{"api_version":"1","generated_at":"2026-08-29T12:13:31+00:00","cve":"CVE-2026-80612","urls":{"html":"https://cve.report/CVE-2026-80612","api":"https://cve.report/api/cve/CVE-2026-80612.json","docs":"https://cve.report/api","cve_org":"https://www.cve.org/CVERecord?id=CVE-2026-80612","nvd":"https://nvd.nist.gov/vuln/detail/CVE-2026-80612"},"summary":{"title":"net: lwtunnel: Drop skb metadata before LWT encapsulation","description":"In the Linux kernel, the following vulnerability has been resolved:\n\nnet: lwtunnel: Drop skb metadata before LWT encapsulation\n\nskb metadata is meant for passing information between XDP and TC. It lives\nin the skb headroom, immediately before skb->data. LWT programs cannot\naccess the __sk_buff->data_meta pseudo-pointer to metadata.\n\nHowever, LWT encapsulation prepends outer headers, moving skb->data back\nover the headroom where the metadata sits. On an RX-originated (forwarded)\npacket that still carries XDP metadata this goes wrong in two different\nways, depending on the encap type:\n\n1. Non-BPF LWT encaps (mpls, seg6, ioam6 ...) call skb_push()/skb_pull()\n   and silently overwrite the metadata that sits in the headroom.\n\n2) BPF LWT xmit calls bpf_skb_change_head(), which uses skb_data_move().\n   That helper expects metadata immediately before skb->data. But since\n   the IP output path runs LWT xmit before neighbour output has built\n   the outgoing L2 header, for forwarded packets skb->data points at the\n   L3 header while skb_mac_header() still points at the old L2 header.\n   skb_data_move() sees metadata ending at skb_mac_header(), not before\n   skb->data, warns and clears metadata:\n\n  WARNING: CPU: 21 PID: 454557 at include/linux/skbuff.h:4609 skb_data_move+0x47/0x90\n  CPU: 21 UID: 0 PID: 454557 Comm: napi/iconduit-g Tainted: G           O        6.18.21 #1\n  RIP: 0010:skb_data_move+0x47/0x90\n  Call Trace:\n   <IRQ>\n   bpf_skb_change_head+0xe6/0x1a0\n   bpf_prog_...+0x213/0x2e3\n   run_lwt_bpf.isra.0+0x1d3/0x360\n   bpf_xmit+0x46/0xe0\n   lwtunnel_xmit+0xa1/0xf0\n   ip_finish_output2+0x1e7/0x5e0\n   ip_output+0x63/0x100\n   __netif_receive_skb_one_core+0x85/0xa0\n   process_backlog+0x9c/0x150\n   __napi_poll+0x2b/0x190\n   net_rx_action+0x40b/0x7f0\n   handle_softirqs+0xd2/0x270\n   do_softirq+0x3f/0x60\n   </IRQ>\n\nThat is what happens, as for how to fix it - a received packet that\ncarries metadata can reach an encap through any of the three LWT\nredirect modes:\n\n  LWTUNNEL_STATE_INPUT_REDIRECT\n   ip6_rcv_finish\n     dst_input\n       lwtunnel_input\n\n  LWTUNNEL_STATE_OUTPUT_REDIRECT\n   ip6_rcv_finish\n     dst_input\n       ip6_forward\n         ip6_forward_finish\n           dst_output\n             lwtunnel_output\n\n  LWTUNNEL_STATE_XMIT_REDIRECT\n   ip6_rcv_finish\n     dst_input\n       ip6_forward\n         ip6_forward_finish\n           dst_output\n             ip6_output\n               ip6_finish_output\n                 ip6_finish_output2\n                   lwtunnel_xmit\n\nEvery encap funnels through the three LWT dispatch helpers, so drop the\nmetadata there, right before handing the skb to the encap op. This\nsingle chokepoint covers all encap types and all three redirect modes:\n\n  - lwtunnel_input():  seg6, rpl, ila, seg6_local\n  - lwtunnel_output(): ioam6\n  - lwtunnel_xmit():   mpls, LWT BPF xmit\n\nAlternatively, we could clear the metadata right after TC ingress hook.