From owner-freebsd-hackers Sat Feb 17 03:07:14 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id DAA06234 for hackers-outgoing; Sat, 17 Feb 1996 03:07:14 -0800 (PST) Received: from hda.com (hda.com [199.232.40.182]) by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id DAA06226 for ; Sat, 17 Feb 1996 03:07:08 -0800 (PST) Received: (from dufault@localhost) by hda.com (8.6.11/8.6.9) id GAA06130; Sat, 17 Feb 1996 06:15:21 -0500 From: Peter Dufault Message-Id: <199602171115.GAA06130@hda.com> Subject: Re: an invite to industrial users of Freebsd. To: julian@TFS.COM (Julian Elischer) Date: Sat, 17 Feb 1996 06:15:20 -0500 (EST) Cc: hackers@freebsd.org In-Reply-To: from "Julian Elischer" at Feb 16, 96 06:01:00 pm X-Mailer: ELM [version 2.4 PL24] Content-Type: text Sender: owner-hackers@freebsd.org Precedence: bulk > real-time issues. (what-ever your definition of real-time > might be this week) I have a list of people interested in realtime issues. There hasn't been a mailing to it in about a year. I'll hand all those names over to Julian under separate copy; I HUMBLY apologize to anyone who tried to send anything to that list since last October, since I now realize that that list was never installed on the new hda.com mail server that went on line back then! I'd like to have that list rehosted at freebsd.org where it gets advertised and run by competent people (just to make it clear who screws up here at hda.com: I set up and "maintain" this mail server with occasional 4:30 AM work on it). If anyone is willing to set it up Julian now has the names. I'm leaving this AM so I'm not going to see the responses to this thread, but I'm going to (after the SCSI clean up with you, Julian, which has to come first since though it ain't that interesting it affects many more people) start in on the skeleton for the real time extensions. My initial plans are modest - add the recently posted real time interface to rtprio (I cleared up the copyright with the author), add rtprio aware resource allocation to the kernel during the SCSI cleanup, and define and add the infrastructure for adding the POSIX real time extensions (no-brainer infrastructure - required header files, C stubs that return that something doesn't work per POSIX, kernel compile options that disable the POSIX functionality mechanistically derived from the POSIX facility test macros so that it is modular and can be added in pieces, etc). This with a few additions will especially help out the multimedia group who will then have a standard way to do some of the sort of things they want to do - they are really the biggest group interested in real time sorts of stuff. If you are interested in this get the latest POSIX spec which includes the realtime extensions. Then you multimedia folk can drool over the direct and async I/O. If anyone wants to actually contract with me to work on this I'll give you a good rate (or on the SCSI cleanup, for that matter). Please defer discussion of realtime until the new list is up. Please no many page follow ups about kernel threads and MP support when I talk about adding POSIX async I/O and user specified memory range locking (you know who you are) - I'm a simple man. -- Peter Dufault Real-Time Machine Control and Simulation HD Associates, Inc. Voice: 508 433 6936 dufault@hda.com Fax: 508 433 5267