In the Linux kernel, the following vulnerability has been resolved:
drm/pagemap: Fix folio allocation fallback and use-after-put
drm_pagemap_migrate_populate_ram_pfn() had two issues when populating
RAM PFNs with higher-order folios:
-
The higher-order vma_alloc_folio()/folio_alloc() calls did not pass
__GFP_NOWARN, so a THP allocation failure under memory pressure
would spam the kernel log, and there was no fallback path despite a
TODO comment stating one was needed. Add __GFP_NOWARN to the
higher-order allocation and, on failure, fall back to order-0
allocations for the entire range originally covered by the failed
higher-order allocation, leaving MIGRATE_PFN_COMPOUND unset for
those PFNs.
-
In the free_pages error path, order was computed via
folio_order(page_folio(page)) after put_page(page) had already
dropped the reference, resulting in a use-after-free/put when that
was the last reference on the page. Compute order before releasing
the page.
Introducing the fallback in 1. also requires the source page array
handed to ->copy_to_ram() to be built differently. Both callers only
populated the entry at the head of each source folio, relying on the
copy callback to derive the rest of the folio from the order recorded
in the matching drm_pagemap_addr. Once the destination has been demoted
to order-0 folios the drm_pagemap_addr entries are per-page, so a source
page is needed for every one of them; leaving them NULL makes the copy
callback stop after the first page and the remainder of the range is
never copied.
The source folio is only split later, by migrate_vma_pages() /
migrate_device_pages(), so its order cannot be used to detect the
demotion - test the destination for MIGRATE_PFN_COMPOUND instead. Factor
the array population out into drm_pagemap_migrate_populate_src_pages()
and use it from both drm_pagemap_evict_to_ram() and
__drm_pagemap_migrate_to_ram().
CVSS Vector: CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
CVSS Score: 7.8
AV:L - The UAF is in drm_pagemap_migrate_populate_ram_pfn on the migrate-to-RAM path, reached from a local CPU fault on device-private SVM pages (do_swap_page → drm_pagemap_migrate_to_ram) or Xe TTM eviction (xe_svm_bo_evict → drm_pagemap_evict_to_ram), not from network packets.
AC:L - The attacker controls SVM setup and can force the free_pages path by failing a later GFP_HIGHUSER/THP allocation under self-induced memory pressure or fragmentation after earlier folios succeeded; the use-after-put then runs deterministically with no victim timing.
PR:L - DRM_IOCTL_XE_VM_CREATE and VM_BIND are DRM_RENDER_ALLOW with no capable() or DRM-master check. An unprivileged user with typical render/video-group access to /dev/dri/renderD* on an Intel Xe discrete GPU can create a fault-mode SVM VM and trigger migration.
UI:N - The attacking process drives mmap, SVM bind/prefetch to VRAM, migrate-back via CPU access or VRAM eviction, and memory pressure itself; no separate victim action is required.
S:U - The UAF corrupts host kernel page-allocator metadata in the same security authority. This is standard local kernel privilege escalation, not a KVM/Xen guest-to-host escape or IOMMU/DMA boundary bypass.
C:H - put_page() drops the last reference on the just-allocated folio, then folio_order(page_folio(page)) reads the freed page struct and may follow stale compound_info tail encoding, yielding an arbitrary kernel read of vmemmap/page flags.
I:H - The same use-after-put lets reused or stale page flags supply an attacker-influenced order that skips remaining puts or corrupts the free loop, and a page-struct UAF is exploitable via buddy reuse for arbitrary kernel writes or control-flow hijack.
A:H - Accessing a freed or reused page struct can oops or panic, and a corrupted order that makes NR_PAGES() return 0 hangs in an infinite free_pages loop; any kernel UAF is high availability impact.
| Attack Vector |
Local |
Scope |
Unchanged |
| Attack Complexity |
Low |
Confidentiality Impact |
High |
| Privileges Required |
Low |
Integrity Impact |
High |
| User Interaction |
None |
Availability Impact |
High |
AV:L - The UAF is in drm_pagemap_migrate_populate_ram_pfn on the migrate-to-RAM path, reached from a local CPU fault on device-private SVM pages (do_swap_page → drm_pagemap_migrate_to_ram) or Xe TTM eviction (xe_svm_bo_evict → drm_pagemap_evict_to_ram), not from network packets.
AC:L - The attacker controls SVM setup and can force the free_pages path by failing a later GFP_HIGHUSER/THP allocation under self-induced memory pressure or fragmentation after earlier folios succeeded; the use-after-put then runs deterministically with no victim timing.
PR:L - DRM_IOCTL_XE_VM_CREATE and VM_BIND are DRM_RENDER_ALLOW with no capable() or DRM-master check. An unprivileged user with typical render/video-group access to /dev/dri/renderD* on an Intel Xe discrete GPU can create a fault-mode SVM VM and trigger migration.
UI:N - The attacking process drives mmap, SVM bind/prefetch to VRAM, migrate-back via CPU access or VRAM eviction, and memory pressure itself; no separate victim action is required.
S:U - The UAF corrupts host kernel page-allocator metadata in the same security authority. This is standard local kernel privilege escalation, not a KVM/Xen guest-to-host escape or IOMMU/DMA boundary bypass.
C:H - put_page() drops the last reference on the just-allocated folio, then folio_order(page_folio(page)) reads the freed page struct and may follow stale compound_info tail encoding, yielding an arbitrary kernel read of vmemmap/page flags.
I:H - The same use-after-put lets reused or stale page flags supply an attacker-influenced order that skips remaining puts or corrupts the free loop, and a page-struct UAF is exploitable via buddy reuse for arbitrary kernel writes or control-flow hijack.
A:H - Accessing a freed or reused page struct can oops or panic, and a corrupted order that makes NR_PAGES() return 0 hangs in an infinite free_pages loop; any kernel UAF is high availability impact.
CVSS 3.1