GHSA-M48G-6CV9-P3RH
Vulnerability from github – Published: 2026-08-28 09:31 – Updated: 2026-08-31 09:30In the Linux kernel, the following vulnerability has been resolved:
inet: frags: strip GSO state from fragments before reassembly
A virtio_net_hdr (tun/tap, or AF_PACKET with PACKET_VNET_HDR) can mark an IPv4 or IPv6 fragment as GSO; nothing relates gso_type to frag_off. inet_frag_reasm_prepare()/inet_frag_reasm_finish() keep the first fragment's skb as the head of the reassembled datagram, including its shinfo->gso_size/gso_type/gso_segs, and chain the remaining fragments on frag_list with whatever linear/paged layout they arrived with.
After ip_defrag() (ip_local_deliver(), nf_defrag_ipv4, ...) the reassembled skb therefore still claims to be GSO (SKB_GSO_DODGY), and the next software segmentation point - udp_rcv_segment() on local delivery, validate_xmit_skb(), or the ip_finish_output_gso() slow path - hands it to skb_segment(). skb_segment()'s frag_list walk assumes GRO-shaped input and hits one of its BUG_ON()s. Two writes to a tap by an unprivileged user in its own userns are enough:
kernel BUG at net/core/skbuff.c:4899! Oops: invalid opcode: 0000 [#1] SMP KASAN NOPTI CPU: 0 UID: 1000 PID: 82 Comm: poc Not tainted 7.2.0-pentest+ #2 RIP: 0010:skb_segment+0x20ca/0x48b0 Call Trace: __udp_gso_segment+0x29a/0x27d0 udp4_ufo_fragment+0x458/0x6c0 inet_gso_segment+0x429/0x1340 skb_mac_gso_segment+0x233/0x4f0 __skb_gso_segment+0x308/0x660 udp_queue_rcv_skb+0x440/0xad0 udp_unicast_rcv_skb+0xc7/0x2c0 udp_rcv+0x16ce/0x2260 ip_protocol_deliver_rcu+0x197/0x2d0 ip_local_deliver+0x430/0x690 ip_rcv+0x16f/0x1f0 __netif_receive_skb_one_core+0x15e/0x1c0 __netif_receive_skb+0x1e/0x110 netif_receive_skb+0xf6/0x5c0 tun_rx_batched.isra.0+0x3ab/0x790 tun_get_user+0x17c3/0x3550 tun_chr_write_iter+0xba/0x1b0 vfs_write+0x646/0x1130 Kernel panic - not syncing: Fatal exception in interrupt
This runs with BH disabled, so it is a panic rather than an oops. The same is reachable with CAP_NET_RAW in a netns where a defrag point precedes a GSO point, and from a guest whose VMM forwards virtio_net_hdr to a tap. The SKB_GSO_DODGY frag_list checks added by commit 3dcbdb134f32 ("net: gso: Fix skb_segment splat when splitting gso_size mangled skb having linear-headed frag_list") and by commit 9e4b7a99a03a ("net: gso: fix panic on frag_list with mixed head alloc types") do not cover it: page-backed heads skip them, and kmalloc heads skip them when gso_size == skb_headlen(head), which the sender controls.
An skb entering a frag queue is an IP fragment by definition and cannot legitimately carry GSO state: GRO does not merge fragments and the stack segments before it fragments, so only untrusted sources are affected. This has been reachable since commit f43798c27684 ("tun: Allow GSO using virtio_net_hdr"), the first path that let userspace attach GSO metadata to an IP fragment. Reset the GSO fields of every fragment as it is queued, in inet_frag_queue_insert(), which IPv4, IPv6, nf_conntrack_reasm and 6lowpan reassembly share; then neither the head nor the frag_list members of the reassembled skb carry them (the members matter too: the ip_do_fragment()/ip6_fragment() fast paths send them out as they are). The head may remain CHECKSUM_PARTIAL; that is already accepted on receive and resolved by skb_checksum_help() in ip_do_fragment()/ip6_fragment() on forward.
Tested on top of net.git (dc4b95b8fee9), x86_64: the tap reproducer above, two further IPv4 frag_list geometries that reach BUG_ON(i >= nfrags) and BUG_ON(!list_skb->head_frag), and an IPv6 fragment-header variant (udp6_ufo_fragment()) each panic the unpatched kernel; with this patch all four datagrams are delivered intact and nothing is logged.
