February 2017 - ghc-tickets

[GHC] #10930: missing empty-Enumeration warning for `Natural`
by GHC 03 Dec '18

03 Dec '18

#10930: missing empty-Enumeration warning for `Natural` -------------------------------------+------------------------------------- Reporter: hvr | Owner: Type: bug | Status: new Priority: low | Milestone: 8.0.1 Component: Compiler | Version: 7.10.2 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: #7881, #10929 Differential Rev(s): | -------------------------------------+------------------------------------- -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10930> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

[GHC] #9505: Bounded instance for Word (and possibly others) uses unboxed literals
by GHC 03 Dec '18

03 Dec '18

#9505: Bounded instance for Word (and possibly others) uses unboxed literals -------------------------------------+------------------------------------- Reporter: schyler | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.8.2 Keywords: | Operating System: Architecture: Unknown/Multiple | Unknown/Multiple Difficulty: Unknown | Type of failure: Blocked By: | None/Unknown Related Tickets: | Test Case: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- There's a comment above saying GHC won't optimise. I think since 7.8 they are unboxed by default, so that comment and the explicit constructors can be removed. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9505> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 7

[GHC] #12643: class declaration works in ghci, but not in a file
by GHC 02 Dec '18

02 Dec '18

#12643: class declaration works in ghci, but not in a file -------------------------------------+------------------------------------- Reporter: dmwit | Owner: Type: bug | Status: new Priority: normal | 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: -------------------------------------+------------------------------------- The following ghci session prints no errors: {{{ > :set -XStandaloneDeriving -XDeriveGeneric > import GHC.Generics > import Control.Monad.Except > deriving instance Generic (ExceptT e m a) > class F a where f :: Rep (Except String a) x }}} However, when I transfer this to a file: {{{ {-# LANGUAGE StandaloneDeriving #-} {-# LANGUAGE DeriveGeneric #-} import GHC.Generics import Control.Monad.Except deriving instance Generic (ExceptT e m a) class F a where f :: Rep (Except String a) x }}} I get a mysterious error: {{{ test.hs:6:17: error: • Couldn't match type ‘Rep (Except String a0)’ with ‘Rep (Except String a)’ Expected type: Rep (Except String a) x Actual type: Rep (Except String a0) x NB: ‘Rep’ is a type function, and may not be injective The type variable ‘a0’ is ambiguous • In the ambiguity check for ‘f’ To defer the ambiguity check to use sites, enable AllowAmbiguousTypes When checking the class method: f :: forall a. F a => forall x. Rep (Except String a) x In the class declaration for ‘F’ }}} If I turn on -fdefer-type-errors, I can verify that the type family indeed reduces far enough for a and x to be arguments to injective types, so I believe GHC should not consider this an error: {{{ > :set -XRankNTypes > :kind! forall a x. Rep (Except String a) x forall a x. Rep (Except String a) x :: * = D1 ('MetaData "ExceptT" "Control.Monad.Trans.Except" "transformers-0.5.2.0" 'True) (C1 ('MetaCons "ExceptT" 'PrefixI 'False) (S1 ('MetaSel 'Nothing 'NoSourceUnpackedness 'NoSourceStrictness 'DecidedLazy) (Rec0 (Data.Functor.Identity.Identity (Either [Char] a))))) x }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12643> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 3

[GHC] #8043: Feature Request : Qualified module exports
by GHC 01 Dec '18

01 Dec '18

#8043: Feature Request : Qualified module exports ------------------------------------+------------------------------------- Reporter: erikd | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.7 Keywords: | Operating System: Unknown/Multiple Architecture: Unknown/Multiple | Type of failure: Other Difficulty: Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | ------------------------------------+------------------------------------- I quite like the practice of using `-XNoImplicitPrelude` and then defining my own (often per-project) Prelude doing things like: {{{ module MyProject.Prelude ( module P ) import Data.Text as P (Text) import Prelude as P hiding (head, init, tail, last) }}} As an extension to the above idea, I would like to add qualified module exports defined so that I can do this: {{{ module MyProject.Prelude ( module P , qualified module T ) import Data.Text as P (Text) import Prelude as P hiding (head, init, tail, last) import qualified Data.Text as T }}} so that any file which imports `MyProject.Prelude` will then have all the identifiers that they would normally get if they did: {{{ import qualified Data.Text as T }}} Is this something that would have to go through the Haskell Next procedure? If so, how does one get that kicked off? -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8043> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 13

[GHC] #8177: Roles for type families
by GHC 29 Nov '18

29 Nov '18

#8177: Roles for type families ------------------------------------+------------------------------------- Reporter: simonpj | 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: | ------------------------------------+------------------------------------- Now that we have [wiki:Roles roles], it might be helpful to be able to give a role signature for a data/type family. At the moment, data/type family constructors have all parameters conservatively assigned to be role `N`. Thus {{{ data family D a -- Parameter conservatively assumed to be N class C a where op :: D a -> a instance C Int where .... newtype N a = MkN a deriving( C ) -- Rejected }}} The generalised-newtype-deriving clause `deriving( C )` is rejected because `D` might use its parameter at role `N` thus: {{{ data instance D [b] = MkD (F b) -- F is a type function }}} It would be strictly more expressive if we could * '''Declare''' the roles of D's arguments (as we can declare their kinds). E.g. {{{ data family D a@R }}} * '''Check''' that each family instance obeys that role signature. E.g. given the preceding role signature, reject this instance: {{{ data instance D [b] = MkD (F b) -- F is a type function }}} I think there is no technical difficulty here. Just a question of doing it. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8177> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 41