\nThat would require a compromise, however. Metadata would become\ninaccessible from TC egress (in setups where it actually reaches the\nhook it tact, that is without any L2 tunnels on path).","state":"PUBLISHED","assigner":"Linux","published_at":"2026-08-28 08:16:45","updated_at":"2026-08-29 07:16:45"},"problem_types":[],"metrics":[{"version":"3.1","source":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","type":"Secondary","score":"9.8","severity":"CRITICAL","vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H","data":{"version":"3.1","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H","baseScore":9.8,"baseSeverity":"CRITICAL","attackVector":"NETWORK","attackComplexity":"LOW","privilegesRequired":"NONE","userInteraction":"NONE","scope":"UNCHANGED","confidentialityImpact":"HIGH","integrityImpact":"HIGH","availabilityImpact":"HIGH"}},{"version":"3.1","source":"CNA","type":"DECLARED","score":"9.8","severity":"CRITICAL","vector":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H","data":{"baseScore":9.8,"baseSeverity":"CRITICAL","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H","version":"3.1"}}],"references":[{"url":"https://git.kernel.org/stable/c/c00320b0e355c4bf0ae4743a53b4180fea237546","name":"https://git.kernel.org/stable/c/c00320b0e355c4bf0ae4743a53b4180fea237546","refsource":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","tags":[],"title":"","mime":"","httpstatus":"","archivestatus":"0"},{"url":"https://git.kernel.org/stable/c/19eec11f3ab5dd29ba58f5f209c24e946c95ef12","name":"https://git.kernel.org/stable/c/19eec11f3ab5dd29ba58f5f209c24e946c95ef12","refsource":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","tags":[],"title":"","mime":"","httpstatus":"","archivestatus":"0"},{"url":"https://www.cve.org/CVERecord?id=CVE-2026-80612","name":"CVE Program record","refsource":"CVE.ORG","tags":["canonical"]},{"url":"https://nvd.nist.gov/vuln/detail/CVE-2026-80612","name":"NVD vulnerability detail","refsource":"NVD","tags":["canonical","analysis"]}],"affected":[{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected 8989d328dfe7c7a3b9f4b9f0ef60006d277f81cc 19eec11f3ab5dd29ba58f5f209c24e946c95ef12 git","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected 8989d328dfe7c7a3b9f4b9f0ef60006d277f81cc c00320b0e355c4bf0ae4743a53b4180fea237546 git","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"affected 6.19","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 6.19 semver","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 7.1.5 7.1.* semver","platforms":[]},{"source":"CNA","vendor":"Linux","product":"Linux","version":"unaffected 7.2 * original_commit_for_fix","platforms":[]}],"timeline":[],"solutions":[],"workarounds":[],"exploits":[],"credits":[],"nvd_cpes":[],"vendor_comments":[],"enrichments":{"kev":null,"epss":{"cve_year":"2026","cve_id":"80612","cve":"CVE-2026-80612","epss":"0.001550000","percentile":"0.049800000","score_date":"2026-08-28","updated_at":"2026-08-29 00:12:17"},"legacy_qids":[]},"source_records":{"cve_program":{"containers":{"cna":{"affected":[{"defaultStatus":"unaffected","product":"Linux","programFiles":["net/core/lwtunnel.c"],"repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","vendor":"Linux","versions":[{"lessThan":"19eec11f3ab5dd29ba58f5f209c24e946c95ef12","status":"affected","version":"8989d328dfe7c7a3b9f4b9f0ef60006d277f81cc","versionType":"git"},{"lessThan":"c00320b0e355c4bf0ae4743a53b4180fea237546","status":"affected","version":"8989d328dfe7c7a3b9f4b9f0ef60006d277f81cc","versionType":"git"}]},{"defaultStatus":"affected","product":"Linux","programFiles":["net/core/lwtunnel.c"],"repo":"https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git","vendor":"Linux","versions":[{"status":"affected","version":"6.19"},{"lessThan":"6.19","status":"unaffected","version":"0","versionType":"semver"},{"lessThanOrEqual":"7.1.