Five Awesome C++ Papers for the Q1 2021 and C++23 Status

by Bartlomiej Filipek

Between 2018 and 2020, I released several articles with top 5 proposals just around a new ISO C++ meeting happened. Since March 2020, this pattern broke as the meeting went online. Why not restart the series? We can look at the recent papers from a whole Quarter.

Let’s start!

C++ coroutines: Allowing the awaiter to be destroyed while suspended

by Raymond Chen

One issue that we dealt with when we created our co_await awaitable signal was the case of the awaiter being destroyed while suspended. We had been ignoring that problem in our coroutine promise, but we can’t keep our head in the sand forever. Let’s take a look around and see where we are.

Pros and Cons of Alternative Function Syntax in C++

by Petr Zemek

C++11 introduced an alternative syntax for writing function declarations. Instead of putting the return type before the name of the function (e.g. int func()), the new syntax allows us to write it after the parameters (e.g. auto func() -> int). This leads to a couple of questions: Why was such an alternative syntax added? Is it meant to be a replacement for the original syntax? To help you with these questions, the present blog post tries to summarize the advantages and disadvantages of this newly added syntax...

C++ coroutines: Getting rid of our reference count

by Raymond Chen

In an earlier installment, we simplified our promise_ptr type, and one of the consequences of this is that there are no remaining caller of increment_ref. This means that we don’t need a reference count at all and can rely on the state changes to tell us when to destroy the promise: When the awaiter has obtained the result or, or when the coroutine completes and discovers that the awaiter has abandoned its effort to obtain the result...

C++ coroutines: Getting rid of our mutex

by Raymond Chen

Our coroutine implementation uses a mutex to guard against the race condition where a coroutine completes at the same time another thread tries to wait for its completion. The race condition in question is this one...

C++ coroutines: Making it impossible to co_await a task twice

by Raymond Chen

One design limitation of the coroutine implementation we’ve been developing is that it supports only one co_await client. We enforce this with a runtime assertion, but what if the problem occurs in the release build?

C++ coroutines: Tradeoffs of making the promise be the shared state

by Raymond Chen

Last time, we traced the lifetimes of the objects involved in the coroutine function. Now we can look at some of the tradeoffs we’ve made in our design and see how the decisions are intertwined...

Automatically Resuming a Job with Coroutines on a Separate Thread

by Rainer Grimm

In my last post "Starting Jobs with Coroutines", I applied co_await to start a job. In this post, I improve the workflow and automatically resume a job if necessary. In my final step, I resume the job on a separate thread...

C++ coroutines: The lifetime of objects involved in the coroutine function

by Raymond Chen

We finally hooked up the last missing piece of our coroutine promise implementation. Before we can look at the tradeoffs we’ve made, let’s step back and follow the lifetime of the various objects involved in the coroutine function.

C++ coroutines: Managing the reference count of the coroutine state

by Raymond Chen

Last time, we hooked up the co_await of the simple_task and we had a brief glimpse into the the management of the reference count on the promise (and therefore also the coroutine state). Today we’ll fill in the rest of the story...