Re: [patch] mkcshadow improvements

[Nathan Myers <ncm at cantrip dot org>]

On Sun, May 14, 2000 at 11:30:19PM -0700, Benjamin Kosnik wrote:
> >  ../../../gcc/libstdc++-v3/inclosure \
> >    -I ../../../v3/lib/gcc-lib/i686-pc-linux-gnu/2.96/include/ \
> >    -I /usr/include \
> >    -G machine/ansi.h \
> >   | ../../../gcc/libstdc++-v3/mkcshadow 
> 
> Okay. Let me try to decipher what you've done:
> 
> GCC_EXEC_PREFIX/include
> /usr/include

Actually, that should probably be 

  -I ../../gcc/include \

because the "v3" above is the installed location for the headers, 
and we haven't necessarily installed them yet.  The rest (represented 
in the example by the single directory /usr/include, which is enough
on Linux) are all the places where a C compiler on the target system 
looks for headers.
 
> but then what's the machine/ansi.h stuff, and how would this translate 
> to solaris, for instance?
> 
> >The appropriate "-I" arguments must be discovered by configure, and 
> >the -G argument(s) depend on the target build environment.
> 
> Can you explain how configure is to discover them? At least what it
> should look for?

The "-G" argument(s), if any, have to be listed explicitly in the 
configure scripts, in config/{aix,gnu-linux,solaris,irix,bsd}/.  
The way to find out what to put there is to run inclosure without 
any "-G" arguments and see what it can't find, then verify that the 
headers it complains about really don't exist or aren't used.  If
they really are used, it means you haven't supplied enough "-I"
options.

I guess that's a good reason to keep the inclosure script separate.

The list of "-I" options (aside from the first, ../../gcc/include) 
is identically the set of non-C++ header directories searched by 
"g++" or "gcc" with this compiler.

I left a task off the list: 

  - make the target g++ search the shadow/ and cshadow/ directories 
    before the other C header directories, if configured with
    --enable-shadow-headers. 

Nathan Myers
ncm at cantrip dot org