[GHC] #12965: String merging broken on Windows
by GHC 26 Nov '18

26 Nov '18

#12965: String merging broken on Windows ----------------------------------------+--------------------------------- Reporter: bgamari | Owner: Type: bug | Status: new Priority: normal | Milestone: 8.2.1 Component: Compiler | Version: 8.0.1 Keywords: | Operating System: Windows Architecture: Unknown/Multiple | Type of failure: None/Unknown Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: ----------------------------------------+--------------------------------- #9577 introduced string literal merging via the linker. Unfortunately I haven't been able to get this to function on Windows, resulting in testsuite failures of #9577. Even `gcc` appears to do the wrong thing here, {{{ $ cat hi.c #include <stdio.h> const char *hi = "hello world"; extern const char *hi2; void main() { printf("%p %p", hi, hi2); } $ cat hi2.c const char *hi2 = "hello world"; $ gcc -c hi.c; gcc -c hi2.c; gcc -fmerge-all-constants hi2.o hi.o $ ./a.exe 0x100403040 0x100403030 $ gcc -s hi2.c $ cat hi2.s .file "hi2.c" .globl hi .section .rdata,"dr" .LC0: .ascii "hello world\0" .data .align 8 hi: .quad .LC0 .ident "GCC: (GNU) 5.3.0" }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12965> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 10

[GHC] #10844: CallStack should not be inlined
by GHC 26 Nov '18

26 Nov '18

#10844: CallStack should not be inlined -------------------------------------+------------------------------------- Reporter: nomeata | Owner: Type: task | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.10.2 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: | -------------------------------------+------------------------------------- The use of CallStack based error messages (since changeset:6740d70d95cb) has led to some code size increase, and I believe this needs to be improved. While investigating the #10788, I was looking at the core produced by that code, and I found this in the code: {{{ -- RHS size: {terms: 2, types: 0, coercions: 0} Main.main17 :: [Char] [GblId, Str=DmdType, Unf=Unf{Src=<vanilla>, TopLvl=True, Value=False, ConLike=False, WorkFree=False, Expandable=False, Guidance=IF_ARGS [] 100 0}] Main.main17 = unpackCString# "vecto_20t0fI93Jj9LUbaZg6e04I"# -- RHS size: {terms: 2, types: 0, coercions: 0} Main.main16 :: [Char] [GblId, Str=DmdType, Unf=Unf{Src=<vanilla>, TopLvl=True, Value=False, ConLike=False, WorkFree=False, Expandable=False, Guidance=IF_ARGS [] 110 0}] Main.main16 = unpackCString# "Data.Vector.Primitive.Mutable"# -- RHS size: {terms: 2, types: 0, coercions: 0} Main.main15 :: [Char] [GblId, Str=DmdType, Unf=Unf{Src=<vanilla>, TopLvl=True, Value=False, ConLike=False, WorkFree=False, Expandable=False, Guidance=IF_ARGS [] 120 0}] Main.main15 = unpackCString# "./Data/Vector/Primitive/Mutable.hs"# -- RHS size: {terms: 2, types: 0, coercions: 0} Main.main14 :: Int [GblId, Caf=NoCafRefs, Str=DmdType, Unf=Unf{Src=<vanilla>, TopLvl=True, Value=True, ConLike=True, WorkFree=True, Expandable=True, Guidance=IF_ARGS [] 10 20}] Main.main14 = I# 97# -- RHS size: {terms: 2, types: 0, coercions: 0} Main.main13 :: Int [GblId, Caf=NoCafRefs, Str=DmdType, Unf=Unf{Src=<vanilla>, TopLvl=True, Value=True, ConLike=True, WorkFree=True, Expandable=True, Guidance=IF_ARGS [] 10 20}] Main.main13 = I# 16# -- RHS size: {terms: 2, types: 0, coercions: 0} Main.main12 :: Int [GblId, Caf=NoCafRefs, Str=DmdType, Unf=Unf{Src=<vanilla>, TopLvl=True, Value=True, ConLike=True, WorkFree=True, Expandable=True, Guidance=IF_ARGS [] 10 20}] Main.main12 = I# 21# -- RHS size: {terms: 8, types: 0, coercions: 0} Main.main11 :: SrcLoc [GblId, Str=DmdType, Unf=Unf{Src=<vanilla>, TopLvl=True, Value=True, ConLike=True, WorkFree=True, Expandable=True, Guidance=IF_ARGS [] 10 80}] Main.main11 = SrcLoc Main.main17 Main.main16 Main.main15 Main.main14 Main.main13 Main.main14 Main.main12 }}} This is clearly a CallStack from a library, and there is no point in copying that into the (every?) use of wherever that came from, as it will only increase code size and slow down compilation. I don’t know how to fix it, though. Float out CallStacks more aggressively? Or is the problem that unfoldings are recorded before that can happen? -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10844> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 33

