feat add event_bus

CMakeLists.txt

@@ -134,8 +134,9 @@ function(sled_add_test)
   add_executable(${SLED_TEST_NAME} ${SLED_TEST_SRCS})
 
   if(CMAKE_CXX_COMPILER_ID STREQUAL "Clang")
-    set(EXTRA_FLAGS -Wthread-safety -g -fsanitize=address
+    set(EXTRA_FLAGS # -Wthread-safety
-                    -fno-omit-frame-pointer -fno-optimize-sibling-calls)
+        -g -fsanitize=address -fno-omit-frame-pointer
+        -fno-optimize-sibling-calls)
     target_compile_options(${SLED_TEST_NAME} PRIVATE ${EXTRA_FLAGS})
 
     target_link_options(${SLED_TEST_NAME} PRIVATE ${EXTRA_FLAGS})
@@ -179,11 +180,13 @@ if(SLED_BUILD_TESTS)
     src/sled/rx_test.cc
     src/sled/uri_test.cc)
 
-  sled_add_test(NAME sled_symbolize_test SRCS
+  sled_add_test(NAME sled_event_bus_test SRCS
-                src/sled/debugging/symbolize_test.cc NO_MAIN)
+                src/sled/event_bus/event_bus_test.cc)
   sled_add_test(NAME sled_lua_test SRCS tests/lua_test.cc)
   sled_add_test(NAME sled_move_on_copy_test SRCS
                 src/sled/utility/move_on_copy_test.cc)
+  sled_add_test(NAME sled_symbolize_test SRCS
+                src/sled/debugging/symbolize_test.cc NO_MAIN)
 endif(SLED_BUILD_TESTS)
 
 if(SLED_BUILD_FUZZ)

