July 2014 - ghc-tickets

[GHC] #8848: Warning: Rule too complicated to desugar
by GHC 01 Dec '16

01 Dec '16

#8848: Warning: Rule too complicated to desugar ------------------------------------+------------------------------------- Reporter: carter | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.8.1-rc2 Keywords: | Operating System: Unknown/Multiple Architecture: Unknown/Multiple | Type of failure: None/Unknown Difficulty: Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | ------------------------------------+------------------------------------- I've a very very modest application of Specialize to fixed sized lists in some of my code which seems to trip up the specialization machinery. Is there any flags I can pass GHC to make sure it doesn't give up on these specialize calls? is the only work around to write my own monomorphic versions and add some hand written rewrite rules?! {{{ rc/Numerical/Types/Shape.hs:225:1: Warning: RULE left-hand side too complicated to desugar let { $dFunctor_a3XB :: Functor (Shape ('S 'Z)) [LclId, Str=DmdType] $dFunctor_a3XB = Numerical.Types.Shape.$fFunctorShape @ 'Z $dFunctor_a3Rn } in map2 @ a @ b @ c @ ('S ('S 'Z)) (Numerical.Types.Shape.$fApplicativeShape @ ('S 'Z) (Numerical.Types.Shape.$fFunctorShape @ ('S 'Z) $dFunctor_a3XB) (Numerical.Types.Shape.$fApplicativeShape @ 'Z $dFunctor_a3XB Numerical.Types.Shape.$fApplicativeShape0)) src/Numerical/Types/Shape.hs:226:1: Warning: RULE left-hand side too complicated to desugar let { $dFunctor_a3XG :: Functor (Shape ('S 'Z)) [LclId, Str=DmdType] $dFunctor_a3XG = Numerical.Types.Shape.$fFunctorShape @ 'Z $dFunctor_a3Rn } in let { $dFunctor_a3XF :: Functor (Shape ('S ('S 'Z))) [LclId, Str=DmdType] $dFunctor_a3XF = Numerical.Types.Shape.$fFunctorShape @ ('S 'Z) $dFunctor_a3XG } in map2 @ a @ b @ c @ ('S ('S ('S 'Z))) (Numerical.Types.Shape.$fApplicativeShape @ ('S ('S 'Z)) (Numerical.Types.Shape.$fFunctorShape @ ('S ('S 'Z)) $dFunctor_a3XF) (Numerical.Types.Shape.$fApplicativeShape @ ('S 'Z) $dFunctor_a3XF (Numerical.Types.Shape.$fApplicativeShape @ 'Z $dFunctor_a3XG Numerical.Types.Shape.$fApplicativeShape0))) }}} the associated code (smashed into a single module ) is {{{ {-# LANGUAGE DataKinds, GADTs, TypeFamilies, ScopedTypeVariables #-} {-# LANGUAGE DeriveDataTypeable #-} {-# LANGUAGE TypeOperators #-} {-# LANGUAGE BangPatterns #-} {-# LANGUAGE FlexibleInstances #-} {-# LANGUAGE FlexibleContexts #-} {-# LANGUAGE FunctionalDependencies #-} {-# LANGUAGE UndecidableInstances #-} {-# LANGUAGE ScopedTypeVariables #-} {-# LANGUAGE StandaloneDeriving #-} {-# LANGUAGE CPP #-} {-# LANGUAGE DeriveFunctor #-} {-# LANGUAGE TemplateHaskell #-} {-# LANGUAGE NoImplicitPrelude #-} module Numerical.Types.Shape where import GHC.Magic import Data.Data import Data.Typeable() import Data.Type.Equality import qualified Data.Monoid as M import qualified Data.Functor as Fun import qualified Data.Foldable as F import qualified Control.Applicative as A import Prelude hiding (foldl,foldr,init,scanl,scanr,scanl1,scanr1) data Nat = S !Nat | Z deriving (Eq,Show,Read,Typeable,Data) #if defined(__GLASGOW_HASKELL_) && (__GLASGOW_HASKELL__ >= 707) deriving instance Typeable 'Z deriving instance Typeable 'S #endif type family n1 + n2 where Z + n2 = n2 (S n1') + n2 = S (n1' + n2) -- singleton for Nat data SNat :: Nat -> * where SZero :: SNat Z SSucc :: SNat n -> SNat (S n) --gcoerce :: (a :~: b) -> ((a ~ b) => r) -> r --gcoerce Refl x = x --gcoerce = gcastWith -- inductive proof of right-identity of + plus_id_r :: SNat n -> ((n + Z) :~: n) plus_id_r SZero = Refl plus_id_r (SSucc