Andrew Reynolds [Wed, 27 Apr 2011 00:49:02 +0000 (00:49 +0000)]
cleaned up some of the hacks in the datatypes theory solver, working on using Transitive Closure to detect cycles, added rewrite rule for disinguished ground terms
Morgan Deters [Mon, 25 Apr 2011 23:44:00 +0000 (23:44 +0000)]
Monday tasks:
* new "well-foundedness" type property (like cardinality) specified in
Theory kinds files; specifies well-foundedness and a ground term
* well-foundedness / finite checks in Datatypes now superseded by type
system isFinite(), isWellFounded(), mkGroundTerm().
* new "RecursionBreaker" template class, a convenient class that keeps
a "seen" trail without you having to pass it around (which is
difficult in cases of mutual recursion) of the idea of passing
around a "seen" trail
Morgan Deters [Mon, 25 Apr 2011 07:59:57 +0000 (07:59 +0000)]
small unit test fix; was broken only in non-assertion, non-CLN builds
Morgan Deters [Mon, 25 Apr 2011 06:56:14 +0000 (06:56 +0000)]
Weekend work. The main points:
* Type::getCardinality() returns the cardinality for for all types.
Theories give a cardinality in the their kinds file. For
cardinalities that depend on a type argument, a "cardinality
computer" function is named in the kinds file, which takes a
TypeNode and returns its cardinality.
* There's a bitmap for the set of "active theories" in the
TheoryEngine. Theories become "active" when a term that is owned by
them, or whose type is owned by them, is pre-registered (run CVC4
with --verbose to see theory activation). Non-active theories don't
get any calls for check() or propagate() or anything, and if we're
running in single-theory mode, the shared term manager doesn't have
to get involved. This is really important for get() performance
(which can only skimp on walking the entire sub-DAG only if the
theory doesn't require it AND the shared term manager doesn't
require it).
* TheoryEngine now does not call presolve(), registerTerm(),
notifyRestart(), etc., on a Theory if that theory doesn't declare
that property in its kinds file. To avoid coding errors,
mktheorytraits greps the theory header and gives warnings if:
+ the theory appears to declare one of the functions (check,
propagate, etc.) that isn't listed among its kinds file properties
(but probably should be)
+ the theory appears NOT to declare one of the functions listed in
its kinds file properties
* some bounded token stream work
Morgan Deters [Sat, 23 Apr 2011 05:36:09 +0000 (05:36 +0000)]
fix for parser/tests for ANTLR 3.2 (it was working fine on 3.3)
Morgan Deters [Sat, 23 Apr 2011 05:15:56 +0000 (05:15 +0000)]
* reviewed BooleanSimplification, added documentation & unit test
* work around a lexer ambiguity in CVC grammar
* add support for tracing antlr parser/lexer
* add parsing support for more language features
* initial parameterized types parsing work to support Andy's work
Morgan Deters [Sat, 23 Apr 2011 04:00:02 +0000 (04:00 +0000)]
make run_regression script robust to DOS newlines :(
Andrew Reynolds [Fri, 22 Apr 2011 22:24:29 +0000 (22:24 +0000)]
added fixes for datatype theory solver to account for rewriting before finite/well-founded check.
Morgan Deters [Fri, 22 Apr 2011 22:02:46 +0000 (22:02 +0000)]
fix to last commit
Morgan Deters [Fri, 22 Apr 2011 22:01:36 +0000 (22:01 +0000)]
Fixing SmtEngine::getValue() by adding a NodeManagerScope (thanks Tim for finding this)
Morgan Deters [Wed, 20 Apr 2011 11:19:50 +0000 (11:19 +0000)]
numerous bugfixes
Morgan Deters [Wed, 20 Apr 2011 08:07:07 +0000 (08:07 +0000)]
incorrect usage of C++ std::string caused a test to fail
Morgan Deters [Wed, 20 Apr 2011 07:57:28 +0000 (07:57 +0000)]
Minor mixed-bag commit. Expected performance impact negligible.
* Fixed hole in arrays typechecking.
* Fixed "make dist".
* Better ouroborous test, and some printer fixes.
