| Age | Commit message (Collapse) | Author |
|---|
|
This intermediate representation serves two purposes:
1- It is a preliminary step for primitive machine integers, as iterators
will be compiled to Clambda.
2- It makes the VM compilation passes closer to the ones of
native_compute. Once we unifiy the representation of values, we should
be able to factorize the lambda-code generation between the two
compilers, as well as the reification code.
This code was written by Benjamin Grégoire and myself.
|
|
|
|
camlp4
|
|
longer use camlp4.
|
|
|
|
The current error mechanism in the core part of Coq is 100% exception
based; there was some confusion in the past as to whether raising and
exception could be replace with `Feedback.msg_error`.
As of today, this is not the case [due to some issues in the layer
that generates error feedbacks in the STM] so all cases of `msg_error`
must raise an exception of print at a different level [for now].
|
|
The old semantics of `with/without_option` allowed the called function
to modify the value of the option. This is an issue mainly with the
`silently/verbose` combinators, as `Set Silent` can be executed under
one of them and thus the modification will be lost in the updated code
introduced in a554519874c15d0a790082e5f15f3dc2419c6c38
IMHO these kind of semantics are quite messy but we have to preserve
them in order for the `Silent` system to work. In fact, note that in
the previous code, `with_options` was not consistent with
`with_option` [maybe that got me confused?]
Ideally we could restore the saner semantics once we clean up the
`Silent` system [that is, we remove the flag altogether], but that'll
have to wait.
Fixes #6645.
|
|
|
|
|
|
|
|
... in favor of having Public/Internal sub modules in each and
every module grouping functions according to their intended client.
|
|
Some code in typeclasses was even breaking the invariant that
use_polymorphic_flag should not be called twice, but that code was
morally dead it seems, so we remove it.
|
|
possibly fixing printing errors (was: Removing failure of coq_makefile on no arguments)
|
|
One less global flag.
|
|
Up to this point the `lib` directory contained two different library
archives, `clib.cma` and `lib.cma`, which a rough splitting between
Coq-specific libraries and general-purpose ones.
We know split the directory in two, as to make the distinction clear:
- `clib`: contains libraries that are not Coq specific and implement
common data structures and programming patterns. These libraries
could be eventually replace with external dependencies and the rest
of the code base wouldn't notice much.
- `lib`: contains Coq-specific common libraries in widespread use
along the codebase, but that are not considered part of other
components. Examples are printing, error handling, or flags.
In some cases we have coupling due to utility files depending on Coq
specific flags, however this commit doesn't modify any files, but only
moves them around, further cleanup is welcome, as indeed a few files
in `lib` should likely be placed in `clib`.
Also note that `Deque` is not used ATM.
|
|
This was apparently either silently doing nothing or failing.
|
|
|
|
|
|
same right-hand side.
|
|
|
|
|
|
|
|
These are convenient to use `command.ml` for example.
We also fix a critical bug in the `fold_left_map` family of functions,
as witnessed by this old behavior.
```ocaml
fold_left2_map (fun c u v -> c+1,u+v) 0 [1;2;3] [1;2;3;];;
- : int * int list = (3, [6; 4; 2])
```
I have opted for a simple fix keeping the tail-recursive nature, I am
not in the mood of writing base libraries, but feel free to improve.
|
|
Adding a space before the bar separating disjunctive patterns.
Removing an extra space after the bar for inner disjunctive patterns.
|
|
And some code simplification.
|
|
We move the main async flags to the STM in preparation for
more state encapsulation.
There is still more work to do, in particular we should make some of
the defaults a parameter instead of a flag.
|
|
This brings us one step closer to actually moving all STM flags to
`stm`.
|
|
We remove coqmktop in favor of a couple of simple makefile rules using
ocamlfind. In order to do that, we introduce a new top-level file that
calls the coqtop main entry.
This is very convenient in order to use other builds systems such as
`ocamlbuild` or `jbuilder`.
An additional consideration is that we must perform a side-effect on
init depending on whether we have an OCaml toplevel available [byte]
or not. We do that by using two different object files, one for the
bytecode version other for the native one, but we may want to review
our choice.
We also perform some smaller cleanups taking profit from ocamlfind.
|
|
New module introduced in OCaml 4.05 I think, can create problems when
linking with the OCaml toplevel for `Drop`.
|
|
|
|
|
|
The patch has three parts:
- Introduction of a configure flag `-bytecode-compiler (yes|no)`
(due to static initialization this is a configure-time option)
- Installing the hooks that register the VM with the pretyper and the
kernel conditionally on the flag.
- Replacing the normalization function in `Redexpr` by compute if the
VM is disabled.
We also rename `Coq_config.no_native_compiler` to `native_compiler`
and `Flags.native_compiler` to `output_native_objects` [see #4607].
|
|
|
|
We also remove some internal implementation details from the mli file,
there due historical reasons.
|
|
|
|
|
|
While we are at it we refactor a few special cases of the helper.
|
|
|
|
|
|
|
|
|
|
This is useful for tools such as `coqchk` or `coq_makefile` that want
to handle feedback on their own.
|
|
|
|
|
|
4.02.3 has been the minimal OCaml version for a while now.
|
|
We make Vernacentries.interp functional wrt state, and thus remove
state-handling from `Future`. Now, a future needs a closure if it
wants to preserve state.
Consequently, `Vernacentries.interp` takes a state, and returns the
new one.
We don't explicitly thread the state in the STM yet, instead, we
recover the state that was used before and pass it explicitly to
`interp`.
I have tested the commit with the files in interactive, but we aware
that some new bugs may appear or old ones be made more apparent.
However, I am confident that this step will improve our understanding
of bugs.
In some cases, we perform a bit more summary wrapping/unwrapping. This
will go away in future commits; informal timings for a full make:
- master:
real 2m11,027s
user 8m30,904s
sys 1m0,000s
- no_futures:
real 2m8,474s
user 8m34,380s
sys 0m59,156s
|
|
STM (BZ#5556)
|
|
We move delicate library/module instillation code to the STM so the
API guarantees that the first state snapshot is correct. That was not
the case in the past, which meant that the STM cache was unsound in
batch mode, however we never used its contents due to not backtracking
to the first state.
This provides quite an improvement in the API, however more work is
needed until the codepath is fully polished.
This is a critical step towards handling the STM functionally.
|
|
The test was abandoned at the time of merging subdirectory browsing
between coqdep and coqtop, and to limit at the same time the
dependency of coqdep in files such as unicode.cmo.
But checking ident validity speeds up browsing in arbitrary directory
structure and we restore it for this reason.
(One could also say that browsing arbitrary directory structures is
not intended, but in practice this may happen, as e.g. reported in
BZ#5734.)
|
|
|
