March 2017 - ghc-tickets

[GHC] #8082: Ordering of assembly blocks affects performance
by GHC 10 May '17

10 May '17

#8082: Ordering of assembly blocks affects performance ----------------------------------+--------------------------------- Reporter: jstolarek | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler (NCG) | Version: 7.6.3 Keywords: | Operating System: Linux Architecture: x86_64 (amd64) | Type of failure: None/Unknown Difficulty: Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | ----------------------------------+--------------------------------- During my work on #6135 I noticed that performance of reverse-complem benchmark in nofib depends highly on the order in which assembly block are laid out. With my patches I am consistently getting a 18-20% speed-up. In theory my patches should not impact performance of existing programs, but for some reason they affect the ordering of generated assembly blocks. On of the earlier versions of my patch I noticed that kahan benchmark suffered a 16% performance hit and again the only difference I noticed in the generated assembly was ordering of blocks. I did a more in-depth investigation in case of kahan and it turned out that this difference results from the way Core is generated: the difference between HEAD and my patches was that a worker function had its three parameters passed in different order. I did not investigate this for reverse-complem because Core is considerably larger, but I could spend some time on it if it might be relevant. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8082> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 8

[GHC] #13150: unknown warning is not reported by GHC
by GHC 09 May '17

09 May '17

#13150: unknown warning is not reported by GHC -------------------------------------+------------------------------------- Reporter: jonaprieto | Owner: Type: bug | Status: new Priority: high | Milestone: Component: Compiler | Version: 8.0.1 Keywords: | Operating System: MacOS X Architecture: | Type of failure: Incorrect Unknown/Multiple | error/warning at compile-time Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- The documented and official flag for the warning when a module, function or type with a WARNING or DEPRECATED pragma is used is `-fwarn-warnings- deprecations`. (See for instance https://downloads.haskell.org/~ghc/7.8.2/docs/html/users_guide/options- sanity.html). If the goal is turn off the warning I use `-fno-warn- warnings-deprecations` flag. By default `-fwarn-warnings-deprecations` on. In the next example shows a warning by using the flag mention above. {{{ $ cat Foo.hs {-# OPTIONS_GHC -fwarn-warnings-deprecations #-} import Data.Version (Version(Version, versionBranch, versionTags)) main :: IO () main = do let v = Version { versionBranch = [1,0,0] , versionTags = ["beta"] } return () }}} and GHC outputs: {{{ $ ghc Foo.hs [1 of 1] Compiling Main ( Foo.hs, Foo.o ) Foo.hs:9:11: warning: [-Wdeprecations] In the use of ‘versionTags’ (imported from Data.Version): Deprecated: "See GHC ticket #2496" Linking Foo ... }}} The first unexcepted behavior is the warning message showed above. Following the convention the warning message should be: {{{ Foo.hs:9:11: warning: [-Wwarnings-deprecations] }}} The second unexcepted behavior happened when I used a pragma to silenced the warning `warnings-deprecations`, I committed the mistake to misspelled it, but surprisingly GHC didn't complaint about the unknown pragma, it seems treated again like a synonymous of `-fno-warn-warnings- deprecations` in this case. {{{ {-# OPTIONS_GHC -fno-warn-deprecations #-} import Data.Version (Version(Version, versionBranch, versionTags)) main :: IO () main = do let v = Version { versionBranch = [1,0,0] , versionTags = ["beta"] } return () }}} Searching a little, I found it is not documented the flag -Wdeprecations in recent versions of GHC ( I tested with 7.8.2 too) but it does for instance available in GHC 6.3.0 (See https://manned.org/ghc-cvs/42a4c961) but it actually continues appearing in the source code of the recent GHC. See https://github.com/mlen/ghc/search?utf8=%E2%9C%93&q=deprecations. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13150> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 1

[GHC] #13239: Phabricator upgrade broke the ticket custom field
by GHC 09 May '17

09 May '17

#13239: Phabricator upgrade broke the ticket custom field -------------------------------------+------------------------------------- Reporter: mpickering | 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: -------------------------------------+------------------------------------- We have one custom field which should also appear in commit messages when running "arc diff". Since the upgrade on 07/02/2017, the field no longer appears when running "arc diff". The upgrade path is detailed in this upstream task. https://secure.phabricator.com/T12085 -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13239> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

[GHC] #8332: hp2ps does not escape parentheses
by GHC 09 May '17

09 May '17

#8332: hp2ps does not escape parentheses ------------------------------------+------------------------------------- Reporter: luite | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Profiling | 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: | ------------------------------------+------------------------------------- If some symbol name in the key includes a (, but is cut off before the ), hp2ps generates an invalid postscript file. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8332> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

[GHC] #13211: NegativeLiterals -0.0 :: Double
by GHC 09 May '17

09 May '17

#13211: NegativeLiterals -0.0 :: Double -------------------------------------+------------------------------------- Reporter: rwbarton | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.1 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: -------------------------------------+------------------------------------- Compare {{{ rwbarton@morphism:~$ ghc-8.0.1 -e '-0.0' -0.0 rwbarton@morphism:~$ ghc-8.0.1 -XNegativeLiterals -e '-0.0' 0.0 }}} This behavior is logical, but unexpected. The logic is that with `NegativeLiterals` the literal `-0.0` means `fromRational ((-0)%0)`, which equals `fromRational 0`, which for `Double` is positive zero, not negative zero. But this seems unfortunate. Perhaps it would be best if `-0.0` (or `-` applied to any other literal mathematically equal to zero) was desugared as `negate 0.0` even under `NegativeLiterals`, since there doesn't seem to be any other reason to write `-0.0`. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13211> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 14