* Continued cleanup in CVC parser, removed some warnings.
* Better output.
Morgan Deters [Wed, 20 Apr 2011 05:37:38 +0000 (05:37 +0000)]
Tuesday end-of-day commit.
Expected performance impact outside of datatypes/CVC parser is
negligible.
* CVC language LAMBDA, functional LET, type LET, precedence fixes,
bitvectors, and arrays, with partial parsing support also for
quantifiers, tuples, subranges, subtypes, and records
* support for complex recursive DATATYPE selectors, e.g.
tree = node(children:ARRAY INT OF tree) | leaf(data:INT)
these are complicated because they have to be left unresolved
at parse time and dealt with in a second pass.
* bugfix for Exprs/Types that occurred when setting them to null
(not Nodes/TypeNodes, just Exprs/Types).
* Cleanup/code review items
Tim King [Mon, 18 Apr 2011 20:19:29 +0000 (20:19 +0000)]
Removing dead code that came in on commit r1740.
Morgan Deters [Mon, 18 Apr 2011 19:08:11 +0000 (19:08 +0000)]
more work on CVC language
Morgan Deters [Mon, 18 Apr 2011 18:05:39 +0000 (18:05 +0000)]
mostly CVC presentation language parsing and printing
Tim King [Mon, 18 Apr 2011 16:48:52 +0000 (16:48 +0000)]
This commit merges the branch arithmetic/propagation-again into trunk.
- This adds code for bounds refinement, and conflict weakening.
- This adds util/boolean_simplification.h.
- This adds a propagation manager to theory of arithmetic.
- Propagation is disabled by default.
- Propagation can be enabled by the command line flag "--enable-arithmetic-propagation"
- Propagation interacts *heavily* with rewriting equalities, and will work best if the command line flag "--rewrite-arithmetic-equalities" is enabled.
Morgan Deters [Mon, 18 Apr 2011 08:59:09 +0000 (08:59 +0000)]
Partial merge from datatypes-merge branch:
1. Defines a new type "DatatypeType", a type-constant that holds a Datatype, describing an inductive data type.
2. CVC language parser supports datatypes.
3. CVC language printer now functional.
4. Minor other cleanups.
No performance impact is expected outside of datatypes. I'm verifying that that is the case with a cluster job this morning.
Christopher L. Conway [Mon, 18 Apr 2011 03:10:55 +0000 (03:10 +0000)]
Fixing output for EOF token in parser errors
Morgan Deters [Sat, 16 Apr 2011 00:12:01 +0000 (00:12 +0000)]
also a fix for a system test related to ParserBuilder
Morgan Deters [Sat, 16 Apr 2011 00:10:57 +0000 (00:10 +0000)]
unit test fixes for new NodeManager constructor (related to previous two trunk commits)
Morgan Deters [Fri, 15 Apr 2011 22:17:22 +0000 (22:17 +0000)]
parser/driver fixes for last commit
Morgan Deters [Fri, 15 Apr 2011 22:11:11 +0000 (22:11 +0000)]
partial merge from portfolio branch, adding conversions (library-internal-only of course) between Exprs and Nodes, Types and TypeNodes, ExprManagers and NodeManagers.
Dejan Jovanović [Thu, 14 Apr 2011 20:57:28 +0000 (20:57 +0000)]
reverting back the minisat code and adding a simpler one that shouldn't change the search
Morgan Deters [Thu, 14 Apr 2011 20:41:19 +0000 (20:41 +0000)]
Three things:
1. Infrastructure for unit T-conflicts added to SAT proxy
(and also the theory output channel documentation);
previously theories could not communicate unit T-conflicts
with the SAT layer because that layer had an implicit
assumption (not asserted) that the conflict nodes were an AND.
2. UF now pre-rewrites trivial equalities to "true". These could
conceivably occur in artificial benchmarks in this form:
(let (?x BIG-HUGE-TERM) ... (= ?x ?x) ... )
3. The SMT-LIBv2 printer now properly prints Bool constants.
Dejan Jovanović [Thu, 14 Apr 2011 19:11:42 +0000 (19:11 +0000)]
fixing an uninitialized literal variable
Dejan Jovanović [Wed, 13 Apr 2011 21:15:48 +0000 (21:15 +0000)]
adding support for unit conflicts in minisat...