*","status":"unaffected","version":"7.1.5","versionType":"semver"},{"lessThanOrEqual":"*","status":"unaffected","version":"7.2","versionType":"original_commit_for_fix"}]}],"cpeApplicability":[{"nodes":[{"cpeMatch":[{"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionEndExcluding":"7.1.5","versionStartIncluding":"6.19","vulnerable":true},{"criteria":"cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*","versionEndExcluding":"7.2","versionStartIncluding":"6.19","vulnerable":true}],"negate":false,"operator":"OR"}]}],"descriptions":[{"lang":"en","value":"In the Linux kernel, the following vulnerability has been resolved:\n\nnet: lwtunnel: Drop skb metadata before LWT encapsulation\n\nskb metadata is meant for passing information between XDP and TC. It lives\nin the skb headroom, immediately before skb->data. LWT programs cannot\naccess the __sk_buff->data_meta pseudo-pointer to metadata.\n\nHowever, LWT encapsulation prepends outer headers, moving skb->data back\nover the headroom where the metadata sits. On an RX-originated (forwarded)\npacket that still carries XDP metadata this goes wrong in two different\nways, depending on the encap type:\n\n1. Non-BPF LWT encaps (mpls, seg6, ioam6 ...) call skb_push()/skb_pull()\n   and silently overwrite the metadata that sits in the headroom.\n\n2) BPF LWT xmit calls bpf_skb_change_head(), which uses skb_data_move().\n   That helper expects metadata immediately before skb->data. But since\n   the IP output path runs LWT xmit before neighbour output has built\n   the outgoing L2 header, for forwarded packets skb->data points at the\n   L3 header while skb_mac_header() still points at the old L2 header.\n   skb_data_move() sees metadata ending at skb_mac_header(), not before\n   skb->data, warns and clears metadata:\n\n  WARNING: CPU: 21 PID: 454557 at include/linux/skbuff.h:4609 skb_data_move+0x47/0x90\n  CPU: 21 UID: 0 PID: 454557 Comm: napi/iconduit-g Tainted: G           O        6.18.21 #1\n  RIP: 0010:skb_data_move+0x47/0x90\n  Call Trace:\n   <IRQ>\n   bpf_skb_change_head+0xe6/0x1a0\n   bpf_prog_...+0x213/0x2e3\n   run_lwt_bpf.isra.0+0x1d3/0x360\n   bpf_xmit+0x46/0xe0\n   lwtunnel_xmit+0xa1/0xf0\n   ip_finish_output2+0x1e7/0x5e0\n   ip_output+0x63/0x100\n   __netif_receive_skb_one_core+0x85/0xa0\n   process_backlog+0x9c/0x150\n   __napi_poll+0x2b/0x190\n   net_rx_action+0x40b/0x7f0\n   handle_softirqs+0xd2/0x270\n   do_softirq+0x3f/0x60\n   </IRQ>\n\nThat is what happens, as for how to fix it - a received packet that\ncarries metadata can reach an encap through any of the three LWT\nredirect modes:\n\n  LWTUNNEL_STATE_INPUT_REDIRECT\n   ip6_rcv_finish\n     dst_input\n       lwtunnel_input\n\n  LWTUNNEL_STATE_OUTPUT_REDIRECT\n   ip6_rcv_finish\n     dst_input\n       ip6_forward\n         ip6_forward_finish\n           dst_output\n             lwtunnel_output\n\n  LWTUNNEL_STATE_XMIT_REDIRECT\n   ip6_rcv_finish\n     dst_input\n       ip6_forward\n         ip6_forward_finish\n           dst_output\n             ip6_output\n               ip6_finish_output\n                 ip6_finish_output2\n                   lwtunnel_xmit\n\nEvery encap funnels through the three LWT dispatch helpers, so drop the\nmetadata there, right before handing the skb to the encap op. This\nsingle chokepoint covers all encap types and all three redirect modes:\n\n  - lwtunnel_input():  seg6, rpl, ila, seg6_local\n  - lwtunnel_output(): ioam6\n  - lwtunnel_xmit():   mpls, LWT BPF xmit\n\nAlternatively, we could clear the metadata right after TC ingress hook.