Implement full filter for block based table.

Summary:
1. Make filter_block.h a base class. Derive block_based_filter_block and full_filter_block. The previous one is the traditional filter block. The full_filter_block is newly added. It would generate a filter block that contain all the keys in SST file.

2. When querying a key, table would first check if full_filter is available. If not, it would go to the exact data block and check using block_based filter.

3. User could choose to use full_filter or tradional(block_based_filter). They would be stored in SST file with different meta index name. "filter.filter_policy" or "full_filter.filter_policy". Then, Table reader is able to know the fllter block type.

4. Some optimizations have been done for full_filter_block, thus it requires a different interface compared to the original one in filter_policy.h.

5. Actual implementation of filter bits coding/decoding is placed in util/bloom_impl.cc

Benchmark: base commit 1d23b5c470
Command:
db_bench --db=/dev/shm/rocksdb --num_levels=6 --key_size=20 --prefix_size=20 --keys_per_prefix=0 --value_size=100 --write_buffer_size=134217728 --max_write_buffer_number=2 --target_file_size_base=33554432 --max_bytes_for_level_base=1073741824 --verify_checksum=false --max_background_compactions=4 --use_plain_table=0 --memtablerep=prefix_hash --open_files=-1 --mmap_read=1 --mmap_write=0 --bloom_bits=10 --bloom_locality=1 --memtable_bloom_bits=500000 --compression_type=lz4 --num=393216000 --use_hash_search=1 --block_size=1024 --block_restart_interval=16 --use_existing_db=1 --threads=1 --benchmarks=readrandom —disable_auto_compactions=1
Read QPS increase for about 30% from 2230002 to 2991411.

Test Plan:
make all check
valgrind db_test
db_stress --use_block_based_filter = 0
./auto_sanity_test.sh

Reviewers: igor, yhchiang, ljin, sdong

Reviewed By: sdong

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D20979

Makefile

@@ -90,7 +90,8 @@ TESTS = \
 	blob_store_test \
 	filelock_test \
 	filename_test \
-	filter_block_test \
+	block_based_filter_block_test \
+	full_filter_block_test \
 	histogram_test \
 	log_test \
 	manual_compaction_test \
@@ -393,8 +394,11 @@ rate_limiter_test: util/rate_limiter_test.o $(LIBOBJECTS) $(TESTHARNESS)
 filename_test: db/filename_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(CXX) db/filename_test.o $(LIBOBJECTS) $(TESTHARNESS) $(EXEC_LDFLAGS) -o $@ $(LDFLAGS) $(COVERAGEFLAGS)
 
-filter_block_test: table/filter_block_test.o $(LIBOBJECTS) $(TESTHARNESS)
+block_based_filter_block_test: table/block_based_filter_block_test.o $(LIBOBJECTS) $(TESTHARNESS)
-	$(CXX) table/filter_block_test.o $(LIBOBJECTS) $(TESTHARNESS) $(EXEC_LDFLAGS) -o $@ $(LDFLAGS) $(COVERAGEFLAGS)
+	$(CXX) table/block_based_filter_block_test.o $(LIBOBJECTS) $(TESTHARNESS) $(EXEC_LDFLAGS) -o $@ $(LDFLAGS) $(COVERAGEFLAGS)
+
+full_filter_block_test: table/full_filter_block_test.o $(LIBOBJECTS) $(TESTHARNESS)
+	$(CXX) table/full_filter_block_test.o $(LIBOBJECTS) $(TESTHARNESS) $(EXEC_LDFLAGS) -o $@ $(LDFLAGS) $(COVERAGEFLAGS)
 
 log_test: db/log_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(CXX) db/log_test.o $(LIBOBJECTS) $(TESTHARNESS) $(EXEC_LDFLAGS) -o $@ $(LDFLAGS) $(COVERAGEFLAGS)

db/c.cc

@@ -118,7 +118,7 @@ struct rocksdb_compactionfilter_t : public CompactionFilter {
       const Slice& existing_value,
       std::string* new_value,
       bool* value_changed) const {
-    char* c_new_value = NULL;
+    char* c_new_value = nullptr;
     size_t new_value_length = 0;
     unsigned char c_value_changed = 0;
     unsigned char result = (*filter_)(

db/db_bench.cc

@@ -39,8 +39,8 @@ int main() {
 #include "rocksdb/memtablerep.h"
 #include "rocksdb/write_batch.h"
 #include "rocksdb/slice.h"
+#include "rocksdb/filter_policy.h"
 #include "rocksdb/slice_transform.h"
-#include "rocksdb/statistics.h"
 #include "rocksdb/perf_context.h"
 #include "port/port.h"
 #include "port/stack_trace.h"
@@ -553,7 +553,9 @@ DEFINE_double(cuckoo_hash_ratio, 0.9, "Hash ratio for Cuckoo SST table.");
 DEFINE_bool(use_hash_search, false, "if use kHashSearch "
             "instead of kBinarySearch. "
             "This is valid if only we use BlockTable");
-
+DEFINE_bool(use_block_based_filter, false, "if use kBlockBasedFilter "
+            "instead of kFullFilter for filter block. "
+            "This is valid if only we use BlockTable");
 DEFINE_string(merge_operator, "", "The merge operator to use with the database."
               "If a new merge operator is specified, be sure to use fresh"
               " database The possible merge operators are defined in"
@@ -1076,9 +1078,9 @@ class Benchmark {
            (FLAGS_cache_numshardbits >= 1 ?
             NewLRUCache(FLAGS_compressed_cache_size, FLAGS_cache_numshardbits) :
             NewLRUCache(FLAGS_compressed_cache_size)) : nullptr),
-    filter_policy_(FLAGS_bloom_bits >= 0
+    filter_policy_(FLAGS_bloom_bits >= 0 ?
-                   ? NewBloomFilterPolicy(FLAGS_bloom_bits)
+        NewBloomFilterPolicy(FLAGS_bloom_bits, FLAGS_use_block_based_filter)
-                   : nullptr),
+        : nullptr),
     prefix_extractor_(NewFixedPrefixTransform(FLAGS_prefix_size)),
     num_(FLAGS_num),
     value_size_(FLAGS_value_size),

db/db_test.cc

@@ -324,21 +324,22 @@ class DBTest {
     kHashCuckoo = 7,
     kMergePut = 8,
     kFilter = 9,
-    kUncompressed = 10,
+    kFullFilter = 10,
-    kNumLevel_3 = 11,
+    kUncompressed = 11,
-    kDBLogDir = 12,
+    kNumLevel_3 = 12,
-    kWalDir = 13,
+    kDBLogDir = 13,
-    kManifestFileSize = 14,
+    kWalDir = 14,
-    kCompactOnFlush = 15,
+    kManifestFileSize = 15,
-    kPerfOptions = 16,
+    kCompactOnFlush = 16,
-    kDeletesFilterFirst = 17,
+    kPerfOptions = 17,
-    kHashSkipList = 18,
+    kDeletesFilterFirst = 18,
-    kUniversalCompaction = 19,
+    kHashSkipList = 19,
-    kCompressedBlockCache = 20,
+    kUniversalCompaction = 20,
-    kInfiniteMaxOpenFiles = 21,
+    kCompressedBlockCache = 21,
-    kxxHashChecksum = 22,
+    kInfiniteMaxOpenFiles = 22,
-    kFIFOCompaction = 23,
+    kxxHashChecksum = 23,
-    kEnd = 24
+    kFIFOCompaction = 24,
+    kEnd = 25
   };
   int option_config_;
 
@@ -448,6 +449,30 @@ class DBTest {
     }
   }
 
+  // Switch between different filter policy
+  // Jump from kDefault to kFilter to kFullFilter
+  bool ChangeFilterOptions(Options* prev_options = nullptr) {
+    if (option_config_ == kDefault) {
+      option_config_ = kFilter;
+      if (prev_options == nullptr) {
+        prev_options = &last_options_;
+      }
+      Destroy(prev_options);
+      TryReopen();
+      return true;
+    } else if (option_config_ == kFilter) {
+      option_config_ = kFullFilter;
+      if (prev_options == nullptr) {
+        prev_options = &last_options_;
+      }
+      Destroy(prev_options);
+      TryReopen();
+      return true;
+    } else {
+      return false;
+    }
+  }
+
   // Return the current option configuration.
   Options CurrentOptions(
       const anon::OptionsOverride& options_override = anon::OptionsOverride()) {
@@ -486,7 +511,10 @@ class DBTest {
         options.merge_operator = MergeOperators::CreatePutOperator();
         break;
       case kFilter:
-        table_options.filter_policy.reset(NewBloomFilterPolicy(10));
+        table_options.filter_policy.reset(NewBloomFilterPolicy(10, true));
+        break;
+      case kFullFilter:
+        table_options.filter_policy.reset(NewBloomFilterPolicy(10, false));
         break;
       case kUncompressed:
         options.compression = kNoCompression;
@@ -5744,6 +5772,92 @@ TEST(DBTest, BloomFilter) {
   } while (ChangeCompactOptions());
 }
 
+TEST(DBTest, BloomFilterRate) {
+  while (ChangeFilterOptions()) {
+    Options options = CurrentOptions();
+    options.statistics = rocksdb::CreateDBStatistics();
+    CreateAndReopenWithCF({"pikachu"}, &options);
+
+    const int maxKey = 10000;
+    for (int i = 0; i < maxKey; i++) {
+      ASSERT_OK(Put(1, Key(i), Key(i)));
+    }
+    // Add a large key to make the file contain wide range
+    ASSERT_OK(Put(1, Key(maxKey + 55555), Key(maxKey + 55555)));
+    Flush(1);
+
+    // Check if they can be found
+    for (int i = 0; i < maxKey; i++) {
+      ASSERT_EQ(Key(i), Get(1, Key(i)));
+    }
+    ASSERT_EQ(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), 0);
+
+    // Check if filter is useful
+    for (int i = 0; i < maxKey; i++) {
+      ASSERT_EQ("NOT_FOUND", Get(1, Key(i+33333)));
+    }
+    ASSERT_GE(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), maxKey*0.98);
+  }
+}
+
+TEST(DBTest, BloomFilterCompatibility) {
+  Options options;
+  options.statistics = rocksdb::CreateDBStatistics();
+  BlockBasedTableOptions table_options;
+  table_options.filter_policy.reset(NewBloomFilterPolicy(10, true));
+  options.table_factory.reset(NewBlockBasedTableFactory(table_options));
+
+  // Create with block based filter
+  CreateAndReopenWithCF({"pikachu"}, &options);
+
+  const int maxKey = 10000;
+  for (int i = 0; i < maxKey; i++) {
+    ASSERT_OK(Put(1, Key(i), Key(i)));
+  }
+  ASSERT_OK(Put(1, Key(maxKey + 55555), Key(maxKey + 55555)));
+  Flush(1);
+
+  // Check db with full filter
+  table_options.filter_policy.reset(NewBloomFilterPolicy(10, false));
+  options.table_factory.reset(NewBlockBasedTableFactory(table_options));
+  ReopenWithColumnFamilies({"default", "pikachu"}, &options);
+
+  // Check if they can be found
+  for (int i = 0; i < maxKey; i++) {
+    ASSERT_EQ(Key(i), Get(1, Key(i)));
+  }
+  ASSERT_EQ(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), 0);
+}
+
+TEST(DBTest, BloomFilterReverseCompatibility) {
+  Options options;
+  options.statistics = rocksdb::CreateDBStatistics();
+  BlockBasedTableOptions table_options;
+  table_options.filter_policy.reset(NewBloomFilterPolicy(10, false));
+  options.table_factory.reset(NewBlockBasedTableFactory(table_options));
+
+  // Create with full filter
+  CreateAndReopenWithCF({"pikachu"}, &options);
+
+  const int maxKey = 10000;
+  for (int i = 0; i < maxKey; i++) {
+    ASSERT_OK(Put(1, Key(i), Key(i)));
+  }
+  ASSERT_OK(Put(1, Key(maxKey + 55555), Key(maxKey + 55555)));
+  Flush(1);
+
+  // Check db with block_based filter
+  table_options.filter_policy.reset(NewBloomFilterPolicy(10, true));
+  options.table_factory.reset(NewBlockBasedTableFactory(table_options));
+  ReopenWithColumnFamilies({"default", "pikachu"}, &options);
+
+  // Check if they can be found
+  for (int i = 0; i < maxKey; i++) {
+    ASSERT_EQ(Key(i), Get(1, Key(i)));
+  }
+  ASSERT_EQ(TestGetTickerCount(options, BLOOM_FILTER_USEFUL), 0);
+}
+
 TEST(DBTest, SnapshotFiles) {
   do {
     Options options = CurrentOptions();
@@ -7194,47 +7308,49 @@ void PrefixScanInit(DBTest *dbtest) {
 }  // namespace
 
 TEST(DBTest, PrefixScan) {
-  int count;
+  while (ChangeFilterOptions()) {
-  Slice prefix;
+    int count;
-  Slice key;
+    Slice prefix;
-  char buf[100];
+    Slice key;
-  Iterator* iter;
+    char buf[100];
-  snprintf(buf, sizeof(buf), "03______:");
+    Iterator* iter;
-  prefix = Slice(buf, 8);
+    snprintf(buf, sizeof(buf), "03______:");
-  key = Slice(buf, 9);
+    prefix = Slice(buf, 8);
-  // db configs
+    key = Slice(buf, 9);
-  env_->count_random_reads_ = true;
+    // db configs
-  Options options = CurrentOptions();
+    env_->count_random_reads_ = true;
-  options.env = env_;
+    Options options = CurrentOptions();
-  options.prefix_extractor.reset(NewFixedPrefixTransform(8));
+    options.env = env_;
-  options.disable_auto_compactions = true;
+    options.prefix_extractor.reset(NewFixedPrefixTransform(8));
-  options.max_background_compactions = 2;
+    options.disable_auto_compactions = true;
-  options.create_if_missing = true;
+    options.max_background_compactions = 2;
-  options.memtable_factory.reset(NewHashSkipListRepFactory(16));
+    options.create_if_missing = true;
+    options.memtable_factory.reset(NewHashSkipListRepFactory(16));
 