{
"affected": [],
"aliases": [
"CVE-2026-80590"
],
"database_specific": {
"cwe_ids": [],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2026-08-28T08:16:42Z",
"severity": "HIGH"
},
"details": "In the Linux kernel, the following vulnerability has been resolved:\n\ninet: frags: strip GSO state from fragments before reassembly\n\nA virtio_net_hdr (tun/tap, or AF_PACKET with PACKET_VNET_HDR) can mark\nan IPv4 or IPv6 fragment as GSO; nothing relates gso_type to frag_off.\ninet_frag_reasm_prepare()/inet_frag_reasm_finish() keep the first\nfragment\u0027s skb as the head of the reassembled datagram, including its\nshinfo-\u003egso_size/gso_type/gso_segs, and chain the remaining fragments\non frag_list with whatever linear/paged layout they arrived with.\n\nAfter ip_defrag() (ip_local_deliver(), nf_defrag_ipv4, ...) the\nreassembled skb therefore still claims to be GSO (SKB_GSO_DODGY), and\nthe next software segmentation point - udp_rcv_segment() on local\ndelivery, validate_xmit_skb(), or the ip_finish_output_gso() slow\npath - hands it to skb_segment(). skb_segment()\u0027s frag_list walk\nassumes GRO-shaped input and hits one of its BUG_ON()s. Two writes to\na tap by an unprivileged user in its own userns are enough:\n\n kernel BUG at net/core/skbuff.c:4899!\n Oops: invalid opcode: 0000 [#1] SMP KASAN NOPTI\n CPU: 0 UID: 1000 PID: 82 Comm: poc Not tainted 7.2.0-pentest+ #2\n RIP: 0010:skb_segment+0x20ca/0x48b0\n Call Trace:\n \u003cTASK\u003e\n __udp_gso_segment+0x29a/0x27d0\n udp4_ufo_fragment+0x458/0x6c0\n inet_gso_segment+0x429/0x1340\n skb_mac_gso_segment+0x233/0x4f0\n __skb_gso_segment+0x308/0x660\n udp_queue_rcv_skb+0x440/0xad0\n udp_unicast_rcv_skb+0xc7/0x2c0\n udp_rcv+0x16ce/0x2260\n ip_protocol_deliver_rcu+0x197/0x2d0\n ip_local_deliver+0x430/0x690\n ip_rcv+0x16f/0x1f0\n __netif_receive_skb_one_core+0x15e/0x1c0\n __netif_receive_skb+0x1e/0x110\n netif_receive_skb+0xf6/0x5c0\n tun_rx_batched.isra.0+0x3ab/0x790\n tun_get_user+0x17c3/0x3550\n tun_chr_write_iter+0xba/0x1b0\n vfs_write+0x646/0x1130\n \u003c/TASK\u003e\n Kernel panic - not syncing: Fatal exception in interrupt\n\nThis runs with BH disabled, so it is a panic rather than an oops. The\nsame is reachable with CAP_NET_RAW in a netns where a defrag point\nprecedes a GSO point, and from a guest whose VMM forwards\nvirtio_net_hdr to a tap. The SKB_GSO_DODGY frag_list checks added by\ncommit 3dcbdb134f32 (\"net: gso: Fix skb_segment splat when splitting\ngso_size mangled skb having linear-headed frag_list\") and by\ncommit 9e4b7a99a03a (\"net: gso: fix panic on frag_list with mixed head\nalloc types\") do not cover it: page-backed heads skip them, and kmalloc\nheads skip them when gso_size == skb_headlen(head), which the sender\ncontrols.\n\nAn skb entering a frag queue is an IP fragment by definition and\ncannot legitimately carry GSO state: GRO does not merge fragments and\nthe stack segments before it fragments, so only untrusted sources are\naffected. This has been reachable since\ncommit f43798c27684 (\"tun: Allow GSO using virtio_net_hdr\"), the first\npath that let userspace attach GSO metadata to an IP fragment. Reset\nthe GSO fields of every fragment as it is queued, in\ninet_frag_queue_insert(), which IPv4, IPv6, nf_conntrack_reasm and\n6lowpan reassembly share; then neither the head nor the frag_list\nmembers of the reassembled skb carry them (the members matter too:\nthe ip_do_fragment()/ip6_fragment() fast paths send them out as they\nare). The head may remain CHECKSUM_PARTIAL; that is already accepted\non receive and resolved by skb_checksum_help() in\nip_do_fragment()/ip6_fragment() on forward.\n\nTested on top of net.git (dc4b95b8fee9), x86_64: the tap reproducer\nabove, two further IPv4 frag_list geometries that reach\nBUG_ON(i \u003e= nfrags) and BUG_ON(!list_skb-\u003ehead_frag), and an IPv6\nfragment-header variant (udp6_ufo_fragment()) each panic the unpatched\nkernel; with this patch all four datagrams are delivered intact and\nnothing is logged.",
"id": "GHSA-m48g-6cv9-p3rh",
"modified": "2026-08-31T09:30:26Z",
"published": "2026-08-28T09:31:45Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-80590"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/14a8f3e10fa9a5abd6cedcdaa0c0b7ea9a09f234"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/29dda278a5ed272f2230ff4eaa23cf403107bba0"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/3edf721bb4b99d272c336631b44e3d8ff9a4f31b"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/69b73b74d9eb45f5560a8fe4fa406ada580e1340"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/c49f04e8d2b94dbb8d9fd99731dd3f00589c8ace"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/cfdbc8c2e6f9ef5d8b8e54859da03dfe682b0bee"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/d5dc1e69fd7258ea605c9952e5d5947539159ae3"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/da857e448322a2e871ce3ecc2900027041160d43"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/dec2edb7aaf12a8878b3a03172ea8fc277b8eaad"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:N/I:N/A:H",
"type": "CVSS_V3"
}
]
}
Sightings
| Author | Source | Type | Date | Other |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.