| Age | Commit message (Collapse) | Author |
|---|
|
Fixes 2887.
|
|
|
|
Before refactoring the preprocessing passes, we were using three
arguments to add assertions to the decision engine. Now all that
information lives in the AssertionPipeline. This commit moves the
AssertionPipeline to its own file and changes the `addAssertions()`
methods related to the decision engine to take an AssertionPipeline as
an arguement instead of three separate ones. Additionally, the
TheoryEngine now uses an AssertionPipeline for lemmas.
|
|
|
|
Adds missing override keywords.
|
|
|
|
* Replacing __gnu_cxx::hash_map with std::unordered_map.
* Replacing __gnu_cxx::hash_set with std::unordered_set.
* Replacing __gnu_cxx::hash with std::hash.
* Adding missing includes.
|
|
|
|
|
|
What to Know As a User:
A number of files have moved. Users that include files in the public API in more refined ways than using #include <cvc4.h> should consult which files have moved. Note though that some files may move again after being cleaned up. A number of small tweaks have been made to the swig interfaces that may cause issues. Please file bug reports for any problems.
The Problem:
The build order of CVC4 used to be [roughly] specified as:
options < expr < util < libcvc4 < parsers < main
Each of these had their own directories and their own Makefile.am files. With the exception of the util/ directory, each of the subdirectories built exactly one convenience library. The util/ directory additionally built a statistics library. While the order above was partially correct, the build order was more complicated as options/Makefile.am executed building the sources for expr/Makefile.am as part of its BUILT_SOURCES phase. This options/Makefile.am also build the options/h and options.cpp files in other directories. There were cyclical library dependencies between the first four above libraries. All of these aspects combined to make options extremely brittle and hard to develop. Maintaining these between clang versus gcc, and bazel versus autotools has become increasing unpredictable.
The Solution:
To address these cyclic build problems, I am simplifying the build process. Here are the main things that have to happen:
1. util/ will be split into 3 separate directories: base, util, and smt_util. Each will have their own library and Makefile.am file.
2. Dependencies for options/ will be moved into options/. If a type appears as an option, this file will be moved into options.
3. All of the old options_handlers.h files have been refactored.
4. Some files have moved from util into expr/ to resolve cycles. Some of these moves are temporary.
5. I am removing the libstatistics library.
The constraints that the CVC4 build system will eventually satisfy are:
- The include order for both the .h and .cpp files for a directory must respect the order libraries are built. For example, a file in options/ cannot include from the expr/ directory. This includes built source files such as those coming from */kinds files and */options files.
- The types definitions must also respect the build order. Forward type declarations will be allowed in exceptional, justified cases.
- The Makefile.am for a directory cannot generate a file outside of the directory it controls. (Or call another Makefile.am except through subdirectory calls.)
- One library per Makefile.am.
- No extra copies of libraries will be built for the purpose of distinguishing between external and internal visibility in libraries for building parser/ or main/ libraries and binaries. Any function used by parser/ and main/ will be labeled with CVC4_PUBLIC and be in a public API. (AFAICT, libstatistics was being built exactly to skirt this.)
The build order of CVC4 can now be [roughly] specified as
base < options < util < expr < smt_util < libcvc4 < parsers < main
The distinction between "base < options < util < expr" are currently clean. The relationship between expr and the subsequent directories/libraries are not yet clean.
More details about the directories:
base/
The new directory base/ contains the shared utilities that are absolutely crucial to starting cvc4. The list currently includes just: cvc4_assert.{h,cpp}, output.{h,cpp}, exception.{h,cpp}, and tls.{h, h.in, cpp}. These are things that are required everywhere.
options/
The options/ directory is self contained.
- It contains all of the enums that appear as options. This includes things like theory/bv/bitblast_mode.h .
- There are exactly 4 classes that handled currently using forward declarations currently to this: LogicInfo, LemmaInputChannel, LemmaOutputChannel, and CommandSequence. These will all be removed from options.
- Functionality of the options_handlers.h files has been moved into smt/smt_options_handler.h. The options library itself only uses an interface class defined in options/options_handler_interface.h. We are now using virtual dispatch to avoid using inlined functions as was previously done.
- The */options_handlers.h files have been removed.
- The generated smt/smt_options.cpp file has been be replaced by pushing the functionality that was generated into: options/options_handler_{get,set}_option_template.cpp . The non-generated functionality was moved into smt_engine.cpp.
