view src/share/vm/gc_implementation/parallelScavenge/parallelScavengeHeap.cpp @ 453:c96030fff130

6684579: SoftReference processing can be made more efficient Summary: For current soft-ref clearing policies, we can decide at marking time if a soft-reference will definitely not be cleared, postponing the decision of whether it will definitely be cleared to the final reference processing phase. This can be especially beneficial in the case of concurrent collectors where the marking is usually concurrent but reference processing is usually not. Reviewed-by: jmasa
author ysr
date Thu, 20 Nov 2008 16:56:09 -0800
parents a4b729f5b611
children 660978a2a31a
line wrap: on
line source

/*
 * Copyright 2001-2008 Sun Microsystems, Inc.  All Rights Reserved.
 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
 *
 * This code is free software; you can redistribute it and/or modify it
 * under the terms of the GNU General Public License version 2 only, as
 * published by the Free Software Foundation.
 *
 * This code is distributed in the hope that it will be useful, but WITHOUT
 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
 * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
 * version 2 for more details (a copy is included in the LICENSE file that
 * accompanied this code).
 *
 * You should have received a copy of the GNU General Public License version
 * 2 along with this work; if not, write to the Free Software Foundation,
 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
 *
 * Please contact Sun Microsystems, Inc., 4150 Network Circle, Santa Clara,
 * CA 95054 USA or visit www.sun.com if you need additional information or
 * have any questions.
 *
 */

# include "incls/_precompiled.incl"
# include "incls/_parallelScavengeHeap.cpp.incl"

PSYoungGen*  ParallelScavengeHeap::_young_gen = NULL;
PSOldGen*    ParallelScavengeHeap::_old_gen = NULL;
PSPermGen*   ParallelScavengeHeap::_perm_gen = NULL;
PSAdaptiveSizePolicy* ParallelScavengeHeap::_size_policy = NULL;
PSGCAdaptivePolicyCounters* ParallelScavengeHeap::_gc_policy_counters = NULL;
ParallelScavengeHeap* ParallelScavengeHeap::_psh = NULL;
GCTaskManager* ParallelScavengeHeap::_gc_task_manager = NULL;

static void trace_gen_sizes(const char* const str,
                            size_t pg_min, size_t pg_max,
                            size_t og_min, size_t og_max,
                            size_t yg_min, size_t yg_max)
{
  if (TracePageSizes) {
    tty->print_cr("%s:  " SIZE_FORMAT "," SIZE_FORMAT " "
                  SIZE_FORMAT "," SIZE_FORMAT " "
                  SIZE_FORMAT "," SIZE_FORMAT " "
                  SIZE_FORMAT,
                  str, pg_min / K, pg_max / K,
                  og_min / K, og_max / K,
                  yg_min / K, yg_max / K,
                  (pg_max + og_max + yg_max) / K);
  }
}

jint ParallelScavengeHeap::initialize() {
  // Cannot be initialized until after the flags are parsed
  GenerationSizer flag_parser;

  size_t yg_min_size = flag_parser.min_young_gen_size();
  size_t yg_max_size = flag_parser.max_young_gen_size();
  size_t og_min_size = flag_parser.min_old_gen_size();
  size_t og_max_size = flag_parser.max_old_gen_size();
  // Why isn't there a min_perm_gen_size()?
  size_t pg_min_size = flag_parser.perm_gen_size();
  size_t pg_max_size = flag_parser.max_perm_gen_size();

  trace_gen_sizes("ps heap raw",
                  pg_min_size, pg_max_size,
                  og_min_size, og_max_size,
                  yg_min_size, yg_max_size);

  // The ReservedSpace ctor used below requires that the page size for the perm
  // gen is <= the page size for the rest of the heap (young + old gens).
  const size_t og_page_sz = os::page_size_for_region(yg_min_size + og_min_size,
                                                     yg_max_size + og_max_size,
                                                     8);
  const size_t pg_page_sz = MIN2(os::page_size_for_region(pg_min_size,
                                                          pg_max_size, 16),
                                 og_page_sz);

  const size_t pg_align = set_alignment(_perm_gen_alignment,  pg_page_sz);
  const size_t og_align = set_alignment(_old_gen_alignment,   og_page_sz);
  const size_t yg_align = set_alignment(_young_gen_alignment, og_page_sz);