-  BlockBasedTableOptions table_options;
+    BlockBasedTableOptions table_options;
-  table_options.no_block_cache = true;
+    table_options.no_block_cache = true;
-  table_options.filter_policy.reset(NewBloomFilterPolicy(10));
+    table_options.filter_policy.reset(NewBloomFilterPolicy(10));
-  table_options.whole_key_filtering = false;
+    table_options.whole_key_filtering = false;
-  options.table_factory.reset(NewBlockBasedTableFactory(table_options));
+    options.table_factory.reset(NewBlockBasedTableFactory(table_options));
 
-  // 11 RAND I/Os
+    // 11 RAND I/Os
-  DestroyAndReopen(&options);
+    DestroyAndReopen(&options);
-  PrefixScanInit(this);
+    PrefixScanInit(this);
-  count = 0;
+    count = 0;
-  env_->random_read_counter_.Reset();
+    env_->random_read_counter_.Reset();
-  iter = db_->NewIterator(ReadOptions());
+    iter = db_->NewIterator(ReadOptions());
-  for (iter->Seek(prefix); iter->Valid(); iter->Next()) {
+    for (iter->Seek(prefix); iter->Valid(); iter->Next()) {
-    if (! iter->key().starts_with(prefix)) {
+      if (! iter->key().starts_with(prefix)) {
-      break;
+        break;
+      }
+      count++;
     }
-    count++;
+    ASSERT_OK(iter->status());
-  }
+    delete iter;
-  ASSERT_OK(iter->status());
+    ASSERT_EQ(count, 2);
-  delete iter;
+    ASSERT_EQ(env_->random_read_counter_.Read(), 2);
-  ASSERT_EQ(count, 2);
+    Close();
-  ASSERT_EQ(env_->random_read_counter_.Read(), 2);
+  }  // end of while
-  Close();
 }
 
 TEST(DBTest, TailingIteratorSingle) {

include/rocksdb/filter_policy.h

@@ -21,11 +21,52 @@
 #define STORAGE_ROCKSDB_INCLUDE_FILTER_POLICY_H_
 
 #include <string>
+#include <memory>
 
 namespace rocksdb {
 
 class Slice;
 
+// A class that takes a bunch of keys, then generates filter
+class FilterBitsBuilder {
+ public:
+  virtual ~FilterBitsBuilder() {}
+
+  // Add Key to filter, you could use any way to store the key.
+  // Such as: storing hashes or original keys
+  // Keys are in sorted order and duplicated keys are possible.
+  virtual void AddKey(const Slice& key) = 0;
+
+  // Generate the filter using the keys that are added
+  // The return value of this function would be the filter bits,
+  // The ownership of actual data is set to buf
+  virtual Slice Finish(std::unique_ptr<const char[]>* buf) = 0;
+};
+
+// A class that checks if a key can be in filter
+// It should be initialized by Slice generated by BitsBuilder
+class FilterBitsReader {
+ public:
+  virtual ~FilterBitsReader() {}
+
+  // Check if the entry match the bits in filter
+  virtual bool MayMatch(const Slice& entry) = 0;
+};
+
+// We add a new format of filter block called full filter block
+// This new interface gives you more space of customization
+//
+// For the full filter block, you can plug in your version by implement
+// the FilterBitsBuilder and FilterBitsReader
+//
+// There are two sets of interface in FilterPolicy
+// Set 1: CreateFilter, KeyMayMatch: used for blockbased filter
+// Set 2: GetFilterBitsBuilder, GetFilterBitsReader, they are used for
+// full filter.
+// Set 1 MUST be implemented correctly, Set 2 is optional
+// RocksDB would first try using functions in Set 2. if they return nullptr,
+// it would use Set 1 instead.
+// You can choose filter type in NewBloomFilterPolicy
 class FilterPolicy {
  public:
   virtual ~FilterPolicy();
@@ -51,11 +92,28 @@ class FilterPolicy {
   // This method may return true or false if the key was not on the
   // list, but it should aim to return false with a high probability.
   virtual bool KeyMayMatch(const Slice& key, const Slice& filter) const = 0;
+
+  // Get the FilterBitsBuilder, which is ONLY used for full filter block
+  // It contains interface to take individual key, then generate filter
+  virtual FilterBitsBuilder* GetFilterBitsBuilder() const {
+    return nullptr;
+  }
+
+  // Get the FilterBitsReader, which is ONLY used for full filter block
+  // It contains interface to tell if key can be in filter
+  // The input slice should NOT be deleted by FilterPolicy
+  virtual FilterBitsReader* GetFilterBitsReader(const Slice& contents) const {
+    return nullptr;
+  }
 };
 
 // Return a new filter policy that uses a bloom filter with approximately
-// the specified number of bits per key.  A good value for bits_per_key
+// the specified number of bits per key.
+//
+// bits_per_key: bits per key in bloom filter. A good value for bits_per_key
 // is 10, which yields a filter with ~ 1% false positive rate.
+// use_block_based_builder: use block based filter rather than full fiter.
+// If you want to builder full filter, it needs to be set to false.
 //
 // Callers must delete the result after any database that is using the
 // result has been closed.
@@ -67,8 +125,8 @@ class FilterPolicy {
 // ignores trailing spaces, it would be incorrect to use a
 // FilterPolicy (like NewBloomFilterPolicy) that does not ignore
 // trailing spaces in keys.
-extern const FilterPolicy* NewBloomFilterPolicy(int bits_per_key);
+extern const FilterPolicy* NewBloomFilterPolicy(int bits_per_key,
-
+    bool use_block_based_builder = true);
 }
 
 #endif  // STORAGE_ROCKSDB_INCLUDE_FILTER_POLICY_H_

include/rocksdb/statistics.h

@@ -115,7 +115,7 @@ enum Tickers : uint32_t {
   // head of the writers queue.
   WRITE_DONE_BY_SELF,
   WRITE_DONE_BY_OTHER,
-  WRITE_TIMEDOUT,        // Number of writes ending up with timed-out.
+  WRITE_TIMEDOUT,       // Number of writes ending up with timed-out.
   WRITE_WITH_WAL,       // Number of Write calls that request WAL
   COMPACT_READ_BYTES,   // Bytes read during compaction
   COMPACT_WRITE_BYTES,  // Bytes written during compaction

table/block_based_filter_block.cc

@@ -7,7 +7,7 @@
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file. See the AUTHORS file for names of contributors.
 
-#include "table/filter_block.h"
+#include "table/block_based_filter_block.h"
 
 #include "db/dbformat.h"
 #include "rocksdb/filter_policy.h"
@@ -15,21 +15,39 @@
 
 namespace rocksdb {
 
+namespace {
+bool SamePrefix(const SliceTransform* prefix_extractor,
+                const Slice& key1, const Slice& key2) {
+  if (!prefix_extractor->InDomain(key1) &&
+      !prefix_extractor->InDomain(key2)) {
+    return true;
+  } else if (!prefix_extractor->InDomain(key1) ||
+             !prefix_extractor->InDomain(key2)) {
+    return false;
+  } else {
+    return (prefix_extractor->Transform(key1) ==
+            prefix_extractor->Transform(key2));
+  }
+}
+}  // namespace
+
+
 // See doc/table_format.txt for an explanation of the filter block format.
 
 // Generate new filter every 2KB of data
 static const size_t kFilterBaseLg = 11;
 static const size_t kFilterBase = 1 << kFilterBaseLg;
 
-FilterBlockBuilder::FilterBlockBuilder(const SliceTransform* prefix_extractor,
+BlockBasedFilterBlockBuilder::BlockBasedFilterBlockBuilder(
-                                       const BlockBasedTableOptions& table_opt,
+    const SliceTransform* prefix_extractor,
-                                       const Comparator* internal_comparator)
+    const BlockBasedTableOptions& table_opt)
     : policy_(table_opt.filter_policy.get()),
       prefix_extractor_(prefix_extractor),
-      whole_key_filtering_(table_opt.whole_key_filtering),
+      whole_key_filtering_(table_opt.whole_key_filtering) {
-      comparator_(internal_comparator) {}
+  assert(policy_);
+}
 
-void FilterBlockBuilder::StartBlock(uint64_t block_offset) {
+void BlockBasedFilterBlockBuilder::StartBlock(uint64_t block_offset) {
   uint64_t filter_index = (block_offset / kFilterBase);
   assert(filter_index >= filter_offsets_.size());
   while (filter_index > filter_offsets_.size()) {
@@ -37,53 +55,45 @@ void FilterBlockBuilder::StartBlock(uint64_t block_offset) {
   }
 }
 
-bool FilterBlockBuilder::SamePrefix(const Slice &key1,
+void BlockBasedFilterBlockBuilder::Add(const Slice& key) {
-                                    const Slice &key2) const {
+  added_to_start_ = 0;
-  if (!prefix_extractor_->InDomain(key1) &&
+  if (whole_key_filtering_) {
-      !prefix_extractor_->InDomain(key2)) {
+    AddKey(key);
-    return true;
+    added_to_start_ = 1;
-  } else if (!prefix_extractor_->InDomain(key1) ||
+  }
-             !prefix_extractor_->InDomain(key2)) {
+  if (prefix_extractor_ && prefix_extractor_->InDomain(key)) {
-    return false;
+    AddPrefix(key);
-  } else {
-    return (prefix_extractor_->Transform(key1) ==
-            prefix_extractor_->Transform(key2));
   }
 }
 
-void FilterBlockBuilder::AddKey(const Slice& key) {
+// Add key to filter if needed
+inline void BlockBasedFilterBlockBuilder::AddKey(const Slice& key) {
+  start_.push_back(entries_.size());
+  entries_.append(key.data(), key.size());
+}
+
+// Add prefix to filter if needed
+inline void BlockBasedFilterBlockBuilder::AddPrefix(const Slice& key) {
   // get slice for most recently added entry
   Slice prev;
-  size_t added_to_start = 0;
+  if (start_.size() > added_to_start_) {
-
+    size_t prev_start = start_[start_.size() - 1 - added_to_start_];
-  // add key to filter if needed
-  if (whole_key_filtering_) {
-    start_.push_back(entries_.size());
-    ++added_to_start;
-    entries_.append(key.data(), key.size());
-  }
-
-  if (start_.size() > added_to_start) {
-    size_t prev_start = start_[start_.size() - 1 - added_to_start];
     const char* base = entries_.data() + prev_start;
     size_t length = entries_.size() - prev_start;
     prev = Slice(base, length);
   }
 
-  // add prefix to filter if needed
+  // this assumes prefix(prefix(key)) == prefix(key), as the last
-  if (prefix_extractor_ && prefix_extractor_->InDomain(key)) {
+  // entry in entries_ may be either a key or prefix, and we use
-    // this assumes prefix(prefix(key)) == prefix(key), as the last
+  // prefix(last entry) to get the prefix of the last key.
-    // entry in entries_ may be either a key or prefix, and we use
+  if (prev.size() == 0 || !SamePrefix(prefix_extractor_, key, prev)) {
-    // prefix(last entry) to get the prefix of the last key.
+    Slice prefix = prefix_extractor_->Transform(key);
-    if (prev.size() == 0 || !SamePrefix(key, prev)) {
+    start_.push_back(entries_.size());
-      Slice prefix = prefix_extractor_->Transform(key);
+    entries_.append(prefix.data(), prefix.size());
-      start_.push_back(entries_.size());
-      entries_.append(prefix.data(), prefix.size());
-    }
   }
 }
 
-Slice FilterBlockBuilder::Finish() {
+Slice BlockBasedFilterBlockBuilder::Finish() {
   if (!start_.empty()) {
     GenerateFilter();
   }
@@ -99,7 +109,7 @@ Slice FilterBlockBuilder::Finish() {
   return Slice(result_);
 }
 
-void FilterBlockBuilder::GenerateFilter() {
+void BlockBasedFilterBlockBuilder::GenerateFilter() {
   const size_t num_entries = start_.size();
   if (num_entries == 0) {
     // Fast path if there are no keys for this filter
@@ -112,7 +122,7 @@ void FilterBlockBuilder::GenerateFilter() {
   tmp_entries_.resize(num_entries);
   for (size_t i = 0; i < num_entries; i++) {
     const char* base = entries_.data() + start_[i];
-    size_t length = start_[i+1] - start_[i];
+    size_t length = start_[i + 1] - start_[i];
     tmp_entries_[i] = Slice(base, length);
   }
 
@@ -125,7 +135,7 @@ void FilterBlockBuilder::GenerateFilter() {
   start_.clear();
 }
 
-FilterBlockReader::FilterBlockReader(
+BlockBasedFilterBlockReader::BlockBasedFilterBlockReader(
     const SliceTransform* prefix_extractor,
     const BlockBasedTableOptions& table_opt,
     const Slice& contents, bool delete_contents_after_use)
@@ -136,9 +146,10 @@ FilterBlockReader::FilterBlockReader(
       offset_(nullptr),
       num_(0),
       base_lg_(0) {
+  assert(policy_);
   size_t n = contents.size();
   if (n < 5) return;  // 1 byte for base_lg_ and 4 for start of offset array
-  base_lg_ = contents[n-1];
+  base_lg_ = contents[n - 1];
   uint32_t last_word = DecodeFixed32(contents.data() + n - 5);
   if (last_word > n - 5) return;
   data_ = contents.data();
@@ -149,27 +160,30 @@ FilterBlockReader::FilterBlockReader(
   }
 }
 
-bool FilterBlockReader::KeyMayMatch(uint64_t block_offset,
+bool BlockBasedFilterBlockReader::KeyMayMatch(const Slice& key,
-                                    const Slice& key) {
+                                              uint64_t block_offset) {
+  assert(block_offset != kNotValid);
   if (!whole_key_filtering_) {
     return true;
   }
-  return MayMatch(block_offset, key);
+  return MayMatch(key, block_offset);
 }
 
-bool FilterBlockReader::PrefixMayMatch(uint64_t block_offset,
+bool BlockBasedFilterBlockReader::PrefixMayMatch(const Slice& prefix,
-                                       const Slice& prefix) {
+                                                 uint64_t block_offset) {
+  assert(block_offset != kNotValid);
   if (!prefix_extractor_) {
     return true;
   }
-  return MayMatch(block_offset, prefix);
+  return MayMatch(prefix, block_offset);
 }
 
