From: "Aneesh Kumar K.V" Subject: Re: [PATCH 2/4] ext3: truncate block allocated on a failed ext3_write_begin Date: Thu, 18 Sep 2008 12:33:40 +0530 Message-ID: <20080918070340.GA7042@skywalker> References: <1221319971-29879-1-git-send-email-tytso@mit.edu> <1221319971-29879-2-git-send-email-tytso@mit.edu> <20080917122254.a704eff1.akpm@linux-foundation.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: "Theodore Ts'o" , linux-kernel@vger.kernel.org, linux-ext4@vger.kernel.org, Nick Piggin To: Andrew Morton Return-path: Received: from e28smtp05.in.ibm.com ([59.145.155.5]:47419 "EHLO e28esmtp05.in.ibm.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1751210AbYIRHEx (ORCPT ); Thu, 18 Sep 2008 03:04:53 -0400 Content-Disposition: inline In-Reply-To: <20080917122254.a704eff1.akpm@linux-foundation.org> Sender: linux-ext4-owner@vger.kernel.org List-ID: On Wed, Sep 17, 2008 at 12:22:54PM -0700, Andrew Morton wrote: > On Sat, 13 Sep 2008 11:32:49 -0400 > "Theodore Ts'o" wrote: > > > From: Aneesh Kumar K.V > > > > For blocksize < pagesize we need to remove blocks that got allocated in > > block_write_begin() if we fail with ENOSPC for later blocks. > > block_write_begin() internally does this if it allocated page > > locally. This makes sure we don't have blocks outside inode.i_size > > during ENOSPC. > > > > Signed-off-by: Aneesh Kumar K.V > > Signed-off-by: "Theodore Ts'o" > > Cc: linux-ext4@vger.kernel.org > > --- > > fs/ext3/inode.c | 7 +++++++ > > 1 files changed, 7 insertions(+), 0 deletions(-) > > > > diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c > > index 507d868..bff22b9 100644 > > --- a/fs/ext3/inode.c > > +++ b/fs/ext3/inode.c > > @@ -1178,6 +1178,13 @@ write_begin_failed: > > ext3_journal_stop(handle); > > unlock_page(page); > > page_cache_release(page); > > + /* > > + * block_write_begin may have instantiated a few blocks > > + * outside i_size. Trim these off again. Don't need > > + * i_size_read because we hold i_mutex. > > + */ > > + if (pos + len > inode->i_size) > > + vmtruncate(inode, inode->i_size); > > } > > if (ret == -ENOSPC && ext3_should_retry_alloc(inode->i_sb, &retries)) > > goto retry; > > Well we used to do this trimming in core VFS, but Nick broke it. We > still do it if the fs doesn't implement ->write_begin(). We still do it in block_write_begin if the pages are allocated by block_write_begin. > > Should we do this trimming in pagecache_write_begin() in both cases? pagecache_write_begin is not used in the write_begin call path for ext3/ext4. generic_file_buffered_write generic_perform_write ext3_write_begin block_write_begin -aneesh