Improve SmallEnumSet (#11178)

Summary:
In anticipation of using this to represent sets of CacheEntryRole for including or excluding kinds of blocks in block cache tiers, add significant new features to SmallEnumSet, including at least:

* List initialization
* Applicative constexpr operations
* copy/move/equality ops
* begin/end/const_iterator for iteration
* Better comments

Pull Request resolved: https://github.com/facebook/rocksdb/pull/11178

Test Plan: unit tests added/expanded

Reviewed By: ltamasi

Differential Revision: D42973723

Pulled By: pdillinger

fbshipit-source-id: 40783486feda931c3f7c6fcc9a300acd6a4b0a0a

CMakeLists.txt

@@ -879,6 +879,7 @@ set(SOURCES
         util/compression_context_cache.cc
         util/concurrent_task_limiter_impl.cc
         util/crc32c.cc
+        util/data_structure.cc
         util/dynamic_bloom.cc
         util/hash.cc
         util/murmurhash.cc

TARGETS

@@ -247,6 +247,7 @@ cpp_library_wrapper(name="rocksdb_lib", srcs=[
         "util/concurrent_task_limiter_impl.cc",
         "util/crc32c.cc",
         "util/crc32c_arm64.cc",
+        "util/data_structure.cc",
         "util/dynamic_bloom.cc",
         "util/file_checksum_helper.cc",
         "util/hash.cc",
@@ -589,6 +590,7 @@ cpp_library_wrapper(name="rocksdb_whole_archive_lib", srcs=[
         "util/concurrent_task_limiter_impl.cc",
         "util/crc32c.cc",
         "util/crc32c_arm64.cc",
+        "util/data_structure.cc",
         "util/dynamic_bloom.cc",
         "util/file_checksum_helper.cc",
         "util/hash.cc",

include/rocksdb/data_structure.h

@@ -15,37 +15,172 @@
 
 namespace ROCKSDB_NAMESPACE {
 
-// This is a data structure specifically designed as a "Set" for a
-// pretty small scale of Enum structure. For now, it can support up
-// to 64 element, and it is expandable in the future.
-template <typename ENUM_TYPE, ENUM_TYPE MAX_VALUE>
+namespace detail {
+int CountTrailingZeroBitsForSmallEnumSet(uint64_t);
+}  // namespace detail
+
+// Represents a set of values of some enum type with a small number of
+// possible enumerators. For now, it supports enums where no enumerator
+// exceeds 63 when converted to int.
+template <typename ENUM_TYPE, ENUM_TYPE MAX_ENUMERATOR>
 class SmallEnumSet {
+ private:
+  using StateT = uint64_t;
+  static constexpr int kStateBits = sizeof(StateT) * 8;
+  static constexpr int kMaxMax = kStateBits - 1;
+  static constexpr int kMaxValue = static_cast<int>(MAX_ENUMERATOR);
+  static_assert(kMaxValue >= 0);
+  static_assert(kMaxValue <= kMaxMax);
+
  public:
+  // construct / create
   SmallEnumSet() : state_(0) {}
 
-  ~SmallEnumSet() {}
-
-  // Return true if the input enum is included in the "Set" (i.e., changes the
-  // internal scalar state successfully), otherwise, it will return false.
-  bool Add(const ENUM_TYPE value) {
-    static_assert(MAX_VALUE <= 63, "Size currently limited to 64");
-    assert(value >= 0 && value <= MAX_VALUE);
-    uint64_t old_state = state_;
-    uint64_t tmp = 1;
-    state_ |= (tmp << value);
-    return old_state != state_;
+  template <class... TRest>
+  /*implicit*/ constexpr SmallEnumSet(const ENUM_TYPE e, TRest... rest) {
+    *this = SmallEnumSet(rest...).With(e);
   }
 
+  // Return the set that includes all valid values, assuming the enum
+  // is "dense" (includes all values converting to 0 through kMaxValue)
+  static constexpr SmallEnumSet All() {
+    StateT tmp = StateT{1} << kMaxValue;
+    return SmallEnumSet(RawStateMarker(), tmp | (tmp - 1));
+  }
+
+  // equality
+  bool operator==(const SmallEnumSet& that) const {
+    return this->state_ == that.state_;
+  }
+  bool operator!=(const SmallEnumSet& that) const { return !(*this == that); }
+
+  // query
+
   // Return true if the input enum is contained in the "Set".
-  bool Contains(const ENUM_TYPE value) {
-    static_assert(MAX_VALUE <= 63, "Size currently limited to 64");
-    assert(value >= 0 && value <= MAX_VALUE);
-    uint64_t tmp = 1;
+  bool Contains(const ENUM_TYPE e) const {
+    int value = static_cast<int>(e);
+    assert(value >= 0 && value <= kMaxValue);
+    StateT tmp = 1;
     return state_ & (tmp << value);
   }
 