  // Update sizes to reflect the selected page size(s).
  //
  // NEEDS_CLEANUP.  The default TwoGenerationCollectorPolicy uses NewRatio; it
  // should check UseAdaptiveSizePolicy.  Changes from generationSizer could
  // move to the common code.
  yg_min_size = align_size_up(yg_min_size, yg_align);
  yg_max_size = align_size_up(yg_max_size, yg_align);
  size_t yg_cur_size = align_size_up(flag_parser.young_gen_size(), yg_align);
  yg_cur_size = MAX2(yg_cur_size, yg_min_size);

  og_min_size = align_size_up(og_min_size, og_align);
  og_max_size = align_size_up(og_max_size, og_align);
  size_t og_cur_size = align_size_up(flag_parser.old_gen_size(), og_align);
  og_cur_size = MAX2(og_cur_size, og_min_size);

  pg_min_size = align_size_up(pg_min_size, pg_align);
  pg_max_size = align_size_up(pg_max_size, pg_align);
  size_t pg_cur_size = pg_min_size;

  trace_gen_sizes("ps heap rnd",
                  pg_min_size, pg_max_size,
                  og_min_size, og_max_size,
                  yg_min_size, yg_max_size);

  // The main part of the heap (old gen + young gen) can often use a larger page
  // size than is needed or wanted for the perm gen.  Use the "compound
  // alignment" ReservedSpace ctor to avoid having to use the same page size for
  // all gens.
  ReservedHeapSpace heap_rs(pg_max_size, pg_align, og_max_size + yg_max_size,
                            og_align);
  os::trace_page_sizes("ps perm", pg_min_size, pg_max_size, pg_page_sz,
                       heap_rs.base(), pg_max_size);
  os::trace_page_sizes("ps main", og_min_size + yg_min_size,
                       og_max_size + yg_max_size, og_page_sz,
                       heap_rs.base() + pg_max_size,
                       heap_rs.size() - pg_max_size);
  if (!heap_rs.is_reserved()) {
    vm_shutdown_during_initialization(
      "Could not reserve enough space for object heap");
    return JNI_ENOMEM;
  }

  _reserved = MemRegion((HeapWord*)heap_rs.base(),
                        (HeapWord*)(heap_rs.base() + heap_rs.size()));

  CardTableExtension* const barrier_set = new CardTableExtension(_reserved, 3);
  _barrier_set = barrier_set;
  oopDesc::set_bs(_barrier_set);
  if (_barrier_set == NULL) {
    vm_shutdown_during_initialization(
      "Could not reserve enough space for barrier set");
    return JNI_ENOMEM;
  }

  // Initial young gen size is 4 Mb
  //
  // XXX - what about flag_parser.young_gen_size()?
  const size_t init_young_size = align_size_up(4 * M, yg_align);
  yg_cur_size = MAX2(MIN2(init_young_size, yg_max_size), yg_cur_size);

  // Split the reserved space into perm gen and the main heap (everything else).
  // The main heap uses a different alignment.
  ReservedSpace perm_rs = heap_rs.first_part(pg_max_size);
  ReservedSpace main_rs = heap_rs.last_part(pg_max_size, og_align);

  // Make up the generations
  // Calculate the maximum size that a generation can grow.  This
  // includes growth into the other generation.  Note that the
  // parameter _max_gen_size is kept as the maximum
  // size of the generation as the boundaries currently stand.
  // _max_gen_size is still used as that value.
  double max_gc_pause_sec = ((double) MaxGCPauseMillis)/1000.0;
  double max_gc_minor_pause_sec = ((double) MaxGCMinorPauseMillis)/1000.0;

  _gens = new AdjoiningGenerations(main_rs,
                                   og_cur_size,
                                   og_min_size,
                                   og_max_size,
                                   yg_cur_size,
                                   yg_min_size,
                                   yg_max_size,
                                   yg_align);

  _old_gen = _gens->old_gen();
  _young_gen = _gens->young_gen();

  const size_t eden_capacity = _young_gen->eden_space()->capacity_in_bytes();
  const size_t old_capacity = _old_gen->capacity_in_bytes();
  const size_t initial_promo_size = MIN2(eden_capacity, old_capacity);
  _size_policy =
    new PSAdaptiveSizePolicy(eden_capacity,
                             initial_promo_size,
                             young_gen()->to_space()->capacity_in_bytes(),
                             intra_heap_alignment(),
                             max_gc_pause_sec,
                             max_gc_minor_pause_sec,
                             GCTimeRatio
                             );

  _perm_gen = new PSPermGen(perm_rs,
                            pg_align,
                            pg_cur_size,
                            pg_cur_size,
                            pg_max_size,
                            "perm", 2);

  assert(!UseAdaptiveGCBoundary ||
    (old_gen()->virtual_space()->high_boundary() ==
     young_gen()->virtual_space()->low_boundary()),
    "Boundaries must meet");
  // initialize the policy counters - 2 collectors, 3 generations
  _gc_policy_counters =
    new PSGCAdaptivePolicyCounters("ParScav:MSC", 2, 3, _size_policy);
  _psh = this;

