Customer Reviews

Writing Windows WDM Device Drivers

5 star:		(2)
4 star:		(2)
3 star:		(0)
2 star:		(3)
1 star:		(1)

 

 

I needed to write a Windows device driver to do parallel port I/O on NT machines and handle hardware interrupts. The DDK is incomprehensible to novices. The Viscarola / Mason book is a great reference, but is not a good how-to. Oney has lots of important information on lots of important topics, but you can quickly get lost in tons of details that don't apply to the task at hand. Chris Cant gets it right in his book. His pedagogical strategy is to actually create a very small device driver, and then study it as you layer on the complexity, and not hit you with it all at once. It exactly addressed my primary needs. Its drawback is that it is not very detailed, nor is it a good advanced reference, so the ideal solution is to get all three books, and use them each for their own strength. But start with Chris Cant's. (He also includes a couple of very useful utilities, one of which is a marvelous debugging tool.)

I purchased this book back in July when no other WDM books were available. Overall, I found the book didn't answer any of my concerns and was considerably lacking details for non-USB drivers. For example, there is no explaination on controling DMA transfers. All my NT driver books, have an entire chapter devoted to the subject! The PNP section was repetive but not clear. Overall, not a very good driver book.

get Oney's book instead - it's more readable and the examples actually work...

I've found this book chaotic, and not very informative. It doesn't cover many topics like DMA and Direct I/O mode. I think this book can be more useful after reading any book covering NT drivers in details.

Chris Cant writes (and draws!) in a user friendly sytle which makes the book easier to read than most. Along his way, he stops to note many important side points which other, more down-to-business books might not (e.g. noting that when running setenv under win98, you will first need to increase your environment size.) The chapter entitled "WDM Driver Environment" gives you a complete guide to setting up your WDM development systems for Win98 and Win2000, with a tremendous amount of detail - and when it comes to setting up, you can never give too much detail. Had I had that chapter when I first set up my systems to compile WDM drivers, I would have saved a good number of hours.

I found no useful information at all in this book. I developed a PCI device driver with DMA capabilities, and hoped that this book could give me some valuable hints on the subject. It did not. Nor did it give any clues on how to handle plug and play or power management. It looks as though the author just put some small WDM driver together, saw that it worked quite good, and then wrote a book about it.

From the point of view of some-one who had never written a Windows device driver I found this an excellent introduction to the problem domain. I found the material to be well organized and easy to follow. The blend of theory and sample code snippets was well done and added considerably to my understanding of the subject matter. The content is dated as it was written in the Windows 2000 era. However I found that I was better able to follow the content in newer text books after reading this book from cover to cover. I was disappointed that the companion CD with full sample drivers was not included nor could I find it on the internet.

Besides being able to open the book and find the exact topic that relates to W2K and WDM (not like those NT4 books that touch on WDM), it's nice to open the book and find a diagram that shows the differences and commonalities between NT and WDM drivers. It's when you can find the in depth details on the major blocks of the big picture and how they relate to each other and then realize that the little tid bits of information that you didn't have time to worry about in the past are all in one place (as in the differences and commonalities mentioned above)that a single book will become your main reference. Device drivers under Windows 2000 are not to be taken lightly. To comment on the previous negative review: I would say that for a person to read and find so many errors, when the book has only been released for a few days means that either they are a seasoned GURU and have nothing better to do than buy books to look for mistakes, or maybe they didn't quite grasp the complexities involved....Or it could be another author trying increase his own sales:)