| Age | Commit message (Collapse) | Author |
|---|
|
|
|
|
|
Based on #4774, this PR adds a new CadSolver class that allows the NonlinearExtension to actually employ the CAD-based method. In more detail:
add --nl-cad option
add CadSolver class that wraps cad::CDCAC with support for checks, model construction and conflict generation
add new Inference types for the NlLemma class
use CadSolver in NonlinearExtension (if --nl-cad is given)
|
|
This is the 4th step of adding interface for SyGuS Interpolation module. The whole change will be adding the API for (get-interpol s B), which is aim for computes an I that A->I and I->B. Here A is the assertions in the stack.
The 4th step finished the implementation of the interpolation module.
|
|
This PR adds new rewrite rules for BV.
None of them is meant to be used by the default BV rewriter.
However, they are planned to be used in bv_to_int preprocessing pass.
In the pass we use FixpointRewriteStrategy to call various rewrite rules.
After consulting @4tXJ7f , we thought that the best way to include more rewrite rules in that call is to implement them using the existing BV rewrites infrastructure.
|
|
This class will be used as a utility in a new algorithm for solution reconstruction and requires a generalized interface.
FYI @abdoo8080
|
|
datatype (#4825)
|
|
Fixes #4820. The performance issue was caused by
0988217562006d3f59e01dc261f39121df6d75f5. That commit introduced an
option (`--strings-len-conc`) that optionally moves the length
constraint for skolems to the conclusion of an inference instead of
making it part of the term registration. However, for
`DEQ_DISL_STRINGS_SPLIT`, we were only asserting that `LENGTH_GEQ_ONE`
in the case where this option was not enabled, instead of asserting that
the length of the skolem is equal to the component on the other side of
the disequality. This lead to an infinite loop of inferences because we
effectively were just splitting a component into two skolems and the
only restriction was that the first one was non-empty. We guessed the
second skolem to be empty, so the first skolem was equal to the
component, the skolem was marked congruent, and we ended up with the
same normal form as before.
The commit fixes the issue by adding an argument to
`getDecomposeConclusion()` that specifies whether to add the length
constraint or not. This option is used to always add the length
constraint in the case of `DEQ_DISL_STRINGS_SPLIT`.
|
|
Fixes #4701. That benchmark now times out.
|
|
Some internal inferences had a non-standard way of providing (disjunctive) reasons and a custom way of explaining them.
This PR simplifies the array solver so that its explanations are analogous to the other equality engine based theory solvers (strings, datatypes, sets, ...). This is done in preparation for the incorporation of the proof equality engine in arrays, which follows a standard design for reasons/explanations.
The performance impact on QF arrays is negligible
|
|
This commit implements the CAD interface added in #4773.
|
|
This PR adds some utilities and, most importantly, the interface of the new
CAD-based solver.
The approach is based on https://arxiv.org/pdf/2003.05633.pdf and the code
structure follows the paper rather closely.
|
|
Avoids crashing on some of our debug traces.
|
|
fixing issue #4808. Remove structural binding, which only supported by c++17.
|
|
This option has model soundness issues (#4771) and moreover is overall worse performance on SMT-LIB {QF_ABV QF_ABVFP QF_ABVFPLRA QF_ALIA QF_ANIA QF_AUFBV QF_AUFLIA QF_AUFNIA QF_AX}:
Configuration #unsat time #sat time #solved #total
CVC4-072720_def 9428 9405.46 24932 16631.6 34360 35399
CVC4-072720_nalr1 9446 9536.41 24924 16146.3 34370 35399
where def = default, nalr1 = --no-arrays-lazy-rintro1.
Fixes #4771.
FYI @barrettcw
|
|
Fixes #4759 .
Also refactors this method.
|
|
This PR adds support for seq.nth operator by eliminating it during expandDefinitions, based on sub-sequences.
Tests that use this operator are also included.
|
|
This is the 3rd step of adding interface for SyGuS Interpolation module. The whole change will be adding the API for (get-interpol s B), which is aim for computes an I that A->I and I->B. Here A is the assertions in the stack.
The 3rd step partially implemented the interpolation module.
|
|
There are 3 Boolean flags for OutputChannel::lemma, and plans to add another for relevance.
This makes them into a enum.
|
|
This adds proof support in the term formula removal pass.
It also refactors this class heavily so that its interface is more intuitive and does not depend on AssertionPipeline.
|
|
This updates the interface for arithmetic operator elimination for the new proof format.
