src/bloom.cpp

   1 // Copyright (c) 2012-2016 The Bitcoin Core developers
   2 // Distributed under the MIT software license, see the accompanying
   3 // file COPYING or http://www.opensource.org/licenses/mit-license.php.
   4
   5 #include "bloom.h"
   6
   7 #include "primitives/transaction.h"
   8 #include "hash.h"
   9 #include "script/script.h"
  10 #include "script/standard.h"
  11 #include "random.h"
  12 #include "streams.h"
  13
  14 #include <math.h>
  15 #include <stdlib.h>
  16
  17 #include <boost/foreach.hpp>
  18
  19 #define LN2SQUARED 0.4804530139182014246671025263266649717305529515945455
  20 #define LN2 0.6931471805599453094172321214581765680755001343602552
  21
  22 CBloomFilter::CBloomFilter(unsigned int nElements, double nFPRate, unsigned int nTweakIn, unsigned char nFlagsIn) :
  23     /**
  24      * The ideal size for a bloom filter with a given number of elements and false positive rate is:
  25      * - nElements * log(fp rate) / ln(2)^2
  26      * We ignore filter parameters which will create a bloom filter larger than the protocol limits
  27      */
  28     vData(std::min((unsigned int)(-1  / LN2SQUARED * nElements * log(nFPRate)), MAX_BLOOM_FILTER_SIZE * 8) / 8),
  29     /**
  30      * The ideal number of hash functions is filter size * ln(2) / number of elements
  31      * Again, we ignore filter parameters which will create a bloom filter with more hash functions than the protocol limits
  32      * See https://en.wikipedia.org/wiki/Bloom_filter for an explanation of these formulas
  33      */
  34     isFull(false),
  35     isEmpty(true),
  36     nHashFuncs(std::min((unsigned int)(vData.size() * 8 / nElements * LN2), MAX_HASH_FUNCS)),
  37     nTweak(nTweakIn),
  38     nFlags(nFlagsIn)
  39 {
  40 }
  41
  42 // Private constructor used by CRollingBloomFilter
  43 CBloomFilter::CBloomFilter(unsigned int nElements, double nFPRate, unsigned int nTweakIn) :
  44     vData((unsigned int)(-1  / LN2SQUARED * nElements * log(nFPRate)) / 8),
  45     isFull(false),
  46     isEmpty(true),
  47     nHashFuncs((unsigned int)(vData.size() * 8 / nElements * LN2)),
  48     nTweak(nTweakIn),
  49     nFlags(BLOOM_UPDATE_NONE)
  50 {
  51 }
  52
  53 inline unsigned int CBloomFilter::Hash(unsigned int nHashNum, const std::vector<unsigned char>& vDataToHash) const
  54 {
  55     // 0xFBA4C795 chosen as it guarantees a reasonable bit difference between nHashNum values.
  56     return MurmurHash3(nHashNum * 0xFBA4C795 + nTweak, vDataToHash) % (vData.size() * 8);
  57 }
  58
  59 void CBloomFilter::insert(const std::vector<unsigned char>& vKey)
  60 {
  61     if (isFull)
  62         return;
  63     for (unsigned int i = 0; i < nHashFuncs; i++)
  64     {
  65         unsigned int nIndex = Hash(i, vKey);
  66         // Sets bit nIndex of vData
  67         vData[nIndex >> 3] |= (1 << (7 & nIndex));
  68     }
  69     isEmpty = false;
  70 }
  71
  72 void CBloomFilter::insert(const COutPoint& outpoint)
  73 {
  74     CDataStream stream(SER_NETWORK, PROTOCOL_VERSION);
  75     stream << outpoint;
  76     std::vector<unsigned char> data(stream.begin(), stream.end());
  77     insert(data);
  78 }
  79
  80 void CBloomFilter::insert(const uint256& hash)
  81 {
  82     std::vector<unsigned char> data(hash.begin(), hash.end());
  83     insert(data);
  84 }
  85
  86 bool CBloomFilter::contains(const std::vector<unsigned char>& vKey) const
  87 {
  88     if (isFull)
  89         return true;
  90     if (isEmpty)
  91         return false;
  92     for (unsigned int i = 0; i < nHashFuncs; i++)
  93     {
  94         unsigned int nIndex = Hash(i, vKey);
  95         // Checks bit nIndex of vData
  96         if (!(vData[nIndex >> 3] & (1 << (7 & nIndex))))
  97             return false;
  98     }
  99     return true;
 100 }
 101
 102 bool CBloomFilter::contains(const COutPoint& outpoint) const
 103 {
 104     CDataStream stream(SER_NETWORK, PROTOCOL_VERSION);
 105     stream << outpoint;
 106     std::vector<unsigned char> data(stream.begin(), stream.end());
 107     return contains(data);
 108 }
 109
 110 bool CBloomFilter::contains(const uint256& hash) const
 111 {
 112     std::vector<unsigned char> data(hash.begin(), hash.end());
 113     return contains(data);
 114 }
 115
 116 void CBloomFilter::clear()
 117 {
 118     vData.assign(vData.size(),0);
 119     isFull = false;
 120     isEmpty = true;
 121 }
 122
 123 void CBloomFilter::reset(unsigned int nNewTweak)
 124 {
 125     clear();
 126     nTweak = nNewTweak;
 127 }
 128
 129 bool CBloomFilter::IsWithinSizeConstraints() const
 130 {
 131     return vData.size() <= MAX_BLOOM_FILTER_SIZE && nHashFuncs <= MAX_HASH_FUNCS;
 132 }
 133
