Files
kotlin-fork/runtime/src/main/cpp/Memory.cpp
T
2018-03-28 17:33:09 +03:00

1526 lines
47 KiB
C++

/*
* Copyright 2010-2017 JetBrains s.r.o.
*
* 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.
*/
#include <string.h>
#include <stdio.h>
#include <cstddef> // for offsetof
#include "Alloc.h"
#include "Assert.h"
#include "Exceptions.h"
#include "Memory.h"
#include "MemoryPrivate.hpp"
#include "Natives.h"
#include "Porting.h"
// If garbage collection algorithm for cyclic garbage to be used.
// We are using the Bacon's algorithm for GC, see
// http://researcher.watson.ibm.com/researcher/files/us-bacon/Bacon03Pure.pdf.
#define USE_GC 1
// Define to 1 to print all memory operations.
#define TRACE_MEMORY 0
// Collect memory manager events statistics.
#define COLLECT_STATISTIC 0
// Auto-adjust GC thresholds.
#define GC_ERGONOMICS 1
// TODO: ensure it it read-only.
ContainerHeader ObjHeader::theStaticObjectsContainer = {
CONTAINER_TAG_PERMANENT | CONTAINER_TAG_INCREMENT
};
namespace {
// Granularity of arena container chunks.
constexpr container_size_t kContainerAlignment = 1024;
// Single object alignment.
constexpr container_size_t kObjectAlignment = 8;
#if TRACE_MEMORY
#define MEMORY_LOG(...) konan::consolePrintf(__VA_ARGS__);
#else
#define MEMORY_LOG(...)
#endif
inline int atomicAdd(int* where, int what) {
#ifndef KONAN_NO_THREADS
return __sync_add_and_fetch(where, what);
#else
return *where += what;
#endif
}
#if USE_GC
// Collection threshold default (collect after having so many elements in the
// release candidates set).
constexpr size_t kGcThreshold = 4 * 1024;
#if GC_ERGONOMICS
// Ergonomic thresholds.
// If GC to computations time ratio is above that value,
// increase GC threshold by 1.5 times.
constexpr double kGcToComputeRatioThreshold = 0.5;
// Never exceed this value when increasing GC threshold.
constexpr size_t kMaxErgonomicThreshold = 1024 * 1024;
#endif // GC_ERGONOMICS
typedef KStdDeque<ContainerHeader*> ContainerHeaderDeque;
#endif
} // namespace
#if TRACE_MEMORY || USE_GC
typedef KStdUnorderedSet<ContainerHeader*> ContainerHeaderSet;
typedef KStdVector<ContainerHeader*> ContainerHeaderList;
typedef KStdVector<KRef*> KRefPtrList;
#endif
struct FrameOverlay {
ArenaContainer* arena;
};
// A little hack that allows to enable -O2 optimizations
// Prevents clang from replacing FrameOverlay struct
// with single pointer.
// Can be removed when FrameOverlay will become more complex
FrameOverlay exportFrameOverlay;
// Current number of allocated containers.
int allocCount = 0;
int aliveMemoryStatesCount = 0;
// Forward declarations.
void FreeContainer(ContainerHeader* header);
#if COLLECT_STATISTIC
class MemoryStatistic {
public:
// UpdateRef per-object type counters.
uint64_t updateCounters[4][4];
// Alloc per container type counters.
uint64_t containerAllocs[4][2];
// Free per container type counters.
uint64_t objectAllocs[4][2];
// Histogram of allocation size distribution.
KStdUnorderedMap<int, int>* allocationHistogram;
// Number of allocation cache hits.
int allocCacheHit;
// Number of allocation cache misses.
int allocCacheMiss;
// Map of array index to human readable name.
static constexpr const char* indexToName[] = { "normal", "stack ", "perm ", "null " };
void init() {
memset(containerAllocs, 0, sizeof(containerAllocs));
memset(objectAllocs, 0, sizeof(objectAllocs));
memset(updateCounters, 0, sizeof(updateCounters));
allocationHistogram = konanConstructInstance<KStdUnorderedMap<int, int>>();
allocCacheHit = 0;
allocCacheMiss = 0;
}
void deinit() {
konanDestructInstance(allocationHistogram);
allocationHistogram = nullptr;
}
void incUpdateRef(const ObjHeader* objOld, const ObjHeader* objNew) {
updateCounters[toIndex(objOld)][toIndex(objNew)]++;
}
void incAlloc(size_t size, const ContainerHeader* header) {
containerAllocs[toIndex(header)][0]++;
++(*allocationHistogram)[size];
#if 0
auto queue = memoryState->finalizerQueue;
bool hit = false;
for (int i = 0; i < queue->size(); i++) {
auto container = (*queue)[i];
if (containerSize(container) == size) {
hit = true;
break;
}
}
if (hit)
allocCacheHit++;
else
allocCacheMiss++;
#endif // USE_GC
}
void incFree(const ContainerHeader* header) {
containerAllocs[toIndex(header)][1]++;
}
void incAlloc(size_t size, const ObjHeader* header) {
objectAllocs[toIndex(header)][0]++;
}
void incFree(const ObjHeader* header) {
objectAllocs[toIndex(header)][1]++;
}
static int toIndex(const ObjHeader* obj) {
if (obj == nullptr) return 3;
return toIndex(obj->container());
}
static int toIndex(const ContainerHeader* header) {
switch (header->tag()) {
case CONTAINER_TAG_NORMAL : return 0;
case CONTAINER_TAG_STACK : return 1;
case CONTAINER_TAG_PERMANENT: return 2;
}
RuntimeAssert(false, "unknown container type");
return -1;
}
void printStatistic() {
konan::consolePrintf("\nMemory manager statistic:\n\n");
for (int i = 0; i < 2; i++) {
konan::consolePrintf("Container %s alloc: %lld, free: %lld\n",
indexToName[i], containerAllocs[i][0],
containerAllocs[i][1]);
}
for (int i = 0; i < 2; i++) {
konan::consolePrintf("Object %s alloc: %lld, free: %lld\n",
indexToName[i], objectAllocs[i][0],
objectAllocs[i][1]);
}
konan::consolePrintf("\n");
for (int i = 0; i < 4; i++) {
for (int j = 0; j < 4; j++) {
konan::consolePrintf("UpdateRef[%s -> %s]: %lld\n",
indexToName[i], indexToName[j], updateCounters[i][j]);
}
}
konan::consolePrintf("\n");
konan::consolePrintf("Allocation histogram:\n");
KStdVector<int> keys(allocationHistogram->size());
int index = 0;
for (auto& it : *allocationHistogram) {
keys[index++] = it.first;
}
std::sort(keys.begin(), keys.end());
for (auto& it : keys) {
konan::consolePrintf(
"%d bytes -> %d times\n", it, (*allocationHistogram)[it]);
}
#if USE_GC
konan::consolePrintf(
"alloc cache: %d hits/%d misses\n", allocCacheHit, allocCacheMiss);
#endif // USE_GC
}
};
constexpr const char* MemoryStatistic::indexToName[];
#endif // COLLECT_STATISTIC
struct MemoryState {
#if TRACE_MEMORY
// Set of all containers.