Morgan Deters [Wed, 13 Apr 2011 06:25:37 +0000 (06:25 +0000)]
fix compiler warning in non-replay builds
Morgan Deters [Wed, 13 Apr 2011 06:10:55 +0000 (06:10 +0000)]
cache the LET rewriting (and defined-function expansion too)---it wasn't before, leading to terrible slowdown on heavily-nested LETs (and defined functions)
Morgan Deters [Wed, 13 Apr 2011 05:27:42 +0000 (05:27 +0000)]
add disequality token ("/=") and rules to CVC parser
Morgan Deters [Tue, 12 Apr 2011 18:45:44 +0000 (18:45 +0000)]
another small fix to "make dist" that can lead to a misconfigured tarball
Clark Barrett [Mon, 11 Apr 2011 17:33:07 +0000 (17:33 +0000)]
Transitive closure module is working
Morgan Deters [Mon, 11 Apr 2011 05:55:37 +0000 (05:55 +0000)]
fix "make dist" issues in makefiles
Morgan Deters [Sun, 10 Apr 2011 00:44:20 +0000 (00:44 +0000)]
merge from replay branch
Morgan Deters [Sun, 10 Apr 2011 00:14:49 +0000 (00:14 +0000)]
Add -lprofiler when --with-google-perftools is offered; also fix some newswire-raised documentation issues.
Dejan Jovanović [Sat, 9 Apr 2011 03:06:53 +0000 (03:06 +0000)]
changing the sat solver to assert propagated literals back to the theories
Clark Barrett [Fri, 8 Apr 2011 13:22:18 +0000 (13:22 +0000)]
Added util class
Tim King [Thu, 7 Apr 2011 19:28:25 +0000 (19:28 +0000)]
Made Valuation::getValue() and Valuation::getSatValue() const.
Morgan Deters [Tue, 5 Apr 2011 18:31:38 +0000 (18:31 +0000)]
Memory fix for congruence closure; affects many UF benchmarks, probably AX too.
Tim King [Tue, 5 Apr 2011 15:58:59 +0000 (15:58 +0000)]
Added options for setting the random decision frequency and random seed for the sat solver. Also added command line options for setting both.
Morgan Deters [Tue, 5 Apr 2011 04:06:10 +0000 (04:06 +0000)]
Minor adjustments to the Registrar commit in 1644, documentation.
Tim King [Mon, 4 Apr 2011 20:42:23 +0000 (20:42 +0000)]
Merging the satliteral-before-prereg branch into trunk. Theory preregistration is now called during the conversion to cnf. This fixes bug 257.
Tim King [Mon, 4 Apr 2011 20:18:11 +0000 (20:18 +0000)]
Reverts previous commit r1636.
Morgan Deters [Mon, 4 Apr 2011 18:59:33 +0000 (18:59 +0000)]
Add documentation to Node and TNode (closes bug #201).
Also, only build doxygen documentation on stuff in src/,
not test/ or contrib/ or anywhere else. Hopefully this
turns our 3000+ page user manual into something a little
more useful!
Tim King [Sat, 2 Apr 2011 21:05:47 +0000 (21:05 +0000)]
Delayed the addition of unate propagation lemmas until propagation is called. The OutputChannel is now untouched by TheoryArith during preregistration.
Morgan Deters [Sat, 2 Apr 2011 05:43:24 +0000 (05:43 +0000)]
with --with-google-perftools, don't just take it on blind faith, require a successful link at configure time
Morgan Deters [Sat, 2 Apr 2011 04:24:17 +0000 (04:24 +0000)]
minor fixes
Morgan Deters [Fri, 1 Apr 2011 21:35:50 +0000 (21:35 +0000)]
minor bugfixes (fixes broken dynamic-library build from last night)
Morgan Deters [Fri, 1 Apr 2011 01:49:09 +0000 (01:49 +0000)]
documentation fix
Morgan Deters [Fri, 1 Apr 2011 00:56:42 +0000 (00:56 +0000)]
This commit is a merge from the "betterstats" branch, which:
* Makes Options an "omnipresent thread-local global" (like the notion
of the "current NodeManager" was already). Options::current() accesses
this structure.
