2013-10-16 23:59:46 +02:00
|
|
|
// 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.
|
|
|
|
//
|
2011-03-18 23:37:00 +01:00
|
|
|
// 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.
|
|
|
|
|
2014-01-31 02:18:17 +01:00
|
|
|
// Arena is an implementation of Arena class. For a request of small size,
|
Make arena block size configurable
Summary:
Add an option for arena block size, default value 4096 bytes. Arena will allocate blocks with such size.
I am not sure about passing parameter to skiplist in the new virtualized framework, though I talked to Jim a bit. So add Jim as reviewer.
Test Plan:
new unit test, I am running db_test.
For passing paramter from configured option to Arena, I tried tests like:
TEST(DBTest, Arena_Option) {
std::string dbname = test::TmpDir() + "/db_arena_option_test";
DestroyDB(dbname, Options());
DB* db = nullptr;
Options opts;
opts.create_if_missing = true;
opts.arena_block_size = 1000000; // tested 99, 999999
Status s = DB::Open(opts, dbname, &db);
db->Put(WriteOptions(), "a", "123");
}
and printed some debug info. The results look good. Any suggestion for such a unit-test?
Reviewers: haobo, dhruba, emayanke, jpaton
Reviewed By: dhruba
CC: leveldb, zshao
Differential Revision: https://reviews.facebook.net/D11799
2013-07-31 21:42:23 +02:00
|
|
|
// it allocates a block with pre-defined block size. For a request of big
|
|
|
|
// size, it uses malloc to directly get the requested size.
|
|
|
|
|
2013-10-05 07:32:05 +02:00
|
|
|
#pragma once
|
2011-03-18 23:37:00 +01:00
|
|
|
#include <cstddef>
|
2014-05-15 06:31:03 +02:00
|
|
|
#include <cerrno>
|
2011-03-18 23:37:00 +01:00
|
|
|
#include <vector>
|
|
|
|
#include <assert.h>
|
|
|
|
#include <stdint.h>
|
2014-01-31 02:18:17 +01:00
|
|
|
#include "util/arena.h"
|
2011-03-18 23:37:00 +01:00
|
|
|
|
2013-10-04 06:49:15 +02:00
|
|
|
namespace rocksdb {
|
2011-03-18 23:37:00 +01:00
|
|
|
|
2014-05-05 00:52:23 +02:00
|
|
|
class Logger;
|
|
|
|
|
In DB::NewIterator(), try to allocate the whole iterator tree in an arena
Summary:
In this patch, try to allocate the whole iterator tree starting from DBIter from an arena
1. ArenaWrappedDBIter is created when serves as the entry point of an iterator tree, with an arena in it.
2. Add an option to create iterator from arena for following iterators: DBIter, MergingIterator, MemtableIterator, all mem table's iterators, all table reader's iterators and two level iterator.
3. MergeIteratorBuilder is created to incrementally build the tree of internal iterators. It is passed to mem table list and version set and add iterators to it.
Limitations:
(1) Only DB::NewIterator() without tailing uses the arena. Other cases, including readonly DB and compactions are still from malloc
(2) Two level iterator itself is allocated in arena, but not iterators inside it.
Test Plan: make all check
Reviewers: ljin, haobo
Reviewed By: haobo
Subscribers: leveldb, dhruba, yhchiang, igor
Differential Revision: https://reviews.facebook.net/D18513
2014-06-03 01:38:00 +02:00
|
|
|
const size_t kInlineSize = 2048;
|
|
|
|
|
2014-01-31 02:18:17 +01:00
|
|
|
class Arena {
|
2011-03-18 23:37:00 +01:00
|
|
|
public:
|
2014-01-09 00:06:07 +01:00
|
|
|
// No copying allowed
|
2014-01-31 02:18:17 +01:00
|
|
|
Arena(const Arena&) = delete;
|
|
|
|
void operator=(const Arena&) = delete;
|
2014-01-09 00:06:07 +01:00
|
|
|
|
2014-05-14 21:36:21 +02:00
|
|
|
static const size_t kInlineSize = 2048;
|
2014-01-09 00:06:07 +01:00
|
|
|
static const size_t kMinBlockSize;
|
|
|
|
static const size_t kMaxBlockSize;
|
|
|
|
|
2014-01-31 02:18:17 +01:00
|
|
|
explicit Arena(size_t block_size = kMinBlockSize);
|
|
|
|
~Arena();
|
2011-03-18 23:37:00 +01:00
|
|
|
|
2014-01-31 02:18:17 +01:00
|
|
|
char* Allocate(size_t bytes);
|
2011-03-18 23:37:00 +01:00
|
|
|
|
2014-05-09 20:01:54 +02:00
|
|
|
// huge_page_size: if >0, will try to allocate from huage page TLB.
