Re: B20: bad support for Mingw?

Re: B20: bad support for Mingw?

[Earnie Boyd]

Is it possible that the configuration process imbeds the physical
paths in the jinclude.h or jconfig.h?

I can't get this to fail after the modification to the specs file.



---Maxim Sobolev <max@vega.pfts.com> wrote:
>
> Here is output produced with -H switch, to me it seems that all
inclides are
> CYGWIN **NOT* MINGW! See my specs file also.
> 
> gcc -mno-cygwin -H -O3  -I.   -c jerror.c -o jerror.o
> jinclude.h
>  jconfig.h
> 
c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\include\stddef.h
> 
> 
c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\..\..\..\..\i586-cygwin32\include\stdlib.h
> 
> 
>
c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\..\..\..\..\i586-cygwin32\include\_ansi.h
> 
> 
>
c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\..\..\..\..\i586-cygwin32\include\sys/config.h
> 
> 
>
c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\include\stddef.h
> 
> 
>
c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\..\..\..\..\i586-cygwin32\include\sys/reent.h
> 
> 
c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\..\..\..\..\i586-cygwin32\include\stdio.h
> 
> 
>
c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\include\stddef.h
> 
> 
>
c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\include\stdarg.h
> 
> 
c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\..\..\..\..\i586-cygwin32\include\string.h
> 
> 
>
c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\include\stddef.h
> 
> jpeglib.h
>  jmorecfg.h
> jversion.h
> jerror.h
> jerror.h
> gcc -mno-cygwin -mdll -o junk.tmp -Wl,--base-file,base.tmp jcapimin.o
> jcapistd.o jctrans.o jcparam.o jdatadst.o jcinit.o jcmaster.o
jcmarker.o
> jcmainct.o jcprep
> ct.o jccoefct.o jccolor.o jcsample.o jchuff.o jcphuff.o jcdctmgr.o
> jfdctfst.o jfdctflt.o jfdctint.o jdapimin.o jdapistd.o jdtrans.o
jdatasrc.o
> jdmaster.o jdinpu
> t.o jdmarker.o jdhuff.o jdphuff.o jdmainct.o jdcoefct.o jdpostct.o
> jddctmgr.o jidctfst.o jidctflt.o jidctint.o jidctred.o jdsample.o
jdcolor.o
> jquant1.o jquant2
> .o jdmerge.o jcomapi.o jutils.o jerror.o jmemmgr.o jmemnobs.o
> jerror.o(.text+0x1a3b):jerror.c: undefined reference to `_impure_ptr'
> collect2: ld returned 1 exit status
> make: *** [jpeg.dll] Error 1
> 
> Earnie Boyd wrote:
> 
> > I didn't see anything particularly incorrect.  Anybody else?
> >
> > Next add a -H to the gcc command so that you can visually see what
> > headers are being included.  One of the headers being included has
the
> > _impure_ptr referenced.  It is undefined because the -mno-cygwin
> > switch doesn't include the libcygwin.a library where it would be
> > defined.  The headers from include and include/sys will be the ones
> > you'll need to be concerned about.  The ones from include/mingw32 or
> > include/mingw32/sys will be ok.  _reent is a structure that uses the
> > _impure_ptr so if you see it in the header you've also got a
problem.
> >
