CVE-2026-99723: vhost-vdpa: fix pinned page leak in vhost_vdpa_pa_map() error path
In the Linux kernel, the following vulnerability has been resolved:
vhost-vdpa: fix pinned page leak in vhostvdpapamap() error path
When pinuserpages() or vhostvdpamap() fails, vhostvdpapamap() is supposed to unpin the pages pinned after the last successfully mapped chunk, i.e. the range [mappfn, lastpfn]. The unpin is guarded by nchunks, but nchunks is reset to 0 every time a chunk is mapped from inside the current page list, so a failure after any such boundary leaves that pending range pinned forever. With physically scattered memory nearly every page list contains a boundary, so the leak happens on practically every failed IOTLB update.
The pending range is non-empty exactly when a page has been pinned, which lastpfn already tells us, so drop nchunks and test lastpfn.
Affected Software
Remediation
Recommended actions to resolve this vulnerability, in priority order.
- Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 5.10 - 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 - Upgrade
Upgrade
debian/linuxto a version that resolves this vulnerability.Fixed in 7.3.5-1 - Compensating control
In vhost_vdpa_pa_map()'s error path, drop nchunks and use last_pfn to determine whether the pending page range must be unpinned.
Event History
Frequently Asked Questions
Under what conditions does the page leak occur?
It occurs when pin_user_pages() or vhost_vdpa_map() fails during an IOTLB update after pages have been pinned. The issue is especially likely with physically scattered memory, where page-list chunk boundaries occur frequently.
What is the operational impact of a failed mapping request?
Pages pinned after the last successfully mapped chunk can remain pinned indefinitely rather than being unpinned on the error path. Repeated failed IOTLB updates can therefore accumulate pinned-page leaks.
How can affected systems be identified from behavior?
Look for failed IOTLB updates involving vhost-vdpa together with pinned pages that are not released afterward. The vulnerable condition is tied to failures in pin_user_pages() or vhost_vdpa_map(), particularly when mappings span page-list boundaries.