The actual proof production of the operator elimination class (providing proofs for
introduced witness terms) will be done in a separate PR.
This also changes the witness terms introduced by this class so their body is in
Skolem form, which simplifies term formula removal.
Co-authored-by: Alex Ozdemir <aozdemir@hmc.edu>
|
|
The current arith proof machinery can prove conflicts which are
explained by assumptions and tightened assumptions.
Previously we verified that the conflict constraint was explained in
terms of assumptions and tightened assumptions, before trying to
save/produce a proof.
We did not verify that the negated constraint was an assumption or
tightened assumption.
This caused us to attempt to prove conflicts around constraints whose
negations where proven by general means (in the case of #4714, by the
equality engine), which the proof machinery was not prepared to handle.
This PR also checks that the negate constraint is an assumption or
tightened assumption, before saving the proof.
Thanks, Gereon, for doing the initial investigation into this!
fixes 4714
Co-authored-by: Gereon Kremer <nafur42@gmail.com>
|
|
|
|
Based on #4679, this PR contains further preparations for a CAD-based arithmetic solver that extends the current NonlinearExtension.
In detail, it provides:
a Constraints class that manages all (polynomial) constraints visible to the cad solver,
a collection of methods used for cad projections,
a VariableOrdering class that provides different variable ordering heuristics tailored to cad,
an extension to the NlModel class, allowing for witness terms,
further conversion methods, in particular between Node and poly::Polynomial, poly::Value and RealAlgebraicNumber.
|
|
To keep track of the reasoning in the equality engine for n-ary kinds, for which partial applications amount to valid fully-applied terms, it's imperative to be able to see the IDs of the internal representation of the equality engine nodes. This commit updates tracing messages to print these IDs. It also improves the tracing for explanation generation.
|
|
This adds the feature of enumerating shapes in the fast sygus enumerator, which is required for a new algorithm for sygus solution reconstruction.
This also adds a builtinToSygus backwards mapping that is required for that algorithm as well.
Note this also changes the implementation of mkFreeVar in sygus term database from skolem to bound variable, which is required to do proper caching for expr::hasBoundVar.
|
|
This updates the core strings solver in preparation for proofs.
The main changes include:
The addition of the strings PfRule enum values.
The definition of a "getConclusion" static method, used by the core solver, and to be used in the future by the strings proof checker. This includes several optional forms of lemmas, which are added as options in this PR.
Major simplifications to our inference schemas for disequality handling (a STRING_DECOMPOSE inference rule). Note this is the only significant intended behavioral change in this PR.
Minor updates to the form of inferences send to inference manager, for instance to orient equalities in the expected way, and to reorder assumptions. These changes are done for uniformity and make the strings proof reconstruction from inference steps easier.
|
|
A wrong change slipped away during the cleaning of the module. This commit fixes the conversion.
|
|
|
|
Previously, we assumed that ExtTheory, the module for doing context-dependent simplification, was one-to-one with Theory. This design is not necessary. This makes this class a utility, which can be used as needed. This makes e.g. the initialization of TheoryStrings much easier, since the ExtTheory object can be created first.
|
|
This PR removes two things from the resource manager:
cumulative time limits
cpu time limits
Cumulative time limiting has been moved to the binary and is (as before) accessible via --tlimit. As per discussion among the devs, we no longer support time limits based on CPU time and thus everything related to that is removed as well.
Note that this includes the option --cpu-time, removes an argument from SmtEngine::setTimeLimit() and the method SmtEngine::getTimeRemaining() .
|
|
Also puts EqProof into its own module. Next will come the implementation of the API.
|
|
This commit changes SmtEngine::assertFormula() to use Nodes rather
than Exprs and changes AssertionList to be Node-based. This is
work towards removing the Expr layer.
|
|
This splits everything related to abducts into its own standalone module, AbductionSolver.
It furthermore converts some of the interfaces of SmtEngine to make them take Node instead of Expr (this will be done for every method eventually).
The code for interpolation should follow a similar pattern, e.g. InterpolantSolver.
|
|
This commit changes UninterpretedConstant to use TypeNode instead of
Type.
|
|
The generality of this interface is unnecessary.
|
|
This removes sygus print callbacks, since they are no longer necessary to support the sygus v1 parser.
This involved generalizing the scope of the datatype utility sygusToBuiltin. Now, printing "as sygus" simply is accomplished by doing sygusToBuiltin conversion and then calling the printer on the builtin term.
