- Glasgow-haskell-users

Is there a way to find out the type of a variable inside a function?
by Andrew Wilcox 18 Aug '06

18 Aug '06

I'm learning Haskell, and I've been reading through various source code examples. I find the ":type" commend in GHCi very helpful in finding out what the type of top-level declarations are. However, it only works on top-level declarations (as far as I know). I find myself looking at code like f x = ... where g = ... wondering what f does, and so I try to figure out what the type of g is... and I get lost pretty quickly! :-) Is there a way to find out the type of g? Thanks! Andrew

2 2

New option -a (append) for hasktags
by Martin Grabmueller 16 Aug '06

16 Aug '06

Hello, when using the hasktags program as advertised in the users manual, e.g. find -name \*.\*hs | xargs hasktags the tags file(s) get repeatedly truncated when there are so many files that xargs calls the hasktags program more than once (because otherwise there would be too many command line arguments). I have patched hasktags to accept a new command line option -a --append append to existing CTAGS and/or ETAGS file(s) which makes hasktags to append to existing tags file(s) instead of truncating it. With this patch, it is possible to index the whole darcs tree for ghc (including all libraries) using the following command: find -name \*.\*hs | xargs hasktags -a I will 'darcs send' the resulting patch and hope this patch will be applied. Thanks, Martin

1 0

class declaration in boot file
by Gaal Yahas 15 Aug '06

15 Aug '06

I'm trying to solve a circularity problem with .hs-boot, but am getting the error: "Illegal class declaration in hs-boot file". The offending declaration is: class (Monad m, Functor m, Eq a, Data a, Typeable a) => ICoercible m a | a -> m I've tried with and without the fundeps. Data and Typeable are in scope, and -fglasgow-exts is set. Any ideas on what else I should be doing? -- Gaal Yahas <gaal(a)forum2.org> http://gaal.livejournal.com/

2 4

recursive import
by Serge D. Mechveliani 15 Aug '06

15 Aug '06

Please, discard my previous request for the help with .hs-boot. I seem to get through the .hs-boot matter. We shall see further. Still it looks strange. The compiler reports errors: this and this thing must not be in .hs-boot. As it understands everything, it could take the .hs copy and filter what is needed for .hs-boot. Why the user needs to create .hs-boot by hand? ----------------- Serge Mechveliani mechvel(a)botik.ru

2 1

behaviour of {-# NOINLINE #-} in where clauses
by Duncan Coutts 14 Aug '06

14 Aug '06

This doesn't have the effect I expected: loop xs = case blah of One thing -> ... loop The other -> ... realloc ... where {-# NOINLINE realloc #-} realloc = do something loop ... My intention here was that the loop would not contain the code for realloc and that it'd be done as a call at the cmm level. My intention is to take the slow and rarely taken realloc path out of the code for the fast path. It seems the {-# NOINLINE realloc #-} pagma did not have the effect I intended. Looking at the -ddump-simpl and -ddump-cmm, the code for the realloc gets expanded in place in a branch of a case statement. In the cmm code we end up with just what I didn't want: loop_info: if (offset != 4096) goto later; ... ... lots of realloc code taking up space ... in the instruction / trace cache ... later: .. do the fast bits, read a byte, write a byte jump loop_info; Not only does the slow path take up space but it's in the location favoured by the hardware's static branch prediction. Reversing the test doesn't help because either way ghc turns it into: case thing of _DEFAULT -> 4096 -> and from that generates CMM: if (thing != 4096) goto much_later; ... much_later: ... The reason I was looking at this is because I've been trying to figure out why our lazy byte string fusion primitives are much slower than the strict versions. It's improving though, it's now only half the speed rather than a tenth of the speed. :-) The ByteString.Lazy code is an interesting mixture of strict and lazy. We must strictly read/write the chunks but lazily generate/consume the list of chunks. I just discovered that I should have been reading STG all along rather than core from the simplifier or CMM. STG takes out all the type annotations which tend to make things quite verbose. Mind you, seeing the types can be handy too to see if/how things are unboxed. Even so, I kind of wish there were a stage between STG and CMM that showed the imperative model of STG with linear layout, control flow and notes to indicate thunk/closure allocations. I expect most of my problem is that I do not understand the STG evaluation model sufficiently well to see how it maps to basic blocks, jumps/calls etc. Duncan

3 3

Holidays
by Simon Peyton-Jones 14 Aug '06

14 Aug '06

GHC enthusiasts, Simon M is on holiday this week (14 Aug), and I'm on holiday next week (21 Aug), so GHC HQ may seem less responsive than usual. For example, thanks for Johan and Esa for recent patches -- but I'll leave them for Simon to review and commit. keep hacking, though! Simon

