Hello,
in a uni class about encryption, we coded a RSA encryption on a tiva (the Launchpad one: EK-TM4C123GXL). Part of that class is breaking our own encryption, using a bellcore attack.
We are inducing a small (adjustable) voltage drop on the power supply rail (for a few microseconds), using two MOSFETs (one turning the power supply off, the other pulling the chips VDD, VDDA and VBAT line low). The FETs are connected in place of R30 on the evaluation board.
We disabled the brown out detection successfully (uC keeps running, when voltage dip goes below 2.8V). The problem is, that its not enough to get the processor to make a mistake when doing the calculations (which is what we want to achieve) and as soon as we go lower than 2.7V drop, the controller resets itself.
I assume that reset is triggered due to the 'power ok levels' (see page 1365/ table 24-10 in the datasheet: www.ti.com/.../tm4c123gh6pm.pdf)
Do you know whether that assumption is sensible? And does anyone know, if the power ok checking can be disabled?
Is attacking that chip using a voltage glitch possible, or is it protected well enough to prevent this form of an attack?
Thanks alot for any replies and let me know, if theres any information lacking,
Sebastian
