Re: Small problems with ghc-2.02 binary distribution for linux

1997-03-28 Thread Sigbjorn Finne
Sigh, the hbc syslib was unintentionally left out of the `sequential' bundle ( ghc-2.02-i386-unknown-linux.tar.gz ), hence the linker complaint. (I'm mildly surprised that you got this far, though). I've updated the bundle and put up a new copy at Glasgow together with a recompiled conc. binary

Re: Profiling error on -hD

1997-04-08 Thread Sigbjorn Finne
Jon Mountjoy writes: Hello People, The -hD profile on ghc-2.02 seems to break the hp2ps. The other profiles seem to work. I suspect it is because somewhere (on line 14) the words "FOREIGN OBJ" are outputted instead of "ForeignObj", because if I hack the former with the latter it seems

Re: Another Interface-file parse error

1997-04-08 Thread Sigbjorn Finne
Jon Mountjoy writes: Hello People, Here is another interface-parse file error. I don't know if it is fixed due to my other bug report or not. Running Solaris ghc-2.02. Compile Dt.hs with:ghc-2.02 -fglasgow-exts -prof -auto -c -O Dt.hs Compile Main.lhs with: ghc-2.02 -fglasgow-exts

Re: ghc-0.29 -prof -concurrent

1997-03-11 Thread Sigbjorn Finne
Alex Ferguson writes: Yes, and no. I reported my inability to get conc and prof to work together on _either_ 0.29 or 2.01, quite a while ago. WDP said the latter likely hadn't been made to work at all; I got no real feedback on the former, so I gave up on both. Fixed in 2.02, I'm

Re: Typo in Socket.lhs

1997-04-10 Thread Sigbjorn Finne
Sven Panne writes: On the hard road of building ghc-2.02 from source, I stumbled across an old friend of mine::-( Change line 110 of fptools/hslibs/ghc/src/Socket.lhs: listeOn (UnixSocket path) = to listenOn (UnixSocket path) = It's a rather traditional bug: It survived at

Re: Building ghc-2.02 from C (.hc) files

1997-04-28 Thread Sigbjorn Finne
Klaus Georg Barthelmann writes: Dear GHC team, yes, I would like to take the route via C files. My machine is a Sun3 (sun-m68k), which obviously is no longer supported. Building version 2.02 from source would be possible theoretically (I have 0.29 running), but would require a week or so, I

Re: -fno-warn-incomplete-patt

1997-05-12 Thread Sigbjorn Finne
[EMAIL PROTECTED] writes: Does the -fno-warn-incomplete-patterns flag actually work? Certainly my use of it in ~satnam/research/vhdl2edn/Makefile (at OGI) does not seem to work for VHDLParser.hs (a happy generated file). Sigh, the command line option is recognised but not put to any

Re: A list of failures

1997-05-13 Thread Sigbjorn Finne
Ralf Hinze writes: [] * decided to build the system in the source tree [though not recommended, but I didn't find `lndir'.] I think Sven answered this one already, I'll change the docs to mention this upfront (i.e., look in fptools/glafp-utils/lndir ) [...] * started with

Re: Installation problems

1997-05-13 Thread Sigbjorn Finne
Arne John Glenstrup writes: To The Haskell-Hackers, I am trying to install ghc version 2.02 on sun4c architechture. 1 In file fptools/literate/mkdependlit I find the following lines at the top:

Re: Major Installation problem, help!

1997-05-13 Thread Sigbjorn Finne
Arne John Glenstrup writes: In my attempts to install the ghc2.02 on the sun4c architechture I have run into the following major problem: 1 When I `gmake all' in the fptools/ghc/lib directory, fairly large (up to 3Mb) temporary files are created and then hsc is run, invoking GNU

Re: System libraries in 2.03

1997-05-19 Thread Sigbjorn Finne
Sven Panne writes: The interface files for the system libraries ghc, hbc, posix and contrib seem to be installed in the wrong place: The lib directory (containing hsc and friends) does not contain a hslibs subdirectory. The ghc, hbc,... directories directly reside in lib. Another strange

Re: GHC2.03: stranal/SaAbsInt.lhs line 664: incomplete pattern(s)

1997-05-20 Thread Sigbjorn Finne
Fail: "stranal/SaAbsInt.lhs", line 664: incomplete pattern(s) to match in function "absApply" Running Solaris 2.5.1 I enclose the verbose listing below. If you need more, please yell. This is a new one (to me anyway), so if you could send the sources where this one shows up, that

Re: Installing ghc2.03

1997-05-21 Thread Sigbjorn Finne
Alex Ferguson writes: Hi again, Vladimiro. VS: Exactly. It fails compiling the parser. I'm trying to compile using (the binaries of) ghc-2.02. Is there something intrinsecally wrong with this? Ah-hah! Yes. You need ghc-0.29 as the builder, I believe. I don't know why this might

Re: ghc2.03: install problems with: install -c , ...

