Mercurial > hg > graal-compiler
view src/os/solaris/vm/osThread_solaris.cpp @ 6812:988bf00cc564
7200261: G1: Liveness counting inconsistencies during marking verification
Summary: The clipping code in the routine that sets the bits for a range of cards, in the liveness accounting verification code was incorrect. It set all the bits in the card bitmap from the given starting index which would lead to spurious marking verification failures.
Reviewed-by: brutisso, jwilhelm, jmasa
author | johnc |
---|---|
date | Thu, 27 Sep 2012 15:44:01 -0700 |
parents | f95d63e2154a |
children | cd3d6a6b95d9 |
line wrap: on
line source
/* * Copyright (c) 1998, 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. * */ // no precompiled headers #include "runtime/atomic.hpp" #include "runtime/handles.inline.hpp" #include "runtime/mutexLocker.hpp" #include "runtime/os.hpp" #include "runtime/osThread.hpp" #include "runtime/safepoint.hpp" #include "runtime/vmThread.hpp" #ifdef TARGET_ARCH_x86 # include "assembler_x86.inline.hpp" #endif #ifdef TARGET_ARCH_sparc # include "assembler_sparc.inline.hpp" #endif # include <signal.h> // *************************************************************** // Platform dependent initialization and cleanup // *************************************************************** void OSThread::pd_initialize() { _thread_id = 0; sigemptyset(&_caller_sigmask); _current_callback = NULL; _current_callback_lock = VM_Version::supports_compare_and_exchange() ? NULL : new Mutex(Mutex::suspend_resume, "Callback_lock", true); _saved_interrupt_thread_state = _thread_new; _vm_created_thread = false; } void OSThread::pd_destroy() { } // Synchronous interrupt support // // _current_callback == NULL no pending callback // == 1 callback_in_progress // == other value pointer to the pending callback // // CAS on v8 is implemented by using a global atomic_memory_operation_lock, // which is shared by other atomic functions. It is OK for normal uses, but // dangerous if used after some thread is suspended or if used in signal // handlers. Instead here we use a special per-thread lock to synchronize // updating _current_callback if we are running on v8. Note in general trying // to grab locks after a thread is suspended is not safe, but it is safe for // updating _current_callback, because synchronous interrupt callbacks are // currently only used in: // 1. GetThreadPC_Callback - used by WatcherThread to profile VM thread // There is no overlap between the callbacks, which means we won't try to // grab a thread's sync lock after the thread has been suspended while holding // the same lock. // used after a thread is suspended static intptr_t compare_and_exchange_current_callback ( intptr_t callback, intptr_t *addr, intptr_t compare_value, Mutex *sync) { if (VM_Version::supports_compare_and_exchange()) { return Atomic::cmpxchg_ptr(callback, addr, compare_value); } else { MutexLockerEx ml(sync, Mutex::_no_safepoint_check_flag); if (*addr == compare_value) { *addr = callback; return compare_value; } else { return callback; } } } // used in signal handler static intptr_t exchange_current_callback(intptr_t callback, intptr_t *addr, Mutex *sync) { if (VM_Version::supports_compare_and_exchange()) { return Atomic::xchg_ptr(callback, addr); } else { MutexLockerEx ml(sync, Mutex::_no_safepoint_check_flag); intptr_t cb = *addr; *addr = callback; return cb; } } // one interrupt at a time. spin if _current_callback != NULL int OSThread::set_interrupt_callback(Sync_Interrupt_Callback * cb) { int count = 0; while (compare_and_exchange_current_callback( (intptr_t)cb, (intptr_t *)&_current_callback, (intptr_t)NULL, _current_callback_lock) != NULL) { while (_current_callback != NULL) { count++; #ifdef ASSERT if ((WarnOnStalledSpinLock > 0) && (count % WarnOnStalledSpinLock == 0)) { warning("_current_callback seems to be stalled: %p", _current_callback); } #endif os::yield_all(count); } } return 0; } // reset _current_callback, spin if _current_callback is callback_in_progress void OSThread::remove_interrupt_callback(Sync_Interrupt_Callback * cb) { int count = 0; while (compare_and_exchange_current_callback( (intptr_t)NULL, (intptr_t *)&_current_callback, (intptr_t)cb, _current_callback_lock) != (intptr_t)cb) { #ifdef ASSERT intptr_t p = (intptr_t)_current_callback; assert(p == (intptr_t)callback_in_progress || p == (intptr_t)cb, "wrong _current_callback value"); #endif while (_current_callback != cb) { count++; #ifdef ASSERT if ((WarnOnStalledSpinLock > 0) && (count % WarnOnStalledSpinLock == 0)) { warning("_current_callback seems to be stalled: %p", _current_callback); } #endif os::yield_all(count); } } } void OSThread::do_interrupt_callbacks_at_interrupt(InterruptArguments *args) { Sync_Interrupt_Callback * cb; cb = (Sync_Interrupt_Callback *)exchange_current_callback( (intptr_t)callback_in_progress, (intptr_t *)&_current_callback, _current_callback_lock); if (cb == NULL) { // signal is delivered too late (thread is masking interrupt signal??). // there is nothing we need to do because requesting thread has given up. } else if ((intptr_t)cb == (intptr_t)callback_in_progress) { fatal("invalid _current_callback state"); } else { assert(cb->target()->osthread() == this, "wrong target"); cb->execute(args); cb->leave_callback(); // notify the requester } // restore original _current_callback value intptr_t p; p = exchange_current_callback((intptr_t)cb, (intptr_t *)&_current_callback, _current_callback_lock); assert(p == (intptr_t)callback_in_progress, "just checking"); } // Called by the requesting thread to send a signal to target thread and // execute "this" callback from the signal handler. int OSThread::Sync_Interrupt_Callback::interrupt(Thread * target, int timeout) { // Let signals to the vm_thread go even if the Threads_lock is not acquired assert(Threads_lock->owned_by_self() || (target == VMThread::vm_thread()), "must have threads lock to call this"); OSThread * osthread = target->osthread(); // may block if target thread already has a pending callback osthread->set_interrupt_callback(this); _target = target; int rslt = thr_kill(osthread->thread_id(), os::Solaris::SIGasync()); assert(rslt == 0, "thr_kill != 0"); bool status = false; jlong t1 = os::javaTimeMillis(); { // don't use safepoint check because we might be the watcher thread. MutexLockerEx ml(_sync, Mutex::_no_safepoint_check_flag); while (!is_done()) { status = _sync->wait(Mutex::_no_safepoint_check_flag, timeout); // status == true if timed out if (status) break; // update timeout jlong t2 = os::javaTimeMillis(); timeout -= t2 - t1; t1 = t2; } } // reset current_callback osthread->remove_interrupt_callback(this); return status; } void OSThread::Sync_Interrupt_Callback::leave_callback() { if (!_sync->owned_by_self()) { // notify requesting thread MutexLockerEx ml(_sync, Mutex::_no_safepoint_check_flag); _is_done = true; _sync->notify_all(); } else { // Current thread is interrupted while it is holding the _sync lock, trying // to grab it again will deadlock. The requester will timeout anyway, // so just return. _is_done = true; } } // copied from synchronizer.cpp void OSThread::handle_spinlock_contention(int tries) { if (NoYieldsInMicrolock) return; if (tries > 10) { os::yield_all(tries); // Yield to threads of any priority } else if (tries > 5) { os::yield(); // Yield to threads of same or higher priority } }