From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from rhirst.linuxcare.com (pc1-hems4-0-cust171.bre.cable.ntl.com [213.105.88.171]) by dsl2.external.hp.com (Postfix) with ESMTP id D3CD6482A for ; Mon, 28 May 2001 07:53:50 -0600 (MDT) Received: by rhirst.linuxcare.com (Postfix, from userid 501) id AD8BFB008; Mon, 28 May 2001 14:51:09 +0100 (BST) Date: Mon, 28 May 2001 14:51:09 +0100 From: Richard Hirst To: parisc-linux@lists.parisc-linux.org Cc: amodra@one.net.au Subject: Re: [parisc-linux] elflink.h (elf_bfd_final_link) Message-ID: <20010528145109.A25745@linuxcare.com> References: <20010528184653.B11349@squeak.one.net.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <20010528184653.B11349@squeak.one.net.au>; from amodra@one.net.au on Mon, May 28, 2001 at 06:46:53PM +0930 List-ID: On Mon, May 28, 2001 at 06:46:53PM +0930, amodra@one.net.au wrote: > Not calling elf_link_output_extsym has the effect that the back-end > finish_dynamic_symbol routine isn't called, which means various hacks > are needed like the following: I built a new binutils from pehc:~amodra/binutils-current, which has this patch, and it produces a working find and make. So, we need a new binutils, and a check for more broken binaries. This check seems to do it, someone can roll it in to plt-checker. for i in /bin/* /sbin/* /usr/bin/* /usr/sbin/*; do if readelf -r $i 2> /dev/null | grep -q R_PARISC_NONE; then echo ===== $i =====; fi; done Things I know have a problem are find, make, modutils. libmagick++5 has also been reported. However, I don't have much installed, so someone need to run the check over the archive, I guess. I still have an issue with the binaries produced from a modutils build though (e.g depmod), which still have a R_PARISC_NONE in them. I've tar-ed up my build tree, as pehc:~rhirst/modutilsbug.tgz The libc, etc are as per the ones in findbug.tgz. The binaries do seem to work, but Alan says R_PARISC_NONE is indicative of a linker bug. Many thanks to Alan for his help! Richard