f2fs: decrease the lock granularity during write_begin

This patch reduces the lock granularity during write_begin.
When the system is under memory pressure, it would be better to reduce
the locking time for the data pages.

Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>

diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index 91ff104b38492b97448f2f8d98f53e62adffbd7c..a7bb98f5831ae2127f3d5e4dfdf6335c3e4ba66b 100644 (file)
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -909,6 +909,10 @@ repeat:
        page = grab_cache_page_write_begin(mapping, index, flags);
        if (!page)
                return -ENOMEM;
+
+       /* to avoid latency during memory pressure */
+       unlock_page(page);
+
        *pagep = page;
 
        if (f2fs_has_inline_data(inode) && (pos + len) <= MAX_INLINE_DATA)
@@ -920,10 +924,18 @@ repeat:
        f2fs_unlock_op(sbi);
 
        if (err) {
-               f2fs_put_page(page, 1);
+               f2fs_put_page(page, 0);
                return err;
        }
 inline_data:
+       lock_page(page);
+       if (unlikely(page->mapping != mapping)) {
+               f2fs_put_page(page, 1);
+               goto repeat;
+       }
+
+       f2fs_wait_on_page_writeback(page, DATA);
+
        if ((len == PAGE_CACHE_SIZE) || PageUptodate(page))
                return 0;