ContainerHeaderSet* containers;
#endif
#if USE_GC
// Finalizer queue.
ContainerHeaderDeque* finalizerQueue;
/*
* Typical scenario for GC is as following:
* we have 90% of objects with refcount = 0 which will be deleted during
* the first phase of the algorithm.
* We could mark them with a bit in order to tell the next two phases to skip them
* and thus requiring only one list, but the downside is that both of the
* next phases would iterate over the whole list of objects instead of only 10%.
*/
ContainerHeaderList* toFree; // List of all cycle candidates.
ContainerHeaderList* roots; // Real candidates excluding those with refcount = 0.
// How many GC suspend requests happened.
int gcSuspendCount;
// How many candidate elements in toFree shall trigger collection.
size_t gcThreshold;
// If collection is in progress.
bool gcInProgress;
#if GC_ERGONOMICS
uint64_t lastGcTimestamp;
#endif
#endif // USE_GC
#if COLLECT_STATISTIC
#define CONTAINER_ALLOC_STAT(state, size, container) state->statistic.incAlloc(size, container);
#define CONTAINER_FREE_STAT(state, container)
#define CONTAINER_DESTROY_STAT(state, container) \
state->statistic.incFree(container);
#define OBJECT_ALLOC_STAT(state, size, object) \
state->statistic.incAlloc(size, object);
#define OBJECT_FREE_STAT(state, size, object) \
state->statistic.incFree(object);
#define UPDATE_REF_STAT(state, oldRef, newRef, slot) \
state->statistic.incUpdateRef(oldRef, newRef);
#define INIT_STAT(state) \
state->statistic.init();
#define DEINIT_STAT(state) \
state->statistic.deinit();
#define PRINT_STAT(state) \
state->statistic.printStatistic();
MemoryStatistic statistic;
#else
#define CONTAINER_ALLOC_STAT(state, size, container)
#define CONTAINER_FREE_STAT(state, container)
#define CONTAINER_DESTROY_STAT(state, container)
#define OBJECT_ALLOC_STAT(state, size, object)
#define OBJECT_FREE_STAT(state, object)
#define UPDATE_REF_STAT(state, oldRef, newRef, slot)
#define INIT_STAT(state)
#define DEINIT_STAT(state)
#define PRINT_STAT(state)
#endif // COLLECT_STATISTIC
};
#if TRACE_MEMORY
#define INIT_TRACE(state) \
memoryState->containers = konanConstructInstance<ContainerHeaderSet>();
#define DEINIT_TRACE(state) \
konanDestructInstance(memoryState->containers); \
memoryState->containers = nullptr;
#else
#define INIT_TRACE(state)
#define DEINIT_TRACE(state)
#endif
#define CONTAINER_ALLOC_TRACE(state, size, container) \
MEMORY_LOG("Container alloc %d at %p\n", size, container)
#define CONTAINER_FREE_TRACE(state, container) \
MEMORY_LOG("Container free %p\n", container)
#define CONTAINER_DESTROY_TRACE(state, container) \
MEMORY_LOG("Container destroy %p\n", container)
#define OBJECT_ALLOC_TRACE(state, size, object) \
MEMORY_LOG("Object alloc %d at %p\n", size, object)
#define OBJECT_FREE_TRACE(state, object) \
MEMORY_LOG("Object free %p\n", object)
#define UPDATE_REF_TRACE(state, oldRef, newRef, slot) \
MEMORY_LOG("UpdateRef *%p: %p -> %p\n", slot, oldRef, newRef)
// Events macro definitions.
// Called on worker's memory init.
#define INIT_EVENT(state) \
INIT_STAT(state) \
INIT_TRACE(state)
// Called on worker's memory deinit.
#define DEINIT_EVENT(state) \
DEINIT_STAT(state)
// Called on container allocation.
#define CONTAINER_ALLOC_EVENT(state, size, container) \
CONTAINER_ALLOC_STAT(state, size, container) \
CONTAINER_ALLOC_TRACE(state, size, container)
// Called on container freeing (memory is still in use).
#define CONTAINER_FREE_EVENT(state, container) \
CONTAINER_FREE_STAT(state, container) \
CONTAINER_FREE_TRACE(state, container)
// Called on container destroy (memory is released to allocator).
#define CONTAINER_DESTROY_EVENT(state, container) \
CONTAINER_DESTROY_STAT(state, container) \
CONTAINER_DESTROY_TRACE(state, container)
// Object was just allocated.
#define OBJECT_ALLOC_EVENT(state, size, object) \
OBJECT_ALLOC_STAT(state, size, object) \
OBJECT_ALLOC_TRACE(state, size, object)
// Object is freed.
#define OBJECT_FREE_EVENT(state, size, object) \
OBJECT_FREE_STAT(state, size, object) \
OBJECT_FREE_TRACE(state, object)
// Reference in memory is being updated.
#define UPDATE_REF_EVENT(state, oldRef, newRef, slot) \
UPDATE_REF_STAT(state, oldRef, newRef, slot) \
UPDATE_REF_TRACE(state, oldRef, newRef, slot)
// Infomation shall be printed as worker is exiting.
#define PRINT_EVENT(state) \
PRINT_STAT(state)
namespace {
// TODO: can we pass this variable as an explicit argument?
THREAD_LOCAL_VARIABLE MemoryState* memoryState = nullptr;
constexpr int kFrameOverlaySlots = sizeof(FrameOverlay) / sizeof(ObjHeader**);
inline bool isFreeable(const ContainerHeader* header) {
return header->tag() < CONTAINER_TAG_PERMANENT;
}
inline bool isArena(const ContainerHeader* header) {
return header->stack();
}
inline container_size_t alignUp(container_size_t size, int alignment) {
return (size + alignment - 1) & ~(alignment - 1);
}
#if KONAN_OBJECTS_CAN_HAVE_RESERVED_TAIL
// Note: defined by a compiler-generated bitcode.
extern "C" const container_size_t kObjectReservedTailSize;
#else
constexpr container_size_t kObjectReservedTailSize = 0;
#endif
// TODO: shall we do padding for alignment?
inline container_size_t objectSize(const ObjHeader* obj) {
const TypeInfo* type_info = obj->type_info();
container_size_t size = kObjectReservedTailSize + (type_info->instanceSize_ < 0 ?
