Return-Path: Received: from one.firstfloor.org ([213.235.205.2]:51090 "EHLO one.firstfloor.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753332Ab0HRSx6 (ORCPT ); Wed, 18 Aug 2010 14:53:58 -0400 Date: Wed, 18 Aug 2010 20:53:54 +0200 From: Andi Kleen To: David Woodhouse Cc: Andi Kleen , "J. Bruce Fields" , "Patrick J. LoPresti" , linux-fsdevel@vger.kernel.org, linux-nfs@vger.kernel.org, linux-kernel Subject: Re: Proposal: Use hi-res clock for file timestamps Message-ID: <20100818185354.GB6567@basil.fritz.box> References: <87aaolwar8.fsf@basil.nowhere.org> <20100817174134.GA23176@fieldses.org> <20100817182920.GD18161@basil.fritz.box> <1282155658.13405.36.camel@localhost> Content-Type: text/plain; charset=us-ascii In-Reply-To: <1282155658.13405.36.camel@localhost> Sender: linux-nfs-owner@vger.kernel.org List-ID: MIME-Version: 1.0 On Wed, Aug 18, 2010 at 07:20:58PM +0100, David Woodhouse wrote: > On Tue, 2010-08-17 at 20:29 +0200, Andi Kleen wrote: > > > > > - Increment mtime by a nanosecond when necessary. > > > > You cannot be more precise than the backing file system: this causes > > non monotonity when the inodes are flushed (has happened in the past) > > Um, can't you? You can't *store* timestamps which are more precise, but > they can be in cache can't they? No you can't. The initial implementation did that and it broke someone's make. After that the VFS was fixed to never be precise than the backing file system. -Andi