1 /*
2 * Copyright (c) 2001, 2015, Oracle and/or its affiliates. All rights reserved.
3 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
4 *
5 * This code is free software; you can redistribute it and/or modify it
6 * under the terms of the GNU General Public License version 2 only, as
7 * published by the Free Software Foundation.
8 *
9 * This code is distributed in the hope that it will be useful, but WITHOUT
10 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
11 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
12 * version 2 for more details (a copy is included in the LICENSE file that
13 * accompanied this code).
14 *
15 * You should have received a copy of the GNU General Public License version
16 * 2 along with this work; if not, write to the Free Software Foundation,
17 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
18 *
19 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
20 * or visit www.oracle.com if you need additional information or have any
21 * questions.
22 *
23 */
24
25 #include "precompiled.hpp"
26 #include "classfile/metadataOnStackMark.hpp"
27 #include "classfile/symbolTable.hpp"
28 #include "code/codeCache.hpp"
29 #include "gc/g1/concurrentMark.inline.hpp"
30 #include "gc/g1/concurrentMarkThread.inline.hpp"
31 #include "gc/g1/g1CollectedHeap.inline.hpp"
32 #include "gc/g1/g1CollectorPolicy.hpp"
33 #include "gc/g1/g1ErgoVerbose.hpp"
34 #include "gc/g1/g1Log.hpp"
35 #include "gc/g1/g1OopClosures.inline.hpp"
36 #include "gc/g1/g1RemSet.hpp"
37 #include "gc/g1/g1StringDedup.hpp"
38 #include "gc/g1/heapRegion.inline.hpp"
39 #include "gc/g1/heapRegionManager.inline.hpp"
40 #include "gc/g1/heapRegionRemSet.hpp"
41 #include "gc/g1/heapRegionSet.inline.hpp"
42 #include "gc/g1/suspendibleThreadSet.hpp"
43 #include "gc/shared/gcTimer.hpp"
44 #include "gc/shared/gcTrace.hpp"
45 #include "gc/shared/gcTraceTime.hpp"
46 #include "gc/shared/genOopClosures.inline.hpp"
47 #include "gc/shared/referencePolicy.hpp"
48 #include "gc/shared/strongRootsScope.hpp"
49 #include "gc/shared/taskqueue.inline.hpp"
50 #include "gc/shared/vmGCOperations.hpp"
51 #include "memory/allocation.hpp"
52 #include "memory/resourceArea.hpp"
53 #include "oops/oop.inline.hpp"
54 #include "runtime/atomic.inline.hpp"
55 #include "runtime/handles.inline.hpp"
56 #include "runtime/java.hpp"
57 #include "runtime/prefetch.inline.hpp"
58 #include "services/memTracker.hpp"
59
60 // Concurrent marking bit map wrapper
61
62 CMBitMapRO::CMBitMapRO(int shifter) :
63 _bm(),
64 _shifter(shifter) {
65 _bmStartWord = 0;
66 _bmWordSize = 0;
67 }
68
69 HeapWord* CMBitMapRO::getNextMarkedWordAddress(const HeapWord* addr,
70 const HeapWord* limit) const {
71 // First we must round addr *up* to a possible object boundary.
72 addr = (HeapWord*)align_size_up((intptr_t)addr,
73 HeapWordSize << _shifter);
74 size_t addrOffset = heapWordToOffset(addr);
75 if (limit == NULL) {
76 limit = _bmStartWord + _bmWordSize;
77 }
78 size_t limitOffset = heapWordToOffset(limit);
79 size_t nextOffset = _bm.get_next_one_offset(addrOffset, limitOffset);
80 HeapWord* nextAddr = offsetToHeapWord(nextOffset);
81 assert(nextAddr >= addr, "get_next_one postcondition");
82 assert(nextAddr == limit || isMarked(nextAddr),
83 "get_next_one postcondition");
84 return nextAddr;
85 }
86
87 HeapWord* CMBitMapRO::getNextUnmarkedWordAddress(const HeapWord* addr,
88 const HeapWord* limit) const {
89 size_t addrOffset = heapWordToOffset(addr);
90 if (limit == NULL) {
91 limit = _bmStartWord + _bmWordSize;
92 }
93 size_t limitOffset = heapWordToOffset(limit);
94 size_t nextOffset = _bm.get_next_zero_offset(addrOffset, limitOffset);
95 HeapWord* nextAddr = offsetToHeapWord(nextOffset);
96 assert(nextAddr >= addr, "get_next_one postcondition");
97 assert(nextAddr == limit || !isMarked(nextAddr),
98 "get_next_one postcondition");
99 return nextAddr;
100 }
101
102 int CMBitMapRO::heapWordDiffToOffsetDiff(size_t diff) const {
103 assert((diff & ((1 << _shifter) - 1)) == 0, "argument check");
104 return (int) (diff >> _shifter);
105 }
106
107 #ifndef PRODUCT
108 bool CMBitMapRO::covers(MemRegion heap_rs) const {
109 // assert(_bm.map() == _virtual_space.low(), "map inconsistency");
110 assert(((size_t)_bm.size() * ((size_t)1 << _shifter)) == _bmWordSize,
111 "size inconsistency");
112 return _bmStartWord == (HeapWord*)(heap_rs.start()) &&
113 _bmWordSize == heap_rs.word_size();
114 }
115 #endif
116
117 void CMBitMapRO::print_on_error(outputStream* st, const char* prefix) const {
118 _bm.print_on_error(st, prefix);
119 }
120
121 size_t CMBitMap::compute_size(size_t heap_size) {
122 return ReservedSpace::allocation_align_size_up(heap_size / mark_distance());
123 }
124
125 size_t CMBitMap::mark_distance() {
126 return MinObjAlignmentInBytes * BitsPerByte;
127 }
128
129 void CMBitMap::initialize(MemRegion heap, G1RegionToSpaceMapper* storage) {
130 _bmStartWord = heap.start();
131 _bmWordSize = heap.word_size();
132
133 _bm.set_map((BitMap::bm_word_t*) storage->reserved().start());
134 _bm.set_size(_bmWordSize >> _shifter);
135
136 storage->set_mapping_changed_listener(&_listener);
137 }
138
139 void CMBitMapMappingChangedListener::on_commit(uint start_region, size_t num_regions, bool zero_filled) {
140 if (zero_filled) {
141 return;
142 }
143 // We need to clear the bitmap on commit, removing any existing information.
144 MemRegion mr(G1CollectedHeap::heap()->bottom_addr_for_region(start_region), num_regions * HeapRegion::GrainWords);
145 _bm->clearRange(mr);
146 }
147
148 // Closure used for clearing the given mark bitmap.
149 class ClearBitmapHRClosure : public HeapRegionClosure {
150 private:
151 ConcurrentMark* _cm;
152 CMBitMap* _bitmap;
153 bool _may_yield; // The closure may yield during iteration. If yielded, abort the iteration.
154 public:
155 ClearBitmapHRClosure(ConcurrentMark* cm, CMBitMap* bitmap, bool may_yield) : HeapRegionClosure(), _cm(cm), _bitmap(bitmap), _may_yield(may_yield) {
156 assert(!may_yield || cm != NULL, "CM must be non-NULL if this closure is expected to yield.");
157 }
158
159 virtual bool doHeapRegion(HeapRegion* r) {
160 size_t const chunk_size_in_words = M / HeapWordSize;
161
162 HeapWord* cur = r->bottom();
163 HeapWord* const end = r->end();
164
165 while (cur < end) {
166 MemRegion mr(cur, MIN2(cur + chunk_size_in_words, end));
167 _bitmap->clearRange(mr);
168
169 cur += chunk_size_in_words;
170
171 // Abort iteration if after yielding the marking has been aborted.
172 if (_may_yield && _cm->do_yield_check() && _cm->has_aborted()) {
173 return true;
174 }
175 // Repeat the asserts from before the start of the closure. We will do them
176 // as asserts here to minimize their overhead on the product. However, we
177 // will have them as guarantees at the beginning / end of the bitmap
178 // clearing to get some checking in the product.
179 assert(!_may_yield || _cm->cmThread()->during_cycle(), "invariant");
180 assert(!_may_yield || !G1CollectedHeap::heap()->mark_in_progress(), "invariant");
181 }
182
183 return false;
184 }
185 };
186
187 class ParClearNextMarkBitmapTask : public AbstractGangTask {
188 ClearBitmapHRClosure* _cl;
189 HeapRegionClaimer _hrclaimer;
190 bool _suspendible; // If the task is suspendible, workers must join the STS.
191
192 public:
193 ParClearNextMarkBitmapTask(ClearBitmapHRClosure *cl, uint n_workers, bool suspendible) :
194 _cl(cl), _suspendible(suspendible), AbstractGangTask("Parallel Clear Bitmap Task"), _hrclaimer(n_workers) {}
195
196 void work(uint worker_id) {
197 SuspendibleThreadSetJoiner sts_join(_suspendible);
198 G1CollectedHeap::heap()->heap_region_par_iterate(_cl, worker_id, &_hrclaimer, true);
199 }
200 };
201
202 void CMBitMap::clearAll() {
203 G1CollectedHeap* g1h = G1CollectedHeap::heap();
204 ClearBitmapHRClosure cl(NULL, this, false /* may_yield */);
205 uint n_workers = g1h->workers()->active_workers();
206 ParClearNextMarkBitmapTask task(&cl, n_workers, false);
207 g1h->workers()->run_task(&task);
208 guarantee(cl.complete(), "Must have completed iteration.");
209 return;
210 }
211
212 void CMBitMap::markRange(MemRegion mr) {
213 mr.intersection(MemRegion(_bmStartWord, _bmWordSize));
214 assert(!mr.is_empty(), "unexpected empty region");
215 assert((offsetToHeapWord(heapWordToOffset(mr.end())) ==
216 ((HeapWord *) mr.end())),
217 "markRange memory region end is not card aligned");
218 // convert address range into offset range
219 _bm.at_put_range(heapWordToOffset(mr.start()),
220 heapWordToOffset(mr.end()), true);
221 }
222
223 void CMBitMap::clearRange(MemRegion mr) {
224 mr.intersection(MemRegion(_bmStartWord, _bmWordSize));
225 assert(!mr.is_empty(), "unexpected empty region");
226 // convert address range into offset range
227 _bm.at_put_range(heapWordToOffset(mr.start()),
228 heapWordToOffset(mr.end()), false);
229 }
230
231 MemRegion CMBitMap::getAndClearMarkedRegion(HeapWord* addr,
232 HeapWord* end_addr) {
233 HeapWord* start = getNextMarkedWordAddress(addr);
234 start = MIN2(start, end_addr);
235 HeapWord* end = getNextUnmarkedWordAddress(start);
236 end = MIN2(end, end_addr);
237 assert(start <= end, "Consistency check");
238 MemRegion mr(start, end);
239 if (!mr.is_empty()) {
240 clearRange(mr);
241 }
242 return mr;
243 }
244
245 CMMarkStack::CMMarkStack(ConcurrentMark* cm) :
246 _base(NULL), _cm(cm)
247 #ifdef ASSERT
248 , _drain_in_progress(false)
249 , _drain_in_progress_yields(false)
250 #endif
251 {}
252
253 bool CMMarkStack::allocate(size_t capacity) {
254 // allocate a stack of the requisite depth
255 ReservedSpace rs(ReservedSpace::allocation_align_size_up(capacity * sizeof(oop)));
256 if (!rs.is_reserved()) {
257 warning("ConcurrentMark MarkStack allocation failure");
258 return false;
259 }
260 MemTracker::record_virtual_memory_type((address)rs.base(), mtGC);
261 if (!_virtual_space.initialize(rs, rs.size())) {
262 warning("ConcurrentMark MarkStack backing store failure");
263 // Release the virtual memory reserved for the marking stack
264 rs.release();
265 return false;
266 }
267 assert(_virtual_space.committed_size() == rs.size(),
268 "Didn't reserve backing store for all of ConcurrentMark stack?");
269 _base = (oop*) _virtual_space.low();
270 setEmpty();
271 _capacity = (jint) capacity;
272 _saved_index = -1;
273 _should_expand = false;
274 return true;
275 }
276
277 void CMMarkStack::expand() {
278 // Called, during remark, if we've overflown the marking stack during marking.
279 assert(isEmpty(), "stack should been emptied while handling overflow");
280 assert(_capacity <= (jint) MarkStackSizeMax, "stack bigger than permitted");
281 // Clear expansion flag
282 _should_expand = false;
283 if (_capacity == (jint) MarkStackSizeMax) {
284 if (PrintGCDetails && Verbose) {
285 gclog_or_tty->print_cr(" (benign) Can't expand marking stack capacity, at max size limit");
286 }
287 return;
288 }
289 // Double capacity if possible
290 jint new_capacity = MIN2(_capacity*2, (jint) MarkStackSizeMax);
291 // Do not give up existing stack until we have managed to
292 // get the double capacity that we desired.
293 ReservedSpace rs(ReservedSpace::allocation_align_size_up(new_capacity *
294 sizeof(oop)));
295 if (rs.is_reserved()) {
296 // Release the backing store associated with old stack
297 _virtual_space.release();
298 // Reinitialize virtual space for new stack
299 if (!_virtual_space.initialize(rs, rs.size())) {
300 fatal("Not enough swap for expanded marking stack capacity");
301 }
302 _base = (oop*)(_virtual_space.low());
303 _index = 0;
304 _capacity = new_capacity;
305 } else {
306 if (PrintGCDetails && Verbose) {
307 // Failed to double capacity, continue;
308 gclog_or_tty->print(" (benign) Failed to expand marking stack capacity from "
309 SIZE_FORMAT"K to " SIZE_FORMAT"K",
310 _capacity / K, new_capacity / K);
311 }
312 }
313 }
314
315 void CMMarkStack::set_should_expand() {
316 // If we're resetting the marking state because of an
317 // marking stack overflow, record that we should, if
318 // possible, expand the stack.
319 _should_expand = _cm->has_overflown();
320 }
321
322 CMMarkStack::~CMMarkStack() {
323 if (_base != NULL) {
324 _base = NULL;
325 _virtual_space.release();
326 }
327 }
328
329 void CMMarkStack::par_push_arr(oop* ptr_arr, int n) {
330 MutexLockerEx x(ParGCRareEvent_lock, Mutex::_no_safepoint_check_flag);
331 jint start = _index;
332 jint next_index = start + n;
333 if (next_index > _capacity) {
334 _overflow = true;
335 return;
336 }
337 // Otherwise.
338 _index = next_index;
339 for (int i = 0; i < n; i++) {
340 int ind = start + i;
341 assert(ind < _capacity, "By overflow test above.");
342 _base[ind] = ptr_arr[i];
343 }
344 }
345
346 bool CMMarkStack::par_pop_arr(oop* ptr_arr, int max, int* n) {
347 MutexLockerEx x(ParGCRareEvent_lock, Mutex::_no_safepoint_check_flag);
348 jint index = _index;
349 if (index == 0) {
350 *n = 0;
351 return false;
352 } else {
353 int k = MIN2(max, index);
354 jint new_ind = index - k;
355 for (int j = 0; j < k; j++) {
356 ptr_arr[j] = _base[new_ind + j];
357 }
358 _index = new_ind;
359 *n = k;
360 return true;
361 }
362 }
363
364 template<class OopClosureClass>
365 bool CMMarkStack::drain(OopClosureClass* cl, CMBitMap* bm, bool yield_after) {
366 assert(!_drain_in_progress || !_drain_in_progress_yields || yield_after
367 || SafepointSynchronize::is_at_safepoint(),
368 "Drain recursion must be yield-safe.");
369 bool res = true;
370 debug_only(_drain_in_progress = true);
371 debug_only(_drain_in_progress_yields = yield_after);
372 while (!isEmpty()) {
373 oop newOop = pop();
374 assert(G1CollectedHeap::heap()->is_in_reserved(newOop), "Bad pop");
375 assert(newOop->is_oop(), "Expected an oop");
376 assert(bm == NULL || bm->isMarked((HeapWord*)newOop),
377 "only grey objects on this stack");
378 newOop->oop_iterate(cl);
379 if (yield_after && _cm->do_yield_check()) {
380 res = false;
381 break;
382 }
383 }
384 debug_only(_drain_in_progress = false);
385 return res;
386 }
387
388 void CMMarkStack::note_start_of_gc() {
389 assert(_saved_index == -1,
390 "note_start_of_gc()/end_of_gc() bracketed incorrectly");
391 _saved_index = _index;
392 }
393
394 void CMMarkStack::note_end_of_gc() {
395 // This is intentionally a guarantee, instead of an assert. If we
396 // accidentally add something to the mark stack during GC, it
397 // will be a correctness issue so it's better if we crash. we'll
398 // only check this once per GC anyway, so it won't be a performance
399 // issue in any way.
400 guarantee(_saved_index == _index,
401 err_msg("saved index: %d index: %d", _saved_index, _index));
402 _saved_index = -1;
403 }
404
405 void CMMarkStack::oops_do(OopClosure* f) {
406 assert(_saved_index == _index,
407 err_msg("saved index: %d index: %d", _saved_index, _index));
408 for (int i = 0; i < _index; i += 1) {
409 f->do_oop(&_base[i]);
410 }
411 }
412
413 CMRootRegions::CMRootRegions() :
414 _young_list(NULL), _cm(NULL), _scan_in_progress(false),
415 _should_abort(false), _next_survivor(NULL) { }
416
417 void CMRootRegions::init(G1CollectedHeap* g1h, ConcurrentMark* cm) {
418 _young_list = g1h->young_list();
419 _cm = cm;
420 }
421
422 void CMRootRegions::prepare_for_scan() {
423 assert(!scan_in_progress(), "pre-condition");
424
425 // Currently, only survivors can be root regions.
426 assert(_next_survivor == NULL, "pre-condition");
427 _next_survivor = _young_list->first_survivor_region();
428 _scan_in_progress = (_next_survivor != NULL);
429 _should_abort = false;
430 }
431
432 HeapRegion* CMRootRegions::claim_next() {
433 if (_should_abort) {
434 // If someone has set the should_abort flag, we return NULL to
435 // force the caller to bail out of their loop.
436 return NULL;
437 }
438
439 // Currently, only survivors can be root regions.
440 HeapRegion* res = _next_survivor;
441 if (res != NULL) {
442 MutexLockerEx x(RootRegionScan_lock, Mutex::_no_safepoint_check_flag);
443 // Read it again in case it changed while we were waiting for the lock.
444 res = _next_survivor;
445 if (res != NULL) {
446 if (res == _young_list->last_survivor_region()) {
447 // We just claimed the last survivor so store NULL to indicate
448 // that we're done.
449 _next_survivor = NULL;
450 } else {
451 _next_survivor = res->get_next_young_region();
452 }
453 } else {
454 // Someone else claimed the last survivor while we were trying
455 // to take the lock so nothing else to do.
456 }
457 }
458 assert(res == NULL || res->is_survivor(), "post-condition");
459
460 return res;
461 }
462
463 void CMRootRegions::scan_finished() {
464 assert(scan_in_progress(), "pre-condition");
465
466 // Currently, only survivors can be root regions.
467 if (!_should_abort) {
468 assert(_next_survivor == NULL, "we should have claimed all survivors");
469 }
470 _next_survivor = NULL;
471
472 {
473 MutexLockerEx x(RootRegionScan_lock, Mutex::_no_safepoint_check_flag);
474 _scan_in_progress = false;
475 RootRegionScan_lock->notify_all();
476 }
477 }
478
479 bool CMRootRegions::wait_until_scan_finished() {
480 if (!scan_in_progress()) return false;
481
482 {
483 MutexLockerEx x(RootRegionScan_lock, Mutex::_no_safepoint_check_flag);
484 while (scan_in_progress()) {
485 RootRegionScan_lock->wait(Mutex::_no_safepoint_check_flag);
486 }
487 }
488 return true;
489 }
490
491 #ifdef _MSC_VER // the use of 'this' below gets a warning, make it go away
492 #pragma warning( disable:4355 ) // 'this' : used in base member initializer list
493 #endif // _MSC_VER
494
495 uint ConcurrentMark::scale_parallel_threads(uint n_par_threads) {
496 return MAX2((n_par_threads + 2) / 4, 1U);
497 }
498
499 ConcurrentMark::ConcurrentMark(G1CollectedHeap* g1h, G1RegionToSpaceMapper* prev_bitmap_storage, G1RegionToSpaceMapper* next_bitmap_storage) :
500 _g1h(g1h),
501 _markBitMap1(),
502 _markBitMap2(),
503 _parallel_marking_threads(0),
504 _max_parallel_marking_threads(0),
505 _sleep_factor(0.0),
506 _marking_task_overhead(1.0),
507 _cleanup_sleep_factor(0.0),
508 _cleanup_task_overhead(1.0),
509 _cleanup_list("Cleanup List"),
510 _region_bm((BitMap::idx_t)(g1h->max_regions()), false /* in_resource_area*/),
511 _card_bm((g1h->reserved_region().byte_size() + CardTableModRefBS::card_size - 1) >>
512 CardTableModRefBS::card_shift,
513 false /* in_resource_area*/),
514
515 _prevMarkBitMap(&_markBitMap1),
516 _nextMarkBitMap(&_markBitMap2),
517
518 _markStack(this),
519 // _finger set in set_non_marking_state
520
521 _max_worker_id((uint)ParallelGCThreads),
522 // _active_tasks set in set_non_marking_state
523 // _tasks set inside the constructor
524 _task_queues(new CMTaskQueueSet((int) _max_worker_id)),
525 _terminator(ParallelTaskTerminator((int) _max_worker_id, _task_queues)),
526
527 _has_overflown(false),
528 _concurrent(false),
529 _has_aborted(false),
530 _aborted_gc_id(GCId::undefined()),
531 _restart_for_overflow(false),
532 _concurrent_marking_in_progress(false),
533
534 // _verbose_level set below
535
536 _init_times(),
537 _remark_times(), _remark_mark_times(), _remark_weak_ref_times(),
538 _cleanup_times(),
539 _total_counting_time(0.0),
540 _total_rs_scrub_time(0.0),
541
542 _parallel_workers(NULL),
543
544 _count_card_bitmaps(NULL),
545 _count_marked_bytes(NULL),
546 _completed_initialization(false) {
547 CMVerboseLevel verbose_level = (CMVerboseLevel) G1MarkingVerboseLevel;
548 if (verbose_level < no_verbose) {
549 verbose_level = no_verbose;
550 }
551 if (verbose_level > high_verbose) {
552 verbose_level = high_verbose;
553 }
554 _verbose_level = verbose_level;
555
556 if (verbose_low()) {
557 gclog_or_tty->print_cr("[global] init, heap start = "PTR_FORMAT", "
558 "heap end = " PTR_FORMAT, p2i(_heap_start), p2i(_heap_end));
559 }
560
561 _markBitMap1.initialize(g1h->reserved_region(), prev_bitmap_storage);
562 _markBitMap2.initialize(g1h->reserved_region(), next_bitmap_storage);
563
564 // Create & start a ConcurrentMark thread.
