From patchwork Wed Sep 20 11:05:04 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Eryu Guan X-Patchwork-Id: 9961293 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork.web.codeaurora.org (Postfix) with ESMTP id D65596056A for ; Wed, 20 Sep 2017 11:05:11 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id C78E529096 for ; Wed, 20 Sep 2017 11:05:11 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id BC0FF2909D; Wed, 20 Sep 2017 11:05:11 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-6.9 required=2.0 tests=BAYES_00,RCVD_IN_DNSWL_HI autolearn=unavailable version=3.3.1 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 5B72129096 for ; Wed, 20 Sep 2017 11:05:11 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751828AbdITLFK (ORCPT ); Wed, 20 Sep 2017 07:05:10 -0400 Received: from mx1.redhat.com ([209.132.183.28]:39980 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751585AbdITLFJ (ORCPT ); Wed, 20 Sep 2017 07:05:09 -0400 Received: from smtp.corp.redhat.com (int-mx05.intmail.prod.int.phx2.redhat.com [10.5.11.15]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 0BE1EC9D2C; Wed, 20 Sep 2017 11:05:09 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mx1.redhat.com 0BE1EC9D2C Authentication-Results: ext-mx05.extmail.prod.ext.phx2.redhat.com; dmarc=none (p=none dis=none) header.from=redhat.com Authentication-Results: ext-mx05.extmail.prod.ext.phx2.redhat.com; spf=fail smtp.mailfrom=eguan@redhat.com Received: from localhost (dhcp-12-147.nay.redhat.com [10.66.12.147]) by smtp.corp.redhat.com (Postfix) with ESMTP id 5806F5D6A3; Wed, 20 Sep 2017 11:05:05 +0000 (UTC) Date: Wed, 20 Sep 2017 19:05:04 +0800 From: Eryu Guan To: Christoph Hellwig Cc: Brian Foster , Eric Sandeen , fstests , Zheng Liu , linux-xfs@vger.kernel.org Subject: Re: [PATCH] test race when checking i_size on direct i/o read Message-ID: <20170920110504.GU8034@eguan.usersys.redhat.com> References: <799f63d9-de03-eab3-1e6e-cd747aa04c36@redhat.com> <20170919073616.GP8034@eguan.usersys.redhat.com> <20170919141351.GB3487@bfoster.bfoster> <20170919143406.GA7437@infradead.org> MIME-Version: 1.0 Content-Disposition: inline In-Reply-To: <20170919143406.GA7437@infradead.org> User-Agent: Mutt/1.8.3 (2017-05-23) X-Scanned-By: MIMEDefang 2.79 on 10.5.11.15 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.29]); Wed, 20 Sep 2017 11:05:09 +0000 (UTC) Sender: linux-xfs-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-xfs@vger.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP On Tue, Sep 19, 2017 at 07:34:06AM -0700, Christoph Hellwig wrote: > On Tue, Sep 19, 2017 at 10:13:52AM -0400, Brian Foster wrote: > > Can we pass a boolean or flag to xfs_iomap_write_unwritten() to have it > > update the incore i_size after unwritten extent conversion? Then move > > (or remove) the associated update from xfs_dio_write_end_io(). > > I don't think we even need a flag - all three callers of > xfs_iomap_write_unwritten want to update the file size. I tried this approach, but seems there's some problem in the buffered aio path, generic/112 (aio fsx) failed quickly. But I haven't digged into the reason (maybe I screwed it up, not the method is wrong..). Then I tried Brian's suggestion, pass a boolean to xfs_iomap_write_unwritten() to tell if we want it to update in-core isize after unwritten extent conversion, and skip the in-core isize update in xfs_dio_write_end_io() accordingly. This approach seems to work, it passed the test Eric posted here, and fstests 'aio' group tests, a run of 