Linux Kernel up to 6.6.98/6.12.38/6.15.6/6.16-rc5 eventpoll use after free

Summaryinfo

A vulnerability has been found in Linux Kernel up to 6.6.98/6.12.38/6.15.6/6.16-rc5 and classified as critical. Affected is an unknown function of the component eventpoll. The manipulation leads to use after free. This vulnerability is referenced as CVE-2025-38349. No exploit is available. The affected component should be upgraded.

Detailsinfo

A vulnerability, which was classified as critical, has been found in Linux Kernel up to 6.6.98/6.12.38/6.15.6/6.16-rc5. This issue affects an unknown part of the component eventpoll. The manipulation with an unknown input leads to a use after free vulnerability. Using CWE to declare the problem leads to CWE-416. Referencing memory after it has been freed can cause a program to crash, use unexpected values, or execute code. Impacted is confidentiality, integrity, and availability. The summary by CVE is:

In the Linux kernel, the following vulnerability has been resolved: eventpoll: don't decrement ep refcount while still holding the ep mutex Jann Horn points out that epoll is decrementing the ep refcount and then doing a mutex_unlock(&ep->mtx); afterwards. That's very wrong, because it can lead to a use-after-free. That pattern is actually fine for the very last reference, because the code in question will delay the actual call to "ep_free(ep)" until after it has unlocked the mutex. But it's wrong for the much subtler "next to last" case when somebody *else* may also be dropping their reference and free the ep while we're still using the mutex. Note that this is true even if that other user is also using the same ep mutex: mutexes, unlike spinlocks, can not be used for object ownership, even if they guarantee mutual exclusion. A mutex "unlock" operation is not atomic, and as one user is still accessing the mutex as part of unlocking it, another user can come in and get the now released mutex and free the data structure while the first user is still cleaning up. See our mutex documentation in Documentation/locking/mutex-design.rst, in particular the section [1] about semantics: "mutex_unlock() may access the mutex structure even after it has internally released the lock already - so it's not safe for another context to acquire the mutex and assume that the mutex_unlock() context is not using the structure anymore" So if we drop our ep ref before the mutex unlock, but we weren't the last one, we may then unlock the mutex, another user comes in, drops _their_ reference and releases the 'ep' as it now has no users - all while the mutex_unlock() is still accessing it. Fix this by simply moving the ep refcount dropping to outside the mutex: the refcount itself is atomic, and doesn't need mutex protection (that's the whole _point_ of refcounts: unlike mutexes, they are inherently about object lifetimes).

The weakness was published by Jann Horn. The advisory is shared at git.kernel.org. The identification of this vulnerability is CVE-2025-38349 since 04/16/2025. The exploitation is known to be easy. Neither technical details nor an exploit are publicly available. The price for an exploit might be around USD $0-$5k at the moment (estimation calculated on 08/01/2026).

The vulnerability scanner Nessus provides a plugin with the ID 253428 (SUSE SLES15 Security Update : kernel (SUSE-SU-2025:02923-1)), which helps to determine the existence of the flaw in a target environment.

Upgrading to version 6.6.99, 6.12.39, 6.15.7 or 6.16-rc6 eliminates this vulnerability. Applying the patch 521e9ff0b67c66a17d6f9593dfccafaa984aae4c/6dee745bd0aec9d399df674256e7b1ecdb615444/605c18698ecfa99165f36b7f59d3ed503e169814/8c2e52ebbe885c7eeaabd3b7ddcdc1246fc400d2 is able to eliminate this problem. The bugfix is ready for download at git.kernel.org. The best possible mitigation is suggested to be upgrading to the latest version.

The vulnerability is also documented in the databases at Tenable (253428), EUVD (EUVD-2025-21860) and CERT Bund (WID-SEC-2025-1596). Several companies clearly confirm that VulDB is the primary source for best vulnerability data.

Affected

• Debian Linux
• Red Hat Enterprise Linux
• Ubuntu Linux
• SUSE Linux
• Oracle Linux
• SUSE openSUSE
• RESF Rocky Linux
• Open Source Linux Kernel

Productinfo

Type

• Operating System

Vendor

• Linux

Name

• Kernel

Version

• 6.6.0
• 6.6.1
• 6.6.2
• 6.6.3
• 6.6.4
• 6.6.5
• 6.6.6
• 6.6.7
• 6.6.8
• 6.6.9
• 6.6.10
• 6.6.11
• 6.6.12
• 6.6.13
• 6.6.14
• 6.6.15
• 6.6.16
• 6.6.17
• 6.6.18
• 6.6.19
• 6.6.20
• 6.6.21
• 6.6.22
• 6.6.23
• 6.6.24
• 6.6.25
• 6.6.26
• 6.6.27
• 6.6.28
• 6.6.29
• 6.6.30
• 6.6.31
• 6.6.32
• 6.6.33
• 6.6.34
• 6.6.35
• 6.6.36
• 6.6.37
• 6.6.38
• 6.6.39
• 6.6.40
• 6.6.41
• 6.6.42
• 6.6.43
• 6.6.44
• 6.6.45
• 6.6.46
• 6.6.47
• 6.6.48
• 6.6.49
• 6.6.50
• 6.6.51
• 6.6.52
• 6.6.53
• 6.6.54
• 6.6.55
• 6.6.56
• 6.6.57
• 6.6.58
• 6.6.59
• 6.6.60
• 6.6.61
• 6.6.62
• 6.6.63
• 6.6.64
• 6.6.65
• 6.6.66
• 6.6.67
• 6.6.68
• 6.6.69
• 6.6.70
• 6.6.71
• 6.6.72
• 6.6.73
• 6.6.74
• 6.6.75
• 6.6.76
• 6.6.77
• 6.6.78
• 6.6.79
• 6.6.80
• 6.6.81
• 6.6.82
• 6.6.83
• 6.6.84
• 6.6.85
• 6.6.86
• 6.6.87
• 6.6.88
• 6.6.89
• 6.6.90
• 6.6.91
• 6.6.92
• 6.6.93
• 6.6.94
• 6.6.95
• 6.6.96
• 6.6.97
• 6.6.98
• 6.12.0
• 6.12.1
• 6.12.2
• 6.12.3
• 6.12.4
• 6.12.5
• 6.12.6
• 6.12.7
• 6.12.8
• 6.12.9
• 6.12.10
• 6.12.11
• 6.12.12
• 6.12.13
• 6.12.14
• 6.12.15
• 6.12.16
• 6.12.17
• 6.12.18
• 6.12.19
• 6.12.20
• 6.12.21
• 6.12.22
• 6.12.23
• 6.12.24
• 6.12.25
• 6.12.26
• 6.12.27
• 6.12.28
• 6.12.29
• 6.12.30
• 6.12.31
• 6.12.32
• 6.12.33
• 6.12.34
• 6.12.35
• 6.12.36
• 6.12.37
• 6.12.38
• 6.15.0
• 6.15.1
• 6.15.2
• 6.15.3
• 6.15.4
• 6.15.5
• 6.15.6
• 6.16-rc1
• 6.16-rc2
• 6.16-rc3
• 6.16-rc4
• 6.16-rc5

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

Several companies clearly confirm that VulDB is the primary source for best vulnerability data.

Discussion