June 2015 - ghc-tickets

[GHC] #9661: Branchless ==# is compiled to branchy code
by GHC 08 Sep '17

08 Sep '17

#9661: Branchless ==# is compiled to branchy code -------------------------------------+------------------------------------- Reporter: dfeuer | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.9 Keywords: | Operating System: Architecture: Unknown/Multiple | Unknown/Multiple Difficulty: Unknown | Type of failure: Runtime Blocked By: | performance bug Related Tickets: | Test Case: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- This started as a comment on #6135, but Jan Stolarek requested that I file a separate report. The branchless tests `<#`, `>#`, `<=#`, and `>=#`, and their wordy cousins, seem to work properly, but `==#` and `eqWord#` don't. If I write {{{#!hs foo x = tagToEnum# ((x <# 3#) `orI#` (x ># 100#) `orI#` (x ==# 12#) `orI#` (x ==# 15#)) }}} I get (in 7.8.3 and in 7.9) {{{#!hs IsDigit.foo :: GHC.Prim.Int# -> GHC.Types.Bool [GblId, Arity=1, Caf=NoCafRefs, Str=DmdType <S,1*U>, Unf=Unf{Src=InlineStable, TopLvl=True, Value=True, ConLike=True, WorkFree=True, Expandable=True, Guidance=ALWAYS_IF(arity=1,unsat_ok=True,boring_ok=False) Tmpl= \ (x3_a2if [Occ=Once!] :: GHC.Prim.Int#) -> case x3_a2if of wild_Xe { __DEFAULT -> GHC.Prim.tagToEnum# @ GHC.Types.Bool (GHC.Prim.orI# (GHC.Prim.<# wild_Xe 3) (GHC.Prim.># wild_Xe 100)); 12 -> GHC.Types.True; 15 -> GHC.Types.True }}] IsDigit.foo = \ (x3_a2if :: GHC.Prim.Int#) -> case x3_a2if of wild_Xe { __DEFAULT -> GHC.Prim.tagToEnum# @ GHC.Types.Bool (GHC.Prim.orI# (GHC.Prim.<# wild_Xe 3) (GHC.Prim.># wild_Xe 100)); 12 -> GHC.Types.True; 15 -> GHC.Types.True }}} and branching assembly to match. Jan Stolarek indicates this is probably a problem with constant folding. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9661> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 30

[GHC] #10365: Make Semigroup as a superclass of Monoid
by GHC 07 Sep '17

07 Sep '17

#10365: Make Semigroup as a superclass of Monoid -------------------------------------+------------------------------------- Reporter: gidyn | Owner: Type: feature | Status: new request | Milestone: Priority: normal | Version: 7.10.1 Component: | Operating System: Unknown/Multiple libraries/base | Type of failure: None/Unknown Keywords: | Blocked By: Architecture: | Related Tickets: Unknown/Multiple | Test Case: | Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- See http://haskell.1045720.n5.nabble.com/Proposal-Make-Semigroup-as-a -superclass-of-Monoid-td5767835.html -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10365> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 45

[GHC] #9797: Investigate rewriting `>>=` to `*>` or `>>` for appropriate types
by GHC 06 Sep '17

06 Sep '17

#9797: Investigate rewriting `>>=` to `*>` or `>>` for appropriate types -------------------------------------+------------------------------------- Reporter: dfeuer | Owner: dfeuer Type: task | Status: new Priority: normal | Milestone: 7.10.1 Component: Core Libraries | Version: 7.9 Keywords: | Operating System: Architecture: Unknown/Multiple | Unknown/Multiple Difficulty: Unknown | Type of failure: Runtime Blocked By: | performance bug Related Tickets: | Test Case: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- When we see `m >>= (\_ -> n)` with a type that admits an optimized `>>` or `<$`, we should try to take advantage of that. I don't currently know if this applies to any types under "GHC HQ" control. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9797> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4

[GHC] #10328: Control.Monad exports lead to weird Haddocks
by GHC 06 Sep '17

06 Sep '17

#10328: Control.Monad exports lead to weird Haddocks -------------------------------------+------------------------------------- Reporter: dfeuer | Owner: ekmett Type: feature | Status: new request | Milestone: Priority: normal | Version: 7.10.1 Component: Core | Operating System: Unknown/Multiple Libraries | Type of failure: Documentation Keywords: | bug Architecture: | Blocked By: Unknown/Multiple | Related Tickets: Test Case: | Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- `Control.Monad` exports, and therefore documents, the `Functor` class (but only the `fmap` member) and the `Monad` class. It does not export, and therefore does not document, the `Applicative` class. This is all pretty weird. Since `Functor` is in the prelude (and probably most/all alternate preludes), is there actually a reason to export it from `Control.Monad`? -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10328> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4

