WSL2-Linux-Kernel/fs
Jan Kara c4a391b53a writeback: do not sync data dirtied after sync start
When there are processes heavily creating small files while sync(2) is
running, it can easily happen that quite some new files are created
between WB_SYNC_NONE and WB_SYNC_ALL pass of sync(2).  That can happen
especially if there are several busy filesystems (remember that sync
traverses filesystems sequentially and waits in WB_SYNC_ALL phase on one
fs before starting it on another fs).  Because WB_SYNC_ALL pass is slow
(e.g.  causes a transaction commit and cache flush for each inode in
ext3), resulting sync(2) times are rather large.

The following script reproduces the problem:

  function run_writers
  {
    for (( i = 0; i < 10; i++ )); do
      mkdir $1/dir$i
      for (( j = 0; j < 40000; j++ )); do
        dd if=/dev/zero of=$1/dir$i/$j bs=4k count=4 &>/dev/null
      done &
    done
  }

  for dir in "$@"; do
    run_writers $dir
  done

  sleep 40
  time sync

Fix the problem by disregarding inodes dirtied after sync(2) was called
in the WB_SYNC_ALL pass.  To allow for this, sync_inodes_sb() now takes
a time stamp when sync has started which is used for setting up work for
flusher threads.

To give some numbers, when above script is run on two ext4 filesystems
on simple SATA drive, the average sync time from 10 runs is 267.549
seconds with standard deviation 104.799426.  With the patched kernel,
the average sync time from 10 runs is 2.995 seconds with standard
deviation 0.096.

