Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Thu, 8 Nov 2001 17:34:21 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Thu, 8 Nov 2001 17:34:11 -0500 Received: from quattro-eth.sventech.com ([205.252.89.20]:30480 "EHLO quattro.sventech.com") by vger.kernel.org with ESMTP id ; Thu, 8 Nov 2001 17:34:04 -0500 Date: Thu, 8 Nov 2001 17:34:04 -0500 From: Johannes Erdfelt To: Tim Pepper Cc: linux-kernel@vger.kernel.org Subject: Re: Big USB speed difference when compiled as module (was Re: speed difference between using hard-linked and modular drives?) Message-ID: <20011108173404.L32355@sventech.com> In-Reply-To: <20011108142412.A19451@vato.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <20011108142412.A19451@vato.org>; from tpepper@vato.org on Thu, Nov 08, 2001 at 02:24:12PM -0800 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Nov 08, 2001, Tim Pepper wrote: > On Thu 08 Nov at 17:01:24 +0100 roy@karlsbakk.net done said: > > > > Are there any speed difference between hard-linked device drivers and > > their modular counterparts? > > I've been wondering this as well. I've got a curious situation: With usbcore > and the uhci driver compiled into the kernel I can initiate transfers to a usb > device I have (Creative Nomad Jukebox) at what I'm eyeballing to be near the > 12Mbps USB speed. These tranfers always fail. With both drivers compiled as > modules the transfers are over an order of magnitude slower, not even > appearing to reach the 1.5Mbps USB speed, but they succeed. I haven't had a > chance to try to figure out what's going on for sure. > > It's been a little while since I was playing with this but I believe I saw > the same thing with the usb-uhci driver as well but it didn't seem to like > talking to the Nomad regardless of it being compiled as a module or not. > This was with the 2.4.13 kernel. I can provide more details upon request. This absolutely should not be the result of compiling into the kernel. It's most likely a bug in the HC which just happened to show up with modules. > BTW: is there a way to do USB sniffing in software in linux? I'd imagine it's > possible, but just can't find anything that does... Not in Linux yet. JE - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/