Re: glib 2.2.3 test failures on Solaris

[Owen Taylor <otaylor redhat com>]

On Mon, 2003-09-15 at 14:02, Albert Chin wrote:
> On Mon, Sep 15, 2003 at 06:28:10AM -0400, Owen Taylor wrote:
> > On Sat, 2003-09-13 at 18:00, Albert Chin wrote:
> > > I'm getting a test failure in strtod-test on glib-2.2.3 on Solaris
> > > 2.5.1-9/SPARC with Sun's C compiler, HP-UX 11.00 with the HP C
> > > compiler, and AIX 5.1 with the IBM C compiler. The problem is:
> > 
> > This is:
> > 
> >  http://bugzilla.gnome.org/show_bug.cgi?id=101874
> > 
> > The problem is that the trio implementation of printf() is not accurate
> > enough for doubles. In glib-2.4, we are using the gnulib printf
> > implementation which uses the system's numeric formatting.
> 
> I just pulled down HEAD and get a coredump on HP-UX 11.x and funky
> bahaviour on AIX. The following is from HP-UX 11i with the HP C
> compiler:
>   gmake[3]: Leaving directory `/opt/build/glib-2.4.0/gobject'
>   echo "#ifndef __G_MARSHAL_H__" > xgen-gmh \
>   && echo "#define __G_MARSHAL_H__" >> xgen-gmh \
>   && ./glib-genmarshal --nostdinc --prefix=g_cclosure_marshal
>   ./gmarshal.list --header >> xgen-gmh \
>   && echo "#endif /* __G_MARSHAL_H__ */" >> xgen-gmh \
>   && (cmp -s xgen-gmh gmarshal.h 2>/dev/null || cp xgen-gmh gmarshal.h) \
>   && rm -f xgen-gmh xgen-gmh~ \
>   && echo timestamp > stamp-gmarshal.h
>   ./glib-genmarshal --nostdinc --prefix=g_cclosure_marshal
>   ./gmarshal.list --body >> xgen-gmc \
>   && cp xgen-gmc gmarshal.c \
>   && rm -f xgen-gmc xgen-gmc~
>   /bin/sh: 29114 Memory fault(coredump)
> 
>   $ gdb .libs/glib-genmarshal core
>   gdb> bt
>   #0  0xc016eaa4 in memset+0x5c () from /usr/lib/libc.2
>   #1  0x4224 in indent (n_spaces=4294967295) at glib-genmarshal.c:313
>   #2  0x45f4 in generate_marshal (signame=0x40008340 "VOID__VOID", 
>       sig=0x7f7f065c) at glib-genmarshal.c:374
>   #3  0x4f50 in process_signature (sig=0x7f7f065c) at glib-genmarshal.c:507
>   #4  0x58a4 in main (argc=2, argv=0x7f7f0464) at glib-genmarshal.c:696
> 
> Should the above bug report be reopened or should I create a new one?
> 
> In gobject/glib-genmarshal.c:
>       /* cfile marhsal header */ 
>       g_fprintf (fout, "void\n");
>       ind = g_fprintf (fout, "%s_%s (", marshaller_prefix, signame);
> g_fprintf is returning -1.

Neither of these have anything to do with the strtod failure. So, no,
don't reopen the bug. Check in bugzilla for dups; if you don't find one,
then you should file a new bug. For any sort of portability problem,
exact diagnosis and/or patches is *much* appreciated, since most bugs
that we can't reproduce will just sit there for years.

Regards,
						Owen