  // Set up the GCTaskManager
  _gc_task_manager = GCTaskManager::create(ParallelGCThreads);

  if (UseParallelOldGC && !PSParallelCompact::initialize()) {
    return JNI_ENOMEM;
  }

  return JNI_OK;
}

void ParallelScavengeHeap::post_initialize() {
  // Need to init the tenuring threshold
  PSScavenge::initialize();
  if (UseParallelOldGC) {
    PSParallelCompact::post_initialize();
  } else {
    PSMarkSweep::initialize();
  }
  PSPromotionManager::initialize();
}

void ParallelScavengeHeap::update_counters() {
  young_gen()->update_counters();
  old_gen()->update_counters();
  perm_gen()->update_counters();
}

size_t ParallelScavengeHeap::capacity() const {
  size_t value = young_gen()->capacity_in_bytes() + old_gen()->capacity_in_bytes();
  return value;
}

size_t ParallelScavengeHeap::used() const {
  size_t value = young_gen()->used_in_bytes() + old_gen()->used_in_bytes();
  return value;
}

bool ParallelScavengeHeap::is_maximal_no_gc() const {
  return old_gen()->is_maximal_no_gc() && young_gen()->is_maximal_no_gc();
}


size_t ParallelScavengeHeap::permanent_capacity() const {
  return perm_gen()->capacity_in_bytes();
}

size_t ParallelScavengeHeap::permanent_used() const {
  return perm_gen()->used_in_bytes();
}

size_t ParallelScavengeHeap::max_capacity() const {
  size_t estimated = reserved_region().byte_size();
  estimated -= perm_gen()->reserved().byte_size();
  if (UseAdaptiveSizePolicy) {
    estimated -= _size_policy->max_survivor_size(young_gen()->max_size());
  } else {
    estimated -= young_gen()->to_space()->capacity_in_bytes();
  }
  return MAX2(estimated, capacity());
}

bool ParallelScavengeHeap::is_in(const void* p) const {
  if (young_gen()->is_in(p)) {
    return true;
  }

  if (old_gen()->is_in(p)) {
    return true;
  }

  if (perm_gen()->is_in(p)) {
    return true;
  }

  return false;
}

bool ParallelScavengeHeap::is_in_reserved(const void* p) const {
  if (young_gen()->is_in_reserved(p)) {
    return true;
  }

  if (old_gen()->is_in_reserved(p)) {
    return true;
  }

  if (perm_gen()->is_in_reserved(p)) {
    return true;
  }

  return false;
}

// Static method
bool ParallelScavengeHeap::is_in_young(oop* p) {
  ParallelScavengeHeap* heap = (ParallelScavengeHeap*)Universe::heap();
  assert(heap->kind() == CollectedHeap::ParallelScavengeHeap,
                                            "Must be ParallelScavengeHeap");

  PSYoungGen* young_gen = heap->young_gen();

  if (young_gen->is_in_reserved(p)) {
    return true;
  }

  return false;
}

// Static method
bool ParallelScavengeHeap::is_in_old_or_perm(oop* p) {
  ParallelScavengeHeap* heap = (ParallelScavengeHeap*)Universe::heap();
  assert(heap->kind() == CollectedHeap::ParallelScavengeHeap,
                                            "Must be ParallelScavengeHeap");

  PSOldGen* old_gen = heap->old_gen();
  PSPermGen* perm_gen = heap->perm_gen();

  if (old_gen->is_in_reserved(p)) {
    return true;
  }

  if (perm_gen->is_in_reserved(p)) {
    return true;
  }

  return false;
}

// There are two levels of allocation policy here.
//
// When an allocation request fails, the requesting thread must invoke a VM
// operation, transfer control to the VM thread, and await the results of a
// garbage collection. That is quite expensive, and we should avoid doing it
// multiple times if possible.
//
// To accomplish this, we have a basic allocation policy, and also a
// failed allocation policy.
//
// The basic allocation policy controls how you allocate memory without
// attempting garbage collection. It is okay to grab locks and
// expand the heap, if that can be done without coming to a safepoint.
// It is likely that the basic allocation policy will not be very
// aggressive.
//
// The failed allocation policy is invoked from the VM thread after
// the basic allocation policy is unable to satisfy a mem_allocate
// request. This policy needs to cover the entire range of collection,
// heap expansion, and out-of-memory conditions. It should make every
// attempt to allocate the requested memory.

