- Nhc-bugs - Haskell.org

Installation Problem
by Dieter Büttner 29 Jul '05

29 Jul '05

Hello, I downloaded an executable tar abd zipped version of nhc98. I unzipped and tared, changed into the new nhc98 directory. and started ./configure and then make install. After that I tried to translate a little haskell program with nhc98 test.hs but got error messages: "/usr/local/bin/nhc98: line 131: /usr/local/lib/nhc98/x86_64-Linux/hmake-PRAGMA: File or Directory not found" or "/usr/local/bin/nhc98: line 131: /usr/local/lib/nhc98/x86_64-Linux/nhc98comp: File or Directory not found" Indeed, those files didn't exist in the given directories. Any idea, what went wrong? Thank you for an answer Dieter

2 1

Misleading error message for underscore patterns
by Tony Sloane 14 Jun '05

14 Jun '05

Not really a bug, but a confusion from a misleading error message. This was seen using nhc98 1.18 on Mac OS X 10.4.1. The following code contains underscore identifier patterns which are accepted by nhc98 when the -98 option is given: foo _2a = 1 bar (T a _2b) = 1 (I don't have the H98 definition handy but I'm presuming that nhc is correct in allowing these patterns in H98.) However, when -98 is not present, errors are reported that are misleading in that they refer to the "foo" and "bar" tokens, not the bad patterns, and they don't complain about the patterns at all: 3:1-3:3 Found foo but expected a {-end-of-definition-or-EOF-} 5:1-5:3 Found bar but expected a {-end-of-definition-or-EOF-} As the bar example illustrates, the actual error could be arbitrarily far away from the position in the message. Tony

2 1

No default for MutualRecursion.PP
by Christian Maeder 07 Jun '05

07 Jun '05

Hi, The program below is accepted by "ghc -fno-monomorphism-restriction" and "hugs +98". Why can't nhc98 (v1.18) handle it? Tossing in something like: instance PP () default (()) looks a bit stupid in order to become "portable", doesn't it? If I replace PP by Show (and pp by show and omit the class declaration) nhc98 has no problem. Also if I replace one occurrence of fA or fB directly with pp it goes through. Cheers Christian > hmake -nhc98 MutualRekursion.hs nhc98 -c -o MutualRekursion.o MutualRekursion.hs ====== Errors after type inference/checking: No default for MutualRecursion.PP at 23:1-24:18.(174,[(148,187)]) No default for MutualRecursion.PP at 19:1-20:18.(183,[(148,186)]) -- -------------------------------------------------------------------- module MutualRecursion where data A a = A a | BasA (B a) data B a = B a | AasB (A a) class PP a where pp :: a -> String instance PP a => PP (A a) where pp (A a) = pp a pp (BasA b) = pp b instance PP a => PP (B a) where pp (B a) = pp a pp (AasB b) = pp b fA :: PP a => A a -> String fA (A a) = pp a fA (BasA b) = fB b fB :: PP a => B a -> String fB (B b) = pp b fB (AasB a) = fA a

1 0

Integer divide bug in nhc98 causes runtime segfault
by Martin Guy 05 Jun '05

05 Jun '05

Hi again! Last november I reported a bug in the nhc runtime code that bombs the program out with Segmentation fault during garbage collection. Since then there have been 2 more releases of nhc98, but the Integer divide bug has not gone away nor even made it into the known bug list. I've been staring at the C code for several days now without finding it yet, so here I am reporting it again to the list. A few example programs that segfault are attached. Any other code that makes repeated calls to the div, mod, quot or rem on Integers bombs equally well :) All run fine under hugs, and their making nhc generate faulty code has been confirmed by one other person on the net (so it isn't my C compiler!). I also have a example program (an earlier version just after its translation from Miranda) that runs correctly under hugs but gives pi as "3.13785478481Segmentation fault" with hugs98 v1.16 and 1.18 (though it's 65K - hardly a test case). I'm not an nhcbugs list reader (sorry, too poor to afford it!) - please CC any replies. Cheers! M

