From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andi Kleen Subject: Re: Tracking actual disk write sources instead of flush thread Date: Wed, 16 Apr 2014 12:33:54 -0700 Message-ID: <871twxhxwd.fsf@tassilo.jf.intel.com> References: <534DE477.5080904@ubuntu.com> <20140416140147.GA15429@linux.intel.com> <534E9E9D.5030600@ubuntu.com> Mime-Version: 1.0 Content-Type: text/plain Cc: Matthew Wilcox , linux-fsdevel@vger.kernel.org To: Phillip Susi Return-path: Received: from mga03.intel.com ([143.182.124.21]:4742 "EHLO mga03.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751305AbaDPTdz (ORCPT ); Wed, 16 Apr 2014 15:33:55 -0400 In-Reply-To: <534E9E9D.5030600@ubuntu.com> (Phillip Susi's message of "Wed, 16 Apr 2014 11:15:41 -0400") Sender: linux-fsdevel-owner@vger.kernel.org List-ID: Phillip Susi writes: > > That seems like a severe defect. How can such a defect be tolerated > in this day and age? Why does the io accounting not track how many > pages the process dirties rather than how many it actually initiates > the writeout for? There is no per process dirty counter anyways. The exiting proces write counter is legacy and generally useless. If you want to track syscalls there are already plenty of ways to do that per process. There's no way to track background mmap flushes and it may be meaningless in this case anyways. -Andi -- ak@linux.intel.com -- Speaking for myself only