From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail202.messagelabs.com (mail202.messagelabs.com [216.82.254.227]) by kanga.kvack.org (Postfix) with ESMTP id D7F296B004D for ; Tue, 2 Feb 2010 01:44:15 -0500 (EST) Date: Tue, 2 Feb 2010 17:44:09 +1100 From: Nick Piggin Subject: Re: [PATCH -mm] remove VM_LOCK_RMAP code Message-ID: <20100202064409.GC6175@laptop> References: <20100128002000.2bf5e365@annuminas.surriel.com> <20100129151423.8b71b88e.akpm@linux-foundation.org> <20100129193410.7ce915d0@annuminas.surriel.com> <20100201061532.GC9085@laptop> <4B66F977.5010708@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <4B66F977.5010708@redhat.com> Sender: owner-linux-mm@kvack.org To: Rik van Riel Cc: Andrew Morton , linux-mm@kvack.org, linux-kernel@vger.kernel.org, KOSAKI Motohiro , lwoodman@redhat.com, Lee Schermerhorn , aarcange@redhat.com List-ID: On Mon, Feb 01, 2010 at 10:55:35AM -0500, Rik van Riel wrote: > On 02/01/2010 01:15 AM, Nick Piggin wrote: > >On Fri, Jan 29, 2010 at 07:34:10PM -0500, Rik van Riel wrote: > >>When a VMA is in an inconsistent state during setup or teardown, the > >>worst that can happen is that the rmap code will not be able to find > >>the page. > > > >OK, but you missed the interesting thing, which is to explain why > >that worst case is not a problem. > > > >rmap of course is not just used for reclaim but also invalidations > >from mappings, and those guys definitely need to know that all > >page table entries have been handled by the time they return. > > This is not a problem, because the mapping is in the process > of being torn down (PTEs just got invalidated by munmap), or > set up (no PTEs have been instantiated yet). > > The third case is split_vma, where we can have one VMA in an > inconsistent state (rmap cannot find the PTEs), while the > other VMA is still in its original state (rmap finds the PTEs > through that VMA). > > That is what makes this safe. OK, that sounds fine then. Your changelog was just a bit strange because you said it would not be able to find the page, which didn't really make sense. -- 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