rocksdb-native 2.6.2 → 2.6.4

package/vendor/librocksdb/vendor/rocksdb/db/log_writer.h

@@ -1,151 +0,0 @@
-//  Copyright (c) 2011-present, Facebook, Inc.  All rights reserved.
-//  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).
-//
-// Copyright (c) 2011 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.
-#pragma once
-
-#include <cstdint>
-#include <memory>
-#include <unordered_map>
-#include <vector>
-
-#include "db/log_format.h"
-#include "rocksdb/compression_type.h"
-#include "rocksdb/env.h"
-#include "rocksdb/io_status.h"
-#include "rocksdb/slice.h"
-#include "rocksdb/status.h"
-#include "util/compression.h"
-#include "util/hash_containers.h"
-
-namespace ROCKSDB_NAMESPACE {
-
-class WritableFileWriter;
-
-namespace log {
-
-/**
- * Writer is a general purpose log stream writer. It provides an append-only
- * abstraction for writing data. The details of the how the data is written is
- * handled by the WritableFile sub-class implementation.
- *
- * File format:
- *
- * File is broken down into variable sized records. The format of each record
- * is described below.
- *       +-----+-------------+--+----+----------+------+-- ... ----+
- * File  | r0  |        r1   |P | r2 |    r3    |  r4  |           |
- *       +-----+-------------+--+----+----------+------+-- ... ----+
- *       <--- kBlockSize ------>|<-- kBlockSize ------>|
- *  rn = variable size records
- *  P = Padding
- *
- * Data is written out in kBlockSize chunks. If next record does not fit
- * into the space left, the leftover space will be padded with \0.
- *
- * Legacy record format:
- *
- * +---------+-----------+-----------+--- ... ---+
- * |CRC (4B) | Size (2B) | Type (1B) | Payload   |
- * +---------+-----------+-----------+--- ... ---+
- *
- * CRC = 32bit hash computed over the record type and payload using CRC
- * Size = Length of the payload data
- * Type = Type of record
- *        (kZeroType, kFullType, kFirstType, kLastType, kMiddleType )
- *        The type is used to group a bunch of records together to represent
- *        blocks that are larger than kBlockSize
- * Payload = Byte stream as long as specified by the payload size
- *
- * Recyclable record format:
- *
- * +---------+-----------+-----------+----------------+--- ... ---+
- * |CRC (4B) | Size (2B) | Type (1B) | Log number (4B)| Payload   |
- * +---------+-----------+-----------+----------------+--- ... ---+
- *
- * Same as above, with the addition of
- * Log number = 32bit log file number, so that we can distinguish between
- * records written by the most recent log writer vs a previous one.
- */
-class Writer {
- public:
-  // Create a writer that will append data to "*dest".
-  // "*dest" must be initially empty.
-  // "*dest" must remain live while this Writer is in use.
-  explicit Writer(std::unique_ptr<WritableFileWriter>&& dest,
-                  uint64_t log_number, bool recycle_log_files,
-                  bool manual_flush = false,
-                  CompressionType compressionType = kNoCompression);
-  // No copying allowed
-  Writer(const Writer&) = delete;
-  void operator=(const Writer&) = delete;
-
-  ~Writer();
-
-  IOStatus AddRecord(const WriteOptions& write_options, const Slice& slice);
-  IOStatus AddCompressionTypeRecord(const WriteOptions& write_options);
-
-  // If there are column families in `cf_to_ts_sz` not included in
-  // `recorded_cf_to_ts_sz_` and its user-defined timestamp size is non-zero,
-  // adds a record of type kUserDefinedTimestampSizeType or
-  // kRecyclableUserDefinedTimestampSizeType for these column families.
-  // This timestamp size record applies to all subsequent records.
-  IOStatus MaybeAddUserDefinedTimestampSizeRecord(
-      const WriteOptions& write_options,
-      const UnorderedMap<uint32_t, size_t>& cf_to_ts_sz);
-
-  WritableFileWriter* file() { return dest_.get(); }
-  const WritableFileWriter* file() const { return dest_.get(); }
-
-  uint64_t get_log_number() const { return log_number_; }
-
-  IOStatus WriteBuffer(const WriteOptions& write_options);
-
-  IOStatus Close(const WriteOptions& write_options);
-
-  // If closing the writer through file(), call this afterwards to modify
-  // this object's state to reflect that. Returns true if the destination file
-  // has been closed. If it hasn't been closed, returns false with no change.
-  bool PublishIfClosed();
-
-  bool BufferIsEmpty();
-
-  size_t TEST_block_offset() const { return block_offset_; }
-
- private:
-  std::unique_ptr<WritableFileWriter> dest_;
-  size_t block_offset_;  // Current offset in block
-  uint64_t log_number_;
-  bool recycle_log_files_;
-  int header_size_;
-
-  // crc32c values for all supported record types.  These are
-  // pre-computed to reduce the overhead of computing the crc of the
-  // record type stored in the header.
-  uint32_t type_crc_[kMaxRecordType + 1];
-
-  IOStatus EmitPhysicalRecord(const WriteOptions& write_options,
-                              RecordType type, const char* ptr, size_t length);
-
-  // If true, it does not flush after each write. Instead it relies on the upper
-  // layer to manually does the flush by calling ::WriteBuffer()
-  bool manual_flush_;
-
-  // Compression Type
-  CompressionType compression_type_;
-  StreamingCompress* compress_;
-  // Reusable compressed output buffer
-  std::unique_ptr<char[]> compressed_buffer_;
-
-  // The recorded user-defined timestamp size that have been written so far.
-  // Since the user-defined timestamp size cannot be changed while the DB is
-  // running, existing entry in this map cannot be updated.
-  UnorderedMap<uint32_t, size_t> recorded_cf_to_ts_sz_;
-};
-
-}  // namespace log
-}  // namespace ROCKSDB_NAMESPACE

