mirrors/z3
3
0
Fork 0
mirror of https://github.com/Z3Prover/z3 synced 2025-04-18 06:39:02 +00:00
Code Activity
10237 commits 23 branches 48 tags 157 MiB
Commit graph

825 commits

Author SHA1 Message Date
Arie Gurfinkel 18e3c7b13d Fix bug introduced by formatting 2018-06-14 16:08:51 -07:00
Nikolaj Bjorner f3466bb3e4 tidy 
Signed-off-by: Nikolaj Bjorner <nbjorner@microsoft.com>
 2018-06-14 16:08:51 -07:00
Nikolaj Bjorner 1920450f98 throttle ite-blasting 
Signed-off-by: Nikolaj Bjorner <nbjorner@microsoft.com>
 2018-06-14 16:08:51 -07:00
Arie Gurfinkel 1f0fd38c99 ground sat refutation from spacer (wip) 2018-06-14 16:08:51 -07:00
Arie Gurfinkel 0534b72c4d sort hypotheses 2018-06-14 16:08:51 -07:00
Arie Gurfinkel c5fb1c1223 Use vector instead of a hashtable to represent a set 2018-06-14 16:08:51 -07:00
Arie Gurfinkel e84ca25f05 Check whether one proof node is an ancestor of another on-demand 
Instead of pre-computing sets
 2018-06-14 16:08:51 -07:00
Arie Gurfinkel a40e0dce0c proof_utils: use expr_mark instead of hashtable 2018-06-14 16:08:51 -07:00
Arie Gurfinkel 2a6b694373 Imrove hypothesis_reducer 2018-06-14 16:08:51 -07:00
Nikolaj Bjorner 4b2196f114 nits 
Signed-off-by: Nikolaj Bjorner <nbjorner@microsoft.com>
 2018-06-14 16:08:51 -07:00
Nikolaj Bjorner 6adaed718f remove pdr 
Signed-off-by: Nikolaj Bjorner <nbjorner@microsoft.com>
 2018-06-14 16:08:51 -07:00
Arie Gurfinkel cefdb8c01d Use reachable cache 2018-06-14 16:08:51 -07:00
Arie Gurfinkel 9fef466c63 Respect children order in spacer/pdr 2018-06-14 16:08:51 -07:00
Arie Gurfinkel f74ca2f0c0 Fix caching bug in mbc 2018-06-14 16:08:51 -07:00
Arie Gurfinkel 54add824e9 Debug print 2018-06-14 16:08:51 -07:00
Nikolaj Bjorner c3fb863ad1 formatting/reviewing 
Signed-off-by: Nikolaj Bjorner <nbjorner@microsoft.com>
 2018-06-14 16:08:51 -07:00
Arie Gurfinkel d2ae3b4025 Create children for pdr in spacer 
This is first working version of gpdr strategy. Passes one test.
 2018-06-14 16:08:51 -07:00
Arie Gurfinkel e1a45671b3 Cleanup spacer options 2018-06-14 16:08:51 -07:00
Arie Gurfinkel 1994f1d7e4 Cleanup of spacer options 2018-06-14 16:08:51 -07:00
Arie Gurfinkel cb683389f6 spacer::context: Factor params into udpt_params 2018-06-14 16:08:51 -07:00
Arie Gurfinkel 521392a8f1 First partially working pdr strategy in spacer 2018-06-14 16:08:51 -07:00
Arie Gurfinkel ab5f579d0b Comments in pdr_context.cpp 2018-06-14 16:08:51 -07:00
Arie Gurfinkel ece2e53c98 Ported model_search and model_node from pdr into spacer 2018-06-14 16:08:51 -07:00
Arie Gurfinkel 478d7c790e mbc: moved code under get_subst() 2018-06-14 16:08:51 -07:00
Arie Gurfinkel e860e4d045 Bug fix for quantified pob generation 2018-06-14 16:08:51 -07:00
Arie Gurfinkel 2a243d38d1 Model based Cartesian decomposition 2018-06-14 16:08:51 -07:00
Arie Gurfinkel 3178f7f86d Add random order of children in spacer 2018-06-14 16:08:51 -07:00
Arie Gurfinkel 5756871738 Always attempt to eliminate all existential variables 
Sometimes variables that cannot be eliminated in one context, can be
eliminated in the other. Pass all available variables to MBP to be
eliminated if possible
 2018-06-14 16:08:51 -07:00
Arie Gurfinkel 7396ad72ab Give up when a lemma is re-discovered too many times 2018-06-14 16:08:51 -07:00
Arie Gurfinkel 6fb6279f07 Cleanup array_eq_generalizer 2018-06-14 16:08:51 -07:00
Arie Gurfinkel da66ad6f80 Cleanup derivation::create_next_child 2018-06-14 16:08:50 -07:00
Arie Gurfinkel c5ff5ac2a1 Clen up spacer::pred_transformer::get_origin_summary 2018-06-14 16:08:50 -07:00
Arie Gurfinkel d7dc10212e Clean up spacer::context::create_children 2018-06-14 16:08:50 -07:00
