runtime/gc/allocator/rosalloc.h

/*
 * Copyright (C) 2013 The Android Open Source Project
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *      http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

#ifndef ART_RUNTIME_GC_ALLOCATOR_ROSALLOC_H_
#define ART_RUNTIME_GC_ALLOCATOR_ROSALLOC_H_

#include <set>
#include <stdint.h>
#include <stdlib.h>
#include <string>
#include <sys/mman.h>
#include <vector>

#include "base/mutex.h"
#include "base/logging.h"
#include "globals.h"
#include "utils.h"

// A boilerplate to use hash_map/hash_set both on host and device.
#ifdef HAVE_ANDROID_OS
#include <hash_map>
#include <hash_set>
using std::hash_map;
using std::hash_set;
#else  // HAVE_ANDROID_OS
#ifdef __DEPRECATED
#define ROSALLOC_OLD__DEPRECATED __DEPRECATED
#undef __DEPRECATED
#endif
#include <ext/hash_map>
#include <ext/hash_set>
#ifdef ROSALLOC_OLD__DEPRECATED
#define __DEPRECATED ROSALLOC_OLD__DEPRECATED
#undef ROSALLOC_OLD__DEPRECATED
#endif
using __gnu_cxx::hash_map;
using __gnu_cxx::hash_set;
#endif  // HAVE_ANDROID_OS

namespace art {
namespace gc {
namespace allocator {

// A Runs-of-slots memory allocator.
class RosAlloc {
 private:
  // Rerepresents a run of free pages.
  class FreePageRun {
   public:
    byte magic_num_;  // The magic number used for debugging only.

    bool IsFree() const {
      if (kIsDebugBuild) {
        return magic_num_ == kMagicNumFree;
      }
      return true;
    }
    size_t ByteSize(RosAlloc* rosalloc) const EXCLUSIVE_LOCKS_REQUIRED(rosalloc->lock_) {
      const byte* fpr_base = reinterpret_cast<const byte*>(this);
      size_t pm_idx = rosalloc->ToPageMapIndex(fpr_base);
      size_t byte_size = rosalloc->free_page_run_size_map_[pm_idx];
      DCHECK_GE(byte_size, static_cast<size_t>(0));
      DCHECK_EQ(byte_size % kPageSize, static_cast<size_t>(0));
      return byte_size;
    }
    void SetByteSize(RosAlloc* rosalloc, size_t byte_size)
        EXCLUSIVE_LOCKS_REQUIRED(rosalloc->lock_) {
      DCHECK_EQ(byte_size % kPageSize, static_cast<size_t>(0));
      byte* fpr_base = reinterpret_cast<byte*>(this);
      size_t pm_idx = rosalloc->ToPageMapIndex(fpr_base);
      rosalloc->free_page_run_size_map_[pm_idx] = byte_size;
    }
    void* Begin() {
      return reinterpret_cast<void*>(this);
    }
    void* End(RosAlloc* rosalloc) EXCLUSIVE_LOCKS_REQUIRED(rosalloc->lock_) {
      byte* fpr_base = reinterpret_cast<byte*>(this);
      byte* end = fpr_base + ByteSize(rosalloc);
      return end;
    }
    bool IsLargerThanPageReleaseThreshold(RosAlloc* rosalloc)
        EXCLUSIVE_LOCKS_REQUIRED(rosalloc->lock_) {
      return ByteSize(rosalloc) >= rosalloc->page_release_size_threshold_;
    }
    bool IsAtEndOfSpace(RosAlloc* rosalloc)
        EXCLUSIVE_LOCKS_REQUIRED(rosalloc->lock_) {
      return reinterpret_cast<byte*>(this) + ByteSize(rosalloc) == rosalloc->base_ + rosalloc->footprint_;
    }
    bool ShouldReleasePages(RosAlloc* rosalloc) EXCLUSIVE_LOCKS_REQUIRED(rosalloc->lock_) {
      switch (rosalloc->page_release_mode_) {
        case kPageReleaseModeNone:
          return false;
        case kPageReleaseModeEnd:
          return IsAtEndOfSpace(rosalloc);
        case kPageReleaseModeSize:
          return IsLargerThanPageReleaseThreshold(rosalloc);
        case kPageReleaseModeSizeAndEnd:
          return IsLargerThanPageReleaseThreshold(rosalloc) && IsAtEndOfSpace(rosalloc);
        case kPageReleaseModeAll:
          return true;
        default:
          LOG(FATAL) << "Unexpected page release mode ";
          return false;
      }
    }
    void ReleasePages(RosAlloc* rosalloc) EXCLUSIVE_LOCKS_REQUIRED(rosalloc->lock_) {
      byte* start = reinterpret_cast<byte*>(this);
      size_t byte_size = ByteSize(rosalloc);
      DCHECK_EQ(byte_size % kPageSize, static_cast<size_t>(0));
      bool release_pages = ShouldReleasePages(rosalloc);
      if (kIsDebugBuild) {
        // Exclude the first page that stores the magic number.
        DCHECK_GE(byte_size, static_cast<size_t>(kPageSize));
        start += kPageSize;
        byte_size -= kPageSize;
        if (byte_size > 0) {
          if (release_pages) {
            madvise(start, byte_size, MADV_DONTNEED);
          }
        }
      } else {
        if (release_pages) {
          madvise(start, byte_size, MADV_DONTNEED);
        }
      }
    }
  };

