https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=276363
Bug ID: 276363
Summary: if_wg: Fix a code bug in calculate_padding() although
impossible to hit
Product: Base System
Version: CURRENT
Hardware: Any
OS: Any
Status: New
Severity: Affects Some People
Priority: ---
Component: kern
Assignee: b...@freebsd.org
Reporter: a...@aaronly.me
The padding length calculation in calculate_padding() is wrong for the special
case of 'p_mtu = 0'. It's calculating the padded length instead of the length
to pad. So it should be:
if (__predict_false(!pkt->p_mtu))
return ((last_unit + (WG_PKT_PADDING - 1)) & ~(WG_PKT_PADDING - 1)) -
last_unit;
However, it's impossible for 'p_mtu' to be zero. So I'd propose the following
fix:
@@ -1461,8 +1461,7 @@ calculate_padding(struct wg_packet *pkt)
{
unsigned int padded_size, last_unit = pkt->p_mbuf->m_pkthdr.len;
- if (__predict_false(!pkt->p_mtu))
- return (last_unit + (WG_PKT_PADDING - 1)) & ~(WG_PKT_PADDING -
1);
+ KASSERT(pkt->p_mtu != 0, ("%s: packet %p has p_mtu = 0", __func__, m));
if (__predict_false(last_unit > pkt->p_mtu))
last_unit %= pkt->p_mtu;
--
You are receiving this mail because:
You are the assignee for the bug.
