boostorg/unordered
1
0
Fork 1
mirror of https://github.com/boostorg/unordered.git synced 2025-07-30 03:17:15 +02:00
Code Issues Packages Projects Releases Wiki Activity

optimized epoch waiting

Browse Source
This commit is contained in:
joaquintides
2023-12-02 20:52:20 +01:00
parent 49ff2a63ea
commit 5af735827f
1 changed files with 27 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files

42
include/boost/unordered/detail/foa/concurrent_table.hpp
View File

@ -971,22 +971,27 @@ private:
using exclusive_bilock_guard= using exclusive_bilock_guard=
reentrancy_bichecked<scoped_bilock<multimutex_type>>; reentrancy_bichecked<scoped_bilock<multimutex_type>>;
#if defined(BOOST_UNORDERED_LATCH_FREE)
using epoch_type=std::atomic<std::size_t>;
using epoch_array=cache_aligned_array<epoch_type,128>; // TODO: adapt 128 to the machine
#endif
#if defined(BOOST_UNORDERED_LATCH_FREE) #if defined(BOOST_UNORDERED_LATCH_FREE)
struct group_shared_lock_guard struct group_shared_lock_guard
{ {
group_shared_lock_guard(std::atomic_int& cs_):cs{cs_}{cs.store(1);} group_shared_lock_guard(epoch_type& e_):e{e_}{++e;}
~group_shared_lock_guard(){cs.store(0);} ~group_shared_lock_guard(){++e;}
std::atomic_int& cs; epoch_type& e;
}; };
struct group_exclusive_lock_guard struct group_exclusive_lock_guard
{ {
group_exclusive_lock_guard( group_exclusive_lock_guard(
std::atomic_int& cs_,group_access::exclusive_lock_guard&& lck_): epoch_type& e_,group_access::exclusive_lock_guard&& lck_):
cs{cs_},lck{std::move(lck_)}{cs.store(1);} e{e_},lck{std::move(lck_)}{++e;}
~group_exclusive_lock_guard(){cs.store(0);} ~group_exclusive_lock_guard(){++e;}
std::atomic_int& cs; epoch_type& e;
group_access::exclusive_lock_guard lck; group_access::exclusive_lock_guard lck;
}; };
#else #else
@ -1047,7 +1052,7 @@ private:
inline group_shared_lock_guard access(group_shared,std::size_t pos)const inline group_shared_lock_guard access(group_shared,std::size_t pos)const
{ {
#if defined(BOOST_UNORDERED_LATCH_FREE) #if defined(BOOST_UNORDERED_LATCH_FREE)
return {csections[thread_id()%csections.size()]}; return {epochs[thread_id()%epochs.size()]};
#else #else
return this->arrays.group_accesses()[pos].shared_access(); return this->arrays.group_accesses()[pos].shared_access();
#endif #endif
@ -1058,7 +1063,7 @@ private:
{ {
#if defined(BOOST_UNORDERED_LATCH_FREE) #if defined(BOOST_UNORDERED_LATCH_FREE)
return { return {
csections[thread_id()%csections.size()], epochs[thread_id()%epochs.size()],
this->arrays.group_accesses()[pos].exclusive_access() this->arrays.group_accesses()[pos].exclusive_access()
}; };
#else #else
@ -1505,7 +1510,7 @@ private:
/* slot wasn't empty */ /* slot wasn't empty */
goto startover; goto startover;
} }
wait_for_csections(); // WHY BEFORE THE FOLLOWING? wait_for_epochs(); // WHY BEFORE THE FOLLOWING?
if(BOOST_UNLIKELY(insert_counter(pos0)++!=counter)){ if(BOOST_UNLIKELY(insert_counter(pos0)++!=counter)){
/* other thread inserted from pos0, need to start over */ /* other thread inserted from pos0, need to start over */
pg->reset(n); pg->reset(n);
@ -1815,13 +1820,16 @@ private:
mutable multimutex_type mutexes; mutable multimutex_type mutexes;
#if defined(BOOST_UNORDERED_LATCH_FREE) #if defined(BOOST_UNORDERED_LATCH_FREE)
mutable cache_aligned_array< static epoch_array epochs;
std::atomic_int,128> csections;
void wait_for_csections()const static void wait_for_epochs()
{ {
for(std::size_t i=0;i<csections.size();++i){ for(std::size_t i=0;i<epochs.size();++i){
while(csections[i].load(std::memory_order_acquire)){} auto e=epochs[i].load(std::memory_order_acquire),
e1=e|1u;
while(e==e1){
e=epochs[i].load(std::memory_order_acquire);
}
} }
} }
#endif #endif
@ -1830,6 +1838,10 @@ private:
template<typename T,typename H,typename P,typename A> template<typename T,typename H,typename P,typename A>
std::atomic<std::size_t> concurrent_table<T,H,P,A>::thread_counter={}; std::atomic<std::size_t> concurrent_table<T,H,P,A>::thread_counter={};
template<typename T,typename H,typename P,typename A>
typename concurrent_table<T,H,P,A>::epoch_array
concurrent_table<T,H,P,A>::epochs={};
#if defined(BOOST_MSVC) #if defined(BOOST_MSVC)
#pragma warning(pop) /* C4714 */ #pragma warning(pop) /* C4714 */
#endif #endif