RE: kern/43958: Can't compile Atari kernel with ksyms pseudo-device enabled

[David Ross <dross%pobox.com@localhost>]

David,
 
Ok, FWIW I'm unblocked because I can just build with DDB.  But yeah, looks like 
there is a bug to fix.  =)
 
Thank you!
 
David Ross
dross%pobox.com@localhost


----------------------------------------
> From: dholland-bugs%netbsd.org@localhost
> To: kern-bug-people%netbsd.org@localhost; gnats-admin%netbsd.org@localhost; 
> netbsd-bugs%netbsd.org@localhost; dross%pobox.com@localhost
> Subject: Re: kern/43958: Can't compile Atari kernel with ksyms pseudo-device 
> enabled
> CC:
> Date: Mon, 8 Nov 2010 03:45:02 +0000
>
> The following reply was made to PR kern/43958; it has been noted by GNATS.
>
> From: David Holland 
> To: gnats-bugs%NetBSD.org@localhost
> Cc:
> Subject: Re: kern/43958: Can't compile Atari kernel with ksyms
> pseudo-device enabled
> Date: Mon, 8 Nov 2010 03:41:58 +0000
>
> On Sun, Nov 07, 2010 at 07:55:03AM +0000, David Ross wrote:
> > I'm not sure if it's completely nonsensical to build the kernel with ksyms
> > but without DDB.
>
> I don't think so.
>
> > If that's the case then this bug should be closed. Otherwise
> > maybe it makes sense to remove "defined(DDB) &&" from the conditional
> > regulating the inclusion of ?
>
> That or what I suggested before...
>
> I notice however that the code that needs exec_elf.h looks kind of
> bogus in its own right:
>
> #ifndef __ELF__
> ksyms_addsyms_elf(*(int *)&end, ((int *)&end) + 1, esym);
> #else
> ksyms_addsyms_elf((int)esym - (int)&end - sizeof(Elf32_Ehdr),
> (void *)&end, esym);
> #endif
>
> Surely it shouldn't be calling ksyms_addsyms_*elf* when *not* __ELF__?
>
> --
> David A. Holland
> dholland%netbsd.org@localhost
>