565 _cmThread = new ConcurrentMarkThread(this);
566 assert(cmThread() != NULL, "CM Thread should have been created");
567 assert(cmThread()->cm() != NULL, "CM Thread should refer to this cm");
568 if (_cmThread->osthread() == NULL) {
569 vm_shutdown_during_initialization("Could not create ConcurrentMarkThread");
570 }
571
572 assert(CGC_lock != NULL, "Where's the CGC_lock?");
573 assert(_markBitMap1.covers(g1h->reserved_region()), "_markBitMap1 inconsistency");
574 assert(_markBitMap2.covers(g1h->reserved_region()), "_markBitMap2 inconsistency");
575
576 SATBMarkQueueSet& satb_qs = JavaThread::satb_mark_queue_set();
577 satb_qs.set_buffer_size(G1SATBBufferSize);
578
579 _root_regions.init(_g1h, this);
580
581 if (ConcGCThreads > ParallelGCThreads) {
582 warning("Can't have more ConcGCThreads (" UINTX_FORMAT ") "
583 "than ParallelGCThreads (" UINTX_FORMAT ").",
584 ConcGCThreads, ParallelGCThreads);
585 return;
586 }
587 if (!FLAG_IS_DEFAULT(ConcGCThreads) && ConcGCThreads > 0) {
588 // Note: ConcGCThreads has precedence over G1MarkingOverheadPercent
589 // if both are set
590 _sleep_factor = 0.0;
591 _marking_task_overhead = 1.0;
592 } else if (G1MarkingOverheadPercent > 0) {
593 // We will calculate the number of parallel marking threads based
594 // on a target overhead with respect to the soft real-time goal
595 double marking_overhead = (double) G1MarkingOverheadPercent / 100.0;
596 double overall_cm_overhead =
597 (double) MaxGCPauseMillis * marking_overhead /
598 (double) GCPauseIntervalMillis;
599 double cpu_ratio = 1.0 / (double) os::processor_count();
600 double marking_thread_num = ceil(overall_cm_overhead / cpu_ratio);
601 double marking_task_overhead =
602 overall_cm_overhead / marking_thread_num *
603 (double) os::processor_count();
604 double sleep_factor =
605 (1.0 - marking_task_overhead) / marking_task_overhead;
606
607 FLAG_SET_ERGO(uintx, ConcGCThreads, (uint) marking_thread_num);
608 _sleep_factor = sleep_factor;
609 _marking_task_overhead = marking_task_overhead;
610 } else {
611 // Calculate the number of parallel marking threads by scaling
612 // the number of parallel GC threads.
613 uint marking_thread_num = scale_parallel_threads((uint) ParallelGCThreads);
614 FLAG_SET_ERGO(uintx, ConcGCThreads, marking_thread_num);
615 _sleep_factor = 0.0;
616 _marking_task_overhead = 1.0;
617 }
618
619 assert(ConcGCThreads > 0, "Should have been set");
620 _parallel_marking_threads = (uint) ConcGCThreads;
621 _max_parallel_marking_threads = _parallel_marking_threads;
622
623 if (parallel_marking_threads() > 1) {
624 _cleanup_task_overhead = 1.0;
625 } else {
626 _cleanup_task_overhead = marking_task_overhead();
627 }
628 _cleanup_sleep_factor =
629 (1.0 - cleanup_task_overhead()) / cleanup_task_overhead();
630
631 #if 0
632 gclog_or_tty->print_cr("Marking Threads %d", parallel_marking_threads());
633 gclog_or_tty->print_cr("CM Marking Task Overhead %1.4lf", marking_task_overhead());
634 gclog_or_tty->print_cr("CM Sleep Factor %1.4lf", sleep_factor());
635 gclog_or_tty->print_cr("CL Marking Task Overhead %1.4lf", cleanup_task_overhead());
636 gclog_or_tty->print_cr("CL Sleep Factor %1.4lf", cleanup_sleep_factor());
637 #endif
638
639 _parallel_workers = new FlexibleWorkGang("G1 Marker",
640 _max_parallel_marking_threads, false, true);
641 if (_parallel_workers == NULL) {
642 vm_exit_during_initialization("Failed necessary allocation.");
643 } else {
644 _parallel_workers->initialize_workers();
645 }
646
647 if (FLAG_IS_DEFAULT(MarkStackSize)) {
648 size_t mark_stack_size =
649 MIN2(MarkStackSizeMax,
650 MAX2(MarkStackSize, (size_t) (parallel_marking_threads() * TASKQUEUE_SIZE)));
651 // Verify that the calculated value for MarkStackSize is in range.
652 // It would be nice to use the private utility routine from Arguments.
653 if (!(mark_stack_size >= 1 && mark_stack_size <= MarkStackSizeMax)) {
654 warning("Invalid value calculated for MarkStackSize (" SIZE_FORMAT "): "
655 "must be between 1 and " SIZE_FORMAT,
656 mark_stack_size, MarkStackSizeMax);
657 return;
658 }
659 FLAG_SET_ERGO(size_t, MarkStackSize, mark_stack_size);
660 } else {
661 // Verify MarkStackSize is in range.
662 if (FLAG_IS_CMDLINE(MarkStackSize)) {
663 if (FLAG_IS_DEFAULT(MarkStackSizeMax)) {
664 if (!(MarkStackSize >= 1 && MarkStackSize <= MarkStackSizeMax)) {
665 warning("Invalid value specified for MarkStackSize (" SIZE_FORMAT "): "
666 "must be between 1 and " SIZE_FORMAT,
667 MarkStackSize, MarkStackSizeMax);
668 return;
669 }
670 } else if (FLAG_IS_CMDLINE(MarkStackSizeMax)) {
671 if (!(MarkStackSize >= 1 && MarkStackSize <= MarkStackSizeMax)) {
672 warning("Invalid value specified for MarkStackSize (" SIZE_FORMAT ")"
673 " or for MarkStackSizeMax (" SIZE_FORMAT ")",
674 MarkStackSize, MarkStackSizeMax);
675 return;
676 }
677 }
678 }
679 }
680
681 if (!_markStack.allocate(MarkStackSize)) {
682 warning("Failed to allocate CM marking stack");
683 return;
684 }
685
686 _tasks = NEW_C_HEAP_ARRAY(CMTask*, _max_worker_id, mtGC);
687 _accum_task_vtime = NEW_C_HEAP_ARRAY(double, _max_worker_id, mtGC);
688
689 _count_card_bitmaps = NEW_C_HEAP_ARRAY(BitMap, _max_worker_id, mtGC);
690 _count_marked_bytes = NEW_C_HEAP_ARRAY(size_t*, _max_worker_id, mtGC);
691
692 BitMap::idx_t card_bm_size = _card_bm.size();
693
694 // so that the assertion in MarkingTaskQueue::task_queue doesn't fail
695 _active_tasks = _max_worker_id;
696
697 uint max_regions = _g1h->max_regions();
698 for (uint i = 0; i < _max_worker_id; ++i) {
699 CMTaskQueue* task_queue = new CMTaskQueue();
700 task_queue->initialize();
701 _task_queues->register_queue(i, task_queue);
702
703 _count_card_bitmaps[i] = BitMap(card_bm_size, false);
704 _count_marked_bytes[i] = NEW_C_HEAP_ARRAY(size_t, max_regions, mtGC);
705
706 _tasks[i] = new CMTask(i, this,
707 _count_marked_bytes[i],
708 &_count_card_bitmaps[i],
709 task_queue, _task_queues);
710
711 _accum_task_vtime[i] = 0.0;
712 }
713
714 // Calculate the card number for the bottom of the heap. Used
715 // in biasing indexes into the accounting card bitmaps.
716 _heap_bottom_card_num =
717 intptr_t(uintptr_t(_g1h->reserved_region().start()) >>
718 CardTableModRefBS::card_shift);
719
720 // Clear all the liveness counting data
721 clear_all_count_data();
722
723 // so that the call below can read a sensible value
724 _heap_start = g1h->reserved_region().start();
725 set_non_marking_state();
726 _completed_initialization = true;
727 }
728
729 void ConcurrentMark::reset() {
730 // Starting values for these two. This should be called in a STW
731 // phase.
732 MemRegion reserved = _g1h->g1_reserved();
733 _heap_start = reserved.start();
734 _heap_end = reserved.end();
735
736 // Separated the asserts so that we know which one fires.
737 assert(_heap_start != NULL, "heap bounds should look ok");
738 assert(_heap_end != NULL, "heap bounds should look ok");
739 assert(_heap_start < _heap_end, "heap bounds should look ok");
740
741 // Reset all the marking data structures and any necessary flags
742 reset_marking_state();
743
744 if (verbose_low()) {
745 gclog_or_tty->print_cr("[global] resetting");
746 }
747
748 // We do reset all of them, since different phases will use
749 // different number of active threads. So, it's easiest to have all
750 // of them ready.
751 for (uint i = 0; i < _max_worker_id; ++i) {
752 _tasks[i]->reset(_nextMarkBitMap);
753 }
754
755 // we need this to make sure that the flag is on during the evac
756 // pause with initial mark piggy-backed
757 set_concurrent_marking_in_progress();
758 }
759
760
761 void ConcurrentMark::reset_marking_state(bool clear_overflow) {
762 _markStack.set_should_expand();
763 _markStack.setEmpty(); // Also clears the _markStack overflow flag
764 if (clear_overflow) {
765 clear_has_overflown();
766 } else {
767 assert(has_overflown(), "pre-condition");
768 }
769 _finger = _heap_start;
770
771 for (uint i = 0; i < _max_worker_id; ++i) {
772 CMTaskQueue* queue = _task_queues->queue(i);
773 queue->set_empty();
774 }
775 }
776
777 void ConcurrentMark::set_concurrency(uint active_tasks) {
778 assert(active_tasks <= _max_worker_id, "we should not have more");
779
780 _active_tasks = active_tasks;
781 // Need to update the three data structures below according to the
782 // number of active threads for this phase.
783 _terminator = ParallelTaskTerminator((int) active_tasks, _task_queues);
784 _first_overflow_barrier_sync.set_n_workers((int) active_tasks);
785 _second_overflow_barrier_sync.set_n_workers((int) active_tasks);
786 }
787
788 void ConcurrentMark::set_concurrency_and_phase(uint active_tasks, bool concurrent) {
789 set_concurrency(active_tasks);
790
791 _concurrent = concurrent;
792 // We propagate this to all tasks, not just the active ones.
793 for (uint i = 0; i < _max_worker_id; ++i)
794 _tasks[i]->set_concurrent(concurrent);
795
796 if (concurrent) {
797 set_concurrent_marking_in_progress();
798 } else {
799 // We currently assume that the concurrent flag has been set to
800 // false before we start remark. At this point we should also be
801 // in a STW phase.
802 assert(!concurrent_marking_in_progress(), "invariant");
803 assert(out_of_regions(),
804 err_msg("only way to get here: _finger: "PTR_FORMAT", _heap_end: "PTR_FORMAT,
805 p2i(_finger), p2i(_heap_end)));
806 }
807 }
808
809 void ConcurrentMark::set_non_marking_state() {
810 // We set the global marking state to some default values when we're
811 // not doing marking.
812 reset_marking_state();
813 _active_tasks = 0;
814 clear_concurrent_marking_in_progress();
815 }
816
817 ConcurrentMark::~ConcurrentMark() {
818 // The ConcurrentMark instance is never freed.
819 ShouldNotReachHere();
820 }
821
822 void ConcurrentMark::clearNextBitmap() {
823 G1CollectedHeap* g1h = G1CollectedHeap::heap();
824
825 // Make sure that the concurrent mark thread looks to still be in
826 // the current cycle.
827 guarantee(cmThread()->during_cycle(), "invariant");
828
829 // We are finishing up the current cycle by clearing the next
830 // marking bitmap and getting it ready for the next cycle. During
831 // this time no other cycle can start. So, let's make sure that this
832 // is the case.
833 guarantee(!g1h->mark_in_progress(), "invariant");
834
835 ClearBitmapHRClosure cl(this, _nextMarkBitMap, true /* may_yield */);
836 ParClearNextMarkBitmapTask task(&cl, parallel_marking_threads(), true);
837 _parallel_workers->run_task(&task);
838
839 // Clear the liveness counting data. If the marking has been aborted, the abort()
840 // call already did that.
841 if (cl.complete()) {
842 clear_all_count_data();
843 }
844
845 // Repeat the asserts from above.
846 guarantee(cmThread()->during_cycle(), "invariant");
847 guarantee(!g1h->mark_in_progress(), "invariant");
848 }
849
850 class CheckBitmapClearHRClosure : public HeapRegionClosure {
851 CMBitMap* _bitmap;
852 bool _error;
853 public:
854 CheckBitmapClearHRClosure(CMBitMap* bitmap) : _bitmap(bitmap) {
855 }
856
857 virtual bool doHeapRegion(HeapRegion* r) {
858 // This closure can be called concurrently to the mutator, so we must make sure
859 // that the result of the getNextMarkedWordAddress() call is compared to the
860 // value passed to it as limit to detect any found bits.
861 // We can use the region's orig_end() for the limit and the comparison value
862 // as it always contains the "real" end of the region that never changes and
863 // has no side effects.
864 // Due to the latter, there can also be no problem with the compiler generating
865 // reloads of the orig_end() call.
866 HeapWord* end = r->orig_end();
867 return _bitmap->getNextMarkedWordAddress(r->bottom(), end) != end;
868 }
869 };
870
871 bool ConcurrentMark::nextMarkBitmapIsClear() {
872 CheckBitmapClearHRClosure cl(_nextMarkBitMap);
873 _g1h->heap_region_iterate(&cl);
874 return cl.complete();
875 }
876
877 class NoteStartOfMarkHRClosure: public HeapRegionClosure {
878 public:
879 bool doHeapRegion(HeapRegion* r) {
880 if (!r->is_continues_humongous()) {
881 r->note_start_of_marking();
882 }
883 return false;
884 }
885 };
886
887 void ConcurrentMark::checkpointRootsInitialPre() {
888 G1CollectedHeap* g1h = G1CollectedHeap::heap();
889 G1CollectorPolicy* g1p = g1h->g1_policy();
890
891 _has_aborted = false;
892
893 // Initialize marking structures. This has to be done in a STW phase.
894 reset();
895
896 // For each region note start of marking.
897 NoteStartOfMarkHRClosure startcl;
898 g1h->heap_region_iterate(&startcl);
899 }
900
901
902 void ConcurrentMark::checkpointRootsInitialPost() {
903 G1CollectedHeap* g1h = G1CollectedHeap::heap();
904
905 // If we force an overflow during remark, the remark operation will
906 // actually abort and we'll restart concurrent marking. If we always
907 // force an overflow during remark we'll never actually complete the
908 // marking phase. So, we initialize this here, at the start of the
909 // cycle, so that at the remaining overflow number will decrease at
910 // every remark and we'll eventually not need to cause one.
911 force_overflow_stw()->init();
912
913 // Start Concurrent Marking weak-reference discovery.
914 ReferenceProcessor* rp = g1h->ref_processor_cm();
915 // enable ("weak") refs discovery
916 rp->enable_discovery();
917 rp->setup_policy(false); // snapshot the soft ref policy to be used in this cycle
918
919 SATBMarkQueueSet& satb_mq_set = JavaThread::satb_mark_queue_set();
920 // This is the start of the marking cycle, we're expected all
921 // threads to have SATB queues with active set to false.
922 satb_mq_set.set_active_all_threads(true, /* new active value */
923 false /* expected_active */);
924
925 _root_regions.prepare_for_scan();
926
927 // update_g1_committed() will be called at the end of an evac pause
928 // when marking is on. So, it's also called at the end of the
929 // initial-mark pause to update the heap end, if the heap expands
930 // during it. No need to call it here.
931 }
932
933 /*
934 * Notice that in the next two methods, we actually leave the STS
935 * during the barrier sync and join it immediately afterwards. If we
936 * do not do this, the following deadlock can occur: one thread could
937 * be in the barrier sync code, waiting for the other thread to also
938 * sync up, whereas another one could be trying to yield, while also
939 * waiting for the other threads to sync up too.
940 *
941 * Note, however, that this code is also used during remark and in
942 * this case we should not attempt to leave / enter the STS, otherwise
943 * we'll either hit an assert (debug / fastdebug) or deadlock
944 * (product). So we should only leave / enter the STS if we are
945 * operating concurrently.
946 *
947 * Because the thread that does the sync barrier has left the STS, it
948 * is possible to be suspended for a Full GC or an evacuation pause
949 * could occur. This is actually safe, since the entering the sync
950 * barrier is one of the last things do_marking_step() does, and it
951 * doesn't manipulate any data structures afterwards.
952 */
953
954 void ConcurrentMark::enter_first_sync_barrier(uint worker_id) {
955 bool barrier_aborted;
956
957 if (verbose_low()) {
958 gclog_or_tty->print_cr("[%u] entering first barrier", worker_id);
959 }
960
961 {
962 SuspendibleThreadSetLeaver sts_leave(concurrent());
963 barrier_aborted = !_first_overflow_barrier_sync.enter();
964 }
965
966 // at this point everyone should have synced up and not be doing any
967 // more work
968
969 if (verbose_low()) {
970 if (barrier_aborted) {
971 gclog_or_tty->print_cr("[%u] aborted first barrier", worker_id);
972 } else {
973 gclog_or_tty->print_cr("[%u] leaving first barrier", worker_id);
974 }
975 }
976
977 if (barrier_aborted) {
978 // If the barrier aborted we ignore the overflow condition and
979 // just abort the whole marking phase as quickly as possible.
980 return;
981 }
982
983 // If we're executing the concurrent phase of marking, reset the marking
984 // state; otherwise the marking state is reset after reference processing,
985 // during the remark pause.
986 // If we reset here as a result of an overflow during the remark we will
987 // see assertion failures from any subsequent set_concurrency_and_phase()
988 // calls.
989 if (concurrent()) {
990 // let the task associated with with worker 0 do this
991 if (worker_id == 0) {
992 // task 0 is responsible for clearing the global data structures
993 // We should be here because of an overflow. During STW we should
994 // not clear the overflow flag since we rely on it being true when
995 // we exit this method to abort the pause and restart concurrent
996 // marking.
997 reset_marking_state(true /* clear_overflow */);
998 force_overflow()->update();
999
1000 if (G1Log::fine()) {
1001 gclog_or_tty->gclog_stamp(concurrent_gc_id());
1002 gclog_or_tty->print_cr("[GC concurrent-mark-reset-for-overflow]");
1003 }
1004 }
1005 }
1006
1007 // after this, each task should reset its own data structures then
1008 // then go into the second barrier
1009 }
1010
1011 void ConcurrentMark::enter_second_sync_barrier(uint worker_id) {
1012 bool barrier_aborted;
1013
1014 if (verbose_low()) {
1015 gclog_or_tty->print_cr("[%u] entering second barrier", worker_id);
1016 }
1017
1018 {
1019 SuspendibleThreadSetLeaver sts_leave(concurrent());
1020 barrier_aborted = !_second_overflow_barrier_sync.enter();
1021 }
1022
1023 // at this point everything should be re-initialized and ready to go
1024
1025 if (verbose_low()) {
1026 if (barrier_aborted) {
1027 gclog_or_tty->print_cr("[%u] aborted second barrier", worker_id);
1028 } else {
1029 gclog_or_tty->print_cr("[%u] leaving second barrier", worker_id);
1030 }
1031 }
1032 }
1033
1034 #ifndef PRODUCT
1035 void ForceOverflowSettings::init() {
1036 _num_remaining = G1ConcMarkForceOverflow;
1037 _force = false;
1038 update();
1039 }
1040
1041 void ForceOverflowSettings::update() {
1042 if (_num_remaining > 0) {
1043 _num_remaining -= 1;
1044 _force = true;
1045 } else {
1046 _force = false;
1047 }
1048 }
1049
1050 bool ForceOverflowSettings::should_force() {
1051 if (_force) {
1052 _force = false;
1053 return true;
1054 } else {
1055 return false;
1056 }
1057 }
1058 #endif // !PRODUCT
1059
1060 class CMConcurrentMarkingTask: public AbstractGangTask {
1061 private:
1062 ConcurrentMark* _cm;
1063 ConcurrentMarkThread* _cmt;
1064
1065 public:
1066 void work(uint worker_id) {
1067 assert(Thread::current()->is_ConcurrentGC_thread(),
1068 "this should only be done by a conc GC thread");
1069 ResourceMark rm;
1070
1071 double start_vtime = os::elapsedVTime();
1072
1073 {
1074 SuspendibleThreadSetJoiner sts_join;
1075
1076 assert(worker_id < _cm->active_tasks(), "invariant");
1077 CMTask* the_task = _cm->task(worker_id);
1078 the_task->record_start_time();
1079 if (!_cm->has_aborted()) {
1080 do {
1081 double start_vtime_sec = os::elapsedVTime();
1082 double mark_step_duration_ms = G1ConcMarkStepDurationMillis;
1083
1084 the_task->do_marking_step(mark_step_duration_ms,
1085 true /* do_termination */,
1086 false /* is_serial*/);
1087
1088 double end_vtime_sec = os::elapsedVTime();
1089 double elapsed_vtime_sec = end_vtime_sec - start_vtime_sec;
1090 _cm->clear_has_overflown();
1091
1092 _cm->do_yield_check(worker_id);
1093
1094 jlong sleep_time_ms;
1095 if (!_cm->has_aborted() && the_task->has_aborted()) {
1096 sleep_time_ms =
1097 (jlong) (elapsed_vtime_sec * _cm->sleep_factor() * 1000.0);
1098 {
1099 SuspendibleThreadSetLeaver sts_leave;
1100 os::sleep(Thread::current(), sleep_time_ms, false);
1101 }
1102 }
1103 } while (!_cm->has_aborted() && the_task->has_aborted());
1104 }
1105 the_task->record_end_time();
1106 guarantee(!the_task->has_aborted() || _cm->has_aborted(), "invariant");
1107 }
1108
1109 double end_vtime = os::elapsedVTime();
1110 _cm->update_accum_task_vtime(worker_id, end_vtime - start_vtime);
1111 }
1112
1113 CMConcurrentMarkingTask(ConcurrentMark* cm,
1114 ConcurrentMarkThread* cmt) :
1115 AbstractGangTask("Concurrent Mark"), _cm(cm), _cmt(cmt) { }
1116
1117 ~CMConcurrentMarkingTask() { }
1118 };
1119
1120 // Calculates the number of active workers for a concurrent
1121 // phase.
1122 uint ConcurrentMark::calc_parallel_marking_threads() {
1123 uint n_conc_workers = 0;
1124 if (!UseDynamicNumberOfGCThreads ||
1125 (!FLAG_IS_DEFAULT(ConcGCThreads) &&
1126 !ForceDynamicNumberOfGCThreads)) {
1127 n_conc_workers = max_parallel_marking_threads();
1128 } else {
1129 n_conc_workers =
1130 AdaptiveSizePolicy::calc_default_active_workers(
1131 max_parallel_marking_threads(),
1132 1, /* Minimum workers */
1133 parallel_marking_threads(),
1134 Threads::number_of_non_daemon_threads());
1135 // Don't scale down "n_conc_workers" by scale_parallel_threads() because
1136 // that scaling has already gone into "_max_parallel_marking_threads".
1137 }
1138 assert(n_conc_workers > 0, "Always need at least 1");
1139 return n_conc_workers;
1140 }
1141
1142 void ConcurrentMark::scanRootRegion(HeapRegion* hr, uint worker_id) {
1143 // Currently, only survivors can be root regions.
