aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2017-03-20[future] Use eager evaluation for chaining values.Emilio Jesus Gallego Arias
The current future system is lazy when "chaining" (*) a resolved future, which implies that chaining with a resolved future will produce a non-resolved one. This misfeature interacts badly with the "purification" optimization, which in turn provokes a swarm of spurious state setting calls in real use. To solve this problem, we revert to the more natural semantics of respecting the evaluation semantics when mapping over a future, indeed respecting the previous resolution status. This commit solves a kind of _critical_ bug in the current system, with the particular bad path origination in `Future.split2` due to the following accumulation of circumstances: ``` split2 x -> chain x (fun x -> fst x) => let y = chain ~pure x (fun x -> fst x) in if is_over x && greedy then ignore(force ~pure y); y => [y <- Closure (fun x -> fst x)] ignore(force (Closure (fun x -> fst x))) => purify_future (force ~pure) (Closure (fun x -> fst x)) ``` and then, the test in `purify_future` fails, triggering the spurious state reset operation. This problem was first noted at https://sympa.inria.fr/sympa/arc/coqdev/2016-02/msg00081.html , and seems related to https://coq.inria.fr/bugs/show_bug.cgi?id=5382 We fix the problem by making chaining eager, but other solutions would be possible. Given that the main user of `chain` is `split2` which does `snd/fst`, I recommend this solution. The difference in calls to `unfreeze_state` is dramatic: ``` | File | Freeze Calls After | Freeze Calls Before | |----------------------------------------+--------------------+---------------------| | theories/Init/Notations.v | 0 | 0 | | theories/Init/Logic.v | 57 | 614 | | theories/Init/Datatypes.v | 13 | 132 | | theories/Init/Logic_Type.v | 7 | 57 | | theories/Init/Specif.v | 5 | 35 | | theories/Init/Nat.v | 0 | 0 | | theories/Init/Peano.v | 22 | 264 | | theories/Init/Wf.v | 8 | 89 | | theories/Init/Tactics.v | 2 | 24 | | theories/Init/Tauto.v | 0 | 0 | | theories/Init/Prelude.v | 0 | 0 | | Bool/Bool.v | 104 | 1220 | | Program/Basics.v | 0 | 0 | | Classes/Init.v | 0 | 0 | | Program/Tactics.v | 0 | 0 | | Relations/Relation_Definitions.v | 0 | 0 | | Classes/RelationClasses.v | 21 | 341 | | Classes/Morphisms.v | 47 | 689 | | Classes/CRelationClasses.v | 18 | 245 | | Classes/CMorphisms.v | 50 | 587 | | Classes/Morphisms_Prop.v | 3 | 127 | | Classes/Equivalence.v | 6 | 105 | | Classes/SetoidTactics.v | 0 | 0 | | Setoids/Setoid.v | 4 | 33 | | Structures/Equalities.v | 8 | 93 | | Relations/Relation_Operators.v | 0 | 0 | | Relations/Operators_Properties.v | 35 | 627 | | Relations/Relations.v | 2 | 24 | | Structures/Orders.v | 12 | 148 | | Numbers/NumPrelude.v | 0 | 0 | | Structures/OrdersTac.v | 13 | 234 | | Structures/OrdersFacts.v | 73 | 931 | | Structures/GenericMinMax.v | 82 | 1294 | | Numbers/NatInt/NZAxioms.v | 0 | 0 | | Numbers/NatInt/NZBase.v | 7 | 87 | | Numbers/NatInt/NZAdd.v | 14 | 168 | | Numbers/NatInt/NZMul.v | 12 | 144 | | Logic/Decidable.v | 