f2fs: fix to keep consistent i_gc_rwsem lock order

Message ID 20230410021140.1825781-1-chao@kernel.org
State New
Headers
Series f2fs: fix to keep consistent i_gc_rwsem lock order |

Commit Message

Chao Yu April 10, 2023, 2:11 a.m. UTC
  i_gc_rwsem[WRITE] and i_gc_rwsem[READ] lock order is reversed
in gc_data_segment() and f2fs_dio_write_iter(), fix to keep
consistent lock order as below:
1. lock i_gc_rwsem[WRITE]
2. lock i_gc_rwsem[READ]

Signed-off-by: Chao Yu <chao@kernel.org>
---
 fs/f2fs/gc.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
  

Comments

patchwork-bot+f2fs@kernel.org April 11, 2023, 5 p.m. UTC | #1
Hello:

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

On Mon, 10 Apr 2023 10:11:40 +0800 you wrote:
> i_gc_rwsem[WRITE] and i_gc_rwsem[READ] lock order is reversed
> in gc_data_segment() and f2fs_dio_write_iter(), fix to keep
> consistent lock order as below:
> 1. lock i_gc_rwsem[WRITE]
> 2. lock i_gc_rwsem[READ]
> 
> Signed-off-by: Chao Yu <chao@kernel.org>
> 
> [...]

Here is the summary with links:
  - [f2fs-dev] f2fs: fix to keep consistent i_gc_rwsem lock order
    https://git.kernel.org/jaegeuk/f2fs/c/db141b729bcc

You are awesome, thank you!
  

Patch

diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c
index 6dc2bc710ba8..91841b0057ba 100644
--- a/fs/f2fs/gc.c
+++ b/fs/f2fs/gc.c
@@ -1616,14 +1616,14 @@  static int gc_data_segment(struct f2fs_sb_info *sbi, struct f2fs_summary *sum,
 			int err;
 
 			if (S_ISREG(inode->i_mode)) {
-				if (!f2fs_down_write_trylock(&fi->i_gc_rwsem[READ])) {
+				if (!f2fs_down_write_trylock(&fi->i_gc_rwsem[WRITE])) {
 					sbi->skipped_gc_rwsem++;
 					continue;
 				}
 				if (!f2fs_down_write_trylock(
-						&fi->i_gc_rwsem[WRITE])) {
+						&fi->i_gc_rwsem[READ])) {
 					sbi->skipped_gc_rwsem++;
-					f2fs_up_write(&fi->i_gc_rwsem[READ]);
+					f2fs_up_write(&fi->i_gc_rwsem[WRITE]);
 					continue;
 				}
 				locked = true;
@@ -1646,8 +1646,8 @@  static int gc_data_segment(struct f2fs_sb_info *sbi, struct f2fs_summary *sum,
 				submitted++;
 
 			if (locked) {
-				f2fs_up_write(&fi->i_gc_rwsem[WRITE]);
 				f2fs_up_write(&fi->i_gc_rwsem[READ]);
+				f2fs_up_write(&fi->i_gc_rwsem[WRITE]);
 			}
 
 			stat_inc_data_blk_count(sbi, 1, gc_type);