- ghc-tickets - Haskell.org

[GHC] #11222: Teach strictness analysis about `catch`-like operations
by GHC 31 Jan '19

31 Jan '19

#11222: Teach strictness analysis about `catch`-like operations -------------------------------------+------------------------------------- Reporter: bgamari | Owner: Type: feature | Status: new request | Priority: normal | Milestone: 8.2.1 Component: Compiler | Version: 7.10.3 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: -------------------------------------+------------------------------------- Consider the `catch#` primop, {{{#!hs catch# :: (State# RealWorld -> (# State# RealWorld, a #) ) -- ^ thing to catch exceptions from -> (b -> State# RealWorld -> (# State# RealWorld, a #) ) -- ^ exception handler -> State# RealWorld -> (# State# RealWorld, a #) }}} Semantically, this operation will always evaluate its first argument. Ideally we would indicate this in the primop's strictness signature in `primops.txt.pp`. Sadly, we can't do this at the moment due to a subtle wrinkle (discovered in #10712): Consider, {{{#!hs let r = \st -> raiseIO# blah st in catch (\st -> ...(r st)..) handler st }}} If we give the first argument of catch a strict signature, we'll get a demand `C(S)` for `r`; that is, `r` is definitely called with one argument, which indeed it is. The trouble comes when we feed `C(S)` into `r`'s RHS as the demand of the body as this will lead us to conclude that the whole `let` will diverge; clearly this isn't right. As Simon noted in ticket:10712#comment:4, > There's something very special about catch: it turns divergence into non-divergence. In order to apply a proper strictness signature to `catch`-like operations we would need to teach the strictness analyzer about this property. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11222> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 12

