RE: cross compiler for Cygwin

[John Andrews <john at triscend dot com>]

Thanks a million for the debug tip Mumit!

I think it's led to the problem.

In the file /usr/include/w32api/winnt.h line 2266 there's a typedef:

typedef struct _IMAGE_IMPORT_BY_NAME {
        WORD Hint;
        BYTE Name[1];
} IMAGE_IMPORT_BY_NAME,*PIMAGE_IMPORT_BY_NAME;

When I found this definition in the prefix.i it had been altered:

typedef struct _IMAGE_IMPORT_BY_NAME {
        WORD int  ;
        BYTE Name[1];
} IMAGE_IMPORT_BY_NAME,*PIMAGE_IMPORT_BY_NAME;

WORD Hint; is changed to WORD int ;

I tried a test program with the native gcc:

typedef unsigned short WORD;
typedef unsigned char BYTE;
main () {
    typedef struct _IMAGE_IMPORT_BY_NAME {
        WORD int  ;
        BYTE Name[1];
    } IMAGE_IMPORT_BY_NAME,*PIMAGE_IMPORT_BY_NAME;
  }

The compile failed with the same error as my cross gcc build:

$ gcc test.c
test.c: In function `main':
test.c:5: two or more data types in declaration of `type name'

I had read several of your other how-tos but somehow I missed cygwin ->
linux. I read through that and the run time libraies and headers are a major
thing I'm missing. I'm trying to find a prebuilt version for my target to
fill in the blanks. Do you think this is the?

Thanks Again to You and everyone else out there reading this stuff...

John E. A.

BTW - It wasn't me complaining about the cost of compilers. Companies put a
ton of time into developing their tools and probably even more on the phone
with customers. In my case this is a personal effort to learn embedded linux
and the GNU cross tools, I'm not in a position to get a professional support
contract. Hopefully I'll be able to contribute back to the list once I've
been through the mill successfully a few times. Thanks again for helping me
out!

-----Original Message-----
From: Mumit Khan [mailto:khan@NanoTech.Wisc.EDU]
Sent: Friday, March 09, 2001 8:21 PM
To: John Andrews
Cc: cygwin@cygwin.com
Subject: RE: cross compiler for Cygwin


On Fri, 9 Mar 2001, John Andrews wrote:

> Well now that you've offered Mumit... 8-)

That's why I wrote up a few howto's in the past, and the latest one is
cygwin->linux. See http://www.nanotech.wisc.edu/~khan/software/gnu-win32/
for details.

The trouble, as you've found out, is that various targets use different
gcc versions (or patches to it), and it gets trickier to build. The only
gcc that will build without tweaking is the gcc distributed by Cygwin
team, but that may not have some of the patches that your target may
require. And this is where the vendors for these embedded systems can
do a lot of good, but of course support costs real money. Then again, so
does time.

[ steps in building look good. However ... ]
> #12) make gcc with:
> make -w all-gcc LANGUAGES="c" 2>&1 | tee make.out
> 
> And that brings us full circle to where I was when I posted my first
message
> a couple of days ago. The make dies with the following error message:

> ---tail of gcc make.out---
> gcc -DCROSS_COMPILE -DIN_GCC     -g -O2  -DHAVE_CONFIG_H    -I.
> -I/usr/src/gcc-2
> .95.2/gcc -I/usr/src/gcc-2.95.2/gcc/config
> -I/usr/src/gcc-2.95.2/gcc/../include
> \
> -DPREFIX=\"/usr/arm-uClinux\" \
>   -c `echo /usr/src/gcc-2.95.2/gcc/prefix.c | sed 's,^\./,,'`
> In file included from /usr/include/w32api/windef.h:143,
>                  from /usr/include/w32api/windows.h:96,
>                  from /usr/src/gcc-2.95.2/gcc/prefix.c:69:
> /usr/include/w32api/winnt.h:2266: two or more data types in declaration of
> `type
>  name'

This is where debugging comes in. Here's what you do:
  
  $ make all-gcc LANGUAGES="c" CFLAGS=-save-temps

and look at prefix.i file and see what gives. And then you'll get to the
next problem, and the same applies, and so on. Fun, isn't it?

Regards,
Mumit


--
Want to unsubscribe from this list?
Check out: http://cygwin.com/ml/#unsubscribe-simple