Laser-induced Single-bit Faults in Flash Memory: Instructions Corruption on a 32-bit Microcontroller

Abstract

Physical attacks are a known threat posed against secure embedded systems. Notable among these is laser fault injection, which is often considered as the most effective fault injection technique. Indeed, laser fault injection provides a high spatial accuracy, which enables an attacker to induce bit-level faults. However, experience gained from attacking 8-bit targets might not be relevant on more advanced micro-architectures, and these attacks become increasingly challenging on 32-bit microcontrollers. In this article, we show that the flash memory area of a 32-bit microcontroller is sensitive to laser fault injection. These faults occur during the instruction fetch process, hence the stored value remains unaltered. After a thorough characterisation of the induced faults and the associated fault model, we provide detailed examples of bitlevel corruption of instructions and demonstrate practical applications in compromising the security of real-life codes. Based on these experimental results, we formulate a hypothesis about the underlying micro-architectural features that explain the observed fault model.

An End-to-End Approach for Multi-Fault Attack Vulnerability Assessment

“…• The fault analysis will miss potential implementation weaknesses. Recent target specific fault models [4], [5], [6], bypass software countermeasures in a single fault. • This challenge becomes even more crucial for multifault attacks analysis as there will be more unnoticed vulnerabilities as well as more unexploitable attacks because of the combinatorial explosion of possible attack paths.…”

Experimental Analysis of the Laser-Induced Instruction Skip Fault Model

“…Instruction modification achieves an instruction skip if the modified instruction has no effect on the code operations (instruction skips are often actual code modification); the same is true for turning an instruction into a nop. An analysis of single instruction skips due to instruction modification induced by laser is reported in [8]. It relates how faulting one bit of an instruction opcode led to two successful fas.…”

“…Laser-induced fault injection may be used to force the identification of an attacker using a wrong PIN. As illustrated in [8], where a laser-induced modification of one instruction of a similar PIN algorithm permits to perform a PIN bypass. [9] describes, on simulation basis, several instruction skip attacks that may result in a successful PIN bypass: most of them targeting one or a few successive code instructions.…”

“…Denomination of the fault model The laser-induced instruction skip fm relies on two possible mechanisms: (1) injecting faults into an instruction opcode [8,11] that turn it into another instruction, or (2) replacing it by a nop instruction (this work). In neither case is the instruction really skipped over.…”

“…This suggest that a similar mechanism may be at work, even though they did not test the feasibility of successive instruction skips. The authors of [8] also induced single instruction skips through laser illumination into a cortex-M3 microcontroller by faulting a single bit of the targeted instruction (hence inducing an instruction modification equivalent to an instruction skip because the modified instruction had no effect on the executed test code). The physical mechanism they revealed as the root cause of this instruction skip (a laser-induced discharge of a bitline of the Flash memory) appears compatible with induction of several skips if a long laser pulse is used.…”