February 2003 - Glasgow-haskell-users

RE: Histogram-building code (was: Re: Yet another weakly defined bug report)
by Simon Marlow 08 May '03

08 May '03

> One question though, about hFlush. I print out the status by > repeatedly putStr'ing "blah blah \r". With NoBuffering set, it works, > but when following the putStr with 'hFlush stdout', it doesn't (only > outputs very sporadically. I guess I'm misunderstanding the function > of hFlush, anybody care to elaborate?) It could be a bug. Can you provide us with a cut-down example? Cheers, Simon

3 2

RE: Yet another weakly defined bug report
by Simon Marlow 13 Mar '03

13 Mar '03

> You should do the counting strictly: > > Just n -> case n+1 of n1 -> addToFM f w n1 Careful - that case expression doesn't actually force n to be evaluated. It's exactly equivalent to let n1 = n+1 in addToFM f w n1 You need to use seq to force evaluation. Cheers, Simon

3 3

child process statistics
by Dean Herington 28 Feb '03

28 Feb '03

Does anyone know how to get the process statistics (cputime, page faults, etc.) for a child process? When the child process has terminated, before doing getAnyProcessStatus on it, I'd like to get its statistics. Browsing the libraries documentation didn't turn anything up. I couldn't even find what I want at the Unix level among man pages. (I'm happy to write some FFI code to access system routines if I can locate the right ones.) Dean

2 3

RE: ghcfilt
by Ashley Yakeley 26 Feb '03

26 Feb '03

Here's a Perl script that does it. -- Ashley Yakeley, Seattle WA

1 0

RE: ghcfilt
by Simon Peyton-Jones 26 Feb '03

26 Feb '03

There isn't a standalone program, no. But why do you want one? The code, and (I believe accurate) accompanying comments are in fptools/ghc/compiler/OccName.lhs. It would be fairly simple to extract this into a standalone program if you wanted to do so. Simon | -----Original Message----- | From: Ashley Yakeley [mailto:ashley@semantic.org] | Sent: 26 February 2003 02:25 | To: glasgow-haskell-users(a)haskell.org | Subject: ghcfilt | | Is there a "ghcfilt" equivalent of c++filt that will unmangle | 'z'-escaped GHC symbols? | | Is the mangling written up somewhere? I might be able to write | something... | | -- | Ashley Yakeley, Seattle WA | | _______________________________________________ | Glasgow-haskell-users mailing list | Glasgow-haskell-users(a)haskell.org | http://www.haskell.org/mailman/listinfo/glasgow-haskell-users

1 0

ghcfilt
by Ashley Yakeley 26 Feb '03

26 Feb '03

Is there a "ghcfilt" equivalent of c++filt that will unmangle 'z'-escaped GHC symbols? Is the mangling written up somewhere? I might be able to write something... -- Ashley Yakeley, Seattle WA

2 1

Wir verkaufen gebrauchte Alu FRAMAX Alu
by 520099746102-0001＠t-online.de 25 Feb '03

25 Feb '03

Wir verkaufen 300 qm DOKA Alu und Stahl, RABATT 82 % von Neuwert. Btte faxen Sie uns Ihre Anfrage unter Fax Nr: 00 48 22 654 50 81 We sell used Doka Framax Steel and Aluminium, 300 sqm. Prce with 82 & RAbat. Our Fax: + 48 22 654 50 81 STABILO S.p.z.o.o. E-MAIL: stabilo(a)stabilo.com.pl Internet Seite: www.stabilo.com.pl

1 0

RE: debug with GHC FFI?
by Simon Marlow 25 Feb '03

25 Feb '03

[ moved to glasgow-haskell-users(a)haskell.org ] > I am using Green Card to make some FFI modules with GHC, but > sometimes it segfaults in the middle of nowhere, leaving > no sensible stack trace in gdb. > > I wonder what is the proper way to debug FFI modules written > in C? Trying to add a "-g" flag doesn't seem help at all... You should be able to debug the normal C code using gdb. If the crash happens somewhere in Haskell code, then gdb is of less use (unless you happen to know the details of GHC's execution model). It sounds like in your case the crash is happening somewhere in Haskell. There are a few possible approaches: - If it is a callback into Haskell that is crashing, you should be able to use gdb to find which call into Haskell is the culprit, which might give you some clues. - use Debug.Trace.trace in the Haskell code to narrow down the bug. Ugly, but very useful. - Robert Ennals is working on a debugger for GHC as part of his speculative evaluation work. It's very much a work in progress though, and I haven't tried it myself yet. It's in ghc/utils/hs-debug on the speceval_2 branch of CVS. Cheers, Simon

1 0

RE: How to build ghci?
by Simon Marlow 25 Feb '03

25 Feb '03

> Thank's very much Andre. Now that I look for it, I did find > this in the FAQ. > (I looked at the building guide but missed the FAQ!) > > I'm still confused however. I though that with the whole two (three?) > stage bootstraping process you get a version of ghc built with itself > anyway, no? Maybe the ghc gurus could enlighten us. :-) > > If it requires an extra time consuming step, could it be a ./configure > option? --enable-ghci In 5.04.x you have to manually bootstrap - that is, you have to have two build trees, and you build the compiler in the second tree using the compiler built in the first tree. In the current CVS sources, the procedure is somewhat easier: a 'make' from the top-level will automatically to a 2-stage bootstrap, and there are options for doing 3-stage or just a 1-stage build too. Cheers, Simon

1 0

How to build ghci?
by Duncan Coutts 24 Feb '03

24 Feb '03

Hi all, I can't seem to get ghc to build so that ghci is enabled. I've tried building the tar and I've tried the Red Hat .src.rpm, which works but doesn't enable ghci on my system. I'm using Mandrake 9.0 which comes with gcc 3.2 and gcc 2.96 I'm building ghc to use gcc 2.96 (since it chokes on Mandrake's gcc 3.2). ghci reports: ghc-5.04.2: not built for interactive use Is there anything special I have to do to enable ghci? Is there something odd / missing from Mandrke's build environment? Duncan

2 2