From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: From: jmk@plan9.bell-labs.com To: 9fans@cse.psu.edu Subject: Re: [9fans] waserror()/nexterror() In-Reply-To: <887916726fef37e3c6950c109038f988@terzarima.net> MIME-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit Date: Fri, 16 Jan 2004 10:37:06 -0500 Topicbox-Message-UUID: bb224be8-eacc-11e9-9e20-41e7f4b1d025 On Fri Jan 16 06:50:39 EST 2004, forsyth@terzarima.net wrote: > >>whereas something very early in the 4ed kernel startup (before any > >>printed announcements) causes the kernel to reset the host. > > i noticed that recently on a 486-class device that did run a few months ago. > i'd wondered if it might have been the changes to devarch.c/l.s but > it wasn't urgent enough for me to debug it yet, nor was > i sure until i saw your remark that it wasn't more likely > some blunder of my own (eg, inconsistent mk) For some time Plan 9 on sources had the 'coherence' variable in pc/devarch.c initialised to to something that wouldn't work on an x86 that didn't have the CPUID instruction, e.g. the 386 and most 486. I attempted to fix it a few weeks ago, if your kernel has the current devarch.c, fns.h and l.s and still doesn't work, let me know. --jim