Re: Problem with gcc 3.0 and 3.0.1 and libstdc++

[Andrew Pinski <pinskia at physics dot uc dot edu>]

This problem is because "-I/usr/include" is being include in g++ 
command line
option and it breaks #include_next.  This has be mention at 
least once before,
I think someone notified QT about it; they should not be using 
"-I/usr/include".

Thanks,
Andrew Pinski


On Tuesday, July 3, 2001, at 10:10 PM, Luke Mauldin wrote:

> Hello, I have downloaded gcc 3.0 and I recently tried 3.0.1 
> snapshots from
> the CVS.  To build gcc I use the gcc_build script that comes 
> with gcc and I
> input my configure options, etc...  When compiling C++ programs 
> that make use
> of libstdc++ I get errors such as these:
> g++ -DHAVE_CONFIG_H -I. -I. -I../../.. -I/usr/include
> -I/usr/lib/qt-2.3.0/include -I/usr/X11R6/include   -O2 -march=athlon
> -mcpu=athlon -D_REENTRANT -I/usr/include/sword -DQT_NO_ASCII_CAST  -O2
> -fno-exceptions -fno-check-new -Wall -pedantic -W -Wpointer-arith
> -Wmissing-prototypes -Wwrite-strings -Wno-long-long -fno-builtin -O2
> -march=athlon -mcpu=athlon -c cswordmodulesearch.cpp
> In file included from /usr/include/g++-v3/bits/char_traits.h:39,
>                  from /usr/include/g++-v3/bits/std_string.h:41,
>                  from /usr/include/g++-v3/string:31,
>                  from /usr/include/sword/swmgr.h:40,
>                  from cswordbackend.h:29,
>                  from cswordmoduleinfo.h:23,
>                  from cswordmodulesearch.cpp:20:
> /usr/include/g++-v3/bits/std_cstring.h:40:25: string.h: No such file or
> directory
>
> I look in /usr/include/g++-v3/bits/std_cstring.h here are lines 
> 39 and 40.
> #pragma GCC system_header
> #include_next <string.h>
>
> If I replace <string.h> with </usr/include/string.h>  
> everything works fine.
> This is not just a problem with std_cstring.h, it is a problem 
> in every file
> in /usr/include/g++-v3/bits/  In order to get programs to 
> compile properly I
> have to replace just #include_next <****.h> with #include_next
> </usr/include****.h>  I have been following the mailing lists 
> for the last
> several months, and I have not seen these problem reported 
> before.  Is it
> something I am doing wrong, or is it a problem with gcc?
>
> Luke
>
>