From mboxrd@z Thu Jan 1 00:00:00 1970 From: Nick Piggin Subject: Re: [patch 1/6] fs: icache RCU free inodes Date: Thu, 11 Nov 2010 15:27:39 +1100 Message-ID: <20101111042739.GA6228@amd> References: <20101109124610.GB11477@amd> <87lj52pokp.fsf@basil.nowhere.org> <20101109213619.GA3246@amd> <20101110144721.GC7205@basil.fritz.box> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Nick Piggin , Al Viro , Linus Torvalds , linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org To: Andi Kleen Return-path: Received: from ipmail06.adl2.internode.on.net ([150.101.137.129]:15862 "EHLO ipmail06.adl2.internode.on.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932340Ab0KKE1q (ORCPT ); Wed, 10 Nov 2010 23:27:46 -0500 Content-Disposition: inline In-Reply-To: <20101110144721.GC7205@basil.fritz.box> Sender: linux-fsdevel-owner@vger.kernel.org List-ID: On Wed, Nov 10, 2010 at 03:47:21PM +0100, Andi Kleen wrote: > > It is in the inode_init_once pile, so yes it has to be returned > > to the allocator initialized. > > The kmem_cache_create()s in the file systems don't pass constructors > today. So I don't see how this could work reliably. They do, grep filesystems for inode_init_once