[GHC] #14998: Sort out the strictness mess for exceptions
by GHC 31 Jan '19

31 Jan '19

#14998: Sort out the strictness mess for exceptions -------------------------------------+------------------------------------- Reporter: simonpj | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.4.3 Component: Compiler | Version: 8.2.2 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: -------------------------------------+------------------------------------- Over time we have evolved into a messy and bad place for the interaction of strictness analysis and exceptions. Here's the amazing history of the strictness of `catch#` * Dec 13: 0558911f91c: catch# gets a strictness signature with `lazyApply1Dmd` * Jul 15: 7c0fff41789: catch# goes from `lazyApply1Dmd` to `strictApply1Dmd` * Dec 15: 28638dfe79e: catch# goes from `strictApply1Dmd` to `lazyApply1Dmd`. Trac #10712 * Jan 16: 9915b656440: catch# goes from `lazyApply1Dmd` to `catchArgDmd`, and result goes from `botRes` to `exnRes` Trac #11222. * Mar 17: 701256df88c: catch# goes from 'catchArgDmd` to `lazyApply1Dmd`. This ticket #13330. See also * `Note [Strictness for mask/unmask/catch]` in `primops.txt/pp`. * `exceptions_and_strictness` in `primops.txt/pp`. * `Note [Exceptions and strictness]` in `Demand.hs` * #11555, #13330, #10712, #11222 '''Item 1: ThrowsExn''' * The strictness analyser has quite a bit of complexity around `ThrowsExn`. * `ThrowsExn` only differs from `Diverges` if the `ExnStr` flag is set to `ExnStr` * And the only place that happens is in `Demand.catchArgDmd` * The only place `catchArgDmd` is used is in `primops.txt.pp`. Originally, in the patch for #11222, `catchArgDmd` was used for the first arg of `catch#`, `catchSTM#` and `catchRetry#`. * But the patch in this ticket removed two of those three; now only `catchRetry#` uses the `catchArgDmd` thing. It looks to me as if `catchRetry#` was left out by mistake; and indeed David says "I left it out on the grounds that I didn't understand it well enough." And if so, that's the last use of `catchArgDmd` and I think that all the tricky `ThrowsExn` machinery can disappear. '''Item 2: strictness of catchException''' As a result of all this to-and-fro we have in GHC.IO {{{ catch (IO io) handler = IO $ catch# io handler' catchException !io handler = catch io handler }}} That is, `catchException` is strict in the IO action itself. But Neil argues in #11555, commment:6 that this is bad because it breaks the monad laws. I believe that there is some claimed performance gain from the strictness of `catchException`, but I don't believe it exists. The perf gain was from when it had a `strictApply1Dmd`; that is, it promised to call its argument. See this note in `primops.txt.pp` {{{ -- Note [Strictness for mask/unmask/catch] -- Consider this example, which comes from GHC.IO.Handle.Internals: -- wantReadableHandle3 f ma b st -- = case ... of -- DEFAULT -> case ma of MVar a -> ... -- 0# -> maskAsynchExceptions# (\st -> case ma of MVar a -> ...) -- The outer case just decides whether to mask exceptions, but we don't want -- thereby to hide the strictness in 'ma'! Hence the use of strictApply1Dmd. }}} I think the above saga was all in pursuit of exposing the strictness of `ma` if we used `catch#` instead of `maskAsynchExceptions#` in this example. But the saga ultimate appears to have failed, and the strictenss annotation in `catchException` is a vestige that carries no benefit, but does requires comments etc. '''Item 3: performance''' If making `catch#` have strictness `strictApply1Dmd` improves perf, it would be good to know where. That would entail making the (tiny) change, recompiling all, and nofibbing. Here is the commit message in 7c0fff41789, which added `strictApply1Dmd`: {{{ commit 7c0fff41789669450b02dc1db7f5d7babba5dee6 Author: Simon Peyton Jones <simonpj(a)microsoft.com> Date: Tue Jul 21 12:28:42 2015 +0100 Improve strictness analysis for exceptions Two things here: * For exceptions-catching primops like catch#, we know that the main argument function will be called, so we can use strictApply1Dmd, rather than lazy Changes in primops.txt.pp * When a 'case' scrutinises a I/O-performing primop, the Note [IO hack in the demand analyser] was throwing away all strictness from the code that followed. I found that this was causing quite a bit of unnecessary reboxing in the (heavily used) function GHC.IO.Handle.Internals.wantReadableHandle So this patch prevents the hack applying when the case scrutinises a primop. See the revised Note [IO hack in the demand analyser] Thse two things buy us quite a lot in programs that do a lot of IO. Program Size Allocs Runtime Elapsed TotalMem -------------------------------------------------------------------------------- hpg -0.4% -2.9% -0.9% -1.0% +0.0% reverse-complem -0.4% -10.9% +10.7% +10.9% +0.0% simple -0.3% -0.0% +26.2% +26.2% +3.7% sphere -0.3% -6.3% 0.09 0.09 +0.0% -------------------------------------------------------------------------------- Min -0.7% -10.9% -4.6% -4.7% -1.7% Max -0.2% +0.0% +26.2% +26.2% +6.5% Geometric Mean -0.4% -0.3% +2.1% +2.1% +0.1% }}} If these gains are still on the table (i.e. moving to `strictApply1Dmd` gives them), perhaps we can add some strictness annotations to the I/O to replicate the effect, even if we'd decided we can't actualy make `catch#` strict? '''Item 4: IO hack in the demand analyser''' In getting up to speed with this, I noticed this comment in the demand analyser: {{{ {- Note [IO hack in the demand analyser] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ There's a hack here for I/O operations. Consider case foo x s of { (# s', r #) -> y } ...omitted... However, consider f x s = case getMaskingState# s of (# s, r #) -> case x of I# x2 -> ... Here it is terribly sad to make 'f' lazy in 's'. After all, getMaskingState# is not going to diverge or throw an exception! This situation actually arises in GHC.IO.Handle.Internals.wantReadableHandle (on an MVar not an Int), and made a material difference. So if the scrutinee is a primop call, we *don't* apply the state hack... }}} Idea: in the nested-CPR work, we have simple termination information. So we can use that, instead of "is a primop call" to deliver on this hack in a much more principled way. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14998> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 41