[GHC] #8101: No pattern match non-exhaustiveness warnings when compiling with -fno-code
by GHC 07 May '17

07 May '17

#8101: No pattern match non-exhaustiveness warnings when compiling with -fno-code -------------------------+------------------------------------------------- Reporter: exbb2 | Owner: Type: bug | Status: new Priority: | Milestone: normal | Version: 7.6.3 Component: | Operating System: Unknown/Multiple Compiler | Type of failure: Incorrect warning at Keywords: | compile-time Architecture: | Test Case: Unknown/Multiple | Blocking: Difficulty: | Unknown | Blocked By: | Related Tickets: | -------------------------+------------------------------------------------- {{{ {-# OPTIONS_GHC -Wall #-} module A where data ABC = A | B | C abc :: ABC -> Int abc x = case x of A -> 1 }}} {{{ $ ghc -fno-code -fforce-recomp /tmp/A.hs [1 of 1] Compiling A ( /tmp/A.hs, nothing ) }}} {{{ $ ghc -fobject-code -fforce-recomp /tmp/A.hs [1 of 1] Compiling A ( /tmp/A.hs, /tmp/A.o ) /tmp/A.hs:7:9: Warning: Pattern match(es) are non-exhaustive In a case alternative: Patterns not matched: B C }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8101> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 9

[GHC] #10431: EqualityConstraints extension?
by GHC 06 May '17

06 May '17

#10431: EqualityConstraints extension? -------------------------------------+------------------------------------- Reporter: adamgundry | Owner: Type: feature | Status: new request | Milestone: 7.12.1 Priority: normal | Version: 7.11 Component: Compiler | Operating System: Unknown/Multiple (Type checker) | Type of failure: None/Unknown Keywords: | Blocked By: Architecture: | Related Tickets: Unknown/Multiple | Test Case: | Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- At the moment, writing an equality constraint in a type requires at least one of the `GADTs` or `TypeFamilies` extensions. However, each of these has other effects. Could we have an `EqualityConstraints` extension to permit equality constraints in types, but neither GADTs nor type families? Presumably this extension should imply `MonoLocalBinds`. The `GADTs` extension could then become precisely the conjunction of `GADTSyntax`, `EqualityConstraints` and `ExistentialQuantification`. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10431> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 8

[GHC] #12947: Core lint error
by GHC 06 May '17

06 May '17

#12947: Core lint error -------------------------------------+------------------------------------- Reporter: Iceland_jack | 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: -------------------------------------+------------------------------------- {{{#!hs import qualified Control.Monad.Fail as Fail newtype P m a = P { unP :: (a -> IO (m ())) -> IO (m ()) } instance Functor (P m) where instance Applicative (P m) where instance Monad (P m) where instance (Fail.MonadFail m) => Fail.MonadFail (P m) where fail msg = ContT $ \ _ -> Fail.fail msg }}} fails when run with `$ ghci -ignore-dot-ghci -fdefer-typed-holes -dcore- lint /tmp/tQZR.hs`, log is attached -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12947> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

[GHC] #12156: -fdefer-typed-holes causes panic on unbound variable
by GHC 06 May '17

06 May '17

#12156: -fdefer-typed-holes causes panic on unbound variable -------------------------------------+------------------------------------- Reporter: osa1 | Owner: Type: bug | Status: new Priority: high | Milestone: Component: Compiler | Version: 8.1 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: -------------------------------------+------------------------------------- {{{#!haskell module Main where main = do print v return () }}} Compile it without {{{-fdefer-typed-holes}}} and it fails as expected: {{{ ➜ hole_bug ghc --make Main.hs -fforce-recomp [1 of 1] Compiling Main ( Main.hs, Main.o ) Main.hs:4:11: error: Variable not in scope: v }}} Adding {{{-fdefer-typed-holes}}} is causing a panic instead of runtime error: {{{ ➜ hole_bug ghc --make Main.hs -fforce-recomp -fdefer-typed-holes [1 of 1] Compiling Main ( Main.hs, Main.o ) Main.hs:4:11: warning: [-Wtyped-holes] Variable not in scope: v ghc: panic! (the 'impossible' happened) (GHC version 8.0.1 for x86_64-unknown-linux): StgCmmEnv: variable not found $dShow_aKo local binds for: $trModule $trModule1_r16k $trModule2_r16s sat_s16x sat_s16z Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug }}} Tried with: 8.0.1, HEAD as of today. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12156> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

[GHC] #9557: Deriving instances is slow
by GHC 06 May '17

06 May '17

#9557: Deriving instances is slow -------------------------------------+------------------------------------- Reporter: Feuerbach | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.8.3 Keywords: | Operating System: Architecture: Unknown/Multiple | Unknown/Multiple Difficulty: Unknown | Type of failure: Blocked By: | None/Unknown Related Tickets: | Test Case: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- Let's take [https://raw.githubusercontent.com/haskell-suite/haskell-src- exts/8e0153d33b66add96fd5606614ce7938d2029510/src/Language/Haskell/Exts/Annotated/Syntax.hs this file] (from haskell-src-exts) and build it with -O0. On my machine it takes 55s. If we remove deriving of all classes except Functor (which is the only one used by the module itself), the time will drop down to 4s. Different classes vary in their compile-time cost, but there's no single culprit. Among the costly ones are Generic, Data, Show, Ord. haskell-src-exts users are very annoyed by its long compile time, about 10 minutes, especially because we don't compile with -O0 by default. It would be nice to get this fixed. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9557> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 24