September 2009 - Glasgow-haskell-users

Re: runghc failing sometimes on Linux/ppc?
by Christopher Lewis 20 Sep '09

20 Sep '09

> On Sun, May 03, 2009 at 03:02:50PM +1000, Jens Petersen wrote: > > 2009/5/2 Jens Petersen <petersen at haskell.org>: > > > So I tried doing an unregisterised ppc build with: > > > > > > echo "GhcUnregisterised=YES" >> mk/build.mk > > > echo "GhcWithNativeCodeGen=NO" >> mk/build.mk > > > echo "SplitObjs=NO" >> mk/build.mk > > > > Are these the right build config settings? > > Yes. > > > > but this failed with: > > : > > > Linking dist-stage2/build/ghc/ghc ... > > : > > > (.text+0x20): relocation truncated to fit: R_PPC_REL24 against symbol > > > `__libc_start_main@@GLIBC_2.0' defined in .glink section in > > > /usr/lib/gcc/ppc64-redhat-linux/4.4.0/../../../../lib/crt1.o > > > /usr/lib/gcc/ppc64-redhat-linux/4.4.0/crtbegin.o:(.fini+0x0): > > > relocation truncated to fit: R_PPC_REL24 against `.text' > > Does using -mlongcall help? e.g.: > > SRC_CC_OPTS += -mlongcall > SRC_HC_OPTS += -optc-mlongcall > > (you can probably get away with something rather more specific, but > hopefully this will work). I haven't seen a message confirming if the suggested solution worked, but I am experiencing a similar problem---failure to link GHC due to relocation truncation errors---on a different environment (Ubuntu, PowerPC 32 bit). I have tried the above workaround, but it did not solve my problem. Furthermore, based on this (http://osdir.com/ml/bug-binutils-gnu/2009-08/msg00068.html) binutils thread (sorry, there is probably a better source to reference), it appears this problem was not fixed until very recently. Any additional information or suggested workarounds would be much appreciated. {c}

1 0

Re: How to move code from GHC to libraries?
by Simon Marlow 17 Sep '09

17 Sep '09

On 17/09/2009 04:43, Norman Ramsey wrote: > > > While we're on this topic, I've had several requests to pull HOOPL > > > (the new dataflow optimization code) out as a library so that we can > > > make it into a Hackage package. I'd like to do this, but the > > > situation is complicated because HOOPL depends on UniqFM, which is > > > GHC-specific. I thought that there was a similar implementation out > > > in the wild (based on Okasaki-Gill rather than Adams), but I can't > > > find it on hackage. > > > > Isn't Data.IntMap ~= UniqFM? > > Yes, thanks --- I couldn't remember the module name, and search did > not help. > > > > Is there an established procedure for moving a part of GHC into a > > > library without breaking everybody's build? ... > > > > The architecture of our library setup is here: > > > > http://hackage.haskell.org/trac/ghc/wiki/Commentary/Libraries > > > You'd be adding a new zero-boot package... > > OK, I would like to propose that I put Data.IntMap into a new > zero-boot package, and that I replace the existing UniqFM with a > compatibility layer over Data.IntMap, with the eventual goal of > expunging UniqFM from the GHC sources. > > Any comments or objections? You can assume that the bootstrapping compiler already has Data.IntMap, no need to put it in a 0-boot package. So in principle using Data.IntMap is a good thing, I wholeheartedly support using the standard libraries instead of in-house duplicates where possible. The only potential problem is performance: we rely heavily on UniqFM in GHC, so we'd have to make sure that any replacement is at least as fast. The only time to measure it is when making the changeover, when you can measure the change in isolation. Cheers, Simon

1 0

C++ libraries and GHCI
by Paolo Losi 14 Sep '09

14 Sep '09

Hi all, I'm working on a tentative implementation of haskell binding to libsvm [1]: http://hackage.haskell.org/package/HSvm The package includes libsvm C++ sources. While it works perfectly with ghc --make (try to install the package and compile http://bitbucket.org/pao/hsvm/raw/6cf7ca91f1e5/test/test.hs) it fails with ghci: (hsvm)(env)paolo@moltrasio:~/trash$ ghci test.hs GHCi, version 6.10.4: http://www.haskell.org/ghc/ :? for help Loading package ghc-prim ... linking ... done. Loading package integer ... linking ... done. Loading package base ... linking ... done. Ok, modules loaded: Main. Prelude Main> main Loading package syb ... linking ... done. Loading package array-0.2.0.0 ... linking ... done. Loading package containers-0.2.0.1 ... linking ... done. Loading package HSvm-0.1.0.2.89 ... linking ... <interactive>: /home/paolo/trash/hsvm/lib/HSvm-0.1.0.2.89/ghc-6.10.4/HSHSvm-0.1.0.2.89.o: unknown symbol `_ZTV6Kernel' ghc: unable to load package `HSvm-0.1.0.2.89' Please note that the symbol is available in libHSHSvm-0.1.0.2.89.a $ nm HSHSvm-0.1.0.2.89.o |grep _ZTV6Kernel 00000000 V _ZTV6Kernel $ nm libHSHSvm-0.1.0.2.89.a |grep _ZTV6Kernel 00000000 V _ZTV6Kernel Can anyone give a look to the problem? Thanks Paolo [1] http://www.csie.ntu.edu.tw/~cjlin/libsvm/

1 0

Re: How to move code from GHC to libraries?
by Simon Marlow 14 Sep '09

14 Sep '09

On 12/09/2009 00:02, Norman Ramsey wrote: > > Incedentally, the reason I'd like us to make a decision on this now is > > because I'm about to add two new boot libraries... > > While we're on this topic, I've had several requests to pull HOOPL > (the new dataflow optimization code) out as a library so that we can > make it into a Hackage package. I'd like to do this, but the > situation is complicated because HOOPL depends on UniqFM, which is > GHC-specific. I thought that there was a similar implementation out > in the wild (based on Okasaki-Gill rather than Adams), but I can't > find it on hackage. Isn't Data.IntMap ~= UniqFM? > Is there an established procedure for moving a part of GHC into a > library without breaking everybody's build? If so, where will I find > it on the developers' wiki? If not, who can advise me so I can create > a page on the developers' wiki? :-) The architecture of our library setup is here: http://hackage.haskell.org/trac/ghc/wiki/Commentary/Libraries Basically you'd be adding a new zero-boot package (that terminology isn't used in the build system, where they're called "boot packages"). The best way to find out how to add another one is to grep for an existing zero-boot package (e.g. Cabal) in the top-level ghc.mk to see the various things you need to add. The new package would need to have a repository created on darcs.haskell.org, and you'd need to add it to $(TOP)/packages too. When the changes are pushed, everyone has to "./darcs-all get" to get the new repository. Cheers, Simon

1 0

GHCi can't see 'atexit', can't load module
by Maurício CA 13 Sep '09

13 Sep '09

(I left this in cafe, but I think it is actually ghc related.) I just checked it seems not to be possible to load any module in GHCi that uses FFI to wrap the standard C function 'atexit'. When trying that, we get a message saying the symbol 'atexit' can't be found. (This is not a problem when building an executable with ghc, though.) It's easy to reproduce the problem. Just try something like this: ----- module XXX where import Foreign import Foreign.C foreign import ccall atexit :: FunPtr (IO ()) -> IO CInt ----- Load it in ghci (after ':set -XForeignFunctionInterface' to get FFI) and you will get the error. If we replace the line below for the above we don't get any error message: foreign import ccall exit :: CInt -> IO () Here, we just replaced 'atexit' by 'exit', which, according to 'nm' tool, is at the same 'libc' library that 'atexit' belongs to. Thanks, Maurício

