December 2015 - ghc-tickets

Re: [GHC] #6024: Allow defining kinds alone, without a datatype
by GHC 12 Dec '15

12 Dec '15

#6024: Allow defining kinds alone, without a datatype -------------------------------------+------------------------------------- Reporter: dreixel | Owner: Type: feature request | Status: infoneeded Priority: normal | Milestone: 8.0.1 Component: Compiler (Type | Version: 7.5 checker) | Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by goldfire): Yes, though it's a shade less useful. Types and kinds are indeed the same now, but terms are not. I see this ticket as a request for an ability to define a datatype whose constructors are defined only in types, never in terms. From a user standpoint, this ticket is largely unaffected by type=kind. It's a shade less useful because you can now say {{{#!hs data Foo = MkFoo1 Type -- Type is a.k.a. * | MkFoo2 Int }}} so some of the motivation for kind-only has gone away. A kind-only definition would still prevent clients from constructing the object at runtime, though. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/6024#comment:12> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #6024: Allow defining kinds alone, without a datatype
by GHC 12 Dec '15

12 Dec '15

#6024: Allow defining kinds alone, without a datatype -------------------------------------+------------------------------------- Reporter: dreixel | Owner: Type: feature request | Status: infoneeded Priority: normal | Milestone: 8.0.1 Component: Compiler (Type | Version: 7.5 checker) | Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Changes (by thomie): * cc: goldfire (added) Comment: Is this feature request still possible to implement, now that "[wiki:DependentHaskell/Phase1#Kindsandtypesarethesame kinds and types are the same]" (6746549772c5cc0ac66c0fce562f297f4d4b80a2). See also comment:3. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/6024#comment:11> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

[GHC] #10902: Refactor type families in Template Haskell
by GHC 12 Dec '15

12 Dec '15

#10902: Refactor type families in Template Haskell -------------------------------------+------------------------------------- Reporter: goldfire | Owner: Type: task | Status: new Priority: normal | Milestone: 8.0.1 Component: Template | Version: 7.11 Haskell | Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: | -------------------------------------+------------------------------------- I see that we now have `OpenTypeFamilyD` and `ClosedTypeFamilyD` in Template Haskell. These constructors have quite a lot in common, and I think it's worth having a `TypeFamilyHead` datatype that contains the bits in common. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10902> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 11

[GHC] #11182: -XStrict doesn't imply -XStrictData
by GHC 12 Dec '15

12 Dec '15

#11182: -XStrict doesn't imply -XStrictData -------------------------------------+------------------------------------- Reporter: RyanGlScott | Owner: Type: bug | Status: new Priority: normal | Milestone: 8.0.1 Component: Compiler | Version: 7.11 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: GHC rejects Unknown/Multiple | valid program Test Case: | Blocked By: Blocking: | Related Tickets: #8347 Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- Compiling this code: {{{#!hs {-# LANGUAGE Strict #-} data Lazy a = Lazy ~a main :: IO () main = case Lazy undefined of Lazy _ -> putStrLn "Lazy" }}} fails with the following error: {{{ $ .\inplace\bin\ghc-stage2.exe ../Lazy.hs [1 of 1] Compiling Main ( ..\Lazy.hs, ..\Lazy.o ) ..\Lazy.hs:3:15: error: * Lazy annotation (~) without StrictData on the first argument of `Lazy' * In the definition of data constructor `Lazy' In the data type declaration for `Lazy' }}} I was under the impression that `-XStrict` implied `-XStrictData`. Is this not the case? -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11182> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

Re: [GHC] #1851: "make install-strip" should work
by GHC 12 Dec '15

12 Dec '15

#1851: "make install-strip" should work -------------------------------------+------------------------------------- Reporter: igloo | Owner: thomie Type: bug | Status: closed Priority: normal | Milestone: 8.0.1 Component: Build System | Version: 7.10.1-rc1 Resolution: fixed | Keywords: Operating System: Unknown/Multiple | Architecture: Type of failure: Building GHC | Unknown/Multiple failed | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): -------------------------------------+------------------------------------- Comment (by Ben Gamari <ben@…>): In [changeset:"aaed24a4e0d8fa0d49aca167fddfb8b606755e05/ghc" aaed24a4/ghc]: {{{ #!CommitTicketReference repository="ghc" revision="aaed24a4e0d8fa0d49aca167fddfb8b606755e05" Build system: fix 'make install-strip' in bindist The INSTALL_PROGRAM variable is set in mk/config.mk, so we have to include that file before using it. Running 'make install' before './configure' in a bindist will now also display a nice message. Reviewers: hvr, austin, bgamari Reviewed By: bgamari Differential Revision: https://phabricator.haskell.org/D1604 GHC Trac Issues: #1851 }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/1851#comment:18> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

