Morgan Deters [Sun, 8 Jul 2012 20:06:22 +0000 (20:06 +0000)]
Minor changes to avoid some warnings on GCC 4.7.1 (Debian wheezy/sid). ANDY - please look at the diff and make sure I didn't do something stupid
Morgan Deters [Sun, 8 Jul 2012 17:36:50 +0000 (17:36 +0000)]
Bugs resolved by this commit: #314, #322, #359, #364, #365.
See below for details.
* Fix the "assert" name-collision bug (resolves bug #364).
Our identifiers should never be named "assert", as that's a preprocessor
definition in <assert.h>, which is often #included indirectly (so simply
having a policy of not including <assert.h> isn't good enough---one of
our dependences might include it). It was once the case that we didn't
have anything named "assert", but "assert()" has now crept back in.
Instead, name things "assertFoo()" or similar. Thanks to Tim for the
report.
To fix this, I've changed some of Dejan's circuit-propagator code from
"assert()" to "assertTrue()". Ditto for Andy's explanation manager.
Guys, if you prefer a different name in your code, please change it.
* Fix the incorrect parsing of lets in SMT-LIBv2 parser (resolves bug #365).
Inner lets now shadow outer lets (previously, they incorrectly gave an
error). Additionally, while looking at this, I found that a sequential let
was implemented rather than a parallel let. This is now fixed. Thanks to
Liana for the report.
* Remove ANTLR parser generation warnings in CVC parser (resolves bug #314).
* There were a lot of Debug lines in bitvectors that had embedded toString()
calls. This wasted a LOT of time in debug builds for BV benchmarks
(like in "make regress"). Added if(Debug.isOn(...)) guards; much faster
now.
* Support for building public-facing interface documentation only (as opposed
to all internals documentation). Now "make doc" does the public-facing and
"make doc-internals" does documentation of everything. (Along with changes
to the nightly build script---which will now build and publish both types
of Doxygen documentation---this resolves bug #359).
* Fix the lambda typechecking bug (resolves bug #322). Thanks to Andy for the
report (a long long time ago--sorry).
* The default output language for all streams is now based on the current set
of Options (if there is one). This has been a constant annoyance, especially
when stringstreams are used to construct output. However, it doesn't work
for calls from outside the library, so it's mainly an annoyance-fixer for
CVC4 library code itself.
* Add some CVC4_UNUSED markers to local variables in theory_arith.cpp that
are used only in assertions-enabled builds (and thus give warnings in
production builds). This was briefly discussed at the meeting this week.
Morgan Deters [Sat, 7 Jul 2012 21:01:33 +0000 (21:01 +0000)]
Various fixes to documentation---typos, some incomplete documentation fixed, \file tags corrected, copyright added to files that had it missing, etc.
I ensured that I didn't change any code with this commit, and even tested on the cluster to be doubly sure:
http://church.cims.nyu.edu/regress-results/compare_jobs.php?job_id=4655&reference_id=4646&p=0
Tim King [Fri, 6 Jul 2012 17:47:49 +0000 (17:47 +0000)]
Adding std namespace to a couple of make_pair instances.
Tim King [Fri, 6 Jul 2012 17:45:59 +0000 (17:45 +0000)]
Added include unistd to main/util.cpp
Tim King [Fri, 6 Jul 2012 17:44:12 +0000 (17:44 +0000)]
Added virtual destructor to PpRewrite.
Morgan Deters [Sun, 1 Jul 2012 00:04:49 +0000 (00:04 +0000)]
Some changes to configure.ac:
1. Includes BOOST_CPPFLAGS during compilation of all files, not just portfolio-relevant files. This is necessary since we now have a general dependence on Boost (not just its threading stuff). This resolves bug 357.
2. Support --disable-thread-support and --enable-thread-support, in an effort to get to the bottom of bug 361.
These changes shouldn't affect performance, though #1 will build the cvc4 libs with a couple of pthread definitions that conceivably could change the behavior of #included standard headers. Let's keep an eye on tonight's regressions.