Arie Gurfinkel b61da6fcc0 Debug print in org-mode format 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 6b82068d8d Bug fix in spacer::derivation::exist_skolemize 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 4ca734528e Formatting 2018-06-14 16:08:50 -07:00
Arie Gurfinkel dd064bd8f9 Bug fix to spacer::sym_mux 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 38c2b56f0e Rewrite spacer::sym_mux 
Simpler implementation that only provides functionality actually used
by spacer
 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 268274911a Fix to cube-and-clause interface in prop_solver 2018-06-14 16:08:50 -07:00
Arie Gurfinkel e0e435582a Minor code cleanup 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 6464468cd8 Remove dead code 2018-06-14 16:08:50 -07:00
Nikolaj Bjorner bfeb15b876 move to list of clauses 
Signed-off-by: Nikolaj Bjorner <nbjorner@microsoft.com>
 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 502e323678 Fixes to pred_tranformer::updt_solver 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 862eef5ec0 Eliminate all existential variables from reach facts 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 70f4674b3a Code to update solver with all constraints of a pred_transformer 2018-06-14 16:08:50 -07:00
Arie Gurfinkel cfcc084688 reach_fact --> rf 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 7a8563a34c spacer: cleaner management of rf tags 2018-06-14 16:08:50 -07:00
Arie Gurfinkel ada548b5ae Removed unused options 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 0b387cd7eb Moved pool_solvers from spacer::manager into spacer::context 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 451d42319b Rename m_reach_ctx into m_reach_solver 2018-06-14 16:08:50 -07:00
Arie Gurfinkel cdba0721e7 Extra stats in iuc_solver 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 16fefe850a Factored mbp into pred_transformer and added stats 2018-06-14 16:08:50 -07:00
Arie Gurfinkel fde58664f6 Moved mk_reach_fact to pred_transformer 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 2a2b21326b Stats on num_proxies in iuc_solver 2018-06-14 16:08:50 -07:00
Arie Gurfinkel e2e1411707 Option to dump SMT queries as benchmarks during Spacer run 2018-06-14 16:08:50 -07:00
Arie Gurfinkel fb52c36210 spacer: switch to new IUC as default 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 3a97451f8c spacer: normalize the cube before creating a lemma 2018-06-14 16:08:50 -07:00
Arie Gurfinkel fce68536d3 spacer: print all lemmas in json 2018-06-14 16:08:50 -07:00
Arie Gurfinkel bfa472faec New style of json dump based on lemmas at pob 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 5072a2a869 spacer: pobs keep track of their lemmas 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 61cd74818f Pin lemmas so that they don't disappear 2018-06-14 16:08:50 -07:00
Nikolaj Bjorner 0c2e3c0894 fixes to clause proof tracking 
Signed-off-by: Nikolaj Bjorner <nbjorner@microsoft.com>
 2018-06-14 16:08:50 -07:00
Arie Gurfinkel f7d015de8d Switch spacer_prop_solver to check_sat_cc 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 1343b272e7 Implement iuc_solver::check_sat_cc 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 723e96175b spacer: prepare to use incremental clause smt_solver interface 2018-06-14 16:08:50 -07:00
Arie Gurfinkel 56a29093d0 Cleanup transition creation in pred_transformer 2018-06-14 16:08:50 -07:00
Arie Gurfinkel af57db0413 Anti-unification of two ground expressions 2018-06-14 16:08:49 -07:00
Arie Gurfinkel a8438e081e Wired qe::mbp into spacer 
use option fixedpoint.spacer.native_mbp=true to use it
 2018-06-14 16:08:49 -07:00
Nikolaj Bjorner 20300bbf94 updates to mbqi 
Signed-off-by: Nikolaj Bjorner <nbjorner@microsoft.com>

