{"id":"CVE-2026-99130","kind":"vulnerability","aliases":["CVE-2026-99130"],"related":[],"title":"Linux kernel: io_uring exit can spin a CPU and panic with panic_on_warn","description":"In the Linux kernel, the following vulnerability has been resolved:\n\nio_uring/cancel: don't spin on local work the task can't run\n\nio_uring_cancel_generic() refuses to sleep while any ring on the exiting\ntask's tctx has local work pending. It assumes the task is that ring's\nsubmitter and that the next io_uring_try_cancel_requests() pass will run\nthe work. This does not hold for a ring that was still\nIORING_SETUP_R_DISABLED when the task got its tctx node from\nio_uring_create() or io_ringfd_register(). The submitter is set later by\nIORING_REGISTER_ENABLE_RINGS, possibly from another task, and only the\nsubmitter may run local work on an IORING_SETUP_DEFER_TASKRUN ring. To\nthe exiting task the work stays pending forever, the loop never reaches\nschedule(), and the task trips the WARN_ON_ONCE() once and then spins in\ndo_exit().\n\nAn unprivileged user can set this up at will by creating a disabled\nring, having a child enable it and stop, posting to the ring with\nIORING_OP_MSG_RING and exiting with a request in flight that only the\nchild can complete. The exiting thread burns a CPU until the child runs\nor dies, and with panic_on_warn set the WARN panics the kernel.\n\nFix by skipping the sleep only when io_allowed_defer_tw_run() allows the\ntask to run the work. This is reasonable since work it cannot run cannot\ncomplete any of its requests either, so nothing is lost by sleeping.\nDrop the WARN_ON_ONCE() since handing a disabled ring to another thread\nis a supported scenario, not a bug state.","summary":"A disabled io_uring ring whose submitter is another task can leave an exiting thread spinning in the kernel instead of sleeping. An unprivileged local user can trigger it on purpose, and with panic_on_warn set the kernel panics.","severity":"unknown","cvss":null,"epss":null,"exploited":false,"exploited_at":null,"products":[],"ecosystems":[],"tags":["io_uring","linux-kernel","multi-user-hosts","runtime","self-hosted"],"not_affected_if":["io_uring-disabled","no-untrusted-local-users"],"how_to_check":"- `uname -r` and compare against the kernel version that carries the io_uring/cancel fix (the sources do not list fixed versions)\n- Check whether io_uring is enabled for unprivileged users: `sysctl kernel.io_uring_disabled`\n- Check for panic_on_warn: `sysctl kernel.panic_on_warn`","fix":"- Update the Linux kernel to a release that includes the io_uring/cancel fix (fixed version: unknown in sources)\n- Until patched, consider `sysctl -w kernel.io_uring_disabled=2` if io_uring is not needed","status":"open","published_at":"2026-10-28T11:17:32+00:00","first_seen_at":"2026-10-28T11:32:35+00:00","triaged_at":"2026-10-28T11:36:19+00:00","updated_at":"2026-10-28T11:36:19+00:00","category":"runtime","confidence":0.5,"triage":{"at":"2026-10-28T11:36:19Z","model":"claude-haiku-5-5","output":{"fix":"- Update the Linux kernel to a release that includes the io_uring/cancel fix (fixed version: unknown in sources)\n- Until patched, consider `sysctl -w kernel.io_uring_disabled=2` if io_uring is not needed","reach":"local","title":"Linux kernel: io_uring exit can spin a CPU and panic with panic_on_warn","impact":"low","summary":"A disabled io_uring ring whose submitter is another task can leave an exiting thread spinning in the kernel instead of sleeping. An unprivileged local user can trigger it on purpose, and with panic_on_warn set the kernel panics.","category":"runtime","nicknames":[],"who_cares":["linux-kernel","io_uring","self-hosted","multi-user-hosts"],"confidence":0.5,"needs_auth":"yes","worst_case":"local user causes a CPU spin and kernel