  // Represents a run of memory slots of the same size.
  //
  // A run's memory layout:
  //
  // +-------------------+
  // | magic_num         |
  // +-------------------+
  // | size_bracket_idx  |
  // +-------------------+
  // | is_thread_local   |
  // +-------------------+
  // | to_be_bulk_freed  |
  // +-------------------+
  // | top_slot_idx      |
  // +-------------------+
  // |                   |
  // | alloc bit map     |
  // |                   |
  // +-------------------+
  // |                   |
  // | bulk free bit map |
  // |                   |
  // +-------------------+
  // |                   |
  // | thread-local free |
  // | bit map           |
  // |                   |
  // +-------------------+
  // | padding due to    |
  // | alignment         |
  // +-------------------+
  // | slot 0            |
  // +-------------------+
  // | slot 1            |
  // +-------------------+
  // | slot 2            |
  // +-------------------+
  // ...
  // +-------------------+
  // | last slot         |
  // +-------------------+
  //
  class Run {
   public:
    byte magic_num_;             // The magic number used for debugging.
    byte size_bracket_idx_;      // The index of the size bracket of this run.
    byte is_thread_local_;       // True if this run is used as a thread-local run.
    byte to_be_bulk_freed_;      // Used within BulkFree() to flag a run that's involved with a bulk free.
    uint32_t top_slot_idx_;      // The top slot index when this run is in bump index mode.
    uint32_t alloc_bit_map_[0];  // The bit map that allocates if each slot is in use.

    // bulk_free_bit_map_[] : The bit map that is used for GC to
    // temporarily mark the slots to free without using a lock. After
    // all the slots to be freed in a run are marked, all those slots
    // get freed in bulk with one locking per run, as opposed to one
    // locking per slot to minimize the lock contention. This is used
    // within BulkFree().

    // thread_local_free_bit_map_[] : The bit map that is used for GC
    // to temporarily mark the slots to free in a thread-local run
    // without using a lock (without synchronizing the thread that
    // owns the thread-local run.) When the thread-local run becomes
    // full, the thread will check this bit map and update the
    // allocation bit map of the run (that is, the slots get freed.)

