netfilter: nf_log: validate MAC header was set before dumping it
Subsystems (4)
PATCH 1/1 netfilter: nf_log: validate MAC header was set before dumping it
From: Xiang Mei <[email protected]>
commit a84b6fedbc97078788be78dbdd7517d143ad1a77 upstream
The fallback path of dump_mac_header() guards the MAC header access only with "skb->mac_header != skb->network_header", without checking skb_mac_header_was_set(). When the MAC header is unset, mac_header is 0xffff, so the test passes and skb_mac_header(skb) returns skb->head + 0xffff, ~64 KiB past the buffer; the loop then reads dev->hard_header_len bytes out of bounds into the kernel log.
This is reachable via the netdev logger: nf_log_unknown_packet() calls dump_mac_header() unconditionally, and an skb sent through AF_PACKET with PACKET_QDISC_BYPASS reaches the egress hook with mac_header still unset (__dev_queue_xmit(), which would reset it, is bypassed).
Add the skb_mac_header_was_set() check the ARPHRD_ETHER path already uses, and replace the open-coded MAC header length test with skb_mac_header_len(). Only skbs with an unset MAC header are affected; valid ones are dumped as before.
BUG: KASAN: slab-out-of-bounds in dump_mac_header (net/netfilter/nf_log_syslog.c:831) Read of size 1 at addr ffff88800ea49d3f by task exploit/148 Call Trace: kasan_report (mm/kasan/report.c:595) dump_mac_header (net/netfilter/nf_log_syslog.c:831) nf_log_netdev_packet (net/netfilter/nf_log_syslog.c:938 net/netfilter/nf_log_syslog.c:963) nf_log_packet (net/netfilter/nf_log.c:260) nft_log_eval (net/netfilter/nft_log.c:60) nft_do_chain (net/netfilter/nf_tables_core.c:285) nft_do_chain_netdev (net/netfilter/nft_chain_filter.c:307) nf_hook_slow (net/netfilter/core.c:619) nf_hook_direct_egress (net/packet/af_packet.c:257) packet_xmit (net/packet/af_packet.c:280) packet_sendmsg (net/packet/af_packet.c:3114) __sys_sendto (net/socket.c:2265)
---
Backport fix for CVE-2026-52942
net/ipv4/netfilter/nf_log_ipv4.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
net/ipv4/netfilter/nf_log_ipv4.c +2 -2
@@ -296,8 +296,8 @@ static void dump_ipv4_mac_header(struct nf_log_buf *m, fallback:nf_log_buf_add (m,"MAC=" ); -if (dev->hard_header_len && - skb->mac_header != skb->network_header) { +if (dev->hard_header_len &&skb_mac_header_was_set (skb) && +skb_mac_header_len (skb) !=0 ) {const unsigned char *p =skb_mac_header (skb);unsigned int i;
Signature · 1 line
2.43.0
Pablo Neira Ayuso
maintainerHi,
Thanks but why only 5.10?
Quoted 66 lines · Alexander Martyniuk
From: Xiang Mei <[email protected]>
commit
a84b6fedbc97078788be78dbdd7517d143ad1a77upstreamThe fallback path of
dump_mac_header()guards the MAC header access only with "skb->mac_header!=skb->network_header", without checkingskb_mac_header_was_set(). When the MAC header is unset,mac_headeris0xffff, so the test passes andskb_mac_header(skb)returnsskb->head+0xffff, ~64 KiB past the buffer; the loop then readsdev->hard_header_lenbytes out of bounds into the kernel log.This is reachable via the netdev logger:
nf_log_unknown_packet()callsdump_mac_header()unconditionally, and an skb sent throughAF_PACKETwithPACKET_QDISC_BYPASSreaches the egress hook withmac_headerstill unset (__dev_queue_xmit(), which would reset it, is bypassed).Add the
skb_mac_header_was_set()check theARPHRD_ETHERpath already uses, and replace the open-coded MAC header length test withskb_mac_header_len(). Only skbs with an unset MAC header are affected; valid ones are dumped as before.BUG: KASAN: slab-out-of-bounds in dump_mac_header (net/netfilter/nf_log_syslog.c:831) Read of size 1 at addr ffff88800ea49d3f by task exploit/148 Call Trace: kasan_report (mm/kasan/report.c:595) dump_mac_header (net/netfilter/nf_log_syslog.c:831) nf_log_netdev_packet (net/netfilter/nf_log_syslog.c:938 net/netfilter/nf_log_syslog.c:963) nf_log_packet (net/netfilter/nf_log.c:260) nft_log_eval (net/netfilter/nft_log.c:60) nft_do_chain (net/netfilter/nf_tables_core.c:285) nft_do_chain_netdev (net/netfilter/nft_chain_filter.c:307) nf_hook_slow (net/netfilter/core.c:619) nf_hook_direct_egress (net/packet/af_packet.c:257) packet_xmit (net/packet/af_packet.c:280) packet_sendmsg (net/packet/af_packet.c:3114) __sys_sendto (net/socket.c:2265)Fixes 7eb9282cd0ef ("netfilter: ipt_LOG/ip6t_LOG: add option to print decoded MAC header")Reported-by Weiming Shi <[email protected]>Assisted-by Claude:claude-opus-4-8Signed-off-by Xiang Mei <[email protected]>Signed-off-by Pablo Neira Ayuso <[email protected]>Signed-off-by Alexander Martyniuk <[email protected]>---
Backport fix for CVE-2026-52942net/ipv4/netfilter/nf_log_ipv4.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)net/ipv4/netfilter/nf_log_ipv4.c +2
-2@@ -296,8 +296,8 @@ static void dump_ipv4_mac_header(struct nf_log_buf *m, fallback:nf_log_buf_add (m,"MAC=" ); -if (dev->hard_header_len && - skb->mac_header != skb->network_header) { +if (dev->hard_header_len &&skb_mac_header_was_set (skb) && +skb_mac_header_len (skb) !=0 ) {const unsigned char *p =skb_mac_header (skb);unsigned int i;Signature · 1 line
2.43.0
Pablo Neira Ayuso
maintainerBTW, fixing Cc: to [email protected]
Quoted 71 lines · Pablo Neira Ayuso
Hi,
Thanks but why only 5.10?
Quoted 66 lines · Alexander Martyniuk
From: Xiang Mei <[email protected]>
commit
a84b6fedbc97078788be78dbdd7517d143ad1a77upstreamThe fallback path of
dump_mac_header()guards the MAC header access only with "skb->mac_header!=skb->network_header", without checkingskb_mac_header_was_set(). When the MAC header is unset,mac_headeris0xffff, so the test passes andskb_mac_header(skb)returnsskb->head+0xffff, ~64 KiB past the buffer; the loop then readsdev->hard_header_lenbytes out of bounds into the kernel log.This is reachable via the netdev logger:
nf_log_unknown_packet()callsdump_mac_header()unconditionally, and an skb sent throughAF_PACKETwithPACKET_QDISC_BYPASSreaches the egress hook withmac_headerstill unset (__dev_queue_xmit(), which would reset it, is bypassed).Add the
skb_mac_header_was_set()check theARPHRD_ETHERpath already uses, and replace the open-coded MAC header length test withskb_mac_header_len(). Only skbs with an unset MAC header are affected; valid ones are dumped as before.BUG: KASAN: slab-out-of-bounds in dump_mac_header (net/netfilter/nf_log_syslog.c:831) Read of size 1 at addr ffff88800ea49d3f by task exploit/148 Call Trace: kasan_report (mm/kasan/report.c:595) dump_mac_header (net/netfilter/nf_log_syslog.c:831) nf_log_netdev_packet (net/netfilter/nf_log_syslog.c:938 net/netfilter/nf_log_syslog.c:963) nf_log_packet (net/netfilter/nf_log.c:260) nft_log_eval (net/netfilter/nft_log.c:60) nft_do_chain (net/netfilter/nf_tables_core.c:285) nft_do_chain_netdev (net/netfilter/nft_chain_filter.c:307) nf_hook_slow (net/netfilter/core.c:619) nf_hook_direct_egress (net/packet/af_packet.c:257) packet_xmit (net/packet/af_packet.c:280) packet_sendmsg (net/packet/af_packet.c:3114) __sys_sendto (net/socket.c:2265)Fixes 7eb9282cd0ef ("netfilter: ipt_LOG/ip6t_LOG: add option to print decoded MAC header")Reported-by Weiming Shi <[email protected]>Assisted-by Claude:claude-opus-4-8Signed-off-by Xiang Mei <[email protected]>Signed-off-by Pablo Neira Ayuso <[email protected]>Signed-off-by Alexander Martyniuk <[email protected]>---
Backport fix for CVE-2026-52942net/ipv4/netfilter/nf_log_ipv4.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)net/ipv4/netfilter/nf_log_ipv4.c +2
-2@@ -296,8 +296,8 @@ static void dump_ipv4_mac_header(struct nf_log_buf *m, fallback:nf_log_buf_add (m,"MAC=" ); -if (dev->hard_header_len && - skb->mac_header != skb->network_header) { +if (dev->hard_header_len &&skb_mac_header_was_set (skb) && +skb_mac_header_len (skb) !=0 ) {const unsigned char *p =skb_mac_header (skb);unsigned int i;Signature · 1 line
2.43.0
Greg Kroah-Hartman
maintainerQuoted 3 lines · Pablo Neira Ayuso
Hi,
Thanks but why only 5.10?