package/vendor/librocksdb/vendor/rocksdb/db/logs_with_prep_tracker.cc

@@ -1,67 +0,0 @@
-//  Copyright (c) 2011-present, Facebook, Inc.  All rights reserved.
-//  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 "db/logs_with_prep_tracker.h"
-
-#include "port/likely.h"
-
-namespace ROCKSDB_NAMESPACE {
-void LogsWithPrepTracker::MarkLogAsHavingPrepSectionFlushed(uint64_t log) {
-  assert(log != 0);
-  std::lock_guard<std::mutex> lock(prepared_section_completed_mutex_);
-  auto it = prepared_section_completed_.find(log);
-  if (UNLIKELY(it == prepared_section_completed_.end())) {
-    prepared_section_completed_[log] = 1;
-  } else {
-    it->second += 1;
-  }
-}
-
-void LogsWithPrepTracker::MarkLogAsContainingPrepSection(uint64_t log) {
-  assert(log != 0);
-  std::lock_guard<std::mutex> lock(logs_with_prep_mutex_);
-
-  auto rit = logs_with_prep_.rbegin();
-  bool updated = false;
-  // Most probably the last log is the one that is being marked for
-  // having a prepare section; so search from the end.
-  for (; rit != logs_with_prep_.rend() && rit->log >= log; ++rit) {
-    if (rit->log == log) {
-      rit->cnt++;
-      updated = true;
-      break;
-    }
-  }
-  if (!updated) {
-    // We are either at the start, or at a position with rit->log < log
-    logs_with_prep_.insert(rit.base(), {log, 1});
-  }
-}
-
-uint64_t LogsWithPrepTracker::FindMinLogContainingOutstandingPrep() {
-  std::lock_guard<std::mutex> lock(logs_with_prep_mutex_);
-  auto it = logs_with_prep_.begin();
-  // start with the smallest log
-  for (; it != logs_with_prep_.end();) {
-    auto min_log = it->log;
-    {
-      std::lock_guard<std::mutex> lock2(prepared_section_completed_mutex_);
-      auto completed_it = prepared_section_completed_.find(min_log);
-      if (completed_it == prepared_section_completed_.end() ||
-          completed_it->second < it->cnt) {
-        return min_log;
-      }
-      assert(completed_it != prepared_section_completed_.end() &&
-             completed_it->second == it->cnt);
-      prepared_section_completed_.erase(completed_it);
-    }
-    // erase from beginning in vector is not efficient but this function is not
-    // on the fast path.
-    it = logs_with_prep_.erase(it);
-  }
-  // no such log found
-  return 0;
-}
-}  // namespace ROCKSDB_NAMESPACE