[GHC] #12092: Wrong error message with TypeApplications
by GHC 31 Jan '19

31 Jan '19

#12092: Wrong error message with TypeApplications -------------------------------------+------------------------------------- Reporter: kwf | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.1 (Type checker) | Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: Incorrect Unknown/Multiple | warning at compile-time Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- When I write the following: {{{#!hs {-# language TypeApplications #-} module Bug where huh = spam @Int }}} I expect to get the error: {{{ • Variable not in scope: spam • Perhaps you meant ‘span’ (imported from Prelude) }}} But instead, I get the error: {{{ • Cannot apply expression of type ‘t0’ to a visible type argument ‘Int’ • In the expression: spam @Int In an equation for ‘huh’: huh = spam @Int }}} While it is //technically// true that I cannot apply an out-of-scope identifier to a visible type argument, this is probably the less useful error to receive. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12092> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[GHC] #16246: GHC HEAD-only Core Lint error with unboxed equality (Non-CoVar has coercion type)
by GHC 31 Jan '19

31 Jan '19

#16246: GHC HEAD-only Core Lint error with unboxed equality (Non-CoVar has coercion type) -------------------------------------+------------------------------------- Reporter: RyanGlScott | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.10.1 Component: Compiler | Version: 8.6.3 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: Compile-time Unknown/Multiple | crash or panic Test Case: | Blocked By: Blocking: | Related Tickets: #15648 Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- {{{#!hs module Foo where import Language.Haskell.TH.Lib import Language.Haskell.TH.Syntax ueqT :: Q Type ueqT = conT $ mkNameG_tc "ghc-prim" "GHC.Prim" "~#" }}} {{{#!hs {-# LANGUAGE KindSignatures #-} {-# LANGUAGE GADTs #-} {-# LANGUAGE TemplateHaskell #-} module Bug where import Data.Kind import Foo (ueqT) data JankyEquality :: Type -> Type -> Type where Jank :: $ueqT a b -> JankyEquality a b unJank :: JankyEquality a b -> $ueqT a b unJank (Jank x) = x }}} {{{ $ ~/Software/ghc5/inplace/bin/ghc-stage2 --interactive Bug.hs -dcore-lint GHCi, version 8.7.20190115: https://www.haskell.org/ghc/ :? for help Loaded GHCi configuration from /home/rgscott/.ghci [1 of 2] Compiling Foo ( Foo.hs, interpreted ) [2 of 2] Compiling Bug ( Bug.hs, interpreted ) *** Core Lint errors : in result of Desugar (before optimization) *** <no location info>: warning: In a case alternative: (Jank x_a4vR :: a_a4wC ~# b_a4wD) Non-CoVar has coercion type x_a4vR :: a_a4wC ~# b_a4wD *** Offending Program *** <elided> unJank :: forall a b. JankyEquality a b -> a ~# b [LclIdX] unJank = \ (@ a_a4wC) (@ b_a4wD) (ds_d4wN :: JankyEquality a_a4wC b_a4wD) -> case ds_d4wN of wild_00 { Jank x_a4vR -> break<0>() x_a4vR } }}} Note that this passes `-dcore-lint` on GHC 8.0.2 through 8.6.3. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/16246> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

[GHC] #16238: Hadrian provides no way to accept output of test
by GHC 31 Jan '19

31 Jan '19

#16238: Hadrian provides no way to accept output of test -------------------------------------+------------------------------------- Reporter: mpickering | Owner: (none) Type: feature | Status: new request | Priority: high | Milestone: Component: Build System | Version: 8.6.3 (Hadrian) | 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: -------------------------------------+------------------------------------- I can't find a way to make Hadrian generate the output for a test. In the old build system this used to be `make accept`. Marking this as high priority as this makes it hard to use for adding test cases which fail. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/16238> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

[GHC] #16249: no runtime error for -fdefer-type-errors with TypeError constraint
by GHC 31 Jan '19