-bool FilterBlockReader::MayMatch(uint64_t block_offset, const Slice& entry) {
+bool BlockBasedFilterBlockReader::MayMatch(const Slice& entry,
+                                           uint64_t block_offset) {
   uint64_t index = block_offset >> base_lg_;
   if (index < num_) {
-    uint32_t start = DecodeFixed32(offset_ + index*4);
+    uint32_t start = DecodeFixed32(offset_ + index * 4);
-    uint32_t limit = DecodeFixed32(offset_ + index*4 + 4);
+    uint32_t limit = DecodeFixed32(offset_ + index * 4 + 4);
     if (start <= limit && limit <= (uint32_t)(offset_ - data_)) {
       Slice filter = Slice(data_ + start, limit - start);
       return policy_->KeyMayMatch(entry, filter);
@@ -181,7 +195,7 @@ bool FilterBlockReader::MayMatch(uint64_t block_offset, const Slice& entry) {
   return true;  // Errors are treated as potential matches
 }
 
-size_t FilterBlockReader::ApproximateMemoryUsage() const {
+size_t BlockBasedFilterBlockReader::ApproximateMemoryUsage() const {
   return num_ * 4 + 5 + (offset_ - data_);
 }
 }

table/block_based_filter_block.h

@@ -0,0 +1,102 @@
+//  Copyright (c) 2013, Facebook, Inc.  All rights reserved.
+//  This source code is licensed under the BSD-style license found in the
+//  LICENSE file in the root directory of this source tree. An additional grant
+//  of patent rights can be found in the PATENTS file in the same directory.
+//
+// Copyright (c) 2012 The LevelDB Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file. See the AUTHORS file for names of contributors.
+//
+// A filter block is stored near the end of a Table file.  It contains
+// filters (e.g., bloom filters) for all data blocks in the table combined
+// into a single filter block.
+
+#pragma once
+
+#include <stddef.h>
+#include <stdint.h>
+#include <string>
+#include <memory>
+#include <vector>
+#include "rocksdb/options.h"
+#include "rocksdb/slice.h"
+#include "rocksdb/slice_transform.h"
+#include "table/filter_block.h"
+#include "util/hash.h"
+
+namespace rocksdb {
+
+
+// A BlockBasedFilterBlockBuilder is used to construct all of the filters for a
+// particular Table.  It generates a single string which is stored as
+// a special block in the Table.
+//
+// The sequence of calls to BlockBasedFilterBlockBuilder must match the regexp:
+//      (StartBlock Add*)* Finish
+class BlockBasedFilterBlockBuilder : public FilterBlockBuilder {
+ public:
+  BlockBasedFilterBlockBuilder(const SliceTransform* prefix_extractor,
+      const BlockBasedTableOptions& table_opt);
+
+  virtual bool IsBlockBased() override { return true; }
+  virtual void StartBlock(uint64_t block_offset) override;
+  virtual void Add(const Slice& key) override;
+  virtual Slice Finish() override;
+
+ private:
+  void AddKey(const Slice& key);
+  void AddPrefix(const Slice& key);
+  void GenerateFilter();
+
+  // important: all of these might point to invalid addresses
+  // at the time of destruction of this filter block. destructor
+  // should NOT dereference them.
+  const FilterPolicy* policy_;
+  const SliceTransform* prefix_extractor_;
+  bool whole_key_filtering_;
+
+  std::string entries_;             // Flattened entry contents
+  std::vector<size_t> start_;       // Starting index in entries_ of each entry
+  uint32_t added_to_start_;         // To indicate if key is added
+  std::string result_;              // Filter data computed so far
+  std::vector<Slice> tmp_entries_;  // policy_->CreateFilter() argument
+  std::vector<uint32_t> filter_offsets_;
+
+  // No copying allowed
+  BlockBasedFilterBlockBuilder(const BlockBasedFilterBlockBuilder&);
+  void operator=(const BlockBasedFilterBlockBuilder&);
+};
+
+// A FilterBlockReader is used to parse filter from SST table.
+// KeyMayMatch and PrefixMayMatch would trigger filter checking
+class BlockBasedFilterBlockReader : public FilterBlockReader {
+ public:
+  // REQUIRES: "contents" and *policy must stay live while *this is live.
+  BlockBasedFilterBlockReader(const SliceTransform* prefix_extractor,
+                              const BlockBasedTableOptions& table_opt,
+                              const Slice& contents,
+                              bool delete_contents_after_use = false);
+  virtual bool IsBlockBased() override { return true; }
+  virtual bool KeyMayMatch(const Slice& key,
+                           uint64_t block_offset = kNotValid) override;
+  virtual bool PrefixMayMatch(const Slice& prefix,
+                              uint64_t block_offset = kNotValid) override;
+  virtual size_t ApproximateMemoryUsage() const override;
+
+ private:
+  const FilterPolicy* policy_;
+  const SliceTransform* prefix_extractor_;
+  bool whole_key_filtering_;
+  const char* data_;    // Pointer to filter data (at block-start)
+  const char* offset_;  // Pointer to beginning of offset array (at block-end)
+  size_t num_;          // Number of entries in offset array
+  size_t base_lg_;      // Encoding parameter (see kFilterBaseLg in .cc file)
+  std::unique_ptr<const char[]> filter_data;
+
+  bool MayMatch(const Slice& entry, uint64_t block_offset);
+
+  // No copying allowed
+  BlockBasedFilterBlockReader(const BlockBasedFilterBlockReader&);
+  void operator=(const BlockBasedFilterBlockReader&);
+};
+}  // namespace rocksdb

table/block_based_filter_block_test.cc

@@ -0,0 +1,242 @@
+//  Copyright (c) 2013, Facebook, Inc.  All rights reserved.
+//  This source code is licensed under the BSD-style license found in the
+//  LICENSE file in the root directory of this source tree. An additional grant
+//  of patent rights can be found in the PATENTS file in the same directory.
+//
+// Copyright (c) 2012 The LevelDB Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file. See the AUTHORS file for names of contributors.
+
+#include "table/block_based_filter_block.h"
+
+#include "rocksdb/filter_policy.h"
+#include "util/coding.h"
+#include "util/hash.h"
+#include "util/logging.h"
+#include "util/testharness.h"
+#include "util/testutil.h"
+
+namespace rocksdb {
+
+// For testing: emit an array with one hash value per key
+class TestHashFilter : public FilterPolicy {
+ public:
+  virtual const char* Name() const {
+    return "TestHashFilter";
+  }
+
+  virtual void CreateFilter(const Slice* keys, int n, std::string* dst) const {
+    for (int i = 0; i < n; i++) {
+      uint32_t h = Hash(keys[i].data(), keys[i].size(), 1);
+      PutFixed32(dst, h);
+    }
+  }
+
+  virtual bool KeyMayMatch(const Slice& key, const Slice& filter) const {
+    uint32_t h = Hash(key.data(), key.size(), 1);
+    for (unsigned int i = 0; i + 4 <= filter.size(); i += 4) {
+      if (h == DecodeFixed32(filter.data() + i)) {
+        return true;
+      }
+    }
+    return false;
+  }
+};
+
+class FilterBlockTest {
+ public:
+  TestHashFilter policy_;
+  BlockBasedTableOptions table_options_;
+
+  FilterBlockTest() {
+    table_options_.filter_policy.reset(new TestHashFilter());
+  }
+};
+
+TEST(FilterBlockTest, EmptyBuilder) {
+  BlockBasedFilterBlockBuilder builder(nullptr, table_options_);
+  Slice block = builder.Finish();
+  ASSERT_EQ("\\x00\\x00\\x00\\x00\\x0b", EscapeString(block));
+  BlockBasedFilterBlockReader reader(nullptr, table_options_, block);
+  ASSERT_TRUE(reader.KeyMayMatch("foo", 0));
+  ASSERT_TRUE(reader.KeyMayMatch("foo", 100000));
+}
+
+TEST(FilterBlockTest, SingleChunk) {
+  BlockBasedFilterBlockBuilder builder(nullptr, table_options_);
+  builder.StartBlock(100);
+  builder.Add("foo");
+  builder.Add("bar");
+  builder.Add("box");
+  builder.StartBlock(200);
+  builder.Add("box");
+  builder.StartBlock(300);
+  builder.Add("hello");
+  Slice block = builder.Finish();
+  BlockBasedFilterBlockReader reader(nullptr, table_options_, block);
+  ASSERT_TRUE(reader.KeyMayMatch("foo", 100));
+  ASSERT_TRUE(reader.KeyMayMatch("bar", 100));
+  ASSERT_TRUE(reader.KeyMayMatch("box", 100));
+  ASSERT_TRUE(reader.KeyMayMatch("hello", 100));
+  ASSERT_TRUE(reader.KeyMayMatch("foo", 100));
+  ASSERT_TRUE(!reader.KeyMayMatch("missing", 100));
+  ASSERT_TRUE(!reader.KeyMayMatch("other", 100));
+}
+
+TEST(FilterBlockTest, MultiChunk) {
+  BlockBasedFilterBlockBuilder builder(nullptr, table_options_);
+
+  // First filter
+  builder.StartBlock(0);
+  builder.Add("foo");
+  builder.StartBlock(2000);
+  builder.Add("bar");
+
+  // Second filter
+  builder.StartBlock(3100);
+  builder.Add("box");
+
+  // Third filter is empty
+
+  // Last filter
+  builder.StartBlock(9000);
+  builder.Add("box");
+  builder.Add("hello");
+
+  Slice block = builder.Finish();
+  BlockBasedFilterBlockReader reader(nullptr, table_options_, block);
+
+  // Check first filter
+  ASSERT_TRUE(reader.KeyMayMatch("foo", 0));
+  ASSERT_TRUE(reader.KeyMayMatch("bar", 2000));
+  ASSERT_TRUE(!reader.KeyMayMatch("box", 0));
+  ASSERT_TRUE(!reader.KeyMayMatch("hello", 0));
+
+  // Check second filter
+  ASSERT_TRUE(reader.KeyMayMatch("box", 3100));
+  ASSERT_TRUE(!reader.KeyMayMatch("foo", 3100));
+  ASSERT_TRUE(!reader.KeyMayMatch("bar", 3100));
+  ASSERT_TRUE(!reader.KeyMayMatch("hello", 3100));
+
+  // Check third filter (empty)
+  ASSERT_TRUE(!reader.KeyMayMatch("foo", 4100));
+  ASSERT_TRUE(!reader.KeyMayMatch("bar", 4100));
+  ASSERT_TRUE(!reader.KeyMayMatch("box", 4100));
+  ASSERT_TRUE(!reader.KeyMayMatch("hello", 4100));
+
+  // Check last filter
+  ASSERT_TRUE(reader.KeyMayMatch("box", 9000));
+  ASSERT_TRUE(reader.KeyMayMatch("hello", 9000));
+  ASSERT_TRUE(!reader.KeyMayMatch("foo", 9000));
+  ASSERT_TRUE(!reader.KeyMayMatch("bar", 9000));
+}
+
+// Test for block based filter block
+// use new interface in FilterPolicy to create filter builder/reader
+class BlockBasedFilterBlockTest {
+ public:
+  BlockBasedTableOptions table_options_;
+
+  BlockBasedFilterBlockTest() {
+    table_options_.filter_policy.reset(NewBloomFilterPolicy(10));
+  }
+
+  ~BlockBasedFilterBlockTest() {}
+};
+
+TEST(BlockBasedFilterBlockTest, BlockBasedEmptyBuilder) {
+  FilterBlockBuilder* builder = new BlockBasedFilterBlockBuilder(
+      nullptr, table_options_);
+  Slice block = builder->Finish();
+  ASSERT_EQ("\\x00\\x00\\x00\\x00\\x0b", EscapeString(block));
+  FilterBlockReader* reader = new BlockBasedFilterBlockReader(
+      nullptr, table_options_, block);
+  ASSERT_TRUE(reader->KeyMayMatch("foo", 0));
+  ASSERT_TRUE(reader->KeyMayMatch("foo", 100000));
+
+  delete builder;
+  delete reader;
+}
+
+TEST(BlockBasedFilterBlockTest, BlockBasedSingleChunk) {
+  FilterBlockBuilder* builder = new BlockBasedFilterBlockBuilder(
+      nullptr, table_options_);
+  builder->StartBlock(100);
+  builder->Add("foo");
+  builder->Add("bar");
+  builder->Add("box");
+  builder->StartBlock(200);
+  builder->Add("box");
+  builder->StartBlock(300);
+  builder->Add("hello");
+  Slice block = builder->Finish();
+  FilterBlockReader* reader = new BlockBasedFilterBlockReader(
+      nullptr, table_options_, block);
+  ASSERT_TRUE(reader->KeyMayMatch("foo", 100));
+  ASSERT_TRUE(reader->KeyMayMatch("bar", 100));
+  ASSERT_TRUE(reader->KeyMayMatch("box", 100));
+  ASSERT_TRUE(reader->KeyMayMatch("hello", 100));
+  ASSERT_TRUE(reader->KeyMayMatch("foo", 100));
+  ASSERT_TRUE(!reader->KeyMayMatch("missing", 100));
+  ASSERT_TRUE(!reader->KeyMayMatch("other", 100));
+
+  delete builder;
+  delete reader;
+}
+
+TEST(BlockBasedFilterBlockTest, BlockBasedMultiChunk) {
+  FilterBlockBuilder* builder = new BlockBasedFilterBlockBuilder(
+      nullptr, table_options_);
+
+  // First filter
+  builder->StartBlock(0);
+  builder->Add("foo");
+  builder->StartBlock(2000);
+  builder->Add("bar");
+
+  // Second filter
+  builder->StartBlock(3100);
+  builder->Add("box");
+
+  // Third filter is empty
+
+  // Last filter
+  builder->StartBlock(9000);
+  builder->Add("box");
+  builder->Add("hello");
+
+  Slice block = builder->Finish();
+  FilterBlockReader* reader = new BlockBasedFilterBlockReader(
+      nullptr, table_options_, block);
+
+  // Check first filter
+  ASSERT_TRUE(reader->KeyMayMatch("foo", 0));
+  ASSERT_TRUE(reader->KeyMayMatch("bar", 2000));
+  ASSERT_TRUE(!reader->KeyMayMatch("box", 0));
+  ASSERT_TRUE(!reader->KeyMayMatch("hello", 0));
+
+  // Check second filter
+  ASSERT_TRUE(reader->KeyMayMatch("box", 3100));
+  ASSERT_TRUE(!reader->KeyMayMatch("foo", 3100));
+  ASSERT_TRUE(!reader->KeyMayMatch("bar", 3100));
+  ASSERT_TRUE(!reader->KeyMayMatch("hello", 3100));
+
+  // Check third filter (empty)
+  ASSERT_TRUE(!reader->KeyMayMatch("foo", 4100));
+  ASSERT_TRUE(!reader->KeyMayMatch("bar", 4100));
+  ASSERT_TRUE(!reader->KeyMayMatch("box", 4100));
+  ASSERT_TRUE(!reader->KeyMayMatch("hello", 4100));
+
+  // Check last filter
+  ASSERT_TRUE(reader->KeyMayMatch("box", 9000));
+  ASSERT_TRUE(reader->KeyMayMatch("hello", 9000));
+  ASSERT_TRUE(!reader->KeyMayMatch("foo", 9000));
+  ASSERT_TRUE(!reader->KeyMayMatch("bar", 9000));
+
+  delete builder;
+  delete reader;
+}
+
+}  // namespace rocksdb
+
+int main(int argc, char** argv) { return rocksdb::test::RunAllTests(); }

table/block_based_table_builder.cc

@@ -31,6 +31,8 @@
 #include "table/block_based_table_reader.h"
 #include "table/block_builder.h"
 #include "table/filter_block.h"
+#include "table/block_based_filter_block.h"
+#include "table/full_filter_block.h"
 #include "table/format.h"
 #include "table/meta_blocks.h"
 #include "table/table_builder.h"
@@ -274,6 +276,21 @@ IndexBuilder* CreateIndexBuilder(IndexType type, const Comparator* comparator,
   return nullptr;
 }
 
