July 2015 - ghc-devs - Haskell.org

behavior change of the pretty printer
by Kazu Yamamoto 31 Jul '15

31 Jul '15

Hi, Testing a package with GHC head on Travis CI, I got the following error: ---- expected: Just (HTTPDate {hdYear = 1994, hdMonth = 11, hdDay = 15, hdHour = 8, hdMinute = 12, hdSecond = 31, hdWkday = 2}) but got: Just HTTPDate {hdYear = 1994, hdMonth = 11, hdDay = 15, hdHour = 8, hdMinute = 12, hdSecond = 31, hdWkday = 2} ---- That is, parentheses are not printed. Is this an intentional change of GHC head? --Kazu

2 2

Re: [GHC] #10250: API Annotations : add Locations in hsSyn were layout occurs
by Alan & Kim Zimmerman 31 Jul '15

31 Jul '15

Oops, my mistake, Sorry On Fri, Jul 31, 2015 at 12:59 PM, GHC <ghc-devs(a)haskell.org> wrote: > #10250: API Annotations : add Locations in hsSyn were layout occurs > -------------------------------------+------------------------------------- > Reporter: alanz | Owner: alanz > Type: task | Status: closed > Priority: normal | Milestone: 7.12.1 > Component: Compiler | Version: 7.10.1 > Resolution: fixed | Keywords: > | ApiAnnotations > Operating System: Unknown/Multiple | Architecture: > | Unknown/Multiple > Type of failure: None/Unknown | Test Case: > Blocked By: | Blocking: > Related Tickets: | Differential Revisions: Phab:D815 > -------------------------------------+------------------------------------- > > Comment (by thomie): > > Phab:D815 was landed on master already, so I'm confused by your message. > > If you're requesting for this to be merged to 7.10.3: just reopen, set the > right milestone, and set it to status=merge. Otherwise it gets lost. > > -- > Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10250#comment:6> > GHC <http://www.haskell.org/ghc/> > The Glasgow Haskell Compiler >

1 0

RE: [Diffusion] [Build Failed] rGHC24afe6d3aae1: Fix missing files
by Simon Peyton Jones 30 Jul '15

30 Jul '15

Apparently it still doesn't validate, but I don't understand Harbourmaster's failure report. Seems to be something to do with Haddock though. Anyone able to help? Simon | -----Original Message----- | From: noreply(a)phabricator.haskell.org | [mailto:noreply@phabricator.haskell.org] | Sent: 30 July 2015 17:14 | To: Simon Peyton Jones | Subject: [Diffusion] [Build Failed] rGHC24afe6d3aae1: Fix missing files | | Harbormaster failed to build B5072: rGHC24afe6d3aae1: Fix missing files! | | BRANCHES | master | | USERS | simonpj (Author) | GHC - Testsuite (Auditor) | | COMMIT | https://phabricator.haskell.org/rGHC24afe6d3aae1 | | EMAIL PREFERENCES | https://phabricator.haskell.org/settings/panel/emailpreferences/ | | To: simonpj, GHC - Testsuite

2 1

Re: [Haskell-cafe] ANNOUNCE: GHC version 7.10.2
by Ben Gamari 30 Jul '15

30 Jul '15

James M <jmartin(a)eecs.berkeley.edu> writes: > There was talk from an earlier email thread of releasing the Haskell > Platform at the same time as 7.10.2. > > I am referring to the weekly news of 2015/05/11: > https://ghc.haskell.org/trac/ghc/blog/weekly20150511 > > and this email thread: > https://mail.haskell.org/pipermail/ghc-devs/2015-May/008911.html > > > Was this plan abandoned? Or was there something unexpected that is delaying > it? > Not at all! There ended up being a bit of timing skew between the Platform and GHC releases but an approximately concurrent release is still the plan. The Platform folks are hard at work as we speak, in the final stages of pushing out their release. I expect it will be announced shortly. Unfortunately due to the late notice of the text issue, the initial release will ship with text-1.2.1.1, which is still affected by the rewrite issue. That being said, the impact of this issue is mostly on compile time, in most cases you shouldn't see much impact on runtime performance. Moreover, they will be working to push out another release with text-1.2.1.3 in the coming days. Cheers, - Ben

1 0

Re: [Haskell-cafe] ANNOUNCE: GHC version 7.10.2
by Brandon Allbery 30 Jul '15

30 Jul '15

On Thu, Jul 30, 2015 at 12:52 PM, James M <jmartin(a)eecs.berkeley.edu> wrote: > There was talk from an earlier email thread of releasing the Haskell > Platform at the same time as 7.10.2. I think the right place to ask this is libraries(a)haskell.org. I would imagine they're in final testing and/or getting release packages in place. (Note as previously mentioned that "text" was just updated, and the Platform needs that; this presumably means they had to start over making release packages.) -- brandon s allbery kf8nh sine nomine associates allbery.b(a)gmail.com ballbery(a)sinenomine.net unix, openafs, kerberos, infrastructure, xmonad http://sinenomine.net