> > ---Maxim Sobolev <max@vega.pfts.com> wrote:
> > >
> > > I'm rebuilding jpeg.dll (on separate Mingw pakage it builds just
> > fine). Here
> > > is Makefile in attachment.
> > >
> > > Earnie Boyd wrote:
> > >
> > > > What library is it from?
> > > >
> > > > What command(s) are you or your package makefile using to
build your
> > > > executable?
> > > >
> >
> > <snip>
> >
> > ==
> > -                  \\||//
> > -------------o0O0--Earnie--0O0o--------------
> > --          earnie_boyd@yahoo.com          --
> > -- http://www.freeyellow.com/members5/gw32 --
> > ----------------ooo0O--O0ooo-----------------
> >
> > PS: Newbie's, you should visit my page.
> > _________________________________________________________
> > DO YOU YAHOO!?
> > Get your free @yahoo.com address at http://mail.yahoo.com
> >
> *asm:
> 
> 
> *asm_final:
> 
> 
> *cpp:
> -remap %(cpp_cpu) %{posix:-D_POSIX_SOURCE}  
%{!mno-cygwin:-D__CYGWIN32__ -D__CYGWIN__}  
%{mno-cygwin:-iwithprefixbefore ../../../../include/mingw32
-D__MINGW32__=0.2}
> 
> *cc1:
> %(cc1_spec) 
> 
> *cc1plus:
> 
> 
> *endfile:
> 
> 
> *link:
> %{mwindows:--subsystem windows}                    %{mdll:--dll -e
_DllMainCRTStartup@12}
> 
> *lib:
> %{pg:-lgmon}                   %{!mno-cygwin:-lcygwin}              
    %{mno-cygwin:-lmingw32 -lmoldname -lcrtdll}                  
%{mwindows:-luser32 -lgdi32 -lcomdlg32} 		  -lkernel32 -ladvapi32
-lshell32
> 
> *libgcc:
> -lgcc
> 
> *startfile:
> %{mdll: %{!mno-cygwin:dllcrt0%O%s}                                
%{mno-cygwin:dllcrt1%O%s}}                         %{!mdll:
%{!mno-cygwin:crt0%O%s}                                 
%{mno-cygwin:crt1%O%s} %{pg:gcrt0%O%s}}
> 
> *switches_need_spaces:
> 
> 
> *signed_char:
> %{funsigned-char:-D__CHAR_UNSIGNED__}
> 
> *predefines:
> -Di386 -D_WIN32   -DWINNT  -D_X86_=1 -D__STDC__=1 
-D__stdcall=__attribute__((__stdcall__))  
-D__cdecl=__attribute__((__cdecl__))  
-D__declspec(x)=__attribute__((x))   -Asystem(winnt) -Acpu(i386)
-Amachine(i386)
> 
> *cross_compile:
> 1
> 
> *version:
> egcs-2.91.57
> 
> *multilib:
> . ;
> 
> *multilib_defaults:
> 
> 
> *multilib_extra:
> 
> 
> *multilib_matches:
> 
> 
> *linker:
> collect2
> 
> *cpp_486:
> %{!ansi:-Di486} -D__i486 -D__i486__
> 
> *cpp_586:
> %{!ansi:-Di586 -Dpentium} 	-D__i586 -D__i586__ -D__pentium
-D__pentium__
> 
> *cpp_686:
> %{!ansi:-Di686 -Dpentiumpro} 	-D__i686 -D__i686__ -D__pentiumpro
-D__pentiumpro__
> 
> *cpp_cpu_default:
> %(cpp_586)
> 
> *cpp_cpu:
> -Acpu(i386) -Amachine(i386) %{!ansi:-Di386} -D__i386 -D__i386__
%{mcpu=i486:%(cpp_486)} %{m486:%(cpp_486)} %{mpentium:%(cpp_586)}
%{mcpu=pentium:%(cpp_586)} %{mpentiumpro:%(cpp_686)}
%{mcpu=pentiumpro:%(cpp_686)}
%{!mcpu*:%{!m486:%{!mpentium*:%(cpp_cpu_default)}}}
> 
> *cc1_cpu:
> %{!mcpu*: %{m386:-mcpu=i386 -march=i386} %{mno-486:-mcpu=i386
-march=i386} %{m486:-mcpu=i486 -march=i486} %{mno-386:-mcpu=i486
-march=i486} %{mno-pentium:-mcpu=i486 -march=i486}
%{mpentium:-mcpu=pentium} %{mno-pentiumpro:-mcpu=pentium}
%{mpentiumpro:-mcpu=pentiumpro}}
> 
> 

