kernelthread.com

contents

Art | Automotive | Computing| Résumé|

meta

book

adsense

The Mac OS X Expert Challenge — 2005.1

© Amit Singh. All Rights Reserved.

The Background Current Challenge Status >>>

I acquired my first Macintosh computer on April 1^st, 2003, thus recently completing two years as a Mac OS X user. Having spent over ten years researching and studying a variety of operating systems, I found Mac OS X to be rather interesting subject matter. It differs from other prevalent systems on many fronts: architecture, philosophy, industry-perception, as well as the size, techno-culture, and nature of the user-base. One aspect that I have no feel for, but have always wanted to somehow quantify, is the system-level expertise spectrum of the Mac OS X user-base.

On a related note, there has been considerable hue and cry recently regarding the purported security and potential insecurity of Mac OS X, with hardly any logically plausible arguments made by any of the camps involved.

Along these lines, I thought of offering a small challenge to the Mac OS X community. "Challenge" is perhaps too strong a word, but I hope it kindles the interest and enterprise of the targeted audience.

The problem posed here simulates a scenario that a Mac OS X system or security expert may face, say, when dealing with malicious or potentially malicious software of unknown origin. "Dealing with" could mean several things, such as:

• Approximating the extent and nature of damage caused by malware (after the fact)
• Defending against existing malware
• Designing and developing generic or "future" defenses against yet unknown malware

In particular, a "security company" — one that creates or sells security solutions (especially host-based, but perhaps even network-based) for Mac OS X — is expected to to be able to tackle such problems and much harder ones.

Finally, I believe the problem posed may also be within the realm of some power-users with programming experience.

The challenge encompasses two partially overlapping areas of expertise:

1. Operating System Internals (to a greater extent)
2. Security (to a lesser extent)

My goals for this endeavor are the following:

• Probe popular interest in system-level Mac OS X topics. Knowledge of such interest is currently valuable to me as I am creating a book on such topics.
• Gauge the initiative and inquisitiveness of the Mac OS X community based on the kind of response generated.
• Use the outcome to roughly quantify, if possible, Internet-wide Mac OS X expertise outside of Apple.
• Facilitate sharing of Mac OS X knowledge.
• At the very least, provide an interesting problem for some people to solve.

The Problem

The problem involves a command line program called "panpipes", which has the following key properties:

• It is a user-space program.
• It runs with normal privileges (that is, does not require "root" access).
• It causes a kernel panic.
• It has been created on Mac OS X 10.3 "Panther". It may produce the same result on newer versions too.
• It uses some crude and intentionally weak measures to cloak its operation. I would rate the difficulty of defeating these measures to be 4 on a scale of 1 to 10, with 10 being the most difficult. My rating is primarily based on my having analyzed more sophisticated measures on other platforms, and the feasibility of similar measures on Mac OS X.

The Challenge

1. Analysis: Describe briefly what "panpipes" does to cause the kernel panic, including how you determined its operation. All other things being equal, a more satisfactory description will be one that includes the determination and deciphering of the cloaking measures used by "panpipes". You must specify the total amount of time you spent in analyzing the problem and describing your analysis. This is only for statistical reasons — whether you took 60 seconds or 60 days will have no explicit bearing on the merit of your submission. However, time is implicitly critical since the first satisfactory entry wins, as described in the rules.
2. Example: Provide your own user-space program — one that is as simple as possible — that triggers the same kernel panic as "panpipes".
3. Fix: Propose a reasonable fix for the Mac OS X flaw that is responsible for this behavior. With the fix implemented, "panpipes" must not be able to trigger a kernel-panic. Note that you do not actually have to implement the fix.

The Reward

1. The winner's solution and analysis will be published in an appropriate section of kernelthread.com.
2. The winner will receive a complimentary copy of my forthcoming book on Mac OS X internals, which is slated for publication (Addison-Wesley) in the second half of 2005. If the winner so desires, I will sign the book.

The Rules

1. Submissions must be emailed to challenge@kernelthread.com.
2. Submissions must be in English.
3. The challenge is not open to Apple employees or their family members. I expect this to be a trivial problem for the relevant people at Apple, and in any case, a goal of this endeavor is to gauge interest and expertise in Mac OS X outside of Apple. However, the challenge is open to friends of Apple employees, provided the former do not receive help of any kind from the latter.
4. Submissions will be evaluated by me (Amit Singh).
5. The sender of the first email that I receive with a satisfactory solution will be the winner.
6. The definition of "first email" will be based on my mail server's perception of time.
7. The definition of "satisfactory solution" will be entirely up to my discretion, and will be binding and final.
8. More than one person may choose to work together as a team. If a team is the winner, each of its members, up to a maximum of five members, will receive an individual copy of the book.
9. The challenge has no explicit deadline, but it will implicitly end as soon as I have accepted a solution as satisfactory, and all subsequent submissions will not be considered. In this sense, the time taken to solve the problem matters. I will make every attempt to announce the challenge's termination as soon as possible. However, depending on the number of satisfactory solutions received after termination and before the result is published, I may post a list of all those whose solutions were satisfactory. Please see below for deadline update.
10. I reserve the right to end the challenge at any arbitrary time for any arbitrary reason, even if no winner exists.
11. Use this discussion thread on Kernelthread Forums for seeking any clarifications regarding the challenge, unless you have good reason for not using a public forum.
12. No purchase necessary.

Challenge Status

• The challenge is CLOSED.
• The challenge was open for 6 days and 15 hours.
• The challenge page was viewed 8030 times during the challenge.
• The "panpipes" program has been downloaded times. It was downloaded 1071 times during the challenge.
• There were 5 submissions.

Result

Result Announced

All contents of this site, unless otherwise noted, are ©1994-2006 Amit Singh. All Rights Reserved. Terms of Use

Hosted at Jaguar Technologies