From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-7.0 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_PATCH,MAILING_LIST_MULTI,SIGNED_OFF_BY,SPF_PASS,URIBL_BLOCKED autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 98080C43387 for ; Tue, 15 Jan 2019 16:20:20 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 6EF3420868 for ; Tue, 15 Jan 2019 16:20:20 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1731434AbfAOQUS (ORCPT ); Tue, 15 Jan 2019 11:20:18 -0500 Received: from relay.sw.ru ([185.231.240.75]:51534 "EHLO relay.sw.ru" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1730451AbfAOQUS (ORCPT ); Tue, 15 Jan 2019 11:20:18 -0500 Received: from [172.16.25.169] by relay.sw.ru with esmtp (Exim 4.91) (envelope-from ) id 1gjRRn-0004Bj-21; Tue, 15 Jan 2019 19:20:15 +0300 Subject: Re: [PATCH RFC] fuse: Prevent background write requests increase inode size To: miklos@szeredi.hu, linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org References: <154151286725.17550.12307253489084032792.stgit@localhost.localdomain> From: Kirill Tkhai Message-ID: <5a653d5f-268b-b2ff-2312-00effb30108f@virtuozzo.com> Date: Tue, 15 Jan 2019 19:20:14 +0300 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Thunderbird/60.4.0 MIME-Version: 1.0 In-Reply-To: <154151286725.17550.12307253489084032792.stgit@localhost.localdomain> Content-Type: text/plain; charset=utf-8 Content-Language: en-US Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi, Miklos, please take a look into two problems this patch points: 1)we have to set "io->async = false" in fuse_direct_IO(), when inode size is changed (we mustn't increase inode size in background); 2)we have to increment fi->writectr on all background requests. Thanks, Kirill On 06.11.2018 17:02, Kirill Tkhai wrote: > Hi, Miklos, > > this is not a well-tested patch, this is a concept, > showing the places, where it looks we have a problem. > > Commit 7879c4e58b7c made io->async careless about inode size, > and this is wrong. Asyncronuos background requests may be sent > to userspace after inode becomes unlocked, when background > queue is throttled. In this case we execute a write request > extending inode size without any protection, and this ruines > everything. Fix that. > > Also, some write background requests do not increment fi->writectr, > e.g.: > fuse_direct_IO() > fuse_direct_io() > fuse_send_write() > fuse_async_req_send() > fuse_request_send_background() > > This is a reason fuse_sync_writes() does not work as expected. > Fix that too. > > Signed-off-by: Kirill Tkhai > --- > fs/fuse/file.c | 20 ++++++++++++++++++-- > 1 file changed, 18 insertions(+), 2 deletions(-) > > diff --git a/fs/fuse/file.c b/fs/fuse/file.c > index cc2121b37bf5..0b71a3c96168 100644 > --- a/fs/fuse/file.c > +++ b/fs/fuse/file.c > @@ -610,6 +610,9 @@ static void fuse_aio_complete(struct fuse_io_priv *io, int err, ssize_t pos) > static void fuse_aio_complete_req(struct fuse_conn *fc, struct fuse_req *req) > { > struct fuse_io_priv *io = req->io; > + struct kiocb *iocb = io->iocb; > + struct file *file = iocb->ki_filp; > + struct fuse_inode *fi = get_fuse_inode(file_inode(file)); > ssize_t pos = -1; > > fuse_release_user_pages(req, io->should_dirty); > @@ -625,6 +628,12 @@ static void fuse_aio_complete_req(struct fuse_conn *fc, struct fuse_req *req) > } > > fuse_aio_complete(io, req->out.h.error, pos); > + > + if (io->write) { > + spin_lock(&fc->lock); > + fi->writectr--; > + spin_unlock(&fc->lock); > + } > } > > static size_t fuse_async_req_send(struct fuse_conn *fc, struct fuse_req *req, > @@ -966,6 +975,7 @@ static size_t fuse_send_write(struct fuse_req *req, struct fuse_io_priv *io, > { > struct kiocb *iocb = io->iocb; > struct file *file = iocb->ki_filp; > + struct fuse_inode *fi = get_fuse_inode(file_inode(file)); > struct fuse_file *ff = file->private_data; > struct fuse_conn *fc = ff->fc; > struct fuse_write_in *inarg = &req->misc.write.in; > @@ -981,8 +991,12 @@ static size_t fuse_send_write(struct fuse_req *req, struct fuse_io_priv *io, > inarg->lock_owner = fuse_lock_owner_id(fc, owner); > } > > - if (io->async) > + if (io->async) { > + spin_lock(&fc->lock); > + fi->writectr++; > + spin_unlock(&fc->lock); > return fuse_async_req_send(fc, req, count, io); > + } > > fuse_request_send(fc, req); > return req->misc.write.out.size; > @@ -2904,8 +2918,10 @@ fuse_direct_IO(struct kiocb *iocb, struct iov_iter *iter) > * We cannot asynchronously extend the size of a file. > * In such case the aio will behave exactly like sync io. > */ > - if ((offset + count > i_size) && iov_iter_rw(iter) == WRITE) > + if ((offset + count > i_size) && iov_iter_rw(iter) == WRITE) { > io->blocking = true; > + io->async = false; > + } > > if (io->async && io->blocking) { > /* >