Signed-off-by: Nikolaj Bjorner <nbjorner@microsoft.com>
 2018-06-14 16:08:49 -07:00
Arie Gurfinkel 14b9dd2cd7 spacer: let pool_solver own the solver 2018-06-14 16:08:49 -07:00
Arie Gurfinkel 15d0fd4b42 spacer: removed virtual_solver 
This commit removes virtual_solver and smt_context_manager that have
been migrated into solver_pool
 2018-06-14 16:08:49 -07:00
Arie Gurfinkel 098e70a9e2 spacer: switched to using solver_pool 2018-06-14 16:08:49 -07:00
Arie Gurfinkel c2304e2636 spacer: Cleanup of smt parameter configuration 2018-06-14 16:08:49 -07:00
Arie Gurfinkel cfeee55d4f spacer: set qi.quick_checker to MC_UNSAT if quantifiers are expected 2018-06-14 16:08:49 -07:00
Arie Gurfinkel c8187886cf spacer: use same params for all solver pools 2018-06-14 16:08:49 -07:00
Arie Gurfinkel c2b8f25cf9 Switch to using solver instead of smt::kernel all around 2018-06-14 16:08:49 -07:00
Arie Gurfinkel 4b09cefb97 Replace smt::kernel with smt_solver 
Replace all ad-hoc uses of smt::kernel with ad-hoc uses of smt_solver
 2018-06-14 16:08:49 -07:00
Arie Gurfinkel 9c37bef553 Fix bug in ctp 2018-06-14 16:08:49 -07:00
Arie Gurfinkel 40781c0b0c Comment on params used in spacer_context 2018-06-14 16:08:49 -07:00
Arie Gurfinkel 55126692c9 spacer: counterexample to pushing (ctp) 
Enable using fixedpoint.spacer.ctp=true

For each lemma L currently at level k, keep a model M that justifies
why L cannot be pushed to (k+1). L is not pushed while the model M
remains valid.
 2018-06-14 16:08:49 -07:00
Arie Gurfinkel 95d820196b Cleanup 2018-06-14 16:08:49 -07:00
Arie Gurfinkel 68b7966254 Use C++11 2018-06-14 16:08:49 -07:00
Arie Gurfinkel 4e9023b8fe Remove dead code 2018-06-14 16:08:49 -07:00
Arie Gurfinkel 00f870b7ff to_mbp_benchmark(): prints an mbp problem in benchmark format 
currently unused. See comment in spacer_util.c:qe_project for example
usage
 2018-06-14 16:08:49 -07:00
Nikolaj Bjorner 402234757e updates to mbqi 
Signed-off-by: Nikolaj Bjorner <nbjorner@microsoft.com>
 2018-06-14 16:08:49 -07:00
Arie Gurfinkel e8cabdc620 Uninitialized variable 2018-06-14 16:08:49 -07:00
Arie Gurfinkel 054c6196a0 Move spacer qe into spacer_qe namespace 
Attempt to solve compilation issues with GCC and current replication
of qe namespace inside and outside spacer
 2018-06-14 16:08:49 -07:00
Arie Gurfinkel 56bce005a0 virtual_solver: debug print 2018-06-14 16:08:49 -07:00
Arie Gurfinkel 891dcd99c2 Use fact-generating version of mk_unit_resolution() 
fact-using version of mk_unit_resolution() requires the fact to be a
literal. Not sure why this restriction is placed there.
 2018-06-14 16:08:49 -07:00
Arie Gurfinkel 8be03f7c1f spacer_context: skolemize quant vars before renaming 
Skolemization has to be done before renaming, otherwise,
can't guarantee that variable names do not clash
 2018-06-14 16:08:49 -07:00
Arie Gurfinkel 4de58a42fe Update initialization order 2018-06-14 16:08:49 -07:00
Arie Gurfinkel aeb2f3c4bb factor out inherit_properties 2018-06-14 16:08:49 -07:00
Arie Gurfinkel 2f369d8d41 Simplify code using C++11 conventions 2018-06-14 16:08:49 -07:00
Arie Gurfinkel 5a6bd5e782 hypothesis_reducer: worked around propositional literals 
propositional formulas (disjunctions) can appear as literals.  This
makes it tricky to recognize whether a formula is a unit clause when
re-building unit resolution.