src/sled/event_bus/event_bus.h

@@ -0,0 +1,154 @@
+#ifndef SLED_EVENT_BUS_EVENT_BUS_H
+#define SLED_EVENT_BUS_EVENT_BUS_H
+
+#include "sled/sigslot.h"
+#include "sled/synchronization/mutex.h"
+#include <typeindex>
+
+namespace sled {
+
+class EventBus;
+class EventSubscriber;
+
+namespace {
+
+template<typename Event>
+class EventRegistry {
+public:
+    using Dispatcher      = sigslot::signal1<Event, sigslot::single_threaded>;
+    using SubscriberTable = std::unordered_map<EventBus *, Dispatcher>;
+
+    static EventRegistry &Instance()
+    {
+        static EventRegistry instance_;
+        return instance_;
+    }
+
+    static std::function<void(EventBus *)> &GetCleanupHandler()
+    {
+        static std::function<void(EventBus *)> cleanup_handler
+            = std::bind(&EventRegistry::OnBusDestroyed, &Instance(), std::placeholders::_1);
+        return cleanup_handler;
+    }
+
+    void Post(EventBus *bus, Event event)
+    {
+        sled::SharedMutexReadLock lock(&shared_mutex_);
+        if (signals_.empty()) { return; }
+        auto iter = signals_.find(bus);
+        if (iter != signals_.end()) { iter->second(event); }
+    }
+
+    template<typename C>
+    void Subscribe(EventBus *bus, C *instance, void (C::*method)(Event))
+    {
+        sled::SharedMutexWriteLock lock(&shared_mutex_);
+        auto iter = signals_.find(bus);
+        if (iter == signals_.end()) {
+            signals_.emplace(bus, Dispatcher());
+            iter = signals_.find(bus);
+        }
+        auto &dispatcher = iter->second;
+        dispatcher.connect(instance, method);
+    }
+
+    template<typename C>
+    void Unsubscribe(EventBus *bus, C *instance)
+    {
+        sled::SharedMutexWriteLock lock(&shared_mutex_);
+        auto iter = signals_.find(bus);
+        if (iter == signals_.end()) { return; }
+        auto &dispatcher = iter->second;
+        dispatcher.disconnect(instance);
+    }
+
+    bool IsEmpty(EventBus *bus) const
+    {
+        sled::SharedMutexReadLock lock(&shared_mutex_);
+        auto iter = signals_.find(bus);
+        if (iter == signals_.end()) { return true; }
+        return iter->second.is_empty();
+    }
+
+    void OnBusDestroyed(EventBus *bus)
+    {
+        sled::SharedMutexWriteLock lock(&shared_mutex_);
+        signals_.erase(bus);
+    }
+
+    template<typename C>
+    void OnSubscriberDestroyed(C *instance)
+    {
+        sled::SharedMutexWriteLock lock(&shared_mutex_);
+        for (auto &entry : signals_) {
+            auto &dispatcher = entry.second;
+            dispatcher.disconnect(instance);
+        }
+    }
+
+private:
+    mutable sled::SharedMutex shared_mutex_;
+
+    SubscriberTable signals_;
+};
+
+}// namespace
+
+class EventSubscriber : public sigslot::has_slots<> {
+public:
+    virtual ~EventSubscriber() {}
+};
+
+class EventBus {
+public:
+    EventBus() = default;
+
+    ~EventBus()
+    {
+        for (const auto &handler : cleanup_handlers_) { handler.second(this); }
+    }
+
+    EventBus(const EventBus &)            = delete;
+    EventBus &operator=(const EventBus &) = delete;
+
+    template<typename Event>
+    void Post(const Event &event)
+    {
+        EventRegistry<Event>::Instance().Post(this, event);
+    }
+
+    // On<Event1> ([](const Event1 &){})
+    template<typename Event, typename C>
+    typename std::enable_if<std::is_base_of<EventSubscriber, C>::value>::type
+    Subscribe(C *instance, void (C::*method)(Event))
+    {
+        {
+            sled::MutexLock lock(&mutex_);
+            cleanup_handlers_[std::type_index(typeid(Event))] = EventRegistry<Event>::GetCleanupHandler();
+        }
+
+        EventRegistry<Event>::Instance().Subscribe(this, instance, method);
+    }
+
+    template<typename Event, typename C>
+    typename std::enable_if<std::is_base_of<EventSubscriber, C>::value>::type Unsubscribe(C *instance)
+    {
+        EventRegistry<Event>::Instance().Unsubscribe(this, instance);
+        {
+            sled::MutexLock lock(&mutex_);
+            if (EventRegistry<Event>::Instance().IsEmpty(this)) {
+                auto iter = cleanup_handlers_.find(std::type_index(typeid(Event)));
+                if (iter != cleanup_handlers_.end()) {
+                    iter->second(this);
+                    cleanup_handlers_.erase(iter);
+                }
+            }
+        }
+    }
+
+private:
+    sled::Mutex mutex_;
+    std::unordered_map<std::type_index, std::function<void(EventBus *)>> cleanup_handlers_ GUARDED_BY(mutex_);
+};
+}// namespace sled
+#endif// SLED_EVENT_BUS_EVENT_BUS_H

src/sled/event_bus/event_bus_test.cc

@@ -0,0 +1,65 @@
+#include <sled/event_bus/event_bus.h>
+#include <sled/system/thread_pool.h>
+
+struct Event1 {
+    int a;
+};
+
+struct Event2 {
+    std::string str;
+};
+
+struct Subscriber : public sled::EventSubscriber {
+    void OnEvent1(Event1 event) { a += event.a; }
+
+    void OnEvent2(Event2 event) { str += event.str; }
+
+    int a           = 0;
+    std::string str = "";
+};
+
+TEST_SUITE("EventBus")
+{
+    TEST_CASE("single thread")
+    {
+        sled::EventBus bus;
+        bus.Post(Event1{1});
+        bus.Post(Event2{"1"});
+
+        Subscriber subscriber;
+        bus.Subscribe<Event1>(&subscriber, &Subscriber::OnEvent1);
+        bus.Subscribe<Event2>(&subscriber, &Subscriber::OnEvent2);
+
+        bus.Post(Event1{1});
+        bus.Post(Event2{"1"});
+
+        CHECK_EQ(subscriber.a, 1);
+        CHECK_EQ(subscriber.str, "1");
+
+        bus.Post(Event1{1});
+        bus.Post(Event2{"1"});
+
+        CHECK_EQ(subscriber.a, 2);
+        CHECK_EQ(subscriber.str, "11");
+    }
+
+    TEST_CASE("multi thread")
+    {
+        auto thread = sled::Thread::Create();
+        thread->Start();
+
+        sled::EventBus bus;
+        Subscriber subscriber;
+
+        bus.Subscribe(&subscriber, &Subscriber::OnEvent1);
+        bus.Subscribe(&subscriber, &Subscriber::OnEvent2);
+
+        thread->BlockingCall([&] {
+            bus.Post(Event1{1});
+            bus.Post(Event2{"1"});
+        });
+
+        CHECK_EQ(subscriber.a, 1);
+        CHECK_EQ(subscriber.str, "1");
+    }
+}

