Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1763381AbYARSAZ (ORCPT ); Fri, 18 Jan 2008 13:00:25 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1757005AbYARSAM (ORCPT ); Fri, 18 Jan 2008 13:00:12 -0500 Received: from smtp2.linux-foundation.org ([207.189.120.14]:46268 "EHLO smtp2.linux-foundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752292AbYARSAK (ORCPT ); Fri, 18 Jan 2008 13:00:10 -0500 Date: Fri, 18 Jan 2008 09:58:04 -0800 (PST) From: Linus Torvalds To: Peter Zijlstra cc: Miklos Szeredi , salikhmetov@gmail.com, linux-mm@kvack.org, jakob@unthought.net, linux-kernel@vger.kernel.org, valdis.kletnieks@vt.edu, riel@redhat.com, ksm@42.dk, staubach@redhat.com, jesper.juhl@gmail.com, akpm@linux-foundation.org, protasnb@gmail.com, r.e.wolff@bitwizard.nl, hidave.darkstar@gmail.com, hch@infradead.org Subject: Re: [PATCH -v6 2/2] Updating ctime and mtime for memory-mapped files In-Reply-To: <1200651958.5920.12.camel@twins> Message-ID: References: <12006091182260-git-send-email-salikhmetov@gmail.com> <12006091211208-git-send-email-salikhmetov@gmail.com> <1200651337.5920.9.camel@twins> <1200651958.5920.12.camel@twins> User-Agent: Alpine 1.00 (LFD 882 2007-12-20) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1153 Lines: 30 On Fri, 18 Jan 2008, Peter Zijlstra wrote: > > Bah, and will break on s390... so we'd need a page_mkclean() variant > that doesn't actually clear dirty. No, we simply want to not play all these very expensive games with dirty in the first place. Guys, mmap access times aren't important enough for this. It's not specified closely enough, and people don't care enough. Of the patches around so far, the best one by far seems to be the simple four-liner from Miklos. And even in that four-liner, I suspect that the *last* two lines are actually incorrect: there's no point in updating the file time when the page *becomes* dirty, we should update the file time when it is marked clean, and "msync(MS_SYNC)" should update it as part of *that*. So I think the file time update should be part of just the page writeout logic, not by msync() or page faulting itself or anything like that. Linus -- 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/