Added work-around that identifies whether a formula is a literal based
on its appearance in previous unit resolution step.
 2018-06-14 16:08:48 -07:00
Arie Gurfinkel b39c532f19 Order of methods in spacer_context.cpp 2018-06-14 16:08:48 -07:00
Nikolaj Bjorner 7931bd1dfc updates to mbqi 
Signed-off-by: Nikolaj Bjorner <nbjorner@microsoft.com>
 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 0fe5e6c2a6 Fix handling of complex literals in hypothesis_reducer 
In Z3, an arbitrary, even propositional, formula can be a literal.

This requires careful handling of restructuring of unit resolution.
 2018-06-14 16:08:48 -07:00
Arie Gurfinkel d1a7c0ceb0 Remove a print 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 05c8067392 Changed pob queue management strategy in spacer_context 2018-06-14 16:08:48 -07:00
Arie Gurfinkel a696a40a3a Refactoring 2018-06-14 16:08:48 -07:00
Arie Gurfinkel ec8f99fee7 Rename expand_node --> expand_pob 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 3f9b5bce99 Remove debug function 2018-06-14 16:08:48 -07:00
Arie Gurfinkel ac3bbed311 Remove dead code in spacer_manager 
- removed bg_assertions. Incompatible with mbp in spacer
  - removed unique number. Unused
  - removed mk_and() and switched to ast_util:mk_and() instead
       spacer_manager::mk_and() uses bool_rewriter to simplify the
       conjunction
 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 33466c75a6 mss loop in prop_solver 
max sat assignment (mss) to replace core-based maxsmt()
 2018-06-14 16:08:48 -07:00
Arie Gurfinkel d379b14942 Cleanup spacer_iuc_solver 2018-06-14 16:08:48 -07:00
Arie Gurfinkel fd13eb9e0e Final cleanup of hypothesis_reducer 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 9d4784baf6 Fix dealloc order in hypotheses_reducer::reset() 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 689414d055 Fix debug printing in iuc_solver 2018-06-14 16:08:48 -07:00
Nikolaj Bjorner ff0f257102 remove iff 
Signed-off-by: Nikolaj Bjorner <nbjorner@microsoft.com>
 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 8d312f9d1f Cleanup of hypothesis_reducer 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 2db38fedd6 Cleanup of theory_axiom_reducer proof trasfomation 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 07ad67ebad Move proof dot printing into iuc_proof 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 45500ff7d3 Cleanup iuc_proof 2018-06-14 16:08:48 -07:00
Arie Gurfinkel abe67705d3 Cleanup iuc_proof 2018-06-14 16:08:48 -07:00
Arie Gurfinkel c893740e13 Fix compilation 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 3bc3b00fdd Post merge compile fixes 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 649bab2f58 Rename itp_solver into iuc_solver 
IUC stands for Interpolanted UNSAT Core
 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 5d3b515a50 Cleanup option names and default values 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 39bdecf9c2 Minor code cleanup 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 83adb6742e Remove whitespace 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 7c727ee922 Fix compiler warnings 2018-06-14 16:08:48 -07:00
Bernhard Gleiss 4b6921dffb removed unnecessary assignment 2018-06-14 16:08:48 -07:00
Bernhard Gleiss 295d16bfae Rewrite hyp-reducer 
This is a new version that conceptually addresses the bugs in
all previous version. However, it had a hard-to-debug memory
corruption. The bug appeared only in optimized compilation under
Linux with GCC.

This code is suspect and should be reviewed and further tested
 2018-06-14 16:08:48 -07:00
Bernhard Gleiss 85c58e344c Add option to use old_hyp_reducer 2018-06-14 16:08:48 -07:00
Arie Gurfinkel f0f75d5254 Wire in arith-axiom-reducer 2018-06-14 16:08:48 -07:00
Bernhard Gleiss 0f25e9e831 Moved farkas stats printing to before and after the hyp-reduction 2018-06-14 16:08:48 -07:00
Bernhard Gleiss de31b07008 arith-theory-axiom reducer to handle arithmetic axioms 2018-06-14 16:08:48 -07:00
Arie Gurfinkel df2eb771ef Fix in spacer_itp_solver: use pr.get() instead of get_proof() 2018-06-14 16:08:48 -07:00
Arie Gurfinkel ab3a6702af Fix several bugs in hyp_reducer 
- compute_marks didn't find all units
  - call to m.mk_unit_resolution expects that there is at least one unit
  - hyp-reduced proof wasn't used
  - bug in early termination
  - any hypothesis was  replaced with the old derivation of the literal
  - handle the case of a single literal premise under hypothesis that is
    replaced by an empty clause under hypothesis
 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 56114a5f6d Refactor iuc_proof as a separate class 
This also adds DOT printing support to interpolating proofs
(color for different parts)