1144 assert(hr->next_top_at_mark_start() == hr->bottom(), "invariant");
1145 G1RootRegionScanClosure cl(_g1h, this, worker_id);
1146
1147 const uintx interval = PrefetchScanIntervalInBytes;
1148 HeapWord* curr = hr->bottom();
1149 const HeapWord* end = hr->top();
1150 while (curr < end) {
1151 Prefetch::read(curr, interval);
1152 oop obj = oop(curr);
1153 int size = obj->oop_iterate(&cl);
1154 assert(size == obj->size(), "sanity");
1155 curr += size;
1156 }
1157 }
1158
1159 class CMRootRegionScanTask : public AbstractGangTask {
1160 private:
1161 ConcurrentMark* _cm;
1162
1163 public:
1164 CMRootRegionScanTask(ConcurrentMark* cm) :
1165 AbstractGangTask("Root Region Scan"), _cm(cm) { }
1166
1167 void work(uint worker_id) {
1168 assert(Thread::current()->is_ConcurrentGC_thread(),
1169 "this should only be done by a conc GC thread");
1170
1171 CMRootRegions* root_regions = _cm->root_regions();
1172 HeapRegion* hr = root_regions->claim_next();
1173 while (hr != NULL) {
1174 _cm->scanRootRegion(hr, worker_id);
1175 hr = root_regions->claim_next();
1176 }
1177 }
1178 };
1179
1180 void ConcurrentMark::scanRootRegions() {
1181 // Start of concurrent marking.
1182 ClassLoaderDataGraph::clear_claimed_marks();
1183
1184 // scan_in_progress() will have been set to true only if there was
1185 // at least one root region to scan. So, if it's false, we
1186 // should not attempt to do any further work.
1187 if (root_regions()->scan_in_progress()) {
1188 _parallel_marking_threads = calc_parallel_marking_threads();
1189 assert(parallel_marking_threads() <= max_parallel_marking_threads(),
1190 "Maximum number of marking threads exceeded");
1191 uint active_workers = MAX2(1U, parallel_marking_threads());
1192
1193 CMRootRegionScanTask task(this);
1194 _parallel_workers->set_active_workers(active_workers);
1195 _parallel_workers->run_task(&task);
1196
1197 // It's possible that has_aborted() is true here without actually
1198 // aborting the survivor scan earlier. This is OK as it's
1199 // mainly used for sanity checking.
1200 root_regions()->scan_finished();
1201 }
1202 }
1203
1204 void ConcurrentMark::markFromRoots() {
1205 // we might be tempted to assert that:
1206 // assert(asynch == !SafepointSynchronize::is_at_safepoint(),
1207 // "inconsistent argument?");
1208 // However that wouldn't be right, because it's possible that
1209 // a safepoint is indeed in progress as a younger generation
1210 // stop-the-world GC happens even as we mark in this generation.
1211
1212 _restart_for_overflow = false;
1213 force_overflow_conc()->init();
1214
1215 // _g1h has _n_par_threads
1216 _parallel_marking_threads = calc_parallel_marking_threads();
1217 assert(parallel_marking_threads() <= max_parallel_marking_threads(),
1218 "Maximum number of marking threads exceeded");
1219
1220 uint active_workers = MAX2(1U, parallel_marking_threads());
1221 assert(active_workers > 0, "Should have been set");
1222
1223 // Parallel task terminator is set in "set_concurrency_and_phase()"
1224 set_concurrency_and_phase(active_workers, true /* concurrent */);
1225
1226 CMConcurrentMarkingTask markingTask(this, cmThread());
1227 _parallel_workers->set_active_workers(active_workers);
1228 _parallel_workers->run_task(&markingTask);
1229 print_stats();
1230 }
1231
1232 // Helper class to get rid of some boilerplate code.
1233 class G1CMTraceTime : public GCTraceTime {
1234 static bool doit_and_prepend(bool doit) {
1235 if (doit) {
1236 gclog_or_tty->put(' ');
1237 }
1238 return doit;
1239 }
1240
1241 public:
1242 G1CMTraceTime(const char* title, bool doit)
1243 : GCTraceTime(title, doit_and_prepend(doit), false, G1CollectedHeap::heap()->gc_timer_cm(),
1244 G1CollectedHeap::heap()->concurrent_mark()->concurrent_gc_id()) {
1245 }
1246 };
1247
1248 void ConcurrentMark::checkpointRootsFinal(bool clear_all_soft_refs) {
1249 // world is stopped at this checkpoint
1250 assert(SafepointSynchronize::is_at_safepoint(),
1251 "world should be stopped");
1252
1253 G1CollectedHeap* g1h = G1CollectedHeap::heap();
1254
1255 // If a full collection has happened, we shouldn't do this.
1256 if (has_aborted()) {
1257 g1h->set_marking_complete(); // So bitmap clearing isn't confused
1258 return;
1259 }
1260
1261 SvcGCMarker sgcm(SvcGCMarker::OTHER);
1262
1263 if (VerifyDuringGC) {
1264 HandleMark hm; // handle scope
1265 g1h->prepare_for_verify();
1266 Universe::verify(VerifyOption_G1UsePrevMarking,
1267 " VerifyDuringGC:(before)");
1268 }
1269 g1h->check_bitmaps("Remark Start");
1270
1271 G1CollectorPolicy* g1p = g1h->g1_policy();
1272 g1p->record_concurrent_mark_remark_start();
1273
1274 double start = os::elapsedTime();
1275
1276 checkpointRootsFinalWork();
1277
1278 double mark_work_end = os::elapsedTime();
1279
1280 weakRefsWork(clear_all_soft_refs);
1281
1282 if (has_overflown()) {
1283 // Oops. We overflowed. Restart concurrent marking.
1284 _restart_for_overflow = true;
1285 if (G1TraceMarkStackOverflow) {
1286 gclog_or_tty->print_cr("\nRemark led to restart for overflow.");
1287 }
1288
1289 // Verify the heap w.r.t. the previous marking bitmap.
1290 if (VerifyDuringGC) {
1291 HandleMark hm; // handle scope
1292 g1h->prepare_for_verify();
1293 Universe::verify(VerifyOption_G1UsePrevMarking,
1294 " VerifyDuringGC:(overflow)");
1295 }
1296
1297 // Clear the marking state because we will be restarting
1298 // marking due to overflowing the global mark stack.
1299 reset_marking_state();
1300 } else {
1301 {
1302 G1CMTraceTime trace("GC aggregate-data", G1Log::finer());
1303
1304 // Aggregate the per-task counting data that we have accumulated
1305 // while marking.
1306 aggregate_count_data();
1307 }
1308
1309 SATBMarkQueueSet& satb_mq_set = JavaThread::satb_mark_queue_set();
1310 // We're done with marking.
1311 // This is the end of the marking cycle, we're expected all
1312 // threads to have SATB queues with active set to true.
1313 satb_mq_set.set_active_all_threads(false, /* new active value */
1314 true /* expected_active */);
1315
1316 if (VerifyDuringGC) {
1317 HandleMark hm; // handle scope
1318 g1h->prepare_for_verify();
1319 Universe::verify(VerifyOption_G1UseNextMarking,
1320 " VerifyDuringGC:(after)");
1321 }
1322 g1h->check_bitmaps("Remark End");
1323 assert(!restart_for_overflow(), "sanity");
1324 // Completely reset the marking state since marking completed
1325 set_non_marking_state();
1326 }
1327
1328 // Expand the marking stack, if we have to and if we can.
1329 if (_markStack.should_expand()) {
1330 _markStack.expand();
1331 }
1332
1333 // Statistics
1334 double now = os::elapsedTime();
1335 _remark_mark_times.add((mark_work_end - start) * 1000.0);
1336 _remark_weak_ref_times.add((now - mark_work_end) * 1000.0);
1337 _remark_times.add((now - start) * 1000.0);
1338
1339 g1p->record_concurrent_mark_remark_end();
1340
1341 G1CMIsAliveClosure is_alive(g1h);
1342 g1h->gc_tracer_cm()->report_object_count_after_gc(&is_alive);
1343 }
1344
1345 // Base class of the closures that finalize and verify the
1346 // liveness counting data.
1347 class CMCountDataClosureBase: public HeapRegionClosure {
1348 protected:
1349 G1CollectedHeap* _g1h;
1350 ConcurrentMark* _cm;
1351 CardTableModRefBS* _ct_bs;
1352
1353 BitMap* _region_bm;
1354 BitMap* _card_bm;
1355
1356 // Takes a region that's not empty (i.e., it has at least one
1357 // live object in it and sets its corresponding bit on the region
1358 // bitmap to 1. If the region is "starts humongous" it will also set
1359 // to 1 the bits on the region bitmap that correspond to its
1360 // associated "continues humongous" regions.
1361 void set_bit_for_region(HeapRegion* hr) {
1362 assert(!hr->is_continues_humongous(), "should have filtered those out");
1363
1364 BitMap::idx_t index = (BitMap::idx_t) hr->hrm_index();
1365 if (!hr->is_starts_humongous()) {
1366 // Normal (non-humongous) case: just set the bit.
1367 _region_bm->par_at_put(index, true);
1368 } else {
1369 // Starts humongous case: calculate how many regions are part of
1370 // this humongous region and then set the bit range.
1371 BitMap::idx_t end_index = (BitMap::idx_t) hr->last_hc_index();
1372 _region_bm->par_at_put_range(index, end_index, true);
1373 }
1374 }
1375
1376 public:
1377 CMCountDataClosureBase(G1CollectedHeap* g1h,
1378 BitMap* region_bm, BitMap* card_bm):
1379 _g1h(g1h), _cm(g1h->concurrent_mark()),
1380 _ct_bs(barrier_set_cast<CardTableModRefBS>(g1h->barrier_set())),
1381 _region_bm(region_bm), _card_bm(card_bm) { }
1382 };
1383
1384 // Closure that calculates the # live objects per region. Used
1385 // for verification purposes during the cleanup pause.
1386 class CalcLiveObjectsClosure: public CMCountDataClosureBase {
1387 CMBitMapRO* _bm;
1388 size_t _region_marked_bytes;
1389
1390 public:
1391 CalcLiveObjectsClosure(CMBitMapRO *bm, G1CollectedHeap* g1h,
1392 BitMap* region_bm, BitMap* card_bm) :
1393 CMCountDataClosureBase(g1h, region_bm, card_bm),
1394 _bm(bm), _region_marked_bytes(0) { }
1395
1396 bool doHeapRegion(HeapRegion* hr) {
1397
1398 if (hr->is_continues_humongous()) {
1399 // We will ignore these here and process them when their
1400 // associated "starts humongous" region is processed (see
1401 // set_bit_for_heap_region()). Note that we cannot rely on their
1402 // associated "starts humongous" region to have their bit set to
1403 // 1 since, due to the region chunking in the parallel region
1404 // iteration, a "continues humongous" region might be visited
1405 // before its associated "starts humongous".
1406 return false;
1407 }
1408
1409 HeapWord* ntams = hr->next_top_at_mark_start();
1410 HeapWord* start = hr->bottom();
1411
1412 assert(start <= hr->end() && start <= ntams && ntams <= hr->end(),
1413 err_msg("Preconditions not met - "
1414 "start: "PTR_FORMAT", ntams: "PTR_FORMAT", end: "PTR_FORMAT,
1415 p2i(start), p2i(ntams), p2i(hr->end())));
1416
1417 // Find the first marked object at or after "start".
1418 start = _bm->getNextMarkedWordAddress(start, ntams);
1419
1420 size_t marked_bytes = 0;
1421
1422 while (start < ntams) {
1423 oop obj = oop(start);
1424 int obj_sz = obj->size();
1425 HeapWord* obj_end = start + obj_sz;
1426
1427 BitMap::idx_t start_idx = _cm->card_bitmap_index_for(start);
1428 BitMap::idx_t end_idx = _cm->card_bitmap_index_for(obj_end);
1429
1430 // Note: if we're looking at the last region in heap - obj_end
1431 // could be actually just beyond the end of the heap; end_idx
1432 // will then correspond to a (non-existent) card that is also
1433 // just beyond the heap.
1434 if (_g1h->is_in_g1_reserved(obj_end) && !_ct_bs->is_card_aligned(obj_end)) {
1435 // end of object is not card aligned - increment to cover
1436 // all the cards spanned by the object
1437 end_idx += 1;
1438 }
1439
1440 // Set the bits in the card BM for the cards spanned by this object.
1441 _cm->set_card_bitmap_range(_card_bm, start_idx, end_idx, true /* is_par */);
1442
1443 // Add the size of this object to the number of marked bytes.
1444 marked_bytes += (size_t)obj_sz * HeapWordSize;
1445
1446 // Find the next marked object after this one.
1447 start = _bm->getNextMarkedWordAddress(obj_end, ntams);
1448 }
1449
1450 // Mark the allocated-since-marking portion...
1451 HeapWord* top = hr->top();
1452 if (ntams < top) {
1453 BitMap::idx_t start_idx = _cm->card_bitmap_index_for(ntams);
1454 BitMap::idx_t end_idx = _cm->card_bitmap_index_for(top);
1455
1456 // Note: if we're looking at the last region in heap - top
1457 // could be actually just beyond the end of the heap; end_idx
1458 // will then correspond to a (non-existent) card that is also
1459 // just beyond the heap.
1460 if (_g1h->is_in_g1_reserved(top) && !_ct_bs->is_card_aligned(top)) {
1461 // end of object is not card aligned - increment to cover
1462 // all the cards spanned by the object
1463 end_idx += 1;
1464 }
1465 _cm->set_card_bitmap_range(_card_bm, start_idx, end_idx, true /* is_par */);
1466
1467 // This definitely means the region has live objects.
1468 set_bit_for_region(hr);
1469 }
1470
1471 // Update the live region bitmap.
1472 if (marked_bytes > 0) {
1473 set_bit_for_region(hr);
1474 }
1475
1476 // Set the marked bytes for the current region so that
1477 // it can be queried by a calling verification routine
1478 _region_marked_bytes = marked_bytes;
1479
1480 return false;
1481 }
1482
1483 size_t region_marked_bytes() const { return _region_marked_bytes; }
1484 };
1485
1486 // Heap region closure used for verifying the counting data
1487 // that was accumulated concurrently and aggregated during
1488 // the remark pause. This closure is applied to the heap
1489 // regions during the STW cleanup pause.
1490
1491 class VerifyLiveObjectDataHRClosure: public HeapRegionClosure {
1492 G1CollectedHeap* _g1h;
1493 ConcurrentMark* _cm;
1494 CalcLiveObjectsClosure _calc_cl;
1495 BitMap* _region_bm; // Region BM to be verified
1496 BitMap* _card_bm; // Card BM to be verified
1497 bool _verbose; // verbose output?
1498
1499 BitMap* _exp_region_bm; // Expected Region BM values
1500 BitMap* _exp_card_bm; // Expected card BM values
1501
1502 int _failures;
1503
1504 public:
1505 VerifyLiveObjectDataHRClosure(G1CollectedHeap* g1h,
1506 BitMap* region_bm,
1507 BitMap* card_bm,
1508 BitMap* exp_region_bm,
1509 BitMap* exp_card_bm,
1510 bool verbose) :
1511 _g1h(g1h), _cm(g1h->concurrent_mark()),
1512 _calc_cl(_cm->nextMarkBitMap(), g1h, exp_region_bm, exp_card_bm),
1513 _region_bm(region_bm), _card_bm(card_bm), _verbose(verbose),
1514 _exp_region_bm(exp_region_bm), _exp_card_bm(exp_card_bm),
1515 _failures(0) { }
1516
1517 int failures() const { return _failures; }
1518
1519 bool doHeapRegion(HeapRegion* hr) {
1520 if (hr->is_continues_humongous()) {
1521 // We will ignore these here and process them when their
1522 // associated "starts humongous" region is processed (see
1523 // set_bit_for_heap_region()). Note that we cannot rely on their
1524 // associated "starts humongous" region to have their bit set to
1525 // 1 since, due to the region chunking in the parallel region
1526 // iteration, a "continues humongous" region might be visited
1527 // before its associated "starts humongous".
1528 return false;
1529 }
1530
1531 int failures = 0;
1532
1533 // Call the CalcLiveObjectsClosure to walk the marking bitmap for
1534 // this region and set the corresponding bits in the expected region
1535 // and card bitmaps.
1536 bool res = _calc_cl.doHeapRegion(hr);
1537 assert(res == false, "should be continuing");
1538
1539 MutexLockerEx x((_verbose ? ParGCRareEvent_lock : NULL),
1540 Mutex::_no_safepoint_check_flag);
1541
1542 // Verify the marked bytes for this region.
1543 size_t exp_marked_bytes = _calc_cl.region_marked_bytes();
1544 size_t act_marked_bytes = hr->next_marked_bytes();
1545
1546 // We're not OK if expected marked bytes > actual marked bytes. It means
1547 // we have missed accounting some objects during the actual marking.
1548 if (exp_marked_bytes > act_marked_bytes) {
1549 if (_verbose) {
1550 gclog_or_tty->print_cr("Region %u: marked bytes mismatch: "
1551 "expected: " SIZE_FORMAT ", actual: " SIZE_FORMAT,
1552 hr->hrm_index(), exp_marked_bytes, act_marked_bytes);
1553 }
1554 failures += 1;
1555 }
1556
1557 // Verify the bit, for this region, in the actual and expected
1558 // (which was just calculated) region bit maps.
1559 // We're not OK if the bit in the calculated expected region
1560 // bitmap is set and the bit in the actual region bitmap is not.
1561 BitMap::idx_t index = (BitMap::idx_t) hr->hrm_index();
1562
1563 bool expected = _exp_region_bm->at(index);
1564 bool actual = _region_bm->at(index);
1565 if (expected && !actual) {
1566 if (_verbose) {
1567 gclog_or_tty->print_cr("Region %u: region bitmap mismatch: "
1568 "expected: %s, actual: %s",
1569 hr->hrm_index(),
1570 BOOL_TO_STR(expected), BOOL_TO_STR(actual));
1571 }
1572 failures += 1;
1573 }
1574
1575 // Verify that the card bit maps for the cards spanned by the current
1576 // region match. We have an error if we have a set bit in the expected
1577 // bit map and the corresponding bit in the actual bitmap is not set.
1578
1579 BitMap::idx_t start_idx = _cm->card_bitmap_index_for(hr->bottom());
1580 BitMap::idx_t end_idx = _cm->card_bitmap_index_for(hr->top());
1581
1582 for (BitMap::idx_t i = start_idx; i < end_idx; i+=1) {
1583 expected = _exp_card_bm->at(i);
1584 actual = _card_bm->at(i);
1585
1586 if (expected && !actual) {
1587 if (_verbose) {
1588 gclog_or_tty->print_cr("Region %u: card bitmap mismatch at " SIZE_FORMAT ": "
1589 "expected: %s, actual: %s",
1590 hr->hrm_index(), i,
1591 BOOL_TO_STR(expected), BOOL_TO_STR(actual));
1592 }
1593 failures += 1;
1594 }
1595 }
1596
1597 if (failures > 0 && _verbose) {
1598 gclog_or_tty->print_cr("Region " HR_FORMAT ", ntams: " PTR_FORMAT ", "
1599 "marked_bytes: calc/actual " SIZE_FORMAT "/" SIZE_FORMAT,
1600 HR_FORMAT_PARAMS(hr), p2i(hr->next_top_at_mark_start()),
1601 _calc_cl.region_marked_bytes(), hr->next_marked_bytes());
1602 }
1603
1604 _failures += failures;
1605
1606 // We could stop iteration over the heap when we
1607 // find the first violating region by returning true.
1608 return false;
1609 }
1610 };
1611
1612 class G1ParVerifyFinalCountTask: public AbstractGangTask {
1613 protected:
1614 G1CollectedHeap* _g1h;
1615 ConcurrentMark* _cm;
1616 BitMap* _actual_region_bm;
1617 BitMap* _actual_card_bm;
1618
1619 uint _n_workers;
1620
1621 BitMap* _expected_region_bm;
1622 BitMap* _expected_card_bm;
1623
1624 int _failures;
1625 bool _verbose;
1626
1627 HeapRegionClaimer _hrclaimer;
1628
1629 public:
1630 G1ParVerifyFinalCountTask(G1CollectedHeap* g1h,
1631 BitMap* region_bm, BitMap* card_bm,
1632 BitMap* expected_region_bm, BitMap* expected_card_bm)
1633 : AbstractGangTask("G1 verify final counting"),
1634 _g1h(g1h), _cm(_g1h->concurrent_mark()),
1635 _actual_region_bm(region_bm), _actual_card_bm(card_bm),
1636 _expected_region_bm(expected_region_bm), _expected_card_bm(expected_card_bm),
1637 _failures(0), _verbose(false),
1638 _n_workers(_g1h->workers()->active_workers()), _hrclaimer(_n_workers) {
1639 assert(VerifyDuringGC, "don't call this otherwise");
1640 assert(_expected_card_bm->size() == _actual_card_bm->size(), "sanity");
1641 assert(_expected_region_bm->size() == _actual_region_bm->size(), "sanity");
1642
1643 _verbose = _cm->verbose_medium();
1644 }
1645
1646 void work(uint worker_id) {
1647 assert(worker_id < _n_workers, "invariant");
1648
1649 VerifyLiveObjectDataHRClosure verify_cl(_g1h,
1650 _actual_region_bm, _actual_card_bm,
1651 _expected_region_bm,
1652 _expected_card_bm,
1653 _verbose);
1654
1655 _g1h->heap_region_par_iterate(&verify_cl, worker_id, &_hrclaimer);
1656
1657 Atomic::add(verify_cl.failures(), &_failures);
1658 }
1659
1660 int failures() const { return _failures; }
1661 };
1662
1663 // Closure that finalizes the liveness counting data.
1664 // Used during the cleanup pause.
1665 // Sets the bits corresponding to the interval [NTAMS, top]
1666 // (which contains the implicitly live objects) in the
1667 // card liveness bitmap. Also sets the bit for each region,
1668 // containing live data, in the region liveness bitmap.
1669
1670 class FinalCountDataUpdateClosure: public CMCountDataClosureBase {
1671 public:
1672 FinalCountDataUpdateClosure(G1CollectedHeap* g1h,
1673 BitMap* region_bm,
1674 BitMap* card_bm) :
1675 CMCountDataClosureBase(g1h, region_bm, card_bm) { }
1676
1677 bool doHeapRegion(HeapRegion* hr) {
1678
1679 if (hr->is_continues_humongous()) {
1680 // We will ignore these here and process them when their
1681 // associated "starts humongous" region is processed (see
1682 // set_bit_for_heap_region()). Note that we cannot rely on their
1683 // associated "starts humongous" region to have their bit set to
1684 // 1 since, due to the region chunking in the parallel region
1685 // iteration, a "continues humongous" region might be visited
1686 // before its associated "starts humongous".
1687 return false;
1688 }
1689
1690 HeapWord* ntams = hr->next_top_at_mark_start();
1691 HeapWord* top = hr->top();
1692
1693 assert(hr->bottom() <= ntams && ntams <= hr->end(), "Preconditions.");
1694
1695 // Mark the allocated-since-marking portion...
1696 if (ntams < top) {
1697 // This definitely means the region has live objects.
1698 set_bit_for_region(hr);
1699
1700 // Now set the bits in the card bitmap for [ntams, top)
1701 BitMap::idx_t start_idx = _cm->card_bitmap_index_for(ntams);
1702 BitMap::idx_t end_idx = _cm->card_bitmap_index_for(top);
1703
1704 // Note: if we're looking at the last region in heap - top
1705 // could be actually just beyond the end of the heap; end_idx
1706 // will then correspond to a (non-existent) card that is also
1707 // just beyond the heap.
