Merge bitcoin/bitcoin#27257: refactor, net: End friendship of CNode, CConnman and ConnmanTestMsg

3566aa7d49 [net] Remove CNode friends (dergoegge)
3eac5e7cd1 [net] Add CNode helper for send byte accounting (dergoegge)
60441a3432 scripted-diff: [net] Rename CNode process queue members (dergoegge)
6693c499f7 [net] Make cs_vProcessMsg a non-recursive mutex (dergoegge)
23d9352654 [net] Make CNode msg process queue members private (dergoegge)
897e342d6e [net] Encapsulate CNode message polling (dergoegge)
cc5cdf8776 [net] Deduplicate marking received message for processing (dergoegge)
ad44aa5c64 [net] Add connection type getter to CNode (dergoegge)

Pull request description:

  We should define clear interfaces between CNode, CConnman and PeerManager. This PR makes a small step in that direction by ending the friendship of CNode, CConnman and ConnmanTestMsg. CNode's message processing queue is made private in the process and its mutex is turned into a non-recursive mutex.

ACKs for top commit:
  jnewbery:
    utACK 3566aa7d49
  vasild:
    ACK 3566aa7d49
  theStack:
    re-ACK 3566aa7d49
  brunoerg:
    re-ACK 3566aa7d49

Tree-SHA512: 26b87da5054e32401b693b2904e9c5f40e35a53937c0b6cf44b8597034ad07bacf27d87cdffc54d3e7ccfebde4231ef30a38d326f88cc18133bbb34688ead567

src/net.cpp

@@ -917,7 +917,7 @@ bool CConnman::AttemptToEvictConnection()
                 .m_is_local = node->addr.IsLocal(),
                 .m_network = node->ConnectedThroughNetwork(),
                 .m_noban = node->HasPermission(NetPermissionFlags::NoBan),
-                .m_conn_type = node->m_conn_type,
+                .m_conn_type = node->GetConnectionType(),
             };
             vEvictionCandidates.push_back(candidate);
         }
@@ -1092,7 +1092,7 @@ bool CConnman::AddConnection(const std::string& address, ConnectionType conn_typ
 
     // Count existing connections
     int existing_connections = WITH_LOCK(m_nodes_mutex,
-                                         return std::count_if(m_nodes.begin(), m_nodes.end(), [conn_type](CNode* node) { return node->m_conn_type == conn_type; }););
+                                         return std::count_if(m_nodes.begin(), m_nodes.end(), [conn_type](CNode* node) { return node->GetConnectionType() == conn_type; }););
 
     // Max connections of specified type already exist
     if (max_connections != std::nullopt && existing_connections >= max_connections) return false;
@@ -1328,18 +1328,7 @@ void CConnman::SocketHandlerConnected(const std::vector<CNode*>& nodes,
                 }
                 RecordBytesRecv(nBytes);
                 if (notify) {
-                    size_t nSizeAdded = 0;
+                    pnode->MarkReceivedMsgsForProcessing(nReceiveFloodSize);
-                    for (const auto& msg : pnode->vRecvMsg) {
-                        // vRecvMsg contains only completed CNetMessage
-                        // the single possible partially deserialized message are held by TransportDeserializer
-                        nSizeAdded += msg.m_raw_message_size;
-                    }
-                    {
-                        LOCK(pnode->cs_vProcessMsg);
-                        pnode->vProcessMsg.splice(pnode->vProcessMsg.end(), pnode->vRecvMsg);
-                        pnode->nProcessQueueSize += nSizeAdded;
-                        pnode->fPauseRecv = pnode->nProcessQueueSize > nReceiveFloodSize;
-                    }
                     WakeMessageHandler();
                 }
             }
@@ -1722,7 +1711,7 @@ void CConnman::ThreadOpenConnections(const std::vector<std::string> connect)
                 if (pnode->IsBlockOnlyConn()) nOutboundBlockRelay++;
 
