From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753371Ab1JLNtl (ORCPT ); Wed, 12 Oct 2011 09:49:41 -0400 Received: from mail.dev.rtsoft.ru ([213.79.90.226]:54453 "HELO mail.dev.rtsoft.ru" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S1753342Ab1JLNth (ORCPT ); Wed, 12 Oct 2011 09:49:37 -0400 Message-ID: <4E959A9B.7060100@ru.mvista.com> Date: Wed, 12 Oct 2011 17:48:11 +0400 From: Sergei Shtylyov User-Agent: Mozilla/5.0 (X11; Linux i686; rv:6.0) Gecko/20110812 Thunderbird/6.0 MIME-Version: 1.0 To: Markus Rechberger CC: Greg KH , Alan Stern , USB list , LKML Subject: Re: [Patch] Increase USBFS Bulk Transfer size References: In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 10/12/2011 04:36 PM, Markus Rechberger wrote: > We have 2 products which can perform better with increased Bulk transfers > Device No. 1: > According to the hardware spec of on of our product > Available Bulk Transfer Size are: > - 188 * n bytes, where n = 1 ~ 256. > Although we can drive that one with 15K as well when setting the HW > register down to it. > Device No. 2 > only creates jitter video with Bulk transfer sizes which are below > 24064 bytes, no such chipfeature is available > to decrease the bulk transfer size. > http://sundtek.de/images/dtvjitter2.jpg > with transfer size of 24064: > http://sundtek.de/images/gooddata.jpg > The patch takes the features of Device No. 1 into account allowing a > maximum buffer of 48128 bytes. > Those issues have been evaluated with MacOSX and a customized patched > Linux version. > Device No. 2 also corrupts on MacOSX with too small packet sizes, > Windows and Mac are using 24064 bytes. You are constantly mixing the packet size and the transfer size. > Default Bulk Transfersize of device No. 1 is around 1-2k which leads > to very high cpu usage, updating it to 15k lowers that one. WBR, Sergei