diff mbox series

[f2fs-dev] f2fs: compress: fix to avoid fragment w/ OPU during f2fs_ioc_compress_file()

Message ID 20230904151242.1786219-1-chao@kernel.org (mailing list archive)
State Accepted
Commit 943f7c6f987f4a2bd081e76cf68d4c089dc66c65
Headers show
Series [f2fs-dev] f2fs: compress: fix to avoid fragment w/ OPU during f2fs_ioc_compress_file() | expand

Commit Message

Chao Yu Sept. 4, 2023, 3:12 p.m. UTC
If file has both cold and compress flag, during f2fs_ioc_compress_file(),
f2fs will trigger IPU for non-compress cluster and OPU for compress
cluster, so that, data of the file may be fragmented.

Fix it by always triggering OPU for IOs from user mode compression.

Signed-off-by: Chao Yu <chao@kernel.org>
---
 fs/f2fs/data.c | 5 +++++
 1 file changed, 5 insertions(+)

Comments

patchwork-bot+f2fs@kernel.org Sept. 20, 2023, 3:50 p.m. UTC | #1
Hello:

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

On Mon,  4 Sep 2023 23:12:42 +0800 you wrote:
> If file has both cold and compress flag, during f2fs_ioc_compress_file(),
> f2fs will trigger IPU for non-compress cluster and OPU for compress
> cluster, so that, data of the file may be fragmented.
> 
> Fix it by always triggering OPU for IOs from user mode compression.
> 
> Signed-off-by: Chao Yu <chao@kernel.org>
> 
> [...]

Here is the summary with links:
  - [f2fs-dev] f2fs: compress: fix to avoid fragment w/ OPU during f2fs_ioc_compress_file()
    https://git.kernel.org/jaegeuk/f2fs/c/943f7c6f987f

You are awesome, thank you!
diff mbox series

Patch

diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index 2a38d2cc2140..d4b59c57d827 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -2665,6 +2665,11 @@  bool f2fs_should_update_outplace(struct inode *inode, struct f2fs_io_info *fio)
 		return true;
 	if (f2fs_is_atomic_file(inode))
 		return true;
+	/* rewrite low ratio compress data w/ OPU mode to avoid fragmentation */
+	if (f2fs_compressed_file(inode) &&
+		F2FS_OPTION(sbi).compress_mode == COMPR_MODE_USER &&
+		is_inode_flag_set(inode, FI_ENABLE_COMPRESS))
+		return true;
 
 	/* swap file is migrating in aligned write mode */
 	if (is_inode_flag_set(inode, FI_ALIGNED_WRITE))