Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Thu, 21 Feb 2002 02:05:11 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Thu, 21 Feb 2002 02:05:01 -0500 Received: from nat-pool-rdu.redhat.com ([66.187.233.200]:30010 "EHLO lacrosse.corp.redhat.com") by vger.kernel.org with ESMTP id ; Thu, 21 Feb 2002 02:04:54 -0500 Date: Thu, 21 Feb 2002 02:04:49 -0500 From: Benjamin LaHaise To: Daniel Phillips Cc: Andries.Brouwer@cwi.nl, torvalds@transmeta.com, linux-kernel@vger.kernel.org Subject: Re: [PATCH] size-in-bytes Message-ID: <20020221020449.B8733@redhat.com> In-Reply-To:

Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5.1i In-Reply-To: ; from phillips@bonn-fries.net on Mon, Feb 18, 2002 at 11:43:38PM +0100 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Feb 18, 2002 at 11:43:38PM +0100, Daniel Phillips wrote: > We want to stay with the shift counts. They should be the primary currency > of size measurement. You can add shift counts together and get nice, compact > code, whereas with absolute size you often have to ugly things - e.g., it's a > pain to divide by blocksize when you have it as an absolute number, it's easy > when you have it as a shift. > > If you are going to the trouble of fixing this, please don't use absolute > size as the primary measure, use a shift count. Most of this is targetted at userland which needs byte counts (size in sectors was a bug introduced after the original BLKGETSIZE64 went in). Using the number of sectors in kernel is perhaps more efficient, but it is a microoptimization that won't show up on any benchmarks. -ben - 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/