July 2002 - Glasgow-haskell-users

RE: Unregisterised build
by Simon Marlow 15 Jul '02

15 Jul '02

> I'm trying to do an unregistered build of ghc, using the cross-port > script. However, compilation aborts at the following point: > > haskell/b1/libraries/base > ../../ghc/compiler/ghc-inplace -ldl > -fglasgow-exts -cpp -Iinclude -funbox-strict-fields > -package-name base -O > -H24m -keep-hc-files-too -optc-DNO_REGS -optc-DUSE_MINIINTERPRETER > -fno-asm-mangling -funregisterised -fvia-C -split-objs -c > GHC/Base.lhs > -o GHC/Base.o > > The problem appears to be a conflict between '-split-objs' and > '-fno-asm-mangling' (which, together with '-funregisterised', > is supposed > to have been replaced by '-unreg' now, if I'm not mistaken). > > However, -split-objs seems to be required for building > libraries, whereas > -fno-asm-mangling is probably needed for an unregisterised > build (I'm just > guessing here...). Turn off splitting by adding the line 'SplitOjbs=NO' to your mk/build.mk file (create one if you don't have it). Splitting is entirely optional. BTW the cross-port script is a little old, so you might find other problems. As far as I know, unregisterised compilation *does* work though. You might run into problems with the libraries, for two reasons: some of them are built via hsc2hs which encodes some platform-specific information into the Haskell source, and some of them rely on configure-generated information, again in the Haskell source. Both of these problems lead to non-portable Haskell and hence non-portable unregisterised .hc files. Your best bet for the .hsc files is to generate the .hs on your Alpha/Linux box by taking the intermediate _hsc.c files over and compiling/running them on the target machine to get the .hs files. Then take the .hs back to the bootstrapping machine and carry on. These are the libraries in stage 1, BTW. For the configure-generated options (eg. GHC/Posix.hs uses a bunch of types from config.h), you might have to temporarily substitute a config.h built on the target box when compiling this file to .hc. The cross-port script also tries to do too much. You won't be able to actually run the b2 compiler on your bootstrap box, because it will be built against the Alpha/Linux-specific libraries, so just stop after you've got a set of .hc files for ghc/compiler in stage 2. Take the library .hc files from stage 1, and the compiler .hc files from stage 2, to your Alpha/Linux box. At some point we'll try to make all this easier, but I'm afraid that's the way it is at the moment :-( Cheers, Simon

1 0

RE: replacing the Prelude (again)
by Bernard James POPE 15 Jul '02

15 Jul '02

Hi All, Thanks to everyone for their comments. I can see that this is quite a difficult design problem, and it is unlikely that one solution will please everyone, which makes me think that the current solution will probably stand. Simon Peyton-Jones writes: > Bernie writes: > | anymore. What I would like is that the defualting rules refer > | to the classes in my version of the Prelude, > | not the Standard Prelude. > > You can always get that (with the -fno-implicit-prelude thing) by adding > > default [Int, Double] > > or whatever to your source module, just after the import that gets your > new standard Prelude. Doesn't that do it? I'll get back to you on this soon. > It would also not be hard to arrange that the "default default" > declaration became scoped with -fno-implicit-prelude (like fromInteger), if > that was useful. That sounds a lot like what I would like to have, but I'm probably being selfish here. > It's a good point; I thought that *all* the numeric > stuff was un-coupled from the Prelude with -fno-implicit-prelude, > but it isn't quite. What I would like to know is the semantics of -fno-implicit-prelude with respect to this bit from the Haskell Report: "4.3.4 Ambiguous Types, and Defaults for Overloaded Numeric Operations Ambiguities in the class Num are most common, so Haskell provides another way to resolve them---with a default declaration: default (t1 , ... , tn) where n>=0, and each ti must be a monotype for which Num ti holds. ^^^^^^^^^^ In situations where an ambiguous type is discovered, an ambiguous type variable is defaultable if at least one of its classes is a numeric class (that is, Num or a subclass of Num) and if all of its classes are defined ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ in the Prelude or a standard library ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Each defaultable variable is replaced by the first type in the default list that is an instance of all the ambiguous variable's classes. It is a static error if no such type is found." Do these become "whatever Num is in scope" when I use -fno-implicit-prelude, and "whatever classes from standard libraries" are in scope? If they don't then I think I'm in a pickle. Ooroo, Bernie.

1 0

RE: replacing the Prelude (again)
by Simon Peyton-Jones 15 Jul '02

15 Jul '02

I'm fond of the idea proposed by Marcin 'Qrczak' Kowalczyk: | May I propose an alternative way of specifying an alternative Prelude? | Instead of having a command line switch, let's say that 3 always means | Prelude.fromInteger 3 - for any *module Prelude* which is in scope! That's a perfectly reasonable idea, but GHC can't make it the default because it's not H98. It *would* be possible to make GHC's -fno-implicit-prelude flag mean that 3 means Prelude.fromInteger 3 rather than fromInteger 3 I quite like the fact that you would then have to say import MyPrelude as Prelude thereby stressing that you are importing the Prelude stuff. But it would be that *plus* -fno-implicit-prelude, I'm afraid. If there were a consensus in favour of this I'll gladly do it (5 min job). | anymore. What I would like is that the defualting rules refer | to the classes in my version of the Prelude, | not the Standard Prelude. You can always get that (with the -fno-implicit-prelude thing) by adding default [Int, Double] or whatever to your source module, just after the import that gets your new standard Prelude. Doesn't that do it? It would also not be hard to arrange that the "default default" declaration became scoped with -fno-implicit-prelude (like fromInteger), if that was useful. It's a good point; I thought that *all* the numeric stuff was un-coupled from the Prelude with -fno-implicit-prelude, but it isn't quite. Simon