// Basic allocation policy. Should never be called at a safepoint, or
// from the VM thread.
//
// This method must handle cases where many mem_allocate requests fail
// simultaneously. When that happens, only one VM operation will succeed,
// and the rest will not be executed. For that reason, this method loops
// during failed allocation attempts. If the java heap becomes exhausted,
// we rely on the size_policy object to force a bail out.
HeapWord* ParallelScavengeHeap::mem_allocate(
                                     size_t size,
                                     bool is_noref,
                                     bool is_tlab,
                                     bool* gc_overhead_limit_was_exceeded) {
  assert(!SafepointSynchronize::is_at_safepoint(), "should not be at safepoint");
  assert(Thread::current() != (Thread*)VMThread::vm_thread(), "should not be in vm thread");
  assert(!Heap_lock->owned_by_self(), "this thread should not own the Heap_lock");

  HeapWord* result = young_gen()->allocate(size, is_tlab);

  uint loop_count = 0;
  uint gc_count = 0;

  while (result == NULL) {
    // We don't want to have multiple collections for a single filled generation.
    // To prevent this, each thread tracks the total_collections() value, and if
    // the count has changed, does not do a new collection.
    //
    // The collection count must be read only while holding the heap lock. VM
    // operations also hold the heap lock during collections. There is a lock
    // contention case where thread A blocks waiting on the Heap_lock, while
    // thread B is holding it doing a collection. When thread A gets the lock,
    // the collection count has already changed. To prevent duplicate collections,
    // The policy MUST attempt allocations during the same period it reads the
    // total_collections() value!
    {
      MutexLocker ml(Heap_lock);
      gc_count = Universe::heap()->total_collections();

      result = young_gen()->allocate(size, is_tlab);

      // (1) If the requested object is too large to easily fit in the
      //     young_gen, or
      // (2) If GC is locked out via GCLocker, young gen is full and
      //     the need for a GC already signalled to GCLocker (done
      //     at a safepoint),
      // ... then, rather than force a safepoint and (a potentially futile)
      // collection (attempt) for each allocation, try allocation directly
      // in old_gen. For case (2) above, we may in the future allow
      // TLAB allocation directly in the old gen.
      if (result != NULL) {
        return result;
      }
      if (!is_tlab &&
          size >= (young_gen()->eden_space()->capacity_in_words(Thread::current()) / 2)) {
        result = old_gen()->allocate(size, is_tlab);
        if (result != NULL) {
          return result;
        }
      }
      if (GC_locker::is_active_and_needs_gc()) {
        // GC is locked out. If this is a TLAB allocation,
        // return NULL; the requestor will retry allocation
        // of an idividual object at a time.
        if (is_tlab) {
          return NULL;
        }

        // If this thread is not in a jni critical section, we stall
        // the requestor until the critical section has cleared and
        // GC allowed. When the critical section clears, a GC is
        // initiated by the last thread exiting the critical section; so
        // we retry the allocation sequence from the beginning of the loop,
        // rather than causing more, now probably unnecessary, GC attempts.
        JavaThread* jthr = JavaThread::current();
        if (!jthr->in_critical()) {
          MutexUnlocker mul(Heap_lock);
          GC_locker::stall_until_clear();
          continue;
        } else {
          if (CheckJNICalls) {
            fatal("Possible deadlock due to allocating while"
                  " in jni critical section");
          }
          return NULL;
        }
      }
    }

    if (result == NULL) {

      // Exit the loop if if the gc time limit has been exceeded.
      // The allocation must have failed above (result must be NULL),
      // and the most recent collection must have exceeded the
      // gc time limit.  Exit the loop so that an out-of-memory
      // will be thrown (returning a NULL will do that), but
      // clear gc_time_limit_exceeded so that the next collection
      // will succeeded if the applications decides to handle the
      // out-of-memory and tries to go on.
      *gc_overhead_limit_was_exceeded = size_policy()->gc_time_limit_exceeded();
      if (size_policy()->gc_time_limit_exceeded()) {
        size_policy()->set_gc_time_limit_exceeded(false);
        if (PrintGCDetails && Verbose) {
        gclog_or_tty->print_cr("ParallelScavengeHeap::mem_allocate: "
          "return NULL because gc_time_limit_exceeded is set");
        }
        return NULL;
      }

      // Generate a VM operation
      VM_ParallelGCFailedAllocation op(size, is_tlab, gc_count);
      VMThread::execute(&op);

      // Did the VM operation execute? If so, return the result directly.
      // This prevents us from looping until time out on requests that can
      // not be satisfied.
      if (op.prologue_succeeded()) {
        assert(Universe::heap()->is_in_or_null(op.result()),
          "result not in heap");