iuc_proof is a proof used for IUC computation
 2018-06-14 16:08:48 -07:00
Bernhard Gleiss 247c570e6b Debug sanity check in spacer_context 
Triggered by bugs in hypothesis remover

only sanitycheck lemmas in debug-mode
 2018-06-14 16:08:48 -07:00
Matteo Marescotti a4e67b8bb6 Wire JSON printing into Spacer 2018-06-14 16:08:48 -07:00
Matteo Marescotti 3248f57434 Add support for printing spacer pobs in JSON 2018-06-14 16:08:48 -07:00
Matteo Marescotti 28ef9ab9d1 User option to enable starting spacer from a given level 2018-06-14 16:08:48 -07:00
Matteo Marescotti ff7c949be8 Fix: call collect_statistics() in virtual_solver 2018-06-14 16:08:48 -07:00
Matteo 65885f7eba add_constraint API 2018-06-14 16:08:48 -07:00
Matteo 3c7165780c Extend spacer with callback events 
Callback events allow the client of spacer to
get events during exection. The events include
new lemmas and unfolding.
 2018-06-14 16:08:48 -07:00
Arie Gurfinkel b51251f394 Move tout under TRACE 2018-06-14 16:08:48 -07:00
Arie Gurfinkel fd7bcc7afc Format 2018-06-14 16:08:48 -07:00
Yakir Vizel 5df7a08d1c A simple version for finding the stride between different indices in a POB 
This current version is very limited.
It assumes a pre-defined structure (namely, an ADDER).
 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 04a778f2fd Option to enable cube normalization in quic generalizer 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 852e181fed New quic3 lemma generalizer 2018-06-14 16:08:48 -07:00
Arie Gurfinkel 9cdb63ae4a Handle conversion of quantified lemma to quantifier free 
When a cube is updated, a lemma might loose all of its quantified
variables. In this case, it is effectively quantifier free
and might be a version of an already existing lemma.

For that reason, we convert it to quantifier free lemma when
this happens.
 2018-06-14 16:08:48 -07:00
Yakir Vizel 23a8e59493 Initial commit of QGen 
Controlled by fixedpoint.spacer.use_quanti_generalizer

measure cumulative time, number of invocations, and number of failed
SMT calls

Relaxing equality in a pattern: if a variable equals a numeral, relax with GE

pob::get_skolems() returns all skolems that might appear in the pob.
New skolems must be added above the largest index in that map,
even if they are not used in the pob itself.

pattern generalization should be done before the pattern is skolemized and
added into the new cube.
 2018-06-14 16:08:47 -07:00
Arie Gurfinkel a1efb88318 Semantic matcher 
Extends matcher with rewrites based on semantics of arithmetic operations

Like matcher, but uses arithmetic and logic rewrites to try to
get a semantic match.
 2018-06-14 16:08:47 -07:00
Arie Gurfinkel 7dee36358d Allow bool_ind_generalizer to skip non-array literals 
Currently a hack to skip generalizing some literals.
Used together with quic generalizer to remove all array terms if possible
before quic generalization
 2018-06-14 16:08:47 -07:00
Arie Gurfinkel 46fb0d928a Fix in spacer_term_graph 2018-06-14 16:08:47 -07:00
Yakir Vizel 068b77d43a Normalizing LE and GE with constants 2018-06-14 16:08:47 -07:00
Arie Gurfinkel 0c1ef7155a Option to rewrite expression in term_graph 
Rewrite expressions to minimize uses of constants 0 and 1
Currently disabled due to interaction with quic
 2018-06-14 16:08:47 -07:00
Arie Gurfinkel ecf9c629b0 Fix binding handling for quantifier free lemmas 2018-06-14 16:08:47 -07:00