{ "schema_version": "1.6.1", "id": "RPI-CVE-2023-54318", "modified": "2026-10-09T11:39:44Z", "upstream": [ "CVE-2023-54318" ], "published": "2025-12-30T12:24:11Z", "summary": "net/smc: use smc_lgr_list.lock to protect smc_lgr_list.list iterate in smcr_port_add", "details": "In the Linux kernel, the following vulnerability has been resolved:\n\nnet/smc: use smc_lgr_list.lock to protect smc_lgr_list.list iterate in smcr_port_add\n\nWhile doing smcr_port_add, there maybe linkgroup add into or delete\nfrom smc_lgr_list.list at the same time, which may result kernel crash.\nSo, use smc_lgr_list.lock to protect smc_lgr_list.list iterate in\nsmcr_port_add.\n\nThe crash calltrace show below:\nBUG: kernel NULL pointer dereference, address: 0000000000000000\nPGD 0 P4D 0\nOops: 0000 [#1] SMP NOPTI\nCPU: 0 PID: 559726 Comm: kworker/0:92 Kdump: loaded Tainted: G\nHardware name: Alibaba Cloud Alibaba Cloud ECS, BIOS 449e491 04/01/2014\nWorkqueue: events smc_ib_port_event_work [smc]\nRIP: 0010:smcr_port_add+0xa6/0xf0 [smc]\nRSP: 0000:ffffa5a2c8f67de0 EFLAGS: 00010297\nRAX: 0000000000000001 RBX: ffff9935e0650000 RCX: 0000000000000000\nRDX: 0000000000000010 RSI: ffff9935e0654290 RDI: ffff9935c8560000\nRBP: 0000000000000000 R08: 0000000000000000 R09: ffff9934c0401918\nR10: 0000000000000000 R11: ffffffffb4a5c278 R12: ffff99364029aae4\nR13: ffff99364029aa00 R14: 00000000ffffffed R15: ffff99364029ab08\nFS: 0000000000000000(0000) GS:ffff994380600000(0000) knlGS:0000000000000000\nCS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033\nCR2: 0000000000000000 CR3: 0000000f06a10003 CR4: 0000000002770ef0\nPKRU: 55555554\nCall Trace:\n smc_ib_port_event_work+0x18f/0x380 [smc]\n process_one_work+0x19b/0x340\n worker_thread+0x30/0x370\n ? process_one_work+0x340/0x340\n kthread+0x114/0x130\n ? __kthread_cancel_work+0x50/0x50\n ret_from_fork+0x1f/0x30", "severity": [ { "type": "CVSS_V3", "score": "CVSS:3.1/AV:A/AC:L/PR:N/UI:N/S:U/C:H/I:H/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": "6.1.54", "compiled_filter": false } } } } ], "references": [ { "type": "ADVISORY", "url": "https://www.cve.org/CVERecord?id=CVE-2023-54318" }, { "type": "FIX", "url": "https://git.kernel.org/stable/c/d1c6c93c27a4bf48006ab16cd9b38d85559d7645" }, { "type": "FIX", "url": "https://git.kernel.org/stable/c/06b4934ab2b534bb92935c7601852066ebb9eab8" }, { "type": "FIX", "url": "https://git.kernel.org/stable/c/70c8d17007dc4a07156b7da44509527990e569b3" }, { "type": "FIX", "url": "https://git.kernel.org/stable/c/b717463610a27fc0b58484cfead7a623d5913e61" }, { "type": "FIX", "url": "https://git.kernel.org/stable/c/f5146e3ef0a9eea405874b36178c19a4863b8989" } ], "database_specific": { "source": "https://git.kernel.org/pub/scm/linux/security/vulns.git", "upstream_modified": "2026-08-05T11:24:39Z", "severity": { "source": "kernel CNA", "cvss_v3_score": 8.8, "cvss_v3_severity": "High", "scenario": "AV:A - The vulnerable walk runs from `smc_ib_port_event_work`, which is scheduled by RoCE async events an adjacent attacker can generate \u2014 IPv6 Router Advertisements churn the netdev's addresses into `IB_EVENT_GID_CHANGE`, and a RoCE fabric peer can raise `IB_EVENT_QP_ACCESS_ERR`; the racing link-group add/delete is driven by SMC-R connection setup/teardown from a peer on the same fabric.\nAC:L - The attacker controls both sides of the race \u2014 repeated address/GID churn to re-arm `smcr_port_add()` and rapid SMC-R connection create/teardown to add and free link groups \u2014 and the window spans the entire unlocked list traversal, so the attempt can be repeated indefinitely until it lands.\nPR:N - SMC link groups are created and torn down during the CLC handshake before any application-level authentication, and injecting Router Advertisements or malformed RDMA operations requires no credentials on the target.\nUI:N - Everything is driven by inbound link-layer events and connection churn handled in kernel workqueue context; no local user has to open, mount, or run anything.\nS:U - The corruption is confined to kernel memory managed by the same kernel security authority; no VM, IOMMU, or sandbox boundary is crossed.\nC:H - The unlocked iterator dereferences a `struct smc_link_group` that has already been unlinked and kfree'd, reading `pnet_id`, `type`, `net`, and `lnk[]` out of reallocated slab memory and following freed/reused `list` pointers, which is an attacker-influenceable kernel memory read.\nI:H - Reaching `smc_llc_add_link_local()` on the freed link leads to `smc_llc_enqueue()`, which takes a spinlock in freed memory, performs `list_add_tail()` into the freed `lgr->llc_event_q` (a write into reclaimed slab), and calls `queue_work()` on a freed `work_struct` \u2014 a use-after-free write usable for control-flow hijack after heap grooming.\nA:H - The bug is a confirmed production kernel oops \u2014 NULL pointer dereference at `smcr_port_add+0xa6` in `smc_ib_port_event_work` \u2014 and any use-after-free on the link-group list reliably panics the machine." } } }