package/vendor/librocksdb/vendor/rocksdb/db/logs_with_prep_tracker.h

@@ -1,62 +0,0 @@
-//  Copyright (c) 2011-present, Facebook, Inc.  All rights reserved.
-//  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).
-//
-#pragma once
-
-#include <cassert>
-#include <cstdint>
-#include <cstdlib>
-#include <mutex>
-#include <unordered_map>
-#include <vector>
-
-#include "rocksdb/rocksdb_namespace.h"
-
-namespace ROCKSDB_NAMESPACE {
-
-// This class is used to track the log files with outstanding prepare entries.
-class LogsWithPrepTracker {
- public:
-  // Called when a transaction prepared in `log` has been committed or aborted.
-  void MarkLogAsHavingPrepSectionFlushed(uint64_t log);
-  // Called when a transaction is prepared in `log`.
-  void MarkLogAsContainingPrepSection(uint64_t log);
-  // Return the earliest log file with outstanding prepare entries.
-  uint64_t FindMinLogContainingOutstandingPrep();
-  size_t TEST_PreparedSectionCompletedSize() {
-    return prepared_section_completed_.size();
-  }
-  size_t TEST_LogsWithPrepSize() { return logs_with_prep_.size(); }
-
- private:
-  // REQUIRES: logs_with_prep_mutex_ held
-  //
-  // sorted list of log numbers still containing prepared data.
-  // this is used by FindObsoleteFiles to determine which
-  // flushed logs we must keep around because they still
-  // contain prepared data which has not been committed or rolled back
-  struct LogCnt {
-    uint64_t log;  // the log number
-    uint64_t cnt;  // number of prepared sections in the log
-  };
-  std::vector<LogCnt> logs_with_prep_;
-  std::mutex logs_with_prep_mutex_;
-
-  // REQUIRES: prepared_section_completed_mutex_ held
-  //
-  // to be used in conjunction with logs_with_prep_.
-  // once a transaction with data in log L is committed or rolled back
-  // rather than updating logs_with_prep_ directly we keep track of that
-  // in prepared_section_completed_ which maps LOG -> instance_count. This helps
-  // avoiding contention between a commit thread and the prepare threads.
-  //
-  // when trying to determine the minimum log still active we first
-  // consult logs_with_prep_. while that root value maps to
-  // an equal value in prepared_section_completed_ we erase the log from
-  // both logs_with_prep_ and prepared_section_completed_.
-  std::unordered_map<uint64_t, uint64_t> prepared_section_completed_;
-  std::mutex prepared_section_completed_mutex_;
-};
-}  // namespace ROCKSDB_NAMESPACE