28 | 336 | | Numbers/NatInt/NZOrder.v | 81 | 1174 | | Numbers/NatInt/NZAddOrder.v | 24 | 288 | | Numbers/NatInt/NZMulOrder.v | 46 | 552 | | Numbers/NatInt/NZParity.v | 35 | 420 | | Numbers/NatInt/NZPow.v | 29 | 348 | | Numbers/NatInt/NZSqrt.v | 54 | 673 | | Numbers/NatInt/NZLog.v | 64 | 797 | | Numbers/NatInt/NZDiv.v | 49 | 588 | | Numbers/NatInt/NZGcd.v | 36 | 432 | | Numbers/NatInt/NZBits.v | 0 | 0 | | Numbers/Natural/Abstract/NAxioms.v | 0 | 0 | | Numbers/NatInt/NZProperties.v | 0 | 0 | | Numbers/Natural/Abstract/NBase.v | 14 | 177 | | Numbers/Natural/Abstract/NAdd.v | 6 | 72 | | Numbers/Natural/Abstract/NOrder.v | 29 | 349 | | Numbers/Natural/Abstract/NAddOrder.v | 5 | 60 | | Numbers/Natural/Abstract/NMulOrder.v | 8 | 96 | | Numbers/Natural/Abstract/NSub.v | 36 | 432 | | Numbers/Natural/Abstract/NMaxMin.v | 18 | 216 | | Numbers/Natural/Abstract/NParity.v | 4 | 48 | | Numbers/Natural/Abstract/NPow.v | 26 | 312 | | Numbers/Natural/Abstract/NSqrt.v | 9 | 108 | | Numbers/Natural/Abstract/NLog.v | 0 | 0 | | Numbers/Natural/Abstract/NDiv.v | 50 | 600 | | Numbers/Natural/Abstract/NGcd.v | 14 | 168 | | Numbers/Natural/Abstract/NLcm.v | 29 | 348 | | Numbers/Natural/Abstract/NBits.v | 168 | 2016 | | Numbers/Natural/Abstract/NProperties.v | 0 | 0 | | Arith/PeanoNat.v | 77 | 990 | | Arith/Le.v | 2 | 57 | | Arith/Lt.v | 14 | 168 | | Arith/Plus.v | 20 | 269 | | Arith/Gt.v | 17 | 248 | | Arith/Minus.v | 11 | 132 | | Arith/Mult.v | 14 | 168 | | Arith/Between.v | 19 | 299 | | Logic/EqdepFacts.v | 26 | 539 | | Logic/Eqdep_dec.v | 13 | 361 | | Arith/Peano_dec.v | 3 | 26 | | Arith/Compare_dec.v | 35 | 360 | | Arith/Factorial.v | 3 | 36 | | Arith/EqNat.v | 10 | 111 | | Arith/Wf_nat.v | 18 | 173 | | Arith/Arith_base.v | 0 | 0 | | Numbers/BinNums.v | 0 | 0 | | PArith/BinPosDef.v | 0 | 0 | | PArith/BinPos.v | 229 | 2810 | | NArith/BinNatDef.v | 0 | 0 | | NArith/BinNat.v | 107 | 1330 | | PArith/Pnat.v | 51 | 688 | | NArith/Nnat.v | 30 | 360 | | setoid_ring/Ring_theory.v | 43 | 756 | | Lists/List.v | 195 | 2908 | | setoid_ring/BinList.v | 6 | 90 | | Numbers/Integer/Abstract/ZAxioms.v | 0 | 0 | | Numbers/Integer/Abstract/ZBase.v | 3 | 36 | | Numbers/Integer/Abstract/ZAdd.v | 46 | 552 | | Numbers/Integer/Abstract/ZMul.v | 8 | 96 | | Numbers/Integer/Abstract/ZLt.v | 21 | 252 | | Numbers/Integer/Abstract/ZAddOrder.v | 45 | 543 | | Numbers/Integer/Abstract/ZMulOrder.v | 24 | 288 | | Numbers/Integer/Abstract/ZMaxMin.v | 22 | 264 | | Numbers/Integer/Abstract/ZSgnAbs.v | 41 | 492 | | Numbers/Integer/Abstract/ZParity.v | 6 | 72 | | Numbers/Integer/Abstract/ZPow.v | 10 | 120 | | Numbers/Integer/Abstract/ZDivTrunc.v | 68 | 816 | | Numbers/Integer/Abstract/ZDivFloor.v | 70 | 840 | | Numbers/Integer/Abstract/ZGcd.v | 29 | 348 | | Numbers/Integer/Abstract/ZLcm.v | 50 | 600 | | Numbers/Integer/Abstract/ZBits.v | 205 | 2460 | | Numbers/Integer/Abstract/ZProperties.v | 0 | 0 | | ZArith/BinIntDef.v | 0 | 0 | | ZArith/BinInt.v | 212 | 2839 | |----------------------------------------+--------------------+---------------------| ``` (*) I would call it `Future.map` better than chain.
