- ghc-tickets - Haskell.org

Re: [GHC] #7953: Segfault on x86 with -O2
by GHC 14 Aug '13

14 Aug '13

#7953: Segfault on x86 with -O2 -------------------------------------+---------------------------- Reporter: igloo | Owner: simonmar Type: bug | Status: new Priority: highest | Milestone: 7.8.1 Component: Compiler | Version: 7.7 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: x86 Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: -------------------------------------+---------------------------- Comment (by kazu-yamamoto): Austin guessed that 3e598fe0b01f4692f50c377ac91010040227e7ce fixed this bug. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/7953#comment:13> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #5902: Cannot tell from an exception handler whether the exception was asynchronous
by GHC 13 Aug '13

13 Aug '13

#5902: Cannot tell from an exception handler whether the exception was asynchronous -------------------------------------+------------------------------------ Reporter: simonmar | Owner: simonmar Type: bug | Status: new Priority: high | Milestone: 7.8.1 Component: Compiler | Version: 7.4.1 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Unknown/Multiple Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: -------------------------------------+------------------------------------ Comment (by exbb2): Replying to [comment:21 simonmar]: > Just never catch `SomeException` unless you intend to re-throw it. The code in my case quickly grew into monstrosity such as {{{ bracket openFile doDangerousStuff close `catches` [\Async -> throw ,\Some -> logFailureAndResume] }}} Even that doesn't work for Timeout, unless we resort to string comparison. > If you're handling an exception, you always know which exception(s) you're handling No. All arguments and values in function can throw any exception whatsoever, there can be no knowing of what exceptions you'll receive. e.g. what if the next version of base introduces a hard-to-track division by zero? We shouldn't ever care about what exception we got. If it had arisen from code enclosed by the handler, — there can be no doubt as to its intended recipient — and no danger in catching it. Asynchronous exceptions have only dimmest relation to code being executed or none at all, there can be no modularity when every exception handler is concerned with whole state of the universe. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/5902#comment:22> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #7953: Segfault on x86 with -O2
by GHC 13 Aug '13

13 Aug '13

#7953: Segfault on x86 with -O2 -------------------------------------+---------------------------- Reporter: igloo | Owner: simonmar Type: bug | Status: new Priority: highest | Milestone: 7.8.1 Component: Compiler | Version: 7.7 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: x86 Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: -------------------------------------+---------------------------- Comment (by kazu-yamamoto): Oh, it's magical. I cannot reproduce this on Linux. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/7953#comment:12> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #5902: Cannot tell from an exception handler whether the exception was asynchronous
by GHC 13 Aug '13

13 Aug '13

#5902: Cannot tell from an exception handler whether the exception was asynchronous -------------------------------------+------------------------------------ Reporter: simonmar | Owner: simonmar Type: bug | Status: new Priority: high | Milestone: 7.8.1 Component: Compiler | Version: 7.4.1 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Unknown/Multiple Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: -------------------------------------+------------------------------------ Comment (by simonmar): Just never catch `SomeException` unless you intend to re-throw it. If you're handling an exception, you always know which exception(s) you're handling, so you can use type-specific handlers. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/5902#comment:21> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #3693: Show stack traces
by GHC 13 Aug '13

13 Aug '13

#3693: Show stack traces -------------------------------------+------------------------------------ Reporter: jpet | Owner: Type: feature request | Status: new Priority: normal | Milestone: 7.6.2 Component: Runtime System | Version: 6.10.4 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Unknown/Multiple Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: -------------------------------------+------------------------------------ Comment (by scpmw): Great, thanks! I will start with splitting things up then. Unless there's other preferences I will submit the patches into a [https://github.com/scpmw/ghc/commits/profiling-import GitHub branch]. That way, we don't have to spam this ticket with individual patches - and can document all interesting design choices in the commit message along the way. Would be good to get feedback on the initial [https://github.com/scpmw/ghc/commit/d9df7b8b0fc46b344b6060800955e8f50110dbbb source notes patch], as that's probably the piece of code with the highest chance of screwing things up in subtle ways. I am currently trying again to hunt down any further Core optimization behaviour changes due to source notes, but this version already works very well in my experience. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/3693#comment:46> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #5902: Cannot tell from an exception handler whether the exception was asynchronous
by GHC 13 Aug '13

