{"id":"AZL-100100","summary":"CVE-2026-89495 affecting package kernel 6.6.150.1-1","details":"In the Linux kernel, the following vulnerability has been resolved:\n\nocfs2: bound namelen in dlm_migrate_request_handler\n\nPatch series \"ocfs2/dlm: bound peer-controlled lengths in the o2dlm\".\n\nThe o2dlm receive handlers trust u8 length and count fields from the wire\nwithout bounding them, so a node in a DLM domain can corrupt or panic any\nother node with a malformed message.  Three defects:\n\n  - dlm_migrate_request_handler() passes migrate-\u003enamelen unchecked to\n    dlm_init_mle(), which memcpy()s it into the 32-byte mname[] of an\n    o2dlm_mle slab object: a heap out-of-bounds write of up to ~215\n    attacker-controlled bytes.\n\n  - dlm_mig_lockres_handler() passes mres-\u003elockname_len unchecked to\n    dlm_init_lockres(), which memcpy()s it into the 32-byte o2dlm_lockname\n    slab object: a heap out-of-bounds write of up to ~223 bytes.\n\n  - the same handler trusts mres-\u003enum_locks without checking that the\n    message is large enough to hold that many entries, so\n    dlm_process_recovery_data() walks mres-\u003eml[] past the kmalloc(data_len)\n    copy and trips a BUG_ON (an out-of-bounds read ending in a panic).\n\nThe other o2dlm receive handlers already reject an oversized name; the\nmigration and recovery handlers have omitted it since the DLM was added\n(see the Fixes tags).  Patch 1 bounds namelen; patch 2 validates\nlockname_len, num_locks, and the payload size.  Conforming recovery and\nmigration traffic is unaffected.\n\no2net authenticates peers only by the DLM domain key, so any node that has\njoined the domain -- including a compromised or malicious member -- can\nsend these messages.  There is no local trigger; the attacker must already\nbe a member of the cluster.\n\nEach sink was confirmed under KASAN with an out-of-tree module mirroring\nit exactly -- a kmem_cache/kmalloc of the real destination size, then the\nsame unclamped memcpy/loop: slab-out-of-bounds Write for the two writes,\nRead for the recovery walk, and a panic.  A userspace AddressSanitizer\nbuild faults identically under -m32 and -m64.  Scrubbed logs are available\non request.\n\nI reported this privately to security@kernel.org and the ocfs2 maintainers\non 2026-06-20; with no response after the standard embargo period I am\nposting the fix publicly.  I have no embargo requirement.\n\n\nThis patch (of 2):\n\nA node receiving a DLM_MIGRATE_REQUEST message trusts the peer-supplied\nname length (migrate-\u003enamelen) without bounding it.  dlm_init_mle() then\ncopies that many bytes into the fixed DLM_LOCKID_NAME_MAX-byte mname[]\narray of an o2dlm_mle slab object, so a malformed message from a cluster\npeer overflows the slab object by up to ~215 bytes: a heap out-of-bounds\nwrite of attacker-controlled data, reachable by any node in the domain.\n\nReject an oversized name, the way dlm_master_request_handler() and the\nother o2dlm receive handlers already do; the migration handler omits the\ncheck entirely.  Conforming messages are unaffected.","modified":"2026-09-13T06:06:36Z","published":"2026-09-11T20:19:31Z","upstream":["CVE-2026-89495"],"references":[{"type":"WEB","url":"https://nvd.nist.gov/vuln/detail/CVE-2026-89495"}],"affected":[{"package":{"name":"kernel","ecosystem":"Azure Linux:3","purl":"pkg:rpm/azure-linux/kernel"},"ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"0"},{"last_affected":"6.6.150.1-1"}]}],"database_specific":{"source":"https://github.com/microsoft/AzureLinuxVulnerabilityData/blob/main/osv/AZL-100100.json"}}],"schema_version":"1.9.0"}