Mercurial > hg > graal-compiler
view src/share/vm/gc_implementation/parallelScavenge/gcTaskThread.cpp @ 2149:7e37af9d69ef
7011379: G1: overly long concurrent marking cycles
Summary: This changeset introduces filtering of SATB buffers at the point when they are about to be enqueued. If this filtering clears enough entries on each buffer, the buffer can then be re-used and not enqueued. This cuts down the number of SATB buffers that need to be processed by the concurrent marking threads.
Reviewed-by: johnc, ysr
author | tonyp |
---|---|
date | Wed, 19 Jan 2011 09:35:17 -0500 |
parents | f95d63e2154a |
children | bca17e38de00 |
line wrap: on
line source
/* * Copyright (c) 2002, 2010, Oracle and/or its affiliates. All rights reserved. * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER. * * This code is free software; you can redistribute it and/or modify it * under the terms of the GNU General Public License version 2 only, as * published by the Free Software Foundation. * * This code is distributed in the hope that it will be useful, but WITHOUT * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License * version 2 for more details (a copy is included in the LICENSE file that * accompanied this code). * * You should have received a copy of the GNU General Public License version * 2 along with this work; if not, write to the Free Software Foundation, * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. * * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA * or visit www.oracle.com if you need additional information or have any * questions. * */ #include "precompiled.hpp" #include "gc_implementation/parallelScavenge/gcTaskManager.hpp" #include "gc_implementation/parallelScavenge/gcTaskThread.hpp" #include "memory/allocation.hpp" #include "memory/allocation.inline.hpp" #include "memory/resourceArea.hpp" #include "runtime/handles.hpp" #include "runtime/handles.inline.hpp" #include "runtime/os.hpp" #include "runtime/thread.hpp" GCTaskThread::GCTaskThread(GCTaskManager* manager, uint which, uint processor_id) : _manager(manager), _processor_id(processor_id), _time_stamps(NULL), _time_stamp_index(0) { if (!os::create_thread(this, os::pgc_thread)) vm_exit_out_of_memory(0, "Cannot create GC thread. Out of system resources."); if (PrintGCTaskTimeStamps) { _time_stamps = NEW_C_HEAP_ARRAY(GCTaskTimeStamp, GCTaskTimeStampEntries ); guarantee(_time_stamps != NULL, "Sanity"); } set_id(which); set_name("GC task thread#%d (ParallelGC)", which); } GCTaskThread::~GCTaskThread() { if (_time_stamps != NULL) { FREE_C_HEAP_ARRAY(GCTaskTimeStamp, _time_stamps); } } void GCTaskThread::start() { os::start_thread(this); } GCTaskTimeStamp* GCTaskThread::time_stamp_at(uint index) { guarantee(index < GCTaskTimeStampEntries, "increase GCTaskTimeStampEntries"); return &(_time_stamps[index]); } void GCTaskThread::print_task_time_stamps() { assert(PrintGCTaskTimeStamps, "Sanity"); assert(_time_stamps != NULL, "Sanity (Probably set PrintGCTaskTimeStamps late)"); tty->print_cr("GC-Thread %u entries: %d", id(), _time_stamp_index); for(uint i=0; i<_time_stamp_index; i++) { GCTaskTimeStamp* time_stamp = time_stamp_at(i); tty->print_cr("\t[ %s " INT64_FORMAT " " INT64_FORMAT " ]", time_stamp->name(), time_stamp->entry_time(), time_stamp->exit_time()); } // Reset after dumping the data _time_stamp_index = 0; } void GCTaskThread::print_on(outputStream* st) const { st->print("\"%s\" ", name()); Thread::print_on(st); st->cr(); } void GCTaskThread::run() { // Set up the thread for stack overflow support this->record_stack_base_and_size(); this->initialize_thread_local_storage(); // Bind yourself to your processor. if (processor_id() != GCTaskManager::sentinel_worker()) { if (TraceGCTaskThread) { tty->print_cr("GCTaskThread::run: " " binding to processor %u", processor_id()); } if (!os::bind_to_processor(processor_id())) { DEBUG_ONLY( warning("Couldn't bind GCTaskThread %u to processor %u", which(), processor_id()); ) } } // Part of thread setup. // ??? Are these set up once here to make subsequent ones fast? HandleMark hm_outer; ResourceMark rm_outer; TimeStamp timer; for (;/* ever */;) { // These are so we can flush the resources allocated in the inner loop. HandleMark hm_inner; ResourceMark rm_inner; for (; /* break */; ) { // This will block until there is a task to be gotten. GCTask* task = manager()->get_task(which()); // In case the update is costly if (PrintGCTaskTimeStamps) { timer.update(); } jlong entry_time = timer.ticks(); char* name = task->name(); task->do_it(manager(), which()); manager()->note_completion(which()); if (PrintGCTaskTimeStamps) { assert(_time_stamps != NULL, "Sanity (PrintGCTaskTimeStamps set late?)"); timer.update(); GCTaskTimeStamp* time_stamp = time_stamp_at(_time_stamp_index++); time_stamp->set_name(name); time_stamp->set_entry_time(entry_time); time_stamp->set_exit_time(timer.ticks()); } // Check if we should release our inner resources. if (manager()->should_release_resources(which())) { manager()->note_release(which()); break; } } } }