July 2016 - ghc-tickets

[GHC] #11717: Way to dump cmm only once
by GHC 18 Jul '16

18 Jul '16

#11717: Way to dump cmm only once -------------------------------------+------------------------------------- Reporter: nomeata | Owner: Type: feature | Status: new request | Priority: normal | Milestone: Component: Compiler | Version: 8.1 (CodeGen) | Keywords: newcomer | 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 current way to dump Cmm output is `-ddump-cmm`, but that dumps Cmm repeatedly, of every stage, similar to `-ddump-verbose-core2core`. This is often confusing. I propose that `-ddump-cmm` dumps only the initial Cmm code, that the flag `-ddump-simpl-cmm` dumps the final Cmm code, and that `-ddump-verbose-cmm` dumps every stage. Shoudn’t be hard to do for a newcomer. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11717> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

[GHC] #12382: Rename clasing type variables more consistently
by GHC 18 Jul '16

18 Jul '16

#12382: Rename clasing type variables more consistently -------------------------------------+------------------------------------- Reporter: nomeata | Owner: Type: feature | Status: new request | Priority: low | Milestone: Component: Compiler | Version: 8.1 (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: -------------------------------------+------------------------------------- This is minor polishing, but polish is nice: Consider {{{ :t (id,id,id) (id,id,id) :: (a -> a, a1 -> a1, a2 -> a2) }}} this looks as if the first `a` is in some way better or more important. What I’d like to see is {{{ :t (id,id,id) (id,id,id) :: (a1 -> a2, a2 -> a2, a3 -> a3) }}} In other words: If two type variables clash and need to be renamed, then rename both (all) of them. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12382> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 9

[GHC] #12378: Not enough inlining happens with single-method type classes
by GHC 18 Jul '16

18 Jul '16

#12378: Not enough inlining happens with single-method type classes -------------------------------------+------------------------------------- Reporter: akio | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.1 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: Runtime Unknown/Multiple | performance bug Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- In the attached file, GHC produces code (with `-O2`) for `foo` that references a top-level definition of type `Small (Either () ())`, despite the fact that all bindings in the module are marked INLINE. If I use `-DOTHER_METHOD` to add another method to the class, this problem goes away. It looks like a ClassOp rule is helping here. I'm not sure if this is a bug, but it was a surprising behavior for me, so I'm reporting it. Please feel free to close this ticket if it's the correct behavior. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12378> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 3

[GHC] #8732: Global big object heap allocator lock causes contention
by GHC 16 Jul '16

16 Jul '16

#8732: Global big object heap allocator lock causes contention ------------------------------------+------------------------------------- Reporter: tibbe | Owner: simonmar Type: bug | Status: new Priority: normal | Milestone: Component: Runtime System | Version: 7.6.3 Keywords: | Operating System: Unknown/Multiple Architecture: Unknown/Multiple | Type of failure: None/Unknown Difficulty: Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | ------------------------------------+------------------------------------- The lock allocate() takes when allocating big objects hurts scalability of I/O bound application. Network.Socket.ByteString.recv is typically called with a buffer size of 4096, which causes a ByteString of that size to be allocate. The size of this ByteString causes it to be allocated from the big object space, which causes contention of the global lock that guards that space. See http://www.yesodweb.com/blog/2014/02/new-warp for a real world example. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8732> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 16

[GHC] #12113: ghc-8.0.1-rc4: unification false positive?
by GHC 14 Jul '16

14 Jul '16

#12113: ghc-8.0.1-rc4: unification false positive? -------------------------------------+------------------------------------- Reporter: _deepfire | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.1-rc4 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: GHC accepts Unknown/Multiple | invalid program Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- {{{#!hs {-# LANGUAGE FunctionalDependencies #-} {-# LANGUAGE ExistentialQuantification #-} {-# LANGUAGE TypeFamilies #-} {-# LANGUAGE UnicodeSyntax #-} module Foo () where type family TF x ∷ * data ADT x type instance TF (ADT x) = x class (a ~ ADT (TF a)) ⇒ TC2 a b | a → b data Forget = ∀ a b. TC2 a b ⇒ Forget a -- ~ Forget (ADT (TF a)) data PhantomF a b = Constr Forget -- ~ Constr (Forget (ADT (TF a))) f ∷ ∀ a b. TC2 a b ⇒ ADT (TF a) → [Forget] f _ = case ((undefined) ∷ (PhantomF a b)) of Constr m → [Forget m] -- Here GHC 8.0.1-rc4 unifies, whereas GHC 7.10.3 (properly?) fails with: -- ghc8-unification-false-positive.hs:20:21: -- Couldn't match type ‘Forget’ with ‘ADT (TF Forget)’ -- In the expression: Forget m -- In the expression: [Forget m] -- In a case alternative: Constr m -> [Forget m] }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12113> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4

