From owner-freebsd-alpha Thu Feb 21 14:26:56 2002 Delivered-To: freebsd-alpha@freebsd.org Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by hub.freebsd.org (Postfix) with ESMTP id C86CC37B402; Thu, 21 Feb 2002 14:26:52 -0800 (PST) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id RAA07448; Thu, 21 Feb 2002 17:26:52 -0500 (EST) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g1LMQL775218; Thu, 21 Feb 2002 17:26:21 -0500 (EST) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15477.29709.908299.287618@grasshopper.cs.duke.edu> Date: Thu, 21 Feb 2002 17:26:21 -0500 (EST) To: Doug Rabson Cc: peter@freebsd.org, jdp@freebsd.org, Subject: Re: alpha klds and the new linker In-Reply-To: References: <15477.3715.6908.110667@grasshopper.cs.duke.edu> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Doug Rabson writes: > On Thu, 21 Feb 2002, Andrew Gallatin wrote: > > > > > Doug Rabson writes: > > > On Wed, 20 Feb 2002, Andrew Gallatin wrote: > > > > > > > > > > > Doug, > > > > > > > > There's something about the new linker (from the latest binutils > > > > import) which breaks klds on alpha. If I attempt to load a module > > > > built with the new binutils, things go kaboom like this: > > > > > > >From looking at this, I can't say much other than that it looks like the > > > symbol lookup for "modmetadata_set" returned something bogus. Either that > > > or maybe the linker sets didn't get relocated properly. What does > > > 'objdump -p' and 'objdump -R' look like for the two files? > > > > > > Due to their size, I've left the objdump -R output on freefall: > > > > http://people.freebsd.org/~gallatin/objdump-R_new > > http://people.freebsd.org/~gallatin/objdump-R_old > > I guess it might be something to do with all those RELATIVE relocs which > have addends. I can't see anything obviously wrong with the code but it > has never been called with non-zero addends so there could easily be a > problem there. Wasn't there something similar with rtld recently? I think I've fixed it. If I remove the '*where' from the addr calculation, all seems to be fine. I don't really know enough about how elf is supposed to work to want to commit this without a review. Anybody care to say this is "OK" ? Index: elf_machdep.c =================================================================== RCS file: /home/ncvs/src/sys/alpha/alpha/elf_machdep.c,v retrieving revision 1.7 diff -u -r1.7 elf_machdep.c --- elf_machdep.c 28 Aug 1999 00:38:25 -0000 1.7 +++ elf_machdep.c 21 Feb 2002 22:22:14 -0000 @@ -101,7 +101,7 @@ break; case R_ALPHA_RELATIVE: - addr = relocbase + addend + *where; + addr = relocbase + addend; if (*where != addr) *where = addr; break; Thanks, Drew To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message