+// Create a index builder based on its type.
+FilterBlockBuilder* CreateFilterBlockBuilder(const ImmutableCFOptions& opt,
+    const BlockBasedTableOptions& table_opt) {
+  if (table_opt.filter_policy == nullptr) return nullptr;
+
+  FilterBitsBuilder* filter_bits_builder =
+      table_opt.filter_policy->GetFilterBitsBuilder();
+  if (filter_bits_builder == nullptr) {
+    return new BlockBasedFilterBlockBuilder(opt.prefix_extractor, table_opt);
+  } else {
+    return new FullFilterBlockBuilder(opt.prefix_extractor, table_opt,
+                                      filter_bits_builder);
+  }
+}
+
 bool GoodCompressionRatio(size_t compressed_size, size_t raw_size) {
   // Check to see if compressed less than 12.5%
   return compressed_size < raw_size - (raw_size / 8u);
@@ -365,7 +382,6 @@ class BlockBasedTableBuilder::BlockBasedTablePropertiesCollector
     std::string val;
     PutFixed32(&val, static_cast<uint32_t>(index_type_));
     properties->insert({BlockBasedTablePropertyNames::kIndexType, val});
-
     return Status::OK();
   }
 
@@ -428,11 +444,7 @@ struct BlockBasedTableBuilder::Rep {
               table_options.index_type, &internal_comparator,
               &this->internal_prefix_transform)),
         compression_type(compression_type),
-        compression_opts(compression_opts),
+        filter_block(CreateFilterBlockBuilder(ioptions, table_options)),
-        filter_block(table_options.filter_policy == nullptr ?
-            nullptr :
-            new FilterBlockBuilder(ioptions.prefix_extractor,
-                                   table_options, &internal_comparator)),
         flush_block_policy(
             table_options.flush_block_policy_factory->NewFlushBlockPolicy(
               table_options, data_block)) {
@@ -497,7 +509,7 @@ void BlockBasedTableBuilder::Add(const Slice& key, const Slice& value) {
   }
 
   if (r->filter_block != nullptr) {
-    r->filter_block->AddKey(ExtractUserKey(key));
+    r->filter_block->Add(ExtractUserKey(key));
   }
 
   r->last_key.assign(key.data(), key.size());
@@ -661,10 +673,7 @@ Status BlockBasedTableBuilder::Finish() {
   assert(!r->closed);
   r->closed = true;
 
-  BlockHandle filter_block_handle,
+  BlockHandle filter_block_handle, metaindex_block_handle, index_block_handle;
-              metaindex_block_handle,
-              index_block_handle;
-
   // Write filter block
   if (ok() && r->filter_block != nullptr) {
     auto filter_contents = r->filter_block->Finish();
@@ -703,7 +712,12 @@ Status BlockBasedTableBuilder::Finish() {
     if (r->filter_block != nullptr) {
       // Add mapping from "<filter_block_prefix>.Name" to location
       // of filter data.
-      std::string key = BlockBasedTable::kFilterBlockPrefix;
+      std::string key;
+      if (r->filter_block->IsBlockBased()) {
+        key = BlockBasedTable::kFilterBlockPrefix;
+      } else {
+        key = BlockBasedTable::kFullFilterBlockPrefix;
+      }
       key.append(r->table_options.filter_policy->Name());
       meta_index_builder.Add(key, filter_block_handle);
     }
@@ -807,5 +821,6 @@ uint64_t BlockBasedTableBuilder::FileSize() const {
 }
 
 const std::string BlockBasedTable::kFilterBlockPrefix = "filter.";
+const std::string BlockBasedTable::kFullFilterBlockPrefix = "fullfilter.";
 
 }  // namespace rocksdb

table/block_based_table_reader.cc

@@ -26,6 +26,8 @@
 
 #include "table/block.h"
 #include "table/filter_block.h"
+#include "table/block_based_filter_block.h"
+#include "table/full_filter_block.h"
 #include "table/block_hash_index.h"
 #include "table/block_prefix_index.h"
 #include "table/format.h"
@@ -46,7 +48,6 @@ using std::unique_ptr;
 typedef BlockBasedTable::IndexReader IndexReader;
 
 namespace {
-
 // The longest the prefix of the cache key used to identify blocks can be.
 // We are using the fact that we know for Posix files the unique ID is three
 // varints.
@@ -527,11 +528,18 @@ Status BlockBasedTable::Open(const ImmutableCFOptions& ioptions,
 
       // Set filter block
       if (rep->filter_policy) {
-        std::string key = kFilterBlockPrefix;
+        // First try reading full_filter, then reading block_based_filter
-        key.append(rep->filter_policy->Name());
+        for (auto filter_block_prefix : { kFullFilterBlockPrefix,
-        BlockHandle handle;
+                                          kFilterBlockPrefix }) {
-        if (FindMetaBlock(meta_iter.get(), key, &handle).ok()) {
+          std::string key = filter_block_prefix;
-          rep->filter.reset(ReadFilter(handle, rep));
+          key.append(rep->filter_policy->Name());
+
+          BlockHandle handle;
+          if (FindMetaBlock(meta_iter.get(), key, &handle).ok()) {
+            rep->filter.reset(ReadFilter(handle, rep,
+                filter_block_prefix, nullptr));
+            break;
+          }
         }
       }
     } else {
@@ -741,9 +749,9 @@ Status BlockBasedTable::PutDataBlockToCache(
   return s;
 }
 
-FilterBlockReader* BlockBasedTable::ReadFilter(const BlockHandle& filter_handle,
+FilterBlockReader* BlockBasedTable::ReadFilter(
-                                               BlockBasedTable::Rep* rep,
+    const BlockHandle& filter_handle, BlockBasedTable::Rep* rep,
-                                               size_t* filter_size) {
+    const std::string& filter_block_prefix, size_t* filter_size) {
   // TODO: We might want to unify with ReadBlockFromFile() if we start
   // requiring checksum verification in Table::Open.
   ReadOptions opt;
@@ -757,13 +765,25 @@ FilterBlockReader* BlockBasedTable::ReadFilter(const BlockHandle& filter_handle,
     *filter_size = block.data.size();
   }
 
-  return new FilterBlockReader(
+  assert(rep->filter_policy);
-       rep->ioptions.prefix_extractor, rep->table_options,
+  if (kFilterBlockPrefix == filter_block_prefix) {
-       block.data, block.heap_allocated);
+    return new BlockBasedFilterBlockReader(rep->ioptions.prefix_extractor,
+        rep->table_options, block.data, block.heap_allocated);
+  } else if (kFullFilterBlockPrefix == filter_block_prefix) {
+    auto filter_bits_reader = rep->filter_policy->
+        GetFilterBitsReader(block.data);
+
+    if (filter_bits_reader != nullptr) {
+      return new FullFilterBlockReader(rep->ioptions.prefix_extractor,
+          rep->table_options, block.data, filter_bits_reader,
+          block.heap_allocated);
+    }
+  }
+  return nullptr;
 }
 
 BlockBasedTable::CachableEntry<FilterBlockReader> BlockBasedTable::GetFilter(
-    bool no_io) const {
+                                                          bool no_io) const {
   // filter pre-populated
   if (rep_->filter != nullptr) {
     return {rep_->filter.get(), nullptr /* cache handle */};
@@ -777,11 +797,9 @@ BlockBasedTable::CachableEntry<FilterBlockReader> BlockBasedTable::GetFilter(
 
   // Fetching from the cache
   char cache_key[kMaxCacheKeyPrefixSize + kMaxVarint64Length];
-  auto key = GetCacheKey(
+  auto key = GetCacheKey(rep_->cache_key_prefix, rep_->cache_key_prefix_size,
-      rep_->cache_key_prefix,
+                         rep_->footer.metaindex_handle(),
-      rep_->cache_key_prefix_size,
+                         cache_key
-      rep_->footer.metaindex_handle(),
-      cache_key
   );
 
   Statistics* statistics = rep_->ioptions.statistics;
@@ -791,8 +809,8 @@ BlockBasedTable::CachableEntry<FilterBlockReader> BlockBasedTable::GetFilter(
 
   FilterBlockReader* filter = nullptr;
   if (cache_handle != nullptr) {
-     filter = reinterpret_cast<FilterBlockReader*>(
+    filter = reinterpret_cast<FilterBlockReader*>(
-         block_cache->Value(cache_handle));
+        block_cache->Value(cache_handle));
   } else if (no_io) {
     // Do not invoke any io.
     return CachableEntry<FilterBlockReader>();
@@ -803,17 +821,22 @@ BlockBasedTable::CachableEntry<FilterBlockReader> BlockBasedTable::GetFilter(
     auto s = ReadMetaBlock(rep_, &meta, &iter);
 
     if (s.ok()) {
-      std::string filter_block_key = kFilterBlockPrefix;
+      // First try reading full_filter, then reading block_based_filter
-      filter_block_key.append(rep_->filter_policy->Name());
+      for (auto filter_block_prefix : {kFullFilterBlockPrefix,
-      BlockHandle handle;
+                                       kFilterBlockPrefix}) {
-      if (FindMetaBlock(iter.get(), filter_block_key, &handle).ok()) {
+        std::string filter_block_key = filter_block_prefix;
-        filter = ReadFilter(handle, rep_, &filter_size);
+        filter_block_key.append(rep_->filter_policy->Name());
-        assert(filter);
+        BlockHandle handle;
-        assert(filter_size > 0);
+        if (FindMetaBlock(iter.get(), filter_block_key, &handle).ok()) {
+          filter = ReadFilter(handle, rep_, filter_block_prefix, &filter_size);
 
-        cache_handle = block_cache->Insert(
+          if (filter == nullptr) break;  // err happen in ReadFilter
-            key, filter, filter_size, &DeleteCachedEntry<FilterBlockReader>);
+          assert(filter_size > 0);
-        RecordTick(statistics, BLOCK_CACHE_ADD);
+          cache_handle = block_cache->Insert(
+              key, filter, filter_size, &DeleteCachedEntry<FilterBlockReader>);
+          RecordTick(statistics, BLOCK_CACHE_ADD);
+          break;
+        }
       }
     }
   }
@@ -918,8 +941,8 @@ Iterator* BlockBasedTable::NewDataBlockIterator(Rep* rep,
 
     // create key for block cache
     if (block_cache != nullptr) {
-      key = GetCacheKey(rep->cache_key_prefix,
+      key = GetCacheKey(rep->cache_key_prefix, rep->cache_key_prefix_size,
-                        rep->cache_key_prefix_size, handle, cache_key);
+                        handle, cache_key);
     }
 
     if (block_cache_compressed != nullptr) {
@@ -1039,42 +1062,50 @@ bool BlockBasedTable::PrefixMayMatch(const Slice& internal_key) {
   // loaded to memory.
   ReadOptions no_io_read_options;
   no_io_read_options.read_tier = kBlockCacheTier;
-  unique_ptr<Iterator> iiter(NewIndexIterator(no_io_read_options));
-  iiter->Seek(internal_prefix);
 
