8186571: Implementation: JEP 307: Parallel Full GC for G1
Summary: Improve G1 worst-case latencies by making the full GC parallel.
Reviewed-by: tschatzl, sangheki, ehelin
/*
* Copyright (c) 2017, 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_G1_G1CONCURRENTMARKBITMAP_INLINE_HPP
#define SHARE_VM_GC_G1_G1CONCURRENTMARKBITMAP_INLINE_HPP
#include "gc/g1/g1ConcurrentMarkBitMap.hpp"
#include "memory/memRegion.hpp"
#include "utilities/align.hpp"
#include "utilities/bitMap.inline.hpp"
inline bool G1CMBitMap::iterate(G1CMBitMapClosure* cl, MemRegion mr) {
assert(!mr.is_empty(), "Does not support empty memregion to iterate over");
assert(_covered.contains(mr),
"Given MemRegion from " PTR_FORMAT " to " PTR_FORMAT " not contained in heap area",
p2i(mr.start()), p2i(mr.end()));
BitMap::idx_t const end_offset = addr_to_offset(mr.end());
BitMap::idx_t offset = _bm.get_next_one_offset(addr_to_offset(mr.start()), end_offset);
while (offset < end_offset) {
HeapWord* const addr = offset_to_addr(offset);
if (!cl->do_addr(addr)) {
return false;
}
size_t const obj_size = (size_t)((oop)addr)->size();
offset = _bm.get_next_one_offset(offset + (obj_size >> _shifter), end_offset);
}
return true;
}
inline HeapWord* G1CMBitMap::get_next_marked_addr(const HeapWord* addr,
const HeapWord* limit) const {
assert(limit != NULL, "limit must not be NULL");
// Round addr up to a possible object boundary to be safe.
size_t const addr_offset = addr_to_offset(align_up(addr, HeapWordSize << _shifter));
size_t const limit_offset = addr_to_offset(limit);
size_t const nextOffset = _bm.get_next_one_offset(addr_offset, limit_offset);
return offset_to_addr(nextOffset);
}
#ifdef ASSERT
inline void G1CMBitMap::check_mark(HeapWord* addr) {
assert(G1CollectedHeap::heap()->is_in_exact(addr),
"Trying to access bitmap " PTR_FORMAT " for address " PTR_FORMAT " not in the heap.",
p2i(this), p2i(addr));
}
#endif
inline void G1CMBitMap::mark(HeapWord* addr) {
check_mark(addr);
_bm.set_bit(addr_to_offset(addr));
}
inline void G1CMBitMap::clear(HeapWord* addr) {
check_mark(addr);
_bm.clear_bit(addr_to_offset(addr));
}
inline bool G1CMBitMap::par_mark(HeapWord* addr) {
check_mark(addr);
return _bm.par_set_bit(addr_to_offset(addr));
}
inline bool G1CMBitMap::par_mark(oop obj) {
return par_mark((HeapWord*) obj);
}
inline bool G1CMBitMap::is_marked(oop obj) const{
return is_marked((HeapWord*) obj);
}
inline void G1CMBitMap::clear(oop obj) {
clear((HeapWord*) obj);
}
#endif // SHARE_VM_GC_G1_G1CONCURRENTMARKBITMAP_INLINE_HPP