This is required for further work towards eliminating the Expr layer.
FYI @4tXJ7f
|
|
This adds support for skeleton proofs in the rewriter (REWRITE -> THEORY_REWRITE).
It adds "extended equality rewrite" as a new method of the rewriter/theory rewriters.
The unit test of this feature should be added on a followup PR.
Co-authored-by: Andres Noetzli <andres.noetzli@gmail.com>
|
|
This commit changes EmptySet to use TypeNode instead of Type.
|
|
This prints # instantiations per round (during solving) for each quantified formula with `--debug-inst`, giving output like this:
```
(num-instantiations myQuant1 1)
(num-instantiations myQuant2 1)
unsat
```
It also changes the default value of print-inst-full to match the behavior of unsat-cores vs unsat-cores-full (by default, nameless quantifiers are ignored).
It fixes an issue with qid, where mkVar was accidentally used instead of mkConst, leading to assertion failures in debug. Marking major since this fixes debug regress1.
|
|
This decouples asserting a formula with initialization (previously it was a complex process to assert a formula due to having to clone/export to a new ExprManager). Now it is trivial.
This commit fixes an unintended consequence of the previous complications. Previously, SmtEngine::setOption would be set after asserting formulas to an SmtEngine subsolver, which is technically incorrect, as options should be finalized before the first assert.
This is required for further cleaning up of options listeners.
|
|
Commit 64a7db86206aa0993f75446a3e7f77f3c0c023c6 introduced a caching
mechanism in `TheoryEngine::getExplanation()`. However, there seem to be
issues related to the timestamps of the explanations. This commit fixes
the issue by making the timestamp part of the cache. The change ensures
that explanations for a certain node never rely on other explanations
for that node with a later timestamp.
|
|
This commit changes ArrayStoreAll to use Node/TypeNode instead of
Expr/Type.
|
|
Fixes regress1.
|
|
This adds the proof generator for storing proofs of preprocessing. It stores assertions added via preprocessing passes and their rewrites. This utility will eventually live in SmtEngine.
It also adds 2 new proof rules, and fixes an issue in ProofNodeUpdater.
|
|
This makes an option --debug-sygus available to the user for tracing the sygus solver. For the classic max2 example the option is:
(sygus-enum 0)
(sygus-candidate (max 0))
(sygus-enum 0)
(sygus-enum 1)
(sygus-enum x)
(sygus-enum x)
(sygus-candidate (max x))
(sygus-enum x)
(sygus-enum y)
(sygus-enum y)
(sygus-candidate (max y))
(sygus-enum y)
(sygus-enum (+ x x))
(sygus-enum (+ x 1))
(sygus-enum (+ 1 1))
...
(sygus-enum (ite (<= x y) y 1))
(sygus-candidate (max (ite (<= x y) y 1)))
(sygus-enum (ite (<= x y) y 1))
(sygus-enum (ite (<= x y) y x))
(sygus-enum (ite (<= x y) y x))
(sygus-enum (ite (<= x y) y x))
(sygus-candidate (max (ite (<= x y) y x)))
unsat
(define-fun max ((x Int) (y Int)) Int (ite (<= x y) y x))
Where sygus-enum denotes enumerated terms and sygus-candidate is one that passes a CEGIS refinement check.
|
|
This adds a new print format for instantiations --print-instantiations=num, which prints the total number of instantations of quantified formulas. This count is user-context-dependent, which is in sync with the existing print-instantiation format (list).
It also simplifies and improves printing of Instantiation Tries.
|
|
This adds basic support for string/sequence updating, which has a semantics that is length preserving.
|
|
Now that we can break the old API, we don't need an ExprSequence
anymore. This commit removes ExprSequence and replaces all of its uses
with Sequence. Note that I had to temporarily make sequence.h public
because we currently include it in a "public" header because we still
generate the code for Expr::mkConst<Sequence>().
|
|
For some hard verification benchmarks from Facebook, TheoryEngine::getExplanation was found to be the bottleneck, where the main loop of TheoryEngine::getExplanation would be run regularly 30k times, sometimes over 1 million times for a single conflict.
This caches explanations in this loop, where now the loop is executed roughly 1k times at max for the same benchmark.
One challenging benchmark that previously solved in 8 min 45 sec now solves in 2 min 45 sec.
FYI @barrettcw .
|