src/sled/sigslot.cc

@@ -4,12 +4,19 @@ namespace sigslot {
 
 #ifdef _SIGSLOT_HAS_POSIX_THREADS
 
-pthread_mutex_t *
+sled::Mutex *
 multi_threaded_global::get_mutex()
 {
-    static pthread_mutex_t g_mutex = PTHREAD_MUTEX_INITIALIZER;
+    static sled::Mutex g_mutex;
     return &g_mutex;
 }
 
+// pthread_mutex_t *
+// multi_threaded_global::get_mutex()
+// {
+//     static pthread_mutex_t g_mutex = PTHREAD_MUTEX_INITIALIZER;
+//     return &g_mutex;
+// }
+
 #endif// _SIGSLOT_HAS_POSIX_THREADS
 }// namespace sigslot

src/sled/sigslot.h

@@ -94,10 +94,11 @@
 // If signalx is single threaded the user must ensure that disconnect, connect
 // or signal is not happening concurrently or data race may occur.
 
-#pragma once
 #ifndef SLED_SIGSLOT_H
 #define SLED_SIGSLOT_H
+#pragma once
 
+#include "sled/synchronization/mutex.h"
 #include <cstring>
 #include <list>
 #include <set>
@@ -105,9 +106,7 @@
 // On our copy of sigslot.h, we set single threading as default.
 #define SIGSLOT_DEFAULT_MT_POLICY single_threaded
 
-#if defined(SIGSLOT_PURE_ISO)                                                  \
+#if defined(SIGSLOT_PURE_ISO) || (!defined(WEBRTC_WIN) && !defined(__GNUG__) && !defined(SIGSLOT_USE_POSIX_THREADS))
-    || (!defined(WEBRTC_WIN) && !defined(__GNUG__)                             \
-        && !defined(SIGSLOT_USE_POSIX_THREADS))
 #define _SIGSLOT_SINGLE_THREADED
 #elif defined(WEBRTC_WIN)
 #define _SIGSLOT_HAS_WIN32_THREADS
@@ -167,10 +166,7 @@ class multi_threaded_local {
 public:
     multi_threaded_local() { InitializeCriticalSection(&m_critsec); }
 
-    multi_threaded_local(const multi_threaded_local &)
+    multi_threaded_local(const multi_threaded_local &) { InitializeCriticalSection(&m_critsec); }
-    {
-        InitializeCriticalSection(&m_critsec);
-    }
 
     ~multi_threaded_local() { DeleteCriticalSection(&m_critsec); }
 
@@ -187,31 +183,49 @@ private:
 // The multi threading policies only get compiled in if they are enabled.
 class multi_threaded_global {
 public:
-    void lock() { pthread_mutex_lock(get_mutex()); }
+    void lock()
+    {
+        get_mutex()->Lock();
+        // pthread_mutex_lock(get_mutex());
+    }
 