1 0

Parallel library
by Simon Peyton Jones 30 Jul '15

30 Jul '15

Simon Are you the maintainer of the 'parallel' library? The core libraries wiki page says yes: https://wiki.haskell.org/Library_submissions#The_Core_Libraries But parallel.cabal says "Maintainer: libraries(a)haskell.org<mailto:libraries@haskell.org>", which I thought we'd gotten rid of entirely. I ask because some should fix this: libraries/parallel/Control/Parallel/Strategies.hs:513:2: warning: Rule "parList/rseq" may never fire because 'rseq' might inline first Probable fix: add an INLINE[n] or NOINLINE[n] pragma for 'rseq' We need a pragma on 'rseq'. Simon

5 10

RE: [Diffusion] [Committed] rGHC24afe6d3aae1: Fix missing files
by Simon Peyton Jones 30 Jul '15

30 Jul '15

Sorry for all the noise recently from me. I _think_ we should be back to clean validation now Simon | -----Original Message----- | From: noreply(a)phabricator.haskell.org | [mailto:noreply@phabricator.haskell.org] | Sent: 30 July 2015 15:53 | To: Simon Peyton Jones | Subject: [Diffusion] [Committed] rGHC24afe6d3aae1: Fix missing files | | simonpj committed rGHC24afe6d3aae1: Fix missing files (authored by | simonpj). | | Fix missing files | | These two testsuite files were correct in my build tree but not in my | source tree; apologies. | | | BRANCHES | master | | AFFECTED FILES | /testsuite/tests/simplCore/should_compile/T7287.stderr | /testsuite/tests/stranal/should_compile/T10694.stdout | | USERS | simonpj (Author) | GHC - Testsuite (Auditor) | | COMMIT | https://phabricator.haskell.org/rGHC24afe6d3aae1 | | EMAIL PREFERENCES | https://phabricator.haskell.org/settings/panel/emailpreferences/ | | To: simonpj, GHC - Testsuite | Cc: thomie

1 0

ANNOUNCE: GHC version 7.10.2
by Ben Gamari 30 Jul '15

30 Jul '15

===================================================================== The (Interactive) Glasgow Haskell Compiler -- version 7.10.2 ===================================================================== The GHC Team is pleased to announce a new bug-fix release of GHC, 7.10.2. There have been a number of significant bug fixes since the 7.10.1 (with over 70 defects fixed). These include fixes affecting type-checker correctness, runtime stability, and compiler performance. For this reason we highly recommend that users of 7.10.1 upgrade quickly. Note that due to changes made in this release to GHC's optimizer, previously fragile-but-working rewrite rules may fail to fire. One example of this was found in the widely-used `text` package late in the release cycle, which manifested in long compilation times and poor code generation for `Text` literals (see [Trac #10528]). Users are advised to use text >=1.2.1.2 which includes more robust rewrite rules. The full release notes including a complete listing of the changes in this release can be found here, https://downloads.haskell.org/~ghc/7.10.2/docs/html/users_guide/release-7-1… [Trac #10528]: https://ghc.haskell.org/trac/ghc/ticket/10528 How to get it ~~~~~~~~~~~~~ Both binary and source tarballs of GHC itself are available on the release download page, https://www.haskell.org/ghc/download_ghc_7_10_2 For a more smoother, better integrated experience users are encouraged to find a Haskell distribution. While these alternatives may not yet include 7.10.2, they offer tools and libraries to quickly get users up and running, as well as potentially better integration with the host operating system and package manager. See, https://www.haskell.org/downloads for more details. Background ~~~~~~~~~~ Haskell is a standard lazy functional programming language. GHC is a state-of-the-art programming suite for Haskell. Included is an optimising compiler generating good code for a variety of platforms, together with an interactive system for convenient, quick development. The distribution includes space and time profiling facilities, a large collection of libraries, and support for various language extensions, including concurrency, exceptions, and foreign language interfaces (C, whatever). GHC is distributed under a BSD-style open source license. A wide variety of Haskell related resources (tutorials, libraries, specifications, documentation, compilers, interpreters, references, contact information, links to research groups) are available from the Haskell home page (see below), https://www.haskell.org/ Supported Platforms ~~~~~~~~~~~~~~~~~~~ The list of platforms supported by GHC and the people responsible for them can be found here, https://ghc.haskell.org/trac/ghc/wiki/Platforms https://ghc.haskell.org/trac/ghc/wiki/CodeOwners Ports to other platforms are possible with varying degrees of difficulty. The Building Guide describes how to go about porting to a new platform: https://ghc.haskell.org/trac/ghc/wiki/Building Developers ~~~~~~~~~~ If you enjoy using GHC, you will likely also enjoy contributing to it! We are always looking for new contributors. Instructions on accessing our source code repository, and getting started with hacking on GHC, are available from the GHC developer site, https://ghc.haskell.org/trac/ghc/ Mailing lists ~~~~~~~~~~~~~ We run mailing lists for GHC users and bug reports; to subscribe, use the web interfaces at https://www.haskell.org/mailman/listinfo/glasgow-haskell-users https://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs There are several other haskell and ghc-related mailing lists on www.haskell.org; for the full list, see https://www.haskell.org/mailman/listinfo/ Some GHC developers hang out on #haskell on IRC, too: https://www.haskell.org/haskellwiki/IRC_channel Please report bugs using our bug tracking system. Instructions on reporting bugs can be found here: https://www.haskell.org/ghc/reportabug Hashes & Signatures ~~~~~~~~~~~~~~~~~ In https://downloads.haskell.org/~ghc/7.10.2/ you will find a signed copy of the SHA256 hashes for the tarballs, using my GPG key, Benjamin Gamari <ben(a)well-typed.com> Fingerprint: FFEB 7CE8 1E16 A36B 3E2D ED6F 2DE0 4D4E 97DB 64AD