// An array.
ArrayDataSizeBytes(obj->array()) + sizeof(ArrayHeader)
:
type_info->instanceSize_ + sizeof(ObjHeader));
return alignUp(size, kObjectAlignment);
}
inline bool isArenaSlot(ObjHeader** slot) {
return (reinterpret_cast<uintptr_t>(slot) & ARENA_BIT) != 0;
}
inline ObjHeader** asArenaSlot(ObjHeader** slot) {
return reinterpret_cast<ObjHeader**>(
reinterpret_cast<uintptr_t>(slot) & ~ARENA_BIT);
}
inline FrameOverlay* asFrameOverlay(ObjHeader** slot) {
return reinterpret_cast<FrameOverlay*>(slot);
}
inline bool isRefCounted(KConstRef object) {
return isFreeable(object->container());
}
} // namespace
extern "C" {
void objc_release(void* ptr);
void Kotlin_ObjCExport_releaseReservedObjectTail(ObjHeader* obj);
RUNTIME_NORETURN void ThrowFreezingException();
RUNTIME_NORETURN void ThrowInvalidMutabilityException();
} // extern "C"
inline void runDeallocationHooks(ObjHeader* obj) {
#if KONAN_OBJC_INTEROP
if (obj->type_info() == theObjCPointerHolderTypeInfo) {
void* objcPtr = *reinterpret_cast<void**>(obj + 1); // TODO: use more reliable layout description
objc_release(objcPtr);
} else {
if (HasReservedObjectTail(obj)) {
Kotlin_ObjCExport_releaseReservedObjectTail(obj);
}
}
#endif
}
inline void runDeallocationHooks(ContainerHeader* container) {
ObjHeader* obj = reinterpret_cast<ObjHeader*>(container + 1);
for (int index = 0; index < container->objectCount(); index++) {
runDeallocationHooks(obj);
obj = reinterpret_cast<ObjHeader*>(
reinterpret_cast<uintptr_t>(obj) + objectSize(obj));
}
}
static inline void DeinitInstanceBodyImpl(const TypeInfo* typeInfo, void* body) {
for (int index = 0; index < typeInfo->objOffsetsCount_; index++) {
ObjHeader** location = reinterpret_cast<ObjHeader**>(
reinterpret_cast<uintptr_t>(body) + typeInfo->objOffsets_[index]);
UpdateRef(location, nullptr);
}
}
void DeinitInstanceBody(const TypeInfo* typeInfo, void* body) {
DeinitInstanceBodyImpl(typeInfo, body);
}
namespace {
template<typename func>
void traverseContainerObjectFields(ContainerHeader* container, func process) {
ObjHeader* obj = reinterpret_cast<ObjHeader*>(container + 1);
for (int object = 0; object < container->objectCount(); object++) {
const TypeInfo* typeInfo = obj->type_info();
for (int index = 0; index < typeInfo->objOffsetsCount_; index++) {
ObjHeader** location = reinterpret_cast<ObjHeader**>(
reinterpret_cast<uintptr_t>(obj + 1) + typeInfo->objOffsets_[index]);
process(location);
}
if (typeInfo == theArrayTypeInfo) {
ArrayHeader* array = obj->array();
for (int index = 0; index < array->count_; index++) {
process(ArrayAddressOfElementAt(array, index));
}
}
obj = reinterpret_cast<ObjHeader*>(
reinterpret_cast<uintptr_t>(obj) + objectSize(obj));
}
}
template<typename func>
void traverseContainerReferredObjects(ContainerHeader* container, func process) {
traverseContainerObjectFields(container, [process](ObjHeader** location) {
ObjHeader* ref = *location;
if (ref != nullptr) process(ref);
});
}
#if USE_GC
inline bool isMarkedAsRemoved(ContainerHeader* container) {
return (reinterpret_cast<uintptr_t>(container) & 1) != 0;
}
inline ContainerHeader* markAsRemoved(ContainerHeader* container) {
return reinterpret_cast<ContainerHeader*>(reinterpret_cast<uintptr_t>(container) | 1);
}
inline void processFinalizerQueue(MemoryState* state) {
// TODO: reuse elements of finalizer queue for new allocations.
while (!state->finalizerQueue->empty()) {
auto container = memoryState->finalizerQueue->back();
state->finalizerQueue->pop_back();
#if TRACE_MEMORY
state->containers->erase(container);
#endif
runDeallocationHooks(container);
CONTAINER_DESTROY_EVENT(state, container)
konanFreeMemory(container);
atomicAdd(&allocCount, -1);
}
}
#endif
inline void scheduleDestroyContainer(
MemoryState* state, ContainerHeader* container, bool clearExternalRefs) {
if (clearExternalRefs) {
traverseContainerObjectFields(container, [](ObjHeader** location) {
ObjHeader* ref = *location;
// Frozen object references do not participate in trial deletion, so shall be explicitly freed.
if (ref != nullptr && ref->container()->frozen())
UpdateRef(location, nullptr);
});
}
#if USE_GC
state->finalizerQueue->push_front(container);
// We cannot clean finalizer queue while in GC.
if (!state->gcInProgress && state->finalizerQueue->size() > 256) {
processFinalizerQueue(state);
}
#else
atomicAdd(&allocCount, -1);
CONTAINER_DESTROY_EVENT(state, header)
konanFreeMemory(header);
#endif
}
#if !USE_GC
template <bool Atomic>
inline void IncrementRC(ContainerHeader* container) {
container->incRefCount<Atomic>();
}
template <bool Atomic>
inline void DecrementRC(ContainerHeader* container, bool useCycleCollector) {
if (container->decRefCount<Atomic>() == 0) {
FreeContainer(container);
}
}
#else // USE_GC
inline uint32_t freeableSize(MemoryState* state) {
return state->toFree->size();
}
template <bool Atomic>
inline void IncrementRC(ContainerHeader* container) {
container->incRefCount<Atomic>();
container->setColor(CONTAINER_TAG_GC_BLACK);
}
template <bool Atomic>
inline void DecrementRC(ContainerHeader* container, bool useCycleCollector) {
if (container->decRefCount<Atomic>() == 0) {
FreeContainer(container);
} else if (!Atomic && useCycleCollector) { // Possible root.