2017-03-20Merge PR#479: [future] Remove unused parameter greedy.Maxime Dénès
2017-03-19Merge PR#487: Add a forgotten (?) line to "theories/Logic/vo.itarget".Maxime Dénès
2017-03-19Add a forgotten (?) line to "theories/Logic/vo.itarget".Matej Kosik
The "theories/Logic/PropExtensionalityFacts.v" file was: - compiled - used in several places but not actually installed. This commit fixes that.
2017-03-17Document Show Match, add ref to that in match variants/extensionsPaul Steckler
2017-03-17Merge PR#429: Don't require printing-only notation to be productiveMaxime Dénès
2017-03-17Merge PR#428: Report missing tactic arguments in error messageMaxime Dénès
2017-03-17Merge PR#437: Improve unification debug trace.Maxime Dénès
2017-03-17Merge PR#445: TACTIC EXTEND now takes an optional level as argument.Maxime Dénès
2017-03-17Merge PR#442: Allow interactive editing of Coq.Init.LogicMaxime Dénès
2017-03-17Merge PR#451: Add η principles for sigma typesMaxime Dénès
2017-03-15Attempt to improve error message when "apply in" fail.Hugo Herbelin
- Adding a better location in the "apply" on the fly pattern. - Printing statement of lemma and of hypothesis. Was suggested by discussion at wish report #5390.
2017-03-15Merge PR#267: Proposal for an update of the recommended style in programming ↵Maxime Dénès
Coq.
2017-03-14doc: fix a French-ismValentin Robert
2017-03-14[safe-string] Enable -safe-string !Emilio Jesus Gallego Arias
We now build Coq with `-safe-string`, which enforces functional use of the `string` datatype. Coq was pretty safe in these regard so only a few tweaks were needed. - coq_makefile: build plugins with -safe-string too. - `names.ml`: we remove `String.copy` uses, as they are not needed.
2017-03-14[safe-string] toolsEmilio Jesus Gallego Arias
No functional changes.
2017-03-14[safe-string] ideEmilio Jesus Gallego Arias
No functional change, one extra copy introduced but it seems hard to avoid.
2017-03-14[safe-string] plugins/extractionEmilio Jesus Gallego Arias
No functional change.
2017-03-14[safe-string] ltac/profile_ltacEmilio Jesus Gallego Arias
No functional change, one extra copy introduced but it seems hard to avoid.
2017-03-14[safe_string] toplevel/vernacEmilio Jesus Gallego Arias
No functional changes, only a minor copy on a deprecated output option.
2017-03-14[safe_string] toplevel/coqloopEmilio Jesus Gallego Arias
No functional change, even if we could optimize `blanch_utf8_string` a bit more by using `String.init`.
2017-03-14[safe-string] parsing/cLexerEmilio Jesus Gallego Arias
No functional change.
2017-03-14[safe_string] interp/dumpglobEmilio Jesus Gallego Arias
No functional change.
2017-03-14[safe_string] library/nameopsEmilio Jesus Gallego Arias
We add a more convenient API to create identifiers from mutable strings. We cannot solve the `String.copy` deprecation problem until we enable `-safe-string`.
2017-03-14[safe_string] kernel/cemitcodesEmilio Jesus Gallego Arias
The `emitcodes` string type was used in both a functional and an imperative way, so we have to handle it with care in order to preserve the previous optimizations and semantics.
