[PATCH v2 net] mac802154: fix uninit-value in ieee802154_header_create()

From: Petr Vaganov

Date: Fri Nov 20 2026 - 03:47:31 EST


During fuzz testing, the following issue was discovered:

BUG: KMSAN: uninit-value in ieee802154_hdr_push_sechdr net/ieee802154/header_ops.c:54 [inline]
BUG: KMSAN: uninit-value in ieee802154_hdr_push+0xa62/0xb20 net/ieee802154/header_ops.c:108
ieee802154_hdr_push_sechdr net/ieee802154/header_ops.c:54 [inline]
ieee802154_hdr_push+0xa62/0xb20 net/ieee802154/header_ops.c:108
ieee802154_header_create+0xa83/0xd20 net/mac802154/iface.c:396
wpan_dev_hard_header include/net/cfg802154.h:525 [inline]
dgram_sendmsg+0xc6d/0x15d0 net/ieee802154/socket.c:677
ieee802154_sock_sendmsg+0x8f/0xd0 net/ieee802154/socket.c:96
sock_sendmsg_nosec net/socket.c:711 [inline]
__sock_sendmsg+0x2f3/0x360 net/socket.c:726
____sys_sendmsg+0x960/0xc30 net/socket.c:2581
___sys_sendmsg+0x285/0x3e0 net/socket.c:2635
__sys_sendmmsg+0x33c/0x950 net/socket.c:2724
__do_sys_sendmmsg net/socket.c:2751 [inline]
__se_sys_sendmmsg net/socket.c:2748 [inline]
__x64_sys_sendmmsg+0xc6/0x150 net/socket.c:2748
x64_sys_call+0x12a8/0x3cf0 arch/x86/include/generated/asm/syscalls_64.h:308
do_syscall_x64 arch/x86/entry/common.c:47 [inline]
do_syscall_64+0xd9/0x1c0 arch/x86/entry/common.c:78
entry_SYSCALL_64_after_hwframe+0x77/0x7f

Uninit was stored to memory at:
ieee802154_hdr_push_sechdr net/ieee802154/header_ops.c:51 [inline]
ieee802154_hdr_push+0xa5b/0xb20 net/ieee802154/header_ops.c:108
ieee802154_header_create+0xa83/0xd20 net/mac802154/iface.c:396
wpan_dev_hard_header include/net/cfg802154.h:525 [inline]
dgram_sendmsg+0xc6d/0x15d0 net/ieee802154/socket.c:677
ieee802154_sock_sendmsg+0x8f/0xd0 net/ieee802154/socket.c:96
sock_sendmsg_nosec net/socket.c:711 [inline]
__sock_sendmsg+0x2f3/0x360 net/socket.c:726
____sys_sendmsg+0x960/0xc30 net/socket.c:2581
___sys_sendmsg+0x285/0x3e0 net/socket.c:2635
__sys_sendmmsg+0x33c/0x950 net/socket.c:2724
__do_sys_sendmmsg net/socket.c:2751 [inline]
__se_sys_sendmmsg net/socket.c:2748 [inline]
__x64_sys_sendmmsg+0xc6/0x150 net/socket.c:2748
x64_sys_call+0x12a8/0x3cf0 arch/x86/include/generated/asm/syscalls_64.h:310
do_syscall_x64 arch/x86/entry/common.c:47 [inline]
do_syscall_64+0xd9/0x1c0 arch/x86/entry/common.c:78
entry_SYSCALL_64_after_hwframe+0x77/0x7f

Both ieee802154_header_create() and mac802154_header_create() declare
struct ieee802154_hdr hdr on the stack and clear only hdr.fc via
memset. This leaves hdr.sec uninitialized.

There are two ways uninitialized hdr.sec fields can be read:

1. ieee802154_header_create() had a pre-existing assignment:

hdr.fc.security_enabled = cb->secen;

This could set security_enabled = 1 from a stale cb->secen (e.g.,
after WPAN_SECURITY_ON followed by WPAN_SECURITY_DEFAULT), even when
mac802154_set_header_security() returns without initializing hdr.sec.
ieee802154_hdr_push_sechdr() would then read uninitialized hdr.sec.

2. mac802154_set_header_security() never writes sec.frame_counter.
ieee802154_hdr_push_sechdr() copies it from the stack into the skb
before mac802154_llsec_encrypt() sets the real value. If encryption
fails, or via AF_PACKET taps in dev_queue_xmit_nit(), the
uninitialized bytes are exposed to userspace.

Fix both issues by replacing memset(&hdr.fc, ...) with
memset(&hdr, ...) in both functions, which zeros hdr.sec entirely
before mac802154_set_header_security() is called. The stale
hdr.fc.security_enabled assignment is also removed as it is now
superseded by the full zero-initialization.

Found by Linux Verification Center (linuxtesting.org) with Syzkaller.

Closes: https://syzkaller.appspot.com/bug?extid=60a66d44892b66b56545
Closes: https://syzkaller.appspot.com/bug?id=23f21e76e9d2d1ed587cb5e71515260d6d025b63
Fixes: f30be4d53cad ("mac802154: integrate llsec with wpan devices")
Cc: stable@xxxxxxxxxxxxxxx
Signed-off-by: Petr Vaganov <petrvaganoff@xxxxxxxxx>
---
Changes in v2:
- Replace memset(&hdr.fc, ...) with memset(&hdr, ...) in both
ieee802154_header_create() and mac802154_header_create() to also
zero hdr.sec, closing the pre-existing uninitialized sec.frame_counter
read on the normal TX path (noted by Sashiko AI review).
- Rename subject to reflect this is a bug fix, not a cleanup.

net/mac802154/iface.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/net/mac802154/iface.c b/net/mac802154/iface.c
index f437b8daa00d..36bdd7558810 100644
--- a/net/mac802154/iface.c
+++ b/net/mac802154/iface.c
@@ -366,9 +366,8 @@ static int ieee802154_header_create(struct sk_buff *skb,
if (!daddr)
return -EINVAL;

- memset(&hdr.fc, 0, sizeof(hdr.fc));
+ memset(&hdr, 0, sizeof(hdr));
hdr.fc.type = cb->type;
- hdr.fc.security_enabled = cb->secen;
hdr.fc.ack_request = cb->ackreq;
hdr.seq = atomic_inc_return(&dev->ieee802154_ptr->dsn) & 0xFF;

@@ -433,7 +432,7 @@ static int mac802154_header_create(struct sk_buff *skb,
if (!daddr)
return -EINVAL;

- memset(&hdr.fc, 0, sizeof(hdr.fc));
+ memset(&hdr, 0, sizeof(hdr));
hdr.fc.type = IEEE802154_FC_TYPE_DATA;
hdr.fc.ack_request = wpan_dev->ackreq;
hdr.seq = atomic_inc_return(&dev->ieee802154_ptr->dsn) & 0xFF;
--
2.55.0