring.")
+		response := appmessage.NewUnbanResponseMessage()
+		response.Error =
+			appmessage.RPCErrorf("Unban RPC command called while node in safe RPC mode")
+		return response, nil
+	}
+
 	unbanRequest := request.(*appmessage.UnbanRequestMessage)
 	ip := net.ParseIP(unbanRequest.IP)
 	if ip == nil {

changelog.txt

@@ -1,3 +1,8 @@
+Kaspad v0.12.3 - 2022-06-29
+===========================
+
+* Fixes a few bugs which can lead to node crashes or out-of-memory errors
+
 Kaspad v0.12.2 - 2022-06-17
 ===========================
 

domain/consensus/model/externalapi/blockinfo.go

@@ -13,6 +13,16 @@ type BlockInfo struct {
 	MergeSetReds   []*DomainHash
 }
 
+// HasHeader returns whether the block exists and has a valid header
+func (bi *BlockInfo) HasHeader() bool {
+	return bi.Exists && bi.BlockStatus != StatusInvalid
+}
+
+// HasBody returns whether the block exists and has a valid body
+func (bi *BlockInfo) HasBody() bool {
+	return bi.Exists && bi.BlockStatus != StatusInvalid && bi.BlockStatus != StatusHeaderOnly
+}
+
 // Clone returns a clone of BlockInfo
 func (bi *BlockInfo) Clone() *BlockInfo {
 	return &BlockInfo{

infrastructure/config/config.go

@@ -98,6 +98,7 @@ type Flags struct {
 	RPCMaxWebsockets                int           `long:"rpcmaxwebsockets" description:"Max number of RPC websocket connections"`
 	RPCMaxConcurrentReqs            int           `long:"rpcmaxconcurrentreqs" description:"Max number of concurrent RPC requests that may be processed concurrently"`
 	DisableRPC                      bool          `long:"norpc" description:"Disable built-in RPC server"`
+	SafeRPC                         bool          `long:"saferpc" description:"Disable RPC commands which affect the state of the node"`
 	DisableDNSSeed                  bool          `long:"nodnsseed" description:"Disable DNS seeding for peers"`
 	DNSSeed                         string        `long:"dnsseed" description:"Override DNS seeds with specified hostname (Only 1 hostname allowed)"`
 	GRPCSeed                        string        `long:"grpcseed" description:"Hostname of gRPC server for seeding peers"`

version/version.go

@@ -11,7 +11,7 @@ const validCharacters = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrs
 const (
 	appMajor uint = 0
 	appMinor uint = 12
-	appPatch uint = 2
+	appPatch uint = 3
 )
 
 // appBuild is defined as a variable so it can be overridden during the build