1708 if (_g1h->is_in_g1_reserved(top) && !_ct_bs->is_card_aligned(top)) {
1709 // end of object is not card aligned - increment to cover
1710 // all the cards spanned by the object
1711 end_idx += 1;
1712 }
1713
1714 assert(end_idx <= _card_bm->size(),
1715 err_msg("oob: end_idx= "SIZE_FORMAT", bitmap size= "SIZE_FORMAT,
1716 end_idx, _card_bm->size()));
1717 assert(start_idx < _card_bm->size(),
1718 err_msg("oob: start_idx= "SIZE_FORMAT", bitmap size= "SIZE_FORMAT,
1719 start_idx, _card_bm->size()));
1720
1721 _cm->set_card_bitmap_range(_card_bm, start_idx, end_idx, true /* is_par */);
1722 }
1723
1724 // Set the bit for the region if it contains live data
1725 if (hr->next_marked_bytes() > 0) {
1726 set_bit_for_region(hr);
1727 }
1728
1729 return false;
1730 }
1731 };
1732
1733 class G1ParFinalCountTask: public AbstractGangTask {
1734 protected:
1735 G1CollectedHeap* _g1h;
1736 ConcurrentMark* _cm;
1737 BitMap* _actual_region_bm;
1738 BitMap* _actual_card_bm;
1739
1740 uint _n_workers;
1741 HeapRegionClaimer _hrclaimer;
1742
1743 public:
1744 G1ParFinalCountTask(G1CollectedHeap* g1h, BitMap* region_bm, BitMap* card_bm)
1745 : AbstractGangTask("G1 final counting"),
1746 _g1h(g1h), _cm(_g1h->concurrent_mark()),
1747 _actual_region_bm(region_bm), _actual_card_bm(card_bm),
1748 _n_workers(_g1h->workers()->active_workers()), _hrclaimer(_n_workers) {
1749 }
1750
1751 void work(uint worker_id) {
1752 assert(worker_id < _n_workers, "invariant");
1753
1754 FinalCountDataUpdateClosure final_update_cl(_g1h,
1755 _actual_region_bm,
1756 _actual_card_bm);
1757
1758 _g1h->heap_region_par_iterate(&final_update_cl, worker_id, &_hrclaimer);
1759 }
1760 };
1761
1762 class G1NoteEndOfConcMarkClosure : public HeapRegionClosure {
1763 G1CollectedHeap* _g1;
1764 size_t _freed_bytes;
1765 FreeRegionList* _local_cleanup_list;
1766 HeapRegionSetCount _old_regions_removed;
1767 HeapRegionSetCount _humongous_regions_removed;
1768 HRRSCleanupTask* _hrrs_cleanup_task;
1769
1770 public:
1771 G1NoteEndOfConcMarkClosure(G1CollectedHeap* g1,
1772 FreeRegionList* local_cleanup_list,
1773 HRRSCleanupTask* hrrs_cleanup_task) :
1774 _g1(g1),
1775 _freed_bytes(0),
1776 _local_cleanup_list(local_cleanup_list),
1777 _old_regions_removed(),
1778 _humongous_regions_removed(),
1779 _hrrs_cleanup_task(hrrs_cleanup_task) { }
1780
1781 size_t freed_bytes() { return _freed_bytes; }
1782 const HeapRegionSetCount& old_regions_removed() { return _old_regions_removed; }
1783 const HeapRegionSetCount& humongous_regions_removed() { return _humongous_regions_removed; }
1784
1785 bool doHeapRegion(HeapRegion *hr) {
1786 if (hr->is_continues_humongous()) {
1787 return false;
1788 }
1789 // We use a claim value of zero here because all regions
1790 // were claimed with value 1 in the FinalCount task.
1791 _g1->reset_gc_time_stamps(hr);
1792 hr->note_end_of_marking();
1793
1794 if (hr->used() > 0 && hr->max_live_bytes() == 0 && !hr->is_young()) {
1795 _freed_bytes += hr->used();
1796 hr->set_containing_set(NULL);
1797 if (hr->is_humongous()) {
1798 assert(hr->is_starts_humongous(), "we should only see starts humongous");
1799 _humongous_regions_removed.increment(1u, hr->capacity());
1800 _g1->free_humongous_region(hr, _local_cleanup_list, true);
1801 } else {
1802 _old_regions_removed.increment(1u, hr->capacity());
1803 _g1->free_region(hr, _local_cleanup_list, true);
1804 }
1805 } else {
1806 hr->rem_set()->do_cleanup_work(_hrrs_cleanup_task);
1807 }
1808
1809 return false;
1810 }
1811 };
1812
1813 class G1ParNoteEndTask: public AbstractGangTask {
1814 friend class G1NoteEndOfConcMarkClosure;
1815
1816 protected:
1817 G1CollectedHeap* _g1h;
1818 FreeRegionList* _cleanup_list;
1819 HeapRegionClaimer _hrclaimer;
1820
1821 public:
1822 G1ParNoteEndTask(G1CollectedHeap* g1h, FreeRegionList* cleanup_list, uint n_workers) :
1823 AbstractGangTask("G1 note end"), _g1h(g1h), _cleanup_list(cleanup_list), _hrclaimer(n_workers) {
1824 }
1825
1826 void work(uint worker_id) {
1827 FreeRegionList local_cleanup_list("Local Cleanup List");
1828 HRRSCleanupTask hrrs_cleanup_task;
1829 G1NoteEndOfConcMarkClosure g1_note_end(_g1h, &local_cleanup_list,
1830 &hrrs_cleanup_task);
1831 _g1h->heap_region_par_iterate(&g1_note_end, worker_id, &_hrclaimer);
1832 assert(g1_note_end.complete(), "Shouldn't have yielded!");
1833
1834 // Now update the lists
1835 _g1h->remove_from_old_sets(g1_note_end.old_regions_removed(), g1_note_end.humongous_regions_removed());
1836 {
1837 MutexLockerEx x(ParGCRareEvent_lock, Mutex::_no_safepoint_check_flag);
1838 _g1h->decrement_summary_bytes(g1_note_end.freed_bytes());
1839
1840 // If we iterate over the global cleanup list at the end of
1841 // cleanup to do this printing we will not guarantee to only
1842 // generate output for the newly-reclaimed regions (the list
1843 // might not be empty at the beginning of cleanup; we might
1844 // still be working on its previous contents). So we do the
1845 // printing here, before we append the new regions to the global
1846 // cleanup list.
1847
1848 G1HRPrinter* hr_printer = _g1h->hr_printer();
1849 if (hr_printer->is_active()) {
1850 FreeRegionListIterator iter(&local_cleanup_list);
1851 while (iter.more_available()) {
1852 HeapRegion* hr = iter.get_next();
1853 hr_printer->cleanup(hr);
1854 }
1855 }
1856
1857 _cleanup_list->add_ordered(&local_cleanup_list);
1858 assert(local_cleanup_list.is_empty(), "post-condition");
1859
1860 HeapRegionRemSet::finish_cleanup_task(&hrrs_cleanup_task);
1861 }
1862 }
1863 };
1864
1865 class G1ParScrubRemSetTask: public AbstractGangTask {
1866 protected:
1867 G1RemSet* _g1rs;
1868 BitMap* _region_bm;
1869 BitMap* _card_bm;
1870 HeapRegionClaimer _hrclaimer;
1871
1872 public:
1873 G1ParScrubRemSetTask(G1CollectedHeap* g1h, BitMap* region_bm, BitMap* card_bm, uint n_workers) :
1874 AbstractGangTask("G1 ScrubRS"), _g1rs(g1h->g1_rem_set()), _region_bm(region_bm), _card_bm(card_bm), _hrclaimer(n_workers) {
1875 }
1876
1877 void work(uint worker_id) {
1878 _g1rs->scrub(_region_bm, _card_bm, worker_id, &_hrclaimer);
1879 }
1880
1881 };
1882
1883 void ConcurrentMark::cleanup() {
1884 // world is stopped at this checkpoint
1885 assert(SafepointSynchronize::is_at_safepoint(),
1886 "world should be stopped");
1887 G1CollectedHeap* g1h = G1CollectedHeap::heap();
1888
1889 // If a full collection has happened, we shouldn't do this.
1890 if (has_aborted()) {
1891 g1h->set_marking_complete(); // So bitmap clearing isn't confused
1892 return;
1893 }
1894
1895 g1h->verify_region_sets_optional();
1896
1897 if (VerifyDuringGC) {
1898 HandleMark hm; // handle scope
1899 g1h->prepare_for_verify();
1900 Universe::verify(VerifyOption_G1UsePrevMarking,
1901 " VerifyDuringGC:(before)");
1902 }
1903 g1h->check_bitmaps("Cleanup Start");
1904
1905 G1CollectorPolicy* g1p = g1h->g1_policy();
1906 g1p->record_concurrent_mark_cleanup_start();
1907
1908 double start = os::elapsedTime();
1909
1910 HeapRegionRemSet::reset_for_cleanup_tasks();
1911
1912 // Do counting once more with the world stopped for good measure.
1913 G1ParFinalCountTask g1_par_count_task(g1h, &_region_bm, &_card_bm);
1914
1915 g1h->workers()->run_task(&g1_par_count_task);
1916
1917 if (VerifyDuringGC) {
1918 // Verify that the counting data accumulated during marking matches
1919 // that calculated by walking the marking bitmap.
1920
1921 // Bitmaps to hold expected values
1922 BitMap expected_region_bm(_region_bm.size(), true);
1923 BitMap expected_card_bm(_card_bm.size(), true);
1924
1925 G1ParVerifyFinalCountTask g1_par_verify_task(g1h,
1926 &_region_bm,
1927 &_card_bm,
1928 &expected_region_bm,
1929 &expected_card_bm);
1930
1931 g1h->workers()->run_task(&g1_par_verify_task);
1932
1933 guarantee(g1_par_verify_task.failures() == 0, "Unexpected accounting failures");
1934 }
1935
1936 size_t start_used_bytes = g1h->used();
1937 g1h->set_marking_complete();
1938
1939 double count_end = os::elapsedTime();
1940 double this_final_counting_time = (count_end - start);
1941 _total_counting_time += this_final_counting_time;
1942
1943 if (G1PrintRegionLivenessInfo) {
1944 G1PrintRegionLivenessInfoClosure cl(gclog_or_tty, "Post-Marking");
1945 _g1h->heap_region_iterate(&cl);
1946 }
1947
1948 // Install newly created mark bitMap as "prev".
1949 swapMarkBitMaps();
1950
1951 g1h->reset_gc_time_stamp();
1952
1953 uint n_workers = _g1h->workers()->active_workers();
1954
1955 // Note end of marking in all heap regions.
1956 G1ParNoteEndTask g1_par_note_end_task(g1h, &_cleanup_list, n_workers);
1957 g1h->workers()->run_task(&g1_par_note_end_task);
1958 g1h->check_gc_time_stamps();
1959
1960 if (!cleanup_list_is_empty()) {
1961 // The cleanup list is not empty, so we'll have to process it
1962 // concurrently. Notify anyone else that might be wanting free
1963 // regions that there will be more free regions coming soon.
1964 g1h->set_free_regions_coming();
1965 }
1966
1967 // call below, since it affects the metric by which we sort the heap
1968 // regions.
1969 if (G1ScrubRemSets) {
1970 double rs_scrub_start = os::elapsedTime();
1971 G1ParScrubRemSetTask g1_par_scrub_rs_task(g1h, &_region_bm, &_card_bm, n_workers);
1972 g1h->workers()->run_task(&g1_par_scrub_rs_task);
1973
1974 double rs_scrub_end = os::elapsedTime();
1975 double this_rs_scrub_time = (rs_scrub_end - rs_scrub_start);
1976 _total_rs_scrub_time += this_rs_scrub_time;
1977 }
1978
1979 // this will also free any regions totally full of garbage objects,
1980 // and sort the regions.
1981 g1h->g1_policy()->record_concurrent_mark_cleanup_end();
1982
1983 // Statistics.
1984 double end = os::elapsedTime();
1985 _cleanup_times.add((end - start) * 1000.0);
1986
1987 if (G1Log::fine()) {
1988 g1h->g1_policy()->print_heap_transition(start_used_bytes);
1989 }
1990
1991 // Clean up will have freed any regions completely full of garbage.
1992 // Update the soft reference policy with the new heap occupancy.
1993 Universe::update_heap_info_at_gc();
1994
1995 if (VerifyDuringGC) {
1996 HandleMark hm; // handle scope
1997 g1h->prepare_for_verify();
1998 Universe::verify(VerifyOption_G1UsePrevMarking,
1999 " VerifyDuringGC:(after)");
2000 }
2001
2002 g1h->check_bitmaps("Cleanup End");
2003
2004 g1h->verify_region_sets_optional();
2005
2006 // We need to make this be a "collection" so any collection pause that
2007 // races with it goes around and waits for completeCleanup to finish.
2008 g1h->increment_total_collections();
2009
2010 // Clean out dead classes and update Metaspace sizes.
2011 if (ClassUnloadingWithConcurrentMark) {
2012 ClassLoaderDataGraph::purge();
2013 }
2014 MetaspaceGC::compute_new_size();
2015
2016 // We reclaimed old regions so we should calculate the sizes to make
2017 // sure we update the old gen/space data.
2018 g1h->g1mm()->update_sizes();
2019 g1h->allocation_context_stats().update_after_mark();
2020
2021 g1h->trace_heap_after_concurrent_cycle();
2022 }
2023
2024 void ConcurrentMark::completeCleanup() {
2025 if (has_aborted()) return;
2026
2027 G1CollectedHeap* g1h = G1CollectedHeap::heap();
2028
2029 _cleanup_list.verify_optional();
2030 FreeRegionList tmp_free_list("Tmp Free List");
2031
2032 if (G1ConcRegionFreeingVerbose) {
2033 gclog_or_tty->print_cr("G1ConcRegionFreeing [complete cleanup] : "
2034 "cleanup list has %u entries",
2035 _cleanup_list.length());
2036 }
2037
2038 // No one else should be accessing the _cleanup_list at this point,
2039 // so it is not necessary to take any locks
2040 while (!_cleanup_list.is_empty()) {
2041 HeapRegion* hr = _cleanup_list.remove_region(true /* from_head */);
2042 assert(hr != NULL, "Got NULL from a non-empty list");
2043 hr->par_clear();
2044 tmp_free_list.add_ordered(hr);
2045
2046 // Instead of adding one region at a time to the secondary_free_list,
2047 // we accumulate them in the local list and move them a few at a
2048 // time. This also cuts down on the number of notify_all() calls
2049 // we do during this process. We'll also append the local list when
2050 // _cleanup_list is empty (which means we just removed the last
2051 // region from the _cleanup_list).
2052 if ((tmp_free_list.length() % G1SecondaryFreeListAppendLength == 0) ||
2053 _cleanup_list.is_empty()) {
2054 if (G1ConcRegionFreeingVerbose) {
2055 gclog_or_tty->print_cr("G1ConcRegionFreeing [complete cleanup] : "
2056 "appending %u entries to the secondary_free_list, "
2057 "cleanup list still has %u entries",
2058 tmp_free_list.length(),
2059 _cleanup_list.length());
2060 }
2061
2062 {
2063 MutexLockerEx x(SecondaryFreeList_lock, Mutex::_no_safepoint_check_flag);
2064 g1h->secondary_free_list_add(&tmp_free_list);
2065 SecondaryFreeList_lock->notify_all();
2066 }
2067 #ifndef PRODUCT
2068 if (G1StressConcRegionFreeing) {
2069 for (uintx i = 0; i < G1StressConcRegionFreeingDelayMillis; ++i) {
2070 os::sleep(Thread::current(), (jlong) 1, false);
2071 }
2072 }
2073 #endif
2074 }
2075 }
2076 assert(tmp_free_list.is_empty(), "post-condition");
2077 }
2078
2079 // Supporting Object and Oop closures for reference discovery
2080 // and processing in during marking
2081
2082 bool G1CMIsAliveClosure::do_object_b(oop obj) {
2083 HeapWord* addr = (HeapWord*)obj;
2084 return addr != NULL &&
2085 (!_g1->is_in_g1_reserved(addr) || !_g1->is_obj_ill(obj));
2086 }
2087
2088 // 'Keep Alive' oop closure used by both serial parallel reference processing.
2089 // Uses the CMTask associated with a worker thread (for serial reference
2090 // processing the CMTask for worker 0 is used) to preserve (mark) and
2091 // trace referent objects.
2092 //
2093 // Using the CMTask and embedded local queues avoids having the worker
2094 // threads operating on the global mark stack. This reduces the risk
2095 // of overflowing the stack - which we would rather avoid at this late
2096 // state. Also using the tasks' local queues removes the potential
2097 // of the workers interfering with each other that could occur if
2098 // operating on the global stack.
2099
2100 class G1CMKeepAliveAndDrainClosure: public OopClosure {
2101 ConcurrentMark* _cm;
2102 CMTask* _task;
2103 int _ref_counter_limit;
2104 int _ref_counter;
2105 bool _is_serial;
2106 public:
2107 G1CMKeepAliveAndDrainClosure(ConcurrentMark* cm, CMTask* task, bool is_serial) :
2108 _cm(cm), _task(task), _is_serial(is_serial),
2109 _ref_counter_limit(G1RefProcDrainInterval) {
2110 assert(_ref_counter_limit > 0, "sanity");
2111 assert(!_is_serial || _task->worker_id() == 0, "only task 0 for serial code");
2112 _ref_counter = _ref_counter_limit;
2113 }
2114
2115 virtual void do_oop(narrowOop* p) { do_oop_work(p); }
2116 virtual void do_oop( oop* p) { do_oop_work(p); }
2117
2118 template <class T> void do_oop_work(T* p) {
2119 if (!_cm->has_overflown()) {
2120 oop obj = oopDesc::load_decode_heap_oop(p);
2121 if (_cm->verbose_high()) {
2122 gclog_or_tty->print_cr("\t[%u] we're looking at location "
2123 "*"PTR_FORMAT" = "PTR_FORMAT,
2124 _task->worker_id(), p2i(p), p2i((void*) obj));
2125 }
2126
2127 _task->deal_with_reference(obj);
2128 _ref_counter--;
2129
2130 if (_ref_counter == 0) {
2131 // We have dealt with _ref_counter_limit references, pushing them
2132 // and objects reachable from them on to the local stack (and
2133 // possibly the global stack). Call CMTask::do_marking_step() to
2134 // process these entries.
2135 //
2136 // We call CMTask::do_marking_step() in a loop, which we'll exit if
2137 // there's nothing more to do (i.e. we're done with the entries that
2138 // were pushed as a result of the CMTask::deal_with_reference() calls
2139 // above) or we overflow.
2140 //
2141 // Note: CMTask::do_marking_step() can set the CMTask::has_aborted()
2142 // flag while there may still be some work to do. (See the comment at
2143 // the beginning of CMTask::do_marking_step() for those conditions -
2144 // one of which is reaching the specified time target.) It is only
2145 // when CMTask::do_marking_step() returns without setting the
2146 // has_aborted() flag that the marking step has completed.
2147 do {
2148 double mark_step_duration_ms = G1ConcMarkStepDurationMillis;
2149 _task->do_marking_step(mark_step_duration_ms,
2150 false /* do_termination */,
2151 _is_serial);
2152 } while (_task->has_aborted() && !_cm->has_overflown());
2153 _ref_counter = _ref_counter_limit;
2154 }
2155 } else {
2156 if (_cm->verbose_high()) {
2157 gclog_or_tty->print_cr("\t[%u] CM Overflow", _task->worker_id());
2158 }
2159 }
2160 }
2161 };
2162
2163 // 'Drain' oop closure used by both serial and parallel reference processing.
2164 // Uses the CMTask associated with a given worker thread (for serial
2165 // reference processing the CMtask for worker 0 is used). Calls the
2166 // do_marking_step routine, with an unbelievably large timeout value,
2167 // to drain the marking data structures of the remaining entries
2168 // added by the 'keep alive' oop closure above.
2169
2170 class G1CMDrainMarkingStackClosure: public VoidClosure {
2171 ConcurrentMark* _cm;
2172 CMTask* _task;
2173 bool _is_serial;
2174 public:
2175 G1CMDrainMarkingStackClosure(ConcurrentMark* cm, CMTask* task, bool is_serial) :
2176 _cm(cm), _task(task), _is_serial(is_serial) {
2177 assert(!_is_serial || _task->worker_id() == 0, "only task 0 for serial code");
2178 }
2179
2180 void do_void() {
2181 do {
2182 if (_cm->verbose_high()) {
2183 gclog_or_tty->print_cr("\t[%u] Drain: Calling do_marking_step - serial: %s",
2184 _task->worker_id(), BOOL_TO_STR(_is_serial));
2185 }
2186
2187 // We call CMTask::do_marking_step() to completely drain the local
2188 // and global marking stacks of entries pushed by the 'keep alive'
2189 // oop closure (an instance of G1CMKeepAliveAndDrainClosure above).
2190 //
2191 // CMTask::do_marking_step() is called in a loop, which we'll exit
2192 // if there's nothing more to do (i.e. we've completely drained the
2193 // entries that were pushed as a a result of applying the 'keep alive'
2194 // closure to the entries on the discovered ref lists) or we overflow
2195 // the global marking stack.
2196 //
2197 // Note: CMTask::do_marking_step() can set the CMTask::has_aborted()
2198 // flag while there may still be some work to do. (See the comment at
2199 // the beginning of CMTask::do_marking_step() for those conditions -
2200 // one of which is reaching the specified time target.) It is only
2201 // when CMTask::do_marking_step() returns without setting the
2202 // has_aborted() flag that the marking step has completed.
2203
2204 _task->do_marking_step(1000000000.0 /* something very large */,
2205 true /* do_termination */,
2206 _is_serial);
2207 } while (_task->has_aborted() && !_cm->has_overflown());
2208 }
2209 };
2210
2211 // Implementation of AbstractRefProcTaskExecutor for parallel
2212 // reference processing at the end of G1 concurrent marking
2213
2214 class G1CMRefProcTaskExecutor: public AbstractRefProcTaskExecutor {
2215 private:
2216 G1CollectedHeap* _g1h;
2217 ConcurrentMark* _cm;
2218 WorkGang* _workers;
2219 uint _active_workers;
2220
2221 public:
2222 G1CMRefProcTaskExecutor(G1CollectedHeap* g1h,
2223 ConcurrentMark* cm,
2224 WorkGang* workers,
2225 uint n_workers) :
2226 _g1h(g1h), _cm(cm),
2227 _workers(workers), _active_workers(n_workers) { }
2228
2229 // Executes the given task using concurrent marking worker threads.
2230 virtual void execute(ProcessTask& task);
2231 virtual void execute(EnqueueTask& task);
2232 };
2233
2234 class G1CMRefProcTaskProxy: public AbstractGangTask {
2235 typedef AbstractRefProcTaskExecutor::ProcessTask ProcessTask;
2236 ProcessTask& _proc_task;
2237 G1CollectedHeap* _g1h;
2238 ConcurrentMark* _cm;
2239
2240 public:
2241 G1CMRefProcTaskProxy(ProcessTask& proc_task,
2242 G1CollectedHeap* g1h,
2243 ConcurrentMark* cm) :
2244 AbstractGangTask("Process reference objects in parallel"),
2245 _proc_task(proc_task), _g1h(g1h), _cm(cm) {
2246 ReferenceProcessor* rp = _g1h->ref_processor_cm();
2247 assert(rp->processing_is_mt(), "shouldn't be here otherwise");
2248 }
2249
2250 virtual void work(uint worker_id) {
2251 ResourceMark rm;
2252 HandleMark hm;
2253 CMTask* task = _cm->task(worker_id);
2254 G1CMIsAliveClosure g1_is_alive(_g1h);
2255 G1CMKeepAliveAndDrainClosure g1_par_keep_alive(_cm, task, false /* is_serial */);
2256 G1CMDrainMarkingStackClosure g1_par_drain(_cm, task, false /* is_serial */);
2257
2258 _proc_task.work(worker_id, g1_is_alive, g1_par_keep_alive, g1_par_drain);
2259 }
2260 };
2261
2262 void G1CMRefProcTaskExecutor::execute(ProcessTask& proc_task) {
2263 assert(_workers != NULL, "Need parallel worker threads.");
2264 assert(_g1h->ref_processor_cm()->processing_is_mt(), "processing is not MT");
2265
2266 G1CMRefProcTaskProxy proc_task_proxy(proc_task, _g1h, _cm);
2267
2268 // We need to reset the concurrency level before each
2269 // proxy task execution, so that the termination protocol
2270 // and overflow handling in CMTask::do_marking_step() knows
2271 // how many workers to wait for.
