1 /*
2 * Copyright (c) 2001, 2017, 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/concurrentMarkThread.inline.hpp"
30 #include "gc/g1/g1CollectedHeap.inline.hpp"
31 #include "gc/g1/g1CollectorState.hpp"
32 #include "gc/g1/g1ConcurrentMark.inline.hpp"
33 #include "gc/g1/g1HeapVerifier.hpp"
34 #include "gc/g1/g1OopClosures.inline.hpp"
35 #include "gc/g1/g1CardLiveData.inline.hpp"
36 #include "gc/g1/g1Policy.hpp"
37 #include "gc/g1/g1StringDedup.hpp"
38 #include "gc/g1/heapRegion.inline.hpp"
39 #include "gc/g1/heapRegionRemSet.hpp"
40 #include "gc/g1/heapRegionSet.inline.hpp"
41 #include "gc/g1/suspendibleThreadSet.hpp"
42 #include "gc/shared/gcId.hpp"
43 #include "gc/shared/gcTimer.hpp"
44 #include "gc/shared/gcTrace.hpp"
45 #include "gc/shared/gcTraceTime.inline.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 "logging/log.hpp"
52 #include "memory/allocation.hpp"
53 #include "memory/resourceArea.hpp"
54 #include "oops/oop.inline.hpp"
55 #include "runtime/atomic.hpp"
56 #include "runtime/handles.inline.hpp"
57 #include "runtime/java.hpp"
58 #include "runtime/prefetch.inline.hpp"
59 #include "services/memTracker.hpp"
60 #include "utilities/align.hpp"
61 #include "utilities/growableArray.hpp"
62
63 // Concurrent marking bit map wrapper
64
65 G1CMBitMapRO::G1CMBitMapRO(int shifter) :
66 _bm(),
67 _shifter(shifter) {
68 _bmStartWord = 0;
69 _bmWordSize = 0;
70 }
71
72 HeapWord* G1CMBitMapRO::getNextMarkedWordAddress(const HeapWord* addr,
73 const HeapWord* limit) const {
74 // First we must round addr *up* to a possible object boundary.
75 addr = align_up(addr, HeapWordSize << _shifter);
76 size_t addrOffset = heapWordToOffset(addr);
77 assert(limit != NULL, "limit must not be NULL");
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 #ifndef PRODUCT
88 bool G1CMBitMapRO::covers(MemRegion heap_rs) const {
89 // assert(_bm.map() == _virtual_space.low(), "map inconsistency");
90 assert(((size_t)_bm.size() * ((size_t)1 << _shifter)) == _bmWordSize,
91 "size inconsistency");
92 return _bmStartWord == (HeapWord*)(heap_rs.start()) &&
93 _bmWordSize == heap_rs.word_size();
94 }
95 #endif
96
97 void G1CMBitMapRO::print_on_error(outputStream* st, const char* prefix) const {
98 _bm.print_on_error(st, prefix);
99 }
100
101 size_t G1CMBitMap::compute_size(size_t heap_size) {
102 return ReservedSpace::allocation_align_size_up(heap_size / mark_distance());
103 }
104
105 size_t G1CMBitMap::mark_distance() {
106 return MinObjAlignmentInBytes * BitsPerByte;
107 }
108
109 void G1CMBitMap::initialize(MemRegion heap, G1RegionToSpaceMapper* storage) {
110 _bmStartWord = heap.start();
111 _bmWordSize = heap.word_size();
112
113 _bm = BitMapView((BitMap::bm_word_t*) storage->reserved().start(), _bmWordSize >> _shifter);
114
115 storage->set_mapping_changed_listener(&_listener);
116 }
117
118 void G1CMBitMapMappingChangedListener::on_commit(uint start_region, size_t num_regions, bool zero_filled) {
119 if (zero_filled) {
120 return;
121 }
122 // We need to clear the bitmap on commit, removing any existing information.
123 MemRegion mr(G1CollectedHeap::heap()->bottom_addr_for_region(start_region), num_regions * HeapRegion::GrainWords);
124 _bm->clear_range(mr);
125 }
126
127 void G1CMBitMap::clear_range(MemRegion mr) {
128 mr.intersection(MemRegion(_bmStartWord, _bmWordSize));
129 assert(!mr.is_empty(), "unexpected empty region");
130 // convert address range into offset range
131 _bm.at_put_range(heapWordToOffset(mr.start()),
132 heapWordToOffset(mr.end()), false);
133 }
134
135 G1CMMarkStack::G1CMMarkStack() :
136 _max_chunk_capacity(0),
137 _base(NULL),
138 _chunk_capacity(0) {
139 set_empty();
140 }
141
142 bool G1CMMarkStack::resize(size_t new_capacity) {
143 assert(is_empty(), "Only resize when stack is empty.");
144 assert(new_capacity <= _max_chunk_capacity,
145 "Trying to resize stack to " SIZE_FORMAT " chunks when the maximum is " SIZE_FORMAT, new_capacity, _max_chunk_capacity);
146
147 TaskQueueEntryChunk* new_base = MmapArrayAllocator<TaskQueueEntryChunk>::allocate_or_null(new_capacity, mtGC);
148
149 if (new_base == NULL) {
150 log_warning(gc)("Failed to reserve memory for new overflow mark stack with " SIZE_FORMAT " chunks and size " SIZE_FORMAT "B.", new_capacity, new_capacity * sizeof(TaskQueueEntryChunk));
151 return false;
152 }
153 // Release old mapping.
154 if (_base != NULL) {
155 MmapArrayAllocator<TaskQueueEntryChunk>::free(_base, _chunk_capacity);
156 }
157
158 _base = new_base;
159 _chunk_capacity = new_capacity;
160 set_empty();
161
162 return true;
163 }
164
165 size_t G1CMMarkStack::capacity_alignment() {
166 return (size_t)lcm(os::vm_allocation_granularity(), sizeof(TaskQueueEntryChunk)) / sizeof(G1TaskQueueEntry);
167 }
168
169 bool G1CMMarkStack::initialize(size_t initial_capacity, size_t max_capacity) {
170 guarantee(_max_chunk_capacity == 0, "G1CMMarkStack already initialized.");
171
172 size_t const TaskEntryChunkSizeInVoidStar = sizeof(TaskQueueEntryChunk) / sizeof(G1TaskQueueEntry);
173
174 _max_chunk_capacity = align_up(max_capacity, capacity_alignment()) / TaskEntryChunkSizeInVoidStar;
175 size_t initial_chunk_capacity = align_up(initial_capacity, capacity_alignment()) / TaskEntryChunkSizeInVoidStar;
176
177 guarantee(initial_chunk_capacity <= _max_chunk_capacity,
178 "Maximum chunk capacity " SIZE_FORMAT " smaller than initial capacity " SIZE_FORMAT,
179 _max_chunk_capacity,
180 initial_chunk_capacity);
181
182 log_debug(gc)("Initialize mark stack with " SIZE_FORMAT " chunks, maximum " SIZE_FORMAT,
183 initial_chunk_capacity, _max_chunk_capacity);
184
185 return resize(initial_chunk_capacity);
186 }
187
188 void G1CMMarkStack::expand() {
189 if (_chunk_capacity == _max_chunk_capacity) {
190 log_debug(gc)("Can not expand overflow mark stack further, already at maximum capacity of " SIZE_FORMAT " chunks.", _chunk_capacity);
191 return;
192 }
193 size_t old_capacity = _chunk_capacity;
194 // Double capacity if possible
195 size_t new_capacity = MIN2(old_capacity * 2, _max_chunk_capacity);
196
197 if (resize(new_capacity)) {
198 log_debug(gc)("Expanded mark stack capacity from " SIZE_FORMAT " to " SIZE_FORMAT " chunks",
199 old_capacity, new_capacity);
200 } else {
201 log_warning(gc)("Failed to expand mark stack capacity from " SIZE_FORMAT " to " SIZE_FORMAT " chunks",
202 old_capacity, new_capacity);
203 }
204 }
205
206 G1CMMarkStack::~G1CMMarkStack() {
207 if (_base != NULL) {
208 MmapArrayAllocator<TaskQueueEntryChunk>::free(_base, _chunk_capacity);
209 }
210 }
211
212 void G1CMMarkStack::add_chunk_to_list(TaskQueueEntryChunk* volatile* list, TaskQueueEntryChunk* elem) {
213 elem->next = *list;
214 *list = elem;
215 }
216
217 void G1CMMarkStack::add_chunk_to_chunk_list(TaskQueueEntryChunk* elem) {
218 MutexLockerEx x(MarkStackChunkList_lock, Mutex::_no_safepoint_check_flag);
219 add_chunk_to_list(&_chunk_list, elem);
220 _chunks_in_chunk_list++;
221 }
222
223 void G1CMMarkStack::add_chunk_to_free_list(TaskQueueEntryChunk* elem) {
224 MutexLockerEx x(MarkStackFreeList_lock, Mutex::_no_safepoint_check_flag);
225 add_chunk_to_list(&_free_list, elem);
226 }
227
228 G1CMMarkStack::TaskQueueEntryChunk* G1CMMarkStack::remove_chunk_from_list(TaskQueueEntryChunk* volatile* list) {
229 TaskQueueEntryChunk* result = *list;
230 if (result != NULL) {
231 *list = (*list)->next;
232 }
233 return result;
234 }
235
236 G1CMMarkStack::TaskQueueEntryChunk* G1CMMarkStack::remove_chunk_from_chunk_list() {
237 MutexLockerEx x(MarkStackChunkList_lock, Mutex::_no_safepoint_check_flag);
238 TaskQueueEntryChunk* result = remove_chunk_from_list(&_chunk_list);
239 if (result != NULL) {
240 _chunks_in_chunk_list--;
241 }
242 return result;
243 }
244
245 G1CMMarkStack::TaskQueueEntryChunk* G1CMMarkStack::remove_chunk_from_free_list() {
246 MutexLockerEx x(MarkStackFreeList_lock, Mutex::_no_safepoint_check_flag);
247 return remove_chunk_from_list(&_free_list);
248 }
249
250 G1CMMarkStack::TaskQueueEntryChunk* G1CMMarkStack::allocate_new_chunk() {
251 // This dirty read of _hwm is okay because we only ever increase the _hwm in parallel code.
252 // Further this limits _hwm to a value of _chunk_capacity + #threads, avoiding
253 // wraparound of _hwm.
254 if (_hwm >= _chunk_capacity) {
255 return NULL;
256 }
257
258 size_t cur_idx = Atomic::add(1, &_hwm) - 1;
259 if (cur_idx >= _chunk_capacity) {
260 return NULL;
261 }
262
263 TaskQueueEntryChunk* result = ::new (&_base[cur_idx]) TaskQueueEntryChunk;
264 result->next = NULL;
265 return result;
266 }
267
268 bool G1CMMarkStack::par_push_chunk(G1TaskQueueEntry* ptr_arr) {
269 // Get a new chunk.
270 TaskQueueEntryChunk* new_chunk = remove_chunk_from_free_list();
271
272 if (new_chunk == NULL) {
273 // Did not get a chunk from the free list. Allocate from backing memory.
274 new_chunk = allocate_new_chunk();
275
276 if (new_chunk == NULL) {
277 return false;
278 }
279 }
280
281 Copy::conjoint_memory_atomic(ptr_arr, new_chunk->data, EntriesPerChunk * sizeof(G1TaskQueueEntry));
282
283 add_chunk_to_chunk_list(new_chunk);
284
285 return true;
286 }
287
288 bool G1CMMarkStack::par_pop_chunk(G1TaskQueueEntry* ptr_arr) {
289 TaskQueueEntryChunk* cur = remove_chunk_from_chunk_list();
290
291 if (cur == NULL) {
292 return false;
293 }
294
295 Copy::conjoint_memory_atomic(cur->data, ptr_arr, EntriesPerChunk * sizeof(G1TaskQueueEntry));
296
297 add_chunk_to_free_list(cur);
298 return true;
299 }
300
301 void G1CMMarkStack::set_empty() {
302 _chunks_in_chunk_list = 0;
303 _hwm = 0;
304 _chunk_list = NULL;
305 _free_list = NULL;
306 }
307
308 G1CMRootRegions::G1CMRootRegions() :
309 _cm(NULL), _scan_in_progress(false),
310 _should_abort(false), _claimed_survivor_index(0) { }
311
312 void G1CMRootRegions::init(const G1SurvivorRegions* survivors, G1ConcurrentMark* cm) {
313 _survivors = survivors;
314 _cm = cm;
315 }
316
317 void G1CMRootRegions::prepare_for_scan() {
318 assert(!scan_in_progress(), "pre-condition");
319
320 // Currently, only survivors can be root regions.
321 _claimed_survivor_index = 0;
322 _scan_in_progress = _survivors->regions()->is_nonempty();
323 _should_abort = false;
324 }
325
326 HeapRegion* G1CMRootRegions::claim_next() {
327 if (_should_abort) {
328 // If someone has set the should_abort flag, we return NULL to
329 // force the caller to bail out of their loop.
330 return NULL;
331 }
332
333 // Currently, only survivors can be root regions.
334 const GrowableArray<HeapRegion*>* survivor_regions = _survivors->regions();
335
336 int claimed_index = Atomic::add(1, &_claimed_survivor_index) - 1;
337 if (claimed_index < survivor_regions->length()) {
338 return survivor_regions->at(claimed_index);
339 }
340 return NULL;
341 }
342
343 uint G1CMRootRegions::num_root_regions() const {
344 return (uint)_survivors->regions()->length();
345 }
346
347 void G1CMRootRegions::notify_scan_done() {
348 MutexLockerEx x(RootRegionScan_lock, Mutex::_no_safepoint_check_flag);
349 _scan_in_progress = false;
350 RootRegionScan_lock->notify_all();
351 }
352
353 void G1CMRootRegions::cancel_scan() {
354 notify_scan_done();
355 }
356
357 void G1CMRootRegions::scan_finished() {
358 assert(scan_in_progress(), "pre-condition");
359
360 // Currently, only survivors can be root regions.
361 if (!_should_abort) {
362 assert(_claimed_survivor_index >= 0, "otherwise comparison is invalid: %d", _claimed_survivor_index);
363 assert((uint)_claimed_survivor_index >= _survivors->length(),
364 "we should have claimed all survivors, claimed index = %u, length = %u",
365 (uint)_claimed_survivor_index, _survivors->length());
366 }
367
368 notify_scan_done();
369 }
370
371 bool G1CMRootRegions::wait_until_scan_finished() {
372 if (!scan_in_progress()) return false;
373
374 {
375 MutexLockerEx x(RootRegionScan_lock, Mutex::_no_safepoint_check_flag);
376 while (scan_in_progress()) {
377 RootRegionScan_lock->wait(Mutex::_no_safepoint_check_flag);
378 }
379 }
380 return true;
381 }
382
383 uint G1ConcurrentMark::scale_parallel_threads(uint n_par_threads) {
384 return MAX2((n_par_threads + 2) / 4, 1U);
385 }
386
387 G1ConcurrentMark::G1ConcurrentMark(G1CollectedHeap* g1h, G1RegionToSpaceMapper* prev_bitmap_storage, G1RegionToSpaceMapper* next_bitmap_storage) :
388 _g1h(g1h),
389 _markBitMap1(),
390 _markBitMap2(),
391 _parallel_marking_threads(0),
392 _max_parallel_marking_threads(0),
393 _sleep_factor(0.0),
394 _marking_task_overhead(1.0),
395 _cleanup_list("Cleanup List"),
396
397 _prevMarkBitMap(&_markBitMap1),
398 _nextMarkBitMap(&_markBitMap2),
399
400 _global_mark_stack(),
401 // _finger set in set_non_marking_state
402
403 _max_worker_id(ParallelGCThreads),
404 // _active_tasks set in set_non_marking_state
405 // _tasks set inside the constructor
406 _task_queues(new G1CMTaskQueueSet((int) _max_worker_id)),
407 _terminator(ParallelTaskTerminator((int) _max_worker_id, _task_queues)),
408
409 _has_overflown(false),
410 _concurrent(false),
411 _has_aborted(false),
412 _restart_for_overflow(false),
413 _concurrent_marking_in_progress(false),
414 _gc_timer_cm(new (ResourceObj::C_HEAP, mtGC) ConcurrentGCTimer()),
415 _gc_tracer_cm(new (ResourceObj::C_HEAP, mtGC) G1OldTracer()),
416
417 // _verbose_level set below
418
419 _init_times(),
420 _remark_times(), _remark_mark_times(), _remark_weak_ref_times(),
421 _cleanup_times(),
422 _total_counting_time(0.0),
423 _total_rs_scrub_time(0.0),
424
425 _parallel_workers(NULL),
426
427 _completed_initialization(false) {
428
429 _markBitMap1.initialize(g1h->reserved_region(), prev_bitmap_storage);
430 _markBitMap2.initialize(g1h->reserved_region(), next_bitmap_storage);
431
432 // Create & start a ConcurrentMark thread.