+  bool empty() const { return state_ == 0; }
+
+  // iterator
+  class const_iterator {
+   public:
+    // copy
+    const_iterator(const const_iterator& that) = default;
+    const_iterator& operator=(const const_iterator& that) = default;
+
+    // move
+    const_iterator(const_iterator&& that) noexcept = default;
+    const_iterator& operator=(const_iterator&& that) noexcept = default;
+
+    // equality
+    bool operator==(const const_iterator& that) const {
+      assert(set_ == that.set_);
+      return this->pos_ == that.pos_;
+    }
+
+    bool operator!=(const const_iterator& that) const {
+      return !(*this == that);
+    }
+
+    // ++iterator
+    const_iterator& operator++() {
+      if (pos_ < kMaxValue) {
+        pos_ = set_->SkipUnset(pos_ + 1);
+      } else {
+        pos_ = kStateBits;
+      }
+      return *this;
+    }
+
+    // iterator++
+    const_iterator operator++(int) {
+      auto old = *this;
+      ++*this;
+      return old;
+    }
+
+    ENUM_TYPE operator*() const {
+      assert(pos_ <= kMaxValue);
+      return static_cast<ENUM_TYPE>(pos_);
+    }
+
    private:
-  uint64_t state_;
+    friend class SmallEnumSet;
+    const_iterator(const SmallEnumSet* set, int pos) : set_(set), pos_(pos) {}
+    const SmallEnumSet* set_;
+    int pos_;
+  };
+
+  const_iterator begin() const { return const_iterator(this, SkipUnset(0)); }
+
+  const_iterator end() const { return const_iterator(this, kStateBits); }
+
+  // mutable ops
+
+  // Modifies the set (if needed) to include the given value. Returns true
+  // iff the set was modified.
+  bool Add(const ENUM_TYPE e) {
+    int value = static_cast<int>(e);
+    assert(value >= 0 && value <= kMaxValue);
+    StateT old_state = state_;
+    state_ |= (StateT{1} << value);
+    return old_state != state_;
+  }
+
+  // Modifies the set (if needed) not to include the given value. Returns true
+  // iff the set was modified.
+  bool Remove(const ENUM_TYPE e) {
+    int value = static_cast<int>(e);
+    assert(value >= 0 && value <= kMaxValue);
+    StateT old_state = state_;
+    state_ &= ~(StateT{1} << value);
+    return old_state != state_;
+  }
+
+  // applicative ops
+
+  // Return a new set based on this one with the additional value(s) inserted
+  constexpr SmallEnumSet With(const ENUM_TYPE e) const {
+    int value = static_cast<int>(e);
+    assert(value >= 0 && value <= kMaxValue);
+    return SmallEnumSet(RawStateMarker(), state_ | (StateT{1} << value));
+  }
+  template <class... TRest>
+  constexpr SmallEnumSet With(const ENUM_TYPE e1, const ENUM_TYPE e2,
+                              TRest... rest) const {
+    return With(e1).With(e2, rest...);
+  }
+
+  // Return a new set based on this one excluding the given value(s)
+  constexpr SmallEnumSet Without(const ENUM_TYPE e) const {
+    int value = static_cast<int>(e);
+    assert(value >= 0 && value <= kMaxValue);
+    return SmallEnumSet(RawStateMarker(), state_ & ~(StateT{1} << value));
+  }
+  template <class... TRest>
+  constexpr SmallEnumSet Without(const ENUM_TYPE e1, const ENUM_TYPE e2,
+                                 TRest... rest) const {
+    return Without(e1).Without(e2, rest...);
+  }
+
+ private:
+  int SkipUnset(int pos) const {
+    StateT tmp = state_ >> pos;
+    if (tmp == 0) {
+      return kStateBits;
+    } else {
+      return pos + detail::CountTrailingZeroBitsForSmallEnumSet(tmp);
+    }
+  }
+  struct RawStateMarker {};
+  explicit SmallEnumSet(RawStateMarker, StateT state) : state_(state) {}
+
+  StateT state_;
 };
 
