From owner-freebsd-usb@FreeBSD.ORG Sun May 27 20:55:51 2007 Return-Path: X-Original-To: freebsd-usb@freebsd.org Delivered-To: freebsd-usb@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [69.147.83.52]) by hub.freebsd.org (Postfix) with ESMTP id B9B6E16A421 for ; Sun, 27 May 2007 20:55:51 +0000 (UTC) (envelope-from imp@bsdimp.com) Received: from harmony.bsdimp.com (bsdimp.com [199.45.160.85]) by mx1.freebsd.org (Postfix) with ESMTP id 7984E13C448 for ; Sun, 27 May 2007 20:55:51 +0000 (UTC) (envelope-from imp@bsdimp.com) Received: from localhost (localhost [127.0.0.1]) by harmony.bsdimp.com (8.13.8/8.13.4) with ESMTP id l4RKrseW003715; Sun, 27 May 2007 14:53:54 -0600 (MDT) (envelope-from imp@bsdimp.com) Date: Sun, 27 May 2007 14:54:12 -0600 (MDT) Message-Id: <20070527.145412.-1597331197.imp@bsdimp.com> To: hselasky@c2i.net From: "M. Warner Losh" In-Reply-To: <200705272235.46048.hselasky@c2i.net> References: <200705272235.46048.hselasky@c2i.net> X-Mailer: Mew version 4.2 on Emacs 21.3 / Mule 5.0 (SAKAKI) Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-2.0 (harmony.bsdimp.com [127.0.0.1]); Sun, 27 May 2007 14:53:54 -0600 (MDT) Cc: freebsd-usb@freebsd.org Subject: Re: 31st address line sometimes not used on EHCI/UHCI/OHCI X-BeenThere: freebsd-usb@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: FreeBSD support for USB List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 27 May 2007 20:55:51 -0000 In message: <200705272235.46048.hselasky@c2i.net> Hans Petter Selasky writes: : I've got some reports back that some USB host controllers do not support : transferring memory from a location higher than 2GB. : : What should we do about this? : : Should we limit all USB DMA allocations to the lower 2GB of the memory? busdma should be managing this behind the scenes. You shouldn't care, as the problematical usb controllers, if any, can do the bouncing as required. We need to get the hierarchical bus tagging stuff more fully integrated, then we'd get this for free. Of course, you'd have to stop using contigmalloc to allocate all the memory for usb. That won't work on some of the embedded platforms we have, for example, because memory on them isn't as fungible as it is on i386 and amd64. Warner