2272 _cm->set_concurrency(_active_workers);
2273 _workers->run_task(&proc_task_proxy);
2274 }
2275
2276 class G1CMRefEnqueueTaskProxy: public AbstractGangTask {
2277 typedef AbstractRefProcTaskExecutor::EnqueueTask EnqueueTask;
2278 EnqueueTask& _enq_task;
2279
2280 public:
2281 G1CMRefEnqueueTaskProxy(EnqueueTask& enq_task) :
2282 AbstractGangTask("Enqueue reference objects in parallel"),
2283 _enq_task(enq_task) { }
2284
2285 virtual void work(uint worker_id) {
2286 _enq_task.work(worker_id);
2287 }
2288 };
2289
2290 void G1CMRefProcTaskExecutor::execute(EnqueueTask& enq_task) {
2291 assert(_workers != NULL, "Need parallel worker threads.");
2292 assert(_g1h->ref_processor_cm()->processing_is_mt(), "processing is not MT");
2293
2294 G1CMRefEnqueueTaskProxy enq_task_proxy(enq_task);
2295
2296 // Not strictly necessary but...
2297 //
2298 // We need to reset the concurrency level before each
2299 // proxy task execution, so that the termination protocol
2300 // and overflow handling in CMTask::do_marking_step() knows
2301 // how many workers to wait for.
2302 _cm->set_concurrency(_active_workers);
2303 _workers->run_task(&enq_task_proxy);
2304 }
2305
2306 void ConcurrentMark::weakRefsWorkParallelPart(BoolObjectClosure* is_alive, bool purged_classes) {
2307 G1CollectedHeap::heap()->parallel_cleaning(is_alive, true, true, purged_classes);
2308 }
2309
2310 void ConcurrentMark::weakRefsWork(bool clear_all_soft_refs) {
2311 if (has_overflown()) {
2312 // Skip processing the discovered references if we have
2313 // overflown the global marking stack. Reference objects
2314 // only get discovered once so it is OK to not
2315 // de-populate the discovered reference lists. We could have,
2316 // but the only benefit would be that, when marking restarts,
2317 // less reference objects are discovered.
2318 return;
2319 }
2320
2321 ResourceMark rm;
2322 HandleMark hm;
2323
2324 G1CollectedHeap* g1h = G1CollectedHeap::heap();
2325
2326 // Is alive closure.
2327 G1CMIsAliveClosure g1_is_alive(g1h);
2328
2329 // Inner scope to exclude the cleaning of the string and symbol
2330 // tables from the displayed time.
2331 {
2332 G1CMTraceTime t("GC ref-proc", G1Log::finer());
2333
2334 ReferenceProcessor* rp = g1h->ref_processor_cm();
2335
2336 // See the comment in G1CollectedHeap::ref_processing_init()
2337 // about how reference processing currently works in G1.
2338
2339 // Set the soft reference policy
2340 rp->setup_policy(clear_all_soft_refs);
2341 assert(_markStack.isEmpty(), "mark stack should be empty");
2342
2343 // Instances of the 'Keep Alive' and 'Complete GC' closures used
2344 // in serial reference processing. Note these closures are also
2345 // used for serially processing (by the the current thread) the
2346 // JNI references during parallel reference processing.
2347 //
2348 // These closures do not need to synchronize with the worker
2349 // threads involved in parallel reference processing as these
2350 // instances are executed serially by the current thread (e.g.
2351 // reference processing is not multi-threaded and is thus
2352 // performed by the current thread instead of a gang worker).
2353 //
2354 // The gang tasks involved in parallel reference processing create
2355 // their own instances of these closures, which do their own
2356 // synchronization among themselves.
2357 G1CMKeepAliveAndDrainClosure g1_keep_alive(this, task(0), true /* is_serial */);
2358 G1CMDrainMarkingStackClosure g1_drain_mark_stack(this, task(0), true /* is_serial */);
2359
2360 // We need at least one active thread. If reference processing
2361 // is not multi-threaded we use the current (VMThread) thread,
2362 // otherwise we use the work gang from the G1CollectedHeap and
2363 // we utilize all the worker threads we can.
2364 bool processing_is_mt = rp->processing_is_mt();
2365 uint active_workers = (processing_is_mt ? g1h->workers()->active_workers() : 1U);
2366 active_workers = MAX2(MIN2(active_workers, _max_worker_id), 1U);
2367
2368 // Parallel processing task executor.
2369 G1CMRefProcTaskExecutor par_task_executor(g1h, this,
2370 g1h->workers(), active_workers);
2371 AbstractRefProcTaskExecutor* executor = (processing_is_mt ? &par_task_executor : NULL);
2372
2373 // Set the concurrency level. The phase was already set prior to
2374 // executing the remark task.
2375 set_concurrency(active_workers);
2376
2377 // Set the degree of MT processing here. If the discovery was done MT,
2378 // the number of threads involved during discovery could differ from
2379 // the number of active workers. This is OK as long as the discovered
2380 // Reference lists are balanced (see balance_all_queues() and balance_queues()).
2381 rp->set_active_mt_degree(active_workers);
2382
2383 // Process the weak references.
2384 const ReferenceProcessorStats& stats =
2385 rp->process_discovered_references(&g1_is_alive,
2386 &g1_keep_alive,
2387 &g1_drain_mark_stack,
2388 executor,
2389 g1h->gc_timer_cm(),
2390 concurrent_gc_id());
2391 g1h->gc_tracer_cm()->report_gc_reference_stats(stats);
2392
2393 // The do_oop work routines of the keep_alive and drain_marking_stack
2394 // oop closures will set the has_overflown flag if we overflow the
2395 // global marking stack.
2396
2397 assert(_markStack.overflow() || _markStack.isEmpty(),
2398 "mark stack should be empty (unless it overflowed)");
2399
2400 if (_markStack.overflow()) {
2401 // This should have been done already when we tried to push an
2402 // entry on to the global mark stack. But let's do it again.
2403 set_has_overflown();
2404 }
2405
2406 assert(rp->num_q() == active_workers, "why not");
2407
2408 rp->enqueue_discovered_references(executor);
2409
2410 rp->verify_no_references_recorded();
2411 assert(!rp->discovery_enabled(), "Post condition");
2412 }
2413
2414 if (has_overflown()) {
2415 // We can not trust g1_is_alive if the marking stack overflowed
2416 return;
2417 }
2418
2419 assert(_markStack.isEmpty(), "Marking should have completed");
2420
2421 // Unload Klasses, String, Symbols, Code Cache, etc.
2422 {
2423 G1CMTraceTime trace("Unloading", G1Log::finer());
2424
2425 if (ClassUnloadingWithConcurrentMark) {
2426 bool purged_classes;
2427
2428 {
2429 G1CMTraceTime trace("System Dictionary Unloading", G1Log::finest());
2430 purged_classes = SystemDictionary::do_unloading(&g1_is_alive, false /* Defer klass cleaning */);
2431 }
2432
2433 {
2434 G1CMTraceTime trace("Parallel Unloading", G1Log::finest());
2435 weakRefsWorkParallelPart(&g1_is_alive, purged_classes);
2436 }
2437 }
2438
2439 if (G1StringDedup::is_enabled()) {
2440 G1CMTraceTime trace("String Deduplication Unlink", G1Log::finest());
2441 G1StringDedup::unlink(&g1_is_alive);
2442 }
2443 }
2444 }
2445
2446 void ConcurrentMark::swapMarkBitMaps() {
2447 CMBitMapRO* temp = _prevMarkBitMap;
2448 _prevMarkBitMap = (CMBitMapRO*)_nextMarkBitMap;
2449 _nextMarkBitMap = (CMBitMap*) temp;
2450 }
2451
2452 // Closure for marking entries in SATB buffers.
2453 class CMSATBBufferClosure : public SATBBufferClosure {
2454 private:
2455 CMTask* _task;
2456 G1CollectedHeap* _g1h;
2457
2458 // This is very similar to CMTask::deal_with_reference, but with
2459 // more relaxed requirements for the argument, so this must be more
2460 // circumspect about treating the argument as an object.
2461 void do_entry(void* entry) const {
2462 _task->increment_refs_reached();
2463 HeapRegion* hr = _g1h->heap_region_containing_raw(entry);
2464 if (entry < hr->next_top_at_mark_start()) {
2465 // Until we get here, we don't know whether entry refers to a valid
2466 // object; it could instead have been a stale reference.
2467 oop obj = static_cast<oop>(entry);
2468 assert(obj->is_oop(true /* ignore mark word */),
2469 err_msg("Invalid oop in SATB buffer: " PTR_FORMAT, p2i(obj)));
2470 _task->make_reference_grey(obj, hr);
2471 }
2472 }
2473
2474 public:
2475 CMSATBBufferClosure(CMTask* task, G1CollectedHeap* g1h)
2476 : _task(task), _g1h(g1h) { }
2477
2478 virtual void do_buffer(void** buffer, size_t size) {
2479 for (size_t i = 0; i < size; ++i) {
2480 do_entry(buffer[i]);
2481 }
2482 }
2483 };
2484
2485 class G1RemarkThreadsClosure : public ThreadClosure {
2486 CMSATBBufferClosure _cm_satb_cl;
2487 G1CMOopClosure _cm_cl;
2488 MarkingCodeBlobClosure _code_cl;
2489 int _thread_parity;
2490
2491 public:
2492 G1RemarkThreadsClosure(G1CollectedHeap* g1h, CMTask* task) :
2493 _cm_satb_cl(task, g1h),
2494 _cm_cl(g1h, g1h->concurrent_mark(), task),
2495 _code_cl(&_cm_cl, !CodeBlobToOopClosure::FixRelocations),
2496 _thread_parity(Threads::thread_claim_parity()) {}
2497
2498 void do_thread(Thread* thread) {
2499 if (thread->is_Java_thread()) {
2500 if (thread->claim_oops_do(true, _thread_parity)) {
2501 JavaThread* jt = (JavaThread*)thread;
2502
2503 // In theory it should not be neccessary to explicitly walk the nmethods to find roots for concurrent marking
2504 // however the liveness of oops reachable from nmethods have very complex lifecycles:
2505 // * Alive if on the stack of an executing method
2506 // * Weakly reachable otherwise
2507 // Some objects reachable from nmethods, such as the class loader (or klass_holder) of the receiver should be
2508 // live by the SATB invariant but other oops recorded in nmethods may behave differently.
2509 jt->nmethods_do(&_code_cl);
2510
2511 jt->satb_mark_queue().apply_closure_and_empty(&_cm_satb_cl);
2512 }
2513 } else if (thread->is_VM_thread()) {
2514 if (thread->claim_oops_do(true, _thread_parity)) {
2515 JavaThread::satb_mark_queue_set().shared_satb_queue()->apply_closure_and_empty(&_cm_satb_cl);
2516 }
2517 }
2518 }
2519 };
2520
2521 class CMRemarkTask: public AbstractGangTask {
2522 private:
2523 ConcurrentMark* _cm;
2524 public:
2525 void work(uint worker_id) {
2526 // Since all available tasks are actually started, we should
2527 // only proceed if we're supposed to be active.
2528 if (worker_id < _cm->active_tasks()) {
2529 CMTask* task = _cm->task(worker_id);
2530 task->record_start_time();
2531 {
2532 ResourceMark rm;
2533 HandleMark hm;
2534
2535 G1RemarkThreadsClosure threads_f(G1CollectedHeap::heap(), task);
2536 Threads::threads_do(&threads_f);
2537 }
2538
2539 do {
2540 task->do_marking_step(1000000000.0 /* something very large */,
2541 true /* do_termination */,
2542 false /* is_serial */);
2543 } while (task->has_aborted() && !_cm->has_overflown());
2544 // If we overflow, then we do not want to restart. We instead
2545 // want to abort remark and do concurrent marking again.
2546 task->record_end_time();
2547 }
2548 }
2549
2550 CMRemarkTask(ConcurrentMark* cm, uint active_workers) :
2551 AbstractGangTask("Par Remark"), _cm(cm) {
2552 _cm->terminator()->reset_for_reuse(active_workers);
2553 }
2554 };
2555
2556 void ConcurrentMark::checkpointRootsFinalWork() {
2557 ResourceMark rm;
2558 HandleMark hm;
2559 G1CollectedHeap* g1h = G1CollectedHeap::heap();
2560
2561 G1CMTraceTime trace("Finalize Marking", G1Log::finer());
2562
2563 g1h->ensure_parsability(false);
2564
2565 // this is remark, so we'll use up all active threads
2566 uint active_workers = g1h->workers()->active_workers();
2567 set_concurrency_and_phase(active_workers, false /* concurrent */);
2568 // Leave _parallel_marking_threads at it's
2569 // value originally calculated in the ConcurrentMark
2570 // constructor and pass values of the active workers
2571 // through the gang in the task.
2572
2573 {
2574 StrongRootsScope srs(active_workers);
2575
2576 CMRemarkTask remarkTask(this, active_workers);
2577 // We will start all available threads, even if we decide that the
2578 // active_workers will be fewer. The extra ones will just bail out
2579 // immediately.
2580 g1h->workers()->run_task(&remarkTask);
2581 }
2582
2583 SATBMarkQueueSet& satb_mq_set = JavaThread::satb_mark_queue_set();
2584 guarantee(has_overflown() ||
2585 satb_mq_set.completed_buffers_num() == 0,
2586 err_msg("Invariant: has_overflown = %s, num buffers = %d",
2587 BOOL_TO_STR(has_overflown()),
2588 satb_mq_set.completed_buffers_num()));
2589
2590 print_stats();
2591 }
2592
2593 void ConcurrentMark::clearRangePrevBitmap(MemRegion mr) {
2594 // Note we are overriding the read-only view of the prev map here, via
2595 // the cast.
2596 ((CMBitMap*)_prevMarkBitMap)->clearRange(mr);
2597 }
2598
2599 void ConcurrentMark::clearRangeNextBitmap(MemRegion mr) {
2600 _nextMarkBitMap->clearRange(mr);
2601 }
2602
2603 HeapRegion*
2604 ConcurrentMark::claim_region(uint worker_id) {
2605 // "checkpoint" the finger
2606 HeapWord* finger = _finger;
2607
2608 // _heap_end will not change underneath our feet; it only changes at
2609 // yield points.
2610 while (finger < _heap_end) {
2611 assert(_g1h->is_in_g1_reserved(finger), "invariant");
2612
2613 // Note on how this code handles humongous regions. In the
2614 // normal case the finger will reach the start of a "starts
2615 // humongous" (SH) region. Its end will either be the end of the
2616 // last "continues humongous" (CH) region in the sequence, or the
2617 // standard end of the SH region (if the SH is the only region in
2618 // the sequence). That way claim_region() will skip over the CH
2619 // regions. However, there is a subtle race between a CM thread
2620 // executing this method and a mutator thread doing a humongous
2621 // object allocation. The two are not mutually exclusive as the CM
2622 // thread does not need to hold the Heap_lock when it gets
2623 // here. So there is a chance that claim_region() will come across
2624 // a free region that's in the progress of becoming a SH or a CH
2625 // region. In the former case, it will either
2626 // a) Miss the update to the region's end, in which case it will
2627 // visit every subsequent CH region, will find their bitmaps
2628 // empty, and do nothing, or
2629 // b) Will observe the update of the region's end (in which case
2630 // it will skip the subsequent CH regions).
2631 // If it comes across a region that suddenly becomes CH, the
2632 // scenario will be similar to b). So, the race between
2633 // claim_region() and a humongous object allocation might force us
2634 // to do a bit of unnecessary work (due to some unnecessary bitmap
2635 // iterations) but it should not introduce and correctness issues.
2636 HeapRegion* curr_region = _g1h->heap_region_containing_raw(finger);
2637
2638 // Above heap_region_containing_raw may return NULL as we always scan claim
2639 // until the end of the heap. In this case, just jump to the next region.
2640 HeapWord* end = curr_region != NULL ? curr_region->end() : finger + HeapRegion::GrainWords;
2641
2642 // Is the gap between reading the finger and doing the CAS too long?
2643 HeapWord* res = (HeapWord*) Atomic::cmpxchg_ptr(end, &_finger, finger);
2644 if (res == finger && curr_region != NULL) {
2645 // we succeeded
2646 HeapWord* bottom = curr_region->bottom();
2647 HeapWord* limit = curr_region->next_top_at_mark_start();
2648
2649 if (verbose_low()) {
2650 gclog_or_tty->print_cr("[%u] curr_region = "PTR_FORMAT" "
2651 "["PTR_FORMAT", "PTR_FORMAT"), "
2652 "limit = "PTR_FORMAT,
2653 worker_id, p2i(curr_region), p2i(bottom), p2i(end), p2i(limit));
2654 }
2655
2656 // notice that _finger == end cannot be guaranteed here since,
2657 // someone else might have moved the finger even further
2658 assert(_finger >= end, "the finger should have moved forward");
2659
2660 if (verbose_low()) {
2661 gclog_or_tty->print_cr("[%u] we were successful with region = "
2662 PTR_FORMAT, worker_id, p2i(curr_region));
2663 }
2664
2665 if (limit > bottom) {
2666 if (verbose_low()) {
2667 gclog_or_tty->print_cr("[%u] region "PTR_FORMAT" is not empty, "
2668 "returning it ", worker_id, p2i(curr_region));
2669 }
2670 return curr_region;
2671 } else {
2672 assert(limit == bottom,
2673 "the region limit should be at bottom");
2674 if (verbose_low()) {
2675 gclog_or_tty->print_cr("[%u] region "PTR_FORMAT" is empty, "
2676 "returning NULL", worker_id, p2i(curr_region));
2677 }
2678 // we return NULL and the caller should try calling
2679 // claim_region() again.
2680 return NULL;
2681 }
2682 } else {
2683 assert(_finger > finger, "the finger should have moved forward");
2684 if (verbose_low()) {
2685 if (curr_region == NULL) {
2686 gclog_or_tty->print_cr("[%u] found uncommitted region, moving finger, "
2687 "global finger = "PTR_FORMAT", "
2688 "our finger = "PTR_FORMAT,
2689 worker_id, p2i(_finger), p2i(finger));
2690 } else {
2691 gclog_or_tty->print_cr("[%u] somebody else moved the finger, "
2692 "global finger = "PTR_FORMAT", "
2693 "our finger = "PTR_FORMAT,
2694 worker_id, p2i(_finger), p2i(finger));
2695 }
2696 }
2697
2698 // read it again
2699 finger = _finger;
2700 }
2701 }
2702
2703 return NULL;
2704 }
2705
2706 #ifndef PRODUCT
2707 enum VerifyNoCSetOopsPhase {
2708 VerifyNoCSetOopsStack,
2709 VerifyNoCSetOopsQueues
2710 };
2711
2712 class VerifyNoCSetOopsClosure : public OopClosure, public ObjectClosure {
2713 private:
2714 G1CollectedHeap* _g1h;
2715 VerifyNoCSetOopsPhase _phase;
2716 int _info;
2717
2718 const char* phase_str() {
2719 switch (_phase) {
2720 case VerifyNoCSetOopsStack: return "Stack";
2721 case VerifyNoCSetOopsQueues: return "Queue";
2722 default: ShouldNotReachHere();
2723 }
2724 return NULL;
2725 }
2726
2727 void do_object_work(oop obj) {
2728 guarantee(!_g1h->obj_in_cs(obj),
2729 err_msg("obj: "PTR_FORMAT" in CSet, phase: %s, info: %d",
2730 p2i((void*) obj), phase_str(), _info));
2731 }
2732
2733 public:
2734 VerifyNoCSetOopsClosure() : _g1h(G1CollectedHeap::heap()) { }
2735
2736 void set_phase(VerifyNoCSetOopsPhase phase, int info = -1) {
2737 _phase = phase;
2738 _info = info;
2739 }
2740
2741 virtual void do_oop(oop* p) {
2742 oop obj = oopDesc::load_decode_heap_oop(p);
2743 do_object_work(obj);
2744 }
2745
2746 virtual void do_oop(narrowOop* p) {
2747 // We should not come across narrow oops while scanning marking
2748 // stacks
2749 ShouldNotReachHere();
2750 }
2751
2752 virtual void do_object(oop obj) {
2753 do_object_work(obj);
2754 }
2755 };
2756
2757 void ConcurrentMark::verify_no_cset_oops() {
2758 assert(SafepointSynchronize::is_at_safepoint(), "should be at a safepoint");
2759 if (!G1CollectedHeap::heap()->mark_in_progress()) {
2760 return;
2761 }
2762
2763 VerifyNoCSetOopsClosure cl;
2764
2765 // Verify entries on the global mark stack
2766 cl.set_phase(VerifyNoCSetOopsStack);
2767 _markStack.oops_do(&cl);
2768
2769 // Verify entries on the task queues
2770 for (uint i = 0; i < _max_worker_id; i += 1) {
2771 cl.set_phase(VerifyNoCSetOopsQueues, i);
2772 CMTaskQueue* queue = _task_queues->queue(i);
2773 queue->oops_do(&cl);
2774 }
2775
2776 // Verify the global finger
2777 HeapWord* global_finger = finger();
2778 if (global_finger != NULL && global_finger < _heap_end) {
2779 // The global finger always points to a heap region boundary. We
2780 // use heap_region_containing_raw() to get the containing region
2781 // given that the global finger could be pointing to a free region
2782 // which subsequently becomes continues humongous. If that
2783 // happens, heap_region_containing() will return the bottom of the
2784 // corresponding starts humongous region and the check below will
2785 // not hold any more.
2786 // Since we always iterate over all regions, we might get a NULL HeapRegion
2787 // here.
2788 HeapRegion* global_hr = _g1h->heap_region_containing_raw(global_finger);
2789 guarantee(global_hr == NULL || global_finger == global_hr->bottom(),
2790 err_msg("global finger: "PTR_FORMAT" region: "HR_FORMAT,
2791 p2i(global_finger), HR_FORMAT_PARAMS(global_hr)));
2792 }
2793
2794 // Verify the task fingers
2795 assert(parallel_marking_threads() <= _max_worker_id, "sanity");
2796 for (int i = 0; i < (int) parallel_marking_threads(); i += 1) {
2797 CMTask* task = _tasks[i];
2798 HeapWord* task_finger = task->finger();
2799 if (task_finger != NULL && task_finger < _heap_end) {
2800 // See above note on the global finger verification.
2801 HeapRegion* task_hr = _g1h->heap_region_containing_raw(task_finger);
2802 guarantee(task_hr == NULL || task_finger == task_hr->bottom() ||
2803 !task_hr->in_collection_set(),
2804 err_msg("task finger: "PTR_FORMAT" region: "HR_FORMAT,
2805 p2i(task_finger), HR_FORMAT_PARAMS(task_hr)));
2806 }
2807 }
2808 }
2809 #endif // PRODUCT
2810
2811 // Aggregate the counting data that was constructed concurrently
2812 // with marking.
