July 2001 - Glasgow-haskell-users

Re: Rationals
by Sigbjorn Finne 30 Jul '01

30 Jul '01

Amanda Clare ajc(a)aber.ac.uk writes: > > Does anyone know how I use Rationals in ghc? In nhc I can just write > > let a = b % c > > But ghc-5.00.2 gives the error message "Variable not in scope: `%'" > > I've looked through the documentation and can't find any reference to a > Rational library, or any other library that might have it. The package > "std" has a Ratio.hi in the lib/imports directory (std isn't described > in the documentation) but including this doesn't help - same error. As you noticed, you need to import Ratio to get at % -- GHC gets this right wrt the Haskell98 report. Hugs exports it from the Prelude (as does NHC, by the looks of it), which used to be right, but isn't in H98. This could be trivially fixed on the Hugs side, but I'm not sure the resulting breakage would be worth it. > The Haskell98 report says: > "In particular, the type Rational is a ratio of two Integer values, as > defined in the Rational library." That's a typo in the Report, it should say the "Ratio library" (Section 6.4, Simon). --sigbjorn

2 1

RE: help needed for adding isWHNF primop to 5.00.2
by Simon Marlow 30 Jul '01

30 Jul '01

> Hi GHC people, > > I would like to add a primitive to GHC 5.00.2 of the form: > > isWHNF :: a -> Bool > > I was able to do (something like) this a while ago in 4.06, > but have got > stuck trying to do the same with 5.00.2. One problem is that you didn't call it 'reallyUnsafeIsWHNF#' :-) (or you could give it an IO type, to follow the current trend). Simon

1 0

Re: help needed for adding isWHNF primop to 5.00.2
by Sigbjorn Finne 29 Jul '01

29 Jul '01

Bernard James POPE bjpop(a)cs.mu.OZ.AU writes: > > I would like to add a primitive to GHC 5.00.2 of the form: > > isWHNF :: a -> Bool > ... list of steps followed deleted... > > If anyone could point me in the right direction to solving my problem > I would be very grateful. Hi, the panic you're getting is due to the fact that you didn't add native code generator support for your primitive (compile your example with -fvia-C to sidestep it). Appended is an NCG implementation for it, which you might find helpful (the impl is based on dataToTagOp's). Whether your primop is precise enough to test WHNFness or not, I'll let others be the judge of.. hth --sigbjorn -- #define isHNFzh(r,a) r=(! closure_HNF((StgClosure *)a)) primCode [res] IsHNFOp [arg] = let res' = amodeToStix res arg' = amodeToStix arg arg_info = StInd PtrRep arg' word_32 = StInd WordRep (StIndex PtrRep arg_info (StInt (-1))) masked_le32 = StPrim SrlOp [word_32, StInt 16] masked_be32 = StPrim AndOp [word_32, StInt 65535] #ifdef WORDS_BIGENDIAN ty_info = masked_le32 #else ty_info = masked_be32 #endif not_a_thunk = StPrim IntEqOp [ StPrim AndOp [ty_info, StInt 0x10] , StInt 0x0 ] -- ToDo: don't hardwire the value of _THUNK from InfoTables.h assign = StAssign IntRep res' not_a_thunk in returnUs (\ xs -> assign : xs)

1 0

RE: wait(2)
by Simon Marlow 29 Jul '01

29 Jul '01

> This is interesting, but not what I want. I want something to wait on > a real, Posix, child, _process_!! Not a GHC thread. Ah. Oops :) There's one other way that Marcin didn't mention: wait for SIGCHLD, which can be done without blocking the whole process. Unfortunately there's no easy way at the moment to tell *which* child generated the SIGCHLD. I guess we should really have a process wait as a threading primitive, like OCaml does. I've been looking at FreeBSD's `kevent' stuff recently - they have a really nice way to wait for several different types of event, including file descriptors, processes and even when a file is modified (eg. tail -f). Does Linux or Solaris have anything like this? I had a quick look around but didn't find anything. Cheers, Simon

