Now IS_DNODE is used in f2fs_flush_inline_data and it has some problems:
1. Just only inodes may include inline data,not all direct nodes
2. When system IO is busy, it is inefficient to lock a direct node page
but not an inode page. Besides, if this direct node page is being
locked by others for IO, f2fs_flush_inline_data will be blocked here,
which will affects the checkpoint process, this is unreasonable.

So IS_INODE should be used in f2fs_flush_inline_data.

Signed-off-by: Zhiguo Niu <zhiguo.niu@unisoc.com>
---
 fs/f2fs/node.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
index 541c4ad..a8fe49a3 100644
--- a/fs/f2fs/node.c
+++ b/fs/f2fs/node.c
@@ -1919,7 +1919,7 @@ void f2fs_flush_inline_data(struct f2fs_sb_info *sbi)
 		for (i = 0; i < nr_folios; i++) {
 			struct page *page = &fbatch.folios[i]->page;
 
-			if (!IS_DNODE(page))
+			if (!IS_INODE(page))
 				continue;
 
 			lock_page(page);
-- 
1.9.1

On 2024/1/17 15:59, Zhiguo Niu wrote:
> Now IS_DNODE is used in f2fs_flush_inline_data and it has some problems:
> 1. Just only inodes may include inline data,not all direct nodes
> 2. When system IO is busy, it is inefficient to lock a direct node page
> but not an inode page. Besides, if this direct node page is being
> locked by others for IO, f2fs_flush_inline_data will be blocked here,
> which will affects the checkpoint process, this is unreasonable.
> 
> So IS_INODE should be used in f2fs_flush_inline_data.
> 
> Signed-off-by: Zhiguo Niu <zhiguo.niu@unisoc.com>

Reviewed-by: Chao Yu <chao@kernel.org>

Thanks,