1997-05-21 Thread Sigbjorn Finne
Doing an install on ghc-2.03 fails. I *think* that I successfully patched the makefile with patch ghc-2.03-make.patch and configure etc. (it failed for one block it said) [...] Doing a gmake -n install in the directory fptools/ghc/utils/mkdependHS I see the lines: echo

Re: ghc-2.03 and green-card

1997-05-25 Thread Sigbjorn Finne
[EMAIL PROTECTED] writes: I have just built ghc-2.03 with green-card 0.9. Whereas green-card used to work fine with ghc-2.02, I now get: ghc: consistency warning: minor version not 1: 080525b0 t cc.StdDis.lhs.36.0.. ghc: consistency error: major version not 32: 080537fd t

Re: Installing ghc-2.03 on HP-UX10 from source

1997-05-23 Thread Sigbjorn Finne
Thanks for the feedback, I'm puzzled as to why you have to grapple with altzone though, since the configure script should look for tm_zone in (struct tm), what does the relevant section of the configure script say when running it? --Sigbjorn Sven Panne writes: Here my promised

Re: Small bug when compiling ghc-2.04 with ghc-2.02

1997-06-08 Thread Sigbjorn Finne
Stefan Westmeier writes: When compiling the ghc-2.04-src.tar.gz (without hi and hc files) with ghc-2.02 "gmake boot" gives the following error. It does not appear when removing the "," from withing the SLIT of this piece of code from TcSimplify.lhs from line 728: \begin{code} reduceErr

Re: Pattern-matching error within GHC

1997-06-11 Thread Sigbjorn Finne
Thanks for chasing this down, Einar - this (mis)behaviour is reproducable. Fixed in the next release. --Sigbjorn The module that caused GHC to fail attempted to reexport a name from a module that was imported as a qualified module, i.e.: module Silly ( Posix.dupTo,

Re: Another ghc-2.04 compiling problem

1997-06-13 Thread Sigbjorn Finne
[Trimmed the References header, original message referred to articles in another thread] Meurig Sage writes: The earlier problem I had with ArrBase disappeared when I used -monly-2-regs. Thanks. But I've got another problem now... When trying to compile a concurrent profiled version of

Re: Makefile bug in ghc-2.04-patchlevel1

1997-06-10 Thread Sigbjorn Finne
Sven Panne writes: There is a small typo in diff fptools/literate/Makefile, resulting in a syntax error from the shell. Patch attached. Urk, I thought I managed to keep that one out of the distribution (the diffs contain the Right Thing here). ghc-2.04-pl2 should be out before the day is

Re: Command line tribulations.

1997-07-18 Thread Sigbjorn Finne
Alex Ferguson writes: Hi all. Some problems I had recently with ghc opts, while trying to twiddle the heap usage of ghc on happy output (see ghc-users, passim). Thanks for the feedback - here's my take on various cmd-line matters: Pedantic observation: contra the user guide,

Re: trying to build ghc-2.04 for linux, i386

1997-07-03 Thread Sigbjorn Finne
S.D.Mechveliani writes: I am trying to build ghc-2.04 on linux, i486 ... using ghc-0.29 installed earlier. The questions arise. Hi, sorry about the delay.. --- I downloaded ghc-2.04-src-pl2.tar.gz,

Re: Bug in ghc perl script

1997-07-03 Thread Sigbjorn Finne
Pradyut Shah writes: on sparc-sun-solaris2, the ghc perl script (fptools/bin/sparc-sun-solaris2/ghc-2.02/ghc) does not work with the -E option. It seems to generate the assembler code instead. Yep, confusing - you'll have to add -fvia-C to make it behave. Thanks for the report,

Re: Pattern-matching error within GHC

1997-06-11 Thread Sigbjorn Finne
While compiling a larger configuration of GHC code (30 KLOC's) with the brand new compiler 2.04 on a Ultra-Sparc, I got the following *** Pattern-matching error within GHC! This is a compiler bug; please report it to [EMAIL PROTECTED] Fail: "rename/RnNames.lhs", line 453:

Re: bind error in decl (2)

1997-06-12 Thread Sigbjorn Finne
Ralf Hinze writes: Dear bug chasers, ghc's latest version sometimes misses conflicting function definitions. Consider the following *wrong* program: Hi, thanks for the report - this is due to a parser problem of signalling when a toplevel value declaration ends. To fix, apply the

