CVE-2026-99655: mm/memory: reject remote COW faults on PFN mappings
In the Linux kernel, the following vulnerability has been resolved:
mm/memory: reject remote COW faults on PFN mappings
When handling a COW fault without a normal source page (PFN mappings), wppagecopyuser() copies from the faulting user address. For remote faults, this accesses the wrong address space, resulting in wrong data getting copied.
For example, userevents updates an enable bit in the registering process's memory. If that memory needs to be faulted in, usereventenablerfaultfixup() calls fixupuserfault() from a work item. Placing the enable bit in a private PFN mapping can trigger this remote COW path.
Reject these faults with VMFAULTSIGBUS when the target mm differs from current->mm: remote faults. Ordinary (non-PFN) mappings are not affected.
The user-address copy dates back to commit 6aab341e0a28 ("mm: re-architect the VMUNPAGED logic"); likely, actual ways to get remote faults in there on COW PFN mappings appeared afterwards. But the fix is simple and straight forward.
Fix generated by an LLM and sent by Chris Mason in private.
[ david: simplify comment and description ]
Affected Software
Remediation
Recommended actions to resolve this vulnerability, in priority order.
- Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 6.12.118 - Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 6.18.64 - Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 7.3.5 - Compensating control
In Linux kernel mm/memory, reject remote COW faults on PFN mappings with VM_FAULT_SIGBUS when the target mm differs from current->mm.
Frequently Asked Questions
Which systems are exposed to the incorrect copy behavior?
The affected condition requires a remote COW fault on a private PFN mapping, where the target memory map differs from the current task's memory map. Ordinary mappings that have a normal source page are not affected.
What must occur for the vulnerable path to be reached?
A fault must be handled remotely rather than in the target process's current address space, and the faulting address must be in a COW PFN mapping. The described example involves user_events faulting in an enable bit in the registering process's memory from a work item.
What happens after the fix when this condition is encountered?
The kernel rejects remote COW faults on PFN mappings with VM_FAULT_SIGBUS instead of copying from the faulting user address in the wrong address space.