        // If GC was locked out during VM operation then retry allocation
        // and/or stall as necessary.
        if (op.gc_locked()) {
          assert(op.result() == NULL, "must be NULL if gc_locked() is true");
          continue;  // retry and/or stall as necessary
        }
        // If a NULL result is being returned, an out-of-memory
        // will be thrown now.  Clear the gc_time_limit_exceeded
        // flag to avoid the following situation.
        //      gc_time_limit_exceeded is set during a collection
        //      the collection fails to return enough space and an OOM is thrown
        //      the next GC is skipped because the gc_time_limit_exceeded
        //        flag is set and another OOM is thrown
        if (op.result() == NULL) {
          size_policy()->set_gc_time_limit_exceeded(false);
        }
        return op.result();
      }
    }

    // The policy object will prevent us from looping forever. If the
    // time spent in gc crosses a threshold, we will bail out.
    loop_count++;
    if ((result == NULL) && (QueuedAllocationWarningCount > 0) &&
        (loop_count % QueuedAllocationWarningCount == 0)) {
      warning("ParallelScavengeHeap::mem_allocate retries %d times \n\t"
              " size=%d %s", loop_count, size, is_tlab ? "(TLAB)" : "");
    }
  }

  return result;
}

// Failed allocation policy. Must be called from the VM thread, and
// only at a safepoint! Note that this method has policy for allocation
// flow, and NOT collection policy. So we do not check for gc collection
// time over limit here, that is the responsibility of the heap specific
// collection methods. This method decides where to attempt allocations,
// and when to attempt collections, but no collection specific policy.
HeapWord* ParallelScavengeHeap::failed_mem_allocate(size_t size, bool is_tlab) {
  assert(SafepointSynchronize::is_at_safepoint(), "should be at safepoint");
  assert(Thread::current() == (Thread*)VMThread::vm_thread(), "should be in vm thread");
  assert(!Universe::heap()->is_gc_active(), "not reentrant");
  assert(!Heap_lock->owned_by_self(), "this thread should not own the Heap_lock");

  size_t mark_sweep_invocation_count = total_invocations();

  // We assume (and assert!) that an allocation at this point will fail
  // unless we collect.

  // First level allocation failure, scavenge and allocate in young gen.
  GCCauseSetter gccs(this, GCCause::_allocation_failure);
  PSScavenge::invoke();
  HeapWord* result = young_gen()->allocate(size, is_tlab);

  // Second level allocation failure.
  //   Mark sweep and allocate in young generation.
  if (result == NULL) {
    // There is some chance the scavenge method decided to invoke mark_sweep.
    // Don't mark sweep twice if so.
    if (mark_sweep_invocation_count == total_invocations()) {
      invoke_full_gc(false);
      result = young_gen()->allocate(size, is_tlab);
    }
  }

  // Third level allocation failure.
  //   After mark sweep and young generation allocation failure,
  //   allocate in old generation.
  if (result == NULL && !is_tlab) {
    result = old_gen()->allocate(size, is_tlab);
  }

  // Fourth level allocation failure. We're running out of memory.
  //   More complete mark sweep and allocate in young generation.
  if (result == NULL) {
    invoke_full_gc(true);
    result = young_gen()->allocate(size, is_tlab);
  }

  // Fifth level allocation failure.
  //   After more complete mark sweep, allocate in old generation.
  if (result == NULL && !is_tlab) {
    result = old_gen()->allocate(size, is_tlab);
  }

  return result;
}

//
// This is the policy loop for allocating in the permanent generation.
// If the initial allocation fails, we create a vm operation which will
// cause a collection.
HeapWord* ParallelScavengeHeap::permanent_mem_allocate(size_t size) {
  assert(!SafepointSynchronize::is_at_safepoint(), "should not be at safepoint");
  assert(Thread::current() != (Thread*)VMThread::vm_thread(), "should not be in vm thread");
  assert(!Heap_lock->owned_by_self(), "this thread should not own the Heap_lock");