1 0

Re: OpenSSL License (was Replacement for GMP: Update)
by Peter Tanski 13 Aug '06

13 Aug '06

John, > Have you carefully investigated the OpenSSL license? We in Debian > have > had repeated problems since the OpenSSL license is, as written, > incompatible with the GPL (even linking to OpenSSL is incompatible > with > the GPL). I would hate to have a situation where all GHC-compiled > programs can't be under the GPL. I have been discussing this very issue with several people in this list. At first I thought it was a peripheral issue because I had not carefully reviewed the GPL license until Florian Weimer pointed out what the problem was. By that time I had already decided to work with another replacement library--I have not yet decided which because I am still testing things out--based on good arguments other people gave for not subjecting users to OpenSSL's advertising clause. The libraries I am working with are either public domain or BSD licenses. Sorry to scare you. Best regards, Peter

1 0

GHC Error in linker "undeined reference"
by Johan Grönqvist 12 Aug '06

12 Aug '06

Hello, I am using ghc on ubuntu 6.06, and it has worked well for my first few attempts (one example pasted below), but I now have a file that works well in ghci, but not in ghc. When loading the file (pasted below) into ghci, it works fine.and the program runs as expected. Compilation (linking, actually) with ghc gives error messages (pasted below), but when I again start ghci adn load the program, I can see that it now uses the compiled ".o"-file instead of interpreting the program. I assume that some files that should be installed by default are not installed. I apologize if I am violating any policy about pasting in messages. Thanks in advance for any help / johan ---------------------------------------------------- -- Working example: sort [] = [] sort (x:xs) = insert x (sort xs) where insert x [] = [x] insert x (y:ys) | x <= y = (x:y:ys) | otherwise = y : (insert x ys) main = print $ sort [1,4,53,45,1,435,45,45,1,435,45,145,45345,3,345] ---------------------------------------------------------- ------------------------------------------------------ -- Non-linkable example {-# OPTIONS_GHC -fglasgow-exts #-} import Control.Monad.State import Control.Monad.ST import Data.Array.ST import Data.List class Stack s a where emptyStack :: (s a) isEmpty :: State (s a) Bool pop :: State (s a) a push :: a -> State (s a) () nTh :: Int -> State (s a) a depth :: State (s a) Int instance Stack [] a where emptyStack = [] :: [a] isEmpty = get >>= \lst -> return (length lst == 0) pop = get >>= \(x:xs) -> put xs >> return x push x = get >>= \xs -> put (x:xs) >> return () nTh n = get >>= \lst -> return $ lst !! n depth = get >>= (return . length) class Mem m a where emptyMem :: Int -> a -> ST s (m s Int a) fetch :: (m s Int a) -> Int -> ST s a store :: (m s Int a) -> Int -> a -> ST s () instance Mem STArray a where emptyMem n val = newArray (0,n) val :: ST s (STArray s Int a) fetch m ix = readArray m ix store m ix val = writeArray m ix val stackTest = evalState doStackTest (emptyStack :: [Int]) where doStackTest = (push 4 >> push 2 >> pop >>= \a -> pop >>= \b -> return (a,b)) memTest = runST doMemTest where doMemTest :: ST s (Int,Int) doMemTest = ((emptyMem 2 0 :: ST s (STArray s Int Int)) >>= \mem -> store mem 0 2 >> store mem 1 4 >> fetch mem 0 >>= \a -> fetch mem 1 >>= \b -> return (a,b)) main = (print stackTest >> print memTest >> return ()) -------------------------------------------------------------------- Error messages frm ghc: ---------------------------------------------------- ubuntu@ubuntu:~/haskell/Forth$ ghc forth.lhs forth.o: In function `s2Q0_info': undefined reference to `ControlziMonadziState_zdfMonadStates_closure' forth.o: In function `s2Q6_info': undefined reference to `ControlziMonadziState_zdfMonadStates_closure' forth.o: In function `s2Qp_info': undefined reference to `ControlziMonadziState_zdfMonadStates_closure' forth.o: In function `s2QA_info': undefined reference to `ControlziMonadziState_zdfMonadStates_closure' forth.o: In function `s2R5_info': undefined reference to `ControlziMonadziState_zdfMonadStates_closure' forth.o: more undefined references to `ControlziMonadziState_zdfMonadStates_closure' follow forth.o: In function `s2TI_info': undefined reference to `ControlziMonadziState_zdfMonadState_closure' forth.o: In function `s2TL_info': undefined reference to `ControlziMonadziState_zdfMonadState_closure' forth.o: In function `s2Tm_info': undefined reference to `ControlziMonadziState_zdfMonadState_closure' forth.o: In function `s2Tp_info': undefined reference to `ControlziMonadziState_zdfMonadState_closure' forth.o: In function `s2TO_info': undefined reference to `ControlziMonadziState_zdfMonadState_closure' forth.o: In function `r2Pv_info': undefined reference to `ControlziMonadziState_evalState_closure' forth.o: In function `__stginit_Main_': undefined reference to `__stginit_ControlziMonadziState_' forth.o: In function `Main_zdfStackZMZN_srt': undefined reference to `ControlziMonadziState_zdfMonadStates_closure' forth.o: In function `s2TO_srt': undefined reference to `ControlziMonadziState_zdfMonadState_closure' forth.o: In function `r2Pv_srt': undefined reference to `ControlziMonadziState_evalState_closure' collect2: ld returned 1 exit status ---------------------------------------------------

