{
  "schema_version": "1.6.1",
  "id": "RPI-CVE-2024-38568",
  "modified": "2026-10-09T11:39:44Z",
  "upstream": [
    "CVE-2024-38568"
  ],
  "published": "2024-06-19T13:36:08Z",
  "summary": "drivers/perf: hisi: hns3: Fix out-of-bound access when valid event group",
  "details": "In the Linux kernel, the following vulnerability has been resolved:\n\ndrivers/perf: hisi: hns3: Fix out-of-bound access when valid event group\n\nThe perf tool allows users to create event groups through following\ncmd [1], but the driver does not check whether the array index is out\nof bounds when writing data to the event_group array. If the number of\nevents in an event_group is greater than HNS3_PMU_MAX_HW_EVENTS, the\nmemory write overflow of event_group array occurs.\n\nAdd array index check to fix the possible array out of bounds violation,\nand return directly when write new events are written to array bounds.\n\nThere are 9 different events in an event_group.\n[1] perf stat -e '{pmu/event1/, ... ,pmu/event9/}",
  "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": [
        "1:6.1.47-1+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": "6.1.93",
            "compiled_filter": false
          }
        }
      }
    }
  ],
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://www.cve.org/CVERecord?id=CVE-2024-38568"
    },
    {
      "type": "FIX",
      "url": "https://git.kernel.org/stable/c/3669baf308308385a2ab391324abdde5682af5aa"
    },
    {
      "type": "FIX",
      "url": "https://git.kernel.org/stable/c/be1fa711e59c874d049f592aef1d4685bdd22bdf"
    },
    {
      "type": "FIX",
      "url": "https://git.kernel.org/stable/c/b5120d322763c15c978bc47beb3b6dff45624304"
    },
    {
      "type": "FIX",
      "url": "https://git.kernel.org/stable/c/aa2d3d678895c8eedd003f1473f87d3f06fe6ec7"
    },
    {
      "type": "FIX",
      "url": "https://git.kernel.org/stable/c/81bdd60a3d1d3b05e6cc6674845afb1694dd3a0e"
    }
  ],
  "database_specific": {
    "source": "https://git.kernel.org/pub/scm/linux/security/vulns.git",
    "upstream_modified": "2026-08-05T11:52:36Z",
    "severity": {
      "source": "kernel CNA",
      "cvss_v3_score": 7.8,
      "cvss_v3_severity": "High",
      "scenario": "AV:L - The vulnerability is reached solely through the local `perf_event_open(2)` syscall against the HNS3 PMU device exposed under /sys/bus/event_source/devices/; no network or remote input path exists. Attack Vector is therefore Local.\nAC:L - Exploitation is fully deterministic and requires no race, no memory-layout guessing, and no victim state \u2014 the attacker simply opens nine events in one group (`perf stat -e '{pmu/event1/,...,pmu/event9/}'`, the reproducer named in the fix), using `attr.disabled=1` to keep the hardware counters unallocated so the `-EBUSY` pre-check is bypassed. Every condition is under the attacker's direct control.\nPR:L - An unprivileged local account suffices on the realistic deployment for this driver \u2014 HiSilicon Kunpeng ARM64 servers with NIC PMU profiling enabled, where `kernel.perf_event_paranoid` is set to 0/-1 or CAP_PERFMON is delegated to an observability agent; CAP_PERFMON is explicitly a de-privileged, delegatable capability rather than real root/admin. No authentication beyond a local login is needed, and the OOB write itself executes in `event_init()` before `perf_allow_cpu()` is ever consulted.\nUI:N - The attacker triggers the entire sequence from its own process via repeated `perf_event_open()` calls. No action by any other user or administrator is involved.\nS:U - The out-of-bounds write corrupts the kernel stack of the calling task within the same kernel security authority; no VM, IOMMU, or sandbox boundary is crossed.\nC:H - The bug is kernel-stack memory corruption whose written value is a slab pointer to an attacker-groomed `struct perf_event`, landing on adjacent frame state (canary/saved-register slot) that can be leveraged to influence subsequent kernel execution and disclose kernel memory. Per the memory-corruption guidance this is scored High.\nI:H - This is an out-of-bounds *write* past a stack array, and the attacker controls which `perf_event` object's address is written by ordering the group's event allocations, giving corruption of control-relevant stack data. Out-of-bounds writes are scored High.\nA:H - On kernels built with CONFIG_STACKPROTECTOR_STRONG (standard for arm64 distro/server kernels) the overwritten slot is the stack canary, so `__stack_chk_fail()` invokes `panic()` unconditionally \u2014 a guaranteed, repeatable full system crash; without the protector it corrupts saved frame state and oopses."
    }
  }
}