From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S262205AbUEQTJj (ORCPT ); Mon, 17 May 2004 15:09:39 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262238AbUEQTJj (ORCPT ); Mon, 17 May 2004 15:09:39 -0400 Received: from mx1.redhat.com ([66.187.233.31]:13290 "EHLO mx1.redhat.com") by vger.kernel.org with ESMTP id S262205AbUEQTJg (ORCPT ); Mon, 17 May 2004 15:09:36 -0400 Date: Mon, 17 May 2004 12:09:25 -0700 From: Pete Zaitcev To: Sebastien Chaumat Cc: linux-usb-devel@lists.sourceforge.net, zaitcev@redhat.com, linux-kernel@vger.kernel.org Subject: Re: please apply the old patch from Georg Acher in 2.4 tree Message-Id: <20040517120925.44ae3af1.zaitcev@redhat.com> In-Reply-To: <1083011181.982.6.camel@muetdhiver> References: <1083011181.982.6.camel@muetdhiver> Organization: Red Hat, Inc. X-Mailer: Sylpheed version 0.9.9 (GTK+ 1.2.10; i686-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Mon, 26 Apr 2004 22:26:21 +0200 Sebastien Chaumat wrote: > I tried the patch from Georg Acher (posted 2001/08). It saved me from a > lot of troubles with a brand news USB2 hub that refuses to take an > address with my nforce2 motherboard. > > @@ -2177,18 +2178,31 @@ > dev->epmaxpacketin [0] = 8; > dev->epmaxpacketout[0] = 8; > > - err = usb_set_address(dev); > - if (err < 0) { > - err("USB device not accepting new address=%d (error=%d)", > - dev->devnum, err); > - clear_bit(dev->devnum, &dev->bus->devmap.devicemap); > - dev->devnum = -1; > - return 1; > - } > + for(m=0;m<2;m++) { > + for(n=0;n<2;n++) { > + err = usb_set_address(dev); > + if (err>=0) > + break; > + wait_ms(200); > + } I see this was admitted to 2.6, with nice symbolic constants, too. But before taking this, I would like someone to do this: 1. posess a device which flakes like the above 2. take the 2.6 version rediff and test it 3. explain to me why he or she cannot just run 2.6 and be content If Sebastien is willing to clear these hurdles I erected (and mentions the vendor and device number of his hub), I'll think about it. But it would be best if someone else came forward. -- Pete