I recently found a case where de->name_len is 0 in f2fs_fill_dentries() easily reproduced,
and finally set the fsck flag.
Thread A Thread B
f2fs_readdir
f2fs_read_inline_dir
ctx->pos = d.max
f2fs_add_dentry
f2fs_add_inline_entry
do_convert_inline_dir
f2fs_add_regular_entry
f2fs_readdir
f2fs_fill_dentries
set_sbi_flag(sbi, SBI_NEED_FSCK)
Process A opens the folder, and has been reading without closing it. During this period,
Process B created a file under the folder (occupying multiple f2fs_dir_entry, exceeding
the d.max of the inline dir). After creation, process A uses the d.max of inline dir to
read it again, and it will read that de->name_len is 0.
And returning early in f2fs_read_inline_dir will cause the process to be unable to see
the changes before reopening the file.
So don't return early and remove the modification of ctx->pos in f2fs_read_inline_dir().
Signed-off-by: Yangtao Li <frank.li@xxxxxxxx>
---
fs/f2fs/inline.c | 5 -----
1 file changed, 5 deletions(-)
diff --git a/fs/f2fs/inline.c b/fs/f2fs/inline.c
index 56a20d5c15da..fc6551139a3d 100644
--- a/fs/f2fs/inline.c
+++ b/fs/f2fs/inline.c
@@ -729,9 +729,6 @@ int f2fs_read_inline_dir(struct file *file, struct dir_context *ctx,
make_dentry_ptr_inline(inode, &d, inline_dentry);
- if (ctx->pos == d.max)
- return 0;
-
ipage = f2fs_get_node_page(F2FS_I_SB(inode), inode->i_ino);
if (IS_ERR(ipage))
return PTR_ERR(ipage);
@@ -747,8 +744,6 @@ int f2fs_read_inline_dir(struct file *file, struct dir_context *ctx,
make_dentry_ptr_inline(inode, &d, inline_dentry);
err = f2fs_fill_dentries(ctx, &d, 0, fstr);
- if (!err)
- ctx->pos = d.max;
f2fs_put_page(ipage, 0);
return err < 0 ? err : 0;