-  if (!iiter->Valid()) {
+  // First, try check with full filter
-    // we're past end of file
+  auto filter_entry = GetFilter(true /* no io */);
-    // if it's incomplete, it means that we avoided I/O
+  FilterBlockReader* filter = filter_entry.value;
-    // and we're not really sure that we're past the end
+  if (filter != nullptr && !filter->IsBlockBased()) {
-    // of the file
+    may_match = filter->PrefixMayMatch(prefix);
-    may_match = iiter->status().IsIncomplete();
+  }
-  } else if (ExtractUserKey(iiter->key()).starts_with(
+
-              ExtractUserKey(internal_prefix))) {
+  // Then, try find it within each block
-    // we need to check for this subtle case because our only
+  if (may_match) {
-    // guarantee is that "the key is a string >= last key in that data
+    unique_ptr<Iterator> iiter(NewIndexIterator(no_io_read_options));
-    // block" according to the doc/table_format.txt spec.
+    iiter->Seek(internal_prefix);
-    //
+
-    // Suppose iiter->key() starts with the desired prefix; it is not
+    if (!iiter->Valid()) {
-    // necessarily the case that the corresponding data block will
+      // we're past end of file
-    // contain the prefix, since iiter->key() need not be in the
+      // if it's incomplete, it means that we avoided I/O
-    // block.  However, the next data block may contain the prefix, so
+      // and we're not really sure that we're past the end
-    // we return true to play it safe.
+      // of the file
-    may_match = true;
+      may_match = iiter->status().IsIncomplete();
-  } else {
+    } else if (ExtractUserKey(iiter->key()).starts_with(
-    // iiter->key() does NOT start with the desired prefix.  Because
+                ExtractUserKey(internal_prefix))) {
-    // Seek() finds the first key that is >= the seek target, this
+      // we need to check for this subtle case because our only
-    // means that iiter->key() > prefix.  Thus, any data blocks coming
+      // guarantee is that "the key is a string >= last key in that data
-    // after the data block corresponding to iiter->key() cannot
+      // block" according to the doc/table_format.txt spec.
-    // possibly contain the key.  Thus, the corresponding data block
+      //
-    // is the only one which could potentially contain the prefix.
+      // Suppose iiter->key() starts with the desired prefix; it is not
-    Slice handle_value = iiter->value();
+      // necessarily the case that the corresponding data block will
-    BlockHandle handle;
+      // contain the prefix, since iiter->key() need not be in the
-    s = handle.DecodeFrom(&handle_value);
+      // block.  However, the next data block may contain the prefix, so
-    assert(s.ok());
+      // we return true to play it safe.
-    auto filter_entry = GetFilter(true /* no io */);
+      may_match = true;
-    may_match = filter_entry.value == nullptr ||
+    } else if (filter != nullptr && filter->IsBlockBased()) {
-                filter_entry.value->PrefixMayMatch(handle.offset(), prefix);
+      // iiter->key() does NOT start with the desired prefix.  Because
-    filter_entry.Release(rep_->table_options.block_cache.get());
+      // Seek() finds the first key that is >= the seek target, this
+      // means that iiter->key() > prefix.  Thus, any data blocks coming
+      // after the data block corresponding to iiter->key() cannot
+      // possibly contain the key.  Thus, the corresponding data block
+      // is the only on could potentially contain the prefix.
+      Slice handle_value = iiter->value();
+      BlockHandle handle;
+      s = handle.DecodeFrom(&handle_value);
+      assert(s.ok());
+      may_match = filter->PrefixMayMatch(prefix, handle.offset());
+    }
   }
 
   Statistics* statistics = rep_->ioptions.statistics;
@@ -1083,6 +1114,7 @@ bool BlockBasedTable::PrefixMayMatch(const Slice& internal_key) {
     RecordTick(statistics, BLOOM_FILTER_PREFIX_USEFUL);
   }
 
+  filter_entry.Release(rep_->table_options.block_cache.get());
   return may_match;
 }
 
@@ -1098,64 +1130,72 @@ Status BlockBasedTable::Get(
                            const Slice& v),
     void (*mark_key_may_exist_handler)(void* handle_context)) {
   Status s;
-  BlockIter iiter;
-  NewIndexIterator(read_options, &iiter);
-
   auto filter_entry = GetFilter(read_options.read_tier == kBlockCacheTier);
   FilterBlockReader* filter = filter_entry.value;
-  bool done = false;
-  for (iiter.Seek(key); iiter.Valid() && !done; iiter.Next()) {
-    Slice handle_value = iiter.value();
 
-    BlockHandle handle;
+  // First check the full filter
-    bool may_not_exist_in_filter =
+  // If full filter not useful, Then go into each block
-        filter != nullptr && handle.DecodeFrom(&handle_value).ok() &&
+  if (filter != nullptr && !filter->IsBlockBased()
-        !filter->KeyMayMatch(handle.offset(), ExtractUserKey(key));
+                        && !filter->KeyMayMatch(ExtractUserKey(key))) {
+    RecordTick(rep_->ioptions.statistics, BLOOM_FILTER_USEFUL);
+  } else {
+    BlockIter iiter;
+    NewIndexIterator(read_options, &iiter);
 
-    if (may_not_exist_in_filter) {
+    bool done = false;
-      // Not found
+    for (iiter.Seek(key); iiter.Valid() && !done; iiter.Next()) {
-      // TODO: think about interaction with Merge. If a user key cannot
+      Slice handle_value = iiter.value();
-      // cross one data block, we should be fine.
-      RecordTick(rep_->ioptions.statistics, BLOOM_FILTER_USEFUL);
-      break;
-    } else {
-      BlockIter biter;
-      NewDataBlockIterator(rep_, read_options, iiter.value(), &biter);
 
-      if (read_options.read_tier && biter.status().IsIncomplete()) {
+      BlockHandle handle;
-        // couldn't get block from block_cache
+      bool not_exist_in_filter =
-        // Update Saver.state to Found because we are only looking for whether
+          filter != nullptr && filter->IsBlockBased() == true &&
-        // we can guarantee the key is not there when "no_io" is set
+          handle.DecodeFrom(&handle_value).ok() &&
-        (*mark_key_may_exist_handler)(handle_context);
+          !filter->KeyMayMatch(ExtractUserKey(key), handle.offset());
+
+      if (not_exist_in_filter) {
+        // Not found
+        // TODO: think about interaction with Merge. If a user key cannot
+        // cross one data block, we should be fine.
+        RecordTick(rep_->ioptions.statistics, BLOOM_FILTER_USEFUL);
         break;
-      }
+      } else {
-      if (!biter.status().ok()) {
+        BlockIter biter;
-        s = biter.status();
+        NewDataBlockIterator(rep_, read_options, iiter.value(), &biter);
-        break;
-      }
 
-      // Call the *saver function on each entry/block until it returns false
+        if (read_options.read_tier && biter.status().IsIncomplete()) {
-      for (biter.Seek(key); biter.Valid(); biter.Next()) {
+          // couldn't get block from block_cache
-        ParsedInternalKey parsed_key;
+          // Update Saver.state to Found because we are only looking for whether
-        if (!ParseInternalKey(biter.key(), &parsed_key)) {
+          // we can guarantee the key is not there when "no_io" is set
-          s = Status::Corruption(Slice());
+          (*mark_key_may_exist_handler)(handle_context);
-        }
-
-        if (!(*result_handler)(handle_context, parsed_key,
-                               biter.value())) {
-          done = true;
           break;
         }
+        if (!biter.status().ok()) {
+          s = biter.status();
+          break;
+        }
+
+        // Call the *saver function on each entry/block until it returns false
+        for (biter.Seek(key); biter.Valid(); biter.Next()) {
+          ParsedInternalKey parsed_key;
+          if (!ParseInternalKey(biter.key(), &parsed_key)) {
+            s = Status::Corruption(Slice());
+          }
+
+          if (!(*result_handler)(handle_context, parsed_key,
+                                 biter.value())) {
+            done = true;
+            break;
+          }
+        }
+        s = biter.status();
       }
-      s = biter.status();
+    }
+    if (s.ok()) {
+      s = iiter.status();
     }
   }
 
   filter_entry.Release(rep_->table_options.block_cache.get());
-  if (s.ok()) {
-    s = iiter.status();
-  }
-
   return s;
 }
 
@@ -1175,8 +1215,8 @@ bool BlockBasedTable::TEST_KeyInCache(const ReadOptions& options,
 
   char cache_key_storage[kMaxCacheKeyPrefixSize + kMaxVarint64Length];
   Slice cache_key =
-      GetCacheKey(rep_->cache_key_prefix, rep_->cache_key_prefix_size, handle,
+      GetCacheKey(rep_->cache_key_prefix, rep_->cache_key_prefix_size,
-                  cache_key_storage);
+                  handle, cache_key_storage);
   Slice ckey;
 
   s = GetDataBlockFromCache(cache_key, ckey, block_cache, nullptr, nullptr,

table/block_based_table_reader.h

@@ -28,6 +28,8 @@ class BlockIter;
 class BlockHandle;
 class Cache;
 class FilterBlockReader;
+class BlockBasedFilterBlockReader;
+class FullFilterBlockReader;
 class Footer;
 class InternalKeyComparator;
 class Iterator;
@@ -47,6 +49,7 @@ using std::unique_ptr;
 class BlockBasedTable : public TableReader {
  public:
   static const std::string kFilterBlockPrefix;
+  static const std::string kFullFilterBlockPrefix;
 
   // Attempt to open the table that is stored in bytes [0..file_size)
   // of "file", and read the metadata entries necessary to allow
@@ -184,7 +187,9 @@ class BlockBasedTable : public TableReader {
 
   // Create the filter from the filter block.
   static FilterBlockReader* ReadFilter(const BlockHandle& filter_handle,
-                                       Rep* rep, size_t* filter_size = nullptr);
+                                       Rep* rep,
+                                       const std::string& filter_block_prefix,
+                                       size_t* filter_size = nullptr);
 
   static void SetupCacheKeyPrefix(Rep* rep);
 

table/filter_block.h

@@ -10,86 +10,70 @@
 // A filter block is stored near the end of a Table file.  It contains
 // filters (e.g., bloom filters) for all data blocks in the table combined
 // into a single filter block.
+//
+// It is a base class for BlockBasedFilter and FullFilter.
+// These two are both used in BlockBasedTable. The first one contain filter
+// For a part of keys in sst file, the second contain filter for all keys
+// in sst file.
 
 #pragma once
 
-#include <memory>
 #include <stddef.h>
 #include <stdint.h>
 #include <string>
 #include <vector>
+#include "rocksdb/options.h"
 #include "rocksdb/slice.h"
-#include "rocksdb/slice_transform.h"
 #include "rocksdb/table.h"
-#include "util/hash.h"
 
 namespace rocksdb {
 
-class FilterPolicy;
+const uint64_t kNotValid = ULLONG_MAX;
 
 // A FilterBlockBuilder is used to construct all of the filters for a
 // particular Table.  It generates a single string which is stored as
 // a special block in the Table.
 //
 // The sequence of calls to FilterBlockBuilder must match the regexp:
-//      (StartBlock AddKey*)* Finish
+//      (StartBlock Add*)* Finish
+//
+// BlockBased/Full FilterBlock would be called in the same way.
 class FilterBlockBuilder {
  public:
-  explicit FilterBlockBuilder(const SliceTransform* prefix_extractor,
+  explicit FilterBlockBuilder() {}
-                              const BlockBasedTableOptions& table_opt,
+  virtual ~FilterBlockBuilder() {}
-                              const Comparator* internal_comparator);
 
-  void StartBlock(uint64_t block_offset);
+  virtual bool IsBlockBased() = 0;                    // If is blockbased filter
-  void AddKey(const Slice& key);
+  virtual void StartBlock(uint64_t block_offset) = 0;  // Start new block filter
-  Slice Finish();
+  virtual void Add(const Slice& key) = 0;      // Add a key to current filter
+  virtual Slice Finish() = 0;                     // Generate Filter
 
  private:
-  bool SamePrefix(const Slice &key1, const Slice &key2) const;
-  void GenerateFilter();
-
-  // important: all of these might point to invalid addresses
-  // at the time of destruction of this filter block. destructor
-  // should NOT dereference them.
-  const FilterPolicy* policy_;
-  const SliceTransform* prefix_extractor_;
-  bool whole_key_filtering_;
-  const Comparator* comparator_;
-
-  std::string entries_;         // Flattened entry contents
-  std::vector<size_t> start_;   // Starting index in entries_ of each entry
-  std::string result_;          // Filter data computed so far
-  std::vector<Slice> tmp_entries_; // policy_->CreateFilter() argument
-  std::vector<uint32_t> filter_offsets_;
-
   // No copying allowed
   FilterBlockBuilder(const FilterBlockBuilder&);
   void operator=(const FilterBlockBuilder&);
 };
 
+// A FilterBlockReader is used to parse filter from SST table.
+// KeyMayMatch and PrefixMayMatch would trigger filter checking
+//
+// BlockBased/Full FilterBlock would be called in the same way.
 class FilterBlockReader {
  public:
- // REQUIRES: "contents" and *policy must stay live while *this is live.
+  explicit FilterBlockReader() {}
-  FilterBlockReader(
+  virtual ~FilterBlockReader() {}
-    const SliceTransform* prefix_extractor,
+
-    const BlockBasedTableOptions& table_opt,
+  virtual bool IsBlockBased() = 0;  // If is blockbased filter
-    const Slice& contents,
+  virtual bool KeyMayMatch(const Slice& key,
-    bool delete_contents_after_use = false);
+                           uint64_t block_offset = kNotValid) = 0;
-  bool KeyMayMatch(uint64_t block_offset, const Slice& key);
+  virtual bool PrefixMayMatch(const Slice& prefix,
-  bool PrefixMayMatch(uint64_t block_offset, const Slice& prefix);
+                              uint64_t block_offset = kNotValid) = 0;
-  size_t ApproximateMemoryUsage() const;
+  virtual size_t ApproximateMemoryUsage() const = 0;
 
  private:
-  const FilterPolicy* policy_;
+  // No copying allowed
-  const SliceTransform* prefix_extractor_;
+  FilterBlockReader(const FilterBlockReader&);
-  bool whole_key_filtering_;
+  void operator=(const FilterBlockReader&);
-  const char* data_;    // Pointer to filter data (at block-start)
-  const char* offset_;  // Pointer to beginning of offset array (at block-end)
-  size_t num_;          // Number of entries in offset array
-  size_t base_lg_;      // Encoding parameter (see kFilterBaseLg in .cc file)
-  std::unique_ptr<const char[]> filter_data;
-
-
-  bool MayMatch(uint64_t block_offset, const Slice& entry);
 };
 