panic with panic_on_warn","analyst_note":"Find the affected kernel version ranges and fixed stable releases from the NVD page or the upstream kernel commit; the sources give none.","how_to_check":"- `uname -r` and compare against the kernel version that carries the io_uring/cancel fix (the sources do not list fixed versions)\n- Check whether io_uring is enabled for unprivileged users: `sysctl kernel.io_uring_disabled`\n- Check for panic_on_warn: `sysctl kernel.panic_on_warn`","needs_analyst":true,"not_affected_if":["io_uring-disabled","no-untrusted-local-users"],"who_is_affected":"Machines running a Linux kernel with io_uring, where unprivileged users can create rings, are exposed. Systems that disable io_uring or do not allow untrusted local users can skip it; the sources do not say which distributions or kernel versions are affected.","user_interaction":"no","needs_special_setup":"unknown","severity_if_unknown":"unknown"},"cost_usd":0.000706,"input_tokens":3768,"output_tokens":659,"severity_before":"unknown"},"analyst_status":null,"analyst_note":null,"search":"'-2026':17A,20A '-99130':18A,21A 'allow':302C,306C 'anoth':43B,182C,345C 'assum':119C 'bug':353C 'burn':274C 'cancel':100C,134C 'cannot':319C,321C 'child':244C,268C,279C 'complet':270C,322C 'cpu':9A,276C 'creat':165C,238C 'cve':16A,19A 'defer':196C,303C 'die':282C 'disabl':36B,154C,240C,342C 'drop':334C 'either':327C 'enabl':178C,245C 'exit':5A,48B,110C,201C,227C,259C,272C 'fix':294C 'flight':264C 'follow':80C 'forev':207C 'generic':101C 'got':158C 'hand':340C 'hold':144C 'host':30B,34B 'instead':54B 'io':3A,22B,37B,85C,98C,131C,163C,167C,301C 'ior':151C,176C,194C,254C 'kernel':2A,26B,53B,73B,78C,293C 'later':174C 'leav':46B 'linux':1A,25B,77C 'linux-kernel':24B 'local':59B,91C,115C,190C 'loop':209C 'lost':331C 'may':188C 'msg':256C 'multi':28B 'multi-user-host':27B 'never':210C 'next':130C 'node':161C 'noth':329C 'op':255C 'panic':11A,13A,68B,74B,285C,291C 'pass':136C 'pend':117C,206C 'possibl':180C 'post':249C 'purpos':65B 'r':153C 'reach':211C 'reason':315C 'refus':102C 'regist':169C,177C 'request':135C,262C,326C 'resolv':84C 'ring':39B,107C,124C,147C,179C,198C,241C,252C,257C,343C 'ringfd':168C 'run':97C,138C,189C,280C,305C,310C,320C 'runtim':31B 'scenario':350C 'schedul':212C 'self':33B 'self-host':32B 'set':71B,173C,232C,288C 'setup':152C,195C 'sinc':316C,339C 'skip':296C 'sleep':56B,104C,298C,333C 'spin':7A,50B,89C,224C 'state':354C 'stay':205C 'still':150C 'stop':248C 'submitt':41B,126C,171C,187C 'support':349C 'task':44B,94C,111C,121C,157C,183C,202C,215C,308C 'taskrun':197C 'tctx':113C,160C 'thread':49B,273C,346C 'tri':133C 'trigger':62B 'trip':216C 'tw':304C 'unprivileg':58B,229C 'ure':4A,23B,38B,99C,132C,164C 'uring/cancel':86C 'user':29B,60B,230C 'vulner':81C 'warn':15A,70B,218C,287C,290C,336C 'whose':40B 'work':92C,116C,140C,191C,204C,312C,317C","nicknames":[],"mitigations":[],"who_affected":"Machines running a Linux kernel with io_uring, where unprivileged users can create rings, are exposed. Systems that disable io_uring or do not allow untrusted local users can skip it; the sources do not say which distributions or kernel versions are affected.","tier":"C","verified":{},"analyst_revisit_at":null,"first_public_at":"2026-10-28T11:17:32+00:00","kev":null,"cvss_vector":null,"cvss_by":null,"poc_count":0,"sources":[{"source":"nvd","external_id":"CVE-2026-99130","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-99130","severity":null,"published_at":"2026-10-28T11:17:32+00:00","modified_at":"2026-10-28T11:17:32+00:00","first_seen_at":"2026-10-28T11:32:35+00:00","fetched_at":"2026-10-28T11:32:35+00:00","note":null}],"documents":[],"notice":"Free to read at a human pace. An API with keys and an MCP server are in preparation; early access: hello@infrastack.ai"}