* Removes Options from constructors and data structures everywhere
(this cleans up a lot of things).
* No longer uses StatisticsRegistry statically. An instance of the
registry is created and linked to a NodeManager.
* StatisticsRegistry::current() is similar to Options::current(), but
the pointer is stowed in the NodeManager (rather than stored)
* The static functions of StatisticsRegistry have been left, for backward
compatibility; they now use the "current" statistics registry.
* SmtEngine::getStatisticsRegistry() is a public accessor for the
registry; this is needed by main() to reach in and get the registry,
for flushing statistics at the end.
Tim King [Thu, 31 Mar 2011 18:35:45 +0000 (18:35 +0000)]
Fixes to Valuation.
Morgan Deters [Wed, 30 Mar 2011 21:53:15 +0000 (21:53 +0000)]
improve recent low-coverage complaints
Dejan Jovanović [Wed, 30 Mar 2011 21:47:12 +0000 (21:47 +0000)]
adding CVC4:: qualifier to the #define for debugging so that it can be used outside of CVC4 namespace (like minisat)
Tim King [Wed, 30 Mar 2011 17:32:50 +0000 (17:32 +0000)]
Moved the constructor for Options out of the header and into the cpp. For people who fiddle with default values set by the Options constructor, this will require significantly less recompiling.
Tim King [Wed, 30 Mar 2011 15:07:02 +0000 (15:07 +0000)]
Added the command line flag --rewrite-arithmetic-equalities. This sets a static flag in Options that the ArithRewriter uses to determine the equality rewriting policy.
Morgan Deters [Wed, 30 Mar 2011 03:59:05 +0000 (03:59 +0000)]
Add Valuation::getSatValue() so that theories can access the current
(propositional) assignment for theory atoms.
Fixed Debug/Trace as discussed in bug ticket #252 and on the mailing list.
This implementation leads to some compiler warnings in production builds,
but these will be corrected in coming days. There appears to be a small
speedup in the parser as a result of this fix:
http://goedel.cims.nyu.edu/regress-results/compare_jobs.php?job_id=1902&reference_id=1886&p=5
Cleaned up a few CD Boolean attribute things.
Various small fixes to coding guidelines / test coverage.
This commit:
* Resolves bug 252 (tracing not disabled in production builds)
* Resolves bug 254 (implement CDAttrHash<>::BitIterator::find())
Tim King [Wed, 30 Mar 2011 01:06:37 +0000 (01:06 +0000)]
Merged the branch sparse-tableau into trunk.
Morgan Deters [Sun, 27 Mar 2011 19:02:02 +0000 (19:02 +0000)]
fixes to attribute-internals warnings on 64-bit; also some GCC function attribute cleanup; nothing major
Dejan Jovanović [Sat, 26 Mar 2011 22:26:06 +0000 (22:26 +0000)]
fix for bug 253, was propagating an asserted literal
also fixing some compile warnings in attributes
Morgan Deters [Sat, 26 Mar 2011 02:38:08 +0000 (02:38 +0000)]
fix typo
Morgan Deters [Fri, 25 Mar 2011 05:32:31 +0000 (05:32 +0000)]
This is a merge from the "theoryfixes+cdattrhash" branch. The changes
are somewhat disparate but belonged on the same branch because they were
held back from trunk all for the same reason (to keep the trunk stable
for furious bitvector development). Dejan has now given me the go-ahead
for a merge.
=========================================
THIS COMMIT CHANGES THE THEORY INTERFACE!
=========================================
Theory constructors are expected to take an additional "Valuation*"
parameter that each Theory should send along to the base class
constructor. The base class Theory keeps the Valuation* in a
d_valuation field for use by it and by its derived classes.
Theory::getValue() no longer takes a Valuation* (it is expected
to use d_valuation instead). This allows other theory functions
to take advantage of getValue() for debugging or heuristic
purposes.
