If you are reading this book, you are probably concerned about the performance of your Rust code. It's known that Rust can offer performance close to that of C/C++ programs, and in some cases, Rust can even win those benchmarks. The main issue, though, is that it's sometimes hard to get that efficiency, especially if you are coming from C/C++. Some concepts don't apply, and some simple efficient approaches in those languages are notably worse in Rust.
In this book, you will learn how to really take advantage of Rust to make it perform at its best while maintaining all the benefits it brings—safety, zero-cost abstractions, and great concurrency. The book can be read from start to finish, and you will probably learn new concepts in every chapter. You can go directly to the chapter that interests you, though, as each chapter contains all the required information for its completion, so that it can be used as a reference.
In this first part of the book, we will start with an introduction on how to improve the performance of your sequential code. You will learn how to avoid common performance pitfalls and how to fix direct translations from other languages. You will then learn how to get better performance from your code, and finally understand memory management in Rust.
In this chapter, we will be looking into:
- Configuration of the compilation process with profiles
- Translation pitfalls—learning how to avoid performance pitfalls with array/slice indexing and master iterators
- New iterator adaptors, both in the standard library and in external crates, and coding any complex behavior at zero cost
- How to use the borrow checker to your advantage
Most of the people that start learning Rust, myself included, tend to bring lessons learned in other languages to Rust. This is usually a great thing, as it will enable you to learn the language faster. The main issue with this approach is that some patterns used in other languages can actually be a trade-off in Rust. We will learn about the most common ones and the not-so-common ones, so that anyone trying to get better performance in Rust can learn how to do it.
