Golem is a solver for Constrained Horn Clauses (CHCs). It accepts the input in the format of (extended) SMT-LIB 2.6, as defined by CHC-COMP.
The easiest way is to download the executables from our release page. This way, all dependencies are already bundled in the executable.
Golem can be compiled on Linux and MacOS.
It uses CMake
for build configuration.
Golem depends on OpenSMT for SMT solving and interpolation.
If you already have OpenSMT installed, you can pass the path using -DOPENSMT_HOME
option to cmake
command.
Note that Golem requires a specific version of OpenSMT, currently v2.7.0.
Otherwise, cmake
will download the latest compatible version of OpenSMT and build it as a subproject.
You can view the usage in the help message after running
$ golem -h
At the moment, you should specified the SMT theory used in the CHC encoding with -l
option. The supported theories are QF_LRA
and QF_LIA
, i.e., the linear arithmetic over reals or integers.
Golem now has limited support to automatically detect the theory from the script, so the option is no longer mandatory, but still recommended.
Golem supports several different backend algorithms for solving CHCs.
Spacer engine represents our own implementation of the Spacer algorithm from this paper. You might be familiar with the original implementation of Spacer inside Z3.
BMC engine implements the simple bounded model checking algorithm which checks for existence of increasingly longer counterexample paths. It uses incremental capibilities of the underlying SMT solver to speed up the process. Works only for linear systems of Horn clauses.
DAR engine implements the algorithm Dual Approximated Reachability described in this paper. It works only for linear systems of Horn clauses.
KIND engine implements very basic k-induction algorithm from this paper. It only supports transition systems.
The implementation of LAWI follows the description of the algorithm in this paper.
The algorithm is also known as Impact
, which was the first tool where the algorithm was implemented.
Works only for linear systems of Horn clauses.
IMC engine implements the original McMillan's interpolation-based model-checking algorithm from this paper. It works on transition system, but it can handle linear systems of Horn clauses by first transforming them into a simple transition system.
The PA engine is a simple prototype of a predicate abstraction with CEGAR. The implementation is still rather naive, but the algorithm can handle all (even nonlinear) CHC systems.
The implementation of PDKIND follows the description of the algorithm in this paper. It works on transition system, but it can handle linear systems of Horn clauses by first transforming them into a simple transition system.
TPA is an algorithm we have developed recently with the goal to detect long counterexample quickly. The description of the algorithm can be found in this paper. TPA directly supports a subset of linear CHC systems which can be mapped to DAGs of transition systems. Transitions that do not fall into this category are handled by transformation into a simple transition system.
split-TPA is a different instantiation of the TPA paradigm and is typically more powerful than basic TPA on satisfiable (safe) CHC systems.
Golem also supports multiprocess run of several engines simultaneously.
You can pass comma-separated list of engines to --engine
options.
For example, the following invocation will run split-TPA, Spacer and LAWI in parallel
golem --engine split-tpa,spacer,lawi --input <file>
Golem supports internal validation of witnesses for its answer using --validate
option.
Witness for sat
is a model, an interpretation of the predicates.
Witness for unsat
is a proof.
This option is still experimental. For example, tpa/split-tpa
does not always produce the witness yet.
To obtain the produced model or proof of unsatisfiability, use --print-witness
.