2 2

Build error in users guide in HEAD of today
by "Philip K.F. Hölzenspies" 11 Sep '09

11 Sep '09

Dear GHC-ers, I'm trying to put together a virtual machine with a default development for GHC hacking. To test it, I was running "sh validate" in the build tree. There were build errors in the doc/users_guide, however. Google gave me an hpaste page with the error on it, but I couldn't find the referring message. Does this ring a bell for anyone: latex failed users_guide_tmp.tex:1380: Undefined control sequence \Documents. users_guide_tmp.tex:1380: leading text: } users_guide_tmp.tex:1380: Undefined control sequence \user. users_guide_tmp.tex:1380: leading text: } users_guide_tmp.tex:5782: Undefined control sequence \Documents. users_guide_tmp.tex:5782: leading text: ...kurl{C: \Documents~And~Settings\user\ghc} users_guide_tmp.tex:5782: Undefined control sequence \user. users_guide_tmp.tex:5782: leading text: ...kurl{C: \Documents~And~Settings\user\ghc} users_guide_tmp.tex:5782: Undefined control sequence \ghc. users_guide_tmp.tex:5782: leading text: ...kurl{C: \Documents~And~Settings\user\ghc} users_guide_tmp.tex: File ended while scanning use of \hyper@n@rmalise. users_guide_tmp.tex: Emergency stop. Error: latex compilation failed make[2]: *** [docs/users_guide/users_guide.ps] Error 1 make[1]: *** [all_docs/users_guide] Error 2 I understand the LaTeX errors, but I'm not so well versed in docbook. Should the path in the ghci.xml and packages.xml be forward slashed, instead of backslashed or does this indicate there's something wrong with my docbook/latex setup? Even when I change the backslashes to forward slashes, the last error (scanning use of \hyper@n@rmalise) remains. Any ideas? Regards, Philip

3 3

Libraries in the repo
by Simon Marlow 11 Sep '09

11 Sep '09

Simon and I have been chatting about how we accommodate libraries in the GHC repository. After previous discussion on this list, GHC has been gradually migrating towards having snapshots of libraries kept as tarballs in the repo (currently only "time" falls into this category), but I don't think we really evaluated the alternatives properly. Here's an attempt to do that, and to my mind the outcome is different: we really want to stick to having all libraries as separate repositories. Background: * Scope: libraries that are needed to build GHC itself (aka "boot libraries") * Boot libraries are of several kinds: - INDEPENDENT: Independently maintained (e.g. time, haskeline) - COUPLED: Tightly coupled to GHC, but used by others (base) - SPECIFIC: Totally specific to GHC (e.g. template-haskell, DPH) * Most boot libraries are INDEPENDENT. INDEPENDENT libraries have a master repository somewhere separate from the GHC repositories. * We need a branch of INDEPENDENT libraries, so that GHC builds don't break when the upstream package is modified. * Sometimes we want to make local modifications to INDEPENDENT libraries: - when GHC adds a new warning, we need to fix instances of the warning in the library to keep the GHC build warning-free. - to check that the changes work, before pushing upstream Choices for how we deal with libraries in the GHC repository: (+) is a pro, (-) is a con. (1) Check out the library from a separate repo, using the darcs-all script. The repo may either be a GHC-specific branch [INDEPENDENT], or the master copy of the package [SPECIFIC/COUPLED]. (+) we can treat every library this way, which gives a consistent story. Consistency is good for developers. (+) [INDEPENDENT] makes it easy to push changes upstream and sync with the upstream repo (unless upstream is using a different VCS). (-) [INDEPENDENT] we have to be careful not to let our branches get too far out of sync with upstream, and we must sync before releasing GHC. (2) Put a snapshot tarball of the library in libraries/tarballs, but allow you to checkout the darcs repo instead. (-) [SPECIFIC/COUPLED] this approach doesn't really make sense, because we expect to be modifying the library often. (-) updating the snapshot is awkward (-) workflow for making a change to the library is awkward: - checkout the darcs repo - make the change, validate it - push the change upstream (bump version?) - make a new snapshot tarball - commit the new snapshot to the GHC repo. (-) having tarballs in the repository is ugly (-) we have no revision history of the library (3) The GHC repo *itself* contains every library unpacked in the tree. You are allowed to check out the darcs repo instead. (+) atomic commits to both the library and GHC. (+) doing this consistently would allow us to remove darcs-all, giving a nice easy development workflow (-) [INDEPENDENT/COUPLED] still need a separate darcs repo. (-) [INDEPENDENT/COUPLED] pushing changes upstream is hard (-) [INDEPENDENT/COUPLED] manual syncing with upstream (-) [COUPLED] (particularly base) syncing with upstream would be painful. (3) works best for SPECIFIC libraries, whereas (1) works best for INDEPENDENT/COUPLED libraries. If we want to treat all libraries the same, then the only real option is (1). Experience with Cabal and bytestring has shown that (1) can work for INDPENDENT libraries, but only if we're careful not to get too out-of-sync (as we did with bytestring). In the case of Cabal, we never have local changes in our branch that aren't in Cabal HEAD, and that works well. Comments/thoughts? Cheers, Simon

