From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Date: Wed, 17 Feb 1999 20:39:10 -0500 From: "Paul J.Y. Lahaie" Sender: pjlahaie@sims-ha.videotron.net To: "hppa-linux@puffingroup.com" Message-id: <36CB6F3E.5860DA3E@ottawa.com> MIME-version: 1.0 Content-type: text/plain; charset=us-ascii Subject: [hppa-linux] binutils / egcs oddity (yet another) Resent-Message-ID: <"HoAQY1.0.Tq6.5zsos"@burrow.puffingroup.com> Reply-To: hppa-linux@puffingroup.com List-Id: linux-parisc.vger.kernel.org I've been trying to get egcs to build a binary that will run on MkLinux, unfortunately, it's (once again) doing odd things. If I take the standard crt0.o shipped with MkLinux and use the MkLinux nm on it, I get the following: 00000000 T $START$ 00000004 D $global$ 00000000 D __environ U _start 00000000 D environ If I run the new x-binutils from the 2.9 series I get the following: 00000000 T $START$ 20000004 D $global$ 20000000 D __environ U _start 20000000 D environ For some reason, it's putting some symbols really far out. If I do a link with the standard MkLinux tools, I get main as my last symbol and it's in the "0000axxx" range (on a program with an empty main file). If I try the same program with egcs/binutils 2.9 I get: 200093c0 D __environ 200093c0 D environ 200093c4 D $global$ 200093c8 D _exit_dummy_ref 200093d0 D _exit_dummy_decl As my last symbols. I've been studying the GNU ld linker scripts but they don't seem to be the culprits. I could run gdb on ld and find out where it's getting that 0x20000000. But this will probably be lots of effort and are we looking for MkLinux binary compatibility? If we aren't then we should just make the Linux/HPPA port loader capable of dealing with the standard egcs HPPA-ELF output. - Paul ---------------------------------------------------------------------- To unsubscribe: send e-mail to hppa-linux-request@puffingroup.com with `unsubscribe' as the subject.