                 // Make sure our persistent outbound slots belong to different netgroups.
-                switch (pnode->m_conn_type) {
+                switch (pnode->GetConnectionType()) {
                     // We currently don't take inbound connections into account. Since they are
                     // free to make, an attacker could make them to prevent us from connecting to
                     // certain peers.
@@ -2806,6 +2795,37 @@ CNode::CNode(NodeId idIn,
     }
 }
 
+void CNode::MarkReceivedMsgsForProcessing(unsigned int recv_flood_size)
+{
+    AssertLockNotHeld(m_msg_process_queue_mutex);
+
+    size_t nSizeAdded = 0;
+    for (const auto& msg : vRecvMsg) {
+        // vRecvMsg contains only completed CNetMessage
+        // the single possible partially deserialized message are held by TransportDeserializer
+        nSizeAdded += msg.m_raw_message_size;
+    }
+
+    LOCK(m_msg_process_queue_mutex);
+    m_msg_process_queue.splice(m_msg_process_queue.end(), vRecvMsg);
+    m_msg_process_queue_size += nSizeAdded;
+    fPauseRecv = m_msg_process_queue_size > recv_flood_size;
+}
+
+std::optional<std::pair<CNetMessage, bool>> CNode::PollMessage(size_t recv_flood_size)
+{
+    LOCK(m_msg_process_queue_mutex);
+    if (m_msg_process_queue.empty()) return std::nullopt;
+
+    std::list<CNetMessage> msgs;
+    // Just take one message
+    msgs.splice(msgs.begin(), m_msg_process_queue, m_msg_process_queue.begin());
+    m_msg_process_queue_size -= msgs.front().m_raw_message_size;
+    fPauseRecv = m_msg_process_queue_size > recv_flood_size;
+
+    return std::make_pair(std::move(msgs.front()), !m_msg_process_queue.empty());
+}
+
 bool CConnman::NodeFullyConnected(const CNode* pnode)
 {
     return pnode && pnode->fSuccessfullyConnected && !pnode->fDisconnect;
@@ -2840,7 +2860,7 @@ void CConnman::PushMessage(CNode* pnode, CSerializedNetMsg&& msg)
         bool optimisticSend(pnode->vSendMsg.empty());
 
         //log total amount of bytes per message type
-        pnode->mapSendBytesPerMsgType[msg.m_type] += nTotalSize;
+        pnode->AccountForSentBytes(msg.m_type, nTotalSize);
         pnode->nSendSize += nTotalSize;
 
         if (pnode->nSendSize > nSendBufferMaxSize) pnode->fPauseSend = true;

src/net.h

@@ -347,9 +347,6 @@ struct CNodeOptions
 /** Information about a peer */
 class CNode
 {
-    friend class CConnman;
-    friend struct ConnmanTestMsg;
-
 public:
     const std::unique_ptr<TransportDeserializer> m_deserializer; // Used only by SocketHandler thread
     const std::unique_ptr<const TransportSerializer> m_serializer;
@@ -376,10 +373,6 @@ public:
     Mutex m_sock_mutex;
     Mutex cs_vRecv;
 
-    RecursiveMutex cs_vProcessMsg;
-    std::list<CNetMessage> vProcessMsg GUARDED_BY(cs_vProcessMsg);
-    size_t nProcessQueueSize GUARDED_BY(cs_vProcessMsg){0};
-
     uint64_t nRecvBytes GUARDED_BY(cs_vRecv){0};
 
     std::atomic<std::chrono::seconds> m_last_send{0s};
@@ -417,6 +410,30 @@ public:
     std::atomic_bool fPauseRecv{false};
     std::atomic_bool fPauseSend{false};
 
