From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752785AbcBPFid (ORCPT ); Tue, 16 Feb 2016 00:38:33 -0500 Received: from ipmail04.adl6.internode.on.net ([150.101.137.141]:12132 "EHLO ipmail04.adl6.internode.on.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751234AbcBPFib (ORCPT ); Tue, 16 Feb 2016 00:38:31 -0500 X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: A2BlBwCatMJWXJbY03ZegzqBP6ZPAQEBAQEBBotvhUaECIYHBAICgTNNAQEBAQEBB0RAhEIBAQQ6HCMQCAMYCSUPBSUDBxoTiBm4HAEBCAIeGIUxhH2IbAEEln2NTo58RI18gmEcgVwoLohPAQEB Date: Tue, 16 Feb 2016 16:38:28 +1100 From: Dave Chinner To: "Nag Avadhanam (nag)" Cc: "Theodore Ts'o" , "Daniel Walker (danielwa)" , Alexander Viro , "Khalid Mughal (khalidm)" , "xe-kernel@external.cisco.com" , "dave.hansen@intel.com" , "hannes@cmpxchg.org" , "riel@redhat.com" , Jonathan Corbet , "linux-doc@vger.kernel.org" , "linux-kernel@vger.kernel.org" , "linux-fsdevel@vger.kernel.org" , "linux-mm@kvack.org" Subject: Re: [PATCH] kernel: fs: drop_caches: add dds drop_caches_count Message-ID: <20160216053827.GX19486@dastard> References: <1455308080-27238-1-git-send-email-danielwa@cisco.com> <20160214211856.GT19486@dastard> <56C216CA.7000703@cisco.com> <20160215230511.GU19486@dastard> <56C264BF.3090100@cisco.com> <20160216004531.GA28260@thunk.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tue, Feb 16, 2016 at 02:58:04AM +0000, Nag Avadhanam (nag) wrote: > Its the calculation of the # of bytes of non-reclaimable file system cache > pages that has been troubling us. We do not want to count inactive file > pages (of programs/binaries) that were once mapped by any process in the > system as reclaimable because that might lead to thrashing under memory > pressure (we want to alert admins before system starts dropping text > pages). The code presented does not match your requirements. It only counts pages that are currently mapped into ptes. hence it will tell you that once-used and now unmapped binary pages are reclaimable, and drop caches will reclaim them. hence they'll need to be fetched from disk again if they are faulted in again after a drop_caches run. Cheers, Dave. -- Dave Chinner david@fromorbit.com