The Rootkit Arsenal 79
Nicola Hahn writes "One of the first things I noticed while flipping through this hefty book is the sheer number of topics covered. Perhaps this is a necessity. As the author puts it, rootkits lie "at the intersection of several related disciplines: computer security, forensics, reverse-engineering, system internals, and device drivers." Upon closer inspection, it becomes clear that great pains have been taken to cover each subject in sufficient depth and to present ideas in a manner that's both articulate and well organized. This accounts for the book's girth; it weighs in at roughly 900 pages." Keep reading for the rest of Nicola's review.
This book is comprehensive enough to appeal to both novices and journeymen. To set the stage, the Rootkit Arsenal begins with a review of foundation material: the IA-32 execution environment, memory management, kernel-mode subtleties, call hooking, detour patching, and so forth. Yet, while the author devotes a significant amount of effort to explaining prerequisites and customary rootkit techniques, there's an abundance of more sophisticated content to engage more experienced members of the audience. For example, his explanation of how to use the WSK API and the most recent incarnation of the NDIS library (version 6.0) to construct covert channels over DNS is worth a read. I also appreciated his meticulous discussion of how to properly install Call Gates and handle the foibles of multi-processor systems.| The Rootkit Arsenal | |
| author | Reverend Bill Blunden |
| pages | 916 |
| publisher | Wordware Publishing |
| rating | 5 Shuriken |
| reviewer | Nicola Hahn |
| ISBN | 1598220616 |
| summary | A solid treatment of rootkits and anti-forensics |
One of the book's strong points is that there's coverage of issues which traditionally haven't appeared in books on this subject. For instance, there are several sections devoted to the Windows startup process and how it relates to the operation of bootkits. Part 3 of the book, which consists of four chapters, focuses on anti-forensics, with an emphasis on defeating file system analysis and the examination of an unknown executable. To this end, Reverend Bill ventures off into the tactics used to implement binary armoring, FISTing, obfuscation, code morphing, file scrubbing, and data contraception.
Not content to merely explain the basic mechanics of a particular scheme, Reverend Bill often illustrates how he derived his results and encourages the reader to verify what they've seen with a kernel debugger. This is a recurring theme throughout the book. Rather than just teach the reader a collection of tricks, the author demonstrates how the reader can identify new ones independently. After all, specific holes come and go, but the art of finding new ones will always have utility. This more than justifies the lengthy discussion of kernel debugging earlier on in the book.
All told, the book is reasonably self-contained. The source code examples are clean, instructive, and have been included in the book's appendix. As Reverend Blunden notes, the "Rootkit Arsenal" isn't about a specific rootkit that someone wrote (though such books exist). It's really about the rootkit that the reader will construct, such that the focus is on the nature of the tactics rather than a proof-of-concept rootkit. In this spirit, examples are long enough to illuminate potential sticking points but not so long that the reader feels like they're wading through mud in search of diamonds.
The author also exhibits good form in terms of giving credit where it's due. In the book's preface he specifically acknowledges a number of researchers who have made lasting contributions to the collective repository of knowledge (Mark Ludwig, Greg Hoglund, the grugq, Sven Schreiber, Joanna Rutkowska, Richard Bejtlich, etc.). While the author admits that many of the book's ideas can be unearthed by skulking about obscure regions of the internet, the real service that this book provides is to consolidate all of this disparate information together into one place, offering working implementations of each concept, and doing so in a remarkably lucid manner.
Yet, is this a responsible thing to do? Is it wise to show aspiring Black Hats how to manipulate forensic evidence so that they can implicate innocent people? Will publicizing the finer points of system modification make life easier for aspiring bad guys? Is he basically handing the reader a loaded gun and teaching them the nuances of a kill shot?
To a degree, Reverend Blunden sidesteps this issue as irrelevant. In the end he claims that he's just a broker of information, and that he doesn't care who uses the information or how they use it. If you asked me, this is a bit of a cop out (he sounds a little like an arms dealer). Furthermore, he accuses other authors (the ones who fall back on the traditional argument that they're bolstering security by encouraging vendors to improve their products) of churching up their books in "ethical window dressing." In the eyes of Reverend Bill, this book is what it is ...without apology: another source of useful data.