n) = gcastWith (plus_id_r n) Refl -- inductive proof of simplification on the rhs of + plus_succ_r :: SNat n1 -> Proxy n2 -> ((n1 + (S n2)) :~: (S (n1 + n2))) plus_succ_r SZero _ = Refl plus_succ_r (SSucc n1) proxy_n2 = gcastWith (plus_succ_r n1 proxy_n2) Refl type N0 = Z type N1= S N0 type N2 = S N1 type N3 = S N2 type N4 = S N3 type N5 = S N4 type N6 = S N5 type N7 = S N6 type N8 = S N7 type N9 = S N8 type N10 = S N9 {- Need to sort out packed+unboxed vs generic approaches see ShapeAlternatives/ for -} infixr 3 :* {- the concern basically boils down to "will it specialize / inline well" -} newtype At a = At a deriving (Eq, Ord, Read, Show, Typeable, Functor) data Shape (rank :: Nat) a where Nil :: Shape Z a (:*) :: !(a) -> !(Shape r a ) -> Shape (S r) a --deriving (Show) #if defined(__GLASGOW_HASKELL_) && (__GLASGOW_HASKELL__ >= 707) deriving instance Typeable Shape #endif instance Eq (Shape Z a) where (==) _ _ = True instance (Eq a,Eq (Shape s a))=> Eq (Shape (S s) a ) where (==) (a:* as) (b:* bs) = (a == b) && (as == bs ) instance Show (Shape Z a) where show _ = "Nil" instance (Show a, Show (Shape s a))=> Show (Shape (S s) a) where show (a:* as) = show a ++ " :* " ++ show as -- at some point also try data model that -- has layout be dynamicly reified, but for now -- keep it phantom typed for sanity / forcing static dispatch. -- NB: may need to make it more general at some future point --data Strided r a lay = Strided { getStrides :: Shape r a } {-# INLINE reverseShape #-} reverseShape :: Shape n a -> Shape n a reverseShape Nil = Nil reverseShape list = go SZero Nil list where go :: SNat n1 -> Shape n1 a-> Shape n2 a -> Shape (n1 + n2) a go snat acc Nil = gcastWith (plus_id_r snat) acc go snat acc (h :* (t :: Shape n3 a)) = gcastWith (plus_succ_r snat (Proxy :: Proxy n3)) (go (SSucc snat) (h :* acc) t) instance Fun.Functor (Shape Z) where fmap = \ _ Nil -> Nil --{-# INLINE fmap #-} instance (Fun.Functor (Shape r)) => Fun.Functor (Shape (S r)) where fmap = \ f (a :* rest) -> f a :* Fun.fmap f rest --{-# INLINE fmap #-} instance A.Applicative (Shape Z) where pure = \ _ -> Nil --{-# INLINE pure #-} (<*>) = \ _ _ -> Nil --{-# INLINE (<*>) #-} instance A.Applicative (Shape r)=> A.Applicative (Shape (S r)) where pure = \ a -> a :* (A.pure a) --{-# INLINE pure #-} (<*>) = \ (f:* fs) (a :* as) -> f a :* (inline (A.<*>)) fs as --{-# INLINE (<*>) #-} instance F.Foldable (Shape Z) where foldMap = \ _ _ -> M.mempty --{-# fold #-} foldl = \ _ init _ -> init foldr = \ _ init _ -> init foldr' = \_ !init _ -> init foldl' = \_ !init _ -> init instance (F.Foldable (Shape r)) => F.Foldable (Shape (S r)) where foldMap = \f (a:* as) -> f a M.<> F.foldMap f as foldl' = \f !init (a :* as) -> let next = f init a in next `seq` F.foldl f next as foldr' = \f !init (a :* as ) -> f a $! F.foldr f init as foldl = \f init (a :* as) -> let next = f init a in F.foldl f next as foldr = \f init (a :* as ) -> f a $ F.foldr f init as -- map2 :: (A.Applicative (Shape r))=> (a->b ->c) -> (Shape r a) -> (Shape r b) -> (Shape r c ) map2 = \f l r -> A.pure f A.<*> l A.<*> r {-# SPECIALIZE map2 :: (a->b->c)-> (Shape Z a )-> Shape Z b -> Shape Z c #-} {-# SPECIALIZE map2 :: (a->b->c)-> (Shape (S Z) a )-> Shape (S Z) b -> Shape (S Z) c #-} {-# SPECIALIZE map2 :: (a->b->c)-> (Shape (S (S Z)) a )-> Shape (S (S Z)) b -> Shape (S (S Z)) c #-} {-# SPECIALIZE map2 :: (a->b->c)-> (Shape (S (S(S Z))) a )-> Shape (S (S (S Z))) b -> Shape (S (S(S Z))) c #-} -- {-# INLINABLE map2 #-} }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8848> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 17