2 1

help needed for adding isWHNF primop to 5.00.2
by Bernard James POPE 28 Jul '01

28 Jul '01

Hi GHC people, I would like to add a primitive to GHC 5.00.2 of the form: isWHNF :: a -> Bool I was able to do (something like) this a while ago in 4.06, but have got stuck trying to do the same with 5.00.2. Here's what I have tried: After reading ghc/compiler/prelude/primops.txt: - added primop IsHNF "isHNF#" GenPrimOp a -> Int# with strictness = { \ arity -> StrictnessInfo [wwLazy] False } to ghc/compiler/prelude/primops.txt - added isHNFzh to ghc/lib/std/PrelGHC.hi-boot - added #define isHNFzh(r,a) r=(! closure_THUNK((StgClosure *)a)) to ghc/includes/PrimOps.h {- although I think this should be the same as: ??? #define isHNFzh(r,a) r=(closure_HNF((StgClosure *) a)) -} - I think I need to do something in: ghc/compiler/nativeGen/StixPrim.lhs but I have no idea what, I looked at the code in there and got scared, so I left it alone, although I suspect that was a bad idea. - I then did: make boot; make all - everything compiled ok, but I get some serious problems when I try to use my new primitive: \begin{code} module Main where import GlaExts import PrelGHC main = print $ g g :: Bool g = fromUnboxedIntAsBoolean (isHNF# ()) fromUnboxedIntAsBoolean :: Int# -> Bool fromUnboxedIntAsBoolean x = case x of 1# -> True _ -> False \end{code} When I compile this code with my newly built compiler using: ghc -ddump-types -fglasgow-exts -package lang I get: ==================== Interface ==================== TYPE SIGNATURES Main.fromUnboxedIntAsBoolean :: PrelGHC.Int# -> PrelBase.Bool Main.g :: PrelBase.Bool Main.main :: PrelIOBase.IO () {-# Generic type constructor details #-} ghc-5.00.2: panic! (the `impossible' happened, GHC version 5.00.2): getRegister(x86,unary primop) (Prim isHNFzh PrelBase.Z0T{-70-}_closure) Please report it as a compiler bug to glasgow-haskell-bugs(a)haskell.org, or http://sourceforge.net/projects/ghc/. The file ghc/docs/rts/rts.tex talks of a predicate "isWHNF", however, I cannot seem to find it anywhere, and suspect that it does not exist anymore. If anyone could point me in the right direction to solving my problem I would be very grateful. Regards, Bernie. PS Unfortunately I clobbered my version for 4.06 with isWHNF added, and now I can't get it to work anymore.

1 0

hopengl
by Johannes Götz 28 Jul '01

28 Jul '01

Hi! Concerning the Hopen-GL-distribution 1.00 on win2000, latest cygwin, greencard 2.0 The way the target is given in the lib-Makefile doesnt work (at least with my version 2.0(1?) of greencard). I changed the options line in /HOpenGL-1.00/lib/Makefile from GREENCARD_OPTS = -t ghc --name-mangling-scheme=classic to GREENCARD_OPTS = --target ghc --name-mangling-scheme=classic (Also possible with my version of greencard is "-tghc" without a space between) That seems to work. The first file here, GLUT_CBGlobal.gc, runs through the parser which correctly generates the output file GLUT_CBGlobal.hs... but then another error occurs while parsing GLUT_CBWindow.gc: bash-2.05$ make depend /cygdrive/c/ghc/ghc-4.08.2/bin/green-card --target ghc --name-mangling-scheme=cl assic GLUT_CBGlobal.gc -o GLUT_CBGlobal.hs /cygdrive/c/ghc/ghc-4.08.2/bin/green-card --target ghc --name-mangling-scheme=cl assic GLUT_CBWindow.gc -o GLUT_CBWindow.hs "GLUT_CBWindow.gc", proc. spec "", line 285: Parse error: Enum, Bounded) CUnsig nedInt [ % GLUT_JOYSTICK_BUTTON_A, % GLUT_JOYSTICK_BUTTON_B, % GLUT_JOYS make: *** [GLUT_CBWindow.hs] Error 1 I begin wondering if it is actually possible to run HOpenGL on my computer. Hm. The HOpenGL-website says that the library has been heavily tested under Linux AND Windows. But it seems that nobody with my configuration has ever compiled the source package (as far as I know, there are no binary packages available, are they?). Has anybody else tried to compile version 1.00 under win2000/cygwin/greencard2.0(1?)? Does someone know where else I can get the necessary libraries? Thanks very much. Johannes