3 6

Question about inliner behaviour with a small function
by Michael Walker 29 Jul '15

29 Jul '15

Hello, I managed to shrink a bizarre memory issue down to this (probably minimal) example: module Main where f :: [Int] -> Int f xs = length is where is = [ i | (_, i) <- pairs ys ] :: [(Int,Int)] ys = zip [0..] xs :: [(Int, Int)] {-# INLINE pairs #-} pairs xs = zip xs $ tail xs main :: IO () main = print $ f xs where xs = replicate 1000000 0 With the INLINE pragma, this allocates 264,049,584 bytes (compiled with `ghc -O2 -prof -fprof-auto inline.hs), without the pragma it allocates 336,049,512 bytes. Dropping the `main` definition (and renaming the module to "Foo") and examining the core, the key difference seems to be in how the list comprehension is compiled. With the INLINE pragma: Rec { Foo.f_go [Occ=LoopBreaker] :: [(Int, Int)] -> [(Int, Int)] -> [(Int, Int)] [GblId, Arity=2, Caf=NoCafRefs, Str=DmdType <S,1*U><L,1*U>] Foo.f_go = \ (ds_a14a :: [(Int, Int)]) (_ys_a14b :: [(Int, Int)]) -> case ds_a14a of _ [Occ=Dead] { [] -> GHC.Types.[] @ (Int, Int); : ipv_a14g ipv1_a14h -> case _ys_a14b of _ [Occ=Dead] { [] -> GHC.Types.[] @ (Int, Int); : ipv2_a14n ipv3_a14o -> GHC.Types.: @ (Int, Int) ipv2_a14n (Foo.f_go ipv1_a14h ipv3_a14o) } } end Rec } And without: Rec { Foo.f_go [Occ=LoopBreaker] :: [((Int, Int), (Int, Int))] -> [(Int, Int)] [GblId, Arity=1, Caf=NoCafRefs, Str=DmdType <S,1*U>] Foo.f_go = \ (ds_a13o :: [((Int, Int), (Int, Int))]) -> case ds_a13o of _ [Occ=Dead] { [] -> GHC.Types.[] @ (Int, Int); : y_a13t ys_a13u -> case y_a13t of _ [Occ=Dead] { (ds1_d12V, i_an2) -> GHC.Types.: @ (Int, Int) i_an2 (Foo.f_go ys_a13u) } } end Rec } Full core is available at https://gist.github.com/barrucadu/a59df62cd16074559e35 In the no-INLINE case, the list comprehension is compiled much like I would expect, it's walking down a zipped list of pairs and producing the result. The INLINE case is rather different, it looks like the `zip` has been inlined and deforestation has happened. This explains the difference in memory usage, a whole intermediary list has been skipped! I assume that GHC's analysis is determining `pairs` is too expensive to inline early enough to allow the further optimisation without the pragma, but the pragma forces it to happen earlier by marking it as really cheap. But `pairs` is a really small definition, syntactically. Why does the analysis consider it expensive? It is simply because it uses its argument multiple times in its body? And if it's not some sort of cost analysis, what's really going on? Thank for your time. -- Michael Walker (http://www.barrucadu.co.uk)

2 1

RE: warnings
by Simon Peyton Jones 28 Jul '15

28 Jul '15

Hopefully done now. Sorry about this -- really busy day S From: Simon Peyton Jones Sent: 28 July 2015 11:01 To: 'ghc-devs' Subject: RE: warnings Validating now, but have to go to London for a few hrs. Meanwhile use -Wwarn! From: Simon Peyton Jones Sent: 27 July 2015 17:08 To: ghc-devs Subject: warnings Yurgh. I committed a patch but my validate had -Wwarn by mistake. So there are warnings in base (and binary actually). I will sort them out tomorrow. Could revert and re-apply but doesn't seem worth it. Sorry Simon

2 2