==
-                  \\||//
-------------o0O0--Earnie--0O0o--------------
--          earnie_boyd@yahoo.com          --
-- http://www.freeyellow.com/members5/gw32 --
----------------ooo0O--O0ooo-----------------

PS: Newbie's, you should visit my page.
_________________________________________________________
DO YOU YAHOO!?
Get your free @yahoo.com address at http://mail.yahoo.com

-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

Re: B20: bad support for Mingw?

[Earnie Boyd]

This is most likely due to cygwin built libraries being used.



---Maxim Sobolev <max@vega.pfts.com> wrote:
>
> There is one problems remaining after patching -iwithprefixbefore.
TIFF
> 3.6  library failed to compile properly on B20 with -mno-cygwin with
many
> "Parse Error" despite the fact that it configured and compiled on
Mingw32
> w/o problems. Unfortunately now I doesn't have a piece of time to
dig into
> this issue dipper, so if anybody wish I can send TIFF sources
configured
> and patched to compile under Mingw32.
> 
> Maxim
> 
> Geoffrey Noer wrote:
> 
> > On Tue, Nov 03, 1998 at 07:10:50AM -0800, Earnie Boyd wrote:
> > [...]
> > > I was able to resolve this by prefixing ../../../../ to the
> > > include/mingw32 in the -mno-cygwin commands for the cpp in the
specs
> > > file.
> > [...]
> >
> > I'm looking into this...
> >
> > --
> > Geoffrey Noer
> > noer@cygnus.com
> 
> 

==
-                  \\||//
-------------o0O0--Earnie--0O0o--------------
--          earnie_boyd@yahoo.com          --
-- http://www.freeyellow.com/members5/gw32 --
----------------ooo0O--O0ooo-----------------

PS: Newbie's, you should visit my page.
_________________________________________________________
DO YOU YAHOO!?
Get your free @yahoo.com address at http://mail.yahoo.com

-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

Re: B20: bad support for Mingw?

[Maxim Sobolev]

There is one problems remaining after patching -iwithprefixbefore. TIFF
3.6  library failed to compile properly on B20 with -mno-cygwin with many
"Parse Error" despite the fact that it configured and compiled on Mingw32
w/o problems. Unfortunately now I doesn't have a piece of time to dig into
this issue dipper, so if anybody wish I can send TIFF sources configured
and patched to compile under Mingw32.

Maxim

Geoffrey Noer wrote:

> On Tue, Nov 03, 1998 at 07:10:50AM -0800, Earnie Boyd wrote:
> [...]
> > I was able to resolve this by prefixing ../../../../ to the
> > include/mingw32 in the -mno-cygwin commands for the cpp in the specs
> > file.
> [...]
>
> I'm looking into this...
>
> --
> Geoffrey Noer
> noer@cygnus.com

-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

Re: B20: bad support for Mingw?

[Earnie Boyd]

---Geoffrey Noer <noer@cygnus.com> wrote:
>
> On Tue, Nov 03, 1998 at 07:10:50AM -0800, Earnie Boyd wrote:
> [...]
> > I was able to resolve this by prefixing ../../../../ to the
> > include/mingw32 in the -mno-cygwin commands for the cpp in the specs
> > file.
> [...]
> 
> I'm looking into this...
> 
> -- 
> Geoffrey Noer
> noer@cygnus.com
> 

Geoffrey,

With the default directory structure this ended up being
../../../../i586-cygwin32/include/mingw32 and since I move things to
more of a unix structure I didn't need the i586-cygwin32 but Maxim did.

==
-                  \\||//
-------------o0O0--Earnie--0O0o--------------
--          earnie_boyd@yahoo.com          --
-- http://www.freeyellow.com/members5/gw32 --
----------------ooo0O--O0ooo-----------------

PS: Newbie's, you should visit my page.
_________________________________________________________
DO YOU YAHOO!?
Get your free @yahoo.com address at http://mail.yahoo.com

-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

Re: B20: bad support for Mingw?

[Geoffrey Noer]

On Tue, Nov 03, 1998 at 07:10:50AM -0800, Earnie Boyd wrote:
[...]
> I was able to resolve this by prefixing ../../../../ to the
> include/mingw32 in the -mno-cygwin commands for the cpp in the specs
> file.
[...]

I'm looking into this...

-- 
Geoffrey Noer
noer@cygnus.com
-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

Re: B20: bad support for Mingw?

[Scott Herndon]

I tried fixing the specs file by prefixing /include/mingw with
"../../../../" as described in an earlier post and it did not work.  What
DID work hower was including the i586-cygwin32 directory in the prefix as
well.  Here is the relevant section from my specs file:
*cpp:
-remap %(cpp_cpu) %{posix:-D_POSIX_SOURCE}
%{!mno-cygwin:-D__CYGWIN32__ -D__CYGWIN__} %{mno-cygwin:-iwithprefixbefore
../../../../i586-cygwin32/include/mingw32 -D__MINGW32__=0.2}

-scott


>I didn't see anything particularly incorrect.  Anybody else?
>
>Next add a -H to the gcc command so that you can visually see what
>headers are being included.  One of the headers being included has the
>_impure_ptr referenced.  It is undefined because the -mno-cygwin
>switch doesn't include the libcygwin.a library where it would be
>defined.  The headers from include and include/sys will be the ones
>you'll need to be concerned about.  The ones from include/mingw32 or
>include/mingw32/sys will be ok.  _reent is a structure that uses the
>_impure_ptr so if you see it in the header you've also got a problem.
>


-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

Re: B20: bad support for Mingw?