Morgan Deters [Thu, 28 Jun 2012 12:32:05 +0000 (12:32 +0000)]
svn:ignore
Clark Barrett [Thu, 28 Jun 2012 11:26:07 +0000 (11:26 +0000)]
Fixed bug in bv rewriter that caused wrong answer in SMT-COMP
Morgan Deters [Thu, 28 Jun 2012 09:42:09 +0000 (09:42 +0000)]
fix a link error on church, due to Antlr #defining "true" and "false" :-( --for now, just #undef them after the #include
Tim King [Wed, 27 Jun 2012 22:45:51 +0000 (22:45 +0000)]
Fixing a bug in proof production for the DioSolver.
Tim King [Wed, 27 Jun 2012 21:06:29 +0000 (21:06 +0000)]
This adds TheoryArith::safeToReset(). This fixes bug 363.
Tim King [Wed, 27 Jun 2012 20:58:37 +0000 (20:58 +0000)]
Adding access to simplex's ArithPriorityQueue to TheoryArith for ArithPriorityQueue::reduce(), ::begin() and ::end().
Tim King [Wed, 27 Jun 2012 20:56:23 +0000 (20:56 +0000)]
Improved debugging output.
Tim King [Wed, 27 Jun 2012 20:56:04 +0000 (20:56 +0000)]
Improved debugging output.
Tim King [Wed, 27 Jun 2012 20:55:17 +0000 (20:55 +0000)]
Adding reduce() to the ArithPriorityQueue. This reduces the queue from a superset of the basic variables that violate a bound to the exact set.
Tim King [Mon, 25 Jun 2012 16:00:21 +0000 (16:00 +0000)]
Added a warning to arithmetic for a known dio solver bug. Somehow the fix never made it to trunk. Do not know how. The fix to the bug is pending the hunt for bug 363.
François Bobot [Fri, 22 Jun 2012 15:11:37 +0000 (15:11 +0000)]
TPTP: add parser for cnf and fof
- include directive works
- no keyword : 'fof', 'cnf', ... can be used for symbols name
- real -> unsorted -> real (for the one that appear, so no bijection bitween real and unsorted)
- same thing for string
But:
- string not distinct by projection to real, not sure if the current state of string theory make them distinct
- filtering in include is not done
- the result is not printed in the TPTP way (currently SMT2 way)
François Bobot [Fri, 22 Jun 2012 15:11:21 +0000 (15:11 +0000)]
parser: add some acces function and recover the original nextToken from antlr3
in order to be able to use the stack of streams.
François Bobot [Fri, 22 Jun 2012 15:11:16 +0000 (15:11 +0000)]
fix : function AntlrInput::tokenTextSubstr
François Bobot [Fri, 22 Jun 2012 15:11:11 +0000 (15:11 +0000)]
Parser: add the possibility to bind at level 0.