1 0

Time profiling on Mac OS X
by Tony Sloane 18 May '05

18 May '05

I've been trying a simple test of time profiling on a Mac OS 10.3.9 machine. It fails during linking on the program: main = putStrLn "Hello World!\n" See the attached typescript. Is this a known problem? I note that the libraries on this machine can't be ranlib -s'ed since they have multiple definitions of some symbols. Is this related or can I ignore these warnings? Tony Sloane

2 1

unicode support
by Antonio Regidor García 09 May '05

09 May '05

Hi, I'm trying to implement unicode support for nhc98, but I'm confused by this code: --------------- module Prelude(Enum(..)) where instance Enum Char where enumFrom n = enumFromTo n (toEnum 255) enumFromThen n n' = enumFromThenTo n n' (toEnum 255) toEnum c = toEnum c -- MAGIC fromEnum c = fromEnum c -- MAGIC --------------- What does this "MAGIC" mean? Do the right hand sides use the prelude functions of the compiler that nhc is built with, or the internal representation of chars is defined in other place? Regards, Antonio Regidor García ______________________________________________ Renovamos el Correo Yahoo! Nuevos servicios, más seguridad http://correo.yahoo.es

2 2

The nhc98 package for NetBSD (update to 1.18)
by Jesper Louis Andersen 02 May '05

02 May '05

I assume this is the correct mailing list to send a report of my workings on nhc98 to get an updated NetBSD port. It is sent CC to since he is the current nhc98 maintainer of the NetBSD package in the NetBSD pkgsrc tree. I am reporting on a number of issues with the package. The rationale for this report is to inform various parties of the problems I have encountered, so they can be thought of in later releases. As a consequence the emphasis of this mail will target the build system, the system infrastructure at york(!) and interfacing with NetBSD in general. Communication often makes the difference between a success and failure when porting software. I also hope Krister does not feel I have hi-jacked his work. A NetBSD package is a Makefile, along with some helping files. The makefile accepts a number of targets which fetches the source code, configures it for NetBSD, patches the sources, compiles the program and installs the program in such a way it can be easily deleted again. The approach is different from binary package distribution, but is equal to other Operating systems, noteably the *BSD family and Gentoo Linux. The first problem with updating the package is the retrieval phase. This is done by a program, ftp(1), which will try to fetch the file ftp://ftp.cs.york.ac.uk/pub/haskell/nhc98/nhc98src-1.18.tar.gz. Unfortunately, this fails because the york ftp-server and ftp(1) agrees on entering Extended Passive Mode (EPRT and EPSV as per RFC 2428) which I then suspect the york firewall to disallow. Note that programs such as FreeBSDs fetch(1) and wget(1) of GNU chooses not entering EPRT/EPSV mode; thus having no problems fetching the source code. The direct solution to this problem would be to fix the firewall. We could add a workaround into the packages fetching process, but this has a number of implications. First, it is not the correct solution to the problem. Second, the fetching is generalized and not necessarily done by ftp(1). Altering the options given to the ${FETCH_COMMAND} is not portable if another retrieval program is chosen. The obvious way to bootstrap nhc98 would be by gcc. Unfortunately, NetBSDs package tree can select the compiler to be something other than gcc. The reason is the portability of NetBSD. If running on another architecture than, say i386, one could choose a native compiler over gcc for compilation. As a consequence $(CC) is set to be ``cc'' when compiling the nhc98 Makefile; not ``gcc'' as expected. My crude solution were to oversteer $(CC) and substitute it for ${BUILDCOMP}, see patches further down. This leads to my question: Can nhc98 be compiled by anything else than nhc98? Have you tried? I can alter the package to constrain itself to certain compilers in the case of impossibility. The next problem I ran into in the upgrade process from 1.16 is a number of old patches we put on the source code before compiling. The patches are all laid out at the following url: http://cvsweb.netbsd.org/bsdweb.cgi/pkgsrc/lang/nhc98/patches/ I will write [aa] for patch-aa, [ab] for patch-ab and so forth. Here is my findings. [aa] concerns an addition of $LDFLAGS to the curses compilation. This is done because the NetBSD package system can set further linker directives in this variable and we need to honour them. [ab] concerns an errornous way of calling abort(3). This is fixed in 1.18 and can go away. [ac, ad, af, ag] are all patches from the nhc98 homepage for nhc1.16. These are obviously not needed for 1.18 and can go away. [ae] does not seem to apply anymore. I am on a -current NetBSD and had no build problems. [ah] adjusts for $(CC) by substituting it for $(BUILDCOMP). I think this is the correct solution, so I have adapted that one. That is all. Unless I get complaints from any department - I will submit a change to the NetBSD bug-tracker in a couple of days. -- jlouis