[jeffdb]

This is from rsxntdj so some may not apply
I haven't found anything better though ;^)

On Sat, 7 Nov 1998 05:53:33 -0800 (PST), you wrote:

>
>I have no idea.  I believe that Mumit said once that it was documented
>well in the source code.  I just found it rather intuitive to read but
>documentation for all semantics would be nice.
>
>
>---Ulrich Lauther <ulrich.lauther@mchp.siemens.de> wrote:
>>
>> > 
>> > *cpp:
>> > -remap %(cpp_cpu) %{posix:-D_POSIX_SOURCE}  
>> > %{!mno-cygwin:-D__CYGWIN32__ -D__CY
>> > GWIN__}   %{mno-cygwin:-iwithprefixbefore
>../../../../include/mingw32
>> > -D__MINGW3
>> > 2__=0.2}
>> > 
>> > 
>> where can I find a description of syntax and semantics of spec-files?
>> 
>> -- 
>> 	-lauther
>> 
>>
>----------------------------------------------------------------------------
>> Ulrich Lauther          ph: +49 89 636 48834 fx: ... 636 42284
>> Siemens ZT SE 4         Internet: Ulrich.Lauther@mchp.siemens.de
>> 
>
>==
>-                  \\||//
>-------------o0O0--Earnie--0O0o--------------
>--          earnie_boyd@yahoo.com          --
>-- http://www.freeyellow.com/members5/gw32 --
>----------------ooo0O--O0ooo-----------------
>
>PS: Newbie's, you should visit my page.
>_________________________________________________________
>DO YOU YAHOO!?
>Get your free @yahoo.com address at http://mail.yahoo.com
>
>-
>For help on using this list (especially unsubscribing), send a message to
>"gnu-win32-request@cygnus.com" with one line of text: "help".

Re: B20: bad support for Mingw?

[Fergus Henderson]

> > where can I find a description of syntax and semantics of spec-files?

Near line 250 in the file gcc.c in the sources to gcc.

-- 
Fergus Henderson <fjh@cs.mu.oz.au>  |  "Binaries may die
WWW: < http://www.cs.mu.oz.au/~fjh >  |   but source code lives forever"
PGP: finger fjh@128.250.37.3        |     -- leaked Microsoft memo.
-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

Re: B20: bad support for Mingw?

[Dmitry Bely]

Hello Earnie,

You wrote:

> For everyone else on the list,
>
> If you plan to use the -mno-cygwin switch you will need to modify the
> specs file.  Add ../../../../ to the include/mingw32 in the cpp
> section.  It will look like this when your done:
>
> *cpp:
> -remap %(cpp_cpu) %{posix:-D_POSIX_SOURCE}
> %{!mno-cygwin:-D__CYGWIN32__ -D__CY
> GWIN__}   %{mno-cygwin:-iwithprefixbefore ../../../../include/mingw32
> -D__MINGW3
> 2__=0.2}

Alas, this won't help in my case too. As far as I understood, cpp
simply ignore -iwithprefixbefore option, as well as any other unknown
option starting with -i :

cpp -iadracadabra hello.c

won't ever generate error message.

I temporarily resolved the problem by changing -iwithprefixbefore <path>
to -isystem <full-path> in the spec file, but it cannot be a long-term
solution.

any other ideas?

Best regards,
 Dmitry                          mailto:dbely@usa.net


-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

Re: B20: bad support for Mingw?

[Earnie Boyd]

I have no idea.  I believe that Mumit said once that it was documented
well in the source code.  I just found it rather intuitive to read but
documentation for all semantics would be nice.


---Ulrich Lauther <ulrich.lauther@mchp.siemens.de> wrote:
>
> > 
> > *cpp:
> > -remap %(cpp_cpu) %{posix:-D_POSIX_SOURCE}  
> > %{!mno-cygwin:-D__CYGWIN32__ -D__CY
> > GWIN__}   %{mno-cygwin:-iwithprefixbefore
../../../../include/mingw32
> > -D__MINGW3
> > 2__=0.2}
> > 
> > 
> where can I find a description of syntax and semantics of spec-files?
> 
> -- 
> 	-lauther
> 
>
----------------------------------------------------------------------------
> Ulrich Lauther          ph: +49 89 636 48834 fx: ... 636 42284
> Siemens ZT SE 4         Internet: Ulrich.Lauther@mchp.siemens.de
> 

==
-                  \\||//
-------------o0O0--Earnie--0O0o--------------
--          earnie_boyd@yahoo.com          --
-- http://www.freeyellow.com/members5/gw32 --
----------------ooo0O--O0ooo-----------------

PS: Newbie's, you should visit my page.
_________________________________________________________
DO YOU YAHOO!?
Get your free @yahoo.com address at http://mail.yahoo.com

-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

Re: B20: bad support for Mingw?