[GHC] #9756: Warn about unnecessary unsafeCoerce
by GHC 06 Sep '17

06 Sep '17

#9756: Warn about unnecessary unsafeCoerce -------------------------------------+------------------------------------- Reporter: dfeuer | Owner: Type: feature request | Status: new Priority: normal | Milestone: Component: Compiler (Type | Version: 7.9 checker) | Operating System: Keywords: | Unknown/Multiple Architecture: Unknown/Multiple | Type of failure: Difficulty: Unknown | None/Unknown Blocked By: | Test Case: Related Tickets: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- `unsafeCoerce` has been around a lot longer than `coerce`, so older code tends to use it even if `coerce` could do the job. It would be nice if the type checker could produce a warning when it detected such a situation. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9756> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4

[GHC] #9574: GHC Panic: No Skolem Info
by GHC 05 Sep '17

05 Sep '17

#9574: GHC Panic: No Skolem Info -------------------------------------+------------------------------------- Reporter: ian_mi | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.9 Keywords: | Operating System: Architecture: Unknown/Multiple | Unknown/Multiple Difficulty: Unknown | Type of failure: Blocked By: | None/Unknown Related Tickets: | Test Case: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- GHC 7.8.3 and from git both panic with Couldn't match kind ‘o1’ with ‘(,) o o’ ‘o1’ is untouchable inside the constraints (Object (c :><: c) a) bound by a type expected by the context: Object (c :><: c) a => Tagged a (c (FMap (ProductF c) a) (FMap (Proj2 c c) a)) at Product.hs:35:9-15ghc: panic! (the 'impossible' happened) (GHC version 7.9.20140908 for x86_64-unknown-linux): No skolem info: o1_a6XS I have not yet attempted to reduce the problem at all. The offending code is the use of fmap here: {{{#!hs proj2 :: forall (c :: o -> o -> *). ProductCategory c => NatTr (c :><: c) c (ProductF c) (Proj2 c c) proj2 = NatTr t where NatTr t = fmap (CompF Proj2 IdentityF) e e :: NatTr (c :><: c) (c :><: c) (Comp ('KProxy :: KProxy o) (Diag c) (ProductF c)) (IdentityF (c :><: c)) e = counit }}} fmap is defined as {{{#!hs fmap :: forall f (a :: o1) (b :: o1). Functor f ('KProxy :: KProxy (o1 -> o2)) => f -> Domain f a b -> Codomain f (FMap f a :: o2) (FMap f b :: o2) fmap _ = proxy morphMap (Proxy :: Proxy f) }}} while CompF is defined by {{{#!hs data CompF f g (k :: KProxy (o3 -> o4, o1 -> o2)) where CompF :: (Functor f ('KProxy :: KProxy (o3 -> o4)), Functor g ('KProxy :: KProxy (o1 -> o2))) => f -> g -> CompF f g ('KProxy :: KProxy (o3 -> o4, o1 -> o2)) }}} This makes some sense because there is a functional dependency between the functor and the kind proxy. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9574> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 9

[GHC] #8317: Optimize tagToEnum# at Core level
by GHC 29 Aug '17

29 Aug '17

#8317: Optimize tagToEnum# at Core level ------------------------------------+------------------------------------- Reporter: jstolarek | Owner: Type: task | 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: #6135 | ------------------------------------+------------------------------------- Old comparison primops that returned Bool were implemented by inserting call to `tagToEnum#` in the code generator. This call to `tagToEnum#` was later optimized away by a special case in the code geenrator (see [[GhcFile(compiler/codeGen/StgCmmExpr.hs)]], Note [case on bool])). Now that we have new comparison primops (see #6135) we no longer insert `tagToEnum#` in the code generator - all uses of `tagToEnum#` come from explicit calls in a source program. But we still have that special case in the code generator to optimize away `tagToEnum#`. We should drop that special case in the code generator and handle scrutinizing of `tagToEnum#` at the Core level by: 1. removing call to `tagToEnum#` in the scrutinee 2. adding calls to `dataToTag#` in case branches 3. constant-folding inserted `dataToTag#` Here is an example. This Haskell code: {{{ if tagToEnum# (a ># b) then E1 else E2 }}} will be translated to this Core: {{{ case tagToEnum# (a ># b) of True -> E1 False -> E2 }}} and optimized like this: {{{ case a ># b of dataToTag# True -> E1 dataToTag# False -> E2 }}} ====> {{{ case a ># b of 1 -> E1 0 -> E2 }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8317> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 8

