aboutsummaryrefslogtreecommitdiff
path: root/proofs
AgeCommit message (Collapse)Author
2013-11-02Fix destruct: nf_evar prior to tactic interpretation.aspiwack
Noticed in CoRN git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@17028 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Add primitives in Goal.V82 to access the goal in nf_evar'd form.aspiwack
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@17025 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Adds a tactic give_up.aspiwack
Gives up on the focused goals. Shows an unsafe status. Unlike the admit tactic, the proof cannot be closed until the users goes back and solves these goals. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@17018 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02A tactic shelve_unifiable.aspiwack
Puts on the shelf every goals under focus on which other goals under focus depend. Useful when we want to solve these goals by unification (as in a first order proof search procedure, for instance). Also meant to be able to recover approximately the semantics of the old refine with the new implementation (use refine t; shelve_unifiable). TODO: bug dans l'example de shelve_unifiable git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@17017 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Corrects a bug on Proofview.Goal.enter whereby sigmas were captured and used ↵aspiwack
at the wrong time. The bug was masked by the fact that Tacinterp uses many superfluous Proofview.Goal.enter, it so happens that the tactic Proofview.Goal.enter (fun _ -> Proofview.Goal.enter fun gl -> t)) had the correct semantics! git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@17014 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Adds a shelve tactic.aspiwack
The shelve tactic puts all the focused goals out of sight. They can be later recalled by the Unshelve command. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@17013 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02bootstrap/Monad.v: implements the writer monad in continuation passing style.aspiwack
Benefits: fewer pair constructed/destructed especially in split. Potential costs: plus and zero now have closures with 11 arguments. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@17012 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02bootstrap/Monad.v: implements the environment monad in continuation passing ↵aspiwack
style. Benefits: the underlying monads are not referenced in the "current" primitive. Potential costs: some extracted functions now have 9 arguments, Ocaml may not be good at handling these. The split primitive, which is called often, now builds one extra closure. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@17011 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Adds an experimental exactly_once tactical.aspiwack
exactly_once t, will have a success if t has exactly once success. There are a few caveats: - The underlying effects of t may happen in an unpredictable order (hence it may be wise to use it only with "pure" tactics) - The second success of a tactic is conditional on the exception thrown. In Ltac it doesn't show, but in the underlying code, the tactical also expects the exception you want to use to produce the second success. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@17009 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Normalise goals with respect to evars in the new tactics.aspiwack
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@17008 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Made Proofview.Goal.hyps a named_context.aspiwack
There was really no point in having it be a named_context val. The tactics are not going to access the vm cache. Only vm_compute will. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@17007 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02A dedicated view type for Proofview_gen.split.aspiwack
It doesn't seem to affect performances. But the generated code is slightly cleaner. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@17005 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Adds a tactical once.aspiwack
[once t] does just as [t] but has exactly one success it [t] has at least one success. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@17004 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Less use of the list-based interface for goal-bound tactics.aspiwack
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@17002 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Tachmach.New is now in Proofview.Goal.enter style.aspiwack
As a result the use of the glist-style interface for manipulating goals has almost been removed. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@17001 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02More Proofview.Goal.enter.aspiwack
Proofview.Goal.enter is meant to eventually replace the Goal.sensitive monad. This commit changes the type of Proofview.Goal.enter from taking a four argument function (environment, evar_map, hyps, concl) from a one argument function of abstract type Proofview.Goal.t. It will be both more extensible and more akin to old-style tactics. This commit also changes the type of Proofview.Goal.{concl,hyps,env} from monadic operations to projection from a Proofview.Goal.t. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@17000 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02A better version of Goal.advance.aspiwack
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16999 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02bootstrap/Monads.v: A more efficient split.aspiwack
Exchanges a IO.bind for a Logic.bind. The latter is better inlined. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16996 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02State monad implemented in CPS.aspiwack
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16995 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02A more principled split.aspiwack
Instead of interleaving the reifications and reflection steps, split is defined as a series of reification followed by a series of reflection. The immediate consequence is that split is hoisted out of the Logic interface, and defined in a single block at the end of the file. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16994 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Set an extraction flag for inling let-s in Monad.v.aspiwack