Re: building 2.04

1997-06-14 Thread Sigbjorn Finne
Mariano Suarez Alvarez writes: Building ghc-2.04-pl2 on a i586 under linux with a `gcc --version` 2.7.2.1 in /fptools/ghc, compiling ArrBase.lhs, ghc-0.29 dies saying /tmp/ghc29172.hc:953: fixed or forbidden register was spilled. This may be due to a compiler bug or to impossible asm

Re: Command line problems with -ohi

1997-07-17 Thread Sigbjorn Finne
make -k ghc -c -O -odir . -ohi Main.hi ../Main.hs ghc: input file doesn't exist: Main.hi ghc: no argument following -ohi option ghc: You can't use -o or -ohi options if you have multiple input files. Perhaps the -odir option will do what you want. Usage: For basic information,

Re: Command line tribulations.

1997-07-18 Thread Sigbjorn Finne
Alex Ferguson writes: Neither -optCrts-Sstderr nor -Rgc-stats seem to do anything useful; the file created by the latter contains only a ghc command line, and the former seems to do nothing at all. Hmm..need more details, I'm not able to reproduce this behaviour. Transcript

Re: A bug in makefiles for 2.04-pl2 ?

1997-06-12 Thread Sigbjorn Finne
Mariano Suarez Alvarez writes: after configuring and `making boot', when I `make all', when make gets to /fptools/ghc/compiler, I get an error I can't fix because I really haven't understood the Makefile structure. "parser/U_binding.hs", line 6, column 22: can't find interface (.hi)

ANNOUNCE: ghc-2.04-patchlevel1 available

1997-06-09 Thread Sigbjorn Finne
A patch release of ghc-2.04 is now available, which fixes up a couple of build-related gaffes in the 2.04 distribution issued on Fri 6/6. To download, please refer to the URL ftp://www.dcs.gla.ac.uk/pub/haskell/glasgow/README.html (2.04-2.04-pl1 diffs are also available) Bug reports to

Re: small bug when compiling ghc-2.04 with ghc-0.29

1997-06-08 Thread Sigbjorn Finne
Stefan Westmeier writes: The following piece of code is from fptools/ghc/compiler/Makefile. When doing "gmake boot" the goal hi-boot is not found and some hi-files are not produced from the hi-boot files. [..Makefile snippet elided..] Oops, a vital target was omitted here for the 0.29

Re: profiling

1997-06-15 Thread Sigbjorn Finne
[redirected to g-h-u, as other GHC users may have a story or two to tell re: 2.04 profiling ] Marc van Dongen writes: Hi all, I have noticed that profiling in ghc-2.03 in combination with -O or -O2 is not possible because of some problems at the linking stage. Also it turns out that

Re: Profiling with -auto-all broken

1997-07-17 Thread Sigbjorn Finne
Sven Panne writes: Once again: The old and sad profiling story... Consider the following modules: ..source code elided.. Compiling and linking with -prof *and* -auto-all: panne@liesl ghc -c -prof -auto-all Foo.hs ghc: module version changed to 1; reason: no old .hi file

Re: panic! (the `impossible' happened): tcInstType:1

1997-06-17 Thread Sigbjorn Finne
Einar Wolfgang Karlsen writes: odd instance decl - is this really the file that shows up the bug? The file is the "correct" one - Yes. I played around with some of the ghc options. It seems as if it is -fglasgow-exts that is causing the problem. [...] Thanks, -fglasgow-exts gives

Re: Haskell and 3D API's

1997-07-30 Thread Sigbjorn Finne
Chris Burdorf writes: I was wondering if anyone out there has built any sort of Haskell interface to OpenInventor or OpenGL? I'm currently looking into the possibly of using one or the other with Haskell and would be interesting if anyone has any design insights they'd like to share. Hi,

Re: I just can't stand it any more

1997-07-30 Thread Sigbjorn Finne
Byron Cook writes: Yes, I agree 70% of the time I dont need lazy state -- and the performance cost _is_ big. I would claim that it is more like 98.7% :-) However, couldn't the same be said of general laziness ? Only a percentage of my code uses laziness. Perhaps, for performance

Re: re_match cannot handle large strings

1997-06-12 Thread Sigbjorn Finne
Einar Wolfgang Karlsen writes: I am using the MatchPS module for loads of tool integration tasks (peeking into the output from those tools looking for important signs and relevant messages). Unfortunately, MatchPS.matchPS seems to be quite restrictive wrt. the size of the string

Re: tcLookupTyVar error (Similar as last one? Don't know.)