433 _cmThread = new ConcurrentMarkThread(this);
434 assert(cmThread() != NULL, "CM Thread should have been created");
435 assert(cmThread()->cm() != NULL, "CM Thread should refer to this cm");
436 if (_cmThread->osthread() == NULL) {
437 vm_shutdown_during_initialization("Could not create ConcurrentMarkThread");
438 }
439
440 assert(CGC_lock != NULL, "Where's the CGC_lock?");
441 assert(_markBitMap1.covers(g1h->reserved_region()), "_markBitMap1 inconsistency");
442 assert(_markBitMap2.covers(g1h->reserved_region()), "_markBitMap2 inconsistency");
443
444 SATBMarkQueueSet& satb_qs = JavaThread::satb_mark_queue_set();
445 satb_qs.set_buffer_size(G1SATBBufferSize);
446
447 _root_regions.init(_g1h->survivor(), this);
448
449 if (ConcGCThreads > ParallelGCThreads) {
450 log_warning(gc)("Can't have more ConcGCThreads (%u) than ParallelGCThreads (%u).",
451 ConcGCThreads, ParallelGCThreads);
452 return;
453 }
454 if (!FLAG_IS_DEFAULT(ConcGCThreads) && ConcGCThreads > 0) {
455 // Note: ConcGCThreads has precedence over G1MarkingOverheadPercent
456 // if both are set
457 _sleep_factor = 0.0;
458 _marking_task_overhead = 1.0;
459 } else if (G1MarkingOverheadPercent > 0) {
460 // We will calculate the number of parallel marking threads based
461 // on a target overhead with respect to the soft real-time goal
462 double marking_overhead = (double) G1MarkingOverheadPercent / 100.0;
463 double overall_cm_overhead =
464 (double) MaxGCPauseMillis * marking_overhead /
465 (double) GCPauseIntervalMillis;
466 double cpu_ratio = 1.0 / os::initial_active_processor_count();
467 double marking_thread_num = ceil(overall_cm_overhead / cpu_ratio);
468 double marking_task_overhead =
469 overall_cm_overhead / marking_thread_num * os::initial_active_processor_count();
470 double sleep_factor =
471 (1.0 - marking_task_overhead) / marking_task_overhead;
472
473 FLAG_SET_ERGO(uint, ConcGCThreads, (uint) marking_thread_num);
474 _sleep_factor = sleep_factor;
475 _marking_task_overhead = marking_task_overhead;
476 } else {
477 // Calculate the number of parallel marking threads by scaling
478 // the number of parallel GC threads.
479 uint marking_thread_num = scale_parallel_threads(ParallelGCThreads);
480 FLAG_SET_ERGO(uint, ConcGCThreads, marking_thread_num);
481 _sleep_factor = 0.0;
482 _marking_task_overhead = 1.0;
483 }
484
485 assert(ConcGCThreads > 0, "Should have been set");
486 log_debug(gc)("ConcGCThreads: %u", ConcGCThreads);
487 log_debug(gc)("ParallelGCThreads: %u", ParallelGCThreads);
488 _parallel_marking_threads = ConcGCThreads;
489 _max_parallel_marking_threads = _parallel_marking_threads;
490
491 _parallel_workers = new WorkGang("G1 Marker",
492 _max_parallel_marking_threads, false, true);
493 if (_parallel_workers == NULL) {
494 vm_exit_during_initialization("Failed necessary allocation.");
495 } else {
496 _parallel_workers->initialize_workers();
497 }
498
499 if (FLAG_IS_DEFAULT(MarkStackSize)) {
500 size_t mark_stack_size =
501 MIN2(MarkStackSizeMax,
502 MAX2(MarkStackSize, (size_t) (parallel_marking_threads() * TASKQUEUE_SIZE)));
503 // Verify that the calculated value for MarkStackSize is in range.
504 // It would be nice to use the private utility routine from Arguments.
505 if (!(mark_stack_size >= 1 && mark_stack_size <= MarkStackSizeMax)) {
506 log_warning(gc)("Invalid value calculated for MarkStackSize (" SIZE_FORMAT "): "
507 "must be between 1 and " SIZE_FORMAT,
508 mark_stack_size, MarkStackSizeMax);
509 return;
510 }
511 FLAG_SET_ERGO(size_t, MarkStackSize, mark_stack_size);
512 } else {
513 // Verify MarkStackSize is in range.
514 if (FLAG_IS_CMDLINE(MarkStackSize)) {
515 if (FLAG_IS_DEFAULT(MarkStackSizeMax)) {
516 if (!(MarkStackSize >= 1 && MarkStackSize <= MarkStackSizeMax)) {
517 log_warning(gc)("Invalid value specified for MarkStackSize (" SIZE_FORMAT "): "
518 "must be between 1 and " SIZE_FORMAT,
519 MarkStackSize, MarkStackSizeMax);
520 return;
521 }
522 } else if (FLAG_IS_CMDLINE(MarkStackSizeMax)) {
523 if (!(MarkStackSize >= 1 && MarkStackSize <= MarkStackSizeMax)) {
524 log_warning(gc)("Invalid value specified for MarkStackSize (" SIZE_FORMAT ")"
525 " or for MarkStackSizeMax (" SIZE_FORMAT ")",
526 MarkStackSize, MarkStackSizeMax);
527 return;
528 }
529 }
530 }
531 }
532
533 if (!_global_mark_stack.initialize(MarkStackSize, MarkStackSizeMax)) {
534 vm_exit_during_initialization("Failed to allocate initial concurrent mark overflow mark stack.");
535 }
536
537 _tasks = NEW_C_HEAP_ARRAY(G1CMTask*, _max_worker_id, mtGC);
538 _accum_task_vtime = NEW_C_HEAP_ARRAY(double, _max_worker_id, mtGC);
539
540 // so that the assertion in MarkingTaskQueue::task_queue doesn't fail
541 _active_tasks = _max_worker_id;
542
543 for (uint i = 0; i < _max_worker_id; ++i) {
544 G1CMTaskQueue* task_queue = new G1CMTaskQueue();
545 task_queue->initialize();
546 _task_queues->register_queue(i, task_queue);
547
548 _tasks[i] = new G1CMTask(i, this, task_queue, _task_queues);
549
550 _accum_task_vtime[i] = 0.0;
551 }
552
553 // so that the call below can read a sensible value
554 _heap_start = g1h->reserved_region().start();
555 set_non_marking_state();
556 _completed_initialization = true;
557 }
558
559 void G1ConcurrentMark::reset() {
560 // Starting values for these two. This should be called in a STW
561 // phase.
562 MemRegion reserved = _g1h->g1_reserved();
563 _heap_start = reserved.start();
564 _heap_end = reserved.end();
565
566 // Separated the asserts so that we know which one fires.
567 assert(_heap_start != NULL, "heap bounds should look ok");
568 assert(_heap_end != NULL, "heap bounds should look ok");
569 assert(_heap_start < _heap_end, "heap bounds should look ok");
570
571 // Reset all the marking data structures and any necessary flags
572 reset_marking_state();
573
574 // We do reset all of them, since different phases will use
575 // different number of active threads. So, it's easiest to have all
576 // of them ready.
577 for (uint i = 0; i < _max_worker_id; ++i) {
578 _tasks[i]->reset(_nextMarkBitMap);
579 }
580
581 // we need this to make sure that the flag is on during the evac
582 // pause with initial mark piggy-backed
583 set_concurrent_marking_in_progress();
584 }
585
586
587 void G1ConcurrentMark::reset_marking_state() {
588 _global_mark_stack.set_empty();
589
590 // Expand the marking stack, if we have to and if we can.
591 if (has_overflown()) {
592 _global_mark_stack.expand();
593 }
594
595 clear_has_overflown();
596 _finger = _heap_start;
597
598 for (uint i = 0; i < _max_worker_id; ++i) {
599 G1CMTaskQueue* queue = _task_queues->queue(i);
600 queue->set_empty();
601 }
602 }
603
604 void G1ConcurrentMark::set_concurrency(uint active_tasks) {
605 assert(active_tasks <= _max_worker_id, "we should not have more");
606
607 _active_tasks = active_tasks;
608 // Need to update the three data structures below according to the
609 // number of active threads for this phase.
610 _terminator = ParallelTaskTerminator((int) active_tasks, _task_queues);
611 _first_overflow_barrier_sync.set_n_workers((int) active_tasks);
612 _second_overflow_barrier_sync.set_n_workers((int) active_tasks);
613 }
614
615 void G1ConcurrentMark::set_concurrency_and_phase(uint active_tasks, bool concurrent) {
616 set_concurrency(active_tasks);
617
618 _concurrent = concurrent;
619 // We propagate this to all tasks, not just the active ones.
620 for (uint i = 0; i < _max_worker_id; ++i)
621 _tasks[i]->set_concurrent(concurrent);
622
623 if (concurrent) {
624 set_concurrent_marking_in_progress();
625 } else {
626 // We currently assume that the concurrent flag has been set to
627 // false before we start remark. At this point we should also be
628 // in a STW phase.
629 assert(!concurrent_marking_in_progress(), "invariant");
630 assert(out_of_regions(),
631 "only way to get here: _finger: " PTR_FORMAT ", _heap_end: " PTR_FORMAT,
632 p2i(_finger), p2i(_heap_end));
633 }
634 }
635
636 void G1ConcurrentMark::set_non_marking_state() {
637 // We set the global marking state to some default values when we're
638 // not doing marking.
639 reset_marking_state();
640 _active_tasks = 0;
641 clear_concurrent_marking_in_progress();
642 }
643
644 G1ConcurrentMark::~G1ConcurrentMark() {
645 // The G1ConcurrentMark instance is never freed.
646 ShouldNotReachHere();
647 }
648
649 class G1ClearBitMapTask : public AbstractGangTask {
650 public:
651 static size_t chunk_size() { return M; }
652
653 private:
654 // Heap region closure used for clearing the given mark bitmap.
655 class G1ClearBitmapHRClosure : public HeapRegionClosure {
656 private:
657 G1CMBitMap* _bitmap;
658 G1ConcurrentMark* _cm;
659 public:
660 G1ClearBitmapHRClosure(G1CMBitMap* bitmap, G1ConcurrentMark* cm) : HeapRegionClosure(), _cm(cm), _bitmap(bitmap) {
661 }
662
663 virtual bool doHeapRegion(HeapRegion* r) {
664 size_t const chunk_size_in_words = G1ClearBitMapTask::chunk_size() / HeapWordSize;
665
666 HeapWord* cur = r->bottom();
667 HeapWord* const end = r->end();
668
669 while (cur < end) {
670 MemRegion mr(cur, MIN2(cur + chunk_size_in_words, end));
671 _bitmap->clear_range(mr);
672
673 cur += chunk_size_in_words;
674
675 // Abort iteration if after yielding the marking has been aborted.
676 if (_cm != NULL && _cm->do_yield_check() && _cm->has_aborted()) {
677 return true;
678 }
679 // Repeat the asserts from before the start of the closure. We will do them
680 // as asserts here to minimize their overhead on the product. However, we
681 // will have them as guarantees at the beginning / end of the bitmap
682 // clearing to get some checking in the product.
683 assert(_cm == NULL || _cm->cmThread()->during_cycle(), "invariant");
684 assert(_cm == NULL || !G1CollectedHeap::heap()->collector_state()->mark_in_progress(), "invariant");
685 }
686 assert(cur == end, "Must have completed iteration over the bitmap for region %u.", r->hrm_index());
687
688 return false;
689 }
690 };
691
692 G1ClearBitmapHRClosure _cl;
693 HeapRegionClaimer _hr_claimer;
694 bool _suspendible; // If the task is suspendible, workers must join the STS.
695
696 public:
697 G1ClearBitMapTask(G1CMBitMap* bitmap, G1ConcurrentMark* cm, uint n_workers, bool suspendible) :
698 AbstractGangTask("G1 Clear Bitmap"),
699 _cl(bitmap, suspendible ? cm : NULL),
700 _hr_claimer(n_workers),
701 _suspendible(suspendible)
702 { }
703
704 void work(uint worker_id) {
705 SuspendibleThreadSetJoiner sts_join(_suspendible);
706 G1CollectedHeap::heap()->heap_region_par_iterate(&_cl, worker_id, &_hr_claimer);
707 }
708
709 bool is_complete() {
710 return _cl.complete();
711 }
712 };
713
714 void G1ConcurrentMark::clear_bitmap(G1CMBitMap* bitmap, WorkGang* workers, bool may_yield) {
715 assert(may_yield || SafepointSynchronize::is_at_safepoint(), "Non-yielding bitmap clear only allowed at safepoint.");
716
717 size_t const num_bytes_to_clear = (HeapRegion::GrainBytes * _g1h->num_regions()) / G1CMBitMap::heap_map_factor();
718 size_t const num_chunks = align_up(num_bytes_to_clear, G1ClearBitMapTask::chunk_size()) / G1ClearBitMapTask::chunk_size();
719
720 uint const num_workers = (uint)MIN2(num_chunks, (size_t)workers->active_workers());
721
722 G1ClearBitMapTask cl(bitmap, this, num_workers, may_yield);
723
724 log_debug(gc, ergo)("Running %s with %u workers for " SIZE_FORMAT " work units.", cl.name(), num_workers, num_chunks);
725 workers->run_task(&cl, num_workers);
726 guarantee(!may_yield || cl.is_complete(), "Must have completed iteration when not yielding.");
727 }
728
729 void G1ConcurrentMark::cleanup_for_next_mark() {
730 // Make sure that the concurrent mark thread looks to still be in
731 // the current cycle.
732 guarantee(cmThread()->during_cycle(), "invariant");
733
734 // We are finishing up the current cycle by clearing the next
735 // marking bitmap and getting it ready for the next cycle. During
736 // this time no other cycle can start. So, let's make sure that this
737 // is the case.
738 guarantee(!_g1h->collector_state()->mark_in_progress(), "invariant");
739
740 clear_bitmap(_nextMarkBitMap, _parallel_workers, true);
741
742 // Clear the live count data. If the marking has been aborted, the abort()
743 // call already did that.
744 if (!has_aborted()) {
745 clear_live_data(_parallel_workers);
746 DEBUG_ONLY(verify_live_data_clear());
747 }
748
749 // Repeat the asserts from above.
750 guarantee(cmThread()->during_cycle(), "invariant");
751 guarantee(!_g1h->collector_state()->mark_in_progress(), "invariant");
752 }
753
754 void G1ConcurrentMark::clear_prev_bitmap(WorkGang* workers) {
755 assert(SafepointSynchronize::is_at_safepoint(), "Should only clear the entire prev bitmap at a safepoint.");
756 clear_bitmap((G1CMBitMap*)_prevMarkBitMap, workers, false);
757 }
758
759 class CheckBitmapClearHRClosure : public HeapRegionClosure {
760 G1CMBitMap* _bitmap;
761 bool _error;
762 public:
763 CheckBitmapClearHRClosure(G1CMBitMap* bitmap) : _bitmap(bitmap) {
764 }
765
766 virtual bool doHeapRegion(HeapRegion* r) {
767 // This closure can be called concurrently to the mutator, so we must make sure
768 // that the result of the getNextMarkedWordAddress() call is compared to the
769 // value passed to it as limit to detect any found bits.
770 // end never changes in G1.
771 HeapWord* end = r->end();
772 return _bitmap->getNextMarkedWordAddress(r->bottom(), end) != end;
773 }
774 };
775
776 bool G1ConcurrentMark::nextMarkBitmapIsClear() {
777 CheckBitmapClearHRClosure cl(_nextMarkBitMap);
778 _g1h->heap_region_iterate(&cl);
779 return cl.complete();
780 }
781
782 class NoteStartOfMarkHRClosure: public HeapRegionClosure {
783 public:
784 bool doHeapRegion(HeapRegion* r) {
785 r->note_start_of_marking();
786 return false;
787 }
788 };
789
790 void G1ConcurrentMark::checkpointRootsInitialPre() {
791 G1CollectedHeap* g1h = G1CollectedHeap::heap();
792 G1Policy* g1p = g1h->g1_policy();
793
794 _has_aborted = false;
795
796 // Initialize marking structures. This has to be done in a STW phase.
797 reset();
798
799 // For each region note start of marking.
800 NoteStartOfMarkHRClosure startcl;
801 g1h->heap_region_iterate(&startcl);
802 }
803
804
805 void G1ConcurrentMark::checkpointRootsInitialPost() {
806 G1CollectedHeap* g1h = G1CollectedHeap::heap();
807
808 // Start Concurrent Marking weak-reference discovery.
