{"cve":"CVE-2025-39821","nvd":{"published":"2025-09-16T13:15:59.300","modified":"2026-10-19T21:10:00.160","status":"Modified","description":"In the Linux kernel, the following vulnerability has been resolved:\n\nperf: Avoid undefined behavior from stopping/starting inactive events\n\nCalling pmu->start()/stop() on perf events in PERF_EVENT_STATE_OFF can\nleave event->hw.idx at -1. When PMU drivers later attempt to use this\nnegative index as a shift exponent in bitwise operations, it leads to UBSAN\nshift-out-of-bounds reports.\n\nThe issue is a logical flaw in how event groups handle throttling when some\nmembers are intentionally disabled. Based on the analysis and the\nreproducer provided by Mark Rutland (this issue on both arm64 and x86-64).\n\nThe scenario unfolds as follows:\n\n 1. A group leader event is configured with a very aggressive sampling\n    period (e.g., sample_period = 1). This causes frequent interrupts and\n    triggers the throttling mechanism.\n 2. A child event in the same group is created in a disabled state\n    (.disabled = 1). This event remains in PERF_EVENT_STATE_OFF.\n    Since it hasn't been scheduled onto the PMU, its event->hw.idx remains\n    initialized at -1.\n 3. When throttling occurs, perf_event_throttle_group() and later\n    perf_event_unthrottle_group() iterate through all siblings, including\n    the disabled child event.\n 4. perf_event_throttle()/unthrottle() are called on this inactive child\n    event, which then call event->pmu->start()/stop().\n 5. The PMU driver receives the event with hw.idx == -1 and attempts to\n    use it as a shift exponent. e.g., in macros like PMCNTENSET(idx),\n    leading to the UBSAN report.\n\nThe throttling mechanism attempts to start/stop events that are not\nactively scheduled on the hardware.\n\nMove the state check into perf_event_throttle()/perf_event_unthrottle() so\nthat inactive events are skipped entirely. This ensures only active events\nwith a valid hw.idx are processed, preventing undefined behavior and\nsilencing UBSAN warnings. The corrected check ensures true before\nproceeding with PMU operations.\n\nThe problem can be reproduced with the syzkaller reproducer:","cvss_v4":null,"cvss_v3":{"version":"3.1","vector":"CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H","score":7.8},"cwes":["CWE-787"],"references":["https://git.kernel.org/stable/c/b64fdd422a85025b5e91ead794db9d3ef970e369","https://git.kernel.org/stable/c/d689135aa9c5e4e0eab5a92bbe35dab0c8d6677f"],"products":["linux:linux_kernel"],"source":"nvd","synced_at":"2026-10-21T03:43:09.190223697+00:00"},"epss":{"cve":"CVE-2025-39821","score":0.00164,"percentile":0.05523,"date":"2026-10-25"},"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."]}