Most Helpful Customer Reviews
|
|
18 of 20 people found the following review helpful:
1.0 out of 5 stars
Lacking substance and focus, June 5, 2004
By A Customer
I work in a field where memory management is absolutely critical and was greatly looking forward to this book greatly. However it turned out to be a massive disappointment. The title is memory management, algorithms and implementations. However, the algorithms don't even really start until the last third of the book, and when they do, they are overly simple, and barely touch on the theory or critical characteristics behind the algorithms. This book is filled with useless tangents, some of the most random of which involving microkernal versus monolithic kernal, scanning of faces at the Super Bowl, and instructions for setting the 80x line of chips from real to protected mode! This book presents no practically useful algorithms, as all of the handling of cases such as growing the memory block are left as 'excercises for the reader' as are performance improvments even though final performance is essential in order to be able to properly compare and contrast the algorithms.In closing, skip this book and pick up either 'Operating System Concepts 6th ed' or 'Operating Systems: Design and Implemenation 2nd ed' which actually have meaningful insight into practical approaches for memory management and concerns.
|
|
|
10 of 10 people found the following review helpful:
5.0 out of 5 stars
Nuts and Bolts Perspective, February 22, 2003
By A Customer
While a lot of books on computer memory end up abstracting their discussion somehow, Blunden's book lies at the opposite end of the spectrum. All aspects of computer memory are examined (hardware level, OS level, application level) and in each instance concrete, non-trivial, examples are presented. Nothing, and I mean nothing, is left to the imagination. This book is nothing short of an exhaustive look at memory management. Engineers interested in this traditionally neglected topic should buy this book!Having read other reviews, I will agree that the material in the last chapter is just a little prophetic.
|
|
|
6 of 6 people found the following review helpful:
4.0 out of 5 stars
Developmental history and some examples of Code, July 21, 2004
This book, is one of the few computer books I have read from cover to cover - code listings are easy to scan/skip, and the book itself is not huge (however the cost seems fairly high for the size).
This book is clearly written, and well researched. It is not for beginners.
This book spends time on the hardware aspects of Memory management based on the Intel 386 and above architecture. Real Mode versus protected mode and how the processor design allows for memory protection in protected mode.
This is then followed by a survey of Operating Systems, from DOS to Linux to Windows - which is moving from the simple to the complex - and how the OS provides Memory Management services.
Then the development of computer languages, and how they allow for memory management starting with COBOL and moving on to Object oriented C++ and Java Virtual Machines.
Finally the last part of the book has a lot of code listings with very simplistic memory management and moving into slightly more complex algorithms for memory management for programs. The focus is on introducing multiple approaches and how to measure the real performance of each - some parts of this part of the book seemed like they were repeating the same text in making comments about the code.
Overall, I liked the book. I read the Pentium Protected Mode architecture book last year, and it prepared me for this book.
I have not done much assembly level x86 programming, but enough to understand what was being shown in the early examples.
The book has a very good bibilography of sources for each chapter - six long chapters. I felt, while looking at these bibilographies, that the time spent in going through all of the items in the bibliography was part of the reason that this books price was set so high. The references here are thorough and identify some turning points in computer sciences to me.
The structure of the book is the layers by which memory managment is accomplshed - the memory management code is more of an intro. I felt that the book might be targeting Computer Science courses as its real market - the books structure lays down a good foundation for further exploration.
The deeper development of Memory Management algorithms is where I too found the book lacking, and the title a bit deceiving - a complaint from another that I would agree with. This is the reason for not giving 5 stars. It would have taken a lot more time to develop the code and write explanations for more complex approaches and the author states this repeatedly in this algorithm section of the book.
I am inspired from this book to learn more about x86 assembly language, and to study the actual code of the Linux Kernel.
If that sounds of interest to you too, then you should buy this book.
|
|
|
Most Recent Customer Reviews
|
by Michael C. Daconta
by Bill Blunden
by Frantisek Franek
by Eldad Eilam
