September 2018 - ghc-tickets

[GHC] #15447: Mark the core libraries' internal modules "not-home" instead of "hide"
by GHC 16 Jan '19

16 Jan '19

#15447: Mark the core libraries' internal modules "not-home" instead of "hide" -------------------------------------+------------------------------------- Reporter: sjakobi | Owner: (none) Type: task | Status: new Priority: normal | Milestone: 8.8.1 Component: Core | Version: 8.4.3 Libraries | Keywords: haddock | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- I'm often a bit annoyed that I cannot browse the haddocks for internal libraries. It would be nice if `base`, `ghc-prim`, `integer-gmp` and `integer-simple` could follow the example of `containers` and other libraries and offer docs for internal libraries too. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15447> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4

[GHC] #13191: Make liftA2 a method of Applicative
by GHC 16 Jan '19

16 Jan '19

#13191: Make liftA2 a method of Applicative -------------------------------------+------------------------------------- Reporter: dfeuer | Owner: dfeuer Type: feature | Status: new request | Priority: normal | Milestone: 8.2.1 Component: Core | Version: 8.0.1 Libraries | Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): Phab:D3031 | Wiki Page: -------------------------------------+------------------------------------- Some functors support implementations of `liftA2` that are considerably faster than the usual definition. Furthermore, some people see `liftA2` as more intuitive and fundamental to the idea of `Applicative` than `<*>` is. This has been discussed on the libraries list in https://mail.haskell.org/pipermail/libraries/2017-January/027579.html and (as an official proposal) in https://mail.haskell.org/pipermail/libraries/2017-January/027612.html. All who expressed an opinion favored the proposal. One (David Menendez) specified that his support was weak. The others (Conal Elliot, Kris Nuttycombe, Wren Romano, Bardur Arantsson, and Mario Blažević) either supported the proposal strongly or merely indicated they were "+1". Conal Elliot wrote: [[BR]] [[BR]] > +1. > > I also sometimes define a specialized `liftA2` and then use it to define > `(<*>)`, which then gets used to define the real `liftA2`. > > I think of `liftA2` as playing a role similar to `foldMap` and `traverse`, while > `(<*>)` corresponds to `fold` and `sequenceA`. The first three self- compose > nicely: `liftA2.liftA2.liftA2`, `foldMap.foldMap.foldMap`, and > `traverse.traverse.traverse`. With functor composition, it's so much nicer to > write `liftA2.liftA2` (in the style of `Functor`, `Foldable`, and `Traversable`) > rather than `liftA2 (<*>)`. [[BR]] Wren Romano wrote [[BR]] [[BR]] > I'm also all for adding liftA2 to the class and have noticed this > inefficiency/asymmetry when working on the class hierarchies for other > languages [[BR]] Kris NuttyCombe wrote [[BR]] [[BR]] > I'm in favor of this change. From my perspecive, liftA2 is actually the > fundamental Applicative operation, an <*> is merely a convenient > isomorphism. When I'm teaching, showing the symmetry between the following > always seems to help students: > > fmap :: (a -> b) -> f a -> f b > liftA2 :: (a -> b -> c) -> f a -> f b -> f c > flip (>>=) :: (a -> f b) -> f a -> f b > > <*> is obviously exceptionally useful in practice. But liftA2 seems like the > more essential shape of that operation. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13191> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

[GHC] #15602: PAP invariant of pointer tagging does not hold
by GHC 16 Jan '19

16 Jan '19

#15602: PAP invariant of pointer tagging does not hold -------------------------------------+------------------------------------- Reporter: osa1 | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Compiler | Version: 8.5 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: #15508, #13767 Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- The PAP invariant of pointer tagging says > the PAP entry code jumps to the function's entry code, so it must have a tagged pointer to the function closure in R1. We therefore assume that a PAP always contains a tagged pointer to the function closure. (from https://ghc.haskell.org/trac/ghc/wiki/Commentary/Rts/HaskellExecution/Point…) As discovered while debugging #15508, this currently does not hold. I tried to fix this in one PAP allocation site in Phab:D5051 but it somehow broke another test. We should review all PAP allocation sites and make sure the invariant holds, and then fix any bugs that this fix reveals. Relevant commits: - 6015a94f9108a502150565577b66c23650796639: Implements pointer tagging - f9c6d53fe997f1c560cda6f346f4b201711df37c: Fixes a PAP allocation site (#13767) -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15602> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 1

