September 2007 - Glasgow-haskell-users

RE: [Haskell] Blocked STM & GC question
by Simon Peyton-Jones 17 Sep '07

17 Sep '07

[Redirecting to GHC Users which is the right place for questions about GHC.] Yes, the garbage collector will (well, certainly should) find and kill such threads. By "kill" I mean that they get sent an asynchronous exception of some kind (I hope the documentation says which) so that the thread has a chance to clean up rather than being forcibly terminated. I'm not quite sure where this fact should be documented -- documentation for GHC itself, or for the STM library? -- but it surely should be somewhere. Simon | -----Original Message----- | From: haskell-bounces(a)haskell.org [mailto:haskell-bounces@haskell.org] | On Behalf Of Ashley Yakeley | Sent: 15 September 2007 21:39 | To: haskell(a)haskell.org | Subject: [Haskell] Blocked STM & GC question | | If I have a thread that's blocked on an STM retry or TChan read, and | none of its TVars are referenced elsewhere, will it get stopped and | garbage-collected? | | I have in mind a pump thread that eternally reads off a TChan and | pushes | the result to some function. If the TChan is forgotten elsewhere, will | the permanently blocked thread still sit around using up some small | amount of memory, or will it be reaped by the garbage collector? | | -- | Ashley Yakeley | Seattle WA | | _______________________________________________ | Haskell mailing list | Haskell(a)haskell.org | http://www.haskell.org/mailman/listinfo/haskell

1 0

Another compilation problem
by David Schonberger 17 Sep '07

17 Sep '07

Another newbie question. To recap, running 6.6.1 on Vista. Fixed the other problem--lexical error, apparently a BOM added by Notepad--by getting Xemacs. Now when I type 'ghc -o hello hello.hs' at the prompt I get gcc: installation problem, cannot exec 'as': no such file ro directory. Any idea what's going on here or how to fix? Thanks. David

2 1

About ReplacingGMPNotes
by Roberto Zunino 16 Sep '07

16 Sep '07

I noticed the following on http://hackage.haskell.org/trac/ghc/wiki/ReplacingGMPNotes === The LGPL licensing situation may be worse: even though The Glasgow Haskell Compiler License is essentially a "free software" license (BSD3), according to paragraph 2 of the LGPL, GHC must be distributed under the terms of the LGPL! === This seems to be inaccurate to me. The LGPLv2.1, section 6, states that, if you statically-link, it is enough to provide the GHC source (even under BSD), o re-linkable object code. I think the newer LGPL has a similar clause. Also: === (Dynamic libraries are currently distributed only for OS X.) === This is for Haskell libs, and does not affect the GMP issues in any way, AFAICS. GMP is indeed dynamically linked in my linux system. Regards, Roberto Zunino.

2 1

Re: ANNOUNCE: GHC 6.8.1 Release Candidate
by Philip K.F. Hölzenspies 14 Sep '07

14 Sep '07

Dear Ian, all, I'm getting a build error that I've seen before. I must be overlooking something. What I do: lzensp@ewi1043:~/local/src> curl http://www.haskell.org/ghc/dist/stable/dist/ghc-6.8.20070912-src.tar.bz2 | tar -xj % Total % Received % Xferd Average Speed Time Time Time % Current Dload Upload Total Spent Left Speed 100 6893k 100 6893k 0 0 682k 0 0:00:10 0:00:10 --:--:-- 743k holzensp@ewi1043:~/local/src> cd ghc-6.8.20070912 holzensp@ewi1043:~/local/src/ghc-6.8.20070912> ./boot Booting . Booting libraries/base Booting libraries/directory /usr/share/aclocal/progsreiserfs.m4:13: warning: underquoted definition of AC_CHECK_LIBREISERFS run info '(automake)Extending aclocal' or see http://sources.redhat.com/automake/automake.html#Extending-aclocal Booting libraries/old-time Booting libraries/process Booting libraries/readline Booting libraries/unix holzensp@ewi1043:~/local/src/ghc-6.8.20070912> ./configure && make [...] make[1]: *** No rule to make target `../includes/DerivedConstants.h', needed by `stage1/codeGen/CgProf.o'. Stop. make[1]: Leaving directory `/local/holzensp/src/ghc-6.8.20070912/compiler' make: *** [stage1] Error 1 Either this is a bugreport or a kind request for advice. Regards, Philip

3 3

getMBlocks: VirtualAlloc MEM_RESERVE 1 blocks failed
by J. Garrett Morris 12 Sep '07

12 Sep '07

Hello everyone, I've been running into the following error: Reproduction.exe: getMBlocks: VirtualAlloc MEM_RESERVE 1 blocks failed: Not enough storage is available to process this command. The confusing part to me is that when the error occurs, Task Manager reports that the process is using around 100K of memory, while I know that the machine has at least 2GB available. Reproduction case is attached. This runs against the Netflix prize training set, which is large enough that I haven't attached it. At the time that the error occurs, it's processed on the order of 15,500 files. Is it possible there's some other resource that I'm exhausting without realizing it? The compiler command line is: ghc --make -O2 -threaded -hidir obj -odir obj -o Reproduction Reproduction.hs /g

