Asynchronous Programming Design Patterns - C/C++

Explore the design and implementation of an event loop, a multipurpose asynchronous data structure that schedules computations, and build an event loop library for various async models.

set up a posix-compliant linux environment, such as ubuntu 20.04 in a vm, fork the course repository on github, and use gcc for c/c++ demonstrations.

Design and implement an event loop mini library by defining task structures as function pointers with arguments, using a linked-list queue, mutexes, a condition variable, and a POSIX thread.

Demonstrates single-threaded concurrency by using an event loop to run upload and download tasks in parallel, illustrating task scheduling, context switching, and that tasks remain invisible to the operating system.

Integrate an event loop in a c/c++ project by creating the loop object, packaging routing table updates as a single job, and updating via a callback.

Explore asynchronous programming design patterns in C/C++ by adding expiration timers to routing table entries, using timer setup and callbacks for millisecond expiration to delete expired entries and prevent leaks.

Explore how the event loop schedules asynchronous computations, dequeueing tasks from its queue to enable concurrency in a single threaded program while avoiding forced multithreading.

Understand premature deletion in asynchronous programming, where deleting a data object with pending tasks creates a dangling pointer and may crash the program; learn a graceful method to handle it.

Implement asynchronous single routing table entry deletion in the STP process, using low priority tasks to prevent premature deletion, detach from the routing table, and safely release resources.

Learn how to safely delete an entire routing table in an asynchronous environment by isolating the routing table, launching a low-priority delete task, and deleting a linked list without interference.

Showcases asynchronous deletion of a routing table (a doubly linked list) via delete tasks scheduled at the lowest event-loop priority, deleting ten entries per preempted batch.

Explore why concurrency remains essential in multithreaded systems by examining how 100 threads share 16 CPUs, achieving a hybrid of concurrency and parallelism for better performance.

Design a multi-threaded process when work can be split into independent chunks, using worker threads to service each client and delegating work for isolation of server and client communications.

Create and start two UDP listener threads on localhost, ports 3000 and 3001, passing received packets to the main application via a packet receive function.