809 ReferenceProcessor* rp = g1h->ref_processor_cm();
810 // enable ("weak") refs discovery
811 rp->enable_discovery();
812 rp->setup_policy(false); // snapshot the soft ref policy to be used in this cycle
813
814 SATBMarkQueueSet& satb_mq_set = JavaThread::satb_mark_queue_set();
815 // This is the start of the marking cycle, we're expected all
816 // threads to have SATB queues with active set to false.
817 satb_mq_set.set_active_all_threads(true, /* new active value */
818 false /* expected_active */);
819
820 _root_regions.prepare_for_scan();
821
822 // update_g1_committed() will be called at the end of an evac pause
823 // when marking is on. So, it's also called at the end of the
824 // initial-mark pause to update the heap end, if the heap expands
825 // during it. No need to call it here.
826 }
827
828 /*
829 * Notice that in the next two methods, we actually leave the STS
830 * during the barrier sync and join it immediately afterwards. If we
831 * do not do this, the following deadlock can occur: one thread could
832 * be in the barrier sync code, waiting for the other thread to also
833 * sync up, whereas another one could be trying to yield, while also
834 * waiting for the other threads to sync up too.
835 *
836 * Note, however, that this code is also used during remark and in
837 * this case we should not attempt to leave / enter the STS, otherwise
838 * we'll either hit an assert (debug / fastdebug) or deadlock
839 * (product). So we should only leave / enter the STS if we are
840 * operating concurrently.
841 *
842 * Because the thread that does the sync barrier has left the STS, it
843 * is possible to be suspended for a Full GC or an evacuation pause
844 * could occur. This is actually safe, since the entering the sync
845 * barrier is one of the last things do_marking_step() does, and it
846 * doesn't manipulate any data structures afterwards.
847 */
848
849 void G1ConcurrentMark::enter_first_sync_barrier(uint worker_id) {
850 bool barrier_aborted;
851 {
852 SuspendibleThreadSetLeaver sts_leave(concurrent());
853 barrier_aborted = !_first_overflow_barrier_sync.enter();
854 }
855
856 // at this point everyone should have synced up and not be doing any
857 // more work
858
859 if (barrier_aborted) {
860 // If the barrier aborted we ignore the overflow condition and
861 // just abort the whole marking phase as quickly as possible.
862 return;
863 }
864
865 // If we're executing the concurrent phase of marking, reset the marking
866 // state; otherwise the marking state is reset after reference processing,
867 // during the remark pause.
868 // If we reset here as a result of an overflow during the remark we will
869 // see assertion failures from any subsequent set_concurrency_and_phase()
870 // calls.
871 if (concurrent()) {
872 // let the task associated with with worker 0 do this
873 if (worker_id == 0) {
874 // task 0 is responsible for clearing the global data structures
875 // We should be here because of an overflow. During STW we should
876 // not clear the overflow flag since we rely on it being true when
877 // we exit this method to abort the pause and restart concurrent
878 // marking.
879 reset_marking_state();
880
881 log_info(gc, marking)("Concurrent Mark reset for overflow");
882 }
883 }
884
885 // after this, each task should reset its own data structures then
886 // then go into the second barrier
887 }
888
889 void G1ConcurrentMark::enter_second_sync_barrier(uint worker_id) {
890 SuspendibleThreadSetLeaver sts_leave(concurrent());
891 _second_overflow_barrier_sync.enter();
892
893 // at this point everything should be re-initialized and ready to go
894 }
895
896 class G1CMConcurrentMarkingTask: public AbstractGangTask {
897 private:
898 G1ConcurrentMark* _cm;
899 ConcurrentMarkThread* _cmt;
900
901 public:
902 void work(uint worker_id) {
903 assert(Thread::current()->is_ConcurrentGC_thread(),
904 "this should only be done by a conc GC thread");
905 ResourceMark rm;
906
907 double start_vtime = os::elapsedVTime();
908
909 {
910 SuspendibleThreadSetJoiner sts_join;
911
912 assert(worker_id < _cm->active_tasks(), "invariant");
913 G1CMTask* the_task = _cm->task(worker_id);
914 the_task->record_start_time();
915 if (!_cm->has_aborted()) {
916 do {
917 double start_vtime_sec = os::elapsedVTime();
918 double mark_step_duration_ms = G1ConcMarkStepDurationMillis;
919
920 the_task->do_marking_step(mark_step_duration_ms,
921 true /* do_termination */,
922 false /* is_serial*/);
923
924 double end_vtime_sec = os::elapsedVTime();
925 double elapsed_vtime_sec = end_vtime_sec - start_vtime_sec;
926 _cm->do_yield_check();
927
928 jlong sleep_time_ms;
929 if (!_cm->has_aborted() && the_task->has_aborted()) {
930 sleep_time_ms =
931 (jlong) (elapsed_vtime_sec * _cm->sleep_factor() * 1000.0);
932 {
933 SuspendibleThreadSetLeaver sts_leave;
934 os::sleep(Thread::current(), sleep_time_ms, false);
935 }
936 }
937 } while (!_cm->has_aborted() && the_task->has_aborted());
938 }
939 the_task->record_end_time();
940 guarantee(!the_task->has_aborted() || _cm->has_aborted(), "invariant");
941 }
942
943 double end_vtime = os::elapsedVTime();
944 _cm->update_accum_task_vtime(worker_id, end_vtime - start_vtime);
945 }
946
947 G1CMConcurrentMarkingTask(G1ConcurrentMark* cm,
948 ConcurrentMarkThread* cmt) :
949 AbstractGangTask("Concurrent Mark"), _cm(cm), _cmt(cmt) { }
950
951 ~G1CMConcurrentMarkingTask() { }
952 };
953
954 // Calculates the number of active workers for a concurrent
955 // phase.
956 uint G1ConcurrentMark::calc_parallel_marking_threads() {
957 uint n_conc_workers = 0;
958 if (!UseDynamicNumberOfGCThreads ||
959 (!FLAG_IS_DEFAULT(ConcGCThreads) &&
960 !ForceDynamicNumberOfGCThreads)) {
961 n_conc_workers = max_parallel_marking_threads();
962 } else {
963 n_conc_workers =
964 AdaptiveSizePolicy::calc_default_active_workers(max_parallel_marking_threads(),
965 1, /* Minimum workers */
966 parallel_marking_threads(),
967 Threads::number_of_non_daemon_threads());
968 // Don't scale down "n_conc_workers" by scale_parallel_threads() because
969 // that scaling has already gone into "_max_parallel_marking_threads".
970 }
971 assert(n_conc_workers > 0 && n_conc_workers <= max_parallel_marking_threads(),
972 "Calculated number of workers must be larger than zero and at most the maximum %u, but is %u",
973 max_parallel_marking_threads(), n_conc_workers);
974 return n_conc_workers;
975 }
976
977 void G1ConcurrentMark::scanRootRegion(HeapRegion* hr) {
978 // Currently, only survivors can be root regions.
979 assert(hr->next_top_at_mark_start() == hr->bottom(), "invariant");
980 G1RootRegionScanClosure cl(_g1h, this);
981
982 const uintx interval = PrefetchScanIntervalInBytes;
983 HeapWord* curr = hr->bottom();
984 const HeapWord* end = hr->top();
985 while (curr < end) {
986 Prefetch::read(curr, interval);
987 oop obj = oop(curr);
988 int size = obj->oop_iterate_size(&cl);
989 assert(size == obj->size(), "sanity");
990 curr += size;
991 }
992 }
993
994 class G1CMRootRegionScanTask : public AbstractGangTask {
995 private:
996 G1ConcurrentMark* _cm;
997
998 public:
999 G1CMRootRegionScanTask(G1ConcurrentMark* cm) :
1000 AbstractGangTask("G1 Root Region Scan"), _cm(cm) { }
1001
1002 void work(uint worker_id) {
1003 assert(Thread::current()->is_ConcurrentGC_thread(),
1004 "this should only be done by a conc GC thread");
1005
1006 G1CMRootRegions* root_regions = _cm->root_regions();
1007 HeapRegion* hr = root_regions->claim_next();
1008 while (hr != NULL) {
1009 _cm->scanRootRegion(hr);
1010 hr = root_regions->claim_next();
1011 }
1012 }
1013 };
1014
1015 void G1ConcurrentMark::scan_root_regions() {
1016 // scan_in_progress() will have been set to true only if there was
1017 // at least one root region to scan. So, if it's false, we
1018 // should not attempt to do any further work.
1019 if (root_regions()->scan_in_progress()) {
1020 assert(!has_aborted(), "Aborting before root region scanning is finished not supported.");
1021
1022 _parallel_marking_threads = MIN2(calc_parallel_marking_threads(),
1023 // We distribute work on a per-region basis, so starting
1024 // more threads than that is useless.
1025 root_regions()->num_root_regions());
1026 assert(parallel_marking_threads() <= max_parallel_marking_threads(),
1027 "Maximum number of marking threads exceeded");
1028
1029 G1CMRootRegionScanTask task(this);
1030 log_debug(gc, ergo)("Running %s using %u workers for %u work units.",
1031 task.name(), _parallel_marking_threads, root_regions()->num_root_regions());
1032 _parallel_workers->run_task(&task, _parallel_marking_threads);
1033
1034 // It's possible that has_aborted() is true here without actually
1035 // aborting the survivor scan earlier. This is OK as it's
1036 // mainly used for sanity checking.
1037 root_regions()->scan_finished();
1038 }
1039 }
1040
1041 void G1ConcurrentMark::concurrent_cycle_start() {
1042 _gc_timer_cm->register_gc_start();
1043
1044 _gc_tracer_cm->report_gc_start(GCCause::_no_gc /* first parameter is not used */, _gc_timer_cm->gc_start());
1045
1046 _g1h->trace_heap_before_gc(_gc_tracer_cm);
1047 }
1048
1049 void G1ConcurrentMark::concurrent_cycle_end() {
1050 _g1h->trace_heap_after_gc(_gc_tracer_cm);
1051
1052 if (has_aborted()) {
1053 _gc_tracer_cm->report_concurrent_mode_failure();
1054 }
1055
1056 _gc_timer_cm->register_gc_end();
1057
1058 _gc_tracer_cm->report_gc_end(_gc_timer_cm->gc_end(), _gc_timer_cm->time_partitions());
1059 }
1060
1061 void G1ConcurrentMark::mark_from_roots() {
1062 // we might be tempted to assert that:
1063 // assert(asynch == !SafepointSynchronize::is_at_safepoint(),
1064 // "inconsistent argument?");
1065 // However that wouldn't be right, because it's possible that
1066 // a safepoint is indeed in progress as a younger generation
1067 // stop-the-world GC happens even as we mark in this generation.
1068
1069 _restart_for_overflow = false;
1070
1071 // _g1h has _n_par_threads
1072 _parallel_marking_threads = calc_parallel_marking_threads();
1073 assert(parallel_marking_threads() <= max_parallel_marking_threads(),
1074 "Maximum number of marking threads exceeded");
1075
1076 uint active_workers = MAX2(1U, parallel_marking_threads());
1077 assert(active_workers > 0, "Should have been set");
1078
1079 // Setting active workers is not guaranteed since fewer
1080 // worker threads may currently exist and more may not be
1081 // available.
1082 active_workers = _parallel_workers->update_active_workers(active_workers);
1083 log_info(gc, task)("Using %u workers of %u for marking", active_workers, _parallel_workers->total_workers());
1084
1085 // Parallel task terminator is set in "set_concurrency_and_phase()"
1086 set_concurrency_and_phase(active_workers, true /* concurrent */);
1087
1088 G1CMConcurrentMarkingTask markingTask(this, cmThread());
1089 _parallel_workers->run_task(&markingTask);
1090 print_stats();
1091 }
1092
1093 void G1ConcurrentMark::checkpointRootsFinal(bool clear_all_soft_refs) {
1094 // world is stopped at this checkpoint
1095 assert(SafepointSynchronize::is_at_safepoint(),
1096 "world should be stopped");
1097
1098 G1CollectedHeap* g1h = G1CollectedHeap::heap();
1099
1100 // If a full collection has happened, we shouldn't do this.
1101 if (has_aborted()) {
1102 g1h->collector_state()->set_mark_in_progress(false); // So bitmap clearing isn't confused
1103 return;
1104 }
1105
1106 SvcGCMarker sgcm(SvcGCMarker::OTHER);
1107
1108 if (VerifyDuringGC) {
1109 HandleMark hm; // handle scope
1110 g1h->prepare_for_verify();
1111 Universe::verify(VerifyOption_G1UsePrevMarking, "During GC (before)");
1112 }
1113 g1h->verifier()->check_bitmaps("Remark Start");
1114
1115 G1Policy* g1p = g1h->g1_policy();
1116 g1p->record_concurrent_mark_remark_start();
1117
1118 double start = os::elapsedTime();
1119
1120 checkpointRootsFinalWork();
1121
1122 double mark_work_end = os::elapsedTime();
1123
1124 weakRefsWork(clear_all_soft_refs);
1125
1126 if (has_overflown()) {
1127 // We overflowed. Restart concurrent marking.
1128 _restart_for_overflow = true;
1129
1130 // Verify the heap w.r.t. the previous marking bitmap.
1131 if (VerifyDuringGC) {
1132 HandleMark hm; // handle scope
1133 g1h->prepare_for_verify();
1134 Universe::verify(VerifyOption_G1UsePrevMarking, "During GC (overflow)");
1135 }
1136
1137 // Clear the marking state because we will be restarting
1138 // marking due to overflowing the global mark stack.
1139 reset_marking_state();
1140 } else {
1141 SATBMarkQueueSet& satb_mq_set = JavaThread::satb_mark_queue_set();
1142 // We're done with marking.
1143 // This is the end of the marking cycle, we're expected all
1144 // threads to have SATB queues with active set to true.
1145 satb_mq_set.set_active_all_threads(false, /* new active value */
1146 true /* expected_active */);
1147
1148 if (VerifyDuringGC) {
1149 HandleMark hm; // handle scope
1150 g1h->prepare_for_verify();
1151 Universe::verify(VerifyOption_G1UseNextMarking, "During GC (after)");
1152 }
1153 g1h->verifier()->check_bitmaps("Remark End");
1154 assert(!restart_for_overflow(), "sanity");
1155 // Completely reset the marking state since marking completed
1156 set_non_marking_state();
1157 }
1158
1159 // Statistics
1160 double now = os::elapsedTime();
1161 _remark_mark_times.add((mark_work_end - start) * 1000.0);
1162 _remark_weak_ref_times.add((now - mark_work_end) * 1000.0);
1163 _remark_times.add((now - start) * 1000.0);
1164
1165 g1p->record_concurrent_mark_remark_end();
1166
1167 G1CMIsAliveClosure is_alive(g1h);
1168 _gc_tracer_cm->report_object_count_after_gc(&is_alive);
1169 }
1170
1171 class G1NoteEndOfConcMarkClosure : public HeapRegionClosure {
1172 G1CollectedHeap* _g1;
1173 size_t _freed_bytes;
1174 FreeRegionList* _local_cleanup_list;
1175 uint _old_regions_removed;
1176 uint _humongous_regions_removed;
1177 HRRSCleanupTask* _hrrs_cleanup_task;
1178
1179 public:
1180 G1NoteEndOfConcMarkClosure(G1CollectedHeap* g1,
1181 FreeRegionList* local_cleanup_list,
1182 HRRSCleanupTask* hrrs_cleanup_task) :
1183 _g1(g1),
1184 _freed_bytes(0),
1185 _local_cleanup_list(local_cleanup_list),
1186 _old_regions_removed(0),
1187 _humongous_regions_removed(0),
1188 _hrrs_cleanup_task(hrrs_cleanup_task) { }
1189
1190 size_t freed_bytes() { return _freed_bytes; }
1191 const uint old_regions_removed() { return _old_regions_removed; }
1192 const uint humongous_regions_removed() { return _humongous_regions_removed; }
1193
1194 bool doHeapRegion(HeapRegion *hr) {
1195 _g1->reset_gc_time_stamps(hr);
1196 hr->note_end_of_marking();
1197
1198 if (hr->used() > 0 && hr->max_live_bytes() == 0 && !hr->is_young() && !hr->is_archive()) {
1199 _freed_bytes += hr->used();
1200 hr->set_containing_set(NULL);
1201 if (hr->is_humongous()) {
1202 _humongous_regions_removed++;
1203 _g1->free_humongous_region(hr, _local_cleanup_list, true /* skip_remset */);
1204 } else {
1205 _old_regions_removed++;
1206 _g1->free_region(hr, _local_cleanup_list, true /* skip_remset */);
1207 }
1208 } else {
1209 hr->rem_set()->do_cleanup_work(_hrrs_cleanup_task);
1210 }
1211
1212 return false;
1213 }
1214 };
1215
1216 class G1ParNoteEndTask: public AbstractGangTask {
1217 friend class G1NoteEndOfConcMarkClosure;
1218
1219 protected:
1220 G1CollectedHeap* _g1h;
1221 FreeRegionList* _cleanup_list;
1222 HeapRegionClaimer _hrclaimer;
1223
1224 public:
1225 G1ParNoteEndTask(G1CollectedHeap* g1h, FreeRegionList* cleanup_list, uint n_workers) :
1226 AbstractGangTask("G1 note end"), _g1h(g1h), _cleanup_list(cleanup_list), _hrclaimer(n_workers) {
1227 }
1228
1229 void work(uint worker_id) {
1230 FreeRegionList local_cleanup_list("Local Cleanup List");
1231 HRRSCleanupTask hrrs_cleanup_task;
1232 G1NoteEndOfConcMarkClosure g1_note_end(_g1h, &local_cleanup_list,
1233 &hrrs_cleanup_task);
1234 _g1h->heap_region_par_iterate(&g1_note_end, worker_id, &_hrclaimer);
1235 assert(g1_note_end.complete(), "Shouldn't have yielded!");
1236
1237 // Now update the lists
1238 _g1h->remove_from_old_sets(g1_note_end.old_regions_removed(), g1_note_end.humongous_regions_removed());
1239 {
1240 MutexLockerEx x(ParGCRareEvent_lock, Mutex::_no_safepoint_check_flag);
1241 _g1h->decrement_summary_bytes(g1_note_end.freed_bytes());
1242
1243 // If we iterate over the global cleanup list at the end of
1244 // cleanup to do this printing we will not guarantee to only
1245 // generate output for the newly-reclaimed regions (the list
1246 // might not be empty at the beginning of cleanup; we might
1247 // still be working on its previous contents). So we do the
1248 // printing here, before we append the new regions to the global
1249 // cleanup list.
