[Gcl-devel] Re: gcl-2.7.0 on FreeBSD

[Camm Maguire]

Greetings!

Gernot Hueber <address@hidden> writes:

> Dear Camm,
> 
> I have removed all notes on dyn.lib loading until gcl-2.7.0 is running.
> Well, gcl still does build. See some snippets below.
> 

OK

> BTW are my options to configure reasonable?
> 
> maxwell# ./configure  --disable-dynsysbfd --disable-statsysbfd
> --enable-locbfd --disable-custreloc --enable-ansi
> 

Yes!


> -------------
> >From you last mail:
> 
> > OK, I need to add a little configure logic to determine what the added
> > function calls are when compiling with -pg, or drop gprof profiling
> > support on freebsd, etc.  Right now you can just not use
> > --enable-gprof (default is off) and comment out these lines from
> > sfasli.c:
> > 
> >     sethash(make_simple_string("_mcount"),
> >         sSAlink_hash_tableA->s.s_dbind,
> >         make_fixnum((fixnum)_mcount));
> >     sethash(make_simple_string("mcount"),
> >         sSAlink_hash_tableA->s.s_dbind,
> >         make_fixnum((fixnum)_mcount));
> > 

Should be automatic now with a fresh cvs checkout, only without
--enable-gprof.  We have to look later at getting profiling working on
bsd.

> > 
> > Please let me know if problems persist.
> > 
> > 
> 
> Furthermore I commented the call to on_exit in o/file.d for the moment?!

We could expand this ifdef:

#if defined(DARWIN)
#define on_exit(a,b)
#else
static void
rmc(int e,void *pid) {

  kill((int)(unsigned long)pid,SIGTERM);

}
#endif
#endif


But does bfd not have an on_exit equivalent?

> ----------
> 
> ;; Loading boot.lisp^M
> ;; Compiling ../lsp/gcl_listlib.lsp.^M
> Warning: NTH-VALUE is being redefined.^M
> ;; End of Pass 1.  ^M
> ;; End of Pass 2.  ^M
> In file included from ../lsp/gcl_listlib.c:2:^M
> /usr/home/ghue/src/gclcvs-2.7.0/unixport/../h/cmpinclude.h:1376: error: 
> syntax error before "if"^M
> ^M
> Correctable error: (SYSTEM "gcc -c -fsigned-char -pipe -Wall   
> -I/usr/home/ghue/src/gclcvs-2.7.0/unixport/../h   -c ../lsp/gcl_listlib.c -o 
> ../lsp/gcl_listlib.o ") returned a non-zero value 0.^M
> Signalled by UNLESS.^M
> If continued: Continues anyway.^M
> (SYSTEM "gcc -c -fsigned-char -pipe -Wall   
> -I/usr/home/ghue/src/gclcvs-2.7.0/unixport/../h   -c ../lsp/gcl_listlib.c -o 
> ../lsp/gcl_listlib.o ") returned a non-zero value 0.^M
> ^M
> Broken at UNLESS.  Type :H for Help.^M
> SYSTEM>>gmake: *** [unixport/saved_pre_gcl] Error 255^M
> 
> ----------
> At this point I did some minor change in h/cmpinclude.h
> 
> 
> #define __sgetc(p) (--(p)->_r < 0 ? __srget(p) : (int)(*(p)->_p++))
> 

This is bizarre, ad this file is just output from cpp -- this file
will not compile -- maybe a cpp bug?

> /* ghue
>  if (--_p->_w >= 0 || (_p->_w >= _p->_lbfsize && (char)_c != '\n'))
>   return (*_p->_p++ = _c);
>  else
>   return (__swbuf(_c, _p));
> }

Don't understand the '}' here.  Reference to '{' where?

> */
> #define __sfeof(p) (((p)->_flags & __SEOF) != 0)
> #define __sferror(p) (((p)->_flags & __SERR) != 0)
> #define __sclearerr(p) ((void)((p)->_flags &= ~(__SERR|__SEOF)))
> #define __sfileno(p) ((p)->_file)
> 
> ----------
> 
> maxwell# gmake^M^M
> 
> 
> 
> [...]
> 
> ;; Compiling ../lsp/gcl_recompile.lsp.
> ;; End of Pass 1.
> ;; End of Pass 2.
> In file included from ../lsp/gcl_recompile.c:2:
> /usr/home/ghue/src/gclcvs-2.7.0/unixport/../h/cmpinclude.h:5057:
> warning: return type defaults to `int'
> /usr/home/ghue/src/gclcvs-2.7.0/unixport/../h/cmpinclude.h:5066:
> warning: return type defaults to `int'

What is this line?

> /usr/home/ghue/src/gclcvs-2.7.0/unixport/../h/cmpinclude.h:5072:
> warning: return type defaults to `int'
> ;; OPTIMIZE levels: Safety=0 (No runtime error checking), Space=0,
> Speed=0, (Debug quality ignored)
> ;; Finished compiling ../lsp/gcl_recompile.o.
> real time       :      0.000 secs
> run-gbc time    :      0.000 secs
> child run time  :      0.055 secs
> gbc time        :      0.000 secs
> ;; Compiling ../lsp/gcl_callhash.lsp.
> ; (DEFUN DO-RECOMPILE ...) is being compiled.
> ;;; The declaration (DECLARE (IGNORE TPN)) was found in a bad place.
> Error: ERROR "The tag (NIL) is undefined."
> Signalled by COMPILER::CMPERR.
> ERROR "The tag (NIL) is undefined."


This should be fixed now in a fresh cvs checkout.
> 
> Broken at COMPILER::CMPERR.  Type :H for Help.
> SYSTEM>>gmake: *** [unixport/saved_pre_gcl] Error 255
> maxwell#
> 
> ---------
> 
> Now I am stuck again - no idea what to comment ;-)
> 

Thanks so much for your report!  bsd is quite close.

Take care,

> Best regards,
> 
> Gernot
> 
> 
> -- 
> Gernot Hueber <address@hidden>
> 
> 
> 

-- 
Camm Maguire                                            address@hidden
==========================================================================
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah