net: change CNetAddr::ip to have flexible size

Before this change `CNetAddr::ip` was a fixed-size array of 16 bytes,
not being able to store larger addresses (e.g. TORv3) and encoded
smaller ones as 16-byte IPv6 addresses.

Change its type to `prevector`, so that it can hold larger addresses and
do not disguise non-IPv6 addresses as IPv6. So the IPv4 address
`1.2.3.4` is now encoded as `01020304` instead of
`00000000000000000000FFFF01020304`.

Rename `CNetAddr::ip` to `CNetAddr::m_addr` because it is not an "IP" or
"IP address" (TOR addresses are not IP addresses).

In order to preserve backward compatibility with serialization (where
e.g. `1.2.3.4` is serialized as `00000000000000000000FFFF01020304`)
introduce `CNetAddr` dedicated legacy serialize/unserialize methods.

Adjust `CSubNet` accordingly. Still use `CSubNet::netmask[]` of fixed 16
bytes, but use the first 4 for IPv4 (not the last 4). Only allow
subnetting for IPv4 and IPv6.

Co-authored-by: Carl Dong <[email protected]>

Author: vasild

src/netaddress.cpp

@@ -9,9 +9,12 @@
 #include <config/bitcoin-config.h>
 #endif
 
+#include <attributes.h>
 #include <compat.h>
+#include <prevector.h>
 #include <serialize.h>
 
+#include <array>
 #include <cstdint>
 #include <string>
 #include <vector>
@@ -39,28 +42,66 @@ enum Network
     /// TORv2
     NET_ONION,
 
-    /// A set of dummy addresses that map a name to an IPv6 address. These
-    /// addresses belong to RFC4193's fc00::/7 subnet (unique-local addresses).
-    /// We use them to map a string or FQDN to an IPv6 address in CAddrMan to
-    /// keep track of which DNS seeds were used.
+    /// A set of addresses that represent the hash of a string or FQDN. We use
+    /// them in CAddrMan to keep track of which DNS seeds were used.
     NET_INTERNAL,
 
     /// Dummy value to indicate the number of NET_* constants.
     NET_MAX,
 };
 
+/// Prefix of an IPv6 address when it contains an embedded IPv4 address.
+/// Used when (un)serializing addresses in ADDRv1 format (pre-BIP155).
+static const std::array<uint8_t, 12> IPV4_IN_IPV6_PREFIX{
+    0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xFF, 0xFF
+};
+
+/// Prefix of an IPv6 address when it contains an embedded TORv2 address.
+/// Used when (un)serializing addresses in ADDRv1 format (pre-BIP155).
+/// Such dummy IPv6 addresses are guaranteed to not be publicly routable as they
+/// fall under RFC4193's fc00::/7 subnet allocated to unique-local addresses.
+static const std::array<uint8_t, 6> TORV2_IN_IPV6_PREFIX{
+    0xFD, 0x87, 0xD8, 0x7E, 0xEB, 0x43
+};
+
+/// Prefix of an IPv6 address when it contains an embedded "internal" address.
+/// Used when (un)serializing addresses in ADDRv1 format (pre-BIP155).
+/// The prefix comes from 0xFD + SHA256("bitcoin")[0:5].
+/// Such dummy IPv6 addresses are guaranteed to not be publicly routable as they
+/// fall under RFC4193's fc00::/7 subnet allocated to unique-local addresses.
+static const std::array<uint8_t, 6> INTERNAL_IN_IPV6_PREFIX{
+    0xFD, 0x6B, 0x88, 0xC0, 0x87, 0x24 // 0xFD + sha256("bitcoin")[0:5].
+};
+
+/// Size of IPv4 address (in bytes).
+static constexpr size_t ADDR_IPV4_SIZE = 4;
+
+/// Size of IPv6 address (in bytes).
+static constexpr size_t ADDR_IPV6_SIZE = 16;
+
+/// Size of TORv2 address (in bytes).
+static constexpr size_t ADDR_TORV2_SIZE = 10;
+
+/// Size of "internal" (NET_INTERNAL) address (in bytes).
+static constexpr size_t ADDR_INTERNAL_SIZE = 10;
+
 /**
  * Network address.
  */
 class CNetAddr
 {
     protected:
+        /**
+         * Raw representation of the network address.
+         * In network byte order (big endian) for IPv4 and IPv6.
+         */
+        prevector<ADDR_IPV6_SIZE, uint8_t> m_addr{ADDR_IPV6_SIZE, 0x0};
+
         /**
          * Network to which this address belongs.
          */
         Network m_net{NET_IPV6};
 
-        unsigned char ip[16]; // in network byte order
         uint32_t scopeId{0}; // for scoped/link-local ipv6 addresses
 
     public:
@@ -74,13 +115,7 @@ class CNetAddr
          * (e.g. IPv4) disguised as IPv6. This encoding is used in the legacy
          * `addr` encoding.
          */
-        void SetLegacyIPv6(const uint8_t ipv6[16]);
-
-        /**
-         * Set raw IPv4 or IPv6 address (in network byte order)
-         * @note Only NET_IPV4 and NET_IPV6 are allowed for network.
-         */
-        void SetRaw(Network network, const uint8_t *data);
+        void SetLegacyIPv6(Span<const uint8_t> ipv6);
 
         bool SetInternal(const std::string& name);
 