1997-06-12 Thread Sigbjorn Finne
Marc van Dongen writes: Sorry if this is caused by exactly the same as my last tcLookupTyVar report. Since I can't tell, I'll report it anyway. panic! (the `impossible' happened): tcLookupTyVar:a_r6F Please report it as a compiler bug to [EMAIL PROTECTED] Fixed in

Re: Ian Collier: ghc-0.29-sparc-sun-sunos4 won't compile ghc-2.04

1997-08-01 Thread Sigbjorn Finne
Alex Ferguson writes: Hi again Ian. Look on the bright side: this time the actual compiler got build successfully... Now there's just the small matter of the libraries... [in the middle of a 2.05 build on a SunOS4 box] ==fptools== make way=p all; PWD =

Re: More questions on installing 2.05

1997-08-05 Thread Sigbjorn Finne
Ian Collier writes: .. I had to add GhcHcOpts=-H12m to my build file to get it to stop complaining (12m was a guess, but the default of 6m was definitely too small). I would have thought this would have been added as necessary by the developers to avoid me having to increase the size

Re: Buglist compiling 2.05 with 2.05 for Linux

1997-08-04 Thread Sigbjorn Finne
after succesfully compiling ghc 2.05 with ghc 0.29 on Linux (with Sven Pannes Patch) I now do try to compile 2.05 with 2.05 and found some small bugs. After doing this corrections it compiled without any further problems. Thanks, I've incorporated the fixes you had to make into the

Re: installing 2.05 problem

1997-08-08 Thread Sigbjorn Finne
[g-h-b listeners: I'm a bit behind on replying to msgs on the GHC lists at the moment - bear with me :)] Justin Cormack writes: Couldn't get Alex's suggestions to work ... so installed 0.29 to install it. However now I notice that when doing make boot I get: ... mkdependHS-1.2:

Bugs page moved

1997-08-08 Thread Sigbjorn Finne
The Frequently Encountered Bugs page has moved back to Glasgow. Old URLs should still work, but the main URL is now http://www.dcs.gla.ac.uk/fp/software/ghc/bugs.html --Sigbjorn

Re: 2.02 newtype problem

1997-08-12 Thread Sigbjorn Finne
Thanks for the report. Since you've apparently made the switch to 2.05, I guess you've given up on 2.02.. It does have a couple of `newtype' related problems, and this is one of them. (Works fine in 2.05). --Sigbjorn Justin Cormack writes: In ghc 2.02 I seem to be having a problem. The

Re: 2.05: make boot problems

1997-08-01 Thread Sigbjorn Finne
Peter Liljenberg writes: I've been trying to compile 2.05 with 2.02 on solaris2, but `make boot' fails when reaching ghc/compiler. It prints out this huge command line, and then only says "make: *** [depend] Error" and stops. I cut out the command line and inserted a -v: ghc -v -M

Re: GHC 2.04 List module: minor 1.4 conformance pbs and a bug (?)

1997-08-12 Thread Sigbjorn Finne
[Sorry for taking so long, had to wait for haskell.org to wake up again] Bjarte Mayanja Oestvold writes: In GHC 2.04, the module List does not conform to the Haskell 1.4 Library Report (but it should?): List needs to export the definitions `unionBy' and `intersectBy'. Thanks, you're

Re: bugg report

1997-08-24 Thread Sigbjorn Finne
Marc van Dongen writes: .. The following program: module Main( main ) where main = putStr (seq bot "1\n") return () where bot = error [] cannot be compiled: Main.lhs:1: Ambiguous context `{PrelBase.Eval tans}' `PrelBase.Eval tans' arising from use

Re: Problems with profiling when using -Ofile

1997-08-04 Thread Sigbjorn Finne
Olaf Chitil writes: I'm using the -Ofile option to be able to call a transformation I added to GHC. I think that Great stuff - when compiling with -prof, one STG options that has to be set is -fmassage-stg-for-profiling. This isn't done by the driver for -Ofile (though it should be), so

Re: GHC 2.05 bug

1997-08-30 Thread Sigbjorn Finne
David Barton writes: Thanks to Sigbjorn Finne's patch, the entire program now compiles. However, the produced program now dies with a segmentation fault. The text (which will probably help not one whit) occurs as follows: Segmentation fault caught, address = 38209318 IOT trap/Abort

Re: 2.05 finally made it

1997-08-28 Thread Sigbjorn Finne
Sven Panne writes: After applying any conceivable patch in this mailing list, ghc-2.05 built (almost) fine on our Linux- and HP-boxes. A few glitches: * As usual, /tmp is much too small on our machines. The hints from installing.lit: The quickest way around it is