[GHC] #9352: Allow `State# s` argument/result types in `ccall` FFI imports
by GHC 16 Aug '17

16 Aug '17

#9352: Allow `State# s` argument/result types in `ccall` FFI imports -------------------------------------+------------------------------------- Reporter: hvr | Owner: Type: feature request | Status: new Priority: normal | Milestone: 7.10.1 Component: Compiler (Type | Version: 7.8.2 checker) | Operating System: Keywords: | Unknown/Multiple Architecture: Unknown/Multiple | Type of failure: Difficulty: Unknown | None/Unknown Blocked By: | Test Case: Related Tickets: #9218 | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- This ticket is to allowing code like {{{#!hs {-# LANGUAGE UnliftedFFITypes #-} {-# LANGUAGE MagicHash #-} module M where import GHC.Exts foreign import ccall unsafe "foo" c_foo :: Int# -> State# s -> State# s foreign import ccall unsafe "bar" c_foo :: Int# -> State# s -> (# State# s, Int# #) }}} See also discussion in #9218 -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9352> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 10

[GHC] #8472: Primitive string literals prevent optimization
by GHC 16 Aug '17

16 Aug '17

#8472: Primitive string literals prevent optimization ------------------------------+-------------------------------------------- Reporter: akio | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.6.3 Keywords: | Operating System: Linux Architecture: x86_64 | Type of failure: Runtime performance bug (amd64) | Test Case: Difficulty: Unknown | Blocking: Blocked By: | Related Tickets: | ------------------------------+-------------------------------------------- Using an Addr# literal seems to result in less aggressive optimization. If I compile the attached program like this: {{{ ghc -O2 -fforce-recomp -ddump-simpl addr.hs }}} the code is optimized nicely. Everything are inlined into {{{t}}}, intermediate pairs are eliminated, etc. However, when I replace the Int# literals in the code with Addr# literals by defining the {{{ADDR}}} macro: {{{ ghc -O2 -fforce-recomp -ddump-simpl -DADDR addr.hs }}} GHC now creates 2 extra top-level bindings, each of which allocates a pair. I don't see why the presence of Addr# literals should prevent inlining, so I'm reporting a bug. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8472> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 24

[GHC] #9577: String literals are wasting space
by GHC 15 Aug '17

15 Aug '17

#9577: String literals are wasting space -------------------------------------+------------------------------------- Reporter: xnyhps | Owner: Type: bug | Status: new Priority: low | Milestone: Component: Compiler (NCG) | Version: 7.8.2 Keywords: | Operating System: Architecture: Unknown/Multiple | Unknown/Multiple Difficulty: Unknown | Type of failure: Runtime Blocked By: | performance bug Related Tickets: | Test Case: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- For [https://phabricator.haskell.org/D199 D199] I looked into how string literals are compiled down by GHC. On 64-bit OS X, a simple string `"AAA"` turns into assembly: {{{ .const .align 3 .align 0 c38E_str: .byte 65 .byte 65 .byte 65 .byte 0 }}} (And also something that invokes `unpackCString#`, but that isn't relevant here.) (`MkCore.mkStringExprFS` -> `CmmUtils.mkByteStringCLit` -> `compiler/nativeGen/X86/Ppr.pprSectionHeader`.) Note how this: * Is 8 byte aligned. * Is a `.const` section. I can't find any reason why string literals would need to be 8-byte aligned on OS X. There might be a small benefit in performance to read data starting 8-byte aligned, but I doubt doing that for string literals would be a meaningful difference. Assembly from both clang and gcc does not align string literals. The trivial program: {{{#!hs main :: IO () main = return () }}} has almost 5kB of wasted space of padding between all strings the Prelude brings in, built with GHC HEAD. The fact that it is a `.const` section, instead of `.cstring` (https://developer.apple.com/library/mac/documentation/DeveloperTools/Refere…) means duplicate strings aren't shared by the assembler. GHC floats out string literals to the top-level and uses CSE to eliminate duplicates, but that only works in a single modules. Strings shared between different modules end up as duplicate strings in an executable. The same program as above also has ~4kB of wasted space due to duplicate Prelude strings (`"base"` occurs 16 times!). Compared to the total binary size (4MB after stripping), removing this redundant data wouldn't be a big improvement (0.2%), but I still think it can be a worthwile optimization. I think this can be solved quite easily by creating a new section header for literal strings, which is unaligned and of type `.cstring`. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9577> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 26