[GHC] #12045: Visible kind application
by GHC 16 Jan '19

16 Jan '19

#12045: Visible kind application -------------------------------------+------------------------------------- Reporter: Iceland_jack | Owner: Type: feature | Status: new request | Priority: normal | Milestone: Component: Compiler | Version: 8.1 Keywords: | Operating System: Unknown/Multiple TypeApplications TypeInType | Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- I've wanted this for a while {{{ ghci> :kind (:~:) (:~:) :: k -> k -> Type }}} {{{ ghci> :kind (:~:) @(Type -> Type) (:~:) @(Type -> Type) :: (Type -> Type) -> (Type -> Type) -> Type ghci> :kind (:~:) @(Type -> Type) [] (:~:) @(Type -> Type) [] :: (Type -> Type) -> Type ghci> :kind (:~:) @(Type -> Type) [] Maybe (:~:) @(Type -> Type) [] Maybe :: Type }}} Working like {{{ ghci> type Same k (a::k) (b::k) = a :~: b ghci> :kind Same Same :: forall k -> k -> k -> * }}} {{{ ghci> :kind Same (Type -> Type) Same (Type -> Type) :: (Type -> Type) -> (Type -> Type) -> * ghci> :kind Same (Type -> Type) [] Same (Type -> Type) [] :: (Type -> Type) -> * ghci> :kind Same (Type -> Type) [] Maybe Same (Type -> Type) [] Maybe :: * }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12045> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 22

[GHC] #15592: Type families without CUSKs cannot be given visible kind variable binders
by GHC 16 Jan '19

16 Jan '19

#15592: Type families without CUSKs cannot be given visible kind variable binders -------------------------------------+------------------------------------- Reporter: RyanGlScott | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Compiler | Version: 8.4.3 Keywords: | Operating System: Unknown/Multiple TypeApplications, TypeFamilies, | CUSKs | Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- Consider the following program and GHCi session which uses it: {{{ {-# LANGUAGE PolyKinds #-} {-# LANGUAGE TypeFamilies #-} module Foo where import Data.Kind type family T1 (x :: Type) (y :: a) :: Type where {} type family T2 x (y :: a) :: Type where {} }}} {{{ $ ghc2/inplace/bin/ghc-stage2 --interactive Foo.hs -fprint-explicit- foralls GHCi, version 8.7.20180831: http://www.haskell.org/ghc/ :? for help Loaded GHCi configuration from /home/rgscott/.ghci [1 of 1] Compiling Foo ( Foo.hs, interpreted ) Ok, one module loaded. λ> :kind T1 T1 :: forall a. * -> a -> * λ> :kind T2 T2 :: forall {k} {a}. k -> a -> * }}} Note that `T1` quantifies `a` visibly whereas `T2` does not. I find this somewhat surprising, since both `T1` and `T2` explicitly mention `a` in their respective definitions. The only difference between the two is that `T1` has a CUSK while `T2` does not. This isn't of much importance at the moment, but it will be once visible kind application lands, as this will prevent anyone from instantiating the `a` in `T2` using a kind application. It's unclear to me whether this is intended behavior or not. I suppose there might be an unwritten rule that you can't use visible kind application on anything that doesn't have a CUSK, but if this really is the case, we should be upfront about it. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15592> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 26

[GHC] #15362: Template Haskell ignores bad type family definitions
by GHC 16 Jan '19

16 Jan '19

#15362: Template Haskell ignores bad type family definitions -------------------------------------+------------------------------------- Reporter: goldfire | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Template | Version: 8.4.3 Haskell | 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: -------------------------------------+------------------------------------- This garbage is accepted: {{{#!hs data Nat = Zero | Succ Nat $( [d| type family a + b where Maybe Zero b = b Succ a + b = Succ (a + b) |] ) }}} Note the `Maybe` photo-bombing in the first equation. The problem is that TH uses the symbol declared in the family head for all the equations, ignoring what's actually there. This is sad. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15362> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 3

[GHC] #15360: Template Haskell splicing drops lots of type arguments
by GHC 16 Jan '19

16 Jan '19

#15360: Template Haskell splicing drops lots of type arguments -------------------------------------+------------------------------------- Reporter: goldfire | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Template | Version: 8.4.3 Haskell | 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 conversion functions (in the `Convert` module) drop type arguments in various places. Here are two examples: {{{ data T a b c = Mk a b c bar :: $( return $ AppT (InfixT (ConT ''Int) ''T (ConT ''Bool)) (ConT ''Double) ) bar = Mk 5 True 3.14 }}} This correct code leads to an error {{{ Expecting one more argument to ‘T Int Bool’ }}} because the `Double` is forgotten. Similarly, {{{ baz :: $( return $ AppT (ParensT (ConT ''Maybe)) (ConT ''Int) ) baz = Just 5 }}} fails because `baz :: (Maybe)` isn't valid. On the other hand, {{{ frob :: forall (a :: $( [t| * Int |] )). a -> a frob x = x }}} is spuriously accepted, because GHC has forgotten about the `Int` argument. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15360> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[GHC] #13648: ApplicativeDo selects "GHC.Base.Monad.return" when actions are used without patterns.
by GHC 15 Jan '19

