June 2015 - ghc-tickets

[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] #9940: GHCi crashes when I hold down control-c for a short time, and then hold down any key that would normally result in typing.
by GHC 16 Nov '16

16 Nov '16

#9940: GHCi crashes when I hold down control-c for a short time, and then hold down any key that would normally result in typing. -------------------------------------------+------------------------------- Reporter: Dingus | Owner: Type: bug | Status: new Priority: low | Milestone: Component: Compiler | Version: 7.6.3 Keywords: crash, control-c | Operating System: Windows Architecture: x86_64 (amd64) | Type of failure: GHCi crash Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: | -------------------------------------------+------------------------------- This is printed: (without the quotes at the beginning or end) " ghc.exe: panic! (the 'impossible' happened) (GHC version 7.6.3 for i386-unknown-mingw32): thread blocked indefinitely in an MVar operation Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug " I am able to reproduce it by: 1. starting ghci (with no arguments, or a file as an argument) 2. holding down contol-c for 2 seconds 3. holding down spacebar/a letter key for 2 seconds The timing varies. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9940> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

[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] #10221: LLVM does not work on OSX
by GHC 31 Oct '16

31 Oct '16

#10221: LLVM does not work on OSX -------------------------------------------+---------------------------- Reporter: yongqli | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.10.1 Keywords: | Operating System: MacOS X Architecture: Unknown/Multiple | Type of failure: Other Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: | -------------------------------------------+---------------------------- I can't seem to be able to use the LLVM backend on OSX. I get the error: {{{ error: unknown directive .macosx_version_min 10, 0 ^ <no location info>: Error running clang! you need clang installed to use theLLVM backend (or GHC tried to execute clang incorrectly) <no location info>: ghc: phase `Clang (Assembler)' failed (exitcode = 1) }}} This is on GHC 7.10.1 {{{ % clang --version Apple LLVM version 6.0 (clang-600.0.57) (based on LLVM 3.5svn) Target: x86_64-apple-darwin14.1.0 Thread model: posix yongqli@auxiliary ~/secrets/projects/stocvol % opt --version LLVM (http://llvm.org/) LLVM version 3.5.1 Optimized build with assertions. Built Jan 15 2015 (18:24:46). Default target: x86_64-apple-darwin14.1.0 Host CPU: core-avx2 }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10221> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 13

[GHC] #9858: Typeable instance for datatype and its promoted constructor is the same
by GHC 26 Oct '16

26 Oct '16

#9858: Typeable instance for datatype and its promoted constructor is the same -------------------------------------+------------------------------------- Reporter: dreixel | Owner: Type: bug | Status: new Priority: normal | Milestone: 7.10.1 Component: Compiler | Version: 7.9 Keywords: | Operating System: Architecture: Unknown/Multiple | Unknown/Multiple Difficulty: Easy (less than 1 | Type of failure: hour) | None/Unknown Blocked By: | Test Case: Related Tickets: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- {{{ {-# LANGUAGE DataKinds #-} {-# LANGUAGE AutoDeriveTypeable #-} import Data.Typeable data A = A main = print $ typeRep (Proxy :: Proxy A) == typeRep (Proxy :: Proxy 'A) }}} This returns `True`, but it should return `False`. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9858> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 119

