From owner-freebsd-hackers Wed Jan 2 16:38:18 2002 Delivered-To: freebsd-hackers@freebsd.org Received: from server.soekris.com (soekris.com [216.15.61.44]) by hub.freebsd.org (Postfix) with ESMTP id ACDA637B427 for ; Wed, 2 Jan 2002 16:38:04 -0800 (PST) Received: from soekris.com (1.4.soekris.com [192.168.1.4] (may be forged)) by server.soekris.com (8.9.2/8.9.2) with ESMTP id QAA21666 for ; Wed, 2 Jan 2002 16:38:11 -0800 (PST) (envelope-from soren@soekris.com) Message-ID: <3C33A842.2C1F54E7@soekris.com> Date: Wed, 02 Jan 2002 16:39:30 -0800 From: Soren Kristensen Organization: Soekris Engineering X-Mailer: Mozilla 4.77 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 To: freebsd-hackers@FreeBSD.ORG Subject: Interrupt Latency / FreeBSD "Realtime" ? Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit 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 Hi Everybody, I'm looking for information about what you can expect in regarding to hardware interrupt latency in FreeBSD.... Assuming that you run without swap, and turn off whatever services that would otherwise complicate things, do anybody have any idea how long you can expect the worst case hardware interrupt latency to be on a FreeBSD 4.x ? Any research available ? Anything you should do or avoid to optimize latency ? I'm asking as I'm looking to implement VoIP hardware on my 133 Mhz 486 net4501 boards, and the voice chips that I'm looking at has limited buffering capabilities. Lost and/or delayed interrupt servicing would degrade the voice performance. That mean that it will not crash if a interrupt is lost, but on the other hand you don't want to miss too many.... My target is in the range of 2 to 10 mS. Regards, Soren To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message