Search icon CANCEL
Subscription
0
Cart icon
Your Cart (0 item)
Close icon
You have no products in your basket yet
Arrow left icon
Explore Products
Best Sellers
New Releases
Books
Videos
Audiobooks
Learning Hub
Free Learning
Arrow right icon
Arrow up icon
GO TO TOP
Linux Device Driver Development Cookbook

You're reading from   Linux Device Driver Development Cookbook Learn kernel programming and build custom drivers for your embedded Linux applications

Arrow left icon
Product type Paperback
Published in May 2019
Publisher Packt
ISBN-13 9781838558802
Length 356 pages
Edition 1st Edition
Languages
Tools
Arrow right icon
Author (1):
Arrow left icon
Rodolfo Giometti Rodolfo Giometti
Author Profile Icon Rodolfo Giometti
Rodolfo Giometti
Arrow right icon
View More author details
Toc

Table of Contents (14) Chapters Close

Preface 1. Installing the Development System FREE CHAPTER 2. A Peek Inside the Kernel 3. Working with Char Drivers 4. Using the Device Tree 5. Managing Interrupts and Concurrency 6. Miscellaneous Kernel Internals 7. Advanced Char Driver Operations 8. Additional Information: Working with Char Drivers 9. Additional Information: Using the Device Tree 10. Additional Information: Managing Interrupts and Concurrency 11. Additional Information: Miscellaneous Kernel Internals 12. Additional Information: Advanced Char Driver Operations 13. Other Books You May Enjoy

Setting up the host machine

As every good device driver developer knows, a host machine is absolutely necessary.
Even if the embedded devices are getting more powerful nowadays (and the ESPRESSObin
is one of these), there are some resource-consuming tasks where a host machine can help.
That's why, in this section, we're going to show how to set up our host machine.

The host machine we decide to use could be a normal PC or a virtualized one—they are equivalent—but the important thing is that it must run a GNU/Linux-based OS.

Getting ready

In this book, I will use an Ubuntu 18.04 LTS based system but you can decide to try to replicate some settings and installation commands that we will use during the course of this book into another major Linux distribution, with little effort for a Debian derivative, or in a bit more of a complicated manner in the case of non-Debian derivative distributions.

I'm not going to show how to install a fresh Ubuntu system on a PC nor on a virtualized machine since it's a really easy task for a real programmer; however, as the last step of this chapter (the Doing native compiling on foreign hardware recipe), I will introduce, with detailed steps about how to install it, an interesting cross-platform environment that proved useful to compile foreign target code on the host machine as we were on the target. This procedure is very useful when we need several different OSes running on your development PC.

So, at this point, the reader should have their own PC running (natively or virtualized) a fresh installed Ubuntu 18.04 LTS OS.

The main usage of a host PC is to edit and then cross-compile our new device drivers and to manage our target device via the serial console, to create its root filesystem, and so on.

In order to do it properly, we need some basic tools; some of them are general while others depend on the specific platform onto which we are going to write our drivers.

General tools are surely an editor, a version control system, and a compiler and its related components, while specific platform tools are essentially the cross-compiler and its related components (on some platforms we may need additional tools but our mileage may vary and, in any case, each manufacturer will give us all of the needed requirements for a comfortable compilation environment).

About the editor: I'm not going to spend any words on it because the reader can use whatever they want (regarding myself, for example, I'm still programming with vi editor) but regarding others tools, I'll have to be more specific.

How to do it...

Now that our GNU/Linux distribution is up and running on our host PC we can start to install some programs we're going to use in this book:

  1. First of all, let's install the basic compiling tools:
$ sudo apt install gcc make pkg-config \
bison flex ncurses-dev libssl-dev \
qemu-user-static debootstrap
As you know already, the sudo command is used to execute a command as a privileged user. It should be already present in your system, otherwise you can install it by using the apt install sudo command as the root user.
  1. Next, we have to test the compiling tools. We should be able to compile a C program. As a simple test, let's use the following standard Hello World code stored in the helloworld.c file:
#include <stdio.h>

int main(int argc, char *argv[])
{
printf("Hello World!\n");

return 0;
}
Remember that code can be downloaded from our GitHub repository.
  1. Now, we should be able to compile it by using the following command:
$ make CFLAGS="-Wall -O2" helloworld
cc -Wall -O2 helloworld.c -o helloworld

In the preceding command, we used both the compiler and the make tool, which is required to compile every Linux driver in a comfortable and reliable manner.

