From owner-freebsd-current Tue Jan 16 21:56:52 1996 Return-Path: owner-current Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id VAA18490 for current-outgoing; Tue, 16 Jan 1996 21:56:52 -0800 (PST) Received: from localhost.cdrom.com (localhost.cdrom.com [127.0.0.1]) by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id VAA18485 Tue, 16 Jan 1996 21:56:50 -0800 (PST) Message-Id: <199601170556.VAA18485@freefall.freebsd.org> X-Authentication-Warning: freefall.freebsd.org: Host localhost.cdrom.com [127.0.0.1] didn't use HELO protocol To: Chien-Ta Lee cc: freebsd-current@FreeBSD.ORG Subject: Re: FreeBSD current is STABLE !! :) In-reply-to: Your message of "Mon, 15 Jan 1996 17:55:53 +0800." <199601150955.RAA29229@linux.csie.nctu.edu.tw> Date: Tue, 16 Jan 1996 21:56:50 -0800 From: "Justin T. Gibbs" Sender: owner-current@FreeBSD.ORG Precedence: bulk > > Hi : > > I am running FreeBSD-current (make world at 1996/01/10), and it IS > very stable with X running more than 3 days. > > Thanks for Justin's new 2940 driver and other's work. > > (To Justin: I defined QUEUE_FULL_SUPPORTED, and it seems to be > ok for 3 days, any change for me to test ?! like change the > QUEUE_FULL_SUPPORTED from 2 to more ?) You can bump it up (I wouldn't go above 8 since there are only 16 SCBs on a 2940), but if you get a QUEUE_FULL message from a target, the behavior is undefined (ie I haven't implemented it yet since I want to do it at the generic SCSI layer). Most modern drives have queue depths in excess of 16 transactions (64 being the norm). -- Justin T. Gibbs =========================================== FreeBSD: Turning PCs into workstations ===========================================