Customer Reviews

The Undocumented PC: A Programmer's Guide to I/O, CPUs, and Fixed Memory Areas (2nd Edition)

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

 

 

This book provides excellent information on hardware ports, CMOS, disk and video systems. It explains the differences between versions, lists known bugs and workarounds and even comes with functioning code examples. Some of the programming leaves a little to be desired, though. :) But if you're a good assembly programmer you'll understand it.

Overall it is an excellent book and a must have for anyone wanting direct hardware-level access (without the save-haven "buffer" provided through an API). This book is for real programmers only. Programmers who want to work at the machine level. It is, quite simply, the most comprehensive book I've found. A one-stop shop for information.

The only problem I found is that this book is slightly dated, originally published in 1996 or 1997. Newer features are not included.

This is the most thorough technical book I've ever read, on a subject only glanced at in other books. You will find every detail about legacy PC hardware and low-level programming for these systems in this book, and it is well organized for low-level system programmers.

PC hardware areas are each given a chapter with an introduction explaining the history and workings of the physical hardware, followed by interrupt and port programming details. There are also chapters devoted solely to the interrupt vector table and BIOS data areas that list all software services, and refer to the other chapters for more details...you can get a good overview of a particular area of PC hardware, or you can just as easily find the particulars of any interrupt service, port usage or BIOS data area entry. Invaluable for low-level programmers!

This book, though having some misprints (minor), is invaluable for those willing to squease every last drop from their PC when writing low-level programs.

I found this book to be something of a disappointment, for two reasons. Firstly, in large part it just documents bits, and doesn't anticipate what's important. And secondly, it is too often wrong on some pretty important points.

An example of the first problem would be the descriptions of INT 9 and INT 15h fcn 4Fh. After the PC/AT, the only safe way to look at a keystroke within a TSR, and then pass it on if it is not a keystroke for the TSR, is to intercept using INT fcn 4Fh -- if a TSR uses INT 9, the keyboard is free to replace the value in the keystoke buffer with the next keystroke as soon as it was read by the TSR, and "passing on" the INT 9 interrupt means that the "passed to" application may never see the keystroke that the TSR first intercepted. This book miss that aspect (which is the reason why it was added) of INT 15h fcn 4F completely.

An example of the second problem is the books description of the error 9 for disk operations. For DMAed disks, this error occurs because the DMA address incrementer doesn't/can't handle a carry out of bit 15 of the address, and hence no operation DMA can crosss a 64K absolute REAL memory address boundary. The book describes it as an error associated where the segment:offset form of the address wrapping within the segment, which is an entirely different thing.

These were a couple of the first things that I checked, and after finding his treatment of these very much lacking, I'm inclinded not to trust the author's interpretation of things very much, particularly when there is subtlety involved.

But the book does have a wealth of detail which is hard to find elsewhere, and on that score, it seems hard to beat.

A more accurate title may be "The Under-Documented PC." There seems to be a lot of information about the PC on a higher level, but very little is applicable to the programmer that wants to access the hardware directly or at the BIOS level. This book fills much of the information gap.
The revised, second, edition has been improved in several areas while retaining the overall format of the original volume.
Highly recommended for those who write operating system and similar software for the PC platform. Also valuable for those who want to know more about "what's under the hood."

This is a great book, or I should say WAS a great book. While the information is still 100% applicable if you are still writing code for a 16-bit operating system (DOS), or writing your own operating system, much of the information becomes somewhat useless when applied to 32-bit operating systems (Windows 2000, Linux).

Still, when writing low-level code it is still a useful reference, one of the few out there.

Gilluwe includes source code samples, and is one of the most comprehensive books on the PC architecture.

Unfortunately, this book is now somewhat dated. The BIOS is no longer the de-facto method of controlling the hardware, and is a big headache under 32 bit protected operating systems (Linux, Windows 2000, etc). Newer features, like PCI configuration space, NVRAM, BIOS 32 directory services, DMI and SMBIOS are not covered.

Of all of the books i have read on the low level aspects of computers this book is by far the most complete. This book has none of the typical "unknown/unused" grey areas found in most books in this area. A must buy for any low level program!

This book is a great book. It covers low level details from the Hard Disk Controller to understanding of the DMA chip. This is a good book for systems programmers. This is my new programming bible.