13 Aug '13

#5902: Cannot tell from an exception handler whether the exception was asynchronous -------------------------------------+------------------------------------ Reporter: simonmar | Owner: simonmar Type: bug | Status: new Priority: high | Milestone: 7.8.1 Component: Compiler | Version: 7.4.1 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Unknown/Multiple Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: -------------------------------------+------------------------------------ Comment (by exbb2): Replying to [comment:19 simonmar]: > > Every innocuous catch or try anywhere at any time in any library can make a thread unintentionally uncancelable > I'm not sure what you mean here. The issue in this ticket only arises when using `unsafePerformIO`. {{{ tryNTimes n | n <= 0 = return () | otherwise = forever (return ()) `catch` \(_::SomeException) -> tryNTimes (n - 1) }}} or {{{ try something :: IO (Either SomeException x) }}} Catching SomeException is sometimes reasonable, but it's unreasonable not to provide ability to discriminate against irrelevant asynchronous exceptions which do not directly follow from handled code. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/5902#comment:20> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #7953: Segfault on x86 with -O2
by GHC 13 Aug '13

13 Aug '13

#7953: Segfault on x86 with -O2 -------------------------------------+---------------------------- Reporter: igloo | Owner: simonmar Type: bug | Status: new Priority: highest | Milestone: 7.8.1 Component: Compiler | Version: 7.7 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: x86 Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: -------------------------------------+---------------------------- Comment (by simonmar): I tried and failed to reproduce this today, with a freshly built GHC on a 32-bit x86 Amazon EC2 instance running Ubuntu 12.04. {{{ ubuntu@ip-10-245-94-31:~/scratch$ uname -a Linux ip-10-245-94-31 3.2.0-40-virtual #64-Ubuntu SMP Mon Mar 25 22:07:32 UTC 2013 i686 i686 i386 GNU/Linux ubuntu@ip-10-245-94-31:~/scratch$ ~/ghc/inplace/bin/ghc-stage2 --version The Glorious Glasgow Haskell Compilation System, version 7.7.20130813 ubuntu@ip-10-245-94-31:~/scratch$ ~/ghc/inplace/bin/ghc-stage2 -O2 7.hs -fforce-recomp [1 of 1] Compiling Main ( 7.hs, 7.o ) Linking 7 ... ubuntu@ip-10-245-94-31:~/scratch$ ./7 Winner (E {_key = 1, prio = 0.7169761920394114} ... After atMost }}} Can anyone else still reproduce it on x86/Linux? -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/7953#comment:11> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #5144: Pattern synonyms
by GHC 13 Aug '13

13 Aug '13

#5144: Pattern synonyms -------------------------------------+------------------------------------ Reporter: simonpj | Owner: cactus Type: feature request | Status: new Priority: normal | Milestone: _|_ Component: Compiler | Version: Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Unknown/Multiple Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: -------------------------------------+------------------------------------ Comment (by cactus): I've implemented infix notation as well, so now you can write this: {{{ {-# LANGUAGE ViewPatterns, PatternSynonyms #-} import qualified Data.Sequence as Seq pattern Empty = (Seq.viewl -> Seq.EmptyL) pattern x :< xs = (Seq.viewl -> x Seq.:< xs) pattern xs :> x = (Seq.viewr -> xs Seq.:> x) zipZag :: Seq.Seq a -> Seq.Seq b -> Seq.Seq (a, b) zipZag (x :< xs) (ys :> y) = (x, y) Seq.<| zipZag xs ys zipZag _ _ = Seq.empty }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/5144#comment:18> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #7978: Profiling broken
by GHC 13 Aug '13

13 Aug '13

#7978: Profiling broken --------------------------------------------+------------------------------ Reporter: gmainland | Owner: Type: bug | Status: closed Priority: highest | Milestone: 7.8.1 Component: Compiler | Version: 7.7 Resolution: fixed | Keywords: Operating System: Unknown/Multiple | Architecture: Type of failure: GHC doesn't work at all | Unknown/Multiple Test Case: | Difficulty: Unknown Blocking: | Blocked By: | Related Tickets: --------------------------------------------+------------------------------ Changes (by simonmar): * status: new => closed * resolution: => fixed Comment: Fixed, thanks to everyone who helped narrow this down, it took me 2 minutes to find and fix the problem. a27895b359db20197d9ec72b03c4e304a20585d8 -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/7978#comment:16> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #5902: Cannot tell from an exception handler whether the exception was asynchronous
by GHC 13 Aug '13

13 Aug '13

#5902: Cannot tell from an exception handler whether the exception was asynchronous -------------------------------------+------------------------------------ Reporter: simonmar | Owner: simonmar Type: bug | Status: new Priority: high | Milestone: 7.8.1 Component: Compiler | Version: 7.4.1 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Unknown/Multiple Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: -------------------------------------+------------------------------------ Comment (by simonmar): @exbb2: we already have `SomeAsyncException`, and the asynchronous exceptions are children of this in the exception hierarchy. The exception hierarchy needs a single root, so `SomeException` remains. > Every innocuous catch or try anywhere at any time in any library can make a thread unintentionally uncancelable I'm not sure what you mean here. The issue in this ticket only arises when using `unsafePerformIO`. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/5902#comment:19> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0