Morgan Deters [Mon, 18 Jun 2012 22:43:09 +0000 (22:43 +0000)]
qf_lra strategy
Clark Barrett [Mon, 18 Jun 2012 22:39:12 +0000 (22:39 +0000)]
Reverting buggy rewriter code
Morgan Deters [Mon, 18 Jun 2012 22:22:04 +0000 (22:22 +0000)]
another qf_lra strategy update
Clark Barrett [Mon, 18 Jun 2012 22:18:27 +0000 (22:18 +0000)]
Fixed bug in rewriter
Morgan Deters [Mon, 18 Jun 2012 21:55:25 +0000 (21:55 +0000)]
unnecessary ^ in regular expression; warning produced on smt-exec
Morgan Deters [Mon, 18 Jun 2012 21:48:02 +0000 (21:48 +0000)]
QF_LRA strategy in run script, now final (?) for smt-comp 2012
Morgan Deters [Mon, 18 Jun 2012 21:33:00 +0000 (21:33 +0000)]
final sources (?) for competition
Clark Barrett [Mon, 18 Jun 2012 19:59:56 +0000 (19:59 +0000)]
Fix for slow array rewrite and minor bug fix in arrays that popped up as a result
Clark Barrett [Mon, 18 Jun 2012 16:19:10 +0000 (16:19 +0000)]
small bug fix and performance fix in ite simplifier
Andrew Reynolds [Mon, 18 Jun 2012 06:04:21 +0000 (06:04 +0000)]
fixed smt version 1 parser for quantifiers
Kshitij Bansal [Mon, 18 Jun 2012 02:52:11 +0000 (02:52 +0000)]
tracing code to make sure decision options are being set correctly
Kshitij Bansal [Mon, 18 Jun 2012 02:04:37 +0000 (02:04 +0000)]
bugfix, enable only QF_LRA, not other arith
Kshitij Bansal [Mon, 18 Jun 2012 01:46:39 +0000 (01:46 +0000)]
QF_LRA, Quantifiers: enable use decision for (only for) stopping search
Morgan Deters [Mon, 18 Jun 2012 00:56:27 +0000 (00:56 +0000)]
Fixing bug 360. The driver wasn't exiting when there was an error (it just plowed ahead to the next command). Now the driver exits on the first error, unless it's in interactive mode.
Kshitij Bansal [Sun, 17 Jun 2012 23:02:01 +0000 (23:02 +0000)]
QF_AUFLIA: enable use decision for (only for) stopping search
Dejan Jovanović [Sun, 17 Jun 2012 22:33:31 +0000 (22:33 +0000)]
fixing a problem due to lemmas produced while backtracking
Kshitij Bansal [Sun, 17 Jun 2012 22:04:41 +0000 (22:04 +0000)]
--decision=justification-stoponly : use decision engine only for stopping
search early, not to make decisions
new options.h :)
Dejan Jovanović [Sun, 17 Jun 2012 18:22:01 +0000 (18:22 +0000)]
enabling theoryof=term for quantifiers with sharing
disableing one test case in equantifiers/decision that runs long
Dejan Jovanović [Sun, 17 Jun 2012 16:08:38 +0000 (16:08 +0000)]
fixing wrong assertion
Clark Barrett [Sun, 17 Jun 2012 02:29:14 +0000 (02:29 +0000)]
Removed assertion that can fail
Dejan Jovanović [Sun, 17 Jun 2012 01:59:53 +0000 (01:59 +0000)]
fixing makefile error that brakes build
Clark Barrett [Sun, 17 Jun 2012 01:49:58 +0000 (01:49 +0000)]
Fix array bug causing incorrect answers
Dejan Jovanović [Sat, 16 Jun 2012 23:58:07 +0000 (23:58 +0000)]
small change to equality assertions so that one doesn't get x = y and y = x
Kshitij Bansal [Sat, 16 Jun 2012 23:04:15 +0000 (23:04 +0000)]
Adding the failing QF_AUFLIA regression mentioned in last commit.
pp-regfile.delta02.smt is the one to look at with
--decision=justificaiton, the delta minimized version of pp-regfile,
which also gives wrong answer. due to various commits/fixes, delta01
gives correct answer currently.
Kshitij Bansal [Sat, 16 Jun 2012 22:44:20 +0000 (22:44 +0000)]
This is an attempt to fix the bug in the justification heuristic. The
other minor change is removing dependency of header file in
theory_engine.h
It fixes all known wrong answers in QF_BV and QF_AUFBV. It doesn't fix a
wrong answer for QF_AUFLIA -- it is currently unclear what is the cause
of this bug, could be sharing.
Performance impact (turns out) is none on QF_BV and QF_AUFBV (except, of
course, those for which the answer was wrong earlier; and also perhaps
one or two off-cases)
The issue was with how the infinite loop in justification stuff was prevented.
To keep it short, I skip what was wrong earlier, and this is what is done
now:
* whenever an atomic formula is seen, a list of pairs of <IteSkolemVariable,
AssertionCorrespondingToIteSkolem> is created for each skolem occuring in
the atom.
