linux/fs
Sandeep Dhavale 3fffb589b9 erofs: add per-cpu threads for decompression as an option
Using per-cpu thread pool we can reduce the scheduling latency compared
to workqueue implementation. With this patch scheduling latency and
variation is reduced as per-cpu threads are high priority kthread_workers.

The results were evaluated on arm64 Android devices running 5.10 kernel.

The table below shows resulting improvements of total scheduling latency
for the same app launch benchmark runs with 50 iterations. Scheduling
latency is the latency between when the task (workqueue kworker vs
kthread_worker) became eligible to run to when it actually started
running.
+-------------------------+-----------+----------------+---------+
|                         | workqueue | kthread_worker |  diff   |
+-------------------------+-----------+----------------+---------+
| Average (us)            |     15253 |           2914 | -80.89% |
| Median (us)             |     14001 |           2912 | -79.20% |
| Minimum (us)            |      3117 |           1027 | -67.05% |
| Maximum (us)            |     30170 |           3805 | -87.39% |
| Standard deviation (us) |      7166 |            359 |         |
+-------------------------+-----------+----------------+---------+

Background: Boot times and cold app launch benchmarks are very
important to the Android ecosystem as they directly translate to
responsiveness from user point of view. While EROFS provides
a lot of important features like space savings, we saw some
performance penalty in cold app launch benchmarks in few scenarios.
Analysis showed that the significant variance was coming from the
scheduling cost while decompression cost was more or less the same.

Having per-cpu thread pool we can see from the above table that this
variation is reduced by ~80% on average. This problem was discussed
at LPC 2022. Link to LPC 2022 slides and talk at [1]

[1] https://lpc.events/event/16/contributions/1338/

[ Gao Xiang: At least, we have to add this until WQ_UNBOUND workqueue
             issue [2] on many arm64 devices is resolved. ]
[2] https://lore.kernel.org/r/CAJkfWY490-m6wNubkxiTPsW59sfsQs37Wey279LmiRxKt7aQYg@mail.gmail.com