-    void unlock() { pthread_mutex_unlock(get_mutex()); }
+    void unlock()
+    {
+        get_mutex()->Unlock();
+        // pthread_mutex_unlock(get_mutex());
+    }
 
 private:
-    static pthread_mutex_t *get_mutex();
+    static sled::Mutex *get_mutex();
+    // static pthread_mutex_t *get_mutex();
 };
 
 class multi_threaded_local {
 public:
-    multi_threaded_local() { pthread_mutex_init(&m_mutex, nullptr); }
+    // multi_threaded_local() { pthread_mutex_init(&m_mutex, nullptr); }
+    //
+    // multi_threaded_local(const multi_threaded_local &)
+    // {
+    //     pthread_mutex_init(&m_mutex, nullptr);
+    // }
+    //
+    // ~multi_threaded_local() { pthread_mutex_destroy(&m_mutex); }
 
-    multi_threaded_local(const multi_threaded_local &)
+    void lock()
     {
-        pthread_mutex_init(&m_mutex, nullptr);
+        mutex_.Lock();
+        // pthread_mutex_lock(&m_mutex);
     }
 
-    ~multi_threaded_local() { pthread_mutex_destroy(&m_mutex); }
+    void unlock()
-
+    {
-    void lock() { pthread_mutex_lock(&m_mutex); }
+        mutex_.Unlock();
-
+        // pthread_mutex_unlock(&m_mutex);
-    void unlock() { pthread_mutex_unlock(&m_mutex); }
+    }
 
 private:
-    pthread_mutex_t m_mutex;
+    sled::Mutex mutex_;
+    // pthread_mutex_t m_mutex;
 };
 #endif// _SIGSLOT_HAS_POSIX_THREADS
 
@@ -229,10 +243,8 @@ class _signal_base_interface;
 
 class has_slots_interface {
 private:
-    typedef void (*signal_connect_t)(has_slots_interface *self,
+    typedef void (*signal_connect_t)(has_slots_interface *self, _signal_base_interface *sender);
-                                     _signal_base_interface *sender);
+    typedef void (*signal_disconnect_t)(has_slots_interface *self, _signal_base_interface *sender);
-    typedef void (*signal_disconnect_t)(has_slots_interface *self,
-                                        _signal_base_interface *sender);
     typedef void (*disconnect_all_t)(has_slots_interface *self);
 
     const signal_connect_t m_signal_connect;
@@ -240,9 +252,7 @@ private:
     const disconnect_all_t m_disconnect_all;
 
 protected:
-    has_slots_interface(signal_connect_t conn,
+    has_slots_interface(signal_connect_t conn, signal_disconnect_t disc, disconnect_all_t disc_all)
-                        signal_disconnect_t disc,
-                        disconnect_all_t disc_all)
         : m_signal_connect(conn),
           m_signal_disconnect(disc),
           m_disconnect_all(disc_all)
@@ -253,23 +263,16 @@ protected:
     virtual ~has_slots_interface() {}
 
 public:
-    void signal_connect(_signal_base_interface *sender)
+    void signal_connect(_signal_base_interface *sender) { m_signal_connect(this, sender); }
-    {
-        m_signal_connect(this, sender);
-    }
 
-    void signal_disconnect(_signal_base_interface *sender)
+    void signal_disconnect(_signal_base_interface *sender) { m_signal_disconnect(this, sender); }
-    {
-        m_signal_disconnect(this, sender);
-    }
 
     void disconnect_all() { m_disconnect_all(this); }
 };
 
 class _signal_base_interface {
 private:
-    typedef void (*slot_disconnect_t)(_signal_base_interface *self,
+    typedef void (*slot_disconnect_t)(_signal_base_interface *self, has_slots_interface *pslot);
-                                      has_slots_interface *pslot);
     typedef void (*slot_duplicate_t)(_signal_base_interface *self,
                                      const has_slots_interface *poldslot,
                                      has_slots_interface *pnewslot);
@@ -286,13 +289,9 @@ protected:
     ~_signal_base_interface() {}
 
 public:
-    void slot_disconnect(has_slots_interface *pslot)
+    void slot_disconnect(has_slots_interface *pslot) { m_slot_disconnect(this, pslot); }
-    {
-        m_slot_disconnect(this, pslot);
-    }
 