    // Returns the byte size of the header except for the bit maps.
    static size_t fixed_header_size() {
      Run temp;
      size_t size = reinterpret_cast<byte*>(&temp.alloc_bit_map_) - reinterpret_cast<byte*>(&temp);
      DCHECK_EQ(size, static_cast<size_t>(8));
      return size;
    }
    // Returns the base address of the free bit map.
    uint32_t* bulk_free_bit_map() {
      return reinterpret_cast<uint32_t*>(reinterpret_cast<byte*>(this) + bulkFreeBitMapOffsets[size_bracket_idx_]);
    }
    // Returns the base address of the thread local free bit map.
    uint32_t* thread_local_free_bit_map() {
      return reinterpret_cast<uint32_t*>(reinterpret_cast<byte*>(this) + threadLocalFreeBitMapOffsets[size_bracket_idx_]);
    }
    void* End() {
      return reinterpret_cast<byte*>(this) + kPageSize * numOfPages[size_bracket_idx_];
    }
    // Frees slots in the allocation bit map with regard to the
    // thread-local free bit map. Used when a thread-local run becomes
    // full.
    bool MergeThreadLocalFreeBitMapToAllocBitMap(bool* is_all_free_after_out);
    // Frees slots in the allocation bit map with regard to the bulk
    // free bit map. Used in a bulk free.
    void MergeBulkFreeBitMapIntoAllocBitMap();
    // Unions the slots to be freed in the free bit map into the
    // thread-local free bit map. In a bulk free, as a two-step
    // process, GC will first record all the slots to free in a run in
    // the free bit map where it can write without a lock, and later
    // acquire a lock once per run to union the bits of the free bit
    // map to the thread-local free bit map.
    void UnionBulkFreeBitMapToThreadLocalFreeBitMap();
    // Allocates a slot in a run.
    void* AllocSlot();
    // Frees a slot in a run. This is used in a non-bulk free.
    void FreeSlot(void* ptr);
    // Marks the slots to free in the bulk free bit map.
    void MarkBulkFreeBitMap(void* ptr);
    // Marks the slots to free in the thread-local free bit map.
    void MarkThreadLocalFreeBitMap(void* ptr);
    // Returns true if all the slots in the run are not in use.
    bool IsAllFree();
    // Returns true if all the slots in the run are in use.
    bool IsFull();
    // Clear all the bit maps.
    void ClearBitMaps();
    // Iterate over all the slots and apply the given function.
    void InspectAllSlots(void (*handler)(void* start, void* end, size_t used_bytes, void* callback_arg), void* arg);
    // Dump the run metadata for debugging.
    void Dump();

   private:
    // The common part of MarkFreeBitMap() and MarkThreadLocalFreeBitMap().
    void MarkFreeBitMapShared(void* ptr, uint32_t* free_bit_map_base, const char* caller_name);
  };

  // The magic number for a run.
  static const byte kMagicNum = 42;
  // The magic number for free pages.
  static const byte kMagicNumFree = 43;
  // The number of size brackets. Sync this with the length of Thread::rosalloc_runs_.
  static const size_t kNumOfSizeBrackets = 34;
  // The number of smaller size brackets that are 16 bytes apart.
  static const size_t kNumOfQuantumSizeBrackets = 32;
  // The sizes (the slot sizes, in bytes) of the size brackets.
  static size_t bracketSizes[kNumOfSizeBrackets];
  // The numbers of pages that are used for runs for each size bracket.
  static size_t numOfPages[kNumOfSizeBrackets];
  // The numbers of slots of the runs for each size bracket.
  static size_t numOfSlots[kNumOfSizeBrackets];
  // The header sizes in bytes of the runs for each size bracket.
  static size_t headerSizes[kNumOfSizeBrackets];
  // The byte offsets of the bulk free bit maps of the runs for each size bracket.
  static size_t bulkFreeBitMapOffsets[kNumOfSizeBrackets];
  // The byte offsets of the thread-local free bit maps of the runs for each size bracket.
  static size_t threadLocalFreeBitMapOffsets[kNumOfSizeBrackets];

  // Initialize the run specs (the above arrays).
  static void Initialize();
  static bool initialized_;