2813 class AggregateCountDataHRClosure: public HeapRegionClosure {
2814 G1CollectedHeap* _g1h;
2815 ConcurrentMark* _cm;
2816 CardTableModRefBS* _ct_bs;
2817 BitMap* _cm_card_bm;
2818 uint _max_worker_id;
2819
2820 public:
2821 AggregateCountDataHRClosure(G1CollectedHeap* g1h,
2822 BitMap* cm_card_bm,
2823 uint max_worker_id) :
2824 _g1h(g1h), _cm(g1h->concurrent_mark()),
2825 _ct_bs(barrier_set_cast<CardTableModRefBS>(g1h->barrier_set())),
2826 _cm_card_bm(cm_card_bm), _max_worker_id(max_worker_id) { }
2827
2828 bool doHeapRegion(HeapRegion* hr) {
2829 if (hr->is_continues_humongous()) {
2830 // We will ignore these here and process them when their
2831 // associated "starts humongous" region is processed.
2832 // Note that we cannot rely on their associated
2833 // "starts humongous" region to have their bit set to 1
2834 // since, due to the region chunking in the parallel region
2835 // iteration, a "continues humongous" region might be visited
2836 // before its associated "starts humongous".
2837 return false;
2838 }
2839
2840 HeapWord* start = hr->bottom();
2841 HeapWord* limit = hr->next_top_at_mark_start();
2842 HeapWord* end = hr->end();
2843
2844 assert(start <= limit && limit <= hr->top() && hr->top() <= hr->end(),
2845 err_msg("Preconditions not met - "
2846 "start: "PTR_FORMAT", limit: "PTR_FORMAT", "
2847 "top: "PTR_FORMAT", end: "PTR_FORMAT,
2848 p2i(start), p2i(limit), p2i(hr->top()), p2i(hr->end())));
2849
2850 assert(hr->next_marked_bytes() == 0, "Precondition");
2851
2852 if (start == limit) {
2853 // NTAMS of this region has not been set so nothing to do.
2854 return false;
2855 }
2856
2857 // 'start' should be in the heap.
2858 assert(_g1h->is_in_g1_reserved(start) && _ct_bs->is_card_aligned(start), "sanity");
2859 // 'end' *may* be just beyond the end of the heap (if hr is the last region)
2860 assert(!_g1h->is_in_g1_reserved(end) || _ct_bs->is_card_aligned(end), "sanity");
2861
2862 BitMap::idx_t start_idx = _cm->card_bitmap_index_for(start);
2863 BitMap::idx_t limit_idx = _cm->card_bitmap_index_for(limit);
2864 BitMap::idx_t end_idx = _cm->card_bitmap_index_for(end);
2865
2866 // If ntams is not card aligned then we bump card bitmap index
2867 // for limit so that we get the all the cards spanned by
2868 // the object ending at ntams.
2869 // Note: if this is the last region in the heap then ntams
2870 // could be actually just beyond the end of the the heap;
2871 // limit_idx will then correspond to a (non-existent) card
2872 // that is also outside the heap.
2873 if (_g1h->is_in_g1_reserved(limit) && !_ct_bs->is_card_aligned(limit)) {
2874 limit_idx += 1;
2875 }
2876
2877 assert(limit_idx <= end_idx, "or else use atomics");
2878
2879 // Aggregate the "stripe" in the count data associated with hr.
2880 uint hrm_index = hr->hrm_index();
2881 size_t marked_bytes = 0;
2882
2883 for (uint i = 0; i < _max_worker_id; i += 1) {
2884 size_t* marked_bytes_array = _cm->count_marked_bytes_array_for(i);
2885 BitMap* task_card_bm = _cm->count_card_bitmap_for(i);
2886
2887 // Fetch the marked_bytes in this region for task i and
2888 // add it to the running total for this region.
2889 marked_bytes += marked_bytes_array[hrm_index];
2890
2891 // Now union the bitmaps[0,max_worker_id)[start_idx..limit_idx)
2892 // into the global card bitmap.
2893 BitMap::idx_t scan_idx = task_card_bm->get_next_one_offset(start_idx, limit_idx);
2894
2895 while (scan_idx < limit_idx) {
2896 assert(task_card_bm->at(scan_idx) == true, "should be");
2897 _cm_card_bm->set_bit(scan_idx);
2898 assert(_cm_card_bm->at(scan_idx) == true, "should be");
2899
2900 // BitMap::get_next_one_offset() can handle the case when
2901 // its left_offset parameter is greater than its right_offset
2902 // parameter. It does, however, have an early exit if
2903 // left_offset == right_offset. So let's limit the value
2904 // passed in for left offset here.
2905 BitMap::idx_t next_idx = MIN2(scan_idx + 1, limit_idx);
2906 scan_idx = task_card_bm->get_next_one_offset(next_idx, limit_idx);
2907 }
2908 }
2909
2910 // Update the marked bytes for this region.
2911 hr->add_to_marked_bytes(marked_bytes);
2912
2913 // Next heap region
2914 return false;
2915 }
2916 };
2917
2918 class G1AggregateCountDataTask: public AbstractGangTask {
2919 protected:
2920 G1CollectedHeap* _g1h;
2921 ConcurrentMark* _cm;
2922 BitMap* _cm_card_bm;
2923 uint _max_worker_id;
2924 uint _active_workers;
2925 HeapRegionClaimer _hrclaimer;
2926
2927 public:
2928 G1AggregateCountDataTask(G1CollectedHeap* g1h,
2929 ConcurrentMark* cm,
2930 BitMap* cm_card_bm,
2931 uint max_worker_id,
2932 uint n_workers) :
2933 AbstractGangTask("Count Aggregation"),
2934 _g1h(g1h), _cm(cm), _cm_card_bm(cm_card_bm),
2935 _max_worker_id(max_worker_id),
2936 _active_workers(n_workers),
2937 _hrclaimer(_active_workers) {
2938 }
2939
2940 void work(uint worker_id) {
2941 AggregateCountDataHRClosure cl(_g1h, _cm_card_bm, _max_worker_id);
2942
2943 _g1h->heap_region_par_iterate(&cl, worker_id, &_hrclaimer);
2944 }
2945 };
2946
2947
2948 void ConcurrentMark::aggregate_count_data() {
2949 uint n_workers = _g1h->workers()->active_workers();
2950
2951 G1AggregateCountDataTask g1_par_agg_task(_g1h, this, &_card_bm,
2952 _max_worker_id, n_workers);
2953
2954 _g1h->workers()->run_task(&g1_par_agg_task);
2955 }
2956
2957 // Clear the per-worker arrays used to store the per-region counting data
2958 void ConcurrentMark::clear_all_count_data() {
2959 // Clear the global card bitmap - it will be filled during
2960 // liveness count aggregation (during remark) and the
2961 // final counting task.
2962 _card_bm.clear();
2963
2964 // Clear the global region bitmap - it will be filled as part
2965 // of the final counting task.
2966 _region_bm.clear();
2967
2968 uint max_regions = _g1h->max_regions();
2969 assert(_max_worker_id > 0, "uninitialized");
2970
2971 for (uint i = 0; i < _max_worker_id; i += 1) {
2972 BitMap* task_card_bm = count_card_bitmap_for(i);
2973 size_t* marked_bytes_array = count_marked_bytes_array_for(i);
2974
2975 assert(task_card_bm->size() == _card_bm.size(), "size mismatch");
2976 assert(marked_bytes_array != NULL, "uninitialized");
2977
2978 memset(marked_bytes_array, 0, (size_t) max_regions * sizeof(size_t));
2979 task_card_bm->clear();
2980 }
2981 }
2982
2983 void ConcurrentMark::print_stats() {
2984 if (verbose_stats()) {
2985 gclog_or_tty->print_cr("---------------------------------------------------------------------");
2986 for (size_t i = 0; i < _active_tasks; ++i) {
2987 _tasks[i]->print_stats();
2988 gclog_or_tty->print_cr("---------------------------------------------------------------------");
2989 }
2990 }
2991 }
2992
2993 // abandon current marking iteration due to a Full GC
2994 void ConcurrentMark::abort() {
2995 // Clear all marks in the next bitmap for the next marking cycle. This will allow us to skip the next
2996 // concurrent bitmap clearing.
2997 _nextMarkBitMap->clearAll();
2998
2999 // Note we cannot clear the previous marking bitmap here
3000 // since VerifyDuringGC verifies the objects marked during
3001 // a full GC against the previous bitmap.
3002
3003 // Clear the liveness counting data
3004 clear_all_count_data();
3005 // Empty mark stack
3006 reset_marking_state();
3007 for (uint i = 0; i < _max_worker_id; ++i) {
3008 _tasks[i]->clear_region_fields();
3009 }
3010 _first_overflow_barrier_sync.abort();
3011 _second_overflow_barrier_sync.abort();
3012 const GCId& gc_id = _g1h->gc_tracer_cm()->gc_id();
3013 if (!gc_id.is_undefined()) {
3014 // We can do multiple full GCs before ConcurrentMarkThread::run() gets a chance
3015 // to detect that it was aborted. Only keep track of the first GC id that we aborted.
3016 _aborted_gc_id = gc_id;
3017 }
3018 _has_aborted = true;
3019
3020 SATBMarkQueueSet& satb_mq_set = JavaThread::satb_mark_queue_set();
3021 satb_mq_set.abandon_partial_marking();
3022 // This can be called either during or outside marking, we'll read
3023 // the expected_active value from the SATB queue set.
3024 satb_mq_set.set_active_all_threads(
3025 false, /* new active value */
3026 satb_mq_set.is_active() /* expected_active */);
3027
3028 _g1h->trace_heap_after_concurrent_cycle();
3029 _g1h->register_concurrent_cycle_end();
3030 }
3031
3032 const GCId& ConcurrentMark::concurrent_gc_id() {
3033 if (has_aborted()) {
3034 return _aborted_gc_id;
3035 }
3036 return _g1h->gc_tracer_cm()->gc_id();
3037 }
3038
3039 static void print_ms_time_info(const char* prefix, const char* name,
3040 NumberSeq& ns) {
3041 gclog_or_tty->print_cr("%s%5d %12s: total time = %8.2f s (avg = %8.2f ms).",
3042 prefix, ns.num(), name, ns.sum()/1000.0, ns.avg());
3043 if (ns.num() > 0) {
3044 gclog_or_tty->print_cr("%s [std. dev = %8.2f ms, max = %8.2f ms]",
3045 prefix, ns.sd(), ns.maximum());
3046 }
3047 }
3048
3049 void ConcurrentMark::print_summary_info() {
3050 gclog_or_tty->print_cr(" Concurrent marking:");
3051 print_ms_time_info(" ", "init marks", _init_times);
3052 print_ms_time_info(" ", "remarks", _remark_times);
3053 {
3054 print_ms_time_info(" ", "final marks", _remark_mark_times);
3055 print_ms_time_info(" ", "weak refs", _remark_weak_ref_times);
3056
3057 }
3058 print_ms_time_info(" ", "cleanups", _cleanup_times);
3059 gclog_or_tty->print_cr(" Final counting total time = %8.2f s (avg = %8.2f ms).",
3060 _total_counting_time,
3061 (_cleanup_times.num() > 0 ? _total_counting_time * 1000.0 /
3062 (double)_cleanup_times.num()
3063 : 0.0));
3064 if (G1ScrubRemSets) {
3065 gclog_or_tty->print_cr(" RS scrub total time = %8.2f s (avg = %8.2f ms).",
3066 _total_rs_scrub_time,
3067 (_cleanup_times.num() > 0 ? _total_rs_scrub_time * 1000.0 /
3068 (double)_cleanup_times.num()
3069 : 0.0));
3070 }
3071 gclog_or_tty->print_cr(" Total stop_world time = %8.2f s.",
3072 (_init_times.sum() + _remark_times.sum() +
3073 _cleanup_times.sum())/1000.0);
3074 gclog_or_tty->print_cr(" Total concurrent time = %8.2f s "
3075 "(%8.2f s marking).",
3076 cmThread()->vtime_accum(),
3077 cmThread()->vtime_mark_accum());
3078 }
3079
3080 void ConcurrentMark::print_worker_threads_on(outputStream* st) const {
3081 _parallel_workers->print_worker_threads_on(st);
3082 }
3083
3084 void ConcurrentMark::print_on_error(outputStream* st) const {
3085 st->print_cr("Marking Bits (Prev, Next): (CMBitMap*) " PTR_FORMAT ", (CMBitMap*) " PTR_FORMAT,
3086 p2i(_prevMarkBitMap), p2i(_nextMarkBitMap));
3087 _prevMarkBitMap->print_on_error(st, " Prev Bits: ");
3088 _nextMarkBitMap->print_on_error(st, " Next Bits: ");
3089 }
3090
3091 // We take a break if someone is trying to stop the world.
3092 bool ConcurrentMark::do_yield_check(uint worker_id) {
3093 if (SuspendibleThreadSet::should_yield()) {
3094 if (worker_id == 0) {
3095 _g1h->g1_policy()->record_concurrent_pause();
3096 }
3097 SuspendibleThreadSet::yield();
3098 return true;
3099 } else {
3100 return false;
3101 }
3102 }
3103
3104 #ifndef PRODUCT
3105 // for debugging purposes
3106 void ConcurrentMark::print_finger() {
3107 gclog_or_tty->print_cr("heap ["PTR_FORMAT", "PTR_FORMAT"), global finger = "PTR_FORMAT,
3108 p2i(_heap_start), p2i(_heap_end), p2i(_finger));
3109 for (uint i = 0; i < _max_worker_id; ++i) {
3110 gclog_or_tty->print(" %u: " PTR_FORMAT, i, p2i(_tasks[i]->finger()));
3111 }
3112 gclog_or_tty->cr();
3113 }
3114 #endif
3115
3116 template<bool scan>
3117 inline void CMTask::process_grey_object(oop obj) {
3118 assert(scan || obj->is_typeArray(), "Skipping scan of grey non-typeArray");
3119 assert(_nextMarkBitMap->isMarked((HeapWord*) obj), "invariant");
3120
3121 if (_cm->verbose_high()) {
3122 gclog_or_tty->print_cr("[%u] processing grey object " PTR_FORMAT,
3123 _worker_id, p2i((void*) obj));
3124 }
3125
3126 size_t obj_size = obj->size();
3127 _words_scanned += obj_size;
3128
3129 if (scan) {
3130 obj->oop_iterate(_cm_oop_closure);
3131 }
3132 statsOnly( ++_objs_scanned );
3133 check_limits();
3134 }
3135
3136 template void CMTask::process_grey_object<true>(oop);
3137 template void CMTask::process_grey_object<false>(oop);
3138
3139 // Closure for iteration over bitmaps
3140 class CMBitMapClosure : public BitMapClosure {
3141 private:
3142 // the bitmap that is being iterated over
3143 CMBitMap* _nextMarkBitMap;
3144 ConcurrentMark* _cm;
3145 CMTask* _task;
3146
3147 public:
3148 CMBitMapClosure(CMTask *task, ConcurrentMark* cm, CMBitMap* nextMarkBitMap) :
3149 _task(task), _cm(cm), _nextMarkBitMap(nextMarkBitMap) { }
3150
3151 bool do_bit(size_t offset) {
3152 HeapWord* addr = _nextMarkBitMap->offsetToHeapWord(offset);
3153 assert(_nextMarkBitMap->isMarked(addr), "invariant");
3154 assert( addr < _cm->finger(), "invariant");
3155
3156 statsOnly( _task->increase_objs_found_on_bitmap() );
3157 assert(addr >= _task->finger(), "invariant");
3158
3159 // We move that task's local finger along.
3160 _task->move_finger_to(addr);
3161
3162 _task->scan_object(oop(addr));
3163 // we only partially drain the local queue and global stack
3164 _task->drain_local_queue(true);
3165 _task->drain_global_stack(true);
3166
3167 // if the has_aborted flag has been raised, we need to bail out of
3168 // the iteration
3169 return !_task->has_aborted();
3170 }
3171 };
3172
3173 G1CMOopClosure::G1CMOopClosure(G1CollectedHeap* g1h,
3174 ConcurrentMark* cm,
3175 CMTask* task)
3176 : _g1h(g1h), _cm(cm), _task(task) {
3177 assert(_ref_processor == NULL, "should be initialized to NULL");
3178
3179 if (G1UseConcMarkReferenceProcessing) {
3180 _ref_processor = g1h->ref_processor_cm();
3181 assert(_ref_processor != NULL, "should not be NULL");
3182 }
3183 }
3184
3185 void CMTask::setup_for_region(HeapRegion* hr) {
3186 assert(hr != NULL,
3187 "claim_region() should have filtered out NULL regions");
3188 assert(!hr->is_continues_humongous(),
3189 "claim_region() should have filtered out continues humongous regions");
3190
3191 if (_cm->verbose_low()) {
3192 gclog_or_tty->print_cr("[%u] setting up for region "PTR_FORMAT,
3193 _worker_id, p2i(hr));
3194 }
3195
3196 _curr_region = hr;
3197 _finger = hr->bottom();
3198 update_region_limit();
3199 }
3200
3201 void CMTask::update_region_limit() {
3202 HeapRegion* hr = _curr_region;
3203 HeapWord* bottom = hr->bottom();
3204 HeapWord* limit = hr->next_top_at_mark_start();
3205
3206 if (limit == bottom) {
3207 if (_cm->verbose_low()) {
3208 gclog_or_tty->print_cr("[%u] found an empty region "
3209 "["PTR_FORMAT", "PTR_FORMAT")",
3210 _worker_id, p2i(bottom), p2i(limit));
3211 }
3212 // The region was collected underneath our feet.
3213 // We set the finger to bottom to ensure that the bitmap
3214 // iteration that will follow this will not do anything.
3215 // (this is not a condition that holds when we set the region up,
3216 // as the region is not supposed to be empty in the first place)
3217 _finger = bottom;
3218 } else if (limit >= _region_limit) {
3219 assert(limit >= _finger, "peace of mind");
3220 } else {
3221 assert(limit < _region_limit, "only way to get here");
3222 // This can happen under some pretty unusual circumstances. An
3223 // evacuation pause empties the region underneath our feet (NTAMS
3224 // at bottom). We then do some allocation in the region (NTAMS
3225 // stays at bottom), followed by the region being used as a GC
3226 // alloc region (NTAMS will move to top() and the objects
3227 // originally below it will be grayed). All objects now marked in
3228 // the region are explicitly grayed, if below the global finger,
3229 // and we do not need in fact to scan anything else. So, we simply
3230 // set _finger to be limit to ensure that the bitmap iteration
3231 // doesn't do anything.
3232 _finger = limit;
3233 }
3234
3235 _region_limit = limit;
3236 }
3237
3238 void CMTask::giveup_current_region() {
3239 assert(_curr_region != NULL, "invariant");
3240 if (_cm->verbose_low()) {
3241 gclog_or_tty->print_cr("[%u] giving up region "PTR_FORMAT,
3242 _worker_id, p2i(_curr_region));
3243 }
3244 clear_region_fields();
3245 }
3246
3247 void CMTask::clear_region_fields() {
3248 // Values for these three fields that indicate that we're not
3249 // holding on to a region.
3250 _curr_region = NULL;
3251 _finger = NULL;
3252 _region_limit = NULL;
3253 }
3254
3255 void CMTask::set_cm_oop_closure(G1CMOopClosure* cm_oop_closure) {
3256 if (cm_oop_closure == NULL) {
3257 assert(_cm_oop_closure != NULL, "invariant");
3258 } else {
3259 assert(_cm_oop_closure == NULL, "invariant");
3260 }
3261 _cm_oop_closure = cm_oop_closure;
3262 }
3263
3264 void CMTask::reset(CMBitMap* nextMarkBitMap) {
3265 guarantee(nextMarkBitMap != NULL, "invariant");
3266
3267 if (_cm->verbose_low()) {
3268 gclog_or_tty->print_cr("[%u] resetting", _worker_id);
3269 }
3270
3271 _nextMarkBitMap = nextMarkBitMap;
3272 clear_region_fields();
3273
3274 _calls = 0;
3275 _elapsed_time_ms = 0.0;
3276 _termination_time_ms = 0.0;
3277 _termination_start_time_ms = 0.0;
3278
3279 #if _MARKING_STATS_
3280 _aborted = 0;
3281 _aborted_overflow = 0;
3282 _aborted_cm_aborted = 0;
3283 _aborted_yield = 0;
3284 _aborted_timed_out = 0;
3285 _aborted_satb = 0;
3286 _aborted_termination = 0;
3287 _steal_attempts = 0;
3288 _steals = 0;
3289 _local_pushes = 0;
3290 _local_pops = 0;
3291 _local_max_size = 0;
3292 _objs_scanned = 0;
3293 _global_pushes = 0;
3294 _global_pops = 0;
3295 _global_max_size = 0;
3296 _global_transfers_to = 0;
3297 _global_transfers_from = 0;
3298 _regions_claimed = 0;
3299 _objs_found_on_bitmap = 0;
3300 _satb_buffers_processed = 0;
3301 #endif // _MARKING_STATS_
3302 }
3303
3304 bool CMTask::should_exit_termination() {
3305 regular_clock_call();
3306 // This is called when we are in the termination protocol. We should
3307 // quit if, for some reason, this task wants to abort or the global
3308 // stack is not empty (this means that we can get work from it).
3309 return !_cm->mark_stack_empty() || has_aborted();
3310 }
3311
3312 void CMTask::reached_limit() {
3313 assert(_words_scanned >= _words_scanned_limit ||
3314 _refs_reached >= _refs_reached_limit ,
3315 "shouldn't have been called otherwise");
3316 regular_clock_call();
3317 }
3318
3319 void CMTask::regular_clock_call() {
3320 if (has_aborted()) return;
3321
3322 // First, we need to recalculate the words scanned and refs reached
3323 // limits for the next clock call.
3324 recalculate_limits();
3325
3326 // During the regular clock call we do the following
3327
3328 // (1) If an overflow has been flagged, then we abort.
3329 if (_cm->has_overflown()) {
3330 set_has_aborted();
3331 return;
3332 }
3333
3334 // If we are not concurrent (i.e. we're doing remark) we don't need
3335 // to check anything else. The other steps are only needed during
3336 // the concurrent marking phase.
3337 if (!concurrent()) return;
3338
3339 // (2) If marking has been aborted for Full GC, then we also abort.
3340 if (_cm->has_aborted()) {
3341 set_has_aborted();
3342 statsOnly( ++_aborted_cm_aborted );
3343 return;
3344 }
3345
3346 double curr_time_ms = os::elapsedVTime() * 1000.0;
3347
3348 // (3) If marking stats are enabled, then we update the step history.
3349 #if _MARKING_STATS_
3350 if (_words_scanned >= _words_scanned_limit) {
3351 ++_clock_due_to_scanning;
3352 }
3353 if (_refs_reached >= _refs_reached_limit) {
3354 ++_clock_due_to_marking;
3355 }
3356
3357 double last_interval_ms = curr_time_ms - _interval_start_time_ms;
3358 _interval_start_time_ms = curr_time_ms;
3359 _all_clock_intervals_ms.add(last_interval_ms);
3360
3361 if (_cm->verbose_medium()) {
3362 gclog_or_tty->print_cr("[%u] regular clock, interval = %1.2lfms, "
3363 "scanned = "SIZE_FORMAT"%s, refs reached = "SIZE_FORMAT"%s",
3364 _worker_id, last_interval_ms,
3365 _words_scanned,
3366 (_words_scanned >= _words_scanned_limit) ? " (*)" : "",
3367 _refs_reached,
3368 (_refs_reached >= _refs_reached_limit) ? " (*)" : "");
3369 }
3370 #endif // _MARKING_STATS_
3371
3372 // (4) We check whether we should yield. If we have to, then we abort.
