Date: Thu, 20 Feb 2003 18:36:49 +1100 From: Tim Robbins <tjr@FreeBSD.org> To: Marcel Moolenaar <marcel@xcllnt.net> Cc: src-committers@FreeBSD.org, cvs-src@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/bin/sh machdep.h memalloc.c nodes.c.pat Message-ID: <20030220183649.A65531@dilbert.robbins.dropbear.id.au> In-Reply-To: <20030218073738.GA3320@athlon.pn.xcllnt.net>; from marcel@xcllnt.net on Mon, Feb 17, 2003 at 11:37:38PM -0800 References: <200302180208.h1I28Kbu046465@repoman.freebsd.org> <20030218022850.GB1973@athlon.pn.xcllnt.net> <20030218143222.A21049@dilbert.robbins.dropbear.id.au> <20030218073738.GA3320@athlon.pn.xcllnt.net>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Feb 17, 2003 at 11:37:38PM -0800, Marcel Moolenaar wrote: > On Tue, Feb 18, 2003 at 02:32:22PM +1100, Tim Robbins wrote: > > > > Try this (ugly) patch: http://people.freebsd.org/~tjr/align.diff > > > > I've hacked up something too. See http://www.xcllnt.net/~marcel/sh.diff > > Some notes: > o For testing purposes I hardcoded 16-byte alignment in machdep.h. This > allows me to test it on any machine I have, not just ia64. In the real > world the header does not exist and <sys/param.h> is included. > o I removed the stackbase completely, because you cannot depend on it > being aligned properly. Hence the initial situation is with stackp > NULL. > o I sprinkled asserts in the code to have sh(1) fail hard at the > first signs of misalignment. > > Give it a spin. I'm currently testing with an i386 buildworld... Seems to work nicely on alpha and sparc64. I haven't had a chance to try a buildworld with it yet, though. Tim To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-src" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030220183649.A65531>