{"cve":"CVE-2023-52587","nvd":{"published":"2024-03-06T07:15:07.633","modified":"2026-06-17T06:43:05.583","status":"Analyzed","description":"In the Linux kernel, the following vulnerability has been resolved:\n\nIB/ipoib: Fix mcast list locking\n\nReleasing the `priv->lock` while iterating the `priv->multicast_list` in\n`ipoib_mcast_join_task()` opens a window for `ipoib_mcast_dev_flush()` to\nremove the items while in the middle of iteration. If the mcast is removed\nwhile the lock was dropped, the for loop spins forever resulting in a hard\nlockup (as was reported on RHEL 4.18.0-372.75.1.el8_6 kernel):\n\n    Task A (kworker/u72:2 below)       | Task B (kworker/u72:0 below)\n    -----------------------------------+-----------------------------------\n    ipoib_mcast_join_task(work)        | ipoib_ib_dev_flush_light(work)\n      spin_lock_irq(&priv->lock)       | __ipoib_ib_dev_flush(priv, ...)\n      list_for_each_entry(mcast,       | ipoib_mcast_dev_flush(dev = priv->dev)\n          &priv->multicast_list, list) |\n        ipoib_mcast_join(dev, mcast)   |\n          spin_unlock_irq(&priv->lock) |\n                                       |   spin_lock_irqsave(&priv->lock, flags)\n                                       |   list_for_each_entry_safe(mcast, tmcast,\n                                       |                  &priv->multicast_list, list)\n                                       |     list_del(&mcast->list);\n                                       |     list_add_tail(&mcast->list, &remove_list)\n                                       |   spin_unlock_irqrestore(&priv->lock, flags)\n          spin_lock_irq(&priv->lock)   |\n                                       |   ipoib_mcast_remove_list(&remove_list)\n   (Here, `mcast` is no longer on the  |     list_for_each_entry_safe(mcast, tmcast,\n    `priv->multicast_list` and we keep |                            remove_list, list)\n    spinning on the `remove_list` of   |  >>>  wait_for_completion(&mcast->done)\n    the other thread which is blocked  |\n    and the list is still valid on     |\n    it's stack.)\n\nFix this by keeping the lock held and changing to GFP_ATOMIC to prevent\neventual sleeps.\nUnfortunately we could not reproduce the lockup and confirm this fix but\nbased on the code review I think this fix should address such lockups.\n\ncrash> bc 31\nPID: 747      TASK: ff1c6a1a007e8000  CPU: 31   COMMAND: \"kworker/u72:2\"\n--\n    [exception RIP: ipoib_mcast_join_task+0x1b1]\n    RIP: ffffffffc0944ac1  RSP: ff646f199a8c7e00  RFLAGS: 00000002\n    RAX: 0000000000000000  RBX: ff1c6a1a04dc82f8  RCX: 0000000000000000\n                                  work (&priv->mcast_task{,.work})\n    RDX: ff1c6a192d60ac68  RSI: 0000000000000286  RDI: ff1c6a1a04dc8000\n           &mcast->list\n    RBP: ff646f199a8c7e90   R8: ff1c699980019420   R9: ff1c6a1920c9a000\n    R10: ff646f199a8c7e00  R11: ff1c6a191a7d9800  R12: ff1c6a192d60ac00\n                                                         mcast\n    R13: ff1c6a1d82200000  R14: ff1c6a1a04dc8000  R15: ff1c6a1a04dc82d8\n           dev                    priv (&priv->lock)     &priv->multicast_list (aka head)\n    ORIG_RAX: ffffffffffffffff  CS: 0010  SS: 0018\n--- <NMI exception stack> ---\n #5 [ff646f199a8c7e00] ipoib_mcast_join_task+0x1b1 at ffffffffc0944ac1 [ib_ipoib]\n #6 [ff646f199a8c7e98] process_one_work+0x1a7 at ffffffff9bf10967\n\ncrash> rx ff646f199a8c7e68\nff646f199a8c7e68:  ff1c6a1a04dc82f8 <<< work = &priv->mcast_task.work\n\ncrash> list -hO ipoib_dev_priv.multicast_list ff1c6a1a04dc8000\n(empty)\n\ncrash> ipoib_dev_priv.mcast_task.work.func,mcast_mutex.owner.counter ff1c6a1a04dc8000\n  mcast_task.work.func = 0xffffffffc0944910 <ipoib_mcast_join_task>,\n  mcast_mutex.owner.counter = 0xff1c69998efec000\n\ncrash> b 8\nPID: 8        TASK: ff1c69998efec000  CPU: 33   COMMAND: \"kworker/u72:0\"\n--\n #3 [ff646f1980153d50] wait_for_completion+0x96 at ffffffff9c7d7646\n #4 [ff646f1980153d90] ipoib_mcast_remove_list+0x56 at ffffffffc0944dc6 [ib_ipoib]\n #5 [ff646f1980153de8] ipoib_mcast_dev_flush+0x1a7 at ffffffffc09455a7 [ib_ipoib]\n #6 [ff646f1980153e58] __ipoib_ib_dev_flush+0x1a4 at ffffffffc09431a4 [ib_ipoib]\n #7 [ff\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/342258fb46d66c1b4c7e2c3717ac01e10c03cf18","https://git.kernel.org/stable/c/4c8922ae8eb8dcc1e4b7d1059d97a8334288d825","https://git.kernel.org/stable/c/4f973e211b3b1c6d36f7c6a19239d258856749f9","https://git.kernel.org/stable/c/5108a2dc2db5630fb6cd58b8be80a0c134bc310a","https://git.kernel.org/stable/c/615e3adc2042b7be4ad122a043fc9135e6342c90"],"products":["linux:linux_kernel"],"source":"nvd","synced_at":"2026-10-19T21:46:54.846886191+00:00"},"epss":{"cve":"CVE-2023-52587","score":0.00309,"percentile":0.23784,"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."]}