  HeapWord* result;

  uint loop_count = 0;
  uint gc_count = 0;
  uint full_gc_count = 0;

  do {
    // We don't want to have multiple collections for a single filled generation.
    // To prevent this, each thread tracks the total_collections() value, and if
    // the count has changed, does not do a new collection.
    //
    // The collection count must be read only while holding the heap lock. VM
    // operations also hold the heap lock during collections. There is a lock
    // contention case where thread A blocks waiting on the Heap_lock, while
    // thread B is holding it doing a collection. When thread A gets the lock,
    // the collection count has already changed. To prevent duplicate collections,
    // The policy MUST attempt allocations during the same period it reads the
    // total_collections() value!
    {
      MutexLocker ml(Heap_lock);
      gc_count      = Universe::heap()->total_collections();
      full_gc_count = Universe::heap()->total_full_collections();

      result = perm_gen()->allocate_permanent(size);

      if (result != NULL) {
        return result;
      }

      if (GC_locker::is_active_and_needs_gc()) {
        // If this thread is not in a jni critical section, we stall
        // the requestor until the critical section has cleared and
        // GC allowed. When the critical section clears, a GC is
        // initiated by the last thread exiting the critical section; so
        // we retry the allocation sequence from the beginning of the loop,
        // rather than causing more, now probably unnecessary, GC attempts.
        JavaThread* jthr = JavaThread::current();
        if (!jthr->in_critical()) {
          MutexUnlocker mul(Heap_lock);
          GC_locker::stall_until_clear();
          continue;
        } else {
          if (CheckJNICalls) {
            fatal("Possible deadlock due to allocating while"
                  " in jni critical section");
          }
          return NULL;
        }
      }
    }

    if (result == NULL) {

      // Exit the loop if the gc time limit has been exceeded.
      // The allocation must have failed above (result must be NULL),
      // and the most recent collection must have exceeded the
      // gc time limit.  Exit the loop so that an out-of-memory
      // will be thrown (returning a NULL will do that), but
      // clear gc_time_limit_exceeded so that the next collection
      // will succeeded if the applications decides to handle the
      // out-of-memory and tries to go on.
      if (size_policy()->gc_time_limit_exceeded()) {
        size_policy()->set_gc_time_limit_exceeded(false);
        if (PrintGCDetails && Verbose) {
        gclog_or_tty->print_cr("ParallelScavengeHeap::permanent_mem_allocate: "
          "return NULL because gc_time_limit_exceeded is set");
        }
        assert(result == NULL, "Allocation did not fail");
        return NULL;
      }

      // Generate a VM operation
      VM_ParallelGCFailedPermanentAllocation op(size, gc_count, full_gc_count);
      VMThread::execute(&op);

      // Did the VM operation execute? If so, return the result directly.
      // This prevents us from looping until time out on requests that can
      // not be satisfied.
      if (op.prologue_succeeded()) {
        assert(Universe::heap()->is_in_permanent_or_null(op.result()),
          "result not in heap");
        // If GC was locked out during VM operation then retry allocation
        // and/or stall as necessary.
        if (op.gc_locked()) {
          assert(op.result() == NULL, "must be NULL if gc_locked() is true");
          continue;  // retry and/or stall as necessary
        }
        // If a NULL results is being returned, an out-of-memory
        // will be thrown now.  Clear the gc_time_limit_exceeded
        // flag to avoid the following situation.
        //      gc_time_limit_exceeded is set during a collection
        //      the collection fails to return enough space and an OOM is thrown
        //      the next GC is skipped because the gc_time_limit_exceeded
        //        flag is set and another OOM is thrown
        if (op.result() == NULL) {
          size_policy()->set_gc_time_limit_exceeded(false);
        }
        return op.result();
      }
    }

    // The policy object will prevent us from looping forever. If the
    // time spent in gc crosses a threshold, we will bail out.
    loop_count++;
    if ((QueuedAllocationWarningCount > 0) &&
        (loop_count % QueuedAllocationWarningCount == 0)) {
      warning("ParallelScavengeHeap::permanent_mem_allocate retries %d times \n\t"
              " size=%d", loop_count, size);
    }
  } while (result == NULL);

  return result;
}

//
// This is the policy code for permanent allocations which have failed
// and require a collection. Note that just as in failed_mem_allocate,
// we do not set collection policy, only where & when to allocate and
// collect.
HeapWord* ParallelScavengeHeap::failed_permanent_mem_allocate(size_t size) {
  assert(SafepointSynchronize::is_at_safepoint(), "should be at safepoint");
  assert(Thread::current() == (Thread*)VMThread::vm_thread(), "should be in vm thread");
  assert(!Universe::heap()->is_gc_active(), "not reentrant");
  assert(!Heap_lock->owned_by_self(), "this thread should not own the Heap_lock");
  assert(size > perm_gen()->free_in_words(), "Allocation should fail");

  // We assume (and assert!) that an allocation at this point will fail
  // unless we collect.

