A Formal Approach to Secure Speculation

Abstract

Transient execution attacks like Spectre, Meltdown and Foreshadow have shown that combinations of microarchitectural side-channels can be exploited to create side-channel leaks that are greater than the sum of their parts. While both hardware and software mitigations have been proposed against these attacks, provable security has remained elusive.This paper introduces a formal methodology for enabling secure speculative execution on modern processors. We propose a new class of information flow security properties called trace property-dependent observational determinism (TPOD). We use this class to formulate a secure speculation property. Our property formulation and associated adversary models help formalize the class of transient execution vulnerabilities. We demonstrate applicability of our methodology by verifying secure speculation for several illustrative programs.

Automatically eliminating speculative leaks from cryptographic code with blade

“…Our operational semantics formalizes the execution of source programs on a pipelined processor and thus enables source-level reasoning about speculation-based information leaks. In contrast to previous semantics for speculative execution Cheang et al 2019;Guarnieri et al 2020;McIlroy et al 2019], our processor abstract machine does not operate directly on fully compiled assembly programs. Instead, our processor translates highlevel commands into low-level instructions just in time, by converting individual commands into corresponding instructions in the first stage of the processor pipeline.…”

“…Speculative Execution Semantics. Several semantics models for speculative execution have been proposed recently Cheang et al 2019;Disselkoen et al 2019;Guanciale et al 2020;Guarnieri et al 2020;McIlroy et al 2019]. Of those, ] is closest to ours, and inspired our semantics (e.g., we share the 3-stages pipeline, attacker-supplied directives and the instruction reorder buffer).…”

“…Spectector [Guarnieri et al 2020] and Pitchfork ] use symbolic execution on x86 binaries to detect speculative vulnerabilities. Cheang et al [2019] and Bloem et al [2019] apply bounded model checking to detect potential speculative vulnerabilities respectively via 4-ways self-composition and taint-tracking. These efforts assume a fixed speculation bound, and they focus on vulnerability detection rather than proposing techniques to repair vulnerable programs.…”

Hyperfuzzing for SoC security validation

“…Many other security properties can be expressed in HyperPLTL. While we cannot describe them in detail due to limited space, a few examples are unique program execution checking [15], speculative non-interference [22], secure speculation [9], security of authenticated load [34], enclave measurement injectivity [44] etc. Note that most of the above properties involve only two traces, therefore in the rest of the paper the presentation will assume 2-trace properties.…”

Verification of Quantitative Hyperproperties Using Trace Enumeration Relations

“…Recent years have seen significant progress in automated and semi-automated techniques for the verification of security requirements of computer systems [4,10,16,19,30,47,50,55]. Much of this progress has built on the theory of hyperproperties [21], and these have been used extensively in analysis of whether systems satisfy secure information flow properties [1,2,6,8,15,28,35,37,39,49,57] such as observational determinism [41,55] and non-interference [32]. Unfortunately, the security specification of several important security primitives cannot be captured by secure information flow properties like observational determinism.…”