Lab 1_1B - AES Loop Skip Fault Attack in Practice with CW Husky

Embedded Security Glitching
1

Hi,

I’m quite a beginner, so I might be overlooking something obvious.

Recently, I’ve been trying to perform a clock glitch using ChipWhisperer Husky with cw308_stm32f303, and I’ve been following the ChipWhisperer tutorial in Jupyter Notebookcourses.

I got stuck on Lab 1_1B, so I would like some advice.

I made the following changes:
・scope.clock.adc_src = “clkgen_x1” to scope.clock.clkgen_src = “system”
・glitch parameters set, scope.glitch.clk_src = “pll”, scope.glitch.width = 2000, scope.glitch.offset = 1000
( I add “volatile” ~aes.c too)

However, although the message “Glitched at 339” was displayed, I intuitively feel that the waveform doesn’t look like it has a glitch.

Additionally, the key_guess returned an empty bytearray.

Could you tell me if this tutorial is executable on Husky and whether there are any mistakes in my modifications or understanding?

2

I’d lean more towards the lab being out of date. This one is pretty dependent on how the compiler chooses to order the code and it’s been quite a while since I’ve had the chance to run through the lab. I’d recommend just skipping this lab if you can’t get it to work, 1_3 and 2_1 are the more important labs anyways.

3

Hi, Alex.

Thank you for the advice.
I think checking the binary or assembly would take quite a bit of time, so I’ll follow your suggestion and skip it for now. If it becomes necessary later, I’ll come back to it.

I really appreciate your help!