Свен шрайбер недокументированные возможности windows 2000 издательство питер 2002 год

Библиографическое описание

I will start will a caveat — the name «Undocumented Windows 2000 Secrets» is clearly modeled on two books written by Matt Pietrek — «Undocumented Windows» and «Windows 95 Programming Secrets». Despite this deliberate title plagiarism, the style and content of Schreiber’s book is quite different to those of Pietrek. Pietrek concentrated on examining the Win32 API functions, disassembling them, exposing many functions not formerly documented, but used internally, and occasionally exploring the Kernel-mode modules which lie beneath the entire Win32 API. Schreiber has almost nothing to say about the Win32 API, exposes no new Win32 API functions, and shows disassembly for none of them.
Caveat aside, Schreiber has done some remarkable work here. Firstly, he knows how to write. Given the generally poor style of most programming books, which are either entirely incomprehensible and poorly edited, or banal, silly, and full of margin pictures etc, this books stands as a shining example of how to present difficult content. The English is flawless, and chapters follow a clear design and progression. Difficult material is presented as such, and I for one, enjoy Schreiber telling me how hard he worked on a certain problem, and how revolutionary is his generic solution. Assembly code is presented with an apology that it could not have been written in straight «C», and clearly explained.
He also avoids the pitfall of ignoring older systems. This book is clearly not about the 9x family, so I have no problem with him ignoring that, but he does an excellent job of pointing out that everything he writes is valid for NT version 4, except in a few cases that he highlights.
He is also very careful about not covering ground already adequately covered by others, and provides clear references to other books where relevant, for example Nebbett’s «Native API Reference» and Solomon & Russinovich’s «Inside Windows 2000».
So, what does he cover? Firstly, how to debug system-level programs, using low-level debuggers, and also shows the internal structure of Microsoft debug files. Then a review of the Native API — not a catalogue like Nebbett, but the mechanism of invoking them, and how the system instructions flow from one module to another.
Then comes a wonderful introduction to writing a simple Kernel Mode device driver. Anybody who has tried this will know that most device driver documentation is written for those already expert, and if all you want is a minimal driver to access some Kernel Mode function, you have to grope around in the dark vainly hoping for somebody to guide you. Well, here it is at last — a clear guide to writing simple device drivers, clearly explained, and even with an MSDev Wizard to whip up a framework! For this alone it is worth getting the book.
Using his device driver, he explores Win2000 memory, and most importantly, develops a wonderful generic mechanism for monitoring and intercepting all calls to the Native API — managed from a user-mode GUI application! He then builds a user-mode application that allows you to make a call to any Kernel function from the user mode, seamlessly calling down to the minimal device driver. Wow!
Finally, he gives a concise description of Kernel Objects, and covers the most important ones.
The appendixes are really huge catalogues of the various WindDBG commands, the Native APIs, and the structures on encounters in the Kernel Mode.
Do I have any criticism? Amazingly, none at all! And this is the first Amazon review I have written in which I have none!
In summary, if you are at all interested in understanding Win2000 internals, monitoring calls to the Kernel Mode, or making your own calls to privileged functions, this book is one of the most important, and most accessible books on the market. Also, the author is available on email — I contacted him, and within a day received a good response. So, after all, Schreiber may go down in history as Peitrek’s successor!