[Earnie Boyd]

---Dmitry Bely <dbely@usa.net> wrote:
>
> Hello Earnie,
> 
> You wrote:
> 
> > For everyone else on the list,
> >
> > If you plan to use the -mno-cygwin switch you will need to modify
the
> > specs file.  Add ../../../../ to the include/mingw32 in the cpp
> > section.  It will look like this when your done:
> >
> > *cpp:
> > -remap %(cpp_cpu) %{posix:-D_POSIX_SOURCE}
> > %{!mno-cygwin:-D__CYGWIN32__ -D__CY
> > GWIN__}   %{mno-cygwin:-iwithprefixbefore
../../../../include/mingw32
> > -D__MINGW3
> > 2__=0.2}
> 
> Alas, this won't help in my case too. As far as I understood, cpp
> simply ignore -iwithprefixbefore option, as well as any other unknown
> option starting with -i :
> 
> cpp -iadracadabra hello.c
> 
> won't ever generate error message.
> 
> I temporarily resolved the problem by changing -iwithprefixbefore
<path>
> to -isystem <full-path> in the spec file, but it cannot be a long-term
> solution.
> 
> any other ideas?

Are you sure that you are using the correct cpp?  The info file for
cpp says that -iwithprefixbefore is a valid option.

What does the output of `cygcheck -s -v' say?  If says that it can't
find cpp then you are ok.  If it does find cpp then you've got a
problem with the PATH variable.

==
-                  \\||//
-------------o0O0--Earnie--0O0o--------------
--          earnie_boyd@yahoo.com          --
-- http://www.freeyellow.com/members5/gw32 --
----------------ooo0O--O0ooo-----------------

PS: Newbie's, you should visit my page.
_________________________________________________________
DO YOU YAHOO!?
Get your free @yahoo.com address at http://mail.yahoo.com

-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

Re: B20: bad support for Mingw?

[Earnie Boyd]

Hi Maxim,

I got your zip package and built it just fine.  It picked up the
mingw32 headers, even built the dll; impressive.

Now let's figure out what's wrong with your setup.

Modify the specs file one more time.  Change the version to something
your going to recognize as your modified specs file.  Now issue the
command "gcc -v" and you should see your modified version number if
this is the specs file being read.

For everyone else on the list,

If you plan to use the -mno-cygwin switch you will need to modify the
specs file.  Add ../../../../ to the include/mingw32 in the cpp
section.  It will look like this when your done:

*cpp:
-remap %(cpp_cpu) %{posix:-D_POSIX_SOURCE}  
%{!mno-cygwin:-D__CYGWIN32__ -D__CY
GWIN__}   %{mno-cygwin:-iwithprefixbefore ../../../../include/mingw32
-D__MINGW3
2__=0.2}



==
-                  \\||//
-------------o0O0--Earnie--0O0o--------------
--          earnie_boyd@yahoo.com          --
-- http://www.freeyellow.com/members5/gw32 --
----------------ooo0O--O0ooo-----------------

PS: Newbie's, you should visit my page.
_________________________________________________________
DO YOU YAHOO!?
Get your free @yahoo.com address at http://mail.yahoo.com

-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

Re: B20: bad support for Mingw?

[Earnie Boyd]

What library is it from?

What command(s) are you or your package makefile using to build your
executable?

I've just succesfully built a native "HelloWorld" type program with no
dependency on cygwin1.dll.

The _impure_ptr means that you are mixing in cygwin1.dll dependencies.