15 Jan '19

#13648: ApplicativeDo selects "GHC.Base.Monad.return" when actions are used without patterns. -------------------------------------+------------------------------------- Reporter: AaronFriel | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.2.1-rc1 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: GHC rejects Unknown/Multiple | valid program Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- GHC 8.0.2 and 8.2.1-rc1 (rc2 not checked) have a bug where -XApplicativeDo causes "GHC.Base.Monad.return" to be used instead of the locally available "return", and a spurious "return ()" shows up. This desugaring is not adhering to the -XRebindableSyntax spec (see: #12490). Example: {{{#!hs {-# LANGUAGE NoImplicitPrelude #-} {-# LANGUAGE RebindableSyntax #-} -- Bug vanishes if this next line is removed: {-# LANGUAGE ApplicativeDo #-} module Main where import Prelude (String, print) class MyFunctor f where fmap :: (a -> b) -> f a -> f b class MyApplicative f where pure :: a -> f a (<*>) :: f (a -> b) -> f a -> f b class MyMonad m where return :: a -> m a (>>) :: m a -> m b -> m b (>>=) :: m a -> (a -> m b) -> m b fail :: String -> m a join :: m (m a) -> m a testCase1 m1 m2 = do m1 m2 return () testCase2 m1 m2 = do _ <- m1 _ <- m2 return () main = print "42" }}} {{{ :t testCase1 testCase1 :: (MyFunctor f, MyApplicative f, MyMonad f, Monad f) => f a2 -> f a1 -> f () :t testCase2 :: testCase2 :: (MyFunctor f, MyApplicative f) => f t -> f a -> f () }}} The desugaring for testCase1 shows the issue: {{{#!hs testCase1' m1 m2 = (<*>) (fmap (\() () -> ()) (m1 >> (GHC.Base.Monad.return ()))) (m2 >> (GHC.Base.Monad.return ())) -- or: testCase1' m1 m2 = (fmap (\() () -> () ) (m1 >> (return ()))) <*> (m2 >> (return ())) }}} I would be able to work on this if someone pointed me in the right direction. It looks like it would be in `compiler/rename/RnEnv` and `compiler/rename/RnExpr`, as with #12490? As a proposed fix, I would want to implement a limited-scope fix before the 8.2.1 release which would not address the thornier issue of #10892. The patch would: 1. Replace `GHC.Base.Monad.return` with local `pure`, removing the `Monad` constraint. 2. Replace `>>` with `*>`, removing the `MyMonad` constraint. This isn't a _complete_ fix, as this would still induce an unnecessary use of the local `fmap`, but it would reduce the desugaring bug in `testCase1` to only that local `fmap`. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13648> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[GHC] #15010: Application (warp) server crashing periodically with "TSO object entered"
by GHC 14 Jan '19

14 Jan '19

#15010: Application (warp) server crashing periodically with "TSO object entered" -------------------------------------+------------------------------------- Reporter: ramirez7 | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.4.3 Component: Runtime | Version: 8.0.2 System | Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: Runtime crash Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- We are seeing one of our warp servers periodically crash with the following error: {{{ Apr 06 10:44:28 stdouterr.log: binary-name-elided: internal error: TSO object entered! Apr 06 10:44:28 stdouterr.log: (GHC version 8.0.2 for x86_64_unknown_linux) }}} We are sometimes seeing CPU/memory spikes concurrently with this error, but not always. We are running very similar servers for other applications and they are '''not''' experiencing this error. But this server is one of our highest- load ones. Other tickets I found with this error: * https://ghc.haskell.org/trac/ghc/ticket/12383 * https://ghc.haskell.org/trac/ghc/ticket/8316 But these seem to be happening in ghci and ghc itself, not in a binary compiled by ghc. Questions I have: * What exactly is this error? * Is there anything else we can do on our side to help debug? Any artifacts that should be produced upon crash? * Should we consider upgrading our GHC version from 8.0.2 to get a bugfix? -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15010> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[GHC] #14645: Allow type family in data family return kind
by GHC 14 Jan '19

14 Jan '19

#14645: Allow type family in data family return kind -------------------------------------+------------------------------------- Reporter: goldfire | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.4.1-alpha1 Keywords: TypeInType, | Operating System: Unknown/Multiple TypeFamilies | Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- GHC currently allows {{{#!hs data family DF1 :: k1 -> k2 }}} where it's expected (and checked) that all data ''instances'' have a return kind of `Type`. (Perhaps `k2` expands to `Type -> Type`, for example.) However, it rejects {{{#!hs type family TF (x :: Type) :: Type data family DF2 :: x -> TF x }}} when that's clearly just as sensible as the first definition. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14645> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 12