// Do not use cycle collector for frozen objects, as we already detected possible cycles during
// freezing.
if (container->color() != CONTAINER_TAG_GC_PURPLE) {
container->setColor(CONTAINER_TAG_GC_PURPLE);
if (!container->buffered()) {
container->setBuffered();
auto state = memoryState;
state->toFree->push_back(container);
if (state->gcSuspendCount == 0 && freeableSize(state) >= state->gcThreshold) {
GarbageCollect();
}
}
}
}
}
inline void initThreshold(MemoryState* state, uint32_t gcThreshold) {
state->gcThreshold = gcThreshold;
state->toFree->reserve(gcThreshold);
}
#endif // USE_GC
#if TRACE_MEMORY || USE_GC
void dumpWorker(const char* prefix, ContainerHeader* header, ContainerHeaderSet* seen) {
MEMORY_LOG("%s: %p (%08x): %d refs\n", prefix, header, header->refCount_,
header->refCount_ >> CONTAINER_TAG_SHIFT)
seen->insert(header);
traverseContainerReferredObjects(header, [prefix, seen](ObjHeader* ref) {
auto child = ref->container();
RuntimeAssert(!isArena(child), "A reference to local object is encountered");
if (!child->permanent() && (seen->count(child) == 0)) {
dumpWorker(prefix, child, seen);
}
});
}
void dumpReachable(const char* prefix, const ContainerHeaderSet* roots) {
ContainerHeaderSet seen;
for (auto container : *roots) {
MEMORY_LOG("%p: %s%s%s\n", container,
container->frozen() ? "frozen " : "",
container->permanent() ? "permanent " : "",
container->stack() ? "stack " : "")
dumpWorker(prefix, container, &seen);
}
}
#endif
void MarkRoots(MemoryState*);
void DeleteCorpses(MemoryState*);
void ScanRoots(MemoryState*);
void CollectRoots(MemoryState*);
template<bool useColor>
void MarkGray(ContainerHeader* container) {
if (useColor) {
if (container->color() == CONTAINER_TAG_GC_GRAY) return;
} else {
if (container->marked()) return;
}
if (useColor) {
container->setColor(CONTAINER_TAG_GC_GRAY);
} else {
container->mark();
}
traverseContainerReferredObjects(container, [](ObjHeader* ref) {
auto childContainer = ref->container();
RuntimeAssert(!isArena(childContainer), "A reference to local object is encountered");
if (!childContainer->permanentOrFrozen()) {
childContainer->decRefCount<false>();
MarkGray<useColor>(childContainer);
}
});
}
void Scan(ContainerHeader* container);
template<bool useColor>
void ScanBlack(ContainerHeader* container) {
if (useColor) {
container->setColor(CONTAINER_TAG_GC_BLACK);
} else {
container->unMark();
}
traverseContainerReferredObjects(container, [](ObjHeader* ref) {
auto childContainer = ref->container();
RuntimeAssert(!isArena(childContainer), "A reference to local object is encountered");
if (!childContainer->permanentOrFrozen()) {
childContainer->incRefCount<false>();
if (useColor) {
if (childContainer->color() != CONTAINER_TAG_GC_BLACK)
ScanBlack<useColor>(childContainer);
} else {
if (childContainer->marked())
ScanBlack<useColor>(childContainer);
}
}
});
}
void CollectWhite(MemoryState*, ContainerHeader* container);
void CollectCycles(MemoryState* state) {
MarkRoots(state);
ScanRoots(state);
CollectRoots(state);
state->toFree->clear();
state->roots->clear();
}
void MarkRoots(MemoryState* state) {
for (auto container : *(state->toFree)) {
if (isMarkedAsRemoved(container))
continue;
auto color = container->color();
auto rcIsZero = container->refCount() == 0;
if (color == CONTAINER_TAG_GC_PURPLE && !rcIsZero) {
MarkGray<true>(container);
state->roots->push_back(container);
} else {
container->resetBuffered();
if (color == CONTAINER_TAG_GC_BLACK && rcIsZero) {
scheduleDestroyContainer(state, container, true);
}
}
}
}
void ScanRoots(MemoryState* state) {
for (auto container : *(state->roots)) {
Scan(container);
}
}
void CollectRoots(MemoryState* state) {
for (auto container : *(state->roots)) {
container->resetBuffered();
CollectWhite(state, container);
}
}
void Scan(ContainerHeader* container) {
if (container->color() != CONTAINER_TAG_GC_GRAY) return;
if (container->refCount() != 0) {
ScanBlack<true>(container);
return;
}
container->setColor(CONTAINER_TAG_GC_WHITE);
traverseContainerReferredObjects(container, [](ObjHeader* ref) {
auto childContainer = ref->container();
RuntimeAssert(!isArena(childContainer), "A reference to local object is encountered");
if (!childContainer->permanentOrFrozen()) {
Scan(childContainer);
}
});
}
void CollectWhite(MemoryState* state, ContainerHeader* container) {
if (container->color() != CONTAINER_TAG_GC_WHITE
|| container->buffered())
return;
container->setColor(CONTAINER_TAG_GC_BLACK);
traverseContainerReferredObjects(container, [state](ObjHeader* ref) {
auto childContainer = ref->container();
RuntimeAssert(!isArena(childContainer), "A reference to local object is encountered");
if (!childContainer->permanentOrFrozen()) {
CollectWhite(state, childContainer);
}
});
scheduleDestroyContainer(state, container, true);
}
inline void AddRef(ContainerHeader* header) {
// Looking at container type we may want to skip AddRef() totally
// (non-escaping stack objects, constant objects).
switch (header->refCount_ & CONTAINER_TAG_MASK) {
case CONTAINER_TAG_STACK:
case CONTAINER_TAG_PERMANENT:
break;
case CONTAINER_TAG_NORMAL:
IncrementRC<false>(header);
break;
case CONTAINER_TAG_FROZEN:
IncrementRC<true>(header);
break;
default:
RuntimeAssert(false, "unknown container type");
break;
}
}
inline void Release(ContainerHeader* header, bool useCycleCollector) {
// Looking at container type we may want to skip Release() totally
// (non-escaping stack objects, constant objects).
switch (header->refCount_ & CONTAINER_TAG_MASK) {
case CONTAINER_TAG_PERMANENT:
case CONTAINER_TAG_STACK:
break;
case CONTAINER_TAG_NORMAL:
DecrementRC<false>(header, useCycleCollector);
break;
case CONTAINER_TAG_FROZEN:
DecrementRC<true>(header, useCycleCollector);
break;
default:
RuntimeAssert(false, "unknown container type");
break;
}
}
// We use first slot as place to store frame-local arena container.
