Dynamic linker under Windows

Sigbjorn Finne sof at galois.com
Mon Jun 20 12:29:24 EDT 2005


Thanks, what did you end up switching to? We've been fortunate
in the past couple of years with gcc/binutils under mingw/cygwin
not to have had any versioning pains, so I really hope this was an
isolated 3.3.1 incident (the nature of the problem does make
it look as if it could be.)

I'll switch over to using gcc-3.4.2 for the nightly builds to see where
takes us.

--sigbjorn

----- Original Message ----- 
From: "Krasimir Angelov" <kr.angelov at gmail.com>
To: "Sigbjorn Finne" <sof at galois.com>
Cc: <cvs-ghc at haskell.org>
Sent: Sunday, June 19, 2005 01:49
Subject: Re: Dynamic linker under Windows


> Yes. This was the problem. My GCC was 3.3.1. Is it documented
> somewhere which gcc,binutils and mingw packages are required for GHC?
> 
> Cheers,
>   Krasimir
> 
> On 6/17/05, Sigbjorn Finne <sof at galois.com> wrote:
>> Odd, works for me using the latest HEAD snapshot.
>> What does "gcc --version | head -1" and "ld --version | head -1"
>> report? The snapshot was compiled using
>> 
>> gcc.exe (GCC) 3.2.3 (mingw special 20030504-1)
>> GNU ld version 2.15.90 20040222
>> 
>> --sigbjorn
>> 
>> ----- Original Message -----
>> From: "Krasimir Angelov" <kr.angelov at gmail.com>
>> To: <cvs-ghc at haskell.org>
>> Sent: Friday, June 17, 2005 06:51
>> Subject: Dynamic linker under Windows
>> 
>> 
>> > Hello Guys,
>> >
>> > GHC from HEAD crashes when I am using getHomeDirectory function from
>> > GHCi. The problem is in __hscore_getFolderPath function. The function
>> > body is:
>> >
>> > int
>> > __hscore_getFolderPath(HWND hwndOwner,
>> >        int nFolder,
>> >        HANDLE hToken,
>> >        DWORD dwFlags,
>> >        char*  pszPath)
>> > {
>> >    static int loaded_dll = 0;
>> >    static HMODULE hMod = (HMODULE)NULL;
>> >    static HSCORE_GETAPPFOLDERFUNTY funcPtr = NULL;
>> > ......
>> >
>> > When the function is called from compiled binary then the inital value
>> > of loaded_dll is 0 but when the function is executed from interpreted
>> > code (i.e. when the function is in dynamically loaded .o file) then
>> > the value is 196940. Any idea what is going wrong?
>> >
>> > Cheers,
>> >  Krasimir
>> > _______________________________________________
>> > Cvs-ghc mailing list
>> > Cvs-ghc at haskell.org
>> > http://www.haskell.org/mailman/listinfo/cvs-ghc
>>


More information about the Cvs-ghc mailing list