|
|
|
|
// The argument will be the size of the page size for huge page TLB. Bytes
|
|
|
|
// will be rounded up to multiple of the page size to allocate through mmap
|
|
|
|
// anonymous option with huge page on. The extra space allocated will be
|
|
|
|
// wasted. If allocation fails, will fall back to normal case. To enable it,
|
|
|
|
// need to reserve huge pages for it to be allocated, like:
|
2014-05-04 22:55:53 +02:00
|
|
|
// sysctl -w vm.nr_hugepages=20
|
|
|
|
// See linux doc Documentation/vm/hugetlbpage.txt for details.
|
2014-05-05 00:52:23 +02:00
|
|
|
// huge page allocation can fail. In this case it will fail back to
|
|
|
|
// normal cases. The messages will be logged to logger. So when calling with
|
|
|
|
// huge_page_tlb_size > 0, we highly recommend a logger is passed in.
|
|
|
|
// Otherwise, the error message will be printed out to stderr directly.
|
2014-05-09 20:01:54 +02:00
|
|
|
char* AllocateAligned(size_t bytes, size_t huge_page_size = 0,
|
2014-05-05 00:52:23 +02:00
|
|
|
Logger* logger = nullptr);
|
2011-03-18 23:37:00 +01:00
|
|
|
|
|
|
|
// Returns an estimate of the total memory usage of data allocated
|
2014-01-09 00:06:07 +01:00
|
|
|
// by the arena (exclude the space allocated but not yet used for future
|
2011-03-18 23:37:00 +01:00
|
|
|
// allocations).
|
2014-02-18 23:58:55 +01:00
|
|
|
size_t ApproximateMemoryUsage() const {
|
2014-01-09 00:06:07 +01:00
|
|
|
return blocks_memory_ + blocks_.capacity() * sizeof(char*) -
|
|
|
|
alloc_bytes_remaining_;
|
2011-03-18 23:37:00 +01:00
|
|
|
}
|
|
|
|
|
2014-02-18 23:58:55 +01:00
|
|
|
size_t MemoryAllocatedBytes() const { return blocks_memory_; }
|
Make arena block size configurable
Summary:
Add an option for arena block size, default value 4096 bytes. Arena will allocate blocks with such size.
I am not sure about passing parameter to skiplist in the new virtualized framework, though I talked to Jim a bit. So add Jim as reviewer.
Test Plan:
new unit test, I am running db_test.
For passing paramter from configured option to Arena, I tried tests like:
TEST(DBTest, Arena_Option) {
std::string dbname = test::TmpDir() + "/db_arena_option_test";
DestroyDB(dbname, Options());
DB* db = nullptr;
Options opts;
opts.create_if_missing = true;
opts.arena_block_size = 1000000; // tested 99, 999999
Status s = DB::Open(opts, dbname, &db);
db->Put(WriteOptions(), "a", "123");
}
and printed some debug info. The results look good. Any suggestion for such a unit-test?
Reviewers: haobo, dhruba, emayanke, jpaton
Reviewed By: dhruba
CC: leveldb, zshao
Differential Revision: https://reviews.facebook.net/D11799
2013-07-31 21:42:23 +02:00
|
|
|
|
2014-03-13 00:40:14 +01:00
|
|
|
size_t AllocatedAndUnused() const { return alloc_bytes_remaining_; }
|
|
|
|
|
|
|
|
// If an allocation is too big, we'll allocate an irregular block with the
|
|
|
|
// same size of that allocation.
|
2014-11-10 23:39:38 +01:00
|
|
|
size_t IrregularBlockNum() const { return irregular_block_num; }
|
2014-03-13 00:40:14 +01:00
|
|
|
|
2014-03-14 18:02:04 +01:00
|
|
|
size_t BlockSize() const { return kBlockSize; }
|
|
|
|
|
2011-03-18 23:37:00 +01:00
|
|
|
private:
|
2014-05-09 20:01:54 +02:00
|
|
|
char inline_block_[kInlineSize];
|
Make arena block size configurable
Summary:
Add an option for arena block size, default value 4096 bytes. Arena will allocate blocks with such size.
I am not sure about passing parameter to skiplist in the new virtualized framework, though I talked to Jim a bit. So add Jim as reviewer.