1 0

building hopengl problem
by Johannes Götz 27 Jul '01

27 Jul '01

Hi there! Somebody help me please! I have a problem concerning HOpenGL. On my win2000 machine I installed the latest cygwin and ghc 4.08.2. I downloaded the GreenCard-installer and copied the greencard-executable to the ghc-bin-directory. I then unzipped the current HOpenGL distribution 1.00 to the ghc-directory and untared it. The configure-script didnt report any errors. Next I tried "make depend", according to the installation instructions, but the script doesnt work. Does somebody know what Im doing wrong? This is what I get: bash-2.05$ make depend ------------------------------------------------------------------------ Recursively making `depend' in lib examples ... PWD = /cygdrive/c/ghc/HOpenGL-1.00 ------------------------------------------------------------------------ ------------------------------------------------------------------------ make depend - --unix; in /cygdrive/c/ghc/HOpenGL-1.00/lib ------------------------------------------------------------------------ /cygdrive/c/ghc/ghc-4.08.2/bin/green-card -t ghc --name-mangling-scheme=classic GLUT_CBGlobal.gc -o GLUT_CBGlobal.hs Unrecognised target: ( `ffi', `hugs', `ghc' and `nhc' recognised.) /cygdrive/c/ghc/ghc-4.08.2/bin/green-card -t ghc --name-mangling-scheme=classic GLUT_CBWindow.gc -o GLUT_CBWindow.hs Unrecognised target: ( `ffi', `hugs', `ghc' and `nhc' recognised.) [...] Thanks very much for your answers. Johannes jgoetz(a)polirami.de

1 0

Re: wait(2) [Slight correction]
by Mike Gunter 27 Jul '01

27 Jul '01

[My previous message omitted a word ("records"). I should take one more look at these things before sending them.] Scsh (Scheme Shell) has a mechanism to turn the Unix process interface into something reasonable. Instead of PIDs, fork returns a process object. The SIGCHLD signal handler reaps processes and records their exit status. The program can retrieve the exit status using the process object. When the process object is GCed, the status can be as well. This seems be right for almost all code (and reaping policy can be changed for the remaining code.) The details of Scsh's mechanism can be found in the manual ftp://ftp-swiss.ai.mit.edu/pub/su/scsh/scsh-manual.ps on pages 55-58. mike

1 0

Re: wait(2)
by Mike Gunter 27 Jul '01

27 Jul '01

Scsh (Scheme Shell) has a mechanism to turn the Unix process interface into something reasonable. Instead of PIDs, fork returns process objects. The SIGCHLD signal handler reaps processes and their exit status. The program can retrieve the exit status using the process object. If the process object is GCed, the status can be as well. This seems be right for almost all code (and reaping policy can be changed for the remaining code.) The details of Scsh's mechanism can be found in the manual ftp://ftp-swiss.ai.mit.edu/pub/su/scsh/scsh-manual.ps on pages 55-58. mike

1 0

gcc & GHC?
by hooh pxw 27 Jul '01

27 Jul '01

h(eye), ghc for win32 isn't cool. it remind me java in gcc 3.0. is there any ghc project as gcc front-end? then, we all happy on the technology of the GCC 3.0. mercury, also the front end of the gcc. http://gcc.gnu.org/frontends.html i am end user, so don't get me heavy/serious! Ah! john backus's FP.. sorry for random/ugly keyboard typing!! __________________________________________________ Do You Yahoo!? Make international calls for as low as $.04/minute with Yahoo! Messenger http://phonecard.yahoo.com/

2 1