3373 if (SuspendibleThreadSet::should_yield()) {
3374 // We should yield. To do this we abort the task. The caller is
3375 // responsible for yielding.
3376 set_has_aborted();
3377 statsOnly( ++_aborted_yield );
3378 return;
3379 }
3380
3381 // (5) We check whether we've reached our time quota. If we have,
3382 // then we abort.
3383 double elapsed_time_ms = curr_time_ms - _start_time_ms;
3384 if (elapsed_time_ms > _time_target_ms) {
3385 set_has_aborted();
3386 _has_timed_out = true;
3387 statsOnly( ++_aborted_timed_out );
3388 return;
3389 }
3390
3391 // (6) Finally, we check whether there are enough completed STAB
3392 // buffers available for processing. If there are, we abort.
3393 SATBMarkQueueSet& satb_mq_set = JavaThread::satb_mark_queue_set();
3394 if (!_draining_satb_buffers && satb_mq_set.process_completed_buffers()) {
3395 if (_cm->verbose_low()) {
3396 gclog_or_tty->print_cr("[%u] aborting to deal with pending SATB buffers",
3397 _worker_id);
3398 }
3399 // we do need to process SATB buffers, we'll abort and restart
3400 // the marking task to do so
3401 set_has_aborted();
3402 statsOnly( ++_aborted_satb );
3403 return;
3404 }
3405 }
3406
3407 void CMTask::recalculate_limits() {
3408 _real_words_scanned_limit = _words_scanned + words_scanned_period;
3409 _words_scanned_limit = _real_words_scanned_limit;
3410
3411 _real_refs_reached_limit = _refs_reached + refs_reached_period;
3412 _refs_reached_limit = _real_refs_reached_limit;
3413 }
3414
3415 void CMTask::decrease_limits() {
3416 // This is called when we believe that we're going to do an infrequent
3417 // operation which will increase the per byte scanned cost (i.e. move
3418 // entries to/from the global stack). It basically tries to decrease the
3419 // scanning limit so that the clock is called earlier.
3420
3421 if (_cm->verbose_medium()) {
3422 gclog_or_tty->print_cr("[%u] decreasing limits", _worker_id);
3423 }
3424
3425 _words_scanned_limit = _real_words_scanned_limit -
3426 3 * words_scanned_period / 4;
3427 _refs_reached_limit = _real_refs_reached_limit -
3428 3 * refs_reached_period / 4;
3429 }
3430
3431 void CMTask::move_entries_to_global_stack() {
3432 // local array where we'll store the entries that will be popped
3433 // from the local queue
3434 oop buffer[global_stack_transfer_size];
3435
3436 int n = 0;
3437 oop obj;
3438 while (n < global_stack_transfer_size && _task_queue->pop_local(obj)) {
3439 buffer[n] = obj;
3440 ++n;
3441 }
3442
3443 if (n > 0) {
3444 // we popped at least one entry from the local queue
3445
3446 statsOnly( ++_global_transfers_to; _local_pops += n );
3447
3448 if (!_cm->mark_stack_push(buffer, n)) {
3449 if (_cm->verbose_low()) {
3450 gclog_or_tty->print_cr("[%u] aborting due to global stack overflow",
3451 _worker_id);
3452 }
3453 set_has_aborted();
3454 } else {
3455 // the transfer was successful
3456
3457 if (_cm->verbose_medium()) {
3458 gclog_or_tty->print_cr("[%u] pushed %d entries to the global stack",
3459 _worker_id, n);
3460 }
3461 statsOnly( size_t tmp_size = _cm->mark_stack_size();
3462 if (tmp_size > _global_max_size) {
3463 _global_max_size = tmp_size;
3464 }
3465 _global_pushes += n );
3466 }
3467 }
3468
3469 // this operation was quite expensive, so decrease the limits
3470 decrease_limits();
3471 }
3472
3473 void CMTask::get_entries_from_global_stack() {
3474 // local array where we'll store the entries that will be popped
3475 // from the global stack.
3476 oop buffer[global_stack_transfer_size];
3477 int n;
3478 _cm->mark_stack_pop(buffer, global_stack_transfer_size, &n);
3479 assert(n <= global_stack_transfer_size,
3480 "we should not pop more than the given limit");
3481 if (n > 0) {
3482 // yes, we did actually pop at least one entry
3483
3484 statsOnly( ++_global_transfers_from; _global_pops += n );
3485 if (_cm->verbose_medium()) {
3486 gclog_or_tty->print_cr("[%u] popped %d entries from the global stack",
3487 _worker_id, n);
3488 }
3489 for (int i = 0; i < n; ++i) {
3490 bool success = _task_queue->push(buffer[i]);
3491 // We only call this when the local queue is empty or under a
3492 // given target limit. So, we do not expect this push to fail.
3493 assert(success, "invariant");
3494 }
3495
3496 statsOnly( size_t tmp_size = (size_t)_task_queue->size();
3497 if (tmp_size > _local_max_size) {
3498 _local_max_size = tmp_size;
3499 }
3500 _local_pushes += n );
3501 }
3502
3503 // this operation was quite expensive, so decrease the limits
3504 decrease_limits();
3505 }
3506
3507 void CMTask::drain_local_queue(bool partially) {
3508 if (has_aborted()) return;
3509
3510 // Decide what the target size is, depending whether we're going to
3511 // drain it partially (so that other tasks can steal if they run out
3512 // of things to do) or totally (at the very end).
3513 size_t target_size;
3514 if (partially) {
3515 target_size = MIN2((size_t)_task_queue->max_elems()/3, GCDrainStackTargetSize);
3516 } else {
3517 target_size = 0;
3518 }
3519
3520 if (_task_queue->size() > target_size) {
3521 if (_cm->verbose_high()) {
3522 gclog_or_tty->print_cr("[%u] draining local queue, target size = " SIZE_FORMAT,
3523 _worker_id, target_size);
3524 }
3525
3526 oop obj;
3527 bool ret = _task_queue->pop_local(obj);
3528 while (ret) {
3529 statsOnly( ++_local_pops );
3530
3531 if (_cm->verbose_high()) {
3532 gclog_or_tty->print_cr("[%u] popped "PTR_FORMAT, _worker_id,
3533 p2i((void*) obj));
3534 }
3535
3536 assert(_g1h->is_in_g1_reserved((HeapWord*) obj), "invariant" );
3537 assert(!_g1h->is_on_master_free_list(
3538 _g1h->heap_region_containing((HeapWord*) obj)), "invariant");
3539
3540 scan_object(obj);
3541
3542 if (_task_queue->size() <= target_size || has_aborted()) {
3543 ret = false;
3544 } else {
3545 ret = _task_queue->pop_local(obj);
3546 }
3547 }
3548
3549 if (_cm->verbose_high()) {
3550 gclog_or_tty->print_cr("[%u] drained local queue, size = %u",
3551 _worker_id, _task_queue->size());
3552 }
3553 }
3554 }
3555
3556 void CMTask::drain_global_stack(bool partially) {
3557 if (has_aborted()) return;
3558
3559 // We have a policy to drain the local queue before we attempt to
3560 // drain the global stack.
3561 assert(partially || _task_queue->size() == 0, "invariant");
3562
3563 // Decide what the target size is, depending whether we're going to
3564 // drain it partially (so that other tasks can steal if they run out
3565 // of things to do) or totally (at the very end). Notice that,
3566 // because we move entries from the global stack in chunks or
3567 // because another task might be doing the same, we might in fact
3568 // drop below the target. But, this is not a problem.
3569 size_t target_size;
3570 if (partially) {
3571 target_size = _cm->partial_mark_stack_size_target();
3572 } else {
3573 target_size = 0;
3574 }
3575
3576 if (_cm->mark_stack_size() > target_size) {
3577 if (_cm->verbose_low()) {
3578 gclog_or_tty->print_cr("[%u] draining global_stack, target size " SIZE_FORMAT,
3579 _worker_id, target_size);
3580 }
3581
3582 while (!has_aborted() && _cm->mark_stack_size() > target_size) {
3583 get_entries_from_global_stack();
3584 drain_local_queue(partially);
3585 }
3586
3587 if (_cm->verbose_low()) {
3588 gclog_or_tty->print_cr("[%u] drained global stack, size = " SIZE_FORMAT,
3589 _worker_id, _cm->mark_stack_size());
3590 }
3591 }
3592 }
3593
3594 // SATB Queue has several assumptions on whether to call the par or
3595 // non-par versions of the methods. this is why some of the code is
3596 // replicated. We should really get rid of the single-threaded version
3597 // of the code to simplify things.
3598 void CMTask::drain_satb_buffers() {
3599 if (has_aborted()) return;
3600
3601 // We set this so that the regular clock knows that we're in the
3602 // middle of draining buffers and doesn't set the abort flag when it
3603 // notices that SATB buffers are available for draining. It'd be
3604 // very counter productive if it did that. :-)
3605 _draining_satb_buffers = true;
3606
3607 CMSATBBufferClosure satb_cl(this, _g1h);
3608 SATBMarkQueueSet& satb_mq_set = JavaThread::satb_mark_queue_set();
3609
3610 // This keeps claiming and applying the closure to completed buffers
3611 // until we run out of buffers or we need to abort.
3612 while (!has_aborted() &&
3613 satb_mq_set.apply_closure_to_completed_buffer(&satb_cl)) {
3614 if (_cm->verbose_medium()) {
3615 gclog_or_tty->print_cr("[%u] processed an SATB buffer", _worker_id);
3616 }
3617 statsOnly( ++_satb_buffers_processed );
3618 regular_clock_call();
3619 }
3620
3621 _draining_satb_buffers = false;
3622
3623 assert(has_aborted() ||
3624 concurrent() ||
3625 satb_mq_set.completed_buffers_num() == 0, "invariant");
3626
3627 // again, this was a potentially expensive operation, decrease the
3628 // limits to get the regular clock call early
3629 decrease_limits();
3630 }
3631
3632 void CMTask::print_stats() {
3633 gclog_or_tty->print_cr("Marking Stats, task = %u, calls = %d",
3634 _worker_id, _calls);
3635 gclog_or_tty->print_cr(" Elapsed time = %1.2lfms, Termination time = %1.2lfms",
3636 _elapsed_time_ms, _termination_time_ms);
3637 gclog_or_tty->print_cr(" Step Times (cum): num = %d, avg = %1.2lfms, sd = %1.2lfms",
3638 _step_times_ms.num(), _step_times_ms.avg(),
3639 _step_times_ms.sd());
3640 gclog_or_tty->print_cr(" max = %1.2lfms, total = %1.2lfms",
3641 _step_times_ms.maximum(), _step_times_ms.sum());
3642
3643 #if _MARKING_STATS_
3644 gclog_or_tty->print_cr(" Clock Intervals (cum): num = %d, avg = %1.2lfms, sd = %1.2lfms",
3645 _all_clock_intervals_ms.num(), _all_clock_intervals_ms.avg(),
3646 _all_clock_intervals_ms.sd());
3647 gclog_or_tty->print_cr(" max = %1.2lfms, total = %1.2lfms",
3648 _all_clock_intervals_ms.maximum(),
3649 _all_clock_intervals_ms.sum());
3650 gclog_or_tty->print_cr(" Clock Causes (cum): scanning = " SIZE_FORMAT ", marking = " SIZE_FORMAT,
3651 _clock_due_to_scanning, _clock_due_to_marking);
3652 gclog_or_tty->print_cr(" Objects: scanned = " SIZE_FORMAT ", found on the bitmap = " SIZE_FORMAT,
3653 _objs_scanned, _objs_found_on_bitmap);
3654 gclog_or_tty->print_cr(" Local Queue: pushes = " SIZE_FORMAT ", pops = " SIZE_FORMAT ", max size = " SIZE_FORMAT,
3655 _local_pushes, _local_pops, _local_max_size);
3656 gclog_or_tty->print_cr(" Global Stack: pushes = " SIZE_FORMAT ", pops = " SIZE_FORMAT ", max size = " SIZE_FORMAT,
3657 _global_pushes, _global_pops, _global_max_size);
3658 gclog_or_tty->print_cr(" transfers to = " SIZE_FORMAT ", transfers from = " SIZE_FORMAT,
3659 _global_transfers_to,_global_transfers_from);
3660 gclog_or_tty->print_cr(" Regions: claimed = " SIZE_FORMAT, _regions_claimed);
3661 gclog_or_tty->print_cr(" SATB buffers: processed = " SIZE_FORMAT, _satb_buffers_processed);
3662 gclog_or_tty->print_cr(" Steals: attempts = " SIZE_FORMAT ", successes = " SIZE_FORMAT,
3663 _steal_attempts, _steals);
3664 gclog_or_tty->print_cr(" Aborted: " SIZE_FORMAT ", due to", _aborted);
3665 gclog_or_tty->print_cr(" overflow: " SIZE_FORMAT ", global abort: " SIZE_FORMAT ", yield: " SIZE_FORMAT,
3666 _aborted_overflow, _aborted_cm_aborted, _aborted_yield);
3667 gclog_or_tty->print_cr(" time out: " SIZE_FORMAT ", SATB: " SIZE_FORMAT ", termination: " SIZE_FORMAT,
3668 _aborted_timed_out, _aborted_satb, _aborted_termination);
3669 #endif // _MARKING_STATS_
3670 }
3671
3672 bool ConcurrentMark::try_stealing(uint worker_id, int* hash_seed, oop& obj) {
3673 return _task_queues->steal(worker_id, hash_seed, obj);
3674 }
3675
3676 /*****************************************************************************
3677
3678 The do_marking_step(time_target_ms, ...) method is the building
3679 block of the parallel marking framework. It can be called in parallel
3680 with other invocations of do_marking_step() on different tasks
3681 (but only one per task, obviously) and concurrently with the
3682 mutator threads, or during remark, hence it eliminates the need
3683 for two versions of the code. When called during remark, it will
3684 pick up from where the task left off during the concurrent marking
3685 phase. Interestingly, tasks are also claimable during evacuation
3686 pauses too, since do_marking_step() ensures that it aborts before
3687 it needs to yield.
3688
3689 The data structures that it uses to do marking work are the
3690 following:
3691
3692 (1) Marking Bitmap. If there are gray objects that appear only
3693 on the bitmap (this happens either when dealing with an overflow
3694 or when the initial marking phase has simply marked the roots
3695 and didn't push them on the stack), then tasks claim heap
3696 regions whose bitmap they then scan to find gray objects. A
3697 global finger indicates where the end of the last claimed region
3698 is. A local finger indicates how far into the region a task has
3699 scanned. The two fingers are used to determine how to gray an
3700 object (i.e. whether simply marking it is OK, as it will be
3701 visited by a task in the future, or whether it needs to be also
3702 pushed on a stack).
3703
3704 (2) Local Queue. The local queue of the task which is accessed
3705 reasonably efficiently by the task. Other tasks can steal from
3706 it when they run out of work. Throughout the marking phase, a
3707 task attempts to keep its local queue short but not totally
3708 empty, so that entries are available for stealing by other
3709 tasks. Only when there is no more work, a task will totally
3710 drain its local queue.
3711
3712 (3) Global Mark Stack. This handles local queue overflow. During
3713 marking only sets of entries are moved between it and the local
3714 queues, as access to it requires a mutex and more fine-grain
3715 interaction with it which might cause contention. If it
3716 overflows, then the marking phase should restart and iterate
3717 over the bitmap to identify gray objects. Throughout the marking
3718 phase, tasks attempt to keep the global mark stack at a small
3719 length but not totally empty, so that entries are available for
3720 popping by other tasks. Only when there is no more work, tasks
3721 will totally drain the global mark stack.
3722
3723 (4) SATB Buffer Queue. This is where completed SATB buffers are
3724 made available. Buffers are regularly removed from this queue
3725 and scanned for roots, so that the queue doesn't get too
3726 long. During remark, all completed buffers are processed, as
3727 well as the filled in parts of any uncompleted buffers.
3728
3729 The do_marking_step() method tries to abort when the time target
3730 has been reached. There are a few other cases when the
3731 do_marking_step() method also aborts:
3732
3733 (1) When the marking phase has been aborted (after a Full GC).
3734
3735 (2) When a global overflow (on the global stack) has been
3736 triggered. Before the task aborts, it will actually sync up with
3737 the other tasks to ensure that all the marking data structures
3738 (local queues, stacks, fingers etc.) are re-initialized so that
3739 when do_marking_step() completes, the marking phase can
3740 immediately restart.
3741
3742 (3) When enough completed SATB buffers are available. The
3743 do_marking_step() method only tries to drain SATB buffers right
3744 at the beginning. So, if enough buffers are available, the
3745 marking step aborts and the SATB buffers are processed at
3746 the beginning of the next invocation.
3747
3748 (4) To yield. when we have to yield then we abort and yield
3749 right at the end of do_marking_step(). This saves us from a lot
3750 of hassle as, by yielding we might allow a Full GC. If this
3751 happens then objects will be compacted underneath our feet, the
3752 heap might shrink, etc. We save checking for this by just
3753 aborting and doing the yield right at the end.
3754
3755 From the above it follows that the do_marking_step() method should
3756 be called in a loop (or, otherwise, regularly) until it completes.
3757
3758 If a marking step completes without its has_aborted() flag being
3759 true, it means it has completed the current marking phase (and
3760 also all other marking tasks have done so and have all synced up).
3761
3762 A method called regular_clock_call() is invoked "regularly" (in
3763 sub ms intervals) throughout marking. It is this clock method that
3764 checks all the abort conditions which were mentioned above and
3765 decides when the task should abort. A work-based scheme is used to
3766 trigger this clock method: when the number of object words the
3767 marking phase has scanned or the number of references the marking
3768 phase has visited reach a given limit. Additional invocations to
3769 the method clock have been planted in a few other strategic places
3770 too. The initial reason for the clock method was to avoid calling
3771 vtime too regularly, as it is quite expensive. So, once it was in
3772 place, it was natural to piggy-back all the other conditions on it
3773 too and not constantly check them throughout the code.
3774
3775 If do_termination is true then do_marking_step will enter its
3776 termination protocol.
3777
3778 The value of is_serial must be true when do_marking_step is being
3779 called serially (i.e. by the VMThread) and do_marking_step should
3780 skip any synchronization in the termination and overflow code.
3781 Examples include the serial remark code and the serial reference
3782 processing closures.
3783
3784 The value of is_serial must be false when do_marking_step is
3785 being called by any of the worker threads in a work gang.
3786 Examples include the concurrent marking code (CMMarkingTask),
3787 the MT remark code, and the MT reference processing closures.
3788
3789 *****************************************************************************/
3790
3791 void CMTask::do_marking_step(double time_target_ms,
3792 bool do_termination,
3793 bool is_serial) {
3794 assert(time_target_ms >= 1.0, "minimum granularity is 1ms");
3795 assert(concurrent() == _cm->concurrent(), "they should be the same");
3796
3797 G1CollectorPolicy* g1_policy = _g1h->g1_policy();
3798 assert(_task_queues != NULL, "invariant");
3799 assert(_task_queue != NULL, "invariant");
3800 assert(_task_queues->queue(_worker_id) == _task_queue, "invariant");
3801
3802 assert(!_claimed,
3803 "only one thread should claim this task at any one time");
3804
3805 // OK, this doesn't safeguard again all possible scenarios, as it is
3806 // possible for two threads to set the _claimed flag at the same
3807 // time. But it is only for debugging purposes anyway and it will
3808 // catch most problems.
3809 _claimed = true;
3810
3811 _start_time_ms = os::elapsedVTime() * 1000.0;
3812 statsOnly( _interval_start_time_ms = _start_time_ms );
3813
3814 // If do_stealing is true then do_marking_step will attempt to
3815 // steal work from the other CMTasks. It only makes sense to
3816 // enable stealing when the termination protocol is enabled
3817 // and do_marking_step() is not being called serially.
3818 bool do_stealing = do_termination && !is_serial;
3819
3820 double diff_prediction_ms =
3821 g1_policy->get_new_prediction(&_marking_step_diffs_ms);
3822 _time_target_ms = time_target_ms - diff_prediction_ms;
3823
3824 // set up the variables that are used in the work-based scheme to
3825 // call the regular clock method
3826 _words_scanned = 0;
3827 _refs_reached = 0;
3828 recalculate_limits();
3829
3830 // clear all flags
3831 clear_has_aborted();
3832 _has_timed_out = false;
3833 _draining_satb_buffers = false;
3834
3835 ++_calls;
3836
3837 if (_cm->verbose_low()) {
3838 gclog_or_tty->print_cr("[%u] >>>>>>>>>> START, call = %d, "
3839 "target = %1.2lfms >>>>>>>>>>",
3840 _worker_id, _calls, _time_target_ms);
3841 }
3842
3843 // Set up the bitmap and oop closures. Anything that uses them is
3844 // eventually called from this method, so it is OK to allocate these
3845 // statically.
3846 CMBitMapClosure bitmap_closure(this, _cm, _nextMarkBitMap);
3847 G1CMOopClosure cm_oop_closure(_g1h, _cm, this);
3848 set_cm_oop_closure(&cm_oop_closure);
3849
3850 if (_cm->has_overflown()) {
3851 // This can happen if the mark stack overflows during a GC pause
3852 // and this task, after a yield point, restarts. We have to abort
3853 // as we need to get into the overflow protocol which happens
3854 // right at the end of this task.
3855 set_has_aborted();
3856 }
3857
3858 // First drain any available SATB buffers. After this, we will not
3859 // look at SATB buffers before the next invocation of this method.
3860 // If enough completed SATB buffers are queued up, the regular clock
3861 // will abort this task so that it restarts.
3862 drain_satb_buffers();
3863 // ...then partially drain the local queue and the global stack
3864 drain_local_queue(true);
3865 drain_global_stack(true);
3866
3867 do {
3868 if (!has_aborted() && _curr_region != NULL) {
3869 // This means that we're already holding on to a region.
3870 assert(_finger != NULL, "if region is not NULL, then the finger "
3871 "should not be NULL either");
3872
3873 // We might have restarted this task after an evacuation pause
3874 // which might have evacuated the region we're holding on to
3875 // underneath our feet. Let's read its limit again to make sure
3876 // that we do not iterate over a region of the heap that
3877 // contains garbage (update_region_limit() will also move
3878 // _finger to the start of the region if it is found empty).
3879 update_region_limit();
3880 // We will start from _finger not from the start of the region,
3881 // as we might be restarting this task after aborting half-way
3882 // through scanning this region. In this case, _finger points to
3883 // the address where we last found a marked object. If this is a
3884 // fresh region, _finger points to start().
3885 MemRegion mr = MemRegion(_finger, _region_limit);
3886
3887 if (_cm->verbose_low()) {
3888 gclog_or_tty->print_cr("[%u] we're scanning part "
3889 "["PTR_FORMAT", "PTR_FORMAT") "
3890 "of region "HR_FORMAT,
3891 _worker_id, p2i(_finger), p2i(_region_limit),
3892 HR_FORMAT_PARAMS(_curr_region));
3893 }
3894
3895 assert(!_curr_region->is_humongous() || mr.start() == _curr_region->bottom(),
3896 "humongous regions should go around loop once only");
3897
3898 // Some special cases:
3899 // If the memory region is empty, we can just give up the region.
3900 // If the current region is humongous then we only need to check
3901 // the bitmap for the bit associated with the start of the object,
3902 // scan the object if it's live, and give up the region.
3903 // Otherwise, let's iterate over the bitmap of the part of the region
3904 // that is left.
3905 // If the iteration is successful, give up the region.