31 Jan '19

#16249: no runtime error for -fdefer-type-errors with TypeError constraint -------------------------------------+------------------------------------- Reporter: guibou | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.6.3 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: -------------------------------------+------------------------------------- Part of a testsuite, I'm using `-fdefer-type-errors` to check if haskell expression raises type error. However the haskell expression may be perfectly valid except for a `TypeError` constraint on a typeclass. The following code: {{{#!haskell {-# LANGUAGE UndecidableInstances #-} {-# LANGUAGE FlexibleInstances #-} {-# LANGUAGE DataKinds #-} import GHC.TypeLits class Foo t where foo :: t -> t foo = id instance Foo Int instance (TypeError (Text "String does not work")) => Foo String main :: IO () main = do putStrLn (show (foo 10 :: Int)) putStrLn (foo "hello") }}} Correctly generates errors when compiled: {{{ $ ghc ./DeferTypes.hs [1 of 1] Compiling Main ( DeferTypes.hs, DeferTypes.o ) [flags changed] DeferTypes.hs:17:13: error: • String does not work • In the first argument of ‘putStrLn’, namely ‘(foo "hello")’ In a stmt of a 'do' block: putStrLn (foo "hello") In the expression: do putStrLn (show (foo 10 :: Int)) putStrLn (foo "hello") | 17 | putStrLn (foo "hello") | }}} And also with `-fdefer-type-erros`, the error is transformed into a warning: {{{ $ ghc -fdefer-type-errors ./DeferTypes.hs [1 of 1] Compiling Main ( DeferTypes.hs, DeferTypes.o ) [flags changed] DeferTypes.hs:17:13: warning: [-Wdeferred-type-errors] • String does not work • In the first argument of ‘putStrLn’, namely ‘(foo "hello")’ In a stmt of a 'do' block: putStrLn (foo "hello") In the expression: do putStrLn (show (foo 10 :: Int)) putStrLn (foo "hello") | 17 | putStrLn (foo "hello") | ^^^^^^^^^^^ Linking DeferTypes ... }}} However, executing the program gives no runtime error: {{{ $ ./DeferTypes 10 hello }}} I was expecting something such as: {{{ $ ./DeferTypes 10 [a defered type error exception] }}} With ghc 8.6.3 from nix. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/16249> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 1

[GHC] #16154: -Wredundant-constraints: False positive
by GHC 30 Jan '19

30 Jan '19

#16154: -Wredundant-constraints: False positive -------------------------------------+------------------------------------- Reporter: fumieval | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.6.3 (Type checker) | Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: Incorrect Unknown/Multiple | error/warning at compile-time Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- GHC produces a warning for the attached source, although it won't compile if the suggested constraints are removed: {{{ redundant-constraint.hs:24:1: warning: [-Wredundant-constraints] • Redundant constraints: (KnownNat 42, Capture 42 p) • In the type signature for: foo :: Foo 42 | 24 | foo :: Foo 42 }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/16154> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 7

[GHC] #15632: Undependable Dependencies
by GHC 30 Jan '19

30 Jan '19

#15632: Undependable Dependencies -------------------------------------+------------------------------------- Reporter: AntC | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Compiler | Version: 8.6.1-beta1 Keywords: | Operating System: Unknown/Multiple FunctionalDependencies, | OverlappingInstances | Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: 10675 Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- Consider {{{#!hs {-# LANGUAGE MultiParamTypeClasses, FunctionalDependencies, GADTs, FlexibleInstances, FlexibleContexts, UndecidableInstances #-} data Hither = Hither deriving (Eq, Show, Read) -- just three data Thither = Thither deriving (Eq, Show, Read) -- distinct data Yon = Yon deriving (Eq, Show, Read) -- types class Whither a b | a -> b where whither :: a -> b -> b -- instance Whither Int Hither where -- rejected: FunDep conflict -- whither _ y = y -- with Thither instance, so instance {-# OVERLAPPING #-} (b ~ Hither) => Whither Int b where whither _ y = y instance {-# OVERLAPS #-} Whither a Thither where whither _ y = y instance {-# OVERLAPPABLE #-} (b ~ Yon) => Whither a b where whither _ y = y f :: Whither Int b => Int -> b -> b f = whither g :: Whither Bool b => Bool -> b -> b g = whither }}} Should those three instances be accepted together? In particular, the published work on FDs (including the FDs through CHRs paper) says the `Thither` instance should behave as if: {{{#!hs instance (beta ~ Thither) => Whither a beta where ... }}} (in which `beta` is a unification variable and the `~` constraint is type improvement under the FD.) But now the instance head is the same as the `Yon` instance, modulo alpha renaming; with the constraints being contrary. That's demonstrated by the inferred/improved type for `g`: {{{#!hs *Whither> :t g ===> g :: Bool -> Thither -> Thither -- and yet *Whither> g True Yon ===> Yon }}} What do I expect here? * At least the `Thither` and `Yon` instances to be rejected as inconsistent under the FunDep. * (The `Hither` instance is also inconsistent, going by the strict interpretation in published work. But GHC's rule here is bogus, as documented in Trac #10675.) Exploiting Overlapping instances is essential to making this go wrong. The published work shies away from considering `FunDeps + Overlap`; and yet specifying the semantics as if the instances used bare unification variables is almost inevitably going give overlaps -- especially if there are multiple FDs. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15632> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 9