Re: GHC 2.05 bug

1997-08-19 Thread Sigbjorn Finne
David Barton writes: Unfortunately, when I continue with the files in that directory (the one I sent you), I get another error. It says: ghc -H16M -c PlotExample.lhs PlotExample.lhs:8: Interface-file parse error:

Re: newline at end of -- comment

1997-09-26 Thread Sigbjorn Finne
[EMAIL PROTECTED] writes: While converting a Hugs program to work with ghc-2.02, I found that if the last line of a file is a -- comment, ghc requires me to add a newline to the end of the line. Thanks for the report - fixed in post-2.02 releases. --Sigbjorn

2.09: cygwin32 binary distrib available

1997-12-01 Thread Sigbjorn Finne
A 2.09 binary bundle for cygwin32 (x86) is now available via the ghc download page, http://www.dcs.gla.ac.uk/fp/software/ghc/download.html NOTE: Its huge (~12M), as it contains sequential, profiled, concurrent and prof-conc libs. --Sigbjorn

Re: Green Card and GHC 2.09 don't work together

1997-12-13 Thread Sigbjorn Finne
David Elworthy writes: I'm finding that Green Card 0.9 seems to have problems with GHC 2.09 (both with the cygwin32 version - I haven't checked the solaris one yet). Essentially functions which return IO types fail type checking. I had a .ss file which worked OK with 2.08 but does not type

Re: 2.09 binary for Alpha/Digital UNIX

1997-12-13 Thread Sigbjorn Finne
Alessandro Vernet writes: By the way, do you know if the quality of the resulting executable is better (faster, smaller, stingier) when using the native code generator or when using the C code generator? One of the main motivations for adding the native code generator(NCG) backends was

Re: ;

1998-01-15 Thread Sigbjorn Finne
Sven Panne writes: [ It seems that my original message only reached me (and not e.g. my colleagues here in Munich), so here's the second try...] Hmm, I saw your msg first time around. Let us know if the problem still persists. Quick hack for the Makefile: Change if test

Re: Profiling error

1998-01-20 Thread Sigbjorn Finne
Lennart Augustsson writes: I'm trying to (time) profile a large program. Here is what I get: spitfire% ./ravec +RTS -H30M -pT -K2M -RTS duty_common_fdca54_apc.crl Entered Forward_Ref 2b1a720: Should never occur! What now? Ouch, try switching over to the 2 space collector (-F2s.)

Re: ghc-2.10 panics at `deriving(Ord)'

1998-01-21 Thread Sigbjorn Finne
S.D.Mechveliani writes: ghc-2.10 cannot compile data D = D1 deriving Eq,Ord) while ghc-2. 08 can. Thanks for a fine report, fixed in the next release. To clarify, derivation of Ord instances on data types with a single nullary constructor will fail with 2.10. We didn't have a

Re: 2.07 binary installation

1997-10-03 Thread Sigbjorn Finne
This suggests to me that you for some reason don't have a #! added to the top of the driver. This turned out to be a bad glitch in the Makefile distributed with the binary bundles. Patch appended. (I've updated the binary distribs to avoid further trouble.) For people with source

Re: -recomp

1997-10-03 Thread Sigbjorn Finne
Alex Ferguson writes: BTW, did there use to be/is there still an option to tell -recomp to be explicit about why it had to do a recompilation? I found -hi-diffs-with-usages, which was of some help, but I thought there was a more directly useful one too, but I couldn't lay my hands on it.

Re: Interacting seq with strictness annotations

1997-10-03 Thread Sigbjorn Finne
Before 2.06, the definition of `seq' was seq x y = y i.e., not terribly useful :) --Sigbjorn Graeme Moss writes: (Usual apologies if this has been noted/solved since 2.05.) The following program (which sums [1..n] by brute force): import System(getArgs) data Test = Test !Int

Re: deriving Read generates incredibly bad code

1997-09-12 Thread Sigbjorn Finne
Julian Seward writes: I have a large enumeration type (example prog below) on which I have done "deriving Read". My problem is, the derived read for lists of these things runs incredibly slowly, so slowly that it useless for reading anything larger than the smallest quantity of text. I

Re: compiling errors using GHC2.10 on iris6 machine

1998-01-28 Thread Sigbjorn Finne
Tina Yu writes: Hi, I am trying out the GHC2.10 compiler on iris6 machine. My programs receive the following error messages. Can you help me to figure out what these mean ? Thank you very much !! ... GNU CPP version 2.7.2.2 [AL 1.1, MM 40] SGI running IRIX 5.x ...