[GHC] #9494: Probable data corruption with GHCi 7.8.* and Zlib
by GHC 15 Oct '16

15 Oct '16

#9494: Probable data corruption with GHCi 7.8.* and Zlib -------------------------------------+------------------------------------- Reporter: nominolo | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: GHCi | Version: 7.8.3 Keywords: | Operating System: Architecture: Unknown/Multiple | Unknown/Multiple Difficulty: Unknown | Type of failure: Incorrect Blocked By: | result at runtime Related Tickets: | Test Case: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- The following program causes Zlib data corruption errors when run from inside GHCi. It launches two threads which then concurrently read a file, compress it, and immediately decompress it. You need libraries `zlib`, `SHA`, and `async`. {{{ module Main where import qualified Codec.Compression.Zlib as Zlib import qualified Data.ByteString.Lazy as BL import qualified Data.ByteString.Internal as BI import Control.Exception (bracket) import Control.Concurrent import Control.Monad import Control.Exception ( evaluate) import Data.Digest.Pure.SHA ( sha1) -- from the 'SHA' package import Control.Concurrent.Async ( mapConcurrently) import System.Mem ( performGC ) import Debug.Trace test :: Int -> IO String test _ = do tid <- myThreadId -- testdata is: dd if=/dev/urandom of=/tmp/testdata bs=100k count=100 -- Could also be replaced by: (BL.take (10^7) "/dev/urandom") dat <- BL.readFile "/tmp/testdata" let cbuf = Zlib.compress $ traceChunks tid $ dat s <- evaluate $ sha1 $ Zlib.decompress $ cbuf return $ show s where -- We used this to check whether buffers were reused by different threads, but that -- doesn't seem to be the case. Removing the call to traceChunks, however, makes it -- harder to reproduce possibly because of scheduler effects. In a much larger program -- it could be reproduced more easily without the trace, but in this small example -- tracing seems to cause the right amount of nondeterminism. traceChunks tid bs = BL.fromChunks $ zipWith (\n x -> trace (show tid ++ ":" ++ showBS x) x) [1..] $ BL.toChunks bs showBS (BI.PS ptr off len) = show ptr main = do r <- withGCThread $ mapConcurrently (test) ([1..2] :: [Int]) putStrLn $ show $ r where -- Regularly forcing the GC makes the test-case more reproducible. withGCThread io = bracket (forkIO $ forever $ performGC >> threadDelay 1000) killThread (const io) }}} The output should be something like: {{{ ... ThreadId 51:0x00000001091ee010 ThreadId 49:0x00000001091a7010 ... ThreadId 49:0x000000010986f010 zlib-test-case.hs: user error (Codec.Compression.Zlib: incorrect data check) }}} You'll get different Zlib errors, depending on where it detects the inconsistency. Sometimes Zlib doesn't throw an error, but the checksums are different. So far we've only been able to reproduce this using GHCi 7.8.3 on both Linux (NixOS) and Mac. We haven't been able to trigger it with a compiled executable, nor with GHCi 7.6.3. It '''was''' reproducable with HEAD from Jan 30 (I had that lying around somewhere). -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9494> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 9

[GHC] #9685: GHC fails to build with mingw32-make on Windows
by GHC 15 Oct '16

15 Oct '16

#9685: GHC fails to build with mingw32-make on Windows -------------------------------------+------------------------------------- Reporter: gintas | Owner: gintas Type: feature request | Status: new Priority: normal | Milestone: Component: Build System | Version: 7.9 Keywords: | Operating System: Windows Architecture: Unknown/Multiple | Type of failure: Difficulty: Unknown | None/Unknown Blocked By: | Test Case: Related Tickets: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- It would be nice to be able to build ghc on Windows with mingw32-make which is bundled with the prebuilt gcc package that we are using for the ghc build. mingw32-make might be faster (due to less overhead associated with POSIX compatibility, forks etc.) and more reliable - msys2's make is a bit glitchy because it needs to fork off a lot of processes and sometimes runs into issues with fork(), which should not be a problem for mingw32-make. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9685> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[GHC] #10437: RunHaskell error in HSExtra on Win64
by GHC 15 Oct '16

15 Oct '16

#10437: RunHaskell error in HSExtra on Win64 -------------------------------------+------------------------------------- Reporter: | Owner: ScottSedgwick | Status: new Type: bug | Milestone: Priority: normal | Version: 7.8.3 Component: Compiler | Operating System: Windows Keywords: runhaskell | Type of failure: Compile-time Architecture: x86_64 | crash (amd64) | Blocked By: Test Case: | Related Tickets: Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- I ran the following command, and got this error: {{{ $ runhaskell shakefile.hs clean shakefile.hs: C:\Users\Scott Sedgwick\AppData\Roaming\cabal\x86_64 -windows-ghc-7.8.3\extra-1.2\HSextra-1.2.o: Not x86_64 PEi386 shakefile.hs: shakefile.hs: panic! (the 'impossible' happened) (GHC version 7.8.3 for x86_64-unknown-mingw32): loadObj "C:\\Users\\Scott Sedgwick\\AppData\\Roaming\\cabal\\x86_64-windows- ghc-7.8.3\\extra-1.2\\HSextra-1.2.o": failed Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug }}} The shakefile in question looks like this: {{{ import Development.Shake import Development.Shake.FilePath --import System.Info srcFiles = [ "source/main.hs", "source/DBDataStructures.hs", "source/FormatDBData.hs", "source/GetDBData.hs", "source/ParseDBData.hs"] exeFile = "dist/build/mssql-er/mssql-er.exe" depFiles = [ "dependencies/erd/erd.exe", "dependencies/dot/Pathplan.dll", "dependencies/dot/config6", "dependencies/dot/gvc.dll", "dependencies/dot/iconv.dll", "dependencies/dot/libfontconfig-1.dll", "dependencies/dot/libgmodule-2.0-0.dll", "dependencies/dot/libpango-1.0-0.dll", "dependencies/dot/libpangowin32-1.0-0.dll", "dependencies/dot/ltdl.dll", "dependencies/dot/cdt.dll", "dependencies/dot/dot.exe", "dependencies/dot/gvplugin_dot_layout.dll", "dependencies/dot/libcairo-2.dll", "dependencies/dot/libfreetype-6.dll", "dependencies/dot/libgobject-2.0-0.dll", "dependencies/dot/libpangocairo-1.0-0.dll", "dependencies/dot/libpng14-14.dll", "dependencies/dot/zlib1.dll", "dependencies/dot/cgraph.dll", "dependencies/dot/freetype6.dll", "dependencies/dot/gvplugin_pango.dll", "dependencies/dot/libexpat.dll", "dependencies/dot/libglib-2.0-0.dll", "dependencies/dot/libgthread-2.0-0.dll", "dependencies/dot/libpangoft2-1.0-0.dll", "dependencies/dot/libxml2.dll"] copyDep :: FilePath -> Action() copyDep src = do let dst = "dist" </> (dropDirectory1 (dropDirectory1 src)) copyFile' src dst main :: IO() main = shakeArgs shakeOptions $ do want ["deploy"] "clean" ~> do cmd "cabal" "clean" "dist/setup-config" *> \_ -> do need ["mssql-er.cabal"] cmd "cabal" "configure" "--enable-tests" "configure" ~> do need ["dist/setup-config"] exeFile *> \_ -> do need (["configure"] ++ srcFiles) cmd "cabal" "build" "build" ~> do need [exeFile, "test"] "test" ~> do need srcFiles cmd "cabal" "test" "deploy" ~> do need ["build"] mapM_ copyDep depFiles }}} Strangely, if I compile that shakefile using ghc, then run it, it works correctly. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10437> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