If I had one complaint about the Rootkit Arsenal, it's that the author sticks primarily to software-based rootkit technology. For instance, he eschews BIOS-based tools. At one point the author states:
"In my opinion, a firmware-based rootkit is essentially a one-shot deal that should only be used in the event of a high-value target where the potential return would justify the R&D required to build it. Also, because of the instance-specific nature of this technique, I'd be hard pressed to offer a single recipe that would a useful to the majority of the reading audience. Though a firmware-related discussion may add a bit of novelty and mystique, in the greater scheme of things it makes much more sense to focus on methods which are transferable from one motherboard to the next."
Last but not least, the author's tendency towards the political arena, which defined a couple of his previous books, rears its head again in The Rootkit Arsenal's final chapter. Here, the good Reverend suggests that if it's possible to control a sprawling operating system like Windows with a relatively small rootkit binary, then perhaps the metaphor carries over into the body politic of the United States. Could a small segment of the population be quietly influencing the trajectory that society takes? Dave Emory and Noam Chomsky look out!
Readers interested in getting a closer look at the book's organization and table to contents can visit the author's web site.
You can purchase The Rootkit Arsenal from amazon.com. Slashdot welcomes readers' book reviews -- to see your own review here, read the book review guidelines, then visit the submission page.
Windows needs a root-kit-cleaner CD (Score:2, Interesting)
Let me rephrase that:
Computers should ship with an "alternative" boot environment that cannot be permanently changed, only toggled to and from the main boot environment.
The job of the alternative boot environment is to allow cleanup tools to delete threats.
An example of how this could be done in Vista:
Boot computer using a back-up, read-only firmware to a Vista CD that had a stripped-down network stack or stripped-down USB-drivers. Having stripped-down software removes some points of vulnerability. From the clean BIOS+Vista boot, load and authenticate security modules. These can be loaded from a web site or external media. The authentication is key: If it's not authenticated it's rejected. The authenticated security modules would then clean up the system as best they could, and would run a heuristic analysis on the non-booted environment to look for remaining suspicious behavior, such as the loading of unsigned device drivers or a BIOS that contains non-authenticated patches.
Why Vista? It's not the best technical solution but in a year or two it will be the most familiar bootable CD out there.
As a side bonus, a similar "clean boot environment" can be used for web-access kiosks. However, these would need a richer network stack, a web browser and plugins, and would need to be re-created almost daily to keep up with security threats. An immutable BIOS with a CD that loads, authenticates, and runs a "boot image" over the network, with a daily reboot to grab the freshest image, might be the way to go here.
Re:not pleased with this review (Score:5, Interesting)
I really would prefer that Slashdot not help promote a book that is intended to help educate would-be malware coders.
So then they should never post any reviews of any books on the topic of security? Pretty much any book that is going to teach you anything of worth in the are of security is going to have information to help those who want to write malicious code.
While I realize the information would still be out there without this review, Slashdot is undoubtedly contributing sales to the author by raising the profile of this book.
So we should just put our fingers in our ears and shut our eyes and pretend it doesn't exist? Yeah that's going to do all of jack and shit.
Re:Is this a responsible thing to do? (Score:3, Interesting)
Besides, not publicizing this information amounts to security through obscurity [wikipedia.org]. Nowadays, all security experts with any credibility consider obscurity to be the opposite of security, at least with respect to computer systems. If a vulnerability exists, some malware author will find it, no matter how many nooks and crannies need to be poked into. Even if there are million nooks and crannies, it's easy to automate the search!
I gotta wonder at the reliability of an author who insists on using his affiliation with a quasi-satirical religion [subgenius.com] as if it were a professional qualification!
I also find it very scary that I have to read 900 pages to become properly acquainted with just one particular kind of malware! Hmm, maybe you do need to be a Dobbshead to deal with that.
Scary (Score:2, Interesting)