March 2005 - Glasgow-haskell-users

comments on the GHC API
by Matthias Neubauer 17 Mar '05

17 Mar '05

Hi folks, I just had a first look at the sketch for the planned GHC API (ghc/compiler/main/GHC.hs), since we are very keen on using it ... Our group would mostly like to make use of a compilation manager/module system implementation for Haskell---so these are the parts I concentrated on for now. More specifically, the scenario how we would like to use a GHC library is the following: So far, we've implemented a static analysis for (single) Haskell modules. To make our tool ready for a wider use, we would like extend it to full Haskell 98 (plus extensions). What we therefore would like to ask a GHC library is, given a specific module A, to compile all the other modules that A is dependent on, load them into a context for A, and start processing A up to the point where the actual type-checking of A would commence. After that, we would like to query the library about the initial (type) environment of A (that is, get out a complete summary of all the imported bindings, classes, methods, etc) to then proceed with our own analysis instead of using GHC's type inferencer. I see two basic ways how a GHC library could help us to achieve this: - The library both offers a separate module dependency analysis and also lets us use the compiler in one-shot mode. We then could first calculate the dependencies of A, compile them separately, and calculate the initial environment of A manually by combining the interfaces of the imported modules. Of course, we would also need the possibility to access the imports of A (including renamings and such) without the need to compile A as a whole. - Loading a set of targets (using *load*) can be controlled in a more fine-grained way. Instead of just offering the possibility to compile a set of targets either completely or not at all, there is an option to compile targets only up to a certain compilation phase (or alternatively, to optionally keep intermediate results occurring between the phases to be able to retrieve them afterwards.) We, for our part, would need parts of the intermediate result (that is, the type environment and class declarations) existing either before parsing A, after parsing A, or, even better, after renaming A. Other users may be interested in other intermediate results. What do others think? Some more, rather random, remarks: - I guess v_CmdLineMode and friends are iorefs that are supposed to live globally in GHC.hs. Couldn't you also pack them into a separate data type created by *init* and later share them between several sessions. Maybe we could even introduce a Session monad? Looking at all the functions again, they all seem to have Session arguments ... - I guess the exported name *GhcSession* should read *Session* (or better the other way around) - Same for *GhcMode* vs. *GhciMode* - What's the purpose of having a mode argument for "newSession"? As I read the rest of the API, *load* is the only way to trigger the compilation of targets. And *load* seems to be intended to work similar to the interactive mode only. - What are Module, ModSummary, ModIface, HsTypecheckedGroup, HsGroup, TyThing, and Type standing for (I know of course ...)? How complicated would it be to return already existing (TH) data types instead of exposing/specifying more of GHC's internals? Cheers, Matthias -- Matthias Neubauer | Universität Freiburg, Institut für Informatik | tel +49 761 203 8060 Georges-Köhler-Allee 79, 79110 Freiburg i. Br., Germany | fax +49 761 203 8052

1 0

RE: [Haskell] ANNOUNCE: GHC version 6.4
by Simon Marlow 17 Mar '05

17 Mar '05

On 17 March 2005 10:35, Graham Klyne wrote: > I've just been browsing the release notes at: > http://haskell.org/ghc/docs/6.4/html/users_guide/release-6-4.html > and I notice some broken links about the end of section 1.4.3: > [[ > Network.URI has been completely rewritten, and now conforms to > RFC2732. > > New module: System.Posix.Error. > > New module: System.Posix.Signals.Exts. > ]] > > Specifically: > > > > http://haskell.org/ghc/docs/6.4/html/libraries/base/Network.URI.html > http://haskell.org/ghc/docs/6.4/html/libraries/base/System.Posix.Error.h tml > http://haskell.org/ghc/docs/6.4/html/libraries/base/System.Posix.Signals .Exts.html Oops, I thought I checked all those. Thanks, should be fixed now. Cheers, Simon

1 0

Re: [Haskell] ANNOUNCE: GHC version 6.4
by Graham Klyne 17 Mar '05

17 Mar '05

Some minor documentation feedback... At 12:25 11/03/05 +0000, Simon Marlow wrote: > ============================================================= > The (Interactive) Glasgow Haskell Compiler -- version 6.4 > ============================================================= I've just been browsing the release notes at: http://haskell.org/ghc/docs/6.4/html/users_guide/release-6-4.html and I notice some broken links about the end of section 1.4.3: [[ Network.URI has been completely rewritten, and now conforms to RFC2732. New module: System.Posix.Error. New module: System.Posix.Signals.Exts. ]] Specifically: http://haskell.org/ghc/docs/6.4/html/libraries/base/Network.URI.html http://haskell.org/ghc/docs/6.4/html/libraries/base/System.Posix.Error.html http://haskell.org/ghc/docs/6.4/html/libraries/base/System.Posix.Signals.Ex… #g ------------ Graham Klyne For email: http://www.ninebynine.org/#Contact

