mbox series

[f2fs-dev,v4,0/4] Speed up f2fs truncate

Message ID 20241223081044.1126291-1-yi.sun@unisoc.com (mailing list archive)
Headers show
Series Speed up f2fs truncate | expand

Message

Yi Sun Dec. 23, 2024, 8:10 a.m. UTC 
Deleting large files is time-consuming, and a large part
of the time is spent in f2fs_invalidate_blocks()
->down_write(sit_info->sentry_lock) and up_write().

If some blocks are continuous, we can process these blocks
at the same time. This can reduce the number of calls to
the down_write() and the up_write(), thereby improving the
overall speed of doing truncate.

Test steps:
Set the CPU and DDR frequencies to the maximum.
dd if=/dev/random of=./test.txt bs=1M count=100000
sync
rm test.txt

Time Comparison of rm:
original        optimization            ratio
7.17s           3.27s                   54.39%

----
v4:
- introduce update_sit_entry_for_alloc().
- [patch 2,3,4 / 4] have no changes compared to v3.

Yi Sun (4):
  f2fs: introduce update_sit_entry_for_release/alloc()
  f2fs: update_sit_entry_for_release() supports consecutive blocks.
  f2fs: add parameter @len to f2fs_invalidate_blocks()
  f2fs: Optimize f2fs_truncate_data_blocks_range()

 fs/f2fs/compress.c |   4 +-
 fs/f2fs/f2fs.h     |   3 +-
 fs/f2fs/file.c     |  78 +++++++++++++++++--
 fs/f2fs/node.c     |   4 +-
 fs/f2fs/segment.c  | 185 +++++++++++++++++++++++++++++----------------
 5 files changed, 198 insertions(+), 76 deletions(-)

Comments

patchwork-bot+f2fs--- via Linux-f2fs-devel Jan. 8, 2025, 6:40 p.m. UTC | #1 
Hello:

This series was applied to jaegeuk/f2fs.git (dev)
by Jaegeuk Kim <jaegeuk@kernel.org>:

On Mon, 23 Dec 2024 16:10:40 +0800 you wrote:
> Deleting large files is time-consuming, and a large part
> of the time is spent in f2fs_invalidate_blocks()
> ->down_write(sit_info->sentry_lock) and up_write().
> 
> If some blocks are continuous, we can process these blocks
> at the same time. This can reduce the number of calls to
> the down_write() and the up_write(), thereby improving the
> overall speed of doing truncate.
> 
> [...]

Here is the summary with links:
  - [f2fs-dev,v4,1/4] f2fs: introduce update_sit_entry_for_release/alloc()
    https://git.kernel.org/jaegeuk/f2fs/c/66baee2b886d
  - [f2fs-dev,v4,2/4] f2fs: update_sit_entry_for_release() supports consecutive blocks.
    https://git.kernel.org/jaegeuk/f2fs/c/81ffbd224e5f
  - [f2fs-dev,v4,3/4] f2fs: add parameter @len to f2fs_invalidate_blocks()
    (no matching commit)
  - [f2fs-dev,v4,4/4] f2fs: Optimize f2fs_truncate_data_blocks_range()
    (no matching commit)

You are awesome, thank you!
patchwork-bot+f2fs--- via Linux-f2fs-devel Jan. 13, 2025, 6:51 p.m. UTC | #2 
Hello:

This series was applied to jaegeuk/f2fs.git (dev)
by Jaegeuk Kim <jaegeuk@kernel.org>:

On Mon, 23 Dec 2024 16:10:40 +0800 you wrote:
> Deleting large files is time-consuming, and a large part
> of the time is spent in f2fs_invalidate_blocks()
> ->down_write(sit_info->sentry_lock) and up_write().
> 
> If some blocks are continuous, we can process these blocks
> at the same time. This can reduce the number of calls to
> the down_write() and the up_write(), thereby improving the
> overall speed of doing truncate.
> 
> [...]

Here is the summary with links:
  - [f2fs-dev,v4,1/4] f2fs: introduce update_sit_entry_for_release/alloc()
    (no matching commit)
  - [f2fs-dev,v4,2/4] f2fs: update_sit_entry_for_release() supports consecutive blocks.
    (no matching commit)
  - [f2fs-dev,v4,3/4] f2fs: add parameter @len to f2fs_invalidate_blocks()
    https://git.kernel.org/jaegeuk/f2fs/c/e53c568f4603
  - [f2fs-dev,v4,4/4] f2fs: Optimize f2fs_truncate_data_blocks_range()
    (no matching commit)

You are awesome, thank you!