* we iterate over this list, doing the following: check if skolem is marked
as visited. if not, mark visited, recurse, when back unmark.
I lied, I will tell you what was being done earlier was -- 1. the check for
not being visited was being done in each recursive call, not just for atoms.
2. The AssertionCorrespondingToIteSkolem was being used to check if something
is visited and not IteSkolemVariable. I don't know if this makes any
difference, but anyhow, I think this is cleaner & clearer, so I keep this.
Morgan Deters [Sat, 16 Jun 2012 22:29:44 +0000 (22:29 +0000)]
updated build script for smt-comp submission
Dejan Jovanović [Sat, 16 Jun 2012 21:35:05 +0000 (21:35 +0000)]
changing theoryOf in shared mode with arrays to move equalities to arrays
disabled in bitvectors due to non-stably infinite problems
the option to enable it is --theoryof-mode=term
Tim King [Sat, 16 Jun 2012 17:37:33 +0000 (17:37 +0000)]
Fixing if condition for trivial equalities in arithmetic. Also some whitespace issues in smt_engine.cpp.
Kshitij Bansal [Fri, 15 Jun 2012 22:15:43 +0000 (22:15 +0000)]
Bug fix in justification heuristic. Had to do with how
a "visited" node in the recursive findSplitter method was
handled (which would lead to infinite loop). Earlier,
they were ignored assuming the ancestor would split on it
later. The right thing to do is to split on it right away.
(Fixes errors from the fuzzer, not the ones from last night's
regression)
Clark Barrett [Fri, 15 Jun 2012 20:51:05 +0000 (20:51 +0000)]
Reverting rewrite rule to working version
Clark Barrett [Fri, 15 Jun 2012 19:05:56 +0000 (19:05 +0000)]
Fixes some assertion failures
Clark Barrett [Fri, 15 Jun 2012 03:24:51 +0000 (03:24 +0000)]
Fix for incompleteness bug with decision engine: repeated simplification
could introduce additional assertions that were not beign processed by the
decision engine. Now these assertions are merged in with pre-ITE-removal
assertions, ensuring the decision engine sees them.
Tim King [Fri, 15 Jun 2012 02:16:41 +0000 (02:16 +0000)]
Fixing mac compilation issues.
Kshitij Bansal [Fri, 15 Jun 2012 00:15:49 +0000 (00:15 +0000)]
one bug fixed
Kshitij Bansal [Thu, 14 Jun 2012 23:49:22 +0000 (23:49 +0000)]
WIP
Dejan Jovanović [Thu, 14 Jun 2012 22:20:15 +0000 (22:20 +0000)]
fixing the problems with the bvminisat. there was a case when things would get bitblasted, it would restart to add the clauses, and loose propagation information.
Kshitij Bansal [Thu, 14 Jun 2012 21:35:18 +0000 (21:35 +0000)]
add failing regression, move error up
Tim King [Thu, 14 Jun 2012 21:24:44 +0000 (21:24 +0000)]
Fixing a case for explanation of non-normal form equalities.
Kshitij Bansal [Thu, 14 Jun 2012 21:08:28 +0000 (21:08 +0000)]
bug fixes in justification heuristic
* remove assert iteSkolemMap gives ite-s (not true with repeatSimp)
* handle a corner case in findSplitter triggered by repeatSimp
Tim King [Thu, 14 Jun 2012 20:59:14 +0000 (20:59 +0000)]
Fixing a bug related to explaining propagations with non-normalized witnesses.