\nThat would require a compromise, however. Metadata would become\ninaccessible from TC egress (in setups where it actually reaches the\nhook it tact, that is without any L2 tunnels on path)."}],"metrics":[{"cvssV3_1":{"baseScore":9.8,"baseSeverity":"CRITICAL","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H","version":"3.1"},"scenarios":[{"lang":"en","value":"AV:N - The bug fires on the network RX datapath in softirq when remotely sourced packets are received, forwarded, and LWT-encapsulated (ip6_rcv_finish to ip6_forward to lwtunnel_xmit as shown in the fix commit stack trace).\nAC:L - Once a forwarding host has XDP/TC metadata plus LWT encap routes, an attacker can repeatedly send matching traffic and deterministically hit the metadata/encap collision each time without races or victim-timed actions.\nPR:N - The attacker only needs to send IP packets to a forwarding system; LWT routes, XDP/TC metadata programs, and forwarding are administrator deployment choices, not privileges the attacker must hold on the victim.\nUI:N - No end-user interaction is required beyond delivering attacker-controlled network packets to an affected router or edge node.\nS:U - Impact is confined to kernel network packet processing on the same host and does not cross VM, hypervisor, or sandbox security boundaries.\nC:H - Non-BPF LWT encap silently overwrites metadata while meta_len can remain set, and BPF LWT xmit mishandles metadata via skb_data_move, enabling corrupted metadata reads by later BPF/TC code and potential information disclosure.\nI:H - LWT encapsulation performs headroom skb_push/bpf_skb_change_head over the metadata region without clearing meta_len, corrupting skb memory state in ways that can misdirect subsequent kernel/BPF packet-buffer writes.\nA:H - The BPF LWT path triggers skb_data_move WARN_ON_ONCE in softirq and non-BPF encap corrupts forwarding skbs, enabling warn-driven failure modes and repeated disruption of packet forwarding on internet-facing edge routers."}]}],"providerMetadata":{"dateUpdated":"2026-08-29T06:21:20.430Z","orgId":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","shortName":"Linux"},"references":[{"url":"https://git.kernel.org/stable/c/19eec11f3ab5dd29ba58f5f209c24e946c95ef12"},{"url":"https://git.kernel.org/stable/c/c00320b0e355c4bf0ae4743a53b4180fea237546"}],"title":"net: lwtunnel: Drop skb metadata before LWT encapsulation","x_generator":{"engine":"bippy-1.2.0"}}},"cveMetadata":{"assignerOrgId":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","assignerShortName":"Linux","cveId":"CVE-2026-80612","datePublished":"2026-08-28T06:48:34.527Z","dateReserved":"2026-08-26T14:34:25.773Z","dateUpdated":"2026-08-29T06:21:20.430Z","state":"PUBLISHED"},"dataType":"CVE_RECORD","dataVersion":"5.2"},"nvd":{"publishedDate":"2026-08-28 08:16:45","lastModifiedDate":"2026-08-29 07:16:45","problem_types":[],"metrics":{"cvssMetricV31":[{"source":"416baaa9-dc9f-4396-8d5f-8c081fb06d67","type":"Secondary","cvssData":{"version":"3.1","vectorString":"CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H","baseScore":9.8,"baseSeverity":"CRITICAL","attackVector":"NETWORK","attackComplexity":"LOW","privilegesRequired":"NONE","userInteraction":"NONE","scope":"UNCHANGED","confidentialityImpact":"HIGH","integrityImpact":"HIGH","availabilityImpact":"HIGH"},"exploitabilityScore":3.9,"impactScore":5.9}]},"configurations":[]},"legacy_mitre":{"record":{"CveYear":"2026","CveId":"80612","Ordinal":"1","Title":"net: lwtunnel: Drop skb metadata before LWT encapsulation","CVE":"CVE-2026-80612","Year":"2026"},"notes":[{"CveYear":"2026","CveId":"80612","Ordinal":"1","NoteData":"In the Linux kernel, the following vulnerability has been resolved:\n\nnet: lwtunnel: Drop skb metadata before LWT encapsulation\n\nskb metadata is meant for passing information between XDP and TC. It lives\nin the skb headroom, immediately before skb->data. LWT programs cannot\naccess the __sk_buff->data_meta pseudo-pointer to metadata.\n\nHowever, LWT encapsulation prepends outer headers, moving skb->data back\nover the headroom where the metadata sits. On an RX-originated (forwarded)\npacket that still carries XDP metadata this goes wrong in two different\nways, depending on the encap type:\n\n1. Non-BPF LWT encaps (mpls, seg6, ioam6 ...) call skb_push()/skb_pull()\n   and silently overwrite the metadata that sits in the headroom.\n\n2) BPF LWT xmit calls bpf_skb_change_head(), which uses skb_data_move().\n   That helper expects metadata immediately before skb->data. But since\n   the IP output path runs LWT xmit before neighbour output has built\n   the outgoing L2 header, for forwarded packets skb->data points at the\n   L3 header while skb_mac_header() still points at the old L2 header.\n   skb_data_move() sees metadata ending at skb_mac_header(), not before\n   skb->data, warns and clears metadata:\n\n  WARNING: CPU: 21 PID: 454557 at include/linux/skbuff.h:4609 skb_data_move+0x47/0x90\n  CPU: 21 UID: 0 PID: 454557 Comm: napi/iconduit-g Tainted: G           O        6.18.21 #1\n  RIP: 0010:skb_data_move+0x47/0x90\n  Call Trace:\n   <IRQ>\n   bpf_skb_change_head+0xe6/0x1a0\n   bpf_prog_...+0x213/0x2e3\n   run_lwt_bpf.isra.0+0x1d3/0x360\n   bpf_xmit+0x46/0xe0\n   lwtunnel_xmit+0xa1/0xf0\n   ip_finish_output2+0x1e7/0x5e0\n   ip_output+0x63/0x100\n   __netif_receive_skb_one_core+0x85/0xa0\n   process_backlog+0x9c/0x150\n   __napi_poll+0x2b/0x190\n   net_rx_action+0x40b/0x7f0\n   handle_softirqs+0xd2/0x270\n   do_softirq+0x3f/0x60\n   </IRQ>\n\nThat is what happens, as for how to fix it - a received packet that\ncarries metadata can reach an encap through any of the three LWT\nredirect modes:\n\n  LWTUNNEL_STATE_INPUT_REDIRECT\n   ip6_rcv_finish\n     dst_input\n       lwtunnel_input\n\n  LWTUNNEL_STATE_OUTPUT_REDIRECT\n   ip6_rcv_finish\n     dst_input\n       ip6_forward\n         ip6_forward_finish\n           dst_output\n             lwtunnel_output\n\n  LWTUNNEL_STATE_XMIT_REDIRECT\n   ip6_rcv_finish\n     dst_input\n       ip6_forward\n         ip6_forward_finish\n           dst_output\n             ip6_output\n               ip6_finish_output\n                 ip6_finish_output2\n                   lwtunnel_xmit\n\nEvery encap funnels through the three LWT dispatch helpers, so drop the\nmetadata there, right before handing the skb to the encap op. This\nsingle chokepoint covers all encap types and all three redirect modes:\n\n  - lwtunnel_input():  seg6, rpl, ila, seg6_local\n  - lwtunnel_output(): ioam6\n  - lwtunnel_xmit():   mpls, LWT BPF xmit\n\nAlternatively, we could clear the metadata right after TC ingress hook.\nThat would require a compromise, however. Metadata would become\ninaccessible from TC egress (in setups where it actually reaches the\nhook it tact, that is without any L2 tunnels on path).","Type":"Description","Title":"net: lwtunnel: Drop skb metadata before LWT encapsulation"}]}}}