1250
1251 G1HRPrinter* hr_printer = _g1h->hr_printer();
1252 if (hr_printer->is_active()) {
1253 FreeRegionListIterator iter(&local_cleanup_list);
1254 while (iter.more_available()) {
1255 HeapRegion* hr = iter.get_next();
1256 hr_printer->cleanup(hr);
1257 }
1258 }
1259
1260 _cleanup_list->add_ordered(&local_cleanup_list);
1261 assert(local_cleanup_list.is_empty(), "post-condition");
1262
1263 HeapRegionRemSet::finish_cleanup_task(&hrrs_cleanup_task);
1264 }
1265 }
1266 };
1267
1268 void G1ConcurrentMark::cleanup() {
1269 // world is stopped at this checkpoint
1270 assert(SafepointSynchronize::is_at_safepoint(),
1271 "world should be stopped");
1272 G1CollectedHeap* g1h = G1CollectedHeap::heap();
1273
1274 // If a full collection has happened, we shouldn't do this.
1275 if (has_aborted()) {
1276 g1h->collector_state()->set_mark_in_progress(false); // So bitmap clearing isn't confused
1277 return;
1278 }
1279
1280 g1h->verifier()->verify_region_sets_optional();
1281
1282 if (VerifyDuringGC) {
1283 HandleMark hm; // handle scope
1284 g1h->prepare_for_verify();
1285 Universe::verify(VerifyOption_G1UsePrevMarking, "During GC (before)");
1286 }
1287 g1h->verifier()->check_bitmaps("Cleanup Start");
1288
1289 G1Policy* g1p = g1h->g1_policy();
1290 g1p->record_concurrent_mark_cleanup_start();
1291
1292 double start = os::elapsedTime();
1293
1294 HeapRegionRemSet::reset_for_cleanup_tasks();
1295
1296 {
1297 GCTraceTime(Debug, gc)("Finalize Live Data");
1298 finalize_live_data();
1299 }
1300
1301 if (VerifyDuringGC) {
1302 GCTraceTime(Debug, gc)("Verify Live Data");
1303 verify_live_data();
1304 }
1305
1306 g1h->collector_state()->set_mark_in_progress(false);
1307
1308 double count_end = os::elapsedTime();
1309 double this_final_counting_time = (count_end - start);
1310 _total_counting_time += this_final_counting_time;
1311
1312 if (log_is_enabled(Trace, gc, liveness)) {
1313 G1PrintRegionLivenessInfoClosure cl("Post-Marking");
1314 _g1h->heap_region_iterate(&cl);
1315 }
1316
1317 // Install newly created mark bitMap as "prev".
1318 swapMarkBitMaps();
1319
1320 g1h->reset_gc_time_stamp();
1321
1322 uint n_workers = _g1h->workers()->active_workers();
1323
1324 // Note end of marking in all heap regions.
1325 G1ParNoteEndTask g1_par_note_end_task(g1h, &_cleanup_list, n_workers);
1326 g1h->workers()->run_task(&g1_par_note_end_task);
1327 g1h->check_gc_time_stamps();
1328
1329 if (!cleanup_list_is_empty()) {
1330 // The cleanup list is not empty, so we'll have to process it
1331 // concurrently. Notify anyone else that might be wanting free
1332 // regions that there will be more free regions coming soon.
1333 g1h->set_free_regions_coming();
1334 }
1335
1336 // call below, since it affects the metric by which we sort the heap
1337 // regions.
1338 if (G1ScrubRemSets) {
1339 double rs_scrub_start = os::elapsedTime();
1340 g1h->scrub_rem_set();
1341 _total_rs_scrub_time += (os::elapsedTime() - rs_scrub_start);
1342 }
1343
1344 // this will also free any regions totally full of garbage objects,
1345 // and sort the regions.
1346 g1h->g1_policy()->record_concurrent_mark_cleanup_end();
1347
1348 // Statistics.
1349 double end = os::elapsedTime();
1350 _cleanup_times.add((end - start) * 1000.0);
1351
1352 // Clean up will have freed any regions completely full of garbage.
1353 // Update the soft reference policy with the new heap occupancy.
1354 Universe::update_heap_info_at_gc();
1355
1356 if (VerifyDuringGC) {
1357 HandleMark hm; // handle scope
1358 g1h->prepare_for_verify();
1359 Universe::verify(VerifyOption_G1UsePrevMarking, "During GC (after)");
1360 }
1361
1362 g1h->verifier()->check_bitmaps("Cleanup End");
1363
1364 g1h->verifier()->verify_region_sets_optional();
1365
1366 // We need to make this be a "collection" so any collection pause that
1367 // races with it goes around and waits for completeCleanup to finish.
1368 g1h->increment_total_collections();
1369
1370 // Clean out dead classes and update Metaspace sizes.
1371 if (ClassUnloadingWithConcurrentMark) {
1372 ClassLoaderDataGraph::purge();
1373 }
1374 MetaspaceGC::compute_new_size();
1375
1376 // We reclaimed old regions so we should calculate the sizes to make
1377 // sure we update the old gen/space data.
1378 g1h->g1mm()->update_sizes();
1379 g1h->allocation_context_stats().update_after_mark();
1380 }
1381
1382 void G1ConcurrentMark::complete_cleanup() {
1383 if (has_aborted()) return;
1384
1385 G1CollectedHeap* g1h = G1CollectedHeap::heap();
1386
1387 _cleanup_list.verify_optional();
1388 FreeRegionList tmp_free_list("Tmp Free List");
1389
1390 log_develop_trace(gc, freelist)("G1ConcRegionFreeing [complete cleanup] : "
1391 "cleanup list has %u entries",
1392 _cleanup_list.length());
1393
1394 // No one else should be accessing the _cleanup_list at this point,
1395 // so it is not necessary to take any locks
1396 while (!_cleanup_list.is_empty()) {
1397 HeapRegion* hr = _cleanup_list.remove_region(true /* from_head */);
1398 assert(hr != NULL, "Got NULL from a non-empty list");
1399 hr->par_clear();
1400 tmp_free_list.add_ordered(hr);
1401
1402 // Instead of adding one region at a time to the secondary_free_list,
1403 // we accumulate them in the local list and move them a few at a
1404 // time. This also cuts down on the number of notify_all() calls
1405 // we do during this process. We'll also append the local list when
1406 // _cleanup_list is empty (which means we just removed the last
1407 // region from the _cleanup_list).
1408 if ((tmp_free_list.length() % G1SecondaryFreeListAppendLength == 0) ||
1409 _cleanup_list.is_empty()) {
1410 log_develop_trace(gc, freelist)("G1ConcRegionFreeing [complete cleanup] : "
1411 "appending %u entries to the secondary_free_list, "
1412 "cleanup list still has %u entries",
1413 tmp_free_list.length(),
1414 _cleanup_list.length());
1415
1416 {
1417 MutexLockerEx x(SecondaryFreeList_lock, Mutex::_no_safepoint_check_flag);
1418 g1h->secondary_free_list_add(&tmp_free_list);
1419 SecondaryFreeList_lock->notify_all();
1420 }
1421 #ifndef PRODUCT
1422 if (G1StressConcRegionFreeing) {
1423 for (uintx i = 0; i < G1StressConcRegionFreeingDelayMillis; ++i) {
1424 os::sleep(Thread::current(), (jlong) 1, false);
1425 }
1426 }
1427 #endif
1428 }
1429 }
1430 assert(tmp_free_list.is_empty(), "post-condition");
1431 }
1432
1433 // Supporting Object and Oop closures for reference discovery
1434 // and processing in during marking
1435
1436 bool G1CMIsAliveClosure::do_object_b(oop obj) {
1437 HeapWord* addr = (HeapWord*)obj;
1438 return addr != NULL &&
1439 (!_g1->is_in_g1_reserved(addr) || !_g1->is_obj_ill(obj));
1440 }
1441
1442 // 'Keep Alive' oop closure used by both serial parallel reference processing.
1443 // Uses the G1CMTask associated with a worker thread (for serial reference
1444 // processing the G1CMTask for worker 0 is used) to preserve (mark) and
1445 // trace referent objects.
1446 //
1447 // Using the G1CMTask and embedded local queues avoids having the worker
1448 // threads operating on the global mark stack. This reduces the risk
1449 // of overflowing the stack - which we would rather avoid at this late
1450 // state. Also using the tasks' local queues removes the potential
1451 // of the workers interfering with each other that could occur if
1452 // operating on the global stack.
1453
1454 class G1CMKeepAliveAndDrainClosure: public OopClosure {
1455 G1ConcurrentMark* _cm;
1456 G1CMTask* _task;
1457 int _ref_counter_limit;
1458 int _ref_counter;
1459 bool _is_serial;
1460 public:
1461 G1CMKeepAliveAndDrainClosure(G1ConcurrentMark* cm, G1CMTask* task, bool is_serial) :
1462 _cm(cm), _task(task), _is_serial(is_serial),
1463 _ref_counter_limit(G1RefProcDrainInterval) {
1464 assert(_ref_counter_limit > 0, "sanity");
1465 assert(!_is_serial || _task->worker_id() == 0, "only task 0 for serial code");
1466 _ref_counter = _ref_counter_limit;
1467 }
1468
1469 virtual void do_oop(narrowOop* p) { do_oop_work(p); }
1470 virtual void do_oop( oop* p) { do_oop_work(p); }
1471
1472 template <class T> void do_oop_work(T* p) {
1473 if (!_cm->has_overflown()) {
1474 oop obj = oopDesc::load_decode_heap_oop(p);
1475 _task->deal_with_reference(obj);
1476 _ref_counter--;
1477
1478 if (_ref_counter == 0) {
1479 // We have dealt with _ref_counter_limit references, pushing them
1480 // and objects reachable from them on to the local stack (and
1481 // possibly the global stack). Call G1CMTask::do_marking_step() to
1482 // process these entries.
1483 //
1484 // We call G1CMTask::do_marking_step() in a loop, which we'll exit if
1485 // there's nothing more to do (i.e. we're done with the entries that
1486 // were pushed as a result of the G1CMTask::deal_with_reference() calls
1487 // above) or we overflow.
1488 //
1489 // Note: G1CMTask::do_marking_step() can set the G1CMTask::has_aborted()
1490 // flag while there may still be some work to do. (See the comment at
1491 // the beginning of G1CMTask::do_marking_step() for those conditions -
1492 // one of which is reaching the specified time target.) It is only
1493 // when G1CMTask::do_marking_step() returns without setting the
1494 // has_aborted() flag that the marking step has completed.
1495 do {
1496 double mark_step_duration_ms = G1ConcMarkStepDurationMillis;
1497 _task->do_marking_step(mark_step_duration_ms,
1498 false /* do_termination */,
1499 _is_serial);
1500 } while (_task->has_aborted() && !_cm->has_overflown());
1501 _ref_counter = _ref_counter_limit;
1502 }
1503 }
1504 }
1505 };
1506
1507 // 'Drain' oop closure used by both serial and parallel reference processing.
1508 // Uses the G1CMTask associated with a given worker thread (for serial
1509 // reference processing the G1CMtask for worker 0 is used). Calls the
1510 // do_marking_step routine, with an unbelievably large timeout value,
1511 // to drain the marking data structures of the remaining entries
1512 // added by the 'keep alive' oop closure above.
1513
1514 class G1CMDrainMarkingStackClosure: public VoidClosure {
1515 G1ConcurrentMark* _cm;
1516 G1CMTask* _task;
1517 bool _is_serial;
1518 public:
1519 G1CMDrainMarkingStackClosure(G1ConcurrentMark* cm, G1CMTask* task, bool is_serial) :
1520 _cm(cm), _task(task), _is_serial(is_serial) {
1521 assert(!_is_serial || _task->worker_id() == 0, "only task 0 for serial code");
1522 }
1523
1524 void do_void() {
1525 do {
1526 // We call G1CMTask::do_marking_step() to completely drain the local
1527 // and global marking stacks of entries pushed by the 'keep alive'
1528 // oop closure (an instance of G1CMKeepAliveAndDrainClosure above).
1529 //
1530 // G1CMTask::do_marking_step() is called in a loop, which we'll exit
1531 // if there's nothing more to do (i.e. we've completely drained the
1532 // entries that were pushed as a a result of applying the 'keep alive'
1533 // closure to the entries on the discovered ref lists) or we overflow
1534 // the global marking stack.
1535 //
1536 // Note: G1CMTask::do_marking_step() can set the G1CMTask::has_aborted()
1537 // flag while there may still be some work to do. (See the comment at
1538 // the beginning of G1CMTask::do_marking_step() for those conditions -
1539 // one of which is reaching the specified time target.) It is only
1540 // when G1CMTask::do_marking_step() returns without setting the
1541 // has_aborted() flag that the marking step has completed.
1542
1543 _task->do_marking_step(1000000000.0 /* something very large */,
1544 true /* do_termination */,
1545 _is_serial);
1546 } while (_task->has_aborted() && !_cm->has_overflown());
1547 }
1548 };
1549
1550 // Implementation of AbstractRefProcTaskExecutor for parallel
1551 // reference processing at the end of G1 concurrent marking
1552
1553 class G1CMRefProcTaskExecutor: public AbstractRefProcTaskExecutor {
1554 private:
1555 G1CollectedHeap* _g1h;
1556 G1ConcurrentMark* _cm;
1557 WorkGang* _workers;
1558 uint _active_workers;
1559
1560 public:
1561 G1CMRefProcTaskExecutor(G1CollectedHeap* g1h,
1562 G1ConcurrentMark* cm,
1563 WorkGang* workers,
1564 uint n_workers) :
1565 _g1h(g1h), _cm(cm),
1566 _workers(workers), _active_workers(n_workers) { }
1567
1568 // Executes the given task using concurrent marking worker threads.
1569 virtual void execute(ProcessTask& task);
1570 virtual void execute(EnqueueTask& task);
1571 };
1572
1573 class G1CMRefProcTaskProxy: public AbstractGangTask {
1574 typedef AbstractRefProcTaskExecutor::ProcessTask ProcessTask;
1575 ProcessTask& _proc_task;
1576 G1CollectedHeap* _g1h;
1577 G1ConcurrentMark* _cm;
1578
1579 public:
1580 G1CMRefProcTaskProxy(ProcessTask& proc_task,
1581 G1CollectedHeap* g1h,
1582 G1ConcurrentMark* cm) :
1583 AbstractGangTask("Process reference objects in parallel"),
1584 _proc_task(proc_task), _g1h(g1h), _cm(cm) {
1585 ReferenceProcessor* rp = _g1h->ref_processor_cm();
1586 assert(rp->processing_is_mt(), "shouldn't be here otherwise");
1587 }
1588
1589 virtual void work(uint worker_id) {
1590 ResourceMark rm;
1591 HandleMark hm;
1592 G1CMTask* task = _cm->task(worker_id);
1593 G1CMIsAliveClosure g1_is_alive(_g1h);
1594 G1CMKeepAliveAndDrainClosure g1_par_keep_alive(_cm, task, false /* is_serial */);
1595 G1CMDrainMarkingStackClosure g1_par_drain(_cm, task, false /* is_serial */);
1596
1597 _proc_task.work(worker_id, g1_is_alive, g1_par_keep_alive, g1_par_drain);
1598 }
1599 };
1600
1601 void G1CMRefProcTaskExecutor::execute(ProcessTask& proc_task) {
1602 assert(_workers != NULL, "Need parallel worker threads.");
1603 assert(_g1h->ref_processor_cm()->processing_is_mt(), "processing is not MT");
1604
1605 G1CMRefProcTaskProxy proc_task_proxy(proc_task, _g1h, _cm);
1606
1607 // We need to reset the concurrency level before each
1608 // proxy task execution, so that the termination protocol
1609 // and overflow handling in G1CMTask::do_marking_step() knows
1610 // how many workers to wait for.
