From mboxrd@z Thu Jan 1 00:00:00 1970 From: Hou Pengyang Subject: Re: [PATCH] f2fs: fix conflict between atomic_write and truncate Date: Tue, 13 Dec 2016 09:11:52 +0800 Message-ID: <584F4AD8.4000008@huawei.com> References: <20161212131618.92800-1-houpengyang@huawei.com> <20161212191851.GC61413@jaegeuk.local> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Received: from sog-mx-1.v43.ch3.sourceforge.com ([172.29.43.191] helo=mx.sourceforge.net) by sfs-ml-1.v29.ch3.sourceforge.com with esmtp (Exim 4.76) (envelope-from ) id 1cGbdn-0001Sj-Iq for linux-f2fs-devel@lists.sourceforge.net; Tue, 13 Dec 2016 01:12:23 +0000 Received: from szxga02-in.huawei.com ([119.145.14.65]) by sog-mx-1.v43.ch3.sourceforge.com with esmtps (TLSv1:RC4-SHA:128) (Exim 4.76) id 1cGbdj-0005BK-GI for linux-f2fs-devel@lists.sourceforge.net; Tue, 13 Dec 2016 01:12:23 +0000 In-Reply-To: <20161212191851.GC61413@jaegeuk.local> List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: linux-f2fs-devel-bounces@lists.sourceforge.net To: Jaegeuk Kim Cc: chao@kernel.org, linux-f2fs-devel@lists.sourceforge.net On 2016/12/13 3:18, Jaegeuk Kim wrote: > Hi Pengyang, > Hi, kim, > On 12/12, Hou Pengyang wrote: >> In fsync_node_pages path, after locking the last_page for setting dirty, >> we should check if the page has been truncated. Or there may be a mem leak, >> as this dirty last_page will NOT be found in next page-cache travese. >> >> This patch adds page->mapping checking, and will NOT rewrite the page if >> it has been truncated. >> >> Signed-off-by: Hou Pengyang >> --- >> fs/f2fs/node.c | 5 +++++ >> 1 file changed, 5 insertions(+) >> >> diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c >> index c1bbfdc..78c5b50 100644 >> --- a/fs/f2fs/node.c >> +++ b/fs/f2fs/node.c >> @@ -1410,6 +1410,11 @@ int fsync_node_pages(struct f2fs_sb_info *sbi, struct inode *inode, >> "Retry to write fsync mark: ino=%u, idx=%lx", >> ino, last_page->index); >> lock_page(last_page); > > We grabbed a reference count for this last_page, and thus, it won't be released. > Something that I missed? > Some path (e.g. shrink) would NOT release the page if we've grabbed it, But truncate operation will invalidate the page from page-cache regardless of the page reference. Patch: 0fac558b965 f2fs: make atomic/volatile operation exclusive seems fix this issue by adding inode_lock(inode) to avoid conflict between atomic/volatile and truncate operation. But this patch makes codes more robust. Thanks. > Thanks, > >> + if (unlikely(last_page->mapping != NODE_MAPPING(sbi))) { >> + unlock_page(last_page); >> + f2fs_put_page(last_page, 0); >> + goto out; >> + } >> f2fs_wait_on_page_writeback(last_page, NODE, true); >> set_page_dirty(last_page); >> unlock_page(last_page); >> -- >> 2.10.1 >> >> >> ------------------------------------------------------------------------------ >> Developer Access Program for Intel Xeon Phi Processors >> Access to Intel Xeon Phi processor-based developer platforms. >> With one year of Intel Parallel Studio XE. >> Training and support from Colfax. >> Order your platform today.http://sdm.link/xeonphi >> _______________________________________________ >> Linux-f2fs-devel mailing list >> Linux-f2fs-devel@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel > > . > ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, SlashDot.org! http://sdm.link/slashdot