Mailing-List: contact cygwin-help AT sourceware DOT cygnus DOT com; run by ezmlm
List-Subscribe: <mailto:cygwin-subscribe AT sources DOT redhat DOT com>
List-Archive: <http://sources.redhat.com/ml/cygwin/>
List-Post: <mailto:cygwin AT sources DOT redhat DOT com>
List-Help: <mailto:cygwin-help AT sources DOT redhat DOT com>, <http://sources.redhat.com/ml/#faqs>
Sender: cygwin-owner AT sources DOT redhat DOT com
Delivered-To: mailing list cygwin AT sources DOT redhat DOT com
Message-ID: <39E43961.6507D018@Informatik.uni-oldenburg.de>
Date: Wed, 11 Oct 2000 11:56:49 +0200
From: "Frank Moehle" <Frank DOT Moehle AT Informatik DOT Uni-Oldenburg DOT DE>
Organization: University of Oldenburg, Germany
X-Mailer: Mozilla 4.61 [en] (X11; U; SunOS 5.6 sun4u)
X-Accept-Language: en
MIME-Version: 1.0
To: cygwin AT sources DOT redhat DOT com
Subject: cooperation of cygwin and NuTCracker ??
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Hello,

We are going to use cygwin to port a set of unix programs from
solaris to Windows NT. (Un)fortunately, we use a library from
am external product. This Product has also been ported from solaris
to NT, but using "NuTCracker".
We only have access to the final .DLL of this external product
(of cource shipped with the NuTCracker runtime enviroment).

We have already made some progress in porting the programs that do not need
the external library. I really want to say that the cygwin team did a
grat job! Everything looks and feels like unix, and we basicially only had to
replace parts that were different between solaris and gnu-linux. All our
Makefiled worked as expected, no need to fiddle with .exe stuff.
(ok, there _are_ some changes, e.g. we removed chown/chgrp lines,
 used "install" instead of "cp" to copy the compiled programs to the 
 bin/ directory and seperated the binaries/libraries by platform 
 by referry to $(OSTYPE))


Now comes the harder part:

The DLL i was talking about is called "anaport.dll" and comed from a very popular
CASE tool for embedded systems.

We have created .a library for this one, using the following lines in
our Makefile:

----
API_OBJ = anaport.a
API_DLL_NAME = anaport.dll
API_DLL = /cygdrive/y/stmm/misc/ANAPORT/$(API_DLL_NAME)
API_DEF = anaport.def

$(API_OBJ): $(API_DEF)
        dlltool --def $< --dllname $(API_DLL_NAME) \
          --output-lib $@ $(API_DLL)
 
$(API_DEF): $(API_DLL)
        echo EXPORTS > $@
        nm $< | grep ' T _stmm' | sed 's/.* T _//' >> $@
-----
 
and the went on as usual:

----
LIBRARIES = -l<our own stuff.a> -lm

$(Our_tool):    $(Our_tool_OBJS)
        gcc -o $(Our_tool) $(Our_tool_OBJS) $(API_OBJ) $(LIBRARIES)
----


The first result was that we could start the resulting binary, and get a
few --help lines out of it (after setting NCDIR to point to the nutcracker root).
But when we tried to really use the  anaport.DLL part,
the NuTCracker side dumped core.


Are there any known issues in mixing cygwin and NuTCracker stuff ??
Or is something important missing in the above procedure ??


Thanks,
	Frank. 

-- 
=======================================================================
         Frank Moehle                              | Understanding is a
Work   : Uni Oldenburg,Computer Architecture Group | three-edged sword.
Inhouse: frankm AT haydn                              | (Vorlon saying)
DOMAIN : Frank DOT Moehle AT Informatik DOT Uni-Oldenburg DOT DE  +-------------------
-----------------------------------------------------------------------
****            Its how you DO it, not how you don't.              ****

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe AT sourceware DOT cygnus DOT com