1611 _cm->set_concurrency(_active_workers);
1612 _workers->run_task(&proc_task_proxy);
1613 }
1614
1615 class G1CMRefEnqueueTaskProxy: public AbstractGangTask {
1616 typedef AbstractRefProcTaskExecutor::EnqueueTask EnqueueTask;
1617 EnqueueTask& _enq_task;
1618
1619 public:
1620 G1CMRefEnqueueTaskProxy(EnqueueTask& enq_task) :
1621 AbstractGangTask("Enqueue reference objects in parallel"),
1622 _enq_task(enq_task) { }
1623
1624 virtual void work(uint worker_id) {
1625 _enq_task.work(worker_id);
1626 }
1627 };
1628
1629 void G1CMRefProcTaskExecutor::execute(EnqueueTask& enq_task) {
1630 assert(_workers != NULL, "Need parallel worker threads.");
1631 assert(_g1h->ref_processor_cm()->processing_is_mt(), "processing is not MT");
1632
1633 G1CMRefEnqueueTaskProxy enq_task_proxy(enq_task);
1634
1635 // Not strictly necessary but...
1636 //
1637 // We need to reset the concurrency level before each
1638 // proxy task execution, so that the termination protocol
1639 // and overflow handling in G1CMTask::do_marking_step() knows
1640 // how many workers to wait for.
1641 _cm->set_concurrency(_active_workers);
1642 _workers->run_task(&enq_task_proxy);
1643 }
1644
1645 void G1ConcurrentMark::weakRefsWork(bool clear_all_soft_refs) {
1646 if (has_overflown()) {
1647 // Skip processing the discovered references if we have
1648 // overflown the global marking stack. Reference objects
1649 // only get discovered once so it is OK to not
1650 // de-populate the discovered reference lists. We could have,
1651 // but the only benefit would be that, when marking restarts,
1652 // less reference objects are discovered.
1653 return;
1654 }
1655
1656 ResourceMark rm;
1657 HandleMark hm;
1658
1659 G1CollectedHeap* g1h = G1CollectedHeap::heap();
1660
1661 // Is alive closure.
1662 G1CMIsAliveClosure g1_is_alive(g1h);
1663
1664 // Inner scope to exclude the cleaning of the string and symbol
1665 // tables from the displayed time.
1666 {
1667 GCTraceTime(Debug, gc, phases) trace("Reference Processing", _gc_timer_cm);
1668
1669 ReferenceProcessor* rp = g1h->ref_processor_cm();
1670
1671 // See the comment in G1CollectedHeap::ref_processing_init()
1672 // about how reference processing currently works in G1.
1673
1674 // Set the soft reference policy
1675 rp->setup_policy(clear_all_soft_refs);
1676 assert(_global_mark_stack.is_empty(), "mark stack should be empty");
1677
1678 // Instances of the 'Keep Alive' and 'Complete GC' closures used
1679 // in serial reference processing. Note these closures are also
1680 // used for serially processing (by the the current thread) the
1681 // JNI references during parallel reference processing.
1682 //
1683 // These closures do not need to synchronize with the worker
1684 // threads involved in parallel reference processing as these
1685 // instances are executed serially by the current thread (e.g.
1686 // reference processing is not multi-threaded and is thus
1687 // performed by the current thread instead of a gang worker).
1688 //
1689 // The gang tasks involved in parallel reference processing create
1690 // their own instances of these closures, which do their own
1691 // synchronization among themselves.
1692 G1CMKeepAliveAndDrainClosure g1_keep_alive(this, task(0), true /* is_serial */);
1693 G1CMDrainMarkingStackClosure g1_drain_mark_stack(this, task(0), true /* is_serial */);
1694
1695 // We need at least one active thread. If reference processing
1696 // is not multi-threaded we use the current (VMThread) thread,
1697 // otherwise we use the work gang from the G1CollectedHeap and
1698 // we utilize all the worker threads we can.
1699 bool processing_is_mt = rp->processing_is_mt();
1700 uint active_workers = (processing_is_mt ? g1h->workers()->active_workers() : 1U);
1701 active_workers = MAX2(MIN2(active_workers, _max_worker_id), 1U);
1702
1703 // Parallel processing task executor.
1704 G1CMRefProcTaskExecutor par_task_executor(g1h, this,
1705 g1h->workers(), active_workers);
1706 AbstractRefProcTaskExecutor* executor = (processing_is_mt ? &par_task_executor : NULL);
1707
1708 // Set the concurrency level. The phase was already set prior to
1709 // executing the remark task.
1710 set_concurrency(active_workers);
1711
1712 // Set the degree of MT processing here. If the discovery was done MT,
1713 // the number of threads involved during discovery could differ from
1714 // the number of active workers. This is OK as long as the discovered
1715 // Reference lists are balanced (see balance_all_queues() and balance_queues()).
1716 rp->set_active_mt_degree(active_workers);
1717
1718 // Process the weak references.
1719 const ReferenceProcessorStats& stats =
1720 rp->process_discovered_references(&g1_is_alive,
1721 &g1_keep_alive,
1722 &g1_drain_mark_stack,
1723 executor,
1724 _gc_timer_cm);
1725 _gc_tracer_cm->report_gc_reference_stats(stats);
1726
1727 // The do_oop work routines of the keep_alive and drain_marking_stack
1728 // oop closures will set the has_overflown flag if we overflow the
1729 // global marking stack.
1730
1731 assert(has_overflown() || _global_mark_stack.is_empty(),
1732 "Mark stack should be empty (unless it has overflown)");
1733
1734 assert(rp->num_q() == active_workers, "why not");
1735
1736 rp->enqueue_discovered_references(executor);
1737
1738 rp->verify_no_references_recorded();
1739 assert(!rp->discovery_enabled(), "Post condition");
1740 }
1741
1742 if (has_overflown()) {
1743 // We can not trust g1_is_alive if the marking stack overflowed
1744 return;
1745 }
1746
1747 assert(_global_mark_stack.is_empty(), "Marking should have completed");
1748
1749 // Unload Klasses, String, Symbols, Code Cache, etc.
1750 if (ClassUnloadingWithConcurrentMark) {
1751 GCTraceTime(Debug, gc, phases) debug("Class Unloading", _gc_timer_cm);
1752 bool purged_classes = SystemDictionary::do_unloading(&g1_is_alive, _gc_timer_cm, false /* Defer cleaning */);
1753 g1h->complete_cleaning(&g1_is_alive, purged_classes);
1754 } else {
1755 GCTraceTime(Debug, gc, phases) debug("Cleanup", _gc_timer_cm);
1756 // No need to clean string table and symbol table as they are treated as strong roots when
1757 // class unloading is disabled.
1758 g1h->partial_cleaning(&g1_is_alive, false, false, G1StringDedup::is_enabled());
1759
1760 }
1761 }
1762
1763 void G1ConcurrentMark::swapMarkBitMaps() {
1764 G1CMBitMapRO* temp = _prevMarkBitMap;
1765 _prevMarkBitMap = (G1CMBitMapRO*)_nextMarkBitMap;
1766 _nextMarkBitMap = (G1CMBitMap*) temp;
1767 }
1768
1769 // Closure for marking entries in SATB buffers.
1770 class G1CMSATBBufferClosure : public SATBBufferClosure {
1771 private:
1772 G1CMTask* _task;
1773 G1CollectedHeap* _g1h;
1774
1775 // This is very similar to G1CMTask::deal_with_reference, but with
1776 // more relaxed requirements for the argument, so this must be more
1777 // circumspect about treating the argument as an object.
1778 void do_entry(void* entry) const {
1779 _task->increment_refs_reached();
1780 HeapRegion* hr = _g1h->heap_region_containing(entry);
1781 if (entry < hr->next_top_at_mark_start()) {
1782 // Until we get here, we don't know whether entry refers to a valid
1783 // object; it could instead have been a stale reference.
1784 oop obj = static_cast<oop>(entry);
1785 assert(obj->is_oop(true /* ignore mark word */),
1786 "Invalid oop in SATB buffer: " PTR_FORMAT, p2i(obj));
1787 _task->make_reference_grey(obj);
1788 }
1789 }
1790
1791 public:
1792 G1CMSATBBufferClosure(G1CMTask* task, G1CollectedHeap* g1h)
1793 : _task(task), _g1h(g1h) { }
1794
1795 virtual void do_buffer(void** buffer, size_t size) {
1796 for (size_t i = 0; i < size; ++i) {
1797 do_entry(buffer[i]);
1798 }
1799 }
1800 };
1801
1802 class G1RemarkThreadsClosure : public ThreadClosure {
1803 G1CMSATBBufferClosure _cm_satb_cl;
1804 G1CMOopClosure _cm_cl;
1805 MarkingCodeBlobClosure _code_cl;
1806 int _thread_parity;
1807
1808 public:
1809 G1RemarkThreadsClosure(G1CollectedHeap* g1h, G1CMTask* task) :
1810 _cm_satb_cl(task, g1h),
1811 _cm_cl(g1h, g1h->concurrent_mark(), task),
1812 _code_cl(&_cm_cl, !CodeBlobToOopClosure::FixRelocations),
1813 _thread_parity(Threads::thread_claim_parity()) {}
1814
1815 void do_thread(Thread* thread) {
1816 if (thread->is_Java_thread()) {
1817 if (thread->claim_oops_do(true, _thread_parity)) {
1818 JavaThread* jt = (JavaThread*)thread;
1819
1820 // In theory it should not be neccessary to explicitly walk the nmethods to find roots for concurrent marking
1821 // however the liveness of oops reachable from nmethods have very complex lifecycles:
1822 // * Alive if on the stack of an executing method
1823 // * Weakly reachable otherwise
1824 // Some objects reachable from nmethods, such as the class loader (or klass_holder) of the receiver should be
1825 // live by the SATB invariant but other oops recorded in nmethods may behave differently.
1826 jt->nmethods_do(&_code_cl);
1827
1828 jt->satb_mark_queue().apply_closure_and_empty(&_cm_satb_cl);
1829 }
1830 } else if (thread->is_VM_thread()) {
1831 if (thread->claim_oops_do(true, _thread_parity)) {
1832 JavaThread::satb_mark_queue_set().shared_satb_queue()->apply_closure_and_empty(&_cm_satb_cl);
1833 }
1834 }
1835 }
1836 };
1837
1838 class G1CMRemarkTask: public AbstractGangTask {
1839 private:
1840 G1ConcurrentMark* _cm;
1841 public:
1842 void work(uint worker_id) {
1843 // Since all available tasks are actually started, we should
1844 // only proceed if we're supposed to be active.
1845 if (worker_id < _cm->active_tasks()) {
1846 G1CMTask* task = _cm->task(worker_id);
1847 task->record_start_time();
1848 {
1849 ResourceMark rm;
1850 HandleMark hm;
1851
1852 G1RemarkThreadsClosure threads_f(G1CollectedHeap::heap(), task);
1853 Threads::threads_do(&threads_f);
1854 }
1855
1856 do {
1857 task->do_marking_step(1000000000.0 /* something very large */,
1858 true /* do_termination */,
1859 false /* is_serial */);
1860 } while (task->has_aborted() && !_cm->has_overflown());
1861 // If we overflow, then we do not want to restart. We instead
1862 // want to abort remark and do concurrent marking again.
1863 task->record_end_time();
1864 }
1865 }
1866
1867 G1CMRemarkTask(G1ConcurrentMark* cm, uint active_workers) :
1868 AbstractGangTask("Par Remark"), _cm(cm) {
1869 _cm->terminator()->reset_for_reuse(active_workers);
1870 }
1871 };
1872
1873 void G1ConcurrentMark::checkpointRootsFinalWork() {
1874 ResourceMark rm;
1875 HandleMark hm;
1876 G1CollectedHeap* g1h = G1CollectedHeap::heap();
1877
1878 GCTraceTime(Debug, gc, phases) trace("Finalize Marking", _gc_timer_cm);
1879
1880 g1h->ensure_parsability(false);
1881
1882 // this is remark, so we'll use up all active threads
1883 uint active_workers = g1h->workers()->active_workers();
1884 set_concurrency_and_phase(active_workers, false /* concurrent */);
1885 // Leave _parallel_marking_threads at it's
1886 // value originally calculated in the G1ConcurrentMark
1887 // constructor and pass values of the active workers
1888 // through the gang in the task.
1889
1890 {
1891 StrongRootsScope srs(active_workers);
1892
1893 G1CMRemarkTask remarkTask(this, active_workers);
1894 // We will start all available threads, even if we decide that the
1895 // active_workers will be fewer. The extra ones will just bail out
1896 // immediately.
1897 g1h->workers()->run_task(&remarkTask);
1898 }
1899
1900 SATBMarkQueueSet& satb_mq_set = JavaThread::satb_mark_queue_set();
1901 guarantee(has_overflown() ||
1902 satb_mq_set.completed_buffers_num() == 0,
1903 "Invariant: has_overflown = %s, num buffers = " SIZE_FORMAT,
1904 BOOL_TO_STR(has_overflown()),
1905 satb_mq_set.completed_buffers_num());
1906
1907 print_stats();
1908 }
1909
1910 void G1ConcurrentMark::clearRangePrevBitmap(MemRegion mr) {
1911 // Note we are overriding the read-only view of the prev map here, via
1912 // the cast.
1913 ((G1CMBitMap*)_prevMarkBitMap)->clear_range(mr);
1914 }
1915
1916 HeapRegion*
1917 G1ConcurrentMark::claim_region(uint worker_id) {
1918 // "checkpoint" the finger
1919 HeapWord* finger = _finger;
1920
1921 // _heap_end will not change underneath our feet; it only changes at
1922 // yield points.
1923 while (finger < _heap_end) {
1924 assert(_g1h->is_in_g1_reserved(finger), "invariant");
1925
1926 HeapRegion* curr_region = _g1h->heap_region_containing(finger);
1927 // Make sure that the reads below do not float before loading curr_region.
1928 OrderAccess::loadload();
1929 // Above heap_region_containing may return NULL as we always scan claim
1930 // until the end of the heap. In this case, just jump to the next region.
1931 HeapWord* end = curr_region != NULL ? curr_region->end() : finger + HeapRegion::GrainWords;
1932
1933 // Is the gap between reading the finger and doing the CAS too long?
1934 HeapWord* res = (HeapWord*) Atomic::cmpxchg_ptr(end, &_finger, finger);
1935 if (res == finger && curr_region != NULL) {
1936 // we succeeded
1937 HeapWord* bottom = curr_region->bottom();
1938 HeapWord* limit = curr_region->next_top_at_mark_start();
1939
1940 // notice that _finger == end cannot be guaranteed here since,
1941 // someone else might have moved the finger even further
1942 assert(_finger >= end, "the finger should have moved forward");
1943
1944 if (limit > bottom) {
1945 return curr_region;
1946 } else {
1947 assert(limit == bottom,
1948 "the region limit should be at bottom");
1949 // we return NULL and the caller should try calling
1950 // claim_region() again.
