{"cve":"CVE-2024-46797","nvd":{"published":"2024-09-18T08:15:06.403","modified":"2026-06-17T07:55:57.253","status":"Modified","description":"In the Linux kernel, the following vulnerability has been resolved:\n\npowerpc/qspinlock: Fix deadlock in MCS queue\n\nIf an interrupt occurs in queued_spin_lock_slowpath() after we increment\nqnodesp->count and before node->lock is initialized, another CPU might\nsee stale lock values in get_tail_qnode(). If the stale lock value happens\nto match the lock on that CPU, then we write to the \"next\" pointer of\nthe wrong qnode. This causes a deadlock as the former CPU, once it becomes\nthe head of the MCS queue, will spin indefinitely until it's \"next\" pointer\nis set by its successor in the queue.\n\nRunning stress-ng on a 16 core (16EC/16VP) shared LPAR, results in\noccasional lockups similar to the following:\n\n   $ stress-ng --all 128 --vm-bytes 80% --aggressive \\\n               --maximize --oomable --verify  --syslog \\\n               --metrics  --times  --timeout 5m\n\n   watchdog: CPU 15 Hard LOCKUP\n   ......\n   NIP [c0000000000b78f4] queued_spin_lock_slowpath+0x1184/0x1490\n   LR [c000000001037c5c] _raw_spin_lock+0x6c/0x90\n   Call Trace:\n    0xc000002cfffa3bf0 (unreliable)\n    _raw_spin_lock+0x6c/0x90\n    raw_spin_rq_lock_nested.part.135+0x4c/0xd0\n    sched_ttwu_pending+0x60/0x1f0\n    __flush_smp_call_function_queue+0x1dc/0x670\n    smp_ipi_demux_relaxed+0xa4/0x100\n    xive_muxed_ipi_action+0x20/0x40\n    __handle_irq_event_percpu+0x80/0x240\n    handle_irq_event_percpu+0x2c/0x80\n    handle_percpu_irq+0x84/0xd0\n    generic_handle_irq+0x54/0x80\n    __do_irq+0xac/0x210\n    __do_IRQ+0x74/0xd0\n    0x0\n    do_IRQ+0x8c/0x170\n    hardware_interrupt_common_virt+0x29c/0x2a0\n   --- interrupt: 500 at queued_spin_lock_slowpath+0x4b8/0x1490\n   ......\n   NIP [c0000000000b6c28] queued_spin_lock_slowpath+0x4b8/0x1490\n   LR [c000000001037c5c] _raw_spin_lock+0x6c/0x90\n   --- interrupt: 500\n    0xc0000029c1a41d00 (unreliable)\n    _raw_spin_lock+0x6c/0x90\n    futex_wake+0x100/0x260\n    do_futex+0x21c/0x2a0\n    sys_futex+0x98/0x270\n    system_call_exception+0x14c/0x2f0\n    system_call_vectored_common+0x15c/0x2ec\n\nThe following code flow illustrates how the deadlock occurs.\nFor the sake of brevity, assume that both locks (A and B) are\ncontended and we call the queued_spin_lock_slowpath() function.\n\n        CPU0                                   CPU1\n        ----                                   ----\n  spin_lock_irqsave(A)                          |\n  spin_unlock_irqrestore(A)                     |\n    spin_lock(B)                                |\n         |                                      |\n         ▼                                      |\n   id = qnodesp->count++;                       |\n  (Note that nodes[0].lock == A)                |\n         |                                      |\n         ▼                                      |\n      Interrupt                                 |\n  (happens before \"nodes[0].lock = B\")          |\n         |                                      |\n         ▼                                      |\n  spin_lock_irqsave(A)                          |\n         |                                      |\n         ▼                                      |\n   id = qnodesp->count++                        |\n   nodes[1].lock = A                            |\n         |                                      |\n         ▼                                      |\n  Tail of MCS queue                             |\n         |                             spin_lock_irqsave(A)\n         ▼                                      |\n  Head of MCS queue                             ▼\n         |                             CPU0 is previous tail\n         ▼                                      |\n   Spin indefinitely                            ▼\n  (until \"nodes[1].next != NULL\")      prev = get_tail_qnode(A, CPU0)\n                                                |\n                                                ▼\n                                       prev == &qnodes[CPU0].nodes[0]\n                                     (as qnodes\n---truncated---","cvss_v4":null,"cvss_v3":{"version":"3.1","vector":"CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H","score":5.5},"cwes":["CWE-667"],"references":["https://git.kernel.org/stable/c/734ad0af3609464f8f93e00b6c0de1e112f44559","https://git.kernel.org/stable/c/d84ab6661e8d09092de9b034b016515ef9b66085","https://git.kernel.org/stable/c/f06af737e4be28c0e926dc25d5f0a111da4e2987"],"products":["linux:linux_kernel"],"source":"nvd","synced_at":"2026-10-19T21:54:38.012257375+00:00"},"epss":{"cve":"CVE-2024-46797","score":0.00161,"percentile":0.05313,"date":"2026-10-31"},"kev":null,"osv":[],"attribution":["This product uses the NVD API but is not endorsed or certified by the NVD.","EPSS scores: FIRST Exploit Prediction Scoring System, https://www.first.org/epss","Known exploited vulnerabilities: CISA Known Exploited Vulnerabilities Catalog, https://www.cisa.gov/known-exploited-vulnerabilities-catalog","Package advisories: OSV (https://osv.dev), from the GitHub Advisory Database, PyPA, Go, RustSec and other ecosystem databases, under CC-BY 4.0 or CC0","CVE records: the CVE Program's CVE List (https://www.cve.org), under its terms of use (https://www.cve.org/Legal/TermsOfUse). CVE is a registered trademark of The MITRE Corporation; this product is not affiliated with the CVE Program."]}