This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
Re: [patch for libg++] Re: 1990314 fails to build
- To: N8TM at aol dot com
- Subject: Re: [patch for libg++] Re: 1990314 fails to build
- From: Jeffrey A Law <law at upchuck dot cygnus dot com>
- Date: Thu, 01 Apr 1999 02:05:35 -0700
- cc: manfred at s-direktnet dot de, mh at exept dot de, oliva at dcc dot unicamp dot br, martin at mira dot isdn dot cs dot tu-berlin dot de, egcs-bugs at egcs dot cygnus dot com, Bruce Korb <korb at datadesign dot com>
- Reply-To: law at cygnus dot com
In message <1191d054.36efb6d0@aol.com>you write:
> In a message dated 3/17/99 1:49:20 AM Pacific Standard Time, manfred@s-
> direktnet.de writes:
>
> > Could it probably be caused by the recent activation of fast-fixincludes?
> > I
> When I built on HPUX10.20 yesterday, fixincludes missed sys/time.h, which I
> had to change manually. Without this, iostream.cc doesn't compile.
It looks like this *should* have been broken for a while, but apparently
there's a subtle difference in how fixinc vs fixincludes is working.
Basically the old fixincludes was supposed to undo a bogus fix to sys/time.h
on hpux11. But it got the wrong match string which caused the match to occur
on hpux10.20 too.
For reasons unknown this did not break the old fixincludes code, but does
break with fast-fixincludes.
Short term I've fixed the trigger string for time.h and one other file it
got wrong. Long term I think those "undo bogus fixes" things can go away
since we're getting better about how handling of '//' appearing in system
header files.
jeff