1 0

RE: [Haskell] RE: ANNOUNCE: GHC version 6.4
by Simon Marlow 16 Mar '05

16 Mar '05

On 15 March 2005 18:26, Sebastian Sylvan wrote: > Ah, that did it! However, some packages seem to be "missing", like HGL > for instance. > rts-1.0, base-1.0, haskell98-1.0, template-haskell-1.0, unix-1.0, > Cabal-1.0, parsec-1.0, haskell-src-1.0, network-1.0, > QuickCheck-1.0, HUnit-1.1, mtl-1.0, fgl-5.2, OpenGL-2.0, stm-1.0, > readline-1.0, (lang-1.0), (concurrent-1.0), (posix-1.0), > (util-1.0), (data-1.0), (text-1.0), (net-1.0), (hssource-1.0) > > How would one go about just installing every damn package availble? It looks like the configure machinery didn't detect the X11 libraries, or for some other reason failed to build the X11 package and hence also omitted the HGL package. Could you send the full ./configure output? Cheers, Simon

1 0

RE: ghc-pkg too happy to create ~/.ghc
by Simon Marlow 16 Mar '05

16 Mar '05

Thanks, I've committed a version of your patch. Cheers, Simon On 16 March 2005 04:07, Ian Lynagh wrote: > The Debian autobuilders don't let you write to ~ (which seems > reasonable, as they are only compiling the software, not running it), > so > my builds are failing with > > ---------- > ==fptools== /usr/bin/make boot -wr; > in /build/buildd/ghc6-6.4/ghc/rts > [...] > ../utils/ghc-pkg/ghc-pkg-inplace --force --update-package > <package.conf.inplace Creating user package database in > /org/buildd/.ghc/sparc-linux-6.4/package.conf > > Fail: createDirectory: permission denied (Permission denied) > ---------- > > The patch below fixes it. I'm not sure I understand why the code is > written as it is, though. It looks to me like if any config file given > by a FlagConfig is missing then the readFile in readParseDatabase is > going to fall over. I don't know what should happen when modifying if > there are -f options, so can't suggest a complete replacement. > > > Thanks > Ian > > > --- ghc6-6.4.orig/ghc/utils/ghc-pkg/Main.hs > +++ ghc6-6.4/ghc/utils/ghc-pkg/Main.hs > @@ -269,10 +269,6 @@ > archdir = appdir `joinFileName` subdir > user_conf = archdir `joinFileName` "package.conf" > b <- doesFileExist user_conf > - when (not b) $ do > - putStrLn ("Creating user package database in " ++ user_conf) > - createDirectoryIfMissing True archdir > - writeFile user_conf emptyPackageConfig > > let > -- The semantics here are slightly strange. If we are > @@ -281,20 +277,23 @@ > -- If we are not modifying (eg. list, describe etc.) then > -- the user database is included by default. > databases > - | modify = foldl addDB [global_conf] flags > - | not modify = foldl addDB [user_conf,global_conf] flags > + | modify || not b = foldl addDB [global_conf] flags > + | not modify = foldl addDB [user_conf,global_conf] flags > > -- implement the following rules: > -- --user means overlap with the user database > -- --global means reset to just the global database > -- -f <file> means overlap with <file> > - addDB dbs FlagUser = if user_conf `elem` dbs > - then dbs > - else user_conf : dbs > + addDB dbs FlagUser > + | (modify || b) && (user_conf `notElem` dbs) = user_conf : dbs > addDB dbs FlagGlobal = [global_conf] > addDB dbs (FlagConfig f) = f : dbs > addDB dbs _ = dbs > > + when (not b && user_conf `elem` databases) $ do > + putStrLn ("Creating user package database in " ++ user_conf) > + createDirectoryIfMissing True archdir > + writeFile user_conf emptyPackageConfig > db_stack <- mapM readParseDatabase databases > return db_stack > > _______________________________________________ > Glasgow-haskell-users mailing list > Glasgow-haskell-users(a)haskell.org > http://www.haskell.org/mailman/listinfo/glasgow-haskell-users

1 0

RE: alpha problems with ghc 6.4
by Simon Marlow 16 Mar '05

16 Mar '05

On 16 March 2005 04:14, Ian Lynagh wrote: > An alpha build of ghc 6.4 quickly fails because of the > > #if alpha_TARGET_ARCH > import PrimRep ( getPrimRepSize, isFloatingRep ) > import Type ( typePrimRep ) > #endif > > in ghc/compiler/typecheck/TcForeign.lhs which no longer exist. > Fortunately, the imported functions aren't used either. > > Unfortunately, the build then fails when it comes to try to compile > this > same file as the typeMachRepRep function, used in this piece of code: > > \begin{code} > #include "nativeGen/NCG.h" > #if alpha_TARGET_ARCH > checkFEDArgs arg_tys > = check (integral_args <= 32) err > where > integral_args = sum [ machRepByteWidth rep > | (rep,hint) <- map typeMachRepRep arg_tys, > hint /= FloatHint ] > err = ptext SLIT("On Alpha, I can only handle 4 > non-floating-point arguments to foreign export dynamic") #else > checkFEDArgs arg_tys = returnM () > #endif > \end{code} > > doesn't exist. Is this fixable? I think you want to use something like typeMachRep in deSugar/DsForeign.lhs. Cheers, Simon