Reduces the size of split significantly. In particular it now uses only 2 matches instead of 4. Patch by Pierre-Marie Pédrot and Pierre Letouzey. Signed-off-by: Arnaud Spiwack <arnaud@spiwack.net> git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16993 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02More optimisations of partial applications.aspiwack
This time in Goal. Patch by Pierre-Marie Pédrot. Signed-off-by: Arnaud Spiwack <arnaud@spiwack.net> git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16992 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Try to remove intermediate allocations when dealing with goal-specific tactics.aspiwack
Introduces a primitive Goal.enter which allows to access the common information needed by goal-specific tactics, avoids a number of monadic binds, and some unnecessary allocations of lists. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16991 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Various rewriting, mostly for speed purposes.aspiwack
- A variant of tclEVARS directly in the language of the monad - A variant of tclDISPATCHGEN (tclINDEPENDENT) hopefully faster in the case there is only one tactic to copy - A better written tclDISPATCHGEN (which may make thing actually a little slower) - A special case in tclDISPATCHGEN and tclINDEPENDENT for the case when they are 0 or 1 goals (adaptation of a patch sent by Pierre-Marie Pédrot) git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16990 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Optimisation of partial applications in the tactic monad.aspiwack
Explanations here: https://ocaml.janestreet.com/?q=node/30 Patch by Pierre-Marie Pédrot, with modifications from Arnaud Spiwack. Signed-off-by: Arnaud Spiwack <arnaud@spiwack.net> git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16989 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Makes the Ltac debugger usable again.aspiwack
This is just a port of the existing design. Basing the tactics on an IO monad may allow to simplify things a bit. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16985 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02New option Default Goal Selector.aspiwack
Set Default Goal Selector "all" prefixes all tactics with "all:" if no selector is specified (it is overridden by 1: for instance). git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16983 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Adds a new goal selector "all:".aspiwack
all:tac applies tac to all the focused subgoals. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16982 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Replaced monads.ml by an essentially equivalent proofview_gen.ml generated ↵aspiwack
by extraction. The goal was to use Coq's partial evaluation capabilities to do manually some inlining that Ocaml couldn't do. It may be critical as we are defining higher order combinators in term of others and no inlining means a lot of unnecessary, short-lived closures built. With this modification we get back some (but not all) of the loss of performance introduced by threading the monadic type all over the place. I still have an estimated 15% longer compilation time for Coq. Makes use of Set Extraction Conservative Types and Set Extraction File Comment to maintain the relationship between the functions and their types. Uses an intermediate layer Proofview_monad between Proofview_gen and Proofview in order to use a hand-written mli to catch potential errors in the generated file (it uses Extract Constant a lot). A bug in the extraction of signatures forces to remove the generated proofview_gen.mli which does not have the correct types. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16981 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Removed spurious try/with in Proofview.Notation.(>>=) and (>>==).aspiwack
They were a hack to avoid looking where exceptions were raised and not caught. Hopefully I produce a cleaner stack now, catching errors when it is needed. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16980 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02The tactic [admit] exits with the "unsafe" status.aspiwack
It is highlighted in yellow in Coqide. The unsafe status is tracked throughout the execution of tactics such that nested calls to admit are caught. Many function (mainly those building constr with tactics such as typeclass related stuff, and Function, and a few other like eauto's use of Hint Extern) drop the unsafe status. This is unfortunate, but a lot of refactoring would be in order. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16977 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Cleanup of comments.aspiwack
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16976 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Small change to the IO monad interface: [val ref : 'a -> 'a ref t]aspiwack
It doesn't matter much, it's simply a matter of aesthetic, so that all functions in the interface of [Monads.IO] are pure ([IO.run] being the only impure function but it's not part of the generic interface). git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16974 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02A whole new implemenation of the refine tactic.aspiwack
It now uses the same algorithm as pretyping does. This produces pretty weird goal when refining pattern matching terms. Modification of the pattern matching compilation algorithm are pending, hence I will let it be so for now. The file Zsqrt_compat.v has two temporary [Admitted] related to this issue. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16973 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Clean-up: removed redundant notations (>>-) and (>>--) from Proofview.Notations.aspiwack
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16971 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Bases tactics on an IO monad.aspiwack
It allowed to restore the timeout tactics. It also prepares for the debugging mechanism to be restored. ['a IO.t] is just [unit -> 'a]. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16970 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Getting rid of Goal.here, and all the related exceptions and combinators.aspiwack
