From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from e28smtp08.in.ibm.com ([122.248.162.8]:43492 "EHLO e28smtp08.in.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751029AbbJBQe3 (ORCPT ); Fri, 2 Oct 2015 12:34:29 -0400 Received: from /spool/local by e28smtp08.in.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Fri, 2 Oct 2015 22:04:26 +0530 Received: from d28relay05.in.ibm.com (d28relay05.in.ibm.com [9.184.220.62]) by d28dlp02.in.ibm.com (Postfix) with ESMTP id 893F7394005A for ; Fri, 2 Oct 2015 22:04:23 +0530 (IST) Received: from d28av05.in.ibm.com (d28av05.in.ibm.com [9.184.220.67]) by d28relay05.in.ibm.com (8.14.9/8.14.9/NCO v10.0) with ESMTP id t92GYLTx6947138 for ; Fri, 2 Oct 2015 22:04:21 +0530 Received: from d28av05.in.ibm.com (localhost [127.0.0.1]) by d28av05.in.ibm.com (8.14.4/8.14.4/NCO v10.0 AVout) with ESMTP id t92GYJJr026840 for ; Fri, 2 Oct 2015 22:04:20 +0530 From: Chandan Rajendra To: Josef Bacik Cc: linux-btrfs@vger.kernel.org, clm@fb.com, bo.li.liu@oracle.com, dsterba@suse.cz, quwenruo@cn.fujitsu.com, chandan@mykolab.com Subject: Re: [PATCH V5 11/13] Btrfs: Clean pte corresponding to page straddling i_size Date: Fri, 02 Oct 2015 22:04:17 +0530 Message-ID: <9501565.vVUm90Rsjs@localhost.localdomain> In-Reply-To: <560D49F0.4040200@fb.com> References: <1443608912-31667-1-git-send-email-chandan@linux.vnet.ibm.com> <1443608912-31667-12-git-send-email-chandan@linux.vnet.ibm.com> <560D49F0.4040200@fb.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: linux-btrfs-owner@vger.kernel.org List-ID: On Thursday 01 Oct 2015 10:57:52 Josef Bacik wrote: > On 09/30/2015 06:28 AM, Chandan Rajendra wrote: > > When extending a file by either "truncate up" or by writing beyond i_size, > > the page which had i_size needs to be marked "read only" so that future > > writes to the page via mmap interface causes btrfs_page_mkwrite() to be > > invoked. If not, a write performed after extending the file via the mmap > > interface will find the page to be writaeable and continue writing to the > > page without invoking btrfs_page_mkwrite() i.e. we end up writing to a > > file without reserving disk space. > > > > Signed-off-by: Chandan Rajendra > > --- > > > > fs/btrfs/file.c | 12 ++++++++++-- > > fs/btrfs/inode.c | 2 +- > > 2 files changed, 11 insertions(+), 3 deletions(-) > > > > diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c > > index 360d56d..5715e29 100644 > > --- a/fs/btrfs/file.c > > +++ b/fs/btrfs/file.c > > @@ -1757,6 +1757,8 @@ static ssize_t btrfs_file_write_iter(struct kiocb > > *iocb,> > > ssize_t err; > > loff_t pos; > > size_t count; > > > > + loff_t oldsize; > > + int clean_page = 0; > > > > mutex_lock(&inode->i_mutex); > > err = generic_write_checks(iocb, from); > > > > @@ -1795,14 +1797,17 @@ static ssize_t btrfs_file_write_iter(struct kiocb > > *iocb,> > > pos = iocb->ki_pos; > > count = iov_iter_count(from); > > start_pos = round_down(pos, root->sectorsize); > > > > - if (start_pos > i_size_read(inode)) { > > + oldsize = i_size_read(inode); > > + if (start_pos > oldsize) { > > > > /* Expand hole size to cover write data, preventing empty gap */ > > end_pos = round_up(pos + count, root->sectorsize); > > > > - err = btrfs_cont_expand(inode, i_size_read(inode), end_pos); > > + err = btrfs_cont_expand(inode, oldsize, end_pos); > > > > if (err) { > > > > mutex_unlock(&inode->i_mutex); > > goto out; > > > > } > > > > + if (start_pos > round_up(oldsize, root->sectorsize)) > > + clean_page = 1; > > > > } > > > > if (sync) > > > > @@ -1814,6 +1819,9 @@ static ssize_t btrfs_file_write_iter(struct kiocb > > *iocb,> > > num_written = __btrfs_buffered_write(file, from, pos); > > if (num_written > 0) > > > > iocb->ki_pos = pos + num_written; > > > > + if (clean_page) > > + pagecache_isize_extended(inode, oldsize, > > + i_size_read(inode)); > > > > } > > > > mutex_unlock(&inode->i_mutex); > > > > diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c > > index c937357..f31da87 100644 > > --- a/fs/btrfs/inode.c > > +++ b/fs/btrfs/inode.c > > @@ -4853,7 +4853,6 @@ static int btrfs_setsize(struct inode *inode, struct > > iattr *attr)> > > } > > > > if (newsize > oldsize) { > > > > - truncate_pagecache(inode, newsize); > > So I don't understand why we are dropping this bit here, could you > explain? Otherwise the patch looks fine to me. Thanks, > Josef, As per our previous discussion on IRC we found that the "truncate_pagecache(inode, newsize)" statement to be a relic of the past. During the test runs, I haven't seen any sort of failure caused by the removal of this statement. -- chandan