szaszm commented on a change in pull request #746: MINIFICPP-1185 - Remove moodycamel::concurrentqueue from threadpool URL: https://github.com/apache/nifi-minifi-cpp/pull/746#discussion_r400057782
########## File path: libminifi/include/utils/MinifiConcurrentQueue.h ########## @@ -0,0 +1,186 @@ +/** + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +#ifndef LIBMINIFI_INCLUDE_CONCURRENT_QUEUE_H +#define LIBMINIFI_INCLUDE_CONCURRENT_QUEUE_H + +#include <chrono> +#include <deque> +#include <mutex> +#include <condition_variable> +#include <stdexcept> + +namespace org { +namespace apache { +namespace nifi { +namespace minifi { +namespace utils { + + +// Provides a queue API and guarantees no race conditions in case of multiple producers and consumers. +template <typename T> +class ConcurrentQueue { + public: + explicit ConcurrentQueue() = default; + + ConcurrentQueue(const ConcurrentQueue& other) = delete; + ConcurrentQueue& operator=(const ConcurrentQueue& other) = delete; + ConcurrentQueue(ConcurrentQueue&& other) + : ConcurrentQueue(std::move(other), std::lock_guard<std::mutex>(other.mutex_)) {} + + ConcurrentQueue& operator=(ConcurrentQueue&& other) { + if (this != &other) { + std::lock(mtx_, other.mtx_); + std::lock_guard<std::mutex> lk1(mtx_, std::adopt_lock); + std::lock_guard<std::mutex> lk2(other.mtx_, std::adopt_lock); + queue_.swap(other.queue_); + } + return *this; + } + + bool tryDequeue(T& out) { + std::unique_lock<std::mutex> lck(mtx_); + return tryDequeueImpl(lck, out); + } + + bool empty() const { + std::unique_lock<std::mutex> lck(mtx_); + return queue_.emptyImpl(lck); + } + + size_t size() const { + std::lock_guard<std::mutex> guard(mtx_); + return queue_.size(); + } + + void clear() { + std::lock_guard<std::mutex> guard(mtx_); + queue_.clear(); + } + + template <typename... Args> + void enqueue(Args&&... args) { + std::lock_guard<std::mutex> guard(mtx_); + queue_.emplace_back(std::forward<Args>(args)...); + } + + private: + ConcurrentQueue(ConcurrentQueue&& other, std::lock_guard<std::mutex>&) + : queue_( std::move(other.queue_) ) {} + + protected: + void checkLock(std::unique_lock<std::mutex>& lck) const { + if (!lck.owns_lock()) { + throw std::logic_error("Caller of protected functions of ConcurrentQueue should own the lock!"); + } + } + + bool tryDequeueImpl(std::unique_lock<std::mutex>& lck, T& out) { + checkLock(lck); + if (queue_.empty()) { + return false; + } + out = std::move(queue_.front()); + queue_.pop_front(); + return true; + } + + bool emptyImpl(std::unique_lock<std::mutex>& lck) const { + checkLock(lck); + return queue_.empty(); + } + + mutable std::mutex mtx_; + private: + std::deque<T> queue_; +}; + + +// A ConcurrentQueue extended with a condition variable to be able to block and wait for incoming data +template <typename T> +class ConditionConcurrentQueue : private ConcurrentQueue<T> { + public: + explicit ConditionConcurrentQueue(bool start = true) : ConcurrentQueue<T>{}, running_{start} {} + + ConditionConcurrentQueue(const ConditionConcurrentQueue& other) = delete; + ConditionConcurrentQueue& operator=(const ConditionConcurrentQueue& other) = delete; + ConditionConcurrentQueue(ConditionConcurrentQueue&& other) = delete; + ConditionConcurrentQueue& operator=(ConditionConcurrentQueue&& other) = delete; + + using ConcurrentQueue<T>::size; + using ConcurrentQueue<T>::empty; + using ConcurrentQueue<T>::clear; + + + template <typename... Args> + void enqueue(Args&&... args) { + ConcurrentQueue<T>::enqueue(std::forward<Args>(args)...); + if (running_) { + cv_.notify_one(); + } + } + + bool dequeueWait(T& out) { + std::unique_lock<std::mutex> lck(this->mtx_); + cv_.wait(lck, [this, &lck]{ return !running_ || !this->emptyImpl(lck); }); // Only wake up if there is something to return or stopped + return running_ && ConcurrentQueue<T>::tryDequeueImpl(lck, out); + } + + template< class Rep, class Period > + bool dequeueWaitFor(T& out, const std::chrono::duration<Rep, Period>& time) { + std::unique_lock<std::mutex> lck(this->mtx_); + cv_.wait_for(lck, time, [this, &lck]{ return !running_ || !this->emptyImpl(lck); }); // Wake up with timeout or in case there is something to do + return running_ && ConcurrentQueue<T>::tryDequeueImpl(lck, out); + } + + bool tryDequeue(T& out) { + std::unique_lock<std::mutex> lck(this->mtx_); + return running_ && ConcurrentQueue<T>::tryDequeueImpl(lck, out); + } + + void stop() { + std::lock_guard<std::mutex> guard(this->mtx_); + running_ = false; + cv_.notify_all(); + } + + void start() { + std::unique_lock<std::mutex> lck(this->mtx_); + if (!running_) { + running_ = true; + if (!this->emptyImpl(lck)) { + cv_.notify_all(); + } Review comment: Since we are transitioning from stopped to started in a critical section, and it's not possible to have consumers in a stopped state, this notification shouldn't be necessary. ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services