-}
+}  // namespace rocksdb

table/filter_block_test.cc

@@ -1,139 +0,0 @@
-//  Copyright (c) 2013, Facebook, Inc.  All rights reserved.
-//  This source code is licensed under the BSD-style license found in the
-//  LICENSE file in the root directory of this source tree. An additional grant
-//  of patent rights can be found in the PATENTS file in the same directory.
-//
-// Copyright (c) 2012 The LevelDB Authors. All rights reserved.
-// Use of this source code is governed by a BSD-style license that can be
-// found in the LICENSE file. See the AUTHORS file for names of contributors.
-
-#include "table/filter_block.h"
-
-#include "rocksdb/filter_policy.h"
-#include "util/coding.h"
-#include "util/hash.h"
-#include "util/logging.h"
-#include "util/testharness.h"
-#include "util/testutil.h"
-
-namespace rocksdb {
-
-// For testing: emit an array with one hash value per key
-class TestHashFilter : public FilterPolicy {
- public:
-  virtual const char* Name() const {
-    return "TestHashFilter";
-  }
-
-  virtual void CreateFilter(const Slice* keys, int n, std::string* dst) const {
-    for (int i = 0; i < n; i++) {
-      uint32_t h = Hash(keys[i].data(), keys[i].size(), 1);
-      PutFixed32(dst, h);
-    }
-  }
-
-  virtual bool KeyMayMatch(const Slice& key, const Slice& filter) const {
-    uint32_t h = Hash(key.data(), key.size(), 1);
-    for (unsigned int i = 0; i + 4 <= filter.size(); i += 4) {
-      if (h == DecodeFixed32(filter.data() + i)) {
-        return true;
-      }
-    }
-    return false;
-  }
-};
-
-class FilterBlockTest {
- public:
-  const Comparator* comparator_;
-  BlockBasedTableOptions table_options_;
-
-  FilterBlockTest()
-    : comparator_(BytewiseComparator()) {
-    table_options_.filter_policy.reset(new TestHashFilter());
-  }
-};
-
-TEST(FilterBlockTest, EmptyBuilder) {
-  FilterBlockBuilder builder(nullptr, table_options_, comparator_);
-  Slice block = builder.Finish();
-  ASSERT_EQ("\\x00\\x00\\x00\\x00\\x0b", EscapeString(block));
-  FilterBlockReader reader(nullptr, table_options_, block);
-  ASSERT_TRUE(reader.KeyMayMatch(0, "foo"));
-  ASSERT_TRUE(reader.KeyMayMatch(100000, "foo"));
-}
-
-TEST(FilterBlockTest, SingleChunk) {
-  FilterBlockBuilder builder(nullptr, table_options_, comparator_);
-  builder.StartBlock(100);
-  builder.AddKey("foo");
-  builder.AddKey("bar");
-  builder.AddKey("box");
-  builder.StartBlock(200);
-  builder.AddKey("box");
-  builder.StartBlock(300);
-  builder.AddKey("hello");
-  Slice block = builder.Finish();
-  FilterBlockReader reader(nullptr, table_options_, block);
-  ASSERT_TRUE(reader.KeyMayMatch(100, "foo"));
-  ASSERT_TRUE(reader.KeyMayMatch(100, "bar"));
-  ASSERT_TRUE(reader.KeyMayMatch(100, "box"));
-  ASSERT_TRUE(reader.KeyMayMatch(100, "hello"));
-  ASSERT_TRUE(reader.KeyMayMatch(100, "foo"));
-  ASSERT_TRUE(! reader.KeyMayMatch(100, "missing"));
-  ASSERT_TRUE(! reader.KeyMayMatch(100, "other"));
-}
-
-TEST(FilterBlockTest, MultiChunk) {
-  FilterBlockBuilder builder(nullptr, table_options_, comparator_);
-
-  // First filter
-  builder.StartBlock(0);
-  builder.AddKey("foo");
-  builder.StartBlock(2000);
-  builder.AddKey("bar");
-
-  // Second filter
-  builder.StartBlock(3100);
-  builder.AddKey("box");
-
-  // Third filter is empty
-
-  // Last filter
-  builder.StartBlock(9000);
-  builder.AddKey("box");
-  builder.AddKey("hello");
-
-  Slice block = builder.Finish();
-  FilterBlockReader reader(nullptr, table_options_, block);
-
-  // Check first filter
-  ASSERT_TRUE(reader.KeyMayMatch(0, "foo"));
-  ASSERT_TRUE(reader.KeyMayMatch(2000, "bar"));
-  ASSERT_TRUE(! reader.KeyMayMatch(0, "box"));
-  ASSERT_TRUE(! reader.KeyMayMatch(0, "hello"));
-
-  // Check second filter
-  ASSERT_TRUE(reader.KeyMayMatch(3100, "box"));
-  ASSERT_TRUE(! reader.KeyMayMatch(3100, "foo"));
-  ASSERT_TRUE(! reader.KeyMayMatch(3100, "bar"));
-  ASSERT_TRUE(! reader.KeyMayMatch(3100, "hello"));
-
-  // Check third filter (empty)
-  ASSERT_TRUE(! reader.KeyMayMatch(4100, "foo"));
-  ASSERT_TRUE(! reader.KeyMayMatch(4100, "bar"));
-  ASSERT_TRUE(! reader.KeyMayMatch(4100, "box"));
-  ASSERT_TRUE(! reader.KeyMayMatch(4100, "hello"));
-
-  // Check last filter
-  ASSERT_TRUE(reader.KeyMayMatch(9000, "box"));
-  ASSERT_TRUE(reader.KeyMayMatch(9000, "hello"));
-  ASSERT_TRUE(! reader.KeyMayMatch(9000, "foo"));
-  ASSERT_TRUE(! reader.KeyMayMatch(9000, "bar"));
-}
-
-}  // namespace rocksdb
-
-int main(int argc, char** argv) {
-  return rocksdb::test::RunAllTests();
-}

table/full_filter_block.cc

@@ -0,0 +1,99 @@
+//  Copyright (c) 2014, Facebook, Inc.  All rights reserved.
+//  This source code is licensed under the BSD-style license found in the
+//  LICENSE file in the root directory of this source tree. An additional grant
+//  of patent rights can be found in the PATENTS file in the same directory.
+
+#include "table/full_filter_block.h"
+
+#include "rocksdb/filter_policy.h"
+#include "port/port.h"
+#include "util/coding.h"
+
+namespace rocksdb {
+
+FullFilterBlockBuilder::FullFilterBlockBuilder(
+    const SliceTransform* prefix_extractor,
+    const BlockBasedTableOptions& table_opt,
+    FilterBitsBuilder* filter_bits_builder)
+    : prefix_extractor_(prefix_extractor),
+      whole_key_filtering_(table_opt.whole_key_filtering),
+      num_added_(0) {
+  assert(filter_bits_builder != nullptr);
+  filter_bits_builder_.reset(filter_bits_builder);
+}
+
+void FullFilterBlockBuilder::Add(const Slice& key) {
+  if (whole_key_filtering_) {
+    AddKey(key);
+  }
+  if (prefix_extractor_ && prefix_extractor_->InDomain(key)) {
+    AddPrefix(key);
+  }
+}
+
+// Add key to filter if needed
+inline void FullFilterBlockBuilder::AddKey(const Slice& key) {
+  filter_bits_builder_->AddKey(key);
+  num_added_++;
+}
+
+// Add prefix to filter if needed
+inline void FullFilterBlockBuilder::AddPrefix(const Slice& key) {
+  Slice prefix = prefix_extractor_->Transform(key);
+  filter_bits_builder_->AddKey(prefix);
+  num_added_++;
+}
+
+Slice FullFilterBlockBuilder::Finish() {
+  if (num_added_ != 0) {
+    num_added_ = 0;
+    return filter_bits_builder_->Finish(&filter_data);
+  }
+  return Slice();
+}
+
+FullFilterBlockReader::FullFilterBlockReader(
+    const SliceTransform* prefix_extractor,
+    const BlockBasedTableOptions& table_opt,
+    const Slice& contents,
+    FilterBitsReader* filter_bits_reader, bool delete_contents_after_use)
+    : prefix_extractor_(prefix_extractor),
+      whole_key_filtering_(table_opt.whole_key_filtering),
+      contents_(contents) {
+  assert(filter_bits_reader != nullptr);
+  filter_bits_reader_.reset(filter_bits_reader);
+
+  if (delete_contents_after_use) {
+    filter_data.reset(contents.data());
+  }
+}
+
+bool FullFilterBlockReader::KeyMayMatch(const Slice& key,
+    uint64_t block_offset) {
+  assert(block_offset == kNotValid);
+  if (!whole_key_filtering_) {
+    return true;
+  }
+  return MayMatch(key);
+}
+
+bool FullFilterBlockReader::PrefixMayMatch(const Slice& prefix,
+                                           uint64_t block_offset) {
+  assert(block_offset == kNotValid);
+  if (!prefix_extractor_) {
+    return true;
+  }
+  return MayMatch(prefix);
+}
+
+bool FullFilterBlockReader::MayMatch(const Slice& entry) {
+  if (contents_.size() != 0)  {
+    return filter_bits_reader_->MayMatch(entry);
+  }
+  return true;  // remain the same with block_based filter
+}
+
+size_t FullFilterBlockReader::ApproximateMemoryUsage() const {
+  return contents_.size();
+}
+}  // namespace rocksdb

table/full_filter_block.h

@@ -0,0 +1,107 @@
+//  Copyright (c) 2014, Facebook, Inc.  All rights reserved.
+//  This source code is licensed under the BSD-style license found in the
+//  LICENSE file in the root directory of this source tree. An additional grant
+//  of patent rights can be found in the PATENTS file in the same directory.
+
+#pragma once
+
+#include <stddef.h>
+#include <stdint.h>
+#include <memory>
+#include <string>
+#include <vector>
+#include "rocksdb/options.h"
+#include "rocksdb/slice.h"
+#include "rocksdb/slice_transform.h"
+#include "db/dbformat.h"
+#include "util/hash.h"
+#include "table/filter_block.h"
+
+namespace rocksdb {
+
+class FilterPolicy;
+class FilterBitsBuilder;
+class FilterBitsReader;
+
+// A FullFilterBlockBuilder is used to construct a full filter for a
+// particular Table.  It generates a single string which is stored as
+// a special block in the Table.
+// The format of full filter block is:
+// +----------------------------------------------------------------+
+// |              full filter for all keys in sst file              |
+// +----------------------------------------------------------------+
+// The full filter can be very large. At the end of it, we put
+// num_probes: how many hash functions are used in bloom filter
+//
+class FullFilterBlockBuilder : public FilterBlockBuilder {
+ public:
+  explicit FullFilterBlockBuilder(const SliceTransform* prefix_extractor,
+                                  const BlockBasedTableOptions& table_opt,
+                                  FilterBitsBuilder* filter_bits_builder);
+  // bits_builder is created in filter_policy, it should be passed in here
+  // directly. and be deleted here
+  ~FullFilterBlockBuilder() {}
+
+  virtual bool IsBlockBased() override { return false; }
+  virtual void StartBlock(uint64_t block_offset) override {}
+  virtual void Add(const Slice& key) override;
+  virtual Slice Finish() override;
+
+ private:
+  // important: all of these might point to invalid addresses
+  // at the time of destruction of this filter block. destructor
+  // should NOT dereference them.
+  const SliceTransform* prefix_extractor_;
+  bool whole_key_filtering_;
+
+  uint32_t num_added_;
+  std::unique_ptr<FilterBitsBuilder> filter_bits_builder_;
+  std::unique_ptr<const char[]> filter_data;
+
+  void AddKey(const Slice& key);
+  void AddPrefix(const Slice& key);
+
+  // No copying allowed
+  FullFilterBlockBuilder(const FullFilterBlockBuilder&);
+  void operator=(const FullFilterBlockBuilder&);
+};
+
+// A FilterBlockReader is used to parse filter from SST table.
+// KeyMayMatch and PrefixMayMatch would trigger filter checking
+class FullFilterBlockReader : public FilterBlockReader {
+ public:
+  // REQUIRES: "contents" and filter_bits_reader must stay live
+  // while *this is live.
+  explicit FullFilterBlockReader(const SliceTransform* prefix_extractor,
+                                 const BlockBasedTableOptions& table_opt,
+                                 const Slice& contents,
+                                 FilterBitsReader* filter_bits_reader,
+                                 bool delete_contents_after_use = false);
+
+  // bits_reader is created in filter_policy, it should be passed in here
+  // directly. and be deleted here
+  ~FullFilterBlockReader() {}
+
+  virtual bool IsBlockBased() override { return false; }
+  virtual bool KeyMayMatch(const Slice& key,
+                           uint64_t block_offset = kNotValid) override;
+  virtual bool PrefixMayMatch(const Slice& prefix,
+                              uint64_t block_offset = kNotValid) override;
+  virtual size_t ApproximateMemoryUsage() const override;
+
+ private:
+  const SliceTransform* prefix_extractor_;
+  bool whole_key_filtering_;
+
+  std::unique_ptr<FilterBitsReader> filter_bits_reader_;
+  Slice contents_;
+  std::unique_ptr<const char[]> filter_data;
+
+  bool MayMatch(const Slice& entry);
+
+  // No copying allowed
+  FullFilterBlockReader(const FullFilterBlockReader&);
+  void operator=(const FullFilterBlockReader&);
+};
+
+}  // namespace rocksdb

