This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
Re: Solaris 8/SPARC libgcj bootstrap broken by recent fix-header change
- From: Neil Booth <neil at daikokuya dot co dot uk>
- To: Rainer Orth <ro at TechFak dot Uni-Bielefeld dot DE>
- Cc: gcc-bugs at gcc dot gnu dot org, "Kaveh R. Ghazi" <ghazi at caip dot rutgers dot edu>
- Date: Wed, 2 Apr 2003 22:01:04 +0100
- Subject: Re: Solaris 8/SPARC libgcj bootstrap broken by recent fix-header change
- References: <16004.34881.805189.404971@xayide.TechFak.Uni-Bielefeld.DE> <20030328211652.GA520@daikokuya.co.uk> <16008.17263.390036.136125@xayide.TechFak.Uni-Bielefeld.DE> <20030331194059.GB520@daikokuya.co.uk> <16010.5921.758682.809862@xayide.TechFak.Uni-Bielefeld.DE>
Rainer Orth wrote:-
> Neil Booth writes:
>
> > Does it work if fix-header doesn't pass any -I options? As I mention
> > in the log above, they were being passed on the command line and
> > ignored. I "fixed" that with my patch. It's the only thing I can see
> > of any relevance.
> >
> > Does this patch fix it? If not, I'd appreciate if you could investigate
> > exactly why it is fixing more than before - the decisions about what to
> > fix aren't directly affected by my patch, and I don't have a system
> > that even uses fix-header.
>
> Unfortunately, the patch doesn't help at all. I've instrumented
> fix-header.c (recognized_function) to print the function declarations
> scan-decls.c (scan_decls) finds: the old version (before the patch in
> question) finds a long list of declarations, kill among them:
>
> % ./fix-header signal.h include/signal.h include/signal.h -D__STDC__=0 -D__cplusplus -Iinclude -I/usr/include
Ah, there's the clue. We need to defer the -D switches like the front
end, since we appear to pass -D options that override built-ins.
I'll fix it in the next day or two; I'm really snowed under at the
moment.
Neil.