1951 return NULL;
1952 }
1953 } else {
1954 assert(_finger > finger, "the finger should have moved forward");
1955 // read it again
1956 finger = _finger;
1957 }
1958 }
1959
1960 return NULL;
1961 }
1962
1963 #ifndef PRODUCT
1964 class VerifyNoCSetOops VALUE_OBJ_CLASS_SPEC {
1965 private:
1966 G1CollectedHeap* _g1h;
1967 const char* _phase;
1968 int _info;
1969
1970 public:
1971 VerifyNoCSetOops(const char* phase, int info = -1) :
1972 _g1h(G1CollectedHeap::heap()),
1973 _phase(phase),
1974 _info(info)
1975 { }
1976
1977 void operator()(G1TaskQueueEntry task_entry) const {
1978 if (task_entry.is_array_slice()) {
1979 guarantee(_g1h->is_in_reserved(task_entry.slice()), "Slice " PTR_FORMAT " must be in heap.", p2i(task_entry.slice()));
1980 return;
1981 }
1982 guarantee(task_entry.obj()->is_oop(),
1983 "Non-oop " PTR_FORMAT ", phase: %s, info: %d",
1984 p2i(task_entry.obj()), _phase, _info);
1985 guarantee(!_g1h->is_in_cset(task_entry.obj()),
1986 "obj: " PTR_FORMAT " in CSet, phase: %s, info: %d",
1987 p2i(task_entry.obj()), _phase, _info);
1988 }
1989 };
1990
1991 void G1ConcurrentMark::verify_no_cset_oops() {
1992 assert(SafepointSynchronize::is_at_safepoint(), "should be at a safepoint");
1993 if (!G1CollectedHeap::heap()->collector_state()->mark_in_progress()) {
1994 return;
1995 }
1996
1997 // Verify entries on the global mark stack
1998 _global_mark_stack.iterate(VerifyNoCSetOops("Stack"));
1999
2000 // Verify entries on the task queues
2001 for (uint i = 0; i < _max_worker_id; ++i) {
2002 G1CMTaskQueue* queue = _task_queues->queue(i);
2003 queue->iterate(VerifyNoCSetOops("Queue", i));
2004 }
2005
2006 // Verify the global finger
2007 HeapWord* global_finger = finger();
2008 if (global_finger != NULL && global_finger < _heap_end) {
2009 // Since we always iterate over all regions, we might get a NULL HeapRegion
2010 // here.
2011 HeapRegion* global_hr = _g1h->heap_region_containing(global_finger);
2012 guarantee(global_hr == NULL || global_finger == global_hr->bottom(),
2013 "global finger: " PTR_FORMAT " region: " HR_FORMAT,
2014 p2i(global_finger), HR_FORMAT_PARAMS(global_hr));
2015 }
2016
2017 // Verify the task fingers
2018 assert(parallel_marking_threads() <= _max_worker_id, "sanity");
2019 for (uint i = 0; i < parallel_marking_threads(); ++i) {
2020 G1CMTask* task = _tasks[i];
2021 HeapWord* task_finger = task->finger();
2022 if (task_finger != NULL && task_finger < _heap_end) {
2023 // See above note on the global finger verification.
2024 HeapRegion* task_hr = _g1h->heap_region_containing(task_finger);
2025 guarantee(task_hr == NULL || task_finger == task_hr->bottom() ||
2026 !task_hr->in_collection_set(),
2027 "task finger: " PTR_FORMAT " region: " HR_FORMAT,
2028 p2i(task_finger), HR_FORMAT_PARAMS(task_hr));
2029 }
2030 }
2031 }
2032 #endif // PRODUCT
2033 void G1ConcurrentMark::create_live_data() {
2034 _g1h->g1_rem_set()->create_card_live_data(_parallel_workers, _nextMarkBitMap);
2035 }
2036
2037 void G1ConcurrentMark::finalize_live_data() {
2038 _g1h->g1_rem_set()->finalize_card_live_data(_g1h->workers(), _nextMarkBitMap);
2039 }
2040
2041 void G1ConcurrentMark::verify_live_data() {
2042 _g1h->g1_rem_set()->verify_card_live_data(_g1h->workers(), _nextMarkBitMap);
2043 }
2044
2045 void G1ConcurrentMark::clear_live_data(WorkGang* workers) {
2046 _g1h->g1_rem_set()->clear_card_live_data(workers);
2047 }
2048
2049 #ifdef ASSERT
2050 void G1ConcurrentMark::verify_live_data_clear() {
2051 _g1h->g1_rem_set()->verify_card_live_data_is_clear();
2052 }
2053 #endif
2054
2055 void G1ConcurrentMark::print_stats() {
2056 if (!log_is_enabled(Debug, gc, stats)) {
2057 return;
2058 }
2059 log_debug(gc, stats)("---------------------------------------------------------------------");
2060 for (size_t i = 0; i < _active_tasks; ++i) {
2061 _tasks[i]->print_stats();
2062 log_debug(gc, stats)("---------------------------------------------------------------------");
2063 }
2064 }
2065
2066 void G1ConcurrentMark::abort() {
2067 if (!cmThread()->during_cycle() || _has_aborted) {
2068 // We haven't started a concurrent cycle or we have already aborted it. No need to do anything.
2069 return;
2070 }
2071
2072 // Clear all marks in the next bitmap for the next marking cycle. This will allow us to skip the next
2073 // concurrent bitmap clearing.
2074 {
2075 GCTraceTime(Debug, gc)("Clear Next Bitmap");
2076 clear_bitmap(_nextMarkBitMap, _g1h->workers(), false);
2077 }
2078 // Note we cannot clear the previous marking bitmap here
2079 // since VerifyDuringGC verifies the objects marked during
2080 // a full GC against the previous bitmap.
2081
2082 {
2083 GCTraceTime(Debug, gc)("Clear Live Data");
2084 clear_live_data(_g1h->workers());
2085 }
2086 DEBUG_ONLY({
2087 GCTraceTime(Debug, gc)("Verify Live Data Clear");
2088 verify_live_data_clear();
2089 })
2090 // Empty mark stack
2091 reset_marking_state();
2092 for (uint i = 0; i < _max_worker_id; ++i) {
2093 _tasks[i]->clear_region_fields();
2094 }
2095 _first_overflow_barrier_sync.abort();
2096 _second_overflow_barrier_sync.abort();
2097 _has_aborted = true;
2098
2099 SATBMarkQueueSet& satb_mq_set = JavaThread::satb_mark_queue_set();
2100 satb_mq_set.abandon_partial_marking();
2101 // This can be called either during or outside marking, we'll read
2102 // the expected_active value from the SATB queue set.
2103 satb_mq_set.set_active_all_threads(
2104 false, /* new active value */
2105 satb_mq_set.is_active() /* expected_active */);
2106 }
2107
2108 static void print_ms_time_info(const char* prefix, const char* name,
2109 NumberSeq& ns) {
2110 log_trace(gc, marking)("%s%5d %12s: total time = %8.2f s (avg = %8.2f ms).",
2111 prefix, ns.num(), name, ns.sum()/1000.0, ns.avg());
2112 if (ns.num() > 0) {
2113 log_trace(gc, marking)("%s [std. dev = %8.2f ms, max = %8.2f ms]",
2114 prefix, ns.sd(), ns.maximum());
2115 }
2116 }
2117
2118 void G1ConcurrentMark::print_summary_info() {
2119 Log(gc, marking) log;
2120 if (!log.is_trace()) {
2121 return;
2122 }
2123
2124 log.trace(" Concurrent marking:");
2125 print_ms_time_info(" ", "init marks", _init_times);
2126 print_ms_time_info(" ", "remarks", _remark_times);
2127 {
2128 print_ms_time_info(" ", "final marks", _remark_mark_times);
2129 print_ms_time_info(" ", "weak refs", _remark_weak_ref_times);
2130
2131 }
2132 print_ms_time_info(" ", "cleanups", _cleanup_times);
2133 log.trace(" Finalize live data total time = %8.2f s (avg = %8.2f ms).",
2134 _total_counting_time, (_cleanup_times.num() > 0 ? _total_counting_time * 1000.0 / (double)_cleanup_times.num() : 0.0));
2135 if (G1ScrubRemSets) {
2136 log.trace(" RS scrub total time = %8.2f s (avg = %8.2f ms).",
2137 _total_rs_scrub_time, (_cleanup_times.num() > 0 ? _total_rs_scrub_time * 1000.0 / (double)_cleanup_times.num() : 0.0));
2138 }
2139 log.trace(" Total stop_world time = %8.2f s.",
2140 (_init_times.sum() + _remark_times.sum() + _cleanup_times.sum())/1000.0);
2141 log.trace(" Total concurrent time = %8.2f s (%8.2f s marking).",
2142 cmThread()->vtime_accum(), cmThread()->vtime_mark_accum());
2143 }
2144
2145 void G1ConcurrentMark::print_worker_threads_on(outputStream* st) const {
2146 _parallel_workers->print_worker_threads_on(st);
2147 }
2148
2149 void G1ConcurrentMark::threads_do(ThreadClosure* tc) const {
2150 _parallel_workers->threads_do(tc);
2151 }
2152
2153 void G1ConcurrentMark::print_on_error(outputStream* st) const {
2154 st->print_cr("Marking Bits (Prev, Next): (CMBitMap*) " PTR_FORMAT ", (CMBitMap*) " PTR_FORMAT,
2155 p2i(_prevMarkBitMap), p2i(_nextMarkBitMap));
2156 _prevMarkBitMap->print_on_error(st, " Prev Bits: ");
2157 _nextMarkBitMap->print_on_error(st, " Next Bits: ");
2158 }
2159
2160 // Closure for iteration over bitmaps
2161 class G1CMBitMapClosure : public BitMapClosure {
2162 private:
2163 // the bitmap that is being iterated over
2164 G1CMBitMap* _nextMarkBitMap;
2165 G1ConcurrentMark* _cm;
2166 G1CMTask* _task;
2167
2168 public:
2169 G1CMBitMapClosure(G1CMTask *task, G1ConcurrentMark* cm, G1CMBitMap* nextMarkBitMap) :
2170 _task(task), _cm(cm), _nextMarkBitMap(nextMarkBitMap) { }
2171
2172 bool do_bit(size_t offset) {
2173 HeapWord* addr = _nextMarkBitMap->offsetToHeapWord(offset);
2174 assert(_nextMarkBitMap->isMarked(addr), "invariant");
2175 assert( addr < _cm->finger(), "invariant");
2176 assert(addr >= _task->finger(), "invariant");
2177
2178 // We move that task's local finger along.
2179 _task->move_finger_to(addr);
2180
2181 _task->scan_task_entry(G1TaskQueueEntry::from_oop(oop(addr)));
2182 // we only partially drain the local queue and global stack
2183 _task->drain_local_queue(true);
2184 _task->drain_global_stack(true);
2185
2186 // if the has_aborted flag has been raised, we need to bail out of
2187 // the iteration
2188 return !_task->has_aborted();
2189 }
2190 };
2191
2192 static ReferenceProcessor* get_cm_oop_closure_ref_processor(G1CollectedHeap* g1h) {
2193 ReferenceProcessor* result = g1h->ref_processor_cm();
2194 assert(result != NULL, "CM reference processor should not be NULL");
2195 return result;
2196 }
2197
2198 G1CMOopClosure::G1CMOopClosure(G1CollectedHeap* g1h,
2199 G1ConcurrentMark* cm,
2200 G1CMTask* task)
2201 : MetadataAwareOopClosure(get_cm_oop_closure_ref_processor(g1h)),
2202 _g1h(g1h), _cm(cm), _task(task)
2203 { }
2204
2205 void G1CMTask::setup_for_region(HeapRegion* hr) {
2206 assert(hr != NULL,
2207 "claim_region() should have filtered out NULL regions");
2208 _curr_region = hr;
2209 _finger = hr->bottom();
2210 update_region_limit();
2211 }
2212
2213 void G1CMTask::update_region_limit() {
2214 HeapRegion* hr = _curr_region;
2215 HeapWord* bottom = hr->bottom();
2216 HeapWord* limit = hr->next_top_at_mark_start();
2217
2218 if (limit == bottom) {
2219 // The region was collected underneath our feet.
2220 // We set the finger to bottom to ensure that the bitmap
2221 // iteration that will follow this will not do anything.
2222 // (this is not a condition that holds when we set the region up,
2223 // as the region is not supposed to be empty in the first place)
2224 _finger = bottom;
2225 } else if (limit >= _region_limit) {
2226 assert(limit >= _finger, "peace of mind");
2227 } else {
2228 assert(limit < _region_limit, "only way to get here");
2229 // This can happen under some pretty unusual circumstances. An
2230 // evacuation pause empties the region underneath our feet (NTAMS
2231 // at bottom). We then do some allocation in the region (NTAMS
2232 // stays at bottom), followed by the region being used as a GC
2233 // alloc region (NTAMS will move to top() and the objects
2234 // originally below it will be grayed). All objects now marked in
2235 // the region are explicitly grayed, if below the global finger,
2236 // and we do not need in fact to scan anything else. So, we simply
2237 // set _finger to be limit to ensure that the bitmap iteration
2238 // doesn't do anything.
2239 _finger = limit;
2240 }
2241
2242 _region_limit = limit;
2243 }
2244
2245 void G1CMTask::giveup_current_region() {
2246 assert(_curr_region != NULL, "invariant");
2247 clear_region_fields();
2248 }
2249
2250 void G1CMTask::clear_region_fields() {
2251 // Values for these three fields that indicate that we're not
2252 // holding on to a region.
2253 _curr_region = NULL;
2254 _finger = NULL;
2255 _region_limit = NULL;
2256 }
2257
2258 void G1CMTask::set_cm_oop_closure(G1CMOopClosure* cm_oop_closure) {
2259 if (cm_oop_closure == NULL) {
2260 assert(_cm_oop_closure != NULL, "invariant");
2261 } else {
2262 assert(_cm_oop_closure == NULL, "invariant");
2263 }
2264 _cm_oop_closure = cm_oop_closure;
2265 }
2266
2267 void G1CMTask::reset(G1CMBitMap* nextMarkBitMap) {
2268 guarantee(nextMarkBitMap != NULL, "invariant");
2269 _nextMarkBitMap = nextMarkBitMap;
2270 clear_region_fields();
2271
2272 _calls = 0;
2273 _elapsed_time_ms = 0.0;
2274 _termination_time_ms = 0.0;
2275 _termination_start_time_ms = 0.0;
2276 }
2277
2278 bool G1CMTask::should_exit_termination() {
2279 regular_clock_call();
2280 // This is called when we are in the termination protocol. We should
2281 // quit if, for some reason, this task wants to abort or the global
2282 // stack is not empty (this means that we can get work from it).
2283 return !_cm->mark_stack_empty() || has_aborted();
2284 }
2285
2286 void G1CMTask::reached_limit() {
2287 assert(_words_scanned >= _words_scanned_limit ||
2288 _refs_reached >= _refs_reached_limit ,
2289 "shouldn't have been called otherwise");
2290 regular_clock_call();
2291 }
2292
2293 void G1CMTask::regular_clock_call() {
2294 if (has_aborted()) return;
2295
2296 // First, we need to recalculate the words scanned and refs reached
2297 // limits for the next clock call.
2298 recalculate_limits();
2299
2300 // During the regular clock call we do the following
2301
2302 // (1) If an overflow has been flagged, then we abort.
2303 if (_cm->has_overflown()) {
2304 set_has_aborted();
2305 return;
2306 }
2307
2308 // If we are not concurrent (i.e. we're doing remark) we don't need
2309 // to check anything else. The other steps are only needed during
2310 // the concurrent marking phase.
2311 if (!concurrent()) return;
2312
2313 // (2) If marking has been aborted for Full GC, then we also abort.
2314 if (_cm->has_aborted()) {
2315 set_has_aborted();
2316 return;
2317 }
2318
2319 double curr_time_ms = os::elapsedVTime() * 1000.0;
2320
2321 // (4) We check whether we should yield. If we have to, then we abort.
2322 if (SuspendibleThreadSet::should_yield()) {
2323 // We should yield. To do this we abort the task. The caller is
2324 // responsible for yielding.
2325 set_has_aborted();
2326 return;
2327 }
2328
2329 // (5) We check whether we've reached our time quota. If we have,
2330 // then we abort.
2331 double elapsed_time_ms = curr_time_ms - _start_time_ms;
2332 if (elapsed_time_ms > _time_target_ms) {
2333 set_has_aborted();
2334 _has_timed_out = true;
2335 return;
2336 }
2337
2338 // (6) Finally, we check whether there are enough completed STAB
2339 // buffers available for processing. If there are, we abort.
2340 SATBMarkQueueSet& satb_mq_set = JavaThread::satb_mark_queue_set();
2341 if (!_draining_satb_buffers && satb_mq_set.process_completed_buffers()) {
2342 // we do need to process SATB buffers, we'll abort and restart
2343 // the marking task to do so
2344 set_has_aborted();
2345 return;
2346 }
2347 }
2348
2349 void G1CMTask::recalculate_limits() {
2350 _real_words_scanned_limit = _words_scanned + words_scanned_period;
2351 _words_scanned_limit = _real_words_scanned_limit;
2352
2353 _real_refs_reached_limit = _refs_reached + refs_reached_period;
2354 _refs_reached_limit = _real_refs_reached_limit;
2355 }
2356
2357 void G1CMTask::decrease_limits() {
2358 // This is called when we believe that we're going to do an infrequent
2359 // operation which will increase the per byte scanned cost (i.e. move
2360 // entries to/from the global stack). It basically tries to decrease the
2361 // scanning limit so that the clock is called earlier.
