Linux Kernel up to 6.18.34/7.0.11 Rmap mm/rmap.c try_to_unmap_one reference count

Summaryinfo

A vulnerability was found in Linux Kernel up to 6.18.34/7.0.11. It has been classified as problematic. Affected by this issue is the function try_to_unmap_one of the file mm/rmap.c of the component Rmap. Performing a manipulation results in reference count. This vulnerability is cataloged as CVE-2026-63950. The attack must be initiated from a local position. There is no exploit available.

Detailsinfo

A vulnerability, which was classified as problematic, was found in Linux Kernel up to 6.18.34/7.0.11. This affects the function try_to_unmap_one of the file mm/rmap.c of the component Rmap. The manipulation with an unknown input leads to a reference count vulnerability. CWE is classifying the issue as CWE-911. The product uses a reference count to manage a resource, but it does not update or incorrectly updates the reference count. This is going to have an impact on integrity, and availability. The summary by CVE is:

In the Linux kernel, the following vulnerability has been resolved: mm/rmap: initialize nr_pages to 1 at loop start in try_to_unmap_one Initialize nr_pages to 1 at the start of each loop iteration, like folio_referenced_one() does. Without this, nr_pages computed by a previous folio_unmap_pte_batch() call can be reused on a later iteration that does not run folio_unmap_pte_batch() again. mmap a 64K large folio with MAP_ANONYMOUS | MAP_DROPPABLE, then call madvise(MADV_FREE), then make the last page device-exclusive via HMM_DMIRROR_EXCLUSIVE. Trigger node reclaim through sysfs. Now, in try_to_unmap_one(), we will first clear the first 15 out of 16 entries mapping the lazyfree folio. This will set nr_pages to 15. In the next pvmw walk, this nr_pages gets reused on a device-exclusive pte, thus potentially corrupting folio refcount/mapcount. At the moment, I have a userspace program which can make the kernel spit out a trace, but the blow up is in folio_referenced_one(), because there are existing bugs in the interaction between device-private and rmap (which too I am investigating). I did a one liner kernel change to avoid going into folio_referenced_one(), and the kernel blows up at folio_remove_rmap_ptes in try_to_unmap_one which is what I wanted. Note that the bug is there not since file folio batching but lazyfree folio batching, since device-exclusive only works for anonymous folios. Userspace visible effect is simply kernel crashing somewhere due to refcount/mapcount corruption.

The advisory is shared at git.kernel.org. This vulnerability is uniquely identified as CVE-2026-63950 since 07/19/2026. The exploitability is told to be easy. An attack has to be approached locally. Technical details are known, but no exploit is available. The price for an exploit might be around USD $0-$5k at the moment (estimation calculated on 07/19/2026).

Upgrading to version 6.18.35 or 7.0.12 eliminates this vulnerability.

If you want to get the best quality for vulnerability data then you always have to consider VulDB.

Productinfo

Type

• Operating System

Vendor

• Linux

Name

• Kernel

Version

• 6.18.0
• 6.18.1
• 6.18.2
• 6.18.3
• 6.18.4
• 6.18.5
• 6.18.6
• 6.18.7
• 6.18.8
• 6.18.9
• 6.18.10
• 6.18.11
• 6.18.12
• 6.18.13
• 6.18.14
• 6.18.15
• 6.18.16
• 6.18.17
• 6.18.18
• 6.18.19
• 6.18.20
• 6.18.21
• 6.18.22
• 6.18.23
• 6.18.24
• 6.18.25
• 6.18.26
• 6.18.27
• 6.18.28
• 6.18.29
• 6.18.30
• 6.18.31
• 6.18.32
• 6.18.33
• 6.18.34
• 7.0.0
• 7.0.1
• 7.0.2
• 7.0.3
• 7.0.4
• 7.0.5
• 7.0.6
• 7.0.7
• 7.0.8
• 7.0.9
• 7.0.10
• 7.0.11

License

• open-source

Website

• Vendor: https://www.kernel.org/

CPE 2.3info

• 🔒
• 🔒
• 🔒

CPE 2.2info

• 🔒
• 🔒
• 🔒

CVSSv4info

CVSSv3info

CVSSv2info

Exploitinginfo

Threat Intelligenceinfo

Countermeasuresinfo

Timelineinfo

Sourcesinfo

Entryinfo

If you want to get the best quality for vulnerability data then you always have to consider VulDB.

Discussion