3906 if (mr.is_empty()) {
3907 giveup_current_region();
3908 regular_clock_call();
3909 } else if (_curr_region->is_humongous() && mr.start() == _curr_region->bottom()) {
3910 if (_nextMarkBitMap->isMarked(mr.start())) {
3911 // The object is marked - apply the closure
3912 BitMap::idx_t offset = _nextMarkBitMap->heapWordToOffset(mr.start());
3913 bitmap_closure.do_bit(offset);
3914 }
3915 // Even if this task aborted while scanning the humongous object
3916 // we can (and should) give up the current region.
3917 giveup_current_region();
3918 regular_clock_call();
3919 } else if (_nextMarkBitMap->iterate(&bitmap_closure, mr)) {
3920 giveup_current_region();
3921 regular_clock_call();
3922 } else {
3923 assert(has_aborted(), "currently the only way to do so");
3924 // The only way to abort the bitmap iteration is to return
3925 // false from the do_bit() method. However, inside the
3926 // do_bit() method we move the _finger to point to the
3927 // object currently being looked at. So, if we bail out, we
3928 // have definitely set _finger to something non-null.
3929 assert(_finger != NULL, "invariant");
3930
3931 // Region iteration was actually aborted. So now _finger
3932 // points to the address of the object we last scanned. If we
3933 // leave it there, when we restart this task, we will rescan
3934 // the object. It is easy to avoid this. We move the finger by
3935 // enough to point to the next possible object header (the
3936 // bitmap knows by how much we need to move it as it knows its
3937 // granularity).
3938 assert(_finger < _region_limit, "invariant");
3939 HeapWord* new_finger = _nextMarkBitMap->nextObject(_finger);
3940 // Check if bitmap iteration was aborted while scanning the last object
3941 if (new_finger >= _region_limit) {
3942 giveup_current_region();
3943 } else {
3944 move_finger_to(new_finger);
3945 }
3946 }
3947 }
3948 // At this point we have either completed iterating over the
3949 // region we were holding on to, or we have aborted.
3950
3951 // We then partially drain the local queue and the global stack.
3952 // (Do we really need this?)
3953 drain_local_queue(true);
3954 drain_global_stack(true);
3955
3956 // Read the note on the claim_region() method on why it might
3957 // return NULL with potentially more regions available for
3958 // claiming and why we have to check out_of_regions() to determine
3959 // whether we're done or not.
3960 while (!has_aborted() && _curr_region == NULL && !_cm->out_of_regions()) {
3961 // We are going to try to claim a new region. We should have
3962 // given up on the previous one.
3963 // Separated the asserts so that we know which one fires.
3964 assert(_curr_region == NULL, "invariant");
3965 assert(_finger == NULL, "invariant");
3966 assert(_region_limit == NULL, "invariant");
3967 if (_cm->verbose_low()) {
3968 gclog_or_tty->print_cr("[%u] trying to claim a new region", _worker_id);
3969 }
3970 HeapRegion* claimed_region = _cm->claim_region(_worker_id);
3971 if (claimed_region != NULL) {
3972 // Yes, we managed to claim one
3973 statsOnly( ++_regions_claimed );
3974
3975 if (_cm->verbose_low()) {
3976 gclog_or_tty->print_cr("[%u] we successfully claimed "
3977 "region "PTR_FORMAT,
3978 _worker_id, p2i(claimed_region));
3979 }
3980
3981 setup_for_region(claimed_region);
3982 assert(_curr_region == claimed_region, "invariant");
3983 }
3984 // It is important to call the regular clock here. It might take
3985 // a while to claim a region if, for example, we hit a large
3986 // block of empty regions. So we need to call the regular clock
3987 // method once round the loop to make sure it's called
3988 // frequently enough.
3989 regular_clock_call();
3990 }
3991
3992 if (!has_aborted() && _curr_region == NULL) {
3993 assert(_cm->out_of_regions(),
3994 "at this point we should be out of regions");
3995 }
3996 } while ( _curr_region != NULL && !has_aborted());
3997
3998 if (!has_aborted()) {
3999 // We cannot check whether the global stack is empty, since other
4000 // tasks might be pushing objects to it concurrently.
4001 assert(_cm->out_of_regions(),
4002 "at this point we should be out of regions");
4003
4004 if (_cm->verbose_low()) {
4005 gclog_or_tty->print_cr("[%u] all regions claimed", _worker_id);
4006 }
4007
4008 // Try to reduce the number of available SATB buffers so that
4009 // remark has less work to do.
4010 drain_satb_buffers();
4011 }
4012
4013 // Since we've done everything else, we can now totally drain the
4014 // local queue and global stack.
4015 drain_local_queue(false);
4016 drain_global_stack(false);
4017
4018 // Attempt at work stealing from other task's queues.
4019 if (do_stealing && !has_aborted()) {
4020 // We have not aborted. This means that we have finished all that
4021 // we could. Let's try to do some stealing...
4022
4023 // We cannot check whether the global stack is empty, since other
4024 // tasks might be pushing objects to it concurrently.
4025 assert(_cm->out_of_regions() && _task_queue->size() == 0,
4026 "only way to reach here");
4027
4028 if (_cm->verbose_low()) {
4029 gclog_or_tty->print_cr("[%u] starting to steal", _worker_id);
4030 }
4031
4032 while (!has_aborted()) {
4033 oop obj;
4034 statsOnly( ++_steal_attempts );
4035
4036 if (_cm->try_stealing(_worker_id, &_hash_seed, obj)) {
4037 if (_cm->verbose_medium()) {
4038 gclog_or_tty->print_cr("[%u] stolen "PTR_FORMAT" successfully",
4039 _worker_id, p2i((void*) obj));
4040 }
4041
4042 statsOnly( ++_steals );
4043
4044 assert(_nextMarkBitMap->isMarked((HeapWord*) obj),
4045 "any stolen object should be marked");
4046 scan_object(obj);
4047
4048 // And since we're towards the end, let's totally drain the
4049 // local queue and global stack.
4050 drain_local_queue(false);
4051 drain_global_stack(false);
4052 } else {
4053 break;
4054 }
4055 }
4056 }
4057
4058 // If we are about to wrap up and go into termination, check if we
4059 // should raise the overflow flag.
4060 if (do_termination && !has_aborted()) {
4061 if (_cm->force_overflow()->should_force()) {
4062 _cm->set_has_overflown();
4063 regular_clock_call();
4064 }
4065 }
4066
4067 // We still haven't aborted. Now, let's try to get into the
4068 // termination protocol.
4069 if (do_termination && !has_aborted()) {
4070 // We cannot check whether the global stack is empty, since other
4071 // tasks might be concurrently pushing objects on it.
4072 // Separated the asserts so that we know which one fires.
4073 assert(_cm->out_of_regions(), "only way to reach here");
4074 assert(_task_queue->size() == 0, "only way to reach here");
4075
4076 if (_cm->verbose_low()) {
4077 gclog_or_tty->print_cr("[%u] starting termination protocol", _worker_id);
4078 }
4079
4080 _termination_start_time_ms = os::elapsedVTime() * 1000.0;
4081
4082 // The CMTask class also extends the TerminatorTerminator class,
4083 // hence its should_exit_termination() method will also decide
4084 // whether to exit the termination protocol or not.
4085 bool finished = (is_serial ||
4086 _cm->terminator()->offer_termination(this));
4087 double termination_end_time_ms = os::elapsedVTime() * 1000.0;
4088 _termination_time_ms +=
4089 termination_end_time_ms - _termination_start_time_ms;
4090
4091 if (finished) {
4092 // We're all done.
4093
4094 if (_worker_id == 0) {
4095 // let's allow task 0 to do this
4096 if (concurrent()) {
4097 assert(_cm->concurrent_marking_in_progress(), "invariant");
4098 // we need to set this to false before the next
4099 // safepoint. This way we ensure that the marking phase
4100 // doesn't observe any more heap expansions.
4101 _cm->clear_concurrent_marking_in_progress();
4102 }
4103 }
4104
4105 // We can now guarantee that the global stack is empty, since
4106 // all other tasks have finished. We separated the guarantees so
4107 // that, if a condition is false, we can immediately find out
4108 // which one.
4109 guarantee(_cm->out_of_regions(), "only way to reach here");
4110 guarantee(_cm->mark_stack_empty(), "only way to reach here");
4111 guarantee(_task_queue->size() == 0, "only way to reach here");
4112 guarantee(!_cm->has_overflown(), "only way to reach here");
4113 guarantee(!_cm->mark_stack_overflow(), "only way to reach here");
4114
4115 if (_cm->verbose_low()) {
4116 gclog_or_tty->print_cr("[%u] all tasks terminated", _worker_id);
4117 }
4118 } else {
4119 // Apparently there's more work to do. Let's abort this task. It
4120 // will restart it and we can hopefully find more things to do.
4121
4122 if (_cm->verbose_low()) {
4123 gclog_or_tty->print_cr("[%u] apparently there is more work to do",
4124 _worker_id);
4125 }
4126
4127 set_has_aborted();
4128 statsOnly( ++_aborted_termination );
4129 }
4130 }
4131
4132 // Mainly for debugging purposes to make sure that a pointer to the
4133 // closure which was statically allocated in this frame doesn't
4134 // escape it by accident.
4135 set_cm_oop_closure(NULL);
4136 double end_time_ms = os::elapsedVTime() * 1000.0;
4137 double elapsed_time_ms = end_time_ms - _start_time_ms;
4138 // Update the step history.
4139 _step_times_ms.add(elapsed_time_ms);
4140
4141 if (has_aborted()) {
4142 // The task was aborted for some reason.
4143
4144 statsOnly( ++_aborted );
4145
4146 if (_has_timed_out) {
4147 double diff_ms = elapsed_time_ms - _time_target_ms;
4148 // Keep statistics of how well we did with respect to hitting
4149 // our target only if we actually timed out (if we aborted for
4150 // other reasons, then the results might get skewed).
4151 _marking_step_diffs_ms.add(diff_ms);
4152 }
4153
4154 if (_cm->has_overflown()) {
4155 // This is the interesting one. We aborted because a global
4156 // overflow was raised. This means we have to restart the
4157 // marking phase and start iterating over regions. However, in
4158 // order to do this we have to make sure that all tasks stop
4159 // what they are doing and re-initialize in a safe manner. We
4160 // will achieve this with the use of two barrier sync points.
4161
4162 if (_cm->verbose_low()) {
4163 gclog_or_tty->print_cr("[%u] detected overflow", _worker_id);
4164 }
4165
4166 if (!is_serial) {
4167 // We only need to enter the sync barrier if being called
4168 // from a parallel context
4169 _cm->enter_first_sync_barrier(_worker_id);
4170
4171 // When we exit this sync barrier we know that all tasks have
4172 // stopped doing marking work. So, it's now safe to
4173 // re-initialize our data structures. At the end of this method,
4174 // task 0 will clear the global data structures.
4175 }
4176
4177 statsOnly( ++_aborted_overflow );
4178
4179 // We clear the local state of this task...
4180 clear_region_fields();
4181
4182 if (!is_serial) {
4183 // ...and enter the second barrier.
4184 _cm->enter_second_sync_barrier(_worker_id);
4185 }
4186 // At this point, if we're during the concurrent phase of
4187 // marking, everything has been re-initialized and we're
4188 // ready to restart.
4189 }
4190
4191 if (_cm->verbose_low()) {
4192 gclog_or_tty->print_cr("[%u] <<<<<<<<<< ABORTING, target = %1.2lfms, "
4193 "elapsed = %1.2lfms <<<<<<<<<<",
4194 _worker_id, _time_target_ms, elapsed_time_ms);
4195 if (_cm->has_aborted()) {
4196 gclog_or_tty->print_cr("[%u] ========== MARKING ABORTED ==========",
4197 _worker_id);
4198 }
4199 }
4200 } else {
4201 if (_cm->verbose_low()) {
4202 gclog_or_tty->print_cr("[%u] <<<<<<<<<< FINISHED, target = %1.2lfms, "
4203 "elapsed = %1.2lfms <<<<<<<<<<",
4204 _worker_id, _time_target_ms, elapsed_time_ms);
4205 }
4206 }
4207
4208 _claimed = false;
4209 }
4210
4211 CMTask::CMTask(uint worker_id,
4212 ConcurrentMark* cm,
4213 size_t* marked_bytes,
4214 BitMap* card_bm,
4215 CMTaskQueue* task_queue,
4216 CMTaskQueueSet* task_queues)
4217 : _g1h(G1CollectedHeap::heap()),
4218 _worker_id(worker_id), _cm(cm),
4219 _claimed(false),
4220 _nextMarkBitMap(NULL), _hash_seed(17),
4221 _task_queue(task_queue),
4222 _task_queues(task_queues),
4223 _cm_oop_closure(NULL),
4224 _marked_bytes_array(marked_bytes),
4225 _card_bm(card_bm) {
4226 guarantee(task_queue != NULL, "invariant");
4227 guarantee(task_queues != NULL, "invariant");
4228
4229 statsOnly( _clock_due_to_scanning = 0;
4230 _clock_due_to_marking = 0 );
4231
4232 _marking_step_diffs_ms.add(0.5);
4233 }
4234
4235 // These are formatting macros that are used below to ensure
4236 // consistent formatting. The *_H_* versions are used to format the
4237 // header for a particular value and they should be kept consistent
4238 // with the corresponding macro. Also note that most of the macros add
4239 // the necessary white space (as a prefix) which makes them a bit
4240 // easier to compose.
4241
4242 // All the output lines are prefixed with this string to be able to
4243 // identify them easily in a large log file.
4244 #define G1PPRL_LINE_PREFIX "###"
4245
4246 #define G1PPRL_ADDR_BASE_FORMAT " "PTR_FORMAT"-"PTR_FORMAT
4247 #ifdef _LP64
4248 #define G1PPRL_ADDR_BASE_H_FORMAT " %37s"
4249 #else // _LP64
4250 #define G1PPRL_ADDR_BASE_H_FORMAT " %21s"
4251 #endif // _LP64
4252
4253 // For per-region info
4254 #define G1PPRL_TYPE_FORMAT " %-4s"
4255 #define G1PPRL_TYPE_H_FORMAT " %4s"
4256 #define G1PPRL_BYTE_FORMAT " "SIZE_FORMAT_W(9)
4257 #define G1PPRL_BYTE_H_FORMAT " %9s"
4258 #define G1PPRL_DOUBLE_FORMAT " %14.1f"
4259 #define G1PPRL_DOUBLE_H_FORMAT " %14s"
4260
4261 // For summary info
4262 #define G1PPRL_SUM_ADDR_FORMAT(tag) " "tag":"G1PPRL_ADDR_BASE_FORMAT
4263 #define G1PPRL_SUM_BYTE_FORMAT(tag) " "tag": "SIZE_FORMAT
4264 #define G1PPRL_SUM_MB_FORMAT(tag) " "tag": %1.2f MB"
4265 #define G1PPRL_SUM_MB_PERC_FORMAT(tag) G1PPRL_SUM_MB_FORMAT(tag)" / %1.2f %%"
4266
4267 G1PrintRegionLivenessInfoClosure::
4268 G1PrintRegionLivenessInfoClosure(outputStream* out, const char* phase_name)
4269 : _out(out),
4270 _total_used_bytes(0), _total_capacity_bytes(0),
4271 _total_prev_live_bytes(0), _total_next_live_bytes(0),
4272 _hum_used_bytes(0), _hum_capacity_bytes(0),
4273 _hum_prev_live_bytes(0), _hum_next_live_bytes(0),
4274 _total_remset_bytes(0), _total_strong_code_roots_bytes(0) {
4275 G1CollectedHeap* g1h = G1CollectedHeap::heap();
4276 MemRegion g1_reserved = g1h->g1_reserved();
4277 double now = os::elapsedTime();
4278
4279 // Print the header of the output.
4280 _out->cr();
4281 _out->print_cr(G1PPRL_LINE_PREFIX" PHASE %s @ %1.3f", phase_name, now);
4282 _out->print_cr(G1PPRL_LINE_PREFIX" HEAP"
4283 G1PPRL_SUM_ADDR_FORMAT("reserved")
4284 G1PPRL_SUM_BYTE_FORMAT("region-size"),
4285 p2i(g1_reserved.start()), p2i(g1_reserved.end()),
4286 HeapRegion::GrainBytes);
4287 _out->print_cr(G1PPRL_LINE_PREFIX);
4288 _out->print_cr(G1PPRL_LINE_PREFIX
4289 G1PPRL_TYPE_H_FORMAT
4290 G1PPRL_ADDR_BASE_H_FORMAT
4291 G1PPRL_BYTE_H_FORMAT
4292 G1PPRL_BYTE_H_FORMAT
4293 G1PPRL_BYTE_H_FORMAT
4294 G1PPRL_DOUBLE_H_FORMAT
4295 G1PPRL_BYTE_H_FORMAT
4296 G1PPRL_BYTE_H_FORMAT,
4297 "type", "address-range",
4298 "used", "prev-live", "next-live", "gc-eff",
4299 "remset", "code-roots");
4300 _out->print_cr(G1PPRL_LINE_PREFIX
4301 G1PPRL_TYPE_H_FORMAT
4302 G1PPRL_ADDR_BASE_H_FORMAT
4303 G1PPRL_BYTE_H_FORMAT
4304 G1PPRL_BYTE_H_FORMAT
4305 G1PPRL_BYTE_H_FORMAT
4306 G1PPRL_DOUBLE_H_FORMAT
4307 G1PPRL_BYTE_H_FORMAT
4308 G1PPRL_BYTE_H_FORMAT,
4309 "", "",
4310 "(bytes)", "(bytes)", "(bytes)", "(bytes/ms)",
4311 "(bytes)", "(bytes)");
4312 }
4313
4314 // It takes as a parameter a reference to one of the _hum_* fields, it
4315 // deduces the corresponding value for a region in a humongous region
4316 // series (either the region size, or what's left if the _hum_* field
4317 // is < the region size), and updates the _hum_* field accordingly.
4318 size_t G1PrintRegionLivenessInfoClosure::get_hum_bytes(size_t* hum_bytes) {
4319 size_t bytes = 0;
4320 // The > 0 check is to deal with the prev and next live bytes which
4321 // could be 0.
4322 if (*hum_bytes > 0) {
4323 bytes = MIN2(HeapRegion::GrainBytes, *hum_bytes);
4324 *hum_bytes -= bytes;
4325 }
4326 return bytes;
4327 }
4328
4329 // It deduces the values for a region in a humongous region series
4330 // from the _hum_* fields and updates those accordingly. It assumes
4331 // that that _hum_* fields have already been set up from the "starts
4332 // humongous" region and we visit the regions in address order.
4333 void G1PrintRegionLivenessInfoClosure::get_hum_bytes(size_t* used_bytes,
4334 size_t* capacity_bytes,
4335 size_t* prev_live_bytes,
4336 size_t* next_live_bytes) {
4337 assert(_hum_used_bytes > 0 && _hum_capacity_bytes > 0, "pre-condition");
4338 *used_bytes = get_hum_bytes(&_hum_used_bytes);
4339 *capacity_bytes = get_hum_bytes(&_hum_capacity_bytes);
4340 *prev_live_bytes = get_hum_bytes(&_hum_prev_live_bytes);
4341 *next_live_bytes = get_hum_bytes(&_hum_next_live_bytes);
4342 }
4343
4344 bool G1PrintRegionLivenessInfoClosure::doHeapRegion(HeapRegion* r) {
4345 const char* type = r->get_type_str();
4346 HeapWord* bottom = r->bottom();
4347 HeapWord* end = r->end();
4348 size_t capacity_bytes = r->capacity();
4349 size_t used_bytes = r->used();
4350 size_t prev_live_bytes = r->live_bytes();
4351 size_t next_live_bytes = r->next_live_bytes();
4352 double gc_eff = r->gc_efficiency();
4353 size_t remset_bytes = r->rem_set()->mem_size();
4354 size_t strong_code_roots_bytes = r->rem_set()->strong_code_roots_mem_size();
4355
4356 if (r->is_starts_humongous()) {
4357 assert(_hum_used_bytes == 0 && _hum_capacity_bytes == 0 &&
4358 _hum_prev_live_bytes == 0 && _hum_next_live_bytes == 0,
4359 "they should have been zeroed after the last time we used them");
4360 // Set up the _hum_* fields.
4361 _hum_capacity_bytes = capacity_bytes;
4362 _hum_used_bytes = used_bytes;
4363 _hum_prev_live_bytes = prev_live_bytes;
4364 _hum_next_live_bytes = next_live_bytes;
4365 get_hum_bytes(&used_bytes, &capacity_bytes,
4366 &prev_live_bytes, &next_live_bytes);
4367 end = bottom + HeapRegion::GrainWords;
4368 } else if (r->is_continues_humongous()) {
4369 get_hum_bytes(&used_bytes, &capacity_bytes,
4370 &prev_live_bytes, &next_live_bytes);
4371 assert(end == bottom + HeapRegion::GrainWords, "invariant");
4372 }
4373
4374 _total_used_bytes += used_bytes;
4375 _total_capacity_bytes += capacity_bytes;
4376 _total_prev_live_bytes += prev_live_bytes;
4377 _total_next_live_bytes += next_live_bytes;
4378 _total_remset_bytes += remset_bytes;
4379 _total_strong_code_roots_bytes += strong_code_roots_bytes;
4380
4381 // Print a line for this particular region.
4382 _out->print_cr(G1PPRL_LINE_PREFIX
4383 G1PPRL_TYPE_FORMAT
4384 G1PPRL_ADDR_BASE_FORMAT
4385 G1PPRL_BYTE_FORMAT
4386 G1PPRL_BYTE_FORMAT
4387 G1PPRL_BYTE_FORMAT
4388 G1PPRL_DOUBLE_FORMAT
4389 G1PPRL_BYTE_FORMAT
4390 G1PPRL_BYTE_FORMAT,
4391 type, p2i(bottom), p2i(end),
4392 used_bytes, prev_live_bytes, next_live_bytes, gc_eff,
4393 remset_bytes, strong_code_roots_bytes);
4394
4395 return false;
4396 }
4397
4398 G1PrintRegionLivenessInfoClosure::~G1PrintRegionLivenessInfoClosure() {
4399 // add static memory usages to remembered set sizes
4400 _total_remset_bytes += HeapRegionRemSet::fl_mem_size() + HeapRegionRemSet::static_mem_size();
4401 // Print the footer of the output.
4402 _out->print_cr(G1PPRL_LINE_PREFIX);
4403 _out->print_cr(G1PPRL_LINE_PREFIX
4404 " SUMMARY"
4405 G1PPRL_SUM_MB_FORMAT("capacity")
4406 G1PPRL_SUM_MB_PERC_FORMAT("used")
4407 G1PPRL_SUM_MB_PERC_FORMAT("prev-live")
4408 G1PPRL_SUM_MB_PERC_FORMAT("next-live")
4409 G1PPRL_SUM_MB_FORMAT("remset")
4410 G1PPRL_SUM_MB_FORMAT("code-roots"),
4411 bytes_to_mb(_total_capacity_bytes),
4412 bytes_to_mb(_total_used_bytes),
4413 perc(_total_used_bytes, _total_capacity_bytes),
4414 bytes_to_mb(_total_prev_live_bytes),
4415 perc(_total_prev_live_bytes, _total_capacity_bytes),
4416 bytes_to_mb(_total_next_live_bytes),
4417 perc(_total_next_live_bytes, _total_capacity_bytes),
4418 bytes_to_mb(_total_remset_bytes),
4419 bytes_to_mb(_total_strong_code_roots_bytes));
4420 _out->cr();
4421 }