You can get more information regarding make by taking a look at https://www.gnu.org/software/make/, and for gcc, you can go to https://www.gnu.org/software/gcc/.
  1. Finally, we can test it on the host PC, as follows:
$ ./helloworld 
Hello World!
  1. The next step is to install the cross-compiler. Since we're going to work with an ARM64 system, we need a cross-compiler and its related tools. To install them, we simply use the following command:
$ sudo apt install gcc-7-aarch64-linux-gnu
Note that we can also use an external toolchain as reported in the ESPRESSObin wiki at http://wiki.espressobin.net/tiki-index.php?page=Build+From+Source+-+Toolchain; however, the Ubuntu toolchain works perfectly!
  1. When the installation is complete, test our new cross-compiler by using the preceding Hello World program, as follows:
$ sudo ln -s /usr/bin/aarch64-linux-gnu-gcc-7 /usr/bin/aarch64-linux-gnu-gcc
$ make CC=aarch64-linux-gnu-gcc CFLAGS="-Wall -O2" helloworld
aarch64-linux-gnu-gcc-7 -Wall -O2 helloworld.c -o helloworld
Note that I've removed the previously compiled helloworld program in order to be able to correctly compile this new version. To do so, I used the mv helloworld helloworld.x86_64 command due to the fact I'll need the x86 version again.

Also, note that since Ubuntu doesn't automatically create the standard cross-compiler name, aarch64-linux-gnu-gcc, we have to do it manually by using the preceding ln command before executing make.
  1. OK, now we can verify that newly created version of the helloworld program for ARM64 by using the following file command. This will point out which platform the program is compiled for:
$ file helloworld
helloworld: ELF 64-bit LSB shared object, ARM aarch64, version 1 (SYSV), dynamically linked, interpreter /lib/ld-linux-aarch64.so.1, for GNU/Linux 3.7.0, BuildID[sha1]=c0d6e9ab89057e8f9101f51ad517a253e5fc4f10, not stripped

If we again use the file command on the previously renamed version, helloworld.x86_64, we get the following:

$ file helloworld.x86_64 
helloworld.x86_64: ELF 64-bit LSB shared object, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, for GNU/Linux 3.2.0, BuildID[sha1]=cf932fab45d36f89c30889df98ed382f6f648203, not stripped
  1. To test whether this new release is really for the ARM64 platform, we can use QEMU, which is an open source and generic machine emulator and virtualizer that is able to execute foreign code on the running platform. To install it, we can use apt command as in the preceding code, specifying the qemu-user-static package:
$ sudo apt install qemu-user-static
  1. Then, we can execute our ARM64 program:
$ qemu-aarch64-static -L /usr/aarch64-linux-gnu/ ./helloworld
Hello World!
To get further information about QEMU, a good staring point is its home page at https://www.qemu.org/.
  1. The next step is to install the version control system. We must install the version control system used for the Linux project, that is, git. To install it, we can use the following command in a similar manner as before:
$ sudo apt install git

If everything works well, we should be able to execute it as follows:

$ git --help
usage: git [--version] [--help] [-C <path>] [-c <name>=<value>]
[--exec-path[=<path>]] [--html-path] [--man-path]
[--info-path] [-p | --paginate | --no-pager]
[--no-replace-objects] [--bare] [--git-dir=<path>]
[--work-tree=<path>] [--namespace=<name>]
<command> [<args>]

These are common Git commands used in various situations:

start a working area (see also: git help tutorial)
clone Clone a repository into a new directory
init Create an empty Git repository or reinitialise an existing one
...
In this book, I'm going to explain every git command used but for complete knowledge of this powerful tool, I suggest you start reading https://git-scm.com/.

See also

You have been reading a chapter from
Linux Device Driver Development Cookbook
Published in: May 2019
Publisher: Packt
ISBN-13: 9781838558802
Register for a free Packt account to unlock a world of extra content!
A free Packt account unlocks extra newsletters, articles, discounted offers, and much more. Start advancing your knowledge today.
Unlock this book and the full library FREE for 7 days
Get unlimited access to 7000+ expert-authored eBooks and videos courses covering every tech area you can think of
Renews at $19.99/month. Cancel anytime
Banner background image