TODO BEFORE MERGE TO TRUNK:
****implement BitIterator find() in CDAttrHash<bool>.
Specifically:
* Added QF_BV support for SMT-LIB v2.
* Two adjustments to the theory interface as requested by Tim King:
1. As described above.
2. Theories now have const access to the fact queue through base
class functions facts_begin() and facts_end(); useful for
debugging.
* Added an "Asserted" attribute so that theories can check if something
has been asserted or not (and therefore not propagate it). However, this
has been disabled for now, pending more data on the overhead of it, and
pending discussion at the 3/25/2011 meeting.
* Do not define NDEBUG in MiniSat in assertion-enabled builds (so
that MiniSat asserts are evaluated).
* As a result of the new MiniSat assertions, some --incremental
regressions had to be disabled; also, some bitvectors ?!!
* Bug 71 is resolved by adding a specialization for CDAttrHash<> in the
attribute package.
* Fixes for some warnings flagged by clang.
* System tests have arrived! So far mainly infrastructure for having
system tests, but there is a system test aimed at improving code
coverage of the printer package.
* Minor other adjustments to documentation and coding to be more
conformant to CVC4 policy.
Tests have been performed to demonstrate that these changes have no or
negligible effect on performance. In particular, changing the
CDAttrHash<> doesn't have any real effect on performance or memory right
now, since there is only one context-dependent boolean flag (as soon
as another is added, the effect is noticeable but probably still slight).
Morgan Deters [Fri, 25 Mar 2011 05:20:56 +0000 (05:20 +0000)]
Fix for a bug Andrew Reynolds found for iterators that affects empty CDList<> objects that allocate from ContextMemoryAllocator<>. Iterators were broken in that begin() != end() for empty lists (again---only those that allocated space from ContextMemoryAllocator<>). Added a unit test for this, too. Thanks Andy!
Tim King [Tue, 22 Mar 2011 13:37:11 +0000 (13:37 +0000)]
Merges the small changes on the queue-period branch into trunk. This branch importantly removes an unintentional line of code that had it pivoting more times than intended before rechecking the queue. Importantly, it does this without losing any examples with rewrite-equality enabled. This adds a parameter NUM_CHECKS which determines how many times the queue chould be checked during difference mode. A value of 10 for NUM_CHECKS has been empirically determined to be good in practice. See jobs 1815, 1824, 1825, 1821, 1814.
Dejan Jovanović [Tue, 22 Mar 2011 02:11:09 +0000 (02:11 +0000)]
updating debug output usage to eliviate impact of bug 252
Dejan Jovanović [Mon, 21 Mar 2011 18:55:05 +0000 (18:55 +0000)]
more bugfixes, some basic propagation, and testcases to cover them
Dejan Jovanović [Mon, 21 Mar 2011 17:43:22 +0000 (17:43 +0000)]
fixing a bug in the BV rewrite, off by one error when merging constants
Dejan Jovanović [Sun, 20 Mar 2011 19:58:51 +0000 (19:58 +0000)]
again a typo
Dejan Jovanović [Sun, 20 Mar 2011 19:50:48 +0000 (19:50 +0000)]
more bugfixes for bitvectors
Dejan Jovanović [Sun, 20 Mar 2011 16:31:19 +0000 (16:31 +0000)]
fixing the failure from last nigth, due to using a reference to an element in a growing vector
Dejan Jovanović [Sun, 20 Mar 2011 01:41:48 +0000 (01:41 +0000)]
missed one case
Dejan Jovanović [Sun, 20 Mar 2011 01:12:31 +0000 (01:12 +0000)]
commit for the version of bitvectors that passes all the unit tests
Tim King [Sat, 19 Mar 2011 22:59:39 +0000 (22:59 +0000)]
Merges the pqueue-set branch into trunk. During VarOrder mode and Collection mode, the arithmetic priority queue is maintained as a set. Compare jobs 1781 and 1782 for the expected performance change.
Tim King [Fri, 18 Mar 2011 22:22:28 +0000 (22:22 +0000)]
- The learned clauses from the miplib trick were being added twice. This was slowing down the search. (The effect can be seen in the difference between jobs 1765 and 1755). This happened during commit -r1480 when adding the ArithStaticLearner. This has been fixed.