'quick' group didn't find any new failure as well. I attached the WIP patch (without proper comments) I was testing, if this looks fine I can format a formal patch and do more testings. Thanks, Eryu --- To unsubscribe from this list: send the line "unsubscribe linux-xfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c index 29172609f2a3..288da47e9ac5 100644 --- a/fs/xfs/xfs_aops.c +++ b/fs/xfs/xfs_aops.c @@ -343,7 +343,7 @@ xfs_end_io( error = xfs_reflink_end_cow(ip, offset, size); break; case XFS_IO_UNWRITTEN: - error = xfs_iomap_write_unwritten(ip, offset, size); + error = xfs_iomap_write_unwritten(ip, offset, size, false); break; default: ASSERT(!xfs_ioend_is_append(ioend) || ioend->io_append_trans); diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c index 350b6d43ba23..f3ad024573e7 100644 --- a/fs/xfs/xfs_file.c +++ b/fs/xfs/xfs_file.c @@ -435,6 +435,7 @@ xfs_dio_write_end_io( struct xfs_inode *ip = XFS_I(inode); loff_t offset = iocb->ki_pos; bool update_size = false; + bool write_unwritten = (flags & IOMAP_DIO_UNWRITTEN); int error = 0; trace_xfs_end_io_direct_write(ip, offset, size); @@ -458,7 +459,8 @@ xfs_dio_write_end_io( */ spin_lock(&ip->i_flags_lock); if (offset + size > i_size_read(inode)) { - i_size_write(inode, offset + size); + if (!write_unwritten) + i_size_write(inode, offset + size); update_size = true; } spin_unlock(&ip->i_flags_lock); @@ -469,8 +471,8 @@ xfs_dio_write_end_io( return error; } - if (flags & IOMAP_DIO_UNWRITTEN) - error = xfs_iomap_write_unwritten(ip, offset, size); + if (write_unwritten) + error = xfs_iomap_write_unwritten(ip, offset, size, update_size); else if (update_size) error = xfs_setfilesize(ip, offset, size); diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c index a1909bc064e9..0a088586371e 100644 --- a/fs/xfs/xfs_iomap.c +++ b/fs/xfs/xfs_iomap.c @@ -829,7 +829,8 @@ int xfs_iomap_write_unwritten( xfs_inode_t *ip, xfs_off_t offset, - xfs_off_t count) + xfs_off_t count, + bool update_size) { xfs_mount_t *mp = ip->i_mount; xfs_fileoff_t offset_fsb; @@ -840,6 +841,7 @@ xfs_iomap_write_unwritten( xfs_trans_t *tp; xfs_bmbt_irec_t imap; struct xfs_defer_ops dfops; + struct inode *inode = VFS_I(ip); xfs_fsize_t i_size; uint resblks; int error; @@ -900,6 +902,13 @@ xfs_iomap_write_unwritten( if (i_size > offset + count) i_size = offset + count; + if (update_size) { + spin_lock(&ip->i_flags_lock); + if (i_size > i_size_read(inode)) + i_size_write(inode, i_size); + spin_unlock(&ip->i_flags_lock); + } + i_size = xfs_new_eof(ip, i_size); if (i_size) { ip->i_d.di_size = i_size; diff --git a/fs/xfs/xfs_iomap.h b/fs/xfs/xfs_iomap.h index 00db3ecea084..ee535065c5d0 100644 --- a/fs/xfs/xfs_iomap.h +++ b/fs/xfs/xfs_iomap.h @@ -27,7 +27,7 @@ int xfs_iomap_write_direct(struct xfs_inode *, xfs_off_t, size_t, struct xfs_bmbt_irec *, int); int xfs_iomap_write_allocate(struct xfs_inode *, int, xfs_off_t, struct xfs_bmbt_irec *); -int xfs_iomap_write_unwritten(struct xfs_inode *, xfs_off_t, xfs_off_t); +int xfs_iomap_write_unwritten(struct xfs_inode *, xfs_off_t, xfs_off_t, bool); void xfs_bmbt_to_iomap(struct xfs_inode *, struct iomap *, struct xfs_bmbt_irec *); diff --git a/fs/xfs/xfs_pnfs.c b/fs/xfs/xfs_pnfs.c index 2f2dc3c09ad0..4246876df7b7 100644 --- a/fs/xfs/xfs_pnfs.c +++ b/fs/xfs/xfs_pnfs.c @@ -274,7 +274,7 @@ xfs_fs_commit_blocks( (end - 1) >> PAGE_SHIFT); WARN_ON_ONCE(error); - error = xfs_iomap_write_unwritten(ip, start, length); + error = xfs_iomap_write_unwritten(ip, start, length, false); if (error) goto out_drop_iolock; }