  // First level allocation failure.  Mark-sweep and allocate in perm gen.
  GCCauseSetter gccs(this, GCCause::_allocation_failure);
  invoke_full_gc(false);
  HeapWord* result = perm_gen()->allocate_permanent(size);

  // Second level allocation failure. We're running out of memory.
  if (result == NULL) {
    invoke_full_gc(true);
    result = perm_gen()->allocate_permanent(size);
  }

  return result;
}

void ParallelScavengeHeap::ensure_parsability(bool retire_tlabs) {
  CollectedHeap::ensure_parsability(retire_tlabs);
  young_gen()->eden_space()->ensure_parsability();
}

size_t ParallelScavengeHeap::unsafe_max_alloc() {
  return young_gen()->eden_space()->free_in_bytes();
}

size_t ParallelScavengeHeap::tlab_capacity(Thread* thr) const {
  return young_gen()->eden_space()->tlab_capacity(thr);
}

size_t ParallelScavengeHeap::unsafe_max_tlab_alloc(Thread* thr) const {
  return young_gen()->eden_space()->unsafe_max_tlab_alloc(thr);
}

HeapWord* ParallelScavengeHeap::allocate_new_tlab(size_t size) {
  return young_gen()->allocate(size, true);
}

void ParallelScavengeHeap::fill_all_tlabs(bool retire) {
  CollectedHeap::fill_all_tlabs(retire);
}

void ParallelScavengeHeap::accumulate_statistics_all_tlabs() {
  CollectedHeap::accumulate_statistics_all_tlabs();
}

void ParallelScavengeHeap::resize_all_tlabs() {
  CollectedHeap::resize_all_tlabs();
}

// This method is used by System.gc() and JVMTI.
void ParallelScavengeHeap::collect(GCCause::Cause cause) {
  assert(!Heap_lock->owned_by_self(),
    "this thread should not own the Heap_lock");

  unsigned int gc_count      = 0;
  unsigned int full_gc_count = 0;
  {
    MutexLocker ml(Heap_lock);
    // This value is guarded by the Heap_lock
    gc_count      = Universe::heap()->total_collections();
    full_gc_count = Universe::heap()->total_full_collections();
  }

  VM_ParallelGCSystemGC op(gc_count, full_gc_count, cause);
  VMThread::execute(&op);
}

// This interface assumes that it's being called by the
// vm thread. It collects the heap assuming that the
// heap lock is already held and that we are executing in
// the context of the vm thread.
void ParallelScavengeHeap::collect_as_vm_thread(GCCause::Cause cause) {
  assert(Thread::current()->is_VM_thread(), "Precondition#1");
  assert(Heap_lock->is_locked(), "Precondition#2");
  GCCauseSetter gcs(this, cause);
  switch (cause) {
    case GCCause::_heap_inspection:
    case GCCause::_heap_dump: {
      HandleMark hm;
      invoke_full_gc(false);
      break;
    }
    default: // XXX FIX ME
      ShouldNotReachHere();
  }
}


void ParallelScavengeHeap::oop_iterate(OopClosure* cl) {
  Unimplemented();
}

void ParallelScavengeHeap::object_iterate(ObjectClosure* cl) {
  young_gen()->object_iterate(cl);
  old_gen()->object_iterate(cl);
  perm_gen()->object_iterate(cl);
}

void ParallelScavengeHeap::permanent_oop_iterate(OopClosure* cl) {
  Unimplemented();
}

void ParallelScavengeHeap::permanent_object_iterate(ObjectClosure* cl) {
  perm_gen()->object_iterate(cl);
}

HeapWord* ParallelScavengeHeap::block_start(const void* addr) const {
  if (young_gen()->is_in_reserved(addr)) {
    assert(young_gen()->is_in(addr),
           "addr should be in allocated part of young gen");
    Unimplemented();
  } else if (old_gen()->is_in_reserved(addr)) {
    assert(old_gen()->is_in(addr),
           "addr should be in allocated part of old gen");
    return old_gen()->start_array()->object_start((HeapWord*)addr);
  } else if (perm_gen()->is_in_reserved(addr)) {
    assert(perm_gen()->is_in(addr),
           "addr should be in allocated part of perm gen");
    return perm_gen()->start_array()->object_start((HeapWord*)addr);
  }
  return 0;
}

size_t ParallelScavengeHeap::block_size(const HeapWord* addr) const {
  return oop(addr)->size();
}

bool ParallelScavengeHeap::block_is_obj(const HeapWord* addr) const {
  return block_start(addr) == addr;
}

jlong ParallelScavengeHeap::millis_since_last_gc() {
  return UseParallelOldGC ?
    PSParallelCompact::millis_since_last_gc() :
    PSMarkSweep::millis_since_last_gc();
}

