Order of libraries
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 the order of $UserLibrary and $SystemLibrary in the ghc driver solves the problem, but I don't know if that is ALWAYS the right behaviour. Stephan -- Stephan Tobies, Student of Computer Science, RWTH Aachen mailto:[EMAIL PROTECTED] ## There is much pleasure to be gained from useless knowledge - B. Russel ##
Re: Order of libraries
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 the order of $UserLibrary and $SystemLibrary in the ghc driver solves the problem, but I don't know if that is ALWAYS the right behaviour. Hi, the present arrangement goes wrong if anything in $SystemLibrary depends on symbols in $UserLibrary. That's not supposed to happen, $SystemLibrary is intended to be `closed'. If that's what you're experiencing, please let us know whatwhere this happens so we can patch up $SystemLibrary to avoid this. (Switching the two is likely to fail if any of the user libraries contain ghc compiled code.) Thx, --Sigbjorn
Re: GranSim/parallel versions of 2.08
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
assembler mangler complains about 5500 times. It's totally
unclear to me what it means: bug/warning/error? Example:
gcc generates code which jumps via %ecx for some reason when compiling
gransim code, something that isn't caught by the mangler. Patch appended.
* The parallel version of the libs can't be compiled because
ghc/lib/glaExts/Foreign.lhs contains lots of
#ifndef__PARALLEL_HASKELL__ lines. Therefore, some exports
are missing later (e.g. for PackedString).
To get the show on the road again here, you'll have to strategically
add #ifndef__PARALLEL_HASKELL__ to avoid referencing stuff that isn't
supported under Parallel Haskell. Sorry, haven't got a patch for this.
--Sigbjorn
*** ghc/driver/ghc-asm.lprl.~1~ Wed Aug 27 15:53:33 1997
--- ghc/driver/ghc-asm.lprl Tue Oct 14 19:40:21 1997
***
*** 976,981
--- 976,982
$c =~
s/^\s+ldil.*\n\s+ldo.*\n\s+bv.*\n(.*\n)?\s+\.EXIT/$1\t.EXIT/;
} elsif ( $TargetPlatform =~ /^i386-/ ) {
$c =~ s/^\tmovl \$${T_US}${symb}_fast\d+,\%edx\n\tjmp
\*\%edx\n//;
+ $c =~ s/^\tmovl \$${T_US}${symb}_fast\d+,\%ecx\n\tjmp
+\*\%ecx\n//;
$c =~ s/^\tmovl \$${T_US}${symb}_fast\d+,\%eax\n\tjmp
\*\%eax\n//;
} elsif ( $TargetPlatform =~ /^mips-/ ) {
$c =~ s/^\tjmp \$31,\(\$27\),0\n\t\.align 4\n\t\.end/\t.align
4\n\t.end/;
Library Random is missing with ghc-2.08 src distribution
The subject says it all... -- Stephan Tobies, Student of Computer Science, RWTH Aachen mailto:[EMAIL PROTECTED] ## There is much pleasure to be gained from useless knowledge - B. Russel ##
Re: ghc Diagnostics
Simon Marlow wrote: I take your point that this isn't very consistent: there should be a way to turn off all warnings easily. What do other people think? The options are: * have all warnings off by default, a standard set of warnings being available by adding the -W command line option. * have a standard set of warnings on by default, with all warnings being turned off by the -Wnot (or something) flag. I'd prefer to have the warnings off by default. And why don't go for a command line option set like that of the `gcc'? Then, `-Wall' activates all warnings and you can get the individual warnings with `-Wwarning name'. Manuel
Re: profiling
Marc I strongly suspect that the names are simply truncated before they get into a .hp file. Doubtless this could be fixed. However, we're now embarking on building a new RTS, designed to support both GHC and Hugs, so I'd rather just make sure that the new system doesn't truncate names. (The new system will take a few months; don't hold your breath.) Is this a show-stopper for you, or can you get by (perhaps by changing your function names)? Simon When I profile my sources (-auto-all) it happens a lot that I see names of the following form SourceFile:ModuleName/descripto in stead of names of the form SourceFile:ModuleName/descriptor_prefix_TYPE_SUFFIX as I would have expected. Is there a way to get the full function-names out of the profiling package? I have looked at the .hp files, but the information does not seem to be available at that level. Any help would be greatly appreciated. Regards, Marc van Dongen [EMAIL PROTECTED]
Re: profiling
At 9:32 am 14/10/97, Simon L Peyton Jones wrote: I strongly suspect that the names are simply truncated before they get into a .hp file. Doubtless this could be fixed. Yes, I came across this when working with the parallel cost-centre code. It would be extremely easy to change the appropriate module in the runtime system (runtime/profiling/CostCentres.lc -- there are some %-16s and %-11s fields that need to be changed to achieve this as far as I can tell), and not hard to add a flag to do this... I'm not sure why the restriction is there -- maybe just to tidy the output?! It is a bit of a pain. Regards, Kevin -- Division of Computer Science, Tel: +44-1334 463241 (Direct) School of Mathematical Fax: +44-1334 463278 and Computational Sciences,URL: http://www.dcs.st-and.ac.uk/~kh/kh.html University of St. Andrews, Fife, KY16 9SS.
Typechecking Monads
Hello,
in the following example, where I define a
state transformer monad "St" parametrized with
the state "c", the ghc-2.08 typechecker reports
the following error message in the type definition of
function "foo" which uses the instantiated "St":
`St' should have no arguments, but has been given 1 .
It works if I comment out the type definition. It also
works in hugs-1.3 with the type definition. Is there a
way to get it being typechecked by given a special compiler option?
Thanks in advance
--
Christoph Herrmann
E-mail: [EMAIL PROTECTED]
WWW: http://brahms.fmi.uni-passau.de/cl/staff/herrmann.html
--
data State c a = State (c - (a,c))
unState :: State c a - (c - (a,c))
unState (State x) = x
unitState :: a - State c a
unitState a = State (\s0 - (a,s0))
bindState :: State c a - (a - State c b) - State c b
bindState m k = State (\s0 - let (a,s1) = (unState m) s0
(b,s2) = (unState (k a)) s1
in (b,s2))
instance Eq c = Monad (State c) where
return = unitState
(=) = bindState
data TS = TS { vs::Int } deriving (Show,Eq)
type St = State TS
foo :: Int - St Int -- it works if this line is not given
foo x = return x
--
END OF MESSAGE
