April 2003 - Cvs-hugs

cvs commit: hugs98/src Makefile.in hugs98/src/unix convert_libraries convert_package
by ross＠glass.cse.ogi.edu 11 Apr '03

11 Apr '03

ross 2003/04/11 17:07:57 PDT Modified files: src Makefile.in src/unix convert_libraries convert_package Log: The conversion script now generates a script of the required ffihugs commands, which are later run by the Makefile. It adds to the ffihugs command line: * a -I option for the package's include directory. * files generated by hsc2hs or greencard. * files from the package's cbits directory mentioned in CBITS "pragmas" in the module. With a bit of tweaking, this could be used for third-party packages too. TODO: some way of specifying C libraries to link with, probably on a per-package basis. Revision Changes Path 1.52 +7 -16 hugs98/src/Makefile.in 1.64 +12 -14 hugs98/src/unix/convert_libraries 1.4 +75 -8 hugs98/src/unix/convert_package

1 0

cvs commit: hugs98/src config.h.in hugs98/src/unix configure configure.in
by ross＠glass.cse.ogi.edu 11 Apr '03

11 Apr '03

ross 2003/04/11 04:30:50 PDT Modified files: src config.h.in src/unix configure configure.in Log: look for a few more headers Revision Changes Path 1.21 +9 -0 hugs98/src/config.h.in 1.53 +404 -403 hugs98/src/unix/configure 1.59 +2 -1 hugs98/src/unix/configure.in

1 0

cvs commit: hugs98/libraries/Hugs Prelude.hs
by ross＠glass.cse.ogi.edu 11 Apr '03

11 Apr '03

ross 2003/04/11 04:29:37 PDT Modified files: libraries/Hugs Prelude.hs Log: add UnsupportedOperation to IOErrorType for GHC compatibility Revision Changes Path 1.26 +3 -0 hugs98/libraries/Hugs/Prelude.hs

1 0

RE: fromInteger :: Integer -> Int/Word
by Simon Marlow 11 Apr '03

11 Apr '03

> Ross Paterson <ross(a)soi.city.ac.uk> writes: > > Under Hugs, fromInteger (and thus fromIntegral) throws an exception > > if the argument is outside the representable range of values, while > > GHC just truncates. > > I would say that GHC implements a performance optimization by dropping > the check and that the behaviour outside the range of Int should be > treated as undefined (which is not the same as 'implementation > defined'). If I remember correctly (this point has come up several times in the past), the reasoning is something like this: since Int arithmetic doesn't do any overflow checking, it's inconsistent for fromInteger to do any. Q: should (0xffffffff :: Int32) == (-1 :: Int32) we take the view that it's useful if this holds. So fromInteger on *sized* integral types doesn't do any overflow checking. Now, if a, b :: Int, should (a + b) :: Int == fromInteger (toInteger a + toInteger b) :: Int (eg. take a to be maxBound :: Int, and b to be 1, and this doesn't hold in Hugs). We take the view that this should hold, i.e. it makes no difference if the addition is done with Int or Integer because the final result is modulo 2^n (where n is the number of bits in Int). If you were to do overflow checking on Int addition, then it would also make sense to do overflow checking on fromInteger, and the above property would still hold, but to do one but not the other is inconsistent. Cheers, Simon

1 0

cvs commit: hugs98/src bignums.c
by ross＠glass.cse.ogi.edu 10 Apr '03

10 Apr '03

ross 2003/04/10 08:07:33 PDT Modified files: src bignums.c Log: Expand the range of values accepted by fromInteger :: Integer -> Int/Word to toInteger(minBound::Int) .. toInteger(maxBound::Word). This is required to provide the semantics advertised in Data.Int and Data.Word, that conversions (using fromIntegral) between signed and unsigned types preserve representation, not sign. Revision Changes Path 1.10 +12 -33 hugs98/src/bignums.c

1 0

cvs commit: hugs98/src machdep.c MkDepend.in hugs98/src/dotnet Makefile hugs98/src/msc Makefile hugs98/src/winhugs Makefile
by ross＠glass.cse.ogi.edu 10 Apr '03

10 Apr '03

ross 2003/04/10 02:26:04 PDT Sender: cvs-hugs-admin(a)haskell.org Errors-To: cvs-hugs-admin(a)haskell.org X-BeenThere: cvs-hugs(a)haskell.org X-Mailman-Version: 2.0.8 Precedence: bulk List-Help: <mailto:cvs-hugs-request@haskell.org?subject=help> List-Post: <mailto:cvs-hugs@haskell.org> List-Subscribe: <http://www.haskell.org/mailman/listinfo/cvs-hugs>, <mailto:cvs-hugs-request@haskell.org?subject=subscribe> List-Id: Hugs CVS commit messages <cvs-hugs.haskell.org> List-Unsubscribe: <http://www.haskell.org/mailman/listinfo/cvs-hugs>, <mailto:cvs-hugs-request@haskell.org?subject=unsubscribe> List-Archive: <http://www.haskell.org/pipermail/cvs-hugs/> Modified files: src machdep.c MkDepend.in src/dotnet Makefile src/msc Makefile src/winhugs Makefile Log: add a #include to avoid a C compiler warning Revision Changes Path 1.87 +3 -2 hugs98/src/machdep.c 1.3 +5 -5 hugs98/src/MkDepend.in 1.9 +5 -5 hugs98/src/dotnet/Makefile 1.34 +5 -5 hugs98/src/msc/Makefile 1.23 +5 -5 hugs98/src/winhugs/Makefile

1 0

cvs commit: hugs98/src MkInstal.in
by ross＠glass.cse.ogi.edu 10 Apr '03

10 Apr '03

ross 2003/04/10 02:20:47 PDT Modified files: src MkInstal.in Log: build libraries before installing Revision Changes Path 1.32 +1 -1 hugs98/src/MkInstal.in

1 0

cvs commit: hugs98/src edit.c
by ross＠glass.cse.ogi.edu 10 Apr '03

10 Apr '03

ross 2003/04/10 02:20:01 PDT Modified files: src edit.c Log: #include <math.h> Revision Changes Path 1.2 +1 -0 hugs98/src/edit.c

1 0

fromInteger :: Integer -> Int/Word
by Ross Paterson 09 Apr '03

09 Apr '03

Under Hugs, fromInteger (and thus fromIntegral) throws an exception if the argument is outside the representable range of values, while GHC just truncates. And some of the libraries rely on the GHC behaviour. It seems it's not a matter of principle with Hugs, since operations on Int or Word silently overflow, as do the built-in conversions between Int and Word types.

2 2

cvs commit: hugs98/libraries/Hugs Observe.hs Observe.lhs
by ross＠glass.cse.ogi.edu 09 Apr '03

09 Apr '03

ross 2003/04/09 01:31:38 PDT Added files: libraries/Hugs Observe.hs Removed files: libraries/Hugs Observe.lhs Log: remove \begin{code} and \end{code}

1 0