[Previous entry: "Ellch and Maynor: the continuing debate"] [Next entry: "Sinking slowly beneath the sea"]

10/02/2006: "More on Ellch/Maynor"

So, last Sunday (not yesterday, but 9/24/06) Maynor calls me on my cell phone. Asks if he can put Johnny Ellch on the phone too. I agree, but warn that I'm taking my family to the beach (I'm teaching Hunter to surf), so my time will be limited.

They explain their version of the story, twice, and then ask if I have any questions.

I had but one, really.

"Does the bug you say you found exist in the FreeBSD ath driver or the madwifi driver?"

Simple, I thought.

Maynor answered that their "bug" does exist (in both), though they would need to write a different exploit for either. I presume this is because they're into smashing the stack or heap (via ETX), and things are different on FreeBSD, than on linux, than on MacOS X/mach, etc.

We politely end the conversation, after Maynor says he'll send me a copy of the exploit code, so i can verify for myself that their bug worked. (I've been keeping the mini away from the updates in order to test their code.)

Well, Friday came, and went, and as everyone knows, No Talk, and thus, I presume, No Code for Jimmy to play with.

And thats all fine... until you know the rest of the story.

Ya see, its quite clear that Maynor and Ellch *know* who wrote the FreeBSD ath driver and net80211 layer, it was Sam Leffler. Sam was also, until quite recently, responsible for the majority of the linux madwifi driver that is the close cousin to the FreeBSD bits.

Maynor and Ellch also know that the FreeBSD ath driver and the net80211 layer are the very heart of Apple's Atheros driver. They took pains to explain this to me, even though I already knew it.

But here's the interesting part.

I asked Sam what he thought of the Maynor/Ellch/Apple/... situation. Sam is obviously close to the situation and while he might be limited in what he is able to say about anything that involves Apple, he's quite free to discuss any issues that surround the linux madwifi driver and the FreeBSD ath / net80211 drivers.

I'll just quote the man:

My only remark about Ellch/Maynor is that I do not understand why
they've not contacted me as the author of the code. Feel free to pass
on that remark should you talk to them again.


I find this both telling and troubling.

Maynor and Ellch told me that their bug also exists in the FreeBSD ath / net80211 driver(s).

Sam says they haven't contacted him.

Maynor and Ellch think there is a bug in FreeBSD.
Maynor and Ellch know how to contact the author of the affected code.
They've failed to do so.

The other shoe drops.