 134 bool CBloomFilter::IsRelevantAndUpdate(const CTransaction& tx)
 135 {
 136     bool fFound = false;
 137     // Match if the filter contains the hash of tx
 138     //  for finding tx when they appear in a block
 139     if (isFull)
 140         return true;
 141     if (isEmpty)
 142         return false;
 143     const uint256& hash = tx.GetHash();
 144     if (contains(hash))
 145         fFound = true;
 146
 147     for (unsigned int i = 0; i < tx.vout.size(); i++)
 148     {
 149         const CTxOut& txout = tx.vout[i];
 150         // Match if the filter contains any arbitrary script data element in any scriptPubKey in tx
 151         // If this matches, also add the specific output that was matched.
 152         // This means clients don't have to update the filter themselves when a new relevant tx
 153         // is discovered in order to find spending transactions, which avoids round-tripping and race conditions.
 154         CScript::const_iterator pc = txout.scriptPubKey.begin();
 155         std::vector<unsigned char> data;
 156         while (pc < txout.scriptPubKey.end())
 157         {
 158             opcodetype opcode;
 159             if (!txout.scriptPubKey.GetOp(pc, opcode, data))
 160                 break;
 161             if (data.size() != 0 && contains(data))
 162             {
 163                 fFound = true;
 164                 if ((nFlags & BLOOM_UPDATE_MASK) == BLOOM_UPDATE_ALL)
 165                     insert(COutPoint(hash, i));
 166                 else if ((nFlags & BLOOM_UPDATE_MASK) == BLOOM_UPDATE_P2PUBKEY_ONLY)
 167                 {
 168                     txnouttype type;
 169                     std::vector<std::vector<unsigned char> > vSolutions;
 170                     if (Solver(txout.scriptPubKey, type, vSolutions) &&
 171                             (type == TX_PUBKEY || type == TX_MULTISIG))
 172                         insert(COutPoint(hash, i));
 173                 }
 174                 break;
 175             }
 176         }
 177     }
 178
 179     if (fFound)
 180         return true;
 181
 182     BOOST_FOREACH(const CTxIn& txin, tx.vin)
 183     {
 184         // Match if the filter contains an outpoint tx spends
 185         if (contains(txin.prevout))
 186             return true;
 187
 188         // Match if the filter contains any arbitrary script data element in any scriptSig in tx
 189         CScript::const_iterator pc = txin.scriptSig.begin();
 190         std::vector<unsigned char> data;
 191         while (pc < txin.scriptSig.end())
 192         {
 193             opcodetype opcode;
 194             if (!txin.scriptSig.GetOp(pc, opcode, data))
 195                 break;
 196             if (data.size() != 0 && contains(data))
 197                 return true;
 198         }
 199     }
 200
 201     return false;
 202 }
 203
 204 void CBloomFilter::UpdateEmptyFull()
 205 {
 206     bool full = true;
 207     bool empty = true;
 208     for (unsigned int i = 0; i < vData.size(); i++)
 209     {
 210         full &= vData[i] == 0xff;
 211         empty &= vData[i] == 0;
 212     }
 213     isFull = full;
 214     isEmpty = empty;
 215 }
 216
 217 CRollingBloomFilter::CRollingBloomFilter(unsigned int nElements, double fpRate)
 218 {
 219     double logFpRate = log(fpRate);
 220     /* The optimal number of hash functions is log(fpRate) / log(0.5), but
 221      * restrict it to the range 1-50. */
 222     nHashFuncs = std::max(1, std::min((int)round(logFpRate / log(0.5)), 50));
 223     /* In this rolling bloom filter, we'll store between 2 and 3 generations of nElements / 2 entries. */
 224     nEntriesPerGeneration = (nElements + 1) / 2;
 225     uint32_t nMaxElements = nEntriesPerGeneration * 3;
 226     /* The maximum fpRate = pow(1.0 - exp(-nHashFuncs * nMaxElements / nFilterBits), nHashFuncs)
 227      * =>          pow(fpRate, 1.0 / nHashFuncs) = 1.0 - exp(-nHashFuncs * nMaxElements / nFilterBits)
 228      * =>          1.0 - pow(fpRate, 1.0 / nHashFuncs) = exp(-nHashFuncs * nMaxElements / nFilterBits)
 229      * =>          log(1.0 - pow(fpRate, 1.0 / nHashFuncs)) = -nHashFuncs * nMaxElements / nFilterBits
 230      * =>          nFilterBits = -nHashFuncs * nMaxElements / log(1.0 - pow(fpRate, 1.0 / nHashFuncs))
 231      * =>          nFilterBits = -nHashFuncs * nMaxElements / log(1.0 - exp(logFpRate / nHashFuncs))
 232      */
 233     uint32_t nFilterBits = (uint32_t)ceil(-1.0 * nHashFuncs * nMaxElements / log(1.0 - exp(logFpRate / nHashFuncs)));
 234     data.clear();
 235     /* For each data element we need to store 2 bits. If both bits are 0, the
 236      * bit is treated as unset. If the bits are (01), (10), or (11), the bit is
 237      * treated as set in generation 1, 2, or 3 respectively.