- All of the options files have been moved from their directories into options/. This means includes like theory/arith/options.h have changed to change to options/arith_options.h .
util/
The util/ directory continues to contain core utility classes that may be used [almost] everywhere. The exception is that these are not used by options/ or base/. This includes things like rational and integer. These may not use anything in expr/ or libcvc4. A number of files have been moved out of this directory as they have cyclic dependencies graph with exprs and types. The build process up to this directory is currently clean.
expr/
The expr/ directory continues to be the home of expressions. The major change is files moving from util/ moving into expr/. The reason for this is that these files form a cycle with files in expr/.
- An example is datatype.h. This includes "expr/expr.h", "expr/type.h" while "expr/command.h" includes datatype.h.
- Another example is predicate.h. This uses expr.h and is also declared in a kinds file and thus appears in kinds.h.
- The rule of thumb is if expr/ pulls it in it needs to be independent of expr/, in which case it is in util/, or it is not, in which case it is pulled into expr/.
- Some files do not have a strong justification currently. Result, ResourceManager and SExpr can be moved back into util/ once the iostream manipulation routines are refactored out of the Node and Expr classes.
- Note the kinds files are expected to remain in the theory/ directories. These are only read in order to build sources.
- This directory is not yet clean. It contains forward references into libcvc4 such as the printer. It also makes some classes used by main/ and parser CVC4_PUBLIC.
smt_util/
The smt_util/ directory contains those utility classes which require exprs, but expr/ does not require them. These are mostly utilities for working with expressions and nodes. Examples include ite_removal.h, LemmaInputChannel and LemmaOutputChannel.
What is up next:
- A number of new #warning "TODO: ..." items have been scattered throughout the code as reminders to myself. Help with these issues is welcomed.
- The expr/ directory needs to be cleaned up in a similar to options/. Before this happens statistics needs to be cleaned up.
|
|
|
|
|
|
|
|
This is done only in "hard" case. Limited testing has not shown
improvement in the "easy" case.
This was triggerred by a benchmark sent by andy/viktor.
performance comparison notes for the change on wiki
http://church.cims.nyu.edu/wiki/User:Kshitij/decisioncacheindex
|
|
None of these are enabled by default, so any performance impact
counts as a bug
Options added are:
--decision-threshold=N :default 0
+ ignore all nodes greater than threshold in first attempt to pick decision
--decision-use-weight bool :default false
+ use the weight nodes (locally, by looking at children) to direct recursive search
--decision-random-weight=N int :default 0
+ assign random weights to nodes between 0 and N-1 (0: disable)
--decision-weight-internal=HOW
+ computer weights of internal nodes using children: off, max, sum, usr1 (meaning evolving)
Squashed commit of the following:
commit 0dbae066c19abde37092517b50f23255398539db
Author: Kshitij Bansal <kshitij@cs.nyu.edu>
Date: Fri Apr 26 16:42:36 2013 -0400
contentless cleanup
commit 62bb99b33deceb803ba5afc563fd322b4b5d1b7e
Author: Kshitij Bansal <kshitij@cs.nyu.edu>
Date: Tue Apr 16 21:43:55 2013 -0400
bugfixes in usr1 auto weight computation
commit 9f039cba805bfd722466734920e758d48ae3b23e
Author: Kshitij Bansal <kshitij@cs.nyu.edu>
Date: Fri Mar 29 15:01:33 2013 -0400
DECISION_WEIGHT_INTERNAL_USR1
commit 744e16d514594e5f1c69b36473b03cf501d9b9d1
Author: Kshitij Bansal <kshitij@cs.nyu.edu>
Date: Wed Mar 27 11:05:43 2013 -0400
split theory and decision requests
commit f379d8a821df31c74b42a7722e891abc5c944f16
Author: Kshitij Bansal <kshitij@cs.nyu.edu>
Date: Wed Mar 27 09:51:58 2013 -0400
fix potential bug with threshold
commit 3dcb45eb5ee648d3edbeddf76b838076afea3d12
Author: Kshitij Bansal <kshitij@cs.nyu.edu>
Date: Wed Feb 27 20:29:38 2013 -0500
stat bv::weightComputationTimer
commit 2ab97d063e221357d2bb017af4589105777fd5a3
Author: Kshitij Bansal <kshitij@cs.nyu.edu>
Date: Sat Feb 23 17:02:43 2013 -0500
decision: option to auto compute weight of boolean structure
commit 0a8c29e699ad96d5f73bc14d31ad9254f6711ae8
Author: Kshitij Bansal <kshitij@cs.nyu.edu>
Date: Sat Feb 23 14:53:50 2013 -0500
decision: fix design to do partial explorations
* make findSplitterRec and all related helper functions' return
type trivalued, to be able to distinguish between
"partial exploration" vs "done exploration but found nothing"
* keep additional data structure to remember to what extent the
partial exploration has been completed so not to repeat it. we
can use this to make multiple passes on formula with arbritrary
order of thresholds for exploration
commit 0815991fc1b0f1d63f0e8124d4672d782e89d671
Author: lianah <lianahady@gmail.com>
Date: Fri Feb 22 17:55:40 2013 -0500
added simple node weight computation for bv.
