| Age | Commit message (Collapse) | Author |
|---|
|
|
|
|
|
(especially as the environment previously used was a bit dodgy)
|
|
|
|
(previously the typechecker did this for all literal patterns, but now
it's only necessary for the rewritten arguments)
|
|
This commit primarily changes how existential types are bound in
letbindings. Essentially, the constraints on both numeric and
existentially quantified types are lifted into the surrounding type
context automatically, so in
```
val f : nat -> nat
let x = f(3)
```
whereas x would have had type nat by default before, it'll now have
type atom('n) with a constraint that 'n >= 0 (where 'n is some fresh
type variable). This has several advantages: x can be passed to
functions expecting an atom argument, such as a vector indexing
operation without any clunky cast functions - ex_int, ex_nat, and
ex_range are no longer required. The let 'x = something() syntax is
also less needed, and is now only really required when we specifically
want a name to refer to x's type. This changes slightly the nature of
the type pattern syntax---whereas previously it was used to cause an
existential to be destructured, it now just provides names for an
automatically destructured binding. Usually however, this just works
the same.
Also:
- Fixed an issue where the rewrite_split_fun_constr_pats rewriting
pass didn't add type paramemters for newly added type variables in
generated function parameters.
- Updated string_of_ functions in ast_util to reflect syntax changes
- Fixed a C compilation issue where elements of union type
constructors were not being coerced between big integers and 64-bit
integers where appropriate
- Type annotations in patterns now generalise, rather than restrict
the type of the pattern. This should be safer and easier to handle
in the various backends. I don't think any code we had was relying
on this behaviour anyway.
- Add inequality operator to lib/flow.sail
- Fix an issue whereby top-level let bindings with annotations were
checked incorrectly
|
|
(when they're not relevant)
|
|
|
|
|
|
Now it just returns the actual arguments and a separate function
calculates the start index when required.
|
|
|
|
|
|
Fix a bug in initial check which caused X() = y to expect an additional parameter.
Some tweaks to sail2 emacs mode
|
|
More work on Latex output
|
|
Now generate commands for each toplevel definition, such that e.g. the
function clause for execute LOAD could be inserted using
\sailexecuteLOAD. Tries to generate fairly intuitive names while
avoiding clashes where possible.
|
|
Rename l2.ott to sail.ott
|
|
definitions from sail/lib.
|
|
|
|
Turn on complex nexp rewriting for mono by default
(NB: solving is currently quite slow, will optimise)
|
|
(also reorder the phases a little)
|
|
|
|
|
|
New function Type_check.solve : Env.t -> nexp -> Big_int.num option.
Takes an environment and an n-expression (nexp), and returns either
Some u, where u is a unique solution such that nexp = u, or None which
indicates that either no unique solution could be found. It is
technically possible that a unique solution could exist, but Z3 may
not find it.
Involves two calls to Z3, one of which cannot be memoised, so should
be used carefully, as over-reliance could lead to performance issues.
|
|
|
|
(for monomorphisation, off for now because the analysis needs extended).
Also tighten up orig_nexp, make Lem backend replace # in type variables.
|
|
|
|
Allows the type checker to deal with
val foo : forall 'm 'n, 'n = 8 * 'm. atom('m) -> bits('n)
for example
|
|
|
|
Added library for simple integer arithmetic functions in lib/arith.sail
WIP TeX file for formatting latex output included in lib/sail.tex
Fixes for bugs in sail_lib
|
|
|
|
|
|
First, the specialisation of option types has been fixed by allowing
the specialisation of constructor return types - this essentially
means that a constructor, such as Some : 'a -> option('a) can get
specialised to int -> option(int), rather than int -> option('a). This
means that these constructors are treated like GADTs internally. Since
this only happens just before the C translation, I haven't put much
effort into making this very robust so far.
Second, there was a bug in C compilation for the typing of return
expressions in non-unit contexts, which has been fixed.
Finally support for vector literals that are non-bitvectors has been
added.
|
|
|
|
In particular, improve indentation of if-expressions, and provide infix syntax
for monadic binds in Isabelle, allowing Lem to preserve source whitespace.
|
|
Can now compile RISCV. Requires some library tweaks before it'll pass any tests,
Also adds hyperlinks to wip latex output
|
|
Fixes #11
|
|
|
|
Added option -latex that outputs input to a latex document.
Added doc comments that can be attached to certain AST nodes - right now just valspecs and function clauses, e.g.
/*!
Documentation for main
*/
val main : unit -> unit
These comments are kept by the sail pretty printer, and used when generating latex
|
|
Comment out partially working optimisation passes for now
|
|
|
|
Doesn't remove them from function bodies because that can produce
more work for the sizeof rewriting.
|
|
find_opt is only available from OCaml 4.05
|
|
- Update Lem bindings and extras files
- Rewrite Nexp_var's if they are bound to a constant, similar to Nexp_id's
(used for cap_size in the CHERI spec)
- Add Lem and Isabelle Makefile targets for CHERI
|
|
Some functions are partial, e.g. converting a bitvector to an integer, which
might fail for the bit list representation due to undefined bits. Undefined
cases can be handled in different ways:
- call Lem's failwith, which maps to undefined/ARB in Isabelle and HOL (the
default so far),
- return an option type,
- raise a failure in the monad, or
- use a bitstream oracle to resolve undefined bits.
This patch adds different versions of partial functions corresponding to those
options. The desired behaviour can be selected by choosing a binding in the
Sail prelude. The naming scheme is that the failwith version is the default,
while the other versions have the suffixes _maybe, _fail, and _oracle,
respectively.
|
|
For example, generates an auxiliary function execute_ADD (rs, rt, rd) for the
clause execute (ADD (rs,rt,rd)) = ...
Without this rewriting, the execute function easily becomes too large to be
handled by Isabelle (e.g., for CHERI-MIPS; for MIPS alone, it seems to be just
about small enough).
This used to be implemented in the pretty-printer, but that code was commented
out recently in order to support a recursive execute function for RISC-V
compressed instructions.
|
|
Effectful expressions are monadic in Lem, causing type errors when binding them
to global immutable variables.
|
|
The state monad currently assumes that tags are written to and read from
properly aligned addresses (since it does not know the capability size used in
the Sail model). This change allows the Sail model to pass in the aligned
address(es) even if data is written to an unaligned address. There might be
better ways to model tag writing, but this approach seems rather general.
|
|
This simplifies reasoning in Isabelle.
|
|
|
|
Fixed an issue whereby an option constructor that was never
constructed, but only matched on, would cause compilation to
fail. Temporarily fixed an issue where union types that can be
entirely stack-allocated were not being treated as such, by simply
heap-allocating all unions. Need to adapt the code generator to handle
this case properly. Fixed a further small issue whereby multiple union
types would confuse the type specialisation pass.
Added a test case for compiling option types.
RISCV now generates C code, but there are still some bugs that need to
be squashed before it compile and work.
|
