From owner-freebsd-alpha Mon Sep 25 1:49:49 2000 Delivered-To: freebsd-alpha@freebsd.org Received: from anchor-post-31.mail.demon.net (anchor-post-31.mail.demon.net [194.217.242.89]) by hub.freebsd.org (Postfix) with ESMTP id F322237B422; Mon, 25 Sep 2000 01:49:44 -0700 (PDT) Received: from nlsys.demon.co.uk ([158.152.125.33] helo=herring.nlsystems.com) by anchor-post-31.mail.demon.net with esmtp (Exim 2.12 #1) id 13dTxD-000PvI-0V; Mon, 25 Sep 2000 09:49:40 +0100 Received: from salmon.nlsystems.com (salmon.nlsystems.com [10.0.0.3]) by herring.nlsystems.com (8.9.3/8.8.8) with ESMTP id JAA69908; Mon, 25 Sep 2000 09:53:55 +0100 (BST) (envelope-from dfr@nlsystems.com) Date: Mon, 25 Sep 2000 09:49:51 +0100 (BST) From: Doug Rabson To: John Baldwin Cc: alpha@freebsd.org Subject: RE: Ithreads kernel weirdness In-Reply-To: <200009242106.OAA53607@john.baldwin.cx> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Sun, 24 Sep 2000, John Baldwin wrote: > > On 24-Sep-00 Doug Rabson wrote: > > On Fri, 22 Sep 2000, John Baldwin wrote: > > > >> > >> On 20-Sep-00 John Baldwin wrote: > >> > I suppose many of you are wondering why there haven't been any > >> > recent commits of ithreads for the alpha as promised. The reason > >> > is that I can't get a plain kernel to boot. My test kernel with > >> > >> ... > >> > >> Ok, status update: I know have an almost fully MI softinterrupt > >> thread which works fine on UP and SMP i386. It also seems to work > >> fine with a normal kernel with interrupt threads on the alpha. I > >> have one last bug to track down (we are grabbing shed_lock with > >> interrupts disabled at some point late in the sysinit right before > >> init forks, probably the first time we grab the sched_lock). > >> Currently my running kernel is hacking around it by always assuming > >> the saved ipl in a spin mutex is ALPHA_PSL_IPL_0. :-P > >> > >> Once I track this last bug down I'll update the alpha.ithreads.patch > >> so you all can test it and help fine other bugs. We also are still > >> in need of some low-level PCI interrupt enable/disable code for two > >> of the PCI chipsets. > > > > Does this fix the hang you were seeing before? I'm hoping its the same > > problem which was hanging the the rawhide when I was trying to get the > > ithreads changes to dxlpx to work. > > Yes. Actually, it looks like the kproc's (syncer, for example) are > starting off with a saved PSR that has an IPL != 0 in them. I'm going > to be looking at this some more today. Hmm. Possibly cpu_fork needs to fiddle with the savepsr's of the child? -- Doug Rabson Mail: dfr@nlsystems.com Phone: +44 20 8348 6160 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message