- Glasgow-haskell-users

Re: Timing Stats in ghci-6.6
by Daniel Fischer 23 Oct '06

23 Oct '06

Repost, seems to have been lost last week: Am Dienstag, 17. Oktober 2006 12:35 schrieb Simon PJ: > What timing? Can you be more explicit pls? > Just plain old :set +s in ghci: dafis@linux:~/Downloads/Euler> ghci Euler44 ___ ___ _ / _ \ /\ /\/ __(_) / /_\// /_/ / / | | GHC Interactive, version 6.6, for Haskell 98. / /_\\/ __ / /___| | http://www.haskell.org/ghc/ \____/\/ /_/\____/|_| Type :? for help. Loading package base ... linking ... done. [1 of 3] Compiling Primes ( Primes.hs, interpreted ) [2 of 3] Compiling ContFrac ( ContFrac.hs, interpreted ) [3 of 3] Compiling Euler44 ( Euler44.hs, interpreted ) Ok, modules loaded: Euler44, ContFrac, Primes. *Euler44> :set +s *Euler44> smallest (2167,1020,2395,1912) (0.02 secs, 910618996 bytes) *Euler44> Actually took 23 secs. Built it yesterday from source bundle on linux. Oh, and by the way, I've never managed to get the users-guide built. Is that intentional or would that just require some xml-processing functionality I haven't installed (if so: which)? Cheers, Daniel > Simon > > | -----Original Message----- > | From: glasgow-haskell-users-bounces(a)haskell.org > > [mailto:glasgow-haskell-users- > > | bounces(a)haskell.org] On Behalf Of Daniel Fischer > | Sent: 17 October 2006 10:04 > | To: GHC Users > | Subject: Timing Stats in ghci-6.6 > | > | Hi, > | > | apparently the timing doesn't work properly in ghci 6.6 (at least for > > me), I > > | always get very small timings, usually 0.01 secs, sometimes 0.00 secs, > > once > > | 0.07 secs, no matter how long it actually took. > | > | Is that a common problem or is it only me? > | > | Cheers, > | Daniel > | _______________________________________________ > | Glasgow-haskell-users mailing list > | Glasgow-haskell-users(a)haskell.org > | http://www.haskell.org/mailman/listinfo/glasgow-haskell-users

1 0

getting the -B topdir for a GHC-API app
by Mathew Mills 23 Oct '06

23 Oct '06

What is the "right" way to get the topDir for use with the GHC API. I have a little app that I would like to cabalize and release, but I am not sure what is the approved way of getting the '-B' option set-up. I think it would be great if I could use template-haskell to insert the topDir from the ghc used to compile the app, but I am not sure how to get the topDir from the compiler. Is this possible? Does cabal already provide a means to install a "driver" script like that used for the ghc suite? Also, I am looking for a code-review. If you have some time, I would love some "style" guidance. The work in progress can be found here: http://mathewmills.com/runTests/runTests.lhs

2 1

Installing a library
by jim burton 21 Oct '06

21 Oct '06

A very basic question re using GHC 6.4.1 on ubuntu linux - how do I go about installing a library? I want to install HUnit and suspected that /usr/lib/haskell-packages/ghc6/lib would be the place to untar it (into the dir HUnit-1.0), but that doesn't work. I was reluctant to move the contents of HUnit-1.0 directly into ...ghc6/lib because it includes examples and docs as well as the module files. I notice that MissingH, which I've installed, is in that dir but inside it's top level there are files like HSMissingH-0.13.1.o, installed-pkg-config, libHSMissingH-0.13.1.a ... does that mean that MissingH is a Package (something which I don't understand to be honest) and HUnit isn't? How do I tell what the load path of GHC is and add something to it? Where should I unpack HUnit for it to be accessible? Thanks, -- View this message in context: http://www.nabble.com/Installing-a-library-tf2483332.html#a6924826 Sent from the Haskell - Glasgow-haskell-users mailing list archive at Nabble.com.

2 1

pinned byte arrays
by Bulat Ziganshin 20 Oct '06

20 Oct '06

Hello glasgow-haskell-users, i have program that holds in memory a lot of strings (filenames) and use my own packed-string library to represent these string. this library uses newByteArray# to allocate strings. in my benchmark run (with 300.000 filenames in memory) program prints the following statistics: 37,502,956 bytes maximum residency (16 sample(s)) 2928 collections in generation 0 ( 6.90s) 16 collections in generation 1 ( 12.20s) 65 Mb total memory in use most of this memory occupied by filenames. note that 28 mb of memory is used just for GC procedure (i use compacting GC) then, i thought that using pinned byte arrays should significantly improve memory usage because these arrays can't be moved around and therefore, i thought, these arrays will not be involved in GC. amount of data which should be compacted by GC will decrease and amount of memory wasted by GC should also decrease so i replaced in my packed str lib newByteArray# with newPinnedByteArray# and seen the following: 53,629,344 bytes maximum residency (14 sample(s)) 2904 collections in generation 0 ( 6.97s) 14 collections in generation 1 ( 13.16s) 100 Mb total memory in use why? why memory usage was increased? why another 47 megs is just wasted out? how GC works with pinned byte arrays? and why GC times was almost the same - i thought that GC for pinned data should be very different than GC for unpinned byte arrays -- Best regards, Bulat mailto:Bulat.Ziganshin@gmail.com

3 2

recursive commentaries :)
by Bulat Ziganshin 19 Oct '06

19 Oct '06

Hello glasgow-haskell-users, http://hackage.haskell.org/trac/ghc/wiki/Commentary contains a link 'The old GHC Commentary' which points to the http://www.cse.unsw.edu.au/~chak/haskell/ghc/comm/ page. guess what this page contains? :) it will be great to restore link to old commentaries -- Best regards, Bulat mailto:Bulat.Ziganshin@gmail.com