[GHC] #9184: Allow the use of Python 3 when building GHC
by GHC 01 Dec '16

01 Dec '16

#9184: Allow the use of Python 3 when building GHC -------------------------------------------+------------------------------- Reporter: aspidites | Owner: Type: feature request | Status: new Priority: normal | Milestone: Component: Build System | Version: 7.8.2 Keywords: python | Operating System: Architecture: Unknown/Multiple | Unknown/Multiple Difficulty: Easy (less than 1 hour) | Type of failure: Blocked By: | None/Unknown Related Tickets: | Test Case: | Blocking: -------------------------------------------+------------------------------- According to the [https://ghc.haskell.org/trac/ghc/wiki/Building/Preparation/Windows wiki], it is preferred that a user have Python 2.7 installed before attempting to build GHC. As I'm on Arch Linux the default Python version is 3.4. As such, rather than simply changing the shebang line to explicitly ask for version 2, I made it so that the scripts all use syntax that works properly for versions 2.7 to 3.4. I don't think I've used anything that disallows 2.6, but don't have that version available to check against. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9184> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 24

[GHC] #8974: 64 bit windows executable built with ghc-7.9.20140405+LLVM segfaults
by GHC 30 Nov '16

30 Nov '16

#8974: 64 bit windows executable built with ghc-7.9.20140405+LLVM segfaults -----------------------------------+---------------------------------- Reporter: awson | Owner: Type: bug | Status: new Priority: high | Milestone: 7.8.2 Component: Compiler (LLVM) | Version: 7.9 Keywords: | Operating System: Windows Architecture: x86_64 (amd64) | Type of failure: Runtime crash Difficulty: Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | -----------------------------------+---------------------------------- {{{ -- test.hs import System.Mem (performMajorGC) main = performMajorGC >> putStrLn "Done" }}} built with `ghc -pgmlo opt -pgmlc llc -fllvm --make test.hs` segfaults both for LLVM 3.4 and 3.5svn (taken from http://sourceforge.net/projects/msys2/files/REPOS/MINGW/x86_64) 32-bit ghc-7.9.20140404+llvm produces good executable. Adding `ArchX86_64` `OSMinGW32` case to `moduleLayout` in `compiler/llvmGen/LlvmCodeGen/Ppr.hs` improves things slightly (some code segfaulting without it starts to work) but still does not cure the code above. Also I've found the Cmm produced for LLVM CG differs from that produced for NCG. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8974> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 52

[GHC] #9007: relocation R_X86_64_32 against `stg_CHARLIKE_closure' can not be used when making a shared object
by GHC 29 Nov '16

29 Nov '16

#9007: relocation R_X86_64_32 against `stg_CHARLIKE_closure' can not be used when making a shared object ----------------------------------+---------------------------------------- Reporter: nomeata | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.9 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: Building GHC failed Unknown/Multiple | Test Case: Difficulty: Unknown | Blocking: Blocked By: | Related Tickets: | ----------------------------------+---------------------------------------- Building HEAD on Debian wheezy with the Debian packaging setup fails: {{{ configure: Building in-tree ghc-pwd /usr/bin/ld: utils/ghc-pwd/dist-boot/Main.o: relocation R_X86_64_32 against `stg_CHARLIKE_closure' can not be used when making a shared object; recompile with -fPIC utils/ghc-pwd/dist-boot/Main.o: could not read symbols: Bad value collect2: error: ld returned 1 exit status configure: error: Building ghc-pwd failed }}} http://deb.haskell.org/dailies/2014-04-16/ghc_7.9.20140416-0.daily_amd64.bu… This is possibly related to the Hardening flags added by Debian: {{{ ~ $ dpkg-buildflags --get LDFLAGS -Wl,-z,relro ~ $ dpkg-buildflags --get CFLAGS -g -O2 -fstack-protector --param=ssp-buffer-size=4 -Wformat -Werror =format-security }}} Other related bug might be #3668 and https://bugs.debian.org/712228 I’ll disable the hardening flags and see if it helps. The error first occurred 2014-03-27. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9007> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 24

