From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S263607AbTKJOJ2 (ORCPT ); Mon, 10 Nov 2003 09:09:28 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S263624AbTKJOJ2 (ORCPT ); Mon, 10 Nov 2003 09:09:28 -0500 Received: from ns.virtualhost.dk ([195.184.98.160]:57048 "EHLO virtualhost.dk") by vger.kernel.org with ESMTP id S263607AbTKJOJ1 (ORCPT ); Mon, 10 Nov 2003 09:09:27 -0500 Date: Mon, 10 Nov 2003 15:09:27 +0100 From: Jens Axboe To: Simon Haynes Cc: linux-kernel@vger.kernel.org, linux-mtd@lists.infradead.org Subject: Re: SFFDC and blksize_size Message-ID: <20031110140927.GE32637@suse.de> References: <37CC93E8710D@baydel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <37CC93E8710D@baydel.com> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Nov 07 2003, Simon Haynes wrote: > > I have been writing a block driver for SSFDC compliant SMC cards. This stuff > allocates 16k blocks. When I get requests the transfers are split into the > size I specifty in the blksize_size{MAJOR] array. It sems that most things Sounds like a bad way to do it. It's much better to prevent builds of bigger requests than you can handle in one go. You don't mention what kernel you are using, but both 2.4 and 2.6 can do this for you. -- Jens Axboe