It's already in the following releases:
5.15.210 6.1.176 6.6.143 6.12.94 6.18.36 7.0.13 7.1
so 5.10.y seems like the only one missing it at the moment.
thanks,
greg k-h
Sasha Levin
maintainernet/ipv4/netfilter/nf_log_ipv4.c
Quoted 5 lines
[PATCH 5.10] netfilter:
nf_log: validate MAC header was set before dumping itnet/ipv4/netfilter/nf_log_ipv4.c +0
-0
Thanks for the backport - the retarget to nf_log_ipv4.c is right for 5.10.
One gap though: upstream fixed both loggers via the consolidated nf_log_syslog.c, but in 5.10 the IPv6 logger (net/ipv6/netfilter/ nf_log_ipv6.c) still has the identical unguarded fallback and is left vulnerable here - which is also Pablo's "why only 5.10?" point.
--
Thanks,
Sasha
PATCH v2 1/1 netfilter: nf_log: validate MAC header was set before dumping it
From: Xiang Mei <[email protected]>
commit a84b6fedbc97078788be78dbdd7517d143ad1a77 upstream.
The fallback path of dump_mac_header() guards the MAC header access only with "skb->mac_header != skb->network_header", without checking skb_mac_header_was_set(). When the MAC header is unset, mac_header is 0xffff, so the test passes and skb_mac_header(skb) returns skb->head + 0xffff, ~64 KiB past the buffer; the loop then reads dev->hard_header_len bytes out of bounds into the kernel log.
This is reachable via the netdev logger: nf_log_unknown_packet() calls dump_mac_header() unconditionally, and an skb sent through AF_PACKET with PACKET_QDISC_BYPASS reaches the egress hook with mac_header still unset (__dev_queue_xmit(), which would reset it, is bypassed).
Add the skb_mac_header_was_set() check the ARPHRD_ETHER path already uses, and replace the open-coded MAC header length test with skb_mac_header_len(). Only skbs with an unset MAC header are affected; valid ones are dumped as before.
BUG: KASAN: slab-out-of-bounds in dump_mac_header (net/netfilter/nf_log_syslog.c:831) Read of size 1 at addr ffff88800ea49d3f by task exploit/148 Call Trace: kasan_report (mm/kasan/report.c:595) dump_mac_header (net/netfilter/nf_log_syslog.c:831) nf_log_netdev_packet (net/netfilter/nf_log_syslog.c:938 net/netfilter/nf_log_syslog.c:963) nf_log_packet (net/netfilter/nf_log.c:260) nft_log_eval (net/netfilter/nft_log.c:60) nft_do_chain (net/netfilter/nf_tables_core.c:285) nft_do_chain_netdev (net/netfilter/nft_chain_filter.c:307) nf_hook_slow (net/netfilter/core.c:619) nf_hook_direct_egress (net/packet/af_packet.c:257) packet_xmit (net/packet/af_packet.c:280) packet_sendmsg (net/packet/af_packet.c:3114) __sys_sendto (net/socket.c:2265)
--- net/ipv4/netfilter/nf_log_ipv4.c | 4 ++-- net/ipv6/netfilter/nf_log_ipv6.c | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-)
net/ipv4/netfilter/nf_log_ipv4.c +2 -2
@@ -296,8 +296,8 @@ static void dump_ipv4_mac_header(struct nf_log_buf *m, fallback:nf_log_buf_add (m,"MAC=" ); -if (dev->hard_header_len && - skb->mac_header != skb->network_header) { +if (dev->hard_header_len &&skb_mac_header_was_set (skb) && +skb_mac_header_len (skb) !=0 ) {const unsigned char *p =skb_mac_header (skb);unsigned int i;
net/ipv6/netfilter/nf_log_ipv6.c +2 -2
@@ -309,8 +309,8 @@ static void dump_ipv6_mac_header(struct nf_log_buf *m, fallback:nf_log_buf_add (m,"MAC=" ); -if (dev->hard_header_len && - skb->mac_header != skb->network_header) { +if (dev->hard_header_len &&skb_mac_header_was_set (skb) && +skb_mac_header_len (skb) !=0 ) {const unsigned char *p =skb_mac_header (skb);unsigned int len = dev->hard_header_len;unsigned int i;
