From owner-freebsd-stable Tue Jul 31 12:17:21 2001 Delivered-To: freebsd-stable@freebsd.org Received: from zibbi.icomtek.csir.co.za (zibbi.icomtek.csir.co.za [146.64.24.58]) by hub.freebsd.org (Postfix) with ESMTP id 1C46937B401 for ; Tue, 31 Jul 2001 12:17:16 -0700 (PDT) (envelope-from jhay@zibbi.icomtek.csir.co.za) Received: (from jhay@localhost) by zibbi.icomtek.csir.co.za (8.11.1/8.11.1) id f6VJH1U90022; Tue, 31 Jul 2001 21:17:01 +0200 (SAT) (envelope-from jhay) From: John Hay Message-Id: <200107311917.f6VJH1U90022@zibbi.icomtek.csir.co.za> Subject: Re: ata timeouts after debug commits In-Reply-To: <20010730173949.A69298@neutrino.bsdhome.com> from Brian Dean at "Jul 30, 2001 05:39:49 pm" To: bsd@bsdhome.com (Brian Dean) Date: Tue, 31 Jul 2001 21:17:01 +0200 (SAT) Cc: freebsd-stable@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL54 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG > > > > At the end I tracked it down to these commits: > > > > > > > > src/sys/i386/i386/db_trace.c,v 1.35.2.1 2001/07/12 02:57:11 bsd > > > > src/sys/i386/i386/support.s,v 1.67.2.4 2001/07/12 02:57:11 bsd > > > > src/sys/i386/include/cpufunc.h,v 1.96.2.2 2001/07/12 02:57:11 bsd > > > > > > > > Maybe it isn't the code itself but just that things moved? I tried the > > snaps on internat and the ones before July 13 worked, but Jul 13 and > > later didn't work. Even a fairly stripped down kernel with later sources > > didn't work until I backed out those 3 files. > > I don't know, I suppose that's possible. But if you didn't have > options DDB in your kernel, the db_trace.c file is not included in the > kernel you built. And the changes to support.s and cpufunc.h merely > converted a real function to an in-line function, and added a few > previously non-existant in-line functions (which are only used by ddb > at the moment), but otherwise there was no change. Even the function > that was converted from real to in-line would only called in certain > circumstances during a debugger trace trap, so I'm having a hard time > pinning the problem on the above commits. > > If you get access to the machine again, see if you can hook the > machine up to use a serial console or something and capture the bootup > messages, a verbose boot would be good. Post that out and somebody > may spot something. I'll do that. > > Also, you mentioned the date of the 13th, how about a more recent > -stable snapshot? What was the latest snapshot you tried? I tried up to the 24th, which was the latest available at that stage. At the end I took the source of the 24th, built a kernel and tried it. It didn't work. I then reversed those 3 files, recompiled and the resulting kernel worked. John -- John Hay -- John.Hay@icomtek.csir.co.za To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message