Tim King [Thu, 17 Mar 2011 21:28:50 +0000 (21:28 +0000)]
Switched SimplexDecisionProcedure::d_delayedLemmas from a vector to a queue.
Tim King [Thu, 17 Mar 2011 21:20:28 +0000 (21:20 +0000)]
SimplexDecisionProcedure no longer takes an OutputChannel as a parameter.
Tim King [Thu, 17 Mar 2011 20:38:32 +0000 (20:38 +0000)]
- Removes arith_constants.h
- Adds ArithStaticLearner. Consolidates and cleans up the code for static learning in arithmetic. Static learning is now associated with a small amount of state between calls. This is used to track the data for the miplib trick. The goal is to make this inference work without relying on the fact that all of the miplib problem is asserted under the same AND node.
- This commit contains miscellaneous other arithmetic cleanup.
Tim King [Thu, 17 Mar 2011 20:24:05 +0000 (20:24 +0000)]
Adds debugging output to EngineOutputChannel::lemma.
Tim King [Thu, 17 Mar 2011 15:23:53 +0000 (15:23 +0000)]
Fix for the bug introduced in 1477. The stuff that was added to CVC4ostream::operator<< should only be executed when d_os != NULL. This was the cause of the NULL pointer dereference in debug builds.
Dejan Jovanović [Thu, 17 Mar 2011 03:48:53 +0000 (03:48 +0000)]
push and pop manipulators for output stream so that one can indent the output
ueful for me, maybe someone else finds it useful also
Tim King [Wed, 16 Mar 2011 22:08:15 +0000 (22:08 +0000)]
- Turns on the excluded middle assertions during the miplibTrick. If it is known that x \in {c_i}, then x is not in the interval (c_{i}, c_{i+1}) (assuming the c_i's are sorted). (Compare jobs 1742 and 1739 for the expected performance change on trunk. Compare jobs 1740 and 1738 for the expected performance change with the rewrite-equality patch.)
Tim King [Wed, 16 Mar 2011 15:36:45 +0000 (15:36 +0000)]
- Turns on the miplibTrick. This detects during the static learning phase a set of nodes that are asserted to the theory of the form (=> p_i (= x c_i)). If (or p_1 p_2 ...) is a tautology, then x \in {c_1, c_2, ...}. (This tautology check currently requires CUDD to be installed.) Right now all this does is assert x \leq max{c_i} and x \geq min{c_i}. (Compare jobs 1728 to 1626 for how this affects the miplib examples.)
Dejan Jovanović [Tue, 15 Mar 2011 22:08:46 +0000 (22:08 +0000)]
real fix for bug 245, previous one was faulty
Morgan Deters [Tue, 15 Mar 2011 21:06:47 +0000 (21:06 +0000)]
small fixes for run_regression script to workaround bug in old mktemp, was causing a hang in bug220.smt2
Dejan Jovanović [Tue, 15 Mar 2011 20:48:57 +0000 (20:48 +0000)]
fix for bug 254, lemmas were propagating at lower levels, and the conflict clauses asserting literal would overwrite the propagated literal
let's see if i break the build again
Morgan Deters [Tue, 15 Mar 2011 20:32:13 +0000 (20:32 +0000)]
Merge from cudd branch. This mostly just adds support for linking
against cudd libraries, the propositional_query class (in util/),
which uses cudd if it's available (and otherwise answers UNKNOWN for
all queries), and the arith theory support for it (currently disabled
per Tim's request, so he can clean it up).
Other changes include:
* contrib/debug-keys - script to print all used keys under Debug(), Trace()
* test/regress/run_regression - minor fix (don't export a variable)
* configure.ac - replace a comment removed by dejan's google perf commit
* some minor copyright/documentation updates, and minor changes to source
text to make 'clang --analyze' happy.