commit e4c507e2e2fdc8794fd04c31093660a80c7f44c3
Author: Kshitij Bansal <kshitij@cs.nyu.edu>
Date: Wed Feb 20 02:35:21 2013 -0500
--decision-use-weight, --decision-random-weight=N
commit 0624177d66d6ed2b3cc7fdb13df775990cfe50c2
Author: Kshitij Bansal <kshitij@cs.nyu.edu>
Date: Tue Feb 19 23:36:49 2013 -0500
decisionThreshold option
commit ac3579a52e452e3118ce116ff1823d6c6885544b
Author: Kshitij Bansal <kshitij@cs.nyu.edu>
Date: Tue Feb 19 20:22:51 2013 -0500
DecisionWeightAttr
|
|
|
|
|
|
|
|
|
|
other minor: cleanup some remaning fragments of GiveUpException(),
hopefully all is gone now.
|
|
|
|
|
|
|
|
|
|
(cherry picked from commit f056522a587d1b080224992355be070b73d97a3b)
|
|
|
|
CDInsertHashMap. CDHashSet<TNode> have been changed to CDHashSet<Node>. Switching CnfStream to use CDInsertSet. Switches a few CDHashMaps in arithmetic to use CDTrailHashMap. Documentation changes to CDHashMap.
|
|
- Fix a destruction order issue this triggered in DE
(this commit was certified error- and warning-free by the test-and-commit script.)
|
|
Add notice/warning when using incremental-mode + decision (it was
already disabled)
Some other minor cleanup
(this commit was certified error- and warning-free by the test-and-commit script.)
|
|
just the header comments at the top, though. Don't update to this rev if
you don't have time for a complete rebuild, and exclude this rev if you
want to see what's new across a range of commits.
(this commit was certified error- and warning-free by the test-and-commit script.)
|
|
Statistics-printing problem, limited to certain benchmarks).
Mark some unlabeled header files "cvc4_private.h".
Other minor cleanup.
(this commit was certified error- and warning-free by the test-and-commit script.)
|
|
|
|
\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
|
|
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.
|
|
|
|
(gets rid of warning)
|
|
Major changes from last merge
* ITEs supported
* Don't share theory lemmas to DE, only assertions
Should probably be noted that 'make regress' doesn't quite
pass with --decision=justification. Throws off search in couple
of arith benchmarks.
No serious performance changes expected. Keep an eye.
|
|
Overview of changes
* command line option --decision={internal,justification}
* justification heuristic handles all operators except ITEs
revelant stats: decision::jh::*
* if decisionEngine has solved the problem PropEngine returns
unknown and smtEngine queries DE to get the answer
relevant stat: smt::resultSource
* there are known bugs
Full list of commits being merged
r3330 use CD data structures in JH
r3329 add command-line option --decision=MODE
r3328 timer stat, other fixes
r3326 more trace
r3325 enable implies, iff, xor (no further regression losses)
r3324 feed decision engine lemmas, changes to quitting mechanism
r3322 In progress
r3321 more fixes...
r3318 bugfix1 (69 more to go)
r3317 Handle other boolean operators in JH (except ITE)
r3316 mechanism for DE to stopSearch
r3315 merge from trunk + JH translation continuation
r3275 change option to enable JH by default[A
|
|
Adds DecisionEngine and an abstract class DecisionStrategy
which other strategies will derive from eventually.
Full revision summary of merged commits:
r3241 merge from trunk
r3240 fix
r3239 WIP
r3238 JH, CVC3 code: 5% done -- 5% translated
r3237 JH groundwork
r3236 make make regrss pass
r3234 hueristic->heuristic
r3229 JustificationHeuristic: EOD-WIP
r3228 DecisionEngine: hookup assetions
r3227 move ITE outside simplifyAssertions
r3226 DecisionStrategy abstract class
r3222 DecisionEngine: begin
|