package/vendor/librocksdb/vendor/rocksdb/db/lookup_key.h

@@ -1,68 +0,0 @@
-//  Copyright (c) 2011-present, Facebook, Inc.  All rights reserved.
-//  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).
-//
-// Copyright (c) 2011 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.
-
-#pragma once
-#include <string>
-#include <utility>
-
-#include "rocksdb/slice.h"
-#include "rocksdb/types.h"
-
-namespace ROCKSDB_NAMESPACE {
-
-// A helper class useful for DBImpl::Get()
-class LookupKey {
- public:
-  // Initialize *this for looking up user_key at a snapshot with
-  // the specified sequence number.
-  LookupKey(const Slice& _user_key, SequenceNumber sequence,
-            const Slice* ts = nullptr);
-
-  ~LookupKey();
-
-  // Return a key suitable for lookup in a MemTable.
-  Slice memtable_key() const {
-    return Slice(start_, static_cast<size_t>(end_ - start_));
-  }
-
-  // Return an internal key (suitable for passing to an internal iterator)
-  Slice internal_key() const {
-    return Slice(kstart_, static_cast<size_t>(end_ - kstart_));
-  }
-
-  // Return the user key.
-  // If user-defined timestamp is enabled, then timestamp is included in the
-  // result.
-  Slice user_key() const {
-    return Slice(kstart_, static_cast<size_t>(end_ - kstart_ - 8));
-  }
-
- private:
-  // We construct a char array of the form:
-  //    klength  varint32               <-- start_
-  //    userkey  char[klength]          <-- kstart_
-  //    tag      uint64
-  //                                    <-- end_
-  // The array is a suitable MemTable key.
-  // The suffix starting with "userkey" can be used as an InternalKey.
-  const char* start_;
-  const char* kstart_;
-  const char* end_;
-  char space_[200];  // Avoid allocation for short keys
-
-  // No copying allowed
-  LookupKey(const LookupKey&);
-  void operator=(const LookupKey&);
-};
-
-inline LookupKey::~LookupKey() {
-  if (start_ != space_) delete[] start_;
-}
-
-}  // namespace ROCKSDB_NAMESPACE

package/vendor/librocksdb/vendor/rocksdb/db/malloc_stats.cc

@@ -1,52 +0,0 @@
-//  Copyright (c) 2011-present, Facebook, Inc.  All rights reserved.
-//  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).
-//
-// Copyright (c) 2011 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 "db/malloc_stats.h"
-
-#include <cstring>
-#include <memory>
-
-#include "port/jemalloc_helper.h"
-
-namespace ROCKSDB_NAMESPACE {
-
-#ifdef ROCKSDB_JEMALLOC
-
-struct MallocStatus {
-  char* cur;
-  char* end;
-};
-
-static void GetJemallocStatus(void* mstat_arg, const char* status) {
-  MallocStatus* mstat = static_cast<MallocStatus*>(mstat_arg);
-  size_t status_len = status ? strlen(status) : 0;
-  size_t buf_size = (size_t)(mstat->end - mstat->cur);
-  if (!status_len || status_len > buf_size) {
-    return;
-  }
-
-  snprintf(mstat->cur, buf_size, "%s", status);
-  mstat->cur += status_len;
-}
-void DumpMallocStats(std::string* stats) {
-  if (!HasJemalloc()) {
-    return;
-  }
-  MallocStatus mstat;
-  const unsigned int kMallocStatusLen = 1000000;
-  std::unique_ptr<char[]> buf{new char[kMallocStatusLen + 1]};
-  mstat.cur = buf.get();
-  mstat.end = buf.get() + kMallocStatusLen;
-  malloc_stats_print(GetJemallocStatus, &mstat, "");
-  stats->append(buf.get());
-}
-#else
-void DumpMallocStats(std::string*) {}
-#endif  // ROCKSDB_JEMALLOC
-}  // namespace ROCKSDB_NAMESPACE

package/vendor/librocksdb/vendor/rocksdb/db/malloc_stats.h

@@ -1,22 +0,0 @@
-//  Copyright (c) 2011-present, Facebook, Inc.  All rights reserved.
-//  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).
-//
-// Copyright (c) 2011 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.
-
-#pragma once
-
-
-#include <string>
-
-#include "rocksdb/rocksdb_namespace.h"
-
-namespace ROCKSDB_NAMESPACE {
-
-void DumpMallocStats(std::string*);
-
-}
-