1 0

alpha problems with ghc 6.4
by Ian Lynagh 16 Mar '05

16 Mar '05

Hi, An alpha build of ghc 6.4 quickly fails because of the #if alpha_TARGET_ARCH import PrimRep ( getPrimRepSize, isFloatingRep ) import Type ( typePrimRep ) #endif in ghc/compiler/typecheck/TcForeign.lhs which no longer exist. Fortunately, the imported functions aren't used either. Unfortunately, the build then fails when it comes to try to compile this same file as the typeMachRepRep function, used in this piece of code: \begin{code} #include "nativeGen/NCG.h" #if alpha_TARGET_ARCH checkFEDArgs arg_tys = check (integral_args <= 32) err where integral_args = sum [ machRepByteWidth rep | (rep,hint) <- map typeMachRepRep arg_tys, hint /= FloatHint ] err = ptext SLIT("On Alpha, I can only handle 4 non-floating-point arguments to foreign export dynamic") #else checkFEDArgs arg_tys = returnM () #endif \end{code} doesn't exist. Is this fixable? Thanks Ian

1 0

ghc-pkg too happy to create ~/.ghc
by Ian Lynagh 16 Mar '05

16 Mar '05

Hi, The Debian autobuilders don't let you write to ~ (which seems reasonable, as they are only compiling the software, not running it), so my builds are failing with ---------- ==fptools== /usr/bin/make boot -wr; in /build/buildd/ghc6-6.4/ghc/rts [...] ../utils/ghc-pkg/ghc-pkg-inplace --force --update-package <package.conf.inplace Creating user package database in /org/buildd/.ghc/sparc-linux-6.4/package.conf Fail: createDirectory: permission denied (Permission denied) ---------- The patch below fixes it. I'm not sure I understand why the code is written as it is, though. It looks to me like if any config file given by a FlagConfig is missing then the readFile in readParseDatabase is going to fall over. I don't know what should happen when modifying if there are -f options, so can't suggest a complete replacement. Thanks Ian --- ghc6-6.4.orig/ghc/utils/ghc-pkg/Main.hs +++ ghc6-6.4/ghc/utils/ghc-pkg/Main.hs @@ -269,10 +269,6 @@ archdir = appdir `joinFileName` subdir user_conf = archdir `joinFileName` "package.conf" b <- doesFileExist user_conf - when (not b) $ do - putStrLn ("Creating user package database in " ++ user_conf) - createDirectoryIfMissing True archdir - writeFile user_conf emptyPackageConfig let -- The semantics here are slightly strange. If we are @@ -281,20 +277,23 @@ -- If we are not modifying (eg. list, describe etc.) then -- the user database is included by default. databases - | modify = foldl addDB [global_conf] flags - | not modify = foldl addDB [user_conf,global_conf] flags + | modify || not b = foldl addDB [global_conf] flags + | not modify = foldl addDB [user_conf,global_conf] flags -- implement the following rules: -- --user means overlap with the user database -- --global means reset to just the global database -- -f <file> means overlap with <file> - addDB dbs FlagUser = if user_conf `elem` dbs - then dbs - else user_conf : dbs + addDB dbs FlagUser + | (modify || b) && (user_conf `notElem` dbs) = user_conf : dbs addDB dbs FlagGlobal = [global_conf] addDB dbs (FlagConfig f) = f : dbs addDB dbs _ = dbs + when (not b && user_conf `elem` databases) $ do + putStrLn ("Creating user package database in " ++ user_conf) + createDirectoryIfMissing True archdir + writeFile user_conf emptyPackageConfig db_stack <- mapM readParseDatabase databases return db_stack

1 0

RE: [Haskell] RE: ANNOUNCE: GHC version 6.4
by Simon Marlow 15 Mar '05

15 Mar '05

On 15 March 2005 00:46, Sebastian Sylvan wrote: > I'm having trouble building this. > I download the source, go to my source dir and run ./configure (which > seems to run to completion without problems), then I run make and get > the following: > > .... > "./mk/config.mk", line 554: Need an operator > "./mk/config.mk", line 554: Need an operator > ..... > Error expanding embedded variable > > This is on freeBSD 5.3 You need to use 'gmake' (GNU make). Cheers, Simon

1 0

RE: [Haskell] RE: ANNOUNCE: GHC version 6.4
by Simon Marlow 15 Mar '05

15 Mar '05

On 14 March 2005 18:08, Cale Gibbard wrote: >>> 4. It looks like Data.FiniteMap has been renamed Data.Map, or >>> something. There's no Data.Map at >>> <http://haskell.org/haddock/libraries/>, do these need to be >>> updated? >> >> The right place for library documentation is >> >> http://haskell.org/ghc/docs/latest/html/libraries >> >> I should probably make the version you're looking at into a link to >> avoid confusion. >> > > Data.Map appears to be broken from there too. I get some sort of page > with a menu as if the html file is a directory containing only the > items "HTML" and "using" both of which are broken links, and then "[an > internal server error occured]". Fixed; it turns out that Apache was trying to interpret a file with a .map extension differently. Cheers, Simon

1 0