From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrea Arcangeli Subject: Re: Corruption with O_DIRECT and unaligned user buffers Date: Sat, 20 Dec 2008 16:55:36 +0100 Message-ID: <20081220155536.GD6383@random.random> References: <491DAF8E.4080506@quantum.com> <200811191526.00036.nickpiggin@yahoo.com.au> <20081119165819.GE19209@random.random> <20081218152952.GW24856@random.random> <20081219161911.dcf15331.kamezawa.hiroyu@jp.fujitsu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Nick Piggin , Tim LaBerge , linux-mm@kvack.org, linux-fsdevel@vger.kernel.org To: KAMEZAWA Hiroyuki Return-path: Received: from mx2.redhat.com ([66.187.237.31]:49593 "EHLO mx2.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752142AbYLTPzt (ORCPT ); Sat, 20 Dec 2008 10:55:49 -0500 Content-Disposition: inline In-Reply-To: <20081219161911.dcf15331.kamezawa.hiroyu@jp.fujitsu.com> Sender: linux-fsdevel-owner@vger.kernel.org List-ID: On Fri, Dec 19, 2008 at 04:19:11PM +0900, KAMEZAWA Hiroyuki wrote: > Result of cost-of-fork() on ia64. > == > size of memory before after > Anon=1M , 0.07ms, 0.08ms > Anon=10M , 0.17ms, 0.22ms > Anon=100M , 1.15ms, 1.64ms > Anon=1000M , 11.5ms, 15.821ms > == > > fork() cost is 135% when the process has 1G of Anon. Not sure where the 135% number comes from. The above number shows a performance decrease of 27% or a time increase of 37% which I hope is inline with the overhead introduced by the TestSetPageLocked in the fast path (which I didn't expect to be so bad), but that it's almost trivial to eliminate with a smb_wmb in add_to_swap_cache and a smb_rmb in fork. So we'll need to repeat this measurement after replacing the TestSetPageLocked with smb_rmb.