f2fs crypto: avoid unneeded memory allocation in ->readdir

commit e06f86e61d7a67fe6e826010f57aa39c674f4b1b upstream.

When decrypting dirents in ->readdir, fscrypt_fname_disk_to_usr won't
change content of original encrypted dirent, we don't need to allocate
additional buffer for storing mirror of it, so get rid of it.

[This backport fixes a regression in 4.4-stable caused by commit
11a6e8f89521 ("f2fs: check memory boundary by insane namelen"), which
depended on this missing commit.  This bad backport broke f2fs
encryption because it moved the incrementing of 'bit_pos' to earlier in
f2fs_fill_dentries() without accounting for it being used in the
encrypted dir case.  This caused readdir() on encrypted directories to
start failing.  Tested with 'kvm-xfstests -c f2fs -g encrypt'.]

Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c
index e2ff0eb16f89ce49956032e607cf0b9636ed17cc..c1130914d6ed7e18ad952538781905195edcf683 100644 (file)
--- a/fs/f2fs/dir.c
+++ b/fs/f2fs/dir.c
@@ -820,15 +820,8 @@ bool f2fs_fill_dentries(struct dir_context *ctx, struct f2fs_dentry_ptr *d,
                        int save_len = fstr->len;
                        int ret;
 
-                       de_name.name = kmalloc(de_name.len, GFP_NOFS);
-                       if (!de_name.name)
-                               return false;
-
-                       memcpy(de_name.name, d->filename[bit_pos], de_name.len);
-
                        ret = f2fs_fname_disk_to_usr(d->inode, &de->hash_code,
                                                        &de_name, fstr);
-                       kfree(de_name.name);
                        if (ret < 0)
                                return true;