From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail172.messagelabs.com (mail172.messagelabs.com [216.82.254.3]) by kanga.kvack.org (Postfix) with ESMTP id 7928A6B005C for ; Wed, 18 Mar 2009 03:13:15 -0400 (EDT) Date: Wed, 18 Mar 2009 03:13:13 -0400 From: Christoph Hellwig Subject: fput under mmap_sem Message-ID: <20090318071313.GA30011@infradead.org> References: <200903151459.01320.denys@visp.net.lb> <20090315221921.GY26138@disturbed> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20090315221921.GY26138@disturbed> Sender: owner-linux-mm@kvack.org To: xfs@oss.sgi.com, linux-mm@kvack.org List-ID: On Mon, Mar 16, 2009 at 09:19:21AM +1100, Dave Chinner wrote: > This is a VM problem where it calls fput() with the mmap_sem() held > in remove_vma(). It makes the incorrect assumption that filesystems > will never use the same lock in the IO path and the inode release path. > > This can deadlock if you are really unlucky. I really wonder why other filesystems haven't hit this yet. Any chance we can get the fput moved out of mmap_sem to get rid of this class of problems? -- 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