---Maxim Sobolev <max@vega.pfts.com> wrote:
>
> Thank you it helped a bit but I still on link phase have an "undefined
> reference to `_impure_ptr'" error.
> 
> Maxim
> 
> Earnie Boyd wrote:
> 
> > ---Maxim Sobolev <max@vega.pfts.com> wrote:
> > >
> > > It seems to me that "-mno-cygwin" switch doesn't make it magick to
> > > include Mingw32 version of header files (program that just fine
> > compiled
> > > with Mingw32 with B20 doesn't). Maybe there is any other switch to
> > > include all needed header (Mingw instead of Cygwin) files? Any
ideas?
> > >
> >
> > I was able to resolve this by prefixing ../../../../ to the
> > include/mingw32 in the -mno-cygwin commands for the cpp in the specs
> > file.
> >
> > Here is the section of the specs file with the switches that work:
> > *cpp:
> > -remap %(cpp_cpu) %{posix:-D_POSIX_SOURCE}
> > %{!mno-cygwin:-D__CYGWIN32__ -D__CY
> > GWIN__}   %{mno-cygwin:-iwithprefixbefore
../../../../include/mingw32
> > -D__MINGW3
> > 2__=0.2}
> >
> > I verifyed that it picked up the correct header file by specifying
the
> > -H switch on the command line verifying that
include/mingw32/dirent.h
> > was included instead of include/dirent.h.
> >
> > ==
> > -                  \\||//
> > -------------o0O0--Earnie--0O0o--------------
> > --          earnie_boyd@yahoo.com          --
> > -- http://www.freeyellow.com/members5/gw32 --
> > ----------------ooo0O--O0ooo-----------------
> >
> > PS: Newbie's, you should visit my page.
> > _________________________________________________________
> > DO YOU YAHOO!?
> > Get your free @yahoo.com address at http://mail.yahoo.com
> 
> 

==
-                  \\||//
-------------o0O0--Earnie--0O0o--------------
--          earnie_boyd@yahoo.com          --
-- http://www.freeyellow.com/members5/gw32 --
----------------ooo0O--O0ooo-----------------

PS: Newbie's, you should visit my page.
_________________________________________________________
DO YOU YAHOO!?
Get your free @yahoo.com address at http://mail.yahoo.com

-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

Re: B20: bad support for Mingw?

[Maxim Sobolev]

Here is output produced with -H switch, to me it seems that all inclides are
CYGWIN **NOT* MINGW! See my specs file also.

gcc -mno-cygwin -H -O3  -I.   -c jerror.c -o jerror.o
jinclude.h
 jconfig.h
 c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\include\stddef.h

 c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\..\..\..\..\i586-cygwin32\include\stdlib.h


c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\..\..\..\..\i586-cygwin32\include\_ansi.h


c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\..\..\..\..\i586-cygwin32\include\sys/config.h


c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\include\stddef.h


c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\..\..\..\..\i586-cygwin32\include\sys/reent.h

 c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\..\..\..\..\i586-cygwin32\include\stdio.h


c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\include\stddef.h


c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\include\stdarg.h

 c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\..\..\..\..\i586-cygwin32\include\string.h


c:\cygnus\CYGWIN~1\H-I586~1\bin\..\lib\gcc-lib\i586-cygwin32\egcs-2.91.57\include\stddef.h

jpeglib.h
 jmorecfg.h
jversion.h
jerror.h
jerror.h
gcc -mno-cygwin -mdll -o junk.tmp -Wl,--base-file,base.tmp jcapimin.o
jcapistd.o jctrans.o jcparam.o jdatadst.o jcinit.o jcmaster.o jcmarker.o
jcmainct.o jcprep
ct.o jccoefct.o jccolor.o jcsample.o jchuff.o jcphuff.o jcdctmgr.o
jfdctfst.o jfdctflt.o jfdctint.o jdapimin.o jdapistd.o jdtrans.o jdatasrc.o
jdmaster.o jdinpu
t.o jdmarker.o jdhuff.o jdphuff.o jdmainct.o jdcoefct.o jdpostct.o
jddctmgr.o jidctfst.o jidctflt.o jidctint.o jidctred.o jdsample.o jdcolor.o
jquant1.o jquant2
.o jdmerge.o jcomapi.o jutils.o jerror.o jmemmgr.o jmemnobs.o
jerror.o(.text+0x1a3b):jerror.c: undefined reference to `_impure_ptr'
collect2: ld returned 1 exit status
make: *** [jpeg.dll] Error 1

Earnie Boyd wrote:

> I didn't see anything particularly incorrect.  Anybody else?
>
> Next add a -H to the gcc command so that you can visually see what
> headers are being included.  One of the headers being included has the
> _impure_ptr referenced.  It is undefined because the -mno-cygwin
> switch doesn't include the libcygwin.a library where it would be
> defined.  The headers from include and include/sys will be the ones
> you'll need to be concerned about.  The ones from include/mingw32 or
> include/mingw32/sys will be ok.  _reent is a structure that uses the
> _impure_ptr so if you see it in the header you've also got a problem.
>
> ---Maxim Sobolev <max@vega.pfts.com> wrote:
> >
> > I'm rebuilding jpeg.dll (on separate Mingw pakage it builds just
> fine). Here
> > is Makefile in attachment.
> >
> > Earnie Boyd wrote:
> >
> > > What library is it from?
> > >
> > > What command(s) are you or your package makefile using to build your
> > > executable?
> > >
>
> <snip>
>
> ==
> -                  \\||//
> -------------o0O0--Earnie--0O0o--------------
> --          earnie_boyd@yahoo.com          --
> -- http://www.freeyellow.com/members5/gw32 --
> ----------------ooo0O--O0ooo-----------------
>
> PS: Newbie's, you should visit my page.
> _________________________________________________________
> DO YOU YAHOO!?
> Get your free @yahoo.com address at http://mail.yahoo.com

Re: B20: bad support for Mingw?

[Maxim Sobolev]

I'm rebuilding jpeg.dll (on separate Mingw pakage it builds just fine). Here
is Makefile in attachment.

Earnie Boyd wrote:

> What library is it from?
>
> What command(s) are you or your package makefile using to build your
> executable?
>
> I've just succesfully built a native "HelloWorld" type program with no
> dependency on cygwin1.dll.
>
> The _impure_ptr means that you are mixing in cygwin1.dll dependencies.
>
> ---Maxim Sobolev <max@vega.pfts.com> wrote:
> >
> > Thank you it helped a bit but I still on link phase have an "undefined
> > reference to `_impure_ptr'" error.
> >
> > Maxim
> >
> > Earnie Boyd wrote:
> >
> > > ---Maxim Sobolev <max@vega.pfts.com> wrote:
> > > >
> > > > It seems to me that "-mno-cygwin" switch doesn't make it magick to
> > > > include Mingw32 version of header files (program that just fine
> > > compiled
> > > > with Mingw32 with B20 doesn't). Maybe there is any other switch to
> > > > include all needed header (Mingw instead of Cygwin) files? Any
> ideas?
> > > >
> > >
> > > I was able to resolve this by prefixing ../../../../ to the
> > > include/mingw32 in the -mno-cygwin commands for the cpp in the specs
> > > file.
> > >
> > > Here is the section of the specs file with the switches that work:
> > > *cpp:
> > > -remap %(cpp_cpu) %{posix:-D_POSIX_SOURCE}
> > > %{!mno-cygwin:-D__CYGWIN32__ -D__CY
> > > GWIN__}   %{mno-cygwin:-iwithprefixbefore
> ../../../../include/mingw32
> > > -D__MINGW3
> > > 2__=0.2}
> > >
> > > I verifyed that it picked up the correct header file by specifying
> the
> > > -H switch on the command line verifying that
> include/mingw32/dirent.h
> > > was included instead of include/dirent.h.
> > >
> > > ==
> > > -                  \\||//
> > > -------------o0O0--Earnie--0O0o--------------
> > > --          earnie_boyd@yahoo.com          --
> > > -- http://www.freeyellow.com/members5/gw32 --
> > > ----------------ooo0O--O0ooo-----------------
> > >
> > > PS: Newbie's, you should visit my page.
> > > _________________________________________________________
> > > DO YOU YAHOO!?
> > > Get your free @yahoo.com address at http://mail.yahoo.com
> >
> >
>
> ==
> -                  \\||//
> -------------o0O0--Earnie--0O0o--------------
> --          earnie_boyd@yahoo.com          --
> -- http://www.freeyellow.com/members5/gw32 --
> ----------------ooo0O--O0ooo-----------------
>
> PS: Newbie's, you should visit my page.
> _________________________________________________________
> DO YOU YAHOO!?
> Get your free @yahoo.com address at http://mail.yahoo.com

Re: B20: bad support for Mingw?

[Maxim Sobolev]

Thank you it helped a bit but I still on link phase have an "undefined
reference to `_impure_ptr'" error.