[GHC] #10803: New TypeSignatureSections extension
by GHC 26 Nov '18

26 Nov '18

#10803: New TypeSignatureSections extension -------------------------------------+------------------------------------- Reporter: hvr | Owner: Type: bug | Status: new Priority: normal | Milestone: 7.12.1 Component: Compiler | Version: 7.10.2 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: | -------------------------------------+------------------------------------- = Preliminary specification Have - `(::` ''type'' `)` be syntax sugar for - `(\x -> x ::` ''type'' `)` This new syntax extension is enabled via a new `TypeSignatureSections` language pragma, i.e. {{{#!hs {#- LANGUAGE TypeSignatureSections #-} }}} TODO: formalise this a bit more -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10803> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 19

[GHC] #12232: Opportunity to do better in register allocations
by GHC 25 Nov '18

25 Nov '18

#12232: Opportunity to do better in register allocations -------------------------------------+------------------------------------- Reporter: harendra | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.1 (NCG) | Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: Runtime Unknown/Multiple | performance bug Test Case: | Blocked By: Blocking: | Related Tickets: #12231 Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- [https://github.com/harendra-kumar/unicode-transforms/tree/GHC_PERF_ISSUE This github repo] demonstrates that introduction of one "if" condition which is never taken, degrades the performance of the fast path loop in this particular use case by around 70%, from 2.6ms to 4.4 ms. On investigation I found two main issues. See #12231(ticket) regarding rdeundant allocations. This one is about the other issue regarding register allocations. [https://github.com/harendra-kumar/unicode- transforms/commit/261dbb35e753376e4ee160dab556051738a0be50#diff- 4be4240cb00d1b13a038d1d45fbae235 See this assembly diff] of instructions executed in both cases. The repo has both versions to play with and simpl/stg/cmm/asm are checked in. The first file in the diff is with the if condition and the second one without. To summarize, the introduction of "if" condition adds the following register manipulations before and after the main logic: {{{ # reassignments # rbx -> r10 -> r9 -> r8 -> rdi -> rsi -> rdx -> rcx -> rbx => 0x408d7e: mov %rbx,0x90(%rsp) => 0x408d86: mov %rcx,%rbx => 0x408d89: mov %rdx,%rcx => 0x408d8c: mov %rsi,%rdx => 0x408d8f: mov %rdi,%rsi => 0x408d92: mov %r8,%rdi => 0x408d95: mov %r9,%r8 => 0x408d98: mov %r10,%r9 => 0x408d9b: mov 0x90(%rsp),%r10 . . . loop logic here which uses only %rax, %r10 and %r9 . . . . # _n4s8: # shuffle back to original assignments => 0x4090dc: mov %r14,%r11 => 0x4090df: mov %r9,%r10 => 0x4090e2: mov %r8,%r9 => 0x4090e5: mov %rdi,%r8 => 0x4090e8: mov %rsi,%rdi => 0x4090eb: mov %rdx,%rsi => 0x4090ee: mov %rcx,%rdx => 0x4090f1: mov %rbx,%rcx => 0x4090f4: mov %rax,%rbx => 0x4090f7: mov 0x88(%rsp),%rax => 0x4090ff: jmpq 0x408d2a }}} The registers seem to be getting reassigned here, data flowing from one to the next. In this particular path a lot of these register movements seem unnecessary and are only undone at the end without being used. My suspicion is that this allocation is being done en bloc keeping several paths in mind, just like the heap allocations I reported in the other issue. Some questions to be answered: * Can we make this smarter? * Will `-fregs-graph` do a better job and if not what improvements can be done to graph coloring allocator to do a good job here? I have try that option yet. LLVM assembly trace is also available in the repo. == References == * [1] mail thread: https://mail.haskell.org/pipermail/ghc- devs/2016-June/012243.html * [2] github code repo: https://github.com/harendra-kumar/unicode- transforms/tree/GHC_PERF_ISSUE -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12232> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[GHC] #13262: Allow type synonym family application in instance head if it reduces away
by GHC 25 Nov '18

25 Nov '18

#13262: Allow type synonym family application in instance head if it reduces away -------------------------------------+------------------------------------- Reporter: ezyang | Owner: Type: feature | Status: new request | Priority: lowest | Milestone: Component: Compiler | Version: 8.0.1 (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: -------------------------------------+------------------------------------- GHC rejects the following program: {{{ {-# LANGUAGE TypeFamilies, FlexibleInstances #-} module F where type family F a type instance F Int = Int -> Int instance Show (F Int) where show _ = "fun" }}} But this doesn't seem fundamentally problematic to me: at the instance declaration, we know that F Int will reduce, and after that reduction there isn't any problem with the instance. Here's a less restrictive instance head test: first, we reduce type families as much as possible. Then, if there are still leftover non- reducing type families, reject the instance. After having typed this up, I have realized that I don't actually need this feature. But I'm filing it here for posterity. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13262> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 14