C++20 three way comparison operator: Part 7

by Gajendra Gulgulia

In the fifth and sixth part of the tutorial series, I explained the comparison category std::strong_ordering and std::weak_ordering respectively with examples and use cases. In this part of the tutorial series, we take a closer look at the third and final comparison category, i.e. std::partial_ordering...

Classic Railroad Field Trip Announced

The CppCon 2021 Field Trip will be an adventure into the mountains to sample classic mountain cuisine from Beau Jo’s, followed by a train trip over the far-famed Georgetown Loop. Spend a fun-filled Sunday on October 24 with fellow attendees as we ascend to 9,101 ft (2,774 m) via air-conditioned buses and railroad coaches. This trip will encompass Geologic, Historic, Natural, and Culinary wonders west of Denver...

Abusing Your Memory Model for Fun and Profit

by Samy Al Bahra, Paul Khuong

The most efficient concurrent C++ data structures used in the wild today usually achieve break-neck performance by either constraining their workload or constraining correctness to a particular memory model. The audience will learn about the Wild West of abusing memory models for performance and simplification, through real world examples. Non-blocking data structures and their benefits often come at the cost of increased latency because they require additional complexity in the common case. There are plenty of exceptions to this if the requirements of the data structure are relaxed, such as supporting only a bounded level of write or read concurrency or if correctness is constrained to a particular memory model. For this reason, well-designed specialized non-blocking data structures guarantee improved resiliency, throughput and latency in all cases compared to alternatives relying on traditional concurrency primitives. Specialized concurrent structures are common place in the Linux kernel and other performance critical systems.

You will learn about foundational concepts to understanding your underlying hardware's memory model and abusing memory models for fun and profit:
* Cache coherency
* Store Buffers
* Pipelines and speculative execution

This talk provides real-world examples that exploit the x86-TSO model to their advantage:
* A general technique to turn literally, any, open-addressed hash table into a concurrent hash table with low to negligible (near 0) cost. The transformation makes your hash table wait-free for writers and mostly wait-free for readers (lock-free in hypothetical worse cases) and is practical for languages such as C++. The mechanism is superior to the previously popular Azure lock-free hash table and even more importantly, practical for any non-garbage-collected environment. The overhead is negligible on TSO and low on non-TSO.
* Blazingly fast event counters. An extremely efficient replacement for condition variables is introduced and faster than any other alternative. This is implemented without requiring any heavy-weight atomic operations on the fast path by exploiting properties of the x86-TSO model.
* Scalable memory management: Exploit the ordering and visibility constraints of the underlying architecture for blazingly fast implementations of RCU and other safe memory reclamation schemes.
* and more.

All rotations of a string with ranges

by Marco Arena

A common programming puzzle consists in printing all rotations of a string. For example, all rotations of abc are:

abc
bca
cab

...

Announcing the second set of AMAs for Meeting C++ 2021

by Jens Weller

This second set of C++ Experts nicely completes the first one. Its now a mix of folks involved with the C++ Standard complimented with experience in C++ and other languages. All of them have made contributions to C++ and our community, very much looking forward to see the set of questions for each of them in November!

A Parallel and Heterogeneous Task Programming System Using Modern C++

by Tsung-Wei Huang

The Taskflow project addresses the long-standing question: "How can we make it easier for developers to write parallel and heterogeneous programs with high performance and simultaneous high productivity?" Modern scientific computing relies on a heterogeneous mix of computational patterns, domain algorithms, and specialized hardware to achieve key scientific milestones that go beyond traditional capabilities. However, programming these applications often requires complex expert-level tools and a deep understanding of software methodologies. Specifically, the lack of a suitable software environment that can overcome the complexity of programming large parallel and heterogeneous systems has posed a significant barrier for many organizations to facilitate transformational discoveries.

Taskflow develops a simple and powerful task programming model to enable efficient implementations of heterogeneous decomposition strategies. Our programming model empowers users with both static and dynamic task graph constructions to incorporate a broad range of computational patterns including hybrid CPU-GPU computing, dynamic control flow, and irregularity. We develop an efficient heterogeneous work-stealing strategy that adapts worker threads to available task parallelism at any time during the graph execution. We have demonstrated promising performance of Taskflow on both micro-benchmark and real-world applications. As an example, we solved a large machine learning workload by up to 1.5× faster, 1.6× less memory, and 1.7× fewer lines of code than two industrial-strength systems, oneTBB and StarPU, on a machine of 40 CPUs and 4 GPUs.

This talk will cover three aspects: (1) heterogeneous task programming model using modern C++, (2) an efficient work-stealing strategy generalizable to arbitrary heterogeneous domains, and (3) user experience we have obtained and suggested roadmap for C++ in face of future heterogeneity.

The Taskflow project is available at https://taskflow.github.io/

C++20 three way comparison operator part 5 and part 6

by Gajendra Gulgulia

In the fourth part of the tutorial, I introduced the theoretical ideas behind the return types of the three way operator in C++20 and demonstrated that there might be semantic restrictions on the program that allow for comparison of equivalence instead of equality. Also I demonstrated by a simple example when can two objects be semantically incomparable, even though the syntax of program allows to compare them and how to deal with such cases with the help of operator<=>...

On the perils of holding a lock across a coroutine suspension point, part 2: Nonrecursive mutexes

On the perils of holding a lock across a coroutine suspension point, part 3: Solutions

by Raymond Chen

Last time, we looked at what can go wrong if you hold a recursive mutex across a coroutine suspension point. Do things get any better if you switch to a nonrecursive mutex?

Small is beautiful: Techniques to minimise memory footprint

by Steven Pigeon

When we code, we often assume that the computer has an infinite amount of memory. This is, of course, /very/ false. It is false for embedded programming, where memory is often limited to a few megabytes or even a few kilobytes, but it is also false for the most powerful workstations and servers. Using memory wisely makes your application possible on small systems as well as prevents, or at least reduces, scaling problems for larger applications. While some programming languages are designed to hide implementation details from the programmer, C++ allows the programmer to specify, with a good level of control, how memory is allocated, structured, and used. In this talk, we will explore what can be done at run-time with negligible cost, what can be done at compile-time with meta-programming, and how we can thwart default compiler behavior to achieve memory-efficient type-safe data representations. We will also extend the discussion to the higher-level reorganization of data structures in order to make a better use of memory.

On the perils of holding a lock across a coroutine suspension point, part 1: The set-up

by Raymond Chen

Say you want to perform a bunch of asynchronous operations involving some object state, but also want to make sure that no other tasks access that object state at the same time. For synchronous code, you would use a traditional synchronization object like a mutex or critical section...