From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail143.messagelabs.com (mail143.messagelabs.com [216.82.254.35]) by kanga.kvack.org (Postfix) with ESMTP id A83BB6B00D9 for ; Tue, 9 Mar 2010 18:14:55 -0500 (EST) Date: Tue, 9 Mar 2010 15:13:34 -0800 From: Andrew Morton Subject: Re: [PATCH -mmotm 7/8] memcg: move charges of anonymous swap Message-Id: <20100309151334.f28d9930.akpm@linux-foundation.org> In-Reply-To: <20100205011602.GA8416@linux-sh.org> References: <20091221143106.6ff3ca15.nishimura@mxp.nes.nec.co.jp> <20091221143816.9794cd17.nishimura@mxp.nes.nec.co.jp> <20100203193127.fe5efa17.akpm@linux-foundation.org> <20100204140942.0ef6d7b1.nishimura@mxp.nes.nec.co.jp> <20100204142736.2a8bec26.kamezawa.hiroyu@jp.fujitsu.com> <20100204071840.GC5574@linux-sh.org> <20100204164441.d012f6fa.kamezawa.hiroyu@jp.fujitsu.com> <20100205093806.5699d406.nishimura@mxp.nes.nec.co.jp> <20100205011602.GA8416@linux-sh.org> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org To: Paul Mundt Cc: Daisuke Nishimura , KAMEZAWA Hiroyuki , Balbir Singh , Li Zefan , Paul Menage , linux-mm List-ID: On Fri, 5 Feb 2010 10:16:02 +0900 Paul Mundt wrote: > On Fri, Feb 05, 2010 at 09:38:06AM +0900, Daisuke Nishimura wrote: > > On Thu, 4 Feb 2010 16:44:41 +0900, KAMEZAWA Hiroyuki wrote: > > > On Thu, 4 Feb 2010 16:18:40 +0900 > > > Paul Mundt wrote: > > > > CONFIG_SWAP itself is configurable even with CONFIG_MMU=y, so having > > > > stubbed out helpers for the CONFIG_SWAP=n case would give the compiler a > > > > chance to optimize things away in those cases, too. Embedded systems > > > > especially will often have MMU=y and BLOCK=n, resulting in SWAP being > > > > unset but swap cache encodings still defined. > > > > > > > > How about just changing the is_swap_pte() definition to depend on SWAP > > > > instead? > > > > > > > I think the new feature as "move task charge" itself depends on CONFIG_MMU > > > because it walks a process's page table. > > > > > > Then, how about this ? (sorry, I can't test this in valid way..) > > > > > I agree to this direction of making "move charge" depend on CONFIG_MMU, > > although I can't test !CONFIG_MMU case either. > > > I'll try to give it a test on nommu today and see how it goes. The patch is still breaking the NOMMU build for me: mm/memcontrol.c: In function `is_target_pte_for_mc': mm/memcontrol.c:3641: error: implicit declaration of function `is_swap_pte' btw, sh allmodconfig gives me a huge spew related to the missing definition of pt_regs: /usr/src/25/arch/sh/include/asm/system.h:166: warning: parameter has incomplete type /usr/src/25/arch/sh/include/asm/system.h:167: warning: parameter has incomplete type /usr/src/25/arch/sh/include/asm/system.h:168: warning: parameter has incomplete type /usr/src/25/arch/sh/include/asm/system.h:169: warning: parameter has incomplete type /usr/src/25/arch/sh/include/asm/system.h:170: warning: parameter has incomplete type /usr/src/25/arch/sh/include/asm/system.h:171: warning: parameter has incomplete type /usr/src/25/arch/sh/include/asm/system.h:172: warning: parameter has incomplete type /usr/src/25/arch/sh/include/asm/system.h:173: warning: parameter has incomplete type The `struct pt_regs;' declararion isn't enough - it wants to see the definition. I think. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org