void ParallelScavengeHeap::prepare_for_verify() {
  ensure_parsability(false);  // no need to retire TLABs for verification
}

void ParallelScavengeHeap::print() const { print_on(tty); }

void ParallelScavengeHeap::print_on(outputStream* st) const {
  young_gen()->print_on(st);
  old_gen()->print_on(st);
  perm_gen()->print_on(st);
}

void ParallelScavengeHeap::gc_threads_do(ThreadClosure* tc) const {
  PSScavenge::gc_task_manager()->threads_do(tc);
}

void ParallelScavengeHeap::print_gc_threads_on(outputStream* st) const {
  PSScavenge::gc_task_manager()->print_threads_on(st);
}

void ParallelScavengeHeap::print_tracing_info() const {
  if (TraceGen0Time) {
    double time = PSScavenge::accumulated_time()->seconds();
    tty->print_cr("[Accumulated GC generation 0 time %3.7f secs]", time);
  }
  if (TraceGen1Time) {
    double time = PSMarkSweep::accumulated_time()->seconds();
    tty->print_cr("[Accumulated GC generation 1 time %3.7f secs]", time);
  }
}


void ParallelScavengeHeap::verify(bool allow_dirty, bool silent) {
  // Why do we need the total_collections()-filter below?
  if (total_collections() > 0) {
    if (!silent) {
      gclog_or_tty->print("permanent ");
    }
    perm_gen()->verify(allow_dirty);

    if (!silent) {
      gclog_or_tty->print("tenured ");
    }
    old_gen()->verify(allow_dirty);

    if (!silent) {
      gclog_or_tty->print("eden ");
    }
    young_gen()->verify(allow_dirty);
  }
  if (!silent) {
    gclog_or_tty->print("ref_proc ");
  }
  ReferenceProcessor::verify();
}

void ParallelScavengeHeap::print_heap_change(size_t prev_used) {
  if (PrintGCDetails && Verbose) {
    gclog_or_tty->print(" "  SIZE_FORMAT
                        "->" SIZE_FORMAT
                        "("  SIZE_FORMAT ")",
                        prev_used, used(), capacity());
  } else {
    gclog_or_tty->print(" "  SIZE_FORMAT "K"
                        "->" SIZE_FORMAT "K"
                        "("  SIZE_FORMAT "K)",
                        prev_used / K, used() / K, capacity() / K);
  }
}

ParallelScavengeHeap* ParallelScavengeHeap::heap() {
  assert(_psh != NULL, "Uninitialized access to ParallelScavengeHeap::heap()");
  assert(_psh->kind() == CollectedHeap::ParallelScavengeHeap, "not a parallel scavenge heap");
  return _psh;
}

// Before delegating the resize to the young generation,
// the reserved space for the young and old generations
// may be changed to accomodate the desired resize.
void ParallelScavengeHeap::resize_young_gen(size_t eden_size,
    size_t survivor_size) {
  if (UseAdaptiveGCBoundary) {
    if (size_policy()->bytes_absorbed_from_eden() != 0) {
      size_policy()->reset_bytes_absorbed_from_eden();
      return;  // The generation changed size already.
    }
    gens()->adjust_boundary_for_young_gen_needs(eden_size, survivor_size);
  }

  // Delegate the resize to the generation.
  _young_gen->resize(eden_size, survivor_size);
}

// Before delegating the resize to the old generation,
// the reserved space for the young and old generations
// may be changed to accomodate the desired resize.
void ParallelScavengeHeap::resize_old_gen(size_t desired_free_space) {
  if (UseAdaptiveGCBoundary) {
    if (size_policy()->bytes_absorbed_from_eden() != 0) {
      size_policy()->reset_bytes_absorbed_from_eden();
      return;  // The generation changed size already.
    }
    gens()->adjust_boundary_for_old_gen_needs(desired_free_space);
  }

  // Delegate the resize to the generation.
  _old_gen->resize(desired_free_space);
}

#ifndef PRODUCT
void ParallelScavengeHeap::record_gen_tops_before_GC() {
  if (ZapUnusedHeapArea) {
    young_gen()->record_spaces_top();
    old_gen()->record_spaces_top();
    perm_gen()->record_spaces_top();
  }
}

void ParallelScavengeHeap::gen_mangle_unused_area() {
  if (ZapUnusedHeapArea) {
    young_gen()->eden_space()->mangle_unused_area();
    young_gen()->to_space()->mangle_unused_area();
    young_gen()->from_space()->mangle_unused_area();
    old_gen()->object_space()->mangle_unused_area();
    perm_gen()->object_space()->mangle_unused_area();
  }
}
#endif