-//$$CDS-header$$
+/*
+ This file is a part of libcds - Concurrent Data Structures library
-#ifndef _CDS_URCU_DETAILS_GP_H
-#define _CDS_URCU_DETAILS_GP_H
+ (C) Copyright Maxim Khizhinsky (libcds.dev@gmail.com) 2006-2017
+
+ Source code repo: http://github.com/khizmax/libcds/
+ Download: http://sourceforge.net/projects/libcds/files/
+
+ Redistribution and use in source and binary forms, with or without
+ modification, are permitted provided that the following conditions are met:
+
+ * Redistributions of source code must retain the above copyright notice, this
+ list of conditions and the following disclaimer.
+
+ * Redistributions in binary form must reproduce the above copyright notice,
+ this list of conditions and the following disclaimer in the documentation
+ and/or other materials provided with the distribution.
+
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
+ FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+ CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+ OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*/
+
+#ifndef CDSLIB_URCU_DETAILS_GP_H
+#define CDSLIB_URCU_DETAILS_GP_H
#include <cds/urcu/details/gp_decl.h>
#include <cds/threading/model.h>
template <typename RCUtag>
inline gp_thread_gc<RCUtag>::gp_thread_gc()
{
- if ( !threading::Manager::isThreadAttached() )
+ if ( !threading::Manager::isThreadAttached())
cds::threading::Manager::attachThread();
}
inline void gp_thread_gc<RCUtag>::access_lock()
{
thread_record * pRec = get_thread_record();
- assert( pRec != null_ptr<thread_record *>());
+ assert( pRec != nullptr );
- uint32_t tmp = pRec->m_nAccessControl.load( CDS_ATOMIC::memory_order_relaxed );
+ uint32_t tmp = pRec->m_nAccessControl.load( atomics::memory_order_relaxed );
if ( (tmp & rcu_class::c_nNestMask) == 0 ) {
- pRec->m_nAccessControl.store( gp_singleton<RCUtag>::instance()->global_control_word(CDS_ATOMIC::memory_order_relaxed),
- CDS_ATOMIC::memory_order_relaxed );
- CDS_ATOMIC::atomic_thread_fence( CDS_ATOMIC::memory_order_acquire );
- //CDS_COMPILER_RW_BARRIER;
+ pRec->m_nAccessControl.store( gp_singleton<RCUtag>::instance()->global_control_word(atomics::memory_order_relaxed),
+ atomics::memory_order_relaxed );
+
+ atomics::atomic_thread_fence( atomics::memory_order_seq_cst );
}
else {
- pRec->m_nAccessControl.fetch_add( 1, CDS_ATOMIC::memory_order_relaxed );
+ // nested lock
+ pRec->m_nAccessControl.store( tmp + 1, atomics::memory_order_relaxed );
}
}
inline void gp_thread_gc<RCUtag>::access_unlock()
{
thread_record * pRec = get_thread_record();
- assert( pRec != null_ptr<thread_record *>());
+ assert( pRec != nullptr );
+
+ uint32_t tmp = pRec->m_nAccessControl.load( atomics::memory_order_relaxed );
+ assert( (tmp & rcu_class::c_nNestMask) > 0 );
- //CDS_COMPILER_RW_BARRIER;
- pRec->m_nAccessControl.fetch_sub( 1, CDS_ATOMIC::memory_order_release );
+#if CDS_COMPILER == CDS_COMPILER_CLANG && CDS_COMPILER_VERSION < 30800
+ // CLang 3.6-3.7: some tests of intrusive::FeldmanHashSet based on general-purpose RCU
+ // are failed even in single-threaded mode (unit tests) without magic compiler barrier below
+ CDS_COMPILER_RW_BARRIER;
+#endif
+ pRec->m_nAccessControl.store( tmp - 1, atomics::memory_order_release );
}
template <typename RCUtag>
inline bool gp_thread_gc<RCUtag>::is_locked()
{
thread_record * pRec = get_thread_record();
- assert( pRec != null_ptr<thread_record *>());
+ assert( pRec != nullptr );
- return (pRec->m_nAccessControl.load( CDS_ATOMIC::memory_order_relaxed ) & rcu_class::c_nNestMask) != 0;
+ return (pRec->m_nAccessControl.load( atomics::memory_order_relaxed ) & rcu_class::c_nNestMask) != 0;
}
template <typename RCUtag>
inline bool gp_singleton<RCUtag>::check_grace_period( typename gp_singleton<RCUtag>::thread_record * pRec ) const
{
- uint32_t const v = pRec->m_nAccessControl.load( CDS_ATOMIC::memory_order_relaxed );
+ uint32_t const v = pRec->m_nAccessControl.load( atomics::memory_order_acquire );
return (v & general_purpose_rcu::c_nNestMask)
- && ((( v ^ m_nGlobalControl.load( CDS_ATOMIC::memory_order_relaxed )) & ~general_purpose_rcu::c_nNestMask ));
+ && (( v ^ m_nGlobalControl.load( atomics::memory_order_relaxed )) & ~general_purpose_rcu::c_nNestMask );
}
template <typename RCUtag>
template <class Backoff>
inline void gp_singleton<RCUtag>::flip_and_wait( Backoff& bkoff )
{
- OS::ThreadId const nullThreadId = OS::nullThreadId();
- m_nGlobalControl.fetch_xor( general_purpose_rcu::c_nControlBit, CDS_ATOMIC::memory_order_seq_cst );
+ OS::ThreadId const nullThreadId = OS::c_NullThreadId;
+ m_nGlobalControl.fetch_xor( general_purpose_rcu::c_nControlBit, atomics::memory_order_seq_cst );
- for ( thread_record * pRec = m_ThreadList.head( CDS_ATOMIC::memory_order_acquire); pRec; pRec = pRec->m_list.m_pNext ) {
- while ( pRec->m_list.m_idOwner.load( CDS_ATOMIC::memory_order_acquire) != nullThreadId && check_grace_period( pRec ) ) {
+ for ( thread_record * pRec = m_ThreadList.head( atomics::memory_order_acquire ); pRec; pRec = pRec->m_list.m_pNext ) {
+ while ( pRec->m_list.m_idOwner.load( atomics::memory_order_acquire ) != nullThreadId && check_grace_period( pRec )) {
bkoff();
CDS_COMPILER_RW_BARRIER;
}
}}} // namespace cds:urcu::details
//@endcond
-#endif // #ifndef _CDS_URCU_DETAILS_GP_H
+#endif // #ifndef CDSLIB_URCU_DETAILS_GP_H
projects / libcds.git / blobdiff