[GHC] #8300: split-objs doesn't split on LLVM
by GHC 22 Nov '16

22 Nov '16

#8300: split-objs doesn't split on LLVM ------------------------------------+------------------------------------- Reporter: rwbarton | Owner: Type: feature request | Status: new Priority: normal | Milestone: Component: Compiler (LLVM) | 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: | ------------------------------------+------------------------------------- `-split-objs` is allowed with `-fllvm`, but it doesn't really do anything. E.g. in my perf-llvm build I have a `libHSbase-4.7.0.0.a` with `Applicative__1.o`, `Arrow__1.o` etc. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8300> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4

[GHC] #8321: improve basic block layout on LLVM backend by predicting stack/heap checks
by GHC 17 Nov '16

17 Nov '16

#8321: improve basic block layout on LLVM backend by predicting stack/heap checks ------------------------------------+------------------------------------- Reporter: rwbarton | Owner: Type: feature request | Status: new Priority: normal | Milestone: 7.10.1 Component: Compiler (LLVM) | 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: | ------------------------------------+------------------------------------- Currently we don't give the LLVM optimizer any information about which branch of an `if` is likely to be taken. As a result, the optimizer is likely to produce a basic block layout which is not optimal. Improving the layout can improve performance through better instruction cache usage and better branch prediction by the hardware. We can control LLVM's idea of what is likely with the `llvm.expect` intrinsic function. Some obvious branches which we can predict accurately are the stack and heap checks that appear near the entry of many functions. Here's a small example of some Cmm code and the output of the LLVM optimizer/compiler. {{{ block_c2Lc_entry() // [R1, R2] { info_tbl: [(c2Lc, label: block_c2Lc_info rep:StackRep [])] stack_info: arg_space: 0 updfr_space: Nothing } {offset c2Lc: Hp = Hp + 24; if (Hp > HpLim) goto c2Lm; else goto c2Ll; c2Lm: HpAlloc = 24; R2 = R2; R1 = R1; call stg_gc_pp(R2, R1) args: 8, res: 8, upd: 24; c2Ll: I64[Hp - 16] = :_con_info; P64[Hp - 8] = R1; P64[Hp] = R2; R1 = Hp - 14; Sp = Sp + 8; call (P64[Sp])(R1) args: 24, res: 0, upd: 24; } }] }}} {{{ 00000000000002b8 <c2Lc_info>: 2b8: 4c 89 e0 mov %r12,%rax 2bb: 4c 8d 60 18 lea 0x18(%rax),%r12 2bf: 4d 3b a5 18 01 00 00 cmp 0x118(%r13),%r12 2c6: 76 10 jbe 2d8 <c2Lc_info+0x20> 2c8: 49 c7 85 48 01 00 00 movq $0x18,0x148(%r13) 2cf: 18 00 00 00 2d3: e9 00 00 00 00 jmpq 2d8 <c2Lc_info+0x20> 2d4: R_X86_64_PC32 stg_gc_pp+0xfffffffffffffffc 2d8: 48 c7 40 08 00 00 00 movq $0x0,0x8(%rax) 2df: 00 2dc: R_X86_64_32S ghczmprim_GHCziTypes_ZC_con_info 2e0: 48 89 58 10 mov %rbx,0x10(%rax) 2e4: 4c 89 70 18 mov %r14,0x18(%rax) 2e8: 48 8b 45 08 mov 0x8(%rbp),%rax 2ec: 48 83 c5 08 add $0x8,%rbp 2f0: 49 8d 5c 24 f2 lea -0xe(%r12),%rbx 2f5: ff e0 jmpq *%rax 2f7: 90 nop }}} It would likely be better to invert the branch at `2c6` to a `ja`, so that the common case is adjacent to the function entry, and when `llvm.expect` is used on the condition in the branch, the LLVM optimizer does produce this alternate layout. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8321> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[GHC] #9010: TemplateHaskell leads to an "unknown symbol" error
by GHC 04 Nov '16