Dejan Jovanović [Thu, 14 Jun 2012 19:54:25 +0000 (19:54 +0000)]
enabling fixed bug345 case
Dejan Jovanović [Thu, 14 Jun 2012 19:46:22 +0000 (19:46 +0000)]
fixes for the hasTerm issues in the shared database under the decision heuristic
Clark Barrett [Thu, 14 Jun 2012 19:37:31 +0000 (19:37 +0000)]
New substitutions implementation - fixes performance issue seen in nonclausal
simplification for some benchmarks
Tim King [Thu, 14 Jun 2012 19:29:25 +0000 (19:29 +0000)]
Fixed arithmetic consistency issue. The simplex conflict variable had to be reenqueued so that the queue was a superset of the failing assertions. This adds a super expensive debug routine unenqueuedVariablesAreConsistent() that catches this bug. This is enabled when -d arith::consistency is turned on. make check passes with this flag enabled.
Dejan Jovanović [Thu, 14 Jun 2012 19:09:37 +0000 (19:09 +0000)]
fix for clark's bug
imporant: theories SHOULD NOT use getSatValue at all, otherwise they might create a conflict with a literal they didn't get()
Morgan Deters [Thu, 14 Jun 2012 18:59:18 +0000 (18:59 +0000)]
don't run rewriterules regressions by default; fixes needed
Kshitij Bansal [Thu, 14 Jun 2012 18:51:34 +0000 (18:51 +0000)]
fix quantifier non-bug
Clark Barrett [Thu, 14 Jun 2012 17:55:08 +0000 (17:55 +0000)]
Removed an assertion, unneeded header file
Morgan Deters [Thu, 14 Jun 2012 17:42:47 +0000 (17:42 +0000)]
some changes to make CVC4 work nicely with trace executor for application track; (set-option :print-success true) supported, (exit) causes immediate exit regardless of EOF, etc.
Morgan Deters [Thu, 14 Jun 2012 17:25:22 +0000 (17:25 +0000)]
making --simplification=none the default for quantified logics; this a request from andy. evidence of performance improvement: church.cims.nyu.edu/regress-results/compare_jobs.php?job_id=4516&reference_id=4475&p=5
Kshitij Bansal [Thu, 14 Jun 2012 17:14:30 +0000 (17:14 +0000)]
bug ifx, mv
Kshitij Bansal [Thu, 14 Jun 2012 17:06:14 +0000 (17:06 +0000)]
restore destruction of stuff in driver
Kshitij Bansal [Thu, 14 Jun 2012 16:46:44 +0000 (16:46 +0000)]
This commit:
* enables decision heuristic (justification) for QF_BV and QF_AUFBV
* disables a failing regression in aufbv (because of equality engine
assert failure trigerred by above change)
* moves around the init procedure smt_engine
* destruction time issues because of moving this -- still to be fixed,
currently get around by not destucting stuff in driver
Kshitij Bansal [Thu, 14 Jun 2012 14:57:24 +0000 (14:57 +0000)]
failing quantifier
Morgan Deters [Thu, 14 Jun 2012 14:32:51 +0000 (14:32 +0000)]
The "no-tears-in-competition-mode" commit. Change all (non-driver, non-SAT-solver) uses of std::cout to the Message stream, and all uses of std::cerr to the Warning stream.
Kshitij Bansal [Thu, 14 Jun 2012 13:47:33 +0000 (13:47 +0000)]
fix cout, fix statname, rm deadcode
Kshitij Bansal [Thu, 14 Jun 2012 09:50:14 +0000 (09:50 +0000)]
add passing regression
Dejan Jovanović [Thu, 14 Jun 2012 07:26:09 +0000 (07:26 +0000)]
changing to a more natural propagation order in uf, seems to pay off
Dejan Jovanović [Thu, 14 Jun 2012 06:44:49 +0000 (06:44 +0000)]
some changes to the uf engine
* dramatically less terms to manage by doing reflexivity semantically
* fixes the problem clark had with not detecting inconsistencies with shared terms
i'm not sure what's the performance impact, but this is so much better and we'll deal with performance later
Tim King [Thu, 14 Jun 2012 04:39:43 +0000 (04:39 +0000)]
Brings the tuning branch into trunk. This includes the changes from restricted-simplex.