Signature · 1 line
2.43.0
Greg KH
maintainerQuoted 82 lines · Alexander Martyniuk
From: Xiang Mei <[email protected]>
commit
a84b6fedbc97078788be78dbdd7517d143ad1a77upstream.The fallback path of
dump_mac_header()guards the MAC header access only with "skb->mac_header!=skb->network_header", without checkingskb_mac_header_was_set(). When the MAC header is unset,mac_headeris0xffff, so the test passes andskb_mac_header(skb)returnsskb->head+0xffff, ~64 KiB past the buffer; the loop then readsdev->hard_header_lenbytes out of bounds into the kernel log.This is reachable via the netdev logger:
nf_log_unknown_packet()callsdump_mac_header()unconditionally, and an skb sent throughAF_PACKETwithPACKET_QDISC_BYPASSreaches the egress hook withmac_headerstill unset (__dev_queue_xmit(), which would reset it, is bypassed).Add the
skb_mac_header_was_set()check theARPHRD_ETHERpath already uses, and replace the open-coded MAC header length test withskb_mac_header_len(). Only skbs with an unset MAC header are affected; valid ones are dumped as before.BUG: KASAN: slab-out-of-bounds in dump_mac_header (net/netfilter/nf_log_syslog.c:831) Read of size 1 at addr ffff88800ea49d3f by task exploit/148 Call Trace: kasan_report (mm/kasan/report.c:595) dump_mac_header (net/netfilter/nf_log_syslog.c:831) nf_log_netdev_packet (net/netfilter/nf_log_syslog.c:938 net/netfilter/nf_log_syslog.c:963) nf_log_packet (net/netfilter/nf_log.c:260) nft_log_eval (net/netfilter/nft_log.c:60) nft_do_chain (net/netfilter/nf_tables_core.c:285) nft_do_chain_netdev (net/netfilter/nft_chain_filter.c:307) nf_hook_slow (net/netfilter/core.c:619) nf_hook_direct_egress (net/packet/af_packet.c:257) packet_xmit (net/packet/af_packet.c:280) packet_sendmsg (net/packet/af_packet.c:3114) __sys_sendto (net/socket.c:2265)Fixes 7eb9282cd0ef ("netfilter: ipt_LOG/ip6t_LOG: add option to print decoded MAC header")Reported-by Weiming Shi <[email protected]>Assisted-by Claude:claude-opus-4-8Signed-off-by Xiang Mei <[email protected]>Signed-off-by Pablo Neira Ayuso <[email protected]>Signed-off-by Alexander Martyniuk <[email protected]>--- net/ipv4/netfilter/nf_log_ipv4.c | 4 ++-- net/ipv6/netfilter/nf_log_ipv6.c | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-)net/ipv4/netfilter/nf_log_ipv4.c +2
-2@@ -296,8 +296,8 @@ static void dump_ipv4_mac_header(struct nf_log_buf *m, fallback:nf_log_buf_add (m,"MAC=" ); -if (dev->hard_header_len && - skb->mac_header != skb->network_header) { +if (dev->hard_header_len &&skb_mac_header_was_set (skb) && +skb_mac_header_len (skb) !=0 ) {const unsigned char *p =skb_mac_header (skb);unsigned int i;net/ipv6/netfilter/nf_log_ipv6.c +2
-2@@ -309,8 +309,8 @@ static void dump_ipv6_mac_header(struct nf_log_buf *m, fallback:nf_log_buf_add (m,"MAC=" ); -if (dev->hard_header_len && - skb->mac_header != skb->network_header) { +if (dev->hard_header_len &&skb_mac_header_was_set (skb) && +skb_mac_header_len (skb) !=0 ) {const unsigned char *p =skb_mac_header (skb);unsigned int len = dev->hard_header_len;unsigned int i;Signature · 1 line
2.43.0
What kernel(s) is this for?
Quoted 1 line
What kernel(s) is this for?
5.10