2017-03-14[toplevel] Remove unusable option -notopEmilio Jesus Gallego Arias
Maxime points out that -notop cannot be used as the kernel requires all constants to belong into a module. Indeed: ``` $ rlwrap ./bin/coqtop -notop Coq < Definition foo := True. Toplevel input, characters 0-23: > Definition foo := True. > ^^^^^^^^^^^^^^^^^^^^^^^ Error: No session module started (use -top dir) Coq < Module M. Definition foo := True. End M. Module M is defined Coq < Locate foo. Constant If you see this, it's a bug.M.foo (shorter name to refer to it in current context is M.foo) ``` My rationale for the removal is that this kind of incomplete features are often confusing to newcomers ─ it has happened to me many times ─ as it can be seen for example in #397 .
2017-03-14Merge PR#444: Simplifying a statement in Hurkens.v + a case study for eautoMaxime Dénès
2017-03-14Report missing tactic arguments in error messageTej Chajed
Augments "A fully applied tactic is expected" with the list of missing arguments to the tactic. Addresses [bug 5344](https://coq.inria.fr/bugs/show_bug.cgi?id=5344).
2017-03-14[safe-string] kernel/nativevaluesEmilio Jesus Gallego Arias
No functional change.
2017-03-14[safe_string] kernel/term_typingEmilio Jesus Gallego Arias
No functional change, we create the new string beforehand and `id_of_string` will become a noop with `-safe-string`.
2017-03-14[safe-string] lib/miscelaneaEmilio Jesus Gallego Arias
No functional change.js
2017-03-14[safe-string] lib/cUnixEmilio Jesus Gallego Arias
No functional change.
2017-03-14[safe_string] lib/cThreadEmilio Jesus Gallego Arias
No functional changes.
2017-03-14Merge PR#438: Fix V7 syntax in refman.Maxime Dénès
2017-03-14Merge PR#412: Remove outdated comment from 2002.Maxime Dénès
2017-03-14Merge PR#432: [cleanup] Change Id.t option to Name.t in TacFunMaxime Dénès
2017-03-14Merge PR#477: [travis] Basic support for overlays.Maxime Dénès
2017-03-14Merge PR#473: [ci] Document that sudo: false is slowerMaxime Dénès
2017-03-14Merge PR#464: [META] More fixesMaxime Dénès
2017-03-14Merge PR#465: Fix #5132: coq_makefile generates incorrect install goalMaxime Dénès
2017-03-14Fix #5132: coq_makefile generates incorrect install goalVadim Zaliva
2017-03-14Merge PR#446: Remove a dead exception catching code.Maxime Dénès
2017-03-14Fix 3 unused-intro-pattern warnings in stdlib.Théo Zimmermann
2017-03-14Show unused-intro-pattern warning.Théo Zimmermann
This warning was shown in CoqIDE but not by coqc.
2017-03-14[future] Remove unused parameter greedy.Emilio Jesus Gallego Arias
It was always set to `greedy:true`.
2017-03-14[library] Refactor state handling.Emilio Jesus Gallego Arias
This part of state is critical. We refactor it and make it into a record to ease handling.
2017-03-14[library] Don't recompute path_prefix on unfreeze.Emilio Jesus Gallego Arias
We instead save the current value.
2017-03-13[travis] Basic support for overlays.Emilio Jesus Gallego Arias
We now allow the user to overlay contribution repositories and branches by adding their own rules to `ci-basic-overlay.sh`. This just provides very basic support.
2017-03-13Remove a dead exception catching code.Théo Zimmermann
The code was assuming that Proofview.tclFOCUS could raise a CList.IndexOutOfRange exception but this isn't the case. The focusing functions already catch this exception and raises an algebraic exception within the tactic mechanism.
2017-03-13Merge PR#456: Proposing improvement to the CI targets for local useMaxime Dénès
generated by cgit v1.2.3 (git 2.25.1) at 2026-06-29 02:27:36 +0000