From owner-freebsd-ports@FreeBSD.ORG Tue Sep 16 23:44:09 2008 Return-Path: Delivered-To: ports@FreeBSD.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id F297E106566B for ; Tue, 16 Sep 2008 23:44:09 +0000 (UTC) (envelope-from linimon@lonesome.com) Received: from mail.soaustin.net (lefty.soaustin.net [66.135.55.46]) by mx1.freebsd.org (Postfix) with ESMTP id D44378FC14 for ; Tue, 16 Sep 2008 23:44:09 +0000 (UTC) (envelope-from linimon@lonesome.com) Received: by mail.soaustin.net (Postfix, from userid 502) id 7F60B8C096; Tue, 16 Sep 2008 18:44:09 -0500 (CDT) Date: Tue, 16 Sep 2008 18:44:09 -0500 To: Daniel Austin MBCS Message-ID: <20080916234409.GL30695@soaustin.net> References: <485EF201.9030906@dan.me.uk> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <485EF201.9030906@dan.me.uk> User-Agent: Mutt/1.5.13 (2006-08-11) From: linimon@lonesome.com (Mark Linimon) Cc: ports@FreeBSD.org Subject: Re: pointyhat sparc64 X-BeenThere: freebsd-ports@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Porting software to FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 16 Sep 2008 23:44:10 -0000 On Mon, Jun 23, 2008 at 01:44:49AM +0100, Daniel Austin MBCS wrote: > Is there anything I can do to assist getting more ports/packages tested > (or the pointyhat cluster running faster) ? I'm not sure I ever answered this email (I am going through the backlog today). I hope you did not get discouraged by the lack of response. We are always on the lookout for some more UltraSparc-II machines to add to the build cluster. The fastest things we have are e4500s right now. (There are 3 of those, 5 Netra 1s, and some other misc. stuff). > I had a quick look at your crash logs for the netras... It seems they > have some issues with the ATA controller (atapci0 interrupt storm) - you > only appear to use it for the cdrom drive - perhaps simply disabling via > the boot loader would prevent the kernel panic (and get the machines > into the build cluster)? No, the machines are set up to pxeboot. acd0: CDRW at ata2-master UDMA33 IOMMU fault virtual address 0xc3008000 panic: pcib: uncorrectable DMA error AFAR 0x1fc5c020 AFSR 0x270000ff80000000 cpuid = 0 KDB: enter: panic [thread pid 11 tid 100003 ] Stopped at kdb_enter+0x80: ta %xcc, 1 db However, you are right to point out ata. The T1 200s use a different chipset than the T1 105s. (There are also other differences between them, so it may not necessarily be the cause.) However, you are right that this is a fairly low-cost thing to look at. I'll try it and let you know. (I've been short-cycled on time to look at this for a while, as you might be able to tell from the fact it took me this long to get around to looking at your email). Again, don't get discouraged, there is progress on the sparc64 front, albeit slow. I am running the 7.1R package builds right now. mcl