table/full_filter_block_test.cc

@@ -0,0 +1,181 @@
+//  Copyright (c) 2014, Facebook, Inc.  All rights reserved.
+//  This source code is licensed under the BSD-style license found in the
+//  LICENSE file in the root directory of this source tree. An additional grant
+//  of patent rights can be found in the PATENTS file in the same directory.
+
+#include "table/full_filter_block.h"
+
+#include "rocksdb/filter_policy.h"
+#include "util/coding.h"
+#include "util/hash.h"
+#include "util/logging.h"
+#include "util/testharness.h"
+#include "util/testutil.h"
+
+namespace rocksdb {
+
+class TestFilterBitsBuilder : public FilterBitsBuilder {
+ public:
+  explicit TestFilterBitsBuilder() {}
+
+  // Add Key to filter
+  virtual void AddKey(const Slice& key) override {
+    hash_entries_.push_back(Hash(key.data(), key.size(), 1));
+  }
+
+  // Generate the filter using the keys that are added
+  virtual Slice Finish(std::unique_ptr<const char[]>* buf) override {
+    uint32_t len = hash_entries_.size() * 4;
+    char* data = new char[len];
+    for (size_t i = 0; i < hash_entries_.size(); i++) {
+      EncodeFixed32(data + i * 4, hash_entries_[i]);
+    }
+    buf->reset(data);
+    return Slice(data, len);
+  }
+
+ private:
+  std::vector<uint32_t> hash_entries_;
+};
+
+class TestFilterBitsReader : public FilterBitsReader {
+ public:
+  explicit TestFilterBitsReader(const Slice& contents)
+    : data_(contents.data()), len_(contents.size()) {}
+
+  virtual bool MayMatch(const Slice& entry) override {
+    uint32_t h = Hash(entry.data(), entry.size(), 1);
+    for (size_t i = 0; i + 4 <= len_; i += 4) {
+      if (h == DecodeFixed32(data_ + i)) {
+        return true;
+      }
+    }
+    return false;
+  }
+
+ private:
+  const char* data_;
+  uint32_t len_;
+};
+
+
+class TestHashFilter : public FilterPolicy {
+ public:
+  virtual const char* Name() const {
+    return "TestHashFilter";
+  }
+
+  virtual void CreateFilter(const Slice* keys, int n, std::string* dst) const {
+    for (int i = 0; i < n; i++) {
+      uint32_t h = Hash(keys[i].data(), keys[i].size(), 1);
+      PutFixed32(dst, h);
+    }
+  }
+
+  virtual bool KeyMayMatch(const Slice& key, const Slice& filter) const {
+    uint32_t h = Hash(key.data(), key.size(), 1);
+    for (unsigned int i = 0; i + 4 <= filter.size(); i += 4) {
+      if (h == DecodeFixed32(filter.data() + i)) {
+        return true;
+      }
+    }
+    return false;
+  }
+
+  virtual FilterBitsBuilder* GetFilterBitsBuilder() const override {
+    return new TestFilterBitsBuilder();
+  }
+
+  virtual FilterBitsReader* GetFilterBitsReader(const Slice& contents)
+      const override {
+    return new TestFilterBitsReader(contents);
+  }
+};
+
+class PluginFullFilterBlockTest {
+ public:
+  BlockBasedTableOptions table_options_;
+
+  PluginFullFilterBlockTest() {
+    table_options_.filter_policy.reset(new TestHashFilter());
+  }
+};
+
+TEST(PluginFullFilterBlockTest, PluginEmptyBuilder) {
+  FullFilterBlockBuilder builder(nullptr, table_options_,
+      table_options_.filter_policy->GetFilterBitsBuilder());
+  Slice block = builder.Finish();
+  ASSERT_EQ("", EscapeString(block));
+
+  FullFilterBlockReader reader(nullptr, table_options_, block,
+      table_options_.filter_policy->GetFilterBitsReader(block));
+  // Remain same symantic with blockbased filter
+  ASSERT_TRUE(reader.KeyMayMatch("foo"));
+}
+
+TEST(PluginFullFilterBlockTest, PluginSingleChunk) {
+  FullFilterBlockBuilder builder(nullptr, table_options_,
+      table_options_.filter_policy->GetFilterBitsBuilder());
+  builder.Add("foo");
+  builder.Add("bar");
+  builder.Add("box");
+  builder.Add("box");
+  builder.Add("hello");
+  Slice block = builder.Finish();
+  FullFilterBlockReader reader(nullptr, table_options_, block,
+      table_options_.filter_policy->GetFilterBitsReader(block));
+  ASSERT_TRUE(reader.KeyMayMatch("foo"));
+  ASSERT_TRUE(reader.KeyMayMatch("bar"));
+  ASSERT_TRUE(reader.KeyMayMatch("box"));
+  ASSERT_TRUE(reader.KeyMayMatch("hello"));
+  ASSERT_TRUE(reader.KeyMayMatch("foo"));
+  ASSERT_TRUE(!reader.KeyMayMatch("missing"));
+  ASSERT_TRUE(!reader.KeyMayMatch("other"));
+}
+
+class FullFilterBlockTest {
+ public:
+  BlockBasedTableOptions table_options_;
+
+  FullFilterBlockTest() {
+    table_options_.filter_policy.reset(NewBloomFilterPolicy(10, false));
+  }
+
+  ~FullFilterBlockTest() {}
+};
+
+TEST(FullFilterBlockTest, EmptyBuilder) {
+  FullFilterBlockBuilder builder(nullptr, table_options_,
+      table_options_.filter_policy->GetFilterBitsBuilder());
+  Slice block = builder.Finish();
+  ASSERT_EQ("", EscapeString(block));
+
+  FullFilterBlockReader reader(nullptr, table_options_, block,
+      table_options_.filter_policy->GetFilterBitsReader(block));
+  // Remain same symantic with blockbased filter
+  ASSERT_TRUE(reader.KeyMayMatch("foo"));
+}
+
+TEST(FullFilterBlockTest, SingleChunk) {
+  FullFilterBlockBuilder builder(nullptr, table_options_,
+      table_options_.filter_policy->GetFilterBitsBuilder());
+  builder.Add("foo");
+  builder.Add("bar");
+  builder.Add("box");
+  builder.Add("box");
+  builder.Add("hello");
+  Slice block = builder.Finish();
+  FullFilterBlockReader reader(nullptr, table_options_, block,
+      table_options_.filter_policy->GetFilterBitsReader(block));
+  ASSERT_TRUE(reader.KeyMayMatch("foo"));
+  ASSERT_TRUE(reader.KeyMayMatch("bar"));
+  ASSERT_TRUE(reader.KeyMayMatch("box"));
+  ASSERT_TRUE(reader.KeyMayMatch("hello"));
+  ASSERT_TRUE(reader.KeyMayMatch("foo"));
+  ASSERT_TRUE(!reader.KeyMayMatch("missing"));
+  ASSERT_TRUE(!reader.KeyMayMatch("other"));
+}
+
+}  // namespace rocksdb
+
+int main(int argc, char** argv) { return rocksdb::test::RunAllTests(); }

table/plain_table_builder.cc

@@ -20,7 +20,6 @@
 #include "table/block_builder.h"
 #include "table/bloom_block.h"
 #include "table/plain_table_index.h"
-#include "table/filter_block.h"
 #include "table/format.h"
 #include "table/meta_blocks.h"
 #include "util/coding.h"

table/plain_table_reader.cc

@@ -20,7 +20,6 @@
 
 #include "table/block.h"
 #include "table/bloom_block.h"
-#include "table/filter_block.h"
 #include "table/format.h"
 #include "table/meta_blocks.h"
 #include "table/two_level_iterator.h"

tools/db_sanity_test.cc

@@ -16,6 +16,7 @@
 #include "include/rocksdb/comparator.h"
 #include "include/rocksdb/table.h"
 #include "include/rocksdb/slice_transform.h"
+#include "include/rocksdb/filter_policy.h"
 
 namespace rocksdb {
 
@@ -146,13 +147,30 @@ class SanityTestPlainTableFactory : public SanityTest {
   Options options_;
 };
 
+class SanityTestBloomFilter : public SanityTest {
+ public:
+  explicit SanityTestBloomFilter(const std::string& path)
+      : SanityTest(path) {
+    table_options_.filter_policy.reset(NewBloomFilterPolicy(10));
+    options_.table_factory.reset(NewBlockBasedTableFactory(table_options_));
+  }
+  ~SanityTestBloomFilter() {}
+  virtual Options GetOptions() const { return options_; }
+  virtual std::string Name() const { return "BloomFilter"; }
+
+ private:
+  Options options_;
+  BlockBasedTableOptions table_options_;
+};
+
 namespace {
 bool RunSanityTests(const std::string& command, const std::string& path) {
   std::vector<SanityTest*> sanity_tests = {
       new SanityTestBasic(path),
       new SanityTestSpecialComparator(path),
       new SanityTestZlibCompression(path),
-      new SanityTestPlainTableFactory(path)};
+      new SanityTestPlainTableFactory(path),
+      new SanityTestBloomFilter(path)};
 
   if (command == "create") {
     fprintf(stderr, "Creating...\n");

tools/db_stress.cc

@@ -209,6 +209,9 @@ static const bool FLAGS_reopen_dummy __attribute__((unused)) =
 DEFINE_int32(bloom_bits, 10, "Bloom filter bits per key. "
              "Negative means use default settings.");
 
+DEFINE_bool(use_block_based_filter, false, "use block based filter"
+              "instead of full filter for block based table");
+
 DEFINE_string(db, "", "Use the db with the following name.");
 
 DEFINE_bool(verify_checksum, false,
@@ -757,8 +760,10 @@ class StressTest {
                               ? NewLRUCache(FLAGS_compressed_cache_size)
                               : nullptr),
         filter_policy_(FLAGS_bloom_bits >= 0
-                           ? NewBloomFilterPolicy(FLAGS_bloom_bits)
+                   ? FLAGS_use_block_based_filter
-                           : nullptr),
+                     ? NewBloomFilterPolicy(FLAGS_bloom_bits, true)
+                     : NewBloomFilterPolicy(FLAGS_bloom_bits, false)
+                   : nullptr),
         db_(nullptr),
         new_column_family_name_(1),
         num_times_reopened_(0) {

util/bloom.cc

@@ -1,4 +1,4 @@
-//  Copyright (c) 2013, Facebook, Inc.  All rights reserved.
+//  Copyright (c) 2014, Facebook, Inc.  All rights reserved.
 //  This source code is licensed under the BSD-style license found in the
 //  LICENSE file in the root directory of this source tree. An additional grant
 //  of patent rights can be found in the PATENTS file in the same directory.
@@ -10,42 +10,266 @@
 #include "rocksdb/filter_policy.h"
 
 #include "rocksdb/slice.h"
+#include "table/block_based_filter_block.h"
+#include "table/full_filter_block.h"
 #include "util/hash.h"
+#include "util/coding.h"
 
 namespace rocksdb {
 
-namespace {
+class BlockBasedFilterBlockBuilder;
+class FullFilterBlockBuilder;
+
+namespace {
+class FullFilterBitsBuilder : public FilterBitsBuilder {
+ public:
+  explicit FullFilterBitsBuilder(const size_t bits_per_key,
+                                 const size_t num_probes)
+      : bits_per_key_(bits_per_key),
+        num_probes_(num_probes) {
+    assert(bits_per_key_);
+  }
+
+  ~FullFilterBitsBuilder() {}
+
+  virtual void AddKey(const Slice& key) override {
+    uint32_t hash = BloomHash(key);
+    if (hash_entries_.size() == 0 || hash != hash_entries_.back()) {
+      hash_entries_.push_back(hash);
+    }
+  }
+
+  // Create a filter that for hashes [0, n-1], the filter is allocated here
+  // When creating filter, it is ensured that
+  // total_bits = num_lines * CACHE_LINE_SIZE * 8
+  // dst len is >= 5, 1 for num_probes, 4 for num_lines
+  // Then total_bits = (len - 5) * 8, and cache_line_size could be calulated
+  // +----------------------------------------------------------------+
+  // |              filter data with length total_bits/8              |
+  // +----------------------------------------------------------------+
+  // |                                                                |
+  // | ...                                                            |
+  // |                                                                |
+  // +----------------------------------------------------------------+
+  // | ...                | num_probes : 1 byte | num_lines : 4 bytes |
+  // +----------------------------------------------------------------+
+  virtual Slice Finish(std::unique_ptr<const char[]>* buf) override {
+    char* data = nullptr;
+    uint32_t total_bits, num_lines;
+    data = ReserveSpace(hash_entries_.size(), &total_bits, &num_lines);
+    assert(data);
+
+    if (total_bits != 0 && num_lines != 0) {
+      for (auto h : hash_entries_) {
+        AddHash(h, data, num_lines, total_bits);
+      }
+    }
+    data[total_bits/8] = static_cast<char>(num_probes_);
+    EncodeFixed32(data + total_bits/8 + 1, static_cast<uint32_t>(num_lines));
+
+    buf->reset(data);
+    hash_entries_.clear();
+
+    return Slice(data, total_bits / 8 + 5);
+  }
 
-class BloomFilterPolicy : public FilterPolicy {
  private:
   size_t bits_per_key_;
-  size_t k_;
+  size_t num_probes_;
-  uint32_t (*hash_func_)(const Slice& key);
+  std::vector<uint32_t> hash_entries_;
 
-  void initialize() {
+  // Get totalbits that optimized for cpu cache line
-    // We intentionally round down to reduce probing cost a little bit
+  uint32_t GetTotalBitsForLocality(uint32_t total_bits);
-    k_ = static_cast<size_t>(bits_per_key_ * 0.69);  // 0.69 =~ ln(2)
+
-    if (k_ < 1) k_ = 1;
+  // Reserve space for new filter
-    if (k_ > 30) k_ = 30;
+  char* ReserveSpace(const int num_entry, uint32_t* total_bits,
+      uint32_t* num_lines);
+
+  // Assuming single threaded access to this function.
+  void AddHash(uint32_t h, char* data, uint32_t num_lines,
+      uint32_t total_bits);
+
+  // No Copy allowed
+  FullFilterBitsBuilder(const FullFilterBitsBuilder&);
+  void operator=(const FullFilterBitsBuilder&);
+};
+
+uint32_t FullFilterBitsBuilder::GetTotalBitsForLocality(uint32_t total_bits) {
+  uint32_t num_lines =
+      (total_bits + CACHE_LINE_SIZE * 8 - 1) / (CACHE_LINE_SIZE * 8);
+
+  // Make num_lines an odd number to make sure more bits are involved
+  // when determining which block.
+  if (num_lines % 2 == 0) {
+    num_lines++;
+  }
+  return num_lines * (CACHE_LINE_SIZE * 8);
+}
+
+char* FullFilterBitsBuilder::ReserveSpace(const int num_entry,
+    uint32_t* total_bits, uint32_t* num_lines) {
+  assert(bits_per_key_);
+  char* data = nullptr;
+  if (num_entry != 0) {
+    uint32_t total_bits_tmp = num_entry * bits_per_key_;
+
+    *total_bits = GetTotalBitsForLocality(total_bits_tmp);
+    *num_lines = *total_bits / (CACHE_LINE_SIZE * 8);
+    assert(*total_bits > 0 && *total_bits % 8 == 0);
+  } else {
+    // filter is empty, just leave space for metadata
+    *total_bits = 0;
+    *num_lines = 0;
   }
 