Signed-off-by: Sandeep Dhavale <dhavale@google.com>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Link: https://lore.kernel.org/r/20230208093322.75816-1-hsiangkao@linux.alibaba.com
2023-02-15 08:11:26 +08:00
..
9p 9p-for-6.2-rc1 2022-12-23 11:39:18 -08:00
adfs fs: Convert block_read_full_page() to block_read_full_folio() 2022-05-09 16:21:44 -04:00
affs affs: initialize fsdata in affs_truncate() 2023-01-10 14:55:20 +01:00
afs rxrpc: Move call state changes from recvmsg to I/O thread 2023-01-06 09:43:33 +00:00
autofs autofs: remove unused ino field inode 2022-07-17 17:31:42 -07:00
befs befs: Convert befs_symlink_read_folio() to use a folio 2022-08-02 12:34:03 -04:00
bfs fs: Convert block_read_full_page() to block_read_full_folio() 2022-05-09 16:21:44 -04:00
btrfs for-6.2-rc4-tag 2023-01-20 11:59:01 -08:00
cachefiles fscache,cachefiles: add prepare_ondemand_read() callback 2022-12-07 10:56:29 +08:00
ceph ceph: avoid use-after-free in ceph_fl_release_lock() 2023-01-02 12:27:25 +01:00
cifs cifs: remove unused function 2023-01-18 14:49:51 -06:00
coda coda: Convert coda_symlink_filler() to use a folio 2022-08-02 12:34:03 -04:00
configfs configfs: fix possible memory leak in configfs_create_dir() 2022-12-02 11:11:22 +01:00
cramfs cramfs: read_mapping_page() is synchronous 2022-08-02 12:34:02 -04:00
crypto for-6.2/block-2022-12-08 2022-12-13 10:43:59 -08:00
debugfs debugfs: fix error when writing negative value to atomic_t debugfs file 2022-11-30 16:13:16 -08:00
devpts
dlm Treewide: Stop corrupting socket's task_frag 2022-12-19 17:28:49 -08:00
ecryptfs ecryptfs: use stub posix acl handlers 2022-10-20 10:13:31 +02:00
efivarfs efi: vars: prohibit reading random seed variables 2022-12-01 09:51:21 +01:00
efs efs: Convert efs symlinks to read_folio 2022-05-09 16:21:45 -04:00
erofs erofs: add per-cpu threads for decompression as an option 2023-02-15 08:11:26 +08:00
exfat Description for this pull request: 2022-12-15 18:14:21 -08:00
exportfs exportfs: use pr_debug for unreachable debug statements 2022-11-28 12:54:45 -05:00
ext2 \n 2022-12-12 20:32:50 -08:00
ext4 ext4: deal with legacy signed xattr name hash values 2023-01-21 10:14:47 -08:00
f2fs f2fs: let's avoid panic if extent_tree is not created 2023-01-03 08:59:06 -08:00
fat MM patches for 6.2-rc1. 2022-12-13 19:29:45 -08:00
freevxfs freevxfs: Convert vxfs_immed_read_folio() to use a folio 2022-08-02 12:34:03 -04:00
fscache iov_iter work; most of that is about getting rid of 2022-12-12 18:29:54 -08:00
fuse MM patches for 6.2-rc1. 2022-12-13 19:29:45 -08:00
gfs2 gfs2 fixes 2022-12-17 08:18:04 -06:00
hfs hfs/hfsplus: avoid WARN_ON() for sanity check, use proper error handling 2023-01-06 14:09:13 -08:00
hfsplus MM patches for 6.2-rc1. 2022-12-13 19:29:45 -08:00
hostfs hostfs: move from strlcpy with unused retval to strscpy 2022-09-19 22:46:25 +02:00
hpfs hpfs: remove ->writepage 2022-12-11 18:12:18 -08:00
hugetlbfs hugetlbfs: inode: remove unnecessary (void*) conversions 2022-11-30 15:58:56 -08:00
iomap New XFS code for 6.2: 2022-12-14 10:11:51 -08:00
isofs - hfs and hfsplus kmap API modernization from Fabio Francesco 2022-10-12 11:00:22 -07:00
jbd2 jbd2: switch jbd2_submit_inode_data() to use fs-provided hook for data writeout 2022-12-08 21:49:25 -05:00
jffs2 fs: rename current get acl method 2022-10-20 10:13:27 +02:00
jfs MM patches for 6.2-rc1. 2022-12-13 19:29:45 -08:00
kernfs kernfs: fix all kernel-doc warnings and multiple typos 2022-11-23 19:28:26 +01:00
ksmbd ksmbd: fix infinite loop in ksmbd_conn_handler_loop() 2023-01-01 22:49:24 -06:00