@@ -111,7 +146,6 @@ class CNetAddr
         enum Network GetNetwork() const;
         std::string ToString() const;
         std::string ToStringIP() const;
-        unsigned int GetByte(int n) const;
         uint64_t GetHash() const;
         bool GetInAddr(struct in_addr* pipv4Addr) const;
         uint32_t GetNetClass() const;
@@ -127,7 +161,7 @@ class CNetAddr
         uint32_t GetMappedAS(const std::vector<bool> &asmap) const;
 
         std::vector<unsigned char> GetGroup(const std::vector<bool> &asmap) const;
-        std::vector<unsigned char> GetAddrBytes() const { return {std::begin(ip), std::end(ip)}; }
+        std::vector<unsigned char> GetAddrBytes() const;
         int GetReachabilityFrom(const CNetAddr *paddrPartner = nullptr) const;
 
         explicit CNetAddr(const struct in6_addr& pipv6Addr, const uint32_t scope = 0);
@@ -143,7 +177,7 @@ class CNetAddr
         template <typename Stream>
         void Serialize(Stream& s) const
         {
-            s << ip;
+            SerializeV1Stream(s);
         }
 
         /**
@@ -152,14 +186,92 @@ class CNetAddr
         template <typename Stream>
         void Unserialize(Stream& s)
         {
-            unsigned char ip_temp[sizeof(ip)];
-            s >> ip_temp;
+            UnserializeV1Stream(s);
+        }
+
+        friend class CSubNet;
+
+    private:
+        /**
+         * Size of CNetAddr when serialized as ADDRv1 (pre-BIP155) (in bytes).
+         */
+        static constexpr size_t V1_SERIALIZATION_SIZE = ADDR_IPV6_SIZE;
+
+        /**
+         * Serialize in pre-ADDRv2/BIP155 format to an array.
+         * Some addresses (e.g. TORv3) cannot be serialized in pre-BIP155 format.
+         */
+        void SerializeV1Array(uint8_t (&arr)[V1_SERIALIZATION_SIZE]) const
+        {
+            size_t prefix_size;
+
+            switch (m_net) {
+            case NET_IPV6:
+                assert(m_addr.size() == sizeof(arr));
+                memcpy(arr, m_addr.data(), m_addr.size());
+                return;
+            case NET_IPV4:
+                prefix_size = sizeof(IPV4_IN_IPV6_PREFIX);
+                assert(prefix_size + m_addr.size() == sizeof(arr));
+                memcpy(arr, IPV4_IN_IPV6_PREFIX.data(), prefix_size);
+                memcpy(arr + prefix_size, m_addr.data(), m_addr.size());
+                return;
+            case NET_ONION:
+                prefix_size = sizeof(TORV2_IN_IPV6_PREFIX);
+                assert(prefix_size + m_addr.size() == sizeof(arr));
+                memcpy(arr, TORV2_IN_IPV6_PREFIX.data(), prefix_size);
+                memcpy(arr + prefix_size, m_addr.data(), m_addr.size());
+                return;
+            case NET_INTERNAL:
+                prefix_size = sizeof(INTERNAL_IN_IPV6_PREFIX);
+                assert(prefix_size + m_addr.size() == sizeof(arr));
+                memcpy(arr, INTERNAL_IN_IPV6_PREFIX.data(), prefix_size);
+                memcpy(arr + prefix_size, m_addr.data(), m_addr.size());
+                return;
+            case NET_UNROUTABLE:
+            case NET_MAX:
+                assert(false);
+            } // no default case, so the compiler can warn about missing cases
+
+            assert(false);
+        }
+
+        /**
+         * Serialize in pre-ADDRv2/BIP155 format to a stream.
+         * Some addresses (e.g. TORv3) cannot be serialized in pre-BIP155 format.
+         */
+        template <typename Stream>
+        void SerializeV1Stream(Stream& s) const
+        {
+            uint8_t serialized[V1_SERIALIZATION_SIZE];
+
+            SerializeV1Array(serialized);
+
+            s << serialized;
+        }
+
+        /**
+         * Unserialize from a pre-ADDRv2/BIP155 format from an array.
+         */
+        void UnserializeV1Array(uint8_t (&arr)[V1_SERIALIZATION_SIZE])
+        {
             // Use SetLegacyIPv6() so that m_net is set correctly. For example
             // ::FFFF:0102:0304 should be set as m_net=NET_IPV4 (1.2.3.4).
-            SetLegacyIPv6(ip_temp);
+            SetLegacyIPv6(arr);
         }
 