[GHC] #12057: TypeFamilyDependencies on GHC.Generic's `Rep`
by GHC 14 Jul '16

14 Jul '16

#12057: TypeFamilyDependencies on GHC.Generic's `Rep` -------------------------------------+------------------------------------- Reporter: Iceland_jack | Owner: Type: feature | Status: new request | Priority: normal | Milestone: Component: | Version: 8.0.1 libraries/base | Keywords: TypeFamilies, | Operating System: Unknown/Multiple Injective | Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- With `TypeFamilyDependencies` (`InjectiveTypeFamilies`) can't `Rep` from `Generic` have a dependency? {{{#!hs class Generic a where type Rep a = (res :: Type -> Type) | res -> a ... }}} Assuming the meta data is always unique. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12057> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 13

Re: [GHC] #7662: Improve GC of mutable objects
by GHC 14 Jul '16

14 Jul '16

#7662: Improve GC of mutable objects -------------------------------------+------------------------------------- Reporter: ezyang | Owner: Type: feature request | Status: new Priority: normal | Milestone: ⊥ Component: Runtime System | Version: 7.7 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 rwbarton): I think this has been discussed elsewhere too, but I don't understand how the above could be consistent with Haskell's semantics. What if I define {{{#!hs upperize :: SomeType -> SomeType upperize s = s { y = toUpper (y s) } }}} Obviously this needs to allocate a new `SomeType` closure on the heap, but if the `IORef Int` is completely unpacked into a mutable field of `SomeType` closures, then the new closure can't share `x` with the old closure. In general just constructing a `SomeType` value would have the observable effect of creating a new mutable cell and so would have to live in IO. In short, ML's records with mutable fields have identity, and so are quite different than just sticking an `IORef` inside a record in Haskell. In Haskell the `IORef` itself has identity (and thus is created and accessed through IO actions) but the record `SomeType` does not. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/7662#comment:10> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #7662: Improve GC of mutable objects
by GHC 14 Jul '16

14 Jul '16

#7662: Improve GC of mutable objects -------------------------------------+------------------------------------- Reporter: ezyang | Owner: Type: feature request | Status: new Priority: normal | Milestone: ⊥ Component: Runtime System | Version: 7.7 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 dobenour): MLton does manage to unpack refs in records though. I think that it can be done in general IF you either: - support interior pointers in the GC (which GHC probably doesn't). - represent mutable refs as "fat pointers", consisting of both a pointer to the object being mutated and a pointer to the cell within the object. In this case {{{#!hs data SomeType = SomeType { x :: {-# UNPACK #-} !(IORef Int) , y :: {-# UNPACK #-} !Char } getRef :: SomeType -> IORef Int getRef = x }}} `x (v :: SomeType)` would produce an object that contains a pointer to `v` (for the GC) and a pointer to field `x` of `v` (used to mutate `x`). -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/7662#comment:9> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #393: functions without implementations
by GHC 14 Jul '16

14 Jul '16

#393: functions without implementations -------------------------------------+------------------------------------- Reporter: c_maeder | Owner: Type: feature request | Status: new Priority: normal | Milestone: ⊥ Component: Compiler (Type | Version: None checker) | Resolution: | Keywords: newcomer 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 Iceland_jack): Should also work for methods. This currently fails: {{{#!hs class FFunctor p where ffmap :: (forall x. f x -> g x) -> (p f -> p g) instance FFunctor (TList l) where ffmap :: (forall x. f x -> g x) -> (TList l f -> TList l g) ffmap = undefined }}} I would like {{{#!hs class FFunctor p where ffmap :: (forall x. f x -> g x) -> (p f -> p g) instance FFunctor (TList l) where ffmap :: (forall x. f x -> g x) -> (TList l f -> TList l g) }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/393#comment:33> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