1 0

FFI and Libraries using GMP
by Benedikt Huber 10 Sep '07

10 Sep '07

Hello, I've been struggling using FFI bindings to libraries which rely on the GNU Mp Bignum library (gmp) - this is apparently a well known problem (http://hackage.haskell.org/trac/ghc/ticket/311) I do want to use some of those libraries however and so I started to get them work on ghc-6.6.1/darwin_8_i368. But as no solution was close to being satisfactory, I'd like to ask for some advice. Those two options worked to some extend: (1) Create or modify the library in question, so gmp is statically linked and its symbols are hidden. When source code is available, this is relatively easy, altough it requires modification of the build process (which can be a hassle). If a static "ar" archive is available, it is cumbersome (at least on Mac Os X I ran into a lot of troubles using nmedit), but possible. Furthermore, the resulting libraries are bloated, as each of them contains a copy of the GMP; left alone portability issues. (2) As suggested in ticket#311, I tried switching the allocator functions when switching to FFI. It worked, but not in GHCi; also, doing it manually is a lot of work, because, as far as I could figure out, it is neccessary to write a wrapper function for every C-function (indirectly) using the gmp. Furthermore, most of the libraries expose gmp datatypes (mpz_t,mpq_t) in their API. I currently use a little haskell module working on GHC.Exts, but that's propably not a good option from a maintainer's point of view. There are certainly other possibilities, but I couldn't find one which is both maintainable and portable. I would be very grateful for any advice, or some information on plans for resolving ticket #311; I'm aware that there are long-term plans for replacing gmp, but I'd prefer a seperate short-term solution for the problems with the FFI. There are some great libraries (like the Parma Polyhedral Library, to pick an example) out there using gmp, and it would be nice if writing bindings to those libs could be simplified. Thanks, Benedikt

1 0

capturing ghci command output; :redir/:grep/:find
by Claus Reinke 08 Sep '07

08 Sep '07

having only recently started to play with ghci's :def and :cmd, i'm still in the process of figuring out what is and what isn't possible, and i've been wondering what wonderful things other ghci users have in their .ghci files? to start with, here are my current favourites (definitions attached, this is using a recent ghc 6.7). i'm looking forward to seeing your gems and favourites:-) -- capturing command output in variables: *Main> :redir out :info Monad *Main> putStrLn out class Monad m where (>>=) :: m a -> (a -> m b) -> m b (>>) :: m a -> m b -> m b return :: a -> m a fail :: String -> m a -- Defined in GHC.Base instance Monad Maybe -- Defined in Data.Maybe instance Monad [] -- Defined in GHC.Base instance Monad IO -- Defined in GHC.IOBase -- filtering command output for lines matching pattern: *Main> :grep break :? :abandon at a breakpoint, abandon current computation :break [<mod>] <l> [<col>] set a breakpoint at the specified location :break <name> set a breakpoint on the specified function :continue resume after a breakpoint :delete <number> delete the specified breakpoint :delete * delete all breakpoints :step single-step after stopping at a breakpoint :trace trace after stopping at a breakpoint :trace <expr> trace into <expr> (remembers breakpoints for :history) :set stop <cmd> set the command to run when a breakpoint is hit :show breaks show the active breakpoints :show context show the breakpoint context *Main> :grep class :browse Control.Monad class Monad m where class Functor f where fmap :: (a -> b) -> f a -> f b class (Monad m) => MonadPlus m where -- Hugs' :find <id>, calling editor to open definition of <id> *Main> :find main probably, there should be a wiki page collecting such :definitions, and other ways of improving the ghci usage experience. but i'm unsure whether to add to the haskell wiki http://haskell.org/haskellwiki/GHC/GHCi or to the ghc developer's wiki http://hackage.haskell.org/trac/ghc opinions? suggestions? tips? tricks?-) claus

2 1

docon-2.10
by Serge D. Mechveliani 06 Sep '07

06 Sep '07

Announcement: I have put docon-2.10 instead of 2.09. What is new: two bugs fixed in installation. One of the fixes is for MacOS X, for its feature concerning the letter case in file names. Regards, ----------------- Serge Mechveliani mechvel(a)botik.ru

1 0

Re: Mac GHC binary
by Christian Maeder 06 Sep '07

06 Sep '07

Hi Jeff, I think, there's only a problem when the GNUreadline framework is only in your home directory and when you want to link your own binary. Correct me if you also have a problem if the framework is globally installed in /Library/Frameworks. Indeed a framework entry is missing in the package.conf. But I doubt if a generic entry using $HOME would work. So this entry could only be set at installation time for a single user (and I don't know how). In fact, for linking we use the additional ghc flags: -optl-F$HOME/Library/Frameworks -optl-framework -optlGNUreadline These flags could be put in top-level script for ghc (in bindir). Does anyone have a better idea? Cheers Christian P.S, I hope it's ok for you, if I forward this to glasgow-haskell-users(a)haskell.org, too Jeff Lewis wrote: > Hi Christian: > > A few people here have tried using your binary Mac OS X build of > GHC. There's one problem we've run into. The readline package doesn't > work out of the box. It seems to be necessary to add `-framework > GNUreadline'. This shouldn't be necessary. I'd guess that there's a > missing `frameworks' entry in the package description. > > --Jeff

1 0

configure option --oldincludedir
by Wolfgang Jeltsch 05 Sep '07

05 Sep '07

Hello, what is the option --oldincludedir of the configure script in the current 6.7 nightly snapshot for? ./configure --help says that its default is /usr/include. Is this directory used to install files into or is it just used for reading files? Best wishes, Wolfgang

2 1