From owner-freebsd-ppc Mon Sep 30 7:46:35 2002 Delivered-To: freebsd-ppc@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id B2E9637B401; Mon, 30 Sep 2002 07:46:33 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by mx1.FreeBSD.org (Postfix) with ESMTP id 76F9443E4A; Mon, 30 Sep 2002 07:46:29 -0700 (PDT) (envelope-from gallatin@cs.duke.edu) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id KAA08439; Mon, 30 Sep 2002 10:46:28 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.6/8.9.1) id g8UEjwW09331; Mon, 30 Sep 2002 10:45:58 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <15768.25510.803465.802205@grasshopper.cs.duke.edu> Date: Mon, 30 Sep 2002 10:45:58 -0400 (EDT) To: Peter Grehan Cc: freebsd-ppc@FreeBSD.ORG Subject: Re: New system binaries available In-Reply-To: <3D9851D8.97999FD1@freebsd.org> References: <3D9851D8.97999FD1@freebsd.org> X-Mailer: VM 6.75 under 21.1 (patch 12) "Channel Islands" XEmacs Lucid Sender: owner-freebsd-ppc@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG Peter Grehan writes: > I've put a mostly complete system up at > > http://people.freebsd.org/~grehan/root.tgz Woo-hoo! A compiler! > In conjunction with a diskless kernel built out of the main CVS tree: > > http://people.freebsd.org/~grehan/kernel.nfs > > .. it should be possible together something quite usable. Cool.. My first problem is that heavy network transmit loads with messages large enough to use mbuf clusters causes the machine to crash rapidly with some variety of mcl bookkepping problem: Transmit via loopback: # ifconfig lo0 inet 127.0.0.1 up # ./netserver Starting netserver at port 12865 # ./netperf TCP STREAM TEST to localhost Out of mbuf address space! Consider increasing NMBCLUSTERS exclusive sleep mutex Giant r = 0 (0x33b680) locked @ ../../../vm/vm_object.c:446 panic: mb_reclaim() called with locks held Transmit via gem0: # ./netperf -H10.0.0.2 TCP STREAM TEST to 10.0.0.2 panic: m_ext refcnt < 0 However, transmits of non cluster mbufs seem to be fine: # ./netperf -H10.0.0.2 -tUDP_STREAM -- -m 1 UDP UNIDIRECTIONAL SEND TEST to 10.0.0.2 Socket Message Elapsed Messages Size Size Time Okay Errors Throughput bytes bytes secs # # 10^6bits/sec 9216 1 10.01 99419 0 0.08 41600 10.01 99309 0.08 And, recieves seem to be fine on real hardare. To refresh, this is a dual G4 800MHz. Using gem0 (but since it crashes on loopback, I dont' think this is hardware related). What's the status of ddb? Drew PS: I've left the powerpc netperf/netserver binaries in ~gallatin/net{perf,server}.ppc on freefall in case you want to play with this. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-ppc" in the body of the message