[MLton-user] Unable to build a Win64 hosted version of MLton

[Nicolas Bertolotti]

Right ! It seems to be specific to the way I build the whole stuff.

I am actually trying to target another version of mingw than the one that is shipped with the MLton installer (taken from http://drangon.org/mingw/). Their 64 bit version appears to be more conformant to what the MS tools do with the symbols (e.g. no extra "_") which requires to change mlton/main/main.fun to set Control.labelsHaveExtra_ to false for the amd64-mingw target.

In my build, I also link to a DLL version of GMP that is already used in some other components by my company.

I guess one of those external components is broken and now I have to identify which one.

May you please tell me where you got the "mingw" version that you included in the MLton installer? As far as I know, the one we can download from www.mingw.org<http://www.mingw.org> does not include 64 bit support which is developed by another team.

Thanks for your help

Nicolas

P.S.: thanks also for your quick response to my question related to the library support. I am afraid this change is too big for my poor knowledge of the compiler internals.

From: Wesley W. Terpstra [mailto:wesley at terpstra.ca]
Sent: Saturday, May 07, 2011 3:32 PM
To: Nicolas Bertolotti
Cc: mlton-user at mlton.org
Subject: Re: [MLton-user] Unable to build a Win64 hosted version of MLton

I can't reproduce this. I built a 64-bit mlton-compile.exe and dropped it into lib/mlton and it worked fine for me.

ie:
cd mlton-20100608/mlton
mlton @MLton ram-slop 0.8 -- -target x86_64-w64-mingw32 -verbose 2 mlton.mlb
cp mlton.exe /lib/mlton/mlton-compile.exe

I also ran the regression suite targetting 64-bit from 32-bit and 64-bit targetting 64-bit. No problems except the expected changes in object and floating point sizes.

On Fri, May 6, 2011 at 4:25 PM, Nicolas Bertolotti <Nicolas.Bertolotti at mathworks.fr<mailto:Nicolas.Bertolotti at mathworks.fr>> wrote:
I implemented the following workaround:
"
structure Int64 =
struct
    open Int64;
    fun fromInt(x) = Int64.fromLarge(LargeInt.fromInt(x));
end;
"

Do you have a test case for the problem?

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mlton.org/pipermail/mlton-user/attachments/20110509/95628596/attachment.htm