// TODO: create ArenaContainer object on the stack, so that we don't
// do two allocations per frame (ArenaContainer + actual container).
inline ArenaContainer* initedArena(ObjHeader** auxSlot) {
auto frame = asFrameOverlay(auxSlot);
auto arena = frame->arena;
if (!arena) {
arena = konanConstructInstance<ArenaContainer>();
MEMORY_LOG("Initializing arena in %p\n", frame)
arena->Init();
frame->arena = arena;
}
return arena;
}
inline size_t containerSize(const ContainerHeader* container) {
size_t result = 0;
const ObjHeader* obj = reinterpret_cast<const ObjHeader*>(container + 1);
for (int object = 0; object < container->objectCount(); object++) {
size_t size = objectSize(obj);
result += size;
obj = reinterpret_cast<ObjHeader*>(
reinterpret_cast<uintptr_t>(obj) + size);
}
return result;
}
} // namespace
ContainerHeader* AllocContainer(size_t size) {
auto state = memoryState;
#if USE_GC
// TODO: try to reuse elements of finalizer queue for new allocations, question
// is how to get actual size of container.
#endif
ContainerHeader* result = konanConstructSizedInstance<ContainerHeader>(alignUp(size, kObjectAlignment));
CONTAINER_ALLOC_EVENT(state, size, result);
#if TRACE_MEMORY
state->containers->insert(result);
#endif
atomicAdd(&allocCount, 1);
return result;
}
void FreeContainer(ContainerHeader* header) {
RuntimeAssert(!header->permanent(), "this kind of container shalln't be freed");
auto state = memoryState;
CONTAINER_FREE_EVENT(state, header)
// Now let's clean all object's fields in this container.
traverseContainerObjectFields(header, [](ObjHeader** location) {
UpdateRef(location, nullptr);
});
// And release underlying memory.
if (!isFreeable(header)) {
runDeallocationHooks(header);
} else {
header->setColor(CONTAINER_TAG_GC_BLACK);
if (!header->buffered())
scheduleDestroyContainer(state, header, false);
}
}
void ObjectContainer::Init(const TypeInfo* type_info) {
RuntimeAssert(type_info->instanceSize_ >= 0, "Must be an object");
uint32_t alloc_size =
sizeof(ContainerHeader) + sizeof(ObjHeader) + type_info->instanceSize_ + kObjectReservedTailSize;
header_ = AllocContainer(alloc_size);
if (header_) {
// One object in this container.
header_->setObjectCount(1);
// header->refCount_ is zero initialized by AllocContainer().
SetMeta(GetPlace(), type_info);
MEMORY_LOG("object at %p\n", GetPlace())
OBJECT_ALLOC_EVENT(memoryState, type_info->instanceSize_, GetPlace())
}
}
void ArrayContainer::Init(const TypeInfo* type_info, uint32_t elements) {
RuntimeAssert(type_info->instanceSize_ < 0, "Must be an array");
uint32_t alloc_size =
sizeof(ContainerHeader) + sizeof(ArrayHeader) -
type_info->instanceSize_ * elements + kObjectReservedTailSize;
header_ = AllocContainer(alloc_size);
RuntimeAssert(header_ != nullptr, "Cannot alloc memory");
if (header_) {
// One object in this container.
header_->setObjectCount(1);
// header->refCount_ is zero initialized by AllocContainer().
GetPlace()->count_ = elements;
SetMeta(GetPlace()->obj(), type_info);
MEMORY_LOG("array at %p\n", GetPlace())
OBJECT_ALLOC_EVENT(
memoryState, -type_info->instanceSize_ * elements, GetPlace()->obj())
}
}
// TODO: store arena containers in some reuseable data structure, similar to
// finalizer queue.
void ArenaContainer::Init() {
allocContainer(1024);
}
void ArenaContainer::Deinit() {
MEMORY_LOG("Arena::Deinit start: %p\n", this)
auto chunk = currentChunk_;
while (chunk != nullptr) {
// FreeContainer() doesn't release memory when CONTAINER_TAG_STACK is set.
MEMORY_LOG("Arena::Deinit free chunk %p\n", chunk)
FreeContainer(chunk->asHeader());
chunk = chunk->next;
}
chunk = currentChunk_;
while (chunk != nullptr) {
auto toRemove = chunk;
chunk = chunk->next;
konanFreeMemory(toRemove);
}
}
bool ArenaContainer::allocContainer(container_size_t minSize) {
auto size = minSize + sizeof(ContainerHeader) + sizeof(ContainerChunk);
size = alignUp(size, kContainerAlignment);
// TODO: keep simple cache of container chunks.
ContainerChunk* result = konanConstructSizedInstance<ContainerChunk>(size);
RuntimeAssert(result != nullptr, "Cannot alloc memory");
if (result == nullptr) return false;
result->next = currentChunk_;
result->arena = this;
result->asHeader()->refCount_ = (CONTAINER_TAG_STACK | CONTAINER_TAG_INCREMENT);
currentChunk_ = result;
current_ = reinterpret_cast<uint8_t*>(result->asHeader() + 1);
end_ = reinterpret_cast<uint8_t*>(result) + size;
return true;
}
void* ArenaContainer::place(container_size_t size) {
size = alignUp(size, kObjectAlignment);
// Fast path.
