Laboratoire des systemes fiables

À propos de ce résultat

Cette page est générée automatiquement et peut contenir des informations qui ne sont pas correctes, complètes, à jour ou pertinentes par rapport à votre recherche. Il en va de même pour toutes les autres pages de ce site. Veillez à vérifier les informations auprès des sources officielles de l'EPFL.

Resource Sharing in Dataflow Circuits

Andrea Guerrieri, Paolo Ienne, Lana Josipovic, Axel Marmet

To achieve resource-efficient hardware designs, HLS tools share (i.e., time-multiplex) functional units among operations of the same type. This optimization is typically performed together with operation scheduling to ensure the best possible unit usage at each point in time. Dataflow circuits have emerged as an alternative HLS approach to efficiently handle irregular and control-dominated code. Yet, these circuits do not have a predetermined schedule-in its absence, it is challenging to determine which operations can share a functional unit without a performance penalty. Furthermore, although sharing seems to imply only some trivial circuitry, time-multiplexing units in dataflow circuits may cause deadlock by blocking certain data transfers and preventing operations from executing. In this paper, we present a technique to automatically identify performance-acceptable resource sharing opportunities in dataflow circuits and we describe a sharing mechanism that achieves deadlock-free dataflow designs. On benchmarks obtained from C code, we show that our approach effectively implements resource sharing: it results in significant area savings (i.e., a DSP reduction of up to 81%) compared to dataflow circuits which do not support this feature and matches the sharing capabilities of a state-of-the-art HLS tool.

IEEE2022

Scaling Language Features for Program Verification

Georg Stefan Schmid

Formal verification of real-world software systems remains challenging for a number of reasons, including lack of automation, friction in specifying properties, and limited support for the diverse programming paradigms used in industry. In this thesis we make progress towards a better verification experience in general-purpose programming languages by contributing improvements both to the automated checking and the specification of safety properties in languages combining functional and imperative features. We present our extensions in two parts -- reasoning about shared mutable data, and types as specifications -- both of which ultimately rely on reductions of expressive surface languages to a functional core. Throughout, we instantiate our techniques for the particular example of Scala, a mixed-paradigm language widely-used in industry.The first part shows how to extend a verifier for higher-order functions and immutable data to support imperative programs with shared mutable data. We build upon Stainless, a contract-based verification system that relies on SMT solvers to automatically verify a large fragment of Scala. Our technique extends Stainless to check general heap-manipulating programs against modular specifications in the style of dynamic frames. A novelty of our approach is the translation of imperative function contracts that encodes frame conditions using quantifier-free formulas in first-order logic, instead of relying on quantifiers or on dedicated separation logic reasoning. Our quantifier-free encoding enables SMT solvers to both prove safety and to report counterexamples relative to the semantics of procedure contracts.In the second part we turn to types and type-level programming as an alternative means of specifying correctness properties. While dependent types have been studied extensively for purely-functional languages, we investigate their applications to languages with subtyping and (abstractions of) imperative features. We first study a calculus that provides type-level computation through singleton types and allows abstraction of state and IO through a non-deterministic choice operator. This allows for modelling interactions with existing imprecisely-typed and impure code. Our calculus is formalized and mechanically proven correct using the Coq proof assistant. In addition, we develop a prototypical implementation in the Scala compiler and study typical type-level programming use cases in the Scala ecosystem.

EPFL2022

Automated Verification of Network Function Binaries

George Candea, Solal Vincenzo Pirelli

Formally verifying the correctness of software network functions (NFs) is necessary for network reliability, yet existing techniques require full source code and mandate the use of specific data structures. We describe an automated technique to verify NF binaries, making verification usable by network operators even on proprietary code. To solve the key challenge of bridging the abstraction levels of NF implementations and specifications without special-casing a set of data structures, we observe that data structures used by NFs can be modeled as maps, and introduce a universal type to specify both NFs and their data structures, the "ghost map". In addition, we observe that the interactions between an NF and its environment are sufficient to infer control flow and types, removing the requirement for source code. We implement our technique in Klint, a tool with which we verify, in minutes, that 7 NF binaries satisfy their specifications, without limiting developers' choices of data structures. The specifications are written in Python and use maps to model state. Klint can also verify an entire NF binary stack, all the way down to the NIC driver, using a minimal operating system. Operators can thus verify NF binaries, without source code or debug symbols, without requiring developers to use specific programming languages or data structures, and without trusting any software except Klint.

USENIX Association2022

Scaling Language Features for Program Verification

Georg Stefan Schmid

EPFL2022