2362
2363 _words_scanned_limit = _real_words_scanned_limit -
2364 3 * words_scanned_period / 4;
2365 _refs_reached_limit = _real_refs_reached_limit -
2366 3 * refs_reached_period / 4;
2367 }
2368
2369 void G1CMTask::move_entries_to_global_stack() {
2370 // Local array where we'll store the entries that will be popped
2371 // from the local queue.
2372 G1TaskQueueEntry buffer[G1CMMarkStack::EntriesPerChunk];
2373
2374 size_t n = 0;
2375 G1TaskQueueEntry task_entry;
2376 while (n < G1CMMarkStack::EntriesPerChunk && _task_queue->pop_local(task_entry)) {
2377 buffer[n] = task_entry;
2378 ++n;
2379 }
2380 if (n < G1CMMarkStack::EntriesPerChunk) {
2381 buffer[n] = G1TaskQueueEntry();
2382 }
2383
2384 if (n > 0) {
2385 if (!_cm->mark_stack_push(buffer)) {
2386 set_has_aborted();
2387 }
2388 }
2389
2390 // This operation was quite expensive, so decrease the limits.
2391 decrease_limits();
2392 }
2393
2394 bool G1CMTask::get_entries_from_global_stack() {
2395 // Local array where we'll store the entries that will be popped
2396 // from the global stack.
2397 G1TaskQueueEntry buffer[G1CMMarkStack::EntriesPerChunk];
2398
2399 if (!_cm->mark_stack_pop(buffer)) {
2400 return false;
2401 }
2402
2403 // We did actually pop at least one entry.
2404 for (size_t i = 0; i < G1CMMarkStack::EntriesPerChunk; ++i) {
2405 G1TaskQueueEntry task_entry = buffer[i];
2406 if (task_entry.is_null()) {
2407 break;
2408 }
2409 assert(task_entry.is_array_slice() || task_entry.obj()->is_oop(), "Element " PTR_FORMAT " must be an array slice or oop", p2i(task_entry.obj()));
2410 bool success = _task_queue->push(task_entry);
2411 // We only call this when the local queue is empty or under a
2412 // given target limit. So, we do not expect this push to fail.
2413 assert(success, "invariant");
2414 }
2415
2416 // This operation was quite expensive, so decrease the limits
2417 decrease_limits();
2418 return true;
2419 }
2420
2421 void G1CMTask::drain_local_queue(bool partially) {
2422 if (has_aborted()) {
2423 return;
2424 }
2425
2426 // Decide what the target size is, depending whether we're going to
2427 // drain it partially (so that other tasks can steal if they run out
2428 // of things to do) or totally (at the very end).
2429 size_t target_size;
2430 if (partially) {
2431 target_size = MIN2((size_t)_task_queue->max_elems()/3, GCDrainStackTargetSize);
2432 } else {
2433 target_size = 0;
2434 }
2435
2436 if (_task_queue->size() > target_size) {
2437 G1TaskQueueEntry entry;
2438 bool ret = _task_queue->pop_local(entry);
2439 while (ret) {
2440 scan_task_entry(entry);
2441 if (_task_queue->size() <= target_size || has_aborted()) {
2442 ret = false;
2443 } else {
2444 ret = _task_queue->pop_local(entry);
2445 }
2446 }
2447 }
2448 }
2449
2450 void G1CMTask::drain_global_stack(bool partially) {
2451 if (has_aborted()) return;
2452
2453 // We have a policy to drain the local queue before we attempt to
2454 // drain the global stack.
2455 assert(partially || _task_queue->size() == 0, "invariant");
2456
2457 // Decide what the target size is, depending whether we're going to
2458 // drain it partially (so that other tasks can steal if they run out
2459 // of things to do) or totally (at the very end).
2460 // Notice that when draining the global mark stack partially, due to the racyness
2461 // of the mark stack size update we might in fact drop below the target. But,
2462 // this is not a problem.
2463 // In case of total draining, we simply process until the global mark stack is
2464 // totally empty, disregarding the size counter.
2465 if (partially) {
2466 size_t const target_size = _cm->partial_mark_stack_size_target();
2467 while (!has_aborted() && _cm->mark_stack_size() > target_size) {
2468 if (get_entries_from_global_stack()) {
2469 drain_local_queue(partially);
2470 }
2471 }
2472 } else {
2473 while (!has_aborted() && get_entries_from_global_stack()) {
2474 drain_local_queue(partially);
2475 }
2476 }
2477 }
2478
2479 // SATB Queue has several assumptions on whether to call the par or
2480 // non-par versions of the methods. this is why some of the code is
2481 // replicated. We should really get rid of the single-threaded version
2482 // of the code to simplify things.
2483 void G1CMTask::drain_satb_buffers() {
2484 if (has_aborted()) return;
2485
2486 // We set this so that the regular clock knows that we're in the
2487 // middle of draining buffers and doesn't set the abort flag when it
2488 // notices that SATB buffers are available for draining. It'd be
2489 // very counter productive if it did that. :-)
2490 _draining_satb_buffers = true;
2491
2492 G1CMSATBBufferClosure satb_cl(this, _g1h);
2493 SATBMarkQueueSet& satb_mq_set = JavaThread::satb_mark_queue_set();
2494
2495 // This keeps claiming and applying the closure to completed buffers
2496 // until we run out of buffers or we need to abort.
2497 while (!has_aborted() &&
2498 satb_mq_set.apply_closure_to_completed_buffer(&satb_cl)) {
2499 regular_clock_call();
2500 }
2501
2502 _draining_satb_buffers = false;
2503
2504 assert(has_aborted() ||
2505 concurrent() ||
2506 satb_mq_set.completed_buffers_num() == 0, "invariant");
2507
2508 // again, this was a potentially expensive operation, decrease the
2509 // limits to get the regular clock call early
2510 decrease_limits();
2511 }
2512
2513 void G1CMTask::print_stats() {
2514 log_debug(gc, stats)("Marking Stats, task = %u, calls = %d",
2515 _worker_id, _calls);
2516 log_debug(gc, stats)(" Elapsed time = %1.2lfms, Termination time = %1.2lfms",
2517 _elapsed_time_ms, _termination_time_ms);
2518 log_debug(gc, stats)(" Step Times (cum): num = %d, avg = %1.2lfms, sd = %1.2lfms",
2519 _step_times_ms.num(), _step_times_ms.avg(),
2520 _step_times_ms.sd());
2521 log_debug(gc, stats)(" max = %1.2lfms, total = %1.2lfms",
2522 _step_times_ms.maximum(), _step_times_ms.sum());
2523 }
2524
2525 bool G1ConcurrentMark::try_stealing(uint worker_id, int* hash_seed, G1TaskQueueEntry& task_entry) {
2526 return _task_queues->steal(worker_id, hash_seed, task_entry);
2527 }
2528
2529 /*****************************************************************************
2530
2531 The do_marking_step(time_target_ms, ...) method is the building
2532 block of the parallel marking framework. It can be called in parallel
2533 with other invocations of do_marking_step() on different tasks
2534 (but only one per task, obviously) and concurrently with the
2535 mutator threads, or during remark, hence it eliminates the need
2536 for two versions of the code. When called during remark, it will
2537 pick up from where the task left off during the concurrent marking
2538 phase. Interestingly, tasks are also claimable during evacuation
2539 pauses too, since do_marking_step() ensures that it aborts before
2540 it needs to yield.
2541
2542 The data structures that it uses to do marking work are the
2543 following:
2544
2545 (1) Marking Bitmap. If there are gray objects that appear only
2546 on the bitmap (this happens either when dealing with an overflow
2547 or when the initial marking phase has simply marked the roots
2548 and didn't push them on the stack), then tasks claim heap
2549 regions whose bitmap they then scan to find gray objects. A
2550 global finger indicates where the end of the last claimed region
2551 is. A local finger indicates how far into the region a task has
2552 scanned. The two fingers are used to determine how to gray an
2553 object (i.e. whether simply marking it is OK, as it will be
2554 visited by a task in the future, or whether it needs to be also
2555 pushed on a stack).
2556
2557 (2) Local Queue. The local queue of the task which is accessed
2558 reasonably efficiently by the task. Other tasks can steal from
2559 it when they run out of work. Throughout the marking phase, a
2560 task attempts to keep its local queue short but not totally
2561 empty, so that entries are available for stealing by other
2562 tasks. Only when there is no more work, a task will totally
2563 drain its local queue.
2564
2565 (3) Global Mark Stack. This handles local queue overflow. During
2566 marking only sets of entries are moved between it and the local
2567 queues, as access to it requires a mutex and more fine-grain
2568 interaction with it which might cause contention. If it
2569 overflows, then the marking phase should restart and iterate
2570 over the bitmap to identify gray objects. Throughout the marking
2571 phase, tasks attempt to keep the global mark stack at a small
2572 length but not totally empty, so that entries are available for
2573 popping by other tasks. Only when there is no more work, tasks
2574 will totally drain the global mark stack.
2575
2576 (4) SATB Buffer Queue. This is where completed SATB buffers are
2577 made available. Buffers are regularly removed from this queue
2578 and scanned for roots, so that the queue doesn't get too
2579 long. During remark, all completed buffers are processed, as
2580 well as the filled in parts of any uncompleted buffers.
2581
2582 The do_marking_step() method tries to abort when the time target
2583 has been reached. There are a few other cases when the
2584 do_marking_step() method also aborts:
2585
2586 (1) When the marking phase has been aborted (after a Full GC).
2587
2588 (2) When a global overflow (on the global stack) has been
2589 triggered. Before the task aborts, it will actually sync up with
2590 the other tasks to ensure that all the marking data structures
2591 (local queues, stacks, fingers etc.) are re-initialized so that
2592 when do_marking_step() completes, the marking phase can
2593 immediately restart.
2594
2595 (3) When enough completed SATB buffers are available. The
2596 do_marking_step() method only tries to drain SATB buffers right
2597 at the beginning. So, if enough buffers are available, the
2598 marking step aborts and the SATB buffers are processed at
2599 the beginning of the next invocation.
2600
2601 (4) To yield. when we have to yield then we abort and yield
2602 right at the end of do_marking_step(). This saves us from a lot
2603 of hassle as, by yielding we might allow a Full GC. If this
2604 happens then objects will be compacted underneath our feet, the
2605 heap might shrink, etc. We save checking for this by just
2606 aborting and doing the yield right at the end.
2607
2608 From the above it follows that the do_marking_step() method should
2609 be called in a loop (or, otherwise, regularly) until it completes.
2610
2611 If a marking step completes without its has_aborted() flag being
2612 true, it means it has completed the current marking phase (and
2613 also all other marking tasks have done so and have all synced up).
2614
2615 A method called regular_clock_call() is invoked "regularly" (in
2616 sub ms intervals) throughout marking. It is this clock method that
2617 checks all the abort conditions which were mentioned above and
2618 decides when the task should abort. A work-based scheme is used to
2619 trigger this clock method: when the number of object words the
2620 marking phase has scanned or the number of references the marking
2621 phase has visited reach a given limit. Additional invocations to
2622 the method clock have been planted in a few other strategic places
2623 too. The initial reason for the clock method was to avoid calling
2624 vtime too regularly, as it is quite expensive. So, once it was in
2625 place, it was natural to piggy-back all the other conditions on it
2626 too and not constantly check them throughout the code.
2627
2628 If do_termination is true then do_marking_step will enter its
2629 termination protocol.
2630
2631 The value of is_serial must be true when do_marking_step is being
2632 called serially (i.e. by the VMThread) and do_marking_step should
2633 skip any synchronization in the termination and overflow code.
2634 Examples include the serial remark code and the serial reference
2635 processing closures.
2636
2637 The value of is_serial must be false when do_marking_step is
2638 being called by any of the worker threads in a work gang.
2639 Examples include the concurrent marking code (CMMarkingTask),
2640 the MT remark code, and the MT reference processing closures.
2641
2642 *****************************************************************************/
2643
2644 void G1CMTask::do_marking_step(double time_target_ms,
2645 bool do_termination,
2646 bool is_serial) {
2647 assert(time_target_ms >= 1.0, "minimum granularity is 1ms");
2648 assert(concurrent() == _cm->concurrent(), "they should be the same");
2649
2650 G1Policy* g1_policy = _g1h->g1_policy();
2651 assert(_task_queues != NULL, "invariant");
2652 assert(_task_queue != NULL, "invariant");
2653 assert(_task_queues->queue(_worker_id) == _task_queue, "invariant");
2654
2655 assert(!_claimed,
2656 "only one thread should claim this task at any one time");
2657
2658 // OK, this doesn't safeguard again all possible scenarios, as it is
2659 // possible for two threads to set the _claimed flag at the same
2660 // time. But it is only for debugging purposes anyway and it will
2661 // catch most problems.
2662 _claimed = true;
2663
2664 _start_time_ms = os::elapsedVTime() * 1000.0;
2665
2666 // If do_stealing is true then do_marking_step will attempt to
2667 // steal work from the other G1CMTasks. It only makes sense to
2668 // enable stealing when the termination protocol is enabled
2669 // and do_marking_step() is not being called serially.
2670 bool do_stealing = do_termination && !is_serial;
2671
2672 double diff_prediction_ms = _g1h->g1_policy()->predictor().get_new_prediction(&_marking_step_diffs_ms);
2673 _time_target_ms = time_target_ms - diff_prediction_ms;
2674
2675 // set up the variables that are used in the work-based scheme to
2676 // call the regular clock method
2677 _words_scanned = 0;
2678 _refs_reached = 0;
2679 recalculate_limits();
2680
2681 // clear all flags
2682 clear_has_aborted();
2683 _has_timed_out = false;
2684 _draining_satb_buffers = false;
2685
2686 ++_calls;
2687
2688 // Set up the bitmap and oop closures. Anything that uses them is
2689 // eventually called from this method, so it is OK to allocate these
2690 // statically.
2691 G1CMBitMapClosure bitmap_closure(this, _cm, _nextMarkBitMap);
2692 G1CMOopClosure cm_oop_closure(_g1h, _cm, this);
2693 set_cm_oop_closure(&cm_oop_closure);
2694
2695 if (_cm->has_overflown()) {
2696 // This can happen if the mark stack overflows during a GC pause
2697 // and this task, after a yield point, restarts. We have to abort
2698 // as we need to get into the overflow protocol which happens
2699 // right at the end of this task.
2700 set_has_aborted();
2701 }
2702
2703 // First drain any available SATB buffers. After this, we will not
2704 // look at SATB buffers before the next invocation of this method.
2705 // If enough completed SATB buffers are queued up, the regular clock
2706 // will abort this task so that it restarts.
2707 drain_satb_buffers();
2708 // ...then partially drain the local queue and the global stack
2709 drain_local_queue(true);
2710 drain_global_stack(true);
2711
2712 do {
2713 if (!has_aborted() && _curr_region != NULL) {
2714 // This means that we're already holding on to a region.
2715 assert(_finger != NULL, "if region is not NULL, then the finger "
2716 "should not be NULL either");
2717
2718 // We might have restarted this task after an evacuation pause
2719 // which might have evacuated the region we're holding on to
2720 // underneath our feet. Let's read its limit again to make sure
2721 // that we do not iterate over a region of the heap that
2722 // contains garbage (update_region_limit() will also move
2723 // _finger to the start of the region if it is found empty).
2724 update_region_limit();
2725 // We will start from _finger not from the start of the region,
2726 // as we might be restarting this task after aborting half-way
2727 // through scanning this region. In this case, _finger points to
2728 // the address where we last found a marked object. If this is a
2729 // fresh region, _finger points to start().
2730 MemRegion mr = MemRegion(_finger, _region_limit);
2731
2732 assert(!_curr_region->is_humongous() || mr.start() == _curr_region->bottom(),
2733 "humongous regions should go around loop once only");
2734
2735 // Some special cases:
2736 // If the memory region is empty, we can just give up the region.
2737 // If the current region is humongous then we only need to check
2738 // the bitmap for the bit associated with the start of the object,
2739 // scan the object if it's live, and give up the region.
2740 // Otherwise, let's iterate over the bitmap of the part of the region
2741 // that is left.
2742 // If the iteration is successful, give up the region.
2743 if (mr.is_empty()) {
2744 giveup_current_region();
2745 regular_clock_call();
2746 } else if (_curr_region->is_humongous() && mr.start() == _curr_region->bottom()) {
2747 if (_nextMarkBitMap->isMarked(mr.start())) {
2748 // The object is marked - apply the closure
2749 BitMap::idx_t offset = _nextMarkBitMap->heapWordToOffset(mr.start());
2750 bitmap_closure.do_bit(offset);
2751 }
2752 // Even if this task aborted while scanning the humongous object
2753 // we can (and should) give up the current region.
