From: Jan Kara Subject: Re: [PATCH] ext4: fix race aio-dio vs freeze_fs Date: Wed, 25 Nov 2015 11:25:00 +0100 Message-ID: <20151125102500.GM25232@quack.suse.cz> References: <1448294568-20892-1-git-send-email-dmonakhov@openvz.org> <20151124132421.GG25232@quack.suse.cz> <20151124160723.GA17466@infradead.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Jan Kara , Dmitry Monakhov , linux-ext4@vger.kernel.org, tytso@mit.edu, xfs@oss.sgi.com, linux-fsdevel@vger.kernel.org To: Christoph Hellwig Return-path: Content-Disposition: inline In-Reply-To: <20151124160723.GA17466@infradead.org> Sender: linux-fsdevel-owner@vger.kernel.org List-Id: linux-ext4.vger.kernel.org On Tue 24-11-15 08:07:23, Christoph Hellwig wrote: > On Tue, Nov 24, 2015 at 02:24:21PM +0100, Jan Kara wrote: > > Well, this problem seems to suggest that we have the freeze protection for > > AIO writes wrong. We should call file_end_write() from aio_complete() and > > not from aio_run_iocb()... I believe XFS and other filesystems may have > > problems with this as well (CCed). Attached patch (so far only compile > > tested since my test machine is pondering on something else) should fix > > this. > > Sounds like one way to do it, but we'd really want a vfs_* helper for > this so that it doesn't have to duplicated in other write_iter users > like the loop driver, which seems to be missing file > file_start_write/file_end_write entirely. That is mostly a separate issue, isn't it? I guess you mean a helper like vfs_write_iter() that would get freeze protection and call ->write_iter()? And what about files which have ->write (or ->splice_write()) and don't end up calling ->write_iter? Also there is stuff like __kernel_write() which ends up calling ->write_iter() but e.g. kernel/acct.c: do_acct_process() wants to do it's own thing... So the current status is not ideal but I don't see how to make it substantially better... Honza -- Jan Kara SUSE Labs, CR