From owner-cvs-all Sat Sep 21 21:53:10 2002 Delivered-To: cvs-all@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 2E02437B401; Sat, 21 Sep 2002 21:53:09 -0700 (PDT) Received: from out5.mx.nwbl.wi.voyager.net (out5.mx.nwbl.wi.voyager.net [169.207.3.123]) by mx1.FreeBSD.org (Postfix) with ESMTP id CC91A43E6A; Sat, 21 Sep 2002 21:53:08 -0700 (PDT) (envelope-from silby@silby.com) Received: from pop5.nwbl.wi.voyager.net (pop5.nwbl.wi.voyager.net [169.207.3.83]) by out5.mx.nwbl.wi.voyager.net (Postfix) with ESMTP id EA50EC6E49; Sat, 21 Sep 2002 23:53:07 -0500 (CDT) Received: from [10.1.1.6] (d81.as14.nwbl0.wi.voyager.net [169.207.134.81]) by pop5.nwbl.wi.voyager.net (8.10.2/8.10.2) with ESMTP id g8M4r6u65208; Sat, 21 Sep 2002 23:53:06 -0500 (CDT) Date: Sat, 21 Sep 2002 23:56:58 -0500 (CDT) From: Mike Silbersack To: Mike Silbersack Cc: cvs-committers@FreeBSD.org, Subject: Re: cvs commit: src/sys/pci if_xl.c In-Reply-To: <200209220448.g8M4mUJZ002542@freefall.freebsd.org> Message-ID: <20020921235350.Q603-100000@patrocles.silby.com> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-cvs-all@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG Unfortunately, this amount of time still trips PHK's timeout check under DIAGNOSTIC. With the other changes Harti has proposed, we may be able to get it down further. However, he's currently on vacation, so that'll have to wait a few weeks. OTOH, at least this ensure that scheduling works normally on systems with HZ=100. I'm rather curious how badly ~10ms timeouts affected general operation if you had multiple xl nics in a box. Mike "Silby" Silbersack On Sat, 21 Sep 2002, Mike Silbersack wrote: > silby 2002/09/21 21:48:30 PDT > > Modified files: > sys/pci if_xl.c > Log: > Remove all DELAY(1) calls around MII operations in the XL driver. > According to the MII specification, the delay produced by our > reads alone are sufficient for correct operation. > > This reduces the time mii_tick takes from 10ms to ~1ms here. That's > still a lot, but much better than before. > > Submitted by: Harti Brandt > MFC after: 3 weeks > > Revision Changes Path > 1.107 +0 -17 src/sys/pci/if_xl.c > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message