if (current_ + size < end_) {
void* result = current_;
current_ += size;
return result;
}
if (!allocContainer(size)) {
return nullptr;
}
void* result = current_;
current_ += size;
RuntimeAssert(current_ <= end_, "Must not overflow");
return result;
}
#define ARENA_SLOTS_CHUNK_SIZE 16
ObjHeader** ArenaContainer::getSlot() {
if (slots_ == nullptr || slotsCount_ >= ARENA_SLOTS_CHUNK_SIZE) {
slots_ = PlaceArray(theArrayTypeInfo, ARENA_SLOTS_CHUNK_SIZE);
slotsCount_ = 0;
}
return ArrayAddressOfElementAt(slots_, slotsCount_++);
}
ObjHeader* ArenaContainer::PlaceObject(const TypeInfo* type_info) {
RuntimeAssert(type_info->instanceSize_ >= 0, "must be an object");
uint32_t size = type_info->instanceSize_ + sizeof(ObjHeader) + kObjectReservedTailSize;
ObjHeader* result = reinterpret_cast<ObjHeader*>(place(size));
if (!result) {
return nullptr;
}
OBJECT_ALLOC_EVENT(memoryState, type_info->instanceSize_, result)
currentChunk_->asHeader()->incObjectCount();
setMeta(result, type_info);
return result;
}
ArrayHeader* ArenaContainer::PlaceArray(const TypeInfo* type_info, uint32_t count) {
RuntimeAssert(type_info->instanceSize_ < 0, "must be an array");
container_size_t size = sizeof(ArrayHeader) - type_info->instanceSize_ * count + kObjectReservedTailSize;
ArrayHeader* result = reinterpret_cast<ArrayHeader*>(place(size));
if (!result) {
return nullptr;
}
OBJECT_ALLOC_EVENT(memoryState, -type_info->instanceSize_ * count, result->obj())
currentChunk_->asHeader()->incObjectCount();
setMeta(result->obj(), type_info);
result->count_ = count;
return result;
}
inline void AddRef(const ObjHeader* object) {
MEMORY_LOG("AddRef on %p in %p\n", object, object->container())
AddRef(object->container());
}
inline void ReleaseRef(const ObjHeader* object) {
MEMORY_LOG("ReleaseRef on %p in %p\n", object, object->container())
// Use cycle collector only for objects having object fields, or if container is multiobject.
auto container = object->container();
Release(container, (object->type_info()->objOffsetsCount_ > 0) || (container->objectCount() > 1));
}
void AddRefFromAssociatedObject(const ObjHeader* object) {
AddRef(object);
}
void ReleaseRefFromAssociatedObject(const ObjHeader* object) {
ReleaseRef(object);
}
extern "C" {
MemoryState* InitMemory() {
RuntimeAssert(offsetof(ArrayHeader, type_info_)
==
offsetof(ObjHeader, type_info_),
"Layout mismatch");
RuntimeAssert(offsetof(ArrayHeader, container_offset_negative_)
==
offsetof(ObjHeader , container_offset_negative_),
"Layout mismatch");
RuntimeAssert(sizeof(FrameOverlay) % sizeof(ObjHeader**) == 0, "Frame overlay should contain only pointers")
RuntimeAssert(memoryState == nullptr, "memory state must be clear");
memoryState = konanConstructInstance<MemoryState>();
INIT_EVENT(memoryState)
#if USE_GC
memoryState->finalizerQueue = konanConstructInstance<ContainerHeaderDeque>();
memoryState->toFree = konanConstructInstance<ContainerHeaderList>();
memoryState->roots = konanConstructInstance<ContainerHeaderList>();
memoryState->gcInProgress = false;
initThreshold(memoryState, kGcThreshold);
memoryState->gcSuspendCount = 0;
#endif
atomicAdd(&aliveMemoryStatesCount, 1);
return memoryState;
}
void DeinitMemory(MemoryState* memoryState) {
#if USE_GC
GarbageCollect();
RuntimeAssert(memoryState->toFree->size() == 0, "Some memory have not been released after GC");
konanDestructInstance(memoryState->toFree);
konanDestructInstance(memoryState->roots);
konanDestructInstance(memoryState->finalizerQueue);
memoryState->finalizerQueue = nullptr;
#endif // USE_GC
bool lastMemoryState = atomicAdd(&aliveMemoryStatesCount, -1) == 0;
#if TRACE_MEMORY
if (lastMemoryState && allocCount > 0) {
MEMORY_LOG("*** Memory leaks, leaked %d containers ***\n", allocCount);
dumpReachable("", memoryState->containers);
}
#else
if (lastMemoryState)
RuntimeAssert(allocCount == 0, "Memory leaks found");
#endif
PRINT_EVENT(memoryState)
DEINIT_EVENT(memoryState)
konanFreeMemory(memoryState);
::memoryState = nullptr;
}
OBJ_GETTER(AllocInstance, const TypeInfo* type_info) {
RuntimeAssert(type_info->instanceSize_ >= 0, "must be an object");
if (isArenaSlot(OBJ_RESULT)) {
auto arena = initedArena(asArenaSlot(OBJ_RESULT));
auto result = arena->PlaceObject(type_info);
MEMORY_LOG("instance %p in arena: %p\n", result, arena)
return result;
}
RETURN_OBJ(ObjectContainer(type_info).GetPlace());
}
OBJ_GETTER(AllocArrayInstance, const TypeInfo* type_info, uint32_t elements) {
RuntimeAssert(type_info->instanceSize_ < 0, "must be an array");
if (isArenaSlot(OBJ_RESULT)) {
auto arena = initedArena(asArenaSlot(OBJ_RESULT));
auto result = arena->PlaceArray(type_info, elements)->obj();
MEMORY_LOG("array[%d] %p in arena: %p\n", elements, result, arena)
return result;
}
RETURN_OBJ(ArrayContainer(type_info, elements).GetPlace()->obj());
}
OBJ_GETTER(InitInstance,
ObjHeader** location, const TypeInfo* type_info, void (*ctor)(ObjHeader*)) {
ObjHeader* value = *location;
if (value != nullptr) {
// OK'ish, inited by someone else.
RETURN_OBJ(value);
}
ObjHeader* object = AllocInstance(type_info, OBJ_RESULT);
MEMORY_LOG("Calling UpdateRef from InitInstance\n")
UpdateRef(location, object);
#if KONAN_NO_EXCEPTIONS
ctor(object);
return object;
#else
try {
ctor(object);
return object;
} catch (...) {
UpdateRef(OBJ_RESULT, nullptr);
UpdateRef(location, nullptr);
throw;
}
#endif
}
OBJ_GETTER(InitSharedInstance,
ObjHeader** location, const TypeInfo* type_info, void (*ctor)(ObjHeader*)) {
#if KONAN_NO_THREADS
return InitInstance(location, type_info, ctor);
#else
ObjHeader* initializing = reinterpret_cast<ObjHeader*>(1);
ObjHeader* value;
// Spin lock.
while ((value = __sync_val_compare_and_swap(location, nullptr, initializing)) == initializing);
if (value != nullptr) {
// OK'ish, inited by someone else.
RETURN_OBJ(value);
}
ObjHeader* object = AllocInstance(type_info, OBJ_RESULT);
MEMORY_LOG("Calling UpdateRef from InitInstance\n")
UpdateRef(location, object);
__sync_synchronize();
#if KONAN_NO_EXCEPTIONS
ctor(object);
// TODO: uncomment as soon as cycles are correctly handled during freezing.
