Crellvm: verified credible compilation for LLVM

Kang, Jeehoon; Kim, Yoonseung; Song, Youngju; Lee, Jun-Young; Park, Sanghoon; Shin, Mark Dongyeon; Kim, Yonghyun; Cho, Sungkeun; Choi, Joonwon; Hur, Chung-Kil; Yi, Kwangkeun

doi:10.1145/3192366.3192377

Cited by 17 publications

(11 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Until recently, building formally verified compilers for realistic languages was seen as a grand challenge. However, the advent of the CompCert [Kästner et al 2018;Leroy 2006Leroy , 2009aLeroy et al 2016], Vellvm [Zhao et al 2012[Zhao et al , 2013, Crellvm [Kang et al 2018], and CakeML [Kumar et al 2014;Lööw et al 2019;Owens et al 2017;Tan et al 2016] verified compilers show that they are now well within reachÐalthough admittedly, every formally verified compiler for a realistic language remains a feat.…”

Section: Introductionmentioning

confidence: 99%

Formal verification of a constant-time preserving C compiler

Barthe

Blazy

Grégoire

et al. 2019

Proc. ACM Program. Lang.

View full text Add to dashboard Cite

Timing side-channels are arguably one of the main sources of vulnerabilities in cryptographic implementations. One effective mitigation against timing side-channels is to write programs that do not perform secret-dependent branches and memory accesses. This mitigation, known as łcryptographic constant-timež, is adopted by several popular cryptographic libraries. This paper focuses on compilation of cryptographic constant-time programs, and more specifically on the following question: is the code generated by a realistic compiler for a constant-time source program itself provably constant-time? Surprisingly, we answer the question positively for a mildly modified version of the CompCert compiler, a formally verified and moderately optimizing compiler for C. Concretely, we modify the CompCert compiler to eliminate sources of potential leakage. Then, we instrument the operational semantics of CompCert intermediate languages so as to be able to capture cryptographic constant-time. Finally, we prove that the modified CompCert compiler preserves constant-time. Our mechanization maximizes reuse of the CompCert correctness proof, through the use of new proof techniques for proving preservation of constant-time. These techniques achieve complementary trade-offs between generality and tractability of proof effort, and are of independent interest. CCS Concepts: • Security and privacy → Logic and verification.

show abstract

Section: Introductionmentioning

confidence: 99%

Formal verification of a constant-time preserving C compiler

Barthe

Blazy

Grégoire

et al. 2019

Proc. ACM Program. Lang.

View full text Add to dashboard Cite

show abstract

“…In order to simplify the work of translation validation tools, it is possible to extend the compiler to produce hints (witnesses) [18,36,38,41]. One of these tools, Crellvm [20], is formally verified in Coq.…”

Section: Related Workmentioning

confidence: 99%

“…This entails describing precisely what each instruction does and how it handles special cases such as integer overflows, division by zero, or dereferencing out-of-bounds pointers [8,24,26,29,47]. There has also been work on automatic verification of classes of optimizations, such as peephole optimizations [25,31], semi-automated proofs [48], translation validation [20,35,42,44], and fuzzing [23,46]. All this work uncovered several hundred bugs in LLVM.…”

Section: Introductionmentioning

confidence: 99%

An SMT Encoding of LLVM’s Memory Model for Bounded Translation Validation

Lee

Kim

Hur

et al. 2021

Computer Aided Verification

Self Cite

View full text Add to dashboard Cite

Several automatic verification tools have been recently developed to verify subsets of LLVM’s optimizations. However, none of these tools has robust support to verify memory optimizations.In this paper, we present the first SMT encoding of LLVM’s memory model that 1) is sufficiently precise to validate all of LLVM’s intra-procedural memory optimizations, and 2) enables bounded translation validation of programs with up to hundreds of thousands of lines of code. We implemented our new encoding in Alive2, a bounded translation validation tool, and used it to uncover 21 new bugs in LLVM memory optimizations, 10 of which have been already fixed. We also found several inconsistencies in LLVM IR’s official specification document (LangRef) and fixed LLVM’s code and the document so they are in agreement.

show abstract

“…Credible compilation [19], or witnessing compiler [16,17], is an approach to improve translation validation by accepting witnesses generated by a compiler. Crellvm [8] is a credible compilation framework for LLVM. It requires modifications to the compiler, which makes it harder to apply and maintain.…”

Section: Translation Validation and Credible Compilationmentioning

confidence: 99%

AliveInLean: A Verified LLVM Peephole Optimization Verifier

Lee

Hur

Lopes

2019

Computer Aided Verification

Self Cite

View full text Add to dashboard Cite

Ensuring that compiler optimizations are correct is important for the reliability of the entire software ecosystem, since all software is compiled. Alive [12] is a tool for verifying LLVM's peephole optimizations. Since Alive was released, it has helped compiler developers proactively find dozens of bugs in LLVM, avoiding potentially hazardous miscompilations. Despite having verified many LLVM optimizations so far, Alive is itself not verified, which has led to at least once declaring an optimization correct when it was not. We introduce AliveInLean, a formally verified peephole optimization verifier for LLVM. As the name suggests, AliveInLean is a reengineered version of Alive developed in the Lean theorem prover [14]. Assuming that the proof obligations are correctly discharged by an SMT solver, AliveInLean gives the same level of correctness guarantees as state-ofthe-art formal frameworks such as CompCert [11], Peek [15], and Vellvm [26], while inheriting the advantages of Alive (significantly more automation and easy adoption by compiler developers).

show abstract

Crellvm: verified credible compilation for LLVM

Cited by 17 publications

References 38 publications

Formal verification of a constant-time preserving C compiler

Formal verification of a constant-time preserving C compiler

An SMT Encoding of LLVM’s Memory Model for Bounded Translation Validation

AliveInLean: A Verified LLVM Peephole Optimization Verifier

Contact Info

Product

Resources

About