August 2003 - Glasgow-haskell-users

RE: fast CString -> IO PackedString fuction
by Hal Daume 22 Aug '03

22 Aug '03

Hrm. On the C side the arrays disappear shortly after my call, so just casting wouldn't really work. I settled for the following monstrosity where (cast :: a -> b) is your standard unsafePerformIO/IORef based casting function: peekPackedCString :: CString -> IO PackedString peekPackedCString cs = do i <- findLength cs 0 (arr :: IOUArray Int Char) <- newArray_ (0,i) writeToArr arr i 0 cs (arr' :: UArray Int Char) <- unsafeFreeze arr return (cast arr') where findLength ptr n = do c <- peek ptr if c == 0 then return n else findLength (plusPtr ptr 1) $! n+1 writeToArr arr len pos ptr | pos >= len = return () | otherwise = do c <- peek ptr writeArray arr pos (castCCharToChar c) writeToArr arr len (len+1) (plusPtr ptr 1) it's pretty fast and works only because the type of PackedString is newtype PackedString = PS (UArray Int Char) so the casting gives us what we want. This does seem like something that belongs in the PackedString library or the Foreign.C.String library, though. - Hal -- Hal Daume III | hdaume(a)isi.edu "Arrest this man, he talks in maths." | www.isi.edu/~hdaume > -----Original Message----- > From: glasgow-haskell-users-admin(a)haskell.org > [mailto:glasgow-haskell-users-admin@haskell.org] On Behalf Of > John Meacham > Sent: Thursday, August 21, 2003 3:10 AM > To: glasgow-haskell-users(a)haskell.org > Subject: Re: fast CString -> IO PackedString fuction > > > On Wed, Aug 20, 2003 at 06:48:58PM -0700, Hal Daume wrote: > > is there a way to go from a CString to a PackedString w/o > going through > > a normal String in the middle? > > or should i write my own? > > If you can get a wchar_t * out of the app then that can be simply cast > to a packedstring, this is quite speedy. > if you use mbsrtowcs(3) and friends to do the conversion from char * > then you get properly localized text via locale settings for free :) > I have used this sort of thing for interfacing haskell to libraries > where honoring locale was important in the past. > John > > -- > -------------------------------------------------------------- > ------------- > John Meacham - California Institute of Technology, Alum. - > john(a)foo.net > -------------------------------------------------------------- > ------------- > _______________________________________________ > Glasgow-haskell-users mailing list > Glasgow-haskell-users(a)haskell.org > http://www.haskell.org/mailman/listinfo/glasgow-haskell-users >

1 0

fast CString -> IO PackedString fuction
by Hal Daume 21 Aug '03

21 Aug '03

is there a way to go from a CString to a PackedString w/o going through a normal String in the middle? or should i write my own? -- Hal Daume III | hdaume(a)isi.edu "Arrest this man, he talks in maths." | www.isi.edu/~hdaume

2 1

RE: Hash functions
by Simon Marlow 20 Aug '03

20 Aug '03

> Many thanks for Data.HashTable, which I am about to use. > Unfortunately > I seem to need an unseemly hack because the key I want, > namely ThreadId's, > don't have a hash function defined, and defining one requires > me to muck > around with GHC internal functions. Could some more hash functions be > provided? > > The logical method would be to have > > class HashKey key where > hash :: key -> Int32 > > and define it for as many types as possible. I think the reason I didn't do this was because the choice of a hash function usually depends on more than just the type of the data you want to hash. For example, there are lots of hash functions over Int, but you'll want to pick one that works well for the distribution of values you have. There really ought to be a way to get an Int from a ThreadId, though. Cheers, Simon

1 0

Hash functions
by George Russell 20 Aug '03

20 Aug '03

Many thanks for Data.HashTable, which I am about to use. Unfortunately I seem to need an unseemly hack because the key I want, namely ThreadId's, don't have a hash function defined, and defining one requires me to muck around with GHC internal functions. Could some more hash functions be provided? The logical method would be to have class HashKey key where hash :: key -> Int32 and define it for as many types as possible.