Test Plan:
new unit test, I am running db_test.
For passing paramter from configured option to Arena, I tried tests like:
TEST(DBTest, Arena_Option) {
std::string dbname = test::TmpDir() + "/db_arena_option_test";
DestroyDB(dbname, Options());
DB* db = nullptr;
Options opts;
opts.create_if_missing = true;
opts.arena_block_size = 1000000; // tested 99, 999999
Status s = DB::Open(opts, dbname, &db);
db->Put(WriteOptions(), "a", "123");
}
and printed some debug info. The results look good. Any suggestion for such a unit-test?
Reviewers: haobo, dhruba, emayanke, jpaton
Reviewed By: dhruba
CC: leveldb, zshao
Differential Revision: https://reviews.facebook.net/D11799
2013-07-31 21:42:23 +02:00
|
|
|
// Number of bytes allocated in one block
|
2014-01-09 00:06:07 +01:00
|
|
|
const size_t kBlockSize;
|
2011-03-18 23:37:00 +01:00
|
|
|
// Array of new[] allocated memory blocks
|
2014-01-09 00:06:07 +01:00
|
|
|
typedef std::vector<char*> Blocks;
|
|
|
|
Blocks blocks_;
|
2014-05-04 22:55:53 +02:00
|
|
|
|
|
|
|
struct MmapInfo {
|
|
|
|
void* addr_;
|
|
|
|
size_t length_;
|
|
|
|
|
|
|
|
MmapInfo(void* addr, size_t length) : addr_(addr), length_(length) {}
|
|
|
|
};
|
|
|
|
std::vector<MmapInfo> huge_blocks_;
|
2014-03-13 00:40:14 +01:00
|
|
|
size_t irregular_block_num = 0;
|
2014-01-09 00:06:07 +01:00
|
|
|
|
|
|
|
// Stats for current active block.
|
|
|
|
// For each block, we allocate aligned memory chucks from one end and
|
|
|
|
// allocate unaligned memory chucks from the other end. Otherwise the
|
|
|
|
// memory waste for alignment will be higher if we allocate both types of
|
|
|
|
// memory from one direction.
|
|
|
|
char* unaligned_alloc_ptr_ = nullptr;
|
|
|
|
char* aligned_alloc_ptr_ = nullptr;
|
|
|
|
// How many bytes left in currently active block?
|
|
|
|
size_t alloc_bytes_remaining_ = 0;
|
|
|
|
|
|
|
|
char* AllocateFallback(size_t bytes, bool aligned);
|
|
|
|
char* AllocateNewBlock(size_t block_bytes);
|
2011-03-18 23:37:00 +01:00
|
|
|
|
|
|
|
// Bytes of memory in blocks allocated so far
|
2014-01-09 00:06:07 +01:00
|
|
|
size_t blocks_memory_ = 0;
|
2011-03-18 23:37:00 +01:00
|
|
|
};
|
|
|
|
|
2014-01-31 02:18:17 +01:00
|
|
|
inline char* Arena::Allocate(size_t bytes) {
|
2011-03-18 23:37:00 +01:00
|
|
|
// The semantics of what to return are a bit messy if we allow
|
|
|
|
// 0-byte allocations, so we disallow them here (we don't need
|
|
|
|
// them for our internal use).
|
|
|
|
assert(bytes > 0);
|
|
|
|
if (bytes <= alloc_bytes_remaining_) {
|
2014-01-09 00:06:07 +01:00
|
|
|
unaligned_alloc_ptr_ -= bytes;
|
2011-03-18 23:37:00 +01:00
|
|
|
alloc_bytes_remaining_ -= bytes;
|
2014-01-09 00:06:07 +01:00
|
|
|
return unaligned_alloc_ptr_;
|
2011-03-18 23:37:00 +01:00
|
|
|
}
|
2014-01-09 00:06:07 +01:00
|
|
|
return AllocateFallback(bytes, false /* unaligned */);
|
2011-03-18 23:37:00 +01:00
|
|
|
}
|
|
|
|
|
2014-01-09 00:06:07 +01:00
|
|
|
// check and adjust the block_size so that the return value is
|
|
|
|
// 1. in the range of [kMinBlockSize, kMaxBlockSize].
|
|
|
|
// 2. the multiple of align unit.
|
|
|
|
extern size_t OptimizeBlockSize(size_t block_size);
|
|
|
|
|
2013-10-04 06:49:15 +02:00
|
|
|
} // namespace rocksdb
|