| Age | Commit message (Collapse) | Author |
|---|
|
|
|
Copied from Coq backend.
|
|
These are the correct versions for div/mod in the SMT solver
|
|
inconsistent with ocaml etc.). Rename div and mod builtins to ediv_int/emod_int and tdiv_int/tmod_int and add corresponding implementations. Add a test with negative operands. This will break existing models but will mean users have to think about which versions they want and won't accidentally use the wrong one.
|
|
Most such patterns are re-written away by various re-writing steps,
but for those that arn't they are fairly easy to handle by just having
as patterns directly in the ANF-patterns.
Fixes #39
|
|
Issue with pretty-printing mapping clauses types fixed by both
cf7eb8b83 and 0e2f1710a, so remove the redundant clauses. In the
parser we allow both
```
| Mapping id
{ mk_sd (SD_mapping ($2, mk_tannotn)) $startpos $endpos }
| Mapping id Colon funcl_typ
{ mk_sd (SD_mapping ($2, $4)) $startpos $endpos }
```
so we should probably use doc_binding to correctly print any
quantifier on the funcl_typ. Although since polymorphism and mappings
don't play nicely together right now this likely doesn't occur very
often in practice.
|
|
|
|
Add a test case to ensure variable types in l-expressions remain the
same with flow-sensitive constraints.
|
|
See comments on e92ff687.
|
|
Fix scattered mapping printing and output message for missing val spec
|
|
|
|
In particular, spot variable shadowing and handle (n as 'm) patterns.
|
|
|
|
Remove unused experimental optimizations
|
|
|
|
Add a function Jib_optimize.inline which can inline functions. To make
this more efficient, we can make identifiers unique on a per-function
basis.
|
|
- add dummy print_bits function
- support int(1) like types in axioms
|
|
- skip a few more that aren't supported yet
- produce better debugging information (in particular, in the right order)
- avoid some autocasts that aren't supported yet and are usually
unnecessary
- Handle more constraints like `8 * n = 8 * ?Goal`
|
|
It helps the Coq backend if the shape of constraints embedded in types
doesn't change too much.
|
|
|
|
These should be preserved for prover backends.
|
|
|
|
|
|
Propagating constants into mutually recursive calls and removing dead branches
might break mutually recursive cycles.
Also make constant propagation use the existing interpreter-based constant
folding to evaluate function calls with only constant arguments (as opposed to
a mixture of inlining and hard-coded rewrite rules).
|
|
Rewrite <> true/false in goals.
Correct implicits in record and variant types.
Use expanded valspecs from the type checker in axioms.
Allow list notations in type definitions.
Skip some not-yet-supported tests.
|
|
|
|
Mostly just a small quality-of-life improvement
|
|
|
|
Avoids duplication between l-expressions and expressions. Also means that
special variables like current_exception and have_exception are treated
normally by functions such as instr_reads and instr_writes etc. Furthermore
we can now easily annotate Jib identifiers in ways that were not previously
possible with plain sail ids.
|
|
:def <definition> evaluates a top-level definition
:(b)ind <id> : <type> creates an identifier within the interactive type-checking environment
:let <id> = <expression> defines an identifier
Using :def the following now works and brings the correct vector
operations into scope.
:def default Order dec
:load lib/prelude.sail
Also fix a type-variable shadowing bug
|
|
|
|
Some SMT goals in unification would generate problems with missing
variables. Turns out the SMT solver would happily ignore this and
return the correct unsat/sat result anyway. However, this does affect
the error code from the solver so if we now check the return code we
must make sure that we don't generate any smtlib files that generate
warnings or errors.
Now that kopts_of_X functions exist in ast_util we can just use those
to get well-kinded variables from the constraint itself rather than
relying on the typechecker to pass in a list of variables which makes
things simpler to boot!
|
|
|
|
|
|
|
|
|
|
This is in order to allow register reads etc to be handled by RMEM
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Add a CL_void l-expression so we don't have redundant unit-typed
variables everywhere, and add an optimization in Jib_optimize called
optimize_unit which introduces these.
Remove the basic control-flow graph in Jib_util and add a new mutable
control-flow graph type in Jib_ssa which allows the IR to be converted
into SSA form. The mutable graph allows for more efficient updates,
and includes both back and forwards references making it much more
convenient to traverse.
Having an SSA representation should make some optimizations much
simpler, and is also probably more natural for SMT generation where
variables have to be defined once using declare-const anyway.
Debug option -ddump_flow_graphs now outputs SSA'd graphs of the
functions in a specification.
|
|
|
|
|
|
Fixes bad precedence issues, removes an out-of-date special case that's
not necessary, and solves more goals.
|
|
Check in a slightly nicer stylesheet for OCamldoc generated
documentation in etc. Most just add a maximum width and increase the
font size because the default looks absolutely terrible on high-DPI
monitors.
Move val_spec_ids out of initial_check and into ast_util where it
probably belongs. Rename some functions in util.ml to better match the
OCaml stdlib.
|