lockd NFSD 6.2 Release Notes 2022-12-12 20:54:39 -08:00
minix vfs: open inside ->tmpfile() 2022-09-24 07:00:00 +02:00
netfs use less confusing names for iov_iter direction initializers 2022-11-25 13:01:55 -05:00
nfs NFS client fixes for Linux 6.2 2023-01-07 10:38:11 -08:00
nfs_common
nfsd nfsd-6.2 fixes: 2023-01-17 09:29:17 -08:00
nilfs2 nilfs2: fix general protection fault in nilfs_btree_insert() 2023-01-11 16:14:21 -08:00
nls
notify Merge tag 'fsnotify-for_v6.1-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs 2022-10-07 08:28:50 -07:00
ntfs - hfs and hfsplus kmap API modernization from Fabio Francesco 2022-10-12 11:00:22 -07:00
ntfs3 fs/ntfs3: don't hold ni_lock when calling truncate_setsize() 2023-01-02 10:31:09 -08:00
ocfs2 Treewide: Stop corrupting socket's task_frag 2022-12-19 17:28:49 -08:00
omfs omfs: remove ->writepage 2022-12-11 18:12:18 -08:00
openpromfs fs: allocate inode by using alloc_inode_sb() 2022-03-22 15:57:03 -07:00
orangefs orangefs: four fixes from Zhang Xiaoxu and two from Colin Ian King 2022-12-14 11:16:33 -08:00
overlayfs overlayfs update for 6.2 2022-12-12 20:18:26 -08:00
proc ARM64: 2022-12-15 11:12:21 -08:00
pstore pstore updates for v6.2-rc1-fixes 2022-12-23 11:55:54 -08:00
qnx4 fs: Convert block_read_full_page() to block_read_full_folio() 2022-05-09 16:21:44 -04:00
qnx6 fs/qnx6: delete unnecessary checks before brelse() 2022-09-11 21:55:07 -07:00
quota ext4: fix bug_on in __es_tree_search caused by bad quota inode 2022-12-08 21:49:23 -05:00
ramfs tmpfile API change 2022-10-10 19:45:17 -07:00
reiserfs lsm/stable-6.2 PR 20221212 2022-12-13 09:47:48 -08:00
romfs romfs: Convert romfs to read_folio 2022-05-09 16:21:46 -04:00
smbfs_common smb3: define missing create contexts 2022-10-05 01:55:27 -05:00
squashfs fs.idmapped.squashfs.v6.2 2022-12-12 20:24:51 -08:00
sysfs kobject: kobj_type: remove default_attrs 2022-04-05 15:39:19 +02:00
sysv fs: sysv: Fix sysv_nblocks() returns wrong value 2022-12-10 14:13:37 -05:00
tracefs tracefs: Only clobber mode/uid/gid on remount if asked 2022-09-08 17:10:54 -04:00
ubifs treewide: use get_random_u32_below() instead of deprecated function 2022-11-18 02:15:15 +01:00
udf udf: initialize newblock to 0 2023-01-06 15:44:32 +01:00
ufs ufs: replace ll_rw_block() 2022-09-11 20:26:07 -07:00
unicode
vboxsf vboxsf: Convert vboxsf to read_folio 2022-05-09 16:21:46 -04:00
verity fsverity: simplify fsverity_get_digest() 2022-11-29 21:07:41 -08:00
xfs xfs: fix extent busy updating 2023-01-05 07:34:21 -08:00
zonefs zonefs: Detect append writes at invalid locations 2023-01-16 08:42:12 +09:00
aio.c use less confusing names for iov_iter direction initializers 2022-11-25 13:01:55 -05:00
anon_inodes.c dynamic_dname(): drop unused dentry argument 2022-08-20 11:34:04 -04:00
attr.c attr: use consistent sgid stripping checks 2022-10-18 10:09:47 +02:00
bad_inode.c fs: rename current get acl method 2022-10-20 10:13:27 +02:00
binfmt_elf_fdpic.c elfcore: Add a cprm parameter to elf_core_extra_{phdrs,data_size} 2023-01-05 15:12:12 +00:00
binfmt_elf_test.c
binfmt_elf.c elfcore: Add a cprm parameter to elf_core_extra_{phdrs,data_size} 2023-01-05 15:12:12 +00:00
binfmt_flat.c binfmt_flat: Remove shared library support 2022-04-22 10:57:18 -07:00
binfmt_misc.c binfmt_misc: fix shift-out-of-bounds in check_special_flags 2022-12-02 13:57:04 -08:00
binfmt_script.c
buffer.c - hfs and hfsplus kmap API modernization from Fabio Francesco 2022-10-12 11:00:22 -07:00
char_dev.c chardev: fix error handling in cdev_device_add() 2022-12-02 17:48:59 +01:00
compat_binfmt_elf.c
coredump.c hardening updates for v6.2-rc1 2022-12-14 12:20:00 -08:00
