On Sat, May 13, 2006 at 04:14:08PM +0200, Julio M. Merino Vidal wrote:
> On 5/3/06, Pavel Cahyna <pavel@netbsd.org> wrote:
> >I see that Nick committed the change to ld which makes the provided script
> >work. (It is necessary to change PROVIDE (_end) = . ; to
> >PROVIDE (_end = .) ; and so on.)
> >
> >Is it appropriate to commit it? I see that the current ldscript for i386
> >is very different, it has many things which apparently make sense only for
> >userland.
> >
> >And, could please somebody test that a kernel made with this ldscript is
> >bootable with the native bootloader? (I have only grub.)
> 
> I just tried and works fine with both the native loader and GRUB.
> It'd be very nice if this was commited to the tree because it'd make
> multiboot Just Work (TM) -- but note that I am not approving it since
> I do not know anything about linker scripts ;-)

yes, Just Work - that's why I want do do it before NetBSD 4.0. I'll
commit it probably today.

But the "Just Work" means "except the symbol table". Please, would you
have time and knowledge to work on that?

(See http://thread.gmane.org/gmane.os.netbsd.devel.kernel/18148/focus=18843 )

Bye	Pavel