It was a bad idea. The new API based on lists seems more sensible. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16969 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Uses Proofview.tclEXTEND more sparingly.aspiwack
It is used where failures can be caughts (tclORELSE, tclTRY, …) rather than at each tclTHEN. Hopefully avoiding making things a bit less slow. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16968 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-11-02Makes the new Proofview.tactic the basic type of Ltac.aspiwack
On the compilation of Coq, we can see an increase of ~20% compile time on my completely non-scientific tests. Hopefully this can be fixed. There are a lot of low hanging fruits, but this is an iso-functionality commit. With a few exceptions which were not necessary for the compilation of the theories: - The declarative mode is not yet ported - The timeout tactical is currently deactivated because it needs some subtle I/O. The framework is ready to handle it, but I haven't done it yet. - For much the same reason, the ltac debugger is unplugged. It will be more difficult, but will eventually be back. A few comments: I occasionnally used a coercion from [unit Proofview.tactic] to the old [Prooftype.tactic]. It should work smoothely, but loses any backtracking information: the coerced tactics has at most one success. - It is used in autorewrite (it shouldn't be a problem there). Autorewrite's code is fairly old and tricky - It is used in eauto, mostly for "Hint Extern". It may be an issue as time goes as we might want to have various success in a "Hint Extern". But it would require a heavy port of eauto.ml4 - It is used in typeclass eauto, but with a little help from Matthieu, it should be easy to port the whole thing to the new tactic engine, actually simplifying the code. - It is used in fourier. I believe it to be inocuous. - It is used in firstorder and congruence. I think it's ok. Their code is somewhat intricate and I'm not sure they would be easy to actually port. - It is used heavily in Function. And honestly, I have no idea whether it can do harm or not. Updates: (11 June 2013) Pierre-Marie Pédrot contributed the rebase over his new stream based architecture for Ltac matching (r16533), which avoid painfully and expensively working around the exception-throwing control flow of the previous API. (11 October 2013) Rebasing over recent commits (somewhere in r16721-r16730) rendered a major bug in my implementation of Tacticals.New.tclREPEAT_MAIN apparent. It caused Field_theory.v to loop. The bug made rewrite !lemma, rewrite ?lemma and autorewrite incorrect (tclREPEAT_MAIN was essentially tclREPEAT, causing rewrites to be tried in the side-conditions of conditional rewrites as well). The new implementation makes Coq faster, but it is pretty much impossible to tell if it is significant at all. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16967 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-10-31Future: better doc + restore ~pure optimizationgareuselesinge
This optimization was undone because the kernel type checking was not a pure functions (it was accessing the conv_oracle state imperatively). Now that the conv_oracle state is part of env, the optimization can be restored. This was the cause of the increase in memory consumption, since it was forcing to keep a copy of the system state for every proof, even the ones that are not delayed/delegated to slaves. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16963 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-10-31Conv_orable made functional and part of pre_envgareuselesinge
But for vm, the kernel should be functional now git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16961 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-10-29Do not generate useless argument arrays in whd_* functions.ppedrot
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16954 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-10-28Useless array to list conversions in proof/logic.ml.ppedrot
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16941 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-10-27Abstracting evar filter away. The API is not perfect, but better than nothing.ppedrot
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16939 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-10-24More monomorphic List.mem + List.assoc + ...letouzey
To reduce the amount of syntactic noise, we now provide a few inner modules Int.List, Id.List, String.List, Sorts.List which contain some monomorphic (or semi-monomorphic) functions such as mem, assoc, ... NB: for Int.List.mem and co we reuse List.memq and so on. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16936 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-10-23cList: a few alternative to hashtbl-based uniquize, distinct, subsetletouzey
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16924 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-10-23cList: set-as-list functions are now with an explicit comparisonletouzey
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16920 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-10-18proof modes: use ephemerons to represent them in proof stategareuselesinge
git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16895 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-10-18declaration_hooks use Ephemerongareuselesinge
Ideally, any component of the global state that is a function or any other unmarshallable data should be stocked as an ephemeron to make the state always marshallable. git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16893 85f007b7-540e-0410-9357-904b9bb8a0f7
2013-10-18Future: ported to Ephemeron + exception enhancinggareuselesinge
A future always carries a fix_exn with it: a function that enriches an exception with the state in which the error occurs and also a safe state close to it where one could backtrack. A future can be in two states: Ongoing or Finished. The latter state is obtained by Future.join and after that the future can be safely marshalled. An Ongoing future can be marshalled, but its value is lost. This makes it possible to send the environment to a slave process without pre-processing it to drop all unfinished proofs (they are dropped automatically in some sense). git-svn-id: svn+ssh://scm.gforge.inria.fr/svn/coq/trunk@16892 85f007b7-540e-0410-9357-904b9bb8a0f7
generated by cgit v1.2.3 (git 2.25.1) at 2026-08-09 13:59:10 +0000