From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S967432Ab2EPLGW (ORCPT ); Wed, 16 May 2012 07:06:22 -0400 Received: from mail-wi0-f172.google.com ([209.85.212.172]:52502 "EHLO mail-wi0-f172.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S966952Ab2EPLGT convert rfc822-to-8bit (ORCPT ); Wed, 16 May 2012 07:06:19 -0400 MIME-Version: 1.0 Reply-To: konrad@darnok.org In-Reply-To: <4FB3048C.20008@kernel.org> References: <4FAB21E7.7020703@kernel.org> <20120510140215.GC26152@phenom.dumpdata.com> <4FABD503.4030808@vflare.org> <4FABDA9F.1000105@linux.vnet.ibm.com> <20120510151941.GA18302@kroah.com> <4FABECF5.8040602@vflare.org> <20120510164418.GC13964@kroah.com> <4FABF9D4.8080303@vflare.org> <20120510173322.GA30481@phenom.dumpdata.com> <4FAC4E3B.3030909@kernel.org> <20120511192831.GC3785@phenom.dumpdata.com> <4FB06B91.1080008@kernel.org> <4FB3048C.20008@kernel.org> Date: Wed, 16 May 2012 07:06:18 -0400 X-Google-Sender-Auth: nnsBm4F9mX8OeP0-9j_tTwy3Fr0 Message-ID: Subject: Re: [PATCH 3/4] zsmalloc use zs_handle instead of void * From: Konrad Rzeszutek Wilk To: Minchan Kim Cc: Konrad Rzeszutek Wilk , linux-mm@kvack.org, linux-kernel@vger.kernel.org Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8BIT Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org >>> It's not good abstraction. >> >> If we want good abstraction, then I don't think 'unsigned long' is >> either? I mean it will do for the conversion from 'void *'. Perhaps I >> am being a bit optimistic here - and I am trying to jam in this >> 'struct zs_handle' in all cases but in reality it needs a more >> iterative process. So first do 'void *' -> 'unsigned long', and then >> later on if we can come up with something more nicely that abstracts >> - then use that? ..snip.. >>> No. What I want is to remove coupling zsallocator's handle with zram/zcache. >>> They shouldn't know internal of handle and assume it's a pointer. >> >> I concur. And hence I was thinking that the 'struct zs_handle *' >> pointer would work. > > > Do you really hate "unsigned long" as handle? ..snip,, >> Well, everything changes over time  so putting a stick in the ground >> and saying 'this must >> be this way' is not really the best way. > > > Hmm, agree on your above statement but I can't imagine better idea. > OK. Lets go with unsigned long. I can prep a patch next week when I am back from vacation unless somebody beats me to it.