-    void slot_duplicate(const has_slots_interface *poldslot,
+    void slot_duplicate(const has_slots_interface *poldslot, has_slots_interface *pnewslot)
-                        has_slots_interface *pnewslot)
     {
         m_slot_duplicate(this, poldslot, pnewslot);
     }
@@ -323,15 +322,14 @@ public:
     _opaque_connection(DestT *pd, void (DestT::*pm)(Args...)) : pdest(pd)
     {
         typedef void (DestT::*pm_t)(Args...);
-        static_assert(sizeof(pm_t) <= sizeof(pmethod),
+        static_assert(sizeof(pm_t) <= sizeof(pmethod), "Size of slot function pointer too large.");
-                      "Size of slot function pointer too large.");
 
         std::memcpy(pmethod, &pm, sizeof(pm_t));
 
         typedef void (*em_t)(const _opaque_connection *self, Args...);
         union_caster<em_t, emit_t> caster2;
         caster2.from = &_opaque_connection::emitter<DestT, Args...>;
-        pemit = caster2.to;
+        pemit        = caster2.to;
     }
 
     has_slots_interface *getdest() const { return pdest; }
@@ -339,7 +337,7 @@ public:
     _opaque_connection duplicate(has_slots_interface *newtarget) const
     {
         _opaque_connection res = *this;
-        res.pdest = newtarget;
+        res.pdest              = newtarget;
         return res;
     }
 
@@ -360,8 +358,7 @@ private:
     {
         typedef void (DestT::*pm_t)(Args...);
         pm_t pm;
-        static_assert(sizeof(pm_t) <= sizeof(pmethod),
+        static_assert(sizeof(pm_t) <= sizeof(pmethod), "Size of slot function pointer too large.");
-                      "Size of slot function pointer too large.");
         std::memcpy(&pm, self->pmethod, sizeof(pm_t));
         (static_cast<DestT *>(self->pdest)->*(pm))(args...);
     }
@@ -373,8 +370,7 @@ protected:
     typedef std::list<_opaque_connection> connections_list;
 
     _signal_base()
-        : _signal_base_interface(&_signal_base::do_slot_disconnect,
+        : _signal_base_interface(&_signal_base::do_slot_disconnect, &_signal_base::do_slot_duplicate),
-                                 &_signal_base::do_slot_duplicate),
           m_current_iterator(m_connected_slots.end())
     {}
 
@@ -385,8 +381,7 @@ private:
 
 public:
     _signal_base(const _signal_base &o)
-        : _signal_base_interface(&_signal_base::do_slot_disconnect,
+        : _signal_base_interface(&_signal_base::do_slot_disconnect, &_signal_base::do_slot_duplicate),
-                                 &_signal_base::do_slot_duplicate),
           m_current_iterator(m_connected_slots.end())
     {
         lock_block<mt_policy> lock(this);
@@ -409,8 +404,7 @@ public:
         while (!m_connected_slots.empty()) {
             has_slots_interface *pdest = m_connected_slots.front().getdest();
             m_connected_slots.pop_front();
-            pdest->signal_disconnect(
+            pdest->signal_disconnect(static_cast<_signal_base_interface *>(this));
-                static_cast<_signal_base_interface *>(this));
         }
         // If disconnect_all is called while the signal is firing, advance the
         // current slot iterator to the end to avoid an invalidated iterator from
@@ -422,7 +416,7 @@ public:
     bool connected(has_slots_interface *pclass)
     {
         lock_block<mt_policy> lock(this);
-        connections_list::const_iterator it = m_connected_slots.begin();
+        connections_list::const_iterator it    = m_connected_slots.begin();
         connections_list::const_iterator itEnd = m_connected_slots.end();
         while (it != itEnd) {
             if (it->getdest() == pclass) return true;
@@ -435,7 +429,7 @@ public:
     void disconnect(has_slots_interface *pclass)
     {
         lock_block<mt_policy> lock(this);
-        connections_list::iterator it = m_connected_slots.begin();
+        connections_list::iterator it    = m_connected_slots.begin();
         connections_list::iterator itEnd = m_connected_slots.end();
 
