From owner-freebsd-current Tue Sep 19 22:39:23 2000 Delivered-To: freebsd-current@freebsd.org Received: from nagual.pp.ru (pobrecita.freebsd.ru [194.87.13.42]) by hub.freebsd.org (Postfix) with ESMTP id F263E37B422 for ; Tue, 19 Sep 2000 22:39:19 -0700 (PDT) Received: (from ache@localhost) by nagual.pp.ru (8.11.0/8.11.0) id e8K5b5t06699; Wed, 20 Sep 2000 09:37:05 +0400 (MSD) (envelope-from ache) Date: Wed, 20 Sep 2000 09:37:03 +0400 From: "Andrey A. Chernov" To: Poul-Henning Kamp Cc: Bruce Evans , current@FreeBSD.ORG Subject: Re: recent kernel, microuptime went backwards Message-ID: <20000920093702.A6671@nagual.pp.ru> References: <92160.969388225@critter> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <92160.969388225@critter>; from phk@critter.freebsd.dk on Tue, Sep 19, 2000 at 08:30:25PM +0200 Organization: Biomechanoid Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Tue, Sep 19, 2000 at 08:30:25PM +0200, Poul-Henning Kamp wrote: > In message , Bruce Eva > ns writes: > >On Tue, 19 Sep 2000, Andrey A. Chernov wrote: > > > >> With very latest kernel I got lots of > >> > >> microuptime() went backwards (1.3624050 -> 1.998840) > >> > >> messages just before > >> > >> Mounting root from ufs:/dev/da0s1a > > > >It really does go backwards. This is caused by the giant lock preventing > >the clock interrupt task from running soon enough. The giant lock can > >also prevent the clock interrupt task from running often enough even > >after booting. E.g., "dd if=/dev/random of=/dev/null bs=large" does > >several bad things. > > Yes, we are royally hosed right now with respect to interrupt service. > > It's a testimony to the robustness of the majority of our drivers that > they don't explode more often than they do. It seems it was SCSI driver in my case waiting for SCSI device settle (or in some another state, I am not sure). BTW, could you please summarize this diagnostic? I.e. istead of tons of messages effectively overflowing dmesg buffer and maybe cause backwards by time consumed by themselfs put only one message summarizing all of them, something like "went backwards N times". -- Andrey A. Chernov http://ache.pp.ru/ To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message