2 1

RE: replacing the Prelude (again)
by Ashley Yakeley 15 Jul '02

15 Jul '02

At 2002-07-15 01:05, Simon Peyton-Jones wrote: >I quite like the fact that you would then have to say > > import MyPrelude as Prelude > >thereby stressing that you are importing the Prelude stuff. Doesn't this assume your Prelude stuff is all in one module? Or can you import several modules "as Prelude"? As a Prelude-replacer, my preference is rather towards keeping it as it is... it seems a bit simpler and easier to understand. -- Ashley Yakeley, Seattle WA

1 0

package name mismatch
by Max Kirillov 13 Jul '02

13 Jul '02

Hello! Building htk, I've got an error: TestGetPut.hs:6: Module `GetPut' is located in package `Main' but its interface file claims it is part of package `uni-events-test' It happened (as far as I got) when ghc saw in current path the *.hi file built for installation. There are some ways to work around it, but, there was a lot of such places, so can I just skip the 'package name' check and go as it is? In http://www.haskell.org/pipermail/glasgow-haskell-users/2002-January/002849.… Simon Marlow said that it is a warning. That wold be nice, but I see an error. that's ghc-5.02 Max.

1 0

Unregisterised build
by Christoph Reichenbach 12 Jul '02

12 Jul '02

Hi, I'm trying to do an unregistered build of ghc, using the cross-port script. However, compilation aborts at the following point: haskell/b1/libraries/base > ../../ghc/compiler/ghc-inplace -ldl -fglasgow-exts -cpp -Iinclude -funbox-strict-fields -package-name base -O -H24m -keep-hc-files-too -optc-DNO_REGS -optc-DUSE_MINIINTERPRETER -fno-asm-mangling -funregisterised -fvia-C -split-objs -c GHC/Base.lhs -o GHC/Base.o The problem appears to be a conflict between '-split-objs' and '-fno-asm-mangling' (which, together with '-funregisterised', is supposed to have been replaced by '-unreg' now, if I'm not mistaken). However, -split-objs seems to be required for building libraries, whereas -fno-asm-mangling is probably needed for an unregisterised build (I'm just guessing here...). How do I compile libraries for an unregisterised build, then? (Alternatively: Does anyone happen to have the resulting .hc files from one lying around?) Background: I'm trying to get ghc up and running on Alpha/Linux (I'm reluctant to call it 'porting', since the hard part has been done by Ken Shan already, as far as I know). Thanks in advance for any help! llap, Christoph

1 0

Updated version of 5.04 installer available
by Sigbjorn Finne 12 Jul '02

12 Jul '02

Hi, should you run into linking problems when/if using the --mk-dll option to build DLLs with ghc-5.04 on a Windows box, a patched version of the 5.04 installer is now available which fixes the underlying problem. That's the only thing different wrt. the original 5.04 installer, so unless you're directly running into this problem, it's probably not worth your bandwidth to download the updated MSI. Thanks to Mark Tehver for reporting this. --sigbjorn

1 0

make -j2 fails on SMP with ghc-5.04
by Shae Matijs Erisson 12 Jul '02

12 Jul '02

I have a dual Athlon box, make -j2 fails, but make works. Is that normal? -- Shae Matijs Erisson - http://www.webwitches.com/~shae/

1 0

RE: ANNOUNCE: GHC version 5.04 released
by Simon Marlow 12 Jul '02

12 Jul '02

> RPMs built on Red Hat 7.3 are now available from Also now available from the GHC download page. Simon

1 0

Re: CATALOG & RH 7.3
by dons＠cse.unsw.edu.au 11 Jul '02

11 Jul '02

On Fri, Jul 12, 2002 at 12:34:39AM +1000, Manuel M T Chakravarty wrote: > Since I updated to RedHat 7.3, configure doesn't seem to be > able to determine a valid DocBook catalog anymore. Anybody > else seen this problem or is there something wrong in my > setup? > > Manuel Same problem on OpenBSD. The "catalog" file is stored in /usr/local/share/sgml on my system, and is not found. Hence: --- configure.orig Thu Jul 11 21:02:09 2002 +++ configure Thu Jul 11 21:02:51 2002 @@ -2550,7 +2550,7 @@ Test. EOF fptools_cv_sgml_catalog=no if test -z "$SGML_CATALOG_FILES" ; then - for fptools_catalog in /etc/sgml/catalog /etc/sgml.catalog /usr/share/sgml/CAT ALOG.docbkdsl glafp-utils/docbook/CATALOG*; do + for fptools_catalog in /usr/local/share/sgml/catalog /etc/sgml/catalog /etc/sg ml.catalog /usr/share/sgml/CATALOG.docbkdsl glafp-utils/docbook/CATALOG*; do ac_try="$JadeCmd -t rtf -d docs/fptools-both.dsl#print -c $fptools_catalog c onftest.sgml" Don

2 1