No. Just look the 2.1 section.
I'm very sorry to read the paper almost throughout with my poor ARM hardware knowledge.
The main clock speed of LPC1768 (Cortex-M3) which mentioned in the paper is 100MHz. However the target CPU in the paper is Marvell Kirkwood 88F6281 (ARM926), which has a maximum main clock speed as 1.2GHz, while there is 88F6282 can archieve 2.0GHz in the family. That sounds good isn't it?
"Our approach neither requires hardware modifications nor expensive test equipment.", "The first step is to stop the SUT by sending a debug request via JTAG. At this point, the OCD takes control over the CPU." However 3DS's MPCore didn't expose a JTAG. I'm getting disappointed.
Also about its section 2.1:
pin-level probes and sockets [ACL89][KF95] - ...
without contact by exposing the circuit to a particle beam [KF95][VKC+92][ELDF92] - Beam?
using lasers [PLF03] or to electromagnetic inferences [KF95][VCG+05] - Orz.
Well.. i recalled something special. (With the f**king cheap STM32 ARM Chips). Chips have a lowest possible working voltage. Lower than that, it would function abnormally. Also, there is a main clock source, or other crystals that generate the clock chips need (even feeded to PLLs lately). When you give it a much slower clock source, that slower than its PLL could adjust, it would have a inproper clock speed, or just function abnormally. I don't know if there are similar faults with the hardware that exploitable. Hope it isn't one including a source as MSP430.
--Too long to be quoted here--
I'm sorry to disturb you. NTR's debugger has some issues that related to multi-core support. I thought of BKPT or a modified HardFault that could trigger. However he is busy playing with iWatch or anything (doesn't matter), so it is much likely to be a finished one currently. I wouldn't ask you for helping him directly, but if you know some techinics that may solve this.. Yes even 44670 isn't a friend of mine.
Even i don't like close-sourced ones, i respect what you chose. I would say thanks to your work. And hope you play well with your research.
You do realize that free doesnt just mean that you dont have to pay for something right? Free as in all the signatures get patched, free as in full control of the OS without all that anti-piracy stuff, free as in full open source.... Jesus Christ
Be Free the "No-Fee" or "Freedom". So i don't consider NTR a Free(dom) Solution now. And well the code is within their hands, so open or not is their choice - at least you can not hit one of them with your fist and threaten him to open the code (yep). I would say nothing if one chose close-source.
Those are all Karma.