Customer Reviews

Developing Drivers with the Windows Driver Foundation (Pro Developer)

5 star:		(1)
4 star:		(3)
3 star:		(2)
2 star:		(1)
1 star:		(2)

 

 

This book does exactly what it says, it provides a practical, sample-oriented introduction to developing drivers the Microsoft Windows Driver Foundation way.

The driver code for the samples used in the book, tools needed for developing drivers, and reference documentation are all downloadable (all 2.5GB of it, but it's free) from Microsoft. If you're like me and spend only a small part of your time working on drivers (I'm trying to interface a USB gadget), this is a great guide to WDF as well as to Windows I/O techniques and interface best practices. To get started, you can just hack the samples provided, as the authors intend. WDF looks after plug-n-play and power management, so it makes it easy to develop a basic user-mode USB driver like mine.

If you're a driver specialist, are writing kernel drivers, or have drivers to port from a different operating system, then the book is a detailed reference for moving to WDF. There's a lot of abstraction in the Windows way of doing drivers, and understanding the abstractions helps you write and debug your driver, so this book does a comprehensive job of explaining the relevant abstractions as you go along.

For example, if you're already an expert in the COM programming model, so that it's obvious to you why you need to implement the IUnknown methods, then you can likely skip most of Chapter 18. For the rest of us, we need the how-to advice and the examples, so there's a good reason the book is close to 900 pages :).

The content of the book feels more accessible than the online WDK documentation. It does cover the material, but each chapter is divided into three parts: stuff common between the kernel driver framework and user mode driver framework, stuff about the kernel driver framework, and stuff about the user mode driver framework. The authors probably had a hard time organizing the material, but the book should have been structured into those three parts. For example, I'm not currently interested in developing a user mode driver and I found the user mode driver material distracting.

This book is more reference than how-to. Maybe the authors should have structured the book like some of the Linux driver books: develop a real device driver.

I found this book well organized and useful for learning the Windows Driver Foundation. This book, however, is not a complete book for learning Windows Driver Development. Unfortunately, the book makes the claim that it is for newbies. This cannot be the case, because it doesn't give much in-depth information about Windows driver and kernel concepts, such as how memory is described (Neither I/O, buffered I/O or MDL's), different execution contexts, IRQL levels and what can and cannot be done in these levels, and basic IRP and I/O Manager concepts. Beginners will still have to start by learning WDM from a book like "Programming the Windows Driver Model". You just can't expect to succeed using WDF if you don't first have a firm grasp on WDM.

I think this book provides an organized approach to learning WDF. It is, however, not a book for leaning Windows driver development basics. I actually thought the book read very well and I'm not a fast reader at all.

A great disappointment, nothing new compared to what is on MSDN.

What is needed are simpler examples to start with, not hundred of pages of references: what are the bare necessities (I insist bare) to write a UMDF Filter driver (the simplest one) for a very basic device these days : a USB device. And there let's take an every day device (who has this FX2 stuff???) : a USB stick!

Also no answers as to what this UMDF Filter driver can do (can it do standard I/O like fprintf ?).

This book may very well describe writing WDF drivers, but it is not necessarily a useful book. It is not the Windows equivalent of the one Writing Linux Device Drivers book, and even that book is only just so useful. First: User-mode drivers are described, but what good are they? They can't do a _lot_ of things, and they are source-code incompatible with kernel-mode drivers. Second: In the book's Forward, a Microsoft "Architect" mentions that 3rd party driver developers find the pre-WDF Windows driver model "complex and difficult to use". Unfortunately neither WDF nor this book has helped me debug real world issues involving Cardbus, inf/driver install failures, and NDIS API failures. Microsoft did _not_ address/document/fix the known "complex and difficult to use" problem.

for those do not have previous driver dev. experience. Too much documentation style "teaching" without a step-by-step practicing process.

It is very good for a beginner to WDF. I mean it starts from the very beginning introducing you to driver development concepts.

If you are stuck making a driver for the lame horse called vista, then this book gets the job done, it is very dense, and has a very microsoft feel to it. The WDM has been abstracted to such a point of obscurity at this point that... well I dont have an analogy, it just makes your head hurt, it has definitely advanced to the .net cobweb model of idiocy at this point. Not to mention that you have next to no control over how windows really decides to interact with your hardware anymore with its compiled/interpreted 9 times before it eats up all of the system resources horse manure. Me and a team of 5 other people fought with this stupid driver model for 9 months before we convinced the company to let us use an embedded linux host that feeds a dummy windows GUI the data for output. If you are going to work on a piece of equipment that is not entirely Microsoft endorsed IEEE I suggest doing the same thing, because at this point, if microsoft does not have it prewriten for you, your SOL!!!
I could say a lot more, but it will do no good just like the rant I just went on, if you need to write a standard driver, you should be able to get the job done with this book, I really only took off one star because it was published by microsoft... I hope developers and engineers continue to get disgusted with them, and they either correct there arrogant and ignorant ways, or they fall by the way side, because this is getting ridiculous.

I took a job that required me to make device drivers for some kind of weird thing with some blinking lights on it. I never made a driver before, but I am good at using Office and stuff so I told them I knew C++ and read a bit about it and was able to BS the interview. So I got this book and tried to understand how to make my drivers, but I couldn't and I got fired after 6 weeks. NOT RECOMMENDED for driver makers.