Book Image

Mastering Embedded Linux Programming

By : Chris Simmonds
Book Image

Mastering Embedded Linux Programming

By: Chris Simmonds

Overview of this book

Mastering Embedded Linux Programming takes you through the product cycle and gives you an in-depth description of the components and options that are available at each stage. You will begin by learning about toolchains, bootloaders, the Linux kernel, and how to configure a root filesystem to create a basic working device. You will then learn how to use the two most commonly used build systems, Buildroot and Yocto, to speed up and simplify the development process. Building on this solid base, the next section considers how to make best use of raw NAND/NOR flash memory and managed flash eMMC chips, including mechanisms for increasing the lifetime of the devices and to perform reliable in-field updates. Next, you need to consider what techniques are best suited to writing applications for your device. We will then see how functions are split between processes and the usage of POSIX threads, which have a big impact on the responsiveness and performance of the final device The closing sections look at the techniques available to developers for profiling and tracing applications and kernel code using perf and ftrace.
Table of Contents (22 chapters)
Mastering Embedded Linux Programming
About the Author
About the Reviewers

The observer effect

Before diving into the tools, let's talk about what the tools will show you. As is the case in many fields, measuring a certain property affects the observation itself. Measuring the electric current in a line requires measuring the voltage drop over a small resistor. However, the resistor itself affects the current. The same is true for profiling: every system observation has a cost in CPU cycles and that resource is no longer spent on the application. Measurement tools also mess up caching behavior, eat memory space, and write to disk, which all make it worse. There is no measurement without overhead.

I've often heard engineers say that the results of a profiling job were totally misleading. That is usually because they were performing the measurements on something approaching a real situation. Always try to measure on the target, using release builds of the software, with a valid data set, using as few extra services as possible.

Symbol tables and compile flags

We will...