Don,

  To avoid any interference from my dev machine, I decided to to a test on
a new machine. I grabbed another Windows machine with a semi pristine
state.  I installed svn, mingw32/msys, mingw64, checked out a copy of
Unicon sources and built; no errors!

So it is a toolchain thing at your end. gcc maybe? you are using a
different build dwarf in your case vs tdm or the "official" mingw in my
case. Here are the links to my tools:

http://sourceforge.net/projects/mingw/files/Installer/
http://tdm-gcc.tdragon.net/download

Since you have a work around I guess this is not a big issue anyway.

Below is the version info of my tools.

Cheers,
Jafar

c:\unicon>gcc --version
gcc (tdm64-1) 5.1.0
Copyright (C) 2015 Free Software Foundation, Inc.

c:\>make -version
GNU Make 3.81
Copyright (C) 2006  Free Software Foundation, Inc.

This program built for i686-pc-msys

c:\>make -p | grep CXX
LINK.cc = $(CXX) $(CXXFLAGS) $(CPPFLAGS) $(LDFLAGS) $(TARGET_ARCH)
COMPILE.cc = $(CXX) $(CXXFLAGS) $(CPPFLAGS) $(TARGET_ARCH) -c
CXX = g++






On Fri, Nov 6, 2015 at 1:41 PM, Don Ward <[email protected]> wrote:

>
> On 6 Nov 2015, at 06:48, Jafar Al-Gharaibeh <[email protected]> wrote:
>
> I did another test; I switched to mingw64 toolchain and did a 32-bit
> build, same result: no build errors. The new gcc reports the following
> version:
>
> c:\unicon>gcc --version
> gcc (tdm64-2) 4.8.1
>
> --Jafar
>
>
> Hello Jafar,
>
> I think we might be better taking this offline to sort out the detailed
> differences in our configurations and, then, to figure out why it works for
> you and not for me.
>
> I downloaded the very latest stuff (rev 4189) and tried it. After the
> initial make WUnicon (which worked), it failed like this
>
> $ make WUnicon32
>
> =============================================
> Running configuration and setup scripts NT 32
> =============================================
>
> cd config/win32/gcc && sh config.sh
>
> cd config/win32/gcc && sh nt-config.sh
>
> Configuring for console build...
>
> cd config/win32/gcc && sh config32.sh
>
> Running 32-bit build configuration...
>
>
> Done!
>
>
> =========================
> Building NT Unicon 32-bit
> =========================
>
> make Unicon
> make[1]: Entering directory 'C:/Users/Don/UniconBuild/unicon-4189'
> cd src/rtt && make
> make[2]: Entering directory 'C:/Users/Don/UniconBuild/unicon-4189/src/rtt'
> g++    -c -o rttparse.o rttparse.c
> In file included from ../preproc/../h/gsupport.h:16:0,
>                  from ../preproc/preproc.h:1,
>                  from rtt1.h:1,
>                  from rttgram.y:7:
> ../preproc/../h/../h/sys.h:360:18: fatal error: ndbm.h: No such file or
> directory
> compilation terminated.
> <builtin>: recipe for target 'rttparse.o' failed
> make[2]: *** [rttparse.o] Error 1
> make[2]: Leaving directory 'C:/Users/Don/UniconBuild/unicon-4189/src/rtt'
> makefile:159: recipe for target 'Icon-icont' failed
> make[1]: *** [Icon-icont] Error 2
> make[1]: Leaving directory 'C:/Users/Don/UniconBuild/unicon-4189'
> makefile:24: recipe for target 'WUnicon32' failed
> make: *** [WUnicon32] Error 2
>
>
> I know how to fix it (with CPPFLAGS and CXX) but the more interesting
> question is how our build tools differ.
> Here, gcc —version gives
>
> $ gcc --version
> gcc.exe (i686-posix-dwarf-rev0, Built by MinGW-W64 project) 5.2.0
>
>
> and make —version gives
>
> $ make --version
> GNU Make 4.1
> Built for i686-w64-mingw32
>
>
> When I run the command below in an empty directory, I get the following
>
> $ make -p 2>/dev/null | grep CXX
> LINK.cc = $(CXX) $(CXXFLAGS) $(CPPFLAGS) $(LDFLAGS) $(TARGET_ARCH)
> COMPILE.cc = $(CXX) $(CXXFLAGS) $(CPPFLAGS) $(TARGET_ARCH) -c
> CXX = g++
>
>
> which at least explains why I need the CXX flag. I’m guessing your make is
> different. We’ll have to confer to find out why I need CPPFLAGS and you
> don’t.  I suggest you send me the output privately from when you run “make
> -p” in an empty directory, and I’ll try to spot any salient differences.
>
> regards
>
> Don
>
> PS. I can build rev 4189 successfully, but only with CPPFLAGS and CXX in
> place. So, your fixes for rusage did the trick.  I will try the other make
> experiments you talked about and get back to with the results.
>
>
------------------------------------------------------------------------------
_______________________________________________
Unicon-group mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/unicon-group

Reply via email to