| Age | Commit message (Collapse) | Author |
|---|
|
This highlights that this is a binary mode changing the interpretation
of "?x" rather than additionally allowing patvar.
|
|
The new function is interp_glob_closure which is basically a renaming
and generalization of interp_uconstr.
Note a change of semantics that I could however not observe in
practice.
Formerly, interp_uconstr discarded ltac variables bound to names for
interning, but interp_constr did not. Now, both discard them.
We also export the new interp_glob_closure.
|
|
This simplifies the API as before, inference of instances of type
classes was iff a type constraint was given.
We then export these both versions of interp_open_constr.
|
|
This module collects the functions of Nameops which are about Name.t
and somehow standardize or improve their name, resulting in particular
from discussions in working group.
Note the use of a dedicated exception rather than a failwith for
Nameops.Name.out.
Drawback of the approach: one needs to open Nameops, or to use long
prefix Nameops.Name.
|
|
On a machine for which ocamlopt is available, the make world will now
perform bytecode compilation only in grammar/ (up to the syntax
extension grammar.cma), and then exclusively use ocamlopt.
In particular, make world do not build bin/coqtop.byte.
A separate rule 'make byte' does it, as well as bytecode plugins and
things like dev/printers.cma.
'make install' deals only with the part built by 'make', while a new
rule 'make install-byte' installs the part built by 'make byte'.
IMPORTANT: PLEASE AVOID doing things like 'make -j world byte' or any
parallel mix of native and byte rules. These are known to crash sometimes,
see below. Instead, do rather 'make -j && make -j byte'.
Indeed, apart from marginal compilation speed-up for users not interested
in byte versions, the main reason for this commit is to discourage any
simultaneous use of OCaml native and byte compilers. Indeed, ocamlopt and
ocamlc will both happily destroy and recreate .cmi for .ml files with no .mli,
and in case of parallel build this may happen at the very moment another
ocaml(c|opt) is accessing this .cmi. Until now, this issue has been
handled via nasty hacks (see the former MLWITHOUTMLI and HACKMLI vars in
Makefile.build). But these hacks weren't obvious to extend to ocamlopt
-pack vs. ocamlopt -pack.
coqdep_boot takes a "-dyndep" option to control precisely how a Declare ML
Module influences the .v.d dependency file. Possible values are:
-dyndep opt : regular situation now, depends only on .cmxs
-dyndep byte : no ocamlopt, or compilation forced to bytecode, depends on .cm(o|a)
-dyndep both : earlier behavior, dependency over both .cm(o|a) and .cmxs
-dyndep none : interesting for coqtop with statically linked plugins
-dyndep var : place Makefile variables $(DYNLIB) and $(DYNOBJ) in .v.d
instead of extensions .cm*, so that the choice is made in the rest of the
makefile (see a future commit about coq_makefile)
NB: two extra mli added to avoid building unecessary .cmo during 'make world',
without having to use the ocamldep -native option.
NB: we should state somewhere that coqmktop -top won't work unless
'make byte' was done first
|
|
This allows a better control on the name to give to an evar and, in
particular, to address the issue about naming produced by "epose
proof" in one of the comment of Zimmi48 at PR #248 (see file names.v).
Incidentally updating output of Show output test (evar numbers shifted).
|
|
evars.
This is for consistency with the rest of the language. For instance,
"eremember" and "epose" are supposed to refer to terms occurring in
the goal, hence not leaving evars, hence in general pointless.
Eventually, I guess that "e" should be a modifier (see e.g. the
discussion at #3872), or the difference is removed.
|
|
a goal with unresolved evars.
|
|
|
|
Two issues in one:
- some focused_simpl were called on the wrong locations
- some focused_simpl were done on whole equations
In the two cases, this could be bad if "simpl" goes too far
with respect to what omega expects: later calls to "occurrence"
might fail. This may happen for instance if an atom isn't a variable,
but a let-in (b:=5:Z in the example).
|
|
|
|
|
|
Constrintern.pf_global returns a global_reference, not a constr,
adapt plugins accordingly, properly registering universes where
necessary.
|
|
|
|
|
|
Deprecations which can't be fixed in 4.02.3 are locally wrapped with
[@@@ocaml.warning "-3"]. The only ones encountered are
- capitalize to capitalize_ascii and variants. Changing to ascii would
break coqdoc -latin1 and maybe other things though.