1 0

GHC 6.0.1 for Mac OS X available through darwinports
by Gregory Wright 20 Aug '03

20 Aug '03

Hello, GHC 6.0.1 for Mac OS X is now available through the darwinports system. Darwinports is similar to the *BSD port system. It downloads and builds both the target and dependencies automatically for your machine. Note that darwinports is still under development. OS X users who want a simple, reliable installation should use Wolfgang Thaller's GHC disk image. The darwinports version is aimed at people who want to build from source. If you have suggestions for particular customizations, let me know. Darwinports has support for 'variant' builds that make customization easy. For more information on darwinports, and to download the infrastructure, see http://www.opendarwin.org/projects/darwinports/ Best Wishes, Greg Wright Gregory Wright Antiope Associates 18 Clay Street Fair Haven, New Jersey 07704 USA gwright(a)antiope.com

1 0

What's difference between Integer and Int?
by Serguey Zefirov 19 Aug '03

19 Aug '03

Hello glasgow-haskell-users, The following program ----------------------------------------------- --main = print (show (fibs!!200000)) main = print (show (nth 200000 fibs)) nth 0 (h:_) = h nth n (h:t) = nth (n-1) t fibs :: [Integer] -- fibs = 1 : (map snd $ iterate (\(x,y) -> (y, x+y)) (1,1)) fibs = [1..] ----------------------------------------------- compiled with ghc 6.0: ghc -O -o a.exe a.hs does not execute correctly. It causes stack overflow in 1M stack. If we change "fibs :: [Integer]" to "fibs :: [Int]" the new program runs Ok. -- Best regards, Serguey mailto:sz@uc.ru

2 1

Üzleti ajánalt ! - Business offer !
by András József 19 Aug '03

19 Aug '03

Törvényes, és bevált kereseti lehetoséget ajánlunk minimális befektetéssel , napi 1 - 2 órai munkával az Interneten ! Havi jövedelem kb. 5.000 - 10.000 USD tisztán ! Amennyiben komolyan érdekli a téma, bovebb tájékoztatást nyújtok e-mailben a következo címen : aiosif(a)xms.ms . We offer you a legal, well working source of income, by just a minimal expenditure, and by a 1 - 2 hour daily work over the Internet, YOUR monthly income will be around 5.000 - 10.000 USD ! In so far as you are interested by the theme, I'll provide further details by e-mail in the following adress : aiosif(a)xms.ms . András József

1 0

Üzleti ajánalt ! - Business offer !
by András Jószef 19 Aug '03

19 Aug '03

Törvényes, és bevált kereseti lehetoséget ajánlunk minimális befektetéssel , napi 1 - 2 órai munkával az Interneten ! Havi jövedelem kb. 5.000 - 10.000 USD tisztán ! Amennyiben komolyan érdekli a téma, bovebb tájékoztatást nyújtok e-mailben a következo címen : aiosif(a)xms.ms . We offer you a legal, well working source of income, by just a minimal expenditure, and by a 1 - 2 hour daily work over the Internet, YOUR monthly income will be around 5.000 - 10.000 USD ! In so far as you are interested by the theme, I'll provide further details by e-mail in the following adress : aiosif(a)xms.ms . András József

1 0

RE: profiling for stack usage
by Simon Marlow 19 Aug '03

19 Aug '03

> >> 1. Is there a way to profile stack usage, so that I can > identify the > >> culprit and deal with the problem at the root? > > > > http://www.haskell.org/ghc/docs/latest/html/users_guide/prof-h eap.html#RTS-OPTIONS-HEAP-PROF+RTS > > > > in particular, the -xt flag. > > Hmm, sorry for being so dense, but I'm having a tough time with this. > It seems profiling by itself tends to blow the stack -- is that > correct/normal behavior? (I'm also using -O2, if that matters) It is possible for profiling to change the stack behaviour, unfortunately. This is because certain optimisations are disabled when profiling is on, so that we can retain useful profiling information (oh the irony :-). > Is there any way to know what kind of data resides on the stack, or > what function generated it? The TSO tells med its size, but isn't > really helpful beyond that (or am I missing something?) I'm afraid not, no. We don't have a way to profile the contents of the stack, only its size. Cheers, Simon

1 0

profiling for stack usage
by ketil＠ii.uib.no 19 Aug '03

19 Aug '03

Hi again, I'm currently in a situation where my program runs easily out of stack. Depending on the input, stack usage often exceeds 10Mb. 1. Is there a way to profile stack usage, so that I can identify the culprit and deal with the problem at the root? Normal (time) profiling tells me how many times a function is called, but it would be interesting to know how many times it was recursively called, or the size of its stack frames. Is that information available? Heap profiling -- well, it doesn't *sound* as if it would incorporate the stack; does it anyway? 2. Is there a way to compile programs to use more than the default stack? I can of course pass +RTS -K10M -RTS on the command line, but I would rather like to change the default, instead of kicking myself for forgetting it all the time. And is there any reason (except excessive resource consumption and postponed failure from infinite loops) not to run with huge stacks? -kzm -- If I haven't seen further, it is by standing in the footprints of giants

3 3