[GHC] #12392: Die unglaublichen Effekte der „Revolyn“-Diättablette!
by GHC 14 Jul '16

14 Jul '16

#12392: Die unglaublichen Effekte der „Revolyn“-Diättablette! -------------------------------------+------------------------------------- Reporter: Milena | Owner: Type: bug | Status: new Priority: lowest | Milestone: Component: Compiler | Version: 8.0.1 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: -------------------------------------+------------------------------------- Seit Januar 2016 erobert ein effektives Diätmittel aus den vereinigten Staaten von Amerika auch die westliche Himmelsphäre! Dabei geht es um ein Abnehmprodukt auf ausschließlich pflanzlicher Grundlage, dass eine extrem fettverbrennende Wirkung besitzt. Erfahrung Revolyn-Diäthelfern: Wir haben die Überraschungs-Pille ausprobiert und herausgefunden, dass die beachtlichen Bestandteile und die Verknüpfung mit dem Ergänzungsprodkut "Pure Cleanse Diät Pillen“ zu unfasbaren Ergebnissen führt. Zwar waren wiranfangs extrem skeptisch, ob die Versprechen der Produkte eingehalten werden, aber die Langzeitwirkung hat mich beeindruckt und parallel fasziniert. Das Team hat zweifellos aber auch neben der Einnahme der Diäthilfen auf die Ernährung acht genommen und Sport getrieben, dies ist zweifellos zu empfehlen! Kombination der beiden Produkte: „Revolyn Diet Tabletten “ und „Pure Cleannse Tabletten“ ergeben zusammen die erstklassige Plattform für eine aussichtsreiche Diät. Während „Revolyn“ die Fettverbrennung gewaltig beschleunigt, erfüllt die „Pure Cleanse Diet Tablette“ andere wichtige Punkte. Es steigert extrem ihre Energie, säubert und entgiftet ihren Organismus und bringt jedermanns Stoffwechsel in Schwung! Es verhindert also die normalen side-effects einer Diät: das Trägheitsgefühl, die 0-Bock-Phase und die Schmerzen in der Bauchgegend. Inhaltsstoffe von Revolyn-Tabletten: Doch wie ist dies alles möglich? Gucken wir uns den Inhalt mal im Einzelnen an. Die Revolyn Diät-Tablette enthält gleich mehrere natürliche und effektive Bestandteile. Der grüne Tee und die enthaltenen Guarana- Bestandteile verbessern gehörig den Metabolismus. Außerdem steigert Guarana den Blutdruck und aus dem Grund auch die Anzahl an verbrannten Kalorien.Der Inhaltsstoff „Damiana“ sichert eine leistungsfähigere Libido, steigert die Fettverbrennung und stärkt das Nerven- und Immunsystem! Ein echtes Allround-Talent!Das außerdem enthaltene Granatapfel-Extrakt wirkt positiv auf die Blutfette (beschleunigt dessen Vernichtung) und beinhaltet extrem viele gute und für die Gesundheit wichtige Vitamine. Einnahme von den Revolyn-Diätpillen: Das Einnehmen von [http://www.abnehmprodukte-test.com Revolyn Diet] Pillen ist mega einfach erklärt. Am Morgen und Am Abend eine Pille mit einem Glas mit zuckerfreier Sprudel einnehmen – und voilà. Das Einnehmen von „Pure Life Cleanse ist noch problemloser und benutzerfreundlicher: Eine Tablette am frühen Nachmittag mit einem Schluck mit zuckerfreier Sprudel zu sich nehmen, voila! Wie teuer sind die Revolyn-Diätpillen? Gegen anderen Abnehmprodukten mit einer sich ähnelnden Wirkung „ist Revolyn “ und „die [http://www.abnehmprodukte-test.com/revolyn-diat-und- pure-cleanse Pure Cleanse] Diet Pillen“ ist Revolyn sehr kostengünstig. Zwar kostet 1 Flasche 39,95 €, allerdings verringert sich der Einzelpreis je nach Menge an bestellten Produkten. Wer 2 Flaschen bestellt (79,95Euro) erhält eine einfach so obendrauf. Vier Flaschen kosten 159,80 Taler, es gibt aber auch vier Flaschen für lau dazu. Im Endeffekt bekommt ihr also 8 Flaschen für nur 159,80 €! Ein einzigartiger Oberknaller! -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12392> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0