CVE-2026-72417 PUBLISHED

netfilter: flowtable: Validate iph->ihl in nf_flow_ip4_tunnel_proto()

Assigner: Linux
Reserved: 09.08.2026 Published: 15.08.2026 Updated: 17.08.2026

In the Linux kernel, the following vulnerability has been resolved:

netfilter: flowtable: Validate iph->ihl in nf_flow_ip4_tunnel_proto()

Add sanity check for iph->ihl field in nf_flow_ip4_tunnel_proto() before using it to compute the header size, avoiding out-of-bounds access with malformed IP headers. While at it, use iph->protocol instead of the hardcoded IPPROTO_IPIP constant when setting ctx->tun.proto and reference ctx->tun.hdr_size when updating ctx->offset.

Metrics

CVSS Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
CVSS Score: 9.8

AV:N - The flaw is hit in nf_flow_offload_ip_hook on NF_NETDEV_INGRESS for IPv4 packets arriving on nftables flowtable devices; a remote attacker can send crafted malformed IP/IPIP headers to that interface without any local access on the victim. AC:L - Once a flowtable ingress hook exists, the attacker fully controls packet headers and can reliably reach nf_flow_ip4_tunnel_proto() by sending IP traffic with an invalid ihl field; no race or victim-specific state beyond admin-enabled flow offload is required. PR:N - Exploitation requires no credentials or capabilities on the victim host—only that an administrator previously configured nftables flow offload on an ingress device; any remote sender can deliver the malformed packets to trigger parsing on the fast path. UI:N - No victim interaction is needed; malicious packets are processed automatically by the netdev ingress netfilter hook during normal receive/forward softirq handling on the configured interface. S:U - Impact is kernel memory corruption and crashes within the host kernel security boundary; this is standard in-kernel compromise territory, not a VM escape, IOMMU bypass, or other cross-authority scope change. C:H - Missing ihl validation lets iph->ihl<<2 compute an incorrect header size and ctx->offset, causing out-of-bounds skb reads in subsequent tuple parsing and forwarding; per kernel guidance, OOB reads warrant High confidentiality impact. I:H - Incorrect ctx->offset propagates into nf_flow_tuple_ip() and nf_flow_offload_forward(), which invoke skb_ensure_writable() and perform in-place NAT/header mutation at attacker-influenced offsets—memory corruption exploitable for write primitives warrants High integrity impact. A:H - Out-of-bounds access in softirq packet processing on the flowtable fast path can trigger KASAN reports, kernel oops, or panics when malformed headers are processed, constituting High availability impact.

Product Status

Vendor Linux
Product Linux
Versions Default: unaffected
  • affected from ab427db17885814069bae891834f20842f0ac3a4 to 025a41e76b51fbc7b8eaa5bacbaa9621d00e6aa7 (excl.)
  • affected from ab427db17885814069bae891834f20842f0ac3a4 to 84460b644329e25809b4a6d9279d6359d7fd8ebc (excl.)
Vendor Linux
Product Linux
Versions Default: affected
  • Version 6.19 is affected
  • unaffected from 0 to 6.19 (excl.)
  • unaffected from 7.1.5 to 7.1.* (incl.)
  • unaffected from 7.2 to * (incl.)

References