From owner-freebsd-hackers Wed Jul 17 14:45:34 2002 Delivered-To: freebsd-hackers@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 06E0F37B400 for ; Wed, 17 Jul 2002 14:45:28 -0700 (PDT) Received: from smtpout.mac.com (smtpout.mac.com [204.179.120.88]) by mx1.FreeBSD.org (Postfix) with ESMTP id 49F9543E5E for ; Wed, 17 Jul 2002 14:45:27 -0700 (PDT) (envelope-from robm@mac.com) Received: from smtp-relay01.mac.com (smtp-relay01-en1 [10.13.10.224]) by smtpout.mac.com (8.12.1/8.10.2/1.0) with ESMTP id g6HLjQ7d005983 for ; Wed, 17 Jul 2002 14:45:26 -0700 (PDT) Received: from asmtp01.mac.com (asmtp01-qfe3 [10.13.10.65]) by smtp-relay01.mac.com (8.12.1/8.12.1/1.0) with ESMTP id g6HLjLWf026340 for ; Wed, 17 Jul 2002 14:45:21 -0700 (PDT) Received: from h209-139-252-250.gtconnect.net ([66.119.176.126]) by asmtp01.mac.com (Netscape Messaging Server 4.15) with ESMTP id GZEXRK00.KBV for ; Wed, 17 Jul 2002 14:45:20 -0700 Date: Wed, 17 Jul 2002 14:45:18 -0700 Mime-Version: 1.0 (Apple Message framework v482) Content-Type: text/plain; charset=US-ASCII; format=flowed Subject: problem getting multiple bktr devices talking through a single HB1 bridge From: Rob McKeever To: freebsd-hackers@freebsd.org Content-Transfer-Encoding: 7bit Message-Id: <7CCB9A8B-99CE-11D6-8D79-0050E4465D5F@mac.com> X-Mailer: Apple Mail (2.482) Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG Hello everyone, I've got a problem I'm hoping that someone out there can help solve (surprise, surprise!). I've got a bunch of capture cards that use 4 Conexant BT878A capture chips connected through a single HiNT HB1 bridge chip. When I've tried connecting multiple BT878's directly to the main PCI bus, the system runs just fine with all slots filled. I've tried the same thing using the multi-chip capture cards but have had much less success - the system will run for a period of time, then spontaneously reboot without warning. It usually does not leave a kernel dump. So far, Ive got: 1) Multiple independent BT878A's work fine (devices bktr0 through bktr7 usually, although I've also tweaked things to test with up to 20 bktr devices) 2) Running only one BT878A per HiNT HB1 bridge works just fine 3) Running multiple BT878A's per HB1 causes reboots The test software I've been using to try and figure this out is fxtv. I'll be using some of my own streaming code in the end, but this provides a nice open source verification of the problem (thus assuring me that it's not something I did wrong). The more simultaneous session started ('cause the bktr driver doesn't automatically disable capture when a connected client process closes), the quicker the system crashes. I've talked to Roger Hardiman, who maintains the bktr driver, who suggested that perhaps some special setup info had to be given to bridge chip (like what is done to support the newer Matrox Meteor cards). The only affect I've been able to have by doing this has been to set the primary and secondary latency higher - this results in somewhat longer times between reboots. I've talked to HiNT who thinks this is likely an issue with the BT878A driver on FreeBSD. I've been trying to get in touch with Conexant, but the closest I've been able to get is phone tag with a local field engineer. Can anyone provide any other suggestions? I can provide specs sheets for all chips involved, kernel configuration info, run tests, etc. if it will help. I's also be willing to compensate someone some for their efforts should we manage to find an answer. Thanks, in advance, Rob McKeever Vancouver, BC Canada robm@mac.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message