data built 2026-10-19T13:16:53Z

Look up another CVE · open CVE-2026-64150 in the interactive tracker

CVE-2026-64150

netfilter: nft_inner: release local_lock before re-enabling softirqs

9.8 EPSS 0.6% AV:N

Published 2026-07-19 (3 months ago) by the Linux kernel CNA. Subsystem net/netfilter. Debian urgency: unimportant.

What to do, per Debian release

Debian's own status for CVE-2026-64150 in every release this site tracks, and what each one means in practice. Every status below is the one the Debian Security Team publishes for this CVE. source

ReleaseStatus VersionWhat this means
Debian 12bookwormnot affected6.1.190-1Nothing to do. bookworm never shipped the vulnerable code.
Debian 13trixienot affected6.12.113-1Nothing to do. trixie never shipped the vulnerable code.
Debian 14forkyfixed7.0.12-1Update the kernel and reboot. Fixed in 7.0.12-1. A running kernel keeps the old code until the machine restarts.
Debian unstablesidfixed7.0.12-1Update the kernel and reboot. Fixed in 7.0.12-1. A running kernel keeps the old code until the machine restarts.

The vulnerability

What the bug is

Quoting sashiko: In the error path, local_bh_enable() is called before local_unlock_nested_bh().

Where it lives

The fix touches these files. If the subsystem is one you do not use, the practical exposure is lower, though the package is still the vulnerable one.

net/netfilter/nft_inner.c

What this is
NETFILTERMAINTAINERS section matched on net/netfilter/ · netfilter-devel@vger.kernel.org · Maintained
Kernel module
nf_tables

Check it on the machine

grep -wE 'CONFIG_NETFILTER|CONFIG_NF_TABLES' /boot/config-$(uname -r)

Reads the booted kernel's own configuration, on the machine and nowhere else. =y is built in, =m is a module, and an is not set line means the code is not there.

lsmod | grep -w nf_tables

Prints a line if that module is loaded right now, and nothing if it is not. One that is not loaded can still be loaded later by anything that needs it.

Show whether Debian builds this, per release

Debian builds this as a loadable module. Every release with an answer below compiles it as a module on every flavour, so whether the code is in the running kernel depends on whether that module is loaded. The build rule could not be established for 1 release, so nothing is claimed about those.

  • bookworm?
  • trixie, forky, sidm
?
not established - no Makefile entry for the affected file
m
a loadable module on every flavour - check lsmod, can be blacklisted

Where a release above reads "not established", no build rule for the affected file was found in that release's kernel tree. That is a gap in this lookup, not a finding that the code is absent.

Build switches
CONFIG_NETFILTER, CONFIG_NF_TABLES
Reads as
all on means the file is compiled; any one off means it is not.

From the configuration Debian publishes for every kernel flavour it builds. It covers only the architectures Debian ships a linux-config package for, and nothing about a kernel you built yourself. Debian's own status above stays authoritative about the source package.MAINTAINERS kbuild Makefiles Debian linux-config packages

Signals

Published
2026-07-19 (3 months ago)source
CVSS
9.8 critical
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:Hsource recompute it
Attack vector
network (AV:N)source
EPSS
0.65%, higher than 50.1% of all CVEssource
CISA KEV
not listedthe KEV catalogue
Debian urgency
unimportantsource
Subsystem
net/netfilter

Upstream fixes

The versions the CNA records as introducing and fixing this, and the commit for each. source

IntroducedFixed inCommit
6.166.18.34df19b6af1716
6.167.0.116fecd39c6401
6.167.1a6cb3ff97985

Why the CNA scored it this way

Quoted from the kernel CNA's own scoring file, word for word. A highlighted metric carries the most severe value CVSS v3.1 defines for it, which is what pushes the score up. source

AV:NAttack vector, network
The bug is hit during packet evaluation in netfilter hooks (prerouting/forward/input) when tunneled traffic (GRE/UDP/Geneve) matching admin-deployed `inner` rules is processed; remote packets reach `nft_inner_restore_tun_ctx()` without any local syscall.
AC:LAttack complexity, low
An attacker can reliably provoke the cookie-mismatch error path by sending concurrent tunneled packets through an `inner`-enabled chain, controlling both sides of the per-CPU context race; no attacker-uncontrollable memory layout or rare timing is required.
PR:NPrivileges required, none
Exploitation requires only the ability to send network packets to a host whose nftables rules already include `inner` expressions (typical on cloud NFV/firewall gateways); no local account, capability, or user-namespace `CAP_NET_ADMIN` is needed by the remote attacker.
UI:NUser interaction, none
No victim interaction beyond normal network traffic delivery to a host running nftables tunnel-inspection rules; the attacker does not need the victim to open files, mount filesystems, or take any deliberate action.
S:UScope, unchanged
Impact is confined to kernel netfilter/nftables processing on the target host; it does not cross VM, container, or IOMMU security boundaries to affect separate security authorities.
C:HConfidentiality impact, high
Re-enabling softirqs before releasing the nested-BH lock breaks serialization of the per-CPU `nft_pcpu_tun_ctx`, allowing concurrent corruption of inner header offsets that `nft_payload_inner_eval()` uses in `skb_copy_bits()`, yielding out-of-bounds kernel memory reads.
I:HIntegrity impact, high
The same unsynchronized per-CPU state corruption can desynchronize inner parsing metadata and, on PREEMPT_RT, leave a per-CPU spinlock held across softirq re-entry causing deadlock; memory corruption of parsing state is exploitable for further control of packet-processing behavior.
A:HAvailability impact, high
On PREEMPT_RT kernels the inverted unlock/BH ordering leaves the per-CPU spinlock held while softirqs run, causing softirq handlers to block on the same lock and producing CPU deadlock or soft lockup; non-RT paths can also crash or hang from the resulting race.