{ "schema_version": "1.6.1", "id": "RPI-CVE-2026-46304", "modified": "2026-10-09T11:39:44Z", "upstream": [ "CVE-2026-46304" ], "published": "2026-06-08T15:46:07Z", "summary": "nvmet: avoid recursive nvmet-wq flush in nvmet_ctrl_free", "details": "In the Linux kernel, the following vulnerability has been resolved:\n\nnvmet: avoid recursive nvmet-wq flush in nvmet_ctrl_free\n\nnvmet_tcp_release_queue_work() runs on nvmet-wq and can drop the\nfinal controller reference through nvmet_cq_put(). If that triggers\nnvmet_ctrl_free(), the teardown path flushes ctrl->async_event_work on\nthe same nvmet-wq.\n\nCall chain:\n\n nvmet_tcp_schedule_release_queue()\n kref_put(&queue->kref, nvmet_tcp_release_queue)\n nvmet_tcp_release_queue()\n queue_work(nvmet_wq, &queue->release_work) <--- nvmet_wq\n process_one_work()\n nvmet_tcp_release_queue_work()\n nvmet_cq_put(&queue->nvme_cq)\n nvmet_cq_destroy()\n nvmet_ctrl_put(cq->ctrl)\n nvmet_ctrl_free()\n flush_work(&ctrl->async_event_work) <--- nvmet_wq\n\n Previously Scheduled by :-\n\t\t nvmet_add_async_event\n\t\t queue_work(nvmet_wq, &ctrl->async_event_work);\n\nThis trips lockdep with a possible recursive locking warning.\n\n[ 5223.015876] run blktests nvme/003 at 2026-04-07 20:53:55\n[ 5223.061801] loop0: detected capacity change from 0 to 2097152\n[ 5223.072206] nvmet: adding nsid 1 to subsystem blktests-subsystem-1\n[ 5223.088368] nvmet_tcp: enabling port 0 (127.0.0.1:4420)\n[ 5223.126086] nvmet: Created discovery controller 1 for subsystem nqn.2014-08.org.nvmexpress.discovery for NQN nqn.2014-08.org.nvmexpress:uuid:0f01fb42-9f7f-4856-b0b3-51e60b8de349.\n[ 5223.128453] nvme nvme1: new ctrl: NQN \"nqn.2014-08.org.nvmexpress.discovery\", addr 127.0.0.1:4420, hostnqn: nqn.2014-08.org.nvmexpress:uuid:0f01fb42-9f7f-4856-b0b3-51e60b8de349\n[ 5233.199447] nvme nvme1: Removing ctrl: NQN \"nqn.2014-08.org.nvmexpress.discovery\"\n\n[ 5233.227718] ============================================\n[ 5233.231283] WARNING: possible recursive locking detected\n[ 5233.234696] 7.0.0-rc3nvme+ #20 Tainted: G O N\n[ 5233.238434] --------------------------------------------\n[ 5233.241852] kworker/u192:6/2413 is trying to acquire lock:\n[ 5233.245429] ffff888111632548 ((wq_completion)nvmet-wq){+.+.}-{0:0}, at: touch_wq_lockdep_map+0x26/0x90\n[ 5233.251438]\n but task is already holding lock:\n[ 5233.255254] ffff888111632548 ((wq_completion)nvmet-wq){+.+.}-{0:0}, at: process_one_work+0x5cc/0x6e0\n[ 5233.261125]\n other info that might help us debug this:\n[ 5233.265333] Possible unsafe locking scenario:\n\n[ 5233.269217] CPU0\n[ 5233.270795] ----\n[ 5233.272436] lock((wq_completion)nvmet-wq);\n[ 5233.275241] lock((wq_completion)nvmet-wq);\n[ 5233.278020]\n *** DEADLOCK ***\n\n[ 5233.281793] May be due to missing lock nesting notation\n\n[ 5233.286195] 3 locks held by kworker/u192:6/2413:\n[ 5233.289192] #0: ffff888111632548 ((wq_completion)nvmet-wq){+.+.}-{0:0}, at: process_one_work+0x5cc/0x6e0\n[ 5233.294569] #1: ffffc9000e2a7e40 ((work_completion)(&queue->release_work)){+.+.}-{0:0}, at: process_one_work+0x1c5/0x6e0\n[ 5233.300128] #2: ffffffff82d7dc40 (rcu_read_lock){....