//if (!object->container()->frozen())
//ThrowFreezingException();
return object;
#else
try {
ctor(object);
//if (!object->container()->frozen())
//ThrowFreezingException();
return object;
} catch (...) {
UpdateRef(OBJ_RESULT, nullptr);
UpdateRef(location, nullptr);
__sync_synchronize();
throw;
}
#endif
#endif
}
bool HasReservedObjectTail(ObjHeader* obj) {
return kObjectReservedTailSize != 0 && !obj->permanent();
}
void* GetReservedObjectTail(ObjHeader* obj) {
return reinterpret_cast<void*>(
reinterpret_cast<uintptr_t>(obj) + objectSize(obj) - kObjectReservedTailSize
);
}
void SetRef(ObjHeader** location, const ObjHeader* object) {
MEMORY_LOG("SetRef *%p: %p\n", location, object)
*const_cast<const ObjHeader**>(location) = object;
AddRef(object);
}
ObjHeader** GetReturnSlotIfArena(ObjHeader** returnSlot, ObjHeader** localSlot) {
return isArenaSlot(returnSlot) ? returnSlot : localSlot;
}
ObjHeader** GetParamSlotIfArena(ObjHeader* param, ObjHeader** localSlot) {
if (param == nullptr) return localSlot;
auto container = param->container();
if ((container->refCount_ & CONTAINER_TAG_MASK) != CONTAINER_TAG_STACK)
return localSlot;
auto chunk = reinterpret_cast<ContainerChunk*>(container) - 1;
return reinterpret_cast<ObjHeader**>(reinterpret_cast<uintptr_t>(&chunk->arena) | ARENA_BIT);
}
void UpdateReturnRef(ObjHeader** returnSlot, const ObjHeader* object) {
if (isArenaSlot(returnSlot)) {
// Not a subject of reference counting.
if (object == nullptr || !isRefCounted(object)) return;
auto arena = initedArena(asArenaSlot(returnSlot));
returnSlot = arena->getSlot();
}
UpdateRef(returnSlot, object);
}
void UpdateRef(ObjHeader** location, const ObjHeader* object) {
RuntimeAssert(!isArenaSlot(location), "must not be a slot");
ObjHeader* old = *location;
UPDATE_REF_EVENT(memoryState, old, object, location)
if (old != object) {
if (object != nullptr) {
AddRef(object);
}
*const_cast<const ObjHeader**>(location) = object;
if (reinterpret_cast<uintptr_t>(old) > 1) {
ReleaseRef(old);
}
}
}
void EnterFrame(ObjHeader** start, int parameters, int count) {
MEMORY_LOG("EnterFrame %p .. %p\n", start, start + count + parameters)
}
void LeaveFrame(ObjHeader** start, int parameters, int count) {
MEMORY_LOG("LeaveFrame %p .. %p\n", start, start + count + parameters)
ReleaseRefs(start + parameters + kFrameOverlaySlots, count - kFrameOverlaySlots - parameters);
if (*start != nullptr) {
auto arena = initedArena(start);
MEMORY_LOG("LeaveFrame: free arena %p\n", arena)
arena->Deinit();
konanFreeMemory(arena);
MEMORY_LOG("LeaveFrame: free arena done %p\n", arena)
}
}
void ReleaseRefs(ObjHeader** start, int count) {
MEMORY_LOG("ReleaseRefs %p .. %p\n", start, start + count)
ObjHeader** current = start;
auto state = memoryState;
while (count-- > 0) {
ObjHeader* object = *current;
if (object != nullptr) {
ReleaseRef(object);
// Just for sanity, optional.
*current = nullptr;
}
current++;
}
}
#if USE_GC
void GarbageCollect() {
MemoryState* state = memoryState;
RuntimeAssert(!state->gcInProgress, "Recursive GC is disallowed");
MEMORY_LOG("Garbage collect\n")
#if GC_ERGONOMICS
auto gcStartTime = konan::getTimeMicros();
#endif
state->gcInProgress = true;
while (state->toFree->size() > 0) {
CollectCycles(state);
processFinalizerQueue(state);
}
state->gcInProgress = false;
#if GC_ERGONOMICS
auto gcEndTime = konan::getTimeMicros();
auto gcToComputeRatio = double(gcEndTime - gcStartTime) / (gcStartTime - state->lastGcTimestamp + 1);
if (gcToComputeRatio > kGcToComputeRatioThreshold) {
auto newThreshold = state->gcThreshold * 3 / 2 + 1;
if (newThreshold < kMaxErgonomicThreshold) {
MEMORY_LOG("Adjusting GC threshold to %d\n", newThreshold);
initThreshold(state, newThreshold);
}
}
MEMORY_LOG("Garbage collect: GC length=%lld sinceLast=%lld\n",
(gcEndTime - gcStartTime), gcStartTime - state->lastGcTimestamp);
state->lastGcTimestamp = gcEndTime;
#endif
}
#endif // USE_GC
void Kotlin_konan_internal_GC_collect(KRef) {
#if USE_GC
GarbageCollect();
#endif
}
void Kotlin_konan_internal_GC_suspend(KRef) {
#if USE_GC
memoryState->gcSuspendCount++;
#endif
}
void Kotlin_konan_internal_GC_resume(KRef) {
#if USE_GC
MemoryState* state = memoryState;
if (state->gcSuspendCount > 0) {
state->gcSuspendCount--;
if (state->toFree != nullptr &&
freeableSize(state) >= state->gcThreshold) {
GarbageCollect();
}
}
#endif
}
void Kotlin_konan_internal_GC_stop(KRef) {
#if USE_GC
if (memoryState->toFree != nullptr) {
GarbageCollect();
konanDestructInstance(memoryState->toFree);
konanDestructInstance(memoryState->roots);
memoryState->toFree = nullptr;
memoryState->roots = nullptr;
}
#endif
}
void Kotlin_konan_internal_GC_start(KRef) {
#if USE_GC
if (memoryState->toFree == nullptr) {
memoryState->toFree = konanConstructInstance<ContainerHeaderList>();
memoryState->roots = konanConstructInstance<ContainerHeaderList>();
}
#endif
}
void Kotlin_konan_internal_GC_setThreshold(KRef, KInt value) {
#if USE_GC
if (value > 0) {
initThreshold(memoryState, value);
}
#endif
}
KInt Kotlin_konan_internal_GC_getThreshold(KRef) {
#if USE_GC
return memoryState->gcThreshold;
#else
return -1;
#endif
}
KNativePtr CreateStablePointer(KRef any) {
if (any == nullptr) return nullptr;
AddRef(any->container());
return reinterpret_cast<KNativePtr>(any);
}
void DisposeStablePointer(KNativePtr pointer) {
if (pointer == nullptr) return;
KRef ref = reinterpret_cast<KRef>(pointer);
ReleaseRef(ref);
}
OBJ_GETTER(DerefStablePointer, KNativePtr pointer) {
KRef ref = reinterpret_cast<KRef>(pointer);
RETURN_OBJ(ref);
}
OBJ_GETTER(AdoptStablePointer, KNativePtr pointer) {
#ifndef KONAN_NO_THREADS
__sync_synchronize();
#endif
KRef ref = reinterpret_cast<KRef>(pointer);
UpdateRef(OBJ_RESULT, nullptr);
// Somewhat hacky.