 238      * These bits are stored in separate integers: position P corresponds to bit
 239      * (P & 63) of the integers data[(P >> 6) * 2] and data[(P >> 6) * 2 + 1]. */
 240     data.resize(((nFilterBits + 63) / 64) << 1);
 241     reset();
 242 }
 243
 244 /* Similar to CBloomFilter::Hash */
 245 static inline uint32_t RollingBloomHash(unsigned int nHashNum, uint32_t nTweak, const std::vector<unsigned char>& vDataToHash) {
 246     return MurmurHash3(nHashNum * 0xFBA4C795 + nTweak, vDataToHash);
 247 }
 248
 249 void CRollingBloomFilter::insert(const std::vector<unsigned char>& vKey)
 250 {
 251     if (nEntriesThisGeneration == nEntriesPerGeneration) {
 252         nEntriesThisGeneration = 0;
 253         nGeneration++;
 254         if (nGeneration == 4) {
 255             nGeneration = 1;
 256         }
 257         uint64_t nGenerationMask1 = 0 - (uint64_t)(nGeneration & 1);
 258         uint64_t nGenerationMask2 = 0 - (uint64_t)(nGeneration >> 1);
 259         /* Wipe old entries that used this generation number. */
 260         for (uint32_t p = 0; p < data.size(); p += 2) {
 261             uint64_t p1 = data[p], p2 = data[p + 1];
 262             uint64_t mask = (p1 ^ nGenerationMask1) | (p2 ^ nGenerationMask2);
 263             data[p] = p1 & mask;
 264             data[p + 1] = p2 & mask;
 265         }
 266     }
 267     nEntriesThisGeneration++;
 268
 269     for (int n = 0; n < nHashFuncs; n++) {
 270         uint32_t h = RollingBloomHash(n, nTweak, vKey);
 271         int bit = h & 0x3F;
 272         uint32_t pos = (h >> 6) % data.size();
 273         /* The lowest bit of pos is ignored, and set to zero for the first bit, and to one for the second. */
 274         data[pos & ~1] = (data[pos & ~1] & ~(((uint64_t)1) << bit)) | ((uint64_t)(nGeneration & 1)) << bit;
 275         data[pos | 1] = (data[pos | 1] & ~(((uint64_t)1) << bit)) | ((uint64_t)(nGeneration >> 1)) << bit;
 276     }
 277 }
 278
 279 void CRollingBloomFilter::insert(const uint256& hash)
 280 {
 281     std::vector<unsigned char> vData(hash.begin(), hash.end());
 282     insert(vData);
 283 }
 284
 285 bool CRollingBloomFilter::contains(const std::vector<unsigned char>& vKey) const
 286 {
 287     for (int n = 0; n < nHashFuncs; n++) {
 288         uint32_t h = RollingBloomHash(n, nTweak, vKey);
 289         int bit = h & 0x3F;
 290         uint32_t pos = (h >> 6) % data.size();
 291         /* If the relevant bit is not set in either data[pos & ~1] or data[pos | 1], the filter does not contain vKey */
 292         if (!(((data[pos & ~1] | data[pos | 1]) >> bit) & 1)) {
 293             return false;
 294         }
 295     }
 296     return true;
 297 }
 298
 299 bool CRollingBloomFilter::contains(const uint256& hash) const
 300 {
 301     std::vector<unsigned char> vData(hash.begin(), hash.end());
 302     return contains(vData);
 303 }
 304
 305 void CRollingBloomFilter::reset()
 306 {
 307     nTweak = GetRand(std::numeric_limits<unsigned int>::max());
 308     nEntriesThisGeneration = 0;
 309     nGeneration = 1;
 310     for (std::vector<uint64_t>::iterator it = data.begin(); it != data.end(); it++) {
 311         *it = 0;
 312     }
 313 }