Signed-off-by: Jan Kara <jack@suse.cz>
Reviewed-by: Fengguang Wu <fengguang.wu@intel.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-11-13 12:09:07 +09:00
..
9p FS-Cache: Provide the ability to enable/disable cookies 2013-09-27 18:40:25 +01:00
adfs truncate: drop 'oldsize' truncate_pagecache() parameter 2013-09-12 15:38:02 -07:00
affs truncate: drop 'oldsize' truncate_pagecache() parameter 2013-09-12 15:38:02 -07:00
afs Merge branch 'fscache' of git://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs into linux-next 2013-10-28 19:36:46 -04:00
autofs4 autofs4: close the races around autofs4_notify_daemon() 2013-09-16 19:16:38 -04:00
befs
bfs truncate: drop 'oldsize' truncate_pagecache() parameter 2013-09-12 15:38:02 -07:00
btrfs Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs 2013-10-18 16:46:21 -07:00
cachefiles FS-Cache: Provide the ability to enable/disable cookies 2013-09-27 18:40:25 +01:00
ceph FS-Cache: Provide the ability to enable/disable cookies 2013-09-27 18:40:25 +01:00
cifs Merge branch 'for-linus' of git://git.samba.org/sfrench/cifs-2.6 2013-11-08 06:01:47 +09:00
coda
configfs Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2013-07-14 11:42:26 -07:00
cramfs
debugfs debugfs: debugfs_remove_recursive() must not rely on list_empty(d_subdirs) 2013-07-31 12:16:31 -04:00
devpts
dlm dlm: remove signal blocking 2013-08-12 15:22:43 -05:00
ecryptfs eCryptfs: fix 32 bit corruption issue 2013-10-24 12:36:30 -07:00
efivarfs efivarfs: we can use simple_lookup() now 2013-07-14 17:48:35 +04:00
efs efs: iget_locked() doesn't return an ERR_PTR() 2013-08-24 12:10:22 -04:00
exofs truncate: drop 'oldsize' truncate_pagecache() parameter 2013-09-12 15:38:02 -07:00
exportfs exportfs: don't assume that ->iterate() won't feed us too long entries 2013-09-07 19:54:55 -04:00
ext2 truncate: drop 'oldsize' truncate_pagecache() parameter 2013-09-12 15:38:02 -07:00
ext3 ext[34]: fix double put in tmpfile 2013-10-15 12:14:06 -04:00
ext4 Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2013-10-16 17:18:18 -07:00
f2fs f2fs: optimize gc for better performance 2013-09-05 13:50:32 +09:00
fat truncate: drop 'oldsize' truncate_pagecache() parameter 2013-09-12 15:38:02 -07:00
freevxfs
fscache FS-Cache: Provide the ability to enable/disable cookies 2013-09-27 18:40:25 +01:00
fuse fuse: no RCU mode in fuse_access() 2013-10-01 16:41:23 +02:00
gfs2 The main feature of interest this time is quota updates. There are 2013-11-11 07:11:00 +09:00
hfs truncate: drop 'oldsize' truncate_pagecache() parameter 2013-09-12 15:38:02 -07:00
hfsplus truncate: drop 'oldsize' truncate_pagecache() parameter 2013-09-12 15:38:02 -07:00
hostfs um: hostfs: Fix writeback 2013-09-07 10:38:29 +02:00
hpfs truncate: drop 'oldsize' truncate_pagecache() parameter 2013-09-12 15:38:02 -07:00
hppfs
hugetlbfs cope with potentially long ->d_dname() output for shmem/hugetlb 2013-08-24 12:10:17 -04:00
isofs isofs: Refuse RW mount of the filesystem instead of making it RO 2013-07-31 22:14:50 +02:00
jbd jbd: use a single printk for jbd_debug() 2013-08-09 10:49:00 +02:00
jbd2 jbd2: Fix endian mixing problems in the checksumming code 2013-08-28 14:59:58 -04:00
jffs2
jfs Just a patch to fix an oops in an error path. 2013-10-22 09:01:11 +01:00
lockd LOCKD: Don't call utsname()->nodename from nlmclnt_setlockargs 2013-08-05 15:03:46 -04:00
logfs
minix fs/minix: Drop dependency on H8300 2013-09-16 18:20:25 -07:00
ncpfs
nfs NFSv4.2: Remove redundant checks in nfs_setsecurity+nfs4_label_init_security 2013-11-04 16:42:52 -05:00
nfs_common
nfsd Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs 2013-09-12 15:01:38 -07:00
nilfs2 nilfs2: fix issue with race condition of competition between segments for dirty blocks 2013-09-30 14:31:02 -07:00
nls
notify
ntfs truncate: drop 'oldsize' truncate_pagecache() parameter 2013-09-12 15:38:02 -07:00
ocfs2 ocfs2: simplify ocfs2_invalidatepage() and ocfs2_releasepage() 2013-11-13 12:09:02 +09:00
omfs truncate: drop 'oldsize' truncate_pagecache() parameter 2013-09-12 15:38:02 -07:00
openpromfs
proc /proc/pid/smaps: show VM_SOFTDIRTY flag in VmFlags line 2013-11-13 12:09:07 +09:00