 }  // namespace ROCKSDB_NAMESPACE

src.mk

@@ -234,6 +234,7 @@ LIB_SOURCES =                                                   \
   util/concurrent_task_limiter_impl.cc                          \
   util/crc32c.cc                                                \
   util/crc32c_arm64.cc                                          \
+  util/data_structure.cc                                        \
   util/dynamic_bloom.cc                                         \
   util/hash.cc                                                  \
   util/murmurhash.cc                                            \

util/data_structure.cc

@@ -0,0 +1,18 @@
+//  Copyright (c) Meta Platforms, Inc. and affiliates.
+//  This source code is licensed under both the GPLv2 (found in the
+//  COPYING file in the root directory) and Apache 2.0 License
+//  (found in the LICENSE.Apache file in the root directory).
+
+#include "rocksdb/data_structure.h"
+
+#include "util/math.h"
+
+namespace ROCKSDB_NAMESPACE {
+namespace detail {
+
+int CountTrailingZeroBitsForSmallEnumSet(uint64_t v) {
+  return CountTrailingZeroBits(v);
+}
+
+}  // namespace detail
+}  // namespace ROCKSDB_NAMESPACE

util/slice_test.cc

@@ -173,13 +173,74 @@ class SmallEnumSetTest : public testing::Test {
   ~SmallEnumSetTest() {}
 };
 
-TEST_F(SmallEnumSetTest, SmallSetTest) {
-  FileTypeSet fs;
+TEST_F(SmallEnumSetTest, SmallEnumSetTest1) {
+  FileTypeSet fs;  // based on a legacy enum type
+  ASSERT_TRUE(fs.empty());
   ASSERT_TRUE(fs.Add(FileType::kIdentityFile));
+  ASSERT_FALSE(fs.empty());
   ASSERT_FALSE(fs.Add(FileType::kIdentityFile));
   ASSERT_TRUE(fs.Add(FileType::kInfoLogFile));
   ASSERT_TRUE(fs.Contains(FileType::kIdentityFile));
   ASSERT_FALSE(fs.Contains(FileType::kDBLockFile));
+  ASSERT_FALSE(fs.empty());
+  ASSERT_FALSE(fs.Remove(FileType::kDBLockFile));
+  ASSERT_TRUE(fs.Remove(FileType::kIdentityFile));
+  ASSERT_FALSE(fs.empty());
+  ASSERT_TRUE(fs.Remove(FileType::kInfoLogFile));
+  ASSERT_TRUE(fs.empty());
+}
+
+namespace {
+enum class MyEnumClass { A, B, C };
+}  // namespace
+
+using MyEnumClassSet = SmallEnumSet<MyEnumClass, MyEnumClass::C>;
+
+TEST_F(SmallEnumSetTest, SmallEnumSetTest2) {
+  MyEnumClassSet s;  // based on an enum class type
+  ASSERT_TRUE(s.Add(MyEnumClass::A));
+  ASSERT_TRUE(s.Contains(MyEnumClass::A));
+  ASSERT_FALSE(s.Contains(MyEnumClass::B));
+  ASSERT_TRUE(s.With(MyEnumClass::B).Contains(MyEnumClass::B));
+  ASSERT_TRUE(s.With(MyEnumClass::A).Contains(MyEnumClass::A));
+  ASSERT_FALSE(s.Contains(MyEnumClass::B));
+  ASSERT_FALSE(s.Without(MyEnumClass::A).Contains(MyEnumClass::A));
+  ASSERT_FALSE(
+      s.With(MyEnumClass::B).Without(MyEnumClass::B).Contains(MyEnumClass::B));
+  ASSERT_TRUE(
+      s.Without(MyEnumClass::B).With(MyEnumClass::B).Contains(MyEnumClass::B));
+  ASSERT_TRUE(s.Contains(MyEnumClass::A));
+
+  const MyEnumClassSet cs = s;
+  ASSERT_TRUE(cs.Contains(MyEnumClass::A));
+  ASSERT_EQ(cs, MyEnumClassSet{MyEnumClass::A});
+  ASSERT_EQ(cs.Without(MyEnumClass::A), MyEnumClassSet{});
+  ASSERT_EQ(cs, MyEnumClassSet::All().Without(MyEnumClass::B, MyEnumClass::C));
+  ASSERT_EQ(cs.With(MyEnumClass::B, MyEnumClass::C), MyEnumClassSet::All());
+  ASSERT_EQ(
+      MyEnumClassSet::All(),
+      MyEnumClassSet{}.With(MyEnumClass::A, MyEnumClass::B, MyEnumClass::C));
+  ASSERT_NE(cs, MyEnumClassSet{MyEnumClass::B});
+  ASSERT_NE(cs, MyEnumClassSet::All());
+
+  int count = 0;
+  for (MyEnumClass e : cs) {
+    ASSERT_EQ(e, MyEnumClass::A);
+    ++count;
+  }
+  ASSERT_EQ(count, 1);
+
+  count = 0;
+  for (MyEnumClass e : MyEnumClassSet::All().Without(MyEnumClass::B)) {
+    ASSERT_NE(e, MyEnumClass::B);
+    ++count;
+  }
+  ASSERT_EQ(count, 2);
+
+  for (MyEnumClass e : MyEnumClassSet{}) {
+    (void)e;
+    assert(false);
+  }
 }
 
 }  // namespace ROCKSDB_NAMESPACE