Received-SPF: pass (google.com: domain of linux-kernel+bounces-180609-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.48.161 as permitted sender) client-ip=147.75.48.161;
From: Yang Yang <yang.yang@vivo.com>
To: Alasdair Kergon <agk@redhat.com>,
	Mike Snitzer <snitzer@kernel.org>,
	Mikulas Patocka <mpatocka@redhat.com>,
	dm-devel@lists.linux.dev,
	linux-kernel@vger.kernel.org
Cc: Yang Yang <yang.yang@vivo.com>
Subject: [PATCH v3 0/5] dm: empty flush optimization
Date: Thu, 16 May 2024 12:02:29 +0800
Message-Id: <20240516040235.115651-1-yang.yang@vivo.com>
Content-Transfer-Encoding: 8bit
Content-Type: text/plain
Precedence: bulk
MIME-Version: 1.0
X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1
X-MS-Exchange-AntiSpam-MessageData-0:
	=?us-ascii?Q?mxohw8KwVCkXE9YQzDsE2BX7vl6NPXqjoj4UpZ0JThUg0Zc1uItnEY/2nXiR?=
 =?us-ascii?Q?ZwiQ+vvwARuF/md8xHl+3pHHieZxoEJ4aQqPnP21yxr8or/UbrpXyLlFS9RG?=
 =?us-ascii?Q?A4ghNPJw2+/myu5kyRU9/yRFP9AAvMPJpIuiiZcsR+fjMDzauKu8nFHL9+S6?=
 =?us-ascii?Q?dmTI/umfnBjZvnBrGIimYKdnyYaH4dQml3AwRPxHdtF4H2xrFtBNtEq4BTda?=
 =?us-ascii?Q?+wm6VqpRsPb2G0dz95qqOVK1R9AaIRWQbz2kBUAM15upB8j9whgYhxVu99AU?=
 =?us-ascii?Q?01uHFqZgaw7MJcxTRu6MWxMOvovyE8Cc8e+CIRluLY4CCdIbjrl3Krjguu9G?=
 =?us-ascii?Q?u/cPyuMr3k4mjFI/QqXvvi1QoAMIRi//R1XVATYDoQEeUhavLVSSt0MNoVMw?=
 =?us-ascii?Q?BPJ8Lc7VRQRAzb0fkz/iCYQWH12l0tL+DoDpXHneVy2xtp6lfGmcAD3pLqJB?=
 =?us-ascii?Q?Z/Pt9Tkora6/goltqXZ8jpbNNj0QrSZ9YExc2E76YxYvDoHDIJlaA0V1HRhD?=
 =?us-ascii?Q?RgO/uaySsPFtb4ZsY0gHnTeVznkBfwQWlp1MIcfrijjHV2/zB9UMA95BIXUL?=
 =?us-ascii?Q?EmnLmuHzVJiq/2FOYZLlCGoJFXbyoGuGwESXCogTDQYSZgwO0AHSvlXegpP5?=
 =?us-ascii?Q?knWhN39x5Sunvi71YCFIgjvuks0wifcB2yj1CwZqdOD4fd9AHVPdOXQssT/J?=
 =?us-ascii?Q?lgpQ3g6e2aag7dvoXAS6AXU0XkvtMCK2PKmdvy357Dir/QnapX+UoxAzGN+n?=
 =?us-ascii?Q?buooontc8sq7xu2Hsmg9VhQ+WerKAGH29lfBYqaOZjcmibT2UCvmFEk94+/1?=
 =?us-ascii?Q?rfQf9+zpH2lBAomVleYolB+uMdLaFqSaWD7vmvvG8jCBCHa8MUsea0gSR4wi?=
 =?us-ascii?Q?m2LfDogBTyUMotS/Au4oGkVFa11jh9xTp11pJKyD5rmc89NuBJQSRKOcb0wT?=
 =?us-ascii?Q?3I4taLL7+Ot/yvN72pNceQSs+u5FBiihEj4rQzZ9OU9S9huGRp7mk3qCUG4w?=
 =?us-ascii?Q?MLaiukJiaMKXxeCFr5loeQDJJKs4/4boSIioYtXWvwivqqXpxbiWtz+VEKVE?=
 =?us-ascii?Q?YMiiU2hm3mEbqbwIsbhRDCNXV9yHwLjmEOo3ZQoZLFGUE1KkxBz/eubEFNd1?=
 =?us-ascii?Q?QvkakTKbhjBsJQo8gUk124dZXwHL5SGBfk5Y+qqIncrXDkjVGCZF0VLYIv1F?=
 =?us-ascii?Q?SwwYRmQgFdDWQnQGPV0TVAKRBFGkt0gWE+2l2/0lC+hl9ZQhmD7fK9ws4WJ+?=
 =?us-ascii?Q?LVo2zfWpHOArJ+Eh1XlQOUNje6QmEkgUCO7jz18ThDE79tfrB3Z+XGdPrn6w?=
 =?us-ascii?Q?PvS1X9bBJgjrHUgBooGTAVqFnDpoukUpK9ZbnXRJtRlO2tIxu1zVbzPreda+?=
 =?us-ascii?Q?pIBlsSLXV0tUF/bURyvLp7hlcC74UJpOGaaSbgP5Kbss51s1/nq0+TEC9rX+?=
 =?us-ascii?Q?I1BvYl1nhF81WFOuKFE4UMWlilu3tTOWrIrmCWXUiSTzvjXESOeZx/7wd4Dz?=
 =?us-ascii?Q?iZNjTcK0qgdE2UuEhDQ1M2dfxKC9PDDJCYluRW+TZ/h0IFWNjgK4LEJFr5Cn?=
 =?us-ascii?Q?tJXJYc4/z0Qf2QmTx7ZbdLXWB+LHJn0J6IQfpOf4?=
X-OriginatorOrg: vivo.com
X-MS-Exchange-CrossTenant-Network-Message-Id: 1fbae2fc-60d8-410b-c005-08dc755d1103
X-MS-Exchange-CrossTenant-AuthSource: TYSPR06MB6411.apcprd06.prod.outlook.com
X-MS-Exchange-CrossTenant-AuthAs: Internal
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 16 May 2024 04:02:55.9240
 (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: bMPuntpMXOEpsuoHncs1KT0rYuFKwI2ZKKMrZrTk2sZIXIJc4XE1cIkn8LQQLyuAeQvgfh1KMjBE/tiU3+1Isw==
X-MS-Exchange-Transport-CrossTenantHeadersStamped: TYZPR06MB6807