2 1

Re: Replacement for GMP: Update
by Peter Tanski 12 Aug '06

12 Aug '06

Einar, >> Summary: I finally settled on modifying OpenSSL, since that would be >> the easiest to work with under GHC's hood (plain C code, not C++). I >> have to: > > Does this cause license problems? > > I think OpenSSL license had the advertising clause which means > problems if Haskell programs want to link against GPLed > libraries too. The relevant portion of the OpenSSL license (reproduced in all source code and documentation) is: * The licence and distribution terms for any publically available version or * derivative of this code cannot be changed. i.e. this code cannot simply be * copied and put under another distribution licence * [including the GNU Public Licence.] (NOTE: this section is actually from the original SSLeay (1995-1998) license, but applies because OpenSSL is distributed under a dual license.) Otherwise the OpenSSL license is a BSD-style license that gives the authors credit (BSD2 and BSD3 do not require names in all the source code and documentation, only the distribution). The copying restriction means that I cannot simply take the BN library out of the OpenSSL distribution and put that code under the GHC license; it does not mean that I cannot distribute the OpenSSL code (still under the OpenSSL license) along with GHC code. For developers (commercial or open source), the OpenSSL license only mentions redistribution of the OpenSSL code in binary form (paragraph 2). In this context "binary form" means the complete program binary, not partial binary as with statically linking to a library, so developers of GHC programs would *not* have to include the whole OpenSSL&SSLeay license in their source code. Paragraph 6 of the OpenSSL (1998-2005) license states that: * 6. Redistributions of any form whatsoever must retain the following * acknowledgment: * "This product includes software developed by the OpenSSL Project * for use in the OpenSSL Toolkit (http://www.openssl.org/)". All developers would have to do is include the acknowledgment stated above. On the other hand, the OpenSSL FAQ at <http://www.openssl.org/support/ faq.html#LEGAL2> mentions that some "GPL" programs do not allow binary combination (static linking) or interoperation (dynamic linking) with OpenSSL. Honestly I have not seen any "GPL" licenses like this. The GNU GPL version 2, at <http://www.gnu.org/licenses/ gpl.html>, does not mention OpenSSL, nor does version 1, nor does the GNU LGPL, at <http://www.gnu.org/licenses/lgpl.html>. In any case, I think it would be more of a restriction to someone *using* the OpenSSL program, not a developer. Best regards, Peter P.S. On a related note, OpenSSL's BN library does not contain any patented algorithms.

6 10

recursive import problems
by Serge D. Mechveliani 12 Aug '06

12 Aug '06

Dear GHC team, Can you, please help me with recursive import? I have a certain large project which compiles under ghc-6.4.1. Then I add a certain import declaration to the module Dumatel.Reduce, which leads to import loop, and `make' (under Cabal) reports --------------- Module imports form a cycle for modules: Bool0 imports: Dumatel.Reduce Parse2 Parse1 Prelude4 Prelude3_1 Prelude3 Prelude2_0 Prelude2 Prelude1 Prelude00 Prelude0 List Maybe Data.Map Data.Set Dumatel.Reduce imports: Prelude5 Debug.Trace NewUnify Parse1 Prelude4 Prelude3_0 Prelude3 Prelude2 Prelude1 Prelude00 Prelude0 List Maybe Data.Map Data.Set Prelude5 imports: Bool0 Dumatel.Reduce Parse1 Prelude4 Prelude3_1 Prelude3_0 Prelude3 Prelude2_0 Prelude2 Prelude1 Prelude00 Prelude0 List Char Maybe Data.Map Data.Set -------------------- I am trying the .hs-boot modules and {-# SOURCE #-} import for each of the above 3 modules. And fail, because the situation is much more complex than in the GHC guide examples. Does GHC provide a tool for automating producing a .hs-boot module from .hs module? Because this is tedious to convert from a large .hs module by hand. It tried short variants for .hs-boot with exporting a single type instead of many items in .hs. But this was not sufficient. If I send you the source project, could you show me the simplest .hs-boot modules and import pragmas that enable to `make' it ? ----------------- Serge Mechveliani mechvel(a)botik.ru

1 0