1 0

Re: Bad control characters in GHCi 6.6 (Mac OS X PPC)
by Björn Buckwalter 19 Oct '06

19 Oct '06

On 10/16/06, Simon Marlow <simonmarhaskell(a)gmail.com> wrote: > Björn Buckwalter wrote: > > > I downloaded the GHC 6.6 tarballs, built and installed them on my mac > > (Tiger, PPC) using the documented procedure: > > ./configure > > make > > make install > > > > The build and install was fairly smooth. However, ghci will not > > recognize "control characters", e.g. delete/backspace, arrow keys, > > ctrl-a, ctrl-d et c. It seems to recognize ctrl-c for cancellin5.2g a > > line but will print the ^C first. A typical session may look like: > > It sound like either you don't have readline installed, or the GHC build didn't > detect it. Do you have readline? (the real GNU readline, not Apple's version). I have GNU readline 5.1 installed through darwinports. As for the GHC build detecting it I wouldn't know. Rerunning ./configure produces the following lines mentioning readline (note that I have already built and installed GHC 6.6 and have it in the path): checking whether ghc has readline package... no checking readline/readline.h usability... yes checking readline/readline.h presence... yes checking for readline/readline.h... yes checking readline/history.h usability... yes checking readline/history.h presence... yes checking for readline/history.h... yes configure: configuring in readline checking for readline in -lreadline... yes checking for rl_readline_version in -lreadline... no config.status: creating readline.buildinfo config.status: creating include/HsReadlineConfig.h config.status: include/HsReadlineConfig.h is unchanged I don't know what to make of the above. Thanks, Bjorn Buckwalter

1 0

Data.Array.MArray "lost" bounds
by Andreas Marth 18 Oct '06

18 Oct '06

With ghc-6.6 there is a change in module Data.Array.MArray which now includes a function getBounds :: (MArray a e m, Ix i) => a i e -> m (i,i) but it misses the good old bounds [which had type bounds :: (HasBounds a, Ix i) => aie -> (i,i)]. What is the rational behind this change? Are MArrays now able to change there size? Thanks for any eplanations, Andreas

3 2

Some potentially useful installers..
by Sigbjorn Finne 17 Oct '06

17 Oct '06

Hi, I've bundled up a bunch of Win32 installers for various tools that come in handy when developing with&for GHC: Alex, Happy, and Haddock (aka "The Marlow Collection") http://galois.com/~sof/msi/alex-2-0-1.msi http://galois.com/~sof/msi/happy-1-15.msi http://galois.com/~sof/msi/haddock-0-7.msi enjoy, --sigbjorn

2 2

typo in http://www.haskell.org/onlinereport/prelude-index.html
by Christian Maeder 17 Oct '06

17 Oct '06

I've just noticed the wrong example entry for break: break :: (a -> Bool) -> [a] -> ([a], [a]) break (<2) [1,2,3] = ([1],[2,3]) this should be changed to: break (>2) [1,2,3] = ([1,2],[3]) C. Prelude> break (<2) [1,2,3] ([],[1,2,3])

1 0

foldl laziness support. Reply
by Serge D. Mechveliani 17 Oct '06

17 Oct '06

Concerning the laziness support problem, I thank people for explanations about foldl and foldr. >> I wonder how to avoid these numerous cost pitfalls. >> Maybe, the complier could do more optimization? Duncan Coutts <duncan.coutts(a)worc.ox.ac.uk> writes > There are important differences between foldl, foldl' and foldr. It is > quite important to choose the right one. I don't think this can be done > automatically. > > In my experience, the choice is almost always between foldl' and foldr. > > [..] I do not see foldl' in the standard library. Is it of the GHC lib extension? has it strictness annotation? > So as Lemmih says, in this case you want to use foldr: > > import List (union) > main = let n = 10^4 :: Int > in > putStr > (shows (take 2 $ unionMany [[1 .. i] | i <- [1 .. n]]) "\n") > > unionMany = foldr union [] I see. Thank you. I have impression that something is here besides the intuition for the foldl/foldr choice. Here is a contrived example which is more close to my real situation. ----------------------------------------------------------------- import qualified Data.Set as Set (Set(..), empty, member, insert) import List (union, find) main = let n = 10^6 :: Int in putStr (shows (g1 n) "\n") f :: Int -> (Set.Set Int, [Int]) f n = -- original version, I write so because it is easy to program -- foldl add (Set.empty, []) [[1 .. i] | i <- [1 .. n]] where add (s, xs) ys = (Set.insert (sum xs) s, union xs ys) {- attempt to optimize (fails) -- h (Set.empty, []) [[1 .. i] | i <- [1 .. n]] where h (s, xs) [] = (s, xs) h (s, xs) (ys: yss) = h (Set.insert (sum xs) s, union xs ys) yss -} g1, g2 :: Int -> Bool -- client functions g1 n = case snd $ f n of x: _ -> even x _ -> False g2 n = let (set, xs) = f n in case find (> 100) xs of Just x -> Set.member (2*x) set _ -> False ----------------------------------------------------------------- Evidently, g1 n must have the cost of O(1). But in ghc-6.6 -O, it has O(n). How to improve f ? I tried foldr, and failed. The situation is so that some clients are as g1, and others are as g2, and, at least, g1 must be O(1). Regards, ----------------- Serge Mechveliani mechvel(a)botik.ru

4 3