3 2

RE: [nhc-bugs] Exporting class methods
by Simon Marlow 19 Apr '05

19 Apr '05

On 19 April 2005 12:15, Malcolm Wallace wrote: >>>> Control.Monad.Reader (Reader, runReader, ask, local) >>>> Control.Monad.Error (instance MonadPlus IO) >> >> These have H98 implementations, as far as I'm aware. > > The mtl package requires multi-param classes and functional > dependencies. Where are the H'98 versions? oops, you're right. But instance MonadPlus IO is ok. Cheers, Simon

1 0

RE: [nhc-bugs] Exporting class methods
by Simon Marlow 19 Apr '05

19 Apr '05

On 15 April 2005 17:26, Malcolm Wallace wrote: > Yes, pretty-much all of these rely on significant extensions to > Haskell'98. Nah, many of these are implementable without changing the compiler: > Ian Lynagh <igloo(a)earth.li> writes: > >> Control.Exception (bracket_, catch, block, unblock >> catchJust, Exception(IOException) >> catchDyn, throwDynTo, bracket, ioErrors, finally ) Needs exception support, of course >> Control.Concurrent ( ThreadId, myThreadId ) >> Control.Concurrent.MVar (MVar, modifyMVar_, swapMVar, newMVar) Easy (no forkIO) >> System.Posix.IO (fdToHandle) nhc98 must have something like this already. >> System.Posix.Types ( Fd(Fd) ) >> System.IO ( hPutBuf, hGetBuf, hIsTerminalDevice ) -- claims to be >> portable! System.Posix (getFileStatus, setFileMode, unionFileModes, >> ownerExecuteMode, groupExecuteMode, fileMode >> EpochTime, modificationTime, sleep, FileOffset, >> fileSize, setFileTimes, epochTime) All implementable using the FFI only >> Control.Monad.Reader (Reader, runReader, ask, local) >> Control.Monad.Error (instance MonadPlus IO) These have H98 implementations, as far as I'm aware. >> Also, unboxed arrays of Int32, Bool, Int and STUArrays of Int and >> Bool (plus runST, mutable array functions, unsafeFreeze, etc). Ok, MArray and IArray need MPTC, but you don't actually need unboxed arrays (Hugs provides these). Cheers, Simon

2 1

Exporting class methods
by Ian Lynagh 15 Apr '05

15 Apr '05

Hi, Andother problem with both nhc98 1.16 and 1.18. $ cat Q.hs module Q where import Data.Bits (shiftL) $ nhc98 -c Q.hs ====== Errors when renaming: Identifier shiftL used at 4:19-4:24 is not defined. If I change the second line to import Data.Bits (Bits(shiftL)) then it compiles fine. ghc and hugs are both happy with it either way. My reading of the report suggests the first form should be allowed (and hence this is a bug in nhc98): http://www.haskell.org/onlinereport/modules.html says: --------8<----------------8<----------------8<-------- 5.2 Export Lists Entities in an export list may be named as follows: 1. A value, field name, or class method, whether declared in the module body or imported, may be named by giving the name of the value as a qvarid, which must be in scope. 5.3.1 What is imported Exactly which entities are to be imported can be specified in one of the following three ways: 1. The imported entities can be specified explicitly by listing them in parentheses. Items in the list have the same form as those in export lists, except qualifiers are not permitted and the `module modid' entity is not permitted. --------8<----------------8<----------------8<-------- By the way, these bugs are coming from looking at compiling darcs with nhc98, but there are also a number of modules/functions missing. Are you interested in a list of these? Here is a (probably incomplete) one: Control.Concurrent.MVar System.IO (hPutBuf, hGetBuf) Control.Exception (catch, block, unblock, bracket_) Thanks Ian

2 4