d_path.c d_path.c: typo fix... 2022-08-20 11:34:33 -04:00
dax.c fsdax,xfs: port unshare to fsdax 2022-12-11 18:12:17 -08:00
dcache.c tmpfile API change 2022-10-10 19:45:17 -07:00
direct-io.c block: remove PSI accounting from the bio layer 2022-09-20 08:24:38 -06:00
drop_caches.c
eventfd.c eventfd: provide a eventfd_signal_mask() helper 2022-11-22 06:07:55 -07:00
eventpoll.c eventpoll: add EPOLL_URING_WAKE poll wakeup flag 2022-11-21 07:45:29 -07:00
exec.c fs.vfsuid.conversion.v6.2 2022-12-12 19:20:05 -08:00
fcntl.c keep iocb_flags() result cached in struct file 2022-06-10 16:10:23 -04:00
fhandle.c do_sys_name_to_handle(): constify path 2022-09-01 17:36:39 -04:00
file_table.c locks: fix TOCTOU race when granting write lease 2022-08-16 10:59:54 -04:00
file.c fs: use acquire ordering in __fget_light() 2022-10-31 15:30:11 -04:00
filesystems.c
fs_context.c
fs_parser.c ext4: journal_path mount options should follow links 2022-12-01 10:46:54 -05:00
fs_pin.c
fs_struct.c
fs_types.c
fs-writeback.c for-6.2/writeback-2022-12-12 2022-12-15 18:09:48 -08:00
fsopen.c uninline may_mount() and don't opencode it in fspick(2)/fsopen(2) 2022-05-19 23:25:10 -04:00
init.c
inode.c fs.vfsuid.conversion.v6.2 2022-12-12 19:20:05 -08:00
internal.h fs.ovl.setgid.v6.2 2022-12-12 19:03:10 -08:00
ioctl.c Fixes for 5.18-rc1: 2022-04-01 19:35:56 -07:00
Kconfig hugetlb: make hugetlb depends on SYSFS or SYSCTL 2022-09-11 20:26:10 -07:00
Kconfig.binfmt Xtensa updates for v6.1 2022-10-10 14:21:11 -07:00
kernel_read_file.c fs/kernel_read_file: allow to read files up-to ssize_t 2022-06-16 19:58:21 -07:00
libfs.c libfs: add DEFINE_SIMPLE_ATTRIBUTE_SIGNED for signed value 2022-11-30 16:13:16 -08:00
locks.c Add process name and pid to locks warning 2022-11-30 05:08:10 -05:00
Makefile a.out: Remove the a.out implementation 2022-09-27 07:11:02 -07:00
mbcache.c ext4: fix deadlock due to mbcache entry corruption 2022-12-08 21:49:25 -05:00
mount.h switch try_to_unlazy_next() to __legitimize_mnt() 2022-07-05 16:18:21 -04:00
mpage.c Folio changes for 6.0 2022-08-03 10:35:43 -07:00
namei.c Landlock updates for v6.2-rc1 2022-12-13 09:14:50 -08:00
namespace.c fs.idmapped.mnt_idmap.v6.2 2022-12-12 19:30:18 -08:00
no-block.c
nsfs.c dynamic_dname(): drop unused dentry argument 2022-08-20 11:34:04 -04:00
open.c Landlock updates for v6.2-rc1 2022-12-13 09:14:50 -08:00
pipe.c dynamic_dname(): drop unused dentry argument 2022-08-20 11:34:04 -04:00
pnode.c pnode: terminate at peers of source 2022-12-21 14:45:25 +01:00
pnode.h
posix_acl.c fs.idmapped.mnt_idmap.v6.2 2022-12-12 19:30:18 -08:00
proc_namespace.c vfs: escape hash as well 2022-06-28 13:58:05 -04:00
read_write.c iov_iter work; most of that is about getting rid of 2022-12-12 18:29:54 -08:00
readdir.c Change calling conventions for filldir_t 2022-08-17 17:25:04 -04:00
remap_range.c New VFS code for 6.2: 2022-12-13 10:26:38 -08:00
select.c
seq_file.c use less confusing names for iov_iter direction initializers 2022-11-25 13:01:55 -05:00
signalfd.c
splice.c use less confusing names for iov_iter direction initializers 2022-11-25 13:01:55 -05:00
stack.c
stat.c fs: use type safe idmapping helpers 2022-10-26 10:02:34 +02:00
statfs.c
super.c misc pile 2022-12-12 18:38:47 -08:00
sync.c riscv: compat: syscall: Add compat_sys_call_table implementation 2022-04-26 13:36:25 -07:00
sysctls.c
timerfd.c
userfaultfd.c mm/userfaultfd: enable writenotify while userfaultfd-wp is enabled for a VMA 2023-01-11 16:14:20 -08:00
utimes.c
xattr.c fs.xattr.simple.rework.rbtree.rwlock.v6.2 2022-12-13 10:08:36 -08:00