__send_empty_flush() sends empty flush bios to every target in the
dm_table. However, if the num_targets exceeds the number of block
devices in the dm_table's device list, it could lead to multiple
invocations of __send_duplicate_bios() for the same block device.
Typically, a single thread sending numerous empty flush bios to one
block device is redundant, as these bios are likely to be merged by the
flush state machine. In scenarios where num_targets significantly
outweighs the number of block devices, such behavior may result in a
noteworthy decrease in performance.

This is a real-world scenario that we have encountered:
1) Call fallocate(file_fd, 0, 0, SZ_8G)
2) Call ioctl(file_fd, FS_IOC_FIEMAP, fiemap). In situations of severe
file system fragmentation, fiemap->fm_mapped_extents may exceed 1000.
3) Create a dm-linear device based on fiemap->fm_extents
4) Create a snapshot-cow device based on the dm-linear device 

Perf diff of fio test:
  fio --group_reporting --name=benchmark --filename=/dev/mapper/example \
      --ioengine=sync --invalidate=1 --numjobs=16 --rw=randwrite \
      --blocksize=4k --size=2G --time_based --runtime=30 --fdatasync=1

Scenario one:
  for i in {0..1023}; do
    echo $((8000*$i)) 8000 linear /dev/sda2 $((16384*$i))
  done | sudo dmsetup create example

  Before: bw=857KiB/
  After:  bw=30.8MiB/s    +3580%

Scenario two:
  for i in {0..1023}; do
    if [[ $i -gt 511 ]]; then
      echo $((8000*$i)) 8000 linear /dev/nvme0n1p6 $((16384*$i))
    else
      echo $((8000*$i)) 8000 linear /dev/sda2 $((16384*$i))
    fi
  done | sudo dmsetup create example

  Before: bw=1470KiB/
  After:  bw=33.9MiB/s    +2261%

Any comments are welcome!

V3:
-- Focus on targets with num_flush_bios equal to 1 to simplify the code
-- Use t->devices_lock to protect the dm_table's device list

V2:
-- Split into smaller pieces that are easier to review
-- Add flush_pass_around, suggested by Mikulas Patocka
-- Handling different target types separately

Yang Yang (5):
  dm: introduce flush_pass_around flag
  dm: add __send_empty_flush_bios() helper
  dm: support retrieving struct dm_target from struct dm_dev
  dm: Avoid sending redundant empty flush bios to the same block device
  dm linear: enable flush optimization function

 drivers/md/dm-core.h          |  3 +++
 drivers/md/dm-ioctl.c         |  4 ++++
 drivers/md/dm-linear.c        |  1 +
 drivers/md/dm-table.c         | 19 +++++++++++++++++++
 drivers/md/dm.c               | 34 +++++++++++++++++++++++++---------
 include/linux/device-mapper.h |  6 ++++++
 6 files changed, 58 insertions(+), 9 deletions(-)

-- 
2.34.1