view src/share/vm/gc_implementation/g1/bufferingOopClosure.hpp @ 1973:631f79e71e90

6974966: G1: unnecessary direct-to-old allocations Summary: This change revamps the slow allocation path of G1. Improvements include the following: a) Allocations directly to old regions are now totally banned. G1 now only allows allocations out of young regions (with the only exception being humongous regions). b) The thread that allocates a new region (which is now guaranteed to be young) does not dirty all its cards. Each thread that successfully allocates out of a young region is now responsible for dirtying the cards that corresponding to the "block" that just got allocated. c) allocate_new_tlab() and mem_allocate() are now implemented differently and TLAB allocations are only done by allocate_new_tlab(). d) If a thread schedules an evacuation pause in order to satisfy an allocation request, it will perform the allocation at the end of the safepoint so that the thread that initiated the GC also gets "first pick" of any space made available by the GC. e) If a thread is unable to allocate a humongous object it will schedule an evacuation pause in case it reclaims enough regions so that the humongous allocation can be satisfied aftewards. f) The G1 policy is more careful to set the young list target length to be the survivor number +1. g) Lots of code tidy up, removal, refactoring to make future changes easier. Reviewed-by: johnc, ysr
author tonyp
date Tue, 24 Aug 2010 17:24:33 -0400
parents f95d63e2154a
children c685ef164975
line wrap: on
line source

/*
 * Copyright (c) 2001, 2010, Oracle and/or its affiliates. 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 Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
 * or visit www.oracle.com if you need additional information or have any
 * questions.
 *
 */

#ifndef SHARE_VM_GC_IMPLEMENTATION_G1_BUFFERINGOOPCLOSURE_HPP
#define SHARE_VM_GC_IMPLEMENTATION_G1_BUFFERINGOOPCLOSURE_HPP

#include "memory/genOopClosures.hpp"
#include "memory/generation.hpp"
#include "runtime/os.hpp"
#include "utilities/taskqueue.hpp"

// A BufferingOops closure tries to separate out the cost of finding roots
// from the cost of applying closures to them.  It maintains an array of
// ref-containing locations.  Until the array is full, applying the closure
// to an oop* merely records that location in the array.  Since this
// closure app cost is small, an elapsed timer can approximately attribute
// all of this cost to the cost of finding the roots.  When the array fills
// up, the wrapped closure is applied to all elements, keeping track of
// this elapsed time of this process, and leaving the array empty.
// The caller must be sure to call "done" to process any unprocessed
// buffered entriess.

class Generation;
class HeapRegion;

class BufferingOopClosure: public OopClosure {
protected:
  enum PrivateConstants {
    BufferLength = 1024
  };

  StarTask  _buffer[BufferLength];
  StarTask* _buffer_top;
  StarTask* _buffer_curr;

  OopClosure* _oc;
  double      _closure_app_seconds;

  void process_buffer () {
    double start = os::elapsedTime();
    for (StarTask* curr = _buffer; curr < _buffer_curr; ++curr) {
      if (curr->is_narrow()) {
        assert(UseCompressedOops, "Error");
        _oc->do_oop((narrowOop*)(*curr));
      } else {
        _oc->do_oop((oop*)(*curr));
      }
    }
    _buffer_curr = _buffer;
    _closure_app_seconds += (os::elapsedTime() - start);
  }

  template <class T> inline void do_oop_work(T* p) {
    if (_buffer_curr == _buffer_top) {
      process_buffer();
    }
    StarTask new_ref(p);
    *_buffer_curr = new_ref;
    ++_buffer_curr;
  }

public:
  virtual void do_oop(narrowOop* p) { do_oop_work(p); }
  virtual void do_oop(oop* p)       { do_oop_work(p); }