+  // Reserve space for Filter
+  uint32_t sz = *total_bits / 8;
+  sz += 5;  // 4 bytes for num_lines, 1 byte for num_probes
+
+  data = new char[sz];
+  memset(data, 0, sz);
+  return data;
+}
+
+inline void FullFilterBitsBuilder::AddHash(uint32_t h, char* data,
+    uint32_t num_lines, uint32_t total_bits) {
+  assert(num_lines > 0 && total_bits > 0);
+
+  const uint32_t delta = (h >> 17) | (h << 15);  // Rotate right 17 bits
+  uint32_t b = (h % num_lines) * (CACHE_LINE_SIZE * 8);
+
+  for (uint32_t i = 0; i < num_probes_; ++i) {
+    // Since CACHE_LINE_SIZE is defined as 2^n, this line will be optimized
+    // to a simple operation by compiler.
+    const uint32_t bitpos = b + (h % (CACHE_LINE_SIZE * 8));
+    data[bitpos / 8] |= (1 << (bitpos % 8));
+
+    h += delta;
+  }
+}
+
+class FullFilterBitsReader : public FilterBitsReader {
  public:
-  explicit BloomFilterPolicy(int bits_per_key,
+  explicit FullFilterBitsReader(const Slice& contents)
-                             uint32_t (*hash_func)(const Slice& key))
+      : data_(const_cast<char*>(contents.data())),
-      : bits_per_key_(bits_per_key), hash_func_(hash_func) {
+        data_len_(contents.size()),
-    initialize();
+        num_probes_(0), num_lines_(0) {
+    assert(data_);
+    GetFilterMeta(contents, &num_probes_, &num_lines_);
+    // Sanitize broken parameter
+    if (num_lines_ != 0 && (data_len_-5) % num_lines_ != 0) {
+      num_lines_ = 0;
+      num_probes_ = 0;
+    }
   }
-  explicit BloomFilterPolicy(int bits_per_key)
+
-      : bits_per_key_(bits_per_key) {
+  ~FullFilterBitsReader() {}
-    hash_func_ = BloomHash;
+
+  virtual bool MayMatch(const Slice& entry) override {
+    if (data_len_ <= 5) {   // remain same with original filter
+      return false;
+    }
+    // Other Error params, including a broken filter, regarded as match
+    if (num_probes_ == 0 || num_lines_ == 0) return true;
+    uint32_t hash = BloomHash(entry);
+    return HashMayMatch(hash, Slice(data_, data_len_),
+                        num_probes_, num_lines_);
+  }
+
+ private:
+  // Filter meta data
+  char* data_;
+  uint32_t data_len_;
+  size_t num_probes_;
+  uint32_t num_lines_;
+
+  // Get num_probes, and num_lines from filter
+  // If filter format broken, set both to 0.
+  void GetFilterMeta(const Slice& filter, size_t* num_probes,
+                             uint32_t* num_lines);
+
+  // "filter" contains the data appended by a preceding call to
+  // CreateFilterFromHash() on this class.  This method must return true if
+  // the key was in the list of keys passed to CreateFilter().
+  // This method may return true or false if the key was not on the
+  // list, but it should aim to return false with a high probability.
+  //
+  // hash: target to be checked
+  // filter: the whole filter, including meta data bytes
+  // num_probes: number of probes, read before hand
+  // num_lines: filter metadata, read before hand
+  // Before calling this function, need to ensure the input meta data
+  // is valid.
+  bool HashMayMatch(const uint32_t& hash, const Slice& filter,
+      const size_t& num_probes, const uint32_t& num_lines);
+
+  // No Copy allowed
+  FullFilterBitsReader(const FullFilterBitsReader&);
+  void operator=(const FullFilterBitsReader&);
+};
+
+void FullFilterBitsReader::GetFilterMeta(const Slice& filter,
+    size_t* num_probes, uint32_t* num_lines) {
+  uint32_t len = filter.size();
+  if (len <= 5) {
+    // filter is empty or broken
+    *num_probes = 0;
+    *num_lines = 0;
+    return;
+  }
+
+  *num_probes = filter.data()[len - 5];
+  *num_lines = DecodeFixed32(filter.data() + len - 4);
+}
+
+bool FullFilterBitsReader::HashMayMatch(const uint32_t& hash,
+    const Slice& filter, const size_t& num_probes,
+    const uint32_t& num_lines) {
+  uint32_t len = filter.size();
+  if (len <= 5) return false;  // remain the same with original filter
+
+  // It is ensured the params are valid before calling it
+  assert(num_probes != 0);
+  assert(num_lines != 0 && (len - 5) % num_lines == 0);
+  uint32_t cache_line_size = (len - 5) / num_lines;
+  const char* data = filter.data();
+
+  uint32_t h = hash;
+  const uint32_t delta = (h >> 17) | (h << 15);  // Rotate right 17 bits
+  uint32_t b = (h % num_lines) * (cache_line_size * 8);
+
+  for (uint32_t i = 0; i < num_probes; ++i) {
+    // Since CACHE_LINE_SIZE is defined as 2^n, this line will be optimized
+    //  to a simple and operation by compiler.
+    const uint32_t bitpos = b + (h % (cache_line_size * 8));
+    if (((data[bitpos / 8]) & (1 << (bitpos % 8))) == 0) {
+      return false;
+    }
+
+    h += delta;
+  }
+
+  return true;
+}
+
+// An implementation of filter policy
+class BloomFilterPolicy : public FilterPolicy {
+ public:
+  explicit BloomFilterPolicy(int bits_per_key, bool use_block_based_builder)
+      : bits_per_key_(bits_per_key), hash_func_(BloomHash),
+        use_block_based_builder_(use_block_based_builder) {
     initialize();
   }
 
-  virtual const char* Name() const {
+  ~BloomFilterPolicy() {
+  }
+
+  virtual const char* Name() const override {
     return "rocksdb.BuiltinBloomFilter";
   }
 
-  virtual void CreateFilter(const Slice* keys, int n, std::string* dst) const {
+  virtual void CreateFilter(const Slice* keys, int n,
+                            std::string* dst) const override {
     // Compute bloom filter size (in both bits and bytes)
     size_t bits = n * bits_per_key_;
 
@@ -58,14 +282,14 @@ class BloomFilterPolicy : public FilterPolicy {
 
     const size_t init_size = dst->size();
     dst->resize(init_size + bytes, 0);
-    dst->push_back(static_cast<char>(k_));  // Remember # of probes in filter
+    dst->push_back(static_cast<char>(num_probes_));  // Remember # of probes
     char* array = &(*dst)[init_size];
     for (size_t i = 0; i < (size_t)n; i++) {
       // Use double-hashing to generate a sequence of hash values.
       // See analysis in [Kirsch,Mitzenmacher 2006].
       uint32_t h = hash_func_(keys[i]);
       const uint32_t delta = (h >> 17) | (h << 15);  // Rotate right 17 bits
-      for (size_t j = 0; j < k_; j++) {
+      for (size_t j = 0; j < num_probes_; j++) {
         const uint32_t bitpos = h % bits;
         array[bitpos/8] |= (1 << (bitpos % 8));
         h += delta;
@@ -73,7 +297,8 @@ class BloomFilterPolicy : public FilterPolicy {
     }
   }
 
-  virtual bool KeyMayMatch(const Slice& key, const Slice& bloom_filter) const {
+  virtual bool KeyMayMatch(const Slice& key,
+                           const Slice& bloom_filter) const override {
     const size_t len = bloom_filter.size();
     if (len < 2) return false;
 
@@ -98,11 +323,43 @@ class BloomFilterPolicy : public FilterPolicy {
     }
     return true;
   }
-};
-}
 
-const FilterPolicy* NewBloomFilterPolicy(int bits_per_key) {
+  virtual FilterBitsBuilder* GetFilterBitsBuilder() const override {
-  return new BloomFilterPolicy(bits_per_key);
+    if (use_block_based_builder_) {
+      return nullptr;
+    }
+
+    return new FullFilterBitsBuilder(bits_per_key_, num_probes_);
+  }
+
+  virtual FilterBitsReader* GetFilterBitsReader(const Slice& contents)
+      const override {
+    return new FullFilterBitsReader(contents);
+  }
+
+  // If choose to use block based builder
+  bool UseBlockBasedBuilder() { return use_block_based_builder_; }
+
+ private:
+  size_t bits_per_key_;
+  size_t num_probes_;
+  uint32_t (*hash_func_)(const Slice& key);
+
+  const bool use_block_based_builder_;
+
+  void initialize() {
+    // We intentionally round down to reduce probing cost a little bit
+    num_probes_ = static_cast<size_t>(bits_per_key_ * 0.69);  // 0.69 =~ ln(2)
+    if (num_probes_ < 1) num_probes_ = 1;
+    if (num_probes_ > 30) num_probes_ = 30;
+  }
+};
+
+}  // namespace
+
+const FilterPolicy* NewBloomFilterPolicy(int bits_per_key,
+                                         bool use_block_based_builder) {
+  return new BloomFilterPolicy(bits_per_key, use_block_based_builder);
 }
 
 }  // namespace rocksdb

util/bloom_test.cc

@@ -16,12 +16,13 @@ int main() {
 #else
 
 #include <gflags/gflags.h>
+#include <vector>
 
 #include "rocksdb/filter_policy.h"
-
 #include "util/logging.h"
 #include "util/testharness.h"
 #include "util/testutil.h"
+#include "util/arena.h"
 
 using GFLAGS::ParseCommandLineFlags;
 
@@ -36,6 +37,19 @@ static Slice Key(int i, char* buffer) {
   return Slice(buffer, sizeof(i));
 }
 
+static int NextLength(int length) {
+  if (length < 10) {
+    length += 1;
+  } else if (length < 100) {
+    length += 10;
+  } else if (length < 1000) {
+    length += 100;
+  } else {
+    length += 1000;
+  }
+  return length;
+}
+
 class BloomTest {
  private:
   const FilterPolicy* policy_;
@@ -43,7 +57,8 @@ class BloomTest {
   std::vector<std::string> keys_;
 
  public:
-  BloomTest() : policy_(NewBloomFilterPolicy(FLAGS_bits_per_key)) { }
+  BloomTest() : policy_(
+      NewBloomFilterPolicy(FLAGS_bits_per_key)) {}
 
   ~BloomTest() {
     delete policy_;
@@ -117,19 +132,6 @@ TEST(BloomTest, Small) {
   ASSERT_TRUE(! Matches("foo"));
 }
 
-static int NextLength(int length) {
-  if (length < 10) {
-    length += 1;
-  } else if (length < 100) {
-    length += 10;
-  } else if (length < 1000) {
-    length += 100;
-  } else {
-    length += 1000;
-  }
-  return length;
-}
-
 TEST(BloomTest, VaryingLengths) {
   char buffer[sizeof(int)];
 
@@ -171,6 +173,121 @@ TEST(BloomTest, VaryingLengths) {
 
 // Different bits-per-byte
 
+class FullBloomTest {
+ private:
+  const FilterPolicy* policy_;
+  std::unique_ptr<FilterBitsBuilder> bits_builder_;
+  std::unique_ptr<FilterBitsReader> bits_reader_;
+  std::unique_ptr<const char[]> buf_;
+  size_t filter_size_;
+
+ public:
+  FullBloomTest() :
+      policy_(NewBloomFilterPolicy(FLAGS_bits_per_key, false)),
+      filter_size_(0) {
+    Reset();
+  }
+
+  ~FullBloomTest() {
+    delete policy_;
+  }
+
+  void Reset() {
+    bits_builder_.reset(policy_->GetFilterBitsBuilder());
+    bits_reader_.reset(nullptr);
+    buf_.reset(nullptr);
+    filter_size_ = 0;
+  }
+
+  void Add(const Slice& s) {
+    bits_builder_->AddKey(s);
+  }
+
+  void Build() {
+    Slice filter = bits_builder_->Finish(&buf_);
+    bits_reader_.reset(policy_->GetFilterBitsReader(filter));
+    filter_size_ = filter.size();
+  }
+
+  size_t FilterSize() const {
+    return filter_size_;
+  }
+
+  bool Matches(const Slice& s) {
+    if (bits_reader_ == nullptr) {
+      Build();
+    }
+    return bits_reader_->MayMatch(s);
+  }
+
+  double FalsePositiveRate() {
+    char buffer[sizeof(int)];
+    int result = 0;
+    for (int i = 0; i < 10000; i++) {
+      if (Matches(Key(i + 1000000000, buffer))) {
+        result++;
+      }
+    }
+    return result / 10000.0;
+  }
+};
+
+TEST(FullBloomTest, FullEmptyFilter) {
+  // Empty filter is not match, at this level
+  ASSERT_TRUE(!Matches("hello"));
+  ASSERT_TRUE(!Matches("world"));
+}
+
+TEST(FullBloomTest, FullSmall) {
+  Add("hello");
+  Add("world");
+  ASSERT_TRUE(Matches("hello"));
+  ASSERT_TRUE(Matches("world"));
+  ASSERT_TRUE(!Matches("x"));
+  ASSERT_TRUE(!Matches("foo"));
+}
+
+TEST(FullBloomTest, FullVaryingLengths) {
+  char buffer[sizeof(int)];
+
+  // Count number of filters that significantly exceed the false positive rate
+  int mediocre_filters = 0;
+  int good_filters = 0;
+
+  for (int length = 1; length <= 10000; length = NextLength(length)) {
+    Reset();
+    for (int i = 0; i < length; i++) {
+      Add(Key(i, buffer));
+    }
+    Build();
+
+    ASSERT_LE(FilterSize(), (size_t)((length * 10 / 8) + 128 + 5)) << length;
+
+    // All added keys must match
+    for (int i = 0; i < length; i++) {
+      ASSERT_TRUE(Matches(Key(i, buffer)))
+          << "Length " << length << "; key " << i;
+    }
+
+    // Check false positive rate
+    double rate = FalsePositiveRate();
+    if (kVerbose >= 1) {
+      fprintf(stderr, "False positives: %5.2f%% @ length = %6d ; bytes = %6d\n",
+              rate*100.0, length, static_cast<int>(FilterSize()));
+    }
+    ASSERT_LE(rate, 0.02);   // Must not be over 2%
+    if (rate > 0.0125)
+      mediocre_filters++;  // Allowed, but not too often
+    else
+      good_filters++;
+  }
+  if (kVerbose >= 1) {
+    fprintf(stderr, "Filters: %d good, %d mediocre\n",
+            good_filters, mediocre_filters);
+  }
+  ASSERT_LE(mediocre_filters, good_filters/5);
+}
+
 }  // namespace rocksdb
 
 int main(int argc, char** argv) {