[GHC] #11194: Frontend plugins
by GHC 12 Dec '15

12 Dec '15

#11194: Frontend plugins -------------------------------------+------------------------------------- Reporter: ezyang | Owner: ezyang Type: feature | Status: new request | Priority: normal | Milestone: Component: GHC API | Version: 7.11 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- A frontend plugin adds a new major mode to GHC, like `--make`, `--abi- hash`, etc. This will be useful for GHC API clients, because GHC's frontend `ghc/Main.hs` handles a lot of plumbing (flag parsing, setting up the session) which otherwise is painful to reimplement in GHC API (worse yet, it rapidly changes between GHC versions). For example, suppose someone wants to reimplement `--make` using the Shake library. They would define a frontend plugin which implements the `--shake` major mode. Here's how it will work: * A frontend plugin is a module which exports a single identifier `frontendPlugin` of type `GhcPlugins.FrontendPlugin`. This data structure has the following form: {{{ type FrontendPluginAction = [CommandLineOption] -> [(String, Maybe Phase)] -> GHC () data FrontendPlugin = FrontendPlugin { frontendPluginAction :: FrontendPluginAction, -- possibly more fields } defaultFrontendPlugin :: FrontendPluginAction -> FrontendPlugin }}} * There is a new major mode, `--frontend Foo.Plugin`, which loads and runs your plugin. You can pass extra flags to the plugin using `-ffrontend- opt`. We reuse the dynamic loading facilities for normal plugin loading. This should be relatively easy to backport. Possible design flex points: 1. Here, you only load the plugin that you actually want to run. You could also imagine loading a plugin adding a new major mode flag. 2. An alternative to having a full-on plugin architecture is to lib-ify most of `ghc/Main.hs`, so that other clients can use it but with different sets of options, etc. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11194> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 3

[GHC] #10785: ghc-pkg appends colon to db path when piped/redirected
by GHC 12 Dec '15

12 Dec '15

#10785: ghc-pkg appends colon to db path when piped/redirected -------------------------------------+------------------------------------- Reporter: jgertm | Owner: jgertm Type: bug | Status: new Priority: normal | Milestone: Component: ghc-pkg | Version: 7.10.2 Keywords: colon pipe | Operating System: Unknown/Multiple redirect ghc-pkg list | Architecture: | Type of failure: Other Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: | -------------------------------------+------------------------------------- Quite simple. Compare the first line of output from {{{#!sh ghc-pkg list ghc-pkg list | head }}} In the second command's output, a colon gets appended at the end of the first line, compared to the first command. This is generally undesired behaviour, since the intent of redirection is to further process the output of the original command. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10785> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4

Re: [GHC] #5615: ghc produces poor code for `div` with constant powers of 2.
by GHC 12 Dec '15

12 Dec '15

#5615: ghc produces poor code for `div` with constant powers of 2. -------------------------------------+------------------------------------- Reporter: Lennart | Owner: | daniel.is.fischer Type: bug | Status: new Priority: normal | Milestone: 8.0.1 Component: Compiler | Version: 7.4.1-rc1 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: x86 Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Changes (by michalt): * cc: michalt (added) -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/5615#comment:23> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

[GHC] #11142: Type-level skolem capture leads to core lint error
by GHC 11 Dec '15

11 Dec '15

#11142: Type-level skolem capture leads to core lint error -------------------------------------+------------------------------------- Reporter: goldfire | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.10.2 (Type checker) | Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- The following ill-typed code is accepted. {{{ {-# LANGUAGE PolyKinds, RankNTypes #-} module Bug where data SameKind :: k -> k -> * foo :: forall b. (forall (a :: k). SameKind a b) -> () foo = undefined }}} Indeed, Core Lint catches this. But the type-checker should. The problem is that `SameKind` forces `a` and `b` to have the same kind. But they can't, because `a`'s kind is out of scope at `b`'s binding site. The program should be rejected. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11142> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 8

[GHC] #8566: Panic with kindFunResult
by GHC 11 Dec '15

11 Dec '15

#8566: Panic with kindFunResult ------------------------------------+------------------------------------- Reporter: dreixel | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.7 Keywords: | Operating System: Unknown/Multiple Architecture: Unknown/Multiple | Type of failure: None/Unknown Difficulty: Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | ------------------------------------+------------------------------------- The following program: {{{ {-# LANGUAGE DataKinds #-} {-# LANGUAGE PolyKinds #-} {-# LANGUAGE GADTs #-} {-# LANGUAGE FlexibleInstances #-} {-# LANGUAGE ScopedTypeVariables #-} {-# LANGUAGE TypeOperators #-} {-# LANGUAGE UndecidableInstances #-} {-# LANGUAGE MultiParamTypeClasses #-} module Bug where data U (s :: *) = forall k. AA k [U s] data I (u :: U *) (r :: [*]) :: * where A :: I (AA t as) r -- fs unused, but needs to be present for the bug class C (u :: U *) (r :: [*]) (fs :: [*]) where c :: I u r -> I u r instance (C (AA (t (I a ps)) as) ps fs) => C (AA t (a ': as)) ps fs where c A = c undefined }}} crashes a fresh copy of GHC HEAD with the following: {{{ ghc-stage2: panic! (the 'impossible' happened) (GHC version 7.7.20131126 for i386-unknown-linux): kindFunResult k1{tv a24f} [ssk] }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8566> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 21