8 20

Resolving overloading loops for circular constraint graph
by Stefan Holdermans 10 Sep '09

10 Sep '09

Manuel, Simon, I've spotted a hopefully small but for us quite annoying bug in GHC's type checker: it loops when overloading resolving involves a circular constraint graph containing type-family applications. The following program (also attached) demonstrates the problem: {-# LANGUAGE FlexibleContexts #-} {-# LANGUAGE TypeFamilies #-} type family F a :: * type instance F Int = (Int, ()) class C a instance C () instance (C (F a), C b) => C (a, b) f :: C (F a) => a -> Int f _ = 2 main :: IO () main = print (f (3 :: Int)) My guess is that the loop is caused by the constraint C (F Int) that arises from the use of f in main: C (F Int) = C (Int, ()) <= C (F Int) Indeed, overloading can be resolved successfully by "black-holing" the initial constraint, but it seems like the type checker refuses to do so. Can you confirm my findings? I'm not sure whether this is a known defect. If it isn't, I'd be more than happy to issue a report. Since this problem arises in a piece of very mission-critical code, I would be pleased to learn about any workarounds. Thanks in advance, Stefan

3 5

Functor <constant>
by Serge D. Mechveliani 09 Sep '09

09 Sep '09

People, I have data DBit = Bit0 | Bit1 deriving (Eq, Ord, Enum) data BNatural = BNat [DBit] deriving (Eq) and want to apply things like fmap reverse (bn :: BNatural). GHC reports an error on this usage of fmap. It also does not allow instance Functor BNatural where fmap f (BNat ds) = BNat $ f ds, -- it reports that a type constructor for Functor must have kind * -> *. Probaly, GHC agrees with Haskell at this point (?). Now, I deceive the compiler by declaring instead newtype BNatAux a = BNat a deriving (Eq) type BNatural = BNatAux [DBit] instance Functor BNatAux where fmap f (BNat ds) = BNat $ f ds So, a parasitic BNatAux is introduced. I wonder: is there a simpler way to have fmap for BNatural ? Thank you in advance for advice, ----------------- Serge Mechveliani mechvel(a)botik.ru

2 1

fromEnum (c :: Char)
by Serge D. Mechveliani 09 Sep '09

09 Sep '09

People, I need to convert Char <-> Int in a possibly _standard_ way for Haskell -- and also in an efficient way. In particular, it must not spend 100 comparisons in a look through the listing of Char. I use ord :: Char -> Int and chr :: Int -> Char. Is this all right? Thank you in advance for advice, ----------------- Serge Mechveliani mechvel(a)botik.ru

2 1