June 2016 - ghc-tickets

[GHC] #8396: cleanup / refactor native code gens
by GHC 10 May '17

10 May '17

#8396: cleanup / refactor native code gens ----------------------------------------+---------------------------------- Reporter: carter | Owner: Type: feature request | Status: new Priority: high | Milestone: 7.10.1 Component: Compiler | Version: Keywords: | Operating System: Architecture: Unknown/Multiple | Unknown/Multiple Difficulty: Unknown | Type of failure: None/Unknown Blocked By: | Test Case: Related Tickets: #8299 , #8287 ,#8272 | Blocking: ----------------------------------------+---------------------------------- the native code gen could use some cleanup / love. 1. Catchup with feature parity of the llvm a. this includes adding SIMD support on applicable native code gen architectures, starting with x86_64 2. make it easier to extend / modify experiment with the native code gen 3. make it easier to add new primops (maybe even a magic inline asm primop?!) 4. other things i'm overlooking right now this also overlaps with a number of related tickets i've been adding to trac. a more ambitious strategic goal that won't be the initial goal, that simon marlow suggested when i chatted with him at ICFP, is to make the native code gen codebase a sort of "mini-llvm". I think thats a bit ambitious for the near term, but maybe thats the point on the horizon we may want to aim towards. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8396> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 7

[GHC] #8573: "evacuate: strange closure type 0" when creating large array
by GHC 10 May '17

10 May '17

#8573: "evacuate: strange closure type 0" when creating large array ----------------------------+---------------------------------- Reporter: nad | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.6.3 Keywords: | Operating System: Linux Architecture: x86 | Type of failure: Runtime crash Difficulty: Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | ----------------------------+---------------------------------- Consider the following code: {{{ module Main where import Data.Array xs :: [Int] xs = [0 .. 64988] crash :: Int -> IO () crash m = array (0, m) [ (x, x) | x <- xs ] `seq` return () strangeClosureType = do print (sum xs) crash (maxBound - 1) segFault1 = crash (maxBound - 1) segFault2 = do print (sum xs) crash (maxBound - 2) }}} If I compile the program using `ghc --make Main.hs -O -main-is strangeClosureType`, then I get the following error message: {{{ Main: internal error: evacuate: strange closure type 0 (GHC version 7.6.3 for i386_unknown_linux) Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug Aborted (core dumped) }}} If I don't use `-O`, or if I let `segFault1` or `segFault2` be `main`, then I get the following error message instead: {{{ Segmentation fault (core dumped) }}} If the number `30000` is replaced by some other number, then the strange closure error may be replaced by a segfault, or even no error at all. Perhaps this is another instance of bug #7762; I have only tested using GHC 7.6.3. I am using base 4.6.0.1 and array 0.4.0.1. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8573> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[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] #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] #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] #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

[GHC] #11799: Legend for hpc markup colors
by GHC 05 May '17

05 May '17

#11799: Legend for hpc markup colors -------------------------------------+------------------------------------- Reporter: drathier | Owner: Type: feature | Status: new request | Priority: normal | Milestone: Component: Test Suite | Version: 7.10.3 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: -------------------------------------+------------------------------------- The hpc markup colors are not documented anywhere. Coming from an imperative background, red means never executed and green means executed at least once. Here, green means always true, red means always false, yellow means never executed. Adding a one-line legend at the top of the output file(s) describing the colours would help a lot. Suggesting `never executed` `always true` and `always false` on a single line at the top with the same yellow, green and red background as the respective highlighted code would have. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11799> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

[GHC] #11616: Kinds aren't instantiated
by GHC 05 May '17

05 May '17

#11616: Kinds aren't instantiated -------------------------------------+------------------------------------- Reporter: Iceland_jack | Owner: Type: task | Status: new Priority: normal | 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: -------------------------------------+------------------------------------- {{{#!hs {-# LANGUAGE TypeInType, AllowAmbiugousTypes #-} class Whoami a where whoami :: String instance Whoami Int where whoami = "int" instance Whoami Bool where whoami = "[y/n]" instance Whoami Maybe where whoami = "call me maybe" }}} we can write {{{ >>> whoami @Type @Int "int" >>> whoami @Type @Bool "[y/n]" >>> whoami @(Type -> Type) @Maybe "call me maybe" }}} Attempting to specialise `whoami` to `Type` we can't simply write `whoamiType = whoami @Type` as one might expect. To start we define a synonym: {{{ $ ghci -ignore-dot-ghci /tmp/t17q.hs GHCi, version 8.1.20160117: http://www.haskell.org/ghc/ :? for help [1 of 1] Compiling Main ( /tmp/t17q.hs, interpreted ) Ok, modules loaded: Main. *Main> whoami2 = whoami <interactive>:1:1: error: • Ambiguous type variable ‘a0’ arising from a use of ‘whoami’ prevents the constraint ‘(Whoami a0)’ from being solved. Probable fix: use a type annotation to specify what ‘k0’, ‘a0’ should be. These potential instances exist: three instance involving out-of-scope typess (use -fprint-potential-instances to see them all) • When instantiating ‘whoami2’, initially inferred to have this overly-general type: forall k (a :: k). Whoami a => String NB: This instantiation can be caused by the monomorphism restriction. }}} Using the “overly-general” inferred type `whoami :: forall k (a :: k). Whoami a => String` fails: {{{#!hs -- • Could not deduce (Whoami a0) arising from a use of ‘whoami’ -- from the context: Whoami a -- bound by the type signature for: -- whoami2 :: Whoami a => String -- at /tmp/t17q.hs:20:1-44 -- The type variable ‘a0’ is ambiguous -- These potential instances exist: -- three instance involving out-of-scope typess -- (use -fprint-potential-instances to see them all) -- • In the expression: whoami -- In an equation for ‘whoami2’: whoami2 = whoami whoami2 :: forall k (a :: k). Whoami a => String whoami2 = whoami }}} so we need `TypeApplications`: {{{#!hs whoami2 :: forall k (a :: k). Whoami a => String whoami2 = whoami @_ @a }}} and then we can write: {{{#!hs whoamiType :: forall (a :: Type). Whoami a => String whoamiType = whoami @_ @a }}} and we can write as intended: {{{ >>> whoamiType @Int "int" >>> whoamiType @Bool "[y/n]" }}} Is there a reason one cannot write: {{{#!hs whoami2 = whoami whoamiType = whoami @Type }}} ---- As a side note, what is the difference between {{{#!hs -- >>> :set -fprintf-explicit-foralls -- >>> :kind Whoami1 -- Whoami1 :: forall {k}. k -> Constraint -- >>> :kind Whoami2 -- Whoami2 :: forall k. k -> Constraint class Whoami1 a class Whoami2 (a :: k) }}} Is there a snippet of code that highlights differences -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11616> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 8