- external "noalloc" to external [@@noalloc]
|
|
|
|
This is the continuation of #244, we now deprecate `CErrors.error`,
the single entry point in Coq is `user_err`.
The rationale is to allow for easier grepping, and to ease a future
cleanup of error messages. In particular, we would like to
systematically classify all error messages raised by Coq and be sure
they are properly documented.
We restore the two functions removed in #244 to improve compatibility,
but mark them deprecated.
|
|
We move Coqlib to library in preparation for the late binding of
Gallina-level references. Placing `Coqlib` in `library/` is convenient
as some components such as pretyping need to depend on it.
By moving we lose the ability to locate references by syntactic
abbreviations, but IMHO it makes to require ML code to refer to
a true constant instead of an abbreviation/notation.
Unfortunately this change means that we break the `Coqlib`
API (providing a compatibility function is not possible), however we
do so for a good reason.
The main changes are:
- move `Coqlib` to `library/`.
- remove reference -> term from `Coqlib`. In particular, clients will
have different needs with regards to universes/evar_maps, so we
force them to call the (not very safe) `Universes.constr_of_global`
explicitly so the users are marked.
- move late binding of impossible case from `Termops` to
`pretying/Evarconv`. Remove hook.
- `Coqlib.find_reference` doesn't support syntactic abbreviations
anymore.
- remove duplication of `Coqlib` code in `Program`.
- remove duplication of `Coqlib` code in `Ltac.Rewrite`.
- A special note about bug 5066 and commit 6e87877 . This case
illustrates the danger of duplication in the code base; the solution
chosen there was to transform the not-found anomaly into an error
message, however the general policy was far from clear. The long
term solution is indeed make `find_reference` emit `Not_found` and
let the client handle the error maybe non-fatally. (so they can test
for constants.
|
|
We remove redundant functions `coq_constant`, `gen_reference`, and
`gen_constant`.
This is a first step towards a lazy binding of libraries references.
We have also chosen to untangle `constr` from `Coqlib`, as how to
instantiate the reference (in particular wrt universes) is a
client-side issue. (The client may want to provide an `evar_map` ?)
c.f. #186
|
|
|
|
resolution trace
|
|
|
|
|
|
|
|
|
|
The trace only mentions the constant k by which we want to divide
the equation, not anymore the equation we obtain after the division.
Shorter trace, and it won't take much more time to perform the few
Z.div than checking as currently the equality of the initial equation
and the final equation multiplied by k.
|
|
Inspired by https://coq.inria.fr/bugs/show_bug.cgi?id=5229 , which
this PR solves, I propose to remove support for non-synchronous
options.
It seems the few uses of `optsync = false` we legacy and shouldn't
have any impact.
Moreover, non synchronous options may create particularly tricky
situations as for instance, they won't be propagated to workers.
|
|
|
|
|
|
|
|
We benefit from the fact that we normalize now *all* hypotheses
even the one defining the "stated" variable: it is produced as
...def of v... = v
and normalized as
-v + ...def of v... = 0
which is precisely what we should add to the initial equation during
a O_STATE.
|
|
Now that O_SUM is properly optimized (cf. the [fusion] function),
we could use it to encode CONTRADICTION and NEGATE_CONTRADICT(_INV).
This way, the trace has almost the same size, but ReflOmegaCore.v
is shorter and easier to maintain.
|
|
|
|
|
|
|
|
In this variant, the proof term produced by romega isn't verified at
the tactic run-time (no vm_compute). In theory, [unsafe_romega] should
behave exactly as [romega], but faster. Now, if there's a bug in romega,
we'll be notified only at the following Qed. This could be interesting
for debugging purpose : you could inspect the produced buggish term
via a Show Proof.
|
|
This is a major change :
- Generated proofs are quite shorter, since only the resolution trace remains.
- No time penalty mesured (it even tends to be slightly faster this way).
- Less infrastructure in ReflOmegaCore and refl_omega.
- Warning: the normalization functions in ML and in Coq should be kept
in sync, as well as the variable order.
- Btw: get rid of ML constructor Oufo
|
|
|
|
In a coming commit, we'll normalize terms by a Coq function
that will compare Tvar's instead of blindly applying a trace,
so let's speed-up these comparisons.
|
|
|
|
|
|
|
|
|
|
|
|
For the bug, see new test test_romega10 in test-suite/success/ROmega0.v.
|
|
|
|
|
|
|
|
|