pstore pstore: Remove the messages related to compression failure 2013-09-16 09:28:29 -07:00
qnx4
qnx6
quota fs: convert fs shrinkers to new scan/count API 2013-09-10 18:56:31 -04:00
ramfs initmpfs: move rootfs code from fs/ramfs/ to init/ 2013-09-11 15:59:37 -07:00
reiserfs reiserfs: fix race with flush_used_journal_lists and flush_journal_list 2013-09-24 11:24:21 +02:00
romfs
squashfs Squashfs: add corruption check for type in squashfs_readdir() 2013-09-06 04:57:54 +01:00
sysfs Revert "sysfs: drop kobj_ns_type handling" 2013-11-07 20:47:28 +09:00
sysv sysv: Add forgotten superblock lock init for v7 fs 2013-09-29 22:02:02 -04:00
ubifs Just one patch which fixes the power-cut recovery testing mode. 2013-09-16 15:36:55 -04:00
udf udf: Fortify LVID loading 2013-09-24 11:23:33 +02:00
ufs truncate: drop 'oldsize' truncate_pagecache() parameter 2013-09-12 15:38:02 -07:00
xfs writeback: do not sync data dirtied after sync start 2013-11-13 12:09:07 +09:00
Kconfig
Kconfig.binfmt
Makefile
aio.c aio: fix use-after-free in aio_migratepage 2013-09-26 20:34:51 -04:00
anon_inodes.c fs/anon_inode: Introduce a new lib function anon_inode_getfile_private() 2013-07-16 09:32:17 -04:00
attr.c
bad_inode.c
binfmt_aout.c mm: remove free_area_cache 2013-07-10 18:11:34 -07:00
binfmt_elf.c fs/binfmt_elf.c: prevent a coredump with a large vm_map_count from Oopsing 2013-09-30 14:31:01 -07:00
binfmt_elf_fdpic.c
binfmt_em86.c
binfmt_flat.c
binfmt_misc.c
binfmt_script.c
binfmt_som.c
bio-integrity.c Merge branch 'for-3.12/core' of git://git.kernel.dk/linux-block 2013-09-22 15:00:11 -07:00
bio.c block: Fix bio_copy_data() 2013-09-24 14:41:42 -07:00
block_dev.c a trivial writeback fix 2013-09-13 23:06:40 -04:00
buffer.c fs: buffer: move allocation failure loop into the allocator 2013-10-16 21:35:53 -07:00
char_dev.c
compat.c
compat_binfmt_elf.c
compat_ioctl.c
coredump.c coredump: add new %P variable in core_pattern 2013-09-11 15:59:01 -07:00
coredump.h
dcache.c vfs: decrapify dput(), fix cache behavior under normal load 2013-10-31 15:43:02 -07:00
dcookies.c
direct-io.c direct-io: Use return from cmpxchg to decide of assignment happened 2013-09-09 10:47:42 -07:00
drop_caches.c shrinker: add node awareness 2013-09-10 18:56:31 -04:00
eventfd.c
eventpoll.c Revert "epoll: use freezable blocking call" 2013-10-30 15:27:53 +01:00
exec.c sched/numa: Call task_numa_free() from do_execve() 2013-10-09 14:48:00 +02:00
fcntl.c vfs: add missing check for __O_TMPFILE in fcntl_init() 2013-08-05 18:25:32 +04:00
fhandle.c
file.c
file_table.c nfsd regression since delayed fput() 2013-10-20 08:44:39 -04:00
filesystems.c
fs-writeback.c writeback: do not sync data dirtied after sync start 2013-11-13 12:09:07 +09:00
fs_struct.c
generic_acl.c
inode.c fs: convert inode and dentry shrinking to be node aware 2013-09-10 18:56:31 -04:00
internal.h fs: convert inode and dentry shrinking to be node aware 2013-09-10 18:56:31 -04:00
ioctl.c
ioprio.c
libfs.c make simple_lookup() usable for filesystems that set ->s_d_op 2013-07-14 17:43:25 +04:00
locks.c
mbcache.c fs: convert fs shrinkers to new scan/count API 2013-09-10 18:56:31 -04:00
mount.h
mpage.c
namei.c fs/namei.c: fix new kernel-doc warning 2013-10-22 12:02:40 +01:00
namespace.c initmpfs: move rootfs code from fs/ramfs/ to init/ 2013-09-11 15:59:37 -07:00
no-block.c
open.c vfs: improve i_op->atomic_open() documentation 2013-09-16 19:17:24 -04:00
pipe.c
pnode.c
pnode.h vfs: Don't copy mount bind mounts of /proc/<pid>/ns/mnt between namespaces 2013-08-26 18:42:15 -07:00
posix_acl.c
proc_namespace.c
read_write.c aio: Kill aio_rw_vect_retry() 2013-07-30 11:53:12 -04:00
readdir.c
select.c Revert "select: use freezable blocking call" 2013-10-30 15:28:35 +01:00
seq_file.c seq_file: always update file->f_pos in seq_lseek() 2013-10-25 10:46:40 -04:00
signalfd.c
splice.c
stack.c
stat.c quota: provide interface for readding allocated space into reserved space 2013-08-17 09:32:32 -04:00
statfs.c vfs: allow O_PATH file descriptors for fstatfs() 2013-10-12 13:12:31 -07:00
super.c fs/super.c: fix lru_list leak for real 2013-10-01 13:11:21 -04:00
sync.c writeback: do not sync data dirtied after sync start 2013-11-13 12:09:07 +09:00
timerfd.c
utimes.c
xattr.c
xattr_acl.c