[GHC] #16256: -fexternal-interpreter slow for profiling with template haskell
by GHC 30 Jan '19

30 Jan '19

#16256: -fexternal-interpreter slow for profiling with template haskell -------------------------------------+------------------------------------- Reporter: guibou | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.6.3 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: -------------------------------------+------------------------------------- I'm trying to build the following program (in `Foo.hs`) with profiling enabled: {{{#!haskell {-# LANGUAGE TemplateHaskell #-} main = print $([| 10 |]) }}} With the "alternate object file suffix" method, I have correct build time: {{{ $ time ghc -prof -osuf o_p Foo.hs [1 of 1] Compiling Main ( Foo.hs, Foo.o_p ) Linking Foo ... real 0m1.698s user 0m1.518s sys 0m0.184s }}} However with `-fexternal-interpreter`, times are abysmally slow: {{{ $ time ghc -prof -fexternal-interpreter -osuf "p_o" Foo.hs [1 of 1] Compiling Main ( Foo.hs, Foo.p_o ) Linking Foo ... real 0m26.077s user 0m1.531s sys 0m0.190s }}} I'm using `ghc-8.6.3` on nixos. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/16256> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 1

[GHC] #16245: GHC panic (No skolem info) with QuantifiedConstraints and strange scoping
by GHC 30 Jan '19

30 Jan '19

#16245: GHC panic (No skolem info) with QuantifiedConstraints and strange scoping -------------------------------------+------------------------------------- Reporter: RyanGlScott | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.6.3 (Type checker) | Keywords: TypeInType, | Operating System: Unknown/Multiple QuantifiedConstraints | Architecture: | Type of failure: Compile-time Unknown/Multiple | crash or panic Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- The following program panics with GHC 8.6.3 and HEAD: {{{#!hs {-# LANGUAGE ConstraintKinds #-} {-# LANGUAGE DataKinds #-} {-# LANGUAGE PolyKinds #-} {-# LANGUAGE QuantifiedConstraints #-} module Bug where import Data.Kind type Const a b = a type SameKind (a :: k) (b :: k) = (() :: Constraint) class (forall (b :: k). SameKind a b) => C (k :: Const Type a) }}} {{{ $ /opt/ghc/8.6.3/bin/ghc Bug.hs [1 of 1] Compiling Bug ( Bug.hs, Bug.o ) Bug.hs:11:36: error:ghc: panic! (the 'impossible' happened) (GHC version 8.6.3 for x86_64-unknown-linux): No skolem info: [k1_a1X4[sk:1]] Call stack: CallStack (from HasCallStack): callStackDoc, called at compiler/utils/Outputable.hs:1160:37 in ghc:Outputable pprPanic, called at compiler/typecheck/TcErrors.hs:2891:5 in ghc:TcErrors }}} As with #16244, I imagine that the real culprit is that `SameKind a b` would force `a :: k`, which would be ill-scoped. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/16245> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 1