  // Returns the byte size of the bracket size from the index.
  static size_t IndexToBracketSize(size_t idx) {
    DCHECK(idx < kNumOfSizeBrackets);
    return bracketSizes[idx];
  }
  // Returns the index of the size bracket from the bracket size.
  static size_t BracketSizeToIndex(size_t size) {
    DCHECK(16 <= size && ((size < 1 * KB && size % 16 == 0) || size == 1 * KB || size == 2 * KB));
    size_t idx;
    if (UNLIKELY(size == 1 * KB)) {
      idx = kNumOfSizeBrackets - 2;
    } else if (UNLIKELY(size == 2 * KB)) {
      idx = kNumOfSizeBrackets - 1;
    } else {
      DCHECK(size < 1 * KB);
      DCHECK_EQ(size % 16, static_cast<size_t>(0));
      idx = size / 16 - 1;
    }
    DCHECK(bracketSizes[idx] == size);
    return idx;
  }
  // Rounds up the size up the nearest bracket size.
  static size_t RoundToBracketSize(size_t size) {
    DCHECK(size <= kLargeSizeThreshold);
    if (LIKELY(size <= 512)) {
      return RoundUp(size, 16);
    } else if (512 < size && size <= 1 * KB) {
      return 1 * KB;
    } else {
      DCHECK(1 * KB < size && size <= 2 * KB);
      return 2 * KB;
    }
  }
  // Returns the size bracket index from the byte size with rounding.
  static size_t SizeToIndex(size_t size) {
    DCHECK(size <= kLargeSizeThreshold);
    if (LIKELY(size <= 512)) {
      return RoundUp(size, 16) / 16 - 1;
    } else if (512 < size && size <= 1 * KB) {
      return kNumOfSizeBrackets - 2;
    } else {
      DCHECK(1 * KB < size && size <= 2 * KB);
      return kNumOfSizeBrackets - 1;
    }
  }
  // A combination of SizeToIndex() and RoundToBracketSize().
  static size_t SizeToIndexAndBracketSize(size_t size, size_t* bracket_size_out) {
    DCHECK(size <= kLargeSizeThreshold);
    if (LIKELY(size <= 512)) {
      size_t bracket_size = RoundUp(size, 16);
      *bracket_size_out = bracket_size;
      size_t idx = bracket_size / 16 - 1;
      DCHECK_EQ(bracket_size, IndexToBracketSize(idx));
      return idx;
    } else if (512 < size && size <= 1 * KB) {
      size_t bracket_size = 1024;
      *bracket_size_out = bracket_size;
      size_t idx = kNumOfSizeBrackets - 2;
      DCHECK_EQ(bracket_size, IndexToBracketSize(idx));
      return idx;
    } else {
      DCHECK(1 * KB < size && size <= 2 * KB);
      size_t bracket_size = 2048;
      *bracket_size_out = bracket_size;
      size_t idx = kNumOfSizeBrackets - 1;
      DCHECK_EQ(bracket_size, IndexToBracketSize(idx));
      return idx;
    }
  }
  // Returns the page map index from an address. Requires that the
  // address is page size aligned.
  size_t ToPageMapIndex(const void* addr) const {
    DCHECK(base_ <= addr && addr < base_ + capacity_);
    size_t byte_offset = reinterpret_cast<const byte*>(addr) - base_;
    DCHECK_EQ(byte_offset % static_cast<size_t>(kPageSize), static_cast<size_t>(0));
    return byte_offset / kPageSize;
  }
  // Returns the page map index from an address with rounding.
  size_t RoundDownToPageMapIndex(void* addr) {
    DCHECK(base_ <= addr && addr < reinterpret_cast<byte*>(base_) + capacity_);
    return (reinterpret_cast<uintptr_t>(addr) - reinterpret_cast<uintptr_t>(base_)) / kPageSize;
  }

  // A memory allocation request larger than this size is treated as a large object and allocated
  // at a page-granularity.
  static const size_t kLargeSizeThreshold = 2048;

  // We use use thread-local runs for the size Brackets whose indexes
  // are less than or equal to this index. We use shared (current)
  // runs for the rest.
  static const size_t kMaxThreadLocalSizeBracketIdx = 10;

