This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the EGCS project.
Re: What causes the 'type and size of dynamic symbol ... not defined' warning?
- To: kai@cats.ms, Richard Frith-Macdonald <richard@brainstorm.co.uk>
- Subject: Re: What causes the 'type and size of dynamic symbol ... not defined' warning?
- From: "Kai Henningsen" <kai@cats.ms>
- Date: Fri, 23 Jul 1999 15:53:44 +0200
- CC: Ronald.Pijnacker@best.ms.philips.com, GNUstep discussion <discuss-gnustep@gnu.org>, gcc-bugs@gcc.gnu.org
- Organization: Spuentrup EDV & CTI
- References: <E117cW7-0000z5-00@charlotte.intern.cats.ms>
On 23 Jul 99, at 11:53, Richard Frith-Macdonald wrote:
> On Fri, 23 Jul 1999 12:25:32 +0200, kai@cats.ms wrote:
> >
> > I'm not so sure that either gcc or ld is at fault. It seems to me the
> > list of clases it complains about, are those GNUstep classes my
> > program uses; it NEVER complains about my own classes.
> >
> > That suggests to me that the fault is in the GNUstep build process.
> >
> > But I don't know what it is, either.
>
> That's an interesting observation - I think it may be true for my code too.
>
> If the problem is only with classes in the GNUstep libraries, that would
> almost certainly make it some problem with the way dynamic libraries are being
> created (or with how ld handles dynamic libraries).
I searched for the NSThread one in nm output.
This might be enlightening:
./gstepsrc/gstep-core-cats.19990723/base/Source/static_obj/ix86/linux-gnu/gnu-gnu-gnu-xgps/libgnustep-base_s.a
U __objc_class_name_NSThread
U __objc_class_name_NSThread
0000009c R __objc_class_name_NSThread
./gstepsrc/gstep-core-cats.19990723/base/Source/shared_obj/ix86/linux-gnu/gnu-gnu-gnu-xgps/libgnustep-base.so.0.5.50017b740
R __objc_class_name_NSThread
./gstepsrc/gstep-core-cats.19990723/debian/gstep-base/usr/lib/GNUstep/Libraries/ix86/linux-gnu/gnu-gnu-gnu-xgps/libgnustep-base.so.0.5.50017b740
R __objc_class_name_NSThread
./gstepsrc/gstep-core-cats.19990723/debian/gstep-base-dev/usr/lib/GNUstep/Libraries/ix86/linux-gnu/gnu-gnu-gnu-xgps/libgnustep-base_s.a
U __objc_class_name_NSThread
U __objc_class_name_NSThread
0000009c R __objc_class_name_NSThread
I think you're right - this points to a ld bug, unless someone is
actually telling ld to drop that length info.
Regards - Kai Henningsen
--
http://www.cats.ms
Spuentrup CTI Fon: +49 251 322311 0
Windbreede 12 Fax: +49 251 322311 99
D-48157 Münster Mob: +49 161 3223111
Germany GSM: +49 171 7755060