-        friend class CSubNet;
+        /**
+         * Unserialize from a pre-ADDRv2/BIP155 format from a stream.
+         */
+        template <typename Stream>
+        void UnserializeV1Stream(Stream& s)
+        {
+            uint8_t serialized[V1_SERIALIZATION_SIZE];
+
+            s >> serialized;
+
+            UnserializeV1Array(serialized);
+        }
 };
 
 class CSubNet
@@ -174,11 +286,11 @@ class CSubNet
 
     public:
         CSubNet();
-        CSubNet(const CNetAddr &addr, int32_t mask);
-        CSubNet(const CNetAddr &addr, const CNetAddr &mask);
+        CSubNet(const CNetAddr& addr, uint8_t mask);
+        CSubNet(const CNetAddr& addr, const CNetAddr& mask);
 
         //constructor for single ip subnet (<ipv4>/32 or <ipv6>/128)
-        explicit CSubNet(const CNetAddr &addr);
+        explicit CSubNet(const CNetAddr& addr);
 
         bool Match(const CNetAddr &addr) const;
 

src/netaddress.h

@@ -13,6 +13,7 @@
 
 #include <atomic>
 #include <cstdint>
+#include <limits>
 
 #ifndef WIN32
 #include <fcntl.h>
@@ -838,8 +839,8 @@ bool LookupSubNet(const std::string& strSubnet, CSubNet& ret)
         if (slash != strSubnet.npos)
         {
             std::string strNetmask = strSubnet.substr(slash + 1);
-            int32_t n;
-            if (ParseInt32(strNetmask, &n)) {
+            uint8_t n;
+            if (ParseUInt8(strNetmask, &n)) {
                 // If valid number, assume CIDR variable-length subnet masking
                 ret = CSubNet(network, n);
                 return ret.IsValid();

src/netbase.cpp

@@ -33,7 +33,7 @@ void test_one_input(const std::vector<uint8_t>& buffer)
     if (buffer.size() < 1 + 3 + 4) return;
     int asmap_size = 3 + (buffer[0] & 127);
     bool ipv6 = buffer[0] & 128;
-    int addr_size = ipv6 ? 16 : 4;
+    const size_t addr_size = ipv6 ? ADDR_IPV6_SIZE : ADDR_IPV4_SIZE;
     if (buffer.size() < size_t(1 + asmap_size + addr_size)) return;
     std::vector<bool> asmap = ipv6 ? IPV6_PREFIX_ASMAP : IPV4_PREFIX_ASMAP;
     asmap.reserve(asmap.size() + 8 * asmap_size);
@@ -43,7 +43,17 @@ void test_one_input(const std::vector<uint8_t>& buffer)
         }
     }
     if (!SanityCheckASMap(asmap)) return;
+
+    const uint8_t* addr_data = buffer.data() + 1 + asmap_size;
     CNetAddr net_addr;
-    net_addr.SetRaw(ipv6 ? NET_IPV6 : NET_IPV4, buffer.data() + 1 + asmap_size);
+    if (ipv6) {
+        assert(addr_size == ADDR_IPV6_SIZE);
+        net_addr.SetLegacyIPv6(Span<const uint8_t>(addr_data, addr_size));
+    } else {
+        assert(addr_size == ADDR_IPV4_SIZE);
+        in_addr ipv4;
+        memcpy(&ipv4, addr_data, addr_size);
+        net_addr.SetIP(CNetAddr{ipv4});
+    }
     (void)net_addr.GetMappedAS(asmap);
 }

src/test/fuzz/asmap.cpp

@@ -17,9 +17,6 @@ void test_one_input(const std::vector<uint8_t>& buffer)
     FuzzedDataProvider fuzzed_data_provider(buffer.data(), buffer.size());
 
     const CNetAddr net_addr = ConsumeNetAddr(fuzzed_data_provider);
-    for (int i = 0; i < 15; ++i) {
-        (void)net_addr.GetByte(i);
-    }
     (void)net_addr.GetHash();
     (void)net_addr.GetNetClass();
     if (net_addr.GetNetwork() == Network::NET_IPV4) {
@@ -78,7 +75,7 @@ void test_one_input(const std::vector<uint8_t>& buffer)
     (void)net_addr.ToString();
     (void)net_addr.ToStringIP();
 
-    const CSubNet sub_net{net_addr, fuzzed_data_provider.ConsumeIntegral<int32_t>()};
+    const CSubNet sub_net{net_addr, fuzzed_data_provider.ConsumeIntegral<uint8_t>()};
     (void)sub_net.IsValid();
     (void)sub_net.ToString();
 

src/test/fuzz/netaddress.cpp

@@ -257,7 +257,7 @@ CNetAddr ConsumeNetAddr(FuzzedDataProvider& fuzzed_data_provider) noexcept
 
 CSubNet ConsumeSubNet(FuzzedDataProvider& fuzzed_data_provider) noexcept
 {
-    return {ConsumeNetAddr(fuzzed_data_provider), fuzzed_data_provider.ConsumeIntegral<int32_t>()};
+    return {ConsumeNetAddr(fuzzed_data_provider), fuzzed_data_provider.ConsumeIntegral<uint8_t>()};
 }
 
 void InitializeFuzzingContext(const std::string& chain_name = CBaseChainParams::REGTEST)