Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:1 as permitted sender) client-ip=2620:137:e000::3:1;
From:   Chunhai Guo <guochunhai@vivo.com>
To:     jack@suse.cz
Cc:     brauner@kernel.org, viro@zeniv.linux.org.uk,
        linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
        Chunhai Guo <guochunhai@vivo.com>
Subject: [PATCH] fs-writeback: do not requeue a clean inode having skipped pages
Date:   Fri, 15 Sep 2023 22:51:31 -0600
Message-Id: <20230916045131.957929-1-guochunhai@vivo.com>
Content-Transfer-Encoding: 8bit
Content-Type: text/plain
MIME-Version: 1.0
X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1
X-MS-Exchange-AntiSpam-MessageData-0: =?us-ascii?Q?qDGpopi8JF+iX8k2A8cJbYksi0DF+lcIGt8efgluxczroKjnV104l0c197qs?=
 =?us-ascii?Q?z0oPJ1j6xXylpwks5VTXmg30DlrNajw5Ygdzga90a1XfIFRnJUxs03u0JEf1?=
 =?us-ascii?Q?/U6XPbCQ6r2hW+2gZwlEdDSW5fPrChOAIrwwvjLPTj4pcCAG4ZkZ0xoBuNWt?=
 =?us-ascii?Q?qGp+uWtu6uXW7PhzZM3yWW9yqNm/INIUUY7F7fENtjJHJ5OrWFES/G3OwZGU?=
 =?us-ascii?Q?ZWlD7nttv3iRPd5OXs5wAuTNSnJZMg//WnnZghyTO9sEdB9dGNOXqNt5+zY8?=
 =?us-ascii?Q?8geegdjcrA+4IsO9W1WftHuxT9C3tD16NIfRpTxgT4y6jndkzw0iorra/qgJ?=
 =?us-ascii?Q?ED3KfBx55vkRXtR68dR4uTKkyJhip5UW5ZdwvvJ3XarYdY2KGLJVp9XCO11e?=
 =?us-ascii?Q?VJnXNPU4SIbNO7RGVt3KVeLjXpdeS3KMSz1FqeB1BVs8lhFnYlTkOwDFEENs?=
 =?us-ascii?Q?epNT+0Fe1OY9Tu5uZW5AVFBkIAql07RPE254+Sw8P9G6Bk0ROXRaRdttdoQg?=
 =?us-ascii?Q?miLRjsefBv/M4i1Chw62TS3Iq7wF8T/yqdiA7BC2srFiZgeFfzPBgALw6ViK?=
 =?us-ascii?Q?Z3795kks+ISaFInL6cJbM2gxrVmEwwa6dovPWDe8S8SMJPCxf980dotbP/od?=
 =?us-ascii?Q?eYIPlDN9624vDQMtnj9YXUE4Wfc5gt3SaTHd55o7sSg34m9kXS6IMbWrfgP8?=
 =?us-ascii?Q?Od+6u8jVQ8BKANXDilpJk/hQ3pFOlj4emyuWSgljrPr9OmGtpit1Cg1lY8k+?=
 =?us-ascii?Q?ReO5QxoHIyJQcsZpgQkiq+weayY2rWI5xkthAxhkPhBQX4B2GdCNxSeaG5Ob?=
 =?us-ascii?Q?fb/8/bcluNeeDTCDVh+2g9BHsa4JkV9woCA0e50qieT1eh3o7yGFH8rll0mW?=
 =?us-ascii?Q?HZSEvLjl0sN+dmTRQ04Ysg1yPcnMDmUwmKq787J29ZKQ7DXwgqfQT2xkQwaG?=
 =?us-ascii?Q?aKyoRYT7/XrxKij9ycLnh+jMVgHT8Ode/+bxD0DkTNDpKzsy6t3jA6A5OnvS?=
 =?us-ascii?Q?dIEDOg6ATxWaYFep59LsDLJAnN1b6oy8kqY8PFoYMYK7cG0wMROEMcveoyaV?=
 =?us-ascii?Q?5oLC/Onlsk0/H+PJY7MsMXSClvzWsMUXYuxZu2rby9793bbh7ifrBOvWUw44?=
 =?us-ascii?Q?F5JcH6lUPDXKUYnSDadVwrcxgoYWWRxMO3we3SjFNyrh7zTuUjs7wiet6fGQ?=
 =?us-ascii?Q?5qfKusmnl1fX1mphIoKeizZxTGxzv7cSrhLdXb50e1Oa366IHelHKmQRBXqH?=
 =?us-ascii?Q?EDaeU3NujTyKV5ufwTqPj/6aR6XmdQOQAL/EcQ+dd8COa7Z8i4WTeUhoTv8z?=
 =?us-ascii?Q?uvC/wJkc12mIN6oYOKlFSQgyP0RcQK9+sqVALzrhKTnK68c2TvGJYApwARbr?=
 =?us-ascii?Q?gkvlFoMr0xwd3JW9pp3MprPZ8Dgh2M8lObkPw7QZMpeyQXQj9TPv9LU4i3I1?=
 =?us-ascii?Q?CA/26+DDYVD4jWMp4UbbxbVi1r99k9yegd4CeQLS4wDpJEic9u/WLNDiT9F4?=
 =?us-ascii?Q?Y6gALjdWqlcI3RGBN621WWg9rqQzOdUbydrb8E2unonLtgjsq7lO59aJKv0g?=
 =?us-ascii?Q?j9Z8T9I160m2VBnlwz7KqSx8G3IZnCL9Hzgf1pxd?=
X-OriginatorOrg: vivo.com
X-MS-Exchange-CrossTenant-Network-Message-Id: e47af366-76b8-4c1d-602e-08dbb67a0120
X-MS-Exchange-CrossTenant-AuthSource: TY2PR06MB3342.apcprd06.prod.outlook.com
X-MS-Exchange-CrossTenant-AuthAs: Internal
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 16 Sep 2023 05:58:52.6948
 (UTC)
X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted
X-MS-Exchange-CrossTenant-Id: 923e42dc-48d5-4cbe-b582-1a797a6412ed
X-MS-Exchange-CrossTenant-MailboxType: HOSTED
X-MS-Exchange-CrossTenant-UserPrincipalName: 4n8aSjGY5OZHHIcmWGuAmxPW/FfYqPX8gXcY4l4JizId39cU7XkDPrbY+YoOCbKq8O07/SxDjJm4GR0hsRVmNQ==
X-MS-Exchange-Transport-CrossTenantHeadersStamped: TYZPR06MB6308
X-Spam-Status: No, score=-0.8 required=5.0 tests=DKIM_SIGNED,DKIM_VALID,
	DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS,MAILING_LIST_MULTI,
	SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED autolearn=unavailable
	autolearn_force=no version=3.4.6
X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on morse.vger.email
Precedence: bulk
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org
X-Greylist: Sender passed SPF test, not delayed by milter-greylist-4.6.4 (morse.vger.email [0.0.0.0]); Fri, 15 Sep 2023 23:03:37 -0700 (PDT)