04 Nov '16

#9010: TemplateHaskell leads to an "unknown symbol" error ------------------------------------+------------------------------------- Reporter: Feuerbach | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | 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: | ------------------------------------+------------------------------------- Code: {{{ {-# LANGUAGE TemplateHaskell #-} module M where import System.Console.Editline import Data.Lens.Template data S = S { _a :: EditLine } makeLens ''S }}} When compiled with `ghc link.hs`, everything is fine; {{{ [1 of 1] Compiling M ( link.hs, link.o ) Loading package ghc-prim ... linking ... done. Loading package integer-gmp ... linking ... done. Loading package base ... linking ... done. Loading package array-0.4.0.1 ... linking ... done. Loading package deepseq-1.3.0.1 ... linking ... done. Loading package containers-0.5.0.0 ... linking ... done. Loading package bytestring-0.10.0.2 ... linking ... done. Loading package tagged-0.7.1 ... linking ... done. Loading package transformers-0.3.0.0 ... linking ... done. Loading package transformers-compat-0.1.1.1 ... linking ... done. Loading package contravariant-0.4.4 ... linking ... done. Loading package distributive-0.4.3.1 ... linking ... done. Loading package mtl-2.1.3.1 ... linking ... done. Loading package text-1.1.0.1 ... linking ... done. Loading package hashable-1.2.1.0 ... linking ... done. Loading package nats-0.1.2 ... linking ... done. Loading package unordered-containers-0.2.3.3 ... linking ... done. Loading package semigroups-0.12.2 ... linking ... done. Loading package comonad-4.0.1 ... linking ... done. Loading package semigroupoids-4.0.1 ... linking ... done. Loading package data-lens-2.10.4 ... linking ... done. Loading package pretty-1.1.1.0 ... linking ... done. Loading package template-haskell ... linking ... done. Loading package data-lens-template-2.1.8 ... linking ... done. }}} However, when compiling with `ghc -package-id editline-0.2.1.1-ef91ae78e7ea1b9142c2fb0b84e85c11 link.hs`, I get {{{ [1 of 1] Compiling M ( link.hs, link.o ) Loading package ghc-prim ... linking ... done. Loading package integer-gmp ... linking ... done. Loading package base ... linking ... done. Loading package editline-0.2.1.1 ... linking ... ghc: /home/feuerbach/work /.cabal-sandbox/lib/i386-linux- ghc-7.6.3/editline-0.2.1.1/HSeditline-0.2.1.1.o: unknown symbol `el_reset' ghc: unable to load package `editline-0.2.1.1' }}} This is a problem because that's how cabal invokes ghc. This looks similar to #4899, except there's nothing "non-standard" here (no profiling or dynamic linking). -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9010> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 24

[GHC] #8870: GHC 7.8.0 RC2 fails when compiling a hello world program on Windows 7 32bits
by GHC 01 Oct '16

01 Oct '16

#8870: GHC 7.8.0 RC2 fails when compiling a hello world program on Windows 7 32bits ----------------------------+--------------------------------------- Reporter: facundoq | Owner: Type: bug | Status: new Priority: highest | Milestone: Component: Compiler | Version: 7.8.1-rc2 Keywords: | Operating System: Windows Architecture: x86 | Type of failure: Compile-time crash Difficulty: Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | ----------------------------+--------------------------------------- Windows shows a window that says "GHC has stopped working" when compiling a simple hello world. Steps to reproduce 1) Download GHC 7.8 RC2 (http://www.haskell.org/ghc/dist/7.8.1-rc2/ghc-7.8.0.20140228-i386 -unknown-mingw32.tar.bz2) 2) Extract the contents of the tar to folder $GHC$ 3) Add $GHC$\bin and $GHC$\mingw\bin folder to the PATH, add env var "LANG=C". 4) Create file test.hs with contents: main = putStrLn "Hello, World!" 5) run "ghc test.hs". Interestingly, the error is not deterministic, ie, after running the command several times the program gets compiled. When that happens, and ff the file test.o is not deleted, further attempts to compile run succesfully. GHCi appears to work normally (when doing simple tests). More details: http://stackoverflow.com/questions/22291739/cant-install-packages-with- cabal-using-ghc-7-8rc2-and-windows-7 -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8870> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 22

