This is the mail archive of the
binutils@sourceware.org
mailing list for the binutils project.
Re: gdb/2045: gdb 6.4 compilation fails on HPUX 10
- From: Daniel Jacobowitz <drow at false dot org>
- To: Mark Kettenis <mark dot kettenis at xs4all dot nl>
- Cc: gdb-gnats at sources dot redhat dot com, binutils at sourceware dot org
- Date: Mon, 5 Dec 2005 17:19:08 -0500
- Subject: Re: gdb/2045: gdb 6.4 compilation fails on HPUX 10
- References: <20051205213801.17245.qmail@sourceware.org> <20051205215612.GA15126@nevyn.them.org> <200512052214.jB5ME79m019072@elgar.sibelius.xs4all.nl>
On Mon, Dec 05, 2005 at 11:14:07PM +0100, Mark Kettenis wrote:
> > Date: Mon, 5 Dec 2005 16:56:12 -0500
> > From: Daniel Jacobowitz <drow@false.org>
> >
> > On Mon, Dec 05, 2005 at 09:38:01PM -0000, Mark Kettenis wrote:
> > > Hi,
> > >
> > > I've been looking into the problem today. I remember stumbling across
> > > the -D_LARGEFILE64_SOURCE issue, and could indeed reproduce it. I'm
> > > working on a proper fix, but it seems the HP-UX 10.20 header files are
> > > seriously borked, so the "right" fix doesn't quite work :(.
> >
> > Can you reproduce this with HEAD? I'm pretty sure one of the HP/UX
> > binutils developers fixed this, possibly just after GDB 6.4 branched.
>
> Oh yes, sorry, I reproduces this with head (didn't try with the
> official gdb 6.4 release). I tried to fix this by using
> AC_SYS_LARGEFILE and dumping the fseeko64/ftello64 stuff.
> Unfortunately the HP-UX 10.20 header files are broken for
> -D_FILE_OFFSET_BITS=64. In <sys/fcntl.h> they do something like:
>
> static open(a,b,c) const char *a; mode_t c; { return __open64(a,b,c); }
>
> without providing a protytype for __open64. Of course that makes gcc
> spit out all kinds of warnings, which are fatal with -Werror.
>
> Not sure what to do about it yet :(.
Steve, I really thought you'd fixed this... any comment?
--
Daniel Jacobowitz
CodeSourcery, LLC