Re: Pattern-matching strings.

1998-01-28 Thread Sigbjorn Finne
Simon Marlow writes: Alex Ferguson [EMAIL PROTECTED] writes: Is pattern-matching short strings (one or two characters) likely to be _vastly_ less efficient than matching against a single level of constructor? (Order of magnitude, plus.) Trying to make sense of some profiling

Re: error: (misc)

1998-01-29 Thread Sigbjorn Finne
Alex Ferguson writes: install-sh does a fine line in unhelpful error messages: well, error message singular, at any rate... for i in hp2ps; do \ /export/home/ferguson/ghc-3.00/build/install-sh -c -g ghc-admin -s $i /usr/local/bin; \ done hp2ps:error reading file This

Re: bug report 3.00

1998-02-06 Thread Sigbjorn Finne
Marc van Dongen= writes: While compiling some happy output with 3.00 I came across the following: NOTE: Simplifier still going after 4 iterations; bailing out. NOTE: Simplifier still going after 4 iterations; bailing out. panic! (the `impossible' happened): fun_result_ty:

Re: -H in ghc-3.00 installing

1998-02-09 Thread Sigbjorn Finne
S.D.Mechveliani writes: Osman Buyukisik [EMAIL PROTECTED] writes: I have obtained ghc 3.0 . I had 2.10. "make all" took 8 hrs! (P120-32MB-linux-2 .0.30-gcc-2.7.2.3.f.1). It failed during making of hslibs/contrib. Needed -H800 to compile not enough heap space. Which file can I edit

Re: The latest in a continuing saga......

1998-02-11 Thread Sigbjorn Finne
David Barton writes: /usr/src/ghc/tmp/ghc8242.hc:35545: structure has no member named `tm_zone' /usr/src/ghc/tmp/ghc8242.hc:36256: structure has no member named `tm_zone' /usr/src/ghc/tmp/ghc8242.hc:36304: structure has no member named `tm_gmtoff' /usr/src/ghc/tmp/ghc8242.hc:37074:

Re: Order of libraries

1997-10-14 Thread Sigbjorn Finne
Stephan Tobies writes: Another little problem: When linking programs which need additional libraries (say -lreadline -ltermcap) the ghc driver puts these after its own libraries when calling the linker. This is not desirable since the linker complains about unresolved names. Switchig

Re: GranSim/parallel versions of 2.08

1997-10-14 Thread Sigbjorn Finne
Sven Panne writes: The normal/profiling/concurrent versions of 2.08 compiled far too smoothly, so I was looking for new installation challenges. ;-) Alas, compiling with mg and mp added to GhcLibWays didn't work: * During the compilation of the _mg-versions of the libs the

Re: Channels and files

1997-10-16 Thread Sigbjorn Finne
Sven writes: -- Strange: hClose does *not* really close a file from fdToHandle!! -- No idea why. So we have to use this function... A bug, I'd say. Fixed in the next release - patch appended for interested parties. I'm not sure how well documented this feature is, but GHC-2.0x Handles have

Re: Problem with assembler on Digital UNIX

1997-12-01 Thread Sigbjorn Finne
Alessandro Vernet writes: We are trying to compile GHC 2.09 on our Digital UNIX 4.0b systems. We had no problem with GHC itself, but rather with the Digital assembler. When compiling "huge" files, our assembler crashes with a segmentation fault. We could solve this problem by spliting one

Re: Fixity of and =

1997-11-24 Thread Sigbjorn Finne
Thanks, typo in ghc's implementation of the prelude. Fixed in the upcoming 2.09. --Sigbjorn Nathan Charles writes: I expect this is already known but I couldn't find anything saying so. Both and = are supposed to be left associative in Haskell 1.4. However, ghc-2.08 treats them as right

Re: Building ghc-2.05 on an HP (or not).