  // If true, check that the returned memory is actually zero.
  static constexpr bool kCheckZeroMemory = kIsDebugBuild;

  // If true, log verbose details of operations.
  static constexpr bool kTraceRosAlloc = false;

  struct hash_run {
    size_t operator()(const RosAlloc::Run* r) const {
      return reinterpret_cast<size_t>(r);
    }
  };

  struct eq_run {
    bool operator()(const RosAlloc::Run* r1, const RosAlloc::Run* r2) const {
      return r1 == r2;
    }
  };

 public:
  // Different page release modes.
  enum PageReleaseMode {
    kPageReleaseModeNone,         // Release no empty pages.
    kPageReleaseModeEnd,          // Release empty pages at the end of the space.
    kPageReleaseModeSize,         // Release empty pages that are larger than the threshold.
    kPageReleaseModeSizeAndEnd,   // Release empty pages that are larger than the threshold or
                                  // at the end of the space.
    kPageReleaseModeAll,          // Release all empty pages.
  };

  // The default value for page_release_size_threshold_.
  static constexpr size_t kDefaultPageReleaseSizeThreshold = 4 * MB;

 private:
  // The base address of the memory region that's managed by this allocator.
  byte* base_;

  // The footprint in bytes of the currently allocated portion of the
  // memory region.
  size_t footprint_;

  // The maximum footprint. The address, base_ + capacity_, indicates
  // the end of the memory region that's managed by this allocator.
  size_t capacity_;

  // The run sets that hold the runs whose slots are not all
  // full. non_full_runs_[i] is guarded by size_bracket_locks_[i].
  std::set<Run*> non_full_runs_[kNumOfSizeBrackets];
  // The run sets that hold the runs whose slots are all full. This is
  // debug only. full_runs_[i] is guarded by size_bracket_locks_[i].
  hash_set<Run*, hash_run, eq_run> full_runs_[kNumOfSizeBrackets];
  // The set of free pages.
  std::set<FreePageRun*> free_page_runs_ GUARDED_BY(lock_);
  // The free page run whose end address is the end of the memory
  // region that's managed by this allocator, if any.
  FreePageRun* last_free_page_run_;
  // The current runs where the allocations are first attempted for
  // the size brackes that do not use thread-local
  // runs. current_runs_[i] is guarded by size_bracket_locks_[i].
  Run* current_runs_[kNumOfSizeBrackets];
  // The mutexes, one per size bracket.
  Mutex* size_bracket_locks_[kNumOfSizeBrackets];
  // The types of page map entries.
  enum {
    kPageMapEmpty           = 0,  // Not allocated.
    kPageMapRun             = 1,  // The beginning of a run.
    kPageMapRunPart         = 2,  // The non-beginning part of a run.
    kPageMapLargeObject     = 3,  // The beginning of a large object.
    kPageMapLargeObjectPart = 4,  // The non-beginning part of a large object.
  };
  // The table that indicates what pages are currently used for.
  std::vector<byte> page_map_ GUARDED_BY(lock_);
  // The table that indicates the size of free page runs. These sizes
  // are stored here to avoid storing in the free page header and
  // release backing pages.
  std::vector<size_t> free_page_run_size_map_ GUARDED_BY(lock_);
  // The global lock. Used to guard the page map, the free page set,
  // and the footprint.
  Mutex lock_ DEFAULT_MUTEX_ACQUIRED_AFTER;
  // The reader-writer lock to allow one bulk free at a time while
  // allowing multiple individual frees at the same time.
  ReaderWriterMutex bulk_free_lock_ DEFAULT_MUTEX_ACQUIRED_AFTER;

  // The page release mode.
  const PageReleaseMode page_release_mode_;
  // Under kPageReleaseModeSize(AndEnd), if the free page run size is
  // greater than or equal to this value, release pages.
  const size_t page_release_size_threshold_;

