nls: utf8_wcstombs: fix buffer overflow

utf8_wcstombs forgot to include one-byte UTF-8 characters when
calculating the output buffer size, i.e., theoretically, it was possible
to overflow the output buffer with an input string that contains enough
ASCII characters.

In practice, this was no problem because the only user so far (VFAT)
always uses a big enough output buffer.

Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

diff --git a/fs/nls/nls_base.c b/fs/nls/nls_base.c
index 000736d..750abf2 100644 (file)
--- a/fs/nls/nls_base.c
+++ b/fs/nls/nls_base.c
@@ -150,6 +150,7 @@ utf8_wcstombs(__u8 *s, const wchar_t *pwcs, int maxlen)
                        }
                } else {
                        *op++ = (__u8) *ip;
+                       maxlen--;
                }
                ip++;
        }