本文整理汇总了C++中GenCollectedHeap::total_full_collections_completed方法的典型用法代码示例。如果您正苦于以下问题：C++ GenCollectedHeap::total_full_collections_completed方法的具体用法？C++ GenCollectedHeap::total_full_collections_completed怎么用？C++ GenCollectedHeap::total_full_collections_completed使用的例子？那么, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类GenCollectedHeap的用法示例。
在下文中一共展示了GenCollectedHeap::total_full_collections_completed方法的2个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于系统推荐出更棒的C++代码示例。
示例1: doit_epilogue
void VM_GenCollectFullConcurrent::doit_epilogue() {
Thread* thr = Thread::current();
assert(thr->is_Java_thread(), "just checking");
JavaThread* jt = (JavaThread*)thr;
// Release the Heap_lock first.
Heap_lock->unlock();
release_and_notify_pending_list_lock();
// It is fine to test whether completed collections has
// exceeded our request count without locking because
// the completion count is monotonically increasing;
// this will break for very long-running apps when the
// count overflows and wraps around. XXX fix me !!!
// e.g. at the rate of 1 full gc per ms, this could
// overflow in about 1000 years.
GenCollectedHeap* gch = GenCollectedHeap::heap();
if (gch->total_full_collections_completed() <= _full_gc_count_before) {
// Now, wait for witnessing concurrent gc cycle to complete,
// but do so in native mode, because we want to lock the
// FullGCEvent_lock, which may be needed by the VM thread
// or by the CMS thread, so we do not want to be suspended
// while holding that lock.
ThreadToNativeFromVM native(jt);
MutexLockerEx ml(FullGCCount_lock, Mutex::_no_safepoint_check_flag);
// Either a concurrent or a stop-world full gc is sufficient
// witness to our request.
while (gch->total_full_collections_completed() <= _full_gc_count_before) {
FullGCCount_lock->wait(Mutex::_no_safepoint_check_flag);
}
}
// Enable iCMS back.
CMSCollector::enable_icms();
}示例2: doit_epilogue
void VM_GenCollectFullConcurrent::doit_epilogue() {
Thread* thr = Thread::current();
assert(thr->is_Java_thread(), "just checking");
JavaThread* jt = (JavaThread*)thr;
if (Universe::has_reference_pending_list()) {
Heap_lock->notify_all();
}
Heap_lock->unlock();
// It is fine to test whether completed collections has
// exceeded our request count without locking because
// the completion count is monotonically increasing;
// this will break for very long-running apps when the
// count overflows and wraps around. XXX fix me !!!
// e.g. at the rate of 1 full gc per ms, this could
// overflow in about 1000 years.
GenCollectedHeap* gch = GenCollectedHeap::heap();
if (_gc_cause != GCCause::_gc_locker &&
gch->total_full_collections_completed() <= _full_gc_count_before) {
// maybe we should change the condition to test _gc_cause ==
// GCCause::_java_lang_system_gc or GCCause::_dcmd_gc_run,
// instead of _gc_cause != GCCause::_gc_locker
assert(GCCause::is_user_requested_gc(_gc_cause),
"the only way to get here if this was a System.gc()-induced GC");
assert(ExplicitGCInvokesConcurrent, "Error");
// Now, wait for witnessing concurrent gc cycle to complete,
// but do so in native mode, because we want to lock the
// FullGCEvent_lock, which may be needed by the VM thread
// or by the CMS thread, so we do not want to be suspended
// while holding that lock.
ThreadToNativeFromVM native(jt);
MutexLockerEx ml(FullGCCount_lock, Mutex::_no_safepoint_check_flag);
// Either a concurrent or a stop-world full gc is sufficient
// witness to our request.
while (gch->total_full_collections_completed() <= _full_gc_count_before) {
FullGCCount_lock->wait(Mutex::_no_safepoint_check_flag);
}
}
}