Morgan Deters [Thu, 14 Jun 2012 01:40:51 +0000 (01:40 +0000)]
bug 346 resolved
Dejan Jovanović [Thu, 14 Jun 2012 01:08:11 +0000 (01:08 +0000)]
* removing rewriteEquality from the rewriter
* theories now get either an assertion from the SAT solver (normalized) or an (dis-)equality between two shared terms that is non-normalized
Dejan Jovanović [Wed, 13 Jun 2012 22:49:16 +0000 (22:49 +0000)]
removing bug233 until morgan commits the actual file
Morgan Deters [Wed, 13 Jun 2012 22:17:47 +0000 (22:17 +0000)]
adding some regressions to the usual regressions runs; several recently-fixed incremental bugs are closed
Tim King [Wed, 13 Jun 2012 21:30:05 +0000 (21:30 +0000)]
Added witnesses to Constraints.
Tim King [Wed, 13 Jun 2012 20:37:43 +0000 (20:37 +0000)]
- Added a loop to internally assert constraints that are marked as true.
- Changes how CongruenceManager reports conflicts.
- ConstraintDatabase can now detect and raise conflicts when doing unate propagation.
Tim King [Wed, 13 Jun 2012 20:32:52 +0000 (20:32 +0000)]
Fixed whitespace warning on Makefile.
Tim King [Wed, 13 Jun 2012 20:32:04 +0000 (20:32 +0000)]
Adds debugging output to theory_engine.cpp.
Morgan Deters [Wed, 13 Jun 2012 20:27:58 +0000 (20:27 +0000)]
revisions to the "make submission" target
Morgan Deters [Wed, 13 Jun 2012 19:50:35 +0000 (19:50 +0000)]
Don't use the "inlined" feature of ANTLR 3.2, which causes a buffer overflow condition when reading from stdin. This should completely resolve bug #319.
However, on large inputs especially (like the stp/testcase benchmarks), this inlining feature can speed parsing by 5-10%, at the cost of not supporting interactive sessions on stdin (like in the SMT-COMP application track).
So I updated the submission script and competition build so that
* a competition build with antlr-inlining is built for the main and parallel tracks
* a competition build without antlr-inlining is built for the application track
Again, the effect is only when reading the stdin stream (but that's how SMT-COMP works). For normal (non-competition) builds, we need to support interactive sessions (from e.g. KIND) on stdin, so this inlining is off for all builds except main- and parallel-track competition builds.
Also added a "get-antlr-3.4" script that automatically downloads and locally installs a copy of libantlr3c and the antlr parser generator inside the CVC4 source tree.
Closing bug #319.
Kshitij Bansal [Wed, 13 Jun 2012 18:24:30 +0000 (18:24 +0000)]
add passing regression
Dejan Jovanović [Wed, 13 Jun 2012 18:08:09 +0000 (18:08 +0000)]
fix for bug 354
Clark Barrett [Wed, 13 Jun 2012 17:06:08 +0000 (17:06 +0000)]
Fixed failing assertion when EqualityEngine is in conflict
Dejan Jovanović [Wed, 13 Jun 2012 15:50:56 +0000 (15:50 +0000)]
enabling regressions from last night, all fixed
Kshitij Bansal [Wed, 13 Jun 2012 15:36:03 +0000 (15:36 +0000)]
enable some decision regressions
Kshitij Bansal [Wed, 13 Jun 2012 15:31:38 +0000 (15:31 +0000)]
Make d_result in DE context dependent
(fixes bugs in bv, others with JH on)
Clark Barrett [Wed, 13 Jun 2012 14:25:53 +0000 (14:25 +0000)]
Fixed definition of bvsmod
Kshitij Bansal [Wed, 13 Jun 2012 14:00:35 +0000 (14:00 +0000)]
decision regressions, all but one fail
Clark Barrett [Wed, 13 Jun 2012 13:54:14 +0000 (13:54 +0000)]
Fixes more problems in bv rewrites