1997-10-16 Thread Sigbjorn Finne
collect2: ld terminated with signal 11 , core dumped /bin/ld: (Warning) Inter-quadrant branch in simplCore/OccurAnal.o Anyone know: a) what the heck that means; and/or b) how to fix? Alternatively, I'll settle for c): any (other) feasible way of obtaining a working ghc-2.0[5-8] on our

Re: building ghc-3.00 using ghc-2.10

1998-02-19 Thread Sigbjorn Finne
Timothy Robin BARBOUR writes: I am trying to use ghc-2.10 (from a Glasgow binary bundle) to build ghc-3.00 from source, on a debian 1.3 linux box. .. configure and make boot go okay, but make all ends thus:

Re: building ghc-3.00 using ghc-2.10

1998-02-20 Thread Sigbjorn Finne
Timothy Robin BARBOUR writes: I had already worked around the problem by making the missing .hi's individually. However, in a spirit of investigation, I lndir'd another build tree, did './configure', 'make boot', 'make all' (with of course the same result). Re-running 'make boot' from

Re: GHC 3.01: glitch in install

1998-02-20 Thread Sigbjorn Finne
Jeffrey R. Lewis writes: On a sparc, Solaris 2.5.1, during make boot, I encounter the following minor glitch: ... Done. ln -s ghc-3.01 ghc ln: cannot create ghc: File exists No wonder, a copy of the driver made it into the distrib (bad interaction between `make dist' and local build

Re: Interface lexer bug in 3.01

1998-02-22 Thread Sigbjorn Finne
Sven Panne writes: After some failed attempts of compiling GreenCard-generated sources with -O, I boilt down the problem to a bug in GHC's interface lexer. Here a failing program: -- Foo.hs module Foo where bar :: IO Int bar = _casm_ ``%r =

Re: GHC3.00:stranal/SaAbsInt.lhs:665: Non-exhaustive patterns in ...

1998-02-23 Thread Sigbjorn Finne
Jon Mountjoy writes: Hello People, Compiling with ghc-3.00 I get: TypeDt.lhs:189: Pattern match(es) are overlapped in the definition of function `showType' x = ... NOTE: Simplifier still going after 4 iterations; bailing out. stranal/SaAbsInt.lhs:665:

Re: Probably not a bug ... rather user manual omission.

1998-02-26 Thread Sigbjorn Finne
David Plume writes: ghc-3.10, Solaris 2.5.1 The compiler doesn't appear to link programs compiled with the -prof -auto options if the -O flag is set. Hi, profiling optimised programs is legal - what's the linker errors you're getting? --Sigbjorn

Re: Probably not a bug ... rather user manual omission.

1998-02-26 Thread Sigbjorn Finne
David Plume writes: Resent: Apologies if this arrives twice... Mailer at dcs.gla.ac.uk didn't like my MIME attached files, so they're just appended to this message. Dave Thanks, will have a look. There's been a couple of people reporting problems with

Re: Probably not a bug ... rather user manual omission.

1998-02-26 Thread Sigbjorn Finne
David Plume writes: The compiler doesn't appear to link programs compiled with the -prof -auto options if the -O flag is set. Hi, profiling optimised programs is legal - what's the linker errors you're getting? I've attached the output of the make process and a

Re: problems with nofib benchmarks

1998-02-19 Thread Sigbjorn Finne
Olaf Chitil writes: I have several problems with using the nofib benchmarks 2.05 with ghc-2.08. Most problems are connected with the fact that I try to compile them for several ways, i.e. my `build.mk' includes: ... 1. `make all' says: ==nofib== anna_a: time to compile Utils

Re: Is this a bug?

1998-03-01 Thread Sigbjorn Finne
Marc van Dongen= writes: I suspect the program included below is incorrect. Nevertheless it compiles fine under ghc-3.01 patchlevel 0. module Main( main ) where import List( genericLength ) main = putStr (show integral) putStr "\n" return ()

Re: Is this a bug?

1998-03-01 Thread Sigbjorn Finne
Marc van Dongen= writes: : of the report), this means resolving `integral' to be a value of type : Int. I think this ``resolving'' may lead to unwanted results. It took me quite some time to discover that Integral was resolved to Int in some program I had. Is there a possibility of

Re: building ghc-3.01

1998-03-02 Thread Sigbjorn Finne
Arthur Gold writes: Now I'm having a problem building 3.01: "No rule to make target `bison.simple', needed by `syntax.tab.o' hi, I'm guessing this happens while trying to do a `make boot' in ghc/utils/ugen. Could you send us the output of doing a `make boot' there? I'm puzzled as to why

Re: ST s a - IO a ???