}-{1:3}, at: __flush_work+0x62/0x530\n[ 5233.304290]\n stack backtrace:\n[ 5233.306520] CPU: 4 UID: 0 PID: 2413 Comm: kworker/u192:6 Tainted: G O N 7.0.0-rc3nvme+ #20 PREEMPT(full)\n[ 5233.306524] Tainted: [O]=OOT_MODULE, [N]=TEST\n[ 5233.306525] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.17.0-0-gb52ca86e094d-prebuilt.qemu.org 04/01/2014\n[ 5233.306527] Workqueue: nvmet-wq nvmet_tcp_release_queue_work [nvmet_tcp]\n[ 5233.306532] Call Trace:\n[ 5233.306534] \n[ 5233.306536] dump_stack_lvl+0x73/0xb0\n[ 5233.306552] print_deadlock_bug+0x225/0x2f0\n[ 5233.306556] __lock_acquire+0x13f0/0x2290\n[ 5233.306563] lock_acquire+0xd0/0x300\n[ 5233.306565] ? touch_wq_lockdep_map+0x26/0x90\n[ 5233.306571] ? __flush_work+0x20b/0x530\n[ 5233.306573] ? touch_wq_lockdep_map+0x26/0x90\n[ 5233.306577] touch_wq_lockdep_map+0x3b/0x90\n[ 5233.306580] ? touch_wq_lockdep_map+0x26/0x90\n[ 52\n---truncated---", "severity": [ { "type": "CVSS_V3", "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/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": "6.1.175", "compiled_filter": false } } } } ], "references": [ { "type": "ADVISORY", "url": "https://www.cve.org/CVERecord?id=CVE-2026-46304" }, { "type": "FIX", "url": "https://git.kernel.org/stable/c/ae5b0cad163833e10b271e9becc05d81dae56e5f" }, { "type": "FIX", "url": "https://git.kernel.org/stable/c/8d66ba89480ff098a58d79003a505f383aa4e920" }, { "type": "FIX", "url": "https://git.kernel.org/stable/c/a696fbbd5240b4ac9b166f7bd4c550882ff543f1" }, { "type": "FIX", "url": "https://git.kernel.org/stable/c/9a4d7222c0955b221e38bb66d10e6bccb672c8a1" }, { "type": "FIX", "url": "https://git.kernel.org/stable/c/ee6e20c4bc9eae542a0954a368449532383169d4" }, { "type": "FIX", "url": "https://git.kernel.org/stable/c/781f47d641432c26c19625b2cdd7f40825097592" }, { "type": "FIX", "url": "https://git.kernel.org/stable/c/551f445a56a11a6457550cddcf39c9ebb8bcacc6" }, { "type": "FIX", "url": "https://git.kernel.org/stable/c/aade8abd8b868b6ffa9697aadaea28ec7f65bee6" } ], "database_specific": { "source": "https://git.kernel.org/pub/scm/linux/security/vulns.git", "upstream_modified": "2026-08-05T13:16:28Z", "severity": { "source": "kernel CNA", "cvss_v3_score": 7.5, "cvss_v3_severity": "High", "scenario": "AV:N - The bug is reached through the NVMe-oF TCP target (nvmet-tcp) disconnect/teardown path; remote peers connect over TCP and closing that connection drives `nvmet_tcp_release_queue_work()` on `nvmet-wq`, which can call the vulnerable `flush_work()` in `nvmet_ctrl_free()`.\nAC:L - An attacker fully controls both sides of the race by connecting, issuing an Async Event Request (which queues `async_event_work` on `nvmet-wq`), and then disconnecting to force controller teardown on the same workqueue.\nPR:N - Exploitation requires only network reachability to an exposed nvmet-tcp port; the discovery subsystem accepts any host NQN without Linux credentials, and no CAP_SYS_ADMIN or other local privileges are needed on the target.\nUI:N - The deadlock is triggered entirely through automated NVMe-oF protocol interaction (connect, admin command, disconnect) with no action required from a local user or administrator.\nS:U - Impact is confined to the kernel NVMe target subsystem (workqueue worker hang and nvmet service disruption); it does not cross a VM, container, or IOMMU security boundary.\nC:N - This is a recursive workqueue flush deadlock, not memory corruption; there is no out-of-bounds access, use-after-free, or information disclosure primitive.\nI:N - The flaw causes a worker to block during teardown but does not modify arbitrary kernel or user data and provides no write or code-execution primitive.\nA:H - Calling `flush_work()` on `nvmet-wq` from `nvmet_tcp_release_queue_work()` running on the same workqueue can permanently hang an nvmet worker (lockdep reports \"DEADLOCK\"), denying NVMe target service and stalling further nvmet workqueue processing." } } }