{
  "schema_version": "1.6.1",
  "id": "RPI-CVE-2023-53065",
  "modified": "2026-10-09T11:39:44Z",
  "upstream": [
    "CVE-2023-53065"
  ],
  "published": "2025-05-02T15:56:53Z",
  "summary": "perf/core: Fix perf_output_begin parameter is incorrectly invoked in perf_event_bpf_output",
  "details": "In the Linux kernel, the following vulnerability has been resolved:\n\nperf/core: Fix perf_output_begin parameter is incorrectly invoked in perf_event_bpf_output\n\nsyzkaller reportes a KASAN issue with stack-out-of-bounds.\nThe call trace is as follows:\n  dump_stack+0x9c/0xd3\n  print_address_description.constprop.0+0x19/0x170\n  __kasan_report.cold+0x6c/0x84\n  kasan_report+0x3a/0x50\n  __perf_event_header__init_id+0x34/0x290\n  perf_event_header__init_id+0x48/0x60\n  perf_output_begin+0x4a4/0x560\n  perf_event_bpf_output+0x161/0x1e0\n  perf_iterate_sb_cpu+0x29e/0x340\n  perf_iterate_sb+0x4c/0xc0\n  perf_event_bpf_event+0x194/0x2c0\n  __bpf_prog_put.constprop.0+0x55/0xf0\n  __cls_bpf_delete_prog+0xea/0x120 [cls_bpf]\n  cls_bpf_delete_prog_work+0x1c/0x30 [cls_bpf]\n  process_one_work+0x3c2/0x730\n  worker_thread+0x93/0x650\n  kthread+0x1b8/0x210\n  ret_from_fork+0x1f/0x30\n\ncommit 267fb27352b6 (\"perf: Reduce stack usage of perf_output_begin()\")\nuse on-stack struct perf_sample_data of the caller function.\n\nHowever, perf_event_bpf_output uses incorrect parameter to convert\nsmall-sized data (struct perf_bpf_event) into large-sized data\n(struct perf_sample_data), which causes memory overwriting occurs in\n__perf_event_header__init_id.",
  "severity": [
    {
      "type": "CVSS_V3",
      "score": "CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H"
    }
  ],
  "affected": [
    {
      "package": {
        "ecosystem": "Raspberry Pi OS:12",
        "name": "linux"
      },
      "versions": [
        "1:6.1.21-1+rpt1"
      ],
      "ecosystem_specific": {
        "binaries": [
          {
            "binary_name": "linux-image-6.1.0-rpi1-rpi-v6",
            "binary_version": "1:6.1.21-1+rpt1"
          },
          {
            "binary_name": "linux-image-6.1.0-rpi1-rpi-v7",
            "binary_version": "1:6.1.21-1+rpt1"
          },
          {
            "binary_name": "linux-image-6.1.0-rpi1-rpi-v7l",
            "binary_version": "1:6.1.21-1+rpt1"
          },
          {
            "binary_name": "linux-image-6.1.0-rpi1-rpi-v8",
            "binary_version": "1:6.1.21-1+rpt1"
          }
        ]
      },
      "database_specific": {
        "per_version": {
          "1:6.1.21-1+rpt1": {
            "linux_commit": "0afb5e98488aed7017b9bf321b575d0177feb7ed",
            "upstream_version": "6.1.21",
            "assessment": "commit ancestry (strak rule)",
            "upstream_fixed": "6.1.22"
          }
        }
      }
    }
  ],
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://www.cve.org/CVERecord?id=CVE-2023-53065"
    },
    {
      "type": "FIX",
      "url": "https://git.kernel.org/stable/c/ddcf8320003638a06eb1e46412e045d0c5701575"
    },
    {
      "type": "FIX",
      "url": "https://git.kernel.org/stable/c/ac5f88642cb211152041f84a985309e9af4baf59"
    },
    {
      "type": "FIX",
      "url": "https://git.kernel.org/stable/c/ff8137727a2af4ad5f6e6c8b9f7ec5e8db9da86c"
    },
    {
      "type": "FIX",
      "url": "https://git.kernel.org/stable/c/3a776fddb4e5598c8bfcd4ad094fba34f9856fc9"
    },
    {
      "type": "FIX",
      "url": "https://git.kernel.org/stable/c/eb81a2ed4f52be831c9fb879752d89645a312c13"
    }
  ],
  "database_specific": {
    "source": "https://git.kernel.org/pub/scm/linux/security/vulns.git",
    "upstream_modified": "2026-08-05T11:24:39Z",
    "severity": {
      "source": "kernel CNA",
      "cvss_v3_score": 7.8,
      "cvss_v3_severity": "High",
      "scenario": "AV:L - The bug is reached via local perf_event_open (CPU-wide bpf_event sideband) plus a local BPF unload/load trigger (bpf() or tc cls_bpf netlink); there is no network-facing path into perf_event_bpf_output.\nAC:L - An attacker who opens the bpf_event with sample_id_all, maps a small ring buffer, stops consuming it to force rb->lost, then loads/unloads a BPF program reliably hits the LOST path that performs the stack overflow; no attacker-uncontrollable race or layout is required.\nPR:L - Opening the required CPU-wide bpf_event needs CAP_PERFMON (perf_allow_cpu), and triggering BPF unload needs CAP_BPF or userns-reachable CAP_NET_ADMIN for tc/cls_bpf; these are delegatable non-root capabilities routinely granted to observability agents/containers, matching established PR:L scoring for CAP_PERFMON/CAP_BPF bugs.\nUI:N - The attacker opens the perf event, fills the ring buffer, and triggers BPF program unload entirely from their own processes; no separate victim action is required.\nS:U - The stack corruption and any resulting privilege escalation occur within the host kernel\u2019s security authority and do not cross a VM, IOMMU, or sandbox boundary.\nC:H - This is an attacker-influenced kernel stack out-of-bounds write (perf_bpf_event treated as the much larger perf_sample_data); such memory corruption can be leveraged for arbitrary kernel read primitives.\nI:H - Overflowing past the on-stack perf_bpf_event in perf_event_bpf_event smashes saved frame/return state under attacker-controlled sample_type write patterns, enabling control-flow hijack and arbitrary kernel write/code execution.\nA:H - The same stack-out-of-bounds write KASAN reported causes kernel oops/panic even when not fully weaponized, which is High availability impact."
    }
  }
}