[GHC] #10122: PolyKinds: inferred type not as polymorphic as possible
by GHC 15 Oct '16

15 Oct '16

#10122: PolyKinds: inferred type not as polymorphic as possible -------------------------------------+------------------------------------- Reporter: thomie | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.10.1-rc2 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Blocked By: Test Case: | Related Tickets: Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- In the user's guide on [https://downloads.haskell.org/~ghc/latest/docs/html/users_guide/kind- polymorphism.html kind polymorphism] the following example is presented: {{{#!hs f1 :: (forall a m. m a -> Int) -> Int -- f1 :: forall (k:BOX). -- (forall (a:k) (m:k->*). m a -> Int) -- -> Int }}} "Here in f1 there is no kind annotation mentioning the polymorphic kind variable, so k is generalised at the top level of the signature for f1, making the signature for f1 is as polymorphic as possible." When I ask GHCi for the type of `f1`, it is however not as polymorphic as possible: {{{ > :t f1 f1 :: (forall a (m :: * -> *). m a -> Int) -> Int }}} Trying to compile the following program: {{{#!hs {-# LANGUAGE PolyKinds #-} {-# LANGUAGE ExplicitForAll #-} {-# LANGUAGE Rank2Types #-} module PolyFail where f :: (forall a m. m a -> Int) -> Int f g = g (Just 3) }}} Results in this error: {{{ [1 of 1] Compiling PolyFail ( PolyFail.hs, PolyFail.o ) PolyFail.hs:8:10: Kind incompatibility when matching types: m0 :: k -> * Maybe :: * -> * Expected type: m0 a0 Actual type: Maybe a1 Relevant bindings include g :: forall (a :: k) (m :: k -> *). m a -> Int (bound at PolyFail.hs:8:3) f :: (forall (a :: k) (m :: k -> *). m a -> Int) -> Int (bound at PolyFail.hs:8:1) In the first argument of ‘g’, namely ‘(Just 3)’ In the expression: g (Just 3) -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10122> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 7

[GHC] #9717: More lazy orphan module loading
by GHC 14 Oct '16

14 Oct '16

#9717: More lazy orphan module loading -------------------------------------+------------------------------------- Reporter: ezyang | Owner: ezyang Type: bug | Status: new Priority: normal | Milestone: 7.10.1 Component: Compiler (Type | Version: 7.9 checker) | Operating System: Keywords: backpack | Unknown/Multiple Architecture: Unknown/Multiple | Type of failure: Difficulty: Unknown | None/Unknown Blocked By: | Test Case: Related Tickets: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- Currently, when we import a module, we eagerly load all of the orphan modules transitively reachable from it (TcRnDriver:tcRnImports). This is a bit of bummer: as module import hierarchies get deeper and deeper, we'll collect more and more orphan instances and load more and more modules eagerly. The idea is to try to arrange for an orphan module not to be loaded, unless we think that it might have a relevant instance. Instead of just recording a list of orphan modules transitively reachable from a module, we will also record a digest of what type classes and types the orphan provides instances for. (This will be similar to the data we record for `ifInstTys` in `IfaceClsInst`.) Now, when we do any operation involving instances, we check and see if there are also orphans which could provide relevant instances (based on the digest) and load those. If an instance could never have been used, we avoid loading its interface file entirely. This doesn't appear to help too much for type families, where we have to load all of the instances anyway in order to check for consistency. A choice to make: Should we record just the class, or also the types involved? It would be best if we could quickly conclude that there are no more instances to load, but this may be difficult if matching against types as well. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9717> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4