Dejan Jovanović [Mon, 14 Mar 2011 22:14:47 +0000 (22:14 +0000)]
adding support for google performance tools to the build sytem, it can be enabled at configure with
--with-google-perftools
to use it on ubuntu, you need to install packages google-perftools and libgoogle-perftools0
to run the profiling of the heap, you can run it for example with
HEAPPROFILE=/tmp/profile ./builds/bin/cvc4 test/regress/regress0/lemmas/clocksynchro_5clocks.main_invar.base.smt
this will create some files /tmp/profile* that you can then
to get the pdf of the profile you can then run
google-pprof --pdf ./builds/bin/cvc4 /tmp/profile.0001.heap > profile.pdf
or for other options check http://goog-perftools.sourceforge.net/doc/
Morgan Deters [Mon, 14 Mar 2011 21:00:34 +0000 (21:00 +0000)]
change to the run_regression script to better manage temporary files; hopefully this will somewhat alleviate the problem with all the junk files in /tmp
Morgan Deters [Mon, 14 Mar 2011 19:51:26 +0000 (19:51 +0000)]
Fix to bug 251 (non-spurious warnings in builds) by shifting metakind array by 1 to handle the UNDEFINED_KIND case.
Morgan Deters [Thu, 10 Mar 2011 21:05:13 +0000 (21:05 +0000)]
Fix bug 246 (occasional buffer overflow related to varargs in assertion-failure string construction) and addition of an assert_white unit test check for the issue
Morgan Deters [Thu, 10 Mar 2011 07:28:32 +0000 (07:28 +0000)]
ITE removal in TheoryEngine was not properly handling PARAMETERIZED kinds. Fixed and added bug regression. Thanks Andrew Reynolds for the bug report!
Morgan Deters [Tue, 8 Mar 2011 21:43:45 +0000 (21:43 +0000)]
Clean up Theory base class as per code review bug #60; also fixes to CodeTimer statistic, and adding a CodeTimer to TheoryEngine::EngineOutputChannel::newFact() for investigation into (possible) slow or redundant theory registration.
Tim King [Tue, 8 Mar 2011 01:46:31 +0000 (01:46 +0000)]
- Merges queue-interrogation branch into the trunk. This branch adds extra phases of looking for additional conflicts during and after the heuristic pivoting stage. (For the expected performance gain, comparing jobs 1676 and 1643 gives a rough idea.)
Tim King [Mon, 7 Mar 2011 19:10:16 +0000 (19:10 +0000)]
Merges branches/arithmetic/tableau-reset into the trunk. The tableau is now heuristically reset to its initial state during restarts.
Tim King [Sat, 5 Mar 2011 23:10:41 +0000 (23:10 +0000)]
- Adds PermissiveBackArithVarSet. This is very similar to ArithVarSet. The difference is that set.isMember(x) for an ArithVar x s.t. x > set.allocated() returns false for PermissiveBackArithVarSet and is an assertion failure for ArithVarSet. This cuts down on the memory usage of the ColumnMatrix slightly.
Tim King [Sat, 5 Mar 2011 19:30:06 +0000 (19:30 +0000)]
Enables the PreferenceFunction minBoundAndRowCount.
Tim King [Sat, 5 Mar 2011 17:33:01 +0000 (17:33 +0000)]
- Adds "PreferenceFunction" to SimplexDecisionProcedure. A PreferenceFunction allows for specifying how to choose between two nonbasic variables for which should become basic during the selectSlack(...) function. This partially addresses a point brought up by Dejan during the Code Review. (Unfortunately, function pointers are involved in the implementation. Because of this, I have had Morgan review this code before check-in.)
Tim King [Sat, 5 Mar 2011 16:59:13 +0000 (16:59 +0000)]
- Made Rational::sgn() function const.
Morgan Deters [Sat, 5 Mar 2011 00:03:08 +0000 (00:03 +0000)]
adding three features to CVC parser that drastically improve its support for the language: LET now supported (but not "type-lets" yet), OPTION now supported, and ^ operator (exponentiation) supported for nonnegative integer exponents. The latter simply expands to MULT, and of course fails assertions in arithmetic if a variable is raised to n >= 2. Also other CVC parser clean-up.
Morgan Deters [Thu, 3 Mar 2011 18:35:17 +0000 (18:35 +0000)]
fix for bug #244, "Segfault if file cannot be found and --stats is on"