1998-03-02 Thread Sigbjorn Finne
Alastair Reid writes: is there a way in Hugs to convert a state transformer into a IO computation? In GHC "IO a" is (or perhaps used to be) defined as "ST RealWorld a", so you can use state transformers in the IO monad. I don't think that works with GHC anymore. Yes it does,

Re: hstags; and an old bug

1998-03-03 Thread Sigbjorn Finne
Timothy Robin BARBOUR writes: Also, the module below (my attempt to clean up Hash.hs) produces this warning when compiled with ghc-3.01: Hash.hs:1: Warning: `Hashable' mentioned 3 times in export list I thought this one had been fixed, but now it seems to be growing. ... module Hash

Re: Stack space leak

1998-03-03 Thread Sigbjorn Finne
Sven Panne writes: The problem with this code is the fact that every callback seems to eat 8 bytes from the stack: Sigh, thought I'd fixed this - patch at the end. --Sigbjorn *** runtime/c-as-asm/PerformIO.lhc 1998/01/09 13:33:03 1.5 --- runtime/c-as-asm/PerformIO.lhc

Re: error msg and unboxing

1998-03-03 Thread Sigbjorn Finne
Fermin Reig writes: Hi, I got the following error message with 2.10: MyFiniteMap.lhs:857: Compiler bug: kind mis-match between `ta1XZ' :: `*' and `Int#' :: `*#' Expected: `Int#' Inferred: `ta1XZ' In the second argument of `put', namely `size'

Re: linking project in ghc-3.00

1998-03-06 Thread Sigbjorn Finne
S.D.Mechveliani writes: Hello, people, Please, why the project does not link? ... Now, remove this -Onot from D/docon/Makefile and prepare the optimized DoCon by cd D/docon; rm export/* ; make all Then the above test with cubeext.hs, it cannot link:

Re: The impossible happened in 3.01

1998-03-05 Thread Sigbjorn Finne
Sven Panne writes: During the compilation of a small project with ghc-3.01 (Linux) I get: panic! (the `impossible' happened): unboxArg: PrelBase.(){-40-} The code leading to this error is available at:

Re: Funny things on HPs

1998-03-11 Thread Sigbjorn Finne
Sven Panne writes: Compiling GreenCard generated programs does not always work on HP-UX. The following program (a short version of some GreenCard output) demonstrates the problem: -- Main.hs -- import Addr main :: IO () main =

Re: Cost Centre Confusion

1998-03-12 Thread Sigbjorn Finne
Sven Panne writes: Traditionally, profiling with optimizations on has been a problem in GHC's history... :-]Here's a new one: .. ghc-3.01 -o blah -O -prof -fglasgow-exts Foo.o Main.o Main.o(.text+0xe9): undefined reference to `CC_FooZdmycc2ZdDUPD_struct' Main.o(.text+0xef):

Re: Obscure warning

1998-03-16 Thread Sigbjorn Finne
Sven Panne writes: ifaceBinds: free vars: PrelPack.unpackCString#{-8D-} ghc-3.01: module version changed to 1; reason: no old .hi file Is this a bug in GHC? Can the warning be ignored safely? Just ignore it... This warning from the interface file generator has been with us for a

Re: Difference between Hugs-1.4 and GHC-2.10

1998-03-16 Thread Sigbjorn Finne
Simon Marlow writes: Alastair Reid [EMAIL PROTECTED] writes: My reading of the report is that Hugs is right and GHC is wrong. Do the Glasgow folks agree? Yep - we're still trying to track down the cause though. We've located the problem now - fixed in the next release. (Cause:

Re: GranSim again...

1998-03-16 Thread Sigbjorn Finne
Sven Panne writes: My previous mail went our a little bit too early: Browsing through the compilation logs, I guess that %edi needs some fixup, too. So the mangler probably needs yet another line... Yes, adding cases for %edi and %esi should account for code that gcc might generate when

Re: GHC/Hugs Extension Library lacks Functor (ST s)

1998-03-19 Thread Sigbjorn Finne
Patrik Jansson writes: [In ghc-2.10 and ghc-3.01] The ghc version of the ST module ghc/lib/exts/ST.lhs makes (ST s) a Monad but not a Functor. In hugs 1.4 February it is both a Monad and a Functor. ST should be an instance in Functor. Will fix for the next version - thanks for

Re: Anomaly in IO in GHC (I think)

1998-03-25 Thread Sigbjorn Finne
Hi, GHC's behaviour is consistent with the report, Hugs' isn't. This issue came up on the haskell mailing list a couple of months ago, see http://www.dcs.gla.ac.uk/mail-www/haskell/threads#00270 for more info. hth, --Sigbjorn David Barton writes: Consider the following (literate)

Re: Anomaly in IO in GHC (I think)

1998-03-25 Thread Sigbjorn Finne
Sorry, forgot the .html extension: http://www.dcs.gla.ac.uk/mail-www/haskell/threads.html#00270 --Sigbjorn Steve Price writes: On Wed, 25 Mar 1998, Sigbjorn Finne wrote: http://www.dcs.gla.ac.uk/mail-www/haskell/threads#00270 I had problems with that address, try: http

  1   2   3   4   5   >