[GHC] #9349: excessive inlining with -fpre-inlining
by GHC 29 Sep '16

29 Sep '16

#9349: excessive inlining with -fpre-inlining --------------------------------------------+------------------------------ Reporter: rwbarton | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.8.3 Keywords: | Differential: Operating System: Unknown/Multiple | Architecture: Type of failure: Runtime performance bug | Unknown/Multiple Test Case: | Difficulty: Unknown Blocking: | Blocked By: | Related Tickets: --------------------------------------------+------------------------------ The program at https://gist.github.com/jorendorff/a3005968adc8f054baf7 runs very slowly when compiled with `-O` or higher. It seems that `arr` and/or `rangeMap` is being inlined into the do block on lines 89-91 and recomputed for each iteration of the loop. The program runs nearly instantly when compiled with `-O0` or with `-O -fno-pre-inlining`. (Of course, this does not mean `-fpre-inlining` is necessary the culprit; it could be enabling some subsequent misoptimization.) -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9349> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 10

[GHC] #8774: Transitivity of Auto-Specialization
by GHC 28 Sep '16

28 Sep '16

#8774: Transitivity of Auto-Specialization -------------------------+------------------------------------------------- Reporter: | Owner: crockeea | Status: new Type: bug | Milestone: Priority: | Version: 7.6.3 normal | Operating System: Linux Component: | Type of failure: Compile-time performance bug Compiler | Test Case: Keywords: | Blocking: Architecture: | Unknown/Multiple | Difficulty: | Unknown | Blocked By: | Related Tickets: 5928, | 8668, 8099 | -------------------------+------------------------------------------------- From [http://www.haskell.org/ghc/docs/7.6.3/html/users_guide/pragmas.html#idp4986… the docs]: [Y]ou often don't even need the SPECIALIZE pragma in the first place. When compiling a module M, GHC's optimiser (with -O) automatically considers each top-level overloaded function declared in M, and specialises it for the different types at which it is called in M. The optimiser also considers each imported INLINABLE overloaded function, and specialises it for the different types at which it is called in M. ... Moreover, given a SPECIALIZE pragma for a function f, GHC will automatically create specialisations for any type-class-overloaded functions called by f, if they are in the same module as the SPECIALIZE pragma, or if they are INLINABLE; and so on, transitively. So GHC should automatically specialize some/most/all(?) functions marked `INLINABLE` ''without'' a pragma, and if I use an explicit pragma, the specialization is transitive. My question is: is the ''auto''-specialization transitive? Either way, I'd like to see the docs updated to answer this question. Specifically, the attached files demonstrate a bug if auto-specialization ''should'' be transitive. Main.hs: {{{ #!haskell import Data.Vector.Unboxed as U import Foo main = let y = Bar $ Qux $ U.replicate 11221184 0 :: Foo (Qux Int) (Bar (Qux ans)) = iterate (plus y) y !! 100 in putStr $ show $ foldl1' (*) ans }}} Foo.hs: {{{ #!haskell module Foo (Qux(..), Foo(..), plus) where import Data.Vector.Unboxed as U newtype Qux r = Qux (Vector r) -- GHC inlines `plus` if I remove the bangs or the Baz constructor data Foo t = Bar !t | Baz !t instance (Num r, Unbox r) => Num (Qux r) where {-# INLINABLE (+) #-} (Qux x) + (Qux y) = Qux $ U.zipWith (+) x y {-# INLINABLE plus #-} plus :: (Num t) => (Foo t) -> (Foo t) -> (Foo t) plus (Bar v1) (Bar v2) = Bar $ v1 + v2 }}} GHC specializes the call to `plus`, but does *not* specialize `(+)` in the `Qux` `Num` instance. (In the attached core excerpt: `main6` calls `iterate main8`. `main8` is just `plus`, specialized for `Int`. So far so good. However, `splus` calls the *polymorphic* `c+`. If auto- specialization is transitive, I expect `c+` to be specialized to `Int`.) This kills performance: an explicit pragma `{-# SPECIALIZE plus :: Foo (Qux Int) -> Foo (Qux Int) -> Foo (Qux Int) #-}` results in ''transitive'' specialization as the docs indicate, so `(+)` is specialized and the code is 30x faster. Is this expected behavior? Should I only expect `(+)` to be specialized transitively with an explicit pragma? Note: this question is different from #5928 for two reasons: 1. I believe that no inlining is occuring, and hence I don't think inlining is interfering with specialization 2. I have `INLINABLE` pragmas on all relevant functions. Note: this question is different from #8668 because I am asking about ''auto''-specialization. This question was originally posted on [http://stackoverflow.com/questions/21502335/transitivity-of-auto- specialization-in-ghc StackOverflow]. As mentioned in the comments of that question, I am intentionally ''not'' fully applying the call to `plus` in Main, contrary to the suggestions in #8099. I'd love to see why I'm getting that behavior as well. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8774> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 17