  void done () {
    if (_buffer_curr > _buffer) {
      process_buffer();
    }
  }
  double closure_app_seconds () {
    return _closure_app_seconds;
  }
  BufferingOopClosure (OopClosure *oc) :
    _oc(oc),
    _buffer_curr(_buffer), _buffer_top(_buffer + BufferLength),
    _closure_app_seconds(0.0) { }
};

class BufferingOopsInGenClosure: public OopsInGenClosure {
  BufferingOopClosure _boc;
  OopsInGenClosure* _oc;
 protected:
  template <class T> inline void do_oop_work(T* p) {
    assert(generation()->is_in_reserved((void*)p), "Must be in!");
    _boc.do_oop(p);
  }
 public:
  BufferingOopsInGenClosure(OopsInGenClosure *oc) :
    _boc(oc), _oc(oc) {}

  virtual void do_oop(narrowOop* p) { do_oop_work(p); }
  virtual void do_oop(oop* p)       { do_oop_work(p); }

  void done() {
    _boc.done();
  }

  double closure_app_seconds () {
    return _boc.closure_app_seconds();
  }

  void set_generation(Generation* gen) {
    OopsInGenClosure::set_generation(gen);
    _oc->set_generation(gen);
  }

  void reset_generation() {
    // Make sure we finish the current work with the current generation.
    _boc.done();
    OopsInGenClosure::reset_generation();
    _oc->reset_generation();
  }

};


class BufferingOopsInHeapRegionClosure: public OopsInHeapRegionClosure {
private:
  enum PrivateConstants {
    BufferLength = 1024
  };

  StarTask     _buffer[BufferLength];
  StarTask*    _buffer_top;
  StarTask*    _buffer_curr;

  HeapRegion*  _hr_buffer[BufferLength];
  HeapRegion** _hr_curr;

  OopsInHeapRegionClosure*  _oc;
  double                    _closure_app_seconds;

  void process_buffer () {

    assert((_hr_curr - _hr_buffer) == (_buffer_curr - _buffer),
           "the two lengths should be the same");

    double start = os::elapsedTime();
    HeapRegion** hr_curr = _hr_buffer;
    HeapRegion*  hr_prev = NULL;
    for (StarTask* curr = _buffer; curr < _buffer_curr; ++curr) {
      HeapRegion* region = *hr_curr;
      if (region != hr_prev) {
        _oc->set_region(region);
        hr_prev = region;
      }
      if (curr->is_narrow()) {
        assert(UseCompressedOops, "Error");
        _oc->do_oop((narrowOop*)(*curr));
      } else {
        _oc->do_oop((oop*)(*curr));
      }
      ++hr_curr;
    }
    _buffer_curr = _buffer;
    _hr_curr = _hr_buffer;
    _closure_app_seconds += (os::elapsedTime() - start);
  }

public:
  virtual void do_oop(narrowOop* p) { do_oop_work(p); }
  virtual void do_oop(      oop* p) { do_oop_work(p); }

  template <class T> void do_oop_work(T* p) {
    if (_buffer_curr == _buffer_top) {
      assert(_hr_curr > _hr_buffer, "_hr_curr should be consistent with _buffer_curr");
      process_buffer();
    }
    StarTask new_ref(p);
    *_buffer_curr = new_ref;
    ++_buffer_curr;
    *_hr_curr = _from;
    ++_hr_curr;
  }
  void done () {
    if (_buffer_curr > _buffer) {
      assert(_hr_curr > _hr_buffer, "_hr_curr should be consistent with _buffer_curr");
      process_buffer();
    }
  }
  double closure_app_seconds () {
    return _closure_app_seconds;
  }
  BufferingOopsInHeapRegionClosure (OopsInHeapRegionClosure *oc) :
    _oc(oc),
    _buffer_curr(_buffer), _buffer_top(_buffer + BufferLength),
    _hr_curr(_hr_buffer),
    _closure_app_seconds(0.0) { }
};

#endif // SHARE_VM_GC_IMPLEMENTATION_G1_BUFFERINGOOPCLOSURE_HPP