2754 giveup_current_region();
2755 regular_clock_call();
2756 } else if (_nextMarkBitMap->iterate(&bitmap_closure, mr)) {
2757 giveup_current_region();
2758 regular_clock_call();
2759 } else {
2760 assert(has_aborted(), "currently the only way to do so");
2761 // The only way to abort the bitmap iteration is to return
2762 // false from the do_bit() method. However, inside the
2763 // do_bit() method we move the _finger to point to the
2764 // object currently being looked at. So, if we bail out, we
2765 // have definitely set _finger to something non-null.
2766 assert(_finger != NULL, "invariant");
2767
2768 // Region iteration was actually aborted. So now _finger
2769 // points to the address of the object we last scanned. If we
2770 // leave it there, when we restart this task, we will rescan
2771 // the object. It is easy to avoid this. We move the finger by
2772 // enough to point to the next possible object header (the
2773 // bitmap knows by how much we need to move it as it knows its
2774 // granularity).
2775 assert(_finger < _region_limit, "invariant");
2776 HeapWord* new_finger = _nextMarkBitMap->nextObject(_finger);
2777 // Check if bitmap iteration was aborted while scanning the last object
2778 if (new_finger >= _region_limit) {
2779 giveup_current_region();
2780 } else {
2781 move_finger_to(new_finger);
2782 }
2783 }
2784 }
2785 // At this point we have either completed iterating over the
2786 // region we were holding on to, or we have aborted.
2787
2788 // We then partially drain the local queue and the global stack.
2789 // (Do we really need this?)
2790 drain_local_queue(true);
2791 drain_global_stack(true);
2792
2793 // Read the note on the claim_region() method on why it might
2794 // return NULL with potentially more regions available for
2795 // claiming and why we have to check out_of_regions() to determine
2796 // whether we're done or not.
2797 while (!has_aborted() && _curr_region == NULL && !_cm->out_of_regions()) {
2798 // We are going to try to claim a new region. We should have
2799 // given up on the previous one.
2800 // Separated the asserts so that we know which one fires.
2801 assert(_curr_region == NULL, "invariant");
2802 assert(_finger == NULL, "invariant");
2803 assert(_region_limit == NULL, "invariant");
2804 HeapRegion* claimed_region = _cm->claim_region(_worker_id);
2805 if (claimed_region != NULL) {
2806 // Yes, we managed to claim one
2807 setup_for_region(claimed_region);
2808 assert(_curr_region == claimed_region, "invariant");
2809 }
2810 // It is important to call the regular clock here. It might take
2811 // a while to claim a region if, for example, we hit a large
2812 // block of empty regions. So we need to call the regular clock
2813 // method once round the loop to make sure it's called
2814 // frequently enough.
2815 regular_clock_call();
2816 }
2817
2818 if (!has_aborted() && _curr_region == NULL) {
2819 assert(_cm->out_of_regions(),
2820 "at this point we should be out of regions");
2821 }
2822 } while ( _curr_region != NULL && !has_aborted());
2823
2824 if (!has_aborted()) {
2825 // We cannot check whether the global stack is empty, since other
2826 // tasks might be pushing objects to it concurrently.
2827 assert(_cm->out_of_regions(),
2828 "at this point we should be out of regions");
2829 // Try to reduce the number of available SATB buffers so that
2830 // remark has less work to do.
2831 drain_satb_buffers();
2832 }
2833
2834 // Since we've done everything else, we can now totally drain the
2835 // local queue and global stack.
2836 drain_local_queue(false);
2837 drain_global_stack(false);
2838
2839 // Attempt at work stealing from other task's queues.
2840 if (do_stealing && !has_aborted()) {
2841 // We have not aborted. This means that we have finished all that
2842 // we could. Let's try to do some stealing...
2843
2844 // We cannot check whether the global stack is empty, since other
2845 // tasks might be pushing objects to it concurrently.
2846 assert(_cm->out_of_regions() && _task_queue->size() == 0,
2847 "only way to reach here");
2848 while (!has_aborted()) {
2849 G1TaskQueueEntry entry;
2850 if (_cm->try_stealing(_worker_id, &_hash_seed, entry)) {
2851 scan_task_entry(entry);
2852
2853 // And since we're towards the end, let's totally drain the
2854 // local queue and global stack.
2855 drain_local_queue(false);
2856 drain_global_stack(false);
2857 } else {
2858 break;
2859 }
2860 }
2861 }
2862
2863 // We still haven't aborted. Now, let's try to get into the
2864 // termination protocol.
2865 if (do_termination && !has_aborted()) {
2866 // We cannot check whether the global stack is empty, since other
2867 // tasks might be concurrently pushing objects on it.
2868 // Separated the asserts so that we know which one fires.
2869 assert(_cm->out_of_regions(), "only way to reach here");
2870 assert(_task_queue->size() == 0, "only way to reach here");
2871 _termination_start_time_ms = os::elapsedVTime() * 1000.0;
2872
2873 // The G1CMTask class also extends the TerminatorTerminator class,
2874 // hence its should_exit_termination() method will also decide
2875 // whether to exit the termination protocol or not.
2876 bool finished = (is_serial ||
2877 _cm->terminator()->offer_termination(this));
2878 double termination_end_time_ms = os::elapsedVTime() * 1000.0;
2879 _termination_time_ms +=
2880 termination_end_time_ms - _termination_start_time_ms;
2881
2882 if (finished) {
2883 // We're all done.
2884
2885 if (_worker_id == 0) {
2886 // let's allow task 0 to do this
2887 if (concurrent()) {
2888 assert(_cm->concurrent_marking_in_progress(), "invariant");
2889 // we need to set this to false before the next
2890 // safepoint. This way we ensure that the marking phase
2891 // doesn't observe any more heap expansions.
2892 _cm->clear_concurrent_marking_in_progress();
2893 }
2894 }
2895
2896 // We can now guarantee that the global stack is empty, since
2897 // all other tasks have finished. We separated the guarantees so
2898 // that, if a condition is false, we can immediately find out
2899 // which one.
2900 guarantee(_cm->out_of_regions(), "only way to reach here");
2901 guarantee(_cm->mark_stack_empty(), "only way to reach here");
2902 guarantee(_task_queue->size() == 0, "only way to reach here");
2903 guarantee(!_cm->has_overflown(), "only way to reach here");
2904 } else {
2905 // Apparently there's more work to do. Let's abort this task. It
2906 // will restart it and we can hopefully find more things to do.
2907 set_has_aborted();
2908 }
2909 }
2910
2911 // Mainly for debugging purposes to make sure that a pointer to the
2912 // closure which was statically allocated in this frame doesn't
2913 // escape it by accident.
2914 set_cm_oop_closure(NULL);
2915 double end_time_ms = os::elapsedVTime() * 1000.0;
2916 double elapsed_time_ms = end_time_ms - _start_time_ms;
2917 // Update the step history.
2918 _step_times_ms.add(elapsed_time_ms);
2919
2920 if (has_aborted()) {
2921 // The task was aborted for some reason.
2922 if (_has_timed_out) {
2923 double diff_ms = elapsed_time_ms - _time_target_ms;
2924 // Keep statistics of how well we did with respect to hitting
2925 // our target only if we actually timed out (if we aborted for
2926 // other reasons, then the results might get skewed).
2927 _marking_step_diffs_ms.add(diff_ms);
2928 }
2929
2930 if (_cm->has_overflown()) {
2931 // This is the interesting one. We aborted because a global
2932 // overflow was raised. This means we have to restart the
2933 // marking phase and start iterating over regions. However, in
2934 // order to do this we have to make sure that all tasks stop
2935 // what they are doing and re-initialize in a safe manner. We
2936 // will achieve this with the use of two barrier sync points.
2937
2938 if (!is_serial) {
2939 // We only need to enter the sync barrier if being called
2940 // from a parallel context
2941 _cm->enter_first_sync_barrier(_worker_id);
2942
2943 // When we exit this sync barrier we know that all tasks have
2944 // stopped doing marking work. So, it's now safe to
2945 // re-initialize our data structures. At the end of this method,
2946 // task 0 will clear the global data structures.
2947 }
2948
2949 // We clear the local state of this task...
2950 clear_region_fields();
2951
2952 if (!is_serial) {
2953 // ...and enter the second barrier.
2954 _cm->enter_second_sync_barrier(_worker_id);
2955 }
2956 // At this point, if we're during the concurrent phase of
2957 // marking, everything has been re-initialized and we're
2958 // ready to restart.
2959 }
2960 }
2961
2962 _claimed = false;
2963 }
2964
2965 G1CMTask::G1CMTask(uint worker_id,
2966 G1ConcurrentMark* cm,
2967 G1CMTaskQueue* task_queue,
2968 G1CMTaskQueueSet* task_queues)
2969 : _g1h(G1CollectedHeap::heap()),
2970 _worker_id(worker_id), _cm(cm),
2971 _objArray_processor(this),
2972 _claimed(false),
2973 _nextMarkBitMap(NULL), _hash_seed(17),
2974 _task_queue(task_queue),
2975 _task_queues(task_queues),
2976 _cm_oop_closure(NULL) {
2977 guarantee(task_queue != NULL, "invariant");
2978 guarantee(task_queues != NULL, "invariant");
2979
2980 _marking_step_diffs_ms.add(0.5);
2981 }
2982
2983 // These are formatting macros that are used below to ensure
2984 // consistent formatting. The *_H_* versions are used to format the
2985 // header for a particular value and they should be kept consistent
2986 // with the corresponding macro. Also note that most of the macros add
2987 // the necessary white space (as a prefix) which makes them a bit
2988 // easier to compose.
2989
2990 // All the output lines are prefixed with this string to be able to
2991 // identify them easily in a large log file.
2992 #define G1PPRL_LINE_PREFIX "###"
2993
2994 #define G1PPRL_ADDR_BASE_FORMAT " " PTR_FORMAT "-" PTR_FORMAT
2995 #ifdef _LP64
2996 #define G1PPRL_ADDR_BASE_H_FORMAT " %37s"
2997 #else // _LP64
2998 #define G1PPRL_ADDR_BASE_H_FORMAT " %21s"
2999 #endif // _LP64
3000
3001 // For per-region info
3002 #define G1PPRL_TYPE_FORMAT " %-4s"
3003 #define G1PPRL_TYPE_H_FORMAT " %4s"
3004 #define G1PPRL_BYTE_FORMAT " " SIZE_FORMAT_W(9)
3005 #define G1PPRL_BYTE_H_FORMAT " %9s"
3006 #define G1PPRL_DOUBLE_FORMAT " %14.1f"
3007 #define G1PPRL_DOUBLE_H_FORMAT " %14s"
3008
3009 // For summary info
3010 #define G1PPRL_SUM_ADDR_FORMAT(tag) " " tag ":" G1PPRL_ADDR_BASE_FORMAT
3011 #define G1PPRL_SUM_BYTE_FORMAT(tag) " " tag ": " SIZE_FORMAT
3012 #define G1PPRL_SUM_MB_FORMAT(tag) " " tag ": %1.2f MB"
3013 #define G1PPRL_SUM_MB_PERC_FORMAT(tag) G1PPRL_SUM_MB_FORMAT(tag) " / %1.2f %%"
3014
3015 G1PrintRegionLivenessInfoClosure::
3016 G1PrintRegionLivenessInfoClosure(const char* phase_name)
3017 : _total_used_bytes(0), _total_capacity_bytes(0),
3018 _total_prev_live_bytes(0), _total_next_live_bytes(0),
3019 _total_remset_bytes(0), _total_strong_code_roots_bytes(0) {
3020 G1CollectedHeap* g1h = G1CollectedHeap::heap();
3021 MemRegion g1_reserved = g1h->g1_reserved();
3022 double now = os::elapsedTime();
3023
3024 // Print the header of the output.
3025 log_trace(gc, liveness)(G1PPRL_LINE_PREFIX" PHASE %s @ %1.3f", phase_name, now);
3026 log_trace(gc, liveness)(G1PPRL_LINE_PREFIX" HEAP"
3027 G1PPRL_SUM_ADDR_FORMAT("reserved")
3028 G1PPRL_SUM_BYTE_FORMAT("region-size"),
3029 p2i(g1_reserved.start()), p2i(g1_reserved.end()),
3030 HeapRegion::GrainBytes);
3031 log_trace(gc, liveness)(G1PPRL_LINE_PREFIX);
3032 log_trace(gc, liveness)(G1PPRL_LINE_PREFIX
3033 G1PPRL_TYPE_H_FORMAT
3034 G1PPRL_ADDR_BASE_H_FORMAT
3035 G1PPRL_BYTE_H_FORMAT
3036 G1PPRL_BYTE_H_FORMAT
3037 G1PPRL_BYTE_H_FORMAT
3038 G1PPRL_DOUBLE_H_FORMAT
3039 G1PPRL_BYTE_H_FORMAT
3040 G1PPRL_BYTE_H_FORMAT,
3041 "type", "address-range",
3042 "used", "prev-live", "next-live", "gc-eff",
3043 "remset", "code-roots");
3044 log_trace(gc, liveness)(G1PPRL_LINE_PREFIX
3045 G1PPRL_TYPE_H_FORMAT
3046 G1PPRL_ADDR_BASE_H_FORMAT
3047 G1PPRL_BYTE_H_FORMAT
3048 G1PPRL_BYTE_H_FORMAT
3049 G1PPRL_BYTE_H_FORMAT
3050 G1PPRL_DOUBLE_H_FORMAT
3051 G1PPRL_BYTE_H_FORMAT
3052 G1PPRL_BYTE_H_FORMAT,
3053 "", "",
3054 "(bytes)", "(bytes)", "(bytes)", "(bytes/ms)",
3055 "(bytes)", "(bytes)");
3056 }
3057
3058 bool G1PrintRegionLivenessInfoClosure::doHeapRegion(HeapRegion* r) {
3059 const char* type = r->get_type_str();
3060 HeapWord* bottom = r->bottom();
3061 HeapWord* end = r->end();
3062 size_t capacity_bytes = r->capacity();
3063 size_t used_bytes = r->used();
3064 size_t prev_live_bytes = r->live_bytes();
3065 size_t next_live_bytes = r->next_live_bytes();
3066 double gc_eff = r->gc_efficiency();
3067 size_t remset_bytes = r->rem_set()->mem_size();
3068 size_t strong_code_roots_bytes = r->rem_set()->strong_code_roots_mem_size();
3069
3070 _total_used_bytes += used_bytes;
3071 _total_capacity_bytes += capacity_bytes;
3072 _total_prev_live_bytes += prev_live_bytes;
3073 _total_next_live_bytes += next_live_bytes;
3074 _total_remset_bytes += remset_bytes;
3075 _total_strong_code_roots_bytes += strong_code_roots_bytes;
3076
3077 // Print a line for this particular region.
3078 log_trace(gc, liveness)(G1PPRL_LINE_PREFIX
3079 G1PPRL_TYPE_FORMAT
3080 G1PPRL_ADDR_BASE_FORMAT
3081 G1PPRL_BYTE_FORMAT
3082 G1PPRL_BYTE_FORMAT
3083 G1PPRL_BYTE_FORMAT
3084 G1PPRL_DOUBLE_FORMAT
3085 G1PPRL_BYTE_FORMAT
3086 G1PPRL_BYTE_FORMAT,
3087 type, p2i(bottom), p2i(end),
3088 used_bytes, prev_live_bytes, next_live_bytes, gc_eff,
3089 remset_bytes, strong_code_roots_bytes);
3090
3091 return false;
3092 }
3093
3094 G1PrintRegionLivenessInfoClosure::~G1PrintRegionLivenessInfoClosure() {
3095 // add static memory usages to remembered set sizes
3096 _total_remset_bytes += HeapRegionRemSet::fl_mem_size() + HeapRegionRemSet::static_mem_size();
3097 // Print the footer of the output.
3098 log_trace(gc, liveness)(G1PPRL_LINE_PREFIX);
3099 log_trace(gc, liveness)(G1PPRL_LINE_PREFIX
3100 " SUMMARY"
3101 G1PPRL_SUM_MB_FORMAT("capacity")
3102 G1PPRL_SUM_MB_PERC_FORMAT("used")
3103 G1PPRL_SUM_MB_PERC_FORMAT("prev-live")
3104 G1PPRL_SUM_MB_PERC_FORMAT("next-live")
3105 G1PPRL_SUM_MB_FORMAT("remset")
3106 G1PPRL_SUM_MB_FORMAT("code-roots"),
3107 bytes_to_mb(_total_capacity_bytes),
3108 bytes_to_mb(_total_used_bytes),
3109 perc(_total_used_bytes, _total_capacity_bytes),
3110 bytes_to_mb(_total_prev_live_bytes),
3111 perc(_total_prev_live_bytes, _total_capacity_bytes),
3112 bytes_to_mb(_total_next_live_bytes),
3113 perc(_total_next_live_bytes, _total_capacity_bytes),
3114 bytes_to_mb(_total_remset_bytes),
3115 bytes_to_mb(_total_strong_code_roots_bytes));
3116 }