+    const ConnectionType& GetConnectionType() const
+    {
+        return m_conn_type;
+    }
+
+    /** Move all messages from the received queue to the processing queue. */
+    void MarkReceivedMsgsForProcessing(unsigned int recv_flood_size)
+        EXCLUSIVE_LOCKS_REQUIRED(!m_msg_process_queue_mutex);
+
+    /** Poll the next message from the processing queue of this connection.
+     *
+     * Returns std::nullopt if the processing queue is empty, or a pair
+     * consisting of the message and a bool that indicates if the processing
+     * queue has more entries. */
+    std::optional<std::pair<CNetMessage, bool>> PollMessage(size_t recv_flood_size)
+        EXCLUSIVE_LOCKS_REQUIRED(!m_msg_process_queue_mutex);
+
+    /** Account for the total size of a sent message in the per msg type connection stats. */
+    void AccountForSentBytes(const std::string& msg_type, size_t sent_bytes)
+        EXCLUSIVE_LOCKS_REQUIRED(cs_vSend)
+    {
+        mapSendBytesPerMsgType[msg_type] += sent_bytes;
+    }
+
     bool IsOutboundOrBlockRelayConn() const {
         switch (m_conn_type) {
             case ConnectionType::OUTBOUND_FULL_RELAY:
@@ -602,6 +619,10 @@ private:
 
     std::list<CNetMessage> vRecvMsg; // Used only by SocketHandler thread
 
+    Mutex m_msg_process_queue_mutex;
+    std::list<CNetMessage> m_msg_process_queue GUARDED_BY(m_msg_process_queue_mutex);
+    size_t m_msg_process_queue_size GUARDED_BY(m_msg_process_queue_mutex){0};
+
     // Our address, as reported by the peer
     CService addrLocal GUARDED_BY(m_addr_local_mutex);
     mutable Mutex m_addr_local_mutex;

src/net_processing.cpp

@@ -4860,8 +4860,6 @@ bool PeerManagerImpl::ProcessMessages(CNode* pfrom, std::atomic<bool>& interrupt
 {
     AssertLockHeld(g_msgproc_mutex);
 
-    bool fMoreWork = false;
-
     PeerRef peer = GetPeerRef(pfrom->GetId());
     if (peer == nullptr) return false;
 
@@ -4889,17 +4887,14 @@ bool PeerManagerImpl::ProcessMessages(CNode* pfrom, std::atomic<bool>& interrupt
     // Don't bother if send buffer is too full to respond anyway
     if (pfrom->fPauseSend) return false;
 
-    std::list<CNetMessage> msgs;
+    auto poll_result{pfrom->PollMessage(m_connman.GetReceiveFloodSize())};
-    {
+    if (!poll_result) {
-        LOCK(pfrom->cs_vProcessMsg);
+        // No message to process
-        if (pfrom->vProcessMsg.empty()) return false;
+        return false;
-        // Just take one message
-        msgs.splice(msgs.begin(), pfrom->vProcessMsg, pfrom->vProcessMsg.begin());
-        pfrom->nProcessQueueSize -= msgs.front().m_raw_message_size;
-        pfrom->fPauseRecv = pfrom->nProcessQueueSize > m_connman.GetReceiveFloodSize();
-        fMoreWork = !pfrom->vProcessMsg.empty();
     }
-    CNetMessage& msg(msgs.front());
+
+    CNetMessage& msg{poll_result->first};
+    bool fMoreWork = poll_result->second;
 
     TRACE6(net, inbound_message,
         pfrom->GetId(),

src/test/util/net.cpp

@@ -66,18 +66,7 @@ void ConnmanTestMsg::NodeReceiveMsgBytes(CNode& node, Span<const uint8_t> msg_by
 {
     assert(node.ReceiveMsgBytes(msg_bytes, complete));
     if (complete) {
-        size_t nSizeAdded = 0;
+        node.MarkReceivedMsgsForProcessing(nReceiveFloodSize);
-        for (const auto& msg : node.vRecvMsg) {
-            // vRecvMsg contains only completed CNetMessage
-            // the single possible partially deserialized message are held by TransportDeserializer
-            nSizeAdded += msg.m_raw_message_size;
-        }
-        {
-            LOCK(node.cs_vProcessMsg);
-            node.vProcessMsg.splice(node.vProcessMsg.end(), node.vRecvMsg);
-            node.nProcessQueueSize += nSizeAdded;
-            node.fPauseRecv = node.nProcessQueueSize > nReceiveFloodSize;
-        }
     }
 }