)]}'
{
  "commit": "787268686c64be60023743b527645a907e38ccca",
  "tree": "ce307ad5d1d35014218ce1a50db26cb9e0884418",
  "parents": [
    "f19f2b308beb158d0e6f6eb806ced636e4ebf6ce"
  ],
  "author": {
    "name": "Eric Dumazet",
    "email": "edumazet@kernel.org",
    "time": "Fri Oct 30 10:23:03 2026 +0100"
  },
  "committer": {
    "name": "Paolo Abeni",
    "email": "pabeni@redhat.com",
    "time": "Tue Nov 03 22:30:55 2026 +0100"
  },
  "message": "ipmr, ip6mr: queue netlink route queries out of band\n\nipmr_get_route() and ip6mr_get_route() queue synthetic skbs on the\nunresolved MFC entry to answer RTM_GETROUTE requests once the route\nis resolved. These skbs carry a fake IP header with version 0, and\nipmr_cache_resolve(), ipmr_destroy_unres(), ip6mr_cache_resolve()\nand ip6mr_destroy_unres() rely on this in-band tag to tell them\napart from regular packets queued by the input and output paths.\n\nThe IP version is only validated once, in ip_rcv_core() and\nip6_rcv_core(). A netfilter hook running later, for instance an\nnftables payload statement in a PREROUTING chain owned by an\nunprivileged user namespace, can rewrite the first header byte to\nzero. The SRv6 End.DX6 action also passes the decapsulated packet\nto ip6_mr_input() without checking that the inner header version\nis 6. If no MFC entry exists, the packet is queued and later handled\nas a netlink request: its payload is treated as an nlmsghdr and\nwritten to, mr_fill_mroute() may call skb_put() on a nonlinear skb\n(hitting SKB_LINEAR_ASSERT()), and the skb is unicast to a netlink\nportid taken from stale IPCB/IP6CB bytes in skb-\u003ecb[].\n\nipmr_cache_report() also relies on the same tag since\ncommit 94a6736bb5ab (\"ipmr: only copy pktinfo to cache reports\").\n\nStop trusting packet contents: queue route queries on a dedicated\nnl_queries list in the unresolved MFC entry, and pass an explicit\nnl_query flag from the callers.\n\nThe per-entry limit of four queued skbs now applies to both lists\ncombined, as before. The synthetic header keeps version 0, since it\nis copied into the NOCACHE report delivered to the mroute socket.\n\nFixes: 1da177e4c3f4 (\"Linux-2.6.12-rc2\")\nFixes: 7bc570c8b4f7 (\"[IPV6] MROUTE: Support multicast forwarding.\")\nReported-by: Katherine Leaver \u003ckleaver@janestreet.com\u003e\nSigned-off-by: Eric Dumazet \u003cedumazet@kernel.org\u003e\nReviewed-by: Ido Schimmel \u003cidosch@nvidia.com\u003e\nLink: https://patch.msgid.link/20261030092303.2749386-1-edumazet@kernel.org\nSigned-off-by: Paolo Abeni \u003cpabeni@redhat.com\u003e\n",
  "tree_diff": [
    {
      "type": "modify",
      "old_id": "4d55827e970537aff6938503a39ee3306c8fde6d",
      "old_mode": 33188,
      "old_path": "include/linux/mroute_base.h",
      "new_id": "2d506966357b2591af4eadef88b27e236a2fe9cf",
      "new_mode": 33188,
      "new_path": "include/linux/mroute_base.h"
    },
    {
      "type": "modify",
      "old_id": "e8f7d3b573d0f4f74ebabc3425993b8d39fdfd25",
      "old_mode": 33188,
      "old_path": "net/ipv4/ipmr.c",
      "new_id": "836f3f95b20c45f70f659379656761a18d220b93",
      "new_mode": 33188,
      "new_path": "net/ipv4/ipmr.c"
    },
    {
      "type": "modify",
      "old_id": "af968a2c2a7847c96d50774729d8aee23df5294a",
      "old_mode": 33188,
      "old_path": "net/ipv6/ip6mr.c",
      "new_id": "9710e8958b3892403e3cac1102554eea4d0a58fe",
      "new_mode": 33188,
      "new_path": "net/ipv6/ip6mr.c"
    }
  ]
}
