-
Add a test that verifies `dafny server` produces the same SMTLib as `dafny verify`
-
I was looking at the following output from Owi: `Assert failure: (bool.eq (i32.add (i32.and symbol_0 symbol_1) (i32.shr (i32.xor symbol_0 symbol_1) (i32 1))) (i32.div (i32.add symbol_0 symbol_1) (i32 …
-
The type would have similar status to the `Integer` type where it would be mostly used for writing specifications rather than generating code. It would enable to us to interface with the correspondin…
-
`bv2nat` is part of the official SMTLib logic for bitvectors: http://smtlib.cs.uiowa.edu/theories-FixedSizeBitVectors.shtml Though, it looks like Yices doesn't support it. Do you plan to support it? I…
-
The current implementation of Alt-Ergo doesn't support the float theory https://smtlib.cs.uiowa.edu/theories-FloatingPoint.shtml but
a theory about round functions from the `real` to subset of `real…
-
In SMTLIB the modulo operator uses Boute's Euclidean definition (http://smtlib.cs.uiowa.edu/theories-Ints.shtml).
In JS, it uses its own interpretation. Critically, for any x, ```x%0 == NaN```. In …
-
The current `Expr` module does not expose smart constructors for FP theory as the `Z3.FloatingPoint` does. I noticed
this lack while writing the interface for AE in the project https://github.com/wa…
-
It would be good to not just run ESBMC with its default solver (Z3 on a Windows CI build, Boolector for other CI builds, depending on configure-time options otherwise) over the solver-agnostic regress…
-
Still not clear on how an automated setup could work, the libs are definitely too big for the free CI on github.
Gbury updated
4 years ago
-
In my attempts to explore issue #304 I used the GNUMakefile's bitops verification step to export the z3v4.5+SMTlib theory output and then tried to get z3 to verify this output from the command line "b…