linux.git/fs/f2fs/data.c, branch v4.1

Revert "f2fs: enhance multi-threads performance"

2015-05-04T21:15:15+00:00

This reports performance regression by Yuanhan Liu.
The basic idea was to reduce one-point mutex, but it turns out this causes
another contention like context swithes.

https://lkml.org/lkml/2015/4/21/11

Until finishing the analysis on this issue, I'd like to revert this for a while.

This reverts commit 78373b7319abdf15050af5b1632c4c8b8b398f33.

Merge tag 'for-f2fs-4.1' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs

2015-04-18T15:17:20+00:00

Pull f2fs updates from Jaegeuk Kim:
 "New features:
   - in-memory extent_cache
   - fs_shutdown to test power-off-recovery
   - use inline_data to store symlink path
   - show f2fs as a non-misc filesystem

  Major fixes:
   - avoid CPU stalls on sync_dirty_dir_inodes
   - fix some power-off-recovery procedure
   - fix handling of broken symlink correctly
   - fix missing dot and dotdot made by sudden power cuts
   - handle wrong data index during roll-forward recovery
   - preallocate data blocks for direct_io

  ... and a bunch of minor bug fixes and cleanups"

* tag 'for-f2fs-4.1' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs: (71 commits)
  f2fs: pass checkpoint reason on roll-forward recovery
  f2fs: avoid abnormal behavior on broken symlink
  f2fs: flush symlink path to avoid broken symlink after POR
  f2fs: change 0 to false for bool type
  f2fs: do not recover wrong data index
  f2fs: do not increase link count during recovery
  f2fs: assign parent's i_mode for empty dir
  f2fs: add F2FS_INLINE_DOTS to recover missing dot dentries
  f2fs: fix mismatching lock and unlock pages for roll-forward recovery
  f2fs: fix sparse warnings
  f2fs: limit b_size of mapped bh in f2fs_map_bh
  f2fs: persist system.advise into on-disk inode
  f2fs: avoid NULL pointer dereference in f2fs_xattr_advise_get
  f2fs: preallocate fallocated blocks for direct IO
  f2fs: enable inline data by default
  f2fs: preserve extent info for extent cache
  f2fs: initialize extent tree with on-disk extent info of inode
  f2fs: introduce __{find,grab}_extent_tree
  f2fs: split set_data_blkaddr from f2fs_update_extent_cache
  f2fs: enable fast symlink by utilizing inline data
  ...

direct_IO: remove rw from a_ops->direct_IO()

2015-04-12T02:29:45+00:00

Now that no one is using rw, remove it completely.

Signed-off-by: Omar Sandoval 
Signed-off-by: Al Viro

direct_IO: use iov_iter_rw() instead of rw everywhere

2015-04-12T02:29:45+00:00

The rw parameter to direct_IO is redundant with iov_iter->type, and
treated slightly differently just about everywhere it's used: some users
do rw & WRITE, and others do rw == WRITE where they should be doing a
bitwise check. Simplify this with the new iov_iter_rw() helper, which
always returns either READ or WRITE.

Signed-off-by: Omar Sandoval 
Signed-off-by: Al Viro

Remove rw from {,__,do_}blockdev_direct_IO()

2015-04-12T02:29:44+00:00

Most filesystems call through to these at some point, so we'll start
here.

Signed-off-by: Omar Sandoval 
Signed-off-by: Al Viro

f2fs: limit b_size of mapped bh in f2fs_map_bh

2015-04-10T22:08:55+00:00

Map bh over max size which caller defined is not needed, limit it in
f2fs_map_bh.

Signed-off-by: Chao Yu 
Signed-off-by: Jaegeuk Kim

f2fs: preallocate fallocated blocks for direct IO

2015-04-10T22:08:52+00:00

Normally, due to DIO_SKIP_HOLES flag is set by default, blockdev_direct_IO in
f2fs_direct_IO tries to skip DIO in holes when writing inside i_size, this
makes us falling back to buffered IO which shows lower performance.

So in commit 59b802e5a453 ("f2fs: allocate data blocks in advance for
f2fs_direct_IO"), we improve perfromance by allocating data blocks in advance
if we meet holes no matter in i_size or not, since with it we can avoid falling
back to buffered IO.

But we forget to consider for unwritten fallocated block in this commit.
This patch tries to fix it for fallocate case, this helps to improve
performance.

Test result:
Storage info: sandisk ultra 64G micro sd card.

touch /mnt/f2fs/file
truncate -s 67108864 /mnt/f2fs/file
fallocate -o 0 -l 67108864 /mnt/f2fs/file
time dd if=/dev/zero of=/mnt/f2fs/file bs=1M count=64 conv=notrunc oflag=direct

Time before applying the patch:
67108864 bytes (67 MB) copied, 36.16 s, 1.9 MB/s
real    0m36.162s
user    0m0.000s
sys     0m0.180s

Time after applying the patch:
67108864 bytes (67 MB) copied, 27.7776 s, 2.4 MB/s
real    0m27.780s
user    0m0.000s
sys     0m0.036s

Signed-off-by: Chao Yu 
Signed-off-by: Jaegeuk Kim

f2fs: preserve extent info for extent cache

2015-04-10T22:08:51+00:00

This patch tries to preserve last extent info in extent tree cache into on-disk
inode, so this can help us to reuse the last extent info next time for
performance.

Signed-off-by: Chao Yu 
Signed-off-by: Jaegeuk Kim

f2fs: initialize extent tree with on-disk extent info of inode

2015-04-10T22:08:50+00:00

With normal extent info cache, we records largest extent mapping between logical
block and physical block into extent info, and we persist extent info in on-disk
inode.

When we enable extent tree cache, if extent info of on-disk inode is exist, and
the extent is not a small fragmented mapping extent. We'd better to load the
extent info into extent tree cache when inode is loaded. By this way we can have
more chance to hit extent tree cache rather than taking more time to read dnode
page for block address.

Signed-off-by: Chao Yu 
Signed-off-by: Jaegeuk Kim

f2fs: introduce __{find,grab}_extent_tree

2015-04-10T22:08:49+00:00

This patch introduces __{find,grab}_extent_tree for reusing by following
patches.

Signed-off-by: Chao Yu 
Signed-off-by: Jaegeuk Kim