Maxim

Earnie Boyd wrote:

> ---Maxim Sobolev <max@vega.pfts.com> wrote:
> >
> > It seems to me that "-mno-cygwin" switch doesn't make it magick to
> > include Mingw32 version of header files (program that just fine
> compiled
> > with Mingw32 with B20 doesn't). Maybe there is any other switch to
> > include all needed header (Mingw instead of Cygwin) files? Any ideas?
> >
>
> I was able to resolve this by prefixing ../../../../ to the
> include/mingw32 in the -mno-cygwin commands for the cpp in the specs
> file.
>
> Here is the section of the specs file with the switches that work:
> *cpp:
> -remap %(cpp_cpu) %{posix:-D_POSIX_SOURCE}
> %{!mno-cygwin:-D__CYGWIN32__ -D__CY
> GWIN__}   %{mno-cygwin:-iwithprefixbefore ../../../../include/mingw32
> -D__MINGW3
> 2__=0.2}
>
> I verifyed that it picked up the correct header file by specifying the
> -H switch on the command line verifying that include/mingw32/dirent.h
> was included instead of include/dirent.h.
>
> ==
> -                  \\||//
> -------------o0O0--Earnie--0O0o--------------
> --          earnie_boyd@yahoo.com          --
> -- http://www.freeyellow.com/members5/gw32 --
> ----------------ooo0O--O0ooo-----------------
>
> PS: Newbie's, you should visit my page.
> _________________________________________________________
> DO YOU YAHOO!?
> Get your free @yahoo.com address at http://mail.yahoo.com

-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

Re: B20: bad support for Mingw?

[Earnie Boyd]

I didn't see anything particularly incorrect.  Anybody else?

Next add a -H to the gcc command so that you can visually see what
headers are being included.  One of the headers being included has the
_impure_ptr referenced.  It is undefined because the -mno-cygwin
switch doesn't include the libcygwin.a library where it would be
defined.  The headers from include and include/sys will be the ones
you'll need to be concerned about.  The ones from include/mingw32 or
include/mingw32/sys will be ok.  _reent is a structure that uses the
_impure_ptr so if you see it in the header you've also got a problem.

---Maxim Sobolev <max@vega.pfts.com> wrote:
>
> I'm rebuilding jpeg.dll (on separate Mingw pakage it builds just
fine). Here
> is Makefile in attachment.
> 
> Earnie Boyd wrote:
> 
> > What library is it from?
> >
> > What command(s) are you or your package makefile using to build your
> > executable?
> >

<snip>

==
-                  \\||//
-------------o0O0--Earnie--0O0o--------------
--          earnie_boyd@yahoo.com          --
-- http://www.freeyellow.com/members5/gw32 --
----------------ooo0O--O0ooo-----------------

PS: Newbie's, you should visit my page.
_________________________________________________________
DO YOU YAHOO!?
Get your free @yahoo.com address at http://mail.yahoo.com

-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

Re: B20: bad support for Mingw?

[Earnie Boyd]

---Maxim Sobolev <max@vega.pfts.com> wrote:
>
> It seems to me that "-mno-cygwin" switch doesn't make it magick to
> include Mingw32 version of header files (program that just fine
compiled
> with Mingw32 with B20 doesn't). Maybe there is any other switch to
> include all needed header (Mingw instead of Cygwin) files? Any ideas?
> 

I was able to resolve this by prefixing ../../../../ to the
include/mingw32 in the -mno-cygwin commands for the cpp in the specs
file.

Here is the section of the specs file with the switches that work:
*cpp:
-remap %(cpp_cpu) %{posix:-D_POSIX_SOURCE}  
%{!mno-cygwin:-D__CYGWIN32__ -D__CY
GWIN__}   %{mno-cygwin:-iwithprefixbefore ../../../../include/mingw32
-D__MINGW3
2__=0.2}

I verifyed that it picked up the correct header file by specifying the
-H switch on the command line verifying that include/mingw32/dirent.h
was included instead of include/dirent.h.

==
-                  \\||//
-------------o0O0--Earnie--0O0o--------------
--          earnie_boyd@yahoo.com          --
-- http://www.freeyellow.com/members5/gw32 --
----------------ooo0O--O0ooo-----------------

PS: Newbie's, you should visit my page.
_________________________________________________________
DO YOU YAHOO!?
Get your free @yahoo.com address at http://mail.yahoo.com

-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

B20: bad support for Mingw?

[Maxim Sobolev]

It seems to me that "-mno-cygwin" switch doesn't make it magick to
include Mingw32 version of header files (program that just fine compiled
with Mingw32 with B20 doesn't). Maybe there is any other switch to
include all needed header (Mingw instead of Cygwin) files? Any ideas?

Maxim

-
For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".