From: Marco Stornelli Subject: Re: Hole punching and mmap races Date: Tue, 5 Jun 2012 08:22:29 +0200 Message-ID: References: <20120515224805.GA25577@quack.suse.cz> <20120516021423.GO25351@dastard> <20120516130445.GA27661@quack.suse.cz> <20120517074308.GQ25351@dastard> <20120517232829.GA31028@quack.suse.cz> <20120518101210.GX25351@dastard> <20120518133250.GC5589@quack.suse.cz> <20120519014024.GZ25351@dastard> <20120524123538.GA5632@quack.suse.cz> <20120605055150.GF4347@dastard> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Jan Kara , linux-fsdevel@vger.kernel.org, xfs@oss.sgi.com, linux-ext4@vger.kernel.org, Hugh Dickins , linux-mm@kvack.org To: Dave Chinner Return-path: Received: from mail-ey0-f174.google.com ([209.85.215.174]:55002 "EHLO mail-ey0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757618Ab2FEGWb convert rfc822-to-8bit (ORCPT ); Tue, 5 Jun 2012 02:22:31 -0400 In-Reply-To: <20120605055150.GF4347@dastard> Sender: linux-ext4-owner@vger.kernel.org List-ID: 2012/6/5 Dave Chinner : > On Thu, May 24, 2012 at 02:35:38PM +0200, Jan Kara wrote: >> On Sat 19-05-12 11:40:24, Dave Chinner wrote: >> > So let's step back a moment and have a look at how we've got here. >> > The problem is that we've optimised ourselves into a corner with t= he >> > way we handle page cache truncation - we don't need mmap >> > serialisation because of the combination of i_size and page locks >> > mean we can detect truncated pages safely at page fault time. With >> > hole punching, we don't have that i_size safety blanket, and so we >> > need some other serialisation mechanism to safely detect whether a >> > page is valid or not at any given point in time. >> > >> > Because it needs to serialise against IO operations, we need a >> > sleeping lock of some kind, and it can't be the existing IO lock. >> > And now we are looking at needing a new lock for hole punching, I'= m >> > really wondering if the i_size/page lock truncation optimisation >> > should even continue to exist. i.e. replace it with a single >> > mechanism that works for both hole punching, truncation and other >> > functions that require exclusive access or exclusion against >> > modifications to the mapping tree. >> > >> > But this is only one of the problems in this area.The way I see it >> > is that we have many kludges in the area of page invalidation w.r.= t. >> > different types of IO, the page cache and mmap, especially when we >> > take into account direct IO. What we are seeing here is we need >> > some level of _mapping tree exclusion_ between: >> > >> > =A0 =A0 1. mmap vs hole punch (broken) >> > =A0 =A0 2. mmap vs truncate (i_size/page lock) >> > =A0 =A0 3. mmap vs direct IO (non-existent) >> > =A0 =A0 4. mmap vs buffered IO (page lock) >> > =A0 =A0 5. writeback vs truncate (i_size/page lock) >> > =A0 =A0 6. writeback vs hole punch (page lock, possibly broken) >> > =A0 =A0 7. direct IO vs buffered IO (racy - flush cache before/aft= er DIO) >> =A0 Yes, this is a nice summary of the most interesting cases. For c= ompleteness, >> here are the remaining cases: >> =A0 8. mmap vs writeback (page lock) >> =A0 9. writeback vs direct IO (as direct IO vs buffered IO) >> =A010. writeback vs buffered IO (page lock) >> =A011. direct IO vs truncate (dio_wait) >> =A012. direct IO vs hole punch (dio_wait) >> =A013. buffered IO vs truncate (i_mutex for writes, i_size/page lock= for reads) >> =A014. buffered IO vs hole punch (fs dependent, broken for ext4) >> =A015. truncate vs hole punch (fs dependent) >> =A016. mmap vs mmap (page lock) >> =A017. writeback vs writeback (page lock) >> =A018. direct IO vs direct IO (i_mutex or fs dependent) >> =A019. buffered IO vs buffered IO (i_mutex for writes, page lock for= reads) >> =A020. truncate vs truncate (i_mutex) >> =A021. punch hole vs punch hole (fs dependent) > I think we have even the xip cases here. Marco -- To unsubscribe from this list: send the line "unsubscribe linux-ext4" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html