From owner-freebsd-hackers Wed Jan 31 16:45:43 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id QAA29480 for hackers-outgoing; Wed, 31 Jan 1996 16:45:43 -0800 (PST) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by freefall.freebsd.org (8.7.3/8.7.3) with ESMTP id QAA29472 for ; Wed, 31 Jan 1996 16:45:39 -0800 (PST) Received: from msmith@localhost by genesis.atrad.adelaide.edu.au (8.6.12/8.6.9) id LAA20495; Thu, 1 Feb 1996 11:29:39 +1030 From: Michael Smith Message-Id: <199602010059.LAA20495@genesis.atrad.adelaide.edu.au> Subject: Re: Watchdog timers (was: Re: Multi-Port Async Cards) To: phk@critter.tfs.com (Poul-Henning Kamp) Date: Thu, 1 Feb 1996 11:29:38 +1030 (CST) Cc: hackers@FreeBSD.org In-Reply-To: <5377.823120360@critter.tfs.com> from "Poul-Henning Kamp" at Jan 31, 96 09:32:40 pm MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-hackers@FreeBSD.org Precedence: bulk Poul-Henning Kamp stands accused of saying: > > > I actually thought about this some time back, and have a suggestion > for a ISA card: > > The "FreeBSD gadget": > > Watchdog, can be set to timouts of 1s, 10s and 60s or > something. Will launch a RESET, NMI, or IRQ (jumper). This is (relatively) easy. Dennis is lying about his markup (he's not completely stupid 8), but anything more than _really_ stupid gets expensive fast. > 10, 100 or 250 Mhz timer, with readable count, > generates IRQ 20/100/250 (jumper) times a second. > (For microtime) (Maybe even with a 10/11 swallow > counter along the lines the XNTP people suggest.) 250MHz? Ouch 8) Still, not hard, and would coexist quite happily with the former. However, it wouldn't solve the basic problem, as we couldn't ask people to buy one before running FreeBSD 8) > NVram for console. 64K of NVram (preferrably FIFO), > for console-messages, not affected by reset. > > Optional: A microcontroller to make a async console > with the above nvram. (incl modem handling and passwd) > Possibly even with a BIOS-eprom to even make the bios > available that way. Hmm. There it starts to get messy. Certainly not impossible though, but making it work _well_ might be tough. I see two models; one, where the kernel handshaked every byte into the card, and another with a FIFO (512 bytes?) decoupling the two. Giving the CPU direct access to the NV on the card would be a bad idea. Would you want to be able to read the messages back? (Silly question 8) > Optional: NVram for acceleration of NFS &c. sockets > and a battery for 4-8 Mb ram. Possibly with a little > clock to do refresh so DRAM could be used. (cheaper) Yeow! That would have to be a PCI card for performance to be acceptable, and you'd basically end up making a Prestoserv card. Sorry, don't have the resources for that 8( > I'm not going to make it, I'm SW only, but I'd buy a couple right away. I see three cards there; the el-cheapo reset card, the buffered console card and the Presto card. > Any takers ? *sigh* > Poul-Henning Kamp | phk@FreeBSD.ORG FreeBSD Core-team. -- ]] Mike Smith, Software Engineer msmith@atrad.adelaide.edu.au [[ ]] Genesis Software genesis@atrad.adelaide.edu.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control (ph/fax) +61-8-267-3039 [[ ]] "wherever you go, there you are" - Buckaroo Banzai [[