         while (it != itEnd) {
@@ -447,8 +441,7 @@ public:
                 } else {
                     m_connected_slots.erase(it);
                 }
-                pclass->signal_disconnect(
+                pclass->signal_disconnect(static_cast<_signal_base_interface *>(this));
-                    static_cast<_signal_base_interface *>(this));
                 return;
             }
             ++it;
@@ -456,12 +449,11 @@ public:
     }
 
 private:
-    static void do_slot_disconnect(_signal_base_interface *p,
+    static void do_slot_disconnect(_signal_base_interface *p, has_slots_interface *pslot)
-                                   has_slots_interface *pslot)
     {
         _signal_base *const self = static_cast<_signal_base *>(p);
         lock_block<mt_policy> lock(self);
-        connections_list::iterator it = self->m_connected_slots.begin();
+        connections_list::iterator it    = self->m_connected_slots.begin();
         connections_list::iterator itEnd = self->m_connected_slots.end();
 
         while (it != itEnd) {
@@ -472,8 +464,7 @@ private:
                 // If we're currently using this iterator because the signal is firing,
                 // advance it to avoid it being invalidated.
                 if (self->m_current_iterator == it) {
-                    self->m_current_iterator =
+                    self->m_current_iterator = self->m_connected_slots.erase(it);
-                        self->m_connected_slots.erase(it);
                 } else {
                     self->m_connected_slots.erase(it);
                 }
@@ -483,19 +474,16 @@ private:
         }
     }
 
-    static void do_slot_duplicate(_signal_base_interface *p,
+    static void
-                                  const has_slots_interface *oldtarget,
+    do_slot_duplicate(_signal_base_interface *p, const has_slots_interface *oldtarget, has_slots_interface *newtarget)
-                                  has_slots_interface *newtarget)
     {
         _signal_base *const self = static_cast<_signal_base *>(p);
         lock_block<mt_policy> lock(self);
-        connections_list::iterator it = self->m_connected_slots.begin();
+        connections_list::iterator it    = self->m_connected_slots.begin();
         connections_list::iterator itEnd = self->m_connected_slots.end();
 
         while (it != itEnd) {
-            if (it->getdest() == oldtarget) {
+            if (it->getdest() == oldtarget) { self->m_connected_slots.push_back(it->duplicate(newtarget)); }
-                self->m_connected_slots.push_back(it->duplicate(newtarget));
-            }
 
             ++it;
         }
@@ -540,16 +528,14 @@ public:
 private:
     has_slots &operator=(has_slots const &);
 
-    static void do_signal_connect(has_slots_interface *p,
+    static void do_signal_connect(has_slots_interface *p, _signal_base_interface *sender)
-                                  _signal_base_interface *sender)
     {
         has_slots *const self = static_cast<has_slots *>(p);
         lock_block<mt_policy> lock(self);
         self->m_senders.insert(sender);
     }
 
-    static void do_signal_disconnect(has_slots_interface *p,
+    static void do_signal_disconnect(has_slots_interface *p, _signal_base_interface *sender)
-                                     _signal_base_interface *sender)
     {
         has_slots *const self = static_cast<has_slots *>(p);
         lock_block<mt_policy> lock(self);
@@ -563,7 +549,7 @@ private:
         while (!self->m_senders.empty()) {
             std::set<_signal_base_interface *> senders;
             senders.swap(self->m_senders);
-            const_iterator it = senders.begin();
+            const_iterator it    = senders.begin();
             const_iterator itEnd = senders.end();
 
             while (it != itEnd) {
@@ -627,22 +613,13 @@ using signal0 = signal_with_thread_policy<mt_policy>;
 template<typename A1, typename mt_policy = SIGSLOT_DEFAULT_MT_POLICY>
 using signal1 = signal_with_thread_policy<mt_policy, A1>;
 