*OBJ_RESULT = ref;
return ref;
}
#if USE_GC
bool hasExternalRefs(ContainerHeader* container, ContainerHeaderSet* visited) {
visited->insert(container);
bool result = container->refCount() != 0;
traverseContainerReferredObjects(container, [&result, visited](ObjHeader* ref) {
auto child = ref->container();
if (!child->permanentOrFrozen() && (visited->find(child) == visited->end())) {
result |= hasExternalRefs(child, visited);
}
});
return result;
}
#endif
bool ClearSubgraphReferences(ObjHeader* root, bool checked) {
#if USE_GC
if (root != nullptr) {
auto state = memoryState;
auto container = root->container();
if (container->frozen())
// We assume, that frozen objects can be safely passed and are already removed
// GC candidate list.
return true;
ContainerHeaderSet visited;
if (!checked) {
hasExternalRefs(container, &visited);
} else {
if (!container->permanentOrFrozen()) {
container->decRefCount<false>();
MarkGray<false>(container);
auto bad = hasExternalRefs(container, &visited);
ScanBlack<false>(container);
container->incRefCount<false>();
if (bad) return false;
}
}
// TODO: not very effecient traversal.
for (auto it = state->toFree->begin(); it != state->toFree->end(); ++it) {
auto container = *it;
if (visited.find(container) != visited.end()) {
container->resetBuffered();
container->setColor(CONTAINER_TAG_GC_BLACK);
*it = markAsRemoved(container);
}
}
}
#endif // USE_GC
return true;
}
/**
* Do DFS cycle detection with three colors:
* - 'marked' bit as BLACK marker (object and its descendants processed)
* - 'seen' bit as GRAY marker (object is being processed)
* - not 'marked' and not 'seen' as WHITE marker (object is unprocessed)
* When we see GREY during DFS, it means we see cycle.
*/
void depthFirstTraversal(ContainerHeader* container, bool* hasCycles) {
// Mark GRAY.
container->setSeen();
traverseContainerObjectFields(container, [&hasCycles](ObjHeader** location) {
ObjHeader* obj = *location;
if (obj != nullptr) {
ContainerHeader* objContainer = obj->container();
if (!objContainer->permanentOrFrozen()) {
// Marked GREY, there's cycle.
if (objContainer->seen()) *hasCycles = true;
// Go deeper if WHITE.
if (!objContainer->seen() && !objContainer->marked()) {
depthFirstTraversal(objContainer, hasCycles);
}
}
}
});
// Mark BLACK.
container->resetSeen();
container->mark();
}
/**
* Theory of operations.
*
* Kotlin/Native supports object graph freezing, allowing to make certain subgraph immutable and thus
* suitable for safe sharing amongs multiple concurrent executors. This operation recursively operates
* on all objects reachable from the given object, and marks them as frozen. In frozen state object's
* fields cannot be modified, and so, lifetime of frozen objects correlates. Practically, it means
* that lifetimes of all strongly connected components are fully controlled by incoming reference
* counters, and so if we place all members of strongly connected component to the single container
* it could be correctly released by just atomic decrement on reference counter, without additional
* cycle collector run.
* So during subgraph freezing operation, we perform the following steps:
* - run Kosoraju-Sharir algorithm to find strongly connected components
* - put all objects in each strongly connected component into an artificial container
* (we assume that they all were in single element containers initially), single-object
* components remain in the same container
* - artifical container sums up outer reference counters of all its objects (i.e.
* incoming references from the same strongly connected component are not counted)
* - mark all object's headers as frozen
*
* Further reference counting on frozen objects is performed with the atomic operations, and so frozen
* references could be passed accross multiple threads.
*/
void FreezeSubgraph(ObjHeader* root) {
// TODO: for now, we just check that passed object graph has no cycles, and throw an exception,
// if it does. Next version will run Kosoraju-Sharir if cycles are found.
ContainerHeader* rootContainer = root->container();
if (rootContainer->permanentOrFrozen()) return;
// Do DFS cycle detection.
bool hasCycles = false;
depthFirstTraversal(rootContainer, &hasCycles);
// Now unmark all marked objects, and freeze them, if no cycles detected.
KStdDeque<ContainerHeader*> stack;
stack.push_back(rootContainer);
while (!stack.empty()) {
ContainerHeader* current = stack.front();
stack.pop_front();
current->unMark();
current->resetSeen();
if (!hasCycles) {
current->resetBuffered();
current->setColor(CONTAINER_TAG_GC_BLACK);
// Note, that once object is frozen, it could be concurrently accessed, so
// color and similar attributes shall not be used.
current->freeze();
}
traverseContainerObjectFields(current, [&hasCycles, &stack](ObjHeader** location) {
ObjHeader* obj = *location;
if (obj != nullptr) {
ContainerHeader* objContainer = obj->container();
if (!objContainer->permanentOrFrozen() && objContainer->marked())
stack.push_back(objContainer);
}
});
}
// Now remove frozen objects from the toFree list.
// TODO: optimize it by keeping ignored (i.e. freshly frozen) objects in the set,
// and use it when analyzing toFree during collection.
auto state = memoryState;
for (auto it = state->toFree->begin(); it != state->toFree->end(); ++it) {
auto container = *it;
if (container->frozen()) {
*it = markAsRemoved(container);
}
}
// For now, just throw an exception here.
if (hasCycles) ThrowFreezingException();
}
// This function is called from field mutators to check if object's header is frozen.
// If object is frozen, an exception is thrown.
void MutationCheck(ObjHeader* obj) {
if (obj->container()->frozen()) ThrowInvalidMutabilityException();
}
} // extern "C"