From owner-freebsd-current@FreeBSD.ORG Sun Feb 19 15:00:18 2006 Return-Path: X-Original-To: freebsd-current@freebsd.org Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id EF02F16A422 for ; Sun, 19 Feb 2006 15:00:18 +0000 (GMT) (envelope-from iedowse@iedowse.com) Received: from nowhere.iedowse.com (nowhere.iedowse.com [82.195.144.75]) by mx1.FreeBSD.org (Postfix) with SMTP id 1A5A543D55 for ; Sun, 19 Feb 2006 15:00:17 +0000 (GMT) (envelope-from iedowse@iedowse.com) Received: from localhost ([127.0.0.1] helo=iedowse.com) by nowhere.iedowse.com via local-iedowse id ; 19 Feb 2006 15:00:13 +0000 (GMT) To: Fredrik Lindberg In-Reply-To: Your message of "Sun, 19 Feb 2006 14:23:33 +0100." <43F87155.3050103@shapeshifter.se> Date: Sun, 19 Feb 2006 15:00:12 +0000 From: Ian Dowse Message-ID: <200602191500.aa87886@nowhere.iedowse.com> Cc: freebsd-current@freebsd.org Subject: Re: USB and clear endpoint stall X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 19 Feb 2006 15:00:19 -0000 In message <43F87155.3050103@shapeshifter.se>, Fredrik Lindberg writes: >After a quite recent -current update, UPEKs touchchip driver began >to run dead slow and by turning on debugging for libusb (which the >driver utilize) I noticed several 5 seconds long read timeouts. > >These timeouts were not present with the same driver on an earlier >current and I think I've tracked it down to the commits on Jan 8 >which dropped usbd_clear_endpoint_stall() from usbd_setup_pipe(). >Apparently this device needs this, because after re-adding this >functionallity the timeout disappears. > >I've attached a patch that works for me, it's basicly the same >as before Jan 8, but with the quirk reversed. >My USB-foo isn't that good so people with better understanding >of the USB stack might have a better fix for this. Committed, thanks! I wasn't sure if such devices would exist, but that's exactly the right way to handle them. Ian