-template<typename A1,
+template<typename A1, typename A2, typename mt_policy = SIGSLOT_DEFAULT_MT_POLICY>
-         typename A2,
-         typename mt_policy = SIGSLOT_DEFAULT_MT_POLICY>
 using signal2 = signal_with_thread_policy<mt_policy, A1, A2>;
 
-template<typename A1,
+template<typename A1, typename A2, typename A3, typename mt_policy = SIGSLOT_DEFAULT_MT_POLICY>
-         typename A2,
-         typename A3,
-         typename mt_policy = SIGSLOT_DEFAULT_MT_POLICY>
 using signal3 = signal_with_thread_policy<mt_policy, A1, A2, A3>;
 
-template<typename A1,
+template<typename A1, typename A2, typename A3, typename A4, typename mt_policy = SIGSLOT_DEFAULT_MT_POLICY>
-         typename A2,
-         typename A3,
-         typename A4,
-         typename mt_policy = SIGSLOT_DEFAULT_MT_POLICY>
 using signal4 = signal_with_thread_policy<mt_policy, A1, A2, A3, A4>;
 
 template<typename A1,
@@ -670,8 +647,7 @@ template<typename A1,
          typename A6,
          typename A7,
          typename mt_policy = SIGSLOT_DEFAULT_MT_POLICY>
-using signal7 =
+using signal7 = signal_with_thread_policy<mt_policy, A1, A2, A3, A4, A5, A6, A7>;
-    signal_with_thread_policy<mt_policy, A1, A2, A3, A4, A5, A6, A7>;
 
 template<typename A1,
          typename A2,
@@ -682,9 +658,8 @@ template<typename A1,
          typename A7,
          typename A8,
          typename mt_policy = SIGSLOT_DEFAULT_MT_POLICY>
-using signal8 =
+using signal8 = signal_with_thread_policy<mt_policy, A1, A2, A3, A4, A5, A6, A7, A8>;
-    signal_with_thread_policy<mt_policy, A1, A2, A3, A4, A5, A6, A7, A8>;
 
 }// namespace sigslot
 
-#endif // SLED_SIGSLOT_H
+#endif// SLED_SIGSLOT_H

src/sled/sled.h

@@ -9,6 +9,9 @@ namespace async {}
 #include "inja.hpp"
 #include "rx.h"
 
+// event_bus
+#include "sled/event_bus/event_bus.h"
+
 // filesystem
 #include "sled/filesystem/path.h"
 #include "sled/filesystem/temporary_file.h"

src/sled/synchronization/mutex.h

@@ -146,46 +146,90 @@ private:
     marl::ConditionVariable cv_;
 };
 
