mm/huge_memory: initialise workingset state before folio split
Linux
Affected versions
| Product | Affected | Fixed in |
|---|---|---|
| Linux Linux |
| No fix yet |
| Linux Linux |
|
|
Linux
| Product | Affected | Fixed in |
|---|---|---|
| Linux Linux |
| No fix yet |
| Linux Linux |
|
|
6.18.567.1.97.2In the Linux kernel, the following vulnerability has been resolved:
mm/huge_memory: initialise workingset state before folio split
xas_try_split() adds __GFP_ACCOUNT for page-cache xa_nodes, but __folio_split() leaves the xa_state's xa_lru unset. That lets a live, memcg-charged xa_node exist without being linked into the mapping's shadow_nodes list_lru; when reclaim later walks the list_lru it trips VM_WARN_ON(!css_is_dying()).
Use mapping_set_update() to install both the workingset update callback and the shadow_nodes list_lru on the xa_state.
From the CVE record · Linux
The Linux kernel can leave a memory-cgroup-charged page-cache node off the mapping’s shadow-node list during folio splitting. When reclaim later walks that list, it can trigger a kernel warning. The input does not specify attacker access requirements or user interaction.
Engineers running Linux kernels should check their vendor’s advisory for affected versions; none are listed in the input.