PostfaceConclusion and Perspectives

Beyond the Scope of This Course

This course has focused on the foundations of Separation Logic for sequential programs, and discussed the set up of tooling for computing weakest preconditions.
There are numerous aspects of Separation Logic that were not covered in this course, including:
  • Concurrent Separation Logic, for reasoning about concurrent behaviors, or even, behaviors of weak memory models.
  • Automated reasoning using Separation Logic, typically for a restricted fragment of Separation Logic.
  • Separation Logic for low-level programs, at the byte level, and with assembly-style code written in continuation-passing style.
  • Separation Logic for object-oriented programs.
  • Reasoning about I/O, integer overflows, floatting-point arithmetics, etc.

Historical Notes

The most relevant historical notes were included near the end of each chapter. Additional information may be found in the ICFP'20 paper Separation Logic for Sequential Programs, by Arthur Charguéraud.
In addition, the reader may be interested in checking out Peter O'Hearn's account of the early days of Separation Logic:

Tools Leveraging Separation Logic

Ideas of Separation Logic have had significant influence on the field of programming languages from various perspectives. For a broad survey of Separation Logic, we refer to Peter O'Hearn's CACM paper 2019. sure to also download the "supplementary material" link at the bottom of the page).
Here is a non-exhaustive list of active projects leveraging Separation Logic.
  • The tool "Infer" leverages for static analysis, that is, fully automated program analysis to produce a list of potential bugs.
  • The programming language "Rust" features a type system with "borrows", a notion directly inspired by Separation Logic.
  • The tool "VeriFast" targets Java programs.
  • The tool "Veriable C" tool, from the VST project, targets the verification of C code, using interactive Coq proofs based on Separation Logic. See also volume 5 of Software Foundations.
  • The tool "Viper" is based on a variant of Separation Logic, its front-end supports several languages, and its backend leverages the SMT solver Z3; it also supports user annotations in the code for guiding proofs.
  • The "Isabelle Refinement Framework" leverages Separation Logic is the process of refining monadic high-order logic definitions into efficient imperative code. It can be used to produce verified ML code, as well as LLVM IR code.
  • The tool "CFML", which targets the verification of OCaml programs. The original version of CFML leverages a version of characteristic formulae slightly more complicated than the one presented in this volume. The new version, CFML 2.0, directly leverages all the ideas from this course. It is still under active development.
  • The verified compiler "CakeML" implements technology similar to that of CFML, and leverages it to verify components from its standard library and from its runtime.
  • The Iris project develops advanced logics for reasoning about concurrent programs using Separation Logic.

Related Courses

The following courses focus on reasoning about programs, and should be of interest to the reader:
Teaser: Arthur Charguéraud has for project to write a follow-up course on practical program verification using CFML, focusing on the specification and verification of classic data structures and algorithms.


The lead author of this volume, Arthur Charguéraud, wishes to thank:
  • Benjamin Pierce, for demonstrating the benefits of Coq-based teaching, for coordinating the Software Foundations series, and for encouraging me to write this volume.
  • The Software Foundations (SF) community, for helping to polish the course material and the tooling, and contributing to the success of teaching using this material.
  • François Pottier, with whom I developed several extensions of Separation Logic, and thereby obtained a deeper understanding of this logic.
  • Jacques-Henri Jourdan, who introduced me to a number of techniques used in the Iris project, and contributed to the definition of mkstruct.
  • Xavier Leroy, with whom I had numerous discussions on mechanized semantics.
  • Jean-Christophe Filliâtre, for numerous discussions on program verification.
  • Andrew Appel, Lars Birkedal, Adam Chlipala, Magnus Myreen, Gerwin Klein, Peter Lammich, and Zhong Shao, who kindly answered questions on related work aspects.
  • Jonathan Leivent for reporting typos on the beta-version of this course.
(* 2021-01-25 13:22 *)