-// class ConditionVariable final {
+class SCOPED_CAPABILITY SharedMutex final {
-// public:
+public:
-//     static constexpr TimeDelta kForever = TimeDelta::PlusInfinity();
+    enum class Mode {
-//     ConditionVariable() = default;
+        kReaderPriority,
-//     ConditionVariable(const ConditionVariable &) = delete;
+        kWriterPriority,
-//     ConditionVariable &operator=(const ConditionVariable &) = delete;
+    };
-//
+
-//     template<typename Predicate>
+    inline SharedMutex(Mode mode = SharedMutex::Mode::kWriterPriority) : mode_(mode) {}
-//     inline bool Wait(LockGuard<Mutex> &guard, Predicate pred)
+
-//     {
+    inline void Lock() SLED_EXCLUSIVE_LOCK_FUNCTION()
-//         std::unique_lock<std::mutex> lock(guard.mutex_->impl_, std::adopt_lock);
+    {
-//         cv_.wait(lock, pred);
+        wait_w_count_.fetch_add(1);
-//         return true;
+
-//     }
+        sled::MutexLock lock(&mutex_);
-//
+        if (Mode::kReaderPriority == mode_) {
-//     template<typename Predicate>
+            // 读取优先，必须在没有任何读取的消费者的情况下才能持有锁
-//     inline bool
+            cv_.Wait(lock, [this] { return r_count_ == 0 && w_count_ == 0 && wait_r_count_.load() == 0; });
-//     WaitFor(LockGuard<Mutex> &guard, TimeDelta timeout, Predicate pred)
+            w_count_++;
-//     {
+        } else {
-//         std::unique_lock<std::mutex> lock(guard.mutex_->impl_, std::adopt_lock);
+            // 写入优先，只要没有持有读锁的消费者，就可以加锁
-//         if (timeout == kForever) {
+            cv_.Wait(lock, [this] { return r_count_ == 0 && w_count_ == 0; });
-//             cv_.wait(lock, pred);
+            w_count_++;
-//             return true;
+            cv_.Wait(lock, [this] { return r_count_ == 0; });
-//         } else {
+        }
-//             return cv_.wait_for(lock, std::chrono::milliseconds(timeout.ms()),
+        wait_w_count_.fetch_sub(1);
-//                                 pred);
+    }
-//         }
+
-//     }
+    inline void Unlock() SLED_UNLOCK_FUNCTION()
-//
+    {
-//     // template<typename Predicate>
+        sled::MutexLock lock(&mutex_);
-//     // bool WaitUntil(Mutex *mutex, TimeDelta timeout, Predicate pred)
+        w_count_--;
-//     // {}
+        if (w_count_ == 0) { cv_.NotifyAll(); }
-//
+    }
-//     inline void NotifyOne() { cv_.notify_one(); }
+
-//
+    inline void LockShared() SLED_SHARED_LOCK_FUNCTION()
-//     inline void NotifyAll() { cv_.notify_all(); }
+    {
-//
+        wait_r_count_.fetch_add(1);
-// private:
+        sled::MutexLock lock(&mutex_);
-//     std::condition_variable cv_;
+        if (Mode::kReaderPriority == mode_) {
-// };
+            cv_.Wait(lock, [this] { return w_count_ == 0; });
+            r_count_++;
+        } else {
+            cv_.Wait(lock, [this] { return w_count_ == 0 && wait_w_count_.load() == 0; });
+            r_count_++;
+        }
+        wait_r_count_.fetch_sub(1);
+    }
+
+    inline void UnlockShared() SLED_UNLOCK_FUNCTION()
+    {
+        sled::MutexLock lock(&mutex_);
+        r_count_--;
+        if (r_count_ == 0) { cv_.NotifyAll(); }
+    }
+
+private:
+    const Mode mode_;
+    sled::Mutex mutex_;
+    sled::ConditionVariable cv_;
+    int r_count_{0};
+    int w_count_{0};
+    std::atomic<int> wait_r_count_{0};
+    std::atomic<int> wait_w_count_{0};
+};
+
+class SharedMutexReadLock final {
+public:
+    explicit SharedMutexReadLock(SharedMutex *mutex) : mutex_(mutex) { mutex_->LockShared(); }
+
+    ~SharedMutexReadLock() { mutex_->UnlockShared(); }
+
+private:
+    SharedMutex *mutex_;
+};
+
+class SharedMutexWriteLock final {
+public:
+    explicit SharedMutexWriteLock(SharedMutex *mutex) : mutex_(mutex) { mutex_->Lock(); }
+
+    ~SharedMutexWriteLock() { mutex_->Unlock(); }
+
+private:
+    SharedMutex *mutex_;
+};
 
 }// namespace sled
 

src/sled/timer/task_queue_timeout.cc

@@ -41,7 +41,7 @@ TaskQueueTimeoutFactory::TaskQueueTimeout::Start(DurationMs duration_ms, Timeout
         precision_,
         SafeTask(safety_flag_,
                  [timeout_id, this]() {
-                     LOGV("timer", "Timeout expired: {}", timeout_id);
+                     LOGV("timer", "Timeout expired id={}", timeout_id);
                      SLED_DCHECK_RUN_ON(&parent_.thread_checker_);
                      SLED_DCHECK(posted_task_expiration_ != std::numeric_limits<TimeMs>::max(), "");
                      posted_task_expiration_ = std::numeric_limits<TimeMs>::max();