Loading vulnerability…
CVE-2026-46274 is a Linux kernel memory-safety flaw in the io-wq subsystem. The bug is in io_wq_remove_pending(), which updates wq->hash_tail[] when removing cancelled work from a hashed bucket. The vulnerable logic compares the cancelled work item with its predecessor in acct->work_list to determine whether the predecessor should become the new bucket tail, but it fails to verify that the predecessor is itself a hashed work item. Because the hash value is derived from flag bits and non-hashed work does not have those bits set, a non-hashed predecessor can be misidentified as belonging to hash bucket 0. As a result, a pointer to a non-hashed io_kiocb can be stored in wq->hash_tail[0]. That non-hashed work is later dequeued through a fast path that does not clear hash_tail[], so once the io_kiocb completes and is freed, hash_tail[0] becomes a dangling pointer. Since the io_wq persists for the lifetime of the task, the stale pointer can survive ring open and close operations. A subsequent enqueue of hashed bucket-0 work dereferences the dangling pointer in io_wq_insert_work(), leading to writes through freed memory.
Mallory correlates every CVE against your assets, your vendors, and active adversary campaigns. Know which vulnerabilities matter for you, not just which ones are loud.
What it means. What to do now. Patch path, mitigations, and the assume-compromise checklist.
What an attacker gets, and what they’ve been doing with it.
If you can’t patch tonight, do this now.
Patch, then assume compromise.
No public exploits tracked yet. Mallory keeps watching.
No public exploit code observed for this vulnerability.
Products and vendors Mallory has correlated with this vulnerability. Open in Mallory to drill down to specific CPE configurations and version ranges.
Vendor-confirmed product mapping. Mallory continuously reconciles this list against your asset inventory.
27 sources tracked across advisories, community write-ups, and news. New activity surfaces here as Mallory finds it.
Linux kernel io-wq flaw involving predecessor hash-state validation in io_wq_remove_pending().
Linux kernel io-wq flaw caused by failure to verify that a predecessor is hashed in io_wq_remove_pending().
Linux kernel io-wq validation flaw in io_wq_remove_pending() concerning whether a predecessor is hashed.
Linux kernel io-wq flaw in which io_wq_remove_pending() did not verify that a predecessor was hashed.
Query your assets running an affected version, and investigate the blast radius.
Every observed campaign linking this CVE to a named adversary.
Malware families riding this exploit, with evidence and IOCs.
YARA, Sigma, Snort, and vendor rules, auto-deployed to your SIEM.
Cross-references every affected SKU, including bundled OEM variants.
Community discussion across Reddit, Mastodon, and other social sources.