Difference between revisions of "Grenoble meeting 25 May 2016"

From vecolib
Jump to: navigation, search
Line 13: Line 13:
 
12h00-12h45 '''Sylvain Boulme (Verimag)'''
 
12h00-12h45 '''Sylvain Boulme (Verimag)'''
  
TBA
+
''How to have a lightweight certification in Coq of your next Ocaml solver ?''
 +
 
 +
The embedding of untrusted Ocaml code into extracted code from Coq -- through a skeptical approach -- significantly simplifies Coq developments of formally proved software. This talks illustrates this lightweight certification approach on two examples: first, an UNSAT-prover -- based on a DPLL oracle -- and second, the Verimag Polyhedra Library -- an abstract domain of polyhedra based on linear programming oracles. VPL is a joint work with Fouihlé, Maréchal, Périn and Monniaux.
  
 
12h45-14h00 '''Lunch (on site)'''
 
12h45-14h00 '''Lunch (on site)'''

Revision as of 09:23, 18 May 2016

Location: Verimag 2nd floor, room 248

10h00-10h30 Coffee+croissants

10h30-11h15 Mihaela Sighireanu (Liafa)

TBA

11h15-12h00 Boris Yakobovski (CEA)

Rebooting Frama-C's abstract interpreter: past, present and future

12h00-12h45 Sylvain Boulme (Verimag)

How to have a lightweight certification in Coq of your next Ocaml solver ?

The embedding of untrusted Ocaml code into extracted code from Coq -- through a skeptical approach -- significantly simplifies Coq developments of formally proved software. This talks illustrates this lightweight certification approach on two examples: first, an UNSAT-prover -- based on a DPLL oracle -- and second, the Verimag Polyhedra Library -- an abstract domain of polyhedra based on linear programming oracles. VPL is a joint work with Fouihlé, Maréchal, Périn and Monniaux.

12h45-14h00 Lunch (on site)

14h00-14h45 Radu Iosif (Verimag)

A Decision Procedure for Separation Logic in SMT

This paper presents a complete decision procedure for the entire quantifier-free fragment of Separation Logic (SL) interpreted over heaplets with data elements ranging over a parametric multi-sorted (possibly infinite) domain. The algorithm uses a combination of theories and is used as a specialized solver inside a DPLL(T) architecture. A prototype was implemented within the CVC4 SMT solver. Preliminary evaluation suggests the possibility of using this procedure as a building block of a more elaborate theorem prover for SL with inductive predicates, or as back-end of a bounded model checker for programs with low-level pointer and data manipulations.

Joint work with Andrew J. Reynolds (U of Iowa), Tim King (Google) and Cristina Serban (Verimag)

14h45-15h00 Coffee break

15h00-17h00 Discussion

Agenda: progress report, mid-term review, etc.