From owner-freebsd-alpha Wed Nov 15 10: 7:55 2000 Delivered-To: freebsd-alpha@freebsd.org Received: from mail.du.gtn.com (mail.du.gtn.com [194.77.9.57]) by hub.freebsd.org (Postfix) with ESMTP id 7BE2037B4CF; Wed, 15 Nov 2000 10:07:52 -0800 (PST) Received: from mail.cicely.de (cicely.de [194.231.9.142]) by mail.du.gtn.com (8.11.0.Beta3/8.11.0.Beta3) with ESMTP id eAFI7ne00210 (using TLSv1/SSLv3 with cipher EDH-RSA-DES-CBC3-SHA (168 bits) verified OK); Wed, 15 Nov 2000 19:07:51 +0100 (MET) Received: from cicely8.cicely.de (cicely8.cicely.de [10.1.2.10]) by mail.cicely.de (8.11.0.Beta1/8.11.0.Beta1) with ESMTP id eAFI7c507206; Wed, 15 Nov 2000 19:07:41 +0100 (CET) Received: (from ticso@localhost) by cicely8.cicely.de (8.11.0/8.9.2) id eAFI7bw10370; Wed, 15 Nov 2000 19:07:37 +0100 (CET) (envelope-from ticso) Date: Wed, 15 Nov 2000 19:07:36 +0100 From: Bernd Walter To: John Baldwin Cc: alpha@FreeBSD.ORG Subject: Re: FW: cvs commit: src/sys/alpha/alpha swtch.s Message-ID: <20001115190736.A10345@cicely8.cicely.de> References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 1.0.1i In-Reply-To: ; from jhb@FreeBSD.ORG on Tue, Nov 14, 2000 at 06:03:55PM -0800 Sender: owner-freebsd-alpha@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On Tue, Nov 14, 2000 at 06:03:55PM -0800, John Baldwin wrote: > A similar fix to the i386 fixed a number of problems that phk was seeing with > the timecounter as well as performance problems. There is a chance that this > will fix some machines that hang during the SCSI/ATA device probes. Also, I am > currently working on getting kernels booting on all archs with WITNESS turned > on (yes, I haven't gone into hiding). On the alpha the kernel panics because > it tries to acquire the malloc mutex when interrupts are disabled during > vfsinit(). No spin mutex is held, so that isn't the problem. On the i386, > interrupts are enabled at this point, and they should be on the alpha > (interrupts are enabled during SI_SUB_CONFIGURE and should stay full on except > for within spin mutexes from then on), so it looks like the alpha is disabling > interrupts somewhere during the bootup. This may also point to the hangs some > people are seeing (on hte PC164's for example). Sorry - It doesn't change anything on the hanging problem with my PC164. What I see with KTR is not that the hardware ints are missing but the threads stop to run. With KTR it stops always working when the first level irq came in which needs to disable the source. It seems to come further without KTR which may be a timing thing because the console output is seriel and limited to 9600. The masking of the irq itself works. And I'm still getting clock ints and the swi clock thread is still scheduled. -- B.Walter COSMO-Project http://www.cosmo-project.de ticso@cicely.de Usergroup info@cosmo-project.de To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-alpha" in the body of the message