{
  "schema_version": "1.6.1",
  "id": "RPI-CVE-2026-31649",
  "modified": "2026-10-09T11:39:44Z",
  "upstream": [
    "CVE-2026-31649"
  ],
  "published": "2026-04-24T14:47:31Z",
  "summary": "net: stmmac: fix integer underflow in chain mode",
  "details": "In the Linux kernel, the following vulnerability has been resolved:\n\nnet: stmmac: fix integer underflow in chain mode\n\nThe jumbo_frm() chain-mode implementation unconditionally computes\n\n    len = nopaged_len - bmax;\n\nwhere nopaged_len = skb_headlen(skb) (linear bytes only) and bmax is\nBUF_SIZE_8KiB or BUF_SIZE_2KiB.  However, the caller stmmac_xmit()\ndecides to invoke jumbo_frm() based on skb->len (total length including\npage fragments):\n\n    is_jumbo = stmmac_is_jumbo_frm(priv, skb->len, enh_desc);\n\nWhen a packet has a small linear portion (nopaged_len <= bmax) but a\nlarge total length due to page fragments (skb->len > bmax), the\nsubtraction wraps as an unsigned integer, producing a huge len value\n(~0xFFFFxxxx).  This causes the while (len != 0) loop to execute\nhundreds of thousands of iterations, passing skb->data + bmax * i\npointers far beyond the skb buffer to dma_map_single().  On IOMMU-less\nSoCs (the typical deployment for stmmac), this maps arbitrary kernel\nmemory to the DMA engine, constituting a kernel memory disclosure and\npotential memory corruption from hardware.\n\nFix this by introducing a buf_len local variable clamped to\nmin(nopaged_len, bmax).  Computing len = nopaged_len - buf_len is then\nalways safe: it is zero when the linear portion fits within a single\ndescriptor, causing the while (len != 0) loop to be skipped naturally,\nand the fragment loop in stmmac_xmit() handles page fragments afterward.",
  "severity": [
    {
      "type": "CVSS_V3",
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H"
    }
  ],
  "affected": [
    {
      "package": {
        "ecosystem": "Raspberry Pi OS:12",
        "name": "linux"
      },
      "versions": [
        "1:6.1.47-1+rpt4"
      ],
      "ecosystem_specific": {
        "binaries": [
          {
            "binary_name": "linux-image-6.1.0-rpi3-rpi-2712",
            "binary_version": "1:6.1.47-1+rpt4"
          },
          {
            "binary_name": "linux-image-6.1.0-rpi3-rpi-v6",
            "binary_version": "1:6.1.47-1+rpt4"
          },
          {
            "binary_name": "linux-image-6.1.0-rpi3-rpi-v7",
            "binary_version": "1:6.1.47-1+rpt4"
          },
          {
            "binary_name": "linux-image-6.1.0-rpi3-rpi-v7l",
            "binary_version": "1:6.1.47-1+rpt4"
          },
          {
            "binary_name": "linux-image-6.1.0-rpi3-rpi-v8",
            "binary_version": "1:6.1.47-1+rpt4"
          }
        ]
      },
      "database_specific": {
        "per_version": {
          "1:6.1.47-1+rpt4": {
            "linux_commit": "655fc658a15ae7a6f37103754adb39ba52a9a14e",
            "upstream_version": "6.1.47",
            "assessment": "commit ancestry (strak rule)",
            "upstream_fixed": "6.1.169",
            "compiled_filter": false
          }
        }
      }
    }
  ],
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://www.cve.org/CVERecord?id=CVE-2026-31649"
    },
    {
      "type": "FIX",
      "url": "https://git.kernel.org/stable/c/513e06735f5be575b409d195822195348b164e48"
    },
    {
      "type": "FIX",
      "url": "https://git.kernel.org/stable/c/275bdf762e82082f064e60a92448fa2ac43cf95b"
    },
    {
      "type": "FIX",
      "url": "https://git.kernel.org/stable/c/a2b68a9a476b9544ff31f1fbcd5d80867a8a5e2f"
    },
    {
      "type": "FIX",
      "url": "https://git.kernel.org/stable/c/b7b8012193fd98236d7ae05d4b553f010a77b2ef"
    },
    {
      "type": "FIX",
      "url": "https://git.kernel.org/stable/c/2c91b39912278d0878f9ba60ba04d2518b18a08d"
    },
    {
      "type": "FIX",
      "url": "https://git.kernel.org/stable/c/6fca757c20396dc2e604dcc61922264e9e3dc803"
    },
    {
      "type": "FIX",
      "url": "https://git.kernel.org/stable/c/10d12b9240ebf96c785f0e2e4228318cd5f3a3eb"
    },
    {
      "type": "FIX",
      "url": "https://git.kernel.org/stable/c/51f4e090b9f87b40c21b6daadb5c06e6c0a07b67"
    }
  ],
  "database_specific": {
    "source": "https://git.kernel.org/pub/scm/linux/security/vulns.git",
    "upstream_modified": "2026-08-05T13:16:28Z",
    "severity": {
      "source": "kernel CNA",
      "cvss_v3_score": 9.8,
      "cvss_v3_severity": "Critical",
      "scenario": "AV:N - The vulnerable stmmac transmit path can be reached over the network when an unauthenticated peer causes a network-facing service or forwarding/router workload to transmit large non-linear skbs through an affected stmmac interface. This is a plausible deployment for embedded/router/NFS/TLS systems using this Ethernet MAC.\nAC:L - Once the system is using affected chain mode, the trigger is a non-linear packet whose total length exceeds the jumbo threshold while the linear head does not, which common scatter-gather/sendpage paths can produce. There is no race or attacker-uncontrolled timing requirement.\nPR:N - In the highest reasonable scenario, a remote unauthenticated client can trigger large page-backed responses or forwarded traffic through the vulnerable device. No local account or administrative capability is required for that network-triggered path.\nUI:N - No victim user action is needed beyond the system already running the affected driver and network service or forwarding configuration. The attacker can trigger transmission by network interaction.\nS:U - The vulnerability is in the host kernel network driver and impacts the same host kernel security authority. It does not cross a VM, IOMMU, or other separate security scope boundary by itself.\nC:H - The underflow makes the driver DMA-map addresses far beyond the skb buffer, exposing arbitrary kernel memory to the DMA engine and potentially onto the wire. That supports high confidentiality impact.\nI:H - The erroneous descriptor programming gives hardware DMA access to unintended kernel memory ranges and the fix description identifies potential memory corruption from hardware. Under the required tie-breaker, this is scored as high integrity impact.\nA:H - The huge wrapped loop and invalid DMA mappings can exhaust/overwrite TX descriptors, hit DMA/IOMMU faults, hang transmission, or crash the kernel. This is high availability impact."
    }
  }
}