  // The base address of the memory region that's managed by this allocator.
  byte* Begin() { return base_; }
  // The end address of the memory region that's managed by this allocator.
  byte* End() { return base_ + capacity_; }

  // Page-granularity alloc/free
  void* AllocPages(Thread* self, size_t num_pages, byte page_map_type)
      EXCLUSIVE_LOCKS_REQUIRED(lock_);
  void FreePages(Thread* self, void* ptr) EXCLUSIVE_LOCKS_REQUIRED(lock_);

  // Allocate/free a run slot.
  void* AllocFromRun(Thread* self, size_t size, size_t* bytes_allocated)
      LOCKS_EXCLUDED(lock_);
  void FreeFromRun(Thread* self, void* ptr, Run* run)
      LOCKS_EXCLUDED(lock_);

  // Used to acquire a new/reused run for a size bracket. Used when a
  // thread-local or current run gets full.
  Run* RefillRun(Thread* self, size_t idx) LOCKS_EXCLUDED(lock_);

  // The internal of non-bulk Free().
  void FreeInternal(Thread* self, void* ptr) LOCKS_EXCLUDED(lock_);

  // Allocates large objects.
  void* AllocLargeObject(Thread* self, size_t size, size_t* bytes_allocated) LOCKS_EXCLUDED(lock_);

 public:
  RosAlloc(void* base, size_t capacity,
           PageReleaseMode page_release_mode,
           size_t page_release_size_threshold = kDefaultPageReleaseSizeThreshold);
  void* Alloc(Thread* self, size_t size, size_t* bytes_allocated)
      LOCKS_EXCLUDED(lock_);
  void Free(Thread* self, void* ptr)
      LOCKS_EXCLUDED(bulk_free_lock_);
  void BulkFree(Thread* self, void** ptrs, size_t num_ptrs)
      LOCKS_EXCLUDED(bulk_free_lock_);
  // Returns the size of the allocated slot for a given allocated memory chunk.
  size_t UsableSize(void* ptr);
  // Returns the size of the allocated slot for a given size.
  size_t UsableSize(size_t bytes) {
    if (UNLIKELY(bytes > kLargeSizeThreshold)) {
      return RoundUp(bytes, kPageSize);
    } else {
      return RoundToBracketSize(bytes);
    }
  }
  // Try to reduce the current footprint by releasing the free page
  // run at the end of the memory region, if any.
  bool Trim();
  // Iterates over all the memory slots and apply the given function.
  void InspectAll(void (*handler)(void* start, void* end, size_t used_bytes, void* callback_arg),
                  void* arg)
      LOCKS_EXCLUDED(lock_);
  // Returns the current footprint.
  size_t Footprint() LOCKS_EXCLUDED(lock_);
  // Returns the current capacity, maximum footprint.
  size_t FootprintLimit() LOCKS_EXCLUDED(lock_);
  // Update the current capacity.
  void SetFootprintLimit(size_t bytes) LOCKS_EXCLUDED(lock_);
  // Releases the thread-local runs assigned to the given thread back to the common set of runs.
  void RevokeThreadLocalRuns(Thread* thread);
  // Releases the thread-local runs assigned to all the threads back to the common set of runs.
  void RevokeAllThreadLocalRuns() LOCKS_EXCLUDED(Locks::thread_list_lock_);
  // Dumps the page map for debugging.
  void DumpPageMap(Thread* self);

  // Callbacks for InspectAll that will count the number of bytes
  // allocated and objects allocated, respectively.
  static void BytesAllocatedCallback(void* start, void* end, size_t used_bytes, void* arg);
  static void ObjectsAllocatedCallback(void* start, void* end, size_t used_bytes, void* arg);

  bool DoesReleaseAllPages() const {
    return page_release_mode_ == kPageReleaseModeAll;
  }
};

}  // namespace allocator
}  // namespace gc
}  // namespace art

#endif  // ART_RUNTIME_GC_ALLOCATOR_ROSALLOC_H_