When writing back an inode and performing an fsync on it concurrently, a
deadlock issue may arise as shown below. In each writeback iteration, a
clean inode is requeued to the wb->b_dirty queue due to non-zero
pages_skipped, without anything actually being written. This causes an
infinite loop and prevents the plug from being flushed, resulting in a
deadlock. We now avoid requeuing the clean inode to prevent this issue.

    wb_writeback        fsync (inode-Y)
blk_start_plug(&plug)
for (;;) {
  iter i-1: some reqs with page-X added into plug->mq_list // f2fs node page-X with PG_writeback
                        filemap_fdatawrite
                          __filemap_fdatawrite_range // write inode-Y with sync_mode WB_SYNC_ALL
                           do_writepages
                            f2fs_write_data_pages
                             __f2fs_write_data_pages // wb_sync_req[DATA]++ for WB_SYNC_ALL
                              f2fs_write_cache_pages
                               f2fs_write_single_data_page
                                f2fs_do_write_data_page
                                 f2fs_outplace_write_data
                                  f2fs_update_data_blkaddr
                                   f2fs_wait_on_page_writeback
                                     wait_on_page_writeback // wait for f2fs node page-X
  iter i:
    progress = __writeback_inodes_wb(wb, work)
    . writeback_sb_inodes
    .   __writeback_single_inode // write inode-Y with sync_mode WB_SYNC_NONE
    .   . do_writepages
    .   .   f2fs_write_data_pages
    .   .   .  __f2fs_write_data_pages // skip writepages due to (wb_sync_req[DATA]>0)
    .   .   .   wbc->pages_skipped += get_dirty_pages(inode) // wbc->pages_skipped = 1
    .   if (!(inode->i_state & I_DIRTY_ALL)) // i_state = I_SYNC | I_SYNC_QUEUED
    .    total_wrote++;  // total_wrote = 1
    .   requeue_inode // requeue inode-Y to wb->b_dirty queue due to non-zero pages_skipped
    if (progress) // progress = 1
      continue;
  iter i+1:
      queue_io
      // similar process with iter i, infinite for-loop !
}
blk_finish_plug(&plug)   // flush plug won't be called

Signed-off-by: Chunhai Guo <guochunhai@vivo.com>
---
 fs/fs-writeback.c | 11 ++++++++---
 1 file changed, 8 insertions(+), 3 deletions(-)

diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
index 969ce991b0b0..c1af01b2c42d 100644
--- a/fs/fs-writeback.c
+++ b/fs/fs-writeback.c
@@ -1535,10 +1535,15 @@ static void requeue_inode(struct inode *inode, struct bdi_writeback *wb,
 
 	if (wbc->pages_skipped) {
 		/*
-		 * writeback is not making progress due to locked
-		 * buffers. Skip this inode for now.
+		 * Writeback is not making progress due to locked buffers.
+		 * Skip this inode for now. Although having skipped pages
+		 * is odd for clean inodes, it can happen for some
+		 * filesystems so handle that gracefully.
 		 */
-		redirty_tail_locked(inode, wb);
+		if (inode->i_state & I_DIRTY_ALL)
+			redirty_tail_locked(inode, wb);
+		else
+			inode_cgwb_move_to_attached(inode, wb);
 		return;
 	}
 
-- 
2.25.1